pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
29130172 | null | s2 | 5,822 | {
"abstract": "Bacteria use chemical molecules called autoinducers as votes to poll their numerical strength in a colony. This polling mechanism, commonly referred to as quorum sensing, enables bacteria to build a social network and provide a collective response for fighting off common threats. In Gram-negative bacteria, AHL synthases synthesize acyl-homoserine lactone (AHL) autoinducers to turn on the expression of several virulent genes including biofilm formation, protease secretion, and toxin production. Therefore, inhibiting AHL signal synthase would limit quorum sensing and virulence. In this chapter, we describe four enzymatic methods that could be adopted to investigate a broad array of AHL synthases. The enzymatic assays described here should accelerate our mechanistic understanding of quorum-sensing signal synthesis that could pave the way for discovery of potent antivirulence compounds."
} | 223 |
23349062 | null | s2 | 5,823 | {
"abstract": "Sericin removal from the core fibroin protein of silkworm silk is a critical first step in the use of silk for biomaterial-related applications, but degumming can affect silk biomaterial properties, including molecular weight, viscosity, diffusivity and degradation behavior. Increasing the degumming time (10, 30, 60, and 90 min) decreases the average molecular weight of silk protein in solution, silk solution viscosity, and silk film glass-transition temperature, and increases the rate of degradation of a silk film by protease. Model compounds spanning a range of physical-chemical properties generally show an inverse relationship between degumming time and release rate through a varied degumming time silk coating. Degumming provides a useful control point to manipulate silk's material properties."
} | 201 |
26826978 | null | s2 | 5,824 | {
"abstract": "Biofilm bacteria have developed escape strategies to avoid stresses associated with biofilm growth, respond to changing environmental conditions, and disseminate to new locations. An ever-expanding body of research suggests that cellular release from biofilms is distinct from a simple reversal of attachment and reversion to a planktonic mode of growth, with biofilm dispersion involving sensing of specific cues, regulatory signal transduction, and consequent physiological alterations. However, dispersion is only one of many ways to escape the biofilm mode of growth. The present review is aimed at distinguishing this active and regulated process of dispersion from the passive processes of desorption and detachment by highlighting the regulatory processes and distinct phenotypes specific to dispersed cells."
} | 203 |
38959041 | PMC11252921 | pmc | 5,826 | {
"abstract": "Significance Nanofluidics has recently revolutionized the traditional solid-state neuromorphic emulations, yet a realization of a robust nanofluidic spiking device remains a challenge. Here, we report a nanofluidic spiking synapse based on a poly (3,4-ethylenedioxythiophene) polystyrene sulfonate membrane confined within a nanopore, which mimics time-dependent plasticity and the biological transmembrane spiking with ionic and dopaminergic tunability. The shared ionic current-induced potential spiking in aqueous environment suggests that this device has the potential to communicate with biological synapses.",
"conclusion": "Conclusion In summary, we have demonstrated a nanofluidic spiking synapse through confinement of a PEDOT: PSS membrane within the nanopipette. Stemming from geometrical asymmetry of the nanopore and capacitance of the PEDOT: PSS membrane, the bionic synapse manifested both ionic rectification and pinched hysteresis, the collected effects of which enabled time-dependent plasticity and especially the ionic current-induced spiking that highly analogous to the biological generation of APs. The corresponding mechanism could be explained by the ICP aside from the PEDOT: PSS membrane. Comparing with existing organic electrochemical transistor spiking devices in aqueous condition, this nanofluidic synapse could be operated with much simplified circuit and lower energy consumption. Moreover, the spiking characteristics could be chemically modulated by different ionic environments and real neurotransmitter dopamine. The shared ionic current-induced spiking in aqueous environment suggests its potential to communicate with biological synapses in a spiking manner. Nevertheless, we acknowledge the difference of this bionic synapse from the biological one, in which the complex coordination among different ion channels within synaptic membranes makes possible the occurrence of APs. Highly close imitation of such rich behaviors remains challenging for current neuromorphic engineering. Using the nanofluidic technique, this work realized the spiking-based signaling, the scalable fabrication of which might facilitate the brain-like neuromorphic computation based on temporally coded spikes using STDP-based learning rules.",
"discussion": "Results and Discussion PEDOT: PSS Nanofluidic Synapse. The confinement of PEDOT: PSS membrane at the nanotip was confirmed by the scanning electron microscope (SEM) and transmission electron microscope (TEM) images ( SI Appendix , Fig. S1 ). The nanofluidic device was then tested upon periodic triangular wave between −1 V and +1 V at a scan rate of 100 mV/s in a 100 mM KCl solution, which manifested both ionic rectification and pinched hysteresis characteristics in the current-potential (I–V) curve ( Fig. 1 C ) with good stability over 1,000 repeated cycles ( SI Appendix , Fig. S2 ). The rectification and hysteresis behavior closely resembled the collected properties of a diode and a memristor, which stemmed from the asymmetrical charge distribution in the confined nanopore ( 14 – 17 ) and the capacitive effect of PEDOT: PSS membrane ( 18 , 19 ), respectively. Note that the hysteresis loop intersected itself near 0 V and exhibited a clockwise/anticlockwise trend in the first/third quadrant, respectively, yielding the monocyclic conductance switching ( Fig. 1 C , Inset ). Incidentally, the pristine nanopipette without PEDOT: PSS membrane ( SI Appendix , Fig. S3 ), the nanopipettes with overoxidized ( SI Appendix , Fig. S4 ) or cross-linked PEDOT: PSS ( SI Appendix , Fig. S5 ) could only produce linear curves without hysteresis, indicating the crucial role of PEDOT: PSS membrane in generating the ionic hysteresis. Essentially, the negative rectification originated from negatively charged sulfonic acid moieties in the PEDOT: PSS membrane ( 20 , 21 ) and geometrical asymmetry of the conical nanotip ( 22 – 24 ). These effects would lead to cations being the main carriers ( SI Appendix , Fig. S6 ) and their accumulation at boundary layer of tip side and depletion at boundary layer of base side ( 25 , 26 ), and thus a built-in ion concentration polarization (ICP) from the tip to base side ( Fig. 1 D ). Based on the reversible redox process in PEDOT: PSS and its capacitive effect, the hysteresis could be observed following several processes ( 27 ): in the first quadrant, the external positive voltage would drive the cations from the lumen to the outside bulk solution, accompanied by an opposite flux of anion that is blocked by the negative charged membrane, leading to the salt depletion/accumulation at the base/tip side and the continuously enhanced ICP. With the scanning from 0 to +1 V and then back to 0 V, the polarization effect was time-accumulated, which would lead to a lower current value of the scanning from +1 to 0 V than that of the scanning from 0 to +1 V. In the third quadrant, the external negative voltage would drive the cations from the outside bulk solution into the lumen, accompanied by an opposite flux of anion that is blocked by the negative charged membrane, leading to the salt accumulation/depletion at the base/tip side and the continuously inhibited ICP. With the scanning from 0 to –1 V and then back to 0 V, the time-accumulated depolarization could lead to a larger current value in the scanning from –1 to 0 V than that in the scanning from 0 to – 1 V. The hysteresis characteristics were further studied at different scan rates ranging from 20 to 5,000 mV/s ( SI Appendix , Fig. S7 ). The geometric areas of the hysteresis loops were derived and plotted against the scan rate, revealing an upward trend before 100 mV/s and a downward trend thereafter ( Fig. 1 E ). Such a phenomenon was due to sufficient ionic equilibrium at low scan rates and the lagged ionic changes at high scan rates ( 28 ), which was consistent with Hodgkin–Huxley ion-channel memristors ( 29 – 31 ). It implied a time-dependent process, which could be demonstrated by sampling transient current and calculating conductance changes under constant potential. Both of current and conductance became smaller/larger under positive/negative potential until reaching a steady state after 10 s ( Fig. 1 F , Inset ), confirming the time-dependent ionic polarization/depolarization processes and thus the memristive behaviors. This property could be also verified by the current responses under different switch conditions ( SI Appendix , Fig. S8 ). Time-Dependent Synaptic Plasticity. Synaptic plasticity plays a significant role in forming learning and memory. As a form of short-term plasticity, PPD/PPF refers to the lower/higher response amplitude upon the second stimulus pulse ( A 2 ) than that of the first pulse ( A 1 ), respectively ( 32 , 33 ). These phenomena could be independently mimicked using our nanofluidic synapse. Two presynaptic pulses with +0.4 V intensity, 10 ms duration, and 20 ms interval ( Δt ) were initially applied, yielding postsynaptic current peaks of 21.70 and 19.84 nA ( Fig. 2 A ). In the scenario of −0.4 V stimuli, peaks of −28.42 and −29.04 nA could be produced ( Fig. 2 B ). The signals originated from the preservation of polarization/depolarization states when the second pulse arrived, aligning well with the time-dependent characteristic. The preservation time in such conditions could be deduced by plotting A 2 / A 1 against the Δt . As the Δt increased, the A 2 / A 1 values increased for PPD and decreased for PPF ( Fig. 2 C ). Time constants for PPD and PPF with different values could be derived, indicating the asymmetric polarization/depolarization dynamics ( Fig. 2 C , Inset ). Pair pulses with 100 ms duration were also applied ( SI Appendix , Fig. S9 ), leading to larger time constants compared to that of 10 ms. These time constants spanned from several to hundreds of milliseconds, coinciding with that of the biological system ( 34 ). Fig. 2. Time-dependent synaptic plasticity. ( A ) PPD and ( B ) PPF behaviors of the synapse under Δt = 20 ms. ( C ) A 2 / A 1 − Δt plots for PPD/PPF and the corresponding fitting curves. The inset shows the derived time constants. ( D ) SRDP signals under 50 successive pulses with +/−0.4 V intensity (+ for upper and – for lower), 10 ms duration, and different intervals, corresponding to the frequency ranging from 67 to 1 Hz (from left to right). ( E ) STDP signals and the calculated Δ Conductance against different ΔT +/− values. The input signals at +/−0.1 V were used for calculation of Δ Conductance. The short-term plasticity contributes to the temporal filtering as the change of synaptic plasticity depends upon multiple spikes with varied rates. This evolution could be simulated by applying 50 presynaptic pulses with +/−0.4 V intensity, 10 ms duration, and different intervals (corresponding to the frequency of 1 to 67 Hz) upon our nanofluidic synapse. The postsynaptic current decreased/increased obviously at high frequency and showed little changes at the low frequency, which was attributable to the accumulation of polarization/depolarization degree at high frequency ( Fig. 2 D ). STDP is another form of synaptic plasticity describing the connection change between the pre- and postsynaptic against their spikes order and interval ( ΔT ). This process could be imitated by treating a positive pulse (+1 V) as the presynaptic spike, a negative pulse (−1 V) as the postsynaptic spike, and small voltage (+/−0.1 V) of 10 ms after the latter pulse as the connection readout. When −1 V pulse first arrived ( ΔT +/− < 0), the conductance variation ( Δ Conductance) increased as the ΔT +/− came near 0 s, which was due to a higher preservation of the predepolarization state that facilitated the ionic transportation ( Fig. 2 E , Left Inset ). When +1 V pulse first arrived ( ΔT +/− > 0), the Δ Conductance decreased as the ΔT +/− came near 0 s, owing to a higher preservation of the prepolarization state that hindered the ionic transportation ( Fig. 2 E , Right Inset ). These phenomena were consistent with the antisymmetric anti-Hebbian learning rule ( 35 ). Bionic Spiking Behaviors. The biological APs operate upon integration of incoming signals and firing output spikes based on predefined thresholds, which includes several main stages ( 29 , 36 – 38 ): i) At rest, the equilibrium with the inherent ionic gradient ( Fig. 3 A , left column) maintains the transmembrane potential ( V mem ) of ca. −70 mV ( Fig. 3B stage 1). ii) Upon stimulus, the dominant sodium ion afflux ( Fig. 3 A , middle column) leads to the depolarization with V mem positive than −70 mV ( Fig. 3 B , stage 2). iii) Then, the dominant potassium ion efflux ( Fig. 3 A , right column) results into repolarization with recovery of V mem to ca. −70 mV, accompanied by a short hyperpolarization with V mem negative than −70 mV before resting ( Fig. 3 B , stage 3). Such a transmembranous electrical dynamics could be well reiterated using our nanofluidic synapse. An initial resting current of −7 nA was applied to induce the high conductance ( Fig. 3 C , left column), which could easily ordinate the V mem to ca. −70 mV ( Fig. 3 D , stage 1). The subsequent switch to the stimulus current of +7 nA allowed the cation accumulation at the boundary layer of tip side and thus the high resistance ( Fig. 3 C , middle column), which could prompt the V mem positive than −70 mV to the predefined threshold of +40 mV that is identical to that of APs ( Fig. 3 D , stage 2). Then, the automatic switch to the resting current of −7 nA led to the fast release of accumulated cations and the recovery of high conductance ( Fig. 3 C , right column), which could decrease the V mem to ca. −70 mV with a short overshoot ( Fig. 3 D , Inset and stage 3). Importantly, the energy consumption of a single potential spiking was around 77 pJ in our nanofluidic device, which was obviously lower than the nJ level of the organic electrochemical transistors ( 2 – 4 ). The spiking durability of this synapse was further studied over 1,000 cycles ( Fig. 3 E ). The little difference between the first and last 10 spikes indicated its good operational stability. Fig. 3. Bionic spiking behaviors. ( A ) Schematic of three main stages in biological generation of APs and ( B ) the corresponding V mem evolution. ( C ) Schematic of three main stages in our bionic spiking synapes and ( D ) the corresponding highly analogous V mem under input ionic currents of −/+7 nA. The Inset shows the magnified view of the V mem signals in dotted box. ( E ) Stable spiking durability over 1,000 cycles. Electrical and Chemical Tunability of the Spiking. The biological generation of APs necessitates a certain intensity of stimulation, and the weak stimuli could not successfully initiate the APs ( 39 ). This phenomenon could be emulated in our nanofluidic device through adjusting stimulus current of −5, −1, +3, and +7 nA ( Fig. 4 A ). As shown, different stimulus currents induced V mem of varying magnitudes and shapes, and only +7 nA could generate the ideal spiking property. Besides, at fixed stimulus current of +7 nA, different threshold values, e.g., +30, +40, and +70 mV, could be realized ( Fig. 4 B ) with more pronounced overshoots ( Fig. 4 B , Inset ), which could be attributed to the accumulation of more cations at the boundary layer of tip side. Moreover, at the predefined threshold of +40 mV, the level of resting potential could also be tuned ( SI Appendix , Fig. S10 ). In addition, biological APs with different frequencies could further be simulated by controlling of the resting time. With the increased time from 0.5 to 10 s, spikes with frequencies ranging from 1.72 to 0.10 Hz could be generated ( Fig. 4 C ). Spikes with larger profiles could be also achieved ( SI Appendix , Fig. S11 ). Fig. 4. Spiking tunability by electrical and chemical inputs. ( A ) Varying spiking behaviors under different stimulus currents with resting current of −7 nA, threshold of +40 mV, and resting time of 2 s. ( B ) The tunable threshold values with stimulus current of +7 nA, resting current of −7 nA and resting time of 2 s. ( C ) Tunable spiking frequencies under varying resting time with stimulus current of +7 nA, resting current of −7 nA and threshold of +40 mV. ( D ) Spiking behaviors under different potassium concentrations. ( E ) The spiking modulation by neurotransmitter dopamine. For ( D and E ), the signals were acquired with stimulus current of +7 nA, resting current of −7 nA, threshold of +40 mV, and resting time of 2 s. Insets show the corresponding I–V signals and the right columns show magnified views of the second spikes. In addition to the types of cations ( SI Appendix , Figs. S6 and S12 ), the spike characteristics of our nanofluidic device could be further modulated by the cation concentrations. For example, upon increased potassium concentration at tip side ( c K + , t ), the resting potential became less negative ( Fig. 4 D ), which was due to the increased conductance state that could be supported by the I–V curves showing larger current level under larger c K + , t ( Fig. 4 D , Inset ). A magnified view of the second stimulus reflected that more time was needed to complete a spike as the c K + , t decreased, corresponding to the decreased spike frequency. Such properties were consistent with the biological hypokalemia and hyperkalemia effects that result in hypoexcitability and hyperexcitability, respectively ( 40 ). Especially, the spiking could be reshaped by the neurotransmitter dopamine. Upon addition of 5 mM dopamine at the tip side ( c Dopa,t ), the resting potential became more negative and the I–V curves exhibited a smaller current level ( Fig. 4 E , Inset ), which was due to the decreased conductance state and the electrostatic binding between the dopamine and the PEDOT: PSS ( 41 ) ( SI Appendix , Fig. S13 ). Magnified view of the second stimulus indicated that more time was needed to complete a spike in the presence of dopamine, corresponding to the decreased spike frequency. Such a property resembled the binding of dopamine with the D2-like receptors on the cell membrane that could inhibit neuronal excitability ( 42 , 43 )."
} | 4,057 |
34372124 | PMC8347837 | pmc | 5,827 | {
"abstract": "In recent work, the thermoreversible Diels–Alder reaction between furan and maleimide functional groups has been studied extensively in the context of self-healing elastomers and thermosets. To elaborate the influence of the stoichiometric ratio between the maleimide and furan reactive groups on the thermomechanical properties and viscoelastic behavior of formed reversible covalent polymer networks, a series of Diels–Alder-based networks with different stoichiometric ratios was synthesized. Differential scanning calorimetry (DSC), dynamic mechanical analysis (DMA) and dynamic rheology measurements were performed on the reversible polymer networks, to relate the reversible network structure to the material properties and reactivity. Such knowledge allows the design and optimization of the thermomechanical behavior of the reversible networks for intended applications. Lowering the maleimide-to-furan ratio creates a deficit of maleimide functional groups, resulting in a decrease in the crosslink density of the system, and a consequent decrease in the glass transition temperature, Young’s modulus, and gel transition temperature. The excess of unreacted furan in the system results in faster reaction and healing kinetics and a shift of the reaction equilibrium.",
"conclusion": "5. Conclusions The effect of the maleimide-to-furan stoichiometric ratio on the crosslink density, glass transition temperature, mechanical properties, thermomechanical behavior (viscoelastic properties, gel transition temperature), Diels–Alder kinetics, and the self-healing behavior of a series of furan-maleimide thermoreversible networks was studied comprehensively by means of thermal analysis instruments (DSC, dynamic rheometry, and DMA) as well as using kinetics simulation. Lowering the stoichiometric ratio led to a decrease in the amount of reversible crosslinks, resulting in more flexible polymer networks with a lower glass transition temperature and Young’s modulus, an increase in the strain at break, and a decrease in the stress at break. According to kinetics simulation, by lowering the stoichiometric ratio, the gelation time decreases, while the reaction equilibrium and the critical gel conversion shift to higher conversions (at the same temperature). As a result, the decreasing stoichiometric ratio results in lower gel transition temperatures. Our work illustrated that starting from only two specific monomers, a bismaleimide (DPBM) and a furan functionalized Jeffamines (F400), a wide variety of polymer networks can be synthesized with mechanical properties ranging from very stiff thermoset to a hyperelastic elastomer that can heal at room temperature.",
"introduction": "1. Introduction In the context of step-growth polymerization of linear polymers, the stoichiometric ratio between reactive groups is well-known as a direct approach to control the number average molar mass, while more generally speaking, the combination of monofunctional and multifunctional monomers and the ratios thereof determine the polymer architecture formed [ 1 ]. For chain-growth polymerizations, the ratios between initiator, monomer and/or chain transfer agent determine the molecular weight [ 2 ]. In both linear cases and in the case of network-forming polymerizations, the thermophysical and thermomechanical properties depend on the formed polymer architectures. The Diels–Alder reaction, described by Otto Diels and Kurt Alder [ 3 ] is one of the most well-known and widely used thermoreversible equilibrium reaction to construct polymer networks with intrinsic self-healing ability [ 4 , 5 , 6 ]. The most studied Diels–Alder (DA) reaction is the cycloaddition of furan, a conjugated electron rich diene, and maleimide, an electron poor dienophile, forming a DA cycloadduct. The reverse process, called the retro Diels–Alder (rDA) reaction, converts the cycloadduct into the starting diene and dienophile [ 7 ]. Their sufficiently fast reaction kinetics and high conversion at room temperature [ 8 ] make them suitable candidates for thermoresponsive materials, such as, thermoremendable and self-healing polymer networks, which can be processed and healed at temperatures between 80 and 140 °C due to the thermoreversible crosslinking. Reversible polymer networks (RPN) constructed by DA reactions consist of covalently reversible chemical crosslinks that can be broken upon an external stimulus, mainly heat or radiation energy, and consequently damage can be healed via a heat-cool cycle below the degelation transition, while reshaping and reprocessing is feasible above degelation transition. These reversible networks have been proven to be valuable in robotics applications [ 9 , 10 , 11 , 12 , 13 ] by increasing the life-time of components through healing of macroscopic damages and as protective coatings [ 14 , 15 , 16 ]. Multiple synthesis design parameters allow tuning the mechanical and processing properties of Diels–Alder networks. To tune the mechanical properties of the dynamic covalent polymer networks for an intended application, different strategies can be used. For elastomers, it is well established how the molecular weight and flexibility of the chain segments between the crosslinking nodes affect the crosslink density and resulting mechanical properties of the polymer networks. The effect of the concentration of the furan and maleimide functional groups on the Diels–Alder reaction kinetics, thermodynamics and viscoelastic properties and thermomechanical behavior have been well studied for stoichiometric ratios by the authors and peers [ 17 ]. Examples in literature exist where the stoichiometry between maleimide and furan functional groups was varied, to study the reaction kinetics [ 8 ], to alter the mechanical properties for multi-material self-healing actuators [ 10 , 11 ], and to improve the healing performance to achieve self-healing at ambient conditions [ 12 ]. In general, increasing the concentration of the furan groups to an excess is expected to speed up the self-healing reactions and to drive the Diels–Alder equilibrium more to the bound state, however, the resulting dangling chain segments of the unreacted furans could negatively affect the mechanical behavior. Therefore, we investigated the influence of the stoichiometric ratio on the viscoelastic properties and thermomechanical behavior, including the glass transition and gel transition temperatures, Young’s modulus and stress–strain behavior, as well as on the reaction kinetics and equilibrium of synthesized reversible covalent polymer networks. It is illustrated that starting from only two specific monomers, a bismaleimide and a furan functionalized Jeffamines, a wide variety of polymer networks can be synthesized with mechanical properties ranging from very stiff to hyper elastic, by altering the stoichiometric ratio. Consequently, using only two monomers, the material properties can be fitted to meet requirements imposed by the manufacturing technique or the application.",
"discussion": "4. Discussion A series of reversible covalent polymer networks was prepared based on a four-functional furan compound (F400) and a bismaleimide (DPBM) mixed in different maleimide-to-furan stoichiometric ratios. The relation between the stoichiometric ratio and the glass transition, the equilibrium gelation temperature and conversion, the mechanical properties at ambient temperature and the thermomechanical properties from the glassy to the rubbery state were experimentally obtained. Lowering the stoichiometric ratio led to a decrease in the amount of Diels–Alder adducts (reversible crosslinks) that could be formed at ambient temperature, resulting in a more flexible polymer networks with a lower glass transition temperature and Young’s modulus, an increase in the strain at break and a decrease in the stress at break. Recently, such off-stoichiometric materials were employed by the authors to create two thermally reversible covalent elastomers with different mechanical properties that could be joined using the same reversible Diels–Alder bonding chemistry to create multi-material tendon-driven [ 11 ] and pneumatic robotic actuators [ 10 ]. In the design of these actuators, the difference in mechanical properties was exploited to achieve the desired actuation behavior. Furthermore, it was shown how the interface between the two RPN was at least as strong as the weakest of the two materials and that applied damage and interfacial failure could be healed successfully. By changing the stoichiometric ratio of the DPBM-F400 system, a wide register of thermomechanical properties can be obtained, ranging from brittle, glassy thermosets to flexible, ductile elastomers at the room temperature. This allows to optimize the mechanical properties of the network to fit specific demands of various applications. Changing the stoichiometric ratio between maleimide and furan further impacts the reaction equilibrium and the (de)gelation behavior. Lowering the stoichiometric ratio results in a shift of the reaction equilibrium towards higher conversions of the maleimide minority component. In parallel, the critical gel conversion increases with decreasing stoichiometric ratio. As the latter change is more pronounced than the shift in the equilibrium conversion, this finally results in lower gel transition temperatures at lower stoichiometric ratios. When comparing simulated and experimental gel transition temperatures, it was shown that at higher stoichiometric ratios the model is able to calculate the gel transition temperature very well, while the error becomes bigger as the stoichiometric ratio decreases, although it is still limited to 4 K at r = 0.4. Nevertheless, the simulation is an excellent tool to estimate the gel transition, proven to be useful in the determination of the temperature-time profile required for manufacturing, as illustrated by the authors [ 9 ] in a fused filament fabrication technique for an RPN based on the Diels–Alder reaction. According to isothermal simulations for the bond formation process at 25 °C, the conversion rate is higher for the lowest stoichiometric ratios, approaching the equilibrium conversion considerably faster than near-stoichiometric reactive systems. This can be explained by the large excess of furan, speeding up the reaction of the maleimide that is in deficit. In contrast, the time to reach gelation increased for lower stoichiometric ratios, due to the higher critical gel conversion. Because of this higher conversion rate, healing is favoured as well in elastomers with low stoichiometry ratio, as illustrated by the healing at 25 °C, resulting in excellent recovery of the mechanical properties. Recently, the authors published the creation and healing evaluation of a pneumatically actuated soft robotic finger that possessed the ability to self-heal many types of damage under ambient conditions [ 12 ]. The ability of this DPBM-F5000 ( r = 0.5) to heal under ambient conditions, as opposed to the stoichiometric ( r = 1) system, can be explained based on the findings of the systematic study presented here. The lower concentration of adduct bonds (reversible crosslink density) provides the higher flexibility and mobility of the polymeric chains needed to achieve efficient healing, while the increased conversion rate due to the high excess of furan groups results in fast healing kinetics to reform the broken bonds at the damage surfaces. Thermosets, with higher stoichiometry, lack mobility and fast conversion rate, making it impossible to heal at ambient conditions. Nonetheless, heating damaged samples, leads to an increase in both mobility and kinetics, resulting in the ability to heal large macroscopic damages."
} | 2,925 |
35711672 | PMC9188022 | pmc | 5,829 | {
"abstract": "Electronic skin (e-skin), a new generation of flexible electronics, has drawn interest in soft robotics, artificial intelligence, and biomedical devices. However, most existing e-skins involve complex preparation procedures and are characterized by single-sensing capability and insufficient scalability. Here, we report on a one-step strategy in which a thermionic source is used for the in situ molecularization of bacterial cellulose polymeric fibers into molecular chains, controllably constructing an ionogel with a scalable mode for e-skin. The synergistic effect of a molecular-scale hydrogen bond interweaving network and a nanoscale fiber skeleton confers a robust tensile strength (up to 7.8 MPa) and high ionic conductivity (up to 62.58 mS/cm) on the as-developed ionogel. Inspired by the tongue to engineer the perceptual patterns in this ionogel, we present a smart e-skin with the perfect combination of excellent ion transport and discriminability, showing six stimulating responses to pressure, touch, temperature, humidity, magnetic force, and even astringency. This study proposes a simple, efficient, controllable, and sustainable approach toward a low-carbon, versatile, and scalable e-skin design and structure–performance development.",
"conclusion": "3. Conclusions In summary, we presented a facile and one-step method for introducing a thermionic source to directly convert biomass materials of BC hydrogel into ionic conductive M-gel materials. This M-gel possessed a designable multiscale structure consisting of a molecular-scale H-bonding topology network and a nanoscale fiber skeleton, which was attributed to the dynamic adjustability of the competitive relationship between molecularization and self-assembly. This endowed the M-gel with superior tunability in mechanical (2.5 and 15 MPa), optical (55.68% and 94.82%), and electronic (0 and 62.58 mS/cm) properties. Inspired by the distinct perceptual structure of the human tongue, we developed a skin-like multisensory e-skin device by constructing various sensing units (similar to the receptor cells of the tongue) in this M-gel. This e-skin demonstrated excellent sensing to pressure, touch, vibration, temperature, humidity, and magnetic force, indicating its broad application prospects in flexible electronics and artificial intelligence. The design method, controllable molecularization process, and scalable bionic e-skin design technique of the proposed ionogel are expected to result in the realization of smart robots with multiple perceptual abilities and the low-carbon sustainable development of flexible electronics.",
"introduction": "1. Introduction Flexible electronics, which are portable and practicable, have flourished in recent decades [ 1 ]. Electronic skin (e-skin), which is flexible, can transduce mechanical or physical stimulations into recognizable electronic data for analysis and readout [ 2 – 4 ], showing great potential application in robotics and bioelectronics [ 5 – 13 ]. In an attempt to realize degrees of softness and comfort close to those of the human skin, chemically synthesized polymer materials with flexibility and stretchability need to be developed via complex machine processing [ 14 – 16 ]. However, these polymer substrates with poor degradability are electronic insulators and possess no ionic conductivity; thus, some conductive materials need to be introduced into the polymers by mixing, layer-by-layer stacking, or three-dimensional printing to achieve signal capture and feedback [ 15 – 18 ]. These processes are undoubtedly tedious and carbon-intensive and lead to problems such as poor interface stability (between the conductive network and the flexible substrate) and environmental pollution. As a polysaccharide polymer cellulose, bacterial cellulose (BC) is characterized by good biocompatibility, adaptability, and air permeability. As such, BC is often used as a wound dressing for human skin and tissue repair. A nanoscale BC fiber endows BC materials, such as hydrogels, with good flexibility and mechanical property (Figure S1 ). However, the internal microstructure of BC hydrogels is still not sufficiently delicate and lacks molecular-scale structure regulation and design. These qualities limit the development of functional BC-based gel materials and their application in e-skin. Utilizing ionic liquids or deep-eutectic solvents to break the hydrogen bonds between cellulose molecules, these interesting works based on hydrogen-bonding (H-bonding) topological network regulation and molecular self-assembly have been conducted on the preparation of cellulosic ionic materials [ 19 – 22 ]. For example, we used a green imidazole-based ionic liquid as solvent and reported a dynamic cellulosic ionic gel with a variable microstructure, which showed its feasibility for application in flexible, self-healing e-skin (with a good sensitivity to touch and humidity) [ 23 ]. However, despite its features of self-healing and skin-friendliness, this dynamic ionic gel involves complex fabrication steps, requires a large amount of energy, and shows poor scalability. Unlike human skin, this e-skin cannot sense temperature and pressure stimulation owing to a lack of heat- or force-sensitive factors. To achieve low-carbon and sustainable development, the design route needs to be simplified, the materials to be used should be green, and multifunctional panels have to be constructed for fabricating multisensory e-skin. Considering the aforementioned challenges, we propose an in situ molecularization strategy for introducing a thermionic liquid of 1-butyl-3-methylimidazolium chloride ([Bmim]Cl) into a monolithic BC hydrogel to convert cellulose fibers into molecular chains ( Figure 1(a) ). This facile and one-step strategy enables the scalable production of molecularized gel materials with both excellent ionic conductivity and high mechanical strength from resource-abundant biomass materials. By controlling the stimulation time of the thermionic source to coordinate the reaction–diffusion relationship between ions, water, and cellulose, this molecularized ionogel (called M-gel) owns a blended multiscale structure: a molecular-scale H-bonding topological network and a nanoscale fiber skeleton ( Figure 1(a) ). This structure endows the M-gel with good transparency ( Figure 1(b) ), flexibility, tunable mechanical performance, and high ionic conductivity reaching 62.58 mS/cm, which is superior to existing ionic gel materials ( Figure 1(c) ) [ 24 – 28 ]. In addition, the degree of molecularization (DM) of M-gel can be quantitatively designed and regulated to be between 0 and 100% by controlling the thermionic treatment time. Inspired by the distinct sensing structure of the human tongue, we developed an e-skin with a multisensory behavior by designing and integrating the respective perception patterns in this M-gel ( Figure 1(d) ). This e-skin device, as a proof-of-concept demonstration, showed ideal multistimulus responsiveness and recognizability to pressure, touch, temperature, humidity, magnetic force, and astringency (this stimulus is only perceived by the tongue).",
"discussion": "2. Results and Discussion 2.1. Design, Construction, and Characterizations of the M-Gel As a cellulose material, the BC hydrogel possesses a nanoscale fiber structure (Figure S1 ) and has water content reaching 40 wt.%. Water molecules (H 2 O) show H-bonding with the hydroxyl of the BC fiber, conferring excellent flexibility and mechanical performance on the BC hydrogel (Figure S1 ). However, when the thermionic sources of [Bmim] + and Cl − are introduced into the BC hydrogel (Figure S2 ), the diffusion-driven Turing instability between cellulose, H 2 O, and ions occurs in the system [ 23 ]. On the one hand, the ions of [Bmim] + and Cl − (as the activator) prefer to interact with BC fibers and form H-bonding interactions with the hydroxyl protons of cellulose to destroy the intermolecular/intramolecular H-bonding network and thus obtain a molecular chain (referred to as a “molecularization”); meanwhile, H 2 O (as the inhibitor) tends to prevent this behavior by inducing the formation of H-bonding between cellulose molecules (referred to as the “self-assembly,” Figure 1(a) ). By controlling the stimulation time, we can expediently adjust the content of H 2 O (from 39.16 wt.% to 21.01 wt.%) and ions (around 70 wt.%, Figure 2(a) ), thus realizing the dynamic regulation of this competitive relationship between molecularization and self-assembly to design the structure and property of the M-gel material ( Figure 2(b) ). To observe the effect of this in situ molecularization process, a comparative experiment was performed on a piece of BC hydrogel. The upper right part of the hydrogel was treated with the thermionic source, unlike the lower left part ( Figure 2(c) ). After heat treatment at 80°C for 10 min, the originally white and opaque BC hydrogel achieved slight transparency. For 30 min, a transparent M-gel appeared, with light transmittance exceeding 85% (Figure S3 ). However, the part of the BC hydrogel without thermionic source treatment showed no intuitive changes. Observation of the X-ray diffraction (XRD) spectrum revealed that the M-gel (treated with a thermionic source for 30 min) exhibited a sharp crystalline peak similar to that of the BC hydrogel ( Figure 2(d) ), indicating that this M-gel retained the ordered crystalline regions. SAXS was performed to detect subtle changes in the topological network. As shown in Figure 2(e) , the peaks from the M-gel appear sharper than those of the BC hydrogel; in Figure 2(f) , the 2D SAXS scattering signal of the M-gel also exhibits a sharper pattern than that of the BC hydrogel. These results indicate that the M-gel had a molecular-scale topological network. Visualization by Fourier transform infrared spectroscopy (FTIR) ( Figure 2(g) ) reveals the sharp peak of the M-gel at 3400 cm −1 , ascribed to the stretching vibration behavior of hydroxyl, which indicates the rich H-bonding network of the M-gel. These rich H-bonds, high crystallization, and molecular-scale topology network can undoubtedly endow the M-gel material with various beneficial properties. 2.2. Morphological Structure Design and Regulation To better demonstrate the multiscale structure of the M-gel, analytical procedures, such as scanning electron microscopy (SEM) and atomic force microscopy (AFM), were conducted to examine the morphology and microstructure of the M-gel. Compared with the BC hydrogel consisting of only cellulose fibers, the M-gel treated via molecularization showed an evident multiscale structural feature, with both a dense interwoven layer and a nanoscale fiber with different diameters ( Figure 3(a) ). Using AFM and images, we also examined this densely layered structure, which was flawlessly embedded in the nanoscale fiber skeleton, forming a seamless whole ( Figure 3(b) ). This combination can confer strong mechanical properties and performance durability on the M-gel material. This distinct morphological structure in the M-gel was derived from the competitive behavior between molecularization (turning fibers into molecular chains) and cellulose-molecule self-assembly (constructing the intermolecular H-bonding topological network, Figure 3(c) ). The thermionic source changed the original and rigid situation between cellulose and H 2 O, prompting the system to exhibit dynamic and tunable behavior patterns and performances. Notably, we can design the morphology and microstructure of the M-gel from a multiscale-coexisting body (both molecular self-assembly and nanofibers) to a fully interwoven dense structure by controlling the molecularization time of the thermionic source (insets of Figure 3(d) ). Meanwhile, by calculating and analyzing the change in the light transmittance value (T %, Figure S3 ) of the M-gel over different molecularization periods, we can quantify and design the degree of molecularization (DM) of the M-gel. As shown in Figure 3(d) , when treated with the thermionic source for 5, 10, 30, 50, and 70 min, the DM of M-gel can be controlled at 7.08%, 32.38%, 81.52%, 90.5%, and around 100%, respectively. In addition, the crystallinity of M-gel also exhibits a designability between 85% and 50.4% (Figure S4 ). So, this manipulation is controllable, customizable, and scalable; in addition, it is not only applicable to the BC hydrogel but can also be extended to other cellulosic materials such as filter paper and printing paper (Figure S5 ). To our knowledge, this structural design strategy for the ionogel has not been reported in the literature. 2.3. Excellent Mechanical Performance and Ionic Conductivity The M-gel presented excellent flexibility, transparency, and stretchability ( Figure 4(a) ), showing free and perfect switchability between stretching and recovery states (Figure S6 ). This M-gel can also be closely attached to human wrists and fingers under various large bending strains ( Figure 4(b) and Figure S7 ). For an ionic gel material used in e-skin, biocompatibility (for human body), self-healing, and noncorrosiveness (for artificial limb) are important performance indicators. As shown in Figure S6 and S8 , our M-gel shows good skin-friendliness (without tissue damage and inflammation when adhered to the human wrists for more than 10 h) and rapid self-healing (80°C for 30 min) and is not invasive to an artificial hand. On the basis of the tensile stress–strain tests ( Figure 4(c) ), the mechanical properties of the M-gel can be regulated, from a strength of 15 MPa (high) to 2.5 MPa. After molecularization for 30 min, the M-gel (referred to as M-gel-30) had a tensile strength of up to 7.8 MPa ( Figure 4(d) ), an elastic modulus of up to 10.2 MPa, and work of fracture of about 1.87 mJ m −3 ( Figure 4(e) ); these properties were superior to those of some high-performance cellulosic ionic gels [ 22 , 23 ]. Meanwhile, M-gel-30 with robust toughness could easily lift a weight of 2 kg (over 50,000 times its weight, Figure 4(e) ). Even with complete molecularization after thermionic treatment for 70 min, the M-gel still achieved a perfect tensile strength of 2.5 MPa, similar to that of human cartilage, and thus exhibited potential for application in artificial tissue for robots. The introduction of the thermionic source endowed the M-gel with an ion-rich environment, which possessed programmable ionic conductivity. On the basis of the electrochemical impedance spectroscopy (EIS) curves in Figure 4(f) , we calculated the ionic conductive behavior of the M-gel during different molecularization processes. As shown in Figure 4(g) , the M-gel-30 achieves the highest ionic conductivity reaching 62.58 mS/cm, which is largely attributed to its multiscale structures. First, the nanoscale fiber skeleton provided a smooth linear path for transporting ions. Second, the molecular scale H-bonding topology network conferred abundant electrostatic-attraction sites for accelerating ion diffusion. Owing to excellent ionic conductivity, M-gel-30 as a flexible conductor can easily illuminate a light-emitting diode (LED) and shows satisfactory antistrain performance (Figure S9 ). In addition, M-gel-30 presented good conductive stability in an open environment with a relative humidity of ≈45% for more than 30 d (Figure S10 ). Even after 50 cyclic folding at 180°, the M-gel-30 still presents the high conductivity of 57 mS/cm and structural stability without any breakage and damage (Figure S11 ). To further demonstrate the excellent mechanical and electronic properties, we compared this M-gel with numerous ionic gel materials and found that both properties were superior ( Figure 4(h) ) [ 26 – 38 ]. 2.4. Applications of the M-Gel in Constructing Multisensory e-Skin The human tongue as a soft and sensitive organ is highly capable of simultaneously sensing various taste stimuli (sour, sweet, bitter, spicy, and astringent) [ 39 ]. This function is attributed to the tongue possessing diverse perception areas (consisting of receptor cells and connective tissue, Figure 5(a) ). Receptor cells buried in connective tissue have a regional structure and perception, which can distinguish the external stimuli and convert them into corresponding ion pulses. The connective tissue as a flexible conductor then transmits these ion signals to the nerve center of the human body, producing accurate behavioral feedback. The proposed M-gel exhibits good flexibility, high ionic conductivity, and performance stability. It has a superior capability to transport ions, similar to that of the connective tissue of the tongue. In addition, owing to its rich active hydroxyl groups freed by molecularization, the M-gel showed an outstanding adhesive performance with strengths of up to 3.93 N (Figure S12 ). Inspired by the structure-feature of the tongue, we innovatively used our ionic conductive M-gel to mimic the connective tissue of the tongue while selecting the sensing materials of poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS, sensing temperature) [ 40 ], carbon nanotubes (CNTs, sensing pressure and deformation), Ag nanofiber (AgNWs, sensing deformation) [ 40 ], ionic gel (sensing humidity) [ 23 ], and nanonickel powder (sensing magnetism and temperature) to mimic the receptor cells (Figure S13 ). We successfully developed a flexible, transparent, and multisensory e-skin device on the basis of a simple brushing process in Figure 5(b) . Notably, this e-skin was easily formed into large sizes, molded into diverse shapes, or imparted with various stimulation receptors, depending on the demand. As shown in Figure 5(c) , this multisensory e-skin exhibits adequate flexibility and seamless interface adhesion to human wrists. Moreover, this e-skin showed excellent structural integrity and stability even after several contacting abrasion and water immersion experiments (Figure S14 ) because of the strong H-bonding and coordination behavior between cellulose molecules and nanomaterials. Owing to its high ion conductivity and structural superiority, the e-skin, as a flexible electronic, presented ideal signal feedback to multistimulations, including touch vibration ( Figure 5(d) ), pressure ( Figure 5(e) ), magnetic force ( Figure 5(f) ), temperature ( Figure 5(g) ), and humidity ( Figure 5(h) ). In addition, this multisensory e-skin exhibited discernible current signal curves with excellent sensitivity, discriminability, and repeatability (Figure S15 ). These attributes indicate that this e-skin can easily distinguish different stimulation behaviors (such as touch, temperature, air flow, humidity, and even magnetic fields) by analyzing the magnitude and area of the produced electrical signal waveform. Real human skin can actively sense varied information from the external environment. Notably, as shown in Figure 5(i) , our e-skin can also simultaneously sense various stimuli, such as vibration, pressure, temperature, magnetic force, and humidity; it also demonstrates good repeatability and recognizability ( Figure 5(i) ). Through the observation of waveform signals (Figure S16 ), we can conveniently identify the intensity such as strain, pressure, temperature, and magnetic force. This is an important feature for an e-skin that is expected to rival human skin. By integrating the porcine gastrointestinal proteins into the proposed M-gel, the obtained e-skin can also clearly capture the astringent stimuli from citric acid ( Figure 5(f) ), which is difficult to achieve with other e-skins [ 41 ]. These features indicate that this e-skin based on M-gel is expected to help robots acquire a soft appearance and exhibit multiple perceptual capabilities for enhancing their uses in real-life settings."
} | 4,946 |
24229321 | PMC4176505 | pmc | 5,831 | {
"abstract": "Background The rapid determination of the release of structural sugars from biomass feedstocks is an important enabling technology for the development of cellulosic biofuels. An assay that is used to determine sugar release for large numbers of samples must be robust, rapid, and easy to perform, and must use modest amounts of the samples to be tested. In this work we present a laboratory-scale combined pretreatment and saccharification assay that can be used as a biomass feedstock screening tool. The assay uses a commercially available automated solvent extraction system for pretreatment followed by a small-scale enzymatic hydrolysis step. The assay allows multiple samples to be screened simultaneously, and uses only ~3 g of biomass per sample. If the composition of the biomass sample is known, the results of the assay can be expressed as reactivity (fraction of structural carbohydrate present in the biomass sample released as monomeric sugars). Results We first present pretreatment and enzymatic hydrolysis experiments on a set of representative biomass feedstock samples (corn stover, poplar, sorghum, switchgrass) in order to put the assay in context, and then show the results of the assay applied to approximately 150 different feedstock samples covering 5 different materials. From the compositional analysis data we identify a positive correlation between lignin and structural carbohydrates, and from the reactivity data we identify a negative correlation between both carbohydrate and lignin content and total reactivity. The negative correlation between lignin content and total reactivity suggests that lignin may interfere with sugar release, or that more mature samples (with higher structural sugars) may have more recalcitrant lignin. Conclusions The assay presented in this work provides a robust and straightforward method to measure the sugar release after pretreatment and saccharification that can be used as a biomass feedstock screening tool. We demonstrated the utility of the assay by identifying correlations between feedstock composition and reactivity in a population of 150 samples.",
"conclusion": "Conclusions In this work we have presented a method to perform a two-step pretreatment-saccharification assay to measure the sugar release (mass sugar released per unit of oven-dry mass of biomass) of biomass samples. The assay uses a commercially available automated solvent extraction system for pretreatment followed by a small-scale enzymatic hydrolysis step. The assay allows multiple samples to be screened simultaneously, and uses only approximately 3 g of biomass per sample. If the composition of the biomass sample is known, the results of the assay can be expressed as sugar yield (fraction of structural carbohydrate present in the biomass sample released as monomeric sugars). We first investigated the effect of pretreatment conditions on sugar release and yield for four representative biomass samples, and then screened more than 150 different biomass samples using the technique. The results of the screening work showed a statistically significant negative correlation between lignin content and total sugar release and reactivity. This result suggests that lignin may interfere with sugar release, or that more mature samples (with higher structural sugars) may have more recalcitrant lignin. Regardless of the cause of this phenomenon, the assay allowed us to identify the correlation in an aggregate of 150 samples, demonstrating that it can be a powerful tool to investigate differences in cellulosic feedstock samples.",
"discussion": "Results and discussion A note on nomenclature and calculations For any assay there is typically more than a single way to express the results of the assay. In this work we report the results in terms of sugar release and sugar yield. We define sugar release as the mass of glucose and/or xylose (the sum of monomeric and oligomeric) released per unit mass of biomass starting material. We define xylan yield as the fraction of xylan released as xylose, including the anhydro correction factor; the ratio of the hydrolyzed sugar monomer to the anhydrous structural repeat unit. We define the glucan yield similarly, except we also include contributions from non-structural glucose (from sucrose) and starch. Thus, both the glucan and xylan yield calculation require knowledge of the composition of the starting material. We define reactivity as the total yield of glucose and xylose from the total structural and nonstructural carbohydrates. For pretreatment experiments, the concept of reactor severity [ 24 , 26 , 29 , 30 ] or severity factor is commonly used to characterize the combination of time and temperature in a given reactor. Severity is a nonlinear combination of reactor temperature and time. We did not use severity in this work. For all experiments the heating and static times in the ASE 350 reactor were held constant (7 minutes and 6 minutes, respectively), while the cell temperature varied. Although the severity factor could be calculated a number of ways (the static time alone, sum of the heating and static times, the sum of the fraction of the heating time above a certain temperature and the static time), the results would be the same for all screening data presented in this work. Because the calculation of severity explicitly excludes any consideration of the ratio of acid to biomass, it is unlikely that other reactor geometries operating at the same severities used in this work (for example, a flow-through reactor with near-instantaneous heating or a microwave system), or even the same reactor system operated in a different flow mode would provide identical conversion data. We believe the concept of severity is most useful for interpreting data in a single-reactor geometry and operating mode. All release, yield, and reactivity calculations are included as Additional file 1 . As mentioned above, the primary results of the enzymatic hydrolysis assay are reported as the mass of sugar released per unit mass of biomass added. The mass of sugar released is calculated as the product of a concentration measurement and a volume measurement. This calculation is an approximation; we assume the liquid volume of the system to be 10 mL, which is an overestimation of the actual liquid volume, as some solid residue remains in the flask at the end of enzymatic hydrolysis. To determine the results of enzymatic hydrolysis for larger-scale systems, we would perform fraction insoluble solids (FIS) measurement [ 31 ] to determine the liquid volume exactly. However, the small amount of material used for this assay, and the need for a high-throughput assay itself, makes the FIS measurement infeasible at this scale. Since the actual liquid volume is less than our estimate of 10 mL (as for example, lignin will remain a solid at the end of the experiment) we recognize that the enzymatic hydrolysis assay slightly overestimates the sugar release values. Accuracy and precision of assay A well-characterized corn stover material was used as an internal method validation sample for both the pretreatment and enzymatic hydrolysis assays. Including replicates, this sample was assayed 21 times in 14 separate pretreatment batches and 12 subsequent enzymatic hydrolysis batches (the batch size for enzymatic hydrolysis was slightly larger than for pretreatment). We can use these repeated data to measure the variability of the method. The data in Figure 1 show the glucose and xylose release from repeated assays of this material. Figure 1 a and b show the results of the pretreatment and enzymatic assay separately, and Figure 1 c shows the results of the combined assay. The xylose release was higher than glucose release during pretreatment and the opposite was true for enzymatic hydrolysis, completely expected given the purpose and biochemistry of the two assays. The mean and standard deviations of the key results from these replicated assays are shown in Table 1 . For the overall assay, the mean xylose release from the corn stover control was 0.228 ± 0.007 g/g and the mean glucose release from the corn stover control was 0.309 ± 0.021 g/g. The overall assay has a higher uncertainty for glucose release than for xylose release because the enzymatic assay, in which most of the glucose is released, has a higher variability than the pretreatment assay. Regardless of the source(s) of the variability in these measurements, they limit the ability of the combined assay to identify differences between and among different sample types. We discuss this issue in detail below. Figure 1 Results of method control experiments using corn stover control material. Measured glucose and xylose release from (a) pretreatment assay, (b) enzymatic hydrolysis assay, and (c) combined pretreatment (PT) + enzymatic hydrolysis (EH) assay. As expected, the majority of xylose was released during pretreatment whereas the majority of glucose was released during enzymatic hydrolysis. The variability of the glucose release (SD = 2.1%) was greater than the variability of the xylose release (SD = 0.7%). The variability of these replicates limits the discriminating power of the assay. Table 1 Mean and standard deviations of replicate measures of xylose release from pretreatment, glucose release from enzymatic hydrolysis, and overall xylose and glucose release from the combined pretreatment and enzymatic hydrolysis assay for the corn stover control sample used for each assay batch, and the pooled standard deviation for all replicated samples Xylose release (PT) Glucose release (EH) Xylose release (PT + EH) Glucose release (PT + EH) Mean SD Mean SD Mean SD Mean SD Corn stover control samples 0.160 0.007 0.476 0.022 0.228 0.007 0.309 0.021 All replicated samples -- 0.009 -- 0.014 -- 0.009 -- 0.011 The mean of all replicated samples is meaningless in this context, and is not shown. PT, pretreatment; EH, enzymatic hydrolysis. The pretreatment/enzymatic hydrolysis assays as variables The results reported here are for dilute acid pretreatment followed by enzymatic hydrolysis. As mentioned previously, other pretreatment chemistries are widely used (for example, steam explosion, ammonia, hot water, and ionic liquid), and different enzyme formulations are typically used for different pretreatment chemistries. For example, Kumar and Wyman [ 32 ] observed that xylanase supplementation during enzymatic hydrolysis was more beneficial for some pretreatments than for others. This is not a limitation of the pretreatment process, but rather a recognition that pretreatment chemistries differ; the greater the amount of xylan removed during pretreatment, the less helpful additional xylanase supplementation during enzymatic hydrolysis. Similarly, enzymatic hydrolysis enzymes are subject to inhibition and deactivation by compounds produced during pretreatment [ 33 ]. For this work, we used a single pretreatment chemistry (dilute sulfuric acid) and a single enzyme formulation applied at a high mass loading in order to focus on the effect of one major pretreatment variable (temperature) on observed sugar yields. The reader is reminded that the selection of the pretreatment chemistry, other pretreatment variables (for example, acid-biomass ratio, reaction time, biomass particle size), the enzyme formulation, loading, and conditions used for enzymatic hydrolysis are in fact variables that could be investigated separately. Such an extensive study is outside the scope of this work. Variable temperature pretreatment/enzymatic hydrolysis experiments The results of the variable temperature pretreatment experiments are summarized in Figures 2 , 3 , 4 , and 5 . The total (monomeric and oligomeric) xylan yields from pretreatment alone and from pretreatment followed by enzymatic hydrolysis for four different biomass feedstocks (corn stover (a), poplar (b), switchgrass (c), and sorghum (d)) across a range of pretreatment temperatures are shown in Figure 2 . The smooth curves through the data are cubic splines added to guide the eye of the reader; they do not represent a theoretical model. There are a large number of points at 130°C for the corn stover plot (Figure 2 a). These are the method controls, shown as a time series in Figure 1 . Figure 2 Total (sum of monomeric and oligomeric) xylan yield (fraction of xylan originally present in biomass feedstock released as xylose) from dilute acid pretreatment alone (hollow symbols) and pretreatment followed by enzymatic hydrolysis (filled symbols) as a function of pretreatment temperature. (a) Corn stover (b) ; poplar; (c) switchgrass; (d) biomass sorghum. All pretreatment experiments were performed with 3 g (dry weight) biomass, 30 mL of 1% sulfuric acid with a 7-minute heating time and a 6-minute static time in a 66-mL zirconium cell. All enzymatic hydrolysis experiments were performed at 10% solids using an enzyme loading of 40 mg/g biomass. For all feedstock types the maximum xylose yield occurs at a temperature of 150°C or 160 C, respectively. However, the maximum difference between the highest and lowest maximum yield values (corn stover and poplar) occurs at a reactor temperature of 130°C. Figure 3 Total (sum of monomeric and oligomeric) glucose release (fraction of biomass feedstock sample dry weight released as glucose) from dilute acid pretreatment alone (hollow symbols) and pretreatment followed by enzymatic hydrolysis (filled symbols) as a function of pretreatment temperature. (a) Corn stover; (b) poplar; (c) switchgrass; (d) biomass sorghum. Experimental conditions were the same as in Figure 2 . Some glucose was released during pretreatment; the horizontal line in each plot shows the amount of non-structural glucose, derived from starch and sucrose (see text). Figure 4 Total (sum of monomeric and oligomeric) glucan yield (fraction of glucan originally present in biomass feedstock, both structural and nonstructural, released as glucose) from dilute acid pretreatment alone (hollow symbols) and pretreatment followed by enzymatic hydrolysis (filled symbols) as a function of pretreatment severity. (a) Corn stover; (b) poplar; (c) switchgrass; (d) biomass sorghum. Experimental conditions were the same as in Figure 2 . The yield calculations are based on both structural and non-structural contributions (for example, cellulose, starch, sucrose; see text). Figure 5 Total reactivity (fraction of glucan and xylan originally present in biomass feedstock, both structural and nonstructural, released as glucose or xylose) from dilute acid pretreatment alone (hollow symbols) and pretreatment followed by enzymatic hydrolysis (filled symbols) as a function of pretreatment severity. (a) Corn stover; (b) poplar; (c) switchgrass; (d) biomass sorghum. Experimental conditions were the same as in Figure 2 . The yield calculations are based on both structural and non-structural contributions (for example, cellulose, starch, sucrose; see text). For all feedstocks, the xylan yield values showed a maximum at a reaction temperature of 150°C or 160°C. Higher reaction temperatures resulted not only in a reduction of xylose release, but in the formation of furfural, indicating nonproductive conversion of the xylose (data not shown). There was variation in the maximum total xylose yield among the four feedstock types, with corn stover and switchgrass having yields over 90% and sorghum and poplar having yields closer to 80%. For pretreatment temperatures below approximately 170°C, enzymatic hydrolysis releases additional xylose. The amount of additional release was greatest at the lowest pretreatment temperature, although the difference was less pronounced for switchgrass and poplar. The total (monomeric and oligomeric) glucose release from pretreatment alone and from pretreatment followed by enzymatic hydrolysis as a function of pretreatment temperature for the four different feedstocks are shown in Figure 3 . As expected, most of the glucose was released during enzymatic hydrolysis. The general trends for glucose release in this figure are similar to the trends for the xylose yield data in Figure 2 , except the maximum glucose release occurred at a pretreatment condition of either 160°C or 170°C. As the pretreatment temperature increased, the amount of glucose released from the pretreatment step alone increased. Note also the repeated data points at 130°C for the corn stover plot (Figure 3 a), showing higher variability than the corresponding xylose release data. The horizontal line in each plot in Figure 3 is the amount of nonstructural glucose present in the materials, the sum of the glucose contributions from soluble sucrose and starch (after anhydro correction). Note that in all cases, the glucose release during pretreatment exceeded this value, suggesting that some structural glucan was released during pretreatment, likely from hemicellulose depolymerization. Figure 4 shows the glucan yield from pretreatment alone and from pretreatment followed by enzymatic hydrolysis as a function of pretreatment temperature for the same four feedstocks. Again, the yield included structural and nonstructural glucose sources. These glucose yield data represent the overall yields from both pretreatment and enzymatic hydrolysis, but must be viewed with caution. First, as discussed previously, the enzymatic hydrolysis assay overestimates the glucose yield; it is unlikely that we have achieved almost 100% glucose yield from the corn stover and over 100% glucose yield from sorghum. Note also that the yield calculations require knowledge of the composition of the feedstocks; these data are unlikely to be available as part of large-scale feedstock screening because of the time-intensive nature of biomass compositional analysis. The combined yield of xylan and glucan is commonly referred to as reactivity. The reactivity of the four feedstocks as a function of pretreatment temperature is shown in Figure 5 . Optimal pretreatment conditions for screening The goal of a high-throughput screening assay is to discriminate among different feedstock samples based on one or more criteria. For this work, we sought an assay that would provide information about the relative merit of different feedstocks with respect to cellulosic sugar production, as measured by the sugar release and yield after pretreatment and enzymatic hydrolysis. The generation of a complete reactivity versus pretreatment temperature profile, such as shown in Figure 5 for each sample to be screened, would multiply any screening effort by an order of magnitude compared to a single-point screening, and varying the enzymatic hydrolysis conditions (for example, enzyme loading) would add another order of magnitude. It seems necessary to pick a single pretreatment and enzymatic hydrolysis condition for screening. How is this condition to be selected, and what should be measured? In Figure 6 we compare the results of the combined assay for the four feedstocks for (a) xylose release, (b) xylan yield, (c) glucose release, (d) glucan yield, (e) total sugar release, and (f) total sugar yield (reactivity) at 130°C and 150°C. These data show that differences among the four model feedstocks depend in part on which of these six values is used to discriminate, and at what temperature the pretreatment portion of the assay is performed. Figure 6 Effect of pretreatment temperature on results from combined pretreatment-enzymatic hydrolysis assay for four representative feedstocks. (a) Total xylose release; (b) total xylan yield; (c) total glucose release; (d) total glucan yield, (e) toal sugar release, (f) toal sugar yield. Larger differences in the experimental results among the four representative feedstocks are apparent at a pretreatment temperature of 130°C than at a pretreatment temperature of 150°C. From the xylose release data in Figure 6 a, we note that xylose release was higher for all four feedstocks at the higher temperature, but that a relative ranking of the four feedstocks for xylose release would be the same at either temperature. The lower pretreatment temperature of 130°C provides the largest differences among the four feedstocks for xylose release, but the higher pretreatment temperature of 150°C provides a higher release, more typical for a biomass conversion process. Thus, if the measured assay value is the xylose release, running the assay at either temperature provides the same sample ranking, and the lower temperature may provide better discrimination among samples. When the measured assay value was the xylan yield (Figure 6 b), the ranking of the samples at 130°C was the same, but at 150°C, the ranking results were slightly different; the samples with the two highest xylose releases now had essentially equivalent xylan yields. It would be much more difficult to identify a statistically significant difference between corn stover and switchgrass at the higher temperature based on xylan yield. When the measured assay value was the glucose release (Figure 6 c) we saw the same relative ranking at the lower temperature (130°C) as in Figure 6 a and Figure 6 b. However, at the higher temperature (150°C), the glucose release from the poplar and switchgrass were equivalent. When the measured assay value was the glucan yield (Figure 6 d) we saw slightly different behavior than for glucose release. At both temperatures, it was difficult to distinguish between the sorghum and switchgrass samples, while the glucan yields from corn stover and poplar were distinct. The total sugar-release data in Figure 6 e show similar discriminating results as in Figure 6 c, and the reactivity data in Figure 6 f show similar results as in Figure 6 d. As glucose is the major sugar in these feedstocks, similar results were expected when the metric was based on glucose and then on total sugar (glucose plus xylose). In general, the data in Figure 6 show that the higher pretreatment temperature (150°C) increased the sugar release and yield data to values more representative of actual process conditions, but also reduced the spread in these data, making discrimination among samples more difficult. We chose to perform the screening experiments using a pretreatment temperature of 130°C, because differences among the four model feedstock samples were greater at this temperature than at 150°C. It could be reasonably argued that a pretreatment temperature of 150°C would have been a better choice for exactly the same reason; differences among feedstock samples would have been harder to identify, making any identified differences more likely to translate to real differences in a full-scale process. However, our goal for the screening experiments was to attempt to identify significant differences in reactivity among different feedstock types, so we believe the lower temperature pretreatment assay was a better choice for this particular screening experiment. Screening experiments The results of the compositional analysis and screening experiments are shown in Figure 7 , with the results separated by the five species tested: corn stover (CS), cool season grasses (CSG), miscanthus (MS), sorghum (SG), and switchgrass (SW). Plots (a) and (b) show the xylan and glucan content of the samples, plots (c) and (d) show the xylose and glucose release, plots (e) and (f) show the xylan and glucan yield, and plots (g) and (h) show the total sugar release and reactivity. Figure 7 Results from combined pretreatment-enzymatic hydrolysis assay for 156 feedstock samples. (a) Xylan content; (b) glucan content; (c) total xylose release; (d) total glucose release; (e) total xylan yield; (f) total glucan yield; (g) total sugar (glucose + xylose) release; (h) reactivity. Samples are grouped by feedstock type: CS, corn stover; CSG, cool season grass; MS, miscanthus; SG, sorghum; SW, switchgrass. There was much variation in structural carbohydrate content and all six measures of reactivity within the feedstock types. Comparing the compositional analysis results in Figure 7 , we see the corn stover, sorghum, and switchgrass samples had larger variability in xylan content compared to the cool season grasses and miscanthus samples. Differences in the population means for xylan content were not statistically significant (Tukey honestly significant difference (HSD) test, P = 0.05) for the cool season grasses and either corn stover and sorghum (CSG-CS, CSG-SG), and for corn stover and miscanthus (CS-MS); all other between-species differences were statistically significant. Similar results are apparent for the glucan data, with the sorghum and corn stover samples being most variable, and the cool season grasses and miscanthus samples being least variable. The miscanthus samples were highest in glucan, and the cool season grasses the lowest (Tukey HSD, P = 0.05). The population means for the glucan content of sorghum, switchgrass, and corn stover were not statistically significantly different (Tukey HSD, P = 0.05); all other species differences were statistically significant. The xylose and glucose release and yield, as well as the total sugar release and reactivity data showed the miscanthus samples to be less variable than all the other species. The yield data were generally (but not uniformly) more variable than the release data. The cool season grasses had the highest glucan yields and the miscanthus samples the lowest ( P = 0.05); other inter-species differences in total sugar release and reactivity are less clear. The generalizations and conclusions from the data (shown in Figure 7 ) discussed above are with respect to the populations we studied. While the miscanthus samples in this study were higher in glucan than the cool season grasses in this study, we do not conclude that this is generally true about miscanthus and cool season grasses. We do not have a detailed understanding of the genotypic or agronomic history of these samples (for example, germplasm, growing conditions). A rapid screening method like the one presented here is most useful when combined with detailed agronomic history of the samples being screened. Looking at the data in aggregate provides further insight regarding these samples. In Figure 8 a we see that the glucan and xylan content of the samples were generally correlated. The miscanthus (MS) samples showed the least compositionally variability, while the corn stover (CS) samples showed the greatest variability in glucan content and the switchgrass samples showed the greatest variability in xylan content. The lignin content of the samples was generally correlated with the sum of glucan and xylan content (Figure 8 b). The miscanthus samples were highest in lignin while the cool season grass samples were lowest. The effect of lignin content on total sugar release and reactivity are shown in Figure 8 c and d. As the lignin content increased the total sugar release and reactivity decreased. Clearly lignin content was positively correlated with glucan and xylan content, but negatively correlated with total sugar release and yield. All of these correlations were statistically significant ( P = 0.05). The positive correlation with the structural sugars was due to the increase in total structural materials and the corresponding decrease in extractives. The negative correlation between lignin and total sugar release and reactivity suggests lignin may interfere with sugar release during the assay, or that more mature samples (with higher structural sugars) may have more recalcitrant lignin. This is consistent with results seen for poplar [ 34 ] but not with results seen for miscanthus [ 35 ]. Since increased lignin content was strongly correlated with increased structural carbohydrates, the decreased yields at higher structural carbohydrate content may be related to product inhibition during enzymatic hydrolysis [ 35 ]. Regardless of the cause of this phenomenon, we were able to identify it only because of the availability of the aggregate data, supporting the utility of a high-throughput pretreatment/enzymatic hydrolysis assay. Figure 8 Correlation of compositional and assay results. (a) Glucan content versus xylan content; (b) sum of glucan and xylan content versus lignin content; (c) total sugar release versus lignin content; (d) reactivity versus lignin content. Samples are colored by feedstock type: CS, corn stover; CSG, cool season grass; MS, miscanthus; SG, sorghum; SW, switchgrass. Increasing lignin content was correlated with increasing xylan and glucan content, but also with decreasing total sugar release and reactivity (see text). We performed triplicate experiments on approximately 10% of the 155 samples, and the standard deviation of these replicates (pooled by feedstock type) is shown in Table 1 . The glucose release and yield data show approximately the same variability as the corn stover control samples discussed earlier, suggesting that we executed the combined assay consistently throughout the course of the work. The use of replicated samples for a high-throughput assay provides another measure of the precision of the assay over time. However, the combined pretreatment and enzymatic hydrolysis assay presented in this work does provide precise and robust data on sugar release for cellulosic biomass feedstocks. With careful selection of experimental conditions in each of the two assays, and with a good understanding of the agronomic backgrounds (pedigree) of the samples, the assay can be a powerful tool to investigate differences in cellulosic feedstock samples. As mentioned previously, there have been a number of reports on the use of secondary spectroscopic methods to predict the release of sugars. Although this work reports only the results of the primary assay itself, we have successfully built near-infrared (NIR) calibration models for both composition and reactivity for the samples presented in this work; we will report on this assay in the near future."
} | 7,517 |
34335645 | PMC8320662 | pmc | 5,832 | {
"abstract": "Most plants living in tropical acid soils depend on the arbuscular mycorrhizal (AM) symbiosis for mobilizing low-accessible phosphorus (P), due to its strong bonding by iron (Fe) oxides. The roots release low-molecular-weight organic acids (LMWOAs) as a mechanism to increase soil P availability by ligand exchange or dissolution. However, little is known on the LMWOA production by AM fungi (AMF), since most studies conducted on AM plants do not discriminate on the LMWOA origin. This study aimed to determine whether AMF release significant amounts of LMWOAs to liberate P bound to Fe oxides, which is otherwise unavailable for the plant. Solanum lycopersicum L. plants mycorrhized with Rhizophagus irregularis were placed in a bicompartmental mesocosm, with P sources only accessible by AMF. Fingerprinting of LMWOAs in compartments containing free and goethite-bound orthophosphate (OP or GOE-OP) and phytic acid (PA or GOE-PA) was done. To assess P mobilization via AM symbiosis, P content, photosynthesis, and the degree of mycorrhization were determined in the plant; whereas, AM hyphae abundance was determined using lipid biomarkers. The results showing a higher shoot P content, along with a lower N:P ratio and a higher photosynthetic capacity, may be indicative of a higher photosynthetic P-use efficiency, when AM plants mobilized P from less-accessible sources. The presence of mono-, di-, and tricarboxylic LMWOAs in compartments containing OP or GOE-OP and phytic acid (PA or GOE-PA) points toward the occurrence of reductive dissolution and ligand exchange/dissolution reactions. Furthermore, hyphae grown in goethite loaded with OP and PA exhibited an increased content of unsaturated lipids, pointing to an increased membrane fluidity in order to maintain optimal hyphal functionality and facilitate the incorporation of P. Our results underpin the centrality of AM symbiosis in soil biogeochemical processes, by highlighting the ability of the AMF and accompanying microbiota in releasing significant amounts of LMWOAs to mobilize P bound to Fe oxides.",
"conclusion": "Conclusion We found that free P sources were earlier acquired by the AM plant compared to their goethite-associated counterparts. Our results on the acquisition of P from GOE-bound sources suggest the AM symbiosis was conducive to greater P-use efficiency. Since we found evidence pointing to a synchronous response of the plant–fungus binomial, by mobilizing P desorbed from GOE to the photosynthetically active tissues and ensuring an adequate photosynthetic capacity for fueling the exploration of hyphae in the soil, as well as the costly production of LMWOAs. The LMWOAs with two and three carboxylic groups (e.g., oxalic, succinic, and citric acids) were more abundant in those FCs where P was mobilized from sources with lower accessibility. This fact suggests that desorption of OP and PA from GOE was mediated either by ligand exchange or by ligand-controlled dissolution. Additionally, the presence of low contents of monocarboxylic acids characteristic of transient anaerobic conditions (i.e., acetic, butyric and lactic acids) before and after the AM plant acquired P from GOE-P associations may be indicative of reductive dissolution processes to release P from goethite surfaces. Finally, the fungal lipid analysis may indicate the AMF modified its membrane lipid composition by increasing the amount of unsaturated lipids when mobilizing the GOE-bound P sources, for maintaining the growth state and functionality. The AM symbiosis with R. irregularis and accompanying microbiota played a central role mobilizing P from GOE-bound sources to the host plant, highlighting the potentially pervasive influence of AMF on key ecosystem processes as the cycling of essential plant nutrients.",
"introduction": "Introduction Phosphorus (P) is an essential plant macronutrient ( Schachtman et al., 1998 ), and its deficiency limits the plant growth in both natural and agricultural systems ( Oberson et al., 2001 ). Particularly, in acidic soils, the high affinity and strong specific adsorption of inorganic (Pi) and organic (Po) phosphorus forms to iron (Fe) oxides determine their accessibility to plants ( He and Zhu, 1998 ). Rhizosphere acidification and the release of low-molecular-weight organic acids (LMWOAs) are the important plant response mechanisms to increase P availability in the soil solution ( Wang et al., 2019 ). The LMWOAs may solubilize P from mineral surfaces either by ligand exchange or by ligand-promoted dissolution of Fe oxides ( Owen et al., 2015 ). The ability of different LMWOAs to desorb P generally decreases with a decrease in the stability constants of Fe (III) acid complexes ( Marschener, 1998 ; Deubel and Merbach, 2005 ). The adsorption of LMWOAs is driven by positively charged oxide surfaces and the negative charge of the carboxylate group and is influenced by the formation of metal complexes in solution, with adsorption generally increasing with their concentration in solution and the number of carboxylic groups ( Oburger et al., 2011 ; Adeleke et al., 2017 ). Thus, tricarboxylic acids such as citrate have a higher efficiency to desorb P from Fe oxides than dicarboxylic or monocarboxylic ones ( Geelhoed et al., 1999 ; Richardson, 2001 ). The association of plants with symbiont organisms is one of the most widespread strategies employed to mobilize P in acidic tropical soils ( Seguel et al., 2013 ). In particular, the association with arbuscular mycorrhizal fungi (AMF) is central to the P cycling, mobilization, and supply to plants adapted to acidic environments ( Klugh and Cumming, 2007 ). The arbuscular mycorrhizal (AM) symbiosis promotes the formation of an extensive mycelium network that operates as functional extensions of the plant root system ( Xu et al., 2007 ), exchanging the acquired P for fresh assimilated photosynthetic carbon (C) from the host plant ( Zhang et al., 2016 ). Furthermore, AMF may act as hub translocating freshly assimilated C to soil microbes on the surfaces of mycorrhizal hyphae, spores, and the hyphosphere, the zone surrounding individual fungal hyphae ( Zhang et al., 2014 ; Manchanda et al., 2017 ). The accompanying AMF microbiota may be functionally diverse and provide essential plant growth-promoting functions, such as phytate mineralization, siderophore production, Pi solubilization, and LMWOA production ( Battini et al., 2016 ). In this way, the association of AMF with bacteria provides a beneficial partnership for accessing and mobilizing soil P pools, which otherwise would not be available to the plant ( Wang et al., 2016 ; Drigo and Donn, 2017 ). Phosphorus mobilization by AMF may involve both Pi ( Smith and Read, 2008 ) and Po forms ( Andrino et al., 2020 ). There is also evidence that AMF can desorb OP from ferrihydrite ( Gogala et al., 1995 ; Rakshit and Bhadoria, 2010 ), and recently, we confirmed the ability of R. irregularis to mobilize Po and Pi bound to goethite (GOE), one of the most abundant Fe (oxy)hydroxides in tropical soils, at differing host plant C cost ( Andrino et al., 2019 ). The release of P bound to pedogenic oxides requires the action of LMWOAs ( Geelhoed et al., 1999 ), but the production of LMWOAs by AMF is still poorly documented ( Bharadwaj et al., 2012 ). Sato et al. (2015) and Burghelea et al. (2018) pointed out that AMF exudates involved in P mobilization from Po and Pi sources may comprise phosphatases, phenolic compounds, protons, siderophores, and an increased root exudation of organic ligands; however, studies on the production of LMWOAs by AMF are scarce ( Tawaraya et al., 2006 ; Toljander et al., 2007 ). Consequently, the present study seeks to understand the role of LMWOAs secreted by the AMF to the P mobilization from GOE-bound P sources. We hypothesize that the suite of LMWOAs produced when mobilizing P from GOE-bound orthophosphate (OP or GOE-OP) and phytic acid (PA or GOE-PA) sources differs from those in the presence of their free P forms, as a consequence of ligand dissolution processes. To this end, we used a bicompartmental mesocosm consisting of a plant compartment (PC) harboring one Solanum lycopersicum L. plant mycorrhized with Rhizophagus irregularis ; however, the fungal compartment (FC) contained free or OP or GOE-OP and PA or GOE-PA only accessible by the AM fungus. To assess P mobilization via AM symbiosis, P and N contents, photosynthesis, and the degree of mycorrhization were determined in the plant; however, LMWOAs fingerprint and the AM hyphae abundance were determined using the FC.",
"discussion": "Discussion In the current study, we investigated the role of LMWOAs secreted by AMF and their accompanying microbiota in the mobilization of GOE-bound P sources. R. irregularis DAOM 197198 seemed not to be a specialist species in terms of mobilizing P bound to Fe oxides, since AM plants did not restore their initial P tissue contents during the time course experiment ( Figure 2 ). It is not surprising, as it is a frequent dweller in agricultural contexts, thus not likely to be a functional specialist ( Köhl et al., 2016 ). Our results show that before any P was acquired by the plant, there was a dilution in the P contents of shoots and roots for all treatments. Phosphorus was preferentially stored in the shoots, showing no P deficiency in case of AM plant mobilizing OP and PA, and only a slight P deficiency in case of those accessing GOE-OP and GOE-PA, as indicated by the N:P ratios ( Figure 3 ). Furthermore, the AM plants with access to a P source exhibited significant higher photosynthetic activities until the end of the experiment, compared to the controls ( Figure 4 ). Phosphorus is a key limiting nutrient and plays an important role in photosynthesis and the production of carbohydrates ( Thuynsma et al., 2016 ). Plants may cycle P more efficiently at low soil P levels, by exhibiting a higher resorption efficiency ( Dalling et al., 2016 ; Rychter et al., 2016 ). Hidaka and Kitayama (2013) proposed that in P-poor soils, plants tend to allocate P to the shoots, for keeping their productivity and growth and reducing the demand for P. Mycorrhizal benefit on the host plant is usually greater when plants are P limited ( Hoeksema et al., 2010 ; Johnson et al., 2014 ) and this is particularly applicable to AM plants, which tend to store more P in the shoots, as compared to non-mycorrhizal plants ( Yang et al., 2014 ; Holste et al., 2016 ). Furthermore, plants establishing AM symbiosis exhibit higher photosynthetic capacity, stomatal conductance, and transpiration rates, compared to non-mycorrhizal ones ( Augé et al., 2016 ). The strength of the C sink in the mycorrhizal roots enhances plant photosynthetic capacity by wider opening of the stomata, allowing for more CO 2 to diffuse into the leaf, which in terms increase the level of sucrose and hexose in roots ( Boldt et al., 2011 ). Furthermore, the C sink to the roots accelerates the utilization of triose phosphate for sucrose synthesis and the export toward the phloem. This increases plant P recycling rates by releasing P back to the chloroplast and activating the regeneration of ribulose 1,5-bisphosphate in the Calvin cycle. By this mechanism, more C is fixed per time and per unit of P, resulting in higher photosynthetic P-use efficiency ( Tuomi et al., 2001 ; Valentine et al., 2001 ; Kaschuk et al., 2009 ). Our results on the reduction of the P dilution in the shoots, coupled with lower N:P ratios and higher photosynthetic capacities over time, in those treatments with access to the GOE-bound P sources, may be an indicator that AM symbiosis was conducive to more efficient use of P mobilized from less-accessible sources. A more efficient photosynthetic P use may also benefit the secretion of LMWOAs, as they entail a substantial C cost which is exclusively supported by the direct supply of photoassimilates ( Plassard and Fransson, 2009 ). Studies on the exudation of LMWOA by AMF are somewhat limited, and we found none where these mediated the mobilization of P bound to iron oxides. Conversely, D’Amico et al. (2020) recently demonstrated that in extremely P-poor environments, ectomycorrhizal fungi were able to release both Pi and Po from associations with goethite. We found two similar studies where AMF growth was isolated and organic acid production was measured: The one by Toljander et al. (2007) detected the presence of acetate and formiate, and other by Tawaraya et al. (2006) observed citrate and malate as part of the AMF hyphal exudates. We also detected three out of the four LMWOAs found in the two previous studies in similar concentration as for citric acid (100 nmol/g FC substrate). Furthermore, we found couple of studies investigating P desorption from goethite with non-mycorrhizal plants and incubation experiments. Parfitt (1979) concluded that one of the mechanisms by which GOE-OP could be solubilized and made available is through the action of the LMWOAs. He performed several extraction cycles on GOE-OP in combination with different LMWOAs, resulting in a higher OP desorption when it was incubated in the presence of citrate. In a more recent study, Martin et al. (2004) studied the effects of LMWOAs which may be released by non-mycorrhizal plant roots, such as citrate, on the desorption of PA and OP bound to GOE. They found a smaller amount of PA desorbed from GOE as compared to OP, which was attributed to the strength of chemical bonds and the high negative charge of the complexes. In our experimental setup, we detected the presence of significantly larger concentrations of LMWOAs in the FCs containing PA, GOE-OP, and GOE-PA before and after any P was allocated to the plant tissues, compared to those ones containing OP ( Figures 2 , 5 ). The FCs where AM plants mobilized GOE-OP and GOE-PA contained the highest concentrations of malic and oxalic acid during the plant P incorporation phase. For citric acid, the trend was opposite to that of the dicarboxylic acids, having a significant larger concentration before any P was acquired in case of GOE-OP and GOE-PA treatments. The LMWOAs detected in the FCs of GOE-OP and GOE-PA treatments belong to the ones with higher chelation capacity, namely, malate, oxalate, and citrate, thus more effective mobilizing P from GOE or amorphous ferric hydroxides, as compared to the ones containing one carboxyl group ( Muthukumar et al., 2014 ; Thorley et al., 2015 ). The release of mono/di/tri LMWOAs by the fungal partner of mycorrhizal plants refers to a possible mechanism involved in the acquisition of P from mineral-bound sources, where organic acids weaken and break the bonds between surface-coordinated P forms and structural metal ions before being mobilized by the AM plant. The presence of low contents of monocarboxylic acids (i.e., acetic, butyric, and lactic acids) before and after the AM plant acquired P from GOE-OP and GOE-PA, may be indicative of fermentation reactions occurring during transient periods of anaerobiosis, when reconstituting the water content to field capacity in the FCs. Although thermodynamically stable, goethite may undergo reductive dissolution in an anoxic environment when the redox potential drops ( Torrent et al., 1987 ). This partial reductive dissolution may have taken place in FCs containing GOE-P compounds caused by the anaerobic respiration of microorganisms, transferring electrons from organic compounds to the Fe(III)-oxides ( Peiffer and Wan, 2016 ). This reaction may have contributed, together with the action of di/tricarboxylic acids, in releasing adsorbed P from the goethite surfaces. However, the presence of the di/tri LMWOAs may suggest a mechanism used by AMF to desorb P from GOE surfaces. This desorption is done either by ligand exchange or by dissolution and subsequent desorption of P through the action of LMWOAs, such as oxalic, succinic, and citric acids. In the case of those plants accessing GOE-PA, the desorbed PA may be mineralized through the action of phosphatases secreted by the AMF ( Tisserant et al., 2012 ). We inoculated the tomato plants with R. irregularis DOAM 197198 grown under xenic conditions; thus, the inoculum carried the microorganisms naturally associated with its hyphae. In this sense, our results of LMWOAs production have to be examined under the possible joint influence of the AMF and its accompanying microbiota. In this regard, Battini et al. (2016) isolated microbiota from Rhizophagus intraradices and found plant growth-promoting activities such as phytate mineralization, siderophore production, Pi solubilization, and LMWOA production in several representatives of Gram-positive (e.g., Streptomyces spp., Arthrobacter spp., Nocardiodes spp., and Bacillus spp.) and Gram-negative bacterial groups (e.g., Sinorhizobium spp.). Furthermore, Lecomte et al. (2011) and Selvakumar et al. (2016) reported bacteria closely associated with the mycelium of R. irregularis involved in the P mobilization from phytic acid. More recently, P transfer from phytate via AMF with the assistance of phytate-mineralizing bacteria was confirmed by Hara and Saito (2016) . They isolated bacteria from the hyphosphere of the cosmopolitan AMF R. irregularis DAOM197198 and found that Claroideoglomus etunicatum can mineralize phytic acid. Taktek et al. (2015 , 2017) and Wang et al. (2016) also isolated bacteria closely attached to the hyphosphere of R. irregularis DAOM197198. They showed that exudates from R. irregularis hyphae supported the growth and activity of bacteria with high potential for LMWOA production and Po mineralization. A possible mechanism used by the AMF and its accompanying microbiota to desorb P from the surface of GOE would involve the release of exudates containing LMWOAs. Following this, the desorbed OP could be taken up directly by the AM hyphae, while the desorbed PA still had to be hydrolyzed by phosphatases. As we observed in our results, it is likely that these previous steps delayed the incorporation of P mobilized from GOE-PA and GOE-OP into the plant tissues, compared to the other treatments ( Figure 2 ). Additionally, several authors ( Otani and Ae, 1999 ; Hayes et al., 2000 ; George et al., 2005 ) have pointed out that some LMWOAs (e.g., citric acid) have a synergistic effect on the secreted phosphatases (e.g., acid phosphatase), by changing the chemical structure or molecular size of the extracted Po and making it more accessible to enzymatic action. Summarizing, we found profiles of LMWOAs differing with the accessibilities of the offered P sources. The LMWOAs with two and three carboxylic groups were more abundant in case of P sources with lower P accessibility, before any P was acquired into the plant tissues. Hence, our results would point to a plant–fungus synchronous functioning that would adapt over time to respond to P accessibility in the soil. In this way, the mycorrhizal symbiosis would favor a more efficient P utilization, by maintaining an adequate photosynthetic capacity to ensure the soil volume exploration, together with the secretion of the LMWOAs. Our results on the presence of AMF lipid biomarkers, together with the P acquisition from the different sources, highlight the central role played by R. irregularis in mobilizing P into the AM plant. This statement is founded on the fact that lipid biomarkers in the FC increased along with P in the plant tissues ( Table 2 ); besides, both parameters positively correlated with the presence of arbuscules ( Table 3 ). The arbuscules are short-lived structures with a turnover rate of 1–2 weeks ( van Aarle and Olsson, 2003 ) and the interface between the plant and AMF ( Wewer et al., 2014 ), where the P and photosynthates are exchanged in the periarbuscular space ( Kobae et al., 2014 ; Saito and Ezawa, 2016 ). Thus, it would be consistent with the interpretation that P mobilization stimulated by the LMWOAs secretion was further supported by fungal growth and the exchange structures at the root level. The second conclusion stems from the correlation between the fungal PLFAs 18:1ω7c and 18:2ω6,9 with the acquired P and the arbuscules (%), for those AM plants mobilizing P from GOE-OP and GOE-PA, respectively. The PLFA are vital components of all biological membranes and play a key role in processes such a signal transduction, cytoskeletal rearrangement, membrane trafficking, etc., and remain at the place where they are synthesized ( van Aarle and Olsson, 2003 ; Debiane et al., 2011 ; Dalpé et al., 2012 ). The AMF lack genes for a de novo biosynthesis of lipids and are enzymatically only able to elongate 16C lipid molecules; they mandatorily receive from their host plant ( Bravo et al., 2017 ; Keymer et al., 2017 ). Since, AMF only elongates 16C lipid molecules, requiring the plant to produce them ( Luginbuehl et al., 2017 ), the presence of the two unsaturated 18C PLFA fungal biomarkers in our experiment (18:1ω7c and 18:2ω6,9) might support the possibility of a modified composition in lipids constituting the hyphal membrane, which might be seen as an adaptation to the accessibility of the different P sources. Plasticity in fatty acid synthesis attributable to nutritional factors is common in filamentous fungi ( Olsson et al., 2002 ). Based on the correlation data, our results point toward a change in the unsaturation level of AMF membrane lipids with the changing quality of the offered P sources. R. irregularis , therefore, might have modified its lipid composition in response to the different P sources. Consequently, the lipid membrane increased its fluidity to keep its integrity compatible with an optimal membrane functionality ( Calonne et al., 2010 ). Membrane fluidity depends on its phospholipid composition of varying length and saturation with unsaturated lipid chains being more fluid than saturated ones. The unsaturated double bonds make it harder for the lipids to pack together by putting kinks into the otherwise straight hydrocarbon chain ( Reichle, 1989 ). For successful adaptation to altered physicochemical environments, the active remodeling of membrane lipid composition is an essential feature and depends on both strain properties and cultivation conditions ( Bentivenga and Morton, 1994 ; Čertík et al., 2005 ). Changes in membrane fluidity influence membrane processes such as transport, enzyme activities, and signal transduction ( Benyagoub et al., 1996 ; Turk et al., 2007 ). In summary, it is plausible to consider that AMF modified its membrane lipid composition when mobilizing the GOE-bound P sources may have modulated the way in which the lipid membrane was organized for maintaining the growth state ( Wang et al., 2017 ), by increasing unsaturated lipids in the case the AMF developed on a P source bound to GOE."
} | 5,711 |
38005773 | PMC10674991 | pmc | 5,833 | {
"abstract": "In modern energy, various technologies for absorbing carbon dioxide from the atmosphere are being considered, including photosynthetic microalgae. An important task is to obtain maximum productivity at high concentrations of CO 2 in gas–air mixtures. In this regard, the aim of the investigation is to study the effect of light intensity on the biomass growth and biochemical composition of five different microalgae strains: Arthrospira platensis , Chlorella ellipsoidea , Chlorella vulgaris , Gloeotila pulchra , and Elliptochloris subsphaerica . To assess the viability of microalgae cells, the method of cytochemical staining with methylene blue, which enables identifying dead cells during microscopy, was used. The microalgae were cultivated at 6% CO 2 and five different intensities: 80, 120, 160, 200, and 245 μmol quanta·m −2 ·s −1 . The maximum growth rate among all strains was obtained for C. vulgaris (0.78 g·L −1 ·d −1 ) at an illumination intensity of 245 µmol quanta·m −2 ·s −1 . For E. subsphaerica and A. platensis , similar results (approximately 0.59 and 0.25 g·L −1 ·d −1 for each strain) were obtained at an illumination intensity of 160 and 245 µmol quanta·m −2 ·s −1 . A decrease in protein content with an increase in illumination was noted for C. vulgaris (from 61.0 to 46.6%) and A. platensis (from 43.8 to 33.6%), and a slight increase in lipid content was shown by A. platensis (from 17.8 to 21.4%). The possibility of increasing microalgae biomass productivity by increasing illumination has been demonstrated. This result can also be considered as showing potential for enhanced lipid microalgae production for biodiesel applications.",
"conclusion": "4. Conclusions This study is devoted to determining the optimal illumination for microalgae strains previously adapted to high CO 2 concentrations to increase the efficiency of CO 2 absorption. To achieve these goals, experiments were conducted on the cultivation of microalgae strains ( A. platensis , C. ellipsoidea , C. vulgaris , E. subsphaerica , and G. pulchra ) at different intensities (80, 120, 160, 200, and 245 µmol quanta·m −2 ·s −1 ) and 6% CO 2 . To optimize the cultivation conditions of microalgae strains and obtain a stable result in biomass productivity, a mode of sequential adaptation to high illumination, using the method of long-term continuous cultivation, was implemented. The effectiveness of this approach has been confirmed by previous studies [ 11 ], where an increase in microalgae biomass was observed under high CO 2 content (up to 9%). The growth rate was different because of species and strain specificity: for C. vulgaris , C. ellipsoidea , and G. pulchra strains, the maximum growth rates were obtained at 245 µmol quanta·m −2 ·s −1 (0.78, 0.72 and 0.47 g·L −1 ·d −1 , respectively); for A. platensis and E. subsphaerica , 0.25 and 0.59 g·L −1 ·d −1 , respectively, at 160 µmol quanta·m −2 ·s −1 . According to the results of biochemical analysis, it is also possible to talk about differences in the response of different types of microalgae to the adaptation process under increased illumination. Thus, E. subsphaerica is the culture least susceptible to an increase in the illumination intensity, i.e., the content of the major organic compounds (proteins, lipids, and carbohydrates) did not change significantly during all the experiments. The results for Chlorella can be considered similar, except for decreasing the number of proteins in the case of C. vulgaris , while protein content for C. ellipsoidea did not change significantly with increasing illumination. For the G. pulchra strain, an increase in the protein content entailed a decrease in the lipid content at the same illumination values (200 and 245 µmol quanta·m −2 ·s −1 ). In the case of A. platensis , a steady increase in the amount of carbohydrates was observed with an increase in illumination intensity. Also, an adapted express method was used to determine the viability of microalgae cells using a light microscope, based on cytochemical staining of living and dead cells with a lifetime dye methylene blue. This method made it possible to determine an insignificant proportion of dead cells at all light intensities, as well as morphological changes in the A. platensis trichomes at an illumination intensity of 245 µmol quanta·m −2 ·s −1 . Thus, the conditions for microalgae cultivation were optimized and the effectiveness of the method, which was used to adapt strains to high illumination intensity at elevated CO 2 content, was demonstrated.",
"introduction": "1. Introduction The development of modern energy is facing various challenges, including the regional limitation of fossil fuel reserves [ 1 ] and the complex impact of the energy production process on the environment. In recent years, the task of reducing greenhouse gas emissions [ 2 ] has acquired a special place in connection with climate change trends. Several technologies have been proposed and tested to prevent CO 2 emissions and capture them from the atmosphere. One of the widely studied and highly effective methods of CO 2 capture is to use microalgae as photosynthetic organisms that absorb CO 2 and can be used for biofuel production [ 3 , 4 , 5 ]. The advantages of microalgae as photosynthetic agents of carbon conversion have been repeatedly discussed. It is worth emphasizing the most important of them: (1) the cultivation of microalgae does not require arable land, unlike terrestrial crops, and the use of wastewater as a nutrient medium reduces the consumption of water and inorganic components of nutrient media [ 6 , 7 , 8 ]; (2) microalgae can synthesize valuable metabolites, vitamins, and various organic compounds in commercially significant volumes [ 9 , 10 ]. For effective CO 2 absorption, it is necessary to select the most productive strains of microalgae tolerant to high concentrations of CO 2 and flue gases [ 11 , 12 , 13 ], as well as to search for optimal cultivation conditions to obtain maximum biomass productivity. A wide range of microalgae is used as objects of research, among which the most common are strains of Chlorella , Nannochloropsis , in some cases Arthrospira ( Spirulina ). The results obtained on the productivity of microalgae biomass vary in a wide range, apparently due to heterogeneous cultivation conditions (type of photobioreactor, temperature, method of supplying gas–air mixtures, gas consumption during bubbling, duration of cultivation, etc.). Generalization and comparative analysis of the results of experiments with increased content of CO 2 in the gas–air mixture (10 and 20%) are given, in particular in [ 14 ], with the absorption rates of CO 2 (from 0.160 to 0.265 g·L −1 ·d −1 , depending on the strains and concentration of CO 2 ), and the achieved content of lipids, chlorophyll, and carotenoids. A similar generalization can be seen in Ho et al. [ 15 ]. Note that both reviews do not indicate illumination levels during cultivation. In their article, Hu Xia et al. [ 16 ] studied the response of ten Chlorella strains to increased CO 2 concentrations, and the maximum productivity for four of them was at 10% CO 2 ; C. vulgaris was identified, demonstrating the maximum accumulation of lipids. It is necessary to note that the cells of both Chlorella sp. and C. vulgaris kept their normal morphologies after 15-day batch culture, while the cells of two other strains were destroyed, which was confirmed by photographs from a scanning electron microscope. This confirms the legitimacy of choosing a method of long-term microalgae cultivation for the most complete characterization of their growth in the face of elevated CO 2 concentrations and optimization of other cultivation conditions. In several studies, A. platensis is used as a convenient object due to the resistance to contamination and the manufacturability of growing and harvesting biomass. Thus, when growing the A. platensis strain in gas–air mixtures with CO 2 concentrations of 1, 5, and 9%, an average biomass growth rate (duration of experiments 15 days) of 0.079, 0.076, and 0.048 g·L −1 ·d −1 , respectively, was achieved [ 17 ]. Adaptation of this strain to elevated concentrations of CO 2 allowed further experiments on its cultivation in a gas–air mixture simulating flue gases with a CO 2 content of 6% [ 18 ], to obtain a growth rate almost twice as high, 0.140 g·L −1 ·d −1 (the duration of the experiments is 14 days). The task of studying the effect of illumination on the growth of microalgae was set earlier and studied in detail. The study of the effect of illumination showed that, as a rule, with an increase in light intensity, the growth rate of microalgae increases until the photoinhibition threshold is reached, but both the result of this effect and the threshold are different in different strains [ 7 , 19 , 20 ]. Thus, the influence of different levels of illumination as well as the duration of periods of illumination (photoperiods) on the growth of a wide range of microalgae strains ( Porphyridium purpureum , Chloromonas reticulata , Parietochloris incisa , Neochloris , Botryococcus braunii , Scenedesmus obliquus , A. platensis , Desmodesmus sp., etc.) were considered [ 7 , 19 , 21 , 22 , 23 , 24 , 25 , 26 ]. The most interesting are the study results of the effect of illumination when cultivating microalgae at high CO 2 concentrations. Thus, the reaction of Euglena gracilis to variations in temperature (25–33 °C), CO 2 concentration (from 2 to 6%), and illumination (20–200 µmol quanta·m −2 ·s −1 ) were experimentally shown [ 27 ]. The maximum productivity values (approximately 0.045 h −1 ) were achieved at 4% CO 2 and an illumination intensity of 100 µmol quanta·m −2 ·s −1 . From those considered in [ 14 ], oleaginous microalgae marine Chlorella sp., freshwater Chlorella sp., Scenedesmus sp., Botryococcus sp., and Nannochloropsis sp., grown at CO 2 concentrations in flue gases of 10 and 20%, the maximum values of biomass growth were obtained at 10% CO 2 and an illumination intensity of 60 µmol quanta·m −2 ·s −1 . The highest lipid content was shown by the strain Nannochloropsis sp.; and with the transition from the 12:12 photoperiod to continuous illumination, the growth rates for biomass and lipids increased. A further increase in the concentration of CO 2 led to a decrease in productivity both in biomass and lipids. In their article, when Shih-Hsin Ho et al. [ 15 ] studied Scenedesmus obliquus CNW-N , a productivity of 0.84 g·L −1 ·d −1 was achieved for the this strain at a much higher illumination intensity of 420 µmol quanta·m −2 ·s −1 (temperature 28 °C, 2.5% CO 2 , illumination constant; the range of studied illumination intensity is 60–540 µmol quanta·m −2 ·s −1 ). Strain Chlorella sp. MTF-7 was grown in a flue gas atmosphere at CO 2 concentrations of 2%, 10%, and 25% at a constant illumination intensity of 300 µmol quanta·m −2 ·s −1 [ 28 ]. With such a significant range of CO 2 concentrations, the achieved biomass productivity was approximately the same in all experiments, and it was approximately 0.350 g·L −1 ·d −1 . It should be noted that both the high level of illumination and CO 2 concentrations did not lead to inhibition of the growth of Chlorella microalgae and provided high biomass productivity. The search for optimal conditions is also conducted in the direction of the spectral composition of light used for microalgae cultivation [ 24 , 26 , 29 , 30 , 31 ] and the optimal duration of the photoperiod [ 25 , 32 , 33 ]. The influence of the spectrum is outside the topic of this work, although it is worth noting that several publications have shown that the daylight spectrum as a whole is close to monochrome lighting in terms of its effect on the productivity of microalgae. Some studies have found a decrease in the rate of carbon fixation in the presence of even a short dark period compared to cultures under continuous illumination [ 25 , 34 ]. This conclusion determined the illumination modes we selected. If LED lighting is used, this does not significantly increase the cost of cultivation. Since the absorption of CO 2 , as a method of reducing emissions into the atmosphere, is associated with the further transformation of microalgae biomass into energy products [ 35 ], the criterion for optimizing the cultivation of microalgae, in addition to the rate of biomass growth and the intensity of CO 2 absorption, is the content of the main biochemical components, especially lipids as raw materials for the production of biodiesel [ 14 , 21 , 22 , 29 , 32 ]. It is important to note that high CO 2 concentrations are a stress for microalgae, and therefore various adaptation methods are used in experimental studies, in particular the Adaptive Laboratory Evolution (ALE) technique described in detail in [ 11 ]. ALE is an innovative stress-induced approach, the main advantage of which is that it can improve the characteristics of the strain and develop the resistance of microalgae to numerous environmental stresses, supporting rapid cell growth [ 35 , 36 ]. The ALE technique was successfully used [ 37 , 38 , 39 ]. In our previous studies [ 11 ], as a result of long-term phased cultivation with adaptation to increasingly high CO 2 concentrations, as indicated above, stable and high productivity of microalgae Arthrospira platensis , Chlorella ellipsoidea , Chlorella vulgaris , Gloeotila pulchra , and Elliptochloris subsphaerica was achieved (CO 2 up to 9% in a gas–air mixture). However, microalgae cultivation in these experiments was carried out at low light values (74.3 µmol quanta·m −2 ·s −1 ). In this regard, the purpose of these studies is to determine the optimal illumination for strains adapted to high concentrations of CO 2 by the criterion of productivity and the growth rate of biomass, which determine the efficiency of CO 2 absorption. Towards this goal, the following work has consistently carried out: optimization of cultivation conditions for microalgae strains previously adapted to elevated CO 2 concentrations by gradually adapting them to a high illumination intensity using the method of long-term continuous cultivation; comparative analysis of the response of the most crucial characteristics of biomass growth (growth rate, biochemical composition, absorption intensity of the main components of nutrient media, morphological characteristics of microalgae) to the process of adaptation to increased illumination; evaluation of the viability of microalgae under various conditions based on the express method of cytochemical staining of cells with a lifetime dye with their subsequent control by light microscopy.",
"discussion": "2. Results and Discussion 2.1. The Biomass Growth Rate and pH of the Culture Medium Figure 1 shows the results of determining the growth rate of microalgae biomass for all conducted experiments with different lighting intensities. C. vulgaris has demonstrated a tendency to increase the growth rate with an increase in illumination intensity from 80 to 120 µmol quanta·m −2 ·s −1 . Further, at an illumination intensity of 120, 160, and 200 µmol quanta·m −2 ·s −1 , the biomass growth rate was the same within the error limits (0.59 g·L −1 ·d −1 ); however, at 245 µmol quanta·m −2 ·s −1 , an active increase in biomass occurred, at which the maximum value of the biomass growth rate among all strains was obtained—0.78 g·L −1 ·d −1 ( Figure 1 ). In C. ellipsoidea , a steady increase in biomass was observed, positively correlating with the intensity of illumination. With an increase in the illumination intensity from 80 to 200 µmol quanta·m −2 ·s −1 , there is a gradual increase in the growth rate. At an illumination intensity of 200 and 240 µmol quanta·m −2 ·s −1 , the growth rate is constant within the error limits, with the maximum value being 0.72 g·L −1 ·d −1 ( Figure 1 ). In the case of G. pulchra , an increase in intensity from 80 to 160 µmol quanta·m −2 ·s −1 led to an increase in the biomass growth rate; at 160 and 200 µmol quanta·m −2 ·s −1 , it is approximately equal; the maximum growth rate is achieved with an illumination intensity of 245 µmol quanta·m −2 ·s −1 and is 0.47 g·L −1 ·d −1 ( Figure 1 ). For E. subsphaerica and A. platensis strains, an increase in the illumination intensity above 160 µmol quanta·m −2 ·s −1 did not lead to a significant increase in the biomass growth rate, which was approximately 0.59 and 0.25 g·L −1 ·d −1 , respectively, while the initial growth rates at 80 µmol quanta·m −2 ·s −1 were 0.32 and 0.13 g·L −1 ·d −1 . Thus, at a light intensity of 160 µmol quanta·m −2 ·s −1 , the same biomass productivity can be obtained as at 245 µmol quanta·m −2 ·s −1 , which may indicate saturation for these strains. Table 1 presents the results of biomass productivity obtained in previous studies of the growth of C. vulgaris and A. platensis as the most frequently considered CO 2 absorption problems. It can be seen that it was possible to achieve higher biomass productivity with an increase in illumination intensity. Note that during all the experiments lasting 14 days each, there was no mass death of microalgae, and all microalgae culture showed biomass gain. Thus, there was no photoinhibition of the strains; and in the case of E. subsphaerica and A. platensis , saturation was reached at 160 µmol quanta·m −2 ·s −1 illumination. For strains of C. vulgaris , C. ellipsoidea , and G. pulchra , it would be advisable to conduct additional studies with a further increase in illumination. In the case of Chlorella strains, a significant pH increase was observed in all experiments ( Table 2 ). At the same time, many works [ 44 , 45 , 46 ] show that, when cultivating microalgae with bubbling air with high CO 2 concentrations, pH decreases, that is, acidification. Our results might be explained by the fact that Chlorella was grown on a Tamiya nutrient medium, which is unstable, since during the growth of microalgae, the physiological absorption of NO 3 − anions prevails compared to Na + cations, which accumulate in the nutrient medium, which leads to an increase in pH. For A. platensis , the pH values in the experiments showed a tendency to a slight increase ( Table 2 ), which confirms the high buffering of the Zarrouk nutrient medium. Acidification during CO 2 bubbling by H + and HCO 3 − ions is leveled by selective absorption of the HCO 3 − anion and accumulation of the Na + cation from NaHCO 3 baking soda from the Zarrouk medium, but, in general, the addition of ions with acidic properties to the medium during CO 2 bubbling prevents strong alkalinization of the nutrient medium during its long-term cultivation. For E. subsphaerica , the pH remained constant for three experiments, but, in the experiment with an illumination intensity of 200 µmol quanta·m −2 ·s −1 , the pH decreased markedly. A similar trend for pH values was obtained for G. pulchra : at 120 µmol quanta·m −2 ·s −1 , the pH is constant, and at 160 and 200 µmol quanta·m −2 ·s −1 , acidification of the medium was noted ( Table 2 ). As described earlier, such a decrease in pH is characteristic when microalgae are grown in an atmosphere with elevated concentrations of CO 2 or flue gases. At the same time, no acidification of pH was observed in the experiment with an illumination intensity of 245 µmol quanta·m −2 ·s −1 . 2.2. The Biochemical Composition of Microalgae Biomass The effect of light on the content of proteins, lipids, and carbohydrates in five different microalgae strains was studied. With increasing illumination, the protein content ( Figure 2 a) showed multidirectional dynamics. For the C. vulgaris strain, a drop in protein content was noted (from 61.0 to 46.4%) with an increase in illumination intensity from 120 to 160 µmol quanta·m −2 ·s −1 ; further, at an illumination intensity of 160, 200, and 245 µmol quanta·m −2 ·s −1 , the number of proteins remained at one level within the margin of error. For strains C. ellipsoidea and E. subsphaerica , the content of proteins did not change significantly during all experiments. In the case of G. pulchra , at illumination intensities of 120 and 160 µmol quanta·m −2 ·s −1 , the protein content is the same within the error limits, with an increase in illumination from 160 to 200 µmol quanta·m −2 ·s −1 , an increase in protein content is observed (from 32.6 to 46.5%), and at 200 and 245 µmol quanta·m −2 ·s −1 protein content again does not change significantly. A similar picture develops for strain A. platensis : at first, the protein content increases with an increase in illumination intensity from 43.8 to 57.2% (from 120 to 160 µmol quanta·m −2 ·s −1 , respectively); with increasing intensity from 160 to 200, there is a noticeable decrease in protein content (from 57.2 to 31.9%), and no further change occurs (at 245 µmol quanta·m −2 ·s −1 ). Thus, a pronounced trend, namely, a decrease in protein content with an increase in illumination, was noted only in C. vulgaris and A. platensis strains. These results positively correlate with those obtained in [ 47 ], where there is also a decrease in the number of proteins with an increase in the intensity of illumination. For lipids, as well as in the case of proteins, there was no unambiguous trend among all strains with an increase in the intensity of illumination. Let us note the main points on the results obtained ( Figure 2 b). For C. vulgaris and C. ellipsoidea strains, the maximum lipid content was recorded at 200 µmol quanta·m −2 ·s −1 (26.5 and 23.9%, respectively). In the margin of error for E. subsphaerica , similar results were obtained at illuminances of 120, 160, and 200 µmol quanta·m −2 ·s −1 , which are 21.1, 20.7, and 20.8%, respectively. G. pulchra showed a tendency to decrease the content of lipids. For this strain, the maximum content was at an illumination intensity of 160 µmol quanta·m −2 ·s −1 (27.3%). At an illumination intensity of 245 µmol quanta·m −2 ·s −1 , the lowest lipid content was recorded for all strains except A. platensis , while for this strain the maximum result was obtained. The results obtained for A. platensis (a slight increase in lipids (from 17.8 to 21.4%) with an increase in illumination from 120 to 245 µmol quanta·m −2 ·s −1 ) were consistent with the results in [ 48 ], where an increase in lipid content from 36.1 to 47.1% was noted with a change in illumination intensity from 55 to 400 µmol quanta·m −2 ·s −1 . In the same work, the growth of lipids with an increase in illumination was also shown for the C. vulgaris strain (from 21.0 to 33.0% at 55 and 450 µmol quanta·m −2 ·s −1 , respectively), but there is no such pronounced trend in our study for Chlorella . Strains C. vulgaris , C. ellipsoidea , and E. subsphaerica were resistant to increased light intensity in terms of carbohydrate content ( Figure 2 c). In Nzayisenga et al. [ 7 ], a similar trend was noted for various strains of microalgae: while increasing illumination, the amount of carbohydrates does not change significantly. The carbohydrate content of G. pulchra strain remained unchanged in all experiments, except for an illumination intensity of 245 µmol quanta·m −2 ·s −1 , at which the amount of carbohydrates increased significantly (25.6%). A. platensis shows a steady accumulation of carbohydrates in cells with increasing illumination intensity (from 9.9 to 23.1% at 120 and 245 µmol quanta·m −2 ·s −1 , respectively). 2.3. Dynamics of the Nutrient Media Components during the Experiments The results of the consumption of nitrates and phosphates by microalgae strains are shown in Figure 3 . In Chlorella strains, the proportion of nitrate consumption from the initial amount ( Figure 3 a) lies in the range of 50–65% according to the results of all experiments with different lighting intensities. It can be said that due to the increase in illumination, there were no significant changes in the consumption of nitrates. In the case of three other strains ( E. subsphaerica , G. pulchra , and A. platensis ), the proportion of nitrate consumption varies from 94% to almost 100%, which could become a limiting factor for the further growth of microalgae biomass under continuous cultivation. According to the results of phosphate consumption ( Figure 3 b), no definite trend has been identified. A major portion of the values (approximately 80%) are from 20 to 60%. However, for G. pulchra , at an illumination intensity of 200 µmol quanta·m −2 ·s −1 , the absorption of phosphates reaches almost 100%; and at 245 µmol quanta·m −2 ·s −1 , decreases to 50%. The content of bicarbonates in the medium showed stable growth during all experiments with different illumination intensities for strains C. vulgaris , C. ellipsoidea , E. subsphaerica , and G. pulchra ( Figure 4 a–d). Note that bicarbonates are the main elements of the Zarrouk medium (for A. platensis ), and initially are not part of the other nutrient media (Tamiya and BG-11). In the case of A. platensis strain, a decrease in HCO 3 − is observed at illumination intensities of 120 and 245 µmol quanta·m −2 ·s −1 , and at 160 and 200 µmol quanta·m −2 ·s −1 ; on the contrary, an increase in the content of bicarbonates. 2.4. The State of Microalgae Cells under the Influence of Different Illumination Intensities Microscopy of strains grown in continuous culture at 120 µmol quanta·m −2 ·s −1 for 12 days showed that visually microalgae cells did not differ in morphometric parameters from culture cells from experiments with 80 µmol quanta·m −2 ·s −1 . Microscopy results of microalgae strains stained with the lifetime dye methylene blue showed the absence or minimal number of dead cells. Only for G. pulchra on the last day of the experiment (120 µmol quanta·m −2 ·s −1 ), a small number of single cells formed as a result of the decay of individual filaments were observed, which were stained with methylene blue dye, dead culture cells ( Figure 5 e), which may be due to a change in growing conditions (increased illumination intensity). Examples of photos are presented in Figure 5 , which demonstrate the absence of mass staining of cells, which means the integrity of the cell walls of microalgae. These photos were taken in conditions close to production conditions using a conventional light microscope, directly at the place of experiments, where an atmospheric gas chamber (AGC) with the photobioreactors (PBRs) were installed. In this experiment (at 120 µmol quanta·m −2 ·s −1 ) with A. platensis and G. pulchra algae, the mucous membranes of microalgae trichomes are partially stained ( Figure 5 d,f), but not algae cells. The photo of G. pulchra shows the cells of E. subsphaerica , which, as a result of partial contamination, were detected in small quantities already in an experiment with an illumination intensity of 120 µmol quanta·m −2 ·s −1 . Note that Figure 5 d shows a cluster of stained E. subsphaerica cells, which indirectly indicates the displacement of this strain by the dominant strain G. pulchra . In the case of two strains, Chlorella and a strain of E. subsphaerica , in all experiments with different illumination, the minimum number of colored cells was noted, and the morphology of the cultures was not visually changed ( Figure 6 ). That is, the microalgae cells remained alive during the experiments, which indicates that the viability of the cultures was preserved under the influence of increased illumination. Microscopy of the strain G. pulchra grown at 160 µmol quanta·m −2 ·s −1 illumination shows that there were significant morphological changes in the cells of the culture grown for 6 days, compared with the results at 120 µmol quanta·m −2 ·s −1 . Figure 7 shows that under the influence of illumination, there was a massive disintegration of G. pulchra filaments into individual cells. However, its subsequent cultivation with an increase in the illumination intensity to 200 µmol quanta·m −2 ·s −1 ( Figure 7 b) revealed that this culture adapted and returned to its original state with the formation of first short 2–4 cells, and then long filaments of cells surrounded by a dense mucous membrane with radial cilia. A further increase in the illumination intensity from 200 to 245 µmol quanta·m −2 ·s −1 did not lead to the mass decay of G. pulchra filaments ( Figure 7 c). Figure 8 shows the trichomes of A. platensis . Very long trichomes and a small number of short trichomes are visible in the field of view, which indirectly indicates a weakening of the process of cell growth and division ( Figure 8 a–d). At an illumination intensity of 245 µmol quanta·m −2 ·s −1 , the trichomes of this culture were found to have disintegrated into separate parts ( Figure 8 e), and these parts are completely colored with methylene blue. Under the influence of lighting, the shell of the cells was destroyed, which led to the release of their contents to the outside, which was also stained with dye."
} | 7,284 |
35904348 | PMC9413408 | pmc | 5,836 | {
"abstract": "Natural, high-performance fibers generally have hierarchically\norganized nanosized building blocks. Inspired by this, whey protein\nnanofibrils (PNFs) are assembled into microfibers, using flow-focusing.\nBy adding genipin as a nontoxic cross-linker to the PNF suspension\nbefore spinning, significantly improved mechanical properties of the\nfinal fiber are obtained. For curved PNFs, with a low content of cross-linker\n(2%) the fiber is almost 3 times stronger and 4 times stiffer than\nthe fiber without a cross-linker. At higher content of genipin (10%),\nthe elongation at break increases by a factor of 2 and the energy\nat break increases by a factor of 5. The cross-linking also enables\nthe spinning of microfibers from long straight PNFs, which has not\nbeen achieved before. These microfibers have higher stiffness and\nstrength but lower ductility and toughness than those made from curved\nPNFs. The fibers spun from the two classes of nanofibrils show clear\nmorphological differences. The study demonstrates the production of\nprotein-based microfibers with mechanical properties similar to natural\nprotein-based fibers and provides insights about the role of the nanostructure\nin the assembly process.",
"conclusion": "Conclusions By introducing a bio-based cross-linker,\nthe morphological and\nmechanical characterization of spun protein microfibers could be expanded\nbeyond previous limitations, including elucidation of the role of\nthe PNF nanostructures on fiber properties. The fibers spun from a\nPNF suspension containing genipin resulted in a dark green color in\nless than 2 h (for 10 wt % genipin), indicating extensive cross-linking\nreactions, confirmed by IR spectroscopy results. An increasing amount\nof added genipin directly correlated with the mechanical integrity\nand properties of the fiber. The Young’s modulus reached ca . 1.4 GPa at 2 wt % genipin, and the energy at break\nincreased to 0.8 MJ/m 3 at 10 wt %, more than 4 times the\nvalues for the fibers from pure PNFs. With 10 wt % of genipin, the\nmodulus and the stress at break of the microfiber from straight PNFs\nreached 1.6 GPa and 20 MPa, respectively, which is ca . 4 times that of the fiber from curved PNFs spun at the same condition.\nThus, it has been demonstrated that the mechanical properties of the\nPNF fibers can be improved by solely adding a bio-based and nontoxic\ncross-linker. In addition, the results herein showed that the fibril\nmorphology and the flow rate used for spinning are also critical for\ndefining the mechanical performances of the spun fibers.",
"discussion": "Results and Discussion Cross-Linking PNFs with Genipin The cross-linking effect\nof genipin on the curved PNF network was studied using rheology and\nIR spectroscopy. Various amounts of genipin powder were mixed into\nthe PNF suspension and incubated at 50 °C for 14 h before the\nmeasurements, as the reaction is more efficient at elevated temperatures\n(50–60 °C). 26 The storage modulus\nof all PNF suspensions was higher than the loss modulus (not shown),\nsuggesting that the PNF suspensions were more elastic than viscous. 27 The storage modulus–oscillation strain\ncurves of these PNF suspensions, with and without genipin, showed\na similar shape ( Figure 1 a), indicating a similar behavior when subjected to an oscillating\nstrain of 0.1% to 100%. The storage modulus of all samples was relatively\nstable at strains lower than ca . 1% and decreased\nwith increasing strain with a similar slope. The addition of 1 wt\n% genipin (with respect to the PNF content) displayed a small effect\non the modulus–strain curve. An increase in the amount of genipin\n(>2 wt %) shifted the curves to higher storage modulus, and the\nmodulus\nincreased with increasing genipin content in the PNF suspension. This\nsuggests the formation of a strong network with a strength dependent\non the amount of genipin present in the system. Figure 1 (a) Storage modulus of\nthe cross-linked and non-cross-linked PNF\nnetworks (curved fibrils, 16 g/L) versus oscillatory\nstrain. (b) IR spectra for these PNF–genipin samples. The amount\nof added genipin in relation to the increase in C–N stretching\nis indicated. (c) Protein cross-linking mechanism by genipin in acidic\nconditions. The generated new C–N bonds in the product were\nhighlighted in yellow as examples. The formation of genipin cross-links between nanofibrils\nwas further\nsupported by the IR spectra of the PNF–genipin samples ( Figure 1 b). It has been reported\nthat the primary amino groups attack the genipin C-3 carbon, forming\nheterocyclic amines, which further associate, generating cross-linked\nnetworks with short genipin oligomer bridges as illustrated in Figure 1 c. 28 − 30 The reaction\nis also accompanied by nucleophilic substitution of the ester group\non genipin by primary amine groups in acidic conditions. Both reaction\nmechanisms generated new C–N bonds, observed in the IR results.\nThe absorbance peak at ca . 1100 cm –1 , stemming from the C–N stretching, is most prominent in the\n10 wt % genipin–PNF sample and but also visible as a shoulder\nin spectra of the PNF samples cross-linked with 2 and 5 wt % of genipin. 29 The presence of the peaks in all spectra with\nadded genipin is confirmed by the second derivative spectra ( Figure S2 ). The increase of the peak intensity\ncorrelated well with the rheology results described above, suggesting\nthe formation of more cross-links. However, it is difficult to determine\nthe amount of genipin reacted with protein to form oligomer bridges\nand the remaining unreacted amount. The amide I and II regions (1500–1700\ncm –1 ) in the IR spectra of the PNFs with and without\ngenipin did not show significant peak shifts or shape variations,\nindicating that the PNF structures remained intact in the cross-linked\nnetwork. Genipin is expected to primarily react with amine groups\n( i . e ., lysine side chains). The\nlysine content in the WPI PNFs is between 5% and 10%, depending on\nwhich parts of the β-lactoglobulin sequence are present in the\nfinal PNFs. This gives a genipin:lysine ratio between 0.5 and 1.5\nfor 10% genipin. Hence increased cross-linking is expected up to 10%\ngenipin. Mechanical Properties of the Cross-Linked PNF Microfiber The microfiber was formed using a double-focusing millimeter-scale\ndevice with a core flow (Q1) of the curved PNF suspension and two\nsheath flows (Q2, Q3) for focusing ( Figure S1a ). Genipin was added directly in the core flow mixed with a PNF suspension\nrather than in the second sheath or bath solution, to maximize the\ninteractions between PNFs and genipin used in the system. The microfiber\nwithout the addition of genipin remained colorless after 14 h of incubation\nat 50 °C; see Figure 2 a. In contrast, the microfibers with genipin obtained a brown/dark\ngreen color after the same treatment ( Figure 2 b–e), which has been previously reported\nand suggested to be a result of the reaction between the primary amines\nand genipin. 26 , 31 The color of the fiber darkened\nwith an increasing amount of genipin added in the PNF suspension,\nindicating that more genipin reacted with the PNFs. This observation\ncorrelated well with the rheology and IR results ( Figure 1 ). Figure 2 Images of the nanostructured\nprotein microfiber assembled from\n20 g/L curved PNFs in acetate buffer solution (pH 5.2) without genipin\n(a) and cross-linked with 1 (b), 2 (c), 5 (d), and 10 (e) wt % of\ngenipin after 14 h of incubation at 50 °C. The scale bar is 300\nμm. The formation of the cross-links within the PNF\nnetworks also improved\nthe microfiber’s mechanical properties (see Figure 3 ). The presence of a low amount\nof genipin (<2 wt %) increased mainly the Young’s modulus\n(from now on referred to as modulus) and the stress at break of the\nfiber, while a higher amount of genipin addition ( i . e ., 5 and 10 wt %) increased the strain at break\nof the fiber ( Figure 3 a and b). The microfiber spun solely from a PNF suspension showed\na modulus of ca . 0.33 GPa (which is in the expected\nrange 17 ) and stress at break of ca . 8.5 MPa. The higher stress at break and elongation at\nbreak (3%) obtained here compared with those reported in the previous\nwork (3 MPa and 1.5%, respectively) could be the effect of the slightly\nhigher PNF concentration used herein and the 4-day conditioning before\nthe tensile tests (in previous work 17 the\nfibers were tested in the dry state). The modulus and the stress at\nbreak increased to 0.51 GPa and 15 MPa, respectively, when 1 wt %\nof genipin was added into the suspension before the spinning process.\nA more significant increase of the modulus and stress at break occurred\nwith the addition of 2 wt % genipin, resulting in 1.4 GPa and 22 MPa\n( Figure 3 a). However,\na higher concentration of genipin did not further increase the modulus\nand stress at break, which remained at 1.4 GPa and 20 MPa with 5 wt\n% genipin and slightly decreased to 1.2 GPa and 17 MPa, respectively,\nat 10 wt %. In contrast, the strain at break of the fiber cross-linked\nwith 10 wt % genipin increased 2 times, reaching ca . 6%; see Figure 3 b. The energy at break of the fiber with 10 wt % genipin (0.81 ±\n0.03 MJ/m 3 ) was 2 and 5 times that of the fiber with 2\nwt % (0.39 ± 0.09 MJ/m 3 ) and without genipin (0.16\n± 0.05 MJ/m 3 ), respectively ( Figure 3 c). Figure 3 (a) Mechanical properties of the microfibers\nspun from curved PNFs\n(20 g/L) with genipin, measured at 50% relative humidity. (b) Representative\nstress–strain curves of microfibers spun from curved PNFs with\n0, 5, and 10 wt % genipin. (c) Plot of energy at break relative to\nthe genipin content in the fibers. The mechanical performance of the produced micrometer-scale\nmaterials\nis governed by the arrangement of the building blocks and the interfaces\nthat join these building blocks. 2 , 3 The addition of genipin\nincreased the cohesion between the nanofibrils ( Figure 1 ), formed stronger fibril interfaces that\nfacilitated load transfer during testing, and resulted in a higher\nmodulus and stress at break ( Figure 3 ). The finding that higher amounts of genipin (>5\nwt\n%) do not lead to a further increase in modulus and strength suggests\na limit for the cohesive forces despite the fact that more genipin\nhas reacted with the PNFs. This behavior differs from the rheology\nof bulk samples ( Figure 1 a). The packing of the PNFs that results from the flow-assisted assembly\nmay not allow further reactions between the bound genipin groups to\nform interfibrillar cross-links. Moreover, the higher extensibility\nof the microfiber at high genipin content suggests increased plasticity\nof the material ( Figure 3 b and c). Since no plasticizer was added, this effect could result\neither from unreacted genipin in the fiber or from an enhanced water\nuptake in the cross-linked fiber. Water has a strong plasticizing\neffect on bio-based materials, which is evident, for example, from\nour previous work on cellulose fibers (see Figure 3 b in Mittal et al . 19 ). Previous work has demonstrated that\nthe addition of genipin significantly increases the water uptake in\nvarious biopolymer materials. 32 − 34 To investigate if the whey PNF\nmaterials show the same behavior, we measured the water uptake in\nsamples with 5% or 10% genipin (reacted under similar conditions to\nthose in the microfibers) and compared it with the material without\ngenipin. The water content was quantified by thermogravimetric analysis\n(TGA) ( Figure S3 ) and was found to be 6.2%,\n5.6%, and 5.1% for the samples with 0%, 5%, and 10% genipin, respectively.\nHence, the water uptake is lower in the samples reacted with genipin,\nwhich is expected for a system with a higher degree of cross-linking.\nThe results show that increased water content cannot be the origin\nof the plastication effect at 5–10% genipin content and suggest\nthat genipin itself is responsible for the observed behavior. Changes\nin mechanical properties similar to what we observe ( i . e ., decreased modulus and stress at break and increased\nstrain at break) have been reported for films made from starch and\npotato protein with genipin added. and one of the explanations for\nthis behavior suggested in that work was that genipin could act as\na plasticizer. 35 Another study on elastin\ncross-linked with genipin also reported that the compressive modulus\nlevels out at 7–10% genipin. 36 Further\nexploration of the mechanism behind this behavior requires detailed\nexperimental analysis using methods that can quantitatively distinguish\nbetween genipin in different states. At least four states may exist\nin the fiber: (i) monomeric, (ii) oligomeric, (iii) reacted with PNFs,\nand (iv) reacted with PNFs and cross-linked. The two first states\nmay act as plasticizers, while the effect of state (iii) on the mechanical\nproperties is difficult to predict. Effect of PNF Morphology and Sheath Flow Rate on the Mechanical\nProperties of the Microfiber The PNF suspension used to assemble\nmicrofibers described so far contained the PNFs that were short and\ncurved with a mean fibril end-to-end length of 0.3 μm ( Figure S4b ). By changing the initial WPI concentration,\nstraight and longer PNFs with a mean fibril end-to-end length of ca . 1.2 μm could be produced ( Figure S4a ). 6 , 17 The persistence length of the\nstraight fibrils was previously reported to be ca . 1960 nm, which was almost 50 times higher than that for the curved\nfibrils (41 nm). 17 However, the hydrogel\nfibers assembled from only straight PNFs were not strong enough to\novercome the surface tension at the liquid–air interface when\npulled out from the acetate bath, potentially due to a lack of fibril–fibril\nentanglements. 17 To strengthen the interactions\nbetween the straight PNFs, 10 wt % of genipin was added into the spinning\nsuspension to cross-link the fibrils and improve the mechanical properties\nof the final fibers. However, the hydrodynamic focusing of the straight\nPNF suspension was not successful using the same flow parameters used\npreviously for the curved fibrils. The core flow of the suspension\ndid not remain as a stable flow parallel to the flow channel but tended\nto twist/oscillate and clog the channel. This could originate from\nthe viscosity difference between the straight and curved PNF suspensions\n( Figure S5 ). Evidently, the suspension\nproperties with the straight PNFs were not optimal for a stable flow\nand a resulting continuous ejected microfiber. 21 , 37 The unstable flow and oscillations were addressed by increasing\nthe flow rate of the second sheath flow (Q3) to 33.9 mL/h (from 24.9\nmL/h). The spinning of the curved PNF suspension was repeated at the\nelevated Q3 sheath flow rate to compare the final fibers. The\nstress–strain curves of the fiber assembled from straight PNFs\nshowed a stiffer behavior compared with that of the fibers spun from\ncurved PNFs ( Figure 4 a). The microfiber from straight PNFs had a modulus as high as 1.6\nGPa and stress at break of ca . 20 MPa, which is around\n4 times higher than that for the microfibers from the curved PNFs\n(0.38 GPa and 6 MPa, respectively); see Figure 4 b. However, fibers from curved PNFs showed\na higher strain at break of ca . 10% and absorbed\ntwice the amount of energy (∼0.68 MJ/m 3 ) compared\nwith fibers from straight PNFs (0.30 kJ/m 3 ) before fracture\n( Figure 4 b). The high\nmodulus of the straight-PNF-derived microfiber is suggested to originate\nfrom the higher modulus of straight PNFs compared to that of curved\nfibrils indicated by persistence length, as described above. The higher\ndegree of alignment of straight PNFs in the microchannel could also\ncontribute to a higher modulus of the final fiber. 16 , 19 It has been shown that straight PNFs form more aligned structures\nalong the flow channel than the curved fibrils at the same focusing\ncondition. 17 In contrast, the higher strain\nat break (10%) observed for the curved-PNF fiber indicates a higher\ntendency of these fibrils to entangle/aggregate than the straight\nfibrils. The different mechanical behavior resulting from the different\nmorphologies of the fibrils also provides ways to harvest desired\nproperties of the microfiber by varying the composition of these two\ntypes of fibrils in the suspension. Figure 4 (a) Stress–strain curves of microfibers\nassembled from straight\nand curved PNFs with 10% genipin at different sheath rate (illustrated\nas the number in the legend, unit: mL/h). (b) Mechanical properties\nof the fibers at a sheath rate of 33.9 mL/h. It is noteworthy to mention that the fibers spun\nfrom curved PNFs\nat the higher Q3 sheath rate (33.9 mL/h) have lower modulus but are\nmore extensible than the fibers assembled at a Q3 rate of 24.9 mL/h\n( Figure 4 a). This is\nin opposition to previous observations of fiber assembly from cellulose\nnanofibrils or straight PNFs, as increased sheath flow rate results\nin higher acceleration of the core and typically results in a higher\ndegree of fibril alignment and a higher fiber modulus. 16 , 21 , 22 , 39 To further address this contradiction, in situ small-angle\nX-ray scattering (SAXS) measurements of the curved fibril alignment\nin the channel, under different flow conditions, were employed. Due\nto the low signal/noise ratio downstream of the second sheath flow\n(Q3), reliable data on the effect of the Q3 flow rate could not be\nobtained. Instead we analyzed the behavior when altering the Q2 flow\nrate. The results in Figure 5 show that the order parameter of PNFs in the channel was\nrelatively low (0.2) and did not increase with increased Q2 flow rate.\nThis indicates that the effects in spinnability and modulus that we\nobserve do not originate from fibril alignment but other, presently\nunknown, aspects of the structure created during hydrodynamic assembly.\nOne explanation could be that the higher ejection rate results in\na less dense hydrogel fiber, which reduces the number of genipin cross-links.\nThe dried fiber obtained at the higher Q3 flow rate is indeed slightly\nthinner ( vide infra ). Figure 5 (a) Schematic of the\nchannel geometry employed for the SAXS experiments.\nWhite circles show the different downstream positions at the center\nof the channel where in situ SAXS measurements were\ncarried out. (b) Example of SAXS pattern at the location where the\nstrongest fibril alignment is observed. (c) Local order parameters\ncalculated from SAXS patterns with Q2 flow rates of 4.7 and 7.1 mL/h.\n(d) The greatest order parameter, i . e ., at position Z = 3 mm, as a function of Q2 flow\nrate. The given order parameter is an average value of 10 measurements. The strength and modulus of the microfiber, illustrated\nas yellow\n(curved PNFs) and red (straight PNFs) triangles in the black area\nin Figure 6 , vary almost\n1 order of magnitude due to the change in the morphology of the PNF\nbuilding blocks, the amount of cross-linker, and the flow-focusing\nparameters ( Figure 6 ). The wide range in mechanical values in the presence of different\namounts of genipin emphasizes the key role of the interactions between\nfibrils in controlling the mechanical properties of microscale materials.\nThe microfibers fabricated in the present work preserved the modulus\nof individual PNFs after the microfluidic spinning process. The fibers\nhave specific moduli in the range similar to ligaments, 40 which is comparable to silk fiber and overall\nhigher than synthetic low-density polyethylene (LDPE); see Figure 6 . 41 , 42 The specific strength is equivalent to that of wood and higher than\nmost natural elastomers, e . g ., muscle,\ncork, and leather. 41 Figure 6 Mechanical properties\nof bio-based and synthetic materials are\ndisplayed as a plot of specific modulus versus specific\nstrength. The data of the PNF microfiber fabricated in the present\nwork are illustrated as triangles in the black region in the figure,\nwhere the yellow and the red refer to the fibers from curved and straight\nfibrils, respectively. The red square represents the mechanical properties\nof the fiber produced by Kamada et al . 17 Morphology of the Dried Microfiber The air-dried fibers\nmade from curved PNFs with (not shown) and without genipin ( Figure 7 a) showed a smooth\nsurface with a constant diameter of 35 ± 2 μm along the\nfiber direction. A closer view of the surface showed that the PNFs\nassembled into closely packed graupel-like units with a size of 50–200\nnm ( Figure 7 a), similar\nto the structure observed previously in the assembly of whey PNFs 17 and recombinant spider silk proteins. 43 The addition of genipin did not significantly\naffect the surface morphology of the final fibers, as granular PNF\naggregates were also observed on the surface of the cross-linked fibers\n( Figure S6a ). The micrograph of the fiber\ncross-section after the tensile test demonstrates a rough fracture\nsurface ( Figures 7 b\nand S5b ). The homogeneous cross-section\nof the cross-linked fiber ( Figure S6b )\nindicates a good distribution of genipin within the fiber, which avoids\ndifferent failure behavior along the transversal direction under stress.\nNanosized fibril-like objects, presumably originating from stretched\nfibrils/fibril aggregates, were also observed in the cross-sections\nof the fibers cross-linked with 5 and 10 wt % genipin ( Figure S6c ), in accordance with the observed\nhigh strain and energy at break of these fibers compared to those\nwith a lower amount of genipin. Figure 7 SEM images of the surface (a) and the\ntensile-fracture cross-section\n(b) of the microfibers assembled from curved PNFs solely. Surface\nof the fiber assembled from curved PNFs (c) and straight PNFs (d)\nwith 10 wt % of genipin at a sheath rate of 33 mL/h. The scale bars\nin the left/inset images are 20 μm. The diameter of the dried fibers assembled at the\nhigher sheath\nflow rate (33 mL/h) was 30 ± 4 μm, slightly smaller than\nthe value of the fiber studied previously (35 ± 2 μm).\nThe surface of the fiber assembled from curved PNFs remained smooth\nregardless of the second sheath flow rate. A magnified view showed\nclose-packed oval-shaped aggregates with the long axis orientated\nparallel to the fiber direction ( Figure 7 c). Compared to the graupel-like units observed\npreviously, the elongated unit (due to the increased sheath rate)\nmay detach more easily from the neighbor units under stress and result\nin a lower stiffness value ( Figure 4 ). In contrast to the fiber composed of curved PNFs,\nthe fiber from straight PNFs did not have a perfect cylinder shape\nafter drying ( Figure 7 d). Moreover, the straight PNFs did not aggregate into graupel-like\nunits during the assembling of the microfiber. Instead it showed a\ntexture in agreement with a fiber with aligned constituents, commonly\nobserved in the fibers assembled from cellulose nanofibrils via a similar microfluidic method. 19 , 21 , 44 The different morphologies of the microfiber\nassembled from the two types of fibrils indicated fundamental differences\nin assembly mechanisms, which resulted in the different mechanical\nperformance of the two fibers described in Figure 4 ."
} | 5,750 |
39847330 | PMC11789028 | pmc | 5,837 | {
"abstract": "Significance Neocortical circuits are characterized by complex oscillatory dynamics. Whether these oscillations serve computations or are an epiphenomenon is still debated. To answer this question, we designed a computational model of a recurrent network that allows control of oscillatory dynamics (harmonic oscillator recurrent network, HORN). When operating in an oscillatory regime, HORNs outperform nonoscillatory recurrent networks in terms of learning speed, noise tolerance, and parameter efficiency. Moreover, they closely replicate the dynamics of neuronal systems, suggesting that biological neural networks are likely to also exploit the unique properties offered by oscillatory dynamics for computing. The interference patterns provided by wave-based responses allow for a holistic representation and highly parallel encoding of both spatial and temporal relations among stimulus features.",
"conclusion": "Concluding Remarks. Taken together, the present results not only unveil the computational principles accessible to HORNs and other oscillator networks but also allow for a functional interpretation of numerous experimentally verified physiological phenomena whose roles in information processing have been elusive or have caused controversial discussions. Plausible functional roles can now be assigned to i) the propensity of nodes to oscillate and the resulting dynamical phenomena such as synchronization, desynchronization, resonance, entrainment, and traveling waves ( 12 – 15 , 17 ), ii) the diversity of preferred oscillation frequencies, their nonstationarity and context dependence ( 3 , 4 ), iii) the heterogeneity of the conduction velocities of the recurrent connections ( 5 , 6 ), iv) the decrease of oscillation frequencies in higher areas of the cortical processing hierarchy ( 11 , 72 ), v) the Hebbian adaptivity of recurrent connections ( 7 , 8 ), vi) the emergence of context-dependent dynamic receptive fields by network interactions ( 60 , 61 ), and vii) the reduction of variance in network dynamics during stimulus presentation ( 63 ). The simulations also suggest a physiologically plausible scenario for the rapid and parallel matching of sensory evidence with stored priors through self-organized convergence of network dynamics to classifiable, stimulus-specific, dynamic substates. These substates consist of highly structured, high-dimensional dynamical landscapes that unfold due to interference of wave patterns in amplitude, frequency, and phase space. In essence, the described networks perform highly parallelized analog computations in high-dimensional state spaces that simultaneously relate a large number of spatially and temporally structured input variables, a capacity ideally suited to accomplish context-dependent feature binding and scene segmentation. Consequently, attempts are made to exploit the principle described in this study in machine learning architectures designed to perform scene segmentation ( 81 ). Moreover, the computational strategy implemented by HORNs is also well suited to overcome challenges requiring the simultaneous evaluation of multiple nested relations as occurring, for example, in language comprehension. Interestingly, biological systems are at ease with solving the binding problem, with the segmentation of cluttered scenes and with the analysis of complex time series (e.g., spoken language), while these tasks are notoriously difficult for digital computer architectures that typically rely on serial feedforward processing. We believe that nature solves such hard problems through analog computations of the kind described in this study. We predict that it will be possible to implement the computational principle presented here in analog hardware that runs at room temperature, is miniaturizable, and is highly energy efficient. Combined with electrical elements mimicking Hebbian synapses, such as memristors, this principle will likely enable the design of self-adapting devices for machine learning applications that can ideally complement existing digital technologies.",
"discussion": "Discussion Controlled Oscillations. Implementing characteristics of the mammalian cerebral cortex in RNNs revealed a powerful computational principle based on oscillatory activity. Although RNNs without oscillating nodes naturally produce oscillations, such emergent oscillations are often transient and difficult to control, hindering their exploitation by gradient-based learning. Enforcing oscillations at each node in HORNs allowed us to study the functional relevance of oscillatory dynamics, identify the computational principle responsible for the increased performance of HORNs, and establish close relationships with the dynamics of natural networks such as the cerebral cortex. In HORNs, individual nodes turn any input into an oscillation, acquiring the ability to extract features through resonance and, more generally, to modulate gain in a frequency-dependent manner. Networks, in turn, generate holistic transient stimulus representations characterized by wave interference patterns ( 37 , 38 , 66 ). Performance testing on standard pattern recognition benchmarks revealed that a gradient-based learning scheme can capitalize on this extended dynamical repertoire, leading to substantially enhanced performance relative to RNNs without oscillatory nodes. These findings of improved task performance are in line with previous studies in the field of machine learning that investigated RNNs with oscillating nodes ( 30 , 37 , 38 ) (while HORNs have 50% less trainable parameters for the same number of nodes). Enforcing oscillatory activity in network nodes serves as an inductive bias in RNNs, enhancing model expressivity ( 48 ). This oscillatory bias has also been shown to improve task performance in spiking networks by allowing subthreshold membrane potentials to oscillate ( 67 , 68 ). These findings suggest a universal computational principle based on coupled oscillators, enabling wave-based representations applicable to both neuronal populations, as well as single neurons ( 69 ). When trained with geometrically organized stimuli, HORNs developed local connectivity patterns. Such locally connected oscillator RNNs ( 37 , 38 ) can be interpreted as a discretization of a neural field model with a specific connection kernel that implements a damped wave equation ( 70 , 71 ). In this sense, local oscillations and global waves are two sides of the same coin, and the dynamics of HORNs and field models are equivalent under certain conditions, a topic left for future study. Reasons for Increased Performance. The good performance of HORNs is due to several reasons, and these are closely related to the propensity of network nodes to engage in oscillations. First, innate preferences for stimulus features (controlled by the parameters ω , γ , α ) allow individual nodes to efficiently extract and encode stimulus features already in an untrained network and contribute to the noise resilience of HORNs. If input signals lack temporal structure, the oscillatory properties of the nodes are still beneficial because they transform sustained inputs into oscillatory responses. This transformation allows computations in the common format of temporally modulated signals which prevail in the communication among nodes; see also the fourth point below. Second, the discretization scheme used for the oscillator differential equations introduces temporal residual connections that stabilize gradients in BPTT learning, and the oscillating dynamics can increase the practical expressivity of the networks by modulating gradients [ SI Appendix and ( 30 , 48 )]. Third, DHOs in a HORN network collectively process stimuli in a fully distributed manner by converting sensory input into waves. Initially, these are standing waves in each oscillator, but then they spread and cause complex interference patterns at the network level ( 16 , 17 ), findings that are compatible with physiological evidence ( 66 ). This representation provides a coding space of massive dimensionality, and, most importantly, permits the superposition of information about multiple spatially and temporally segregated events. This allows HORNs to analyze and encode simultaneously not only spatial but also temporal relations between a large number of stimulus features and to generate holistic representations of the correlation structure of complex input constellations. The Virtues of Heterogeneity. Heterogeneity improves the performance of RNNs because it increases the dimensionality of the state spaces of the networks. Having oscillatory nodes allowed us to increase heterogeneity by varying preferred oscillation frequencies, which increased task performance. In addition to enhancing heterogeneity by varying the preferred oscillation frequencies of the nodes, we induced heterogeneous conduction delays to deliberately induce phase shifts. This further increases the dimensionality of the networks’ coding space, which can be exploited for computation. The advantages of heterogeneity are also documented by simulations of two-layer networks. These networks showed enhanced performance at the same number of parameters, in particular when the higher layer operated at lower preferred frequencies than the lower layer. This allows the multilayer network to operate in different frequency bands and to perform parallel analyses of input patterns at different temporal scales in each layer. This finding was the result of a grid search for optimal parameter settings in two-layer networks and shares similarities with the organization of the cerebral cortex. Here, too, oscillation frequencies decrease as one progresses from lower to higher processing levels ( 11 , 72 ). Slower oscillations at higher levels can establish relations among temporally segregated stimuli over longer time intervals, which could support chunking. Interestingly, according to basic physics, waves with slower frequencies tend to travel over longer spatial distances, in our case over a larger number of network nodes. In the cerebral cortex, higher areas integrate information from increasingly diverse and spatially remote processing streams, as reflected by their large, often polymodal, and multiselective receptive fields. Assuming a wave-based representation ( 66 ), operating at lower oscillation frequencies would allow these higher areas to integrate information over larger temporal and spatial scales, favoring holistic processing of information and multimodal binding. Another advantage of heterogeneity is that it brings network dynamics closer to criticality ( 48 , 53 , 73 ). Dynamics close to criticality are a hallmark of cortical networks and provide computational benefits summarized in the “critical brain” hypothesis ( 74 ). These benefits are due to the emergence of long-lived transient and metastable states ( 54 ). HORNs also encode information in transients and therefore are capable of coding with sequences of metastable states and ghost attractors ( 75 ) when in a regime close to criticality. This distinguishes their dynamics from that of attractor networks ( 76 ) for which critical slowing down limits computational power near criticality ( 77 ). More studies are needed to better understand these transient dynamics in HORNs and to identify related activity in biological networks ( 75 , 78 ). In summary, we found that the implementation of physiologically plausible heterogeneity typically increases performance without increasing the number of trainable parameters. Heterogeneity i) gives even untrained networks sensitivity to diverse correlation structures, thereby accelerating learning; ii) enhances the processing of novel or noisy stimuli with varying spectra; iii) expands coding dimensions; iv) allows networks to utilize computational benefits resulting from dynamics closer to criticality, while v) at the same time reducing the need for costly parameter tuning. The gain of function by heterogeneity was particularly pronounced in larger networks, suggesting that one-shot learning, a hallmark of biological systems, is facilitated in large, heterogeneous networks such as the cerebral cortex. This lets us conclude that the apparent heterogeneity in natural neuronal systems is likely not a reflection of nature’s imprecision but rather an efficient solution to computational challenges. Relations to Neurobiological Systems. HORNs reproduced several characteristic features of the dynamics and organization of natural neuronal systems, particularly the cerebral cortex and probably also the hippocampus. In addition, the simulations allowed us to assign concrete functions to features of natural networks whose role in information processing is still a matter of discussion. Our simulations show that learning-dependent complex, transient, and stimulus-specific synchronization patterns benefit information processing and identify the oscillatory properties of network nodes as an underlying mechanism. This supports the hypothesis that oscillations and synchrony, also observable in neuronal systems ( 79 ), are functionally relevant and not epiphenomena. Simulations with geometrically organized input patterns processed by the visual system yielded results similar to those obtained with time-series data that do not contain geometric information. Thus, the identified computational principles can handle spatial and temporal relations among input signals similarly and represent computations in the same format. This benefits computations in sensory cortices receiving both temporally and spatially structured input and aiding cross-modal and interareal communication. For spatially structured stimuli, learning led to synaptic weight configurations that decay with distance and capture the Gestalt criteria of continuity and vicinity, a property also known from natural systems ( 2 , 65 ). In the visual cortex, the basic layout of recurrent connections is genetically determined, but experience-dependent pruning of these connections further enhances their selectivity through a Hebbian mechanism ( 80 ). Stimulation of locally connected HORNs led to traveling waves that closely resemble those observed in natural neuronal networks ( 16 , 66 ). Traveling waves are also a hallmark of oscillatory RNNs in which local connectivity was enforced by design ( 37 , 38 ). Wave-based representations allow for very high-dimensional representations and manifold coding strategies. Consequently, numerous hypotheses have been proposed regarding the functional role of traveling waves ( 17 , 69 ). In a wave-based model of the motor cortex, the direction and wavelength of traveling waves are used to structure commands in a way that is easily decodable by the dendritic arbors of neurons in the descending motor system ( 69 ). However, the exact function of traveling waves in the sensory cortices is still not fully understood. Another similarity between the dynamics of HORNs and the cerebral cortex is the temporal evolution of responses in simulations with geometrically structured stimuli. The initial transient responses were amplified by reverberation, increasing the decodability of the dynamic state due to better segregation of stimulus-specific principal components of the population vector ( 65 ). This state can be seen as a highly parallelized search for the best match between sensory evidence and learned priors ( 2 ). Thus, one of the core functions of predictive coding, the matching of sensory evidence with stored priors, can be realized through self-organizing dynamic interactions in oscillatory recurrent networks. During learning, nodes activated by semantically related features increase their mutual coupling, and during recall, these nodes self-organize into a stimulus-specific assembly with synchronized and jointly enhanced responses. This dynamic association of nodes is also observed in the visual cortex for neurons tuned to perceptually bound features ( 10 , 65 ) and is at the core of the binding by synchrony hypothesis (BBS) ( 20 ). HORNs, by exploiting the resonance properties of coupled oscillators, reproduce this important feature of natural cortical networks. The dynamics of spontaneously active HORNs resemble those of natural cortical networks in which stimulation decreases variance ( 63 ) and temporarily aligns dynamics to stimulus-specific substates. These substates exist within the subspace of spontaneous activity and arise from comparisons of sensory evidence and stored priors ( 2 ). Therefore, spontaneous activity can be seen as a blend of fragments of learned stimulus-specific representations. In addition to reproducing many physiological phenomena, additional physiological experiments can now be designed to examine specific predictions derived from the present study. These experiments will require massive parallel recordings of neuronal activity both within and across cortical areas with high spatial and temporal resolution to capture the spatiotemporal dynamics of traveling waves and their resulting interference patterns. Concluding Remarks. Taken together, the present results not only unveil the computational principles accessible to HORNs and other oscillator networks but also allow for a functional interpretation of numerous experimentally verified physiological phenomena whose roles in information processing have been elusive or have caused controversial discussions. Plausible functional roles can now be assigned to i) the propensity of nodes to oscillate and the resulting dynamical phenomena such as synchronization, desynchronization, resonance, entrainment, and traveling waves ( 12 – 15 , 17 ), ii) the diversity of preferred oscillation frequencies, their nonstationarity and context dependence ( 3 , 4 ), iii) the heterogeneity of the conduction velocities of the recurrent connections ( 5 , 6 ), iv) the decrease of oscillation frequencies in higher areas of the cortical processing hierarchy ( 11 , 72 ), v) the Hebbian adaptivity of recurrent connections ( 7 , 8 ), vi) the emergence of context-dependent dynamic receptive fields by network interactions ( 60 , 61 ), and vii) the reduction of variance in network dynamics during stimulus presentation ( 63 ). The simulations also suggest a physiologically plausible scenario for the rapid and parallel matching of sensory evidence with stored priors through self-organized convergence of network dynamics to classifiable, stimulus-specific, dynamic substates. These substates consist of highly structured, high-dimensional dynamical landscapes that unfold due to interference of wave patterns in amplitude, frequency, and phase space. In essence, the described networks perform highly parallelized analog computations in high-dimensional state spaces that simultaneously relate a large number of spatially and temporally structured input variables, a capacity ideally suited to accomplish context-dependent feature binding and scene segmentation. Consequently, attempts are made to exploit the principle described in this study in machine learning architectures designed to perform scene segmentation ( 81 ). Moreover, the computational strategy implemented by HORNs is also well suited to overcome challenges requiring the simultaneous evaluation of multiple nested relations as occurring, for example, in language comprehension. Interestingly, biological systems are at ease with solving the binding problem, with the segmentation of cluttered scenes and with the analysis of complex time series (e.g., spoken language), while these tasks are notoriously difficult for digital computer architectures that typically rely on serial feedforward processing. We believe that nature solves such hard problems through analog computations of the kind described in this study. We predict that it will be possible to implement the computational principle presented here in analog hardware that runs at room temperature, is miniaturizable, and is highly energy efficient. Combined with electrical elements mimicking Hebbian synapses, such as memristors, this principle will likely enable the design of self-adapting devices for machine learning applications that can ideally complement existing digital technologies."
} | 5,066 |
28607345 | PMC5468245 | pmc | 5,838 | {
"abstract": "Lipid bodies (LBs) in the coral gastrodermal tissues are key organelles in the regulation of endosymbiosis and exhibit a diel rhythmicity. Using the scleractinian Euphyllia glabrescens collected across the diel cycle, we observed temporally dynamic lipid profiles in three cellular compartments: host coral gastrodermal cells, LBs, and in hospite Symbiodinium . Particularly, the lipidome varied over time, demonstrating the temporally variable nature of the coral– Symbiodinium endosymbiosis. The lipidome-scale data highlight the dynamic, light-driven metabolism of such associations and reveal that LBs are not only lipid storage organelles but also act as a relay center in metabolic trafficking. Furthermore, lipogenesis in LBs is significantly regulated by coral hosts and the lipid metabolites within holobionts featured predominantly triacylglycerols, sterol esters, and free fatty acids. Given these findings through a time-varied lipidome status, the present study provided valuable insights likely to be crucial to understand the cellular biology of the coral– Symbiodinium endosymbiosis.",
"introduction": "Introduction The coral–dinoflagellate endosymbiosis is generally defined as mutualistic because both partners benefit from the relationship 1 . Coral hosts provide certain carbon and nitrogen skeletons for the photosynthetic reactions of Symbiodinium , and Symbiodinium translocate most of their photosynthetically fixed carbon, including glycerol, glucose, amino acids, and lipids, to the coral host 2 – 4 . Symbiodinium may incorporate various lipids, sugars, and amino acids into the host cells or tissues 5 , 6 . Irrespective of the species of Symbiodinium , translocated metabolites affect the host lipid levels, thereby affecting reproduction, growth, and ultimately fitness of the host 7 , 8 . Lipids are the primary long-term source of stored energy in corals, and several studies have indicated that lipids play a pivotal role during endosymbiosis regulation 9 , 10 . Moreover, fatty acid profiles, including the concentration and composition of the saturated fatty acid (SFAs) and polyunsaturated fatty acids (PUFAs) pools, are reflective of lipid biogenesis in hospite and may be used to highlight lipid flux within the holobiont (host and endosymbionts) 11 . During endosymbiosis, various lipids such as wax esters (WEs), triacylglycerides (TAGs), sterol esters (SEs), cholesterols (Cols), phospholipids (PLs), and free fatty acids (FFAs) accumulate within organelle-like structures known as LBs 12 – 16 . From archaea to mammals, one of the main functions of LBs is intracellular trafficking of lipids, thereby emphasizing their critical role in cell biology 17 , 18 . To date, the functional formation of coral LBs during endosymbiosis remains to be elucidated 16 . The relationship between LB formation and the regulation of the coral–dinoflagellate endosymbiosis is evidently quite intimate, and studies describing the cellular mechanisms underlying the regulation of lipid biogenesis have been mostly cursory and preliminary. Previous studies on protein sequencing data and microscopic observations suggest that both Symbiodinium and host organelles are involved in LB biogenesis 13 , 19 . Furthermore, Luo et al . and Chen et al . have demonstrated that coral endosymbiotic status positively correlated with symbiont population and variable changes of LBs, and implied that endosymbiosis may be dynamic 12 , 14 . These studies also revealed that dynamic LBs may highly correlate with potentially critical consequences for symbiosis function. In particular, coral gastrodermal LBs change in density, morphology, composition, and intracellular distribution over diel cycles 14 . However, no study has investigated diel lipidomic patterns of various cellular compartments in corals to determine the relative importance of diel changes. Therefore, “time and composition” information regarding LB biogenesis within the host cell cytoplasm is crucial. Because coral metabolism is easily influenced by light or photosynthesis 20 , it was particularly hypothesized that the lipidomes within different cellular compartments vary over diel cycles. The present study investigating the temporal lipidome of LB biogenesis, metabolism, and trafficking within the coral host– Symbiodinium association across the diel cycle may provide an insight into endosymbiotic regulation. We hope that by shedding light on the lipid composition of these cellular fractions at various time scales, a more efficiently defined working model for intra-holobiont lipid trafficking can be established.",
"discussion": "Discussion Coral lipids profiles and FAs composition are related to the mode of nutrient acquisition, and lipidomes may be strongly affected by symbiotic status and environmental conditions 7 , 12 , 21 , 22 . Understanding the dynamic change of the lipidome enables the elucidation of endosymbiosis regulation. However, lipid profiling in anthozoan–dinoflagellate mutualisms is complicated by the dual-compartmental nature of the holobiont, an entity whose molecular regulation remains unclear 23 . Furthermore, temporal changes in LB distribution, composition, and morphology, all of which appear to suggest that the host- Symbiodinium endosymbiosis is quite dynamic over the diel cycle with respect to metabolite production and catabolism 14 . In the present research, we investigated the nature of endosymbiosis through lipidomic analysis using a time-varied compartmental approach, and revealed the changes in marker lipids and LBs as regulators during metabolism pathways. Temporal dynamic of endosymbiosis status In this study, we sought to unravel a portion of the lipidome molecular regulation of an anthozoan–dinoflagellate endosymbiosis in three cellular fractions: the host gastrodermal cells, LBs, and in hospite Symbiodinium populations, at four points of the diel cycle. Total lipids and FAs pools exhibited diel variation in all three cellular compartments; particularly, the overall lipid profile of the host showed a distinct diel fluctuation pattern compared with that of the other two compartments (Fig. 1 ). Essentially, the diel variation in holobiont lipidome (discussed in detail in further sections) is likely to have critical implications in the maintenance of symbiotic homeostasis 24 , 25 . In particular, there is a relatively diel pattern of FA moiety abundance which differs between host cells and LBs (Fig. 1 ). SFAs and PUFAs are essential for biological membrane synthesis and play a critical role in the regulation of coral metabolism 26 . The proportion of SFAs and PUFAs revealed a distinct temporal pattern, indicating that the lipogenesis in individual compartments of the holobiont varies over the diel cycle. As shown in Fig. 1C , Symbiodinium FA moieties accumulated during the day and were consumed at night, exhibiting a circadian lipogenesis pattern. Acetyl-CoA carboxylase is a critical light-dependent enzyme that regulates the FA synthesis rates 27 , 28 . Light-driven lipogenesis by microalgae, which has been thoroughly studied by Crossland and Mortillaro et al ., may have contributed to the temporal variation in relative FA levels 29 , 30 . Previous studies have shown that marine invertebrates and algae typically synthesize more SFAs during the day and more PUFAs at night 31 , 32 . This finding may indicate that acyl chains of host cells and Symbiodinium , which are necessary for the construction of more complex lipid species, are involved in LB lipogenesis during the daylight hours, accounting for the decrease in host coral SFAs during this period. Thus, the lipid metabolites of the holobiont are accumulated into LBs and influenced by the diel cycle and internal cellular metabolic balance. This hypothesis suggests that lipid trafficking and lipogenesis associated with LB formation causes temporal variation of lipid metabolism in animal cells. Previous studies have determined that many LBs were typically only found within host cells housing endosymbionts, but not in non-symbiotic host cells; this finding highlights the importance of these organelles in endosymbiosis 12 , 14 . Specifically, LB density and lipid concentrations decreased to baseline levels during the night, which may imply that corals are highly susceptible and less resistant to environmental stress. In the present study, we characterized the diel fluctuations of lipid profiles in healthy coral LBs under normal illumination (Figs 1 – 3 ). We observed that the FAs pools in each compartment varied from sunrise to midnight (Fig. 2 ), and identified the prominent FAs that contributed most to the variation in the PCA plot (Fig. 2B–D ). Our results of a dynamic lipidome in the animal coral–plant algae interaction clearly suggest that coral– Symbiodinium endosymbiosis is a temporally variable status. In addition to shedding light on the underlying biology of this environmentally crucial endosymbiosis, previous studies 33 , 34 have evaluated the suitability of various approaches for estimating coral health based on a variety of parameters. However, no study has investigated the diel changes in lipid metabolic indices (e.g., lipids, FAs, SFAs, and PUFAs) among various cellular compartments simultaneously. Our results provide valuable lipidomic indices for evaluating coral endosymbiotic status at various timescales and light levels. Temporal lipogenesis changes result in diel rhythmicity of LB formation The diel variability in LB formation observed by Chen et al . is closely related to the dynamic changes in lipid composition observed in the present study 14 . Many lipidomic studies have revealed that TAGs and SEs serve as intracellular storage molecules accumulated in LBs 35 , 36 . The coral hosts accumulated these storage lipids during the light period (Fig. 3A and C ) and synthesized the structural lipids PLs and Cols at night (Fig. 3D and F ), as previously reported 37 . Nevertheless, the temporal contribution of individual lipids pools to LB formation revealed that Symbiodinium derived TAGs were involved in the lipogenesis of LBs after sunrise, while the involvement of host derived TAGs was highest at sunset (Figs 4A and 6 , [3]). Furthermore, SEs are essential storage molecules of cellular membrane components and are synthesized by cholesterol acyltransferase within the endoplasmic reticulum, or through an acyl-CoA-independent reaction by lecithin 35 , 38 . Circadian changes in SEs metabolism have been related to sterol esterase activity, which appears to be activated by sterols and acyl donors 39 , 40 . In the present study, FFAs and PLs concentrations of host cells showed a diel fluctuation trend antagonistic to that of SEs in LBs and host cells (Fig. 3 ). Also, it is highly implied that host SEs play a role in lipid metabolism of LBs throughout the day (Figs 4B and 6 ). This suggests that host cells may use FFAs and PLs as acyl donors to synthesize more SEs through LB formation during the day. Moreover, SEs synthesizing by Symbiodinium are likely, and involved in LB formation in the afternoon (Figs 3C , 4B and 6 ). Figure 6 Schematic summary of the temporal LB lipogenesis in coral- Symbiodinium endosymbiosis during diel cycle. The circle size of bold black open circles indicative of the total lipid abundance of LBs, and the contained pie charts show percentage of individual lipid species. Furthermore, the extending gray dashed line and gray circle of pie chart presented the proportion of saturated and polyunsaturated FAs in total lipids of LBs. The gray disc among LBs, host and Symbiodinium shows the inter-compartmental lipid involvement of lipid central axis. The filled and blank arrows indicate lipids likely to be involved in LB biogenesis with high and low level, respectively. The double-headed arrow presented lipids were exchanged between host and LBs. [1]–[4] indicated four time periods. \n Our finding that TAGs and FAs synthesis occurred primarily during the light period followed by a rapid decrease in the dark is in agreement with other studies which demonstrated that some key enzymes of the TAGs and FAs synthesis pathways are regulated by light and are involved in lipid-droplet or LB biosynthesis 31 , 41 . Furthermore, FA moieties of the host FFAs showed a strong similarity with those of LBs at sunset (Fig. 4C ). This may indicated that the host contributes FFAs to LB biogenesis at this time (Fig. 6 , [3]). This may be attributed to the importance of FFAs as building blocks of other lipids, and likely suggests a rapid lipid turnover and degradation in host cells, which mobilized implicated lipid metabolites (TAGs and SEs) from Symbiodinium and involved in LB formation. Apart from this, PLs, which are typically structural components in membranes, exhibited significant differences among the three compartments across the diel cycle (Fig. 4D ). As shown in other species, LBs originate from the endoplasmic reticulum in other eukaryotic cells, and the FAs composition of PLs varies among organelles 42 , 43 . WEs of LBs, specifically saturated WEs, reached the highest concentrations at sunset (Table 1 ), coinciding with the highest density of electron-transparent inclusion bodies (ETIs) visualized in Chen et al . 14 . It had been hypothesized that the predominant lipid species of these ETIs are WEs. WEs are not detected in Symbiodinium and are only present in host cells and LBs 16 , 44 . Previous studies have demonstrated that Symbiodinium may provide unesterified FAs, which are then reduced to fatty alcohol in the host and esterified with FAs to form WEs during the day 41 , 45 . Our data clearly indicate that coral animal cells contain substantial and dynamic quantities of WEs across the diel cycle. Nevertheless, LB migration across the gastroderm and the consequent density changes might imply that coral host cells utilize and metabolize the lipid pools of the LBs at night 14 . Moreover, host WEs may be used to generate LBs at sunrise (Figs 4E and 6 , [1]). These findings mean that lipids synthesized in the host cells by endoplasmic reticulum mediated lipogenesis are accumulated within LBs through the lipogenesis mediated by coral host during the light period 13 , 14 , 43 . Collectively, our data and previous studies suggest that the diel change in the lipid composition of LBs, in particular, may be intricately linked to the diel rhythm in their cellular abundance documented in previous studies. LBs as a lipid exchange point for lipid metabolites in endosymbiosis We observed that each compartment possesses a unique lipidome (Fig. 1 ); furthermore, PCA was used to gain insight into the diel variability in the three lipidomes of the E. glabrescent – Symbiodinium holobiont, and several FAs were especially important in explaining variation (Supplementary Figs S1 , S2 , S3 and Table S11 ). Some of these prominent FAs were only synthesized by Symbiodinium , including C18:2n–6, C18:3n–6, and C18:4n–3. as verified in previous studies 10 , 16 , 46 , 47 . Moreover, C20:3n–6, C20:4n–6, and C22:4n–6 were produced by the host only, which is also supported by previous studies 16 , 48 . This demonstrates that TAGs, SEs, and FFAs are among the most trafficked lipid species among different cellular compartments. Moreover, such trafficking, as the lipid central axis, may occur through the LBs (Figs 5 and 6 ). Nevertheless, some FAs were only detectable in specific lipid species of either host cells and LBs, host cells and Symbiodinium , or only in LBs and Symbiodinium . Possible trafficked lipids and FA moieties from Symbiodinium to the host 10 , 44 , or host to Symbiodinium \n 49 have been studied in hard corals for decades. However, the direct translocation of lipids and lipogenesis among LBs, coral host and Symbiodinium has not been reported 9 , 12 , 47 . It has been hypothesized that Symbiodinium synthesize sugars, lipids, and carbohydrates, which are then transported into the host (and LBs) during endosymbiosis 7 , 45 . In this study, we presented a hypothetical model of schematic summary of the temporal LB lipogenesis (Fig. 6 ) during a diel cycle. First, LB formation is endosymbiosis-dependent. According to the present LB lipidome, a previous conclusion based on LB morphological and distribution changes was further confirmed 12 , 14 : the endosymbiosis is dynamic rather static, and shows a diel pattern. For example, the marker organelles of endosymbiotic status, LBs, in total lipid concentration, lipid species composition, and FA moieties change dynamically from sunrise to midnight (Fig. 6 , [1]–[4]). Second, we hypothesize that there are also temporal changes of lipid flow within holobiont lipogenesis, according to diel examination which leads to two major discoveries: (a) Lipid transport occurred in both host and symbiont, and LB was the relay center across the day (Fig. 6 , [1]–[4]). Inter-compartmental lipids transport featured predominantly TAGs, SEs, and FFAs, which as the lipid central axis among the LBs, host, and Symbiodinium . Furthermore, WEs were only present in the host cells and LBs, suggesting that WEs trafficking only occurred between these compartments. (b) LB formation, particularly, the lipid compositions were temporally regulated by the activation of various pathways. For example, TAGs of Symbiodinium were involved in LB lipogenesis from noon to midnight (Fig. 6 , [2]–[4]); the host were highly contributed TAGs to LB lipogenesis at sunset (Fig. 6 , [3]); SEs of Symbiodinium may have participated in LB biogenesis from sunset to midnight and subsequently again at sunrise (Fig. 6 , [3]–[4] followed by [1]); SEs and FFAs of the host were associated with LB formation during the light period (Fig. 6 , [1]–[3]); WEs of the host were initially implicated in the biogenesis of LB at sunrise (Figs 6 , [1]). Yet, the WEs species, palmityl oleate, was only present in LBs from noon to midnight (Fig. 6 , [2]–[4]). In our study, compartment-specific and temporal changes in the lipidome of a hermatypic coral were analyzed, which provided experimental evidence of the temporally variable status of coral– Symbiodinium endosymbiosis. The diurnal rhythmicity of LB formation documented in prior studies may have driven by such the temporal lipidomes of dynamic lipogenesis. Furthermore, our data show that LBs serve as a critical relay center for lipid metabolites trafficking between host and endosymbionts, meaning that they are key regulators of this ecologically important endosymbiosis. According to our review of relevant literature, the present investigation is the first to explore diel fluctuations in the lipidomes of all three compartments in the endosymbiotic coral holobiont. Future studies must investigate the role of lipids trafficking and monitor the lipidomics of host gastrodermal cells under various photosynthetic states."
} | 4,745 |
40124227 | PMC11930379 | pmc | 5,839 | {
"abstract": "ABSTRACT Soil ecosystems host diverse microbial communities, which are influenced by various environmental factors, soil properties, vegetation characteristics, and anthropogenic activities, such as livestock grazing. Grazing serves as a critical management practice in the alpine grasslands of the Qinghai‐Tibet Plateau, affecting soil microbial communities and their functions through processes such as forage consumption, trampling, and the deposition of feces and urine. In this study, we utilized the scientific and technological platform “Alpine Grassland‐Livestock Adaptive Management Technology Platform” in Qinghai Province to examine the effects of grazing intensity on soil microbial communities and functions. Experimental treatments included different grazing intensities (light grazing, moderate grazing, and heavy grazing), along with a no‐grazing control. Metagenomic sequencing technology was employed to investigate the impact of these grazing intensities on the microbial community composition and functional attributes in alpine grasslands. The results indicated that: (1) Actinobacteria, Proteobacteria, and Chloroflexi were the dominant bacterial communities in the soil, while Ascomycota, Mucoromycota, and Basidiomycota represented the primary fungal communities. (2) Grazing had a greater impact on soil fungal communities than on bacterial communities, altering the Shannon diversity index and Simpson index of soil fungal communities. (3) Soil pH and soil moisture were important factors influencing changes in soil microbial communities. (4) Functional analysis focusing on the “nitrogen metabolism” pathway indicated that under light grazing conditions, the relative abundance of multiple functional genes, particularly those involved in denitrification, decreased.",
"conclusion": "5 Conclusion Grazing livestock exerted direct and indirect effects on soil microbial communities through foraging, trampling, and fecal and urinary excretion, thereby influencing the stability of alpine grassland ecosystems. The results of this study indicated that: (1) Grazing exerted a more significant impact on soil fungal communities than on soil bacterial communities. (2) Grazing increased the relative abundance of Cyanobacteria in the soil. (3) Soil pH and moisture content were the primary factors driving changes in soil bacterial and fungal communities. (4) Grazing altered the composition and structure of soil microbial communities but had a minimal impact on microbial functions. This study revealed the effects of different grazing intensities on the structure and function of soil microbial communities in alpine grasslands around Qinghai Lake on the Qinghai‐Tibet Plateau, providing a reliable theoretical foundation for the sustainable management of the region.",
"introduction": "1 Introduction The Qinghai‐Tibet Plateau not only possesses a unique geographical location but also abundant biological resources. Alpine grassland, accounting for approximately 60% of the Plateau's total area, is one of the most unique grassland ecosystems globally and plays a crucial role in maintaining biodiversity and regional economic development on the Qinghai‐Tibet Plateau (H. K. Wang et al. 2022 ). Grazing is a primary method of utilizing alpine grasslands on the Qinghai‐Tibet Plateau (Yang et al. 2013 ). However, the alpine grassland on the Qinghai‐Tibet Plateau has experienced degradation to varying degrees as a result of climate change, grazing, and other factors (Zhang et al. 2016 ). Long‐term overgrazing can cause irreversible damage to alpine grasslands, but solely restricting grazing would affect the economic development of the Qinghai‐Tibet Plateau region. Studies have shown that the degree of grassland degradation and the decline in soil nutrient quality were influenced by different grazing intensities (Ren 2012 ). Additionally, Hong et al. ( 2020 ) Study showed that forbidding grazing reduced soil acidity, especially in topsoil. Different grazing intensities altered the soil carbon‐nitrogen ratio, with light grazing increased the ratio, and heavy grazing decreased it (He et al. 2019 ). In fact, the impact of grazing on ecosystems was very complex, with different climatic conditions and grazing times both influencing the ultimate outcome (Wang et al. 2018 ; Eldridge et al. 2016 ). So far, numerous scholars have conducted extensive research on the impact of grazing on soil physical and chemical properties (Török et al. 2018 ), but relatively few studies have investigated the effects of grazing on soil microbial community. Evaluating the optimal grazing intensity that promotes soil ecosystem health and sustainable development is of critical importance for maintaining ecological stability and ensuring sustainable development in the Qinghai‐Tibet Plateau region. Soil microorganisms are an indispensable component of the grassland ecosystem, playing an important regulatory role in energy flow, material input, and output processes within the system (Moll et al. 2017 ). As an important biological medium at the plant–soil interface, soil microorganisms influence the cycling process of soil nutrients through decomposition and directly interact with plant roots, thereby influencing plant growth and community succession (Li et al. 2016 ). Grazing livestock's hoof impacts compact the soil and alter conditions such as water potential and ventilation. Meanwhile, their feces and urine excretion enhances soil nutrient availability and increase easily decomposable carbon content, affecting the diversity and functionality of soil microbial communities (Eldridge and Delgado‐Baquerizo 2017 ). Moreover, the foraging activities of grazing livestock may cause plant biomass to be allocated towards belowground and increase root exudation, impacting soil microbial community structure (Mueller et al. 2017 ). As grazing intensity increased, grassland degradation intensified, soil quality declined, and the living environment of soil microorganisms was threatened, often leading to a decrease in their numbers (Kooch et al. 2020 ). Furthermore, some microorganisms were unable to adapt to the new living environment and their numbers decreased or even disappeared completely, while some new microorganisms were introduced into the soil through livestock excretion, enhancing soil microorganism diversity (Qu et al. 2016 ). Studies showed that moderate grazing enhances soil bacterial community diversity, while light grazing promotes an increase in the diversity of soil arbuscular mycorrhizal fungi community (Ba et al. 2012 ). A global‐scale study on grazing and soil microbial communities found that light grazing promoted soil microbial richness, with a more pronounced increase in fungal community richness that in bacterial communities, while heavy grazing significantly reduced microbial richness (Wang and Tang 2019 ). Yang's study indicated that heavy grazing favored fast‐growing plant species, which promoted bacterial community growth, while light grazing favored slow‐growing plant species, which were beneficial to fungal community growth (F. Yang et al. 2019 ). Grazing altered the soil microbial composition, shifting from a fungal‐dominated to a bacterial‐dominated community and from slow‐growing to fast‐growing microorganisms, leading to a transition from a fungal‐dominated food web, based on recalcitrant organic carbon, to a bacterial‐dominated food web, reliant on unstable organic carbon (Xun et al. 2018 ). In fact, soil microorganisms are highly diverse; however, only a small fraction of them had been cultivated. Although the quantity and types of soil microorganisms in alpine grassland ecosystems provide some indication of the degree and type of grazing disturbance, their high diversity and largely unknown nature add complexity to related research. Furthermore, human interference has led to ongoing debate regarding the responses of soil microorganisms to grazing (Bardgett et al. 1997 ; Moussa et al. 2007 ). To enhance understanding of soil microorganisms response to grazing intensity, this study conducted field experiments to explore: (1) whether different grazing intensities have distinct impacts on soil microbial communities; (2) which biological or non‐biological factors are influenced by varying grazing intensities, subsequently affecting soil bacterial and fungal communities; and (3) whether soil nitrogen metabolic processes change with different grazing intensities, given that the primary components of grazing livestock's feces and urine are water and nitrogen. This research provides a new perspective for the scientific utilization and management of alpine grasslands.",
"discussion": "4 Discussion 4.1 Effects of Different Grazing Intensities on Soil Microbial Communities Soil microorganisms play a central role in processes such as soil nutrient cycling and organic matter decomposition and are essential for maintaining the health of grassland ecosystems (Wei et al. 2017 ). This study, utilizing metagenomic sequencing technology, identified differences in the composition of soil bacterial and fungal communities under different grazing intensities (Figures 1 and 2 ). The dominant phyla of soil bacteria were mainly composed of Actinobacteria and Proteobacteria, which were consistent with the results of previous studies (Gong et al. 2019 ; Qin et al. 2021 ). Actinobacteria was a gram‐positive bacterium, and most of their taxa were symbiotic. They usually dominated in nutrient‐rich environments, promoting the decomposition of organic matter and being closely related to the nitrogen cycle in nature (Górska et al. 2022 ). Proteobacteria was a major group of soil bacteria, including various nitrogen‐fixing bacteria, and were the dominant group of microorganisms in the plant rhizosphere. Both were beneficial bacterial phyla in the soil, facilitating plant nitrogen absorption and utilization (Jiang et al. 2016 ). The relative abundances of these dominant bacterial phyla varied somewhat but not significantly among different grazing intensities. A possible explanation is that litter and plant root exudates varied under different grazing treatments, with different plants producing distinct litter and root exudates, which subsequently enhanced the competitive ability among bacterial communities and resulted in differences in the relative abundances of soil bacteria. Soil fungi at the phylum level mainly included Ascomycota, Mucoromycota, and Basidiomycota, which were important groups of soil fungi and participated in the soil carbon cycle by degrading organic matter. Additionally, some groups within the Basidiomycota could form symbiotic relationships with algae and mosses and were also important decomposers of plant lignin (Kabuyah et al. 2012 ). The results of this study indicated that grazing had no significant effect on the relative abundances of dominant fungal groups within the Ascomycota, Mucoromycota, and Basidiomycota. This also suggested that despite different grazing intensities, the dominant groups of soil fungi exhibited a certain level of similarity. This, in turn, reflected that the dominant groups of soil fungi had good adaptability to different habitats and possessed strong anti‐interference ability against grazing. Grazing was a potential factor affecting soil microbial diversity in grassland ecosystems. In this study, the Simpson index and Shannon‐Wiener index were used to reflect the overall diversity of soil microbial communities. It was found that the diversity indices of soil bacteria were not significantly different among different grazing intensities, while the MG treatment decreased the Simpson index and Shannon‐Wiener index of soil fungi. The possible reasons were speculated as follows: Firstly, compared to bacteria, fungi had larger individuals and were more susceptible to grazing disturbances. Secondly, according to disturbance theory, there was a nonlinear relationship between disturbance and resource and environmental heterogeneity. Moderate disturbance could increase species diversity, whereas excessive disturbance would decrease it. Under moderate grazing intensity, during the vigorous growth period of plants, a greater proportion of soil nutrients was allocated to plant growth and reproduction, resulting in fewer nutrients available for soil fungi. Under the HG treatment, both aboveground and belowground plant biomass were significantly reduced (Table 2 ), leading to a decrease in nutrients required for plant growth. However, the excretion process of livestock manure returned nutrients to the ecosystem, thereby providing more survival resources for soil fungi. The impact of grazing on soil microbial community dynamics was a highly complex process that was closely related to plant types and also influenced by soil physicochemical properties and environmental factors (Wu et al. 2022 ). Variations in grassland types and soil texture contributed to the complexity of the relationship between soil microbial communities and soil biotic and abiotic characteristics under grazing. The results of this study indicated that soil pH and moisture were key factors affecting soil bacterial communities, while soil pH was a critical factor influencing soil fungal communities (Figure 5 and Table 3 ). This suggested that different factors mediated the effects of grazing on soil bacteria and fungi, which were consistent with some research findings under different grazing intensities (Cao et al. 2011 ; Wang et al. 2021 ). The possible reason was speculated to be that soil microorganisms were sensitive to changes in soil pH and water availability. Specifically, the alteration of soil acidity and alkalinity caused by the excretion of livestock manure and urine due to grazing led to changes in soil microbial community composition, potentially shifting from one dominant species to another (Levy‐Booth et al. 2016 ). 4.2 Impact of Different Grazing Intensities on Soil Microbial Community Functions Microorganisms were not only an essential part of soil ecosystems but also possessed various functions, such as metabolism and nutrient cycling (Gao et al. 2021 ). In this study, a differential analysis of soil microbial functions was conducted. A total of 368 pathways were annotated using the KEGG database. A differential analysis of the top 10 metabolic pathways with the highest relative abundance revealed no significant differences among different grazing intensities. Jiao et al. ( 2022 ) conducted research on restored cropland ecosystems using metagenomic sequencing technology and found that core microbial communities played a crucial role in maintaining the functional stability of afforested ecosystems, with the “core functional characteristics” of microorganisms being relatively conserved. In this study, although soil microbial communities changed with different grazing intensities, their metabolic functions remained relatively conserved. This indicated that the soil microorganisms under different grazing intensities in this study could still maintain stable ecological functions when subjected to external environmental disturbances (Gao et al. 2021 ). Pang et al. ( 2021 ) demonstrated that the metabolic pathways of soil microorganisms were similar under different treatments, but there were slight differences in gene sequences within specific metabolic pathways. In this study, by comparing the relative abundances of different ko numbers in the nitrogen metabolism pathway, it was found that the relative abundances of some functional genes related to nitrogen metabolism (ko00910) were downregulated in the LG treatment, indicating that the LG treatment reduced soil nitrogen metabolism processes. This further reflected the research findings that the ecological functions of soil microbial communities were primarily metabolic functions (Hao et al. 2019 ). Specific functional genes of soil microorganisms regulate nitrogen‐cycling processes, and this characteristic also indirectly determines that microbial functional genes act as drivers for the conversion of various forms of nitrogen (Y. F. Wang et al. 2022 ). In this study, metagenomic‐sequencing technology was used to investigate the response of soil nitrogen metabolism functional genes to different grazing intensities. In this study, the soil nitrogen metabolism involved four pathways: dissimilatory nitrate reduction, assimilatory nitrate reduction, denitrification and nitrification. Among them, the denitrification pathway was detected with a higher frequency in the regional microorganisms, while nitrogen fixation ( nifH ) and anammox ( hzsA ) were not detected. This research outcome was consistent with the findings of Nelson et al. ( 2016 ), indicating similar characteristics of soil nitrogen metabolism pathways in the study area. The anammox process refers to the process in which related microorganisms use NO 2− as an electron acceptor to convert NH 4+ into N 2 through oxidation in an anaerobic environment, this process is considered an important pathway for nitrogen removal in natural environments and contributes to accelerating the global nitrogen‐cycling rate (Mulder et al. 1995 ). The anammox process mainly occurs in anaerobic environments, but the soil in the study area was unsuitable for the survival and development of such microorganisms (Humbert et al. 2010 ). Additionally, the nitrogen metabolism functional gene narG was detected under various grazing intensities. Nitrate reductase has many origins, such as periplasmic nitrate reductase and nitrite oxidoreductase. Among them, narG is often used as a marker gene for denitrifying microorganisms in the detected environment (John and Partha 2002 ). As an important structural gene of nitrogen‐fixing microorganisms, nifH was often used as a key indicator for detecting nitrogen‐fixing bacteria (Xu et al. 2014 ). This study did not detect biological nitrogen fixation pathways, possibly because the soil in the study area was inhospitable to the survival and reproduction of nitrogen‐fixing microorganisms. For instance, the total nitrogen content in the soil was relatively high, falling into the high‐nitrogen level. Excessive nitrogen content inhibits nitrogenase activity in nitrogen‐fixing microorganisms, as the nitrogen fixation process requires substantial energy, which in negatively impacts the functional genes associated with biological nitrogen fixation pathways (Zhang et al. 2013 )."
} | 4,618 |
31769782 | null | s2 | 5,840 | {
"abstract": "Bacteria frequently cooperate by sharing secreted metabolites such as enzymes and siderophores. The expression of such 'public good' traits can be interdependent, and studies on laboratory systems have shown that trait linkage affects eco-evolutionary dynamics within bacterial communities. Here, we examine whether linkage among social traits occurs in natural habitats by examining investment levels and correlations between five public goods (biosurfactants, biofilm components, proteases, pyoverdines and toxic compounds) in 315 Pseudomonas isolates from soil and freshwater communities. Our phenotypic assays revealed that (i) social trait expression profiles varied dramatically; (ii) correlations between traits were frequent, exclusively positive and sometimes habitat-specific; and (iii) heterogeneous (specialised) trait repertoires were rarer than homogeneous (unspecialised) repertoires. Our results show that most isolates lie on a continuum between a 'social' type producing multiple public goods, and an 'asocial' type showing low investment into social traits. This segregation could reflect local adaptation to different microhabitats, or emerge from interactions between different social strategies. In the latter case, our findings suggest that the scope for competition among unspecialised isolates exceeds the scope for mutualistic exchange of different public goods between specialised isolates. Overall, our results indicate that complex interdependencies among social traits shape microbial lifestyles in nature."
} | 384 |
29255672 | PMC5725212 | pmc | 5,841 | {
"abstract": "Production of propionic acid by fermentation of propionibacteria has gained increasing attention in the past few years. However, biomanufacturing of propionic acid cannot compete with the current oxo-petrochemical synthesis process due to its well-established infrastructure, low oil prices and the high downstream purification costs of microbial production. Strain improvement to increase propionic acid yield is the best alternative to reduce downstream purification costs. The recent generation of genome-scale models for a number of Propionibacterium species facilitates the rational design of metabolic engineering strategies and provides a new opportunity to explore the metabolic potential of the Wood-Werkman cycle. Previous strategies for strain improvement have individually targeted acid tolerance, rate of propionate production or minimisation of by-products. Here we used the P. freudenreichii subsp . shermanii and the pan- Propionibacterium genome-scale metabolic models (GEMs) to simultaneously target these combined issues. This was achieved by focussing on strategies which yield higher energies and directly suppress acetate formation. Using P. freudenreichii subsp . shermanii , two strategies were assessed. The first tested the ability to manipulate the redox balance to favour propionate production by over-expressing the first two enzymes of the pentose-phosphate pathway (PPP), Zwf (glucose-6-phosphate 1-dehydrogenase) and Pgl (6-phosphogluconolactonase). Results showed a 4-fold increase in propionate to acetate ratio during the exponential growth phase. Secondly, the ability to enhance the energy yield from propionate production by over-expressing an ATP-dependent phosphoenolpyruvate carboxykinase (PEPCK) and sodium-pumping methylmalonyl-CoA decarboxylase (MMD) was tested, which extended the exponential growth phase. Together, these strategies demonstrate that in silico design strategies are predictive and can be used to reduce by-product formation in Propionibacterium . We also describe the benefit of carbon dioxide to propionibacteria growth, substrate conversion and propionate yield.",
"conclusion": "5 Conclusion In this work, we proposed several genetic engineering strategies for the improvement of propionic acid production in propionibacteria using genome-scale modelling. Two fundamental strategies were identified by exploring pathways encoded within the pan-genome of propionibacteria that lead to the generation of reduced cofactors or enhanced energy production. The first two enzymes of the oxidative branch of the PPP were over-expressed, resulting in a 4-fold increase in the propionate to acetate ratio during the exponential growth phase and demonstrating the ability for NADPH to couple to the Wood-Werkman cycle. Furthermore, our results from the over-expression of the ATP-dependent PEPCK and the sodium-pumping MMD suggest that MMD activity may be limiting and that the ATP-PEPCK operates in the direction of oxaloacetate and ATP synthesis in Propionibacterium. The growth rate of this mutant was higher than that of the empty plasmid strain as predicted by the model. Sparging CO 2 into the fermenter was surprisingly beneficial for cell growth and propionate production. These findings suggest a CO 2 sparge may be a cheap, novel method to promote the production of propionate, offering new opportunities to influence propionate production in efficient native producers.",
"introduction": "1 Introduction Propionic acid is a three-carbon compound with a broad range of applications in the food, pharmaceutical and chemical industries. In addition to its use as an intermediate for the synthesis of cellulose fibres, herbicides, perfumes and pharmaceuticals, propionic acid is a strong antimicrobial agent used in animal feed and as a food preservative ( Guan et al., 2015b ). Propionic acid is mainly produced by petrochemical processes. However, the demand for green chemicals has renewed attention to fermenting propionibacteria for propionic acid production. Propionibacteria are gram-positive, rod-shaped, facultative anaerobes. Dairy propionibacteria are important industrial GRAS (generally recognized as safe) strains used in the food and cheese industry ( Meile et al., 2008 ). Propionibacteria have long been employed as starter and ripening cultures, responsible for flavour development and participating in the formation of cheese eyes ( Britz and Riedel, 1994 , Thierry and Maillard, 2002 ). Dairy strains are also used industrially to produce trehalose and vitamin B12 ( Kośmider et al., 2012 , Ruhal and Choudhury, 2012 ). The bioprocess for propionate production still suffers from low productivity, low yield, and expensive downstream purification costs; acetic and succinic acids are produced alongside propionic acid making the downstream separation costly ( Liu et al., 2012 , Rodriguez et al., 2014 ). Metabolic engineering strategies aiming to reduce by-products are thus needed. Improved yields can theoretically be achieved by deleting by-product genes, but metabolic manipulation of propionibacteria has developed slowly, and only a few examples exist in the literature engineering the best propionic acid producer, P. acidipropionici. Metabolic engineering strategies have focussed on enhancing flux through the Wood-Werkman cycle, either by directing carbon into the cycle through over-expression of pyruvate and phosphoenolpyruvate carboxylases or the pyruvate:methylmalonyl-CoA carboxyltransferase ( Ammar et al., 2014 , Liu et al., 2016 , Wang et al., 2015b ). Other strategies that have been tested include the over-expression of components of the Wood-Werkman cycle itself ( Liu et al., 2015 , Wang et al., 2015a ) or the targeting of acid tolerance associated processes such as trehalose synthesis ( Jiang et al., 2015 ) and the glutamate decarboxylase or arginine deiminase pathways ( Guan et al., 2015a ). While P. acidipropionici remains resistant to engineering, unique metabolic features of this species were over-expressed in lower producers ( Wang et al., 2015b ). Few studies have targeted by-product production, and these have ultimately had limited success ( Liu et al., 2016 , Suwannakham et al., 2006 ). Ultimately, the limited phenotypic improvements in current studies and limitations in genetic engineering tools necessitate a genome-scale model to further aid design ( Wang et al., 2015b ). Here, we utilise a recently developed genome-scale model for P. freudenreichii subsp . shermanii and the pan- Propionibacterium GEM to design genetic modification strategies. Because the production of propionic acid is already growth-coupled, classic growth-coupling strain design algorithms such as OptKnock ( Burgard et al., 2003 ) favour sub-optimal strategies that limit energy generation, leaving cells more susceptible to acid stress. Given the depletion of the pH gradient is regarded as a major mechanism by which weak acids exert their toxicity on cells, we explored strategies favouring propionate production while maintaining high levels of energy generation or even exceeding those of wild-type strains. Strategies probed pathways consistent with the maximum production of energy and propionate in the native metabolism. By simultaneously trying to improve both energy output and propionate production, strains are expected to be readily evolvable to a higher producing phenotype while also improving acid tolerance. Such approaches facilitate the production of high producing mutants with minimum genetic perturbations; a requirement enforced by the few genetic modification tools available. Given the limited success of reducing acetate production in previous studies ( Liu et al., 2016 , Suwannakham et al., 2006 ), we over-expressed the PPP to test in silico simulation predictions and show that such a strategy results in lower acetate production without sacrificing growth. We then tested an alternative strategy consisting of over-expressing an ATP-PEPCK and MMD to enhance energy available to the cell and analyse how this affects the phenotype. We observed an extended exponential growth phase in the mutant strain. Because this strategy was reliant on the addition of CO 2 , the influence of CO 2 on the Propionibacterium fermentation was also assessed, revealing undescribed biomass and propionate stimulating effects in propionibacteria.",
"discussion": "4 Discussion 4.1 Feasibility of the proposed genetic engineering strategies The biological production of propionic acid would be economically feasible if propionibacteria could be engineered to reach near maximum theoretical yields while maintaining high titre and volumetric productivity ( Rodriguez et al., 2014 ). Current strategies in the literature have focussed on directly upregulating enzymatic components of the propionate production pathway ( Ammar et al., 2014 , Liu et al., 2016 , Wang et al., 2015b ) or by improving acid tolerance by upregulating the arginine deiminase, glutamate decarboxylase amino acid systems, or trehalose synthesis ( Guan et al., 2015a , Jiang et al., 2015 ). These strategies compete with, rather than suppress, acetate production and have relatively minor improvements in phenotype, or represent a carbon drain and reduce the yield. Metabolic modelling was therefore used to identify more effective strategies. The classic strain design algorithm, OptKnock ( Burgard et al., 2003 ), was originally used which works on a principle of growth coupling the production of a target product by eliminating alternative, energetically superior pathways. Because propionate is already growth coupled, these strategies inherently either targeted the elimination of other by-products (such as an acetate kinase knockout) or energy generating steps (such as pgk knockout to prevent a net energy yield through glycolysis). Because the fermentation titre is ultimately limited by product inhibition ( Blanc and Goma, 1987 ), primarily thought to be caused by the futile cycling of protons across the membrane, the availability of energy for proton extrusion should dictate acid stress tolerance. Therefore, reducing the energy yield risks trading titre off against fermentation yield and the most successful strategies will enhance propionate production while maintaining, or even improving, the high energy yield of Propionibacterium . Here, we present a number of engineering strategies, which in silico, achieve higher yield ( Table 2 ). The strategies vary in complexity and the limited capacity to implement these strategies with molecular biology means the testing of any strategy is inherently difficult. A strong focus was therefore placed on ensuring these strategies are robust in their implementation. Each strategy proposed is analysed in detail in the supplementary material Table S2 . While P. freudenreichii subsp. shermanii serves as a good propionibacteria model, the ultimate goal is to translate these strategies into strains most suited to industrial production, such as P. acidipropionici , where genetic engineering remains elusive. It is likely that future work needs to address the large variety of restriction modification (RM) systems which have been identified in Propionibacterium , compiled in the REBASE database ( Roberts et al., 2015 ), before this is possible. In other non-model organisms, RM systems have been by-passed by the expression of native methylation components in the cloning host ( O'Connell Motherway et al., 2009 , Zhang et al., 2012 ). In the current study, we explored two (PPP and linear PA) of the strategies identified ( Table 2 ) in the more readily engineered P. freudenreichii subsp. shermanii . The PGK strategy was excluded because it cripples energy production, while the ACK knockout strategy was excluded because P. freudenreichii subsp. shermanii utilises different pathways for acetate production to P. acidipropionici (manuscript in preparation). The TCA and citramalate pathways are energetically advantageous over the PPP but rely on the promiscuity of enzymes associated with acetate metabolism; thus, both pathways would require extensive protein engineering to change enzyme selectivity. Finally, the glycine cleavage pathway is particularly promising on its own or coupled to the linear pathway due to its ability to couple the regeneration of ATP to the production of reduced cofactors. However, it involves at least nine enzymes. 4.2 Influence of PPP over-expression on the metabolism of P. freudenreichii subsp. shermanii A strategy to manipulate the redox metabolism of Propionibacterium using the PPP is presented. Over-expression of the PPP by upregulating the oxidative branch is a proven strategy ( Ahmad et al., 2012 ) and was performed here by upregulating the first two steps of the oxidative branch of the PPP. The PPP is a favourable engineering target given it is compatible with other engineering strategies already implemented such as upregulation of the Wood-Werkman cycle ( Liu et al., 2015 , Wang et al., 2015a ) and is tractable within the limits of current engineering tools. The PPP has long been thought to be the primary pathway responsible for the increased ratio of propionate to acetate in higher producing propionibacteria ( Papoutsakis and Meyer, 1985 ) but this has never been tested. Consistent with the modelling, the strategy resulted in a 4-fold improvement in the propionate to acetate ratio and a 13% improvement in the rate of propionate production without increased production of succinate. Our findings demonstrate that over-expression of the PPP is a valid alternative to supplementing the media with glycerol without inhibiting growth, an initial concern of Wang and colleagues ( Wang et al., 2015b ). This is despite a reduced energy yield from the pathway; suggesting this is partially compensated for by either the increase in NADPH or activity of the PPP which has been shown to combat oxidative stress ( Krüger et al., 2011 ). Importantly, these results show that NADPH can be coupled to the Wood-Werkman cycle, although the exact mechanism remains unclear, and supports the hypothesis that a higher flux through the PPP may be responsible for the divergence in the propionate to acetate ratio from 2 ( Papoutsakis and Meyer, 1985 ). In Pf-EMP, a molar ratio of propionate to acetate of ~2:1 was observed during exponential growth. This ratio is consistent with the simulated l 1 -norm flux distribution for the maximisation of energy. At such a ratio, it is expected that flux through the PPP is minimal. In comparison, Pf-ZOP achieved a molar propionate to acetate ratio of ~8. If all additional reduced cofactors required to achieve such a ratio are assumed to come from the PPP, the model predicts that the split ratio of flux between glycolysis and the PPP is approximately 2:3, implying 60% of the glucose uptake was directed towards the PPP. This compares to the 86% required in order to achieve the maximum theoretical yield. Pf-ZOP appears to experience a partial release of the over-expression phenotype at the onset of the stationary phase. In particular, the propionate production rate reduces with respect to Pf-EMP ( Table 3 ), which ultimately reduces the propionate to acetate ratio from 9.2 (g/g) at the end of the exponential growth phase to 6 (g/g) at the end of the fermentation; although this is still a 55% improvement over Pf-EMP. Conversely, Pf-EMP showed an improvement in the ratio of propionate to acetate at the onset of stationary phase. We postulate this different behaviour may be due to the activity of the Bifidobacterium shunt recently identified in P. freudenreichii subsp. shermanii (manuscript in preparation). The Bifidobacterium shunt catabolises sugars via the PPP and a phosphoketolase enzyme to produce primarily acetate and lactate in Bifidobacterium , where it is over-expressed in response to stress ( Sánchez et al., 2005 ). Expression in response to cold stress ( Dalmasso et al., 2012 ) and apparent activation at the onset of stationary phase suggests it may play a similar role in P. freudenreichii subsp. shermanii . 4.3 Influence of carbon dioxide sparging on the metabolism of P. freudenreichii subsp. shermanii Fermenters were sparged with CO 2 to test the linear propionate pathway strategy. Sparging improved biomass production, which can be observed by comparisons of the strains harbouring the empty plasmid ( Fig. 6 ). While biomass was similar at the end of the exponential growth phase, propionibacteria supplied with CO 2 grew for longer, resulting in 25% higher biomass. The rate of acetate production across the exponential growth phase is reduced by one-third and halved in the stationary phase ( Table 3 , Table 4 ). This suggests CO 2 addition somehow alters metabolism to produce more reduced cofactors and suppresses acetate production. Fig. 6 Comparison of fermentations of Pf-EMP with a CO 2 sparge (solid line, squares) and without (dotted line, triangles). Sparging with CO 2 increased biomass production and allowed increased glucose catabolism. Fig. 6 Our results indicate that in the presence of CO 2 , glucose is catabolised at a slower rate to all three major fermentation products and propionate is favoured over acetate. Such a phenotype is identical to that observed when acid stress is applied by reduction in the external pH ( Feng et al., 2010 ). Likewise, CO 2 can also trigger acid stress ( Baez et al., 2009 , Sun et al., 2005 ). Despite the fact that near equal amounts of dissolved CO 2 and carbonate will be present in the fermenter at pH 6.4, a lack of homologues for bacterial carbonate transporters ( Price, 2011 ) and a higher permeability of dissolved CO 2 ( Jones and Greenfield, 1982 ) leads to cytoplasmic acidification. Metabolic modelling suggests CO 2 fixation does not directly promote propionate production and does not enhance energy. Fixation occurred only when flux was constrained through the native PPi-PEPCK ( Siu et al., 1961 ), resulting in an energetic penalty and the equimolar production of succinate, consistent with experimental results ( Wood and Werkman, 1938 , Wood and Werkman, 1940 ). Further conversion to propionate was only viable in the presence of MMD. The above results lead us to propose a novel mechanism by which CO 2 can modify Propionibacterium metabolism by overcoming thermodynamic limitations. Assuming a mechanism of transport by diffusion, intracellular propionate is expected to increase across the course of the fermentation as extracellular propionate accumulates and the pH gradient diminishes. In addition to decoupling the proton gradient, accumulating propionate is expected to inhibit metabolism kinetically and may eventually form a thermodynamic barrier. This is particularly possible given the Wood-Werkman cycle is entirely reversible ( Rosner and Schink, 1990 ). The fixation of CO 2 through the PPi-dependent PEPCK would help build the concentrations of key metabolic intermediates in the Wood-Werkman cycle allowing further flux through the pathway and resulting in the increased consumption of substrate observed. 4.4 Influence of the linear pathway for propionate production on the metabolism of P. freudenreichii subsp. shermanii The linear pathway for propionate production is a promising, energetically beneficial route to propionate identified by studying the pan Propionibacterium genome (manuscript in preparation). The strategy consists of over-expression of the ATP-dependent PEPCK, which improves the growth rate and succinate yield in E. coli strains lacking the PEP carboxylase ( Kim et al., 2004 ) along with the sodium-pumping MMD, the key energy-yielding step in the fermentation of succinate to propionate for organisms such as Propionigenium modestum ( Schink and Pfennig, 1982 ). Expression of the linear propionate pathway resulted in a continuation of the exponential growth phase at a reduced growth rate. Both glucose catabolism and the production of all fermentation products are stimulated compared to CO 2 -sparged Pf-EMP. This appears to come at a trade-off for propionate specificity; compared to CO 2 -sparged Pf-EMP the propionate to acetate yield is lower in the transition and stationary phases. Increased succinate production suggests that the ATP-PEPCK operates in the direction of oxaloacetate synthesis, but suggests MMD activity may be limiting. While an increased growth rate was expected, it is possible that additional energy was captured as polyphosphate and utilised to fuel further growth at the end of the exponential growth phase instead ( Clark et al., 1986 ). A more pronounced phenotype is expected if more PEP is available, such as during the catabolism of disaccharides."
} | 5,185 |
35193962 | PMC8872734 | pmc | 5,844 | {
"abstract": "Significance The enormous complexity of metabolic pathways, in both their regulation and propensity for metabolite cross-talk, represents a major obstacle for metabolic engineering. Self-assembling, catalytically programmable and genetically transferable bacterial microcompartments (BMCs) offer solutions to decrease this complexity through compartmentalization of enzymes within a selectively permeable protein shell. Synthetic BMCs can operate as autonomous metabolic modules decoupled from the cell’s regulatory network, only interfacing with the cell’s metabolism via the highly engineerable proteinaceous shell. Here, we build a synthetic, modular, multienzyme BMC. It functions not only as a proof-of-concept for next-generation metabolic engineering, but also provides the foundation for subsequent tuning, with the goal to create a microanaerobic environment protecting an oxygen-sensitive reaction in aerobic growth conditions that could be deployed.",
"discussion": "Discussion The overarching goal of the sFUT BMC design is to create a platform metabolic module that can integrate oxygen-sensitive metabolism into an aerobic host to produce, from a cheap feedstock, a central biosynthetic intermediate for the production of high-value compounds. We produced a prototype synthetic microcompartment core for the oxygen-sensitive enzyme PFL based on the HO-shell because of the availability of molecular tools to load the synthetic HO-shell with cargo and its potential to form wiffleball architectures. We modified this architecture with a purification tag that facilitates rapid isolation and the testing of enzymatic core designs. This decouples the design-test-refine cycle for the catalytic core, from optimization of shell permeability, by allowing unrestricted substrates and product exchange for activity measurements. However, for future designs, the BMC-P protein can be added to the shell system to form complete HO-shells, and therefore complete sFUT BMCs. This concept can be applied broadly, streamlining other engineering efforts aiming to design and optimize synthetic BMCs. Core design and assembly constitute the first phase toward constructing a completely functional synthetic “organelle.” This required expanding the technique to specifically encapsulate cargo by covalent linkage developed by Hagen et al. ( 25 ) to include the addition of a new adaptor system (SnoopTag/SnoopCatcher). Heterologous enzymes have been targeted to the lumen of the BMC before ( 38 , 39 ) by using encapsulation peptides ( 40 , 41 ) on the cargo; however, this approach suffers from low efficiency, is hampered by aggregation, and it is unknown how the encapsulation peptides associate with the shell, making this method nonquantitative ( 27 , 42 – 45 ). Our strategy enabled us to reliably, specifically, and independently target two different cargo proteins into the lumen of a BMC. Additionally, the expansion of the adaptor system effectively doubled the number proteins that can be encapsulated from 60 to 120 per BMC, compared to the previously described method ( 25 ). This can increase the overall efficiency of the synthetic BMC by increasing the metabolite flux and permitting faster substrate channeling due to a greater enzyme density. This first-generation sFUT prototypes aimed to encapsulate the minimal number of proteins needed (PFL and EutD) to cycle CoA ( Fig. 1 C ), which is a natural property of BMCs ( 46 ). Tagging the PFL with SpyCatcher for encapsulation was challenging because of low expression level, and if expressed it produced mostly inactive ( Fig. 3 ) and insoluble protein ( SI Appendix , Fig. S4 A ). However, we succeeded in creating an active version of the PFL with a C-terminal encapsulation adapter (PFL-Cspyc). EPR revealed the presence of a glycyl radical signal consistent with activated PFL, suggesting that PFL-AE was able to bind PFL-Cspyc and successfully activate the enzyme. Furthermore, our E. coli growth-based activity assays show that the PFL-Cspyc retains activity comparable to the unmodified PFL in both forward and reverse direction, making it suitable for constructing the sFUT wiffleball. It should be noted that because of the versatility of the SpyCatcher/SpyTag system, this C-terminal–tagged version of the PFL can potentially be used in other scaffolding or engineering efforts. Considering that both PFL and EutD form homodimers ( 35 , 47 – 49 ), and the relatively large size of the PFL-Cspyc (calculated molecular mass of 95 kDa), we anticipated that sFUT wiffleballs cannot fully assemble when all 120 contact points are used for conjugating cargo, because of localized overcrowding of the sFUT wiffleball lumen. The consequent fragility may explain why we couldn’t isolate the minimal sFUT wiffleball architecture. To prevent steric hindrance while BMC self-assembly takes place, we diluted out the presence of T 1-spyt-snpt-6xHis in the wiffleballs by coexpressing T 2 and T 3 , which occupy the same geometric positions in the shell as T 1-spyt-snpt-6xHis but don’t recruit cargo into the lumen. Using this approach, we were able to isolate completely assembled sFUT wiffleballs loaded with PFL and EutD. Our activity measurements of the forward reaction show that the enzymes are active and can undergo multiple turnovers cycling CoA between the enzymes. Although CoA is known to associate with cargo proteins before encapsulation in BMCs ( 45 ), we note that the sFUT wiffleball activity can be improved by adding CoA or acetyl-CoA to the reaction. Considering that isolation of sFUT wiffleballs is a lengthy process, including washes of the streptavidin column, we hypothesize that some CoA was released from the shell in this process; thus, its addition was able to slightly increase sFUT wiffleball activity. We have focused on the forward reaction for the activity assay because this is readily measurable, only requiring the addition of pyruvate to the sample, and we could follow the formation of formate. It should be noted that our in vivo experiments show that the modified enzymes have activity in both forward and reverse directions. However, we noted that the conversion of the provided pyruvate to formate does not reach the published equilibrium of the enzymatic reaction ( 50 ), suggesting that over time the enzymes get inactivated and there is not sufficient PFL-AE or Ado-Met available to reactivate the PFL. A k cat for the encapsulated PFL can be calculated roughly from the activity assay; using the estimated 40 PFL enzymes per sFUT wiffleball, it yields about 1.2 s −1 . However, given that we could not time-resolve the enzymatic reaction because it presumably ends within seconds, and we could only measure on a minute timescale with our experimental setup, the k cat of the encapsulated PFL could be much closer to the published PFL k cat of 105–770 s −1 ( 51 , 52 ). In summary, we have built a synthetic BMC, directly targeting two enzymes to be encapsulated, one of which is extremely oxygen sensitive, and expressing three auxiliary enzymes (PFL-AE, METK, ACK), to enable its function in the cell. In contrast to the synthetic BMCs first pioneered ( 38 , 39 , 53 ), which were used for ethanol production, polyphosphate storage and hydrogen production, the sFUT prototype can be used as a platform in ambitious engineering projects to compartmentalize entire metabolic pathway for the production of a biomolecule of interest starting at the easily accessible feedstocks acetate and formate. Moreover, our prototype metabolic module is poised for shell permeability engineering to address the grand challenge of constructing devices for to compartmentalize oxygen-sensitive reactions for use in aerobic growth conditions."
} | 1,934 |
29468120 | PMC5779721 | pmc | 5,845 | {
"abstract": "The growth characteristics and underlying metabolism of microbial production hosts are critical to the productivity of metabolically engineered pathways. Production in parallel with growth often leads to biomass/bio-product competition for carbon. The growth arrest phenotype associated with the Saccharomyces cerevisiae pheromone-response is potentially an attractive production phase because it offers the possibility of decoupling production from population growth. However, little is known about the metabolic phenotype associated with the pheromone-response, which has not been tested for suitability as a production phase. Analysis of extracellular metabolite fluxes, available transcriptomic data, and heterologous compound production (para-hydroxybenzoic acid) demonstrate that a highly active and distinct metabolism underlies the pheromone-response. These results indicate that the pheromone-response is a suitable production phase, and that it may be useful for informing synthetic biology design principles for engineering productive stationary phase phenotypes.",
"conclusion": "4 Conclusions The S. cerevisiae pheromone-response leads to a distinct and active metabolic phenotype that is suitable for the production of PHBA, and for bioprocesses in general. The key metabolic differences which result from pheromone treatment were identified as a glucose uptake rate that is comparable to exponentially growing populations, increased by-product formation, up-regulation of storage carbohydrate synthesis genes associated with osmolarity and oxidative stress responses, and increased respiratory activity. It is conceivable that the pheromone-response could be used with particularly good effect to increase flux towards metabolites of interest that are associated with the metabolic changes that underlie the phenotype. Although the pheromone-response shows promise as a production phase, it is associated with a complex mating phenotype. The concept of using a growth-arrest as a production phase would be far more effective if it could be reverse engineered and streamlined so that it is decoupled from the unnecessary aspects of the mating phenotype. The ability to engineer a switch from rapid growth, to growth-arrested production in the presence of abundant carbon and nitrogen sources would become an essential design feature of industrial microorganisms. Further investigation of the S. cerevisiae pheromone-response may provide a means for understanding how to coordinate and engineer such a powerful synthetic biology module.",
"introduction": "1 Introduction Microorganisms can be used to manufacture products ranging from therapeutic proteins and industrial enzymes through to metabolites for the replacement of existing petrochemicals and fossil fuels ( Woolston et al., 2013 ). A major challenge for all bioprocesses is balancing resources between biomass accumulation and product formation, as both outcomes require the same cellular resources such as carbon precursors, energy in the form of ATP, and reducing power in the form of NADH and NADPH. Biomass accumulation is essential to achieve the volumetric productivity required for commercial processes; however, excess biomass accumulation limits product yields. Moreover, the product or its intermediates may be toxic to the host organism, again limiting biomass production. Dynamic regulatory systems can be used to trigger the expression of a production pathway after the completion of a growth phase ( Venayak et al., 2015 ). However most non-growth associated phenotypes are poor production phases due to the depletion of available resources and the subsequent induction of stress response mechanisms ( Albers et al., 2007 , Chubukov and Sauer, 2014 ). The yeast Saccharomyces cerevisiae is a widely used industrial host microbe, and has growth characteristics that typify the limitations of normal growth based physiology in industrial microorganisms. S. cerevisiae populations undergo an exponential growth phase where carbon and nitrogen resources are rapidly consumed until they limit biomass production. During exponential growth, approximately 90% of cellular energy is directed towards ribosome biogenesis ( Warner et al., 2001 ). Carbon- or nitrogen-limited populations cease rapid growth and enter a ‘stationary phase’, which is characterised by the induction of stress survival mechanisms and a drastic reduction in the overall rate of protein synthesis relative to the exponential phase ( Werner-Washburne et al., 1993 ). In the case of carbon starvation, there is no substrate left for conversion into product; and under nitrogen starvation, stress signalling severely limits metabolic productivity even in the presence of excess carbon ( Albers et al., 2007 ). An ideal scenario for bio-production would involve a rapid growth phase where biomass (or ‘catalyst’) accumulates to a level that enables high volumetric productivity, before switching to a metabolically active stationary phase. This phase would then be maintained even in the presence of high concentrations of cellular resources such as carbon and nitrogen. With cells metabolically active but not growing and dividing, a much greater proportion of carbon could be directed towards target metabolites. Such a strategy would also open up the possibility of implementing growth limiting genetic modifications such as the silencing of essential genes using dynamic regulatory mechanisms ( Williams et al., 2015a , Williams et al., 2015b ). Stationary phase production is also very attractive because it enables the formation of products that are normally toxic to growth, and therefore limiting to production ( Holtz and Keasling, 2010 , Keasling, 2008 ). The cell-cycle arrest phenotype of the yeast mating system represents a unique phase in the life-cycle of S. cerevisiae , which could be useful as a production phase for metabolic engineering where metabolic productivity is decoupled from growth-based physiology. The mating system has evolved to facilitate the synchronisation of the cell cycle and the fusion of two cells of opposite mating type to form a diploid. Briefly, haploid yeast cells of each mating type (a or α) secrete specific peptide pheromones that they use to detect the proximity of a potential mating partner of the opposite mating type. Binding of pheromone to specific G-protein coupled membrane receptors triggers an intracellular mitogen activated protein kinase signalling event which results in the de-repression of the Ste12p transcription factor and the initiation of the pheromone-response ( Bardwell, 2005 ). Activation of the mating phenotype results in polarized growth, remodelling of cellular morphology and global transcription patterns, and arrest of growth in the G1 phase of the cell cycle ( Bardwell, 2005 ), similarly to entry into stationary phase during carbon or nitrogen starvation. This phenotype can also be triggered via the addition of purified mating peptide to laboratory yeast cultures. The S. cerevisiae mating system has become a cornerstone of eukaryotic synthetic biology ( Furukawa and Hohmann, 2013 ). The pheromone communication system has been utilised for synthetic quorum sensing ( Williams et al., 2015a , Williams et al., 2013 ), signal amplification ( Groß et al., 2011 ), intercellular and interspecies communication ( Hennig et al., 2015 , Jahn et al., 2013 ), and biological computation ( Regot et al., 2011 ). Furthermore, the depth of knowledge surrounding the mitogen activated protein kinase (MAPK) signal transduction machinery has enabled the construction and fine-tuning of a multitude of synthetic regulatory circuits ( Bashor et al., 2008 , Ingolia and Murray, 2007 , O’Shaughnessy et al., 2011 , Tanaka and Yi, 2009 ). In addition to relevance as a potential production phase, knowledge of the pheromone-response metabolism will be invaluable for future design of MAPK related synthetic regulatory systems. However, despite extensive utilisation of the mating system in synthetic biology, almost nothing is known about aspects of the phenotype that are not specifically related to mating. Activation of the pheromone-response could result in a number of different scenarios with respect to metabolic engineering outcomes for a specific product. These include: an unproductive phenotype similar to the G1 arrest of the carbon- or nitrogen-limited stationary phases; higher productivity due to the limitation of carbon flux towards biomass; or no overall effect on cellular productivity due to the diversion of cellular resources towards the mating phenotype. In addition to considerations of general metabolic productivity, it is also important to identify any fundamental differences in metabolism, as they can help to decide which heterologous products will be favoured by the natural fluxes in the network. For example, specific anabolic pathways could be up-regulated in response to mating pheromone, suggesting that industrial products which are derived from these pathways would have higher yields during the pheromone-response. The concept of limiting biomass formation to enhance cellular productivity has received some attention in the field of therapeutic protein production in mammalian cell cultures ( Kumar et al., 2007 ). In particular, the manipulation of the eukaryotic cell cycle to induce a growth arrest phenotype has been successfully used to improve heterologous protein production. For example, the over-expression of the cyclin dependent kinase inhibitor p21and its inducer C/EBPα in a Chinese Hamster Ovary cell line resulted in stable cell-cycle arrest in the G1 phase and a 10–15 fold higher protein productivity per cell ( Fussenegger et al., 1998 ). Similarly, the overexpression of the p21 cyclin inhibitor in an NS0 mouse myeloma cell line increased protein productivity ~4 fold ( Watanabe et al., 2002 ). The increased productivity due to p21 mediated cell-cycle arrest has been attributed to higher mitochondrial membrane potential providing more ATP for peptide bond formation, and increased ribosomal biogenesis ( Bi et al., 2004 , Khoo and Al-Rubeai, 2009 ). It is possible that the cell cycle arrest phenotype of the S. cerevisiae pheromone-response could result in similar productivity improvements. In this work, we have investigated the pheromone-response in S. cerevisiae as a growth arrest phase for metabolic engineering and synthetic biology applications. Fundamental metabolic differences in pheromone-treated populations were identified by comparing external metabolite fluxes, metabolic and global gene expression patterns, and the production capacity of a heterologous compound of industrial importance, para-hydroxybenzoic acid (PHBA).",
"discussion": "3 Results and discussion 3.1 Gene expression capacity and growth characteristics of the pheromone-response An ideal bio-production phase has a high general capacity for gene expression such that heterologous enzymes can be expressed to a high level. A production phase should also have high metabolic activity through central carbon metabolism to allow efficient supply of carbon precursors to products. These factors are particularly important to assess for the pheromone-response because all of the other known types of growth arrest in yeast result in a ‘stationary phase’ which is characterised by low gene expression and metabolic activity ( Albers et al., 2007 , Herman, 2002 , Werner-Washburne et al., 1993 ). To address the question of gene expression capacity of the cell-cycle arrest associated with the pheromone-response, GFP expression levels between cultures treated with and without synthetic α-pheromone were compared (strain PSP03, Table 2 ). The TEF1 promoter was used to control GFP expression because the promoter is constitutively active ( Da Silva and Srikrishnan, 2012 ) and TEF1 mRNA levels are unaffected by pheromone treatment ( Roberts et al., 2000 ). Consequently, the level of GFP should reflect the overall gene expression capacity of the populations. To induce the mating response, alpha-pheromone was added to one set of parallel cultures at 4.5 h. A rapid reduction in growth rate was observed within two hours relative to the non-pheromone-treated control, consistent with the cell cycle arrest phenotype of the pheromone-response ( Bardwell, 2005 ) ( Fig. 1 a). GFP fluorescence showed a similar pattern in both treatments throughout the experiment ( Fig. 1 a). Fluorescence initially increased up until 4 h, then declined over the remainder of the culture. The decreasing expression from the TEF1 promoter is consistent with our recent findings ( Peng et al., 2015 , Williams et al., 2015b ), and is explained by the use of a highly destabilized version of GFP with a 20-min half-life ( Mateus and Avery, 2000 ). This destabilized GFP is highly sensitive to decreases in expression, which may have been masked by the stability (7 h half-life) of the GFP protein used in previous analyses of the TEF1 promoter in yeast ( Partow et al., 2010 , Sun et al., 2012 ). Although this decline occurred after pheromone addition in the treated culture, the fluorescence response was the same in the untreated culture, indicating that this is a generic pattern during cultivation rather than being pheromone-specific. The declining GFP expression rate in the treated strain (−782±104 au/cell/hr) was very similar to the control strain (−934±54 au/cell/hr) over the remainder of the cultivation time following initiation of culture growth arrest. These data suggest that gene expression capacity remains as active during G1 phase growth arrest as it is in exponentially growing populations, thus demonstrating the potential for this phase to be useful for production. Fig. 1 A strain with constitutive expression of GFP (PSP03, a ), and a strain with no GFP expression (PSP04, b-e ) were grown for 15 h in shake-flasks with (squares, purple lines) and without (triangles, black lines) 1 µM alpha pheromone added at 4.5 h (vertical dashed lines). GFP expression per cell (dashed lines) and population density (solid lines) were measured ( a ) along with extracellular glucose (dashed lines in b ) and extracellular metabolite concentrations ( b - e ). Relative carbon concentrations were calculated as described in ( Stephanopoulos et al., 1998 ) after 15 h of growth ( f ), where an S. cerevisiae carbon content of 24.6 g/C-mol with 7% ash ( Stephanopoulos et al., 1998 ) was used along with a conversion factor of 0.243 to convert OD 660 nm to grams dry cell weight. All measurements were carried out in biological triplicate with error bars representing ±1 standard deviation. Fig. 1 In addition to gene expression capacity, another principle requirement for metabolic engineering applications is to have an active central carbon metabolic network. This can be investigated generally by examining biomass accumulation in parallel with carbon source consumption and the accumulation of the predominant extracellular metabolites produced by S. cerevisiae . Parallel measurements of glucose uptake, along with biomass, acetate, ethanol, and glycerol production throughout 15 h of shake-flask cultivation clearly demonstrated that populations treated with pheromone were at least as metabolically active as the non-treated control populations ( Fig. 1 b-h). As with the GFP expression strains, pheromone addition at 4.5 h resulted in a characteristic decrease in growth rate ( Fig. 1 b), but surprisingly a very similar glucose consumption profile ( Fig. 1 c) with rates of 15.40±1.52 mmol g −1 h −1 and 12.55±1.47 mmol g −1 h −1 for pheromone treated and non-treated respectively ( Table 3 ). In contrast to the relatively small differences in glucose consumption and ethanol production, glycerol and acetate production were markedly increased in pheromone-treated cultures ( Fig. 1 , Table 3 ). Table 3 Summary of external metabolite flux rates. Table 3 µ (h −1 )* Glucose uptake (mmol g −1 h −1 ) Ethanol (mmol g −1 h −1 ) Glycerol* (mmol g −1 h −1 ) Acetate* (mmol g −1 h −1 ) Respiratory quotient Control 0.26±0.03 12.55±1.47 20.52±2.58 0.53±0.17 0.16±0.06 4 Treated 0.17±0.02 15.40±1.52 19.65±2.85 2.17±0.29 0.89±0.11 2 Average growth and extracellular product secretion rates are shown for cultures treated with alpha-pheromone (treated) and without (control). All rates were calculated using data points after the addition of pheromone at 4.5 h, including in the control populations. Mean values of biological triplicates are presented with errors representing ±standard error. * denotes significant difference between control and pheromone treated groups (two sided students t -test with equal variance with p≤0.05). When summarized with end-point metabolite concentration values converted to carbon moles ( Fig. 1 f) it is clear that the carbon not used for biomass production is directed towards side-product formation in the form of glycerol, acetate, and CO 2 (calculated as 296 and 273 carbon moles for pheromone treated and non-treated respectively). When carbon balance values were used to infer CO 2 and O 2 production and consumption rates, the respiratory quotients obtained ( Table 3 ) strongly suggest that the pheromone-response entails a shift towards a more respiratory metabolism compared to the fermentative metabolism of the control cultures. The current understanding of the shift between fermentative and respiratory metabolism in yeast is that a reduced glucose uptake rate results in de-repression of TCA cycle enzymes and a corresponding increase in TCA cycle flux, oxidative phosphorylation, and a decrease in fermentative flux towards ethanol ( Blank and Sauer, 2004 , Dijken et al., 1993 , Heyland et al., 2009 ). The slightly reduced ethanol production rate of pheromone treated populations is consistent with this understanding, but it is interesting that the specific glucose uptake rates are not significantly different between the groups. These data support the concept that cell cycle arrest in response to pheromone results in an active and distinct metabolic phenotype, as compared to ‘standard’ carbon/nitrogen-limited stationary phases ( Albers et al., 2007 ). 3.2 Transcriptome analysis In addition to assessing the general gene expression capacity and external metabolic fluxes during the pheromone-response, it is also important to consider global changes in gene expression. Most starvation based stationary phases are characterised by the induction of stress resistance modules at the transcriptional level, and the hypothesis that the pheromone-response growth arrest phenotype is distinct from these phases can be tested using transcriptomics. Global changes in S. cerevisiae gene expression in response to pheromone treatment have previously been reported ( Roberts et al., 2000 ). This study elegantly demonstrated the complexity of the pheromone-response pathway and the degree to which its signalling components are related to other MAPK modules. However, the data interpretation/analysis did not include transcriptional changes outside of the signalling and effector components of the pheromone-response and related MAPK modules. To gain insight into other changes of relevance to metabolic engineering, the data were re-analysed after excluding any genes primarily involved in the pheromone-response (GO terms ‘sexual reproduction’ and ‘conjugation’). A 99% confidence interval and minimum 2-fold change were used as selection criteria to identify up-regulated ( Table S2 ) and down-regulated ( Table S3 ) genes. Structural processes, including cell wall and cytoskeletal organization, were up-regulated in pheromone-treated populations. This likely relates to the characteristic ‘shmoo’ cell morphology of the mating phenotype. Many up-regulated genes were assigned to categories associated with control of the cell cycle, mitosis, budding, and cytokinesis. These genes regulate the characteristic G1 phase cell-cycle arrest of the mating phenotype. The same categories involving cell-cycle related genes that were up-regulated also featured in the down-regulated gene GO terms ( Table S3 ). This reflects the complexity of regulation required to arrest cell division, with the coordinated up- and down-regulation of a multitude of genes needed to elicit such fine control. The GO term ‘Transposition’ refers to the movement of DNA between non-homologous sites and includes many retrotransposon genes. It has previously been shown that Ty3 retrotransposons are up-regulated in mating populations ( Kinsey and Sandmeyer, 1995 ), and it was interesting to see that genes more generally involved in transposition were significantly up-regulated ( Table S2 ). Transposition during the pheromone-response might provide a mechanism to increase genetic variation in the population prior to mating and, given that the process appears to be regulated by the host (not the transposons), could serve as an example of symbiotic retrotransposition. The most notable down-regulated genes were involved in ribosomal RNA biogenesis and processing. This is consistent with the fact that much of the cellular resources of an exponentially growing population are directed towards ribosome synthesis ( Warner et al., 2001 ), and demonstrates a down-regulation of this process during the pheromone-response phase. Although a decrease in ribosome biogenesis could be thought of as limiting the protein expression capacity of the cell, there are still very high expression levels of genes which are switched on during the pheromone-response ( Bardwell, 2005 ), and we have previously demonstrated sustained induction of a pheromone regulated metabolic pathway ( Williams et al., 2015a ). Given the drain that ribosome biogenesis imposes on ATP supply ( Warner et al., 2001 ), and the fact that pheromone regulated genes can still be highly expressed, a reduction in ribosome synthesis can be viewed as a principle requirement for a ‘productive stationary phase’. Many down-regulated genes were involved in DNA replication/repair and chromosome segregation, again reflecting the arrested state of cell division. It is interesting to note that these processes are also typical of the starvation responses associated with stress induced stationary phases ( Wu et al., 2004 ), but that the yeast which were used for the pheromone-response transcriptome analysis were cultured in rich YPD media, and were not starving ( Roberts et al., 2000 ). This observation highlights the unique nature of the pheromone mediated growth-arrest, and validates the idea of attempting to use it as a production phase where flux towards biomass is limited while nutrients are still abundant. Changes in the expression levels of metabolic genes are of direct relevance to the metabolic component of the pheromone-response. Therefore in addition to analysing global transcriptional changes, central carbon metabolism-specific changes were analysed by mapping the expression levels of significantly changed (95% confidence intervals) metabolic genes along with the reactions they encode ( Fig. 2 ). Significant increases in the transcript levels of a multitude of metabolic enzymes were observed, further suggesting that the mating phenotype has a distinct and active metabolism. In particular, expression levels of genes involved trehalose and glycerol synthesis, the TCA cycle, and the pentose phosphate pathway were significantly up-regulated ( Fig. 2 ). The high expression levels observed for multiple central carbon metabolic genes, along with the active secretion of metabolic side-products ( Fig. 1 ) act as a strong indication that central carbon metabolism is active during the growth-arrest phenotype. There were also a number of interesting trends in regards to specific metabolic processes that are worth speculating on. Fig. 2 Transcript levels for genes encoding central metabolic enzymes are shown with up-regulation in green and down-regulation in red. Metabolites are connected by arrows representing reactions catalysed by enzymes (gene names in boxes). Single headed arrows represent one way reactions and double headed denote reversible reactions. The metabolic map and abbreviations were adapted from Oliveira et al. (2012) . Fig. 2 Trehalose acts as a major storage carbohydrate in yeast, and increased synthesis has been associated with exposure to thermal, osmotic, and ethanol stress ( Pereira et al., 2001 ). Given that these experiments were carried out at 30 °C ( Roberts et al., 2000 ) and considering that the hyper osmolarity glycerol (HOG) response to osmotic stress and the pheromone-response are insulated from one another ( O’Rourke and Herskowitz, 1998 ) it is possible that ethanol stress during the pheromone-response may be linked to the observed up-regulation of trehalose synthesis genes ( Fig. 2 ). It has previously been reported that trehalose synthesis is required to enable endocytosis at relatively low ethanol concentrations ( Lucero et al., 2000 ). Endocytosis is a process where cells internalise their plasma membrane proteins from the extracellular environment, and is integral to the pheromone-response in yeast where pheromone bound membrane receptor proteins are internalised ( Marsh et al., 1991 ). Consistent with this idea, genes involved in endocytosis were highly up-regulated in response to pheromone ( Table S2 ). It is therefore possible that the up-regulation of trehalose biosynthetic genes in response to pheromone evolved as a mechanism to protect cells from ethanol stress during pheromone bound receptor endocytosis. An alternative explanation is that there is actually a low level of osmotic stress associated with the morphological changes that occur during the pheromone-response, which is responsible for the up-regulation of storage carbohydrate synthesis (see below). Up-regulation of glycerol synthesis genes ( Fig. 2 ) is consistent with the much higher levels of glycerol accumulation observed in pheromone-treated cells ( Fig. 1 ). It has been proposed that mating yeast cells require a precise osmotic balance prior to cell wall degradation and membrane fusion, and that this balance is achieved with the export of glycerol from the cell via the FPS1 transporter ( Philips and Herskowitz, 1997 ). Recent work has further demonstrated the capacity of yeast responding to pheromone to excrete glycerol, demonstrating that the HOG pathway is actually partially activated by the pheromone-response ( Baltanas et al., 2013 ). The results presented here strongly support these findings, and support the role of trehalose synthesis as an osmoprotectant rather than solely to enable endocytosis under ethanol stress. Respiratory metabolism requires a greater flux through the TCA cycle for the generation of reducing power to drive oxidative phosphorylation through the electron transport chain. The strong up-regulation of TCA cycle genes suggests that pheromone treatment results in a more highly respiratory metabolism. The extracellular metabolite accumulation rates and respiratory quotients independently support this finding ( Table 3 ). The oxidative branch of the pentose phosphate pathway (PPP) is initiated by glucose-6-phosphate dehydrogenase (ZWF1) in an irreversible step ( Nogae and Johnston, 1990 ). The PPP is responsible for producing NADPH, a critical source of reduction potential required by many anabolic pathways ( Minard et al., 1998 ). The PPP also plays a major role in mitigating the effects of oxidative stress by supplying NADPH to glutathione- and thioredoxin-dependent enzymes ( Slekar et al., 1996 ). The strong up-regulation of PPP genes (ZWF1, GND2, RPE1) in response to pheromone ( Fig. 2 ) could occur as a consequence of the increased rate of respiration initiated by pheromone ( Table 3 ) and the subsequent increase in oxygen radicals. In concordance with this idea was the up-regulation of a multitude of genes involved in both oxidative stress and respiration ( Table S2 ). It is important to note that the cultures which were used for RNA extraction in the original transcriptome study ( Roberts et al., 2000 ) were carried out under different conditions than the cultures used here to calculate extracellular flux rates (rich YPD rather than minimal medium, Fig. 1 , Table 3 ). Therefore the trends in metabolic gene expression levels which correspond to the flux data should be considered as independent, but consistent observations. 3.3 Engineered pathway productivity during the pheromone-response To directly test the hypothesis that the pheromone-response is suitable as a production phase, heterologous compound titers from a strain engineered to produce PHBA±pheromone were compared. PHBA is an important industrial chemical used in liquid crystal polymers ( Krömer et al., 2012 ) that can be derived from the shikimate pathway ( Averesch and Krömer, 2014 , Winter et al., 2014 ), which is now also being developed for the synthesis of important products such as codeine and morphine ( DeLoache et al., 2015 , Galanie et al., 2015 , Thodey et al., 2014 ). PHBA productivity under pheromone exposure was here used to assess the suitability of the mating phenotype as a production phase. A minimally engineered PHBA producing strain was constructed as in Fig. 3 using manipulations previously shown to be effective at increasing shikimate pathway flux ( Curran et al., 2013 ). The constitutive TEF1 promoter was used to drive heterologous expression of the UBiC gene and over-expression of both the TKL1 and the feedback resistant ARO4 K229L ( Luttik et al., 2008 ) genes in the engineered pathway; this ensured that gene expression levels would be consistent between treated and untreated populations. Fig. 3 Engineered PHBA production pathway. Gene expression constructs (a) and the role of the expressed enzymes ( TKL1 , ARO4 , UBiC ) are shown for PHBA production from the shikimate pathway (b). Fig. 3 The PHBA producing strain (PSP05) was grown with and without 1 µM alpha pheromone treatment in early log phase (OD 660 nm ~1, 4 h). As expected, growth rate slowed significantly after α-pheromone treatment, indicating that the pheromone-response had initiated ( Fig. 4 a). After 48 h, population densities were significantly different, with the control group at an OD 660 nm of 12.93±0.40, and the pheromone-treated group at 10.13±0.40. PHBA concentrations were measured during 48 h of fermentation ( Fig. 4 b), with no significant difference (p=0.11) between the two groups after 48 h (control=274±16 μM, alpha=217±46 μM). In order to more directly assess the capacity for metabolic productivity, the biomass-specific PHBA production yields and rates between pheromone-mediated growth-arrest at 6 h and glucose exhaustion at 48 h were compared with the control population rate and yield during the glucose consumption phase between 0 and 24 h. There was a significant decrease in the biomass specific PHBA production rate during the glucose consumption phase with the control group producing 9.4±1.95 μM g −1 h −1 and the pheromone treated group producing 3.35±0.58 μM g −1 h −1 . Interestingly, the biomass specific PHBA yield was significantly higher in the pheromone treated group during this phase at 93.2±16.2 μM gDCW −1 compared to 62.3±1.3 μM gDCW −1 (p=0.03). Pheromone treatment also resulted in a significantly higher glucose specific PHBA yield during the respective glucose consumption phases (1.64±0.11 μM PHBA \n m M g l u cos e − 1 compared to 1.19±0.05 μM PHBA \n m M g l u cos e − 1 p=0.003). Although the PHBA production rate was lower during pheromone-mediated growth-arrest, both the biomass- and glucose-specific PHBA yields were higher. This suggests that the growth arrest phenotype of the S. cerevisiae pheromone response system results in a more productive glucose consumption phase where carbon otherwise directed towards cell growth becomes available for metabolite production. The titer was not greater in the pheromone treated group due to the lower amount of biomass, and the fact that the growth-arrest phenotype was not initiated until the mid-exponential growth phase. Fig. 4 PHBA production in pheromone treated cultures. A strain minimally engineered to produce PHBA (strain PHBA03) was grown with (purple lines, squares) and without (black lines, triangles) 1 µM alpha pheromone treatment at early exponential phase (OD of 1, at 4 h, indicated by vertical dashed lines). (a) Population density (OD 660 nm ), (b) extracellular PHBA concentration, and (c) extracellular glucose concentrations were measured for shake-flask growth with and without pheromone over 48 h.(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 4 3.4 Practical considerations for growth-arrested metabolic productivity The current data are promising in that we demonstrate the pheromone mediated cell-cycle arrest phenotype is an active phase, but further optimisation of the system would be required to move towards the idealised ‘productive stationary phase’ that was outlined in the introduction. Cells responding to pheromone expend resources executing the mating program, which involves extensive cell-wall remodelling and cell morphology changes. Minimising this energy/resource drain through further metabolic engineering might improve overall pheromone-phase productivity of specific engineered pathways. Moreover, use of a fed-batch approach along with titration of pheromone induction time might demonstrate greater improvements in the pheromone-response productivity compared to the control. A potential issue with using the pheromone-response as a production phase is the capacity for cells to become desensitized to pheromone after prolonged exposure through a variety of well-understood mechanisms ( Bardwell, 2005 ). This phenomenon was not observed during the time-frame of these experiments (population density was still reduced after 48 h of cultivation; Fig. 4 ), but may need to be prevented through engineering for any longer-term applications. Another consideration is that although it is possible to use the mating phenotype as a production phase, adding purified pheromone to a large-scale fermentation is likely to be prohibitively expensive, depending on the value of the fermentation product. To circumvent this limitation, we recently developed a synthetic quorum sensing circuit where cells produce and respond to their own pheromone in a population density dependent manner ( Williams et al., 2013 ). This quorum sensing network has since been coupled to a recently developed yeast RNA interference module ( Crook et al., 2014 , Drinnenberg et al., 2009 , Si et al., 2014 , Williams et al., 2015b ) to successfully dynamically regulate the production of PHBA ( Williams et al., 2015a ). This auto-response system could be applied to ensure sufficient pheromone is available during cultivation. There are many other methods for arresting the cell cycle under nutrient rich conditions, and it is possible that they could be used as inducers of stationary phase metabolic productivity. For example hydroxyurea can be used to inhibit dNTP synthesis and arrest the cell cycle in S phase ( Koç et al., 2004 ), and nocodazole to disrupt microtubule polymerisation and arrest cells in metaphase ( Arber, 2000 ). While these methods have proved invaluable in elucidating cell-cycle related processes in basic research, they are yet to be explored as inducers of productive growth-arrest phenotypes in biotechnology. These inducers would have the advantage of not involving the initiation of a complex mating phenotype and the associated energetic cost. However, they indirectly induce stress response mechanisms that are likely to impose their own limitations on metabolic productivity. Furthermore, unlike with our synthetic pheromone quorum-sensing circuits ( Williams et al., 2015a , Williams et al., 2013 ) there are no available mechanisms for autoinduction using chemical inducers such as hydroxyurea and nocodazole. In an excellent recent example of chemically induced growth-limitation, pantothenate (vitamin B12) was used to regulate growth and farnesene (jet-fuel) productivity in yeast ( Sandoval et al., 2014 ). In this system the removal of pantothenate from the growth medium of a producer strain resulted in a 70% reduction in farnesene yield and a concomitant increase in growth rate. This enabled pantothenate to be used as an inducer whereby a rapid biomass formation phase in the absence of pantothenate was followed by a switch to a growth-limited production mode upon pantothenate addition to the medium. There are likely to be many analogous systems for other engineered pathways and metabolic networks that hold great promise as inducers of stationary phase metabolic productivity. It is possible that the growth arrest induced by pheromone in S. cerevisiae could also be a useful production phase in other yeast species that use sexual pheromones to coordinate mating. However, although some species such as Schizosaccharomyces pombe use pheromones, they only mate under nutrient starvation conditions ( Yamamoto et al., 1997 ). Clearly the growth arrest phenotype would not be effective as a production phase under these conditions. Given the drawbacks of using a naturally evolved mating phenotype for production, a much grander solution is to try and reverse engineer the principle features of a ‘productive stationary phase’. In this scenario a synthetic regulatory circuit could be employed as a ‘master controller’ that can simultaneously arrest the cell cycle, reduce ribosomal biogenesis, and switch on the expression of production pathway enzymes while maintaining high central carbon metabolism fluxes. The fact that growth fluxes are no longer required in this scenario opens up the possibility of silencing the expression of ‘essential’ genes, and producing metabolites or proteins that are normally highly toxic to growth. Although the specific mechanisms required to implement such a dramatic regulatory and metabolic shift would differ between organisms, it is possible that the general features are universal."
} | 9,558 |
33841754 | PMC8010324 | pmc | 5,846 | {
"abstract": "Advances in sequencing technology have led to the increased availability of genomes and metagenomes, which has greatly facilitated microbial pan-genome and metagenome analysis in the community. In line with this trend, studies on microbial genomes and phenotypes have gradually shifted from individuals to environmental communities. Pan-genomics and metagenomics are powerful strategies for in-depth profiling study of microbial communities. Pan-genomics focuses on genetic diversity, dynamics, and phylogeny at the multi-genome level, while metagenomics profiles the distribution and function of culture-free microbial communities in special environments. Combining pan-genome and metagenome analysis can reveal the microbial complicated connections from an individual complete genome to a mixture of genomes, thereby extending the catalog of traditional individual genomic profile to community microbial profile. Therefore, the combination of pan-genome and metagenome approaches has become a promising method to track the sources of various microbes and decipher the population-level evolution and ecosystem functions. This review summarized the pan-genome and metagenome approaches, the combined strategies of pan-genome and metagenome, and applications of these combined strategies in studies of microbial dynamics, evolution, and function in communities. We discussed emerging strategies for the study of microbial communities that integrate information in both pan-genome and metagenome. We emphasized studies in which the integrating pan-genome with metagenome approach improved the understanding of models of microbial community profiles, both structural and functional. Finally, we illustrated future perspectives of microbial community profile: more advanced analytical techniques, including big-data based artificial intelligence, will lead to an even better understanding of the patterns of microbial communities.",
"conclusion": "5 Conclusion and future perspectives of microbial community profile The diversity of microorganisms at the individual level indicates the significance of cataloging a pan-genome in the community, which can be used in subsequent applications in metagenomics. Metagenomics constitutes an effective approach for studying intact microbial communities, especially when incorporating pan-genomics. In this review, we summarized the applications of the pan-genome approaches to metagenomics, for comprehensively describing the microbial communities in specific habitats and conditions. We discussed the application of the integration of pan-genomics and metagenomics in the community to characterize the genetic content in a specific environment and obtain ecologically meaningful views of different ecosystems. Extending the concept of pan-genome to incorporate metagenome has advanced our understanding of microbial diversity and metabolism. Pan-genomics and metagenomics are complementary. Pan-genomics provides unique insights for a given taxon of microbial genomes, while metagenomics identifies the composition and metabolic patterns of microorganisms in the environment. Metagenomes allow estimating the abundance of variation of specific genes in environmental populations, thus avoiding the limitations of cultivation and genomic assembly. Extending from the traditional pan-genome concept to metagenome, it overcomes many biases related to whole-genome sequencing and provides a comprehensive description of the microbial genetic diversity in specific habitats. The combination of pan-genome and metagenome strategy can not only estimate the abundance and distribution of gene clusters in the environment, but also link them with the distribution of microbial populations, thereby providing a solution for the expansion of genome-wide conventional analysis. In addition, to clarify the basic biology of microbial communities, broader pan-genome and metagenome studies should include the regulatory relationship profile, signal transduction and interactions between microbes. The strategy of integrating the pan-genome into the metagenome has led to the discovery of remarkable genomic diversity and the characterization of novel membership in communities. However, for genomes with incomplete metagenome assembly, this strategy should always be used with caution, as it may reduce the quality of genome repositories. Due to the limited coverage of metagenomes and genomes in the environmental communities, coupled with the inherent complexity of assembly algorithms, determining the characteristics of accessory genes in the community is still a challenging task. The integration of pan-genomes derived from metagenomes needs to be careful to avoid deviations caused by spurious expansion of gene content. Efforts need to be made to improve the accuracy of assembly, pan-genome and metagenome analysis tools, and integrate them into the study of microbial communities. In addition, future research on pan-genomes and metagenomes should not be limited to those gene sets with coding functions. The regulatory regions of genomes also have a decisive influence on microbial characteristics. The focus of future research should expand the definition of pan-genome to include non-coding sequences that may perform regulatory or other important functions. This combination strategy also can be applied to broader taxonomic groups as well as other eukaryotes such as fungus in communities. Besides, integrated amplicon-based taxonomies and pan-genomes will give new insights into complex ecosystems. Moreover, the rapid development of microbial single-cell sequencing reveals the intra-population structure or host-microbe interactions of complex microbial communities [80] , and short-reading and assembly-based strategies may not be effective. The method that combines microbial single-cell sequencing, pan-genomics and metagenomics will allow us to understand complex dynamics of population, gene expression and metabolic functions of the microbial genomes and their communities. With recent technological advances, the integration of pan-genomics and metagenomics data with other complex omics data has become increasingly popular. Applying more advanced analytical techniques to large-scale microbial datasets, including big-data analysis based on artificial intelligence, such as meta-analysis of large metagenomic datasets by machine learning and metagenomic signatures analysis by deep learning, will result in an even better understanding of the patterns and dynamics of microbial communities.",
"introduction": "1 Introduction In recent years, genome analysis of microbial organisms has gradually shifted from focusing on single selected individuals or a few genomes to the large-scale comparative analysis of a set of related isolates. Since the gene pool of a species or community is typically much larger than that of any individual strain, the genetic dynamics and diversity of the genomes of different strains in the same species cannot be represented by a single individual genome. The genomic variation observed at the species and community level leads to the expansion of the pan-genome and metagenome concepts. Whole-genome sequencing lays a foundation for the powerful strategy of identifying core and accessory genes shared among close microbes through pan-genome. Pan-genome represents the entire gene repertoire of a group of isolates (e.g. strains from one species), which can characterize the dynamics and diversity of genomes in a given taxonomy, while individual genome usually only accounts for a small part of the pan-genome [1] , [2] . The pan-genome approach is typically used to evaluate the microbial genetic composition in three ways: core genome profiling, accessory genome profiling, and specific genome profiling, revealing the characteristics of homology, diversity, and specialization between genomes [3] , [4] . For example, the core genome found in all individuals is often used to evaluate the relatedness between strains in the same species [5] , [6] . And the pan-genome analysis has revealed extensive horizontal transfer in microbial accessory genomes [7] . Pan-genome analysis has been used to study the genetic diversity of a group of related microbial genomes, including gene composition of individual strains, strain tracking, evolutionary impact, niche specialization, antimicrobial target screening, and diagnostic marker identification [5] , [8] , [9] , [10] , [11] , [12] , [13] . Although pan-genomics is highly informative in profiling microbial diversity and function, it still has limitations. The widely used pan-genome analysis methods [14] , [15] , such as PGAT [16] , PGAP [17] , and Panseq [18] , provide a good strain-level pan-genome profile for isolates, but they cannot resolve the species relationships at the community level. Natural microbial communities have rich biodiversity, and these pan-genome analysis methods have limitations in studying the genetic variation and interaction of communities, hence fail to capture the dynamic behavior of microbial communities. Besides, when defining the size and content of pan-genome, these pan-genome analysis methods can only be applied to a limited number of species that cannot accurately reflect microbial ecology. Overcoming these issues requires new considerations at the community level, such as cross-community evaluation of pan-genome profile. Metagenome refers to the entire genetic content of all microorganisms in a specific environment, which has been widely used to study microbial diversity in various habitats, such as air, soil, water, plants, and humans [19] . Metagenomics is a method to characterize the taxonomic and functional diversity of microbial communities by isolating genome sequences directly from the environment without prior cultivation [20] . Metagenomics can be used for taxonomic analysis and functional analysis to track community dynamics [21] . For example, it has been used to study the shifts in microbiome composition and function of the human body such as oral, skin, gut [22] , [23] . Microbial composition analysis tools, such as MetaPhlAn2 [24] and Kraken2 [25] , can characterize microbial community structure in the environment and human body by identifying microbial species and estimating their relative abundance. Also, HUMAnN2 [26] estimates the abundance of microbial pathways in terms of metagenomes to detect the metabolic potential of microbial communities. Metagenomics has great advantages in the taxonomic analysis at the species level, and some taxonomic studies have even reached the strain level [27] , [28] . However, due to the lack of high-quality reference genomes, a higher taxonomic resolution is still challenging. For some microbial communities in complex environments such as marine sediments and soils, due to underrepresented reference genomes in databases, it is even difficult to distinguish microbes at the species level using metagenomics methods. Besides, metagenomics methods still have limitations in analyzing genomic heterogeneity, and additional efforts are needed to identify all accessory genes and their functions in the microbial communities. Thus, additional complete genomic information is needed to characterize the microbial community in detail. Pan-genomics and metagenomics have made independent breakthroughs in the study of microbial evolution and function in a given taxon and community respectively, but the limited number of cultivation-based microbes and the limited taxonomic resolution respectively restricts their development. Pan-genomics deciphers genomic heterogeneity and diversification of any given taxon, while metagenomics obtains the taxonomic and functional profiles in communities. Thus, combining complementary metagenome and pan-genome analysis strategies can break their limitations in the study of microbial communities, leading to an unprecedented opportunity to study the microbial interactions with their environments and hosts in communities. Recently, some microbiome studies have adopted this complementary strategy, and such integrated pan-genomics with metagenomics techniques have enabled researchers to study the diversity and dynamics of populations in microbial communities [29] ( Fig. 1 ). Such a method takes advantage of both genome and metagenome to connect these two important genetic profiles to microbiome. For example, IMG/M [30] is a comparative data analysis system that integrates genomes and metagenomes. This system provides genome annotation such as COG clusters, Pfam, InterPro domains, and KEGG pathways. It allows comparative analysis of isolated genomes and metagenomes, including the determination of the phylogenetic and functional characteristics of individual genomes, as well as metabolic comparisons across microbial communities. PanPhlAn [31] is a tool that focuses on identifying the genetic composition of individual strains from metagenomes. This tool can characterize pan-genome and metagenome at the strain level, and determine the taxonomic profile, strain-level profile, functional profile, and phylogenetic profile from metagenomic samples. PanFP [32] is a method based on pan-genome reconstruction, which can describe the functions of KEGG Orthology, Gene Ontology, Pfam, TIGRFAMs of microbial community according to 16 s rRNA gene. Microbial communities can be explored not only by sequencing shotgun metagenome, but also by sequencing amplicon. The amplicon sequencing, such as 16S rRNA gene sequencing, is an economical and effective method for microbial abundance and diversity screening. At present, public databases have accumulated a large amount of amplicon survey data from different environments. Therefore, extending the application of amplicon survey to the pan-genome for the exploration of microbial communities is a cost-effective new method for studying complex ecosystems. Considering that 97.9% of the 16 s rRNA sequence divergence can be used as a representative of species boundaries [33] , [34] , pan-genome analysis with amplicon survey data may lead to hybrid genomes, so the application of amplicon survey to the pan-genome should pay close attention to the sequence identity thresholds. Fig. 1 The strategy that integrated pan-genomics with metagenomics for studying microbial diversity and dynamics in microbial communities. The genetic contents characterized in pan-genomes and metagenomes of the community can be collectively pooled to establish a community profile. The tools available for each workflow are shown in brackets. Integrated pan-genome and metagenome approaches are critical to a system-level understanding of the relevant population-level genetic content of a particular habitat. In this review, we discussed the existing and new integration directions of pan-genomics and metagenomics, and highlighted the strategies and applications of such an integrative approach in the study of microbial community. This integrated strategy promotes the mechanism and model for the evolution and function of microbial communities to a higher level."
} | 3,773 |
25665577 | PMC4381525 | pmc | 5,847 | {
"abstract": "Accurate evaluation of microbial communities is essential for understanding global biogeochemical processes and can guide bioremediation and medical treatments. Metagenomics is most commonly used to analyze microbial diversity and metabolic potential, but assemblies of the short reads generated by current sequencing platforms may fail to recover heterogeneous strain populations and rare organisms. Here we used short (150-bp) and long (multi-kb) synthetic reads to evaluate strain heterogeneity and study microorganisms at low abundance in complex microbial communities from terrestrial sediments. The long-read data revealed multiple (probably dozens of) closely related species and strains from previously undescribed Deltaproteobacteria and Aminicenantes (candidate phylum OP8). Notably, these are the most abundant organisms in the communities, yet short-read assemblies achieved only partial genome coverage, mostly in the form of short scaffolds (N50 = ∼2200 bp). Genome architecture and metabolic potential for these lineages were reconstructed using a new synteny-based method. Analysis of long-read data also revealed thousands of species whose abundances were <0.1% in all samples. Most of the organisms in this “long tail” of rare organisms belong to phyla that are also represented by abundant organisms. Genes encoding glycosyl hydrolases are significantly more abundant than expected in rare genomes, suggesting that rare species may augment the capability for carbon turnover and confer resilience to changing environmental conditions. Overall, the study showed that a diversity of closely related strains and rare organisms account for a major portion of the communities. These are probably common features of many microbial communities and can be effectively studied using a combination of long and short reads.",
"discussion": "Discussion Previously, short-read data from aquifer sediments enabled the recovery of one complete genome (RBG-1), many partial Chloroflexi genomes, and many long fragments that are currently being binned to genomes that will be reported separately (K Anantharaman, CT Brown, I Sharon, BC Thomas, A Singh, LA Hug, CJ Castelle, KH Williams, EL Brodie, JF Banfield, et al., unpubl.). Alignment of synthetic long reads to the manually curated RBG-1 genome confirmed the assembly of the genome and revealed two local misassemblies. This result confirms that high-quality genomes can be recovered from short-read metagenomic data when the proper quality control steps are taken. The short-read assembly fails to reconstruct rare genomes and genomes from groups of closely related organisms. Notably, the most abundant organisms in all the samples were missed in short-read data assemblies. This outcome demonstrates one of the major limitations of de Bruijn graph assemblers for metagenomics data. These assemblers require a high degree of sequence conservation for the assembled genomes, and therefore may fail in the presence of closely related heterogeneous genomes. The majority of short reads in all our samples remained unassembled, suggesting that rare and closely related genomes account for significant portions of the studied communities. However, given the high number of synthetic long reads with low coverage, we estimate that the majority of unassembled reads in the short-read data were left unassembled because of low coverage. Similar or lower levels of assemblies were also reported in other metagenomic studies of complex environments such as the ocean ( Iverson et al. 2012 ) and soil ( Howe et al. 2014 ). Mapping of short reads to the synthetic long reads shows that the coverage for 8%–17% of the synthetic long reads in the three samples is lower than 0.1×. We therefore estimate that ∼100 times more short-read sequencing data than available for this study will be required for the recovery of the genomes for 80%–90% of community members in the three samples. The increase in sequencing throughput is expected to improve the recovery of rare genomes from short-read data, but is not expected to improve the assembly of closely related genomes. The TrueSeq Synthetic Long-Read technology currently offers ∼8 kbp reads at almost two orders of magnitude less throughput (bp per lane), compared to the Illumina HiSeq platform. The study of rare genomes is facilitated by the synthetic long-read technology due to the read size, which is usually sufficient for reliable gene prediction of unassembled reads. In addition, the presence of multiple genes on each synthetic long read enabled the development of a synteny-based approach for the recovery of genome architecture for groups of closely related genomes. The outcome of this process is a gene-centric description of the regions shared in the genomes of organisms in the population. This approach allowed us to recover metabolic features that are common to all related genomes despite not recovering any specific genome. On the other hand, no significant assembly of synthetic long reads was achieved for any single genome in our samples due to the relatively low throughput of the technology and the high complexity of the samples. Scaffolding of short-read assemblies using the long reads was also limited for the same reason: Abundant genomes in the samples typically had sufficient coverage by the short-read data for extensive assembly, and the numerous synthetic long reads from each genome rarely matched a region that was required for scaffolding. Overall, short- and long-read data provide complementary advantages for metagenomics studies, thus making the use of both technologies together more powerful than use of one alone. Analysis of the long-read data revealed that the most abundant species in all three samples belong to populations of multiple (possibly dozens of) different closely related strains and species. As previously described for sediment-associated RBG-1 populations, the abundance of the Aminicenantes (4-m) and Deltaproteobacteria (5- and 6-m) populations in sediment close to and below the water table may be attributed to metabolic flexibility. The Aminicenantes and Deltaproteobacteria populations are numerous and fairly closely related. In addition to their detection in the acetate-amended groundwater, organisms closely related to the Deltaproteobacteria identified here were also recovered from other locations in the same aquifer (LA Hug, BC Thomas, I Sharon, CT Brown, MJ Wilkins, KH Williams, A Singh, and JF Banfield, in prep.), suggesting that this lineage is a key player in the Rifle sediment environment. The presence of many strains from this lineage may increase the niches occupied by these organisms. It may also improve the overall resistance of the population to threats such as phages ( Sharon et al. 2013 ). RBG-1, on the other hand, was found to be abundant only in the samples studied here, which may suggest that specific environmental conditions at the time of sampling contributed to its high abundance. Our results show that microbial communities in sediment consist of a few abundant species and a “long tail” of thousands of rare species whose abundance is <0.1%. The majority of these rare organisms belong to phyla also represented by abundant organisms. Our results show that significantly more glycosyl hydrolases than expected are found in the rare organisms' fraction and that these glycosyl hydrolases cluster into more families than glycosyl hydrolases detected in abundant organisms. Theses enzymes can, therefore, increase the repertoire of nutrients degraded by community members, thus increasing the community's ability to adjust to changing environmental conditions."
} | 1,910 |
28378827 | PMC5381106 | pmc | 5,848 | {
"abstract": "To understand the post-transcriptional molecular mechanisms attributing to oleaginousness in microalgae challenged with nitrogen starvation (N-starvation), the longitudinal proteome dynamics of Chlorella sp. FC2 IITG was investigated using multipronged quantitative proteomics and multiple reaction monitoring assays. Physiological data suggested a remarkably enhanced lipid accumulation with concomitant reduction in carbon flux towards carbohydrate, protein and chlorophyll biosynthesis. The proteomics-based investigations identified the down-regulation of enzymes involved in chlorophyll biosynthesis (porphobilinogen deaminase) and photosynthetic carbon fixation (sedoheptulose-1,7 bisphosphate and phosphoribulokinase). Profound up-regulation of hydroxyacyl-ACP dehydrogenase and enoyl-ACP reductase ascertained lipid accumulation. The carbon skeletons to be integrated into lipid precursors were regenerated by glycolysis, β-oxidation and TCA cycle. The enhanced expression of glycolysis and pentose phosphate pathway enzymes indicates heightened energy needs of FC2 cells for the sustenance of N-starvation. FC2 cells strategically reserved nitrogen by incorporating it into the TCA-cycle intermediates to form amino acids; particularly the enzymes involved in the biosynthesis of glutamate, aspartate and arginine were up-regulated. Regulation of arginine, superoxide dismutase, thioredoxin-peroxiredoxin, lipocalin, serine-hydroxymethyltransferase, cysteine synthase, and octanoyltransferase play a critical role in maintaining cellular homeostasis during N-starvation. These findings may provide a rationale for genetic engineering of microalgae, which may enable synchronized biomass and lipid synthesis.",
"discussion": "Discussion Switching the fuel source from fossils to sustainable bioenergy resources is the need of the hour. Algae-based biofuels have gained much attention recently owing to their superiority over terrestrial biofuel crop sources. Several proteomics investigation of algae has been performed 4 10 36 37 . Gao and co-workers performed comprehensive comparative genomics, transcriptomics and proteomics analysis of Chlorella protothecoides sp. 0710 to determine the oil accumulation mechanisms 36 . The study distinguished the autotrophic and heterotrophic growth conditions using gel-based comparative proteomics followed by LC-MS/MS analysis. Ma and group reported varying inoculum size in a non-model green microalga Chlorella sorokiniana greatly affects cell density, an essential criteria for industrial production of biodiesel from microalgae. The proteins participating in photosynthesis (light reaction) and Calvin cycle (carbon reaction pathway) displayed highest levels of differential expression under inoculum size of 1 × 10 6 cells mL −1 , and lowest under 1 × 10 7 cells mL −1 4 . Guarnieri et al . 37 investigated the global proteome profile of N-starved Chlorella vulgaris ; briefly two different conditions (N-sufficient and N-starved) were considered using gel-based liquid chromatography-mass spectrometry (GeLC/MS). The data indicated enhanced fatty acid and triacylglycerol biosynthetic machinery under N-depletion condition. However, the proteins identified in the study remained unvalidated 37 . The present study was thus undertaken to identify the proteins expressed differentially in longitudinal manner in a non-model oleaginous green microalga Chlorella sp. FC2 IITG. Nitrogen and phosphate starvation are identified as the two major triggering factor for neutral lipid accumulation in FC2 cells; however N-starvation resulted in rapid changeover in neutral lipid content from 1% to 54.4% (w/w, DCW) 28 . The time point 0 h designates the N-sufficient condition while other three time points viz. 40, 88 and 120 h were taken into account to get a comprehensive insight into the N-starvation induced lipid accumulation pathways. A sharp increase in the neutral lipid content was observed post 40 h followed by an exponential increase in neutral lipid content for up to 88 h, which subsequently dropped following 120 h of starvation ( Fig. 1A ). These time points thus represent the critical stages for neutral lipid accumulation and hence were selected for proteomic analysis. The preferential utilization of intracellular nitrogen attributed to the constant growth in the N-deprived FC2 cells up until the initial 40 h of starvation 38 . Advanced stages of N-starvation attributed to drastic drop-down in growth rate with concomitant elevation in neutral lipid content (15.48% to 50.34% w/w, DCW). Indeed, N-starvation tends to shut down housekeeping functions; the same is imitated in the present study where protein, carbohydrate and chlorophyll contents were progressively down-regulated in FC2 with prolonged N-starvation duration. The global protein expressions of FC2 as a function of time under N-starvation conditions were evaluated by two complementary proteomics technique; iTRAQ and DIGE, with three biological replicates for each time-point. The iTRAQ data thus obtained was quantile normalized. Furthermore, the time-resolved proteomes were validated by MRM of 6 selected proteins and western blotting for 3 such proteins ( Fig. 1 ). The expression patterns of PRK and MDH determined using western blotting and MRM were in sink, but showed inconsistency with iTRAQ data ( Fig. 4 ). The reason for this discrepancy might be the low abundance of peptides, smaller sample size, and complicated experimental procedures and data analysis lacking suitable internal standards 39 40 . Despite several disadvantages, iTRAQ is the most amenable technique to orthogonal separation due to multiplexing that ensures its value for many analysis schemes and the reduction of costly LC–MS runtime 41 42 43 . The targeted proteomics based on MRM has emerged as a technology to complement the discovery capabilities and overcome the technical pitfalls, such as incomplete protein extraction, proteolysis, and artifactual protein modifications 44 . Comparative temporal proteomic analysis of FC2 cells indicated regulation of diverse protein classes under the following sub-classes: (a) N-assimilation, amino acid biosynthesis and protein degradation; (b) photosynthesis; (c) energy pathways; (d) fatty acid metabolism; and (e) stress-responsive mechanisms. The discussion hereafter focuses on the specific proteins under these sub-headings identified in our study and their associated role. N assimilation, amino acid biosynthesis, and protein degradation Global proteome re-adjustment in terms of both amino acid biosynthesis and protein degradation (involving proteasomes) pathways was observed in FC2 cells as a feedback mechanism to N-starvation. Overall, the enzymes involved in the biosynthesis of glutamate, aspartate and arginine were elevated. Plastidal ferrodoxin-dependent glutamate synthase (Fd-GOGAT) was 1.5 folds up-regulated following 88 h of N-starvation, suggesting glutamate accumulation in FC2 cells. The GS-GOGAT pathway although energetically expensive, is often triggered during low N- concentrations 45 46 . In parallel, the elevated levels of MDH (TCA cycle) and isocitrate lyase (ICL; glyoxylate cycle), suggests enhanced oxaloacetic acid (OAA) accumulation in FC2 cells. Moreover, OAA may be subsequently transaminated to aspartate using the Aspartate aminotransferase (AAT), which again was up-regulated by more than 2 folds in the advanced N-starvation stages. Two of the arginine biosynthetic enzymes namely Arginosuccinate synthase (AsuS) and Arginine biosynthesis bi-functional protein (ArgJ) were up-regulated at 88 and 120 h of N-starvation ( Fig. 3 , Table 1 ). Previous studies displayed a positive correlation of transcription factor bHLH6 with AsuS in Chlamydomonas during N-starvation 47 . Arginine accumulation is reported during N-deprivation conditions; arginine catabolism assists in mobilizing the stored nitrogen based on the nutritional status of the cells. Besides, arginine plays role in augmenting stress-responsive mechanisms thereby reducing the overall effect of oxidative and other abiotic stresses as reported in Arabidopsis thaliana 48 . Arginine accumulation was also mirrored in our parallel metabolomics investigation (data unpublished). The accumulation of arginine, having highest nitrogen to carbon ratio, could be a strategy adopted by the algal cell to store organic nitrogen and combat abiotic stress. Simultaneously, degradation of proteins using various catabolic enzymes including peptidase, proteasomes and ubiquitin ( Table 1 ) augmented the recycling of nitrogen and several TCA cycle intermediates. These findings are consistent with the transcriptomics and metabolomics 49 , and label-free proteomics 50 data in P. tricornutum (model diatom) following N-starvation. Our data clearly indicates a tight regulation of several enzymes associated with carbon and nitrogen metabolism in response to N-starvation. Photosynthesis N-starvation is often linked to reduction in photosynthetic efficiency, primarily due to chlorosis. Porphobilinogen deaminase, involved in chlorophyll biosynthetic process, was down-regulated by 1.3 folds during initial N-starvation phase (40 h) and corresponds well to the physiology ( Fig. 1A ). The decline in chlorophyll levels during N-starvation is often associated with rapid cessation in its synthesis and dilution by cellular growth rather than its degradation as reported in C. reinhardtii 51 . Interestingly, several of the photosynthetic proteins including photosystem II assembly, E1ZPZ7, E1ZQR2 (KEGG: PSII oxygen evolving enhancer protein 1 and 2; psbp), ferredoxin-NADP reductase, E1ZFB3 (thylakoid lumenal protein), and cytochrome 3 were up-regulated during the process. It has been already shown that the novel isolate FC2 derives energy and carbon for de novo TAG synthesis from photosynthesis during N-starvation 19 , although photosynthetic yields are compromised due to reduced chlorophyll content 52 , and photosynthetic carbon fixation 53 . The later is primarily due to curbed regeneration of Ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) 54 ; reduced expressions of PRK and SBP is perhaps the rate limiting step although other photosynthetic related enzymes are up-regulated. Elevated levels of some of these photosynthetic proteins is in corroboration with the previously described plant omics study; for instance psbp was up-regulated in two different maize cultivars grown in low N-conditions 55 ; possibly released as a degradation product of oxygen evolving complex proteins that assist FC2 cells in adapting to the adverse conditions 56 . Energy pathways The enzymes ferredoxin-NADP reductase and that of PSII are involved in photophosphorylation that fulfils the energy requirements of the cell. However, under unfavourable conditions the energy is re-directed towards lipid-accumulation which serves as energy-reserve for the cell over the prolonged stress durations 57 . The increased energy requirements to synthesise high-energy compounds are attained by glycolysis, TCA cycle and non-oxidative pentose phosphate pathway (PPP). Likewise, FC2 cells displayed elevated energy-metabolism activities with significant coverage of the proteins linked to glycolysis, TCA cycle and non-oxidative pentose phosphate pathways (PPP). The non-oxidative PPP are primarily involved in the inter-conversion of sugars that can re-enter glycolysis or oxidative PPP for generation of reducing equivalents (NADPH). These reductants, apart from their role in maintaining redox (particularly overcoming oxidative stress), also find function in supporting de novo fatty acid biosynthesis 58 59 and N-assimilation 20 . The synergistic actions of PPP and glycolytic pathways using a different combination of their respective enzyme sets provide cells with flexibility to modulate energy levels, reducing power or a combination of these functions 60 . The involvement of glycolytic enzymes namely GPDH, ALDO, TPI and PK has been reported to be the major regulators for lipid induction in oleaginous Scenedesmus dimorphus, S. quadricauda and Mucor circinelloides 61 62 . Co-ordinated expression of the enzymes during N-starvation redirects the carbon-flux from carbohydrate towards neutral lipid biosynthesis via pyruvate, which is the key precursor for acetyl CoA. Likewise, ribose-5-phosphate isomerase (RPI) and ribulose-phosphate-3-epimerase (RPE) functions in sink to supply the carbon and limiting NADH required for lipid biosynthesis. Transcriptomic analysis of N-deprived Neochloris oleoabundans revealed over-expression of PPP 63 . In the present study GPDH, PGK, ENL, Fructose-1,6-bisphosphatase (FBP), ALDO, TPI, and PK belonging to glycolytic pathway, and RPI and RPE from non-oxidative PPP work in a cordial manner to generate NADPH and pyruvate that may be converted to ATP ( Fig. S5 ), and fulfil the energy needs of FC2 cells for sustenance of N-starvation. Furthermore, up-regulation of MDH and phosphoenolpyruvate carboxykinase (PEPC) suggests conversion of malate to OAA to phosphoenolpyruvate (PEP), which is subsequently converted to pyruvate via PK. The pyruvate can then enter the TCA cycle and contribute to the production of mitochondrial citrate, which can then feed into the de novo fatty acid synthesis upon its export to the cytoplasm 64 . Our data is consistent with the enriched gluconeogenesis transcripts in a starchless mutant of C. reinhardtii grown in N-starvation conditions 14 . The FC2 cells ascertained that the PEP is not a limiting factor for ATP and pyruvate generation by elevating the levels of a complementary glycolytic enzyme; ENL. PEP may serve as a precursor molecule for isoprenoid and glycerolipid biosynthesis through pyruvate and acetyl-CoA 65 . Interestingly, the abundance of ENL and PK is in accordance with the temporal lipid accumulation in N-starved FC2 cells as confirmed by nile red staining. An elevated level of ENL is also reported in N-starved rice 66 and Arabidopsis 67 . Proteins associated with fatty acid metabolism Earlier findings from our group and others have highlighted the complex links between lipid accumulation and N-starvation in microalgae. Transcriptomics study in Nannochloropsis oceanica IMET 1 revealed up-regulation of Acyl carrier protein (ACP) by more than 2 folds, while the levels of 3-hydroxyacyl-[ACP]-dehydratase (HAD) and ENR were down-regulated at 48 h under N-deprived conditions 68 . Contrarily in the present study, gradual accumulation of several fatty acid biosynthesis and storage proteins was observed with progressive N-starvation duration. Several components of fatty acid synthase (FAS) including E1Z5W8 (ACP), E1Z8J0 (HAD), and E1Z2Y2 (ENR) were significantly accumulated following 88 h of N-starvation ( Table 1 ). In algae, ACP tethers the growing fatty acid chain as it goes through the elongation step. Dehydratase and reductase catalyzes the third and fourth step of fatty acid elongation, wherein enoyl-ACP losses one molecule of water and is reduced to fully saturated acyl-ACP. Additionally, the involvement of two of the enzymes E1Z2Y2 (ENR) and E1Z8J0 (HAD) in biotin metabolism is highlighted in the KEGG pathway. Regulation of biotin metabolism is correlated with lipid accumulation in C. reinhardtii 11 . The significant upsurge in the fatty acid accumulation during N-starvation may arise either due to (a) the accelerated partitioning of new-photosynthetically fixed carbon through glycolysis, which produces pyruvate from glucose while generating the high-energy compounds ATP and NADH 69 , or (b) the recycling of carbon-molecules into the precursors for lipid biosynthesis. In vascular plants and algae, glycolysis occurs in both plastids and the cytosol. However, in the present study, most of the differentially expressed glycolysis-associated proteins were plastid-bound, except for TPI and ENL (present in the cytosol) ( Fig. 3 ). This suggests that plastids possess the enzymatic machinery of the payoff phase of the glycolysis pathway (from G3P to pyruvate) along with plastidic ALDO that converts F1, 6P to G3P and thus the plastid glycolysis pathway generates pyruvate in N-starved condition, which regulate the elongation of both long-chain saturated and unsaturated fatty acids. This observation is in sync with the transcriptomics studies on N-starved Nannochloropsis oceanica IMET 1 by Li et al . 68 . Interestingly, we observed perpetual up-regulation of E1ZIL0 (KEGG: acyl-CoA dehydrogenase activity), a component of complex I of oxidative phosphorylation which suggests that knock out of this enzyme could be detrimental to the organism 70 . E1ZIL0 is also involved in β-oxidation pathway; transcript analysis of N. oceanica IMET 1 revealed concomitant up-regulation of several components of β-oxidation pathways in response to N-deprivation 68 . Besides, carbon skeletons to be integrated into the neutral lipids may be salvaged from the degradation of membrane-bound glycerolipids via β-oxidation, suggesting that these enzymes could be sensible targets for maximizing neutral lipid biosynthesis. Knock-out of several lipases involved in β-oxidation pathway has been reported to result in synchronized growth and lipid accumulation 71 . Present study thus highlights the fact that up-regulation in TCA cycle coupled with increased fatty acid degradation in the mitochondria may enhance recycling of carbon skeleton for neutral lipid accumulation under N-starvation. Mitochondria thus serve as an auxiliary organelle for bulk fatty acid biosynthesis, in a fashion similar to vascular plants 72 . Our data correlates the previously reported transcriptomics studies with the proteome regulation during N-starvation. Notably, ACCase, a key enzyme catalyzing the irreversible step in fatty acid biosynthesis is not expressed differentially in our study. This may be due to the iTRAQ method employed for the comparative proteomics analysis, which apply differences throughout the measurements 73 74 75 . Suggesting, the difference in the expression level of ACCase in the test and control sample is not significant. Findings from our proteomics study and transcript analysis by Li et al . 68 well explains the failed attempts to enhance lipid content in diatom Cyclotella cryptica and Navicula saprophila 76 77 and plants 78 by overexpressing ACCase. Unfortunately, even up to two to three-fold elevated ACCase activity in the transformed algae did not led to any enhancement of lipid production 79 . Proteins in handling cellular stress N-starvation is characterized by oxidative stress in green alga Chlorella sorokiniana C3 and Scenedesmus sp., and by oxidative stress induced lipid accumulation in Dunaliella salina . In the present study, over-production of three stress-responsive enzymes namely SOD, E1ZG17 (KEGG: thioredoxin peroxiredoxin activity) and E1ZCK9 (KEGG: Chloroplastic lipocalin) are involved in the oxidative stress alleviation of FC2 during N-starvation. SOD (E1Z580) catalyzes dismutation of superoxide (O 2− ) into molecular oxygen and hydrogen peroxide, thus contributing to the tolerance towards ROS damage caused by N-starvation conditions. E1ZG17 has a proven anti-oxidant role. Thioredoxin peroxidases from Synechocystis sp. PCC 6803 is capable of reducing H 2 O 2 and its activities are coupled to the photosynthetic electron transport system 80 . The homolog of E1ZCK9 in Arabidopsis was reported to be accumulated during stress-conditions and has an evident role in the protection of thylakoidal membrane lipids against ROS 81 . Besides, the involvement of photorespiration to combat the redox stress during N-starvation condition cannot be ruled out. Correspondingly in the present study, four of the enzymes involved in photorespiration pathways namely; Serine hydroxymethyltransferase (SHMT), RuBisCO, Fd-GOGAT and Serine-glyoxylate aminotransferase showed up-regulation with the progression of N-starvation duration. SHMT has been known previously for armoring abiotic stress-triggered cell damage in Arabidopsis 82 . Likewise, up-regulation of SHMT during later stages of N-starvation may aid in balancing the cellular redox in N-starved FC2 cells. SHMT reversibly catalyze glycine to serine, which in turn, serves as a precursor for cysteine biosynthesis via cysteine synthase 83 . Cysteine synthase levels were expressed co-ordinately to SHMT, affirming that the cysteine levels are not a limiting factor for glutathione synthesis activity (component of redox homeostasis and detoxification machinery) under N-starved conditions in FC2. On the other hand, cysteine is a limiting factor for glutathione synthesis under N+ conditions. Additionally, enhanced accumulation of Octanoyltransferase, involved in the biosynthesis of lipoic acid has a prominent role in ROS scavenging. Therefore, we hypothesize that the levels of SOD, thioredoxin peroxiredoxin activity, chloroplastic lipocalin, SHMT, cysteine synthase, and octanoyltransferase play a critical role in maintaining cellular homeostasis during N-starvation. To the best of our knowledge, this is the pioneering systemic study of its kind, where the temporal proteomics analysis of a novel green oleaginous algae Chlorella sp. FC2 IITG at lipid induction phase is performed. Overall, the present knowledge laid the background of the post-transcriptional metabolic networks involved in N-starvation linked lipid induction in microalgae. Many of the proteins viz. SBP, PRK (involved in photosynthetic carbon fixation) identified in our study may serve as potential targets for strain improvement (i.e. synchronous growth and lipid accumulation), and pave way for economically viable and sustainable algal-based biofuel. PRK is the rate limiting step in Calvin cycle and down-regulation of this enzyme reduces the regeneration of RuBP, suggesting that enhancing its expression during stress may improve the photosynthetic yields 84 . Interestingly, increased levels of β-oxidation pathway enzymes may function in either way; which may aid in carbon-recycling for lipid biosynthesis or be involved in catabolism of neutral lipids. Thus, selective knock-down of the enzymes may synchronize biomass and lipid accumulation 71 . The impact of SBP in enhancing the plant growth has already been reported by several researchers 85 86 . SBP activity is modulated by the environmental factors 87 88 89 90 . The findings is well justified by the fact that transgenic rice over-expressing SBP did not show any significant change in biomass when grown at ambient conditions while enhanced biomass and photosynthesis was recorded when grown under abiotic stress conditions, particularly high salt 89 and temperature stresses 90 . This is the first-ever report of the MRM-based targeted validation at proteome-level in any algal species, and may be expanded to other microalgae. Although the shotgun data is validated by the western blotting and MRM assays, a more direct proof concerning the role of these proteins towards N-starvation induced lipid accumulation could be reinforced by genetic manipulation of algae for further strain improvement."
} | 5,795 |
17431711 | PMC1915594 | pmc | 5,851 | {
"abstract": "This study provides data on the diversities of bacterial and archaeal communities in an active methane seep at the Kazan mud volcano in the deep Eastern Mediterranean sea. Layers of varying depths in the Kazan sediments were investigated in terms of (1) chemical parameters and (2) DNA-based microbial population structures. The latter was accomplished by analyzing the sequences of directly amplified 16S rRNA genes, resulting in the phylogenetic analysis of the prokaryotic communities. Sequences of organisms potentially associated with processes such as anaerobic methane oxidation and sulfate reduction were thus identified. Overall, the sediment layers revealed the presence of sequences of quite diverse bacterial and archaeal communities, which varied considerably with depth. Dominant types revealed in these communities are known as key organisms involved in the following processes: (1) anaerobic methane oxidation and sulfate reduction, (2) sulfide oxidation, and (3) a range of (aerobic) heterotrophic processes. In the communities in the lowest sediment layer sampled (22–34 cm), sulfate-reducing bacteria and archaea of the ANME-2 cluster (likely involved in anaerobic methane oxidation) were prevalent, whereas heterotrophic organisms abounded in the top sediment layer (0–6 cm). Communities in the middle layer (6–22 cm) contained organisms that could be linked to either of the aforementioned processes. We discuss how these phylogeny (sequence)-based findings can support the ongoing molecular work aimed at unraveling both the functioning and the functional diversities of the communities under study.",
"introduction": "Introduction In the past decade, domelike structures called “mud volcanoes” have been discovered in several deep-sea sites, including the Mediterranean Sea [ 32 ]. Mud volcanoes emit large quantities of methane (an important greeenhouse gas) and hydrogen sulfide, yielding marine sediment environments—also called cold seeps—that can be highly enriched in these compounds [ 7 , 35 ]. For instance, the Kazan mud volcano sediment is known to be rich in methane and to contain reduced as well as oxidized forms of sulfur [ 54 ]. Life at such mud volcanoes (cold seeps) is probably largely sustained by primary producers that use methane and hydrogen sulfide as energy sources. This is reminiscent of hydrothermal vents [ 8 , 19 ], where life is primarily sustained by the microbial oxidation of sulfide [ 3 , 27 , 47 ]. Stable carbon isotope measurements of cold seep sediments and carbonate crusts have indicated that the anaerobic oxidation of methane (AOM), rather than of sulfide, is the basis of chemoautotrophy in these deep-sea ecosystems [ 1 , 44 ]. Specific methanotrophic archaeal groups, denominated ANME-1 and ANME-2, have been implicated as the main constituents of the microbial communities involved in AOM [ 6 , 40 , 49 ]. These organisms are either distantly (ANME-1) or closely (ANME-2) related to the (methanogenic) Methanosarcinales group. However, AOM by other archaea, e.g., the novel ANME-3 group, cannot been excluded [ 20 , 30 ]. The ANME-1 and ANME-2 group archaea consume methane in a close relationship with sulfate-reducing bacteria [ 5 , 6 , 37 ]. To date, these organisms have not been obtained in pure culture, and hence molecular approaches are indicated to assess their diversity, prevalence, and potential function. The occurrence of AOM in the sediments of the Kazan mud volcano has been indicated [ 54 ] by data on membrane lipids of the local microbial communities, but DNA-based studies have not yet been performed. In this work we analyze the diversity and nature of the bacterial and archaeal communities present in different Kazan sediment layers by using direct DNA-based molecular analyses. The 16S ribosomal RNA (rRNA) gene was employed as a marker for diversity. The microbial community structures found are evaluated against the chemical background of the system, in order to estimate the functional dominance of microbial AOM in the sediment layers. From this analysis, the validity of inferring ecosystem function purely from clonal phylogeny will be discussed and possible avenues toward a more sound, ecogenomics-based approach will be put forward.",
"discussion": "Discussion What Do the Data Tell Us? Linking Chemical Environment to Phylogeny-Based Microbial Community Composition We present a study based on directly obtained clone libraries of the microbial communities found at various sediment depths in the deep-sea cold seep Kazan. Much like other similar habitats [ 6 , 23 , 30 , 36 , 40 , 51 ], the Kazan sediment can be characterized by counterposed gradients of several chemical compounds, under methane and sulfate, as drivers of microbial community structures and processes [ 54 ]. We refer to this publication for an extensive description of the dynamics of mud volcano chemistry. Specifically, the porewater analysis for the Kazan sediments showed clear changes in the concentrations of ammonia, nitrate, nitrite, sulfate, sulfide, and methane (Table 1 ), similar to those found in previous studies in marine sediments [ 46 ]. The pH values measured may have been affected by degassing upon sample retrieval; the differential values indicate higher CO 2 levels in the lower layers. We surmised that, in microbial terms, the sediment actually represented a large conglomerate of niches that are determined by the interactions of the varying gradients over sediment length. Considering that all of the aforementioned chemical factors may drive the local microbial communities, it is difficult to ascertain which drivers are the most dominant. However, we hypothesized that the presence of the energy sources methane, sulfide, and (possibly) organic matter, in conjunction with that of the electron acceptors molecular oxygen, nitrate, and sulfate (and less oxidized forms of the latter two compounds), would be the primary drivers of the communities. Hence, we assessed (1) whether the microbial diversity per layer would actually report on the presence of such major driving forces, and (2) whether evidence for the presence of specific organisms potentially driven by a process such as AOM could be found. We will further bring forward a strategy to advance the phylogeny-based diversity versus function issue in this structured habitat. The 16S rRNA gene sequence data generated in this study provided an overall picture of clear heterogeneity between the communities in the different sediment layers. One of the possible caveats in such 16S rRNA sequence-based studies is undersampling, i.e., the recovery and analysis of simply too few sequences to adequately describe the microbial community in its major facets. Recent statistical studies that addressed this issue for soil are actually ominous, in that unrealistically high numbers of sequences seem to be required to adequately sample a soil community [ 10 , 11 ]. It is obvious that the habitat sampled by our team, deep-sea sediment, differs from soil in the extent and nature of its microbial diversity, which were actually unknown at the onset of this study. Hence, any prior estimation of the number of sequences required to obtain reasonable coverage of the total diversity could not be sensibly made. On the basis of the clone number analysed, the data seem to tell us that, grossly speaking, only dominant members of the microbial communities in the different sediment layers were sampled. On this basis, a picture of complex communities that are definitely unique per layer (although the full extent of diversity is as yet invisible from the data obtained) emerges. However, within these—still fragmentary—data, we can clearly discern the confines of the functional capabilities of such communities. The main piece of evidence (Table 2 )—i.e., the finding that sequences related to those of ANME-2-type archaea and to δ-proteobacterial sequences were abundant in Kazan-3 and to a lesser extent in Kazan-2—points in the direction of the AOM process taking place in these layers. This supports the conclusion that the local microbial communities dominated by ANME-2-type archaea and δ-proteobacterial sulfate reducers related to Desulfosarcina sp. are, to a large extent, driven by AOM. We tentatively link the lack of evidence for the presence of these organisms in Kazan-1 to them being less dominant in the light of the strong dominance of heterotrophs in this layer. This was consistent with the absence of sulfide from this layer. Other studies also indicated that the ANME-2 group archaea, next to those of ANME-1, are prominent parts of communities involved in AOM [ 6 , 23 , 40 , 51 ]. On the other hand, the absence from, or low abundance of ANME-1 in, our samples was indicated by the absence of characteristic membrane lipids [ 53 , 55 ]. A recent comparison between AOM communities from different “methane-rich” deep-sea areas also showed differences in dominance of ANME1 or ANME2 [ 30 ]. These areas showed chemical conditions favorable for AOM (i.e., sulfate and methane concentrations of up to 2 and 60 mM, respectively in the zone of AOM) and a relatively low diversity in archaeal ANME phylotypes, which is consistent with our findings. ANME-2-type archaea were also abundant in Kazan-2, and evidence for the occurrence of δ- Proteobacteria capable of sulfate reduction was also found, albeit at low abundance. Direct microscopical observation of the sediment material revealed the presence of microbial aggregates that resemble those implicated in AOM [ 6 ] in these sediments (not shown, shown in an earlier work [ 54 ]). The depth profiles of sulfate, sulfide, methane, and dissolved inorganic carbon (DIC) reflect the environmental conditions at the time of sampling, whereas by nature the microbial community data based on 16S rRNA gene sequences provide an integrated view of historical events [ 54 ]. Moreover, the data on salinity should also be interpreted in a similar manner. One could speculate that the differences observed between Kazan-1 and Kazan-2 on one hand, and Kazan-3 on the other, which might be indicative of recent gas hydrate dissolution, might have selected for specific microbial communities. However, direct evidence for this contention is lacking in this study. The sediment samples used in this study integrate depth intervals of 6, 16, and 12 cm, whereas the porewater chemical data had a resolution of 1 cm. As a consequence, the microbial communities as determined by the DNA analysis provide a more integrated view (and therefore less specific per smaller unit volume) than the porewater data. Support from Lipid Biomarker Data Lipid biomarker data were used as reference data that reflect functionally important microorganisms [ 54 ]. The summed concentrations of archaeol sn -3-hydroxyarchaeol and five unsaturated (PMI:x) compounds were used to indicate AOM archaea [ 53 , 54 ]. Nonisoprenoidal dialkylethers were used as specific indicators of sulfate-reducing bacteria [ 42 ]. Isotopically depleted diploptene/diplopterol and bishomohopanol were used as lipids indicating aerobic methanotrophs or methylotrophs [ 16 , 44 ] and aerobic sulfide-oxidizing prokaryotes [ 43 , 54 ], respectively. These lipids were set in relation to the amount of total lipids extracted from the sediment layers studied to illustrate the relative importance of the organisms presumably involved in the metabolic processes identified. The abundance of characteristic bacterial and archaeal membrane lipids in the top 30 cm of the sediments of the Kazan mud volcano are presented in Table 1 . Lipids assigned to archaea involved in AOM (i.e., archaeal, hydroxyarchaeol, and PMI:x) showed a strong increase in abundance with depth. A maximum abundance of this group was observed in Kazan-2. Previous studies showed strongly depleted δ 13 C carbon isotopic values of these compounds, which were consistent with values for methane-derived carbon in AOM environments [ 54 ]. Lipids presumed to derive from aerobic methane oxidizing (or methylotrophic) bacteria were found in layers Kazan-1 and Kazan-2, with decreasing relative abundance downcore (Table 4 ). β,β-Bishomohopanol, tentatively attributed to sulfide-oxidizing bacteria [ 55 ], was most abundant in the top layer Kazan-1, but it was also present in layer Kazan-2. The carbon isotopic values of these compounds showed δ 13 C values between −46‰ and −53‰, which is consistent with values for chemotrophic processes [ 54 ]. Most characteristic bacterial lipids identified in the three sediment layers were indicative for sulfate-reducing bacteria (e.g., dialkyl ethers). These compounds were detected in the lowest sediment layers, be it that they were most abundant in Kazan-2. We are puzzled by the merely partial support for these organisms from the phylogeny-based data (Table 2 ); however, it is known that PCR-based detection of specific sequences from natural samples can be hampered. Nevertheless, their occurrence is likely, as previously reported carbon isotope measurements from the same sediment showed strongly negative δ 13 C values of DIC and lipids derived from AOM archaea and sulfate-reducing bacteria [ 22 , 54 ]. Linking Phylogeny to Function—Caveats and Approaches The high diversities and low similarities of the phylotypes in the microbial communities between the three Kazan sediment layers may reflect a range of different metabolic processes taking place in these layers resulting from, as well as yielding, different habitat chemistries. However, this study, like virtually all other studies on microbial communities, only assessed a subset of the total microorganisms present. Furthermore, it has solely relied on the commonly used phylogenetic marker, the 16S ribosomal RNA gene sequence, to unravel the microbial communities. Use of this marker to indicate function assumes that functional properties are conserved among phylogenetically related populations, and that the function of “novel” organisms can be inferred by comparison with species that have been previously cultivated and characterized [ 21 , 50 ]. It is known that there are pitfalls in these assumptions, and a cautious approach is therefore indicated [ 9 , 21 , 26 , 45 , 50 ]. For instance, organisms occurring in the different niches in the sediment layers may have diversified at either the 16S rRNA gene or the functional gene levels, processes which both disentangle the link between phylogeny and function [ 17 ]. In addition, lateral gene transfer among prokaryotes is a significant force in microbial evolution, resulting in metabolic functions that show a paraphyletic distribution [ 48 ]. This has created uncertainty about the interpretation of taxonomies based on gene sequences in relation to phenetic taxonomy [ 17 , 48 ]. One way to resolve the apparent riddle posed by the large diversity of phylogenetic types in the different sediment layers and the difficulty of firmly linking such types to function would obviously be to take a metagenomics approach that is possibly coupled to either stable isotope probing or DNA-based preenrichment methodology. These strategies strongly depend on the production of sufficient quantities of highly pure microbial community DNA clonable in, e.g., fosmid vectors. The 16S rRNA-based sequences found to be most dominant in this study, e.g., the ANME-2 related sequences, might be useful as probes that can either “fish out” the underlying organisms from a sediment DNA pool or detect these at the cell level. Once DNA has been obtained (enriched), one might apply metagenomic cloning and sequencing to directly investigate the genetic link between phylogeny and potential function (on the basis of the sequences of the genes involved) and to ascertain whether the underlying organisms are indeed functionally dominant. The genetic information thus unlocked may then yield suitable probes or primer systems applicable in mRNA-based measurements. Using such approaches, a direct link can theoretically be made between the presence and the activities of the populations that play key roles in deep-sea sediments. To accomplish this and at the theoretical level, the identities and activities of the numerically dominant microorganisms in cold seeps should be further examined with respect to their importance in the consumption of methane or sulfide. An emphasis should be placed on the in situ detection of numerically abundant populations and their respective activities, for instance, via experiments in which the metabolism of labeled substrates (either 13 C- or 14 C-labeled) is combined with fluorescent in situ hybridization combined with stable carbon isotope measurement (FISH-SIMS) or substrate-tracking autoradiography fluorescent in situ hybridization (STAR-FISH). These activity measurements could be combined with molecular detection methods such as “real-time” reverse transcription-PCR (RT-PCR) with sediment 16S rRNA as a template, DNA microarrays for functional gene analysis, and FISH to identify the abundance, activity, and viability of microbial populations."
} | 4,275 |
32596121 | PMC7312437 | pmc | 5,852 | {
"abstract": "Abstract Wake‐up circuits in smart microsystems make huge contributions to energy conservation of electronic networks in unmanned areas, which still require higher pressure‐triggering sensitivity and lower power consumption. In this work, a bionic triboelectric nanogenerator (bTENG) is developed to serve as a self‐powered motion sensor in the wake‐up circuit, which captures slight mechanical disturbances and overcomes the drawback of conventional self‐powered motion sensors in the wake‐up circuit that the circuit can only be triggered when a considerable pressure is applied on the sensor. The bTENG mimics the structure of plants and the addition of the leaf‐shaped tentacle structures can increase the electrical outputs by four times, which largely extends the detection range of the wake‐up circuit. The bTENG can detect both noncontact and contact mechanical disturbances; and voltages generated from both situations can trigger the wake‐up system. Moreover, the specially designed circuit that is compatible with the bTENG can help more accurately control the wake‐up system and prolong the battery life of the electronic networks to 12.4 times. An intrusion detection system is established in the wake‐up circuit to distinguish human motion and judge the scene. This work opens new horizons for wake‐up technologies, and provides new routes for persistent sensing.",
"conclusion": "3 Conclusions In summary, a bTENG that mimics the plant with lotus‐leaf‐like nanostructures and leaf‐shaped tentacles is developed to serve as a self‐powered motion sensor in the wake‐up circuit with the property of capturing slight external mechanical disturbances. The bTENG with leaf‐shaped tentacle structures on the base can increase the voltage output by four times compared with the TENG with only the base. The systematic optimization of tentacle length, tentacle number, silicone rubber thickness, and metal wire diameter is also investigated to prove the superiority of the leaf‐shaped tentacle. The wake‐up system based on the bTENG has an increased detection range towards slight external mechanical disturbances and extended distance range, which can work in both the contact and noncontact modes. A wake‐up circuit is specially designed to be compatible with the bTENG, which integrates a switch component that can more accurately control the wake‐up system. Moreover, an intrusion detection system based on audio recognition of motion has also been established to judge the scene and distinguish human motion. Our work has the characteristics of low quiescent power consumption (125 nA theoretical quiescent current consumption), large wake‐up range that can be awakened by human striding over or slight touch, low cost, and environmental concealment. This work overcomes a bottleneck problem of the power limitations of persistent sensing and provides a new wake‐up method with low quiescent power consumption, which can be widely applied in the areas such as Internet of things and remote environmental monitoring.",
"introduction": "1 Introduction With the advent of the information age, the smart microsystem, which is a combination of micro‐electro‐mechanical system and microelectronics technology, has become a global research hotspot. It integrates sensors, actuators, signal collectors, data processors, and control circuits for applications in unattended security, infrastructure monitoring, and the Internet of things. To enhance the service life of the sensor network in the area where the battery cannot be replaced, the defense advanced research projects agency proposed the concept of “Near Zero Power RF and Sensor Operations (N‐ZERO)” that is, wake‐up systems sense the environment 100% of the time with ultralow quiescent power consumption and use the energy in signals to activate an internal signal processing circuit. [ \n \n 1 \n , \n 2 \n \n ] However, the problems faced in the research are as follows. First of all, since most of the common sensors require power supplies, it is a great challenge to reduce quiescent power consumption. Second, the current motion sensors of wake‐up circuits generally require large amplitude vibrations in order to generate large enough voltages to wake up the circuit. Third, the fabrication processes of the existing sensors used in wake‐up systems are generally complex and expensive. To further reduce the quiescent power consumption, one possible solution is to replace the power‐consuming sensors with self‐powered sensors that can work without system power supply; and to increase the wake‐up system's wake‐up range and sensitivity towards slight mechanical action, new sensors that are capable of generating large enough voltages to trigger the wake‐up circuit under slight touch or even no direct touch need to be developed. In addition, sensors with simple and scalable fabrication processes are desirable, from the perspectives of economy and large‐scale production. At present, there are self‐powered sensors that have been used in the wake‐up circuit for detection of mechanical disturbance, which are mostly piezoelectric nanogenerators (PENGs). [ \n \n 3 \n , \n 4 \n , \n 5 \n \n ] However, the PENGs, which work based on the piezoelectric effect, have low output power density and strict applied conditions. The PENGs can only generate electrical outputs when an applied pressure induces strain inside the piezoelectric material, and the small voltage generated by a slight touch cannot easily overcome the high switching threshold in the wake‐up circuit. These characteristics impose limits on the sensing ability of slight touch and wake‐up range of the wake‐up system. Therefore, low power‐consumption wake‐up systems with self‐powered motion sensors that can sense weak mechanical disturbances and have long sensing distances are in urgent demand. The triboelectric nanogenerator (TENG), which is based on the conjunction of contact electrification and electrostatic induction, [ \n \n 6 \n , \n 7 \n , \n 8 \n \n ] has attracted intense attention since its invention. [ \n \n 9 \n , \n 10 \n , \n 11 \n , \n 12 \n , \n 13 \n , \n 14 \n , \n 15 \n , \n 16 \n , \n 17 \n , \n 18 \n , \n 19 \n , \n 20 \n , \n 21 \n , \n 22 \n , \n 23 \n , \n 24 \n \n ] Since the TENG has the property of zero power consumption and can output high voltages even when triggered by noncontact mechanical movements, the TENG is a possible solution to meet the demands of self‐powered sensors for wake‐up systems that may enable the system to wake up under slight touch and enlarge the wake‐up range. In addition, the TENG has the advantages including low cost, lightweight property and simple fabrication. In this work, a plant‐shaped bionic TENG with no power consumption is developed in the wake‐up system, which can trigger the wake‐up system under slight touch or even no touch and largely extends the battery life of the system. Compared with the PENGs that are commonly used in the wake‐up circuits, the bTENG is superior for slight touch sensing. Moreover, the addition of leaf‐shaped tentacles and nanostructures in the bTENG increases the output voltage by four times when sensing slight touch, which improves the detection range of the wake‐up system in the process of environmental monitoring. Also, a wake‐up circuit is specially designed to be compatible with the bTENG and achieve ultralow quiescent power consumption. The wake‐up system with the bTENG can extend the battery life of the electronic network to 12.4 times. To highlight the characteristics of intelligence in our system, a computer program is elaborated to distinguish human motion and judge the wake‐up scene according to the audio signal recorded by the microcontroller after awaking. This work endows the TENG with a new role as a self‐powered sensor for slight touch in the wake‐up circuit and opens up new ideas for the design of wake‐up systems with ultralow quiescent power consumption.",
"discussion": "2 Results and Discussion 2.1 Overview of the bTENG and Wake‐Up System The typical structure of the bTENG is illustrated in Figure \n 1 a , which consists of a base plane and plenty of flat leaf‐shaped tentacles. The base is composed of an internal metal plate as the electrode and an external silicone rubber layer as the dielectric layer. Plenty of flat leaf‐shaped tentacles are attached to the internal metal plate, which enlarge the contact area in three‐dimensional space and lead to higher electrical outputs. The leaf‐shaped tentacle is composed of an internal vein‐like metal wire and an external leaf‐shaped silicone rubber layer. The vein‐like wires endow the bTENG with desirable mechanical properties such as high mechanical strength, hardness, and plasticity. Figure 1 Overview of the bTENG and wake‐up system. a) Schematic diagram showing the structure of the plant‐shaped bTENG. b) SEM images showing the micro/nanostructures on the bTENG leaves’ surface. c) Schematic illustrations of the operation mechanism for the TENG. d) Schematic diagrams showing the bTENG as a self‐powered sensor for wake‐up circuits and the photographs of the bTENG. e) Schematic diagrams of the bTENGs in unattended security that will activate the alarm system when intruders enter the core area. The micro/nanostructures imitating the nanostructures of lotic leaves (Figure 1b ) are created on the silicone rubber surface to increase the contact surface area using a modeling method. Note that the detailed fabrication process of the surface nanostructures can be found in the Experimental Section and Figure S1, Supporting Information. The silicone rubber usually becomes negatively charged after contact with another material because of its strong ability to attract electrons and the triboelectric charges on the silicone rubber's surface can be maintained for a long period of time. [ \n \n 25 \n , \n 26 \n \n ] The periodical contact and separation between the single‐electrode‐mode bTENG and the approaching object (feather, fur, etc.) can causes electrical potential variation between the copper electrode and the ground in the open‐circuit condition (Figure 1c ). [ \n \n 20 \n , \n 27 \n \n ] There will also be generated electrical outputs when the moving object do not directly touch the bTENG because of the electrostatic effect. This bTENG is a trump card for capturing slight external mechanical disturbances with the characteristic of no power consumption. Based on these features, the bTENG is applied to the wake‐up system suitable for the bTENG innovatively as a slight touch sensor, which is the core technology to monitor unmanned areas. When an intruder passing by, the system can wake up, process the collected sound, and judge the safety of the scene (Figure 1d ). Besides, the bTENG's unique structure enhances its concealment in natural environment and makes it more useful in myriad unmanned monitoring fields such as automated factory detections, intelligent home systems, and border inspections (Figure 1e ). 2.2 Optimized Structure Design of the bTENG The TENG exhibits excellent performance for the detection of noncontact and slight‐touch external mechanical disturbances. Note that the term slight touch discussed in this article is that the two surfaces come into contact and the applied pressure during contact is near zero. In the experiments, we divide the external mechanical motion into two types: noncontact motion and contact motion that exerts pressure during contact. For the experimental setup in the testing of the noncontact motion, the TENG was fixed in a slot and was approached by a polyester that was attached to a working plate and driven by a stepping motor at a uniform speed ( Figure \n 2 a ; Movie S1, Supporting Information). The relationship between the nearest distance from the TENG and the TENG's output voltage is recorded (Figure 2b ). It is found that the output voltage of the bTENG decreases with the increasing nearest distance of the moving object from the bTENG. When the nearest distance is 4 mm, the TENG can produce a voltage of up to 20 V, which demonstrates TENGs’ sensing ability of noncontact motion. Figure 2 Detection of slight mechanical disturbances and structure optimization of the bTENG. a) Photograph showing the experimental setup for the noncontact experiment. b) The relationship between the output voltage of the TENG and the nearest noncontact distance. The inset shows the output voltage of the TENG when the nearest noncontact distance is 4 mm. c) Photograph is showing the experimental setup for the contact experiment. d) The relationship between the output voltage of the TENG and the exerted pressure during contact. The inset shows the output voltage of the TENG when exerted by a pressure of 95 Pa. e) Photograph showing the experimental setup for the feather‐touch experiment. f) The output voltage of the TENG in the feather‐touch experiment. g) Schematic diagrams of the TENG with/without tentacles. h) Effects of the TENG's tentacles structure on output voltage. i) The relationship between the output voltage of the TENG and the number of tentacles. j) The relationship between the output voltage and the length of tentacles. k) The relationship between the output voltage and the layer thickness of the external silicone rubber tentacles. l) The relationship between the output voltage and the radius of the internal wire tentacles. In the experiment of contact motion that exerts pressure during contact, we used a gauge to control the force applied on the sensors’ surface (Figure 2c ). The applied pressure is obtained by dividing the applied force by the contact area and the relationship between the output voltage and the pressure is recorded (Figure 2d ). In this case, the pressure sensitivity is 34.4 mV Pa –1 when the pressure is less than 160 Pa and the pressure sensitivity is 8.3 mV Pa –1 when the pressure is higher than 160 Pa (Figure S2, Supporting Information). The TENG produces continuous pulse voltage signals of 23 V when exerted by a pressure of 95 Pa. Moreover, it is found that when objects made of different materials exert pressure on the TENG, the maximum output voltages are all more than 6 V when the pressure is less than 700 Pa (Figure S3, Supporting Information). These two experiments demonstrate TENG's superiority as a self‐powered sensor for detecting noncontact and slight touch disturbances. To further confirm the TENG's superior performance in detecting slight touch, we compared the performance of the TENG with that of the PENG, which is the most commonly used sensor in low‐power wake‐up circuits at present. [ \n \n 28 \n , \n 29 \n , \n 30 \n \n ] It is found that the electrical output of the TENG is 20 times higher than that of the PENG in the slight‐touch experiment; and the PENG generates no electrical outputs in the noncontact experiment (Figure S4a,b, Supporting Information). Note that the detailed theoretical explanation for the TENG's superior performance to the PENG's in slight‐touch sensing can be found in Note S1, Figures S4c and S5, Supporting Information. A feather‐touch experiment is also conducted to demonstrate the TENG's advantage in detecting slight touch. When a feather wavered to the bTENG, the bTENG generates a voltage of 1.2 V (Figure 2e , f ; Movie S2, Supporting Information). However, the PENG generates no electrical outputs when a feather wavered to the PENG (Figure S4d, Supporting Information). It can be seen from the above experiments that TENGs are superior in slight touch detections for wake‐up systems. To further enhance the output voltage of the bTENG and thus extend the detection range towards slight mechanical disturbances of the wake‐up circuit, we add many blade‐shaped tentacles to the basic flat‐shaped bTENG (Figure 2g ). We first simplified the shape of the tentacles into cylindrical structures (Figure S6a, Supporting Information) to investigate the effect of the tentacle on the output voltage. bTENGs with and without a tentacle were placed on the experimental setup and approached by the same material, respectively. The output voltage signals show that the addition of just one tentacle structure can double the output voltage (Figure 2h ). The simulation results also prove the feasibility of the bTENG with the tentacle structure (Figure S6b,c, Supporting Information). Note that the cylindrical tentacle applied in the experiment is to investigate the general impact of the device's structure parameters on the electrical outputs; and the shape of the tentacle could also influence the electrical outputs, that is, larger surface area can accumulate more triboelectric charges and thereby results in higher electrical outputs. The effects of the number of the tentacle (Figure 2i ), length of the tentacle (Figure 2j ), layer thickness of the external silicone rubber (Figure 2k ), and radius of the internal wire (Figure 2l ) on the output voltage are studied systematically. The experiments indicate that the addition of more and longer tentacles with thicker external silicone rubber and finer internal wire can increase the output voltage of the bTENG. Note that the simulations about the effects of parameters of device structure on the electrical outputs have also been done, which have the same trend as the experimental results (Figure S7, Supporting Information). The simulation results also prove that the uniform distribution of tentacles is also conducive to enhance the output voltage (Figure S8, Supporting Information). Basic characteristics of the tentacles in the bTENG (large leaf surface area, long branch length, a large number of leaves, fine veins, uniform distribution) conform to the above requirements. The study on the shape of tentacles lays the foundation for our bionic design theory. At the same time, the bTENG has the advantages of simple fabrication process, low cost, high repeatability, and strong plasticity. The bTENGs also can imitate different species of plants according to various application scenarios. 2.3 Design of the Wake‐Up Circuit Since the bTENG has the property of zero power consumption and can output high voltages even when triggered by noncontact mechanical movements, it can be applied in the wake‐up system as a self‐powered sensor. A wake‐up circuit suitable for the bTENG that is composed of a switch component and a data processing microcontroller has been designed, which together with the bTENG form the wake‐up system ( Figure \n 3 a ). Considering that the bTENG has the characteristics of high voltage output and low current output, and that the output voltage of the TENG under load is impacted by the load impedance, a metal‐oxide‐semiconductor field‐effect transistor (MOSFET) is attached to the bTENG as a switch element (Figure 3b ), which has the characteristics of easy conduction, no restriction on the input current, and low power consumption. [ \n \n 31 \n , \n 32 \n \n ] Since the actual output of the bTENG in a circuit will be influenced by the load impedance, the wake‐up system can be more accurately controlled by connecting the bTENG to the gate electrode of the MOSFET with a known impedance. Figure 3 The design of wake‐up circuits using bTENGs as sensors. a) Schematic diagrams showing the wake‐up circuit's working principle. b) Circuit diagram of the switch element MOSFET connected to the bTENG. c) The output voltage of the switching element triggered by the bTENG. d) Comparison of the power consumptions of the wake‐up systems and nondormant systems. Two parameters should be paid attention to when choosing the MOSFET suitable for the bTENG: the threshold voltage and the parasitic capacitance. The input voltage controls the charge carrier channel between the source and drain electrodes in a MOSFET. When the input voltage is higher than the threshold voltage of the MOSFET, the switch will turn on and give a trigger voltage to the microcontroller to start the whole circuit, which indicates that we can control the wake‐up voltage of the wake‐up system by selecting a MOSFET with a certain threshold voltage. When the bTENG is connected with the MOSFET, the parasitic capacitance of the MOSFET represents the load impedance of the bTENG, which is the input capacitance minus the reverse transmission capacitance in the datasheet. To study how the parasitic capacitance of the MOSFET affects the wake‐up system, two types of MOSFET with different parasitic capacitances (PMZ600UNEL and IRF1404) are connected to the bTENG, respectively. It is found that when the open‐circuit voltages of the bTENG are the same, the smaller the MOSFET's parasitic capacitance, the larger the actual output voltage of the bTENG (Figure S9, Supporting Information). Therefore, for practical applications, the bTENG based wake‐up system can be more precisely controlled by adding a MOSFET with a certain threshold voltage and a certain parasitic capacitance according to practical requirements. In addition, when choosing a MOSFET for the wake‐up system, it should be kept in mind that the MOSFET needs to have the characteristic of low drain leakage current in order to reduce the power consumption of the system and high breakdown voltage in order to ensure the safety of the circuit. In this article, when the voltage signal generated by bTENG turns on the MOSFET, the MOSFET will send a pulse signal to the connected dormant microcontroller and wake up the CPU with short delay (<10 µs) to start the subsequent work. The output signal of the MOSFET with a low threshold voltage (0.45 V) when triggered by the bTENG is illustrated in Figure 3c . Note that the detailed theoretical foundation of the role that the MOSFET plays in the bTENG based wake‐up system can be found in Note S2, Supporting Information. To distinguish human motion and rule out false triggers, an intrusion detection system is also developed for the scenario that an alarm alert is released only when human intrusion happens. The intrusion detection system is established by analyzing the sound recorded by the peripheral recording device of the microcontroller when unexpected event happens. The specific implementation method is as follows. The triggering signal wakes up the microcontroller, and the microcontroller turns on the recorder to record the audio signal of the wake‐up scene. Then the sound signal is sent to the remote computer through peripheral Bluetooth. [ \n \n 33 \n , \n 34 \n \n ] The detailed explanation of the basic description of MSP430 can be found in Note S3, Supporting Information. Using MATLAB software, we write a program extracting the mel‐frequency cepstrum coefficient [ \n \n 35 \n , \n 36 \n \n ] (MFCC) of the recorded signal to facilitate the subsequent comparative analysis which greatly reduce data quantity by 2%. Methods of extracting the MFCC and the key code can be found in Note S4, Supporting Information. After learning 50 groups of characteristic parameters of environmental noise and ambient sounds (human footsteps) by support vector machine learning method, [ \n \n 37 \n , \n 38 \n \n ] which can be changed according to the factual requests, the program can compare the received recording with the previous learning result and judge the wake‐up scene (Note S5, Supporting Information). Note that the specific processes of judging scene based on the analysis of the audio signal are introduced in Figure S10, Supporting Information. When an intrusion occurs, the intrusion detection system will judge the scene: if the intrusion is not initiated by a person, the computer will display “SAFETY;” and if the intrusion is initiated by a person, the wake‐up system will be triggered and the computer will display an “ALARM” alert. 2.4 The Power Consumption of the Wake‐Up System Since the purpose of the N‐ZERO project is to reduce quiescent energy consumption and thereby increasing the battery life, the characteristic of low quiescent power consumption of our system is examined. The wake‐up system consists of a self‐powered bTENG, a switching MOSFET with a leakage current of only 25 nA, and a microcontroller MSP430 with quiescent current consumption of 100 nA in the dormant state according to their data sheets, which is much less than the current micro energy output power (mW level). [ \n \n 39 \n , \n 40 \n \n ] Compared with the nondormant system with the bTENG, the wake‐up system with the bTENG that is set to be woke up every 10 min consumes only one‐seventh of the voltage supply, after the same operation time of 140 min. When the wake‐up time interval is set to be 10 min, the voltage supply consumed by the wake‐up system with the bTENG is only half of that consumed by the wake‐up system with a commercial motion/vibration sensor, after the same operation time of 140 min. The wake‐up system with the bTENG has a battery life that is two times longer than the wake‐up system with a commercial motion/vibration sensor; and the battery life of the wake‐up system with the bTENG is 12.4 times longer than that of the nondormant system with a commercial motion/vibration sensor (Figure 3d ). These experiments demonstrate that the bTENG‐based wake‐up circuit has ultralow quiescent power consumption, which could meet the requirements of lower power consumption. 2.5 The Contact and Noncontact Wake‐Up Modes To demonstrate the practicability of the bTENG‐based wake‐up system, two types of bTENGs are fabricated to meet the requirements of different wake‐up scenarios: bTENGs shaped like shepherd's purse in the noncontact wake‐up mode and bTENGs shaped like philodendron in the contact wake‐up mode. In the noncontact mode, a bTENG with only the base structure and a bTENG shaped like shepherd's purse structured with a base structure of the same size and seven leaf‐shaped tentacles were fabricated for testing ( Figure \n 4 a ). The bTENG shaped like shepherd's purse can generate a V \n oc of ≈4.8 V when triggered by noncontact approaching/departing motion, which is 2.4 times higher than the V \n oc of the bTENG with only the base structure (Figure 4b ). In addition, the influence of the distance of the noncontact material from the bTENG on the output voltage is studied (Figure 4c ). The experimental results show that the bTENG can detect motion within 50 mm, and its output voltage can overcome the threshold voltage of the connected wake‐up circuit, which demonstrates that the bTENG extends the detection range of the wake‐up circuit. In the noncontact mode, the bTENG is connected to a MOSFET with a lower threshold voltage to make it easier to wake up the system. In order to trigger the wake‐up system, the actual output voltage of the bTENG applied on the MOSFET must be higher than the threshold voltage of the MOSFET that is 0.45 V in the noncontact mode. Figure 4 Output characteristics of the bTENG in the noncontact and contact wake‐up modes. a) Schematic diagram of the noncontact wake‐up mode. b) The output voltage of the plant‐shaped bTENG triggered by horizontal contact/separation motion in the noncontact wake‐up mode. c) The relationship between the output voltage and the distance of the noncontact material from the bTENG. d) Schematic diagram of the contact wake‐up mode. e) The output voltage of the plant‐shaped bTENG triggered by horizontal contact/separation motion in the contact wake‐up mode. f) The voltage generated by the bTENG when approached by the fur material. g) The voltage generated by the bTENG when approached by the polyester material. Since in some application scenarios, the intensity of the wake‐up motion needs to be increased because of the strong environmental interference. We design a taller bTENG to test the output signal in the contact mode. The bTENG shaped like philodendron with a base and four tentacles (Figure 4d ) generates a V \n OC of ≈8 V when triggered by slight touches, which is four times higher than the V \n OC of the bTENG with only the base structure (Figure 4e ). Both polyester and fur materials can produce high output voltage by slightly touching the bTENG shaped like philodendron (Figure 4f , \n g \n ). In the contact mode, the bTENG is connected to a MOSFET with a higher threshold voltage to avoid false triggering by strong environmental interference. In order to trigger the wake‐up system, the actual output voltage of the bTENG applied on the MOSFET must be higher than the threshold voltage of the MOSFET that is 2 V in the contact mode. The good performance of these two different work modes demonstrates that our bTENG is not limited by appearance and working patterns, which can be changed according to different scenarios. 2.6 Demonstrations of the Wake‐Up Modes and Judging the Scene Verification experiments have been demonstrated by utilizing this wake‐up system to sustainably monitor the safety of the environment, judge the wake‐up scene, and distinguish false triggers in the two types of wake‐up modes. In the experiments, human motion is set as the true trigger and motion of objects is set as the false trigger. In the contact wake‐up mode, when a person passed by the bTENG with contact, this action caused the warning light to illuminate and the remote computer raised alarm based on the analysis of audio signals ( Figure \n 5 a ; Movie S3, Supporting Information). When a plastic bag fell onto the bTENG, this action also caused the warning light to illuminate, but the remote computer concluded that the condition was safe based on the analysis of audio signals. This indicated that the system successfully woke up the system, judged the scene, and distinguished the intruder (Figure 5b ; Movie S4, Supporting Information). Similarly, in the noncontact wake‐up mode, an alarm was generated when a person passed by the bTENG shaped like shepherd's purse without contact, whereas a safety notice was displayed when a plastic bag moved above the bTENG to simulate the false trigger (Figure 5c , d ; Movies S5 and S6, Supporting Information). These experiments reveal that our wake‐up system can achieve self‐awakening and scene judgment in both contact and noncontact situations. Figure 5 Demonstration and verification of the wake‐up system. a) Photographs show that the intruder touched the bTENG, woke up the circuit, and the computer correctly judged the scene and raised alarm. b) Distinguishing false triggers by the wake‐up system in the contact mode. When a false trigger occurs, for example, a plastic bag contacts the bTENG, the scene is judged to be safe. c) Photographs show that the intruder passed by without touching the bTENG, woke up the circuit, and the computer correctly judged the scene. d) Distinguishing false triggers by the wake‐up system in the noncontact mode. When a false trigger occurs in the noncontact mode, for example, a plastic bag moves above the bTENG, the scene was judged and no alarm was generated."
} | 7,684 |
25373400 | PMC4221787 | pmc | 5,854 | {
"abstract": "As an important method for building blocks synthesis, whole cell biocatalysis is hindered by some shortcomings such as unpredictability of reactions, utilization of opportunistic pathogen, and side reactions. Due to its biological and extensively studied genetic background, Pseudomonas putida KT2440 is viewed as a promising host for construction of efficient biocatalysts. After analysis and reconstruction of the lactate utilization system in the P. putida strain, a novel biocatalyst that only exhibited NAD-independent d -lactate dehydrogenase activity was prepared and used in l -2-hydroxy-carboxylates production. Since the side reaction catalyzed by the NAD-independent l -lactate dehydrogenase was eliminated in whole cells of recombinant P. putida KT2440, two important l -2-hydroxy-carboxylates ( l -lactate and l -2-hydroxybutyrate) were produced in high yield and high optical purity by kinetic resolution of racemic 2-hydroxy carboxylic acids. The results highlight the promise in biocatalysis by the biotechnologically important organism P. putida KT2440 through genomic analysis and recombination.",
"discussion": "Discussion Optically active 2-hydroxy-carboxylates are important building blocks for glycols, halo esters, and epoxides compounds, which are important intermediates of pharmaceuticals 22 23 . Chemical processes for 2-hydroxy-carboxylates production result in a racemic mixture of both stereospecific forms. Many routes, such as high-performance ligand exchange chromatography 24 , enzymatic resolution 25 , and asymmetric hydrolysis 26 , have been developed for the resolution of the racemic 2-hydroxy-carboxylates. Due to its excellent stereo-selectivity, high product yield and environmental friendly process, biocatalytic oxidative resolution of the racemate has emerged as a desirable technique for the production of optically active 2-hydroxy-carboxylates 22 25 . Until now, all of the reported 2-hydroxy-carboxylates resolution processes utilized NAD-independent l -2-hydroxy-carboxylate dehydrogenases such as glycolate oxidase 25 , l -lactate oxidase 27 and membrane bound l- iLDH 28 29 as the biocatalysts, and could only produce d -2-hydroxy-carboxylates. Thus, there is a demand for searching of biocatalysts that could be utilized in the resolution of racemic 2-hydroxy-carboxylates to produce l -2-hydroxy-carboxylates. Although d- iLDH might catalyze the oxidation of d -2-hydroxy-carboxylates, the co-present l- iLDH would also catalyze the l -2-hydroxy-carboxylates into 2-oxo-carboxylates 30 31 32 . After analysis of the lactate utilization system through comparative genomics, the lldPDE operon was identified in P. putida KT2440. l- iLDH encoding gene lldD was disrupted using homologous recombination. As we expect, the resulting recombinant strain P. putida KTM exhibited only d- iLDH activity. The specific activity of d- iLDH in P. putida KTM was 65.0 ± 3.7 nmol min −1 mg −1 , which was higher than that of P. aeruginosa XMG (20 ± 4 nmol min −1 mg -1 ) 32 but lower than that of P. stutzeri SDM (132 ± 5 nmol min −1 mg −1 ) 20 . Similar to d- iLDH in other Pseudomonas strains such as P. stutzeri SDM 20 , d- iLDH in P. putida KTM also exhibited narrow substrate specificity, and only catalyzed the oxidation of d -lactate and d -2-hydroxybutyrate. Optically active l -lactate and l -2-hydroxybutyrate could be used in production of polylactate and poly(2-hydroxybutyrate), which can be utilized as a biodegradable material for biomedical, pharmaceutical, and environmental applications 33 34 . Then, whole-cells of P. putida KTM were used to catalyze the resolution of racemic 2-hydroxybutyrate and lactate. The d -lactate and d -2-hydroxybutyrate could be oxidized into pyruvate and 2-oxobutyrate, respectively. l -Lactate and l -2-hydroxybutyrate, which were not oxidized in the biocatalytic process, accumulated with enantiomeric excess higher than of 99% ( Fig. S2 and Fig. S3 ). In a previous study, l -iLDH in P. stutzeri SDM was rationally re-designed on the basis of sequence alignment and the active site structure of a homologous enzyme; a new biocatalyst with high catalytic efficiency toward an unnatural substrate ( l -mandelate) was successfully constructed 35 . Although the crystal structure of d- iLDH in E. coli was described 36 , the position of the active site is still unknown. On the other hand, there is only 28% sequence identity between d- iLDHs from P. putida KT2440 and E. coli . Thus, at this stage, we did not attempt to expand the application range of d -iLDH in P. putida KTM due to the difficulty in the rational re-design of this membrane bound enzyme. However, if its structure is clarified, reconstruction of d -iLDH to improve its activity towards other straight long aliphatic or aromatic 2-hydroxy-carboxylic acids and synthesis of other valuable l -2-hydroxy-carboxylates might be successful in the future. In summary, racemic 2-hydroxy carboxylic acids were first utilized in the production of l -2-hydroxy-carboxylates through enantioselective oxidation. A novel catalyst with lactate utilization system in P. putida KT2440 was reconstructed; it exhibited high bio-catalytic activities for production of l -2-hydroxy-carboxylates. The reconstruction of the lactate utilization system in P. putida KT2440 resulted in a novel catalyst that has exhibited high biocatalytic activities for l -2-hydroxy-carboxylates production. This study is a good example for the application of the GRAS P. putida KT2440 in the bio-catalysis through genomic analysis and recombination. Other applications of P. putida KT2440 would also be possible through further understanding and reconstruction of this biotechnologically important organism."
} | 1,447 |
36855625 | PMC9928405 | pmc | 5,855 | {
"abstract": "The lack of freshwater\nhas been threatening many people who are\nliving in Africa, the Middle East, and Oceania, while the discovery\nof freshwater harvesting technology is considered a promising solution.\nRecent advances in structured surface materials, metal–organic\nframeworks, hygroscopic inorganic compounds (and derivative materials),\nand functional hydrogels have demonstrated their potential as platform\ntechnologies for atmospheric water (i.e., supersaturated fog and unsaturated\nwater) harvesting due to their cheap price, zero second energy requirement,\nhigh water capture capacity, and easy installation and operation compared\nwith traditional water harvesting methods, such as long-distance water\ntransportation, seawater desalination, and electrical dew collection\ndevices in rural areas or individual-scale emergent usage. In this\ncontribution, we highlight recent developments in functional materials\nfor “passive” atmospheric water harvesting application,\nfocusing on the structure–property relationship (SPR) to illustrate\nthe transport mechanism of water capture and release. We also discuss\ntechnical challenges in the practical applications of the water harvesting\nmaterials, including low adaptability in a harsh environment, low\ncapacity under low humidity, self-desorption, and insufficient solar-thermal\nconversion. Finally, we provide insightful perspectives on the design\nand fabrication of atmospheric water harvesting materials.",
"conclusion": "5 Conclusions In this Review, recent advances in materials chemistry for air–water\nharvesting applications have been critically discussed. Compared with\ntraditional freshwater collection methods such as reverse osmosis\nin membrane seawater desalination, long-distance water delivery, and\npower-based active condensation devices, the new AWH materials do\nnot need external mechanical equipment and external energy sources.\nThe current high-efficiency AWH materials can produce about 6.5 kg\nof freshwater per day using 1 kg of hydrogel under optimal conditions,\nmeeting the daily drinking water requirement of three adults, with\nno additional energy consumption. This makes these materials extremely\ncompetitive in the field of freshwater harvesting. Due to the broad\nprospects of AWH materials, it is expected that a technology based\non them will play a major role in alleviating the shortage of freshwater\nresources in dry or water-deficient areas in the foreseeable future.",
"introduction": "1 Introduction Water is the basis for\nthe existence and continuation of all life\non Earth. For the human body, it not only is essential for maintaining\nthe electrolyte balance but also is the only carrier for the excretion\nof unwanted metabolites from the human body. The water resources on\nEarth are extremely abundant (ca. 1.46 × 10 16 cubic\nmeters). However, 99.97% of the water exists in the form of seawater\nor deep groundwater that is difficult to collect, and only less than\n0.03% can be easily used by humans. 1 − 3 This part of our water\nresources mainly includes surface freshwater, water vapor in the air,\nand shallow groundwater. Today, due to geographical and climate constraints,\nthe shortage of freshwater is of concern to 2.8 billion people in\n48 countries all over the world especially in Africa, the Middle East,\nand Oceania, and the affected population may rise to 4 billion based\non reasonable predictions. 4 , 5 Consequently, the discovery\nof next-generation freshwater harvesting technologies with low-cost,\nhigh-water adsorption capacity, and ease of installation and use is\nconsidered a promising solution to this global challenge and has attracted\nincreasing attention all over the world. There are three forms\nof water existing in the air: solid (i.e.,\nsnow, hail, frost, ice crystals etc.), liquid (i.e., rain, dew and\nfog), and gaseous (i.e., vapor, steam etc.). Humans have a long history\nof directly utilizing solid and liquid water, yet the capture of gaseous\nwater has been overlooked. Currently, due to the geographical and\ntechnological limitations and climate changes, water conservancy approaches\nin deserts or wastelands, including energy required “active”\ndew collection (e.g., electricity driven condenser), liquid water\ntransfer facilities (i.e., river diversion, long-distance water transportation,\ndam and reservoir construction), and seawater desalination systems\nusually require high investments, 6 , 7 high maintenance\ncosts 8 and/or high operating costs, 9 , 10 high environmental impacts, inflexible installation (fixed location),\nand time-consuming and seasonal intermittent water supply issues. 11 , 12 In recent years, “passive” water production technologies\nhave attracted considerable attention because they do not consume\nsecond energy. For instance, researchers have developed various solar\nvapor generators (SVGs) composed of hydrophilic matrix and photothermal\nmaterials. 13 , 14 The SVG devices can float on\nthe sea/water surface and utilize solar energy to produce clean water.\nLikewise, “passive” technologies for directly atmospheric\nwater harvesting using synthetic materials are considered as promising\nalternatives to conventional methods due to their low cost, low energy\nconsumption, ease of installation, and high harvesting performance\nwhen rapid water harvesting is required or electricity support is\ninsufficient. 15 Recently, chemically designed\natmospheric water harvesting (AWH) materials, including structured\nsurface materials, metal–organic frameworks (MOFs), hygroscopic\ninorganic compounds and derivatives (HICs and derivatives), and functional\nhydrogels have attracted increasing attention. Different from the\ntraditional water harvesting methods such as reverse osmosis desalination\nwhich uses secondary energy (e.g., electricity) or phase change condensation\nusing cold fluids, the water harvesting mechanism of chemically designed\nAWH materials is based on their physical and chemical properties,\nincluding Laplace pressure gradients, chemical modifications of polymer\nchains, tunable intermolecular space, and open metal sites. In this Review, we focus on the research output of passive technologies\nfor atmospheric (gaseous) water harvesting in the past decade and\nprovide comprehensive discussion on the state-of-the-art materials\nfrom the perspective of materials chemistry. According to the degree\nof water saturation, these materials are divided into two categories:\n(a) structured surface used for saturated atmospheric water collection\nand (b) MOFs, HICs and derivatives, ionic liquids (ILs), and functional\nhydrogels used for unsaturated vapor capture and water production.\nAt the end of the discussion mentioned above, future perspectives\nand potential optimization approaches are presented which can further\nimprove the water production performance so that it can meet energy\nsustainability and carbon neutral economy requirements."
} | 1,714 |
34335526 | PMC8317133 | pmc | 5,856 | {
"abstract": "Previous work demonstrated that microbial Fe(III)-reduction contributes to void formation, and potentially cave formation within Fe(III)-rich rocks, such as banded iron formation (BIF), iron ore and canga (a surficial duricrust), based on field observations and static batch cultures. Microbiological Fe(III) reduction is often limited when biogenic Fe(II) passivates further Fe(III) reduction, although subsurface groundwater flow and the export of biogenic Fe(II) could alleviate this passivation process, and thus accelerate cave formation. Given that static batch cultures are unlikely to reflect the dynamics of groundwater flow conditions in situ , we carried out comparative batch and column experiments to extend our understanding of the mass transport of iron and other solutes under flow conditions, and its effect on community structure dynamics and Fe(III)-reduction. A solution with chemistry approximating cave-associated porewater was amended with 5.0 mM lactate as a carbon source and added to columns packed with canga and inoculated with an assemblage of microorganisms associated with the interior of cave walls. Under anaerobic conditions, microbial Fe(III) reduction was enhanced in flow-through column incubations, compared to static batch incubations. During incubation, the microbial community profile in both batch culture and columns shifted from a Proteobacterial dominance to the Firmicutes, including Clostridiaceae, Peptococcaceae, and Veillonellaceae, the latter of which has not previously been shown to reduce Fe(III). The bacterial Fe(III) reduction altered the advective properties of canga-packed columns and enhanced permeability. Our results demonstrate that removing inhibitory Fe(II) via mimicking hydrologic flow of groundwater increases reduction rates and overall Fe-oxide dissolution, which in turn alters the hydrology of the Fe(III)-rich rocks. Our results also suggest that reductive weathering of Fe(III)-rich rocks such as canga, BIF, and iron ores may be more substantial than previously understood.",
"introduction": "Introduction The Southern Espinhaço Mountain Range (SE) of southeastern Brazil contains commercially important, high-grade iron ore hosted by the Serra da Serpentina Group, a stratigraphic unit which includes the iron-rich Serra do Sapo Formation ( Auler et al., 2019 ). These sedimentary units were formed by the precipitation of Fe(III) and Si phases from solution during the Proterozoic Eon ( Weber et al., 2006 ; Rosière et al., 2019 ; Silveira Braga et al., 2021 ). Iron ores can include magnetite (Fe 3 O 4 ), hematite (α-Fe 2 O 3 ), or a ferric oxyhydroxide like goethite (α-FeOOH) or limonite (FeO(OH)⋅n(H 2 O), and high-grade ore averages between 60 and 67% Fe ( Dorr, 1964 ; Beukes et al., 2003 ; Rolim et al., 2016 ). The SE and Quadrilátero Ferrífero (Iron Quadrangle; QF) located ∼150 km south of SE, contain abundant Fe(III)-rich minerals, which can be found in intact banded iron formation (BIF), Si-depleted ore, and canga rock ( Beukes et al., 2003 ; Souza et al., 2015 ; Auler et al., 2019 ). BIF contains alternating bands of quartz (SiO 2 ) and either magnetite or hematite that range from a few millimeters to a few centimeters in thickness, and averages between 15 and 38% iron ( Smith, 2015 ; Rolim et al., 2016 ). Canga is a brecciated duricrust containing clasts of iron oxide (usually BIF) with an iron-oxide cement matrix that averages between 57 and 62% iron ( Dorr, 1964 ; Spier et al., 2007 ; Gagen et al., 2019 ). Canga contains the most poorly crystalline Fe(III) of the three major phases (i.e. canga, BIF, and ore), with goethite the most prominent mineral phase ( Parker et al., 2013 ). Canga and BIF are generally considered highly resistant to both mechanical and chemical weathering at pH ≥ 3 ( Dorr, 1964 ; Johnson et al., 2012 ; Auler et al., 2014 ; Spier et al., 2018 ; Gagen et al., 2019 ), and canga covers the slopes and valleys of the SE region. Yet this area is also associated with hundreds of caves (iron formation caves; IFCs) that form mostly at the BIF/canga boundary ( Auler et al., 2019 ). The identification of these caves suggests that processes leading to Fe(III) weathering and removal increase porosity at the canga/BIF interface, despite the resistance of both types of rocks to dissolution. At circumneutral pH, Fe solubility can be enhanced by microbially mediated reductive dissolution of Fe(III) phases to relatively soluble Fe(II). This activity may facilitate the mass transport necessary for the increased porosity and the formation of the observed IFCs ( Parker et al., 2013 , 2018 ). In support of this hypothesis, while the walls of the IFCs are lined with a hard, oxidized layer of canga, the interior (approximately 3 cm behind) of the wall surface contains a soft, gooey, water-saturated material that contains abundant microbial cells ( Parker et al., 2018 ). Given the inhibition of Fe-reduction by oxygen, we wondered whether this material was involved in promoting Fe-reduction and increased porosity, leading to formation of the IFCs. Canga is a rather porous media, and active vertical percolation of water occurs during rainfall, the patterns of which can be irregular, depending on season ( Mesquita et al., 2017 ; Parker et al., 2018 ). As such, intermittent periods of extensive water circulation around and within caves and their hosting rocks can occur, followed by water stagnation or dry periods. Prior enrichment of canga-associated microorganisms from IFCs demonstrated that the microbial communities present were capable of Fe(III) reduction to extents that could contribute to IFC formation ( Parker et al., 2018 ), but Fe(II) that accumulates during Fe(III) (hydr)oxide reduction can adsorb to Fe(III) phase surfaces and induce mineral (trans)formations ( Roden et al., 2000 ; Benner et al., 2002 ; Hansel et al., 2003 , 2005 ; Gonzalez-Gil et al., 2005 ). These consequences of Fe(III) reduction could self-limit further Fe(III) (hydr)oxide reduction, although subsurface water flow could help overcome these limitations by advective transport of Fe(II) ( Gonzalez-Gil et al., 2005 ; Minyard and Burgos, 2007 ; Wefer-Roehl and Kübeck, 2014 ). Additionally, it remained unclear if the extents of microbiological Fe(III) (hydr)oxide dissolution observed were sufficient to induce hydrologic alterations that would culminate in cave formation. To understand whether this hydrologic flow could influence Fe-reduction rates and enhance IFC formation we compared batch cultures [where Fe(II) will accumulate] to columns [where Fe(II) is removed via flow] to evaluate how water flow influenced microbiologically mediated Fe(III) reduction, and whether such activity could influence the hydraulic properties of canga.",
"discussion": "Results and Discussion Fe(III) Reduction in Static Incubations To evaluate the canga-Fe(III) reducing activities of the microbial communities in the sub muric material, static batch incubations were conducted. Canga was provided as the Fe(III) source with SPW at a pH of 4.75 or 6.8, which matched the measured pH values in situ ( Parker et al., 2018 ). Minimal Fe(II) was generated in uninoculated incubations, but accumulated in the sub muros -inoculated incubations at both pH 4.75 and 6.8 ( Figure 1A ). The concentration of dissolved Fe(II) that accumulated in the sub muros- inoculated batch incubations (approximately 5 mM) exceeded previous batch incubation work in which Shewanella oneidensis MR-1 was used to catalyze canga-Fe(III) reduction (less than 0.6 mM; Parker et al., 2013 ). Indeed, mean total Fe(II) concentration of sub muros -inoculated incubations exceeded 80 mmol/L ( Figure 1B ). In previous work, a maximum of 3% of canga-Fe(III) could be reduced by S. oneidensis MR-1 ( Parker et al., 2013 ); however, greater extents of Fe(III) reduction have been observed by fermentative enrichments and isolates from canga by ourselves and other researchers ( Parker et al., 2018 ; Gagen et al., 2019 ). The drivers of this enchanced reduction remain unclear at this time and represent a good target for future research. FIGURE 1 Batch cultures of Fe(III) reduction in SPW canga. The concentration of dissolved Fe(II) (A) and total Fe(II) (B) were measured under static conditions over 3 months. Comparisons were made between sterile canga (open circles) or canga inoculated with sub muros material (closed circles), with a basal SPW medium pH of 4.75 (red) or pH 6.8 (blue). Error bars represent the standard deviation of triplicate incubations. In previous cultures using various Fe(III) mineral phases (including canga; Parker et al., 2018 ) we used a PIPES-buffered (also at pH 4.75 and 6.8) mineral salts medium for growth. In these cultures we saw a shift in microbial community structure from the Proteobacteria-dominated sub muros to one dominated by the Firmicutes, representing >97% of sequences ( Figure 2 ). In these previous experiments, we had assumed that the shift in community structure had been driven in part by the high amount of organic carbon, while the closed nature of the experiment allowed H 2 to accumulate and drive fermentative Fe-reduction by members of the Clostridia ( Shah et al., 2014 ; Parker et al., 2018 ). We tested this hypothesis in this study, using a basal medium (SPW) and 5 mM lactate as a carbon source. Analysis of partial 16S rRNA gene sequences in the batch incubations after 85 days revealed a similar dominance by fermentative Firmicutes ( Figure 2 ). Nonetheless, using SPW/lactate, the Proteobacteria remained abundant, comprising 23 and 15% of the sequences recovered from pH 4.75 and 6.8 incubations, respectively. We also saw a small, but significant population of Actinobacteria (6% at pH 4.75 and 3% at pH 6.8) and Bacteriodetes (2.5% only at pH 6.8) that had not been observed previously ( Figure 2 ). The pH of the uninoculated controls averaged 5.44, regardless of whether the pH 4.75 or 6.8 SPW was used to initiate the experiment, suggesting that canga buffered the pH; however, in the inoculated batch cultures, the SPW/lactate pH 4.75 culture increased to pH 6.10, while the SPW pH 6.8 culture remained reasonably constant at pH 6.67. There was no dramatic change in pH of the cultures following the addition of sub muros at day 0. This suggests Fe(III) reduction through microbial activity likely raises the pH (Eq 1): (1) Fe ( OH ) 3 + 3 H + + e - → Fe 2 + + 3 H 2 O FIGURE 2 Illumina sequencing results of phylum-level community diversity in batch and column cultures. Illumina sequencing of sub muros inoculated samples at day 0 are shown (inoculum). The diversity in our previous batch culture experiments, where the basal media was buffered with PIPES is shown (indicated as PIPES BATCH; Parker et al., 2018 ), followed by the cultures presented here using SPW with lactate (SPW BATCH). Illumina data is also provided for each of the individual columns in the flow-through experiments (COLUMN). The basal pH of each media formulation at day 0 is shown (Media Chemistry pH). The similarity in pH of the final culture conditions may explain the similarity of the final observed community profiles ( Figure 2 ). At the genus level within the dominant Firmicutes ( Figure 3 ), the SPW/lactate batch cultures displayed a different structural diversity to our previous work. Previously, at pH 4.75 the batch cultures were dominated by members of the genus Clostridium (Family Clostridiales; 71%), with a small but significant representation by the Desulfosporosinus (Family Peptococcaceae; 6%), while at pH 6.8, the PIPES batch cultures were dominated by both the Desulfosporosinus (38%) and Clostridium (36%) ( Parker et al., 2018 ). Both cultures also contained minor populations of the Paenibacilli (Family Bacillales; 3% at both pH 4.75 and 6.8). In the batch cultures presented here, we saw a similar dominance by members of the Clostridia (31% at pH 4.75 and 47% at pH 6.8), and Desulfosporosinus (20 and 24% at pH 4.75 and 6.8, respectively). The Desulfosporosinus sp. are normally associated with sulfate reduction, but have also been shown to reduce Fe(III) enzymatically ( Senko et al., 2009 ; Sato et al., 2019 ). If not enzymatic, the production of a minor amount of sulfide could be sufficient to enable Fe(III) reduction via S as an electron shuttle ( Hansel et al., 2015 ). Members of the Paenibacilli , which have recently been demonstrated to play an important role in iron oxide weathering in soils (including in Brazil; Loyaux-Lawniczak et al., 2019 ) were also represented at both pH 4.75 and 6.8 (4% of total diversity; Figure 3 ). Interestingly, we saw a higher percentage of members of the Coprococcus (Family Lachnospiraceae ; 2%) at both pH 4.75 and 6.8. The genus Coprococcus includes strict anaerobes that play an important role in carbohydrate fermentation in the mammalian rumen, including lactate ( Rainey, 2009 ). It is unclear as to why members of this genus would be enriched under the batch culture conditions; however, their growth is stimulated by fermentable carbohydrates, suggesting that the use of lactate may have enhanced their growth ( Cotta and Forster, 2006 ; Rainey, 2009 ). Members of this genus have not been associated with Fe-reduction, or isolated from iron-rich environments, although the production of H 2 during fermentation may contribute to the overall culture Fe-reduction conditions ( Cotta and Forster, 2006 ; Parker et al., 2018 ). We also observed a significant representation by members of the Family Veillonellaceae, with 11% at pH 4.75 and 5% at pH 6.8 ( Figure 3 ). Recently, genera within the Veillonellaceae, such as Sporomusa spp. and Propionispora spp., have been shown to carry out Fe(III) reduction ( Sass et al., 2004 ; Kato et al., 2015 ); however, rather using respiratory Fe(III) reduction, the Sporomusa appear to use Fe(III) as an electron sink in acetogenesis ( Igarashi and Kato, 2021 ). FIGURE 3 Illumina sequencing results of genus-level community diversity within the Firmicutes from the batch and column cultures. Only the SPW/lactate results are shown. The distribution of genera in the batch cultures (BATCH) and individual columns (COLUMN) are shown, along with the basal pH of the SPW at day 0 is shown. Given the myriad of Family- and Genera-level distributions within the Firmicutes, the Order/Family/Genus classification is provided for each identified species. Fe(III) Reduction in Column Incubations The underlying hypothesis of our work is that microbiological Fe(III) reducing activities are sufficient to induce porosity generation within the host rocks (i.e., canga, BIF, and iron ore); however, we have not observed hydrologic alterations of cave hosting rocks. Biogenic Fe(II) can limit the extent of Fe(III) (hydr)oxide reduction ( Roden and Zachara, 1996 ; Urrutia et al., 1999 ; Roden and Urrutia, 2002 ; Roden, 2004 , 2006 ), and induce mineralogical changes that would otherwise limit further Fe(III) reduction or limit the export of soluble Fe(II) (i.e., the formation of secondary minerals; Benner et al., 2002 ; Hansel et al., 2003 , 2005 ). Nonetheless, the advective removal of biogenic Fe(II) as water flows through Fe(III) (hydr)oxide-rich rocks could enhance their reduction ( Roden and Urrutia, 1999 ; Roden et al., 2000 ; Royer et al., 2004 ; Minyard and Burgos, 2007 ). For example, 95% of Fe(III) coating on sand was reduced over six months by Shewanella putrefaciens CN32 in flow-through columns, compared to 13% of the Fe(III) in batch incubations ( Roden et al., 2000 ). The climate regime in the SE area is highly seasonal, with over 80% of the ∼1,400 mm/year rainfall concentrated in November-March. Canga is a highly porous rock, with values between 24 and 29% ( Costa and Sá, 2018 ), while the friable ore underneath the canga is highly impermeable with values as low as 10 –8 m/s ( Mesquita et al., 2017 ). Thus, rainfall infiltrates quickly through the canga towards the caves and then drains rapidly toward the surface, with very little retention of water, except in a few shallow internal ponds. Despite the robust Fe(III) reducing activity observed in the batch incubations ( Figure 1 ; Parker et al., 2018 ), they do not mimic the hydrologic flow associated with the rocks of the SE or QF in which cave formation occurs with Fe(II) accumulating in the cultures. To mimic flow conditions in a laboratory setting, we packed canga into columns under conditions analogous to the batch incubations, and introduced flow into the system. This approach allowed us to answer the two major questions of this work: (1) does advective removal of biogenic Fe(II) enhance further canga-Fe(III) reduction and (2) are the Fe(III) reducing microbial activities associated with the sub muric material sufficient to induce hydrologic alterations to the host rock. The columns were packed with crushed canga alone, or with crushed canga mixed with sub muric material. The columns were incubated statically for 14 days, allowing Fe(III) reduction to initiate, before four column volumes of SPW were then passed through the column and collected separately for analysis of effluent chemistry ( Figure 4 ). This process of static incubation followed by introduced flow was then repeated at 7 day intervals. Minimal dissolved Fe(II) was detected in the effluent of uninoculated control columns throughout the incubation ( Figure 4 ), and effluent pH was ∼4.5–5.0, regardless of influent SPW pH. This is slightly lower than the values obtained in the batch experiments shown in Figure 1 . In the inoculated columns, progressively higher concentrations of dissolved Fe(II) accumulated over the course of the incubation, with maximum Fe(II) concentrations of approximately 3 mM Fe(II) detected after the fifth round of static incubation at both pH 4.75 and 6.8 ( Figures 4A,B ). FIGURE 4 Fe(III) reduction and changes in sulfate and pH in the column experiments. (A,B) Columns were operated semi-continuously, and sulfate and pH were measured in each pore volume (four volumes) recovered after each static incubation. The column was disassembled after two column volumes at day 63 for post mortem analysis. Error bars represent one standard deviation of triplicate columns. The concentration of dissolved Fe(II) is shown at pH 4.75 (A) and 6.8 (B) . Sulfate concentrations (black) and pH (blue) in column effluents are shown in panels (C) (pH 4.75) and (D) (pH 6.8). The values for uninoculated columns are shown with open circles, with the sub muros -inoculated columns represented by closed circles. The concentration of total Fe(II) (dissolved and solid-associated) produced in the columns incubated at either pH 4.75 or 6.8 SPW are shown in Table 1 . A two-sample t -test assuming unequal variances suggested that there was no significant difference ( P = 0.97) in total dissolved Fe(II) accumulation in the columns receiving SPW with pH 4.75 and 6.8 ( Figure 4 ). The pH of the column effluent suggests that there was an increase in pH above 6.0 ( Figure 4C ), similar to the batch cultures. The increase in Fe(III) reduction in the pH 4.75 column as the experiment progressed may reflect a change in column community structure as the cultures move toward similar pH conditions. The increase in pH conditions is correlated with the increasing observation of Clostridium kluyveri ( Figure 3 ) and may suggest either the selection of this species und these pH conditions, or a role in driving Fe(III)-reduction. TABLE 1 Post mortem analysis of column contents. pH 4.75 with sub muros pH 4.75 uninoculated pH 6.8 with sub muros pH 6.8 uninoculated Total Fe(II) (μmol/g) 60 ± 15 4.4 ± 0.2 89 ± 19 4.3 ± 0.1 Cell abundances t = 0 (cell/g wet) 9.3 × 10 7 ± 2.3 × 10 5 N/D 9.1 × 10 7 ± 1.8 × 10 5 N/D Cell abundances t = 63 (cell/g wet) 4.0 × 10 8 ± 8.1 × 10 7 N/D 4.2 × 10 8 ± 4.7 × 10 7 N/D Fe(OH) 3 removed as Fe 2+ (mg) 38 ± 18 2.3 ± 0.2 40 ± 3.5 2.3 ± 0.01 While canga is composed mostly of goethite and poorly crystalline Fe(III) (hydr)oxides ( Parker et al., 2013 ), if we assume dissolved Fe(II) is derived from Fe(OH) 3 , approximately 40 mg of Fe(OH) 3 were reductively dissolved and exported as Fe(II) from the packing material of inoculated columns, with minimal export of Fe from uninoculated columns ( Table 1 ). In the batch incubations, only ∼30 mg of Fe(OH) 3 were reductively dissolved ( Figure 1 ). After two porewater replacement events (21 days), the dissolved Fe(II) concentration in effluent from sub muros -inoculated columns exceeded 6 mM in total in pH 6.8 columns and over 5 mM total in pH 4.75 columns; iron reduction levels which only accumulated after 60 days continuous culture in batch incubations ( Figures 1A , 4A,B ). These results indicate that water flow enhances the reductive solubilization of Fe from canga and separation of the Fe(II) products from solid phases. Microbial Communities in Column Incubations The microbial community composition in the batch incubations suggested that non-respiratory Fe(III) reduction could play a role in the observed iron reduction ( Figure 3 ). To determine the extent of growth during column operation, we counted cells associated with the sub muros inoculum and at the conclusion of the column experiments. All the columns seeded with sub muros were initially inoculated at ∼9.2 × 10 7 cells/g. At 63 days, the population had increased in the columns at pH 4.75 by 4.3×, with the cell number in the pH 6.8 column increasing 4.6×. These data suggested an increase in microbial growth, and indeed the higher cell number is the pH 6.8 columns matches a higher-level of Fe(III) reduction. No microbial cells were detected in the uninoculated controls ( Table 1 ). DNA extraction from the inoculated columns produced sufficient DNA for Illumina sequencing, but repeated attempts to extract DNA from the uninoculated columns failed, matching the observations by direct cell counting. Illumina sequencing of the microbial communities in the columns matched our observations in batch culture ( Figures 2 , 3 ); there had been a shift from dominance by the Proteobacteria, to dominance by members of the Firmicutes. At the genus level, the columns were similarly dominated by members of Clostridium, Desulfosporosinus , and Veillonella , which represented ≥90% of the identified partial 16S rRNA gene sequences ( Figure 3 ); however, members of the Paenibacilli were not observed. There was some inter-column variability under each of the pH conditions, particularly in regard to the dominance of Clostridium relative to Desulfosporosinus ( Figure 3 ). In the Desulfosporosinus- dominated columns, we saw a darkening of the column material, which could indicate sulfidogenesis, but there was no decrease in the effluent sulfate concentration over the course of the incubations ( Figures 4C,D ). This suggests that while members of the Desulfosporosinus are accumulating in these columns, they may be functioning as Fe(III) reducers. Indeed, members of this genus have been shown to be the primary Fe(III) reducers under oligotrophic conditions ( Nixon et al., 2017 ; Bomberg et al., 2019 ). Fe(III) reduction is widespread among the Clostridia, including a strain of Clostridium beijerinckii ( Dobbin et al., 1999 ; Lehours et al., 2010 ; Shah et al., 2014 ; List et al., 2019 ). Indeed, in our previous batch cultures were capable of extensive (in some cases, complete) Fe(III) reduction ( Parker et al., 2018 ), and Lentini et al. (2012) have demonstrated that Clostridium- enriched cultures are capable of extensive reduction of goethite- and hematite-Fe(III). Microbially Induced Hydrologic Alterations of Canga Columns Based on dissolved Fe(II) in column effluents, approximately 40 mg of Fe(OH) 3 were removed from the columns due to microbiological Fe(III) reduction ( Table 1 ). To determine if this export of mass impacted the hydraulic properties of the columns, we pumped bromide-amended SPW through the columns. Bromide breakthrough in the sub muros -inoculated columns preceded that of the uninoculated columns, and breakthrough was spread out in comparison to that of the uninoculated columns, which had a sharper curve ( Figure 5 ). These observations indicate that flow through the uninoculated columns did not experience the same mass transfer resistance seen in the columns in which microbiological Fe(III) reduction occurred ( Lassabatere et al., 2004 ; Koestel et al., 2011 ; Safadoust et al., 2016 ). The porosity that allowed earlier bromide breakthrough is due to reductive dissolution of Fe(III) phases and export of dissolved Fe(II). In similar column experiments, Liang et al. (2019) found that bioreduction of sediment-associated Fe(III) led to the structural breakdown of particles in the columns and led to the earlier breakthrough of poorly-diffusible 2,6-difluorobenzoate. No change in more diffusible bromide breakthrough was observed after Fe(III) bioreduction ( Liang et al., 2019 ). In the work presented here, Fe(III) bioreduction was more extensive, with maximal effluent Fe(II) concentrations of approximately 3 mM, in comparison to the maximal Fe(II) concentration of 0.3 mM observed by Liang et al. (2019) . Taken together, the extensive Fe(III) bioreduction observed in these column experiments induced changes to the water flow paths in the packed canga. FIGURE 5 Bromide breakthrough curves of sub muros -inoculated (closed shapes) and uninoculated (open shapes) columns after 63 days of operation. The columns that received the basal SPW pH 4.75 media are in red, with the SPW pH 6.8 in blue. Mass transfer zone (MTZ) lines represent initial breakthrough point where bromide-amended SPW is mixing with bromide-free SPW and adsorption exhaustion point where column is saturated with bromide-amended SPW. SPW was fed to columns at a rate of 0.2 mL/min. Error bars represent one standard deviation of triplicate columns. Biogeochemical Implications The results of our experiments indicate that the Fe(III) reducing activities of microorganisms associated with IFCs can induce reductive dissolution of Fe(III) phases, resulting in the transport of dissolved Fe(II) and hydrologic changes that are consistent with cave formation. While the Fe(III)-rich rocks of this region were generally considered to be resistant to weathering ( Schuster et al., 2012 ; Monteiro et al., 2014 ), it is becoming increasingly clear that microbiological activities may induce extensive transformations to these rocks, especially canga ( Parker et al., 2013 , 2018 ; Levett et al., 2016 , 2020 ; Gagen et al., 2018 , 2019 ; Paz et al., 2020 ). Previous work has focused on the transformations of canga-Fe as a mechanism of canga permanence, whereby the weathering resistance of canga is owed to the alternating reductive dissolution of Fe(III) (hydr)oxides and abiotic or microbiological reoxidation of Fe(II) back to Fe(III) ( Levett et al., 2016 , 2020 ; Gagen et al., 2018 , 2019 , 2020 ; Paz et al., 2020 ). In this way, canga appears to be continuously weathering and reforming. The work here indicates that the Fe(III) rich phases could be more extensively weathered and removed from the systems, driven by the increased rates of Fe-reduction induced by water flow. Thus, Fe may be extensively mobilized from rocks in the SE and QF by microbiological weathering via microbial Fe(III) reduction (either through respiratory activity or as an electron sink) and separation, which can be enhanced by groundwater flow. These results should be applicable to other iron formation areas in Brazil and help explain why caves are larger in the iron deposits of Carajás, in the wetter Amazon Basin ( Auler et al., 2019 ). A positive feedback mechanism, in which fast infiltration water would lead to increased porosity and thus even faster water percolation could operate, enhancing the mass transfer mechanisms required to mobilize Fe(II). Our observations indicate that microorganisms associated with these systems are capable of robust Fe(III) reducing activity, which could induce sufficient reductive dissolution of Fe(III) phases to form a cave. The numerous caves of the SE and QF (>3,000; Auler et al., 2019 ) indicate that the activity is extensive and continuously occurring. Indeed, we have observed remarkably high dissolved Fe concentrations in water circulating around caves in the QF ( Parker et al., 2018 ). This extensive weathering of SE and QF Fe(III) phases may represent a previously underappreciated component of regional, and perhaps global Fe budgets."
} | 7,164 |
29724835 | PMC5940957 | pmc | 5,857 | {
"abstract": "ABSTRACT The spore-forming, thermophilic, and obligate anaerobic bacterium Moorella stamsii was isolated from digester sludge. Apart from its ability to use carbon monoxide for growth, M. stamsii harbors several enzymes for the use of different sugars. The draft genome has a size of 3,329 Mb and contains 3,306 predicted protein-encoding genes."
} | 87 |
27287198 | PMC4901517 | pmc | 5,858 | {
"abstract": "Background The production and employment of cellulases still represents an economic bottleneck in the conversion of lignocellulosic biomass to biofuels and other biocommodities. This process could be simplified by displaying the necessary enzymes on a microbial cell surface. Such an approach, however, requires an appropriate host organism which on the one hand can withstand the rough environment coming along with lignocellulose hydrolysis, and on the other hand does not consume the generated glucose so that it remains available for subsequent fermentation steps. Results The robust soil bacterium Pseudomonas putida showed a strongly reduced uptake of glucose above a temperature of 50 °C, while remaining structurally intact hence recyclable, which makes it suitable for cellulose hydrolysis at elevated temperatures. Consequently, three complementary, thermophilic cellulases from Ruminiclostridium thermocellum were displayed on the surface of the bacterium. All three enzymes retained their activity on the cell surface. A mixture of three strains displaying each one of these enzymes was able to synergistically hydrolyze filter paper at 55 °C, producing 20 μg glucose per mL cell suspension in 24 h. Conclusion We could establish Pseudomonas putida as host for the surface display of cellulases, and provided proof-of-concept for a fast and simple cellulose breakdown process at elevated temperatures. This study opens up new perspectives for the application of P. putida in the production of biofuels and other biotechnological products. Electronic supplementary material The online version of this article (doi:10.1186/s12934-016-0505-8) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions We provided proof-of-concept for the application of the industrially promising bacterium P. putida as host for the surface display of cellulases, and showed that it is possible to combine bacteria with different cellulases on their surface to achieve a synergistic hydrolysis of cellulosic substrates. The achieved activities still have a rather conceptional than industrially applicable value and require optimization. Nevertheless, the herein proposed approach can serve as a starting point for the creation of a fast, simple and modularly expandable cellulose degradation system.",
"discussion": "Discussion Displaying cellulases on the surface of microbial cells offers some advantages that could make the degradation of lignocellulose more cost-efficient. Therefore several groups have studied various approaches, which basically differ in the used host organism, surface display method, type and origin of the displayed cellulases, and connected to that, their expression as a separate enzyme or as a complexed part of a cellulosome [ 6 ]. The host organism used for expressing and displaying the desired enzymes determines important characteristics of the generated whole cell catalyst such as growth rate, enzyme activity and stability. The availability of an appropriate host organism thus is crucial for the transfer of the cellulase display concept to industrial applications. Yeast, due to its advanced establishment in industry, has been reported most often as host species for cellulase display applications. Examples for the display of multiple individual cellulases [ 17 , 43 ] as well as the display of cellulosomes [ 16 , 18 ] have been reported. Surface display of cellulases on bacteria has to date mostly been approached with the Gram-positive Bacillus subtilis and the Gram-negative E. coli as hosts. B. subtilis has especially been used for the display of cellulosomes [ 19 , 26 ]. In contrast, E. coli has so far only been reported as host for the display of one or more non-complexed cellulases using various outer membrane anchoring motifs [ 20 , 21 , 24 , 25 ], among them also the AIDA-I autotransporter [ 23 ]. The only exception from using these established model organisms was reported by Kojima et al., who displayed an endocellulase on the surface of the ethanologenic bacterium Zymobacter palmae using the ice nucleation protein anchoring motif from Pseudomonas syringae [ 44 ]. Because of its versatile metabolism, ease of genetic manipulation and resistance towards various adverse conditions, P. putida has in recent years strongly developed towards industrial applications [ 27 ]. However, it has never before been used as a platform for surface-displayed cellulases. In this study, we therefore aimed to create a simple and flexible cellulolytic system based on this bacterium. For a cellulose hydrolysis process in which the generated glucose is not directly converted to the desired product, but instead is intended to be utilized in a separate fermentation step, it is of primary interest that the sugar can be recovered in high yields from the hydrolysate. For a whole cell catalytic approach as presented here, this means that the employed host cells must not take up and metabolize glucose at the chosen reaction conditions. We found that at 50 °C and above nearly no glucose was consumed by P. putida , presumably because its glucose transporters and/or metabolic enzymes were denatured. After 1 day of incubation at a temperature of 55 °C, the largest part of the cells was still separable from the reaction mixture by centrifugation. This is particularly important for industrial scale applications, in which a recovery of the cells for repeated hydrolysis cycles is desirable, and emphasizes one of the advantages of using P. putida for the described process. The MATE system used for displaying the cellulases is based on the EhaA autotransporter from E. coli , which has previously been shown to be applicable for surface display in E. coli itself [ 32 , 45 ] and also in the ethanologenic bacteria Z. palmae and Zymomonas mobilis [ 46 ]. In this study, the expressed MATE fusion proteins could be found in the outer membrane of P. putida , providing evidence that their N-terminal CtxB signal peptide was properly recognized by the organism’s Sec machinery and consequently triggered their translocation into the periplasmic space. Flow cytometry measurements showed that all three cellulases were successfully exposed to the extracellular space of P. putida , which proves that the EhaA transporter domains are functional in this species and allow the display of recombinant proteins on its cell surface. Previously reported incompatibilities between AT and host organism [ 47 , 48 ] and the disability of an AT to display heterologous proteins [ 49 ] therefore seem not to be a problem in this case, and suggest that MATE can be used in a broad range of host bacteria. Whole cell activity assays showed that all three enzymes retained their functionality on the cell surfaces and exhibited hydrolytic activity towards their substrates. The cells could be recycled for consecutive reactions and retained between 67 and 96 % of their initial activity after five cycles. These are very high residual activities when compared to previous studies, in which a prenyltransferase retained 23–46 % [ 40 ] of its activity after three repeated uses and a nitrilase 55 % after five reaction cycles [ 39 ]. The chosen reaction temperature of 55 °C reflects the consideration of the previously mentioned uptake of glucose and structural stability of P. putida as well as the enzymes temperature optima, which are reported to be 60 °C for BglA [ 50 ], 65 °C for CelK [ 51 ] and 75 °C for CelA [ 52 ]. According to these optima, we experienced higher catalytic activities of the whole cell catalysts when increasing the temperature. For example, the activity of P. putida cells with surface displayed CelA was more than double as high at 75 °C than at 55 °C (92.78 mU/mL OD1 compared to 41.89 mU/mL OD1 , endpoint measurement; data not shown). Thus, from the viewpoint of enzyme activity a raise of temperature beyond 55 °C appears beneficial, but could impair the host cells structural integrity. To resolve this mismatch, one could apply enzymes with a lower temperature optimum, or establish a more thermostable bacterium as expression host for the MATE system. The first approach would require bacterial cellulases that are highly efficient at low temperatures; enzymes with such characteristics are currently not available. Following the second approach, it has to be taken into account that the cultivation of thermophilic bacteria and their use for protein expression is difficult and energy-intensive. This could collide with the intended simplicity and cost-efficiency of the process. As enzymes represent the largest cost contributors to the conversion of cellulosic biomass to fuels or other chemicals, the commercial success of such a process depends on the ability to produce very large amounts of enzymes at reasonable costs [ 53 ]. When employing microbes instead of purified enzyme cocktails to achieve a cost reduction, it could be problematic to express satisfactory amounts of the at least three necessary types of enzymes in a single microbial cell. In contrast, a separated expression of the enzymes keeps the metabolic burden on the microbial cells low and, more importantly, makes it possible to adjust the quantity of each enzyme in the mixture. Such a strategy has been evaluated previously by Baek et al, who displayed cellulases on yeast cells and determined an optimal cell ratio for a maximized hydrolysis efficiency [ 54 ]. An approach like this has not been followed with bacteria so far. In this study, we found that a mixture of exocellulase CelK- and endocellulase CelA-displaying P. putida cells produced twice the amount of reducing sugars from filter paper than cells displaying only one of these cellulases. This result showed, in accordance with the study of Baek et al., that the enzymes do not necessarily have to be displayed on the same cell surface to obtain a synergistic activity. Unexpectedly, CelA-displaying cells alone produced a relatively large amount of glucose from the filter paper. It is thinkable that the endocellulase preferably hydrolyzed the very ends of the cellulose chains because they were better accessible for the enzyme, resulting in an increased release of glucose. Another explanation could be that CelA, beside its endocellulase activity, also possesses a β-glucosidase functionality, however no reports to this issue are available so far. The addition of β-glucosidase BglA-displaying cells resulted in a significant increase of glucose in the reaction mixture, demonstrating that BglA also participated in the hydrolysis process. However, while the mixture of all three strains was able to produce 300 μg/mL glucose equivalents, the actual yield of glucose was only 20 μg/mL. Although a direct comparison of these two quantities cannot be drawn, their dimensions make obvious that the BglA-displaying cells represented the limiting factor in the complete hydrolysis of filter paper. This could be due to a poor expression level as evidenced by a comparably low amount of MATE-BglA found in the outer membrane of the expressing bacteria. However, it is also conceivable that the activity of BglA was impaired due to its expression as a MATE-fusion protein. According to our experience, the fusion of an enzyme with an AT does in most cases not interfere with its activity, still there are examples in which the fusion lead to a reduced enzyme activity [ 39 , 55 ]. For verification of both hypotheses a comparison between the specific activities of free and displayed enzyme would be necessary, the latter requiring the number of enzymes on the bacterial cell surface for calculation. When using E. coli as host, this number can be determined approximately by performing an SDS-PAGE analysis of outer membrane protein isolates and comparing the intensity of the autotransporter fusion protein band with the intensity of the OmpA protein band, which is known to be present in a constant number in the outer membrane of E. coli . Since an analogous reference protein has not been established for P. putida yet, the quantification of displayed enzymes on the surface of this host, and hence a comparison between the activity of free and displayed enzyme, is currently not feasible. β-Glucosidases represent a known bottleneck in cellulose hydrolysis processes [ 56 ]. To solve this, BglA either has to be substituted by an enzyme with higher activity, or the amount of BglA-displaying cells in the mixture has to be increased. For further development of the presented concept, we are planning to (1) optimize the expression of the MATE-cellulases, e.g. in terms of used plasmid backbone, promoter, culture conditions, in order to achieve a higher number of enzymes on the cell surfaces, and (2) to systematically vary the cell mixtures in their composition to find out an optimal ratio. Beyond that, the replacement of filter paper with an industrially relevant, lignocellulosic substrate would be desirable to have a more realistic valuation basis for the concept at hand."
} | 3,263 |
29349039 | PMC5767561 | pmc | 5,859 | {
"abstract": "Recent studies have revealed that caryophyllene and its stereoisomers not only exhibit multiple biological activities but also have desired properties as renewable candidates for ground transportation and jet fuel applications. This study presents the first significant production of caryophyllene and caryolan-1-ol by an engineered E. coli with heterologous expression of mevalonate pathway genes with a caryophyllene synthase and a caryolan-1-ol synthase. By optimizing metabolic flux and fermentation parameters, the engineered strains yielded 449 mg/L of total terpene, including 406 mg/L sesquiterpene with 100 mg/L caryophyllene and 10 mg/L caryolan-1-ol. Furthermore, a marine microalgae hydrolysate was used as the sole carbon source for the production of caryophyllene and other terpene compounds. Under the optimal fermentation conditions, 360 mg/L of total terpene, 322 mg/L of sesquiterpene, and 75 mg/L caryophyllene were obtained from the pretreated algae hydrolysates. The highest yields achieved on the biomass basis were 48 mg total terpene/g algae and 10 mg caryophyllene/g algae and the caryophyllene yield is approximately ten times higher than that from plant tissues by solvent extraction. The study provides a sustainable alternative for production of caryophyllene and its alcohol from microalgae biomass as potential candidates for next generation aviation fuels.",
"introduction": "1 Introduction Caryophyllene, a natural bicyclical sesquiterpene (C15) compound, is a common component present in the essential oils of various plants ( Kpadonou Kpoviessi et al., 2012 , Meccia et al., 2009 , Rodrigues et al., 2012 ). It is widely used for flavoring and personal healthcare applications ( Sabulal et al., 2006 , Sköld et al., 2006 ). Several studies have revealed that beta-caryophyllene and caryophyllene essential oils exhibit a wide range of potential therapeutic applications ( Alvarez-Gonzalez et al., 2014 , Klauke et al., 2014 , Paula-Freire et al., 2014 , Rufino et al., 2015 ; Cheng et al., 2014 ). Caryolan-1-ol (caryophyllene alcohol) is a hydroxylation product of caryophyllene ( Nakano et al., 2011 ). Similar to caryophyllene, caryolan-1-ol is also a fragrance ingredient widely used in personal and house healthcare ( Bhatia et al., 2008 ). In addition to their biological activities, recent studies suggest that the hydrocarbons derived from terpenes are structurally similar to the compounds in petroleum distillate fuels, and often share similar combustion properties ( Edwards et al., 2010 ). Blending of hydrogenated sesquiterpanes, carophyllane and its stereoisomers, in particular, which have a moderate cetane number and only moderately high viscosity, with synthetic branched paraffins has been found to raise cetane number and reduce viscosity for biosynthetic fuels that meet applicable jet and diesel specifications ( Harvey et al., 2015 ). Additionally, current bio-based fuel molecules such as ethanol have very high oxygen content (up to 2:1 of C:O), which introduces significant cost and material property hurdles for blending into the petroleum-derived fuels infrastructure ( http://www1.eere.energy.gov/bioenergy/pdfs/algal_biofuels_roadmap.pdf ). Compared with fuel alcohol, the terpene mixtures of caryophyllene, caryophyllene alcohol and other caryophyllene stereoisomers are either oxygen free or have very low oxygen hydrocarbon and hydrocarbon-like compounds (C:O>10), making them particularly attractive candidates as “drop-in” replacements for non-renewable ground transportation and aviation fuels. Based on this observation, caryophyllene and its isomers have been deemed to be among the top three most promising jet fuel compounds with high energy density ( http://www.biofuelsdigest.com/bdigest//06/18/9-advanced-molecules-that-could-revolutionize-jet-and-missile-fuel/, 2014 ). It often requires blending of different types of hydrocarbons to achieve satisfactory combustion properties of the fuel. Harvey et al. ( Harvey et al., 2015 ) recently determined that blending hydrogenated heterogeneous sesquiterpenes (particularly, caryophyllene and its stereoisomers) with synthetic branched paraffins could yield fuel with improved combustion properties. The high heterogeneity of the sesquiterpene profile produced by the endophytic caryophyllene synthase makes it a desirable terpene synthase (TPS) for the production of terpene mixture enriched with caryophyllene and its stereoisomers as the next generation renewable aviation and diesel fuels. In our previous studies ( Gladden et al., 2013 , Wu et al., 2016 , Wu et al., 2017 ), we discovered a suite of novel caryophyllene synthases from endophytes, which produced a wide spectrum of terpenes, of which caryophyllene accounted for 40%, indicating low product specificity and high heterogeneity of the terpene profile. Therefore, in this study, we selected a highly non-specific caryophyllene synthase and a caryolan-1-ol synthase with high product heterogeneity and reported the production of caryophyllene and caryophyllene alcohol enriched terpene mixture as potential compounds for aviation fuel through the heterologous expression of the mevalonate pathway with a geranyl diphosphate (GPP) synthase ( Burke and Croteau, 2002 ), an endophytic caryophyllene synthase ( Gladden et al., 2013 , Wu et al., 2016 ), and a caryolan-1-ol synthase ( Nakano et al., 2011 ). Microalgae, due to its high lipid content, high growth rate, no competition for agriculture land use, and the ability to utilize CO 2 , represents a promising renewable resource for the sustainable production of bioproducts and fuels. Under conditions supporting robust growth, microalgae accumulate carbohydrate and protein as major components, up to 80% of total biomass, with lipid contents typically less than 20% ( Laurens et al., 2014 , Luque, 2010 ). Recently, a few studies have reported bioconversion of the carbohydrates or proteins from algae biomass into ethanol ( de la Cruz et al., 2014 , El-Mashad, 2015 , Fasahati et al., 2015 , Martin and Grossmann, 2014 ) and isobutanol ( Huo et al., 2011 ). In this study, we demonstrated the viability of bioconversion of the algae hydrolysate into caryophyllene enriched terpene mixture using the newly developed bioconversion strain, which paves the way for the cost-effective and sustainable production of fuels from algae biomass.",
"discussion": "4 Discussion To optimize the metabolic flux driving force for increased total terpene production, three expression strategies were applied and the constructs were co-expressed into E. coli strains. Among the four engineered strains, DH1-CS2 yielded the highest titer of terpenes while DH1-CS1 produced the lowest concentration of terpenes. The lower terpene titer of DH1-CS1 is probably due to metabolic flux imbalance, likely resulting from the non-optimized expression levels of GPPS and TPS as well as the higher metabolic burden induced by the three plasmids. In construct 1, both GPPS and caryophyllene synthase were expressed in two plasmids containing the same origin of replication ColE1. The incompatibility of these two plasmids in the strain may cause the low expression levels of either GPPS Ag or caryophyllene synthase, which can further contribute to a low flux ( Selzer et al., 1983 ; Tolia and Joshua-Tor, 2006 ). In constructs 2, 3, and 4, the pathway enzymes were expressed in two plasmids with different but compatible origin of replications - pJBEI3122 with p15A and of pBbE1a with ColE1origin. In addition, in construct 2, the GPPS was expressed in the vector pJBEI3122 (p15A) which can yield 10–12 copies of the plasmid in the strain, while the plasmid pBbE1a (ColE1) containing caryophyllene synthase can produce 15–20 copies of the plasmid ( Tolia and Joshua-Tor, 2006 ). pJBEI3122 expressing all intermediate pathway enzymes with relatively lower copy numbers could reduce the potential toxicity of intermediate metabolites on the cell ( Jones et al., 2000 ). Furthermore, the higher-copy-number plasmid pBbE1a expressing caryophyllene synthase could possibly generate a large metabolic flux driving force to the final product formation, which in turn reduced the intermediate accumulation. Most likely, the combination of these two factors resulted in the highest production titer of terpene from the DH1-CS2 strain. In addition, DH1-CS2 accumulated the lowest level of mevalonate ( Fig. 1 (C)) in vivo , indirectly indicating the highest metabolic flux driving force to final products formation. The construct 3 and 4 have similar expression strategies for the pathway enzymes. Consequently, strains DH1-CS3 and 4 accumulated similar levels of mevalonate in vivo and produced similar titers of terpene in the culture. Optimization of fermentation conditions is imperative for improving the product yield of the engineered terpene producing strains. Generally, the chemical reaction rate increases as the reaction temperature rises. However, previous studies showed that lower induction temperatures reduce the formation of inclusion bodies in metabolically engineered strains and can enhance the activities of pathway enzymes and improve the product yield ( de Groot and Ventura, 2006 , Vera et al., 2007 ). Therefore, optimization of inducer concentration and induction temperature for pathway enzyme express was identified as important means to achieve the maximal terpene production yield. In this study, DH1-CS2 produced the highest amount of total terpene at 37 ° C while DH1-gcoA produced the highest amounts of terpenes at 25 °C, indicating that the properties of the expressed terpene synthase was dependent on the production temperature. In most cases, exogenously introduced pathways increase the metabolic burden on the cells ( Glick, 1995 ; Neubauer et al., 2003 ; Sharma et al., 2011 ). The levels of inducer (in this case, IPTG) can modulate the transcription of the heterologous pathway genes and, therefore, the optimal inducer concentration can minimize the extent of metabolic burden on the engineered strains and therefore maximize the product yield ( Donovan et al., 1996 ; Huang et al., 2014 ; Liu et al., 2011 ; Zhang et al., 2009 ). The different optimal inducer concentrations for different TPSs also suggested distinct properties among the expressed terpene synthases. Ohnishi and collaborators ( Nakano et al., 2011 ) previously reported that gcoA catalyzes the biosynthesis of caryolan-1-ol via formation of β-caryophyllene but that the major portion of β-caryophyllene does not interact with the binding site of gcoA during the catalysis process. This indicates that the gcoA has faster kinetics for the conversion of FPP into caryophyllene than the hydroxylation of caryophyllene into caryophyllene alcohol, which results in the major portion of β-caryophyllene being released into the medium before being hydroxylated into caryophyllene alcohol. This is likely the underlying reason for the higher terpene titer from DH1-gcoA strain expressing caryophyllene compared with the one expressing caryolan-1-ol. In DH1-CS-gcoA, it was expected that the caryophyllene synthesized by CI4A-CS could be hydroxylated by caryolan-1-ol synthase (gcoA) to form caryophyllene alcohol, resulting in the increase of the caryophyllene alcohol titer by supplying gcoA with more caryophyllene as substrates. However, DH1-CS-gcoA produced lower titers of caryolan-1ol and caryophyllene than DH1-gcoA, which is probably due to the competition of CI4A-CS against caryolan-1-ol synthase in construct 6 for FPP as substrate. The competition most likely resulted in the formation of other sesquiterpene compounds by CI4A-CS such as humulen, which was indicated by the higher heterogeneity of the terpene mixture profile, and therefore lower amount of caryophyllene and caryophyllene alcohol could be converted from FPP. In this study the terpene titer from algal hydrolysate was reduced compared with the EZ-rich medium. This is likely due to the fact that algae hydrolysate without any supplements is nutrient limited compared with rich EZ-rich medium or other commercial media formulation. We observed that the culture required over double the time to reach the induction cell density than in EZ-rich medium. Compared with terpene production from sugars, the current yield of total terpene based on algae biomass from DH1-CI4A-CS2 is approximately one order of magnitude lower. However, compared with the current state-of-the-art technology of essential oil production where the extraction yields of essential oils ranged from 0.1% to 1% of plant tissues (corresponding to 1–10 mg essential oil/ g plant tissue) ( Gong et al., 2014 , Moncada et al., 2016 ), the engineered strain in this study increased the terpene yield by 4–40 folds (algae biomass based), which makes it a promising alternative pathway for terpene production. In summary, caryophyllene, caryolan-1-ol, and their stereoisomers were produced by the engineered E. coli strains as potential candidates for aviation fuel. The total terpene and caryophyllene titers were increased after improving the metabolic flux through four different biochemical pathway regulation strategies as well as the optimization of fermentation parameters. Moreover, this is the first demonstration of caryophyllene-enriched terpene mixture production from an algae biomass hydrolysate. This study provides an alternative pathway for caryophyllene caryolan-1-ol enriched terpene production as potential aviation fuel compounds from renewable sources."
} | 3,379 |
29196631 | PMC5711802 | pmc | 5,860 | {
"abstract": "Web spiders synthesize silk fibres, nature’s toughest biomaterial, through the controlled assembly of fibroin proteins, so-called spidroins. The highly conserved spidroin N-terminal domain (NTD) is a pH-driven self-assembly device that connects spidroins to super-molecules in fibres. The degree to which forces of self-assembly is conserved across spider glands and species is currently unknown because quantitative measures are missing. Here, we report the comparative investigation of spidroin NTDs originating from the major ampullate glands of the spider species Euprosthenops australis , Nephila clavipes , Latrodectus hesperus , and Latrodectus geometricus . We characterized equilibrium thermodynamics and kinetics of folding and self-association using dynamic light scattering, stopped-flow fluorescence and circular dichroism spectroscopy in combination with thermal and chemical denaturation experiments. We found cooperative two-state folding on a sub-millisecond time scale through a late transition state of all four domains. Stability was compromised by repulsive electrostatic forces originating from clustering of point charges on the NTD surface required for function. pH-driven dimerization proceeded with characteristic fast kinetics yielding high affinities. Results showed that energetics and kinetics of NTD self-assembly are highly conserved across spider species despite the different silk mechanical properties and web geometries they produce.",
"introduction": "Introduction The question of how a linear chain of amino acids spontaneously folds into a highly ordered three-dimensional structure continues to be a central topic in molecular biology 1 . Important insights come from comparative studies of homologous proteins, where differences in folding mechanisms can be traced back to minor sequence changes 2 – 4 . Such studies allow the dissection of the roles of sequence and topology in folding. A class of highly evolved proteins that exhibit rather unusual amino acid composition are the silk proteins produced by web spiders. Spider silk is produced by the controlled assembly of spider fibroins, so-called spidroins, within the spinning gland of the animal yielding threads of outstanding mechanical properties tailored for distinct functionalities. Material scientists are trying to decrypt the assembly process and to reproduce it in the laboratory 5 – 7 . The bulk of a spidroin sequence consists of repetitive poly-alanine and glycine-rich peptide motifs of simple amino acid composition, which are unstructured under storage conditions in the gland and form mainly β-sheet secondary structure in solid fibres. The repetitive central segments are terminated by the globular folded N- and C-terminal domains (NTD and CTD), which provide water-solubility on the one hand and connectivity in response to mechanical and chemical stimuli on the other 8 , 9 . NTD and CTD are five-helix bundles that form homo-dimers. While the CTD is a covalent homo-dimer stabilized by a disulfide linkage, the NTD is monomeric under storage conditions in the gland and undergoes self-association in the spider’s spinning duct in response to changes of solution pH and salt composition, thus polymerizing spidroins to form super-molecules 7 , 10 . pH-triggered NTD self-association is ultrafast and involves site-specific protonation events and conformational change 11 – 15 . NTD and CTD represent the most conserved sequence areas of spidroins, with no structural homologues identified so far, underscoring their importance in the process of silk formation 8 , 10 . Interestingly, the unusual amino acid composition of the central spidroin segments extends into the terminal domains: in the NTD, alanine is the most frequently found amino acid followed by serine; alanine and serine together take up ~30% of an NTD sequence. By contrast, the number of charged side chains is rather low. This is surprising considering the high water-solubility of the domain, which is commonly provided by side chain charges. Unusual amino acid composition and high degree of sequence conservation make spidroin NTDs an interesting system both from the viewpoint of fundamental folding research and material science. Open questions are: does the unusual amino acid composition of NTDs translate into an unusual mechanism of folding? Are energetics and kinetics of folding and self-association conserved? This question is in particular interesting in light of species-dependent differences in strengths and structures of silks 16 . Here, we report the comparative investigation of folding and association of NTDs originating from major ampullate spidroin 1 (MaSp1) of four different spider species, namely the black widow ( Latrodectus hesperus , Lh) the brown widow ( Latrodectus geometricus , Lg), the golden orb spider ( Nephila clavipes , Nc), and the nursery web spider ( Euprosthenops australis , Ea). The Ma gland forms the toughest fiber used to build the web frame or a lifeline and is a focus of current material science. We found that all four domains folded on a similar sub-millisecond time scale via a conventional two-state mechanism. The Nc homologue, however, was significantly less stable and folded more slowly compared with the other domains. Rate constants of pH-triggered domain self-association and dissociation were similar. Results showed that, despite species-dependent differences in web geometries and silk mechanical properties, the energetics and kinetics of NTD self-assembly are conserved.",
"discussion": "Discussion Web spiders use up to seven specialized glands to synthesize silk fibres for various tasks including prey capture, reproduction and shelter 6 , 7 . The basic principles of synthesis are thought to be conserved across glands and species. Conservation of mechanism is reflected in the conserved, modular sequence architecture of spidroins, which are the building blocks of silk. Yet, sizes, geometries, and mechanical properties of webs built by different spider species vary strongly. These differences may arise from sequence variations in the central, repetitive spidroin segments that make up the bulk of interactions in silk and possibly from different processing conditions in various spinning ducts 5 , 6 , 16 . Little is known about the contribution of sequence modulations in the terminal domains to modulation of silk. Structural studies show that MaSp NTD folds from Ea, Lh, and Nc, as well as from a homologue of the minor ampullate gland of Araneus ventricosus \n 8 , 15 , 20 , 21 are conserved, although minor differences in helix packing are observed 20 . Biophysical measurements show similar signatures of pH- and salt-dependent dimerization of the Euprosthenops , Latrodectus , Nephila and Araneus NTDs 13 – 15 , 21 . However, quantitative measures of folding and self-association were missing and are reported here. Our comparative study of MaSp1 NTDs from Ea, Nc, Lh, and Lg shows similar intensity-loss and red-shift of Trp fluorescence emission upon dimerization (Fig. 1b ). Stopped-flow fluorescence experiments showed that the high speed of pH-triggered self-association was similar in all four homologues and responsible for tight binding, with K \n d values in the low nM range (Table 1 ). Structural studies report subtle rearrangement of helices and a consequently altered dimer interface of the Nc homologue 20 . We found that these structural rearrangements did apparently not translate into a modulation of dimerization kinetics or strength of association. Our results suggest that the MaSp1 NTD evolved as a pH-driven module that connects spidroins in the distal part of the gland by same mechanism and energetics irrespective of species. Findings underscore the importance of species-dependent sequence modulations in the central, repetitive segments for modulation of silk mechanical properties. The folding of all four homologues appeared cooperative and was well described by two-state transitions - a behaviour that is frequently observed for small, single-domain proteins 22 . Cooperativity of folding was evident from the good agreement of thermodynamic quantities measured using two different structural probes and from the overall good agreement of quantities derived from equilibrium and kinetic experiments. We found remarkably fast sub-millisecond kinetics of folding of all four homologues (Table 4 ). Chevron analysis of folding kinetics supported the two-state model with no indications of populated folding intermediates. The kinetic m -value and extrapolated rate constant of folding of the Ea NTD was within error of the values previously estimated from temperature-jump experiments 23 . Fast kinetics of folding suggest MaSp1 NTDs as an interesting family for future combined experimental and computational studies that can access overlapping time scales 24 . Such studies yield atomic-detailed insights into pathways of folding 25 . The fold of spidroin NTDs is stabilized by an extensive hydrophobic core (Fig. 6a ). Observed similarity of stabilities may be explained by the high degree of conservation of residue side chains that form the core: 15 of 24 core side chains are identical and the remaining 9 are of high similarity (Fig. 6b ). This raises the question as to the origin of reduced stability of the Nc homologue compared with its family members (Tables 3 and 4 ). The subtle rearrangement of helices found in structural studies 20 and the resulting change of the tertiary interaction network may explain the finding. Figure 6 Structure and sequence alignment of MaSp1 NTDs. ( a ) Solution structure of the monomeric Ea NTD (pdb id 2LPJ). Amino acid side chains that form the hydrophobic core are shown as van-der-Waals spheres (black). Acidic and basic side chains are highlighted in red and blue stick representation. The C-terminus of the domain is indicated. The panel on the right hand side shows a top view on the domain highlighting the conserved N-terminal Trp buried in core position (W10, cyan spheres). ( b ) Sequence alignment of Ea, Nc, Lh, and Lg NTD (from top to bottom). Acidic and basic side chains are highlighted red and blue. Side chains that form the hydrophobic core are highlighted black and bold. Identical (*), very similar (:) and similar (.) side chains are indicated at the bottom. \n The number of ionisable side chains in NTDs is low compared with the average content of 29% found in proteins 26 (Fig. 6 ). The few side chain charges, however, fulfil critical roles in the mechanism of pH-triggered self-association. Basic and acidic residue side chains cluster on opposing poles of the domain generating a macromolecular dipole that steers anti-parallel orientation in the dimeric assembly 8 . Some specific acidic side chains are involved in the pH-relay mechanism that locks the dimer 8 , 11 , 12 , 20 . Clustering of point charges, however, compromises stability of the domains through intramolecular, repulsive electrostatic forces. There is thus a trade-off of stability of the monomeric fold versus stability of the dimeric assembly, both modulated by electrostatics. Whilst high concentrations of salt stabilize the monomeric fold, they destabilizes the dimeric assembly. Decreasing concentrations of sodium chloride found along the spinning duct towards the tapering end where the fibre is formed 27 thus stabilizes the dimer. Salt ions in solution can alleviate intramolecular electrostatic strain either through direct binding or by the effect of Debye-Hückel screening. The latter is explained by the formation of clouds of counter-ions around point charges that screen Coulombic forces. This screening is related to the thickness of the ion cloud, which depends on the square root of I \n 18 , 19 . We found a linear dependence of the free energy of folding on the square root of I for all NTDs and thus identified Debye-Hückel screening as the mechanism of salt action on this domain. Although the Nc homologue exhibits the highest number of side chain charges (Fig. 6b ) additional electrostatic strain appears not to be the origin of the reduced stability compared with the other homologues: m \n eq ’ of all four domains was similar. The mean m \n eq ’ of 3.5 ± 0.1 kcal mol −1 M −0.5 was higher than the value reported for the small protein FynSH3 ( m \n eq ’ = 2.90 ± 0.07 kcal mol −1 M −0.5 ) 19 . The higher m \n eq ’ of NTDs likely arises from the highly localized clustering of point charges that steers anti-parallel orientation during dimerization at the expense of a reduced stability of the fold. In conclusion, our comparative study of folding and association within a family of MaSp1 NTDs reveals kinetics and energetics of self-assembly that are conserved across species. Small size, unusual sequence properties and high speed of folding and binding suggests the NTD family as an interesting system for future combined experimental and computational studies that may elucidate mechanisms of conformational change and self-assembly at atomic detail. Moreover, it will be interesting to see if conservation of folding and association also holds for spidroin NTDs originating from glands other than those used to spin dragline silk."
} | 3,299 |
37669363 | PMC10500273 | pmc | 5,861 | {
"abstract": "Significance Microbial communities are found throughout the biosphere, from human guts to glaciers, from soil to activated sludge. Understanding the statistical properties of such diverse communities can pave the way to elucidate the common mechanisms behind their patterns of variability, stability, and resilience. In particular, shedding light on how bacteria correlate as a function of their genetic similarity is extremely relevant both at fundamental and practical levels. Using data from natural communities and mathematical modeling, we identify a macroecological law relating mean pairwise correlation with genetic similarity, revealing that correlation goes from positive to null values as species dissimilarity increases. Fluctuations of shared environmental factors, such as temperature or resources, are responsible for such a universal pattern.",
"discussion": "Discussion We have considered both cross-sectional (across communities) and longitudinal (across time) empirical data for the species abundances in microbial communities from many different environments and studied their species-abundance pairwise correlations as a function of pairwise phylogenetic distance, revealing the emergence of an universal macroecological law. This empirical law states in quantitative terms that the average correlation function decays from positive to null values as the phylogenetic distance (or dissimilarity) increases, approximately following a stretched-exponential decay function. We explored the possible ecological forces shaping species correlations from a theoretical standpoint. In particular, by scrutinizing different ecological models, each one implementing a diverse set of ecological forces between species, we found that the universal correlation pattern cannot possibly be reproduced by competition or exclusion principles. Instead, temporal environmental filtering—i.e., the presence of correlated noise stemming from shared fluctuating factors—as modeled by a correlated stochastic-logistic model (CSLM), explains quantitatively empirical data. Furthermore, time-dependent (delayed) correlations in longitudinal data are also well reproduced by the model. The ecological pattern identified in this paper gives a quantification at the level of phylogenetic signals detectable in taxa–taxa abundance correlation. The pattern, as also shown in SI Appendix , Figs. S5–S7 , does not recapitulate the full range of correlations observed in natural communities. In this context, our work complements the research aiming at inferring ecological interactions from correlations, by showing how phylogenetic similarity can be used to disentangle the effects of environmental fluctuations and interactions (such as, e.g., competition). These results are based on multiple assumptions and their limitations give opportunities for extensions of the current work. First, at a theoretical level, the CSLM reproduces the average correlation at each discrete phylogenetic distance, but not the full distribution around such a mean value ( SI Appendix , Fig. S33 ). This is because, to be able to connect genetic and preference similarities, we enforced a “mean-field” type of relationship, Eq. 12 , neglecting variability across pairs of species in the phenotypic-distance-to-preference-distance mapping. On the other hand, in SI Appendix , Fig. S5 , we show that the variance of the distribution of the empirically measured pairwise correlations within each distance bin seems to follow a weak decaying power-law pattern with phylogenetic distance, with a diverse decaying exponent characteristic for each analyzed biome. Possibly, these patterns could be used to generate the preference vectors of the model in a more general way, allowing for more variability. Empirical data are not informative enough at the moment to proceed in this direction, and further analyses are required. It is however important to stress that both the empirical analysis and the model assume a certain degree of niche conservatism. One important assumption of our modeling framework is that ecological similarities are fixed in time and environmentally dependent ( 48 , 49 ). In the extreme scenario, in which the ecological strategy is strongly conserved on the phylogenetic tree there would be a 1 : 1 mapping between ecological similarity and phylogenetic distance. This strong assumption is however not needed for our analysis, which requires of a much weaker condition: namely, that ecological similarity correlates with phylogenetic similarity. The variability of correlations around the expected one from phylogenetic distance (shown in SI Appendix , Fig. S33 ) should be interpreted in this way. Note that two interpretations of our results are possible. On the most pessimistic side, one could argue that the pattern we identify and the model we propose serve only to describe the phylogenetic signal observed in the correlations, leaving the variation unexplained. Instead, on the most optimistic side, one could argue that the variability observed in the correlations is not a signal of other ecological mechanisms not included in the model but rather the consequence of the lack of a perfect match between preference similarity and phylogenetic similarity. Recent theoretical works, e.g., in the context of consumer-resource models ( 50 ) explored the case of dynamic ecological preferences, where species’ preferences are dynamically optimized given an environment. One could envision extensions of our model including dynamical preferences. In fact, these changes in ecological strategies might contribute to the large variation observed around the phylogenetic trend by they should be constrained by the robust pattern of mean correlations reported here. It is also important to stress that the origin of the stretched exponential behavior and, in particular, its exponent value close to a value 1 / 3 in the universal pattern of correlations (i.e., Eq. 1 ) remains unexplained. This type of scaling could be influenced by the scale-invariant, i.e., fractal, structure of phylogenetic trees ( 51 – 54 ). Further investigations, beyond the scope of the present work, are needed to shed light onto this empirical finding. Furthermore, it is known that a vast class of competitive models can lead to species clustering in trait space ( 55 , 56 ). Even if such models produce an “oscillating” pattern of positive and negative correlation, and hence are not sufficient to explain the behavior here reported, their possible extension could be relevant for explaining the phylogenetic distance distribution observed in data ( SI Appendix , Fig. S1 ). Although environmental filtering has been found to dominate the pattern of species-abundance correlations, the above-mentioned variability could be the result of the complex interplay of other ecological forces. To identify which further forces are relevant and to discriminate their effects, it will be important to analyze time-dependent data in a more detailed way as well as to analyze differences in carrying capacities and correlations between different hosts ( 27 ). Furthermore, an exhaustive analysis of the variations of the correlation pattern across environments and phyla is also needed. Interestingly, SI Appendix , Figs. S8–S10 show that some phyla (e.g., Bacteroidetes) follow robustly the pattern, while some others, such as Actinobacteria, exhibit wild fluctuations. Indeed, the non-monotonic deviation in the soil biome around distance 0.1 seems to be caused by the actinobacteria phylum and, in particular, by the Actinomycetales and Gaiellales orders ( SI Appendix , Fig. S9 ). The fact that the trend of correlation and phylogeny holds across very different environments strongly suggests that the pattern captures an underlying general ecological process, linking phylogeny with ecological similarity and ecological similarity with correlations. Specific environments and specific taxa might have different behaviors, which is reflected in the deviations from the average patterns and in the variability of the fitted parameters of the stretched-exponential. We leave for future work the promising study of deviations across taxa, that could reveal more information on additional interactions responsible for the observed residual correlations. The general decay pattern of correlations with phylogenetic distance implies a quite universal value of the typical distance above which taxa are on average decorrelated. This scale (determined by the parameter λ ) corresponds roughly to the one of different families, and it is conserved across environments, suggesting that its origin is a consequence of a general biological mechanism. The value of λ could descend from the scale of ecological dissimilarity at which species fluctuations become on average not correlated. Alternatively, the scale λ could derive from the phylogenetic scale at which the signal of ecological similarity disappears. Supporting one of these alternatives would require identifying the proper variables to infer ecological similarity. Another relevant caveat is that our analyses here are limited to the taxonomic resolution of OTUs, clustering together individuals with more than 97 % similarity. Recent results suggest that ecological dynamics starts to decouple at much finer phylogenetic resolutions ( 57 ). Moreover, strains seem to still obey the three macroecological laws of variation and diversity valid at species level ( 58 ). These results leave open the question of how ecological forces shape the variation of community composition at finer phylogenetic scales. On the other hand, from a complementary viewpoint, we analyzed the behavior of correlations at the coarse-grained resolution of phyla. In particular, SI Appendix , Fig. S11 illustrates that by considering just interphyla correlations, one cannot observe the stretched exponential decay, that is determined by intraphyla OTU pairs. Analogously, by extending our analyses to finer phylogenetic resolutions, it could be possible to reveal the nature of intraspecific interactions, eventually elucidating the emergence of competition as a key player in determining correlations. Actually, in our view, one should not fix a characteristic taxonomic resolution to have a complete description of complex communities, but, instead, start from individuals (or functional units) and progressively cluster them together at larger and larger coarse-grained scales, i.e., moving across observational scales as customarily done in physics using “renormalization group” tools in statistical physics ( 59 , 60 ) as different ecological forces may shape communities at diverse resolution levels ( 61 )."
} | 2,677 |
30356700 | PMC6189367 | pmc | 5,862 | {
"abstract": "Methane is a potent greenhouse gas, 25 times more efficient at trapping heat than carbon dioxide. Ruminant methane emissions contribute almost 30% to anthropogenic sources of global atmospheric methane levels and a reduction in methane emissions would significantly contribute to slowing global temperature rises. Here we demonstrate the use of a lytic enyzme, PeiR, from a methanogen virus that infects Methanobrevibacter ruminantium M1 as an effective agent inhibiting a range of rumen methanogen strains in pure culture. We determined the substrate specificity of soluble PeiR and demonstrated that the enzyme is capable of hydrolysing the pseudomurein cell walls of methanogens. Subsequently, peiR was fused to the polyhydroxyalkanoate (PHA) synthase gene phaC and displayed on the surface of PHA bionanoparticles (BNPs) expressed in Eschericia coli via one-step biosynthesis. These tailored BNPs were capable of lysing not only the original methanogen host strain, but a wide range of other rumen methanogen strains in vitro. Methane production was reduced by up to 97% for 5 days post-inoculation in the in vitro assay. We propose that tailored BNPs carrying anti-methanogen enzymes represent a new class of methane inhibitors. Tailored BNPs can be rapidly developed and may be able to modulate the methanogen community in vivo with the aim to lower ruminant methane emissions without impacting animal productivity.",
"conclusion": "Conclusion In this initial proof-of-concept in vitro work, we have shown that the lytic enzyme of a methanogen integrated provirus, PeiR, is biologically active both as a free enzyme and immobilized on the surface of PHA BNPs. PhaC-PeiR tailored BNPs were capable of inhibiting an exceptionally broad range of different rumen methanogen strains in pure culture, while significantly reducing methane production for several days ( Supplementary Text 2 ). The current limitation of this study, using pure cultures, will be overcome in the future by validating PeiR-BNPs in increasingly complex rumen simulations, such as batch and fed-batch rumen fermenters, to deliver convincing evidence of efficacy and to optimize BNP manufacture processes before moving to animal experiments. Once suitable anti-methanogen enzymes have been selected, the development cycle of new anti-methanogen BNPs can be measured in weeks. This is of particular importance in engineering the next generation of tailored beads that may feature multiple proteins and enzymes that inhibit or specifically bind to rumen methanogens. The one-step biological synthesis of tailored beads and the economical purification are attractive protocols for future commercialization.",
"introduction": "Introduction Agriculture contributes between 3 and 4% to the global gross domestic product (GDP) 1 . A growing world population dictates an increasing demand on food suppliers and both crop and livestock production indices have been steadily rising. Such an intensification in production comes at the cost of a growing carbon footprint. Agricultural emissions are a recognized contribution to anthropological climate change and emissions continue to grow annually ( Tubiello et al., 2015 ). In particular, meat, wool, and dairy production relies on ruminants which annually produce ∼80,000,000 metric tons of methane, contributing almost 30% of global methane emissions 2 . Methane in ruminants is produced predominantly in the forestomach, commonly known as the rumen, as part of ruminal feed by microbes which creates nutrients and metabolites for the ruminants. One of the major by-products of this fermentation process is hydrogen and a specialized group of archaea, known as methanogens, remove free hydrogen through the reduction of carbon dioxide to methane. Previous research has shown that methanogens can be inhibited in the rumen using chlorinated inhibitors such as bromochloromethane, where a 30% decrease in methane production was observed ( Denman et al., 2007 ; Bagi et al., 2017 ). However, the use of such compounds is prohibited in food production and other avenues must be found to effectively reduce ruminant methane emissions. More recently, the non-chlorinated inhibitor 3-nitrooxypropanol (3-NOP) has been tested in cows where it successfully reduced methane emissions ( Hristov et al., 2015 ). 3-NOP is metabolized by methanogens and the inhibited methyl-coenzyme M reductase (MCR) enzyme activity is restored over time ( Hristov et al., 2015 ; Duin et al., 2016 ). A potentially limiting factor of 3-NOP is the varying susceptibility of different methanogen groups with as much as a 100-fold variation in growth inhibition ( Ragsdale, 2016 ). More detailed information on 3-NOP and other small molecule inhibitors can be found in the comprehensive review article published by Patra et al. (2017) . Changing feed and feed composition has also been shown to reduce methane emissions. For example, exclusively feeding rape over 15 weeks resulted in a reduction in methane emissions by up to 30% in sheep, whereas feeding chicory or white clover lead to inconsistent results when compared to diets based on rye grass ( Sun et al., 2015 ). However, rape may comprise a much higher nitrogen and sulfur content than rye grass, possibly leading to elevated nitrous oxide emissions of the ruminants through fecal matter, depending on seasonal variations and growth conditions. Alternative feed additives, such as tannins, oils or fats may reduce methane emissions by up to 18 – 20% but must be carefully monitored to avoid negative health effects due to sudden introduction or over ingestion and availability may be limited or seasonal ( Liu et al., 2011 ; Kolling et al., 2018 ; Wang et al., 2018 ). Another promising feed additive is the seaweed Asparagopsis that has been shown to reduce methane emissions by up to 80% ( Machado et al., 2014 ). However, the active compound bromoform, is a brominated organic solvent and known animal carcinogen. In all cases, changing animal diets or adding supplements needs to be carefully assessed with regards to seasonal availability, total emission balances and possible toxicity to the animals and humans. For a comprehensive review on nutritional strategies in ruminants and their impacts on methane emissions, please refer to these review articles ( Buddle et al., 2011 ; Hristov et al., 2013 ; Hatew, 2015 ; McGrath et al., 2018 ). Alternative methane reduction strategies require effective, cost-efficient and non-toxic (environmentally friendly) mechanisms that specifically target rumen methanogen cells without negatively impacting on the microbial plant fiber degradation or on animal production parameters. One such strategy, phage therapy, has been used in biomedical applications since 1920 ( McAuliffe et al., 2007 ). However, phage therapy using intact virions is host-strain specific and the production of virus tail-like structures or even purified enzymes is costly and complex. Further, the cell wall makeup of some methanogenic archaea, in particular within the order Methanobacteriales, differs significantly to the peptidoglycan of bacteria ( Kandler and König, 1993 ) ( Supplementary Figure 5 ). Pseudomurein, the major compound of the cell wall of members of Methanobacteriales is composed of N -acetyl- L -talosaminuronic acid and N -acetyl-D-glucosamine connected through β(1–3) glycosidic linkages. These chemical differences make pseudomurein inert to known bacterial cell wall hydrolases and to lysozyme ( Kandler and König, 1993 ; Visweswaran et al., 2010 ). To date, only a few methanogen viruses have been described in some detail ( Luo et al., 2001 , 2002 ; Krupovic and Bamford, 2008 ; Chien et al., 2013 ; Weidenbach et al., 2017 ) (for an excellent review on archaeal viruses in general refer to Krupovic ( Krupovic et al., 2018 ), while more insights are emerging through the analysis of rumen (viral) metagenomes ( Berg Miller et al., 2012 ; Ross et al., 2013 ; Yutin et al., 2015 ; Anderson et al., 2017 ). Similarly, only two methanogen phage lytic enzymes have been described in detail previously; PeiP of the Methanothermobacter marburgensis phage ψM2 ( Pfister et al., 1998 ) and PeiW of the Methanothermobacter wolfeii phage ψM100 ( Luo et al., 2001 ). Both are capable of hydrolysing the ε-(alanine)-lysine bond of the peptide linker between layers of pseudomurein ( Luo et al., 2002 ). Sequence analyses carried out for PeiP and PeiW highlighted that both enzymes contain four N-terminal pseudomurein binding repeat (PMBR) domains, essential for cell wall binding ( Visweswaran et al., 2010 ), and a C-terminal catalytic domain ( Visweswaran et al., 2010 ), which is in opposite orientation to the structural makeup of lysins of Gram-positive bacteria. Both PeiP and PeiW share a significant level of sequence similarity to each other (50% identity, 62% positives) with individual domains being particularly highly conserved. Genome sequencing the rumen methanogen Methanobrevibacter ruminantium M1, a third methanogen lysin was discovered, PeiR of the Methanobrevibacterium prophage φmru ( Leahy et al., 2010 ). To our knowledge, only a single rumen methanogen provirus has been described in detail, Methanobrevibacter ruminantium M1 virus Φmru and evidence was provided that the lytic enzyme PeiR was active against the host strain in pure culture ( Leahy et al., 2010 ). A novel delivery and enzyme production system in the form of polyhydroxyalkanoate (PHA) bionanoparticles (BNPs) has been developed ( Peters and Rehm, 2005 , 2006 ). A number of microbes are capable of synthesizing PHA as carbon and energy reservoirs ( Keshavarz and Roy, 2010 ). The key enzyme for PHA biosynthesis is the PHA synthase PhaC ( Yuan et al., 2001 ) which forms the polyester by linking (R) -3-hydroxyacyl-CoA thioester. The resulting biopolymer remains covalently bound to PhaC. The hydrophobic polyester strands aggregate into spherical BNPs with the strands in the core and attached proteins forming the surface. BNPs have proven to be effective, cost-efficient ( Choi and Lee, 1997 ) and non-toxic in previous studies and can make up to 80–90% of the cell’s dry weight ( Philip et al., 2007 ). These structures currently receive significant biotechnological interest in a wide range of applications, including agriculture ( Keshavarz and Roy, 2010 ). Polyhydroxyalkanoate (PHA), the main component of one type of BNPs produced by recombinant cell factories, has gained FDA approval as a food additive ( Philip et al., 2007 ). It has also been shown that the key enzyme, PhaC, tolerates protein fusions to both its C- and N-termini and the resulting tailored BNPs have the unique ability to display proteins and enzymes on their surface in an oriented fashion while retaining their respective native enzyme activities. For a comprehensive review on tailored BNPs and their applications refer to Rehm ( Rehm, 2010 ). In the proof-of-concept work presented here we investigated the predicted domain structure of the lytic enzyme PeiR from Methanobrevibacter ruminantium M1 and determined the enzyme characterisitcs of the free enzyme to understand its relation to other, known archaeal lytic enzymes, and its substrate specificity. To evaluate the potential of tailored BNPs active against rumen methanogen strains in vitro , we created fusion proteins and displayed PeiR on the surface of BNPs. We demonstrated that PeiR-tailored BNPs were active against a wide range of rumen methanogen strains in pure culture and provided effective growth and methane inhibition for several days. Combining phage therapy with the accessibility and scalability of tailored BNP production has resulted in a conceptually new approach to mitigate ruminant methane emissions that may also be used in many other applications.",
"discussion": "Discussion The first lytic enzyme tested as an effective anti-methanogen agent was PeiR from the M. ruminantium M1 integrated provirus φmru. The overall enzyme architecture of PeiP, PeiW and PeiR is similar, but the detailed makeup does differ significantly. It has been previously shown that a minimum of three PMBR domains were required for the surface (S)-layer protein MTH719 of Methanothermobacter thermautotrophicus to bind to the pseudomurein cell wall ( Visweswaran et al., 2011 ). Interestingly, the presence of only two PMBR domains in the active PeiR enzyme may point to a more effective binding mechanism than previously known. Compared to the known methanogen lytic enzymes PeiW and PeiP, we propose that PeiR represents a new type of archaeal virus endopeptidase. The presence of only two novel PMBR domains, in combination with a clan CA protease, is sufficient for effective cell wall binding and subsequent cell lysis of the host strain M1 ( Figure 3 ). Despite the difference in peptidase domains, the biological activity of PeiR appears similar to that reported for PeiW and PeiP which cleave the ε-isopeptide bond between alanine and lysine ( Visweswaran et al., 2010 ). In M1, the alanine is replaced by threonine in the pseudomurein peptide side chain and PeiR is able to hydrolyse a synthetic substrate that mimics the glutamate-threonine peptide bond. No activity was detected on other substrates where threonine is replaced. The free PeiR enzyme was active against a much wider range of methanogens than implied by the synthetic substrates. It is noteworthy that PeiR is active against cell walls of Methanothermobacter thermautotrophicus – which is sensitive to PeiW and PeiP - at the same level as tested Methanobrevibacter spp. PeiR was also active against Methanosphaera stadtmanae which contains serine in place of alanine in the pseudomurein cell wall ( Biavati et al., 1988 ), while the corresponding synthetic substrate was not recognized. In contrast to PeiW and PeiP, PeiR does not require divalent metal ions for activity and is not inhibited when treated with EDTA ( Luo et al., 2002 ; Schofield et al., 2015 ). At present, the seemingly wide range of biological activity of PeiR free enzyme cannot be mechanistically explained and further structural and biochemical studies are required to unravel the possible recognition sites and enzymatic mechanisms on native methanogen cell substrates. Industrial production of rumen methanogen inhibitors ideally comprises a simple in situ production process that does not require any complex and expensive post-processing procedures. Microbial cells produce tailored PHA BNPs displaying functional proteins in a one-step process that does not require any physical or biochemical modification beyond a basic cell disruption to free the BNPs ( Draper and Rehm, 2012 ). This implies cost effective large-scale production of individual BNP-types and offers an attractive system for initial in vitro models and subsequent upscale production. The enzyme that polymerises (R) -3-hydroxyacyl-CoA thioester monomers into polyester, PHA synthase PhaC, tolerates both N- and C-terminal fusions ( Grage et al., 2009 ) and a C-terminal fusion between PhaC and PeiR was shown to be biologically active. The effectiveness of a BNP-based anti-methanogen product may also be influenced by the range of sensitive rumen methanogens. Both the free lytic enzyme PeiR and the enzyme immobilized on PHA BNPs exhibited a remarkable versatility against a wide range of different methanogen strains. Effective inhibition and reduction in methane was achieved for Methanobrevibacter spp., Methanobacterium formicicum and a Methanosphaera sp. All four Methanobrevibacter spp. tested were sensitive to PhaC-PeiR BNPs, albeit with a graduated response. This decrease in sensitivity continued for strains A4 and BRM9. A published phylogenetic framework for rumen methanogens ( Jeyanathan et al., 2011 ) inferred relationships between type strains of the methanogen clades investigated here. The lytic enzyme PeiR was identified in M1 and provided the most potent effect on its host strain. Selecting strain M1 as reference point in the inferred phylogenetic tree, the level and duration of PhaC-PeiR inhibition correlated well with the respective phylogenetic distance, making a compelling argument for an increasingly different cell wall makeup that even within the same genus contributes to varying phenotypes. Interestingly, the presence of cells of BRM9 led to aggregate formation of PhaC and PhaC-PeiR BNPs. These aggregates resembled the rod-like cell morphology of BRM9, suggesting that BNPs may attach to an extracellular matrix produced by BRM9 cells. Veiga et al. (1997) reported that Mb. formicicum is able to produce extracellular polymers (ECP) composed of polysaccharides and polypeptides that play an important function in granule formation and cell-to-cell adhesion. The production of ECP by rumen methanogens has not been reported previously and it is tempting to speculate that such an interaction may provide an additional mechanism of methanogen inhibition by interference with ECP-mediated cell-to-cell adhesion or communication between methanogens and/or other syntrophic bacteria. The use of lytic enzymes as antimicrobial agents has been described in other fields and is a particularly promising alternative to the use of antibiotics. A comprehensive assay on the development of the lysin CF-301, active against Staphylococcus aureus (MRSA), toward a novel therapeutic class has been recently published by Fischetti (2018) . While lysins and related enzymes may successfully inhibit rumen methanogens today, they still face the same problem of emerging microbial resistance. However, their modular makeup enables a unique approach by creating chimeric enzymes with different modes of action, thereby bypassing emerging resistances ( Manoharadas et al., 2009 ; Fernandes et al., 2012 ; Mao et al., 2013 ). In fact, libraries of lytic enzymes and their chimera could be established and displayed on BNPs, ready to replace current variations when resistance is detected. The short development cycle for creating new lytic-enzyme displaying BNP variants represents one of the major advantages of this technology and makes it much more flexible and future proof than other approaches that rely on a single compound and/or require much longer development cycles. PhaC-PeiR BNPs offer effective methanogen inhibition for up to 5 days post-bead addition in pure culture. The average turnover of the rumen is between 7 h for liquids ( Evans, 1981 ) and 14 h for solid particles ( Owens et al., 1979 ). By inhibiting methanogens beyond this retention time, it is likely that the methanogen population may be washed out of the rumen, creating a longer lasting methane inhibition effect than can be measured in a static system. In combination with a continuous dose delivery system, recurrence of rumen methanogens may be prevented for extended periods of time. While the tailored BNP platform holds much promise, there are also some known bottlenecks that need to be addressed before the technology can be applied in large scale and on-farm. The current two-plasmid system requires the application of two different antibiotics and is limiting scalability to fed-batch fermentations. Similarly, the current production host E. coli may not be the optimal vehicle for large scale fermentation processes, due to its requirements for oxygenation. Because the bacterial host is a genetically modified organism (GMO), application on-farm will require the removal of host cell material through separation processes."
} | 4,877 |
24847883 | PMC4079543 | pmc | 5,863 | {
"abstract": "Summary Ancient and diverse antibiotic resistance genes (ARGs) have previously been identified from soil 1 – 3 , including genes identical to those in human pathogens 4 . Despite the apparent overlap between soil and clinical resistomes 4 – 6 , factors influencing ARG composition in soil and their movement between genomes and habitats remain largely unknown 3 . General metagenome functions often correlate with the underlying structure of bacterial communities 7 – 12 . However, ARGs are hypothesized to be highly mobile 4 , 5 , 13 , prompting speculation that resistomes may not correlate with phylogenetic signatures or ecological divisions 13 , 14 . To investigate these relationships, we performed functional metagenomic selections for resistance to 18 antibiotics from 18 agricultural and grassland soils. The 2895 ARGs we discovered were predominantly novel, and represent all major resistance mechanisms 15 . We demonstrate that distinct soil types harbor distinct resistomes, and that nitrogen fertilizer amendments strongly influenced soil ARG content. Resistome composition also correlated with microbial phylogenetic and taxonomic structure, both across and within soil types. Consistent with this strong correlation, mobility elements syntenic with ARGs were rare in soil compared to sequenced pathogens, suggesting that ARGs in the soil may not transfer between bacteria as readily as is observed in the clinic. Together, our results indicate that bacterial community composition is the primary determinant of soil ARG content, challenging previous hypotheses that horizontal gene transfer effectively decouples resistomes from phylogeny 13 , 14 ."
} | 415 |
32700346 | PMC11468650 | pmc | 5,864 | {
"abstract": "Abstract The possibility of structuring material at the nanoscale is essential to control light–matter interactions and therefore fabricate next‐generation paints and coatings. In this context, nature can serve not only as a source of inspiration for the design of such novel optical structures, but also as a primary source of materials. Here, some of the strategies used in nature to optimize light–matter interaction are reviewed and some of the recent progress in the production of optical materials made solely of plant‐derived building blocks is highlighted. In nature, nano‐ to micrometer‐sized structured materials made from biopolymers are at the origin of most of the light‐transport effects. How natural photonic systems manage light scattering and what can be learned from plants and animals to produce photonic materials from biopolymers are discussed. Tuning the light‐scattering properties via structural variations allows a wide range of appearances to be obtained, from whiteness to transparency, using the same renewable and biodegradable building blocks. Here, various transparent and white cellulose‐based materials produced so far are highlighted.",
"introduction": "Introduction Scattering is the optical effect at the basis of light propagation in non‐absorbing media. This phenomenon occurs when light encounters refractive index inhomogeneities that can be defined as scattering centers or, in the case of granular media, scattering particles. [ \n \n 1 \n , \n 2 \n , \n 3 \n , \n 4 \n , \n 5 \n \n ] At every scattering event, light deviates from its initial trajectory depending on the geometrical characteristics and the refractive index of the scatterer (parameters that are described by the scattering cross‐section). [ \n \n 1 \n , \n 5 \n \n ] \n Non‐absorbing homogeneous materials scatter (reflect) light only at their interfaces (surfaces). Therefore, their appearance is determined by the refractive index contrast between the external medium and the homogeneous system. Conversely, for inhomogeneous materials, light transport is a bulk phenomenon. Therefore, their appearance results from the interplay between: 1) the single‐scatterer properties as the scattering cross‐section; 2) the ensemble properties as the filling fraction (i.e., the volume percentage occupied by the scatterers), the average orientation of the scatterers, and the structure factor (i.e., the spatial organization of the scatterers). [ \n \n 6 \n , \n 7 \n \n ] \n The appearance of a disordered material (which is defined as system where the structure factor does not exhibit long‐range order) is directly determined by the optical thickness (OT). This parameter is defined as the ratio between the physical thickness of a medium and the transport mean free path, that is, the average distance over which light loses memory of its initial direction. In fact, light transport can undergo different regimes when propagating in disordered media: from ballistic propagation, where most of the intensity is transmitted in the same direction of the incoming beam, to multiple scattering, where the initial propagation direction is completely scrambled ( Figure \n 1 \n ). In the ballistic regime, at low OT, light propagation is almost unperturbed, resulting in a transparent appearance as for homogeneous media; in contrast in the multiple scattering, for high OT (>8), [ \n \n 8 \n , \n 9 \n \n ] the material is opaque white. Systems in the intermediate scattering regime exhibit interesting optical properties as they strongly affect the light propagation direction while showing high transmission values. This property is called haze. Figure 1 Illustration of different light‐propagation regimes in disordered media: from ballistic to multiple scattering. For simplicity, the transition between different regimes is depicted by varying the filling fraction of the system. Ballistic beam, specular reflection, and scattered light are represented by solid black, dotted black, and dotted gray lines, respectively. Line thickness qualitatively represents the difference in intensity between ballistic and scattered light in every regime. In this progress report article, we first describe the strategies used by nature to manage light propagation in biological systems to either produce white or transparent appearances. Then, we review the recently developed strategies for fabricating cellulose‐based optical materials to meet the demand for more sustainable, biocompatible products. Finally, an overview of the applications for bio‐inspired materials in both transparent, low‐scattering and opaque, high‐scattering systems is presented."
} | 1,151 |
21544267 | null | s2 | 5,867 | {
"abstract": "The chemistry of mussel adhesion has commanded the focus of much recent research activity on wet adhesion. By comparison, the equally critical adhesive processing by marine organisms has been little examined. Using a mussel-inspired coacervate formed by mixing a recombinant mussel adhesive protein (fp-151-RGD) with hyaluronic acid (HA), we have examined the nanostructure, viscosity, friction, and interfacial energy of fluid-fluid phase-separated coacervates using the surface forces apparatus and microscopic techniques. At mixing ratios of fp-151-RGD:HA resulting in marginal coacervation, the coacervates showed shear-thickening viscosity and no structure by cryo-transmission electron microscopy (cryo-TEM). However, at the mixing ratio producing maximum coacervation, the coacervate showed shear-thinning viscosity and a transition to a bicontinuous phase by cryo-TEM. The shear-thinning viscosity, high friction coefficient (>1.2), and low interfacial energy (<1 mJ m(-2)) observed at the optimal mixing ratio for coacervation are promising delivery, spreading and adhesion properties for future wet adhesive and coating technologies."
} | 285 |
32341341 | PMC7184735 | pmc | 5,869 | {
"abstract": "How complex, multi-component macromolecular machines evolved remains poorly understood. Here we reveal the evolutionary origins of the chemosensory machinery that controls flagellar motility in Escherichia coli . We first identify ancestral forms still present in Vibrio cholerae , Pseudomonas aeruginosa , Shewanella oneidensis and Methylomicrobium alcaliphilum , characterizing their structures by electron cryotomography and finding evidence that they function in a stress response pathway. Using bioinformatics, we trace the evolution of the system through γ-Proteobacteria, pinpointing key evolutionary events that led to the machine now seen in E. coli . Our results suggest that two ancient chemosensory systems with different inputs and outputs (F6 and F7) existed contemporaneously, with one (F7) ultimately taking over the inputs and outputs of the other (F6), which was subsequently lost.",
"introduction": "Introduction Cells are full of complex, multi-component macromolecular machines with amazingly sophisticated activities. In most cases, how these machines evolved remains mysterious. Presumably, they arose through a long series of small steps in which new components and functions accreted onto, or replaced, original ones. Throughout this process, each new function provided a fitness advantage and was thus retained. The chemosensory pathway in bacteria and archaea is one such multi-component system. It integrates environmental signals to control cellular functions ranging from flagellum- and pilus-mediated motility to biofilm formation. Also, chemosensory proteins are key virulence factors for many pathogens. The best-understood function of chemosensory systems is their control of the rotational bias of the flagellar motor, guiding bacteria toward attractants and away from repellents 1 , 2 . The molecular basis of this activity has been the object of intense study in Escherichia coli , where transmembrane methyl-accepting chemotaxis proteins, or MCPs, form large arrays at the cell pole 3 . These chemoreceptors bind attractants or repellents in the periplasm and relay signals to a histidine kinase (CheA) in the cytoplasm 4 . When activated, CheA first autophosphorylates and then transfers the phosphoryl group to the response regulators CheY and CheB, a methylesterase. Phosphorylated CheY binds to the flagellar motor, changing the direction of flagellar rotation. This allows the cells to switch from swimming forward smoothly (so-called runs) to tumbling randomly. Changes in the duration and frequency of run and tumble phases drive a biased random walk that moves the cells towards favorable environments 5 . The signal is terminated by a phosphatase, CheZ, that dephosphorylates free CheY 6 . Phosphorylated CheB tunes the sensitivity of the system by changing the methylation state of the chemoreceptors, opposing the constitutive activity of the methyltransferase CheR 7 , 8 . While the chemosensory system in E. coli is well understood, the structure and function of many others is not. Chemosensory systems have been classified on the basis of evolutionary history into 17 so-called flagellar classes (F1–17), one type IV pili class (TFP) and one class of alternative cellular functions (ACF) 9 . Because this classification system is based on phylogenomic analysis, the evolutionary relationship between the classes is generally known. Later, by analyzing chemosensory systems in archaeal genomes, we showed evidence that class F1 is the most ancient of the chemosensory classes 10 . Understanding this classification system and its evolution allows for a temporal directionality of the evolution and diversification of this system. However, the class names are not reliable predictors of biological role. In E. coli , the system that controls the flagellar motor is a member of the F7 class, but in many other bacteria this is not the case. Conversely, in Rhodospirillum centenum a member of the F9 class controls biosynthesis of flagella 11 . Historically, all these pathways have been called chemotaxis pathways in reference to their homology to the biological pathway that gives rise to the chemotaxis phenotype in a diverse set of organisms including E. coli . Here, we will refer to them instead as chemosensory pathways, to reflect the diversity of outputs that these pathways modulate in response to chemical cues in the environment. In previous work, we and others have used electron cryotomography (cryo-ET) to reveal the in situ macromolecular organization of several chemosensory systems 8 – 10 , 12 – 14 . This method allows the study of bacterial cells in a near-native state in three dimensions at macromolecular resolution. Cryo-ET revealed that all the chemosensory systems controlling flagellar motors that have been imaged so far, including the F6 systems of various γ-proteobacteria and the F7 system of E. coli , look very similar 12 . Here, imaging some of these same species under stress, we observed a new kind of chemosensory array. Surprisingly, we identify it as another form of F7, but with a remarkably different structural architecture compared to that of the canonical E. coli F7 system. Tracing its evolutionary history, we find that this unusual F7 system actually represents the ancestral form, which in a series of defined steps acquired both the input and output domains of the ancient F6 system to take over control of the flagellar motor, leading to the system seen in modern E. coli . The result is a fascinating example of the evolutionary repurposing of complex cellular machinery.",
"discussion": "Discussion Here, using a combination of cryo-ET and bioinformatics, we have characterized and dissected the evolution of the F7 chemosensory array in γ-proteobacteria. We find that the ancient F7 system, still present in nonenteric γ-Proteobacteria, took control of the flagellar motor from the F6 system in a series of clear evolutionary steps (Fig. 6 ). Thus, the well-studied chemosensory model system of E. coli is a chimera of two other, more widespread systems: the F6 flagellar-control system and an ancient F7 system of still-unknown biological function. These results provide a striking example of how evolution can repurpose macromolecular complexes for new functions. Fig. 6 Evolution of the F7 chemosensory array in non-enteric γ-proteobacteria. F7 chemosensory systems acquire F6-like ultrastructure and function. a Tomographic slices showing F6 and F7 stage 1 chemosensory arrays in the same P. aeruginosa cell (left) and an F7 stage 5 chemosensory array in E. coli (right). Over the course of evolution, the F6 system is lost and the F7 system evolves similar ultrastructure and function to the F6 system. Features to identify chemoreceptors are highlighted: periplasmic domain (PD), inner membrane (IM), and CheA/CheW layer (AW). b Molecular models of F7 chemosensory arrays in P. aeruginosa (left) and E. coli (right) built based on 64 , 65 . Proteins displayed in this representation are: CheA (A), CheB(B), CheR(R), CheD(D), F7-CheY (Y 7 ), F6-like CheY (Y 6 ), and F6-like CheZ (Z 6 ). Models are colored according to their hypothetical original class: F7 (red) and F6 (yellow). c Working model of the evolution of the F7 chemosensory system in γ-proteobacteria and β-proteobacteria. Scale bars are 50 nm. We identified four sequential evolutionary steps, each of which produced a stable, modern subtype of chemosensory array. In the step in which flagellar control moved from the F6 to the F7 system, two major evolutionary events were required: (i) the Aer2-like F7 receptor became Tar-like, swapping its input (Fig. 6B ); and (ii) the F7 CheA began signaling through the remaining F6 CheY, adding an output. We speculate that this receptor transformation may have occurred via a domain swap that replaced the multiple PAS-HAMP domains of an Aer2-like receptor with the sensor domain of an F6 Tar-like receptor. These changes were accompanied by eventual loss of the remaining F6 components, as well as F7 components no longer needed for its new function (Fig. 6C ). Thus, we hypothesize that intermediate stages of the F7 system, present in extant β-proteobacteria, retain both the older and younger functions. Together, these findings suggest several hypotheses on the biological role of individual components and the function of the F7 system in γ- and β-proteobacteria, see Supplementary Discussion ."
} | 2,102 |
39937908 | PMC11817925 | pmc | 5,870 | {
"abstract": "Color spiking encoding and opponent preprocessing are critical for energy-efficient object perception in the human visual system. Emulating the retina and brain’s integration of spatial and chromatic spiking signals holds promise for enhancing the efficiency of vision sensors. Here, we introduce an artificial visual neuron array that generates excitatory or inhibitory spiking responses to specific wavelengths with orientation selectivity. The neuron array can function as double-opponent receptive fields for spatial-chromatic opponent preprocessing to color signals, emulating the neural pathway from the retina to the cortex. With the color spiking preprocessing function of the neuron array, the recognition accuracy is improved almost twofold compared to direct perception of underexposure objects, and the noise robustness is also strengthened. This architecture leverages biological mechanisms for simultaneous spike encoding and antagonistic preprocessing of color information, offering the potential for highly efficient neuromorphic vision systems.",
"introduction": "INTRODUCTION Color vision, a vital complement to luminance information, is a fundamental component of visual perception ( 1 – 3 ). The human visual system achieves highly efficient color spiking encoding and preprocessing through hierarchical steps for feature extraction and redundancy discarding ( 4 – 7 ). Following light absorbance by photoreceptors, color signals are encoded through bipolar and ganglion cells with spiking dynamics in the retina. Then, the spikes are subsequently relayed to the brain for preprocessing in the lateral geniculate nucleus (LGN) and primary visual cortex (V1) ( 1 , 8 – 11 ). In V1, double-opponent (DO) cells form spatially and chromatically antagonistic receptive fields (RFs) that integrate and preprocess color signals. These RFs exhibit preferred spatial orientation and DO color opponency through coordinated color inhibition and excitation. This preprocessing is the physiological basis for color feature extraction, color constancy, and precise object perception in biological systems ( 12 , 13 ). However, contemporary artificial visual systems based on silicon capture color information through filters followed by analog-to-digital conversion, and they process it by sequential binary operations within memory and processing units. These separate sensing, conversion, and processing systems constrain computing efficiency and induce substantial energy consumption ( 5 , 14 ). These limitations have led to the development of artificial visual neurons for color edge computing, combining spiking encoding and preprocessing, and these artificial neurons are desired to be integrated into spiking neural networks (SNNs) for highly efficient artificial vision systems ( 15 – 18 ). Recent developments in advanced materials, hybrid structures, and integrated devices have focused on emulating the spiking encoding functions of the retina and interfacing them with SNNs, thereby enhancing perception efficiency and reducing the power consumption of machine vision systems ( 19 – 23 ). For instance, vertically stacked oxide films have been used for wavelength-dependent coding ( 24 ), integrated transistor-memristor systems have been applied for rate and time-to-first-spike coding ( 25 ), and the silicon neuron transistors by single transistor latch have been explored for intensity or color coding ( 20 ). Despite substantial advancements, most research has focused on optimizing spike-encoding dynamics of color information within the retina. The absence of preprocessing functionality still results in lower system efficiency, as large volumes of raw data are offloaded onto backend processing units. Progress in emulating the neural preprocessing, as performed by the LGN and V1, remains limited despite their critical roles in feature extraction and efficient object perception. These gaps underscore the need for further investigation into harnessing biological light encoding and preprocessing mechanisms within artificial visual systems. Here, we propose an artificial spiking neuron array for simultaneous color spike encoding and preprocessing. Each pixel in the array comprises a bidirectionally responsive synaptic phototransistor (BPR PT) and a threshold switching (TS) NbO x Mott neuron in series. Integrating the artificial synapse and neuron achieves a higher degree of biomimetic functionality of converting dual-band light into configurable excitatory or inhibitory spiking responses. This feature enables the array to operate as an oriented color DO RF, allowing for spatial and chromatic antagonistic preprocessing of near-infrared (NIR) and ultraviolet (UV) light inputs, similar to the mechanisms found in the LGN and visual cortex V1. This neuron array’s color spiking preprocessing improves recognition accuracy almost twofold compared to direct perception of underexposure objects, and the noise robustness is also strengthened. These achievements demonstrate a promising and biologically plausible neuromorphic hardware solution for low-power, efficient artificial vision systems.",
"discussion": "DISCUSSION We propose an artificial neuron array for simultaneous color spiking encoding and preprocessing by integrating the BPR PT and TS. Each neuron pixel can perform wavelength-dependent excitatory and inhibitory spiking encoding. Moreover, the neuron array can be designed to function as an artificial color DO RF, facilitating spatial and chromatic preprocessing in a manner analogous to the biological processing pathway from the LGN to the visual cortex V1. The color boundary feature can be extracted through preprocessing even in harsh lighting environments, and the SNN perception efficiency can be increased nearly twofold compared with the unprocessed counterpart. This method demonstrates the potential for simultaneous spiking encoding and wavelength preprocessing in artificial visual neurons, providing a foundational step toward developing highly efficient neuromorphic vision systems."
} | 1,504 |
36997724 | PMC10203304 | pmc | 5,871 | {
"abstract": "Photogranules are spherical aggregates formed of complex phototrophic ecosystems with potential for “aeration-free” wastewater treatment. Photogranules from a sequencing batch reactor were investigated by fluorescence microscopy, 16S/18S rRNA gene amplicon sequencing, microsensors, and stable- and radioisotope incubations to determine the granules’ composition, nutrient distribution, and light, carbon, and nitrogen budgets. The photogranules were biologically and chemically stratified, with filamentous cyanobacteria arranged in discrete layers and forming a scaffold to which other organisms were attached. Oxygen, nitrate, and light gradients were also detectable. Photosynthetic activity and nitrification were both predominantly restricted to the outer 500 µm, but while photosynthesis was relatively insensitive to the oxygen and nutrient (ammonium, phosphate, acetate) concentrations tested, nitrification was highly sensitive. Oxygen was cycled internally, with oxygen produced through photosynthesis rapidly consumed by aerobic respiration and nitrification. Oxygen production and consumption were well balanced. Similarly, nitrogen was cycled through paired nitrification and denitrification, and carbon was exchanged through photosynthesis and respiration. Our findings highlight that photogranules are complete, complex ecosystems with multiple linked nutrient cycles and will aid engineering decisions in photogranular wastewater treatment.",
"conclusion": "Conclusions Photogranules contain complete ecosystems in only a few cubic millimetres. A scaffold of motile filamentous cyanobacteria and EPS supports a diverse array of phototrophs and heterotrophs capable of nitrification, denitrification, photosynthesis, aerobic respiration, and phosphate uptake. This diverse community internally cycles at least three nutrients important in wastewater treatment: oxygen, nitrogen, and carbon. Oxygen from photosynthesis is immediately consumed by aerobic respiration and nitrification in the presence of acetate or ammonium, respectively. Nitrifiers convert ammonium to nitrate and nitrite, which is subsequently rapidly denitrified. Organic carbon fixed through photosynthesis is used to fuel denitrification after acetate is depleted. Conditions within the photogranule vary dramatically over time, from completely anoxic to oxygen concentrations three times above saturation, with even larger variation in nitrate and organic carbon availability. Nevertheless, photogranules maintain high, robust activity levels, supported by a dense network of interconnected cells that exchange nutrients. This dense network and nutrient exchange allow for denitrification in the absence of external organic carbon, small amounts of nitrification without the provision of external ammonium, and nitrification under nearly anoxic conditions. Photogranules thus offer an opportunity to utilize the sun’s energy to clean wastewater and decrease the energy requirement of conventional wastewater treatment, by reducing or eliminating the need to aerate treatment reactors.",
"introduction": "Introduction Wastewater treatment reactors continuously select for functional traits in their microbial communities. Manipulating operating conditions allows for the selection of desired traits, including conversion processes to purify water and self-aggregation of microbial biomass. Self-aggregation (i.e., forming biogranules, spherical aggregates of microorganisms) enables stratification (oxic and anoxic zones) and facilitates biomass harvesting because dense aggregates rapidly sink once reactor mixing is stopped. Generally, biogranules are formed in reactors where the liquid residence time is shorter than the doubling time of the microorganisms. This washes out suspended cells and generates selective pressure for biomass retention [ 1 , 2 ]. Although the aggregated biomass in biogranules is subject to mass transfer resistances that reduce the activity per cell, efficient biomass retention and the ensuing increased biomass assures strongly elevated volumetric conversion rates in biogranule reactors. Phototrophic biogranules, called photogranules, were first observed in cultures of photosynthetic mats from the North Sea [ 3 ]. The granules were composed of filamentous cyanobacteria, diatoms, and heterotrophic bacteria. Spherical geometry is rare in phototrophic communities, but some examples of photogranules in nature have been reported. For example, photogranules composed of cyanobacteria and heterotrophic bacteria called cryoconites are found in glaciers [ 4 , 5 ]. Green and pink microbial “berries” are found in salt marshes, formed through a symbiosis of cyanobacteria and diatoms, and of communities of sulfur-oxidizing purple sulfur bacteria and sulfur-reducing bacteria, respectively [ 6 , 7 ]. In phototrophic biofilms in nature, as in photogranules, concentration gradients of various dissolved chemical species (e.g., oxygen, substrates) are formed due to diffusional limitation. Similarly, light intensity gradients are formed by light absorption and scattering [ 8 ]. These intersecting gradients create a varied environment that can support the simultaneous growth of diverse microorganisms filling different niches, such as photoautotrophs, chemoautotrophs, and heterotrophs, exhibiting aerobic, and anaerobic metabolisms [ 9 ]. Complex interactions between these microbial groups take place, that may stabilize the functioning of the consortium [ 10 ]. For example, heterotrophs may grow on extracellular organic compounds excreted by phototrophs. The latter, in turn, may fix the inorganic CO 2 produced in heterotrophic growth. Aerobic chemoautotrophs, such as nitrifiers, may benefit from photosynthesis-enhanced oxygen levels while simultaneously competing with phototrophs for inorganic carbon and nitrogen. Unlike in biofilms, photogranules are free living, so their biomass is not limited to the surface area of their container. Furthermore, the surface to volume ratio of a sphere is such that for small spheres, there is capacity for a higher biomass-water exchange than in a plane of the same volume. This relationship holds so long as the radius of the sphere is less than one third of the thickness of the plane. Recently, photogranules were cultivated with selection pressure to remove and recover nitrogen, phosphorus, and carbon from wastewater [ 11 – 17 ]. This involved a regular exchange of the medium after settling of biomass, thus with a selection pressure towards formation of fast-settling granules. The granules exhibited indeed excellent settling properties, and their in-situ photosynthetic oxygen production fuelled oxygen-demanding microbial processes such as nitrification and respiration. This linked the O 2, and CO 2 cycles within the treatment process, making progress towards “aeration-free” wastewater treatment. Initial studies have investigated the physical structure and metabolic functions of single photogranules from wastewater, but were limited in focus to phototrophic organisms and oxygen profiles [ 18 – 20 ]. A modelling approach predicted the distribution of microorganisms and extracellular polymeric substances (EPS) within photogranules and the bulk turnover of chemical compounds and reactor functions based on varying nutrient inputs, but has not yet been tested in real photogranules [ 21 , 22 ]. Photogranules experience both varying external environmental conditions (i.e., light intensity, nutrient concentrations) during reactor operation, and internal conditions, as nutrient gradients are created and eliminated in response to microbial activity and external variation. Therefore, a full understanding of the microbial ecology within photogranules requires detailed investigation under various and varying conditions in vivo. Here, we studied the physical and biological stratification and functioning of these same photogranules with microscopic imaging, metataxonomics, microsensors, and incubations with radio- and stable-isotope labels. Our findings provide insight into the spatial and temporal distribution of functional activity (photosynthesis, nitrification, and denitrification) within photogranules and their dependency on external factors (light, nutrients). Further, the results can be used to support engineering decisions in photogranular wastewater treatment.",
"discussion": "Discussion Physical, biological, and structural features of photogranules The stratification pattern in photogranules closely resembled that of photosynthetic microbial mats ( Fig. 1 ). Motile filamentous cyanobacteria (e.g., Alkalinema pantanalense , Leptolyngbya boryana , Cephalothrix komarekiana and Limnothrix sp.) and excreted extracellular polymeric substances (EPS) generated a complex net of filaments that provided structural rigidity, similar to that observed in cyanobacterial mats [ 42 , 43 ]. The proliferation of microbial communities within biofilms is dependent on the EPS matrix [ 44 ]. The lectin stain revealed that a large fraction of glycoconjugates in the EPS matrix was attributed to d -glucose and d -mannose. The glucose constituents were likely produced by filamentous cyanobacteria, as it is the dominant monosaccharide in the glycoconjugate fraction of their EPS matrix [ 23 , 45 , 46 ]. Both glucose- (especially α(1–4) glucans) and mannose-containing polysaccharides have been shown to play a key role in biofilm cohesion in both aerobic granules and phototrophic microbial mats [ 47 , 48 ]. The lectin-specific BAN glycoconjugates reported show only a part of the total glycoconjugates present. Nevertheless, we anticipate that there are other types of glycoconjugates and matrix compounds present such as extracellular proteins and eDNA [ 49 ]. The importance of filamentous cyanobacteria in photogranule formation was highlighted previously [ 12 , 19 , 50 ]. It was proposed that initially, filamentous, and motile cyanobacteria form a nucleus of filaments (a bundle) that can harbour other organisms. As it grows this structure becomes more and more physically and biologically stratified and finally results in a photogranule, with a large diversity of microenvironments and associated microbial processes. In natural systems, as well as in photogranules, physical and biological stratification occurs according to the availability and gradients of light and nutrients [ 8 ]. Thus, while young and small photogranules (<0.5 mm) do not have a defined structure, photogranules grown to sizes of several millimetres show a clear stratification. Others have also observed microbial stratification in large (>2.5 mm) photogranules, with cyanobacteria forming a layer close to the surface [ 19 ]. In our study we found that independent of photogranule size, the filamentous cyanobacteria were present throughout the whole photogranule. However, they showed different arrangements of the filaments from surface to centre. While the cyanobacterial filaments were densely packed and jumbled at the surface and centre, in the area in between they aligned themselves radially. One explanation for this phenomenon is the competition for space within the photogranule, as radially arranging the filaments would facilitate movement between regions with higher exposure to light (phototaxis) and higher levels of other substrates (chemotaxis) [ 43 ]. This would be especially useful to “pierce” through the thick shell of non-phototrophic organisms in the first 500 µm of the photogranule, the region with the highest light availability. Such radial movement was previously observed in biogranules derived from subcultures of microbial mats composed of cyanobacteria ( Cephalotrix sp. formerly known as Phormidium sp.) diatoms and heterotrophic bacteria originating from the North Sea [ 3 ]. Their movement was shown to be triggered by light and substrates. Functional stratification in photogranules The photogranules also exhibited a functional stratification similar to phototrophic biofilms in nature. The photogranule surface was exposed to the highest light and substrate levels and supported the highest photosynthetic and nitrifying activities. In the presence of acetate, the photogranule harboured anoxic zones that allowed anaerobic processes such as denitrification to occur. Most of the microbial activity was concentrated in the outer 500 µm of the photogranule. In that region phototrophs (cyanobacteria and eukaryotic algae) attenuated about 90% of all incoming light and showed the highest photosynthetic activity at around 400–500 µm (max. 100 nmol/h). In addition, to the high concentration of cyanobacteria in the outer part of the photogranule, there was also a large population of non-phototrophic organisms, including nitrifiers ( Nitrosomonas sp., Nitrobacter sp. and Nitrospira sp.) and chemoheterotrophic bacteria/denitrifiers ( Thauera sp. and Zoogloea sp.). This high microbial density supported high oxygen production/consumption and nitrate production, as well as carbon fixation (Figs. 4 and 5 ). However, the 14 C incubations may underestimate the nitrification activity as they were performed in the dark (in the light carbon fixation by photosynthesis would be orders of magnitude higher than by nitrification, Figs. 4E , 5B ) and nitrification was much higher in the light (Fig. 6 ). Nitrate profiles within the granules did indicate nitrification within the centre of the granule, although the highest rates were confined to the outer edge (peak ~200 µm) of the granule (Fig. 5D ), which is slightly closer to the surface than in similar-sized cryoconites (peak ~400 µm) [ 5 ]. Similarly, the oxygen profiles in the absence of acetate but presence of ammonium also indicate some nitrification (i.e., oxygen consumption) in the centre of the granules, although again the highest rates were at the outer portion of the granule (Figs. 4 B and 5D ). Nevertheless, the nitrate accumulation observed in the profile (Fig. 5C ) was not caused by increased nitrification rates at lower depth but mostly due to the reducing rate of nitrate consumption towards the centre. The peak in nitrification just below the surface also aligns with a model of photogranules grown without DOC, but the same level of ammonium as in this study. Given that DOC is rapidly consumed in the bulk liquid, of the scenarios modelled, this scenario most closely fits our data [ 21 ]. Although the centre of the photogranule seemed to be relatively inactive, it may serve important functions. These could include substrate storage (e.g., as lipids, starch or EPS), fermentative processes, or decomposition of dead organic constituents, which all contribute to internal nutrient cycling in the photogranule. The centre has a lower oxygen concentration and may shelter less oxygen-tolerant microbes. Anaerobic prokaryotes such as Anaerolineaceae and Caldilineaceae and the fungi Trichosporon sp. found in the photogranule can ferment carbohydrates and mineralize organic phosphorus and nitrogen, comprised in EPS or necrotic biomass, to H 2 , alcohols (e.g., butanol, ethanol), ketones (e.g., acetone), PO 4 3− , and NH 4 − , which in turn can be reused within the photogranule [ 51 – 53 ]. Cyanobacteria are also adapted to anoxic conditions and can ferment six-carbon sugars (e.g., glucose) to e.g., lactate or ethanol [ 54 ]. Such internal nutrient cycling likely increases in significance the larger the photogranule becomes and may fuel metabolic processes despite external substrate limitation. Nutrient removal and microbial activity during a sequencing batch cycle The photosynthetic activity of the photogranule was insensitive to the short-term nutrient fluctuations typical of the wastewater batch process, although other microbial activities were substantially impacted. Combining our results with data collected from the wastewater reactor allowed us to build a timeline of the shifting nutrient limitations and metabolic processes that occurred over the course of a batch cycle (Fig. S9 ). At the beginning of a batch, nutrients such as ammonium (NH 4 + ), phosphate (PO 4 3− ), and acetate were available in high concentrations. In this phase, the photogranules could maintain high photosynthetic and heterotrophic activities, resulting in ammonium, phosphate, acetate, carbon dioxide, and oxygen consumption, and oxygen production. Only limited nitrification occurred during this phase, due to the low oxygen concentrations generated by aerobic respiration of acetate. Rapid growth of microorganisms during this phase fuelled ammonium and phosphorous uptake. As acetate concentrations decreased and oxygen concentrations increased, nitrification and subsequently denitrification kicked off. In the dark phase, the phototrophs switched from photosynthesis to respiration on internally stored photosynthates. Nitrification continued until all ammonium was converted to nitrate, even though nitrifiers made up only 2% of the total community. Since all acetate was already consumed, denitrification was fuelled by intracellularly stored carbon (e.g., as polyhydroxyalkanoates) [ 55 ] or internally recycled carbon by decomposition of organic matter [ 56 ] but was not able to completely remove all nitrate, due to carbon limitation. The 15 N incubations showed that denitrification can also occur under oxic conditions. Thus, denitrification was likely performed in part by the aerobic denitrifier Thauera sp. [ 57 ]. Further, the incubations showed that even in the absence of acetate the internally cycled organic carbon (e.g., organic carbon from phototrophs, or internally stored and recycled carbon) can sustain denitrification (Fig. 6A ). Compared to the maximum denitrification rate with addition of acetate in the dark, the denitrification rate sustained on internally cycled carbon was about 25x lower (Fig. 6B ). Nevertheless, these rates were sufficient to keep removing nitrogen from the bioreactor in the dark period. From ecology to application The fundamental knowledge of how photogranules are structured and function will allow engineers to replicate this novel ecosystem for wastewater treatment. In our study we investigated an active phototrophic, nitrifying, and denitrifying community that exhibited high nitrogen and carbon removal/conversion rates and showed internally linked processes (exchange of oxygen, nitrogen, and carbon dioxide). Introducing a phototrophic community to the wastewater treatment process will ultimately allow closure of the CO 2 and oxygen cycles of the conventional activated sludge process. Oxygen would be produced through photosynthesis and CO 2 by the heterotrophic conversion of organic matter, making an external supply of oxygen unnecessary. The oxygen production and consumption in the aggregate was calculated to be 13.6 mmol L −1 d −1 and 13.1 mmol L −1 d −1 , respectively (Equation S1 – S4 ), resulting in a net oxygen production. The oxygen produced by photosynthesis is quickly used by consumption processes, demonstrating closely coupled processes in the aggregate. In the beginning of the batch cycle when acetate was still present, the photogranule was oxygen limited, i.e., photosynthesis could not supply heterotrophic respiration and nitrification with sufficient oxygen to achieve maximal rates (Fig. 4A ). This was also apparent from the hourly oxygen production rate of 1.1 mmol L −1 h −1 and oxygen consumption rate of 1.4 mmol L −1 h −1 in the first 4 h of the cycle (considering full removal of acetate after 4 h into the cycle and high nitrification activity). After acetate was fully consumed, photosynthesis would likely become carbon limited without external CO 2 supply. To optimize oxygen demand versus oxygen production in a system without external oxygen or CO 2 supply, the light supply and specific acetate load could be altered. This will be especially important when wastewater characteristics (N, P, COD) or light conditions change. As stated before, most activity was in the edge (outer 500 µm) of the photogranule due to light penetration and nutrient diffusion limitation. The photogranules (2–4 mm) investigated in this study thereby contained a considerable relatively inactive zone, which reduced the overall conversion rate of an individual photogranule. Smaller sizes would optimize the conversion rate per individual photogranule and could consequently increase the maximal conversion rate of a photogranule treatment system. In previous studies, an ideal granule size was determined to be 1.25–1.5 mm for aerobic granules [ 58 , 59 ] and 0.5–1.7 mm for photogranules [ 19 ]. While the light and carbon load of the system influence photogranule size, the solid retention time (SRT) ultimately controls the size of photogranules as it provides an upper ceiling to the age of a photogranule. A shorter SRT would result in smaller granules and higher photogranule-specific conversion rates but might be detrimental for some slow growing organisms (e.g. nitrifiers) and compromise settleability of the photogranule if granule sizes become too small (<0.5 mm) [ 60 ]. Therefore, granule size must be carefully evaluated with all these different aspects in mind. Finally, our results indicate the need to change the operating conditions of the reactor to further increase nitrification rates. Nitrification was strongly controlled by the availability of oxygen, where rates were 3–4 times higher in the light than in the dark (Fig. 6A ). In the presence of acetate, nitrification is expected to be completely inhibited, as the entire photogranule was anoxic under this condition. Nevertheless, during reactor operation, photogranules always completely removed ammonium from the supernatant, both when the batch started with a light cycle and when it started with a dark cycle. Since acetate consumption takes ~4 h (Fig. S8 ) the conditions for optimal nitrification would only be available for ~2 h in a cycle starting with a light phase. In contrast, in a cycle starting with a dark phase that consumes all acetate, nitrification would be optimized for a full 6 h. This indicates that there is substantial unused nitrification capacity in the cycle starting with a dark phase. An additional, partial exchange of the reactor liquid could be supported at this time, to increase reactor activity. This would also allow for the denitrification of more of the converted nitrate, which is otherwise carbon-limited, as most nitrate is produced after the removal of the acetate (Figs. 6B , S8 ). Another option would be to pulse an external carbon source (e.g., methanol, acetate, or molasses) to promote carbon-limited denitrification, which is a common practice in conventional wastewater treatment plants [ 61 ]."
} | 5,683 |
21070012 | null | s2 | 5,872 | {
"abstract": "Scaffolded DNA origami has recently emerged as a versatile, programmable method to fold DNA into arbitrarily shaped nanostructures that are spatially addressable, with sub-10-nm resolution. Toward functional DNA nanotechnology, one of the key challenges is to integrate the bottom-up self-assembly of DNA origami with the top-down lithographic methods used to generate surface patterning. In this report we demonstrate that fixed length DNA origami nanotubes, modified with multiple thiol groups near both ends, can be used to connect surface patterned gold islands (tens of nanometers in diameter) fabricated by electron beam lithography (EBL). Atomic force microscopic imaging verified that the DNA origami nanotubes can be efficiently aligned between gold islands with various interisland distances and relative locations. This development represents progress toward the goal of bridging bottom-up and top-down assembly approaches."
} | 233 |
19661424 | null | s2 | 5,874 | {
"abstract": "We demonstrate the ability to engineer complex shapes that twist and curve at the nanoscale from DNA. Through programmable self-assembly, strands of DNA are directed to form a custom-shaped bundle of tightly cross-linked double helices, arrayed in parallel to their helical axes. Targeted insertions and deletions of base pairs cause the DNA bundles to develop twist of either handedness or to curve. The degree of curvature could be quantitatively controlled, and a radius of curvature as tight as 6 nanometers was achieved. We also combined multiple curved elements to build several different types of intricate nanostructures, such as a wireframe beach ball or square-toothed gears."
} | 171 |
37478925 | null | s2 | 5,876 | {
"abstract": "Successful phytoremediation of acidic metal-contaminated mine tailings requires amendments to condition tailings properties prior to plant establishment. This conditioning process is complex and includes multiple changes in tailings bio-physico-chemical properties. The objective of this project is to identify relationships between tailings properties, the soil microbiome, and plant stress response genes during growth of Atriplex lentiformis in compost-amended (10 %, 15 %, 20 % w/w) mine tailings. Analyses include RNA-Seq for plant root gene expression, 16S rRNA amplicon sequencing for bacterial/archaeal communities, metal concentrations in both tailings and plant organs, and phenotypic measures of plant stress. Zn accumulation in A. lentiformis leaves varied with compost levels and was the highest in the intermediate treatment (15 %, TC15). Microbial analysis identified Alicyclobacillus, Hydrotalea, and Pseudolabrys taxa with the highest relative abundance in TC15, and these taxa were strongly associated with Zn accumulation. Furthermore, we identified 190 root genes with significant gene expression changes. These root genes were associated with different pathways including, abscisic acid and auxin signaling, defense responses, ion channels, metal ion binding, oxidative stress, transcription regulation, and transmembrane transport. However, root gene expression changes were not driven by the increasing levels of compost. For example, there were 15 genes that were up-regulated in TC15, whereas 106 genes were down-regulated in TC15. The variables analyzed explained 86 % of the variance in Zn accumulation in A. lentiformis leaves. Importantly, Zn accumulation was driven by Zn shoot concentrations, leaf stress symptoms, plant root genes, and microbial taxa. Therefore, our results suggest there are strong plant-microbiome associations that drive Zn accumulation in A. lentiformis and different plant gene pathways are involved in alleviating varying levels of metal stress. Future work is needed to gain a mechanistic understanding of these plant-microbiome interactions to optimize phytoremediation strategies as they will govern the success or failure of the revegetation process."
} | 552 |
28595344 | PMC5850603 | pmc | 5,881 | {
"abstract": "Abstract How does metabolism influence social behavior? This fundamental question at the interface of molecular biology and social evolution is hard to address with experiments in animals, and therefore, we turned to a simple microbial system: swarming in the bacterium Pseudomonas aeruginosa . Using genetic engineering, we excised a locus encoding a key metabolic regulator and disrupted P. aeruginosa ’s metabolic prudence, the regulatory mechanism that controls expression of swarming public goods and protects this social behavior from exploitation by cheaters. Then, using experimental evolution, we followed the joint evolution of the genome, the metabolome and the social behavior as swarming re-evolved. New variants emerged spontaneously with mutations that reorganized the metabolome and compensated in distinct ways for the disrupted metabolic prudence. These experiments with a unicellular organism provide a detailed view of how metabolism—currency of all physiological processes—can determine the costs and benefits of a social behavior and ultimately influence how an organism behaves towards other organisms of the same species.",
"introduction": "Introduction Metabolism influences the way individuals behave toward others. In all species, from bacteria to animals including humans, social behavior appears to be a function of metabolic state ( Robinson et al. 2008 ; Biro and Stamps 2010 ; Boyle et al. 2015 ; Sih et al. 2015 ). Organisms tend to be more cooperative when their metabolic reserves are full. Vampire bats share blood with their starving roost-mates, but they share more when they have fed well ( Carter and Wilkinson 2013 ); bacteria send and receive more chemical quorum-sensing signals to one another when they have more intracellular metabolites ( Xavier and Bassler 2005 ); and judges give more favorable parole decisions when they resume their work after a meal ( Danziger et al. 2011 ). Why is it that—across the tree of life—metabolism seems to condition social behavior? A social behavior is any behavior that involves interactions among members of the same species and influences their reproduction and survival ( Robinson et al. 2008 ). For a social behavior to evolve by natural selection it must have a low cost-to-benefit ratio, where the fitness cost on the actor’s genotype takes into account the indirect fitness benefit given to the recipient ( Hamilton 1964 ; Nowak 2006 ; Gardner et al. 2011 ). For the same benefit to the recipient, natural selection should reduce the cost on the actor. Since metabolism is the currency of all physiological processes that support life ( Smith and Morowitz 2004 ) and a major determinant of behavioral cost ( Biro and Stamps 2010 ), natural selection should favor a regulation of social behavior that reduces metabolic burden on the actor. Here, we investigated this problem experimentally in a microbial model of social behavior: swarming motility in the bacterium Pseudomonas aeruginosa . In order to swarm, P. aeruginosa must synthesize and secrete massive amounts of rhamnolipid biosurfactants that make a thin lubricating film on which billions of bacterial cells rotating their flagella can move collectively ( Deziel et al. 2003 ; Caiazza et al. 2005 ). Swarming helps the colony forage more nutrients; in the laboratory, colonies of wild-type bacteria swarm to become 7 times more populous than colonies of nonswarming mutants lacking rhamnolipid synthesis such as the genetically engineered Δ rhlA mutant ( Xavier et al. 2011 ). Rhamnolipid secretion could be a significant metabolic burden because bacteria can produce >20% of their own biomass in these surfactants ( Guerra-Santos et al. 1984 ) that once secreted become a public good. Wild-type P. aeruginosa avoid wasting valuable metabolic resources, and potential exploitation by Δ rhlA cheaters, by expressing rhlA only when the colony is large enough and individuals have more carbon than they can possibly use for growth. This regulatory strategy is called metabolic prudence ( Xavier et al. 2011 ) and it stabilizes swarming against cheating because it enables bacteria to delay the production of the carbon-rich rhamnolipids to times when their growth is limited by lack of another essential nutrient, such as nitrogen or iron ( Mellbye and Schuster 2014 ). Regulation of rhlA integrates quorum-sensing signals with information on the metabolic state of the cell to turn on expression only when population density is high and there is an excess of carbon, and expression shuts off immediately when carbon is low ( Boyle et al. 2015 ). Metabolic prudence reduces the cost-to-benefit ratio of swarming, ensuring that this social behavior is protected from exploitation by nonrhamnolipid cheaters across a wide range of conditions ( de Vargas Roditi et al. 2013 ). We used the swarming system to conduct a perturbation experiment on organismal social behavior: We disrupted metabolic prudence by genetically engineering a deletion in the key metabolic-regulatory locus, cbrA . Then, we used metabolomics and genome sequencing to follow the concerted evolution of the metabolome, the genome and the social behavior as spontaneous mutants emerged to take over the population. Our experiments—to the best of our knowledge, the first to apply metabolomics to social evolution—revealed that in the absence of proper metabolic regulation the recovery of the original metabolomic state does not necessarily guarantee the best social behavior. More generally, experimental evolution in microbes provides a unique view on how metabolism plays a key role in the intricate feedback between genes and social behavior ( fig. 1 A ).\n Fig . 1. Swarming in Pseudomonas aeruginosa : A model of metabolism and social behavior. ( A ) Metabolism is the currency of all cellular processes, but its role in the feedback between genes and social behavior remains poorly studied. ( B ) Swarming in P. aeruginosa requires the production of rhamnolipid surfactants. Left: detail of the leading edge in a swarming colony with black arrow head indicating the edge of the colony and the white arrow head indicating the edge of the rhamnolipids produced by the bacteria—the public good required for swarming (scale 100 µm, see supplementary video S1, Supplementary Material online). Right: swarming in wild-type, lack of swarming in the strain Δ cbrA which we engineered to perturb the prudent regulation of rhamnolipid synthesis and the Δ rhlA strain which lacks rhamnolipid synthesis entirely (9 cm petri dishes shown). ( C ) Social behavior assays with the Δ rhlA nonswarmer reveal that Δ cbrA strain can produce biosurfactants that help Δ rhlA swarm.",
"discussion": "Discussion We showed that genetic perturbation of a key metabolic regulator, cbrA , profoundly altered the metabolome of P. aeruginosa affecting the evolution of its social behavior. The genetic excision executed in this catabolite repressor produced bacteria that grew slower on casamino acid media, expressed higher levels of public good rhamnolipids and could not swarm by themselves but, altruistically, helped others swarm. Experimental evolution revealed a rapid recovery of social behavior. First, we observed a rapid expansion of crc mutants in all populations. crc were >80% of the mutants at day 3, indicating their important role in the dynamics of the parallel phase. Whole genome sequencing of 20 isolates from the final day of the experiment revealed an alternative path to crc mutations because each mutant had a mutation in one of two loci downstream of cbrA — crc or hfq . crc mutations were still the majority (in 14/20) and one possible explanation is that the crc locus is longer: 787 bp compared with the 251 bp of hfq (found in 6/20). Perhaps more importantly, the mutations found in crc show that a loss-of-function at that locus was sufficient to recover from the slow growth of Δ cbrA , and there were many possibilities for mutations to cause loss-of-function. In hfq , we only found amino acid substitutions and mutations in the promoter region suggesting a reduction of Hfq expression or reduction of Hfq activity, but not Hfq loss-of-function. Mutations in hfq were less frequent, possibly because the chaperon Hfq has many other functions ( Sonnleitner et al. 2006 ); there should be fewer possibilities for hfq to mutate and suppress the effect of Δ cbrA without deleterious pleiotropic effects. We had previously experimented with swarming evolution starting from wild-type bacteria; those experiments led to a remarkably parallel evolution of mutliflagellated hyperswarmers ( van Ditmarsch et al. 2013 ) and provided new insights on the regulation of bacterial flagella ( Kearns 2010 ). Those experiments had started with a metabolically prudent wild-type and thus taught us little on the role of metabolism in swarming behavior. Here, starting with an engineered strain lacking proper metabolic prudence—the Δ cbrA strain—provided a unique view on how the internal metabolic state of an organism changes while it recovers from a perturbed social behavior. We saw that crc mutants recovered the wild-type’s metabolomic state and its growth rate, but retained a fitness cost of swarming behavior. hfq mutants recovered social behavior similar to the wild-type even if their metabolome remained perturbed. This unexpected trend suggests that in order to recover low-cost social behavior in the absence of proper metabolic regulation organisms may need to reorganize their metabolome. Hfq is an abundant regulator of with many physiological roles, which makes it harder to determine exactly how hfq mutations compensate for the fitness cost of swarming cooperation. Despite its metabolome being more similar to Δ cbrA than to the wild-type, some metabolites—including valine, lysine, citrulline and citrate—did recover wild-type levels in hfq . The connection between these metabolites and swarming fitness is unclear, but they provide targets for future investigation into the mechanism of metabolic prudence. A deeper study of the pleiotropic effects of hfq on the metabolome could also shed light on why hfq mutations affect some metabolites and not others, and on the role of Hfq in the ability of bacteria to divert resources away from central metabolism and into public goods essential for social behavior. We also saw that while all six populations recovered swarming by day 4, by day 6, two populations decreased swarming again. Those two populations contained crc -mutated lineages with additional mutations in flagella genes that caused loss of flagella motility, and those mutants blocked others from swarming. The emergence of flagellar mutations after recovery of growth rate by mutations in either crc or hfq suggests that after an initial competition for metabolic advantage, the evolution centers again on competition between motility strategies, which we had observed previously in the evolution of hyperswarmers ( van Ditmarsch et al. 2013 ). Interestingly, here we found two crc -mutated isolates from population 6 had mutations in fleN that turned them into hyperswarmers and were immune to blocking. That population continued swarming robustly until the end of the experiment. Bacteria have many social behaviors besides swarming—they communicate by quorum sensing, produce extracellular polymers to form biofilms, and share nutrient scavenging molecules ( Crespi 2001 ; West et al. 2006 )—and they can implement sophisticated social strategies even without having a brain ( Axelrod and Hamilton 1981 ). Bacteria rely on these social interactions but they live in chaotic worlds of ever-changing environments and random encounters with other strains and species ( Nadell et al. 2013 ). Horizontal gene transfer ( Niehus et al. 2015 ), frequent mutants ( Rainey and Rainey 2003 ; Boles et al. 2004 ; Kim et al. 2014 ; Flynn et al. 2016 ), dispersal, and environmental mixing ( Griffin et al. 2004 ; Nadell et al. 2013 ) constantly alter genetic relatedness and bacteria cannot always count on a stable social structure. Evolution should favor strategies that regulate social behaviors and lower their metabolic cost to the extent possible ( Xavier et al. 2011 ; Dandekar et al. 2012 ; Gupta and Schuster 2013 ). Beyond bacteria, nature shows an astounding diversity of social behaviors ( Robinson et al. 2008 ). Wolfs hunt in packs, termites build colonies of thousands, even plants communicate with chemicals secreted from their roots ( Bais et al. 2004 ). Can we ever identify general principles among them? Our experiments suggest a common premise: metabolism is the currency of all physiological processes and its influence on social behavior is likely conserved in all species, including humans. Proteins are the products of genes and they control metabolism directly, which make metabolic regulation a proximate mean for genes to command social behavior. All else being equal, natural selection should favor expression of a social behavior only when its metabolic cost is low, which closes the loop from genes to social behavior ( fig. 1 A ). Genetic tools to manipulate the metabolism of animals already exist ( Robinson et al. 2008 ; LeBoeuf et al. 2013 ) but it would be difficult to investigate the role of metabolism on the evolution of animal social behavior with the same detail as we did here in bacteria. Experimental evolution requires many generations and even the quick C. elegans can only reproduce once every ∼3.5 days. Bacteria have generations of <1 h, small genomes and large population sizes with billions of organisms which make them unique models for experimental testing of evolution theories ( Lenski et al. 1991 ; Fiegna et al. 2006 ; van Ditmarsch and Xavier 2011 ; Lee and Marx 2012 ; Lind et al. 2015 ; Kim et al. 2016 ). Understanding how metabolism affects social behavior can ultimately help us explain how organisms—ourselves included—govern their behavior towards others. And it could open the door to explain why factors that alter metabolism like the gut microbiomes of animals ( Turnbaugh et al. 2006 ; Ezenwa et al. 2012 ; McFall-Ngai et al. 2013 ) can influence social behavior."
} | 3,555 |
37961710 | PMC10634959 | pmc | 5,882 | {
"abstract": "Archaea belonging to the DPANN superphylum have been found within an expanding number of environments and perform a variety of biogeochemical roles, including contributing to carbon, sulfur, and nitrogen cycling. Generally characterized by ultrasmall cell sizes and reduced genomes, DPANN archaea may form mutualistic, commensal, or parasitic interactions with various archaeal and bacterial hosts, influencing the ecology and functioning of microbial communities. While DPANN archaea reportedly comprise 15–26% of the archaeal community within marine oxygen deficient zone (ODZ) water columns, little is known about their metabolic capabilities in these ecosystems. We report 33 novel metagenome-assembled genomes belonging to DPANN phyla Nanoarchaeota, Pacearchaeota, Woesarchaeota, Undinarchaeota, Iainarchaeota, and SpSt-1190 from pelagic ODZs in the Eastern Tropical North Pacific and Arabian Sea. We find these archaea to be permanent, stable residents of all 3 major ODZs only within anoxic depths, comprising up to 1% of the total microbial community and up to 25–50% of archaea. ODZ DPANN appear capable of diverse metabolic functions, including fermentation, organic carbon scavenging, and the cycling of sulfur, hydrogen, and methane. Within a majority of ODZ DPANN, we identify a gene homologous to nitrous oxide reductase. Modeling analyses indicate the feasibility of a nitrous oxide reduction metabolism for host-attached symbionts, and the small genome sizes and reduced metabolic capabilities of most DPANN MAGs suggest host-associated lifestyles within ODZs.",
"introduction": "Introduction In recent years, metagenomics has enabled the discovery of several prokaryotic superphyla lacking pure culture representatives ( 1 - 3 ). One of these novel groups is the DPANN archaea, named after the first members of the expanding superphylum (Diapherotrites, Parvarchaeota, Aenigmarchaeota, Nanoarchaeota and Nanohaloarchaeota) which has come to include at least ten putative phyla ( 4 , 5 ). The DPANN archaea are characterized by ultrasmall cell sizes (~0.1–1.5 μm), reduced genomes (~1.5 Mb), and limited metabolic capacities ( 6 ). These features, along with several enrichments and visualizations of DPANN archaeal-host associations ( 7 - 9 ), suggest a symbiotic or commensal lifestyle of DPANN archaea with diverse microbial hosts. If DPANN indeed exist in partnership with others, this would explain why they have been challenging to cultivate in isolation. Since their discovery, DPANN archaea have been found in a variety of diverse environments, including hydrothermal vents ( 10 ), freshwater and hypersaline lakes ( 11 , 12 ), groundwater ( 13 , 14 ), terrestrial hot springs ( 15 ), marine sediments and water columns ( 10 , 16 , 17 ), and the Black Sea ( 18 ). Archaea writ large play crucial roles in global biogeochemical cycles, such as in ammonia oxidation ( 19 ), methane cycling ( 20 ), and organic carbon scavenging ( 21 ), and DPANN archaea have been found to possess genes for sulfur cycling and organic substrate degradation ( 13 , 16 ). Additionally, DPANN archaea in anoxic environments may form consortia with methanogens and contribute to anaerobic carbon cycling ( 22 ). However, despite their widespread abundance, distribution, and diversity (accounting for about half of all archaeal diversity ( 6 )), the ecological and biogeochemical roles of DPANN archaea are not fully understood. Culture-independent techniques have only begun to unravel the importance of these previously-overlooked microorganisms within their expanding list of habitats. Amplicon surveys have detected the presence of DPANN archaea within both sediments beneath oxygen deficient zones (ODZs) ( 23 ) and the ODZ water column itself ( 24 ). The three major oceanic ODZs are located in the eastern tropical North Pacific (ETNP), the eastern tropical South Pacific (ETSP), and the Arabian Sea. Oxygen profiles in these regions display rapid decreases from surface saturation to below the detection limit of trace oxygen sensors (<10 nmol L −1 ) between 50–100 m depth, a region termed the oxycline ( 25 , 26 ). Oxygen concentrations then remain below detection and with no vertical gradient for approximately 200–800 m ( 27 ), although the ODZ thickness varies greatly across each basin ( 28 - 30 ). Due to these unique features, ODZ water columns contain multiple biogeochemical gradients that support diverse microbial assemblages performing nitrogen, carbon, and sulfur cycling ( 31 ). In particular, these regions disproportionately contribute to marine nitrogen cycling, accounting for about 30% of marine fixed nitrogen loss despite containing only 0.1–0.2% of oceanic volume ( 32 , 33 ). ODZs are characterized by prevalent denitrification, i.e. the microbially-mediated stepwise reduction of nitrate to dinitrogen gas. This anaerobic respiratory metabolism occurs via reductases encoded by a suite of widely distributed genes ( 34 ). The last step of denitrification, the reduction of N 2 O to N 2 , is catalyzed by nitrous oxide reductase encoded by nos. Two clades of the nos catalytic subunit nosZ have been found, a typical clade I nosZ associated with complete denitrifiers defined by an N-terminal twin-arginine translocation (TAT) motif, and an atypical clade II nosZ associated with partial denitrifiers defined by an N-terminal Sec-type motif ( 35 ). Both variants contain conserved copper-binding sites Cu A and Cu Z , although Cu Z sites of clade II nosZ homologs exhibit greater variability and less conservation ( 36 ). Recent studies reveal clade II nosZ predominates within ODZs, occurs within diverse marine taxa including archaea, and may be associated with low oxygen and enhanced N 2 O affinity ( 36 ). Because N 2 O depletes ozone and is a potent greenhouse gas, organisms with atypical nosZ variants, including archaea, merit interest as potential N 2 O sinks. Increasing attention has been focused on ODZ archaeal communities ( 37 - 39 ), such as members of Thermoproteota (including former Marine Group I Thaumarchaeota) and Thermoplasmatota (including former Marine Group II archaea) ( 40 , 41 ). However, little is known about ODZ DPANN archaea, despite reports that they may comprise up to 15–26% of total archaeal reads in these regions ( 24 ). Challenges in cultivation of these environmental microbes limit our understanding of the metabolic capabilities of clades such as DPANN that lack cultured representatives. Accordingly, the contribution of DPANN archaea in ODZ microbial assemblages and biogeochemical cycling, as well as the abundance, distribution, metabolism, ecology, and phylogeny of these archaea remain open questions. Using genome-resolved metagenomics, we recover 33 genomes belonging to DPANN phyla Nanoarchaeota, Pacearchaeota, Woesarchaeota, Undinarchaeota, and Iainarchaeota from the ETNP and Arabian Sea ODZs. We characterize the metabolic capabilities of these archaea, place them within the existing phylogeny of known DPANN, and determine their relative abundances and distributions within and across global ODZs. Our results demonstrate that DPANN are a ubiquitous portion of the microbial community within ODZs and comprise several lineages with diverse metabolic potential.",
"discussion": "Discussion DPANN archaea were found to be a stable resident population within all 3 permanent pelagic ODZs. Abundances of DPANN archaea, including Nanoarchaeota, SpSt-1190, Iainarchaeota, Woesarchaeota, and Undinarchaeota, increase as oxygen decreases, while few or no DPANN archaea were found in the surface oceans ( Figure 1B ). While a few population differences are found between ODZs, Woesarchaeota are the dominant phylum within all three ODZs, with Nanoarchaeota in the Arabian Sea and Pacearchaeota, Undinarchaeota, and SpSt-1190 in the ETNP and ETSP forming the second most abundant groups ( Figure 1B ). ODZ DPANN archaea are phylogenetically and metabolically diverse and group together with other DPANN from non-ODZ environments, although several Woesarchaeota cluster within the same clade ( Figure 2 ). Similar to DPANN across various environments ( 4 , 10 , 12 , 13 ), most ODZ DPANN have small genome sizes and limited capacity for biosynthesis of essential amino acids and nucleotides, limited energetic capabilities, and partial or absent pentose phosphate pathways despite overall high MAG completion estimates ( Figure 3 ). Additionally, completion metrics may underestimate the completeness of DPANN MAGs due to their limited genomes and high number of absent genes considered essential in other organisms. Numerous studies have reported microscopy images of environmental DPANN attached to host cells ( 7 , 13 , 81 ). While most ODZ DPANN genomes suggest a host-associated rather than free-living lifestyle, SpSt-1190 genomes average 4 Mb in size, possess a number of biosynthesis pathways, and carry pathways for methanogenesis. These unique archaea may represent free-living DPANN organisms ( 82 - 84 ) involved in methane cycling. The identification of ODZ DPANN hosts, whether a single host, various hosts, or a community, may hold keys to their distribution and survival within ODZ environments. Studies have suggested a role of DPANN archaea for carbon cycling, such as by scavenging organic carbon in the form of nucleotides, lipids, and amino acids ( 21 , 85 ), participating in the exchange of carbon compounds with hosts, and even directly parasitizing upon hosts ( 12 ). In addition, some may perform fermentation and consume or produce acetate ( 6 ). We find conserved pathways for amino acid salvage and fermentation across ODZ DPANN genomes ( Figure 3 ). While various sugar, protein, and DNA transporters indicate potential resource exchange with host cells, the existence of a peptidoglycan-degrading enzyme and secreted peptidases within several MAGs may point to a potentially parasitic relationship between host and DPANN cell. Future experimental tests will be needed to clarify these metagenomic predictions. While other nitrogen cycling genes are absent, a majority of ODZ DPANN carry a gene similar to the nitrous oxide reductase gene nosZ that catalyzes the reduction of nitrous oxide (N 2 O) to N 2 . Further investigation of this gene, annotated as nitrous oxide reductase, indicates the presence of a conserved Cu A copper-binding site typical of nosZ and cytochrome c oxidase subunit II ( 75 , 76 ) ( Figure 4 ). The cellular location of the protein product of the DPANN nosZ -like gene is postulated as outside of the membrane, possibly in the periplasmic space ( Figure S2 ). DPANN archaea are thought to possess two membranes ( 86 ), and canonical nosZ is a periplasmic protein unlike the membrane-bound cytochrome c oxidase subunit II ( 74 ). Cytochrome c oxidase performs the last step of aerobic respiration, but no other elements of aerobic respiration were found within these archaea ( Figure 3 ). The Cu Z catalytic center, typically found upstream of the Cu A center in nosZ , is absent within DPANN nosZ -like genes. The Cu Z center lacks a consensus motif, but is characterized by 7 histidine residues that bind copper ions ( 87 ). While the majority of DPANN MAGs possess several acyl carrier protein genes for fatty acid biosynthesis surrounding the nosZ -like gene, 2 DPANN MAGs encode a protein containing 5 histidine residues directly upstream of the nosZ -like gene. This protein, annotated to the same family as nosZ , groups phylogenetically with clade II nosZ sequences ( Figure 4 ) and may perform a function related to that of the Cu Z site. This hypothetical histidine-rich region was absent within other DPANN MAGs, and was not included within the complementation test. The activity of these or other proteins within these genomes may be required for N 2 O reduction. While the function of putative nosZ -like genes within DPANN archaea remain hypothetical, evidence suggests the involvement of these genes in N 2 O reduction or another redox process with metabolic or physiological importance due to their conservation within these small, streamlined genomes. Complementation of P. aeruginosa Δ nosZ with DPANN nosZ -like genes did not result in significant N 2 O consumption. While heterologous complementation may offer convincing evidence for the function of unknown genes, negative results are difficult to interpret. Large evolutionary distances between DPANN archaea and the gram-negative bacterium P. aeruginosa , likely resulting in different intracellular conditions, may inhibit the proper transcription, translation, or maturation of the DPANN NosZ-like protein. The protein may also be adapted to specific environmental conditions necessary for its activity, which differ from those used during standard cultivation of P. aeruginosa . Deletion and complementation of the nosZ -like gene within native DPANN archaea would be an ideal functional test, but currently no cultured representatives or genetic toolkits are available for these organisms, limiting our knowledge of many of their metabolic features to predictions from gene annotations. N 2 O exists in nanomolar concentrations in ODZs compared to the higher concentrations of nitrate and nitrite ( 78 ), posing challenges for N 2 O-reducing specialists lacking upstream denitrification genes. However, an N 2 O-consuming lifestyle may be feasible if local N 2 O concentrations are elevated in proximity to an N 2 O source, such as a partial denitrifier lacking nosZ . Previous studies have indicated widespread occurrence of partial denitrifiers lacking nosZ within ODZ regions ( 39 , 46 ). We tested this scenario by modeling the local flux of N 2 O from a producer (the source) to an N 2 O consumer ( Figure 5 ). N 2 O uptake rate of the consumer is elevated 100-fold when the two cells are in physical contact vs. when they are a short distance of 2 μm away ( Figure 5C ). This increase in N 2 O uptake rate drops off steeply, however, as the consumer-to-producer cell size ratio increases ( Figure 5 ). Under low bulk N 2 O concentrations, partial denitrifiers may provide elevated local N 2 O only to much smaller surface-attached episymbiotic N 2 O consumers. DPANN archaea within ODZs, similar to those found within other environments ( 7 , 8 , 13 ), potentially exist as host-associated episymbionts and likely possess small cell sizes. The average cell volume of DPANN archaea has been reported as 0.004 μm 3 ( 13 ) while average marine bacterial cell volume has been reported at up to 0.096 μm 3 ( 88 ), resulting in a consumer-to-producer cell size ratio of < 0.05. Thus, DPANN archaea may be uniquely adapted to consume N 2 O and other resources that are scarce under bulk conditions but locally elevated in proximity to host cells. DPANN archaea possess a high number of unknown or unannotated genes, representing “microbial dark matter.” Within our ODZ DPANN, we found over 20,000 hypothetical proteins across all MAGs. Further studies, possibly using genetic manipulations, isolation or enrichment cultures, imaging, and computational proteomics approaches are required to characterize the functions of putative or hypothetical proteins. The expanding knowledge of these organisms may make these questions more tractable in the near future. At a large scale, the scavenging of carbon, potential nitrogen, sulfur, and hydrogen cycling capabilities, and ecological effects on host populations via symbiosis or parasitism by DPANN archaea in the ODZs warrants future investigation."
} | 3,885 |
28368020 | PMC5377364 | pmc | 5,883 | {
"abstract": "We introduce multi-gradients including Laplace pressure gradient, wettable gradient and wettable different gradient on a high adhesion surface via special wedge-pattern and improved anodic oxidation method. As a result of the cooperative effect mentioned above, controlled directional motion of a droplet on a high adhesion surface is realized, even when the surface is turned upside down. The droplet motion can be predicted and the movement distances can be controlled by simply adjusting the wedge angle and droplet volume. More interestingly, when Laplace pressure gradient is introduced on a V-shaped wettable gradient surface, two droplets can move toward one another as designed.",
"discussion": "Results and Discussion Liquid transport on wedge-pattern surfaces was first studied by Khoo and Tseng. They presented liquid transport on nanotextured surface with wedge-shaped track 27 . However, the fabrication process was complex and droplet movement distance was uncontrolled as the wedge angle was too small to control precisely. Alheshibr et al . demonstrated droplet spontaneous droplet movement on hydrophilic aluminium surface containing a hydrophobic background 28 . Although droplet spreading distance could be controlled through the change in wedge angle, the droplets tended to spread instead of moving. Here, in order to realize the controlled self-propelling of droplet on a high adhesion surface (droplet can’t slide even when the plate is placed vertically), we combine wettable gradient and Laplace pressure simultaneously on a surface. Surfaces with wettable gradient and Laplace pressure gradient were fabricated as shown in Fig. 1a . At first, graphite plates were prepared by gradient anodic oxidation to form wettable gradient via the change in the surface chemical composition (see Figure S1 ), as our previous report 21 . The difference of SEM image in different regions is not obvious (see Figure S2 ), from which we can conclude that the wettability gradient is caused by the difference of chemical composition. By controlling the volume flow of electrolyte, the wettable gradient can be controlled between 1.2°/mm and 9°/mm, as shown in Figure S3 . Then, the prepared graphite plates were coated with paraffin wax to ensure a hydrophobic background (the coated substrate exhibits average sessile contact angle (CA) of 116°), and a wedge-pattern area was removed to induce Laplace pressure gradient on the wettable gradient surface. The SEM image shows distinct wedge-pattern and clear dividing line, whereas the roughness between covered portion and uncovered portion is similar (see Figure S4 ). This is a simple but effective way to produce wettable gradient and Laplace pressure gradient simultaneously on a surface of conductive substrates such as carbon and some metal materials. The wettable gradient and Laplace pressure gradient can be adjusted via oxidation conditions and wedge angle, respectively. Here, we choose the graphite surface with a wettable gradient of 9.0°/mm (the change of CAs is shown in Figure S5 ) to investigate the movement behaviours of droplets on it. In order to present the necessity of the combination of a wettable gradient and a wedge-pattern, experiments of the graphite plates only with wettable gradient or with wedge-pattern were carried out comparing with the sample with wettable gradient and wedge-pattern simultaneously. As shown in Fig. 1b , longitudinal spreading can be observed on the surface with wettable gradient due to the existence of the gradient force (the wettable gradient is 9.0°/mm). Droplet on the sample with wedge-pattern also spreads a little bit as the result of Laplace pressure (wedge angle is 20°) ( Fig. 1c ). Clearly, droplets can not move, whether on the sample with wettable gradient or wedge-pattern. However, with the combination of wettable gradient and wedge-pattern (wedge angle is 20° and wettable gradient is 9.0°/mm) (see Fig. 1d ), droplet moves unidirectionally as designed, from which we can confirm the necessity of combining wettable gradient and wedge-pattern to drive droplets. The wedge angle is one of the most critical influences on the droplet motion 29 . The movement behaviour of water droplet of 5 μL on the wettable gradient surfaces (9.0°/mm) with different wedge angles is shown in Figs 2a and S6 . When we add a wedge angle of 6° on wettable gradient surface, droplet stays still due to less Laplace pressure gradient. Even the angle increases to 10°, droplet remains still. When the angle is between 12° and 32°, the obvious self-propelling of droplets is observed and the movement distance gradually increases at first and then decreases remarkably, showing a peak of 2.6 mm when the angle is 20° ( Figure S7 ). When the angle is above 35° (including 35°), droplet just spreads along the direction of wettable gradient and no movement behaviour is observed. Figure S8 shows pictures of 5 μL droplet as it moves on the surface with wedge pattern of 20°. The droplet driven by the wettable gradient and Laplace pressure moves along the wedge track. The droplet evolves from a sphere droplet into cone shape. Clearly, we realize the self-propelling of droplets on a high adhesion surface via the collaborative effect of Laplace pressure and wettable gradient and the movement distance can be adjusted by the angle of wedge-patterns. For a thorough understanding of the movement behaviour of water droplets on surfaces with wettable gradient and Laplace pressure, we analyze the forces exerted on the droplet ( Fig. 2b ). The droplet motion on the gradient surface involves the interaction effect between the driving forces and resistance force. The driving forces consist of wettable gradient force ( F wg ), Laplace pressure force ( F L ) and wettable different force ( F wd ). The resistance force is hysteresis force ( F H ). On wettable gradient surface, a driving force F wg arising from the wettable gradient, due to different CAs at the both ends of the droplet, can be described by refs 21 and 26 : In which α is the half wedge angle, γ is the surface tension of water, L is the length of droplet along the wettable gradient direction and k is the wettable gradient (here, k = 9°/mm) and θ M is the CA at more wettable side (M) of droplet. Due to wedge-pattern on wettable gradient surface, Laplace pressure force ( F L ) in the lengthwise direction is generated, which push the droplet from a smaller wettable footprint (left) to a larger one (right). Megaridis et al . 30 followed the approach of Lorenceau and Quéré 31 to calculate the Laplace pressure force ( F L ) on the droplet sat on the wedge-pattern and found out the linear nature of the plot of F L against α . Accordingly, we can evaluate the magnitude of the F L as: here, A, B is constant coefficient for a given volume of droplet. In addition, the difference of wettability between inside and outside of wedge-pattern also provides another driving force ( F wd ) in the x direction, which can be described by ref. 26 : In which θ avg is the average contact angle of droplet over the length of the drop and θ o is the contact angle of hydrophobic area outside of wedge-pattern. Here, we use simple method to introduce three driving forces on a high adhesion surface, i.e., wettable gradient force, Laplace pressure force and wettability different force ( Fig. 2b ). The Hysteresis force, which is always opposite to the moving direction, is composed of three parts (an arc and two straight-line edges) 21 : Here θ aM is the advancing CA at more wettable side of water, φ is the polar angle and θ ro is the apparent receding CA of hydrophobic area outside of wedge-pattern ( θ ro = 105°). Obviously, there seem to be four main forces influencing the motion of droplets: wettable gradient force ( F wg ), Laplace pressure force ( F L ), wettable different force ( F wd ) and hysteresis force ( F H ). On the basis of the equations (1 ), ( 2 ), ( 3 ) and ( 4 ), the total force ( F T ) is described as: Here, . For a given volume of droplet ( v ), the values of A, B, C and L can be seen as constant. Therefore, total force ( F T ) can be regarded as a function of the volume of droplet ( v ) and the half wedge angle ( α ). For a given volume, we can measure the value of the length of droplet ( L ). Then, according to the Figure S9 , the values of θ M , θ avg , θ aM can be read right off the diagram. So, the value of C can be obtained by calculating. Finally, we observe the motion behaviors of droplet at different wedge angles. Accoridng to the critical condition under which droplets can self-propel, we can get the values of A and B for given volume of droplet. For example, for droplet of 5 μL, the contact length L is 2.24 mm. From Figure S9 , We can get the the values of θ M , θ avg , θ aM , i.e., 93°, 103.3°, 112.6°. Accordingly, we can figure out the value of C (C = 0.008749) easily. For droplet of 5 μL, the droplet can self-propell when the wedge angle is between 12° and 32°, i.e., F T (6°) = F T (16°) = 0. So, we can get Values of A = −0.000082689 and B = 0.00000015826. By the same method, the values of A, B, C for droplets of 2 μL, 8 μL, 10 μL can be achieved by calculating and the results are list in Table 1 . As metioned in the previous part, the A, B, C and L are function of volume ( v ), so we use polynomial equations to get the relations between A, B, C, L and v , as shown in Fig. 3a–d . So, we could further simplify express the equation of total force ( F T ): According to equation (6) , a mathematical model based on an analysis of the forces applied on liquid drops is set up (The force acted on the droplet is determined by the droplet volume ( v ) and half pattern wedge angle ( α )). As shown in Fig. 3e , the movement behaviour of droplets may be determined from the value F T and the conditions, under which the value F T is above zero, are marked (area S). The results indicate that actuation range reduces monotonically with droplet volume, which matches well with our experiment results listed in the Table 1 . Otherwise, as can be seen from the simulation diagram, when the droplet volume is around 6.3 μL, the driving force is the largest at certain wedge angle. And when the droplet volume reaches about 11 μL, the driving force reduces to zero at all wedge angle. Besides that, as shown in the graph, when the droplet volume is fixed, the force applied on the droplet increases firstly and then decreases with the increase of wedge angle, showing a peak value. All of results are accorded with observation results, i.e., the droplet exhibits its maximum average motion velocity when the volume is 6.5 μL (in Figure S10 ). When droplets volume reaches 11 μL, the droplet can not be driven at any wedge angle. The movement distance reaches its peak value when the half wedge angle is 10° for 5 μL droplet ( Figure S7 ). Not unexpected, the simulation result corresponds well with our experimental result, which revealed the validity of numerical simulation. Clearly, we can predict droplet movement behaviour via the droplet volume ( v ) and half pattern wedge angle ( α ) ( Fig. 3e ), and to control the droplets motion by adjusting the droplet volume and wedge angle. The optimized wedge angle is depend on the droplet volume, and the optimized amount droplet volume can be discussed in two aspects. One is the amount of 2 uL which can be actuated among the widest range of wedge angle (from 7° to 30°). The other is the amount of 5 uL which exhibited the longest motion distance among the maximum movement distance of different droplet volume (see Figure S11 ). Apart from the single droplet motion, we also fabricated a surface with V-shape wettable gradient and Laplace pressure to realize the controlled motion of two droplets on the basis of the aforementioned method. Firstly, we took a one-step method to form V-shape gradient. Next, the graphite plate was coated by the wax and a rhombus area was removed. By doing so, V-shape wettable gradient and Laplace pressure gradient are formed on one surface. Figure S12 shows that both ends of the graphite plate have CAs about 108° as the oxidation time is the shortest. The smaller the distance between the measurement point and the centre, the more hydrophilic the position exhibited. i.e., CAs change gradually from 108° to 31° along the direction from the boundary to the centre. As shown in Figure S13 , we can see apparent rhombus pattern and the dividing line is distinct in the optical images. In order to observe the movement behaviour of two droplets within limited field of vision, here we chose 2 μL droplet in the following experiment. As shown in Fig. 4a , on the surface only with a V-shape wettable gradient, two separate droplets placed on the either side of the centre spread closely to each other and merge into one if the distance between two droplets is short (below 3.5 mm), while orientation movement of whole droplets can’t be seen. Here, by one-step gradient oxidation method, we can realize unidirectional spreading of two droplets and two droplets spread toward each other and merge into one droplet if the distance is short. The main driving forces arise from wettable gradient alone. However, once the distance of the droplets is above the critical distance (3.5 mm), droplets can’t merge into one (see Figure S14 ). After Laplace pressure is introduced via rhombus-pattern, the movement behaviour of droplets is different. Two droplets move toward to each other and then merge quickly into one ( Fig. 4b ). Both of two droplets move about 1.5 mm before merging into one due to the cooperation of three driving forces: wettable gradient force, Laplace pressure and wettability different force, as mentioned above. Because of the movement behaviour, two droplets can merged into one droplet even when they are placed farther apart (7 mm). The critical distance that droplets can merge into one is much larger than that on the surface merely of V-shape wettable gradient. This phenomenon is very important for micro reactor. By adjusting the angle and the volume of droplets, as well we can control the movement distance of the two droplets. On this basis, efforts are currently underway to fabricate more interesting pattern to realize individual or multiple droplets mobilization as design. In addition, due to the high adhesion of graphite plate, even when the graphite plates are inverted, similar movement behaviour can be observed, as shown in Fig. 4c,d . Comparing movement behaviour of droplets in different conditions, when the plates are inverted, the profile of the droplet is more obvious due to the gravitational force and movement distance of droplets on the lower surface is longer than that on upper surface, which may be due to the fact that adhesion force is reduced under effect of gravity. In conclusion, by combing wettable gradient and Laplace pressure gradient, unidirectional or bidirectional movement of droplets on a high adhesive surface can be realized, even when the surface is turned upside down. More interestingly, the movement distance can be controlled easily via adjusting the angle of the wedge-pattern. On this basis, more interesting shape pattern could be designed to manipulate individual droplet or multiple droplets, which has great potential application prospect in pharmaceutical detection and microfluidic tools 32 33 34 ."
} | 3,871 |
25914604 | null | s2 | 5,884 | {
"abstract": "Plant-plant interactions are driven by environmental conditions, evolutionary relationships (ER) and the functional traits of the plants involved. However, studies addressing the relative importance of these drivers are rare, but crucial to improve our predictions of the effects of plant-plant interactions on plant communities and of how they respond to differing environmental conditions. To analyze the relative importance of -and interrelationships among- these factors as drivers of plant-plant interactions, we analyzed perennial plant co-occurrence at 106 dryland plant communities established across rainfall gradients in nine countries. We used structural equation modeling to disentangle the relationships between environmental conditions (aridity and soil fertility), functional traits extracted from the literature, and ER, and to assess their relative importance as drivers of the 929 pairwise plant-plant co-occurrence levels measured. Functional traits, specifically facilitated plants' height and nurse growth form, were of primary importance, and modulated the effect of the environment and ER on plant-plant interactions. Environmental conditions and ER were important mainly for those interactions involving woody and graminoid nurses, respectively. The relative importance of different plant-plant interaction drivers (ER, functional traits, and the environment) varied depending on the region considered, illustrating the difficulty of predicting the outcome of plant-plant interactions at broader spatial scales. In our global-scale study on drylands, plant-plant interactions were more strongly related to functional traits of the species involved than to the environmental variables considered. Thus, moving to a trait-based facilitation/competition approach help to predict that: 1) positive plant-plant interactions are more likely to occur for taller facilitated species in drylands, and 2) plant-plant interactions within woody-dominated ecosystems might be more sensitive to changing environmental conditions than those within grasslands. By providing insights on which species are likely to better perform beneath a given neighbour, our results will also help to succeed in restoration practices involving the use of nurse plants."
} | 565 |
36267191 | PMC9577173 | pmc | 5,885 | {
"abstract": "Biofilm plays important roles in the life cycle of Bacillus species, such as promoting host and object surface colonization and resisting heavy metal stress. This study utilized transcriptomics to evaluate the impacts of cadmium on the components, morphology, and function of biofilms of Bacillus subtilis strain 1JN2. Under cadmium ion stress, the morphology of the B. subtilis 1JN2 biofilm was flattened, and its mobility increased. Moreover, differential gene expression analysis showed that the main regulator of biofilm formation, Spo0A, decreased in expression under cadmium ion stress, thereby inhibiting extracellular polysaccharide synthesis through the SinI/SinR two-component regulatory system and the AbrB pathway. Cadmium ion treatment also increased the SigD content significantly, thereby increasing the expression of the flagella encoding and assembly genes in the strain. This promoted poly-γ-glutamic acid production via the DegS/DegU two-component regulatory system and the conversion of biofilm extracellular polysaccharide to poly-γ-glutamic acid. This conferred cadmium stress tolerance in the strain. Additionally, the cadmium ion-mediated changes in the biofilm composition affected the colonization of the strain on the host plant root surface. Cadmium ions also induced surfactin synthesis. These findings illustrate the potential of Bacillus species as biocontrol strains that can mitigate plant pathogenic infections and heavy metal stress. The results also provide a basis for the screening of multifunctional biocontrol strains.",
"discussion": "Discussion Biofilms play important roles in the life cycle of Bacillus ( Branda et al., 2001 ; Vlamakis et al., 2013 ), such as protection from antibiotics ( Høiby et al., 2010 ) and adherence to the surface of objects. Bacillus is a preferable biocontrol agent of plant pathogens because of its strong survivability; therefore, various Bacillus species have been screened from the laboratory and applied in the field ( Chen et al., 2012 ). Colonization is a very important factor in biocontrol efficacy since the level of colonization directly affects the inhibition capacity against plant pathogens of biocontrol agents ( Chen et al., 2013 ). Various factors such as light, pH, and metal ions affect the colonization ability of biocontrol strains in the field. These factors also represent bottlenecks limiting the application of laboratory-generated biocontrol strains in the field. Among the main components of Bacillus biofilm, both EPS and γ-PGA were reported to help the strain to colonize the host roots, thus inhibiting plant pathogens effectively ( Yu et al., 2016 ). Wild-type Bacillus species exhibit different biofilm morphologies owing to differences in their extracellular secretions. Moreover, the role of these components in biofilm formation and morphological changes has been demonstrated using deletion mutants of different extracellular secretory components ( Yan et al., 2016 ; Yu et al., 2016 ). Our previous work demonstrated that Cd 2+ can change the biofilm morphology of B. subtilis 1JN2 ( Yang et al., 2015 ). High Cd 2+ concentrations increased the viscosity and flattened the surface of the 1JN2 biofilm, showing that exogenous heavy metal ions can affect the composition of the extracellular secretion of the strain. Accordingly, it is necessary to evaluate the mechanisms involved in the extracellular component change and determine whether the biological functions of the strain are affected by the change. B. subtilis 1JN2 has proven to be a significant biocontrol agent against Ralstonia via host colonization ( Yang et al., 2012 ). In a follow-up study, we found that the strain can effectively adsorb cadmium ions in the environment to relieve the stress to host plants ( Yang et al., 2015 ). After cadmium ion treatment, the biofilm morphology of the strain changed significantly, but its biocontrol effect on bacterial wilt did not change significantly. Thus,we assessed whether the biofilm morphological changes caused by Cd 2+ impact the colonization ability of this strain. The results show that the change in extracellular components caused by Cd 2+ treatment reduced tomato root surface colonization by the strain. However, as mentioned above, the effect of Cd 2+ on the biocontrol efficiency of B. subtilis 1JN2 against Ralstonia was not significant. Therefore, we further explored the mechanism by which this strain can simultaneously control bacterial wilt and assist the host to alleviate cadmium stress. The transcriptome sequencing results show that the expression levels of the EPS-encoding genes were significantly reduced, but that of γ-PGA genes increased after Cd 2+ treatment. Further analysis revealed that KinC is the receptor of exogenous Cd 2+ in B. subtilis 1JN2, which activates the main regulatory factor Spo0A through phosphorylation. This activation process is regulated by the critical protein Sda, which also controls sporulation or biofilm formation in bacterial cells. The decline in Spo0A levels inhibited the EPS synthesis through the SinI/SinR two-component regulatory system and increased AbrB levels. According to previous reports, metal ions generally positively regulate biofilm formation. For example, in Salmonella enterica and Erwinia amylovora , the regulator Zur specifically regulates the uptake of zinc ions and promotes the formation of bacterial biofilms ( Ammendola et al., 2016 ; Kharadi and Sundin, 2020 ). For some pathogenic microorganisms, ion acquisition is critical for survival within the host. Therefore, specialized ion uptake regulators such as Fur and XibR can help these pathogens obtain enough metal ions and can regulate their adhesion and pathogenicity ( Pandey et al., 2016 ; Teper et al., 2021 ). Cadmium ion treatment significantly increased the expression of SigD, thus enhancing the expression of the flagella encoding and assembly genes of the strain. This further activated the DegS/DegU two-component regulatory system for γ-PGA production, enhancing the mobility of the strain and conversion of EPS to γ-PGA. Both EPS and γ-PGA have been reported to improve the colonization ability of Bacillus strains. However, in our results, we found that the bacterial population that colonized the tomato root surface was decreased. Accordingly, we speculate that the roles of γ-PGA and EPS in colonization are interchangeable in some Bacillus strains but not for strain 1JN2 on tomato root surfaces. As such, there may be other reasons why this strain can simultaneously control bacterial wilt and assist the host in alleviating cadmium stress. The expression of surfactin synthesis-related genes was also significantly increased after Cd 2+ treatment. Previous studies also reported that metal ions could induce surfactant secretion by Bacillus ( López et al., 2009 ). As an important active protein product, surfactin plays a key role in alleviating heavy metal stress and controlling plant diseases ( Almoneafy et al., 2014 ).According to our results, increased surfactin levels may improve the biocontrol ability of the strain by resisting metal ion stress and exerting inhibitory activity against Ralstonia solanacearum . Generally, the main changes caused by Cd 2+ treatment in B. subtilis 1JN2 biofilm were decreased EPS content, increased γ-PGA and surfactin levels, and enhanced cell mobility. Based on this, we deduce that the adaptive changes by Bacillus do not affect its disease control activity within a certain range of cadmium ion concentrations. Through surfactin secretion, B. subtilis can alleviate pathogen infections and heavy metal stress for the host plant. Exploring additional mechanisms involved in B. subtilis biofilm changes and screening of multifunctional biocontrol strains will be the focus of our future work."
} | 1,965 |
39020132 | PMC11458647 | pmc | 5,886 | {
"abstract": "Current evidence suggests that some form of cellular organization arose well before the time of the last universal common ancestor (LUCA). Standard phylogenetic analyses have shown that several protein families associated with membrane translocation, membrane transport, and membrane bioenergetics were very likely present in the proteome of the LUCA. Despite these cellular systems emerging prior to the LUCA, extant archaea, bacteria, and eukaryotes have significant differences in cellular infrastructure and the molecular functions that support it, leading some researchers to argue that true cellularity did not evolve until after the LUCA. Here, we use recently reconstructed minimal proteomes of the LUCA as well as the last archaeal common ancestor (LACA) and the last bacterial common ancestor (LBCA) to characterize the evolution of cellular systems along the first branches of the tree of life. We find that a broad set of functions associated with cellular organization were already present by the time of the LUCA. The functional repertoires of the LACA and LBCA related to cellular organization nearly doubled along each branch following the divergence of the LUCA. These evolutionary trends created the foundation for similarities and differences in cellular organization between the taxonomic domains that are still observed today. Supplementary Information The online version contains supplementary material available at 10.1007/s00239-024-10188-7.",
"introduction": "Introduction Cellular organization is a defining characteristic of all life on Earth. A detailed and accurate understanding of the emergence of cellularity is, therefore, crucial to our broader account of early evolutionary history. The subject, however, remains somewhat contentious due to countervailing lines of evidence, some of which indicates an early evolution of cellularity, perhaps even coincident with the origin of life, itself, and some of which indicates a late evolution of cellularity, perhaps even following the divergence of the LUCA into the separate ancestors of bacteria and archaea. Protocell experiments have demonstrated that membranes can form spontaneously from prebiotically available compounds such as decanoic acid (Namani and Walde 2005 ) as well as the lipid fractions of carbonaceous chondrite meteorites (Deamer and Pashley 1989 ), suggesting a possible role for membrane compartmentalization as early as the origin of life, itself. Artificial life simulations have shown that protocell encapsulation could have protected early replicator genomes from parasites (Hogeweg and Takeuchi 2003 ; Takeuchi and Hogeweg 2009 ; Shah, et al. 2019 ) and supported genomic stability in general (Takagi, et al. 2020 ). Other artificial life simulations have also shown that, in the kind of rich environment often associated with the origin of life, selection would have acted against protocell encapsulation even if that encapsulation was imposed by the environment rather than produced by the life form, but that protocell encapsulation would have eventually co-evolved along with the first metabolic pathways (Szathmary 2007 ; Takagi, et al. 2020 ). Phylogenetic reconstructions of universal paralog protein families have shown that cell membrane-associated proteins such as ABC transporters, ATP synthase enzymes, and the signal recognition particle system, all underwent gene duplications prior to the time of the last universal common ancestor (LUCA) (Gribaldo and Cammarano 1998 ; Kollman and Doolittle 2000 ; Zhaxybayeva, et al. 2005 ; Harris and Goldman 2021 ). This evidence would indicate that cellular organization was well in place by the time of the LUCA. However, the primary constituents of cell membranes, phospholipids, are radically different between archaea and bacteria (Pereto, et al. 2004 ; Sojo 2019 ). Typical bacterial phospholipids contain fatty acid chains that are connected to the phosphate head group by an ester linkage, while typical archaeal phospholipids contain isoprenoid chains that are connected to the head group by an ether linkage. Though all bacterial and most archaeal phospholipids are composed of a head group and two hydrophobic tails that form a bilayer membrane, some archaeal phospholipids contain hydrophobic tails that bridge across the bilayer with a phosphate head group on either side. These differences in the main constituents of cell membranes, as well as the metabolic pathways that produce them, have been taken as evidence by some researchers that true cellular organization evolved after the time of the LUCA (Wachtershauser 1988 ; Koga, et al. 1998 ; Martin and Russell 2003 ; Weiss, et al. 2016 ). Given the different lines of evidence, some researchers have argued for a central role of protocell compartmentalization as early as the origin of life itself (for example, (Saha and Chen 2015 ; West, et al. 2017 ; Damer and Deamer 2020 ; Nunes Palmeira, et al. 2022 ; Goldman 2023 )) and that the last universal common ancestor was a fully cellular organism (Becerra, et al. 2007 ; Goldman, et al. 2023 ), while others have argued that even by the relatively later stage of the LUCA, life forms were still not fully cellular (Wachtershauser 1988 ; Koga, et al. 1998 ; Martin and Russell 2003 ; Koonin and Martin 2005 ; Weiss, et al. 2016 ). Here, we seek to reconcile the evidence for and against cellular organization by the time of the LUCA by identifying protein families associated with cellular organization in minimal proteome reconstructions of the LUCA as well as its successors, the last archaeal common ancestor (LACA) and the last bacterial common ancestor (LBCA). In doing so, we portray, in broad terms, the kinds of cellular functions that likely were and were not encoded in the LUCA and how these functions appear to have expanded along the first branches of the tree of life. By complementing an analysis of LUCA cellular functions with an analysis of the subsequent evolution of those functions following the LUCA, we aim to provide context for major differences in cellular organization between the bacteria and archaea that can still be observed today.",
"discussion": "Discussion Inferring the proteomes of organisms that lived approximately 3.5-4Gya is inherently difficult (Crapitto, et al. 2022 ). EggNOG clusters may be incorrectly included or excluded from one of the ancestral proteomes due to limitations of the methodologies. For example, a protein family may have been present in the LUCA but lost to such an extent in subsequent lineages that it cannot be reconstructed as such. For simplicity, we describe differences in the presence and absence of protein functions between the three ancestral proteomes as gains and losses, but these results can also be explained by methodological shortcomings. For this reason, we caution against interpreting the results of an individual protein family being present in one of the ancestral proteomes as definitive. Instead, we portray the evolution of cellular functions in the LUCA, LACA, and LBCA in broad terms that both address the competing hypotheses about when cellularity first evolved and also provide a roadmap for future research. Taken together, these results suggest that the LUCA represents a population of cellular organisms. By the time of the LUCA, many different cellular functions had evolved, and these appear to have expanded during the subsequent evolution of the LACA and the LBCA. The small number of phospholipid biosynthesis enzyme families found in all three datasets agrees with the observation that bacterial (and eukaryotic) phospholipids differ chemically from archaea. Recent evidence suggests that phospholipid chemistry is diverse even within the archaeal (Caforio and Driessen 2017 ) and bacterial (Sohlenkamp and Geiger 2016 ) domains, which explains the lack of conserved phospholipid biosynthesis enzymes even within the LACA and LBCA proteomes. However, despite the lack of a clear signal of conserved phospholipid biosynthesis in any of these ancestors, other cellular systems were clearly in place at the time of the LUCA (Lombard, et al. 2012 ) and expanded upon during the evolution of the LACA and LBCA. The cellular functions associated with the minimal LUCA proteome depict a cellular organism capable of embedding proteins within the membrane and controlling, to some extent, the translocation of ions and biomolecules across that membrane. Importantly, the LUCA also seems to have been capable of controlling its cell division rather than relying on spontaneous growth and division as is observed in protocells (Berclaz, et al. 2001 ; Hanczyc, et al. 2003 ). The LUCA also appears to have had at least some form of a cell wall even though cell wall composition is not universal across the bacteria, archaea, and eukaryotes. The LBCA and LACA appear to have evolved additional functions related to transmembrane transport and cell reproduction and exhibit a parallel evolution of cytoskeletal elements. However, it is also possible that any or all of these features were present in the LUCA as well, but were not reconstructed as such by our methods. The LBCA also evolved several other traits including cell killing through toxins and the ability to actively facilitate horizontal gene transfer, suggesting that it lived within a complex microbial ecology (Goldman and Kacar 2023 ). Perhaps most intriguingly, LUCA appears to have more cellular functions in common with the LBCA than the LACA, suggesting that cellular organization in the LUCA was more like that of bacteria than archaea. If this trend is also true for phospholipid biosynthesis, it would imply that the LUCA membrane was composed of bacteria-like phospholipids, i.e., fatty acid tails and an ester-linked phosphate head group, and that the archaeal phospholipids with isoprenoid tails and an ether-linked phosphate head group were derived in the LACA lineage following the divergence of the LUCA. Future studies pairing phylogenetic analysis with ancestral reconstruction will provide greater detail about specific protein families present in the LUCA, LACA, and LBCA, as well as the molecular functions that those proteins were performing in ancient life."
} | 2,552 |
33361324 | PMC7762795 | pmc | 5,887 | {
"abstract": "Phototrophic organisms are key components of many natural environments. There exist two main phototrophic groups: species that collect light energy using various kinds of (bacterio)chlorophylls and species that utilize rhodopsins.",
"introduction": "INTRODUCTION The ability to use light energy is an important trait widespread within aquatic microbial communities. Photoautotrophic phytoplankton harvest light using chlorophyll, evolve oxygen, and fix inorganic carbon using RubisCO ( 1 ). In addition to these dominant oxygenic phototrophs, there exist a large number of photoheterotrophic organisms, which harvest light to supplement their mostly heterotrophic metabolism. There are two main groups of aquatic photoheterotrophic bacteria: aerobic anoxygenic phototrophic (AAP) bacteria and rhodopsin-containing bacteria. Both groups are commonly retrieved from euphotic zones of the world oceans ( 2 – 5 ) and from limnic environments ( 6 , 7 ). AAP bacteria harvest light using bacteriochlorophyll (BChl), but in contrast to purple nonsulfur photosynthetic bacteria, they are obligate aerobes requiring oxygen for their metabolism and growth ( 8 ). Upon illumination, they drive electron transport and pump protons across the membrane, which are subsequently utilized for ATP synthesis. The metabolic utilization of harvested energy has been demonstrated under laboratory conditions ( 9 , 10 ) and in field experiments ( 11 ). Rhodopsins represent a diverse family of molecules that serve multiple functions. While bacteriorhodopsins, proteorhodopsins (PR), and xanthorhodopsins (XR) serve as proton membrane pumps in Proteobacteria ( 12 ), XR is a PR-like proton pump containing in addition to retinal another chromophore, salinixanthin, which serves as a light-harvesting antenna ( 13 , 14 ). Sensory rhodopsins serve as photoreceptors in vertebrates, including humans. In contrast to bacteriorhodopsin containing Archaea , the role of proteorhodopsin in bacteria remains ambiguous. The first experiments showed no growth stimulation by light in Pelagibacter ubique strain HTCC1062 ( 15 ). In contrast, the illumination of Dokdonia sp. strain MED134 ( Bacteriodetes ) and Vibrio sp. strain AND4 ( Gammaproteobacteria ) enhanced growth and increased survival under starvation conditions, which indicates that PR provided energy for growth ( 16 – 18 ). The potential coexistence of two different phototrophic mechanisms in a single AAP bacterium was suggested for Fulvimarina pelagi (order Rhizobiales ), whose genome sequence contains a XR gene as well as photosynthetic genes ( 19 ). Recently, a co-occurrence of the pufM gene, which encodes the M subunit of the bacterial reaction centers, and XR-like genes was found in three Roseiflexus (phylum Chloroflexi ) genomes ( 20 , 21 ). In Cyanobacteria , sensory rhodopsins were found to accompany chlorophyll-based photosynthetic machinery ( 22 – 25 ). Bacteria of the genus Sphingomonas ( Alphaproteobacteria ) are common in many environments, such as soils, fresh waters, or phyllospheres ( 26 – 31 ). While most of the cultured Sphingomonas species are heterotrophs, there also exist species employing BChl-based reaction centers ( 32 , 33 ) and species containing rhodopsin genes ( 29 , 34 , 35 ). Culture-independent studies documented that Sphingomonas with BChl genes are very common in freshwater photoheterotrophic communities ( 11 , 36 – 38 ). Analysis of freshwater bacterioplankton in the oligotrophic alpine lake Gossenköllesee (Tyrolean Alps, Austria) revealed that phototrophic Sphingomonas dominates the local AAP community ( 30 ). Since no AAP Sphingomonas has been characterized in the laboratory, we revisited the Gossenköllesee and cultured novel Sphingomonas species. We characterized their photosynthetic apparatus and its gene expression to better understand how these organisms use photosynthesis in their natural environment.",
"discussion": "DISCUSSION Control of AAP expression. While PGC or rhodopsin genes have been found in many bacterial species, the simultaneous presence of both systems for harvesting light energy in one organism is unique. Our results document that under low-glucose conditions, the cells assemble fully functional photosynthetic core complexes containing BChl a and spirilloxanthin as an auxiliary pigment. The absence of the peripheral light-harvesting complex LH2 is relatively common among AAP bacteria ( 8 ). In contrast to the closely related phototrophic purple nonsulfur bacteria performing anaerobic anoxygenic photosynthesis, AAP bacteria do not respond to oxygen by shutting down their PS apparatus. On the contrary, oxygen is strictly required for BChl a synthesis in these bacteria ( 8 ). Our transcriptomic data suggest that expression of PS genes in AAP bacteria may be inhibited not only by light, as documented earlier ( 42 , 51 ), but also by higher concentrations of OC. A similar result was reported for the freshwater AAP bacterium Roseateles depolymerans ( 52 ). However, this study was focused only on the expression of the puf operon. These authors concluded that transcription of the puf genes is controlled by changes not only in light intensity and oxygen tension but also in carbon sources ( 52 ). While Suyama and coworkers ( 52 ) argued that Roseateles depolymerans upregulates its puf operon under low-OC conditions, our study shows that in photoheterotrophic AAP5, the photosynthetic apparatus is specifically repressed by glucose and galactose but not rhamnose, pyruvate, or complex C sources. The presence of monosaccharides might be indicative of the presence of OC in general and signals to the bacterium that costly biosynthesis of the photo-apparatus is not needed. Under low-OC availability, photoheterotrophic bacteria generate energy from light to decrease its carbon demand ( 9 ). This trophic strategy may be beneficial especially for opportunistic species such as Sphingomonas . The marine AAP strain Dinoroseobacter shibae uses this energy to generate ATP ( 53 ) to increase its biomass yield ( 10 ). The simultaneous transcriptional activation of the PGC and TonB transport system in AAP5 suggests a potential additional utilization of light energy. TonB-dependent transporters (TBDTs) exploit the proton motive force for the import of nutrients across the outer membrane into the periplasmatic space. TBDTs were initially discovered as transporters for iron-siderophore complexes. However, it is now clear that some representatives of TBDTs can import a variety of nutrients, including vitamins but also carbohydrates (reviewed by Noinaj and coworkers [ 54 ]). Recently, lignin-derived aromatic compounds have been identified as novel substrates for a TBDT of Sphingobium sp. strain SYK-6 ( 55 ). Sphingomonadaceae from extreme oligotrophic environments often encode large numbers (up to 134) TBDTs in their genomes ( 56 ). The genome of AAP5 contains 64 TonB-dependent receptor proteins. To thrive in an oligotrophic environment such as Gossenköllesee, AAP5 may use the proton gradient to concentrate scarce nutrients in the periplasm from where they can be imported into the cell. Light-driven import of thiamine via a TBDT has indeed been suggested for proteorhodopsin-containing flavobacterium Dokdonia sp. MED134 and DSW1 ( 18 ). Rhodopsin functionality. The XR gene seemed to be fully functional, and the complete pathway for the synthesis of its chromophore retinal is present in the genome. However, it did not contain a crtO gene, coding for the carotenoid antenna of the XR. Transmembrane domain analysis of the XR gene predicts that the putative protein contains seven transmembrane domains, which is a conserved hallmark of all rhodopsins. Its amino acid sequence clustered with other known XR genes with strong statistical support (78% maximum likelihood [ML] bootstrap). The presence of characteristic conserved amino acid residues ( 2 , 47 , 48 ) in the sequence suggested that the identified rhodopsin absorbs green light and can tentatively interact with the keto-carotenoid. The former was confirmed when transformed E. coli cells overexpressing this rhodopsin gene exhibited an absorption maximum in the green region of the absorption spectrum, and the latter when E. coli displayed a characteristic pink color after the retinal amendment. Both changes demonstrated that the heterologously expressed rhodopsin protein was properly folded and functional. In contrast to the recombinant E. coli strain overexpressing the rhodopsin gene, we were not able to detect any absorption peak of rhodopsin in the membrane fraction during the purification of PS complexes. This is in agreement with the fact that the identified rhodopsin gene was virtually not expressed under experimental conditions. The flash-induced transient absorption data confirm that the product of the XR gene is capable of performing the photocycle. Our analysis of the kinetic data resolved four components with time constants in the microsecond-to-millisecond range. The fastest process resolved had a time constant of 7.3 μs, which is in perfect agreement with the data from XR of S. ruber ( 13 ). The slowest phase of the XR photocycle was found to be ∼100 ms, also in agreement with the cited work. In our case, the data do not support the resolution with a total of six kinetic components, as suggested by Balashov and coworkers ( 13 ). However, it should be noted that unlike our XR sample, the system studied in the cited work also contained the salinixanthin antenna pigment whose electrochromic response contributes to the complexity of the absorption data. Why are there two systems for light harvesting? The main question of why an organism keeps in its genome two different systems for capturing light energy remains. One possibility is that these systems work together, and the metabolic benefit for the organisms is higher than when using only one of the systems. One such cooperation is found during oxygenic photosynthesis, where the coupled action of two photosystems makes it possible to bridge large redox potential necessary for extracting electrons from water ( 1 ). However, there is no support for this hypothesis in the case of Sphingomonas sp. AAP5, since no xanthorhodopsin was found in the membranes together with BChl a -containing photosystems. The other option is that AAP5 utilizes the two systems under different conditions. We showed that BChl a -containing photosystems are used under low-glucose conditions. There may be some very specific conditions, which are not suitable for bacteriochlorophyll and where the bacterium could use the rhodopsin system, but it is very difficult to test this hypothesis under laboratory conditions. Although we did not observe XR expression in our experiments, it is certainly possible that there is a specific physiological condition or external stimulus that induces its synthesis. On the other hand, there is good reason to believe that XR in AAP5 may have another function than light harvesting. Phylogenetic inference placed the XR sequences on the same branch as nonbacterial dinoflagellate-type rhodopsins. In the marine dinoflagellate Prorocentrum donghaiense , this phylogenetic affiliation was recently confirmed, although it seems these two groups of rhodopsins have a distinct evolutionary origin. Sensory rhodopsins were found to accompany chlorophyll-based photosynthesis in cyanobacteria ( 22 – 25 ). It is hard to imagine that XR can provide a significant energy benefit for algae because they can rely on an effective energy source such as oxygenic photosynthesis. Moreover, AAP5 does not contain the crtO gene, coding for the carotenoid antenna of the XR, which significantly restricts its light absorption properties. The XR gene of AAP5 is in one operon with a histidine kinase gene. Interestingly, the existence of histidine kinase rhodopsin (HKR) was recently reported for the marine alga Chlamydomonas reinhardtii ( 57 ). HKRs are modular proteins containing rhodopsin, a His kinase, a response regulator, and in some cases, an effector domain. All these lines of evidence indicate that the XR gene may have a function other than light energy harvesting. Therefore, we hypothesize that the rhodopsin might represent part of a light-sensing system, rather than a light-harvesting system covering cellular energy needs. The XR together with other numerous photoreceptors existing in this bacterium allows it to react to changes in incident light, an important environmental factor in the sunlit waters from which this strain was isolated."
} | 3,152 |
37810646 | PMC10551918 | pmc | 5,888 | {
"abstract": "Synthetic hydrogels\nstruggle to match the high strength, toughness,\nand recoverability of biological tissues under periodic mechanical\nloading. Although the hydrophobic polymer chain of polystyrene (PS)\nmay initially collapse into a nanosphere upon contact with water,\nit has the ability to be elongated when it is subjected to an external\nforce. To address this challenge, we employ the reversible addition–fragmentation\nchain transfer (RAFT) method to design a carboxyl-substituted polystyrene\n(CPS) which can form a covalently cross-linked network with four-armed\namino-terminated polyethylene glycol (4-armed-PEG-NH 2 ),\nand a ductile polyacrylamide network is introduced in order to prepare\na double-network (DN) hydrogel. Our results demonstrate that the DN\nhydrogel exhibits exceptional mechanical properties (0.62 kJ m –2 fracture energy, 2510.89 kJ m –3 toughness, 0.43 MPa strength, and 820% elongation) when a sufficient\nexternal force is applied to fracture it. Moreover, when the DN hydrogel\nis subjected to a 200% strain, it displays superior recoverability\n(94.5%). This holds a significant potential in enhancing the mechanical\nperformance of synthetic hydrogels and can have wide-ranging applications\nin fields such as tissue engineering for hydrophobic polymers.",
"conclusion": "4 Conclusions Our experimental findings provide\ninsights into the mechanisms\nof deformation and energy dissipation in these hydrogels. When the\nPS/PAAm DN hydrogel is subjected to a small stretch, the Young’s\nmodulus of the PS/PAAm DN hydrogel is nearly close to that of PS/PEG\nSN. Thus, the polystyrene chains bear loads at the beginning of stretching\nin the PS/PAAm DN hydrogel. The load sharing of the two networks may\nbe achieved by entanglements of the polymers and by the covalent cross-links\nformed between the amino groups on PEG chains and the carboxyl groups\non PS chains ( Figure 1 ). Our data demonstrate that the PS/PAAm DN hydrogels consisting\nof polystyrene nanospheres have the ability to be elongated under\nan external force. When the external force is removed, it can be restored\nto approximately its original state (94.5%). The chemically cross-linked\nnetwork structures make the DN hydrogels have high mechanical properties.\nOur experiment provides a new idea for the design of a fully chemically\ncross-linked DN hydrogel with recoverability and also provides a wider\nplatform for the application of hydrophobic polymers.",
"introduction": "1 Introduction Hydrogel is a classical\ntype of polymer with a three-dimensional\nnetwork structure, which can be water-swollen and absorbs amounts\nof water, but insoluble in water. 1 , 2 Therefore,\nhydrogels are used for tissue engineering, 3 − 5 as vehicles\nfor drug delivery, 6 − 8 as biosensors, 9 − 11 and so forth. However, hydrogels\ndisplay brittleness and ineffective energy dissipation due to the\nexistence of water, 12 resulting in poor\nmechanical properties and limiting their wide applications. Many efforts\nhave been devoted to developing high-strength hydrogels by introducing\nnovel network microstructures and cross-linking strategies, such as\ntetra-polyethylene glycol (PEG) hydrogel, 13 − 16 slide-ring hydrogel, 17 nanocomposite hydrogel, 18 hydrophobically associated (HA) hydrogel, 19 macromolecular microsphere composite hydrogel, 20 double-network (DN) hydrogel, 2 , 21 , 22 and so forth. In order to address these issues,\nDN hydrogels are feasible examples\nthat have been used to combine the high mechanical strength and toughness\nowing to the contrasting network structure and effective energy dissipation. 2 , 23 , 24 Specifically, DN hydrogels comprise\ntwo covalently linked networks, that is, the first one is rigid and\nbrittle, which maintains the hydrogel shape, and the second is loose\nand ductile, which fills in the rigid network and absorbs external\nstress. 23 , 25 , 26 For example,\nalginate–polyacrylamide DN hydrogel can be stretched more than\n20 times its original length and has a fracture energy of 9000 J m –2 . 27 The hydrogel cannot\nrecover its mechanical properties upon cyclic mechanical loading,\nif the rigid network, which allows to efficiently dissipate mechanical\nwork and gives rise to high mechanical strength, is ruptured. 23 To overcome such limitations, since the first\nfully chemical poly(2-acrylamido-2-methylpropanesulfonic acid)/poly(acrylamide)\nDN hydrogel was invented by Gong et al., 2 various fully chemical DN hydrogels have obtained attention due\nto their good mechanical stability, 23 including\nmicrogel-reinforced DN hydrogels, 28 void-DN\nhydrogels, 29 and triple-network hydrogels. 30 These fully chemical DN hydrogels, composed\nof chemically cross-linked networks, showed excellent tensile stress\n(0.1–3 MPa), tensile strain (1000–2000%), and high fracture\nenergy (100–1000 J m –2 ). 26 , 31 However, due to the irreversible chain breakage of covalent bonds\nin the rigid and brittle first network, they cannot recover their\noriginal conformation after the first loading, which is one of the\nmain limitations for their further applications. Compared to\nfully chemical DN hydrogels, hybrid DN hydrogels, 32 − 35 which consist of physical first\nnetwork and chemical second network,\nalso display high strength and toughness as well as excellent recovery\nproperty from the loading owing to the reversible physical first network.\nHA hydrogels refer to physically cross-linked hydrogels formed by\nhydrophobic interactions and exhibit the enhancement of mechanical\nstrength; 19 hydrophobic interactions are\nused to construct a sacrificial network to render the DN hydrogels\nof great mechanical recovery properties. 24 Liu and co-workers first prepared HA hydrogels via using acrylamide\n(AAm) as the main component and octyl phenol polyethoxy ether as hydrophobic\nsegments. 36 PEG-based polymers have been\nwidely used to prepare covalently cross-linked hydrogels with hydrophobic\ndomains. 13 , 37 , 38 Inspired by\nthese hybrid DN hydrogels, we expect to use hydrophobic polymers to\ndesign a fully chemically cross-linked DN hydrogel with high mechanical\nand recovery properties. In this work, we propose a new kind\nof method to design a fully\nchemical DN hydrogel by introducing functional polystyrene (PS). To\nour knowledge, PS has long been used to process elastomers or textiles. 39 − 41 Wang has developed a novel nanoparticle-reinforced polyacrylamide-based\nhydrogel with high mechanical strength (compression strength and tensile\nstrength up to 7.0 and 2.0 MPa), but it has an apparent shortcoming,\nthat is, low stretchability. 42 We have\nsynthesized carboxyl-substituted polystyrene (CPS, 5762 Da) by employing\nthe reversible addition–fragmentation chain transfer (RAFT) 43 , 44 polymerization method to synthesize the target polymer that possesses\nthe carboxyl group at two termini ( Scheme 1 , Supporting Information). Weight-average molecular weight ( M w ), number-average molecular weight ( M n ), and polydispersity of CPS\nare quantified by gel permeation chromatography ( Figure S1 , Supporting Information). CPS and four-armed amino-terminated\nPEG (4-armed-PEG-NH 2 , 10000 Da) are chosen to construct\nthe first network ( Figure 1 a and Scheme 2 , Supporting Information);\nthe corresponding network is named PS/PEG single network (SN) hereinafter.\nThen, by introducing the second network of covalently cross-linked\npolyacrylamide (PAAm) into the first network ( Figure 1 b), the stretchable fully chemically cross-linked\nDN hydrogel is synthesized ( Figure 1 c), and the corresponding hydrogel is named the PS/PAAm\nDN hydrogel hereinafter. The chain of the CPS polymer would collapse\ninto a nanosphere when the DN hydrogel is immersed in an amount of\ndeionized water ( Figure 1 d). The tetra-functionalized PEG is chosen due to its high cross-linking\nefficiency than that of the bifunctional PEG. 24 By contrast, AAm and N, N ′-methylenebis(acrylamide)\n(MBAA) are chosen to build the second network ( Figure 1 b and Scheme 3 , Supporting Information). Moreover, we have to mention that the\ncross-linking density for the first and second networks is critical\nfor a notable increase in the mechanical strength of DN hydrogels. 2 A remarkable increase in mechanical strength\noccurs when the first network is highly cross-linked, and the second\none is loosely cross-linked. The chemical structures of AAm, MBAA,\nCPS, and 4-armed-PEG-NH 2 are shown in Figure 1 . Figure 1 Design and preparation\nof the PS/PAAm DN hydrogel. (a) In the first\nnetwork precursor solution, the amino groups (gray balls) on 4-armed-PEG-NH 2 polymer chains form covalent cross-links through the carboxyl\ngroups (yellow balls) on CPS polymer chains. (b) In the second network\nprecursor solution, the first network is joined by covalent cross-links\n(purple cycles), and the AAm (black points) polymer chains form covalent\ncross-links through MBAA (gray points). (c) In the PS/PAAm DN hydrogel,\nthe second network is joined by covalent cross-links (green squares),\nand the two types of polymer network were intertwined. (d) CPSs collapse\ninto nanospheres (blue balls) when the PS/PAAm DN hydrogel is immersed\nin an amount of deionized water.",
"discussion": "3 Results and Discussion 3.1 Characterization of the PS/PAAm DN Hydrogels We first\nemploy the FTIR spectra to analyze the chemical structure\nof the PS/PAAm DN hydrogel, PAAm SN hydrogel, and PS ( Figure 2 a). As we can see from Figure 2 a, the bands around\nthe PS/PAAm DN hydrogel and PAAm SN hydrogel appear at 1655 cm –1 , corresponding to −C=O– on the\npeptide bond, and the bands around PS/PAAm DN hydrogel and PS appear\nat 696 cm –1 , corresponding to −C–H–\non the benzene ring. We consider that the carboxyl group on CPS and\nthe amino group on 4-armed-PEG-NH 2 can covalently cross-link\nto form a rigid network for the PS/PAAm DN hydrogel. In addition,\nthe 1 H NMR spectrum could also confirm our results ( Figure S2 , Supporting Information). All the peaks\ncan be assigned as follows: δ: 0.9 (m, 3H, −CH 3 ), 1.3 (m, 2H, −CH 2 −), 1.5 (m, 1H, −CH≡\n), 2.7 (m, 2H, −CH 2 –COOH), 3.4 (m, 2H, −S–CH 2 −), 5.2 (m, 1H, – S–C 6 H 5 –CH−), and 6.6–7.2 (m, 5H, −C 6 H 5 −). Based on the above data, we demonstrate\nthat we have successfully added the carboxyl groups to PS. To characterize\nhow the stretching process of CPS affects the property of the PS/PAAm\nDN hydrogels, second, we measure the force–extension curve\nof the CPS using atomic force spectroscopy (AFM)-based SMFS in air\nat room temperature. AFM-based SMFS technology is helpful to establish\nthe relationship between the chain structure, chain composition, and\nsingle-chain elasticity of synthetic polymers and the interaction\nbetween chains and their macromechanical properties and to understand\nthe relationship between the structure and interaction of biological\nmacromolecules and their biological functions. 47 − 50 Figure 2 b shows the typical force–extension\ncurve of the unfolding CPS nanosphere in water (blue line). Driven\nby hydrophobic interactions, the CPS collapses into a compact nanospherical\nstructure at the beginning of the force–extension curves. 51 The long force plateau indicates that the unfolding\nnanosphere is under constant force before the collapsed nanosphere\ntransforms into a fully extended chain. The rupture peak shows that\nthe extended chain is further stretched until a certain chemical bond\nbreaks. This stretching process can be well fitted by the worm-like\nchain (WLC) model (red line). 52 , 53 The CPS attaches to\nthe monocrystalline silicon wafer in water, showing a nanosphere structure\nwith a height of 2–8 nm by AFM imaging ( Figure 2 c). In order to clarify the microstructure,\nwe next employ the SEM observation of a PS/PAAm DN hydrogel after\nfreeze-drying. The high glass-transition temperature ( T g ) of CPS and PAAm (80, 188 °C) can maintain the\nstructure after the dehydration of the DN hydrogel, which makes it\nsuitable for structure observation. The SEM images reveal that some\nporous 3D networks exist in the PS/PAAm DN hydrogel ( Figure 2 d). The PS/PAAm DN hydrogel\ncan be stretched to 8 times its original length without rupture ( Figure 2 e,f); mechanical\ntests are performed in air at room temperature using a tensile machine\nwith a 10 N load cell. In both loading and unloading, the rate of\nstretching is kept constant at 1 mm min –1 . Figure 2 Characterization\nof the PS/PAAm DN hydrogels. (a) FTIR spectra\nof the PS/PAAm DN hydrogel and PS and PAAm SN hydrogels. (b) Force–extension\ncurves of unfolding CPS nanospheres in deionized water. The blue line\nis the force–extension curve. The red curve is the WLC fitting\nto the elastic stretching part. The black line is the baseline. (c)\nAFM image shows the CPS nanosphere structure with a height of 2–8\nnm. (d) SEM images of the PS/PAAm DN hydrogel at different magnifications.\nOptical images of the PS/PAAm DN hydrogel before (e) and after (f)\nthe stretching. 3.2 Mechanical\nProperties of the PS/PAAm DN Hydrogels We chose 25 mM CPS\nto construct the PS/PAAm DN hydrogels because\nit has superior mechanical properties than that of 10 and 30 mM CPS\n( Figure S3 , Supporting Information). The\nPS/PAAm DN hydrogels are even more remarkable when compared with their\nparents, the PS/PEG SN and PAAm SN hydrogels ( Figure 3 a). The molar concentrations of CPS and AAm\nin the PS/PAAm DN hydrogels are kept the same as those in the PS/PEG\nSN and PAAm SN hydrogels, respectively. When the elongation is small\n(3.9% strain), the Young’s modulus of the PS/PAAm DN hydrogel\nis 222.05 kPa, which is close to the Young’s modulus of PS/PEG\nSN (219.27 kPa). The maximum strength and fracture strain are, respectively,\n143 and 820% for the PS/PAAm DN hydrogel, 177 and 280% for the PS/PEG\nSN hydrogel, and 5 and 670% for the PAAm SN hydrogel. Thus, the fracture\nstrain of the PS/PAAm DN hydrogels exceeds those of either of its\nparents, and the fracture stress far surpasses the PAAm SN hydrogels. Figure 3 Mechanical\nproperties of the PS/PAAm DN hydrogels. (a) Stress–strain\ncurves of the three types of hydrogels, with each stretching to rupture.\nIllustration shows the strain at 10%. (b) Each hydrogel is loaded\nto a strain of 200%, just below the value that ruptures PS/PEG SN,\nand is then unloaded. Samples of the PS/PEG SN (c) and PS/PAAm DN\nhydrogels (d) are subjected to a cycle of loading and unloading of\nvarying maximum strain (50, 100, 150, and 200%). Illustration shows\nthe energy dissipation of the PS/PEG SN and PS/PAAm DN hydrogels under\nthe corresponding strain, respectively. Frequency sweep of the storage\nmodulus ( G ′) and loss modulus ( G ″) of the PS/PEG SN (e) and PS/PAAm DN hydrogels (f), respectively. The PS/PEG SN hydrogel effectively dissipates more\nenergy than\ndoes the PS/PAAm DN hydrogel. The area between the loading and unloading\ncurves of a hydrogel gives the energy dissipated per unit volume. Figure 3 b shows that the\nPS/PEG SN and PS/PAAm DN hydrogels have different energy dissipation\nat 200% strain, that is, 33.89 and 17.81 kJ m –3 ,\nrespectively. Next, we study the energy dissipation of PS/PEG SN and\nPS/PAAm DN hydrogels at different strains (50, 100, 150, and 200%). Figure 3 c shows the stress–strain\ncurves of PS/PEG SN at different strain rates. It is clear that the\nenergy dissipation of PS/PEG SN increases with the increase of strain\n(from 3.21 to 33.89 kJ m –3 in the illustration).\nPS/PEG SN exhibits pronounced hysteresis and retains significant permanent\ndeformation after unloading. In contrast, the PAAm SN hydrogel fully\nrecovers its original length after unloading due to the low energy\ndissipation ( Figure 3 b). The PS/PAAm DN hydrogel also shows hysteresis, but the permanent\ndeformation after unloading is significantly smaller than that of\nthe PS/PEG SN hydrogel. Figure 3 d shows the stress–strain\ncurves of the PS/PAAm DN hydrogel at different strain rates. It is\nclear that the energy dissipation of PS/PAAm DN hydrogels also increases\nwith the increase of strain (from 2.82 to 17.81 kJ m –3 in the illustration) but is smaller than that of the PS/PEG SN hydrogel.\nWe suppose that PAAm SN makes the PS/PAAm DN hydrogel tougher and\ncloser to the elastomer than PS/PEG SN, and the energy dissipation\ngradually increases with the strain, suggesting that more cross-linkers\nwould rupture at a large strain. Rheological studies ( Figure 3 e,f) show that both PS/PEG\nSN and PS/PAAm DN hydrogels have the storage modulus ( G ′) and loss modulus ( G ″) in which G ′ > G ″, and G ′ of PS/PEG SN is twice as that of the PS/PAAm DN hydrogel.\nWe consider that the PS/PAAm DN hydrogel is a softer hydrogel. Furthermore, the cross-linking density of the second network, that\nis, the concentration of MBAA ( Figure S4 , Supporting Information), and the stretch rate ( Figure S5 , Supporting Information) also strongly affect the\nmechanical behavior of the PS/PAAm DN hydrogels. We consider that\nthe mechanical properties of PS/PAAm DN hydrogels are weakened due\nto the low density when the concentration of CPS is 10 mM because\nthe first network provides rigidity to the PS/PAAm DN hydrogels; 2 on the contrary, when the concentration of CPS\nis 30 mM, the mechanical properties of PS/PAAm DN hydrogels are also\nweakened due to the “wind” or “knot” of\nthe redundant polymer molecular chains. Meanwhile, the strength of\nPS/PAAm DN hydrogels is enhanced with the increased cross-linking\ndensity of the second network, which provides ductility to the PS/PAAm\nDN hydrogels. 2 The higher the cross-linking\ndensity of the second network of PS/PAAm DN hydrogels, the greater\nis its strength and the smaller its elongation. If the cross-linking\ndensity of the second network is higher than 1 mol %, the strength\nof PS/PAAm DN hydrogels increases while the elongation decreases.\nTill now, we can make a conclusion that the PS/PAAm DN hydrogels have\nthe best mechanical properties when the concentration of CPS is 25\nmM and the cross-linking density of the second network is 1 mol %. Besides, we prepare the PS/PAAm DN hydrogels containing various\nconcentrations of AAm (2, 3, 4, and 5 M) to study why the DN hydrogels\nare much more stretchable than either of their parents. When the concentration\nof AAm is increased from 2 to 5 M, the strength of PS/PAAm DN hydrogels\nincreases from 0.14 to 0.43 MPa, respectively ( Figure 4 a). We consider that the mechanical properties\nof PS/PAAm DN hydrogels are enhanced owing to the raised concentration\nof AAm. 2 However, the water content of\nPS/PAAm DN hydrogels decreases with the increase in AAm ( Figure 4 b). The PS/PAAm DN\nhydrogel is soaked in water for 24 h to reach the swelling equilibrium,\nand the water content decreases from 84.39% at 2 M to 81.64% at 5\nM. We suppose that the three-dimensional structure of PS/PAAm DN hydrogels\nis too small with the increased cross-linking density to contain water.\nThe fracture energy reaches a maximum value of 0.62 kJ m –2 at 5 M acrylamide ( Figure 4 c), and the toughness reaches 2510.89 kJ m –3 ( Figure 4 d). When\nthe concentration of CPS increases from 10 to 30 mM, the fracture\nenergy of the PS/PAAm DN hydrogels increases from 0.079 to 0.151 kJ\nm –2 ( Table 1 , Supporting\nInformation). At the same time, when the covalent cross-linking density\nof the second network changes from 0.5 to 2 mol %, the fracture energy\nof the PS/PAAm DN hydrogels changes from 0.127 to 0.190 kJ m –2 ( Table 2 , Supporting Information). Figure 4 Stress–strain curves (a), water content (b), fracture\nenergy (c), and toughness (d) of the PS/PAAm DN hydrogels with different\nconcentrations of AAm. The cross-linking density remains the same. 3.3 Mechanical Recoverability\nof the PS/PAAm DN\nHydrogels Next, we studied the recoverability of the PS/PAAm\nDN hydrogels. The PS/PEG SN ( Figure 5 a) and PS/PAAm DN hydrogels ( Figure 5 b) are subjected to 20 stretching–relaxation\ncycles without interval. The results show that the conformation of\nPS/PEG SN becomes 92.7% of its initial state after 20 consecutive\nstretching–relaxation cycles ( Figure S6 , Supporting Information), its energy dissipation is reduced by 21.8%\n(from 41.5 to 32.45 kJ m –3 ), and the maximum strength\nis reduced by 7.3% (from 0.178 to 0.165 MPa), which is caused by the\ncovalently cross-linked network structure, while for the PS/PAAm DN\nhydrogel, the conformation becomes 94.5% of its initial state ( Figure S7 , Supporting Information), its energy\ndissipation is reduced by 10.5% (from 18.79 to 16.82 kJ m –3 ), and the maximum strength is reduced by 5.5% (from 0.144 to 0.136\nMPa). Figure 5 Recoverability of the PS/PAAm DN hydrogels. After the first cycle\nof loading and unloading of the PS/PEG SN (a) and PS/PAAm DN hydrogels\n(b), one sample is immediately reloaded 20 times. Illustration shows\nthe energy dissipation of the PS/PEG SN and PS/PAAm DN hydrogels under\nthe corresponding number of cycles, respectively."
} | 5,211 |
34255041 | PMC8350355 | pmc | 5,890 | {
"abstract": "Abstract Interest and controversy surrounding the evolutionary origins of extremely halophilic Archaea has increased in recent years, due to the discovery and characterization of the Nanohaloarchaea and the Methanonatronarchaeia. Initial attempts in explaining the evolutionary placement of the two new lineages in relation to the classical Halobacteria (also referred to as Haloarchaea) resulted in hypotheses that imply the new groups share a common ancestor with the Haloarchaea. However, more recent analyses have led to a shift: the Nanohaloarchaea have been largely accepted as being a member of the DPANN superphylum, outside of the euryarchaeota; whereas the Methanonatronarchaeia have been placed near the base of the Methanotecta (composed of the class II methanogens, the Halobacteriales, and Archaeoglobales). These opposing hypotheses have far-reaching implications on the concepts of convergent evolution (distantly related groups evolve similar strategies for survival), genome reduction, and gene transfer. In this work, we attempt to resolve these conflicts with phylogenetic and phylogenomic data. We provide a robust taxonomic sampling of Archaeal genomes that spans the Asgardarchaea, TACK Group, euryarchaeota, and the DPANN superphylum. In addition, we assembled draft genomes from seven new representatives of the Nanohaloarchaea from distinct geographic locations. Phylogenies derived from these data imply that the highly conserved ATP synthase catalytic/noncatalytic subunits of Nanohaloarchaea share a sisterhood relationship with the Haloarchaea. We also employ a novel gene family distance clustering strategy which shows this sisterhood relationship is not likely the result of a recent gene transfer. In addition, we present and evaluate data that argue for and against the monophyly of the DPANN superphylum, in particular, the inclusion of the Nanohaloarchaea in DPANN.",
"introduction": "Introduction Recent studies discovered several new archaeal lineages in hypersaline environments, including the nanosized Nanohaloarchaea and the methanogenic Methanonatronarchaeia. The exact placement of these lineages within the archaeal phylogeny remains controversial; consequently, the number of independent acquisitions of key adaptations to a halophilic lifestyle remains to be determined. Dissecting the evolutionary relationships between these new lineages and the Haloarchaea may inform on the origins of halophily and the role of genome streamlining. To thrive in extreme hypersaline environments (>150 g/l), Haloarchaea employ a “salt-in” strategy through the import of potassium ions, in which the intracellular salt concentration equalizes with the external environmental condition ( Oren 2008 ). This acts to balance the cellular osmotic pressure but also has caused significant changes in amino acid usage, leading to an overabundance of acidic residues, aspartate and glutamate (D/E) in all Haloarchaea ( Lanyi 1974 ; Madern et al. 2000 ). The evolutionary origins of the Nanohaloarchaea have remained uncertain since their discovery ( Ghai et al. 2011 ; Narasingarao et al. 2012 ). The composition of their proteome indicates that Nanohaloarchaea also use the “salt-in” strategy similar to Haloarchaea ( Narasingarao et al. 2012 ). It was originally suggested that the Nanohaloarchaea are euryarchaeota that form a clade with the Haloarchaea, based on phylogenies of the 16S rRNA gene and ribosomal proteins ( Narasingarao et al. 2012 ; Petitjean et al. 2015 ). Additional data obtained from individual cells via cell sorting followed by genome amplification and 16S rRNA sequencing analysis confirmed the original observations of the Nanohaloarchaea as a sister taxon to the Haloarchaea ( Zhaxybayeva et al. 2013 ). More recently, based on analyses of concatenated conserved protein sequences, the Nanohaloarchaea were placed in a group together with similarly nanosized organisms, the Diapherotrites, Parvarachaeota, Aenigmarchaeota, and Nanoarchaeota, forming the DPANN superphylum ( Rinke et al. 2013 ; Andrade et al. 2015 ; Castelle et al. 2015 ). Past analyses of this superphylum ( Brochier-Armanet et al. 2011 ; Raymann et al. 2014 ; Petitjean et al. 2015 ; Williams et al. 2015 ) suggested that the DPANN grouping may not reflect shared ancestry but rather an artifact due to long branches and/or small genomes. However, more recent analyses supported a monophyletic DPANN clade ( Williams et al. 2017 ). Aouad et al. performed a multilocus analysis using various models, which did not include DPANN sequences, and placed the Nanohaloarchaea with the Methanocellales and the Haloarchaea with the Methanomicrobiales ( Aouad et al. 2018 ); that is, the Nanohaloarchaea were recovered as a member of the euryarchaeota, but not as a sister group to the Haloarchaea. We note that a similar controversy surrounds the phylogenetic position of the Nanoarchaeota. Nanoarchaeum equitans was first considered a representative of a new deep branching archaeal phylum ( Huber et al. 2002 ), that is, an archaeon not a member of the euryarchaeotes or crenarchaeotes. However, later analyses of ribosomal proteins, phylogenetically informative HGTs, and signature genes led to the conclusion that N. equitans may represent a fast-evolving euryarchaeote instead of an early branching novel phylum ( Brochier et al. 2005 ; Dutilh et al. 2008 ; Urbonavičius et al. 2008 ). Several more recent analyses placed the Nanoarchaeota inside of the DPANN ( Adam et al. 2017 ; Dombrowski et al. 2019 ; Spang et al. 2017 ), reflecting the ongoing controversy in the phylogenetic placement of these groups. Recently, another group of extreme halophiles, the Methanonatronarchaeia (also spelled as Methanonatronarcheia), were discovered and predicted to also use the “salt-in” strategy ( Sorokin et al. 2017 ). Initial multilocus phylogenetic analyses placed these methanogenic halophiles in a monophyletic clade with the Haloarchaea, suggesting they are an evolutionary intermediate between methanogens and modern halophiles. However, several recent studies have contested this placement: a multilocus data set placed the Methanonatronarchaeia basal to a superclass named Methanotecta, a group that includes the Archaeoglobales, class II methanogens and Haloarchaea ( Adam et al. 2017 ; Aouad et al. 2019 ; Martijn et al. 2020 ). In addition, to the three extreme halophiles mentioned, the recently characterized Hikarchaeia has been identified as a nonhalophilic sister group to the Haloarchea ( Martijn et al. 2020 ). Temporal analysis of the Hikarchaeia divergence from the Haloarchaea may shed light on the genomic events that prelude the Haloarchaea’s adaptation to hypersalinity (see Discussion). Several conclusions can be drawn from these latter results with regard to adaptation to a halophilic lifestyle, most noteworthy of which is the convergent evolution of the “salt-in” strategy among these three lineages. Independent adaptation to hypersalinity in extreme halophiles is certainly a viable evolutionary hypothesis; this is seen in the case of the Salinibacter and Salinicoccus ( Mongodin et al. 2005 ). However, if Nanohaloarchaea, Haloarchaea, and Methanonatronarcheia form a monophyletic group, as seen with some analyses of 16S rRNA and ribosomal proteins, the hypothesis of common ancestral origins can more easily account for the evolutionary development of the salt-in strategy. The evolutionary relationships of the three extreme halophilic archaeal lineages remain unresolved; figure 1 summarizes the current controversies. This lack of resolution can, at least in part, be due to biases that are known to complicate phylogenetics. The genomes of the Methanonatronarchaeia and Nanohaloarchaea are comparatively small with average genome sizes of <2.1 and ∼1.1 Mb, respectively. Furthermore, most genome entries in public databases are incomplete. The Haloarchaea are known to be highly recombinogenic ( Boucher et al. 2004 ; Naor et al. 2012 ; Williams et al. 2012 ; Mohan et al. 2014 ; Méheust et al. 2018 ) and are physically associated with at least some of the Nanohaloarchaea ( Andrade et al. 2015 ; Cono et al. 2020 ; Hamm et al. 2019 ). Fig. 1. Summary of proposed placements of halophilic lineages mapped on an Archaeal reference tree. This reference tree mostly depicts the positions of various euryarchaea. Individual taxa have been collapsed into higher taxonomic groups. The red (R) indicators represent the different placements proposed for the Nanohaloarchaea, whereas the purple (P) indicators are used for the Methanonatronarchaeia. Sources for each placement: R1 ( Narasingarao et al. 2012 ), R2 ( Andrade et al. 2015 ), R3 ( Aouad et al. 2018 ); P1 ( Sorokin et al. 2017 ) and P2 ( Aouad et al. 2019 ). Phylogenies based on the concatenation of many genes face many problems: 1) genes have different evolutionary histories (e.g., duplication and transfer) and forcing the histories of all the genes on a single tree does not reflect the complex evolutionary history of the genomes ( Lapierre et al. 2014 ). In particular, genes acquired from outside the group under consideration may create a strong signal for placing the recipient of the transferred gene at the base of the group. 2) Genes experience differing levels of purifying selection, especially between different lineages. This can lead to long branch attraction (LBA) artifacts ( Felsenstein 1978 ), even if the individual genes evolved along the same history as the host species ( Philippe et al. 2005 ). 3) Substitution bias may create convergent signals in distantly related groups. The work reported here was guided by the hypothesis that the phylogenetic reconstruction of a single, slowly evolving gene might be more robust against artifacts of phylogenetic reconstructions compared with analyses that are based on large sets of genes that may represent different evolutionary histories, include missing data, and contain genes with high substitution rates. We reconstruct single gene alongside multilocus phylogenies to correct for these sources of bias and to critically assess the evolutionary relationships of the Haloarchaea, Nanohaloarchaea, and Methanonatronarchaeia. We also cluster and dissect the evolutionary relationships of the gene families in the Nanohaloarchaeal core genome, using a gene family clustering technique. The ATP synthase catalytic and noncatalytic subunits, AtpA and AtpB, represent extremely slow evolving genes ( Gogarten 1994 ) conserved throughout Archaea and are among the slowest evolving genes in cellular organisms ( supplementary table S5 , Supplementary Material online ). The evolution of these subunits may be slow enough to ameliorate rate signal bias and minimize compositional heterogeneity that otherwise plague reconstructions that includes DPANN and Haloarchaeal sequences. ATP synthase subunits have been used successfully as a phylogenetic marker for large-scale reconstructions ( Gogarten and Taiz 1992 ); however, a drawback of the ATPases is that they are known to have been transferred between divergent phyla ( Olendzenski et al. 2000 ). Recently, Wang et al. convincingly showed the transfer of this operon lead to the adaptation of Thaumarchaeota to more acidic environments ( Wang et al. 2019 ). The same authors drew a similar conclusion when the Nanohaloarchaea–Haloarchaea sister group was recovered, which the authors interpreted as suggesting HGT of the ATPase genes in the Nanohaloarchaea–Haloarchaea. To shed light on this HGT hypothesis, we cluster and correlate the gene families in the Nanohaloarchaea and contrast the position of the ATPase genes in these clusters to the same genes in the Thaumarchaeota. We also provide a more robust sampling of the Nanohaloarchaea; we include seven newly sequenced and assembled nanohaloarchaeal genomes together with existing genomes mined from the NCBI database. Robust sampling of the taxa of interest, like the one offered here, has the potential to improve the recovery of evolutionary relationships without adding more sites (genes) ( Graybeal 1998 ). In maximum likelihood and Bayesian phylogenies, we find that the Nanohaloarchaea group robustly with the Haloarchaea in the single gene phylogenies, whereas the Methanonatronarchaeia were placed as a deeper branching euryarchaeal lineage, most likely at the base of the Methanotecta superclass. In large, concatenated data sets, we recover a monophyletic DPANN (including the Nanohaloarchaea). We also provide evidence that the ATPase genes have likely not been transferred in the case of the Nanohaloarchaea–Haloarchaea, and contrast this specific relationship with the clearly transferred ATPases in the Thaumarchaeota.",
"discussion": "Discussion Sisterhood of Nanohaloarchaea and Haloarchaea Analysis of the catalytic and noncatalytic subunits of the archaeal ATP synthase group the enzyme from Nanohaloarchaea as a sister group to the Halobacterial (Haloarchaeal) subunits ( fig. 2 ; Wang et al. 2019 ). This strongly supported grouping is also recovered when the data are recoded to reduce compositional bias, when alignment columns containing acidic residues in both the Nanohaloarchaea and the Haloarchaea are deleted, and when the CAT-GTR model (a model that is less sensitive to compositional effects and long branch attraction artifacts) is used in phylogenetic reconstruction. None of these analyses recovered the DPANN clan, however, this may not be strong evidence against the existence of DPANN, as HGT in members of DPANN is largely unexplored in this work. Placement of the individual DPANN groups in the ATP synthase phylogenies and the absence of an ATPase in some ectoparasitic Nanoarachaeota ( Wurch et al. 2016 ) may be interpreted as evidence questioning whether the ancestor of the DPANN even possessed a functional ATP synthase. The N. equitans genome is an example of a DPANN member which encodes the typical archaeal ATPase headgroup (3A and 3B subunits), although it may lack ATP synthesis activity ( Mohanty et al. 2015 ). However, homology between the F (from Bacteria) and A type (from Archaea) ATPases demonstrates that these ATPases are older than the origin of Archaea ( Gogarten et al. 1989 ; Gogarten and Taiz 1992 ). Although ATPases are known to have been horizontally transferred, the described findings suggest that the Nanoarchaeal ancestor possessed an ATPase. It is certainly possible that modern DPANN genomes replaced their ATP synthases with homologs from their hosts or from other organisms occupying the same environment. Given the consistent support for the Nanohaloarchaea–Haloarchaea clade in the AtpA and AtpB phylogenies, it is unlikely that this finding is due to compositional bias or long branch attraction. Two conflicting hypotheses can reconcile our findings with those of previous analyses based on concatenation of several genes or on gene tree/species tree reconciliations: 1) the ATP synthase was acquired by the ancestor of the Nanohaloarchaea from a relative of the Haloarchaea or 2) the previous multilocus analyses do not reflect evolutionary history, but are artifacts due to high substitution rates, gene transfer, and small genomes; and the Nanohaloarchaea and Haloarchaea share a common ancestor. The recent study by Wang et al. (2019) includes a phylogeny derived from the entire ATPase operon in Archaea, that also recovered the sisterhood between the Nanohaloarchaea and Haloarchaea. Wang et al. consider horizontal transfer of the operon as explanation for this grouping, and also observe an identical operon structure in both groups, which supports the monophyly of nanohaloarchaeal and haloarchaeal ATPases. Those authors recognized clear conflicts between a DPANN supergroup and the ATPases phylogeny, and reconciled this conflict by invoking ATPase horizontal gene transfer. In correlations of Nanohaloarchaea gene families, it was revealed the ATPase genes are more closely clustered to a large set of genes; in contrast Thaumarchaeota ATPase genes form their own evolutionary cluster distant from other genes in the genome. These results can be reconciled by considering that ATPase genes were indeed transferred into Thaumarchaeota, but not in the case of Nanohaloarchaea. The gene family correlation method would be highly sensitive to single and multiple recent gene transfers, as the distance matrices analyzed for each gene family are vectorized (i.e., taxon-specific information is saved and kept consistent throughout the entire clustering process, so if a single gene is transferred into only a single taxon, it will be recorded and thus affect the clustering of the entire gene family). Although the clustering based on gene distance correlations does well in recovering ATPase transfer in Thaumarchaeota, the absence of detected transfer events in Nanohaloarchaea cannot be considered proof that no transfer has happened. A suspected transfer would have occurred greater than ∼1 Ba before the deepest splits within the Nanohaloarchaea and Haloarchaea, respectively. Within-niche transfer of ATPase operons is certainly possible and is supported in the case of the Thaumarchaeota ( Wang et al. 2019 , supplementary fig. S5 , Supplementary Material online ) and Deinococcaceae ( Lapierre et al. 2006 ); however, we are not aware of any evidence that extends this logic to the Nanohaloarchaea. We provide evidence that ATPases do not stand out as atypical in their evolutionary history as compared with other genes found in Nanohaloarchaea ( figs. 3 and 4 and supplementary figs. S4 and S5 , Supplementary Material online ). Our interpretation of these data is that the sisterhood relationship of the Nanohaloarchaea and Haloarchaea ATPases should not be immediately discarded as resulting from HGT. The analyses presented here and by Raymann et al. (2014 ) and Aouad et al. ( 2018 , 2019 ) suggest that Nanohaloarchaeal genomes have been shaped by a complex evolutionary history. Many gene families support inclusion of the Nanohaloarchaea into DPANN, whereas the aforementioned studies suggest of a placement within Haloarchaea or other euryarchaeota. The totality of support for a Nanohaloarchaea–Haloarchaea sister group contained within our analyses include: ATPase phylogenies (and that these gene are unlikely to have been transferred [ figs. 2–4 and supplementary figs. S4 and S5 , Supplementary Material online ]); recoded nanohaloarchaeal Large core genome; and gCF analyses which identified several slowly evolving genes also supporting a Nanohaloarchaea-Haloarchaea sister group relation. Due to the observation of radically different phylogenetic signals present in the nanohaloarchaeal core, we consider an analogy between Nanohaloarchaea and Thermotoga. The Thermotoga core genome is extremely chimeric: its evolutionary history indicates genes comprising the “informational” functionality (i.e., genes involved in replication, repair, etc.) are bacterial in origin, whereas genes that contribute to metabolism are of Archaeal or Clostridial origin ( Logsdon and Faguy 1999 ; Zhaxybayeva et al. 2009 ). In comparison, genes that compose the Left cluster of genes in Nanohaloarchaea are enriched for proteins that encode for translational functions, whereas the Right cluster is enriched for proteins that serve transcription purposes ( supplementary table S5 , Supplementary Material online , K = transcription, J = translation), indicating that Nanohaloarchaea are highly chimeric too. Phylogenies calculated from concatenated data sets support the existence and monophyly of the DPANN superphylum (including the Nanohaloarchaea). When genomes from DPANN members were included, the Nanohaloarchaea were recovered as part of the DPANN group. In the absence of the other DPANN genomes, Nanohaloarchaea formed a clade with Haloarchaea ( supplementary fig. S8 , Supplementary Material online ), even after removing potential biases. However, the sister group moved out of Methanotecta, and possibly the euryarchaeota too. As to whether this sister group is located in the euryarchaeota depends on where one places the archaeal root. If one expects the root to be inside the euryarchaeota ( Raymann et al. 2015 ), this sister group has a possibility of falling outside the euryarchaeota, as it falls outside Methanotecta and may lead to a branch where the DPANN superphylum could attach. The observation that a monophyletic Nanohaolarchaeal–Haloarchaeal grouping is recovered from the Large core concatenation but at the base or even outside the euryarchaeota illustrates observed evolutionary relationships between archaeal classes obtained from gene family concatenations have to be interpreted with caution. Phylogenetic reconstruction that constrained Nanohaloarchaea to group with Haloarchaea resulted in a maximum-likelihood phylogeny that the AU test (Shimodaira 2002) evaluated as incompatible with the best tree for this Large Core genome data set, revealing a strong phylogenetic signal, either due to shared ancestry or systematic artifact, that does contradict the sister group relationship between Nanohaloarchaea and Haloarchaea. Radically different placements of Nanohaloarchaea ( fig. 1 , red indicators) can be at least partially attributed to the taxonomic sampling of the DPANN superphylum. In instances where the Nanohaloarchaea were recovered inside the euryarchaeota ( Narasingarao et al. 2012 ; Zhaxybayeva et al. 2013 ; Aouad et al. 2018 , 2019 ), DPANN sequences were not included in the tree. However, including a robust sampling of DPANN sequences in the alignment ( Andrade et al. 2015 ; Sorokin et al. 2017 ; Wang et al. 2019 ; fig. 5 ) generally attracts the Nanohaloarchaea into that superphylum. The gCF analysis revealed 16 core genes in support of Nanohaloarchaea–Haloarchaea sister group; however, 15 genes support Nanohaloarchaea inclusion in DPANN. In support of Nanohalo + Haloarchaea group are 16 genes that evolve significantly slower than those in support of the opposing hypothesis ( supplementary table S6 , Supplementary Material online ). Previous analyses have indicated high bootstrap support for including the Nanohaloarchaea within DPANN ( Sorokin et al. 2017 ; Wang et al. 2019 ). This support may reflect the strong but artifactual signal in fast evolving genes, phylogenetic signals created through gene transfers, and forcing all genes with different histories onto the same tree—conflicting signals are likely abundant of in all concatenated marker sets. Our gCF analysis dissected the concatenation based on individual gene trees, revealing opposing phylogenetic signals present in the original concatenated data set. It is important to supplement the sampling variance measure for the singular branch (i.e., bootstrap), with a measure of variance in the overall data set with metrics like the concordance factors. The concordance factors can reveal variance (conflict) within the multilocus alignment data sets. In an attempt to dissect the Large Core genome concatenation even further, we subdivided it into three subdivisions; the Left, Right, and Center supermatrices ( fig. 4 ). These subdivisions are based on pairwise distance matrices of each individual gene family. Phylogenies of the Right supermatrix reveals that a large ensemble of genes (94) that are a part of the Nanohaloarchaea core genome may have been transferred from the Nanohaloarchaea to Haloarchaea, as Haloarchaea moved from euryarchaeota into DPANN. Concatenations involving these genes may calculate a phylogeny with high artifact potential due to their possible transfer or unconventional evolutionary trajectory. It is worth noting again, that the AtpA+B genes fall into the Left supermatrix, even though their individual signal robustly groups the Nanohaloarchaea and Haloarchaea as sister groups. This demonstrates that genes evolving through a similar evolutionary trajectory (Left cluster), can recover evolutionary placements and may be convoluted by concatenating gene families with disparate rates of evolution. Monophyly of Extreme Halophilic Archaea The Methanonatronarchaeia did not reveal a well-supported association with any particular Archaeal group in any of these phylogenies. In the ATP synthase-based phylogenies, homologs from three members of this group were recovered as a deeper branching euryarchaeal lineage without well-supported affinity to any other euryarchaeal group. Sequences from the Methanonatronarchaeia were, however, separated by at least one well-supported bipartition from other halophilic archaea grouping with nonhalophilic methanogens ( fig. 2 and supplementary fig. S3 , Supplementary Material online ). A concatenation of Nanohaloarchaeal core genes reliably placed Methanonatronarchaeia ( fig. 5 b and c ) basal to the Methanotecta super-class, as proposed by Aouad et al. 2019 . When using the entire Large Core genome supermatrix ( fig. 5 a ), Methanonatronarchaeia appeared as a sister group to Haloarchaea (BV = 88). Aouad et al. provided evidence for three independent adaptations to high salt environments (through the salt-in strategy) in Haloarchaea, Nanohaloarchaea, and Methanonatronarchaeia (Aouad et al. 2018 , 2019 ). Although we consider convergent evolution events rare, independent adaptations to hypersalinity resulting from salt-in strategy pressures and revealed through shifts in protein isoelectric points ( Oren 2008 ) have been observed in Salinibacter (Bacteroidetes) and Salinicoccus (Firmicutes) (see supplementary fig. S13 , Supplementary Material online ), with minimal reliance on HGT from haloarchaea ( Mongodin et al. 2005 ). Methanonatronarchaeia have been deduced to employ a salt-in strategy, using intracellular potassium ion concentrations ( Sorokin et al. 2017 ), the same adaptation present in Nanohaloarchaea and Haloarchaea. However, a proteomic analysis of theoretical isoelectric point (pI) distributions reveals a less biased distribution of pIs in these methanogens compared with other proteomes of organisms that use a salt-in strategy (Haloarchaea, Nanohaloarchaea, etc.) ( supplementary fig. S13 , Supplementary Material online ). This distribution of theoretical pIs in Methanonatronarchaeia resembles that found in marine archaea ( supplementary fig. S13 , Supplementary Material online ), and Halanaerobiales . The Halanaerobiales follow an experimentally confirmed salt-in strategy without an acidic proteome. Instead, they hydrolyze Glutamine (Q) and Asparagine (N) to compensate for the lack of acidic amino acids ( Bardavid and Oren 2012 ). However, the genome of Acethalobium arabaticum , a member of the Halanaerobiales , encodes a more acidic proteome, similar to Salinibacter and Salinicoccus ( supplementary fig. S13 , Supplementary Material online ). Methanonatronarchaeia may be a similar example of independent adaptation to hypersalinity. In some Methanonatronarchaeia, the concentration of intracellular potassium did not yet have a significant impact on the distribution of pIs of encoded proteins or possibly, they may also hydrolyze their N/Q residues to make their acidic conjugates, like Halanaerobiales . The AtpAB data set robustly recovers the Nanohaloarchaea–Haloarchaea sister group. Furthermore, we provide evidence that these genes are slow evolving ( supplementary table S5 , Supplementary Material online ), and unlikely to have been transferred recently between the groups ( figs. 3–5 and supplementary figs. S4–S5 , Supplementary Material online ). An obvious caveat is that the better-resolved single-gene phylogeny represents only a single gene or operon, and that its phylogeny is embedded in the net-like, reticulated genome phylogeny. Data from concatenated data sets robustly recovers the Nanohaloarchaea group within DPANN (the exception being recoded Large core phylogeny, supplementary fig. S7 a , Supplementary Material online ). However, these data sets are rife with conflict (transferred genes, genes with differing rates of evolution; gCF and fig. 4 ) and forcing them on a single tree likely is inappropriate. We consider phylogenetic placement of the Nanohaloarchaea an open question. A plethora of analyses using large concatenates support inclusion of Nanohaloarchaea in DPANN ( Rinke et al. 2013 ; Andrade et al. 2015 ; Castelle et al. 2015 ; Dombrowski et al. 2020 ), but the same can be said for the opposite ( Brochier-Armanet et al. 2011 ; Raymann et al. 2014 ; Petitjean et al. 2015 ; Williams et al. 2015 ; Aouad et al. 2018 , 2019 ). Conflict between these analyses ( fig. 1 ) may, at least in part, be due to reliance on large concatenates and forcing disparate evolutionary signals onto the same tree. Dissecting these evolutionary signals and evaluating their suitability for such analyses could be a way forward for resolving the debate of the nanohaloarchaeal placement and the existence of DPANN. Recently, Hikarchaeia ( Martijn et al. 2020 ) were found to be more closely related to Haloarchaea (than Nanohaloarchaea) in the ATPase phylogenies. A hallmark of a salt-in strategist can be found in the Hikarchaeia’s proteomes, as they decidedly favor acidic residues found in other salt-in strategists ( supplementary fig. S13 m and n , Supplementary Material online ; Oren 2008 ; Paul et al. 2008 ). Results of ATPase phylogenetic analysis ( fig. 2 and supplementary fig. S3 , Supplementary Material online ) also raises the possibility that both the Hikarchaeia and Haloarchaea evolved from an extreme halophilic ancestor, and that Hikarchaeia lost this adaptation after their divergence. If the ATPase phylogeny topology results from HGT or gene sharing, acquisition of the Haloarchaeal ATPase by Nanohaloarchaea predates the split between Hikarchaeia and Haloarchaea, again suggesting that extreme halophily might have been an ancestral character of the Hikarchaeia. The alternative assumption that the Hikarchaeia and Haloarchaea ancestor was not an extreme halophile would imply that transfer of the ATPase operon occurred before Haloarchaea and Nanohaloarchaea convergently adapted to hypersalinity. The inclusion of the Hikarchaeia in future phylogenetic analyses (once there is a larger sampling of this lineage) may further elucidate the genomic events that lead to hypersaline adaptation. Although our analyses do not prove that Nanohaloarchaea are not part of a DPANN grouping, our findings indicate that when they are strongly supported in concatenated data sets this might be the result of an artifact, and that the phylogenies of conserved slowly evolving genes (ATPases, ribosomal proteins, and an elongation initiation factor) may better reflect the origin of the Nanohaloarchaea. In most gene families, the phylogenetic signal regarding relationships between different archaeal classes is weak, and single gene phylogenies are poorly resolved. The popular solution of data set concatenation ( Lapierre et al. 2014 ) to amplify a weak phylogenetic signal comes with the possibility that systematic artifacts and not a combined phylogenetic signal dominate the resulting phylogenies ( Bapteste et al. 2008 ). Resolving deep divergences remains a hard problem. Due to horizontal gene transfer and phylogenetic reconstruction artifacts, the placement of divergent archaeal classes into larger groups remains uncertain."
} | 7,845 |
26457129 | PMC4600277 | pmc | 5,892 | {
"abstract": "“Geoglobus ahangari” strain 234 T is an obligate Fe(III)-reducing member of the Archaeoglobales , within the archaeal phylum Euryarchaeota , isolated from the Guaymas Basin hydrothermal system. It grows optimally at 88 °C by coupling the reduction of Fe(III) oxides to the oxidation of a wide range of compounds, including long-chain fatty acids, and also grows autotrophically with hydrogen and Fe(III). It is the first archaeon reported to use a direct contact mechanism for Fe(III) oxide reduction, relying on a single archaellum for locomotion, numerous curled extracellular appendages for attachment, and outer-surface heme-containing proteins for electron transfer to the insoluble Fe(III) oxides. Here we describe the annotation of the genome of “G. ahangari” strain 234 T and identify components critical to its versatility in electron donor utilization and obligate Fe(III) respiratory metabolism at high temperatures. The genome comprises a single, circular chromosome of 1,770,093 base pairs containing 2034 protein-coding genes and 52 RNA genes. In addition, emended descriptions of the genus “ Geoglobus” and species “G. ahangari” are described. Electronic supplementary material The online version of this article (doi:10.1186/s40793-015-0035-8) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions “G. ahangari” strain 234 T is only one of three members of the Archaeoglobales capable of dissimilatory Fe(III) respiration. Furthermore, it is an obligate Fe(III) reducer that grows better with insoluble than soluble Fe(III) species. Consistent with this, the genome contains a large number of c -type cytochromes within and on the cell surface, as well as other redox-active proteins such as thermostable ferredoxin and Fe-S proteins. The paucity of c -type cytochromes within non-Fe(III) respiring members of the Archaeoglobales ( Archaeoglobus species) is consistent with the physiological separation between these archaea and F. placidus , G. acetivorans , and “G. ahangari” , which can gain energy for growth from the reduction of Fe(III) electron acceptors. Additionally, some genes required for both dissimilatory sulfate and nitrate metabolisms are absent in “G. ahangari” and G. acetivorans . This supports the physiological separation of Geoglobus spp. from F. placidus , which is capable of Fe(III)-, thiosulfate-, and nitrate respiration, and from Archaeoglobus species which are primarily sulfur-respiring organisms. Genomic data also support the reported physiological similarities between “G. ahangari” and other Archaeoglobales such as autotrophic growth with H 2 via the reductive acetyl-CoA/Wood-Ljungdahl pathway and the use of similar electron donors, including short- and long-chain fatty acids. Noteworthy is the fact that genomic evidence supports the synthesis of the methanogenic coenzyme-F 420 in “G. ahangari” , which is responsible for the characteristic fluorescence detected in all Archaeoglobus spp. except for “G. ahangari” or F. placidus . Hence, the genome sequence of “G. ahangari” provides valuable insights into its physiology and ecology as well as into the evolution of respiration within the Archaeoglobales . Taxonomic note The initial publication [ 1 ] of the “ Geoglobus ” genus and “ Geoglobus ahangari ” species was accepted for publication with extenuating circumstances at several culture-collection agencies. Thus, upon the original publication “ G. ahangari ” strain 234 T was accepted only at a single agency. In addition, the G + C mol% determined from the complete genome sequence (53.1 mol%) differs from that originally published (58.7 mol%), representing a discrepancy of over 5 mol%. This publication thus warrants an emended description of the genus Geoglobus and the type species, “ Geoglobus ahangari ” . Emended description of “ Geoglobus ” Kashefi et al. The description of the genus “ Geoglobus ” is the one provided by Kashefi et al. [ 1 ], with the following modifications. In addition to the single monopolar flagellum, numerous curled filaments can be seen per cell [ 14 ]. The G + C content of the genomic DNA of the type species is 53.1 mol%. Emended description of “ Geoglobus ahangari ” Kashefi et al. The description of the species “ Geoglobus ahangari ” is the one provided by Kashefi et al. [ 1 , 2 ], with the following modifications. The type strain is strain 234 T and has been deposited at three culture collection agencies, which include the Deutsche Sammlung von Mikroorganismen und Zellkulturen ( DSM-27542 ), the Japan Collection of Microorganisms ( JCM 12378 ), and the American Type Culture Collection ( BAA-425 ).",
"introduction": "Introduction “Geoglobus ahangari” strain 234 T is the type strain and one of only two known members of the Geoglobus genus within the order Archaeoglobales and the family Archaeoglobaceae . It is an obligate Fe(III)-reducing archaeon isolated from the Guaymas Basin hydrothermal system and grows at temperatures ranging from 65–90 °C, with an optimum at about 88 °C [ 1 ]. It was the first isolate in a novel genus within the Archaeoglobales and the first example of a dissimilatory Fe(III)-reducer able to grow autotrophically with H 2 [ 1 ], a metabolic trait later shown to be conserved in many hyperthermophilic Fe(III) reducers [ 2 ]. “G. ahangari” can also couple the reduction of soluble and insoluble Fe(III) acceptors to the oxidation of a wide range of carbon compounds including long-chain fatty acids such as stearate and palmitate, which were previously not known to be used as electron donors by archaea [ 1 ]. It was also the first hyperthermophile reported to fully oxidize acetate to CO 2 , a metabolic function once thought to occur solely in mesophilic environments [ 3 ]. Unlike the other two genera in the order Archaeoglobales ( Archaeoglobus and Ferroglobus ), which can utilize acceptors such as sulfate and nitrate [ 1 , 4 – 10 ], the two cultured members of the genus Geoglobus can only use Fe(III) as an electron acceptor [ 1 , 4 ]. The obligate nature of Fe(III) respiration in Geoglobus spp. makes the genus an attractive model to gain insights into the evolutionary mechanisms that may have led to the loss and/or gain of genes involved in the respiration of iron and other electron acceptors such as sulfur- and nitrogen-containing compounds within the Archaeoglobales . “G. ahangari” strain 234 T also serves as a model organism for mechanistic studies of iron reduction at high (>85 °C) temperatures. Dissimilatory Fe(III) reduction has been extensively studied in mesophilic bacteria (reviewed in references [ 11 , 12 ]). By contrast, little is known about the mechanisms that allow (hyper)thermophilic organisms to respire Fe(III) acceptors [ 13 – 18 ]. As previously observed in the thermophilic Gram-positive bacterium Carboxydothermus ferrireducens [ 13 ], “G. ahangari” also needs to directly contact the insoluble Fe(III) oxides to transfer respiratory electrons [ 14 ]. In “G. ahangari” , cells are motile via a single archaellum, which could help in locating the oxides, and also express numerous curled extracellular appendages, which bind the mineral particles and position them close to heme-containing proteins on the outer surface of the cell to facilitate electron transfer [ 14 ]. A direct contact mechanism such as this is predicted to confer on these organisms a competitive advantage over other organisms relying on soluble mediators such as metal chelators [ 19 ] and electron shuttles [ 20 , 21 ], which are energetically expensive to synthesize and are easily diluted or lost in the environment once excreted [ 22 ]. This is particularly important in hydrothermal vent systems such as the Guaymas basin chimney where “G. ahangari” strain 234 T was isolated, as vent fluids in these systems can flow through at rates as high as 2 m/s [ 23 ]. Here, we report the complete genome sequence of “G. ahangari” strain 234 T and summarize the physiological features that make this organism a good model system to study Fe(III) reduction in hot environments and to gain insights into the evolution of Fe(III) respiration in the family Archaeoglobales ."
} | 2,060 |
19591463 | null | s2 | 5,894 | {
"abstract": "A new hybrid hydrogel based on poly[N-(2-hydroxypropyl)methacrylamide] grafted with a beta-sheet peptide, Beta11, was designed. Circular dichroism spectroscopy indicated that the folding ability of beta-sheet peptide was retained in the hybrid system, whereas the sensitivity of the peptide toward temperature and pH variations was hindered. The polymer backbone also prevented the twisting of the fibrils that resulted from the antiparallel arrangement of the beta-strands, as proved by Fourier transform infrared spectroscopy. Thioflavin T binding experiments and transmission electron microscopy showed fibril formation with minimal lateral aggregation. As a consequence, the graft copolymer self-assembled into a hydrogel in aqueous environment. This process was mediated by association of beta-sheet domains. Scanning electron microscopy revealed a particular morphology of the network characterized by long-range order and uniformly aligned lamellae. Microrheology results confirmed that concentration-dependent gelation occurred."
} | 259 |
18650539 | null | s2 | 5,895 | {
"abstract": "The byssus of marine mussels has attracted attention as a paradigm of strong and versatile underwater adhesion. As the first of the 3,4-dihydroxyphenylalanine (Dopa)-containing byssal precursors to be purified, Mytilus edulis foot protein-1 (mefp-1) has been much investigated with respect to its molecular structure, physical properties, and adsorption to surfaces. Although mefp-1 undoubtedly contributes to the durability of byssus, it is not directly involved in adhesion. Rather, it provides a robust coating that is 4-5 times stiffer and harder than the byssal collagens that it covers. Protective coatings for compliant tissues and materials are highly appealing to technology, notwithstanding the conventional wisdom that coating extensibility can be increased only at the expense of hardness and stiffness. The byssal cuticle is the only known coating in which high compliance and hardness co-exist without mutual detriment; thus, the role of mefp-1 in accommodating both parameters deserves further study."
} | 253 |
35869541 | PMC9306079 | pmc | 5,896 | {
"abstract": "Obtaining efficient autotrophic ammonia removal ( aka partial nitritation-anammox, or PNA) requires a balanced microbiome with abundant aerobic and anaerobic ammonia oxidizing bacteria and scarce nitrite oxidizing bacteria. Here, we analyzed the microbiome of an efficient PNA process that was obtained by sequential feeding and periodic aeration. The genomes of the dominant community members were inferred from metagenomes obtained over a 6 month period. Three Brocadia spp. genomes and three Nitrosomonas spp. genomes dominated the autotrophic community; no NOB genomes were retrieved. Two of the Brocadia spp. genomes lacked the genomic potential for nitrite reduction. A diverse set of heterotrophic genomes was retrieved, each with genomic potential for only a fraction of the denitrification pathway. A mutual dependency in amino acid and vitamin synthesis was noted between autotrophic and heterotrophic community members. Our analysis suggests a highly-reticulated nitrogen cycle in the examined PNA microbiome with nitric oxide exchange between the heterotrophs and the anammox guild. Supplementary Information The online version contains supplementary material available at 10.1186/s40793-022-00432-2.",
"introduction": "Introduction Predicting and managing the composition and function of microbial communities is heralded as the holy grail of microbial ecology [ 1 ]. Composition can be imposed when working under aseptic conditions, but is challenging when communities are open, the typical scenario for most ecology-relevant and technology-interesting microbial communities. While rational engineering of microbiomes is an active field of research, it is recognized that we still lack the required mechanistic understandings to design microbiomes based on a priori knowledge [ 2 ]. On the other hand, heuristic approaches based on manipulation of environmental conditions, have a long history of success in managing microbial communities towards specific functions in food, agriculture, and environmental applications. While those management practices do not rely on first principle understanding of microbial community assembly, unravelling those communities, and seeking links with management constraints might yield insights on which to build new testable theories [ 2 ]. The common, and intuitively simple, conditions that one can impose on a microbial community are the provision/fluxes of specific electron donors versus electron acceptors and of macro-or micronutrients; provision or limitation of these will impose selective pressures and enrich communities with the desired phenotypes. In addition, community control may be further facilitated by spatio-temporal gradients that naturally occur or are artificially imposed. Gradients are especially valuable when they permit the establishment of multiple redox conditions over short spatial or temporal scales that allow the co-occurrence of functional groups that require mutually exclusive environmental conditions; the characteristic of microbial aggregates and biofilms [ 3 , 4 ]. Here, we examine the management of a microbial community with the functional property of complete autotrophic conversion of ammonia (NH 3 ) to dinitrogen gas (N 2 ), also known as the partial nitritation/anammox process (PNA) [ 5 ]. The simplest functional PNA community would consist of only two functional groups: one that performs aerobic NH 3 oxidation (using O 2 as terminal electron acceptor (TEA) and forming NO 2 − , aerobic ammonia oxidizing prokaryotes or AOBs) and one that performs anoxic NH 3 oxidation (using NO 2 − as TEA and forming N 2 , anaerobic ammonia oxidizing prokaryotes or AnAOBs). The PNA process relies on the provision of controlled (limited) supply of oxygen and the spatial or temporal variation in redox conditions [ 6 , 7 ]. Yet, the PNA process remains tested by microbial community management, especially the suppression of the aerobic NO 2 − oxidation guild (aerobic nitrite oxidizing prokaryotes or NOBs) [ 7 , 8 ]. Indeed, all metagenetic and metagenomic analyses of PNA communities to date indicate the persistent presence of NOBs; and their excessive presence would deteriorate or collapse the PNA process [ 8 , 9 ]. Selection for AOB against NOBs has variously relied on growth inhibition by free ammonia or nitrous acid [ 10 ] or oxygen limitation driven by the low O 2 affinity of the NOB [ 9 , 11 – 13 ]. In addition, while autotrophic NH 3 and NO 2 − oxidizers are the only microbes that can grow on the influent devoid of organic carbon and are essential for a functional PNA process, their metabolism and decay will result in release of organic byproducts. As a result, heterotrophic microbes are inevitable and their abundance and potential symbiosis with autotrophs has previously been found [ 14 , 15 ], yet their functional contribution is ill-documented. We document here that periodic and limiting provision of oxygen (via intermittent aeration) to a granule-based reactor can result in complete elimination of NOBs from the community performing the PNA process. We examine the consequences of this operation on the community structure, with specific attention on alternate pathways for NO x (nitrogen oxides) metabolism. We identified a representative set of high-quality MAGs representing most of the metagenome, a few highly abundant AOB and AnAOB, the absence of NOB, and a diversity in NOx respiratory abilities across the heterotrophic MAGs. Genome analysis suggests a highly reticulated network with possibility for NO exchange between autotrophs and heterotrophs and strong evidence of auxotrophies distributed across the community members.",
"discussion": "Discussion Performance and NOB suppression PNA has been successfully attained using various reactor configurations/operations and resulting biomass morphologies [ 5 ]. In general, attached growth (biofilms, aggregates) seems necessary to obtain sufficient AnAOB in the system, yet the same systems are challenged by the retention of unwanted NOB. NOB control is easier to obtain in suspended growth (or hybrid suspended/attached growth) systems [ 11 ]. Here we document that granular systems—obtained by sequential feeding and subjected to intermittent aeration—permit both retention of AOB and AnAOB (regulated by diffusional transport of O 2 , NH 4 + and NO 2 − [ 24 ])) but also control NOB density. While causes for repression of NOB by aeration switching have been proposed [ 12 , 25 – 27 ], direct proof has remained elusive and this operational control towards microbiome engineering remains heuristic. Presence and diversity of AOB and AnAOB In the resulting PNA community, operated on a synthetic feed with NH 4 + as the sole energy source, AOB and AnAOB amounted to 13% (stdev: 5%) and 22% (stdev: 6%) of the community (as fractions of the metagenome). This is similar to the fractions observed in other PNA communities by Speth et al. (AOB max 4% AnAOB max 20%) and Wang et al. (AOB ca. 25% AnAOB ca. 40%) [ 28 ]. While Speth et al. [ 29 ] and Wang et al. [ 28 ] identified only one MAG as AOB and AnAOB, we identified 3 MAGs each that could be classified as AOB and AnAOB. This may, in part, be due to the higher fraction of the MG that could be assigned to the different MAGs (79% here vs only 59% in [ 29 ]); even though the distribution across MAGs was more equitable in our study compared to Wang et al. [ 28 ]. The dominant AnAOB and AOB in the current system were closely related to Candidatus Brocadia sp., and Nitrosomonas europaea and N. eutropha as typically found in these highly loaded synthetic PNA or (for AnAOB) anammox communities [ 14 , 19 , 28 , 29 ]. Absence of NOB Although 16S rRNA gene targeted qPCR ([ 16 , 30 ]) and 16S rRNA gene amplicon targeted community analysis (Additional file 1 : Fig. S1) indicated a small Nitrospira presence (< 0.5%), no MAGs encoding autotrophic nitrite oxidation were recovered, and the presence in the whole metagenome was also minimal (ca. 4 RPM mapped to canonical NOB nxr ). Speth et al. [ 29 ], on the other hand, detected a Nitrospira MAG (at 1.6 to 2.8%); yet also Wang et al. did not detect NOB MAGs in their PNA community MG [ 28 ]. As both our study and [ 28 ] applied sequential (instead of continuous low-rate) aeration to support the PNA community, this may be an effective strategy for NOB counter selection. We also note that, in fact, Nitrospira detection based on 16S rRNA gene or on nxrA , as done here, is not necessarily indicative of the presence of strict NOB since comammox Nitrospira have also been recovered from PNA systems [ 31 ]. Therefore the low fraction of Nitrospira we detect might be an overestimate of NOB abundance. The types of HB and the role of auxotrophy vs. prototrophy of HB Heterotrophic bacteria were abundant in this study (57% of MG vs autotrophs 43%); consistent with other studies [ 14 , 15 , 29 ]. They were distributed across a diverse set of phylotypes but with notable abundance in Chloroflexi (MAG CFX 1-15 at 23%), Ignavibacteriales/Chlorobi (MAG CLB 1-4 at 5%), Armatimonadates (MAG ARM1 at 4%), Bacteroidetes/Flavobacteria (MAG BCD 1,2 at 4%), and a few Proteobacteria (MAG PRO 2 and 6 at 3 and 2%). These taxa are all typically found in PNA and/or anammox communities. Heterotrophs are assumed to be supported by soluble microbial products actively or passively released by autotrophic PNA members [ 32 ]. In addition, others have suggested heterotrophs as essential in providing growth factors to autotrophs [ 14 , 33 ], a claim not consistent with our findings. Auxotrophy for AA and vitamin biosynthesis were present in both autotrophs as heterotrophs; the most auxotrophic MAGs were heterotrophs, and AnAOB MAGs were the only MAGs encoding potential for cobalamin biosynthesis (Fig. 4 ). Clearly mutual dependencies beyond exchange of N species drive the composition of the PNA microbiome [ 15 , 34 ]. Denitrification pathways The current metagenome analysis indicates that heterotrophic MAGs have varying abilities for NOx respiration, with only one MAG encoding a complete denitrification pathway. Similar observations were made before: rare MAGs encoding full denitrification, but wide potential (and expression) for NO 3 − to NO 2 − respiration across multiple MAGs [ 15 , 19 , 29 ]. These observations have supported the notion that heterotrophs in PNA systems support a nitrite loop [ 35 ]. Our analysis results suggest, in addition, an abundance of MAGs with NO as the predicted end product of NO x respiration. In combination with the fact that the two AnAOB MAGs AMX2 and AMX3 (Additional file 3 : Table S2) lack NIR encoding genes, this raises the possibility for NO cycling between autotrophic and heterotrophic MAGs. While the ability to support anaerobic ammonium oxidation supported by NO reduction (instead of NO 2 − ) has been shown in pure culture [ 23 ], direct proof in a PNA microbiome awaits confirmation. In conclusion, our metagenomics analysis indicates that intermittent aeration is a highly-efficient control strategy to suppress NOB presence in a PNA process. The resulting microbiome presents mutual dependencies between the AOB and AnAOB autotrophs and heterotrophs, and a N cycle network that involves NO exchange."
} | 2,815 |
36276034 | PMC9511690 | pmc | 5,898 | {
"abstract": "Brain-inspired neuromorphic computing has become one of the critical technologies to overcome the bottleneck of von Neumann architecture. It is a vital step to construct a brain-like neuromorphic computing system at the hardware level by utilizing artificial synaptic devices. Compared with electronic synaptic devices, optoelectronic synaptic devices have the advantages of low power consumption, low crosstalk, and high bandwidth. Artificial optoelectronic synapses, analogous to retinal structure, can directly respond to and process light signal information to mimic the neuromorphic visual system. As high-level nerve impulses, both generated and regulated, emotions affect the strength and persistence of memory. Ambient illumination can provide visual perception to distinguish the size, color, and other characteristics of objects as well as affect the nonvisual functions of individuals, such as emotional states, thereby affecting learning and memory function. Herein, an artificial optoelectronic synapse composed of ITO/TiO 2− x / p -Si was proposed. A variety of biologically dependent synaptic plasticity relating to learning and memory, including short-term synaptic plasticity, long-term synaptic plasticity, and learning-forgetting-relearning multifunctional advanced synaptic activity, was successfully simulated. A 3 × 3 artificial optoelectronic synapse array based on 9 devices was constructed to mimic the functions of visual learning and memory affected by internal emotion and ambient illumination. The proposed artificial optoelectronic synapse will exhibit great potential in visual and image information perception and memory.",
"conclusion": "4. Conclusions In conclusion, a single optoelectronic synaptic based on ITO/TIO 2− x / p -Si device was successfully fabricated and demonstrated. The light-induced short-term synaptic plasticity of EPSC and PPF behaviors were simulated. The long-term synaptic plasticity transferred from short-term synaptic plasticity was demonstrated by adjusting the intensity, duration, and number of optical signal pulses. Based on excellent light perception and memory behavior, the high-level neural activity of learning-forgetting-relearning was realized. A 3 × 3 optoelectronic synapse array was constructed to reflect the effects of emotion and ambient illumination on visual learning and memory. It is a positive relationship between light memory and emotion or ambient illumination. The proposed artificial optoelectronic synapse offers a bright prospect for the realization of multifunctional electronic eye and artificial vision systems.",
"introduction": "1. Introduction With the rapid development of artificial intelligence (AI) technology, the breakthrough of advanced intelligent electronic devices is urgently needed in the fields of bionic perception systems, human–computer interaction and intelligent robots. 1,2 The neural network based on the traditional von Neumann architecture has the separation bottleneck of the memory and processing units, which leads to insufficient bandwidth and enormous energy. 3,4 The computing method in the human brain has high parallelism, high efficiency, and low power consumption. 5 It is an important step in the hardware construction of brain-like computing systems and artificial intelligence systems to simulate human brain neuromorphic systems by electronic devices. 6,7 Human beings can perceive and respond to external stimuli (like light, 8 pressure, 9 and chemicals 10 ). A variety of biological synaptic behaviors have been simulated. In particular, 80% of the detected information is perceived through the visual system. 11,12 Perceptual neurons in the retina and the visual system can perceive and preprocess complex light signals. 13 Therefore, the artificial neuromorphic vision system established by simulating the biological structure and function of the human retina can realize a more efficient artificial vision system, which has application prospects in the fields of intelligent robots, human–computer interaction, and military target detection. 14 Nowadays, unstructured tasks including image recognition 15 and adaptive detection 16 have been implemented on optoelectronic neuromorphic devices. Taking advantage of the advantages of metal oxides such as visible light transparency, 17 high electron mobility, 18 and intrinsic stability, 19 we fabricated titanium dioxide-based two-terminal optoelectronic synaptic devices that are highly compatible with high-density cross-array structures to simulate visual perception and visual memory functions. The factors that affect biological visual perception and memory are diverse and complex. Here, we discuss two influencing factors, intrinsic and extrinsic. Emotion as an intrinsic factor is a characteristic mental activity of higher-grade animals, which belongs to the nerve impulses projected by nerve nuclei such as cingulate gyrus and amygdala to the various higher cortex. 20,21 The main function of the amygdala is to generate emotions corresponding to various external information transmitted to the neocortex of the brain. Information about emotions can affect the activity state of the higher cortex to complete multi-component intricate functions such as perception, complex movement, learning, and memory. 22 Memory relies heavily on emotions and is consistent with internal emotions. Events that trigger strong emotional responses leave long-term memory. 23 Furthermore, ambient illumination as an intrinsic factor produces visual effects by stimulating the rod photoreceptor cells and cone photoreceptor cells in the retina, which provides a visual perception guarantee for distinguishing the size, color, and spatial orientation of objects, and simultaneously participates in learning, memory, and emotion regulation through various ways. 24,25 Most studies have shown that the physiological characteristics and neuronal activity of the nucleus reuniens (Re) of the midline thalamus can be affected by bright light and contribute to the regulation of neuronal activity and spatial memory in the hippocampus (HPC), thereby affecting synaptic plasticity and playing a prominent role in learning and memory processing. 26 For instance, phototherapy can significantly improve neurodegeneration in Alzheimer's disease, which enhances synaptic function. 27 Interestingly, ambient illumination as well as has direct or indirect effects on the physical and mental states of emotional cognition and physiological rhythm through non-visual forming system due to the presence of intrinsic light-sensitive ganglion cells (ipRGCs) in the retina. 28,29 Emotion changes further have an impact on memory and learning behavior. Phototherapy is also used to treat many mental health problems, including seasonal affective disorder and chronic depressive disorder. 30 That is, the internal emotional state and ambient illumination interact to regulate and affect the individual visual perception, memory, and behavior. Here, the effects of emotion and ambient illumination on visual perception and memory are discussed for the first time based on optoelectronic synaptic devices. In this paper, light-stimulated synaptic devices (optoelectronic synapses) based on ITO/TiO 2− x / p -Si all-metal oxides were successfully fabricated. The adaptive optics information detection and excellent visual memory capability were exhibited on a single optoelectronic synaptic device combining the perception and synaptic function. Several critical biological synaptic behaviors were simulated by perceiving the different amplitude, duration, and quantity information of light signals, realizing the conversion of short-term synaptic plasticity (STP) to long-term synaptic plasticity (LTP) and multi-learning advanced synaptic behaviors of learning-forgetting and relearning. Most importantly, based on such visual memory properties, a 3 × 3 synapse array was developed to emulate internal emotions and ambient illumination effects, which are mapped into image “H”. This more realistically simulates the human visual perception and visual memory based on the artificial optoelectronic synaptic system, which illustrated the potential application prospects in the visual information processing and construction of advanced machine vision memory systems.",
"discussion": "3. Results and discussion Herein, we constructed a simple sandwich-structured photodetector for bio-synaptic performance. The TiO 2− x was chosen for photosensitive layer because the Ti element has a strong ability to absorb oxygen. 31,32 TiO 2− x is widely used in photosensitive devices due to its fluorite crystalline structure with abundant defect energy levels in the bandgap. 33,34 ITO was chosen as the top electrode because of its outstanding transparent conducting feature and rich surface defects. The band gap of TiO 2− x is estimated to be 3.56 eV (Fig. S1 † ), while ITO has a large work function (∼4.5 eV). A Schottky barrier naturally exists at the ITO/TiO 2− x interface. 35 The necessary charge-trapping interface is formed between the two conductive oxide layers. The detailed optoelectronic working mechanism is described in Fig. 1a . Under dark and unbiased conditions, most of the oxygen vacancies in the TiO 2− x layer are in the neutral state (VO 0 ), and it can be considered that the oxygen vacancies are uniformly distributed in the TiO 2− x layer. 36 When exposed to visible light and under positive V bias , some parts of the neutral oxygen vacancies are ionized (VO + or VO 2+ ), providing excess electrons to the bulk TiO 2− x layer. 37 When positive V bias is applied to the ITO electrode, the positively charged ionized oxygen vacancies will tend to migrate to the ITO/TiO 2− x interface and be trapped by the interface trap sites. The energy barrier near the interface will be lowered to increase the probability of electron tunneling, manifesting as an increase in photogenerated current. When the light is off, partially ionized oxygen vacancies still exist at the ITO/TiO 2− x interface due to the relatively high activation energy for neutralizing ionized oxygen vacancies, resulting in a slow decrease of the photogenerated current. 38,39 Fig. 1 (a) Schematic operation mechanism of TiO 2− x artificial optoelectronic synapse under dark and illumination conditions. (b) Cross-sectional and surface morphology image of TiO 2− x films grown on p -Si substrate. (c) Ti 2p core level and (d) O 1s core level XPS spectrum of TiO 2− x film. The surface topography information and cross-sectional of the TiO 2− x thin films were examined by scanning electron microscope (SEM) ( Fig. 1b ). It can be seen from the surface topography that the grain size is small and the grains will form clusters. The thickness of the as-prepared TiO 2− x film is ∼95 nm from the cross-sectional image, which is consistent with that measured in step-height image of AFM (Fig. S2a † ). The roughness of ∼0.7 nm shown in the three-dimensional morphology (Fig. S2b † ) illustrates the smoothness of the film surface. The XPS survey spectrum shown in Fig. S3 † reveals that the samples mainly contain Ti, O, and C element. Fig. 1c shows the XPS spectra of Ti 2p. The Ti 2p peaks at 464.22 eV and 458.42 eV can be attributed to Ti 2p 1/2 and Ti 2p 3/2, respectively, which is consistent with the previous reports. 40,41 The O1 peak can generally be decomposed into two sub-peaks at binding energy of 529.92 eV and 531.82 eV corresponding to lattice oxygen and non-lattice oxygen, respectively ( Fig. 1d ). 42,43 The lower binding energy peak is attributed to Ti–O bonds, while the higher binding energy peak is associated with oxygen vacancies (defects). These defects play an important role in fabricated optoelectronic synaptic devices. Moreover, the XRD analysis images (Fig. S4 † ) also illustrate that the sample films have many defect states. Humans acquire external information mainly through vision, so simulating the human visual system is crucial for the development of artificial intelligence. The synaptic transmission properties were mimicked on the proposed two-terminal optoelectronic synaptic devices more approximate to biological synapse structure. 44,45 The TiO 2− x optoelectronic synaptic devices were fabricated on a p -Si substrate, and top and bottom electrodes can be treated as the presynaptic and postsynaptic neurons, which are shown in Fig. 2a . Fig. 2b illustrated a typical light-induced excitatory postsynaptic current (EPSC) of the optoelectronic synaptic device under 1 V read voltage. The wavelength of the light pulse is 365 nm, and the intensity is 3.0 mW cm −2 . More specifically, the value of EPSC is increased to 1.75 nA with an obvious current amplitude of ∼0.6 nA when the light is turned on for 3 s. The gradual decay property in EPSC (the fitted curve shown in Fig. S5 † ) indicates that short-term plasticity (STP) plays a central role in the realization of visual learning and memory. 46 Paired-pulse facilitation (PPF) as a typical characteristic of STP reflects the capability of biological synapses for encoding temporal information in auditory and visual signals. 47 The optoelectronic synaptic device exhibits an obvious PPF phenomenon as stimulated by two successive 365 nm light pulses with an interval time (Δ t ) of 2 s, which is shown in Fig. 2c . The amplitude of the EPSC inspired by the second pulse ( A 2 ) is significantly larger than the first one ( A 1 ). Furthermore, the PPF index is defined as 100% × ( A 2 − A 1 )/ A 1 . Fig. 2d illustrates that the PPF index decreases with increasing pulse interval time, implying a gradual weakening of the effect between consecutive pulses. PPF index reaches a maximum of nearly 137% with an interval time of 2 s. The functional relationship of PPF index and interval time can be fitted with a biexponential function: 48 1 PPF = C 1 exp(− t / τ 1 ) + C 2 exp(− t / τ 2 ) where C 1 and C 2 as the initial facilitation magnitudes are estimated to be ∼37.0% and ∼114.4%, respectively. τ 1 and τ 2 are the characteristic relaxation times that correspond to the fast and slow decaying terms. The value of τ 1 and τ 2 stimulated to be ∼22.4 ms and ∼238.4 ms, respectively, which are equivalent to those in biological synapses. Fig. 2e shows that the light-induced EPSC amplitudes gradually increase with the increasing width and intensity of the 365 nm light pulse. Moreover, Fig. 2f exhibits that the optoelectronic synaptic devices have excellent current responses stimulated by light pulses under three different wavelengths, and the optoelectronic synaptic devices are relatively more sensitive under 365 nm illumination. Fig. 2 The basic short-term visual memory properties. (a) Schematic diagram of the prepared ITO/TiO 2− x / p -Si devices and synapse structure. (b) EPSC triggered by a single light pulse of 365 nm with the intensity of 3.0 mW cm −2 (c) EPSCs triggered by two continuous light pulses with an interval time of 2 s. (d) PPF index varies with pulse interval time. (e) Current as a function of light intensity and light pulse width. (f) Current variation with light intensity and wavelength. The important information in short-term memory (STM) is easily forgotten over time. While long-term memory (LTM) through continuous practice and stimulation is a state in which synaptic efficiency can be maintained for a long time, which is the basis for enhancing learning and memory ability. 49,50 Fig. 3a shows the gradual increase in the level and the retention time of memory as the light width of the light pulse increases. The value of Δ G (Δ G = G light − G resting , G light corresponds to the maximum conductance under light pulse and G resting is the resting conductance in the dark state before the light pulse) increases from ∼0.18 nS to ∼1.5 nS when the optical pulse width increases from 0.5 s to 10 s. It is worth noting that the transition from STP to LTP can also be turned by modulating the light intensity and the number of light pulses, as shown in Fig. 3b and c , which can significantly enhance and consolidate visual learning and memory. Fig. 3d is a schematic diagram showing the biological multiple memory model in the human brain. The memory can be strengthened by enhancing or continuously repeating the input of external information stimulation, turning SM (Sensory memory) into STM and further into LTM. 51 When learning new events, partial memory information will be gradually forgotten over time, as shown in Fig. 3e . The 31 light pulses were required to reach the conductance level of ∼5.8 nS from the conductance level of ∼3.8 nS, and this process takes 150 s. While, it only takes 21 light pulses to achieve the same conductance level when relearning from the same conductance lever of ∼3.8 nS, and the required time is reduced to 100 s. This phenomenon of conductance change illustrates the realization of the learning-forgetting-relearning process for external information on the proposed devices. This process is consistent with the Ebbinghaus forgetting curve, which describes the discipline of the human brain for forgetting new things. 52 Fig. 3 Long-term memory simulated based on TiO 2− x optoelectronic synaptic device by modulating the width (a), optical power density (b), and the number (c) of the light pulses. (d) The schematic diagram of the human brain “multiple memory”. (e) The “learning-forgetting-relearning” process mimicked on TiO 2− x optoelectronic synaptic device. Human memory can be regulated and influenced by internal emotions and ambient illumination. Changes in memory are consistent with internal emotions. The higher the internal emotion, the higher the level of memory. The same goes for memory retention time. Furthermore, the brightness of the ambient illumination can also affect learning and memory ability. In this work, the memory levels and retention characteristics (response conductance) of the same event (input optical pulses information) under different internal emotion states (different bias voltages) and ambient illumination (different brightness) are mapped to images “H” identification (predefined into a 3 × 3 array). Seven optoelectronic synaptic devices were stimulated with light pulses to obtain response current. The other two optoelectronic synaptic devices were tested for response current without light stimulation. The detailed setting methods are shown in Fig. S6 and S10. † The fabricated optoelectronic devices exhibit excellent sensitivity to the event memory under different emotional states and different ambient illumination, which more realistically simulate the memory properties of human visual information. The light intensity of the input optical pulse (event) information is 3.0 mW cm −2 , the pulse width is 2 s, the pulse interval is 2 s, and the number of pulses is 8. The different internal emotion states are triggered by different biases, which are set to 0.5 V, 0.7 V, and 1.0 V, respectively. As shown in Fig. 4 , the fabricated optoelectronic synaptic devices exhibit the behavior of internal emotions modulating visual memory. The visual memory based on relatively low emotion (resting conductance of ∼0.12 nS) at 0.5 V ( Fig. 4a ) was insufficient for “H” image recognition. The memory of the image almost disappears as the decay time increases to 200 s. The detailed current response curves were shown in Fig. S7. † As the bias voltage increases to 0.7 V, the value of the resting response conductance increases to ∼0.54 nS, indicating the raise of internal emotion (Fig. S8 † ). At the initial time, the image “H” can be recognized ( Fig. 4d ), but the image cannot be recognized as the decay time increases to 200 s ( Fig. 4f ). It is illustrated that insufficient emotional promotion has a weak effect on memory retention. While the response current can clearly identify the image “H” under relatively higher internal emotion with 1.0 V bias ( Fig. 4g ). The resting response conductance increases to ∼1.47 nS under 1.0 V bias (Fig. S9 † ). And memory similarly decays slowly as decay time increases. The response current still had an image-recognizable level at a decay time of 200 s under higher internal emotion ( Fig. 4i ). This phenomenon verifies the regulatory effect of internal emotion on visual memory. Fig. 4 The change of visual memory of the optoelectronic synapse array with decay time at 0 s, 100 s and 200 s under three levers of internal emotions. Additionally, the level of visual memory can be affected by ambient illumination. The spectral distribution, brightness, and duration of ambient illumination have a certain degree of influence on memory and emotion. The regulation of ipRGCs on the individual non-visual effects is mainly accomplished by melanopsin, which has photosensitive properties. Studies have found that melanopsin is most sensitive to the blue light spectrum with wavelengths in the range of 460–480 nm. 53,54 Therefore, a light source with a wavelength of 465 nm is adopted as ambient illumination in this paper. The same event was memorized after 20 s of ambient illumination. The proposed optoelectronic synaptic devices were in the same internal emotion. Fig. S10 † shows the specific setting form in this process. As shown in Fig. 5a, d and g , the memory abilities for the same event were slightly improved with the increase of the brightness of the ambient illumination. While the memory retention ability is significantly enhanced with the ambient brightness (Fig. S11–S13 † ). The image “H” can still be distinctly recognizable when reaching the same forgetting time of 200 s. Obviously, the brighter the ambient illumination, the more pronounced the memory effect. In addition, the “H” image can be discernible at a decay time of 500 s with the light intensity of 4.0 mW cm −2 , which clearly showed the characteristics of long-term memory (Fig. S14 † ). Ambient illumination has a non-visual effect on emotional states. The resting conductance, the emotional state, increases with the increasing brightness of the ambient illumination. At the same time, the maximum conductance triggered by the same event was also increased, indicating the enhancement of memory level. Moreover appropriately prolonging the duration time of ambient illumination can also enhance the level of memory-related proteins. 55 The increase in resting response conductance and the maximum conductance under light pulses were observed on the optoelectronic synaptic device as the illumination duration increased from 5 s to 50 s (Fig. S15 † ). Thus, memory under lower emotions can be complemented by prolonged duration or enhanced brightness of the ambient illumination. Fig. 5 The change of visual memory of the optoelectronic synapse array with decay time at 0 s, 100 s and 200 s under different brightness of the ambient illumination. Human visual memory was realistically mimicked in a designed 3 × 3 optoelectronic synapse array by adjusting the level of internal emotions and the brightness of the ambient illumination. It shows that the memory level can be significantly affected by internal emotions, and the memory maintenance ability is relatively more affected by the brightness of the ambient illumination. This behavior effectively simulates superior visual perception and visual memory performance. A detailed comparison between previously reported photoelectronic artificial synapse devices and our ITO/TiO 2− x / p -Si devices is provided in Table S1. † Herein, the image information detection and memory function that simulates the artificial visual system are achieved on our relatively simple device structure. More importantly, the effects of internal emotion and ambient illumination on visual perception and memory are discussed for the first time based on an optoelectronic synaptic array, which may lend it to a wider range of applications in complex conditions."
} | 5,985 |
39435740 | PMC11633514 | pmc | 5,899 | {
"abstract": "Abstract Presently described is a sliding‐kernel computation‐in‐memory (SKCIM) architecture conceptually involving two overlapping layers of functional arrays, one containing memory elements and artificial synapses for neuromorphic computation, the other is used for storing and sliding convolutional kernel matrices. A low‐temperature metal‐oxide thin‐film transistor (TFT) technology capable of monolithically integrating single‐gate TFTs, dual‐gate TFTs, and memory capacitors is deployed for the construction of a physical SKCIM system. Exhibiting an 88% reduction in memory access operations compared to state‐of‐the‐art systems, a 32 × 32 SKCIM system is applied to execute common convolution tasks. A more involved demonstration is the application of a 5‐layer, SKCIM‐based convolutional neural network to the classification of the modified national institute of standards and technology (MNIST) dataset of handwritten numerals, achieving an accuracy rate of over 95%.",
"conclusion": "2 Conclusion Consisting of two interconnected layers of memory arrays, a SKCIM architecture is proposed and demonstrated. One layer capable of in‐memory neuromorphic computation is used to store the data array; and the other is used to store and to shift the weight parameters. SKCIM can be applied to realize both convolutional and fully connected artificial neural networks. A MO TFT technology is used to physically implement a SKCIM system, allowing monolithic integration of the two layers in an array consisting of 5‐TFT‐3‐capacitor cells. The use of a capacitor as a memory element is made possible by the extremely low leakage current of a MO TFT. Dual‐gate TFT allowing simultaneous modulation of the channel current by its two separate gate biases is deployed as an artificial synapse for neuromorphic computation. Physical SKCIM systems have been constructed and deployed to execute convolution tasks and to classify the MNIST dataset of handwritten numerals.",
"introduction": "1 Introduction Convolutional neural networks (CNNs) [ \n \n 1 \n , \n 2 \n \n ] for feature identification are extensively deployed in applications involving pattern recognition, such as face detection [ \n \n 3 \n \n ] and autonomous driving. [ \n \n 4 \n \n ] The convolution operation can be visualized as the “sliding” of a small kernel matrix across a typically much larger data array while executing neuromorphic computation involving “vector‐matrix multiplication” (VMM) of the overlapping elements of the kernel matrix and the data array. [ \n \n 5 \n \n ] The result produced at each step is stored in a convolution feature map. When a CNN is implemented on a von Neumann computing machine with separate processor and memory units, [ \n \n 6 \n , \n 7 \n \n ] the small kernel matrix is held in the processor and repeated transfer of the elements of the data array from the memory to the processor and the elements of the feature map from the processor back to the memory is required while completing the convolution ( Figure \n 1 a ). Encountering what are often described as the energy and memory walls associated with neuromorphic computation, [ \n \n 8 \n , \n 9 \n \n ] the costs measured in energy and memory access time grow geometrically with increasing size of the data array. Figure 1 Hardware implementation of a CNN: a) von Neumann architecture, b) state‐of‐the‐art architecture with CIM‐based VMM processor, and c) the proposed SKCIM architecture. Implementing computation‐in‐memory (CIM) processors [ \n \n 10 \n , \n 11 \n , \n 12 \n , \n 13 \n \n ] dedicated to speeding up VMM operations for convolution, device technologies for realizing the artificial synapses required for neuromorphic computation have been reported. These technologies are based on either nonvolatile or volatile memories. The former includes resistive random‐access memory (RRAM), [ \n \n 14 \n , \n 15 \n \n ] floating‐gate transistor memory (Flash), [ \n \n 16 \n , \n 17 \n \n ] ferroelectric random‐access memory (FeRAM), [ \n \n 18 \n , \n 19 \n \n ] phase change memory (PCM), [ \n \n 20 \n \n ] and magnetoresistive random‐access memory (MRAM), [ \n \n 21 \n \n ] while the latter encompasses charge‐based dynamic random‐access memory (DRAM), [ \n \n 22 \n \n ] static random‐access memory (SRAM), [ \n \n 23 \n \n ] and embedded DRAM (eDRAM). [ \n \n 24 \n , \n 25 \n \n ] For instance, Yao et al. [ \n \n 26 \n \n ] employed a RRAM synaptic array, storing convolution kernels in the resistive elements and sequentially transferring blocks of the data array to the RRAM for computation, thereby implementing a CNN. This widely adopted scheme on CIM‐based VMM is well‐suited for fully connected (FC) neural networks, where the impact of transferring elements of the data array across the memory wall (Figure 1b ) is negligible due to the size of input vector being significantly smaller than that of the weight matrix. By contrast, convolution operations involve numerous small matrix multiplications, and the substantial ratio of data volume to the size of the convolution kernels creates a notable distinction from conventional FC layer computations. Therefore, existing schemes for CIM‐based VMM for convolution calculations will exacerbate the previously encountered effects of transferring data array elements across the memory wall. A more effective approach of overcoming the memory wall is to deploy CIM for executing VMM operations within the memory storing the data array, thus eliminating data transfer but demanding a mechanism for sliding a kernel matrix across the memory storing the data array. Presently proposed and displayed in Figure 1c is a “sliding kernel computation‐in‐memory” (SKCIM) system, which involves two overlapping layers of functional arrays, dubbed “DV” and “SS.” DV contains both memory elements for storing the data array and artificial synapses for neuromorphic VMM computation, whereas SS is used for storing and sliding the kernel matrices by sequentially erasing and writing of the components of a kernel matrix in the memory elements of SS. The frequent write–erase operations of convolution kernels present significant challenges to typical nonvolatile memory elements employed in the implementation of VMM. Specifically, RRAM and Flash suffer from low endurance and suboptimal write/read speeds; PCM exhibits deficiencies in both endurance and write power; while FeRAM and MRAM exhibit enhanced read/write capabilities and endurance, they encounter challenges related to cost and reliability. [ \n \n 27 \n \n ] In contrast to nonvolatile memories, volatile memories such as DRAM, eDRAM, and SRAM typically offer superior read/write performance and endurance. Metal‐oxide (MO)‐based eDRAM capitalizes on the ultralow leakage current [ \n \n 28 \n \n ] of MO thin‐film transistor (TFT) to achieve a pseudo‐nonvolatile storage behavior. With performance fitting the purpose of electronic displays, MO TFTs have been commercially deployed on an industrial scale. [ \n \n 29 \n , \n 30 \n \n ] Enabled by their exceptionally low leakage current, they are being explored as switching elements for the construction of embedded memory [ \n \n 31 \n \n ] based on capacitors as memory elements. Recently, Hu et al. [ \n \n 32 \n \n ] reported artificial neural networks (ANNs) composed of arrays of dual‐gate (DG) TFTs as artificial synapses and capacitors as memory elements. A physical SKCIM system based on MO TFTs is presently described. A 5‐layer CNN image‐processing system is demonstrated, consisting of two pairs of convolutional and pooling layers, and one fully connected ANN output layer. All 5 layers are implemented using 32 × 32 SKCIM arrays. Combining in and ex situ training methodologies, the system achieves a prediction accuracy of 95% and a reduction of 88% in memory‐access operations when applied to classification of the modified national institute of standards and technology (MNIST) dataset of handwritten numerals. Theoretical estimation is given, comparing how the memory‐access operations scale with the sizes of the data array and the kernel matrix when executing CNN on a state‐of‐the‐art system with dedicated CIM‐based VMM processor or a SKCIM system. 1.1 Device Fabrication and Characterization A low‐temperature MO TFT technology [ \n \n 33 \n , \n 34 \n \n ] capable of monolithically integrating single‐gate (SG) TFTs, DG TFTs, and memory capacitors (see Device Fabrication subsection under the Experimental Section and Figure S1 in the Supporting Information) has been deployed for the construction of a SKCIM system. Shown in Figure \n 2 a are the schematic rendering and scanning electron microscopy (SEM) images of a DG TFT. With at least a portion of its active layer (AC) sandwiched between its bottom‐gate (BG) and top‐gate (TG) electrodes, a DG TFT as an artificial synapse [ \n \n 35 \n , \n 36 \n \n ] is illustrated in Figure 2b . Its drain current I \n d emulating a postsynaptic current is simultaneously modulated by the presynaptic input ( V \n pre ) and the weight signals ( V \n Wt ) placed, respectively, on its high‐impedance TG and BG electrodes. Figure 2 a) The 3D structural schematic and cross‐sectional SEM images of a DG TFT. b) A biological synapse and a DG TFT as its electronic counterpart. c) Transfer and d) output characteristics of a SG TFT. e) Transfer characteristics of a DG TFT. With the source voltage V \n s of a SG TFT set to 0 V, the I \n d versus gate voltage V \n g transfer characteristics at drain voltage V \n d = 0.1 and 1.8 V are shown in Figure 2c . The I \n d versus V \n d output characteristics at 5 values of V \n g between 2 and 10 V are displayed in Figure 2d . The channel width ( W ) and length ( L ) of the TFTs are 5 µm. A threshold voltage V \n T of −1.3 V and a field‐effect mobility of ≈15 cm 2 V −1 s −1 are extracted from these characteristics. With the BG voltage V \n bg set at various values between − 4 and 4 V, the I \n d versus TG voltage V \n tg transfer characteristics of a DG TFT presented in Figure 2e clearly exhibit simultaneous modulation of I \n d by V \n bg and V \n tg . The DG TFTs exhibit negligible threshold drift in their transfer characteristics across a range of operating temperatures, thereby demonstrating its exceptional thermal stability (Figure S2 , Supporting Information). The uniformity of the TFTs has been characterized and shown in Figure S3 (Supporting Information). 1.2 The SKCIM Architecture Shown in Figure \n 3 a are the circuit schematics of cells making up the arrays in DV and SS. Each “3‐TFT‐1‐capacitor” (3T1C) cell at the i th i = 1 , … , m row and j th j = 1 , … , n column of an m × n DV array consists of a capacitor memory element C1 \n ij \n for storing a component DT \n ij \n of the data array, a SG TFT M1 \n ij \n for controlling access to the storage electrode of C1 \n ij \n , and two DG TFTs D1 \n ij \n and D2 \n ij \n as artificial synapses for neuromorphic computation. M1 \n ij \n is switched using a control signal placed on the row‐wise word line WL1 \n i \n to store on C1 \n ij \n the signal placed on the column‐wise bit line BL \n j \n . The TG electrodes of D1 \n ij \n and D2 \n ij \n are biased at DT \n ij \n stored on C1 \n ij \n . The I \n d s of D1 \n ij \n and D2 \n ij \n are accumulated, respectively, on the row‐wise “excitatory” and “inhibitory” source lines SE \n i \n and SI \n i \n . Figure 3 a) Two layers of a 5T3C cell. b) Schematic circuit diagram of a 5T3C cell, including 2 DG TFT artificial synapses, 3 memory capacitors, and 3 memory‐access SG TFTs. c) Schematic diagram of an m × n circuit array based on 5T3C cells. d) Schematic diagram of the peripheral circuit used to drive the 5T3C array. Schematic diagram illustrating the principles of data e) writing in and f) reading from the 5T3C array. The corresponding “2T2C” cell making up the SS array consists of two capacitor memory elements C2 \n ij \n and C3 \n ij \n for storing the respective excitatory and inhibitory weight signals WE \n ij \n and WI \n ij \n , and two SG TFTs M2 \n ij \n and M3 \n ij \n for controlling access to the respective storage electrodes of C2 \n ij \n and C3 \n ij \n . M2 \n ij \n and M3 \n ij \n are switched using the control signals placed on the respective row‐wise word lines WL2 \n i \n and WL3 \n i \n to store the weight signals placed on the shared bit line BL \n j \n . The corresponding cells in DV and SS are coupled by connecting the BG electrodes of D1 \n ij \n and D2 \n ij \n of a DV cell to the respective storage electrodes of C2 \n ij \n and C3 \n ij \n of a SS cell. When physically implemented, the 2 cells are monolithically integrated in a 5T3C circuit, as depicted in Figure 3b . Shown in Figure 3c is a schematic array showing the signals DT \n ij \n , WE \n ij \n , and WI \n ij \n stored in a SKCIM cell. Presented in Figure S4 (Supporting Information) are photographic images taken of a 5T3C array, a SG TFT, and a DG TFT. For the present demonstration, m = n = 32 and the physical sizes of the various components are summarized in Table S1 (Supporting Information). A schematic block diagram of the peripheral system for operating a SKCIM system is depicted in Figure 3d . The system consists of four VMM processing elements (PEs), each based on a 32 × 32 array of 1024 5T3C cells and three shift registers. Photographs of the experimental setup are shown in Figure S5 (Supporting Information). Data and weight signals up to 8 bits resolution are generated using an ARM microprocessor run at a clock frequency of 168 MHz. They are converted to their analog counterparts DT \n ij \n , WE \n ij \n , and WI \n ij \n using digital‐to‐analog converters (DACs) and distributed on the bit lines BL \n j \n using demultiplexers. Shift registers are connected to the word lines WL1 \n i \n , WL2 \n i \n , and WL3 \n i \n to sequentially turn on the respective access SG TFTs M1 \n ij \n , M2 \n ij \n , and M3 \n ij \n , thus passing the signals on BL \n j \n to the respective memory capacitors C1 \n ij \n , C2 \n ij \n , and C3 \n ij \n . Respectively, using transimpedance circuits and analog‐to‐digital converters (ADCs), the accumulated current signals on the source lines SE \n i \n and SI \n i \n are converted to a pair of voltage signals and digitized (Figure S6 , Supporting Information). The resulting signals are selected using multiplexers and passed on to the microprocessor for the computation of a difference voltage. The active‐matrix row‐scanning scheme deployed for placing DT \n ij \n , WE \n ij \n , or WI \n ij \n on C1 \n ij \n , C2 \n ij \n , or C3 \n ij \n by sequentially turning on a row of M1 \n ij \n , M2 \n ij \n , or M3 \n ij \n using control signals placed on the word lines WL1 \n i \n , WL2 \n i \n , or WL3 \n i \n is schematically illustrated in Figure 3e for DT \n ij \n . The sequential scanning stops when the entire set of signals is loaded. The reading of the data stored on C1 \n ij \n is accomplished using active‐matrix column scanning, as demonstrated in Figure 3f . Except for one selected column, the D1 \n ij \n s in the DV array are turned off by loading 0 V on the C2 \n ij \n s. The I \n d of the D1 \n ij \n s along the selected column are available on the source lines SE \n i \n . The corresponding data stored on C1 \n ij \n can be inferred using a measured DT versus I \n d calibration curve (Figure S7 , Supporting Information). The sequential scanning stops when the entire set of DT \n ij \n is read. Reading after writing of DT allows characterization of the storage performance of a capacitor memory element. 1.3 Long‐Term Memory Characterization of a 5T3C Cell The same test circuit of Figure S3a (Supporting Information) was characterized using the custom‐built measurement system shown in Figure S5 (Supporting Information), with V \n bg , V \n SS , and V \n DD , respectively, set at 5, 3.2, and 5 V. An 8 bits input V \n in between 1.8 and 5 V was fed on the bit line through a DAC to generate the V \n BL to be written as V \n tg on capacitor C1. The I \n d of TFT D1 was converted to a voltage using a transimpedance circuit and digitized using an ADC. The output of the ADC is converted to a current I \n out using a 1 MΩ load resistor R E i \n or R I i \n (Figure S6 , Supporting Information). More than 2000 scans of the I \n out versus V \n in characteristics were recorded, and a representative set is shown in Figure \n 4 a . It is clear I \n out could not be resolved with the same precision as V \n in , due to the significant overlap of the I \n out corresponding to adjacent levels of V \n in . The scans were repeated with a 5 bits V \n in , with Levels 0 and 31 corresponding, respectively, to 1.8 and 5 V. It can be seen from the dependence shown in Figure 4b,c that I \n out could be clearly resolved for all levels of V \n in . This reduction in precision from 8 to 5 bits is caused by the higher noise interference of the system compared to that of probe station. Figure 4 \n I \n out versus V \n in over 1000–2000 test cycles. a) Partially overlapping I \n out s of adjacent pairs of V \n in in 8 bits resolution. b) Clearly resolved I \n out s of pairs of adjacent levels of V \n in in 5 bits resolution even down to c) the lowest pair of adjacent Levels 0 and 1. d) Statistical distribution of the I \n out s of 3 groups of 3 consecutive levels of V \n in in decimal codes. e) Time decay of I \n out starting from Level 31 when the address TFT M1 is switched off. Hold (τ H ) and decay (τ D ) times are schematically indicated. f) Measured τ H and τ D corresponding to V \n in in decimal codes of 16–31. Extracted from the 2000 scans and shown in Figure 4d are the distribution of I \n out obtained at 9 levels of V \n in corresponding to decimal codes of 31, 30, 29; 17, 16, 15; and 7, 6, 5. For each level of V \n in , one can extract an upper bound I \n up and a lower bound I \n low of its corresponding I \n out distribution and compute a bandwidth Δ I \n BW ≡ I \n up − I \n low ; between any two adjacent levels, one extracts a lower bound I low _ h of the higher level and an upper bound I up _ l of the lower level and computes a band separation Δ I SP ≡ I low _ h − I up _ l . When a V \n in corresponding to Level 31 is stored on C1 and TFT M1 is turned off, the resulting V \n tg (hence I \n out ) decays due to leakage to Level 27 in about 170 min. The time dependence of I \n out is exhibited in Figure 4e . The time it takes to traverse the Δ I \n BW of a given level is defined as the “hold time” τ H and the time it takes to traverse the Δ I \n SP between adjacent bands is defined as the “decay time” τ D . The dependence of τ H and τ D on V \n in are extracted and shown in Figure 4f . τ H is shorter at a higher level of V \n in and it is about 15 min at Level 31. This is sufficiently long for completing the convolution tasks in the present study. The three capacitors in the 5T3C cell, in conjunction with the addressing transistors, enable long‐term signal storage. The two DG TFTs perform computations that emulate the behavior of synapses, including potentiation, depression, and both excitatory and inhibitory actions (see Characterization of Excitatory, Inhibitory, Potentiation, and Depression Behavior subsection in the Experimental Section and Figure S8 in the Supporting Information). 1.4 Operation of the SKCIM PEs The sliding of a kernel matrix in the SS layer during convolution is schematically illustrated in Figure \n 5 a , using an 8 × 8 data array and a 3 × 3 kernel matrix as an example. The entire set of DT \n ij \n is first loaded in the DV array; the WE \n ij \n and WI \n ij \n on C2 \n ij \n and C3 \n ij \n are initialized to 0 V, thus turning off all D1 \n ij \n and D2 \n ij \n . Since 3 is not a factor of 8, only two 3 × 3 kernel matrices can be stacked in an array with 8 rows. The operation starts with the sequential loading of the first, second, and last rows of the kernel matrix, respectively, in Rows 1 and 4, 2 and 5, and 3 and 6 of the SS array. Consequently, a 6 × 3 portion of Columns 1, 2, and 3 of the SS array, containing 2 copies of the kernel matrices, is created (Figure 5b ). Figure 5 a) Schematic illustration of the storage of a data array in a DV layer and the “sliding” of a kernel matrix in a SS layer. b) Writing of two kernel matrices in an 8 × 8 SS array, starting with their first rows. c) Schematic illustration of the simultaneous “sliding” of two kernel matrices. d) Schematic trajectory of the sliding of two kernel matrices in the SS layer. e) Original image. f) Examples of convolutional operations: impulse response, smoothing, sharpening, and embossing. For each copy of the kernel matrix, the difference voltage values from the three rows corresponding to a kernel matrix are summed to generate a convolution feature value. In the next step of the operation, the vertically stacked 6 × 3 kernel matrices are shifted down by 1 row, thus occupying Rows 2–7 of the SS layer. The respective WE \n ij \n and WI \n ij \n on C2 \n ij \n and C3 \n ij \n of Row 1 of the SS layer are reset to 0 V and a new pair of feature values are computed and recorded in their corresponding locations of the convolution feature map. This vertical sliding operation continues until the bottom edge of the SS array is reached. In the present example, 6 convolution feature values are generated and recorded. The respective WE \n ij \n and WI \n ij \n on C2 \n ij \n and C3 \n ij \n of the SS layer are reset to 0 V before the sequence of vertical sliding operations is repeated after laterally sliding the kernel matrices to the right by 1 column, now involving Columns 2, 3, and 4 (Figure 5c ). This combination of lateral followed by vertical sliding operations continues until the right edge of the SS array is reached and a 6 × 6 convolution feature map is obtained (Figure 5d ). This SKCIM‐based implementation of CNN can be readily generalized to the combination of any m × n data array and any p × p kernel matrix. Four PEs were selected to demonstrate the implementation of a CNN. The global uniformity characterization of each PE, along with the periodic scanning of PE1, is presented in Figure S9 (Supporting Information) (see Stability and Uniformity Characterization of SKCIM PEs subsection in the Experimental Section and Figure S9 in the Supporting Information). Exhibiting the best uniformity, PE1 is chosen for the demonstration of the convolution operations. The utility of a SKCIM system for convolution is verified using a 30 × 30 image (Figure 5e ), with each pixel in the image consisting of 3 subpixels of the primary colors of “red,” “green,” and “blue.” Four convolution kernels K1, K2, K3, and K4 (Figure S10 , Supporting Information) are deployed, producing convoluted images showing the respective effects of “K1: impulse response” for largely preserving the original image, “K2: smoothing” for blurring noise and details in the image, “K3: sharpening” for enhancing details and edges, and “K4: embossing” for directional edge enhancement (Figure 5f ). 1.5 Comparison of von Neumann, State‐of‐the‐Art VMM Processor and SKCIM The three components of main memory, cache memory, and the processor required for convolution are considered. The processor is either a conventional processor for a von Neuman machine or a CIM‐based VMM processor. The corresponding channels ( Figure \n 6 a ) of memory‐access operations (MAO) are T1 for loading of the data array from the main memory to the cache, T2 for sequential loading of the components of the data array or the kernel matrix from the cache to the processor, T3 for transferring components of the convolution feature map from the processor to the cache, and T4 for saving of the feature map from the cache to the main memory. Figure 6 a) Schematic illustration of data transfer channels of T1–T4. Theoretical η T2 in kilo‐operations (KOP) of the three architectures with kernel matrix sizes of b) 3 × 3 and c) 5 × 5. d) Emulation of state‐of‐the‐art CIM‐based VMM system using a SKCIM system. e) Measured (solid line) and extrapolated (dotted line) time to complete η T2 data transfers of a state‐of‐the‐art CIM‐based VMM and a SKCIM system when executing convolution of a 28 × 28 image with different kernel matrix sizes. Since T1, T3, and T4 are deployed in all three architectures of von Neumann, state‐of‐the‐art VMM processor, and SKCIM, comparison will be made only of the MAO associated with T2. Consider as an example an m × m data array convoluting with a p × p kernel matrix and producing a ( m − p + 1) × ( m − p + 1) feature map. The number of MAO is m \n 2 for T1, and ( m − p + 1) 2 for T3 and T4. For T2 of the first two architectures, the data array is divided into p × p component blocks, each with p \n 2 elements. Since there are ( m − p + 1) 2 blocks that need to be transferred from the cache to the processor, the resulting number of MAO is η T2 = ( m − p + 1) 2 · p \n 2 . Note that η T2 changes quadratically with m when m ≫ p . For the SKCIM architecture, since it is the elements of the kernel matrix rather than those of the data array that need to be transferred, there are ( m − p + 1) lateral sliding operations and p vertical sliding operations of p \n 2 elements of the kernel matrix. The product of the three terms gives rise to an η T2 = ( m − p + 1) · p \n 3 . Note that η T2 changes linearly with m when m ≫ p . Compared in Figure 6b,c are the η T2 of the three architectures for m = 28 and p = 3 and 5, respectively. It can be seen that η T2 for the SKCIM architecture is significantly reduced compared to those for the other two architectures, respectively, by 88% and 79% for p = 3 and 5. The state‐of‐the‐art VMM processor for implementing convolution can be emulated using the SKCIM system (Figure 3d ), as illustrated in Figure 6d . In this system, SKCIM‐based convolution is performed by storing data in the DV layer of the 5T3C array while sliding the convolution kernel across the SS layer. Whereas the 5T3C array emulates the state‐of‐the‐art VMM processor for convolution calculations by positioning the unfolded convolution kernel in the first row of the SS layer and transferring image data from the cache in blocks to the first column of the DV layer for computation with the SS layer. For example, A p × p kernel matrix is flattened and loaded on the weight‐storing capacitors in the first row of the SS array. The M1 \n ij \n s are turned on and p × p blocks of the data array to be convoluted is prepared and sequentially moved to the first row of the DV array for computation. Presented in Figure 6e is the measured data transfer time from the cache to the processor to complete a convolution. For both p = 3 and 5, the transfer times are measured for m = 8, 16, 24, and 32. η T2 is used to estimate the number of transfer operations and applied to extrapolate the total transfer times for data array with m > 32 and up to 128. Both emulated state‐of‐the‐art VMM processor and SKCIM are investigated. The latter exhibits obvious reduction in the required transfer time for a given m , consistent with the different dependences of η T2 on m for the two architectures. 1.6 Application of SKCIM to MNIST Recognition Consisting of two pairs of convolution followed by pooling layers and a fully connected ANN output layer, a 5‐layer system ( Figure \n 7 a ) for classification of the MNIST dataset of handwritten numerals has been constructed. The layers are implemented using four SKCIM PEs. At 28 × 28, the size of the image is smaller, thus it can be fully accommodated by a larger 32 × 32 SKCIM PE. Three 5 × 5 convolution kernel matrices are deployed to generate a 24 × 24 × 3 feature map C1 using the first layer. C1 is downsampled by pooling using a 3 × 3 filter to generate an 8 × 8 × 3 feature map S2 using the second layer. Six 3 × 3 × 3 convolution kernel matrices are applied to S2 to generate a 6 × 6 × 6 feature map C3 using the third layer. Subsequently, C3 is downsampled by pooling using a 2 × 2 filter to generate a 3 × 3 × 6 feature map S4 using the fourth layer. Finally, S4 is flattened and fed to the fully connected 56 × 10 ANN output layer. Figure 7 a) Structure of the 5‐layer CNN used for recognition of the MNIST dataset of handwritten numerals: 2 convolutional layers, 2 pooling layers, and a fully connected ANN layer. b) Comparison of the accuracy rates of the SKCIM systems with two different initial weight values and the software implementation. c) Evolution of the accuracy rates obtained using the training and the inference sets. d) Kernel matrices and ANN weight values after training. e) Statistical distribution of the weight values corresponding to initial conditions of zero and random weights. The implementation of the ANN layer using a 32 × 32 SKCIM PE is accomplished by splitting the 54 inputs of the flattened P4 into two rows of 27 inputs. A 2 × 28 input data array is generated after each row is augmented with a bias value. The first 20 rows of the SKCIM PE are used, with the first and second rows of the data array, respectively, loaded in the odd and even numbered rows of the SKCIM PE. Starting from the top of the SKCIM PE, the outputs of every consecutive pair of odd and even numbered rows are accumulated to generate a total of 10 outputs from the 20 rows of the ANN. While the training of a CNN requires external computers for ex situ training to improve training speed, the inference relies entirely on hardware implementation with enhanced operational efficiency. A sequence of ex situ followed by in situ methodologies is adopted [ \n \n 14 \n \n ] for training the SKCIM‐based physical system. The ex situ methodology involves the construction using Python and TensorFlow of an approximate software model that closely resembles the physical system. Rectified linear unit (ReLU) is used as an activation function. The convolutional kernel matrices and the weight parameters for the ANN are obtained. An accuracy of 97.4% (Figure 7b ) is achieved on a test set comprising 10 000 samples using an optimally trained model. After appropriate scaling to match the electrical bias adopted to operate the physical system, the transformed kernel matrices are deployed for convolutional operations (see System Training subsection in the Experimental Section). The weight parameters of the ANN were initialized either to 0 V or random values before in situ training of the ANN is carried out to account for the difference between the software and the physical systems, and to accommodate the nonuniformity in TFT parameters in the latter. An approximate gradient of the cost function is estimated and used to update the weight parameters. This process is repeated until convergence of the weight parameters is achieved. More details are given in Note S1 (Supporting Information). Depicted in Figure 7c are the evolution of the inference accuracy versus training epochs for a system with kernel matrices obtained from ex situ training and ANN weights initialized to random values. Starting with a relatively low ≈15%, the predicted accuracy quickly saturates after 4 or 5 epochs of training and reaches ≈95% after 16 training epochs. Shown in Figure 7d are the kernel matrices and the trained ANN weights. As a comparison, the statistical distribution of the weights corresponding to initial ANN weights of 0 V and random values are shown in Figure 7e , showing largely similar trends. Summarized in Table \n \n 1 \n is a comparison of the proposed SKCIM and other systems based both on TFTs and conventional Si transistors. For the present system implemented using a 5 µm indium–tin–zinc oxide (ITZO) TFT technology, a conservative working clock frequency of 1 MHz was used. The maximum number of parallel neuromorphic instructions computed per clock cycle is 32 × 32 × 2 = 2048, resulting in a maximum throughput of 8.192 giga operations (GOP) s −1 (Table S2 , Supporting Information). Since the operating speed of a TFT scales with L \n −β for 1 ≤ β ≤ 2, the throughput could be increased by tenfold to 100‐fold when length L is scaled by tenfold from 5 to 0.5 µm. [ \n \n 37 \n \n ] It should be noted that MO TFT with L below 10 nm [ \n \n 24 \n \n ] and operating frequencies over GHz [ \n \n 37 \n , \n 38 \n \n ] have been reported. Table 1 Comparison with other published SNNs. This work IEEE Transactions on Circuits and Systems I ’23 [ \n \n 24 \n \n ] \n Japanese Journal of Applied Physics 20 [ \n \n 22 \n \n ] \n International Solid‐State Circuits Conference 21 [ \n \n 25 \n \n ] \n International Conference for High Performance Computing 23 [ \n \n 31 \n \n ] \n Technology \n TFT \n 5 mm \n \n TFT \n 45 nm \n \n TFT \n 4 mm \n TFT 350 nm/Si 110 nm \n Si \n 65 nm \n \n Si \n 40 nm \n Frequency 1 MHz N.A. N.A. 25 MHz 50 MHz N.A. CIM scheme eDRAM eDRAM eDRAM eDRAM DRAM DRAM Retention ≈40 h \n a) \n \n 20 s \n b) \n \n ≈10 h \n b) \n \n >30 h \n c) \n \n N.A. N.A. Precision [bit] 5 4 Analog N.A. 8 8 Power 348 mW N.A. N.A. 9.95 mV 0.99 mW 1.06 W Efficiency (Tera Operations Per Second W −1 ) 23.5 d) \n 686 d) \n 0.728 \n d) \n \n 5 4.76 2.41 \n a) \n Defined as median state degradation; \n b) \n Defined as the time at 0.5 least significant bit output error with evaluated number of rows in a multiply accumulate; \n c) \n Defined as 5% degradation for analog multiplication; \n d) \n CIM array only. John Wiley & Sons, Ltd."
} | 8,367 |
35252877 | null | s2 | 5,900 | {
"abstract": "No abstract available"
} | 5 |
30467278 | PMC6308720 | pmc | 5,901 | {
"abstract": "The management and proper use of the Urban Public Transport Systems (UPTS) constitutes a critical field that has not been investigated in accordance to its relevance and urgent idiosyncrasy within the Smart Cities realm. Swarm Intelligence is a very promising paradigm to deal with such complex and dynamic systems. It presents robust, scalable, and self-organized behavior to deal with dynamic and fast changing systems. The intelligence of cities can be modelled as a swarm of digital telecommunication networks (the nerves), ubiquitously embedded intelligence, sensors and tags, and software. In this paper, a new approach based on the use of the Natural Computing paradigm and Collective Computation is shown, more concretely taking advantage of an Ant Colony Optimization algorithm variation and Fireworks algorithms to build a system that makes the complete control of the UPTS a tangible reality.",
"introduction": "1. Introduction Since their pioneering conception in 1829, underground trains have changed in in so many ways. From the 47 km/h that Stephenson’s train reached in the aforementioned year, to the 310 km/h that the Spanish AVE is capable of obtaining, trains have experienced an evident impact regarding their technology. However, these changes have not been applied to the management system and conception of the underground itself as it is nowadays. On the one hand, the rapidly growing massification of the world’s urban cores together with the intensive use by citizens of the underground, is pushing the transition of these cores to the Smart City purest concept, where every single element within the city has ratiocination enough for it to be called intelligent. In the year 2050, 66% of the world’s population is expected to be living in urban cores (United Nations, Department of Economic and Social Affairs, New York, NY, USA), increasing the current percentage of 54% by 12%. In other words, the current estimations show that the continuous urbanization process that the world is facing, along with the overall growth of the world’s population, will add another 2.5 billion people to urban populations by 2050, with close to 90% of the increase concentrated in Africa and Asia, according to a new United Nations report. To sum up, 66% of a world population of 9 Billion (5.94 Billion) will be living in urban cores in 2050. The aforementioned massification can be seen in Figure 1 . On the other hand, it is important to note that this need has been outlined by organisms such as C.E.O.E (Spanish Confederation of Business Organizations, Madrid, Spain). In fact, as described in CEOE [ 3 ]:\n This frame of sustainability and efficiency that must involve the Smart Cities, has a direct relationship with other key areas, such as […] the efficient management of the mobility of people […] [Cities are lacking] Indicators for the collection appropriate measures […] [Cities systems need] real-time knowledge about incidents, and an improved efficiency and management of the public transport. It is, therefore, evident that cities nowadays need a deep improvement on their IT systems and infrastructure, evolving to new schemes where data is seen as a binder for the city. To contribute to this goal, a gathering and management system based on Natural Computing is presented on this paper. Even the concept of Smart City is still being under constant redefinition, most authors agree that many different individuals, agents, and devices, operate with their environment within the Smart City realm [ 4 , 5 , 6 , 7 ]. Therefore, as R. G. Hollands [ 8 ] points out, the relation among all these elements will define the behavior of the Smart City itself. It is easy to realize that an important area of the Smart City will be based on the interaction between its different components with their environment. This fact disembogues in a Socio-Collective Interaction, where the Smart City in general terms, and specially the underground system beneath, can be seen as a huge swarm, where agents collaborate with each other [ 9 ]. The aforementioned approach justifies the present investigation project, based on a change in the way of tackling the management processes of any underground system, using Collective Computation algorithms [ 10 , 11 , 12 ] instead of the classical, graph-oriented ones [ 13 ].",
"discussion": "7. Conclusions and Discussion In this paper, a new scheme for endowing intelligence to a city UPTS is given, chasing the transition of the city to a Smart City. In this approach, Natural Computing paradigm will be applied to the system, after a deep investigation that aims to improve the involved paradigms, if possible. Despite the investigation still being in an early stage, the system is likely to improve the data gathering related to the UPTS, allowing the pertinent authorities to improve the system and even monetize the information gathered by the system under development. Moreover, users will be able to enjoy a better use of UPTS, knowing alternative routes in case of systems breakdown and being able to travel in an efficient way. Swarm intelligence holds a great potential to totally transform the way traffic patterns affect our daily lives and our daily commutes. It is now up to city planners throughout our major metropolises to recognize the benefits of these simulations and to create the necessary infrastructure."
} | 1,335 |
39463925 | PMC11507797 | pmc | 5,902 | {
"abstract": "Population cycles are prevalent in ecosystems and play key roles in determining their functions 1 , 2 . While multiple mechanisms have been theoretically shown to generate population cycles 3 – 6 , there are limited examples of mutualisms driving self-sustained oscillations. Using an engineered microbial community that cross-feeds essential amino acids, we experimentally demonstrate cycles in strain abundance that are robust across environmental conditions. A nonlinear dynamical model that incorporates the experimentally observed cross-inhibition of amino acid production recapitulates the population cycles. The model shows that the cycles represent internally generated relaxation oscillations, which emerge when fast resource dynamics with positive feedback drive slow changes in strain abundance. Our findings highlight the critical role of resource dynamics and feedback in shaping population cycles in microbial communities and have implications for biotechnology.",
"discussion": "Discussion Previous models of mutualisms in general, and cross-feeding in particular, do not predict self-sustained oscillations in the absence of external forcing. The salient difference in our system is the feedback generated by amino acid release in response to amino acid limitation of the producer auxotroph ( Fig. 2b ). This represents a strategy of resource management involving the release of excess produced amino acid during periods of limitation by the required amino acid. This strategy could be explained by an endogenous stress response mechanism in E. coli in which specific amino acids, including the aromatics, accumulate during starvation because of protein degradation 25 , 37 , 38 . As an alternative explanation, the observed resource dynamics could be an indirect consequence of the auxotrophic gene knockouts. Metabolic flux intended to remedy starvation of the required amino acid could easily leak into an overlapping pathway (see Supplementary Information ). Indeed, the anabolic pathways for tyrosine and phenylalanine are nearly identical with only the final two reactions being distinct, suggesting that dysregulation of one pathway could impact the metabolic fluxes of the other pathway. Beyond the specific mechanism, the resulting positive feedback loop captured in our models is essential to the oscillations in our system. Further, similar features of our system may occur more broadly in ecological systems shaped by resource exchange beyond microbial communities. A plant growth model with separate root and shoot compartments that share excess resources (nitrogen and carbon) shows similar dynamics, alternating between two metabolic equilibria 39 , 40 . Although the interaction between strains is mutualistic (+/+) when averaged over an entire cycle, at most instants one auxotroph benefits without concurrently returning the favor, representing a transient commensalism. Due to the oscillations, species take turns as they alternate out of phase between disjoint periods of growth and amino acid resource production ( Fig. 4a ). Such dynamics mirror reciprocal altruism, where each strain temporarily sacrifices its growth to benefit the other, ensuring long-term mutual benefit 41 . One consequence of these oscillations is that they can prevent the disruption of the mutualism by a cheater that consumes the exchanged resources without producing either of these resources (e.g. a dual auxotroph) in model simulations. In conditions where the community converges to a stable equilibrium, such a cheater can invade a community of two cross-feeding auxotrophs that are each deficient in producing a single amino acid. In this case, the community collapses as the mutualists go extinct in response to depletion of the exchanged resources by the cheater 23 . Because the cheater requires both exchanged resources, any mechanism that separates their simultaneous availability can prevent a successful invasion. Previously, we have shown that local interactions between cross-feeding auxotrophs can exclude cheaters by spatially separating the exchanged resources 42 . Notably, temporal separation of resources, due to the alternating production of individual amino acids in the oscillating community ( Fig. 4a ), provides another mechanism to exclude an otherwise successful cheater ( Extended Data Fig. 6 ). This implies that oscillations may represent an ecological strategy to resist invasion by cheaters, which could be leveraged for engineering stable microbial communities for biotechnology applications."
} | 1,131 |
20523735 | PMC2877717 | pmc | 5,903 | {
"abstract": "Background Coral reefs are hotspots of biodiversity, yet processes of diversification in these ecosystems are poorly understood. The environmental heterogeneity of coral reef environments could be an important contributor to diversification, however, evidence supporting ecological speciation in corals is sparse. Here, we present data from a widespread coral species that reveals a strong association of host and symbiont lineages with specific habitats, consistent with distinct, sympatric gene pools that are maintained through ecologically-based selection. Methodology/Principal Findings Populations of a common brooding coral, Seriatopora hystrix , were sampled from three adjacent reef habitats (spanning a ∼30 m depth range) at three locations on the Great Barrier Reef (n = 336). The populations were assessed for genetic structure using a combination of mitochondrial (putative control region) and nuclear (three microsatellites) markers for the coral host, and the ITS2 region of the ribosomal DNA for the algal symbionts ( Symbiodinium ). Our results show concordant genetic partitioning of both the coral host and its symbionts across the different habitats, independent of sampling location. Conclusions/Significance This study demonstrates that coral populations and their associated symbionts can be highly structured across habitats on a single reef. Coral populations from adjacent habitats were found to be genetically isolated from each other, whereas genetic similarity was maintained across similar habitat types at different locations. The most parsimonious explanation for the observed genetic partitioning across habitats is that adaptation to the local environment has caused ecological divergence of distinct genetic groups within S. hystrix .",
"conclusion": "Conclusions Even though genetic variability between habitats has been previously demonstrated [46] , this study clearly indicates that habitats within a reef can be genetically isolated from each other, whereas the same habitat types separated by up to ∼20 km can exhibit high levels of genetic similarity. Furthermore, the observed genetic partitioning demonstrates that the cryptic diversity previously detected in S. hystrix \n [40] , [41] may be a reflection of lineages associated with distinct reef environments. Habitat-associated cryptic diversity may explain some of the “stochastic” results and high levels of genetic structuring over short geographic distances commonly observed in genetic studies of scleractinian corals. This study highlights the need to further explore genetic diversity over environmental gradients in other coral species, preferably encompassing species with a variety of life history strategies and broad ecological distributions. This is particularly important in the context of local reef connectivity and the general conception that deeper sections of reefs [18] , [79] may act as a reproductive source for shallow reef areas following disturbance [23] . The strong association of host and symbiont genotypes with particular reef environments presents a compelling case for ecological speciation, corroborating previous evidence [13] – [15] that ecologically-based divergent selection may be an important mechanism for diversification on coral reefs. Overall, it underscores the need for understanding processes that shape diversity, which will allow for more accurate predictions on the persistence and community structure of coral reefs in a future of increasing anthropogenic and climate pressures.",
"introduction": "Introduction The tropical marine realm harbors an incredible array of species, with coral reef ecosystems being the iconic epitome of this diversity. Classically, this diversity has been explained through allopatric models of speciation, in which reproductive isolation arises through the physical separation of populations [1] . However, speciation has also been demonstrated to occur sympatrically or parapatrically, where divergence originates in the absence of physical barriers and is driven by ecological sources of divergent selection (i.e. selection occurring in opposing directions) (reviewed in [2] ). A classic marine example of incipient speciation in sympatry is that of the intertidal snail Littorina saxatilis comprising genetically distinct ecotypes (with different shell morphologies), which are partitioned over a gradient of tidal height [3] – [5] . Eventually, divergent selection can lead to complete reproductive isolation (i.e. ecological speciation), either as a by-product of divergent selection (via linkage disequilibrium [6] ), or directly when the genes involved in reproductive isolation are under ecologically-based divergent selection (via pleiotropy [2] ). Over the past decade, ecological speciation has been suggested as an explanation for the diversification of various terrestrial and freshwater taxa [7] – [12] , yet only three examples have been proposed for tropical reef organisms: wrasses of the genus Halichoeres \n [13] , sponges of the genus Chondrilla , [14] , [15] ), and the scleractinian coral Favia fragum \n [16] . Ecological diversification can arise through various sources of divergent selection (reviewed in [2] ), including sexual selection (e.g. selection on mate recognition traits) and ecological interactions (e.g. interspecific competition). However, divergent selection between distinct environments is probably the best understood cause of ecological speciation [2] and has been proposed as a major contributing factor to the diversification of species in environmentally heterogeneous ecosystems such as tropical rainforests [7] . Coral reefs provide a similarly heterogeneous environment, with large variability in abiotic factors such as light [17] , temperature [18] , nutrient variability, and wave action [19] , [20] between locations and also across depths at a single location. Although such environmental variability may favor a certain degree of plasticity [21] , there is also the potential for locally adapted “ecotypes” to evolve through divergent selection [22] . Surprisingly, little is known about genetic structuring of coral reef populations across distinct habitats [23] and the potential role of environmental heterogeneity in species diversification. Numerous observations indicate that ecological diversification may be important in coral evolution. Firstly, there are many examples where closely related, sympatric species occupy only part of the available habitat, i.e. they occupy a distinct environmental niche [24] – [26] . The Caribbean coral genus Madracis provides a good example in point; it consists of six closely related morpho-species [27] , each of which has a distinct depth-distribution [28] . Other examples of closely related species exhibiting similar habitat partitioning are members of the genus Agaricia \n [29] , [30] , the Montastraea annularis species complex [31] , [32] , and the acroporids, Acropora palifera and A. cuneata \n [33] , [34] . Secondly, on an intra-specific level, there are several observations suggestive of local adaptation (reviewed in [35] ), such as local dominance of certain genets [36] , [37] and variation between populations in thermal tolerance [38] , and natural disease resistance [39] . Thirdly, many studies have observed cryptic diversity (e.g., in the genera Seriatopora \n [40] , [41] , Pocillopora \n [42] , and Porites \n [43] ) and genetic differentiation over small spatial scales in the absence of physical barriers (e.g., in population genetics studies on Seriatopora \n [44] , [45] ). Despite these lines of evidence, the specific hypothesis of ecological diversification remains largely untested for corals. Most genetic assessments have focused on concordance of observed patterns with morphology or geography rather than physiological characteristics or habitat. Exceptions are the studies by Carlon and Budd [16] , which established that morpho-types of the coral F. fragum are genetically distinct and partitioned over a small depth gradient (∼3 m), and Ayre et al. [46] , which demonstrated that a proportion (16%) of the genetic variability of Seriatopora hystrix within reefs could be explained by distributions among five shallow reef habitats (reef slope, reef crest, reef flat, lagoon, and back reef). Habitat partitioning and ecological diversification have also been observed for the photosymbiotic partners ( Symbiodinium ) of scleractinian corals. Various coral species harbor distinct depth-specific symbiont types across their distribution range (e.g. [47] – [49] ). Despite this apparent flexibility to associate with various symbiont types over depth [50] , the coral-algal symbiosis is generally characterized by a high degree of host-symbiont specificity [51] , in that coral species usually only associate with certain types of Symbiodinium and vice versa. This specificity is especially apparent in corals with a vertical symbiont transmission strategy, in which symbionts are passed directly from the maternal colony to the offspring [52] – [54] . Thus, corals with vertical symbiont transmission are most likely to codiversify with their algal symbionts but this process is poorly understood since studies considering both host and symbiont identity with a fine-scale genetic resolution are rare [53] . The scleractinian coral S. hystrix represents an ideal candidate to examine processes of ecological diversification and local adaptation in corals, as it occurs in most habitats [55] , and is geographically widespread [26] . S. hystrix exhibits a brooding reproductive strategy and vertically transmits associated Symbiodinium . Furthermore, S. hystrix has been the subject of several genetic studies [44] – [46] , [56] – [60] , with previous allozyme work indicating that genetic structuring of the coral host may occur among shallow habitats [46] . Here, we specifically test the extent of genetic structuring over a large depth range for both symbiotic partners, and evaluate genetic differentiation between the same habitat types at different locations. Focusing on three adjacent, environmentally distinct habitats (spanning a depth range of ∼30 m) at three locations (Yonge Reef, Day Reef and Lizard Island; Figure 1 ) on the northern Great Barrier Reef, we used a combination of mitochondrial (putative control region) and nuclear (microsatellites) markers for the host, and the ITS2 region of the nuclear ribosomal DNA for Symbiodinium to assess genetic differentiation. Results indicate that adjacent habitats within a single reef can be genetically isolated from each other, whereas genetic similarity is maintained between the same habitat types at different locations. The strong partitioning of both host and symbiont lineages occurs between directly adjacent habitats in the absence of physical dispersal barriers, and thus provides a compelling case for divergence due to ecologically-based divergent selection. 10.1371/journal.pone.0010871.g001 Figure 1 Sample design and locations. (A) Map showing the geographic location of the study area on the northern Great Barrier Reef; (B) the reef locations within the study area; and (C) the different habitats sampled.",
"discussion": "Discussion This study demonstrates that S. hystrix and its associated Symbiodinium form genetically isolated clusters across distinct reef habitats ( Figures 2 , 3 , 4 ). The association of host lineages (mtDNA) and genetic clusters (nDNA) with particular reef-environments rather than geographic location is consistent with divergence occurring through ecologically-based selection. Furthermore, the observed coupling of host and symbiont genotypes points to codiversification at a fine taxonomical level. Habitat partitioning of coral host populations/genotypes The three habitats sampled in this study ( ‘Back Reef’ , ‘Upper Slope’ , ‘Deep Slope’ ; Figure 1 ) differ greatly in exposure to wave action, temperature regimes and light availability (see methods for detailed description). Across these habitats, strong partitioning of host mtDNA haplotypes is observed, with the ‘Back Reef’ and ‘Upper Slope’ habitats each containing a single haplotype and the ‘Deep Slope’ habitat containing four different haplotypes. Similarly, Bayesian analysis of three microsatellite loci revealed four genetic clusters, with each cluster corresponding to one of the four common mtDNA haplotypes (i.e., individuals in each cluster share the same mtDNA genotype). There was also strong genetic differentiation across habitats based on microsatellites under the AMOVA framework. Replication of these striking genetic patterns across two distinct reef locations (∼20 km apart) is consistent with local adaptation to distinct habitats followed by non-random mating, as the genetic structure was observed in putative neutral loci. Detectable genetic structure based on linkage disequilibria among microsatellite loci can develop over relatively short timescales, however partitioning based on mitochondrial loci should reflect longer (evolutionary) timescales. As such, the observed partitioning likely reflects long-standing adaptations to the unique environmental conditions of each habitat, such as strong wave action (e.g. ‘Upper Slope’ ), extreme temperature fluctuations (e.g. ‘Back Reef’ ) or low-light conditions and cold-water influxes (e.g. ‘Deep Slope’ ). However, given that various abiotic factors covary between habitats, it is impossible at this time to assess the likely contribution of specific environmental variables to the observed genetic partitioning. Although diversity generally declines with depth in coral species [63] and Symbiodinium types [48] , [62] , here we observed the highest diversity of host and symbiont genotypes in the deeper habitat. Even though populations of S. hystrix are generally highly structured geographically across the Great Barrier Reef (GBR) [45] , [46] , [56] , [58] , van Oppen et al. [45] reported high genetic similarity between populations on the Ribbon Reefs (including Yonge Reef) of the northern GBR with pairwise F ST values ranging from 0.009–0.026 for reefs up to ∼80 km apart. Additionally, genetic similarity was observed between the Ribbon Reefs and a population at Lizard Island (F ST = 0.065–0.090). These results are concordant with the genetic similarity observed in this study between Yonge Reef and Day Reef (within habitats) using a subset of the microsatellite loci used by van Oppen et al. [45] (F ST = 0.012–0.060) and between Yonge Reef, Day Reef and Lizard Island (within habitats) for the mtDNA locus. Thus, despite the highly localized recruitment of S. hystrix \n [44] , larval exchange between directly adjacent habitats (∼50–500 m apart) is unlikely to be hampered by physical barriers. Rather, the differentiation across habitats seems to be driven by non-allopatric diversification processes. Ayre and Dufty [46] were the first to identify an effect of habitat on genetic differentiation in S. hystrix . In their allozyme study they reported that a proportion of the within reef variability was explained by variation among five shallow habitat types (F HR = 0.05) on the central GBR. A later study by Sherman [60] at a single location on the southern GBR reported little differentiation between habitats (F HR = 0.009), but did find different levels of inbreeding between habitats (also observed by Ayre and Dufty [46] ). The study by van Oppen et al. [45] did not specifically assess differences between habitats, but they did report one population in the Ribbon Reefs (which was sampled at a different depth and during a different year compared to the other populations) that was highly divergent from the other Ribbon Reef populations, leading them to suggest that this was either a reflection of temporal variability or was driven by habitat. A similar pattern of differentiation was found for a population sampled on the exposed side of Davies Reef [45] . In this study, we reconfirm the effect of habitat on genetic differentiation, first detected by Ayre and Dufty [46] , but also demonstrate that the extent of differentiation between adjacent habitats can entail fixed differences. Significant F IS and genetic structuring within populations are commonly observed among corals and have previously been attributed to local inbreeding or Wahlund effects [44] . High levels of inbreeding were only detected in the ‘Upper Slope’ habitat at Yonge Reef, possibly due to the lower densities of Seriatopora colonies in the ‘Upper Slope’ habitat (Bongaerts et al. unpublished data). In contrast to most previous studies [44 and references therein], allele frequencies in all other populations approached expectations under HWE. Although local mating would lead to inbreeding, it would not create the replicated associations (at two different reefs) of genotypes and habitats that we observed. By sampling within distinct habitats we seem to have avoided any sign of a Wahlund effect in our data, which may have affected previous studies if distinct ecotypes were present in sample locations. The exception in our study is the ‘Deep Slope’ habitat that does contain multiple distinct genetic groups ( Figures 2 , 4 ). Depth zonation of Symbiodinium \n The observed partitioning of Symbiodinium types across habitats ( Figure 2 ) matches numerous reports on symbiont zonation over depth (e.g. [47] – [49] , [62] , [64] ) and could reflect adaptation to depth-related environmental conditions such as low-light conditions [65] – [67] . Although some overlap existed, the common shallow symbiont, Symbiodinium C120, was rarely encountered in the ‘Deep Slope’ habitat and neither of the deep symbiont types, C3n-t or C3-ff, were found in the shallow habitats. This zonation of symbionts in S. hystrix differs from results on the southern GBR, where Symbiodinium C3n-t was found to occur in colonies from 3 to 18 meters depth [48] . The differences in depth range of Symbiodinium C3n-t between these studies (southern GBR, depth generalist; northern GBR, deep specialist) may reflect latitudinal variation in surface irradiance and light attenuation (with lower irradiance levels recorded on the southern GBR [68] ). However, various abiotic factors other than light change with increasing depth (e.g. spectral quality, temperature, nutrient availability), and as with the host, it is therefore difficult to assess the individual contribution of each factor to the observed partitioning of symbionts [49] . Alternatively, different host-symbiont associations may predominate at different latitudes on the GBR. Codiversification in the coral-algal symbiosis The coral-algal symbiosis has received much attention over the past decade (reviewed in [50] , [69] ), and host-symbiont specificity and stability are tightly linked to the ability of corals to respond to environmental change (i.e. the ability to change symbiotic partners as a mechanism to cope with change). Whereas many studies have focused on the genetic identity of the symbiont, this study is one of the few to evaluate host-symbiont specificity using molecular markers for both symbiotic partners [70] – [72] . The most striking finding was the habitat partitioning of linked symbiotic partners, which suggests adaptation of the holobiont (host plus symbiont) to distinct environmental niches and/or linkage disequilibrium on a genomic level. The ‘Upper Slope’ and ‘Back Reef’ host mtDNA genotypes (‘HostU’ and ‘HostB’) were found in symbiosis with two closely related shallow symbionts types, C120 and C120a, as well as the rare ‘Back Reef’ symbiont C1m-aa. The two common host genotypes associated with the ‘Deep Slope’ habitat occurred with Symbiodinium types C3n-t and C3-ff ( Figure 2 ), with C3-ff occurring exclusively in individuals with the ‘HostD2’ genotype. The observed correlation reinforces that high levels of specificity occur, even among closely related host species [48] , [49] , [54] , [73] and potentially at an intra-specific level ( Figures 2 , 3 ). As such, our data underlines the potential importance of co-speciation processes in the diversification of both symbiotic partners, and this may be particularly important in corals with a vertical symbiont acquisition mode such as the brooding coral S. hystrix \n [52] , [54] , [73] , [74] . It is noteworthy that in the few instances where holobiont genotypes seem ‘misplaced’ with regards to habitat, the host-symbiont genotype associations were maintained with reference to each other. For example, individuals sampled in the ‘Deep Slope’ habitat with the common ‘Upper Slope’ host genotype ’HostU’ (mtDNA) also contained the shallow symbiont C120 instead of any of the deep symbionts ( Figures 2 , 4 ). These colonies may therefore be occurring near the lower depth limit of the ‘shallow’ population, and the ‘Deep Slope’ habitat may be encompassing a contact zone with mixed environmental conditions [13] that marks a transition from ‘shallow’ to ‘deep’ haplotypes. The observation of holobiont-habitat ‘mismatches’ reinforces the status of the mtDNA haplotypes as distinct host lineages. These lineages probably represent ecotypes or potentially incipient/cryptic species that differ in their depth distribution and symbiont types, thus phenotypic plasticity alone is unlikely to be the only mechanism by which S. hystrix can thrive under a broad range of environmental conditions. Ecological speciation Ecological speciation describes a process of diversification that can occur in the absence of extrinsic barriers, and has therefore been proposed as an alternative to allopatric speciation in the tropical marine realm [13] . In many instances, ecological speciation is driven by divergent selection between environments and eventually results in habitat partitioning between closely related lineages. However, as divergent selection between environments is equally consistent with allopatric speciation [75] , it is important to identify the geographic context in which speciation has occurred. Coyne and Orr [76] argue that divergence in sympatry must be demonstrated through a present-day sympatric distribution of the most closely related sister species and an ecological setting in which allopatric differentiation is unlikely. Yet, excluding any scenario of historical allopatry is impossible for most taxa [76] , so that the most convincing cases of sympatric speciation have been limited to unique isolated terrestrial and freshwater settings, such as a crater lake [77] and a remote oceanic island [78] . In the coral reef environment, ecological speciation has been suggested in a few instances, where the general expectation of genetic partitioning according to habitat rather than biogeographical barriers was met. For example, Rocha et al. [13] observed strong genetic differentiation of several congeneric species of tropical reef fish (genus Halichoeres ) across habitats, but not geographic locations. Similarly, Duran and Rützler [14] found partitioning of mtDNA haplotypes of a Caribbean marine sponge (genus Chondrilla ) across mangrove and reef habitats, but not geographically distant locations. On a more local scale, Carlon and Budd [16] identified distinct depth distributions of Favia fragum morpho-types across three sites (up to ∼2 km apart) in the Bocas del Toro region (Panama), consistent with a ‘divergence with gene flow’ model [16] . In a similar fashion, we observe strong genetic segregation of the coral S. hystrix across environmentally distinct habitats, but not between the same habitats at different locations (∼20 km apart; Figure 2 ). As gene flow does not seem to be limited by physical barriers, the observed partitioning of S. hystrix in this study supports the notion of reduced gene flow through divergent selection between distinct reef habitats. Due to the limited geographic range and small number of sampled reefs, however, it is unclear whether the observed patterns of genetic differentiation are part of a broad-scale pattern in S. hystrix . Two of our mtDNA haplotypes match published S. hystrix sequences from other localities in the Indo-Pacific (Okinawa, New Caledonia, Taiwan) [40] , [41] ( Figure 3 ), suggesting a widespread occurrence of these lineages. Furthermore, as sampling was performed in discrete habitats rather than over a bathymetric gradient, it is unclear whether the observed partitioning reflects a step function (i.e. microallopatry) or distributions with zones of overlap. Further studies across a bathymetric gradient covering a broad geographic range could provide insights into the specifics and geographical context of the observed partitioning. Additionally, future studies should test whether genetic segregation is maintained through mainly pre- or post-settlement processes (e.g., reproductive isolation or selection against ecotypes that settle in the ‘wrong’ habitat). Morphological features were not characterized, but gross morphology was observed to vary between habitats, similar to descriptions of ecotypes described by Veron and Pichot [55] . Colonies of S. hystrix in the ‘Back Reef’ and especially the ‘Upper Slope’ habitats seemed to have thicker branches (perhaps related to the greater extent of wave action in these habitats) in comparison to ‘Deep Slope’ individuals. Additionally, colonies in the ‘Upper Slope’ habitat were more compacted with shorter and more frequently dividing branches. Previous work by Flot et al. [34] in Okinawa, New Caledonia and the Philippines showed little congruence between mitochondrial sequences and morphological species delimitations, however they focused on genetic variability between various Seriatopora spp. ( S. hystrix , S. caliendrum , S. aculeata , S. guttatus, and S. stellata ; the latter three are not reported for the GBR) and specifically report distinct genetic lineages within S. hystrix. Although the taxonomic status of the observed mtDNA lineages will need to be resolved in future molecular and morphological studies (in order to assess whether they represent intra-specific diversity, subspecies (ecotypes), or cryptic species [40] , [41] ), at present, incipient ecological speciation seems to provide the most parsimonious explanation for the strong association of closely-related, sympatrically-occurring host lineages with habitat. Conclusions Even though genetic variability between habitats has been previously demonstrated [46] , this study clearly indicates that habitats within a reef can be genetically isolated from each other, whereas the same habitat types separated by up to ∼20 km can exhibit high levels of genetic similarity. Furthermore, the observed genetic partitioning demonstrates that the cryptic diversity previously detected in S. hystrix \n [40] , [41] may be a reflection of lineages associated with distinct reef environments. Habitat-associated cryptic diversity may explain some of the “stochastic” results and high levels of genetic structuring over short geographic distances commonly observed in genetic studies of scleractinian corals. This study highlights the need to further explore genetic diversity over environmental gradients in other coral species, preferably encompassing species with a variety of life history strategies and broad ecological distributions. This is particularly important in the context of local reef connectivity and the general conception that deeper sections of reefs [18] , [79] may act as a reproductive source for shallow reef areas following disturbance [23] . The strong association of host and symbiont genotypes with particular reef environments presents a compelling case for ecological speciation, corroborating previous evidence [13] – [15] that ecologically-based divergent selection may be an important mechanism for diversification on coral reefs. Overall, it underscores the need for understanding processes that shape diversity, which will allow for more accurate predictions on the persistence and community structure of coral reefs in a future of increasing anthropogenic and climate pressures."
} | 7,064 |
31381557 | PMC6681937 | pmc | 5,904 | {
"abstract": "Allocation of goods is a key feature in defining the connection between the individual and the collective scale in any society. Both the process by which goods are to be distributed, and the resulting allocation to the members of the society may affect the success of the population as a whole. One of the most striking natural examples of a highly successful cooperative society is the ant colony which often acts as a single superorganism. In particular, each individual within the ant colony has a “communal stomach” which is used to store and share food with the other colony members by mouth to mouth feeding. Sharing food between communal stomachs allows the colony as a whole to get its food requirements and, more so, allows each individual within the colony to reach its nutritional intake target. The vast majority of colony members do not forage independently but obtain their food through secondary interactions in which food is exchanged between individuals. The global effect of this exchange is not well understood. To gain better understanding into this process we used fluorescence imaging to measure how food from a single external source is distributed and mixed within a Camponotus sanctus ant colony. Using entropic measures to quantify food-blending, we show that while collected food flows into all parts of the colony it mixes only partly. We show that mixing is controlled by the ants’ interaction rule which implies that only a fraction of the maximal potential is actually transferred. This rule leads to a robust blending process: i.e ., neither the exact food volume that is transferred, nor the interaction schedule are essential to generate the global outcome. Finally, we show how the ants’ interaction rules may optimize a trade-off between fast dissemination and efficient mixing. Our results regarding the distribution of a single food source provide a baseline for future studies on distributed regulation of multiple food sources in social insect colonies.",
"introduction": "Introduction Food sharing in social insects is a compelling example of cooperation within a population [ 1 – 7 ]. Ants and bees can store a considerable amount of liquids in a pre-digestion storage organ called the ‘crop’ [ 8 – 10 ]. The stored food can later be regurgitated and passed on to others by mouth-to-mouth feeding (oral trophallaxis) [ 10 – 12 ]. Trophallaxis is a principal mechanism of food-transfer between individuals and therefore, the crop is often referred to as a “social stomach” [ 8 ]. When food is exchanged through trophallaxis, it is stored within the crop of the recipient workers and mixed with the rest of food in the crop [ 13 – 17 ]. Food blending is therefore an important factor in any process mediated by trophallaxis: from nutrient transfer and the maintenance of gestalt odor to hormonal regulation and information sharing [ 8 , 13 , 18 , 19 ]. The extent to which food is blended in the colony has only been partially addressed before [ 3 , 14 , 20 – 22 ] and is still an open question. Food blending is especially interesting in light of the fact that most colony members do not leave the nest [ 5 , 14 , 16 , 23 , 24 ], and all food is brought in by a a small fraction of workers called the foragers [ 16 , 25 ]. The inter-relations between food-supplies brought in by different foragers can be expected to have an important role in the nutritional regulation of the colony. Social insect colonies have a documented ability to tightly regulate both the global nutritional intake [ 15 , 21 ] and the dissemination of food to various sub-populations (such as nurses, larvae and brood) which may have different nutritional needs [ 5 , 14 , 16 , 23 , 26 , 27 ]. The mechanisms that underlie this regulation are, however, not fully understood [ 28 ]. Trophallactic food exchange requires physical contact between ants. The dissemination process is therefore conveniently described by a time ordered network, in which ants are the nodes and the food transfers are the (directed) edges. The topology of this network provides the underlying infrastructure of the food-sharing process [ 17 , 29 – 31 ]. In the study of social insects and other real-world networks, the topology of the network can frequently be traced while the details of particular interactions are concealed [ 32 , 33 ]. Indeed, previous studies that traced individuals in a colony have mainly focused either on the network topology [ 29 , 31 ] or on coarse grained descriptions of food dissemination [ 1 , 16 , 22 , 26 , 27 , 34 ]. In this study we use single ant identification and fluorescently-labeled food ( Fig 1 ) to measure not only the interaction network but also the flow of food over this network. For technical reasons, these experiments are conducted with a single food source. Characterization of this basic case is a first but necessary step towards more complex scenarios which include multiple sources. 10.1371/journal.pcbi.1006925.g001 Fig 1 Quantifying food distribution within an ant colony by combining single ant tracking with fluorescent imaging. a) Two tagged workers engaged in trophallaxis. The identity of ants (orange numbers) was determined using Bugtag barcodes. The volume of food in the ants’ crop is measured using fluorescence imaging and overlaid in red. b-d) Food distribution across the colony and at different stages of the experiment. Markers (round: non-forager, square: forager) overlaid on ants depict their crop contents. Marker size is proportional to the food load held by each ant: P a (small markers were set to a minimal size for clarity). Color division in markers of all ants depicts the computationally derived proportions of food in their crops according to the forager that first collected it (‘food-types’): ( P ( f | A = a )). Scale bars are 1cm. See also supplemental movie “Food dissemination in ant colony”. The flow of food is limited by capacity: As the crop of ants is of finite size, this imposes a constraint on the amount of food that can be transferred in an interaction. This physical constraint limits the rate of mixing as ants become more and more full. Therefore, a potential trade-off between fast rate of food accumulation and well mixed outcome is expected. The main objective of our study is using single ant measurement techniques to quantify how food brought in by different foragers blends as it is being disseminated across an ant colony. To this end, we use Shannon entropy to quantify the quality of mixing in an ant’s crop. The Shannon entropy provides a single quantity that reflects the relative abundances of multiple constituents [ 35 ] and therefore sets a common scale by which food homogenization can be evaluated from our empirical data. Using our detailed measurements we characterize the interaction network and the rules by which food flows across this network. We then use hybrid simulations to identify which of these characteristics function as regulators of food mixing, and which might play a lesser role. Finally, we employ a theoretical model to study the trade-offs between food dissemination and nutritional homogenization.",
"discussion": "Discussion It is well known that social insects manage their nutrient resources on the collective level and also on finer scales because the colony channels foods with different nutritional composition to different sub-populations. In this paper, we put forward the idea that this intricate regulation relates to the interplay between food dissemination and food mixing within the colony. High levels of dissemination are important as they ensure that any food type is available to any ant. On the other hand, high dissemination induces mixing and this reduces the required variety of nutritional choices within the colony. A main finding of this work is that, despite repeated trophallactic interactions between the ants, food in the colony does not become evenly mixed. Quantifying mixing using entropy measures we showed that, compared to what was theoretically possible, mixing is slow to rise and levels up at around 80% of the full mixing potential. The logarithms in the definition of entropy make the significance of this number difficult to assess. For intuition, in the case of only two food sources, the maximal mixing entropy (1 bit) corresponds to each crop holding equal parts of the food sources (1: 1) while 80% of this (0.8 bits) corresponds to, a far from perfect, 3: 1 partition of food sources. This imperfect mixing offers the possibility for receiving ants to choose from a wide spectrum of nutritional compositions when the donors provide different blends. Such choices can allow ants within the nest to reach their nutritional target using feeding schemes similar to those described by the geometrical framework for food foraging [ 42 ]. We further explored the mechanisms that allow for intermediate levels of food blending. Using hybrid simulations, we found that the interaction network over which food flows does not pose any limits on mixing levels. Rather, it is the interaction rule employed by the ants that regulates the extent to which food blends. This is reminiscent of several examples in which cellular pathways with identical architecture can achieve starkly different regulatory behaviors depending on actual rate coefficients [ 43 , 44 ]. Regulation by interaction rules rather than by meeting patterns is an intriguing possibility for social insects in which different collective functions often reside over very similar interaction networks [ 29 ]. For example, while proximity is required for both food sharing and disease transmission [ 45 ] different interaction rules may ensure that one of these is enhanced while the other is suppressed. Quantifying a large number of trophallactic interactions, we directly measured the food-transfer rule (see also [ 36 ]) used by the ants. We stress several important aspects of this rule. First, the rule respects the physical limits on crop size of the ants. Broadly speaking, this limit along with the fact that ants receive a substantial fraction of their free crop space per each interaction imply that an ant may become relatively full following her first few interactions. Thus, an ant’s mixing entropy is, to a large extent, determined by a small number of large events. Since these events are random both in order and in volume it is likely that mixing entropies will not saturate their maximal upper limit (see SI, ‘Entropy by largest events’, S5 Fig ). Second, we show that the interaction rule is most likely stochastic in nature and, therefore, does not entail any strong requirements on ant cognition or communication. Finally, the fact that in trophallactic interactions the recipients fill only partially ( Fig 3 ) is in agreement with a model in which, similar to animals foraging in the environment, ants in the nest regulate their nutritional income by feeding off of multiple partners each with a different mixture of the available ‘food types’. We explored the interplay between food dissemination and mixing using a simple model of food flow that is based on our empirical observations. We find that the intermediate levels of mixing, as measured, can viewed as a compromise between the requirements to quickly unload incoming food and the requirement to disseminate different food types to all parts of the colony. We show that this process is robust over a wide range of δ values and that the actual measured parameter ensures that all ants in the colony are equally well mixed (although each holds a different particular mixture). Finally, we wish to highlight the limitations of this study. Due to current technological availability, this work was performed using a single food source labeled by a single dye. The ants may behave differently in terms of both interaction network and food-transfer-rule when several food sources with different nutritional values are available [ 4 ]. For example, ants may modulate the amount of food they receive in a trophallactic interaction according to its nutritional value. Such modulation, which can be captured in an extension of our current model, can allow the ants to differentially regulate the flow of different nutritional types across the colony. Further, our artificial setup contained a single chamber nest. More realistic, multi-chambered, nest structure may induce interaction networks that are more clustered than the one measured here. This may hold important consequences for nutrition dissemination. Last, is our choice to measure mixing by labeling food types by foragers. While arbitrary, this is a reasonable choice since, as we have shown, foragers are responsible for a large part of the mixing ( Fig 4b ). Taking all these limitations into account we view our findings as a baseline to which future results, where multiple food sources are provided and tracked may be compared to. Overall, our finding that the interaction rule takes precedence over the interaction schedule manifests both the robustness of collective processes within the ant colony and the large extent to which individual behaviors may modulate global outcomes."
} | 3,289 |
32785287 | PMC7423320 | pmc | 5,905 | {
"abstract": "The microbial oxidation of metal sulfides plays a major role in the formation of acid rock drainage (ARD). We aimed to broadly characterize the ARD at Ely Brook, which drains the Ely Copper Mine Superfund site in Vermont, USA, using metagenomics and metatranscriptomics to assess the metabolic potential and seasonal ecological roles of microorganisms in water and sediment. Using Centrifuge against the NCBI “nt” database, ~25% of reads in sediment and water samples were classified as acid-tolerant Proteobacteria (61 ± 4%) belonging to the genera Pseudomonas (2.6–3.3%), Bradyrhizobium (1.7–4.1%), and Streptomyces (2.9–5.0%). Numerous genes (12%) were differentially expressed between seasons and played significant roles in iron, sulfur, carbon, and nitrogen cycling. The most abundant RNA transcript encoded the multidrug resistance protein Stp, and most expressed KEGG-annotated transcripts were involved in amino acid metabolism. Biosynthetic gene clusters involved in secondary metabolism (BGCs, 449) as well as metal- (133) and antibiotic-resistance (8501) genes were identified across the entire dataset. Several antibiotic and metal resistance genes were colocalized and coexpressed with putative BGCs, providing insight into the protective roles of the molecules BGCs produce. Our study shows that ecological stimuli, such as metal concentrations and seasonal variations, can drive ARD taxa to produce novel bioactive metabolites.",
"conclusion": "Conclusions The present study is the first seasonal characterization of a metagenome and metatranscriptome at the Ely Copper Mine Superfund site, providing insight into the microbial community as well as the genes and metabolites they use to adapt to ARD in Ely Brook. Acid-tolerant Proteobacteria were the dominant annotated taxa, varying with season and sample type. Several RNA transcripts were differentially abundant between seasons and the most abundant transcript was involved in antibiotic resistance. KO analysis of Prokka-annotated ORFs identified several differentially expressed genes involved in iron and sulfur, nitrogen, and carbon metabolism, providing insight into seasonal gene function, as taxonomy could not be assigned to many ORFs using our pipeline. Genes involved in metal resistance and secondary metabolism were also annotated and differentially expressed. Future work will involve using additional taxonomic classifiers to assign taxonomy to functionally annotated transcripts as well as experimentally validating the differential expression of selected genes. Importantly, several resistance genes (metal and antibiotic) colocalized with BGCs and, in some instances, were coexpressed, revealing putative antibiotic-producing BGCs and their ecological roles that can be exploited for bioremediation or pharmacological purposes.",
"introduction": "Introduction During the 19 th and 20 th centuries, the mining industry exploited Vermont’s copper belt in Orange County ( Fig 1 ), after which several copper mines were abandoned and left to accumulate acid rock drainage (ARD) [ 1 ]. ARD is the outflow of acidic water from mining regions containing metal-sulfide-rich rocks. When metal sulfides are exposed to water and oxygen, hydronium and sulfate ions are produced, lowering the pH of the water. Toxic levels of Cu, Fe, Zn, and Pb, leaching from pyrrhotite-rich, Besshi-type sulfide deposits [ 2 ] have adversely affected the water quality and aquatic biodiversity in the copper belt [ 3 ]. This process is further accelerated by the presence of acidophilic, sulfur and/or iron-oxidizing bacteria, which quickly convert insoluble sulfides to soluble sulfate ions and Fe 2+ to Fe 3+ , the predominant, soluble form of iron at acidic pH. Due to metals and acidic waters contaminating local streams, mines in this region have been placed on the Superfund National Priorities List by the Environmental Protection Agency (EPA). 10.1371/journal.pone.0237599.g001 Fig 1 Vermont copper belt. Map of Elizabeth Mine, Ely Copper Mine, and Pike Hill Copper Mine (represented by stars) [ 4 ]. Circles represent nearby towns and the Ely Brook (EB-90M) study site is indicated by a red arrow. The pictures on the right show the EB-90M study site on July 28 th , 2017 and January 14 th , 2018. Microorganisms in metal-contaminated environments evolve unique genes conferring resistance to heavy metals [ 5 , 6 ] and/or antibiotics [ 7 – 9 ] to maintain cellular homeostasis. Metal resistance genes (MRGs), some of which are antibiotic-resistant based on having similar mechanisms of action [ 8 ], can induce the biosynthesis of secondary metabolites to scavenge metals. For example, Cupriavidus metallidurans , originally isolated from industrial sludge [ 10 ], is both heavy metal- and antibiotic-resistant and expresses biosynthetic gene clusters (BGCs) involved in the production of a variety of secondary metabolites, including Fe 3+ -binding staphyloferrin B [ 11 , 12 ]. Several bioactive microbial natural products have been isolated from mining environments [ 13 ], such as the berkeleylactones, potent fungal antibiotics isolated from the copper-rich Berkeley Pit in Butte, MT [ 14 ]. MRGs and antibiotic resistance genes have also been identified within biosynthetic gene clusters (BGCs) dedicated to secondary metabolism [ 15 ]. Thus, the coclustering of these resistance genes can be used to bioprospect metal-polluted environments for novel secondary metabolites and understand the stressors that trigger their production, providing insight into their bioactivity. To assess the potential of ARD to produce bioactive secondary metabolites, we characterized the water and sediment associated with ARD at the Superfund site Ely Copper Mine. Both water and sediment at this mine have been affected by ARD (pH > 3) and high metal concentrations (e.g., up to 1,560 μg/L Cu) [ 3 ]. In 2010, dissolved Cu concentrations in the water and sediment at Ely Copper Mine exceeded the aquatic health criteria by 45–222 and 7–40 times, respectively [ 3 ]. Thus, we sampled water and sediment at the Ely Brook, a confluence of clean water and upstream tributaries that drain the mine [ 3 ], and used shotgun metagenomics and metatranscriptomics to characterize the ARD microbiome, including community structure and diversity, and the genes involved in secondary metabolism as well as heavy metal and antibiotic resistance. Acid rock drainage sites rich in copper tailings are generally inhabited by acidophilic iron- and sulfur-oxidizing microorganisms, such as species of Leptospirillum , Acidithiobacillus , Acidiphilium , and Thiobacillus [ 16 – 18 ]. Based on the high acidity and metal concentrations in Ely Brook [ 3 ], we hypothesized that similar species would dominate in Ely Brook. In this study, we aimed to 1) describe the acidophilic, iron- and sulfur-oxidizing chemolithoautotrophs and heterotrophs that likely dominate the water and sediment at Ely Brook and 2) link the microbiome to actively expressed genes, especially those involved in metal transport and the production of bioactive secondary metabolites in this metal-rich extreme environment. Samples were collected in summer and winter to identify seasonal differences that inform community dynamics as well as how environmental stimuli affect gene expression. This work represents the first metagenomic and metatranscriptomic study of an acid rock microbiome within the Vermont copper belt.",
"discussion": "Results and discussion Physicochemical characterization The physicochemical properties of all samples varied between seasons ( S1 Table ). The water temperature was -0.36°C in January (winter) and 16.4°C in July (summer), with a pH of 3.86 and 3.59, respectively. The sediment pH was within the pH range of water, but more acidic in winter (pH 3.56) than summer (pH 3.78), possibly due to how the sediment accumulated protons [ 43 ], reducing their dissociation rates. While there was variable pH among samples from different seasons, more data has to be collected to evaluate the significance of this difference. High redox potentials (423–451 mV) were measured, indicating that EB-90M water was oxidized (aerobic environments have redox potentials ≥-100 mV; [ 44 ]). Water sulfate levels (95–126 mg/L) were within EPA-recommended concentrations (<250 mg/L) and consistent with former Ely Brook geological studies [ 3 ] but less than those reported in other ARD studies [ 45 ]. Most nutrients, including nitrate and nitrite (<0.02 mg/L), total Kjeldahl nitrogen (<0.7 mg/L), and reactive and total phosphorus (<0.15 mg/L) in water were below the detection limit. Low levels of total and dissolved organic carbon (1.4–3.1 mg/L) were also detected in water, which is characteristic of ARD due to competition between species and the inability of the environment to retain nutrients [ 46 , 47 ]. High metal concentrations were detected in all EB-90M samples. The most abundant elements in water were Mg, Al, and Fe (3.07–5.89 mg/L; Table 1 ), and the amounts of total and dissolved elements were the same across water samples. Silica (SiO 2 , 49%), Fe 2 O 3 (27%), and Al 2 O 3 (13%) were the major components of sediment ( Table 2 ), which is also supported by the high levels of Si, Al, and Fe detected by ICAP-MS ( Table 1 ). The weight percent of Fe and Al has increased by 17% and 4%, respectively, since the EPA last analyzed the geochemical properties of EB-90M sediment in 2006 [ 4 , 48 ], underscoring the long-term detrimental effects of ARD on un-remediated sites. 10.1371/journal.pone.0237599.t001 Table 1 Chemical composition of samples. Sample Na Mg Al Cr Mn Fe Co Ni Cu Zn As Cd Ba Pb Sb July water (D) 1.73 (0.04) 4.17 (0.08) 4.93 (0.12) <0.01 0.404 (0.003) 5.22 (0.15) 0.0882 (0.0022) 0.0243 (0.0002) 1.85 (0.01) 0.369 (0.006) <0.01 <0.01 0.0170 (0.0002) <0.01 <0.01 July water (T) 1.76 (0.02) 4.24 (0.08) 5.03 (0.05) <0.01 0.403 (0.004) 5.62 (0.06) 0.0918 (0.0010) 0.0253 (0.0003) 1.87 (0.02) 0.360 (0.003) <0.01 <0.01 0.0172 (0.0002) <0.01 <0.01 January water (D) 1.33 (0.01) 3.19 (0.02) 4.31 (0.25) <0.01 0.286 (0.001) 5.80 (0.01) 0.101 (0.001) 0.0230 (0.0002) 2.27 (0.01) 0.321 (0.004) <0.01 <0.01 0.0119 (0.0002) <0.01 <0.01 January water (T) 1.31 (0.02) 3.07 (0.06) 4.33 (0.08) <0.01 0.278 (0.003) 5.89 (0.08) 0.0982 (0.0014) 0.0223 (0.0004) 2.22 (0.00) 0.305 (0.004) <0.01 <0.01 0.0113 (0.0001) <0.01 <0.01 July sediment 14.3 × 10 3 * 11.3 × 10 3 * 66.1 × 10 3 * 137 (1) 852* 194 × 10 3 * 20.2 (0.2) 28.5 (0.3) 2.21 × 10 3 (26) 323 (5) 2.99 (0.21) 0.195 (0.027) 471 (1) 54.6 (0.2) 4.23 (0.09) January sediment 17.4 × 10 3 * 10.7 × 10 3 * 64.3 × 10 3 * 118 (3) 1.04 × 10 3 * 181 × 10 3 * 15.2 (0. 2) 25.6 (0.6) 1.90 × 10 3 (32) 609 (10) 2.44 (0.38) 0.211 (0.045) 806 (4) 52.9 (0.3) 6.49 (0.65) Selected elemental analysis for water and sediment at EB-90M collected in July 2017 and January 2018 in mg/L and mg/kg, respectively. (D) and (T) represent dissolved and total elements in water samples, respectively. Values with an asterisk (*) were determined by X-ray fluorescence, a more accurate method for the analysis of selected elements in sediment. All other values were determined in triplicate by ICAP-MS. 10.1371/journal.pone.0237599.t002 Table 2 Chemical composition of EB-90M sediment in weight percentages. Chemical composition, wt % (mg/kg) Summer sediment Winter sediment Na 2 O 1.88 (14.3 × 10 3 Na) 2.11 (17.4 × 10 3 Na) MgO 1.98 (11.3 × 10 3 Mg) 1.85 (10.7 × 10 3 Mg) Al 2 O 3 12.9 (66.1 × 10 3 Al) 12.2 (64.3 × 10 3 Al) SiO 2 48.0 (230 × 10 3 Si) 51.7 (241 × 10 3 Si) P 2 O 5 0.156 (697 P) 0.168 (748 P) K 2 O 2.17 (17.5 × 10 3 K) 2.25 (18.7 × 10 3 K) CaO 1.56 (122 × 10 3 Ca) 1.93 (137 × 10 3 Ca) TiO 2 0.763 (4.68 × 10 3 Ti) 0.803 (4.87 × 10 3 Ti) MnO 0.0870 (852 Mn) 0.132 (1.04 × 10 3 Mn) Fe 2 O 3 28.4 (194 × 10 3 Fe) 26.0 (181 × 10 3 Fe) Values in the parentheses represent concentrations of selected elements in units of mg/kg. Microbial community diversity and composition Approximately 25% of paired-end reads from 11 out of 16 samples ( S2 Table ) were taxonomically annotated via Centrifuge and 141 phyla [ 42 ], including candidate phyla, were detected across all domains. Data from winter water samples were excluded, as there was considerably lower sequencing coverage due to low DNA yields. All taxonomic annotation data from Centrifuge are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 1). Of the annotated taxa, Bacteria dominated the entire EB-90M community followed by Eukaryota and Archaea ( Fig 2A ), respectively, and viruses were also detected. Proteobacteria (50 ± 4%) was the most dominant phylum followed by Actinobacteria (19 ± 4%), Chordata (7.6 ± 0.2%), unclassified sequences (19 ± 2%), and Streptophyta (3.4 ± 1%). When only considering microorganisms, Proteobacteria represented 61 ± 4% of the community followed by 23 ± 3% Actinobacteria ( Fig 2B ). Proteobacteria commonly dominate ARD [ 49 , 50 ] due to their metabolic plasticity [ 51 ] and they include iron and sulfur oxidizers that grow under metal-rich and less-restrictive pH conditions. Similarly, Actinobacteria have been reported in other ARD environments, including 90 microbial communities in a copper tailing impoundment in Anhui Province, China [ 50 , 52 , 53 ]. Both Proteobacteria and Actinobacteria thrive in diverse sediments and have evolved mechanisms to inhabit metal-rich environments [ 54 ]. 10.1371/journal.pone.0237599.g002 Fig 2 Taxonomic annotation. Seasonal profile of A) all phyla and B) microbial phyla across all samples at EB-90M. Taxa are annotated by superkingdom followed by phylum. “Other” represents phyla that are less than 2% and 1% percent of the data for all organisms and microorganisms, respectively. Shannon diversity indices (H) of taxa were similar for species of bacteria (H = 6.8–7.0), archaea (H = 4.0–4.9), and fungi (H = 5.4–5.5) regardless of sample type ( S3 – S5 Tables) and indicated greater bacterial diversity. The beta diversity of bacteria, archaea, and eukaryota in samples was significant between summer water and sediment (p < 0.05). However, there was no significant dissimilarity between sediment from different seasons ( S6 – S8 Tables). Non-metric multidimensional scaling (NMDS) plots and principal component analyses (PCAs) of DNA in water and sediment also showed no clear difference at the genus level between taxa in winter and summer sediment [ 42 ]. Season explained 66–92% of the dissimilarity of species between sediment, but the p-value was 0.1 at an alpha level of 0.05 ( S6 – S8 Tables). Thus, more samples need to be evaluated to confirm this dissimilarity, as season can impact ARD taxonomic diversity due to changes in temperature, pH, and metal concentrations [ 55 – 57 ]. Nevertheless, the beta diversity of organisms in summer sediment and water differed at all taxonomic levels (p < 0.05), with high variation (70–87%) based on sample type ( S9 – S11 Tables). Pseudomonas (2.6–3.3%), Bradyrhizobium (1.7–4.1%), and Streptomyces (2.9–5.0%) were the most abundantly annotated microbial genera in all samples ( Fig 3 ). Although not consistent with our hypothesis, acid-tolerant bacteria from these genera have been isolated from other mines [ 17 , 58 – 64 ], producing nutrients and mediating the flux of metal ions [ 65 , 66 ]. Species of Leptospirillum , Acidithiobacillus , Acidiphilium , and Thiobacillus were also present in the metagenome but at significantly lower abundance (<0.7%). Considering not all paired-end reads were taxonomically annotated, other sulfur and iron oxidizers may dominate this site. 10.1371/journal.pone.0237599.g003 Fig 3 Seasonal microbial diversity. Relative abundances of the most abundant genera of archaea (top twelve), microbial eukaryota (top eleven), and viroid/viruses (top eleven) as well as the 24 most abundant genera for bacteria per sample. The most abundant taxa per sample varied within each grouping (archaea, eukaryota, viroid/viruses, and bacteria). Red and blue represent high and low abundance, respectively. Incertae sedis corresponds to a taxonomic group with unknown broader relationships to other taxa. Most microorganisms were differentially abundant between seasons. For example, of the 1177 annotated bacterial genera, 660 were differentially represented between seasons (p-season < 0.05). Data for the differential analyses of annotated bacteria, archaea, and fungi are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 2). Bradyrhizobium , nitrogen-fixing plant endophytes, were slightly more abundant in the summer regardless of sample type (p-season < 0.05). Archaea were also present but less abundant, as Euryarchaeota represented 1.3 ± 0.03% of the observed microbial phyla ( Fig 2B ). Euryarchaeota have also been identified in low abundance in the Baiyin open-pit copper mine in China [ 18 ]. The most abundant archaeal genus was ‘ Candidatus Nitrosotalea’, an acidophile involved in nitrification, oxidizing ammonia to nitrite in acidic sediment [ 67 ]. This genus was also more prevalent in the summer regardless of sample type ( Fig 3 ). Of the 93 annotated archaeal genera, 55 were differentially represented between seasons (p-season < 0.05). Chordata was the most abundant eukaryotic phylum ( Fig 2A ) in the EB-90M metagenome. Within the eukaryotic microbial community, the fungi Ascomycota followed by Basidiomycota were the most abundant eukaryotic phyla ( Fig 2B ). The following genera were the most represented: Aspergillus , Rhodotorula , and Colletotrichum , with Rhototorula being slightly more abundant in the summer (p-season < 0.05) ( Fig 3 ). Ascomycota have been found at mining sites [ 68 ], particularly in biofilms at the Richmond Mine at Iron Mountain [ 69 – 71 ], where fungal hyphae provide a surface for symbionts to attach to pyrite sediment [ 72 , 73 ]. Furthermore, several phyla were algae that inhabit ARD or metal-rich environments (i.e., Bacillariophyta [ 74 ], Xanthophyceae [ 75 ], and Euglenida [ 76 ]). However, all subsequent analyses focused on prokaryotes and fungi, the largest population of unicellular eukaryotes in this dataset. Of 174 fungal genera identified, 89 were differentially represented between seasons (p-season < 0.05). Seasonal metabolic and functional activities of taxa In addition to seasonal variations among taxa, significant seasonal differences in gene expression were observed at EB-90M based on the relative abundance of predicted ORFs. PCA demonstrated uniform but distinct molecular phenotypes across sample type (DNA) and season (ORFs) [ 42 ]. Active gene expression in sediment was quantified by comparing the abundance of Prokka-annotated ORFs ( S12 Table ) to that of the DNA used to assemble the metagenome ( S13 Table ). All differentially expressed, functionally annotated ORF data are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 3). Approximately 104,772 out of 296,476 genes were significantly differentially expressed based on a p-interaction ≤ 0.05. Many predicted ORFs (35,037) had a p-interaction value and an FDR-corrected p-value (q-winter/summer RNA-value) ≤ 0.05, indicating several genes were differentially expressed between seasons, which is consistent with other ARD studies in which physicochemical properties were found to impact gene expression [ 77 ]. The predicted functions and relative abundance of RNA transcripts provided insight into the roles of taxa, as many contigs did not align to NCBI “nt” sequences in our assembly-based taxonomic annotation pipeline. In the future, additional taxonomic classifiers should be used to increase the taxonomic annotations of functionally annotated genes. Most differentially abundant transcripts encoded hypothetical proteins. Table 3 lists the top ten differentially expressed annotated genes between seasons as well as their producing taxa. These genes were involved in amino acid and cofactor metabolism, protein synthesis, transport, virulence [ 78 ], cell wall homeostasis/organization, nucleotide, carbohydrate, and lipid metabolism, cell signaling, and transcription, which are important for survival and have been reported in ARD [ 45 , 49 , 79 ]. Interestingly, comEC , a gene involved in horizontal gene transfer (HGT), was highly expressed in winter. HGT or DNA uptake via cellular membranes is involved in the evolution and adaptation of species, which is largely driven by environment and community composition [ 80 ] and likely plays a significant role in adaptation to this harsh environment. Furthermore, most genes in Table 3 were expressed by species of Bradyrhizobium , Streptomyces , Aromatoleum , Methylococcus , and ‘ Candidatus Solibacter’, which are common to polluted environments [ 81 – 84 ]. The most abundant gene in the entire dataset, especially in winter, was stp , encoding a spectinomycin tetracycline efflux pump belonging to the major facilitator superfamily [ 85 ], which Acidimicrobium has been reported to express in wastewater [ 86 ]. Efflux pumps aid in cellular detoxification and homeostasis, and their expression can be triggered by heavy metal ions, which are abundant at EB-90M (Tables 1 and 2 ) [ 87 , 88 ]. 10.1371/journal.pone.0237599.t003 Table 3 Seasonal gene expression. Protein function, Gene Biological process Organism Highly expressed in Winter Malto-oligosyltrehalose trehalohydrolase, treZ Trehalose/glycan biosynthesis, virulence Aromatoleum aromaticum tRNA-2-methylthio- N (6)-dimethylallyladenosine synthase, miaB tRNA methylation Macrolide export protein, macA Antibiotic resistance Methylococcus capsulatus Multidrug resistance protein, stp Regulation of EF-tu, virulence, cell wall synthesis, and multidrug resistance Acidimicrobium ferrooxidans dITP/XTP pyrophosphatase, rdgB Purine nucleoside catabolism Toxin, fitB Virulence, stress response Enamidase, Ena Cofactor catabolism Bradyrhizobium sp. S23321 ComE operon protein, comEC Competence for transformation Putative sugar transferase, epsL Cell wall organization Sensor histidine kinase, tmoS Two component regulatory system Underexpressed in Winter Leucine-, isoleucine-, valine-, threonine-, and alanine-binding protein, braC Amino acid transport 6-Oxocyclohex-1-ene-1-carbonyl-CoA hydrolase, bamA Benzoyl CoA catabolism Putative phosphoserine phosphatase 2, pspB Serine biosynthesis Coenzyme PQQ synthesis protein, pqqD Cofactor biosynthesis UDP-N-acetyl-D-glucosamine 6-dehydrogenase, wbpA Cell wall organization Ferric uptake regulation protein, fur Transcription regulation Streptomyces avermitilis Glycerophosphodiester phosphodiesterase, glpQ Glycerol and lipid metabolism ABC transporter permease, ytrF ABC transporter involved in acetoin utilization 3-Dehydroquinate dehydratase, aroQ Aromatic amino acid biosynthesis Succinate-acetate/proton symporter, satP Acetate-uptake transporter Candidatus Solibacter usitatus Top ten Prokka-annotated differentially expressed genes in winter sediment paired with taxonomic annotations. Data were sorted by a p-interaction value ≤ 0.05 followed by q-winter/summer RNA value ≤ 0.05, respectively. All RNA transcripts were at least 8 to 20-fold higher or lower compared to DNA in samples. No taxon was annotated if a contig did not align to sequences in “nt”. Functional analysis of metagenome and metatranscriptome KEGG annotated 1,048,574 protein-coding reads with KOs, providing insight into the ecological and metabolic roles of active taxa in sediment. All KEGG annotation data are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 4). While the function of most ORFs were unknown, 442,447 annotated ORFs were assigned to 6,997 KOs, which were then assigned to KEGG pathways, BRITE hierarchies, and modules. KEGG reference pathway maps and BRITE reference hierarchies are applicable to any organism by functional orthologs being defined by K numbers, which can be used to reconstruct pathways from explicitly incomplete datasets [ 89 ]. The most abundant RNA transcripts were involved in BRITE hierarchies and metabolism [ 42 ]. Four hundred and fifteen metabolic pathways were identified and mainly involved in carbohydrate metabolism, energy metabolism, and amino acid metabolism, similar to Prokka-identified ORFs in EB-90M sediment as well as those in other ARD sites [ 45 , 90 ]. Out of 6,997 KOs, representing annotated ORFs, 2,532 were differentially expressed between seasons based on a p-value related to the significance of season (p-season) < 0.05. In the winter, some of the most differentially expressed KEGG pathways were related to protein digestion and absorption as well as phenazine biosynthesis [ 42 ]. While KOs can belong to several pathways, modules, and BRITE hierarchies, most differentially expressed KOs were related to protein families: signaling and cellular processes (BRITE Level 2, 09183). Most transcript levels decreased in winter/increased in summer, possibly due to summer temperatures increasing the metabolic demands of the community, as shown in model ARD biofilms grown at different temperatures [ 91 ]. Several pathways, including those involved in sulfur ( Fig 4 ), nitrogen, and carbon metabolism (see FigShare 5–6; https://doi.org/10.6084/m9.figshare.c.4864863 ), were differentially expressed (p-season < 0.05) [ 42 ]. Experimental investigation of these orthologous functions and individual genes is required to confirm the level of gene expression with respect to season. 10.1371/journal.pone.0237599.g004 Fig 4 Sulfur metabolism gene expression. Sulfur metabolism KEGG reference pathway map diagram ( https://www.kegg.jp/pathway/map00920 ) with color gradation highlighting KEGG-annotated gene expression that changes between seasons. Blue and red colors denote decreased and increased abundance of RNA transcripts in the winter, respectively. Genes that did not change are light gray and undetected genes are white. Significantly differentially expressed genes are indicated by a star and met the following criteria: p-interaction ≤ 0.05 in combination with a q-winter/summer RNA ≤ 0.05, respectively. Actively expressed genes involved in iron and sulfur cycling While common iron and sulfur oxidizers did not dominate EB-90M, several ORFs from EB-90M sediment were involved in iron and sulfur cycling. The expression data for differentially expressed KOs involved in iron and sulfur metabolism are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 7–8). KOs associated with key iron oxidation and reduction genes, such as cox and mtr genes, were identified. Unique ORFs (584) mapped to KOs encoding coxABC genes (K02274, K02275, and K02276) involved in Fe 2+ oxidation were slightly more expressed in winter. These data are consistent with the presence of Fe 3+ reducers, such as species of Granulicella , in sediment. While a total of 483 transcripts were involved in reducing Fe 3+ to Fe 2+ , specifically mtrCAB (K03585, K07670, and K07654), the KOs were not differentially expressed. However, Prokka-annotated mtrCAB ORFs were differentially expressed in the summer. Some acidophiles reduce iron under suboxic conditions [ 92 ], are facultative anaerobes (e.g., Bradyrhizobium ), or inhabit anoxic microzones in sediment where biotic Fe 3+ reduction occurs. Several organisms were involved in sulfur cycling, a process involving the movement of sulfur between rocks, water, and organisms. Fig 4 shows the differentially expressed KOs associated with the sulfur metabolism KEGG pathway. KOs (24 out of 104) were differentially expressed in the summer, specifically those involved in dissimilatory sulfate reduction and oxidation pathways, such as aprBA (68 ORFs mapped to K00394 and K00395). Organisms expressing these genes could be used for bioremediation by removing excess toxic sulfate from the environment via this reversible pathway [ 93 ]. Dissimilatory pathways generate energy and either produce sulfides anaerobically or sulfate aerobically, whereas assimilatory pathways can reduce inorganic sulfate to sulfide to synthesize sulfur-containing amino acids and metabolites in the presence of oxygen. KOs involved in assimilatory sulfate reduction, such as cysH (184 ORFs mapped to K00390), were differentially expressed in the winter. Seventy-five sox genes (i.e. soxAB ; K00301, K00302, and K00303) were involved in sulfur oxidation and differentially expressed in the summer. sox genes oxidize thiosulfate, a product of metal sulfide dissolution, to a sulfate intermediate to generate energy or reduce carbon. Thus, the sox gene may be expressed in species that thrive in warmer temperatures, as other variables, such as pH and metal levels, were similar between seasons. Altogether, these KO data demonstrate that while common iron and sulfur oxidizers did not dominate EB-90M, organisms are actively involved in iron and sulfur cycling. Actively expressed genes involved in nitrogen metabolism The abundance of transcripts involved in nitrogen metabolism is consistent with nitrogen-fixing Bradyrhizobium being the most abundant microbial taxa detected at EB-90M. The expression data for all KOs involved in nitrogen metabolism are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 7–8). Species of Bradyrhizobium fix nitrogen in root nodules for plant growth, especially in acidic [ 94 ] and pyrite-rich [ 63 ] soil. In summer sediment, species of Bradyrhizobium differentially expressed the nitrogenase gene nifH (K02588; 28 ORFs), a biomarker for nitrogen fixation, the conversion of molecular nitrogen to ammonia [ 42 ]. Given that a subset of RNA transcripts annotated as nifH lacked taxonomic annotation, nifH may also be expressed by plants detected in our dataset, which require more nitrogen for growth at this time of year [ 95 ]. Ammonia can also be made from or assimilated via the reduction of nitrate. Assimilatory genes, such as narB /1.7.7.2 (K00367) and NR (K010534) (52 ORFs in total), which convert nitrate to nitrite, were differentially expressed [ 42 ]. Transcripts encoding narB (K00367) were differentially abundant in winter, whereas those encoding NR (K010534) were differentially abundant in summer. Thus, selected microorganisms along with other plants likely regulate nitrate levels in sediment between seasons. This differential expression may be related to the abundance of nitrogen-fixing bacteria and the abundance of transcripts encoding nifH , resulting in the production of excess ammonia in the summer. Actively expressed genes involved in carbon metabolism in photosynthetic organisms Carbon fixation can occur via the Calvin-Benson-Bassham cycle in plants, algae, and phyla of bacteria, such as Cyanobacteria, Chlorobi, Proteobacteria, Firmicutes, Acidobacteria, and Chloroflexi. Some of these bacterial phyla were dominant at EB-90M and reduced the expression of key genes in this cycle, such as rbcL and rbcS ( RuBisCo , 4.1.1.39 [ 42 ]) as well as prkB (2.7.1.19 [ 42 ]), in winter likely due to less sunlight [ 96 ]. The expression data for all KOs involved in carbon fixation are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 7–8). In the summer, there were significantly more transcripts of RuBisCo (144 ORFs) and prkB (51 ORFs), encoding enzymes that fix carbon dioxide to ribulose-1,5-bisphosphate to form 3-phosphoglycerate and phosphorylate ribulose-5-phosphate [ 97 ], respectively. Other carbon fixation pathways were also represented in EB-90M ARD. For example, organisms expressed reductive tricarboxylic acid genes belonging to the C4-dicarboxylic acid pathway [ 42 ], which converts carbon dioxide to acetyl-CoA. ORFs (271 of mdh , 1.1.1.82) encoding malate dehydrogenase, an enzyme that converts oxaloacetate to malate in this pathway [ 42 ], were expressed in the summer to convert oxaloacetate into glucose for energy. Methanogens, prokaryotes that reduce carbon dioxide to produce methane, were also identified based on the expression of the methanogenic genes mcrA [ 98 ] (five ORFs; K07451) and mcrB (11 ORFs; K07452) in both seasons. Secondary metabolic pathways A total of 1589 BGCs were annotated by antiSMASH 5.0 [ 39 ], but only 449 met the filtering criteria [ 42 ]. The antiSMASH annotation and associated expression data are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 9–10). Most BGCs were involved in the biosynthesis of nonribosomal peptides (33%) followed by terpenes (27%) ( Fig 5 ). Since Actinomycetes are present in this dataset, particularly Streptomyces , which are prolific producers of secondary metabolites [ 99 ], we expected to find BGCs dedicated to secondary metabolism. However, most BGCs were found in contigs without taxonomic annotation. A subset of BGCs is identical to those involved in producing carotenoids, anabaenopeptin NZ 857/nostamide A, rhizomide A–C (cytotoxic [ 100 ]), xenotetrapeptide, n -acyl alanine, alkyl resorcinol, 1-heptadecene, bicornutin, patellazole (cytotoxic [ 101 ]), micromonolactam, geosmin, and phomopsin (tubulin polymerization inhibitor [ 102 ]). These data are consistent with our hypothesis that EB-90M can be a source of bioactive metabolites, as there are BGCs involved in producing bioactive compounds. Notably, there is the potential to find more, as many BGC products are unknown. 10.1371/journal.pone.0237599.g005 Fig 5 Percentage of genes dedicated to producing various classes of secondary metabolites. Pi chart of antiSMASH-annotated BGCs involved in secondary metabolite production in all EB-90M DNA samples (sediment and water). Classes that represent <2% of 449 BGCs are classified as “Other”. Within the metatranscriptomic datasets, 65 out of 449 transcripts encoding genes within BGCs were differentially abundant in summer (39 genes) than winter (26 genes) based on p-interaction and q-winter/summer RNA values < 0.05 ( Fig 6 ). The expression of the phytoene synthase gene, crtB , in the summer increased in some organisms and decreased in other organisms ( S14 Table ), suggesting that selected microorganisms may alter their metabolism based on the presence of other taxa and ecological factors that require specific metabolites. Overall, there were more annotated terpenes and nonribosomal peptides produced in the summer. 10.1371/journal.pone.0237599.g006 Fig 6 Seasonal expression of terpene BGCs terpene- and NRP-annotated BGCs. Gradient plot demonstrating the expression of BGCs in winter (light blue) and summer (orange) sediment samples. NRP, nonribosomal peptide; PK, polyketide. All data met the following criteria: p-interaction ≤ 0.05 followed by q-winter/summer RNA ≤ 0.05. Metal resistance genes (MRGs) Of the 285 experimentally validated MRGs in the BacMet database, 133 were identified in the metagenome, representing 7,021 unique ORFs out of 161,984. All BacMet annotation and expression data are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 11). These data are consistent with our hypothesis, as microorganisms conferring metal resistance are commonly found in ARD for the sequestration or chemical conversion of toxic heavy metals. Eight of the 133 BacMet genes were not expressed [ 42 ] and 719 of the 7,021 ORFs were differentially expressed in the summer [ 42 ]. Mostly transcripts without taxonomic annotation and Proteobacteria of the genus Burkholderia conferred heavy metal resistance to Cu, Cd, Co, Zn, Fe, Ag, Pb, Hg, As, Sb, and Ni. Proteobacteria have expressed MRGs in acidic environments [ 103 – 106 ] and been explored for the bioremediation of metal-contaminated environments via efflux pumps, bioabsorption, and transforming metals into less toxic forms [ 107 ]. Some of the most differentially expressed MRGs were involved in Cu ( dnaK ) [ 104 ], Cu/Te ( actP ) [ 108 ], Cd/Co/Zn/Cu ( czcA/B ; actP) [ 105 , 109 ], Cu/Zn/As ( pstA ) [ 110 , 111 ], Mn/Zn/Fe/Cd/Co ( mntH ) [ 112 ], and Zn ( zraR ) [ 113 ] resistance. Genes related to As- and Sb-containing compound ( pgpA; acr3 ) resistance were differentially expressed, even though Sb and As were nearly undetectable ( Table 2 ). Based on the large number of genes without functional annotation, more proteins with unique mechanisms and metal-binding properties likely exist at EB-90M. Resistance mechanisms and secondary metabolism Metal-rich environments select for MRGs and also co-select for antibiotic resistance genes based on having similar genetic mechanisms [ 8 ]. These resistance genes can be used to bioprospect metal-polluted environments for new chemistry and bioactivity. Strategies for prioritizing antibiotic-producing BGCs are based on finding BGCs containing duplicated essential genes, resistance genes, or genetic evidence for HGT [ 114 ]. Additionally, there are MRGs encoding proteins (e.g., resistance-nodulation-division family transporters, such as CzcA) that catalyze the efflux of antibiotics and chemotherapeutics [ 115 ], and a subset also function as antibiotic resistance genes that can be exploited to find new bioactive compounds [ 114 ]. Using the ARTS web server [ 41 ], a platform that prioritizes antibiotic-producing BGCs based on these hypotheses, a range of essential (6358–8289) and duplicated (5595–7395) genes, as well as known resistance models (8501), mostly resistance to biotin-lipoyl domains (1585 out of 8501), were annotated [ 42 ]. The ARTs annotation data are available at https://doi.org/10.6084/m9.figshare.c.4864863 (FigShare 12). Some duplicated genes colocalized with BGCs, mostly (5–12 out of 258–325) RNA polymerase sigma factor 70 and trigger factor, demonstrating the potential of the ARD microbiome to be a source of antibiotics. MRGs can colocalize with BGCs and play a role in antibiotic resistance [ 116 ]. We identified six BGCs that colocalized with BacMet-annotated MRGs on contigs ( Fig 7 ) [ 42 ]. Fig 7 shows the colocalization and coexpression of phosphate transport gene pitA with a homoserine lactone-nonribosomal peptide-annotated BGC in summer ( Fig 7 ). PitA transports phosphate with other cations, such as toxic metal ions [ 117 ], and may be differentially expressed in the summer in response to excess phosphate or metal. Excess metal ions can trigger the production of secondary metabolites, such as metal-binding homoserine lactones [ 100 ] or nonribosomal peptides called siderophores [ 118 ]. The Cu- and Zn-expressed antibiotic resistance gene mdtA , encoding a membrane fusion protein of the multidrug efflux complex MdtABC, also colocalized and coexpressed with the polyketide synthase gene, ppsE [ 42 ]. Both genes were differentially expressed in the summer; however, both MRG and BGC did not meet q-winter/summer RNA ≤ 0.05 for, but they met a p-winter/summer RNA ≤ 0.05 [ 42 ]. A subset of ARTS-annotated resistance genes were also MRGs (i.e., metallopeptidases, HflB [ 119 ] and RseP [ 120 ]) that colocalized with BGCs. Thus, colocalization and coexpression of MRGs and BGCs could be used to prioritize antibiotic-producing BGCs (e.g., ppsE ) and find new regulatory mechanisms of secondary metabolism. 10.1371/journal.pone.0237599.g007 Fig 7 Colocalization and coexpression of metal resistance and secondary metabolite genes. Gradient plot of the differential coexpression of pitA , an MRG encoding a phosphate-uptake transport protein, with genes annotated to be involved in the biosynthesis of a secondary metabolite, homoserine lactone-nonribosomal peptide in contig 4689 (20406 nucleotides long), in summer (orange) and winter (blue). All data met the following criteria: p-interaction ≤ 0.05 in combination with a q-winter/summer RNA ≤ 0.05, respectively. Nucleotide positions in contig are shown in parentheses."
} | 9,951 |
29459733 | PMC5876133 | pmc | 5,906 | {
"abstract": "Bacteria in polymicrobial habitats contend with a persistent barrage of competitors, often under rapidly changing environmental conditions 1 . The direct antagonism of competitor cells is thus an important bacterial survival strategy 2 . Towards this end, many bacterial species employ an arsenal of antimicrobial effectors with multiple activities; however, the benefits conferred by the simultaneous deployment of diverse toxins are unknown. Here we show that the multiple effectors delivered to competitor bacteria by the type VI secretion system (T6SS) of Pseudomonas aeruginosa display conditional efficacy and act synergistically. One of these effectors, Tse4, is most active in high salinity environments and synergizes with effectors that degrade the cell wall or inactivate intracellular electron carriers. We find Tse4 synergizes with these disparate mechanisms by forming pores that disrupt the ΔΨ component of the proton motive force. Our results provide evidence that the concomitant delivery of a cocktail of effectors serves as a bet-hedging strategy to promote bacterial competitiveness in the face of unpredictable and variable environmental conditions."
} | 293 |
36770342 | PMC9919625 | pmc | 5,908 | {
"abstract": "Due to its several economic and ecological consequences, biofouling is a widely recognized concern in the marine sector. The search for non-biocide-release antifouling coatings has been on the rise, with carbon-nanocoated surfaces showing promising activity. This work aimed to study the impact of pristine graphene nanoplatelets (GNP) on biofilm development through the representative marine bacteria Cobetia marina and to investigate the antibacterial mechanisms of action of this material. For this purpose, a flow cytometric analysis was performed and a GNP/polydimethylsiloxane (PDMS) surface containing 5 wt% GNP (G5/PDMS) was produced, characterized, and assessed regarding its biofilm mitigation potential over 42 days in controlled hydrodynamic conditions that mimic marine environments. Flow cytometry revealed membrane damage, greater metabolic activity, and endogenous reactive oxygen species (ROS) production by C. marina when exposed to GNP 5% ( w / v ) for 24 h. In addition, C. marina biofilms formed on G5/PDMS showed consistently lower cell count and thickness (up to 43% reductions) than PDMS. Biofilm architecture analysis indicated that mature biofilms developed on the graphene-based surface had fewer empty spaces (34% reduction) and reduced biovolume (25% reduction) compared to PDMS. Overall, the GNP-based surface inhibited C. marina biofilm development, showing promising potential as a marine antifouling coating.",
"conclusion": "4. Conclusions In this study, the long-term antifouling performance of GNP-based surfaces for inhibiting C. marina biofilm development was demonstrated. In fact, GNP/PDMS coatings, due to their surface properties, allied with GNP antimicrobial activity impacted not only biofilm composition but also its structure. Biofilms developed on the GNP composite displayed significantly reduced cell count, thickness, biovolume, and porosity in the maturation stage when compared to the control surface (bare PDMS). In addition, the comprehensive analysis of graphene’s mechanisms of action carried out in this study showed that these carbon nanomaterials cause bacterial membrane damage and induce oxidative stress mediated by ROS production, leading to cell inactivation. Overall, these results demonstrated the potential of incorporating GNP in marine paints to mitigate marine biofouling and its negative consequences.",
"introduction": "1. Introduction Marine environments host numerous surfaces, both naturally occurring and manmade, with distinct physical and chemical properties. These submerged surfaces are subject to the rapid accumulation of organisms and organic matter, in a process known as biofouling. Marine biofouling is a widespread concern with severe economic and environmental consequences [ 1 , 2 ]. Its nefarious effects are particularly clear in the naval industry, where the attachment and colonization of ship hulls by fouling organisms contribute to surface deterioration and significantly increase the watercraft’s drag force, leading to higher fuel consumption and greater maintenance costs [ 3 , 4 ]. Furthermore, greater fuel consumption caused by marine biofouling implies increased levels of greenhouse gas emissions [ 5 ]. Biofouling on oceangoing vessels can also promote the introduction of invasive exotic species into non-native environments as well as the contamination of aquaculture facilities, therefore harming global biodiversity and raising public health concerns, respectively [ 6 , 7 , 8 ]. Lastly, the adhesion of fouling organisms on marine surfaces can also interfere with the measurements of underwater sensors, and contribute to the deterioration of submerged infrastructures [ 9 , 10 , 11 ]. Marine biofouling involves a wide variety of organisms [ 12 , 13 ]. According to their dimensions and level of complexity, fouling organisms can be divided into micro- and macrofoulers. Even though biofouling does not follow a fixed order of events, it is generally considered that surface colonization by microfoulers precedes and promotes attachment by macrofoulers, which highlights biofilm formation as a crucial step in the marine biofouling process [ 14 , 15 ]. Among the wide range of organisms that comprise marine biofilms, Cobetia marina (formerly Deleya marina ) has been extensively used in marine biofouling research as a model marine biofilm-forming bacterium [ 16 ]. This Gram-negative, rod-shaped bacterium possesses several features that make it relevant when studying bacteria–surface interactions, such as extracellular polymeric substances (EPS) production and gliding motility, which allow it to promptly colonize surfaces and form stable biofilms [ 17 , 18 , 19 ]. In fact, C. marina is known to produce large quantities of EPS [ 20 , 21 , 22 ]. This is particularly relevant since it has been demonstrated that one of the most prominent challenges in marine biofouling mitigation is linked to microorganisms that are both capable of initiating biofilm development, as well as ensuring its cohesion, by excreting large amounts of EPS [ 23 ]. Thanks to these unique features, C. marina can be considered a representative marine microfouler. Marine biofilm formation is a nanoscale interfacial phenomenon. As such, innovative surface engineering techniques that modify surface attributes at the nanoscale can not only optimize the properties of certain materials [ 24 , 25 ], but also provide them with significant antifouling potential [ 26 , 27 ]. In fact, the incorporation of nanomaterials into marine antifouling paints has been reported to greatly impact a surface’s charge potential, hydrophobicity, and topography, as well as its antibacterial and anticorrosion properties [ 28 , 29 ]. Among these emerging nanoengineered antifouling paints, those containing carbon nanomaterials, such as carbon nanotubes and graphene, have shown promising results [ 30 , 31 , 32 ]. Graphene is one of the strongest and thinnest materials available [ 33 ]. It consists of a single-layer sheet of sp 2 -hybridized carbon atoms with a two-dimensional hexagonal structure. Due to their high specific surface area, electrical conductivity, and thermal stability, graphene-based materials are very appealing for multiple applications [ 34 , 35 ]. Furthermore, graphene is recognized for its antimicrobial and anti-adhesive properties [ 36 , 37 ]. Even though this carbon material’s mechanisms of action are not yet fully understood, it is postulated that the sharp edges of graphene sheets can lead to membrane damage and bacterial cell entrapment. Additionally, graphene is assumed to induce oxidative stress through the formation of reactive oxygen species (ROS), which disrupt microorganisms’ DNA and proteins [ 38 ]. As such, the incorporation of graphene in marine paints can not only improve their mechanical strength and durability, but also provide them with enhanced antifouling attributes [ 32 , 39 ]. The main objective of this study was to assess the effect of a graphene-based coating on biofilm development by C. marina over a long-term in vitro assay performed under hydrodynamic conditions mimicking a real marine setting. Furthermore, the mechanism of action of pristine graphene nanoplatelets (GNP) was investigated.",
"discussion": "3. Results and Discussion In this study, the antifouling performance of a graphene-based coating was evaluated using Cobetia marina , a model biofilm-forming marine bacterium, under conditions mimicking those of a real marine setting. Furthermore, graphene’s mechanisms of action were investigated, to gain a further understanding of its antibacterial and antiadhesive properties. C. marina cells were exposed to 5% GNP ( w / v ) for 24 h in order to characterize the mechanisms of action of GNP. After this period, the cells were stained with PI (a membrane integrity marker), 5-CFDA (a metabolic activity marker), and DCFH-DA (a ROS production indicator) and analyzed by flow cytometry ( Figure 1 ). Figure 1 a,b shows the fluorescence intensity (FI) and the percentage of PI-positive cells of C. marina cells non-treated and treated with GNP 5% and stained with PI. Results indicated that GNP exposure caused membrane damage in about 40% of the cell population, as demonstrated by the percentage of PI(+) cells. Additionally, epifluorescence microscopy and SEM analysis further demonstrated the cell membrane damage caused by GNP exposure ( Figures S3 and S4 ). This finding is consistent with the proposed GNP mechanism of action which postulates that the direct contact between the bacterial membranes and the sharp edges of graphene sheets or wrapping and trapping of bacterial membranes by the nanosheets induce cell membrane damage [ 38 , 51 ]. In addition, the analysis of non-treated C. marina cells versus those treated with GNP 5% and stained with 5-CFDA ( Figure 1 c,d) suggested that cells exposed to GNP displayed higher metabolic activity (approximately 1.5-fold) than non-treated cells (control). The changes in bacteria metabolism can be a consequence of the oxidative stress triggered by GNP exposure [ 67 ]. Bacteria can adapt to unfavorable environmental stresses through the activation of protective mechanisms. The regulation of bacterial stress responses occurs at different cellular levels, leading to changes in gene expression, protein activity, and cellular metabolism [ 68 ]. In parallel, the assessment of endogenous ROS production by staining the cells with DCFH-DA ( Figure 1 e,f) showed that GNP exposure led to ROS production, as demonstrated by the higher fluorescence intensity (3.6-fold) of treated cells compared to the control. It is known that ROS production depends on the physiological state of the cells, specifically on changes in metabolism as a result of stress [ 69 ]. Hence, these results suggest that oxidative stress imposed by GNP occurs through a ROS-dependent pathway, because of highly cumulated intracellular ROS [ 67 ]. This finding is also in accordance with the hypothesized mechanism of action for graphene, which involves ROS generation [ 38 ]. Since laboratory assays commonly used to test the antifouling effectiveness of marine surfaces and the dynamics of biofilm formation are laborious and time-consuming [ 70 ], a range of surfaces: PDMS (control) and GNP/PDMS composites containing different GNP loadings—1 wt% GNP/PDMS (G1/PDMS), 2 wt% GNP/PDMS (G2/PDMS), 3 wt% GNP/PDMS (G3/PDMS), 4 wt% GNP/PDMS (G4/PDMS), and 5 wt% GNP/PDMS (G5/PDMS)—were first screened for their anti-adhesion potential through the method proposed by Faria et al. [ 62 ]. According to this method, initial adhesion assays (7.5 h) can be used to estimate marine biofilm development. Although no statistically significant differences were registered, results indicate that there is a tendency for the G5/PDMS surface (mean value of 1.97 × 10 7 cells·cm −2 ) to induce a higher reduction in the number of adhered cells compared to bare PDMS control (mean value of 4.23 × 10 7 cells·cm −2 ) ( Figure S1 ). Additionally, C. marina showed a MIC value to GNP of 5% ( w / v ). Hence, based on these data, G5/PDMS composite was selected for surface characterization and long-term biofilm formation assays. Early cell adhesion, subsequent biofilm formation, and its proliferation rely heavily on surface properties, namely wettability, topography, and roughness [ 71 , 72 ]. As such, the aforementioned surface characterization methods were employed to analyze the two surfaces selected for the long-term biofilm formation assays: PDMS and G5/PDMS. Following the approach proposed by van Oss [ 55 ], the hydrophobicity of the two surfaces and the tested microorganism was determined. Contact angle measurements ( θ l ) and free energy of interaction (∆ G ) values are presented in Table 1 . Both surfaces displayed contact angles with water higher than 90° ( Table 1 and Figure S2 ), as well as negative free energy of interaction values (∆ G ), suggesting that they are hydrophobic. Furthermore, since the free energy of interaction values determined for each surface are very close (−63.1 mJ·m −2 for PDMS and −67.5 mJ·m −2 for G5/PDMS), it can be inferred that they display a similar degree of hydrophobicity. According to the literature, the model bacterium used in these biofilm formation assays, C. marina , shows a preference for adhesion on hydrophobic surfaces [ 17 , 20 ]. As such, biofilm formation can be expected on both surfaces. C. marina ’s contact angle value with water ( θ W = 37.4° ± 1.7°; θ W < 90°) and free energy of interaction (∆ G = 14.9 mJ·m −2 , ∆ G > 0) indicate that this microorganism is hydrophilic, which is confirmed by the literature [ 20 ]. In order to predict the extent of C. marina ’s adhesion to the tested surfaces, the free energy of adhesion (∆ G Adh ) values between the microorganism and each surface were determined ( Table 1 ). The results indicate that, in theory, the adhesion of C. marina to both tested surfaces is thermodynamically favorable (∆ G Adh < 0). To further assess the surfaces’ properties and how they might impact biofilm formation, surface topography and roughness were determined by AFM in tapping mode ( Figure 2 ). The analysis of surface topography revealed that G5/PDMS displayed an average roughness ( R a ) value about 100 times higher than that of bare PDMS. Surface characterization was complemented by SEM analysis ( Figure 3 ). This microscopic technique allows the evaluation of the morphological details of surfaces at nanometer resolution [ 73 ]. PDMS stood out as more homogeneous and smoother than the G5/PDMS composite, which displayed a rough appearance, with the presence of uniformly dispersed graphene agglomerates forming small elevations on the surface of the coating ( Figure 3 b). These results corroborate the AFM analysis, as well as those reported by Oliveira et al. [ 51 ]. As a result of the van der Waals forces and strong π–π interactions between individual GNP sheets, the dispersion of GNP is often challenging, leading to the formation of graphene agglomerates [ 51 , 74 ]. In these clusters, the carbon material is more exposed, promoting its direct contact with bacteria, which, in turn, potentiates graphene’s antimicrobial activity [ 68 , 75 ]. Following surface characterization, C. marina biofilm formation on PDMS and G5/PDMS was evaluated at 185 rpm (average shear rate of 40 s −1 ) for 42 days through the analysis of the number of biofilm cells and biofilm thickness ( Figure 4 ). As expected, considering the results of the thermodynamic analysis, C. marina showed biofilm development on both surfaces. Overall, the results obtained for these two parameters are in accordance. Both biofilm cells and biofilm thickness displayed an increasing trend for each surface throughout the total incubation period. Concerning total biofilm cells, this increase was particularly noticeable between incubation days 21 (on average, 2.46 × 10 7 cells·cm −2 for PDMS and 1.80 × 10 7 cells·cm −2 for G5/PDMS) and 28 (on average, 1.11 × 10 8 cells·cm −2 for PDMS and 8.20 × 10 7 cells·cm −2 for G5/PDMS). As for biofilm thickness, the most noticeable growth spurt was between incubation days 28 (on average, 46 µm for PDMS and 34 µm for G5/PDMS) and 35 (on average, 113 µm for PDMS and 70 µm for G5/PDMS). These increments may indicate that C. marina biofilm maturation occurred between the 3- and 5-week marks. More importantly, from incubation days 14 to 42, G5/PDMS showed consistently lower biofilm cell count and thickness values than bare PDMS (approximately 22% reduction on incubation day 14; 26% reduction on incubation day 28; 38% reduction on incubation day 35; 43% reduction on incubation day 42 for both analyzed parameters; p < 0.05). For the last four incubation weeks, the G5/PDMS reduction percentages were similar for both tested parameters, indicating that the GNP surface was not only able to effectively reduce the total cells but also the thickness of biofilms formed by C. marina , which is of extreme importance, given the role of biofilm control on marine biofouling. Altogether, these results indicate that the graphene-based polymeric coating showed significant antibacterial and antibiofilm performance. This might be a result of the aforementioned GNP agglomerates, which trigger membrane damage and oxidative stress in C. marina cells, as demonstrated by the flow cytometric analysis. In fact, GNP clusters are likely to promote the membrane-piercing effect of exposed graphene particles, consequently affecting cell integrity [ 37 , 67 ]. It is hypothesized that the first layer of adhered cells is particularly affected by this dart-like effect, impairing the subsequent adhesion of cell layers, and, therefore, curbing long-term biofilm formation. Overall, these results corroborate the previously reported antibacterial and antifouling properties of graphene materials [ 32 , 51 , 75 , 76 ]. Figure 5 includes representative 3D images retrieved through OCT from biofilms formed by C. marina on PDMS ( Figure 5 a,b) and G5/PDMS ( Figure 5 c,d) at incubation days 21 ( Figure 5 a,c) and 42 ( Figure 5 b,d). More than the evolution of their thickness between the two incubation days, which is in accordance with the quantitative biofilm thickness results ( Figure 5 b), these representative 3D images allow us to observe the spatial distribution of the biofilm structures across the tested surfaces. Moreover, the biovolume of biofilms formed on both surfaces was calculated for incubation day 42 ( Figure 6 a). This parameter provides an estimate of the biomass in the biofilm (µm 3 ) per area of the region of interest (mm 2 ) [ 31 ]. Results showed that the biofilms formed on G5/PDMS had significantly lower biovolume than those developed on the control surface (on average, 1.80 × 10 8 µm 3 ·mm −2 for PDMS versus 8.89 × 10 7 µm 3 ·mm −2 for G5/PDMS; p ≤ 0.01). These results corroborate the ones obtained for both biofilm total cells and biofilm thickness ( Figure 4 ). Biofilm porosity values were also determined for both surfaces on incubation day 42 ( Figure 6 b). Results showed significantly lower biofilm porosity on G5/PDMS in comparison to the control surface (on average, 61% porosity on PDMS versus 27% on G5/PDMS; p ≤ 0.001). Figure 6 c,d consists of representative 2D cross-sectional images acquired through OCT of biofilms developed on PDMS ( Figure 6 c) and G5/PDMS ( Figure 6 d) on incubation day 42. The comparison of these two images allows us to visually assess the differences in structure and porosity between biofilms formed on each tested surface. It is possible to observe that biofilms formed on G5/PDMS possess fewer empty spaces than those formed on bare PDMS, which corroborates the quantitative porosity percentage results and is in accordance with a previous study performed with GNP-based surfaces using cyanobacteria [ 32 ]. In biofilms, empty spaces, such as pores and channels, can be beneficial to ensure that nutrients and oxygen reach the cells, as well as to dilute waste products or antimicrobials [ 77 ]. Hence, biofilms with a higher percentage of empty spaces, such as those developed on the PDMS surfaces, have an advantage in terms of mass transfer and storage, and greater potential to grow when compared with less porous biofilms [ 78 ], such as the ones formed on the G5/PDMS surface. All in all, this work demonstrates the antibiofilm potential of GNP/PDMS surfaces against C. marina biofilm development. The results obtained are promising, especially considering that the incorporation of pristine graphene into polymeric marine antifouling coatings is poorly documented [ 30 ]. Furthermore, the cytometric assessment of the effects of GNP on C. marina cells can contribute to a better understanding of graphene’s antibacterial mechanisms of action."
} | 4,959 |
25575307 | PMC4478705 | pmc | 5,909 | {
"abstract": "Stratified sulfurous lakes are appropriate environments for studying the links between composition and functionality in microbial communities and are potentially modern analogs of anoxic conditions prevailing in the ancient ocean. We explored these aspects in the Lake Banyoles karstic area (NE Spain) through metagenomics and in silico reconstruction of carbon, nitrogen and sulfur metabolic pathways that were tightly coupled through a few bacterial groups. The potential for nitrogen fixation and denitrification was detected in both autotrophs and heterotrophs, with a major role for nitrogen and carbon fixations in Chlorobiaceae . Campylobacterales accounted for a large percentage of denitrification genes, while Gallionellales were putatively involved in denitrification, iron oxidation and carbon fixation and may have a major role in the biogeochemistry of the iron cycle. Bacteroidales were also abundant and showed potential for dissimilatory nitrate reduction to ammonium. The very low abundance of genes for nitrification, the minor presence of anammox genes, the high potential for nitrogen fixation and mineralization and the potential for chemotrophic CO 2 fixation and CO oxidation all provide potential clues on the anoxic zones functioning. We observed higher gene abundance of ammonia-oxidizing bacteria than ammonia-oxidizing archaea that may have a geochemical and evolutionary link related to the dominance of Fe in these environments. Overall, these results offer a more detailed perspective on the microbial ecology of anoxic environments and may help to develop new geochemical proxies to infer biology and chemistry interactions in ancient ecosystems.",
"introduction": "Introduction Linking microbial community composition and ecological processes such as carbon (CO 2 fixation and respiration), nitrogen (nitrification, denitrification and N 2 fixation) and sulfur cycling (sulfur assimilation, anaerobic sulfate respiration and sulfide oxidation) is a primary goal for microbial ecologists. This information is needed to improve our understanding on the structure and functioning of microbial communities, to properly guide experimental research efforts, to promote our ability to understand fundamental mechanisms controlling microbial processes and interactions in situ ( Prosser, 2012 ) and to approach the study of earlier interactions of biosphere–hydrosphere–geosphere ( Severmann and Anbar, 2009 ). However, a detailed comprehension of biological interactions in highly complex systems is however difficult ( Bascompte and Sole, 1995 ). Stratified lakes with euxinic (anoxic and sulfurous) bottom waters are simplified study systems to explore current biodiversity–biogeochemistry interactions because of their high activity, large biomass and low microbial diversity ( Guerrero et al., 1985 ). Usually, oxic–anoxic interfaces contain conspicuous blooms of photosynthetic bacteria that are often macroscopically visible because of the high intracellular content of pigments, and additional microbial populations also tend to accumulate ( Pedrós-Alió and Guerrero, 1993 ). These blooms are, in fact, natural enrichment cultures that facilitate physiological studies in situ ( Van Gemerden et al., 1985 ). At such interfaces fine gradients of physicochemical conditions are present and tight coupling between different biogeochemical cycles (mainly carbon, nitrogen and sulfur) are established. Microbes adapted to such gradients are difficult to culture because in situ conditions are very difficult to mimic in the laboratory, and their study has improved perceptibly by culture-independent methods ( Casamayor et al., 2000 ). Stratified euxinic lake systems may also provide potential modern day analogue ecosystems for the oceans during long periods of Earth history. The planet was essentially anoxic until 2.7–2.4 billion years ago, with a ferruginous ocean ( Anbar, 2008 ; Reinhard et al., 2013 ). With the advent of oxygenic photosynthesis, atmospheric oxygen began to rise, as did the oxygen content in the surface oceans. The deep oceans remained anoxic, but entered a period of temporal and spatial heterogeneity. Strong euxinic conditions might be expected in ancient coastal areas, with merely anoxic conditions in the open ocean, although high Fe deep ocean conditions would have been maintained ( Reinhard et al., 2013 ). In contrast, Fe is low in the deep waters of the modern ocean and, therefore, it is difficult to find appropriate ancient ocean analogue in the current marine realm. With this in mind, stratified aquatic systems with high Fe concentrations in deep waters could be more appropriate modern day analogues of the Proterozoic ocean. Karstic lacustrine systems with a gradient of organic carbon delivery and sulfide concentrations generated by sulfate reduction, and usually rich in iron, would provide reasonable biogeochemical analogues for ancient coastal to open ocean gradients. In this study, we explored the oxic–anoxic interface (metalimnion) and bottom waters (hypolimnion) from two sulfurous lakes in the Banyoles karstic area (NE Spain) through shotgun metagenomics and in silico analysis of several metabolic pathways. In the framework of paleoreconstruction of anoxic conditions in ancient marine systems, one lake would be representative of strong euxinic conditions (Lake Cisó) and the other of low euxinia and an active iron cycle (basin III of Lake Banyoles). We explore the links between microbial composition and functionality for the carbon, nitrogen and sulfur cycling after phylogenetic and functional identification. The taxonomic identity assigned to each functional step was determined by the closest match in databases, and the relative abundance and distribution of marker genes was comparatively analyzed among samples as a proxy of the potential in situ relevance of these pathways under the specific environmental conditions studied. Because of the lack of oxygen, large microbial biomass and high contribution of deep dark fixation processes to overall CO 2 incorporation ( Casamayor et al., 2008 , 2012 ; Casamayor, 2010 ), we hypothesized a high genetic potential for chemotrophic CO 2 fixation and a tight redox coupling between carbon, nitrogen and sulfur biogeochemical cycling. In addition, because of its euxinic nature we also expected a low contribution of both methanogens and ammonia oxidizers in the biogeochemical cycles prevailing in these environments.",
"discussion": "Discussion Stratified planktonic environments with sharp chemical gradients and sulfide-rich bottom waters are valuable current windows on past Earth conditions. Anoxic and euxinic conditions were common but spatially and temporally heterogeneous in ancient oceans during Proterozoic ( Reinhard et al., 2013 ; Lyons et al., 2014 ), and may have played an important role in mass extinctions during Phanerozoic ( Meyer and Kump, 2008 ). The presence of marker pigments for photosynthetic sulfur bacteria (that is, isorenieratene and okenone) have been often reported as evidence of euxinic conditions in ancient oceans ( Damsté and Köster, 1998 ; Brocks et al., 2005 ). These conditions are not common nowadays, although persistent euxinia can be found in deep silled basins such as the Black Sea, Baltic Sea and Cariaco Basin ( Millero, 1991 ; Stewart et al., 2007 ). Future climate change scenarios predict, however, an increasing of euxinia phenomena, mainly in coastal marine ecosystems ( Diaz and Rosenberg, 2008 ). The study of stratified sulfurous lakes has, therefore, an additional interest to predict biogeochemical functioning and microbial interactions in such future scenarios. In the present study, we explored the community composition and functional gene content along a gradient of redox conditions in a karstic sulfurous area. Continental systems are cheaper and easier to sample than marine basins, and a large variety of photo- and chemolithotrophs organisms, sulfide-oxidizing and sulfate-reducing bacteria, fermenters, denitrifying microbes, methanogens and methane oxidizers are expected to be found, among others, according to previous studies (see, for example, Casamayor et al., 2000 ; Barberán and Casamayor, 2011 ). The metabolisms harbored by these microorganisms have the potential to provide insights into the ecosystems operating in euxinic early stages of Earth. The strong euxinic conditions found in Lake Cisó may match biogeochemistry in ancient coastal areas, whereas basin C-III in Lake Banyoles may represent the transition from euxinic coastal areas to merely anoxic and rich Fe conditions in the ancient open ocean ( Figure 7 ). The very low abundance of genes for nitrification, the minor presence of anammox genes, the high potential for nitrogen fixation and mineralization and the potential for chemotrophic CO 2 fixation and CO oxidation all provide potential clues on the ancient oceanic anoxic zones functioning. The low abundance of ammonia oxidizers (AOA and AOB) agrees with the high ammonia accumulation in the anoxic bottom of the lakes, the lack of oxygen and presence of potentially toxic sulfide. We observed, however, a higher gene abundance of AOB relative to AOA in the metagenomic pool that may have a geochemical link related to the abundance of Fe in these environments. AOA have a highly copper-dependent system for ammonia oxidation and electron transport ( Walker et al., 2010 ), completely different from the iron-dependent system present in AOB. The tradeoff in Fe- vs Cu-rich ammonia oxidation enzymatic systems would suggest that AOA evolved relatively recently (<550 million years ago) and that the Proterozoic oceans, which would have been Fe rich, would have been AOB dominated. Interestingly, the evolutionary dynamics of the amoA gene cladogenesis events visualized using lineage through time plots displays a different scenario for AOA and AOB, with AOB showing a more constant cladogenesis through the evolutionary time, whereas AOA experienced two fast diversification events separated by a long steady-state period ( Fernàndez-Guerra and Casamayor, 2012 ). The potential for nitrogen fixation and denitrification was detected in both autotroph and heterotroph microbial lineages, suggesting a diverse range of potential overlaps between carbon and nitrogen cycling in the ancient ocean, and an active nitrogen cycle in anoxic systems. Our results show a potential major contribution to nitrogen fixation by Chlorobiaceae under euxinic conditions. Chlorobiaceae were also the major contributors to carbon fixation in Banyoles C-III coupled to sulfide oxidation through the Arnon cycle. Therefore, the reported presence of Chlorobiaceae in the ancient ocean ( Damsté and Köster, 1998 ; Brocks et al., 2005 ) would have been of major relevance not only for the carbon but also for the nitrogen cycling. Campylobacterales ( Epsilonproteobacteria ) accounted for a large percentage of the denitrification genes in the anaerobic layers of both lakes, but were taxonomically segregated ( Arcobacter dominated in Cisó, Sulfurimonas was present in C-III). Both genera respire nitrate coupled to C fixation in the dark through the reverse tricarboxylic acid cycle ( Labrenz et al., 2005 ; Burgin and Hamilton, 2007 ; Grote et al., 2012 ), being potentially able to couple denitrification to sulfur oxidation ( Ghosh and Dam, 2009 ). The other important group involved in denitrification was the chemolithoautotrophic Gallionellales oxidizing sulfide or Fe 2+ while respiring nitrate and producing NH 4 + or N 2 . The presence of Gallionellales exclusively in C-III is probably because of their close relation with the iron cycle ( Weber et al., 2006 ), and by the fact that an active Fe 2+ cycle has been previously detected in Lake Banyoles ( Garcia-Gil et al., 1990 ). The potential role of Gallionellales in ancient oceans with an active iron cycle is therefore of major interest. The case of Bacteroidales also deserves to be mentioned. Bacteroidales have been typically considered aerobic or microaerophilic chemoorganoheterotrophic bacteria ( Reichenbach, 2006 ), and have been recurrently detected in the Banyoles area ( Casamayor et al., 2000 ; Casamayor et al., 2002 ; Casamayor et al., 2012 ) and in the marine realm ( Fernández-Gómez et al. , 2013 ). However, their role in anaerobic, sulfide-rich layers was not elucidated. Here, we assigned Bacteroidales as potentially catalyzing DNRA (dissimilatory nitrate reduction to ammonium), coupling the electron flow from organic matter to the reduction of nitrate. Thus, we would expect a potential gradient of distribution for anaerobic Bacteroidales in the ancient ocean being more abundant in the organic carbon- and sulfide-rich coastal zones ( Figure 7 ) than in the anoxic and more oligotrophic open ocean. We also noticed the low abundance of key processes in the anaerobic carbon cycle such as CH 4 cycling, probably because in the presence of high levels of sulfate, methanogens are generally poor competitors with sulfate reducers in stratified natural environments ( Raskin et al., 1996 ). Sulfate reduction normally occurs in fully anoxic sediments by SRB ( Holmer and Storkholm, 2001 ). However, as shown here, a water column with euxinic conditions and a high availability of organic carbon is also suitable for the growth of an important community of planktonic SRB. Previous studies in Banyoles area measured unexpected high rates of dark carbon fixation at the oxic–anoxic interface and the hypolimnetic waters, accounting for 58% of total annual fixed carbon in Lake Cisó ( Garcia-Cantizano et al., 2005 ). It was proposed that photosynthetic bacteria could be partly carrying out dark carbon incorporation in situ ( Casamayor et al., 2008 ), and Thiobacilli may actively fix CO 2 at certain depths ( Casamayor, 2010 ). However, the ecological factors modulating the process and the microbial populations performing dark carbon fixation are still not well understood ( Casamayor et al., 2012 ). In the present investigation, we detected the potential for chemotrophic CO 2 fixation mainly through the reverse tricarboxylic acid cycle (K00174, K00175 and K00244 from KEGG Orthology) in Bacteroidales , Campylobacterales and Desulfarculales . In addition, other SRB such as Desulfobacterales may also participate through the anaerobic C 1 -pathway (Wood–Ljungdahl pathway, K00194 and K00197) yielding formate assimilation and CO 2 fixation ( Hugler et al., 2003 ; Sun et al., 2010 ; Fuchs, 2011 ). Interestingly, the diversity of taxa potentially participating in carbon fixation in the dark was larger in Lake Cisó than in C-III, in agreement with in situ measurements carried out in former investigations ( Garcia-Cantizano et al., 2005 ; Casamayor, 2010 ). These findings would suggest an active carbon-fixing activity in ancient euxinic oceans beyond the euphotic zone that certainly deserves further investigations. In addition, the oxidation of CO generates adenosine triphosphate and CO 2 that may be further processed through one of the reductive CO 2 fixation pathways to be used as C source ( Ragsdale, 2004 ; King and Weber, 2007 ). Some studies indicate that organisms using CO as both energy and C source can be viewed as the extant survivors of early metabolic processes ( Huber and Wächtershäuser, 1997 ). In the hypoxic layers we found that the heterotrophic group of Actinomycetales accounted for most of monooxygenase CO genes in agreement with their mixotrophic lifestyle ( Schmidt and Conrad, 1993 ). More interestingly, in the anoxic depths of Lake Cisó we found that CO oxidation genes were mainly related to SRB from Deltaproteobacteria group ( Geobacter and deltaproteobacterium NaphS2) and to Firmicutes ( Carboxidothermus hydrogenoformans, Moorella thermoacetica and Clostridium spp.). This finding suggests that the fate of the reducing equivalents from CO oxidation in anaerobic conditions could be coupled to sulfate reduction (carried out by SRB) to produce sulfide, or to CO 2 reduction to produce acetate (SRB and Firmicutes) ( Roberts et al., 2004 ; King and Weber, 2007 ). To check whether CO oxidation could be coupled to CO 2 reduction to yield acetate ( Ragsdale, 2004 ; Roberts et al., 2004 ), we identified the phylogenetic affiliation of acetyl-CoA synthase genes (ACS, K14138), and found that Desulfobacterales and Firmicutes had the potential to use the Wood–Ljungdahl pathway to obtain energy and fix carbon from CO in the hypolimnion of Lake Cisó. However, although the CO-oxidizing genes were detected, we cannot assess their relevance in the lake or the ancient oceans because CO-oxidizing bacteria carry out a facultative mixotrophic metabolism ( Gadkari et al., 1990 ). Overall, the metagenomics approach unveiled the interrelationships between microbes and biogeochemical cycling in a comparative framework of two lakes that are modern analogs of ancient ocean conditions. These results may also help to develop new geochemical proxies to infer ancient ocean biology and chemistry. A major pitfall in our metagenomic approach is the reliance on the assumption that the genes come from a particular bacteria or archaea according to phylogentic annotation; lateral gene transfer would compromise the direct link of phylogeny to a metabolic pathway. In most of the cases we found the 16S rRNA gene counterpart present in the metagenomic data pool, giving additional support to such links. Obvious next steps include an experimental quantification of the energy and matter fluxes involved in each of the metabolic pathways to get a complete picture on the tight coupling between microbes and biogeochemical cycling in anoxic ecosystems."
} | 4,458 |
38747391 | PMC11808451 | pmc | 5,910 | {
"abstract": "Abstract Arbuscular mycorrhizal (AM) symbiosis, the nutritional partnership between AM fungi and most plant species, is globally ubiquitous and of great ecological and agricultural importance. Studying the processes of AM symbiosis is confounded by its highly spatiotemporally dynamic nature. While microscopy methods exist to probe the spatial side of this plant–fungal interaction, the temporal side remains more challenging, as reliable deep‐tissue time‐lapse imaging requires both symbiotic partners to remain undisturbed over prolonged time periods. Here, we introduce the AMSlide: a noninvasive, high‐resolution, live‐imaging system optimised for AM symbiosis research. We demonstrate the AMSlide's applications in confocal microscopy of mycorrhizal roots, from whole colonisation zones to subcellular structures, over timeframes from minutes to weeks. The AMSlide's versatility for different microscope set‐ups, imaging techniques, and plant and fungal species is also outlined. It is hoped that the AMSlide will be applied in future research to fill in the temporal blanks in our understanding of AM symbiosis, as well as broader root and rhizosphere processes.",
"introduction": "1 INTRODUCTION Arbuscular mycorrhizal (AM) symbiosis is a nutritional mutualism between a group of filamentous soil fungi (Glomeromycotina) and plants. \n 1 \n AM symbiosis occurs on all continents, in all soils, and in nearly all plant species, with roles in nutrient cycling, carbon storage, soil stability and plant health and nutrition. \n 2 \n , \n 3 \n , \n 4 \n , \n 5 \n , \n 6 \n , \n 7 \n Aligned with this global abundance and importance, researchers strive to better understand the molecular, cellular, physiological and ecological aspects of AM functioning. \n 8 \n , \n 9 \n , \n 10 \n \n During AM symbiosis, the fungal partner extends a network of hyphae into the soil while simultaneously colonising plant roots, growing between and within plant root cells. Inside the root cortical cells, AM fungal hyphae form specialised nutrient exchange structures: tree‐like ‘arbuscules’ and/or knot‐like ‘coils’. \n 11 \n Here, the AM fungi receive lipids and sugars from their plant host, necessary to their survival as obligate biotrophs, and reciprocate with water and mineral nutrients gathered from beyond the reach of the plant's own roots. \n 12 \n Later in the symbiosis, fungal storage organs (vesicles) and reproductive structures (spores) develop. Spatially, this partnership is highly heterogeneous: there can be a mosaic of symbiotic structures and developmental stages in a single colonised zone of a root, multiple colonised zones and fungal individuals per root, and networks made up of different root and hypha types. \n 11 \n , \n 13 \n , \n 14 \n , \n 15 \n , \n 16 \n Additionally, there is temporal dynamism: the arbuscules develop and collapse within just a few days, there is constant progression or contraction of hyphal colonisation throughout the roots as well as out into the soil, there can be turnover in fungal symbionts, and all this takes place in the fluid backdrop of a growing root system. \n 16 \n , \n 17 \n , \n 18 \n , \n 19 \n , \n 20 \n \n Microscopy has been a fundamental tool in AM research. Since compound light microscopy aided the discovery of AM fungi at the turn of the 19th century, a multitude of microscopy techniques have been employed to reveal the spatial aspects of AM symbiosis. \n 21 \n , \n 22 \n , \n 23 \n , \n 24 \n To highlight a few applications, brightfield microscopy of stained roots is routinely used to quantify the extent of AM fungal colonisation in plant roots, while fluorescence microscopy coupled with dyes has revealed the morphologies and chemical aspects of symbiotic structures. \n 25 \n , \n 26 \n , \n 27 \n , \n 28 \n , \n 29 \n , \n 30 \n Increasingly, fluorescence‐based microscopy of genetically encoded reporters is used to characterise where AM‐relevant genes are expressed and proteins are localised during the symbiosis. \n 18 \n , \n 31 \n , \n 32 \n , \n 33 \n , \n 34 \n Further, the development of fluorescence‐based biosensors allows distributions of molecules and processes to be mapped. \n 35 \n , \n 36 \n , \n 37 \n , \n 38 \n More recently applied microscopy techniques include X‐ray microscopy for in situ elucidation of fungal networks inside and outside roots, \n 39 \n and spatial transcriptomics, mapping sequenced transcripts onto brightfield micrographs of roots hosting AM fungi. \n 40 \n \n However, the temporal side of AM symbiosis remains challenging. To accurately follow AM processes, plant and fungal partners must remain alive and undisturbed over lengthy timeframes, such as the multiple days of arbuscule lifespan, and weeks of fungal life cycle. \n 17 \n , \n 18 \n But this symbiosis manifests within and around roots, all concealed in opaque soil, poorly compatible with live, light‐based microscopy. Time‐lapses have been performed on excised roots, but they remain limited in duration and reliability due to the substantial root damage. \n 38 \n , \n 41 \n Longer imaging has been achieved via intact roots grown in ‘membrane sandwiches’ or pots, but the disturbance caused by uprooting and transferring them to microscope slides curtails the time‐lapse timeframe to minutes to hours. \n 35 \n In situ imaging negates the need to disrupt plant or fungus, but the bulk of such studies make use of artificial biological systems, such as root organ cultures (in which the plant‐partner lacks a shoot), or do not include a plant host at all. \n 16 \n , \n 42 \n , \n 43 \n , \n 44 \n , \n 45 \n , \n 46 \n , \n 47 \n , \n 48 \n It remains to be seen how representative these systems are of a natural underground environment. To our knowledge, the pioneering works of Kobae et al. \n 18 \n , \n 27 \n , \n 49 \n represent the only examples of long‐term, noninvasive imaging of AM symbiosis within soil‐grown plant roots. Using a glass‐bottomed Petri dish, rice plants expressing arbuscule‐localised fluorescent reporters were cocultivated with AM fungi. Imaging through the glass ‘window’ captured the appearance and collapse of arbuscules and dynamics of lipid droplets. \n 18 \n , \n 27 \n , \n 49 \n Widefield epifluorescence was predominantly used, capturing long‐term colonisation dynamics but with limited subcellular resolution. The few instances employing confocal laser scanning microscopy ran for shorter periods, up to 19 h, with resolution restricted by the available microscope hardware of the time (over a decade ago). \n 27 \n , \n 49 \n Consequently, at the resolution now possible and desirable in AM microscopy, there remains a temporal terra incognita . This technical note presents the AMSlide: a noninvasive, high‐resolution, live‐imaging system optimised for studying AM symbiosis. We demonstrate how AMSlides can be used to:\n reliably monitor mycorrhizal symbiosis in rice roots without impacting plant growth or AM colonisation, obtain a high‐resolution view of roots colonised by AM fungi, from large scale colonisation zones to subcellular domains of the arbuscules, follow AM symbiosis over the timeframes of seconds, hours, and days to weeks, unveiling arbuscule development and collapse dynamics, arbuscule lifespan, and the rate of colonisation progression. \n We also show how, on the technical side, AMSlides:\n are cheap, adaptable, and functional with most confocal set‐ups, open the door to time‐intrinsic imaging techniques such as fluorescence recovery after photobleaching (FRAP), are compatible for studies involving other plant and AM fungal species, can be used in non‐AM applications. \n It is hoped the AMSlides will be adopted to bring the important temporal context to AM research, including AM signalling, arbuscule development, and nutrient exchange. They could facilitate characterisation of symbiotic dynamics with different plant–fungal partnerships or environmental conditions, enabling a more representative view into complex AM symbioses. Further, AMSlides could allow broader root‐ and rhizosphere‐related processes, including root development, immunity and soil community interactions, to be temporally resolved.",
"discussion": "4 DISCUSSION 4.1 An affordable and adaptable live‐imaging system Due to the similar shape and dimensions of the AMSlides to a standard microscope slide, they are compatible with commonly used microscope apparatus. Their versatility in coverslip position, growth substrate, and chamber orientation makes them adaptable to many different modes of microscopy, such as upright or inverted systems, transmission or epifluorescence, confocal or widefield (Figure 1 , File S6 ), unlike the only previous noninvasive systems which were restricted to inverted, epifluorescence‐based hardware. \n 18 \n , \n 49 \n While not tested, AMSlides would also be compatible with multiphoton microscopy for deeper‐root imaging, or spinning disk microscopy for capturing faster dynamics. The 3D‐printable design, easy‐to‐obtain materials, and no requirement for sterile conditions make the AMSlide a simple and affordable system for live‐imaging AM symbiosis. However, the time lapses of nonsymbiotic roots in this work, from the scale of chromatin dynamics to root system development, highlights the applicability of AMSlides beyond AM symbiosis research (Figure 6 ). The inclusion of natural growth substrates, such as sand or soil, may offer a more representative system for studying root development, plant–pathogen interactions \n 59 \n and rhizosphere dynamics \n 60 \n than currently available devices. \n 61 \n , \n 62 \n , \n 63 \n , \n 64 \n \n 4.2 AM symbiosis can be live‐imaged, noninvasively at high resolution The AMSlides enable truly noninvasive live imaging of AM colonisation. The in situ imaging allows more reliable observations than high‐disturbance methods involving uprooting the plants, excising or sectioning. Despite being small, the AMSlides facilitated comparable plant growth and AM colonisation to conventional pot systems and no effects of photo‐toxicity were observed (Figures S1 and S2 ). They likely represent a more ‘natural’ scenario than previously used growth pouches or agar, offering the heterogeneous structure and moisture properties of soil/sand. This is particularly relevant for rice, which on agar rarely develops the large lateral roots that preferentially host AM fungi. \n 65 \n Additionally, using an entire, intact plant is likely more physiologically relevant than root organ cultures, which lack a shoot, or hairy roots, considering the known roles of shoot‐borne signals in AM regulation. \n 66 \n \n The AMSlides achieve better imaging resolution than previously reported noninvasive imaging set‐ups. Subcellular details such as arbuscule domains can be resolved (Figure 2B ), which are intrinsic to understanding arbuscule development, nutrient exchange and colonisation dynamics. The improved resolution is partly because many previous systems were only compatible with widefield fluorescence microscopy, comparatively lower in resolution than confocal laser scanning microscopy. \n 18 \n \n The image resolution achieved with AMSlides is even comparable to standard root excision and slide‐mounting methods (Figure 2B and C ). While the roots of pot‐grown plants grow in three dimensions and therefore rarely sit flat on a microscope slide, the roots of AMSlide plants grow adjacent to the coverslip, reducing the focal distance and required working distance of the objectives. 4.3 AMSlide brings the (spatio‐)temporal context to AM cell biology studies The AMSlides enable AM processes to be monitored over a wide range of timeframes. This is important for understanding AM symbiosis, where it can be similarly desirable to capture fast processes, such as trafficking of proteins to the symbiotic interface or changes in nutrient levels over minutes to hours, as it is to follow the much slower progression of colonisation over days to weeks. The observations made using AMSlides are largely consistent with previous studies, supporting their reliability. For example, the mobility of the SCAMP‐puntae and transvacuolar strands around collapsed arbuscules mirrors time‐lapses made by Kobae and Fujiwara. \n 49 \n The timing of arbuscule development and collapse recorded in AMSlides – commonly 1–2 days to grow, 1–2 days (but sometimes up to 4 days) in cell‐filling state, then rapid collapse followed by 3–5 days to fully disappear – concurs with time‐lapse measurements made by Kobae et al., \n 18 \n , \n 49 \n as well as estimations made by Alexander et al. \n 67 \n , \n 68 \n The measurement of colonisation progression, ∼400 µm/day in the example here, also concurs with earlier observations. \n 18 \n But unlike previous studies, subcellular detail is visible at every timepoint. Consequently, precise developmental stages and fine‐scale processes could be monitored throughout AM colonisation, revealing for example the different temporal dynamics of coils compared to arbuscules of G. margarita (Movie 3 , Figure S5b ), or different eGFP‐SCAMP expression dynamics throughout arbuscule development (Figure 4 ). 4.4 Future applications of the AMSlide The AMSlide was developed and optimised using eGFP‐SCAMP ‐expressing rice as a useful highlighter of symbiotic structures. However, the future applications are multitude. Combining the AMSlide system with genetically encoded reporters for genes relating to symbiotic signalling, intracellular colonisation, nutrient exchange or the ever‐expanding suite of biosensors should allow the temporal context of these processes to be uncovered. Vital stains further broaden the cellular processes that could be monitored. The selection of plants and fungi successfully imaged in the AMSlides include monocot and dicot plant examples, and AM fungi from the Glomerales and Gigasporales , which show distinct growth and development (Figure S5 ). It is therefore likely that AMSlides are compatible with many further plant–AM fungal partnerships, potentially shedding light on different symbiotic strategies, functions and nutritional outcomes. The ability to image root colonisation and external hyphae over time in AMSlides should allow investigation of the relationship between internal and external hyphal activities (Figure 5 ). For example, how do pre‐ and postsymbiotic fungal behaviours, recently revealed via the Obstacle chip and AMF‐SporeChip , \n 16 \n , \n 48 \n relate to within‐root colonisation dynamics? Monitoring AM symbiosis over time could unearth previously hidden differences in AM colonisation dynamics, such as arbuscule development, lifetime and rate of colonisation. How do these parameters change with different biotic situations (e.g. different plant or fungal species and genotypes) or abiotic conditions (e.g. stresses, nutrient levels, light availability)? Using the AMSlides, we can better uncover the temporal ‘black box’ of AM symbiosis, which has been taking place in the roots of most plant species across the globe for at least the past 450 million years. \n 3 \n It is hoped the ability to observe the interaction between plants and AM fungi in a more natural state, nondisruptively, over time, will improve our fundamental understanding of the intrinsically dynamic AM symbiosis."
} | 3,797 |
28401530 | null | s2 | 5,911 | {
"abstract": "Pathway refactoring serves as an invaluable synthetic biology tool for natural product discovery, characterization, and engineering. However, the complicated and laborious molecular biology techniques largely hinder its application in natural product research, especially in a high-throughput manner. Here we report a plug-and-play pathway refactoring workflow for high-throughput, flexible pathway construction, and expression in both Escherichia coli and Saccharomyces cerevisiae. Biosynthetic genes were firstly cloned into pre-assembled helper plasmids with promoters and terminators, resulting in a series of expression cassettes. These expression cassettes were further assembled using Golden Gate reaction to generate fully refactored pathways. The inclusion of spacer plasmids in this system would not only increase the flexibility for refactoring pathways with different number of genes, but also facilitate gene deletion and replacement. As proof of concept, a total of 96 pathways for combinatorial carotenoid biosynthesis were built successfully. This workflow should be generally applicable to different classes of natural products produced by various organisms. Biotechnol. Bioeng. 2017;114: 1847-1854. © 2017 Wiley Periodicals, Inc."
} | 311 |
34863018 | PMC9299808 | pmc | 5,913 | {
"abstract": "Summary Urbanised environments have been identified as hotspots of anthropogenic methane emissions. Especially urban aquatic ecosystems are increasingly recognised as important sources of methane. However, the microbiology behind these emissions remains unexplored. Here, we applied microcosm incubations and molecular analyses to investigate the methane‐cycling community of the Amsterdam canal system in the Netherlands. The sediment methanogenic communities were dominated by Methanoregulaceae and Methanosaetaceae , with co‐occurring methanotrophic Methanoperedenaceae and Methylomirabilaceae indicating the potential for anaerobic methane oxidation. Methane was readily produced after substrate amendment, suggesting an active but substrate‐limited methanogenic community. Bacterial 16S rRNA gene amplicon sequencing of the sediment revealed a high relative abundance of Thermodesulfovibrionia. Canal wall biofilms showed the highest initial methanotrophic potential under oxic conditions compared to the sediment. During prolonged incubations the maximum methanotrophic rate increased to 8.08 mmol g DW \n −1 d −1 that was concomitant with an enrichment of Methylomonadaceae bacteria. Metagenomic analysis of the canal wall biofilm lead to the recovery of a single methanotroph metagenome‐assembled genome. Taxonomic analysis showed that this methanotroph belongs to the genus Methyloglobulus . Our results underline the importance of previously unidentified and specialised environmental niches at the nexus of the natural and human‐impacted carbon cycle.",
"introduction": "Introduction Since the Industrial Revolution, atmospheric greenhouse gas (GHG) concentrations have been steadily increasing due to human activities like cattle farming, intensive agriculture, use of synthetic fertilisers, waste management and fossil fuel burning (Schaefer et al ., 2016 ; Saunois et al ., 2020 ). Even though the current atmospheric methane (CH 4 ) concentration of >1.87 ppm is lower than the >416 ppm carbon dioxide (CO 2 ) concentration (Dlugokencky, 2020 ), CH 4 accounts for the equivalent of 60% of the radiative forcing of CO 2 due to its 86 times higher global warming potential over a 20‐year time‐scale (Myhre et al ., 2013 ; Dean et al ., 2018 ; Nisbet et al ., 2019 ). A total of 306 Tg yr −1 of CH 4 is emitted by freshwater ecosystems such as lakes, ponds and wetlands globally (Kirschke et al ., 2013 ; Saunois et al ., 2020 ). Wetlands comprise 40% of natural CH 4 emissions, whereas other freshwater systems are now thought to be as high as 159 Tg yr −1 or 43% of global natural CH 4 emissions (Bastviken et al ., 2011 ; Saunois et al ., 2020 ). An important understudied aspect of freshwater CH 4 emissions is the influence of urbanisation on the GHG emissions of the surrounding aquatic systems. Many freshwater sources have been attractive locations for human settlements, which led to the majority of cities containing waterways. The United Nations report that currently an estimated 54% of the human population is living in cities and this percentage is estimated to grow to 66% by 2050 (United Nations, 2015 ). Microorganisms tend to be more abundant in urban waters due to the combined sewer overflows or discharge from wastewater treatment plants (Young and Thackston, 1999 ; Hladilek et al ., 2016 ; Price et al ., 2018; Mansfeldt et al ., 2020 ). In addition, leaking natural gas and sewer pipes, as well as stormwater, influences available substrates for microbial communities in cities (Lamb et al ., 2016 ; Smith et al ., 2017 ; McLellan and Roguet, 2019 ). All these changes are consistent across waterways, so the term ‘urban stream syndrome’ was coined to describe these changes (Meyer et al ., 2005 ). A recent analysis of published CH 4 emission data from streams and rivers revealed that CH 4 concentrations within urban waters rival those of wetlands and agricultural streams (Stanley et al ., 2016 ). Furthermore, an analysis of diffusive CH 4 fluxes from various ecosystems revealed that, like wetlands, urban waterways have higher CH 4 emissions than non‐urbanised rivers and streams. Changes in nutrient loading caused by human activity, together with increased CH 4 concentrations, suggest that urbanisation leads to an imbalance between CH 4 production and consumption resulting in net emissions of CH 4 . CH 4 concentrations and emissions from freshwaters have been reported for several riverine systems in Europe (Alshboul et al ., 2016 ; Borges et al ., 2018 ; Marescaux et al ., 2018 ; Brown and Hershey, 2019 ; Herrero Ortega et al ., 2019 ), China (Wang et al ., 2018 ; Wang et al ., 2021 ) and the United States (Brigham et al ., 2019 ). The majority of studies find a positive correlation with temperature and dissolved CH 4 during summer. However, other environmental parameters, like degree of eutrophication, are not always correlated to increased CH 4 concentrations or emissions (Herrero Ortega et al ., 2019 ). Several studies posit that the increased concentrations within cities are due to wastewater treatment plant effluent and not due to production in the river sediment (Alshboul et al ., 2016 ; Wang et al ., 2018 ). A recent study of built canals in urban and agricultural environments showed CH 4 emissions for these systems as high as tropical wetlands, more than freshwater lakes (Peacock et al ., 2021 ). Thus, urban environments can be considered understudied hotspots of microbial CH 4 cycling. Most of the CH 4 from riverine and urban aquatic ecosystems is thought to be biogenic (Schaefer et al ., 2016 ; Zazzeri et al ., 2017 ). Biological CH 4 production is considered the last step in the anaerobic fermentative degradation of organic matter and is performed by methanogenic archaea (Conrad, 2009 ). Not all CH 4 produced in anaerobic environments enters the atmosphere. A majority is converted to CO 2 by aerobic and anaerobic methanotrophs, diminishing the climate impact (Knittel and Boetius, 2009 ; Knief, 2015 ). Therefore, insight into the microbial CH 4 cycle is paramount to understanding balances in CH 4 emissions. Until now urban microbiome research has mainly focused on planktonic cells in the water column (Savio et al ., 2015 ; Medeiros et al ., 2016 ; Cannon et al ., 2017 ; Hosen et al ., 2017 ; Fresia et al ., 2019 ), whereas methanogens reside in the anoxic sediments of urban waters. However, the studies outlined above reported differences in microbial community structure in urban waters compared to rural waters. Studies that also took samples of sediments observe a similar trend, with sediment microbial communities changing in response to increased nutrient input associated with urbanisation (Saxena et al ., 2015 ; Hosen et al ., 2017 ; Saxena et al ., 2018 ). So far, no investigation into the community structure of CH 4 ‐cycling microorganisms in urban waterways has been undertaken. Here, we describe the urban microbial community of the Amsterdam canals, in the Netherlands, to investigate the local CH 4 cycle of these heavily urbanised waterways. We provide a general description of the microbial community accompanied by microcosm‐based rate measurements of the methane‐cycling bacteria and archaea. Our study reveals that canal wall biofilms, a niche for aerobic Methyloglobulus methanotrophs, might form an as yet underestimated CH 4 filter in urbanised environments.",
"discussion": "Discussion The Netherlands is a densely populated river delta, with large parts of the country lying below sea level (Wong et al ., 2007 ). During the development of Dutch cities the canals served to optimise land use while allowing for water drainage, thereby preventing flooding of the cities (Hoeksema, 2007 ). Furthermore, transport of goods using waterways is efficient and access to trade routes was vital for economic development (Klitgaard, 2019 ). Therefore, man‐made urban canals became ubiquitous in the larger cities and iconic for the Dutch cityscape, and indeed in many cities around the world. At the same time, urban aquatic systems like these canals are implicated to emit CH 4 (Zazzeri et al ., 2017 ; Wang et al ., 2018 ; Brigham et al ., 2019 ; Herrero Ortega et al ., 2019 ; R. Wang et al ., 2020 ). Understanding the microbiology behind CH 4 emissions provides vital information about ecosystem carbon cycling and can aid in designing adequate measures to reduce CH 4 emissions. We set out to describe the microbial community in the urban canals of Amsterdam, determine the potential for both CH 4 production and consumption, and to identify an urban niche for CH 4 ‐cycling microorganisms. Urbanisation is linked to eutrophication, with an increasing number of studies reporting increased nutrient load caused by anthropogenic land use (Harrison et al ., 2012 ; Gessner et al ., 2014 ; Brown and Hershey, 2019 ; Herrero Ortega et al ., 2019 ). Increased nutrient loads can lead to algal blooms in freshwater due to increased net primary production (Huettel et al ., 2014 ; Martinez‐Cruz et al ., 2017 ; Van Bergen et al ., 2019 ). Consequently, the potential for CH 4 production increases as excess carbon is available, especially in highly eutrophic systems. The data presented here suggest that the Amsterdam canal waters are oligotrophic and oxygenated during summer. Moreover, the lack of CH 4 production over 100 days from unamended sediments indicates that the top layer canal sediment was depleted of easy‐to‐use carbon. However, the amount of time the sediments were in storage prior to the start of the incubation could have been sufficient to deplete most of the organic matter. The observed CH 4 production within a week in amended sediment microcosms shows a metabolically active and adaptable methanogenic community. Due to the oxygenated water column and active production of CH 4 after substrate amendment, the upper layer of the sediment could be an environmental niche for aerobic methanotrophs. Taken together, both the methanogenic and methanotrophic communities are able to respond rapidly to changes in substrate availability and show high potential for being a CH 4 source and filter respectively. Methanotrophy in freshwaters has been extensively studied for stratified lakes, while knowledge on riverine systems and shallow lakes is limited (Deutzmann et al ., 2014b ; Oswald et al ., 2017 ; Crevecoeur et al ., 2019 ; Cabrol et al ., 2020 ; Reis et al ., 2020 ). The canals of Amsterdam are well‐mixed due to boat traffic, especially in the city centre. Moreover, no floating vegetation was observed which is an important habitat for plant‐associated methanotrophic bacteria in other waters (Kip et al ., 2011 ; Faußer et al ., 2012 ; Yoshida et al ., 2014 ). Instead, we observed that the biofilm alongside the canal wall was capable of rapid oxidation of CH 4 compared to the water column samples. The brick canal wall is a unique, man‐made structure that is unlike the littoral zone of natural waters and is most commonly found in urban waterways. The rough surface of a clay brick provides opportunity for microorganisms to attach and colonise. Moreover, this brick could be the source of the rare earth elements required for the XoxF‐type methanol dehydrogenase found in the MAG. The canal wall biofilm has the capability of providing niches for diverse microbial metabolisms, niches that might be smaller in more natural settings (Battin et al ., 2016 ). In the environmental biofilm sample, 16S rRNA gene sequencing and metagenomics revealed that a Methyloglobulus sp. constituted about 0.2%–0.7% of the bacterial community. This low abundance led to low coverage in our metagenome and an incomplete MAG. However, pmoA and 16S rRNA phylogeny as well as two separate classification tools placed it within the Methyloglobulus genus. Previous studies have found these methanotrophs in lakes (Deutzmann et al ., 2014a ) and sand filters of drinking water treatment plants (Parks et al ., 2017 ; Poghosyan et al ., 2020 ). Thus, we are the first to report a Methyloglobulus sp. in an urban aquatic system and our microcosm experiments showed that these bacteria are active or highly adaptable. We posit that the canal wall biofilms could play an important role in an urban waterway as a niche habitat for CH 4 ‐cycling microorganisms. The initial rates of CH 4 oxidation in the biofilm were 70 times greater per g DW than the sediment. From the metabolic potentials, the canal wall biofilm seemed to be an environment most suitable for aerobic methanotrophs in our incubation experiments, more so than the sediment or the water column. The biofilm's rates are much higher due to the nature of our drying methods and the normalisation as a biofilm is high in microbial mass, whereas the sediment is higher in non‐microbial mass. The sediment CH 4 oxidation rates were similar to lakes in Northern Germany (Eller et al ., 2005 ). CH 4 oxidation rates of the sediment were also in line with restored peatland sediment incubations (Reumer et al ., 2018 ). However, oxidation potential measured for permafrost wetlands in Siberia exhibited initial rates that were 10 times higher than our sediment incubations (Knoblauch et al ., 2008 ). Taken together, aerobic methanotrophic rates in Amsterdam sediments were in the expected range for methanogenic sediments. To our knowledge, this is the first study where a canal wall biofilm has been identified as a habitat with high methanotrophic potential. Another aspect to the biofilm is its apparent versatility to changes in substrate availability. In theory, many urban surfaces have the potential for biofilm development. Within Amsterdam, this might not be limited to the brick canal wall as there are wooden poles for boat signs, houseboats, concrete walls and steel sheet piles. Consequently, there may be more unique urban habitats where methanotrophs could reside. Methanotrophic biofilms could be a way to mitigate CH 4 emissions in urban waterways, for example in areas impacted by diffuse pollution from wastewater. However, ebullition could contribute significantly to net CH 4 emissions in urban waterways as it has been shown to become the dominant emission pathway of methane in natural freshwater ecosystems under warming scenarios (Aben et al ., 2017 ). CH 4 bubbles will not be accessible to the biofilm community in such shallow waters as canals. Indeed, in situ measurements indicate that there was excess dissolved CH 4 (Table 1 ). Whether due to ebullition or diffusive transport limitations from the water to the canal wall, the biofilm's metabolic capacity was not great enough to mitigate CH 4 emissions entirely. We conclude that the biofilm community could be a novel CH 4 filter in urban waters for which stimulation could lead to a greater filter capacity. We used two different primer sets for archaeal and bacterial 16S rRNA gene amplicon sequencing respectively, to eliminate potential biases and obtain an accurate view of the microbial diversity. In the archaeal domain, the most abundant class was Bathyarchaeia with 31%–41% relative abundance. Due to improvements in sequencing technologies, Bathyarchaeia have been observed in many soils and sediments but their ecological role remains elusive (Zhou et al ., 2018 ). These putative organic matter degraders were shown to be able to grow on lignin (Yu et al ., 2018b ). Bathyarchaeia were detected in freshwater lakes and wetlands with similar relative abundances compared to the Amsterdam canals (Yang et al ., 2016 ; Narrowe et al ., 2017 ). Furthermore, the canal sediment archaeal communities harboured up to 33% CH 4 ‐cycling archaea (Fig. 2 ). The methanogenic community in the canal sediment consisted of a mix of hydrogenotrophic and acetoclastic families. Methanoregulaceae were most abundant which is expected due to their ubiquity in freshwater sediments (Wen et al ., 2017 ). This family consists of hydrogenotrophic methanogens but was not enriched during our microcosm incubations with H 2 and CO 2 . Instead, several Methanobacterium spp. were enriched, probably favouring the high substrate conditions created in the microcosm incubations. Methanosaetaceae were the second most abundant methanogenic family in the Amsterdam canal sediment. They were enriched in microcosms amended with acetate, which is their sole carbon and energy source (Jetten et al ., 1992 ; Smith and Ingram‐Smith, 2007 ). Methanosaetaceae have been found in other freshwater sediments like thermokarst lakes and rivers (De Jong et al ., 2018 ; Wilkinson et al ., 2019 ). The microcosms amended with methanol showed an archaeal community dominated by Methanosarcinaceae . Methanogens of this methylotrophic family comprised less than 1% of archaeal sequences in the environmental sediment but were revived quickly in our incubations. Curiously, the community present at the end of the unamended sediment incubations was highly similar to the environmental sediment. This could indicate a carbon‐starved but active methanogenic archaeal community in the sediments because CH 4 production was observed quickly and their relative abundance did not change over a period of 100 days of incubation in the controls. Importantly, it shows that the incubation strategy employed is relevant to the real‐world situation. Therefore, we hypothesise that acetoclastic and hydrogenotrophic methanogenesis are the dominant CH 4 production pathways in these urban sediments based on the abundance and activity of the Methanosaetaceae family and the presence of Methanoregulaceae . Initial methanogenic rates of the amended sediments were comparable to those of amended Arctic sediments at 20°C (Blake et al ., 2015 ). Furthermore, lake sediment from Northern Germany showed similar production rates after acetate amendment (Eller et al ., 2005 ). Interestingly, our microcosm incubations had higher initial CH 4 production than the observed maximum for thermokarst lake sediment (De Jong et al ., 2018 ). Thus, our determined methanogenic rates are within the range expected for freshwater sediment after substrate amendment. Unamended sediment incubations did not show CH 4 production so identifying the source of sediment carbon is a point for further research. The bacterial community of the environmental sediment was highly diverse, with approximately 40% of the community consisting of sequences with a relative abundance below 1%. Sulfate‐reducing bacteria were abundant, with members of the uncultivated Thermodesulfovibrionia class making up 8%–10% of the total bacterial community (Fig. S4 ). Sulfate is a byproduct of organic matter degradation and is most likely naturally available in canal sediments (Table S1 ). The canals of Amsterdam receive brackish water from the IJ, which would increase the sulfate load and, in turn, explain the presence of sulfate reducers. Since the community in the sediment did not change during the microcosm incubations it is likely that the top layer prokaryotic community is probably starved for nutrients. The sediment did not harbour many nitrogen‐cycling microorganisms, with ammonium oxidisers ( Nitrosomonadaceae ) being the most abundant with 1.9%–3.1% relative abundance. Anammox bacteria of the Brocadiaceae family comprised less than 0.05% of the total community while no Nitrobacter reads were obtained. Nitrite‐oxidising bacteria of the Nitrospira genus were detected at 0.8% relative abundance on average, but only in the canal sediment. In summary, nitrogen compounds seem to be present in low amounts indicating that there is little nitrogen pollution even in the Amsterdam city centre. The genomic potential for anaerobic oxidation of methane was striking. 9% of the archaeal community was classified as Ca . Methanoperedens, a methanotroph capable of oxidising CH 4 anaerobically using NO 3 \n − , Fe(III), or Mn(IV) (Haroon et al ., 2013 ; Cai et al ., 2018 ; Leu et al ., 2020 ). In addition, members of the Methylomirabilota that are known to perform nitrite‐dependent anaerobic methane oxidation were detected to be as much as 1% of the bacterial community (Raghoebarsing et al ., 2006 ; Ettwig et al ., 2010 ). Linking these two domains of life with the qPCR results (Fig. S8 ) and metagenome sequencing (Supplementary Tables S4 – S6 ) showed that nitrate‐ and nitrite‐dependent anaerobic methanotrophs occurred at the same approximate absolute abundance. It has been shown that these two anaerobic methanotrophs co‐occur in freshwater sediments and together perform CH 4 ‐dependent denitrification (Shen et al ., 2017 ; Shen et al ., 2019 ). They could fill a niche in the sediment oxidising CH 4 anaerobically while competing for nitrate with nitrogen‐cycling microorganisms like anaerobic denitrifiers. Our community analysis and microcosm incubation experiments showed little variation between the sampling sites. The biofilm community was highly similar between the five biological samples and a similar result was observed for the three sediment communities. Even though the environmental samples were taken on opposite sides of the city centre (Fig. S1 ), their core microbial communities remained comparable. This finding indicates that our studied waterways are spatially homogeneous. Consequently, we propose that our findings are representative for the entire canal network of the Amsterdam city centre. More importantly, our data have the potential to be applicable to other cities with similar canal networks. Cities with eutrophic waterways or agricultural ditches rich in nitrogen and phosphorus will likely have different CH 4 dynamics from the studied Amsterdam canals. Therefore, investment in efficient wastewater treatment, and the separation of sewer and stormwater systems, could lead to oligotrophic waters and thus lower GHG emissions. However, the exact impact on the microbial community of urban land use compared to other land use types requires further study. Due to the widespread nature of urban waterways not only in the Netherlands but globally, understanding this ecosystem's response to climate warming and human activity is crucial. Moreover, ecological niches present in urban waterways will likely become more important as more land area will become urbanised. Within this man‐made environment, we found that the biofilm attached to the canal walls has the potential to act as a CH 4 filter. The activity of the methanogenic community and metabolic potential emphasised that the canals can be a significant source of atmospheric CH 4 . Further research is required to determine if net GHG fluxes and the prokaryotic community changes temporally, especially between summer and winter, and the implications for CH 4 ‐cycling and net emissions."
} | 5,733 |
36605799 | PMC9765428 | pmc | 5,915 | {
"abstract": "Low efficiency of extracellular electron transfer (EET) is a major bottleneck in developing high-performance microbial fuel cells (MFCs). Herein, we construct Shewanella oneidensis MR-1@Au for the bioanode of MFCs. Through performance recovery experiments of mutants, we proved that abundant Au nanoparticles not only tightly covered the bacteria surface, but were also distributed in the periplasm and cytoplasm, and even embedded in the outer and inner membranes of the cell. These Au nanoparticles could act as electron conduits to enable highly efficient electron transfer between S. oneidensis MR-1 and electrodes. Strikingly, the maximum power density of the S. oneidensis MR-1@Au bioanode reached up to 3749 mW m −2 , which was 17.4 times higher than that with the native bacteria, reaching the highest performance yet reported in MFCs using Au or Au-based nanocomposites as the anode. This work elucidates the role of Au nanoparticles in promoting transmembrane and extracellular electron transfer from the perspective of molecular biology and electrochemistry, while alleviating bottlenecks in MFC performances.",
"conclusion": "3 Conclusions In summary, a S. oneidensis MR-1@Au bioanode was constructed by in situ biomineralization. Transmembrane electron channels formed by Au NPs enabled an extraordinarily enhanced electron transfer compared with that in native bacteria, thus exhibiting promising applications in MFCs. Both the greatly enhanced current output in electrochemical half-cells and the power output in MFCs demonstrated the significant improvement in bioelectricity production. Notably, the maximum power density of the S. oneidensis MR-1@Au bioanode reached up to 3749 mW m −2 , reaching the highest performance yet reported in MFCs using Au or Au-based nanocomposites as the anode. This work provides proof of mechanism for the enhancement of Au-facilitated bioelectricity generation from the perspectives of molecular biology and electrochemistry, and points out the direction for the construction of a high-performance bioanode in MFCs.",
"introduction": "1 Introduction Microbial fuel cells (MFCs), which can directly convert chemical energy stored in biodegradable organic wastes or biomass into electrical energy through microbial metabolism, are applicable for wastewater remediation, desalination, and the removal of toxic chemicals from the environment. 1–4 However, the low power density remains a fundamental bottleneck for the practical application of MFCs and the MFC performance relies heavily on extracellular electron transfer (EET) between the intracellular respiratory chains of exoelectrogens and electron acceptors. 5–8 Among all the exoelectrogens, Shewanella oneidensis is one of the well-studied exoelectrogens type strains due to their robust growth in aerobic and anaerobic environments and their abundant distribution in soil and seawater. 9 Two EET mechanisms of S. oneidensis have been demonstrated through extensive research. 10 One is indirect electron transfer mediated by endogenously secreted soluble redox molecules. 11 Another EET mechanism is the contact-based direct extracellular electron transfer (DET), in which electrons are directly transferred to the anode via a number of conductive outer-membrane c-type cytochromes. 12–15 Notably, DET efficiency is generally considered as a key factor in improving MFC performance, 16 but is currently limited by inefficient interfacial contact between exoelectrogens and the electrode surface. 17 In recent years, substantial efforts have been centered on improving the DET efficiency through modifying electrodes with various functional nanomaterials, 18,19 including carbonaceous material, 20–22 metals, 23 metal oxides 24 and conducting polymers. 25 Highly conductive nanomaterials can act as electron transport channels for bacteria and thus significantly improve the EET efficiency. 26 Moreover, the additional active sites introduced by them can improve interfacial electron transfer between bacteria and the electrode, leading to efficient biocatalysis and electrocatalysis. 27 However, most of the bacterial cells inside the natural biofilm formed in this way are far from being functional nanomaterials, and can only transfer electrons to the electrodes through slow electron hopping of multiple redox centers between bacteria, limiting the improvement of energy output in MFCs. 28,29 In this context, a single-bacterial surface modification strategy was proposed to construct an interconnected intact conductive layer on and across the individual cell membranes for creating highly conductive and stable catalytic interfaces for exoelectrogens and electrodes. 30–33 Conductive polymers, including polypyrrole (PPy), polydopamine (PDA) and their composites, have been used to coat the bacterial surface to reduce charge transfer resistance. 34,35 However, these materials were mainly attached to the outer membrane of bacterial cells and were difficult to embed into the periplasm or inner membrane. In view of this, if the transmembrane electron transport can be further facilitated, the MFC performance would be greatly improved. 36 In addition, efficient interfacial electron transfer requires extremely close contact between the transmembrane electron transfer conduits and conductive abiotic surface. 37 The biomineralization mechanism of exoelectrogens provides a possibility for the realization of this strategy. 16 Several metal nanoparticles (NPs) and their complexes have been studied to improve the transmembrane electron transport efficiency and the cell viability. 38–40 Compared with metal oxides, pure metal NPs possess the advantages of higher conductivity, higher stability, better nanostructure manipulation ability and catalytic activity. 41 Gold nanoparticles (Au NPs) are the most stable metal nanoparticles and have widely served as an ideal anode surface modification material to improve EET efficiency and bacterial adhesion, due to their good biocompatibility, high conductivity and tunable surface charge. Au NPs can accelerate the growth of the Shewanella oneidensis MR-1 biofilm, and MFCs based on a carbon paper-Au anode generate 47% higher total electric charges than MFCs with a carbon paper anode. 42 Furthermore, the coulombic efficiency and power generation could be increased with the Au density increasing on the anode surface. By depositing carbon paper with an Au thickness of 100 nm on each side, the maximum power density was enhanced by 188% and the stabilization time of maximum power generation was increased by 122%. 43 In most cases, Au NPs are fabricated using various chemical methods under heat or sonication conditions. In contrast, Au NPs fabricated by microbial methods not only have the advantage of requiring fewer chemical reagents and reactions under mild conditions, but also have better biocompatibility and higher catalytic activity. Wu et al. tested biogenic Au NPs for anode modification in MFCs, which resulted in a 23% increase in maximum power density compared to a bare carbon cloth control. 44 In previous studies, Au NPs were mainly used as electrode modification materials. Although single-bacterial surface modification technology has been introduced in MFCs, the enhancement of the power generation performance of exoelectrogens by Au NPs has not been studied and the exact location and basic role of these Au NPs in exoelectrogens remain elusive. 45 Therefore, it is necessary to elucidate the specific mechanism by which metal NPs enhance the EET efficiency through in-depth mechanistic studies, and to design a high-performance bioanode to fundamentally address the EET limitations. Herein, we fabricated S. oneidensis MR-1@Au by in situ biomineralization, and deeply studied the mechanism of Au NPs for the enhancement of EET efficiency from the perspective of molecular biology and electrochemistry ( Fig. 1 ). Our systematic studies demonstrated that Au NPs not only tightly covered the bacteria surface, but were also distributed in the periplasm, cytoplasm and cell membrane. These Au NPs could act as electron conduits to provide additional electron channels for membrane cytochromes to facilitate transmembrane and extracellular electron transport, thereby enhancing overall MFC performance. This work not only relieves the bottleneck of MFC performance, but also provides guidance for the design of high-performance bioanodes. Fig. 1 (a) Schematic illustration of the synthesis of S. oneidensis MR-1@Au for enhanced bioelectricity generation. (b) The hypothetical electron transfer pathway in S. oneidensis MR-1@Au.",
"discussion": "2 Results and discussion 2.1 Assembly of S. oneidensis MR-1@Au As shown in Fig. 2a , native S. oneidensis MR-1 cells were rod-shaped with a relatively smooth surface. S. oneidensis MR-1@Au were fabricated using an in situ biomineralization synthesis strategy ( Fig. 1a ). 46 By simply adjusting the polymerization time (the optimum polymerization time is 16 h), a uniformly covered Au nanoshell was assembled on the cell surface ( Fig. 2b and S1 † ). To evaluate the spatial distribution of Au NPs within an individual bacterium, the transmission electron microscope (TEM) image, high-angle annular dark field-scanning transmission electron microscope (HAADF-STEM) image and energy dispersive X-ray (EDX) elemental mapping image of S. oneidensis MR-1@Au were further recorded. The TEM images of cross-section slices of pristine S. oneidensis MR-1 cells showed a relatively smooth profile and uniform contrast, whereas distinct black spots were found in S. oneidensis MR-1@Au ( Fig. 2c, d and S2 † ). Strikingly, abundant Au NPs were aligned in the periplasm and cytoplasm, and even embedded in the cell membrane ( Fig. 2e and S3 † ). The size of these transmembrane and extracellular NPs was approximately 10–30 nm, while the intracellular NPs had a diameter between 5 and 10 nm. Due to the limitation of intracellular space, nanoparticles formed intracellularly are smaller in size than those formed extracellularly. 12 A possible mechanism for this phenomenon is that AuCl 4 − diffused into the bacteria and was then reduced in situ by electrons generated by metabolism, resulting in the formation of Au NPs. 47 Elemental mapping revealed a uniform distribution of the Au element in and across the cell membrane ( Fig. 2f ). Moreover, the nanoparticles were characterized by X-ray diffraction (XRD) analyses (Fig. S4 † ). The characteristic peaks in the XRD spectrum of S. oneidensis MR-1@Au can correspond to the (111), (200), (220) and (311) planes of Au NPs, proving the formation of Au NPs on S. oneidensis MR-1. 48 This result indicated that the nanoparticles present on the cell surface were indeed Au NPs. All these results demonstrated the successful assembly of S. oneidensis MR-1@Au. Fig. 2 (a and b) SEM images of S. oneidensis MR-1 (a) and S. oneidensis MR-1@Au (b). (c and d) TEM images of cross-section slices of S. oneidensis MR-1 (c) and S. oneidensis MR-1@Au (d). The insets of the image (c and d) are the corresponding high-magnification TEM images. (e) HAADF-STEM image of cross-section slices of S. oneidensis MR-1@Au. (f) Element mapping image of cross-section slices of S. oneidensis MR-1@Au. OM: outer membrane of S. oneidensis MR-1 and IM: inner membrane of S. oneidensis MR-1. (g and h) CLSM images of S. oneidensis MR-1 (g) and S. oneidensis MR-1@Au (h), respectively. The insets of the image (g and h) are the corresponding high-magnification CLSM images. Moreover, based on the differential permeability between an intact or compromised cell membrane, we assessed the cell viability after biomineralization. While propidium iodide (PI, red fluorescence) penetrates into cells with damaged membranes exclusively and indicates the dead cells, SYTO 9 (green signal) can stain both living and dead cells. The confocal laser scanning microscope (CLSM) images of S. oneidensis MR-1 and S. oneidensis MR-1@Au both showed strong green fluorescence and slight red fluorescence (the proportion of red fluorescence in both is less than 5%, Fig. S5a † ), indicating that neither the biomineralization process nor the Au NPs were detrimental to bacterial activity ( Fig. 2g and h ). Although some studies have proved that Au NPs can impair bacterial activity by altering the permeability of cell membrane, the concentration of Au NPs used in these studies were much higher than that used in our study. In addition, the difference in the synthesis method and particle size is also a main reason for cytotoxicity. Considering the heavy metal tolerance of Shewanella and the biomineralization synthesis method, it is not surprising that S. oneidensis MR-1@Au maintained such high cell viability. 49 We further explored the effect of Au NPs on bacterial activity after 120 h cultivation. Compared with the initial CLSM image ( Fig. 2g ), the CLSM image of native bacteria showed more red fluorescence (Fig. S6a † ), indicating an increased percentage of dead cells. In contrast, the functionalized bacteria retained higher activity (Fig. S6b † ). The proportion of red fluorescence in native bacteria after 120 h cultivation is 24.3 ± 6.9%, which is 4.5 times higher than that in S. oneidensis MR-1@Au (Fig. S5b † ). These results suggested that Au NPs are beneficial for maintaining bacterial viability during long-term operation. In addition, the cell growth curves showed that the biomineralization of Au NPs significantly prolonged the lag phase of S. oneidensis MR-1 and the modified cells could retain the capability of dividing themselves under aerobic conditions (Fig. S7 † ). 38 2.2 Electricity generation capability of S. oneidensis MR-1@Au After successful assembly of S. oneidensis MR-1@Au, the output current density of the functionalized bacteria was evaluated with a three-electrode system in an electrochemical half-cell. 50 The current output of S. oneidensis MR-1@Au increased continuously with incubation time and reaches a nearly constant value within 45 h ( Fig. 3a ), indicating the successful establishment of a stable biofilm. It then dropped sharply after 35 hours of stable operation, which can be attributed to the substrate consumption and metabolite accumulation. After the replacement of the fresh medium ( Fig. 3a , t = 130 min and t = 230 min), the maximum current density can be restored to a parallel level. Strikingly, S. oneidensis MR-1@Au delivered a current density approximately 4.3 times higher than that of the native bacteria (159 μA cm −2 vs. 37 μA cm −2 , Fig. 3a ), suggesting that the Au NPs formed by in situ mineralization could significantly improve the electricity generation. Moreover, the electrochemical half-cells with dead S. oneidensis MR-1@Au did not deliver a significant current output, indicating that the electrons were only derived from the bacteria (Fig. S8 † ). Fig. 3 (a) Time profile of electricity generation of different bioanodes in electrochemical half-cells. Arrows represent the replacement of a fresh medium. (b) Current output of native or functionalized cells ( n = 3) in electrochemical half-cells. Error bars represent standard error (s.e.) determined by three independent experiments. (c) Polarization (hollow symbols) and power density output curves (solid symbols) of different bioanodes in MFCs. Considering that S. oneidensis MR-1 are not easily apt to generate bionanowires to achieve electron transfer in nutrient-rich environments, membrane cytochromes might play a main role in EET. In such a process, the electrons transferred extracellularly mostly originate from lactate oxidation by lactate dehydrogenase. 51,52 The electrons from menaquinone are hopping from the CymA cytochromes redox centers on the inner membrane to the periplasm. Subsequently, the electrons are further transferred through outer membrane proteins MtrA, MtrB, MtrC, and OmcA to the electrode surface ( Fig. 1b ). To further explore the role of Au NPs in transmembrane and extracellular electron transport, we also investigated the corresponding mutants ΔCymA and ΔMtrC/OmcA . As expected, the disruption of CymA or MtrC/OmcA greatly suppressed (by over 95% and 81%, respectively) the current output of the native S. oneidensis MR-1 ( Fig. 3b ). It has been reported that Au NPs could participate in catalyzing the oxidation of organics and repairing cell damage in electron transfer to some extent. 46 Compared with ΔCymA , which can hardly generate current, the current of ΔCymA @Au can be recovered to 42 ± 8 μA cm −2 , indicating that Au NPs played a role in electron transport across the inner membrane similar to CymA ( Fig. 3b ). Strikingly, the current of ΔMtrC/OmcA @Au reached 125 ± 7 μA cm −2 , which was 3.4 times higher than that of native S. oneidensis MR-1, proving that the Au NPs embedded in the periplasm and outer membrane made a major contribution to transmembrane electron transfer. Taken together, these results demonstrated that Au NPs could act as electron conduits to provide additional electron channels for membrane cytochromes to facilitate transmembrane and extracellular electron transport. To probe the power output of S. oneidensis MR-1@Au, we constructed a double-chamber MFC and measured the polarization curves when the MFC was stably discharged ( Fig. 3c ). Impressively, the maximum power density of the MFC with S. oneidensis MR-1@Au bioanode reached up to 3749 mW m −2 , which was 17.4 times higher than that with the native S. oneidensis MR-1 (216 mW m −2 ). This superior power density is clearly higher than those of previously reported MFCs using Au and Au-based nanocomposites as anodes ( Table 1 ). All these results demonstrate that the Au NPs not only improved the EET efficiency of the individual cell but also facilitated electron transfer across the biofilm, thereby enhancing the power output. Comparison of the performance of previous MFCs using Au and Au-based nanocomposites as anodes Electrode substrates Anode materials Microbe type Power density (mW m −2 ) Ref. Carbon paper CNT/Au/TiO 2 \n E. coli \n 2.4 \n 53 \n Carbon felt MWCNT-Au-Pt/osmium redox polymer \n Gluconobacter oxydans \n 32.1 \n 54 \n Carbon cloth BioAu/MWCNT Mixed bacteria 178.34 ± 4.79 \n 44 \n Carbon paper Au Mixed bacteria 346 \n 55 \n Carbon paper Au Mixed bacteria 461.6 \n 43 \n Carbon paper G/Au \n Shewanella oneidensis \n 508 \n 56 \n Carbon cloth Au@PANI \n E. coli \n 804 ± 73 \n 57 \n Carbon paper Au Mixed bacteria 990 \n 58 \n Carbon cloth — Au and Fe 3 O 4 -coated Shewanella oneidensis 1792 \n 45 \n Carbon felt — \n E. coli @Au 1 @CdS 1 2300.4 \n 40 \n — Fe 3 O 4 /Au NCs-3DGF \n Shewanella oneidensis \n 2980 ± 54 \n 59 \n \n Carbon felt \n \n — \n \n \n S. oneidensis MR-1@Au \n \n 3749 \n \n This work \n 2.3 Mechanism investigation To explore the reasons for the excellent performance of the S. oneidensis MR-1@Au bioanode, we need to further understand the role of Au NPs in the charge transfer process. EET efficiency is the primary factor that is highly associated with electricity generation and electron utilization capability. 60 Thus, the electrochemical impedance spectroscopy (EIS) technique was employed to evaluate the interfacial charge transfer behaviors of different bioanodes in 10 mM Fe(CN) 6 3+ /Fe(CN) 6 4+ containing 100 mM KCl. In the Nyquist curve, the semicircle portion at higher frequencies corresponds to charge transfer resistance ( R ct ) at the liquid–solid interface, which represents the resistance of electrochemical reactions on the electrode and determines the electron transfer kinetics of bioanodes. 61 As shown in Fig. 4a , the S. oneidensis MR-1@Au bioanode reduced the interfacial charge transfer resistance by approximately 19.5 times (65 Ω vs. 1270 Ω), suggesting that Au NPs could significantly facilitate electron transport inside the biofilm and help to establish a favorable interface between the bacteria and the extracellular solid conductive surface. Fig. 4 (a) Nyquist plots of electrochemical impedance spectroscopy of S. oneidensis MR-1 and S. oneidensis MR-1@Au bioanodes in 10 mM Fe(CN) 6 3+ /Fe(CN) 6 4+ containing 100 mM KCl. (b) CV curves of S. oneidensis MR-1 and S. oneidensis MR-1@Au in electrochemical half-cells without a carbon source at a scan rate of 5 mV s −1 . Cyclic voltammograms of S. oneidensis MR-1 (c) and S. oneidensis MR-1@Au (e) biofilms at different scan rates (arrows showed scan rates at 5, 10, 25, 50, 75 and 100 mV s −1 , respectively). Dependence of reduction current density ( j p ) versus scan rate ( v ) on S. oneidensis MR-1 (d) and S. oneidensis MR-1@Au (f) biofilms, separately; inset: linear dependence of j p versus v 1/2 . Furthermore, the redox reaction kinetics at cell–electrode interfaces at the pseudo-steady state were analyzed in detail by cyclic voltammetry (CV). In native S. oneidensis MR-1, electron transport from the cytoplasmic membrane across the periplasm and outer membrane to the electrode is strongly dependent on the efficient pathway of cytochrome chains. Before performing the CV analyses, the medium was changed to fresh M9 buffer to remove the effect of flavin and lactate. As shown in Fig. 4b , the CV curve of S. oneidensis MR-1 bioanode showed a couple of peaks at −0.31 V and −0.03 V, which are relevant to the electrochemical response of outer membrane c-type cytochromes. 34,36 Since the peak separation (Δ E p ) between the oxidation and reduction peaks is inversely proportional to the electron transfer rate, a smaller Δ E p represents an increase in the EET rate. 55 Compared with the S. oneidensis MR-1 bioanode, the Δ E p of the S. oneidensis MR-1@Au bioanode was significantly reduced (0.28 V vs. 0.22 V), illustrating that Au NPs could obviously facilitate the DET efficiency and promote the electron exchange at the electrode surface. To get deeper into the redox reaction kinetics at the cell–electrode interface, we collected the CVs of different bioanodes at a series of scan rates (from 5 to 400 mV s −1 ). It is noted that the peak current of the S. oneidensis MR-1 bioanode depended linearly on the square root of the scan rates, indicating that the redox reaction of c-type cytochromes of S. oneidensis MR-1 was a typical diffusion-controlled process ( Fig. 4c and d ). In contrast, the peak current of the S. oneidensis MR-1@Au bioanode showed a good linear relationship with the scan rate, implying that the surface-controlled electron transfer process was prominent ( Fig. 4e and f ). 62 These results further confirmed the promoted EET after the modification of Au NPs on the cell surface. 2.4 Biomass and activity determination of the bioanode The bacteria-loading amount on the electrode surface and cell viability are also important factors affecting the MFC performance. We investigated the morphology of the bioanode when the MFCs reached the highest voltage. A compact biofilm consisting of densely packed bacteria was found in the S. oneidensis MR-1@Au bioanode ( Fig. 5a and b ), whereas only a few native S. oneidensis MR-1 were attached to the carbon felt anode. Furthermore, the biomass of these bioanodes was quantitatively determined by using a Detergent Compatible Bradford Protein Assay Kit. As expected, the bacteria loading mass of the S. oneidensis MR-1@Au bioanode (276.2 ± 21.2 μg cm −2 ) was significantly higher than that of the native S. oneidensis MR-1 bioanode (158.5 ± 36.7 μg cm −2 ). This biomass determination result coincided well with that from scanning electron microscope (SEM) analysis, indicating that the presence of the Au NPs is beneficial to the formation of a dense biofilm. We have also evaluated the cell viability of these bioanodes in MFCs over 120 h operation ( Fig. 5c and d ). The percentage of dead bacteria on the surface of S. oneidensis MR-1 is 41.7 ± 11.2%, which is 8.9 times higher than that of S. oneidensis MR-1@Au (Fig. S9 † ), proving that Au NPs exhibit terrific biocompatibility which helps the bacterial proteins to retain their native structure and enzymatic activity and thus increase bacterial stability in long-term operation. These results demonstrated that the S. oneidensis MR-1@Au bioanode exhibits higher EET efficiency, better bioactivity and more biological attachment, which enables the rapid transfer of electrons from the bacteria to the electrode and results in an enormous increase in MFC performance. Fig. 5 SEM (a and b) and CLSM (c and d) images of S. oneidensis MR-1 (a and c) and S. oneidensis MR-1@Au (b and d), respectively. The insets of the image (a–d) are the corresponding high-magnification images."
} | 6,175 |
27014323 | PMC4791364 | pmc | 5,917 | {
"abstract": "Global mechanization, urbanization, and various natural processes have led to the increased release of toxic compounds into the biosphere. These hazardous toxic pollutants include a variety of organic and inorganic compounds, which pose a serious threat to the ecosystem. The contamination of soil and water are the major environmental concerns in the present scenario. This leads to a greater need for remediation of contaminated soils and water with suitable approaches and mechanisms. The conventional remediation of contaminated sites commonly involves the physical removal of contaminants, and their disposition. Physical remediation strategies are expensive, non-specific and often make the soil unsuitable for agriculture and other uses by disturbing the microenvironment. Owing to these concerns, there has been increased interest in eco-friendly and sustainable approaches such as bioremediation, phytoremediation and rhizoremediation for the cleanup of contaminated sites. This review lays particular emphasis on biotechnological approaches and strategies for heavy metal and metalloid containment removal from the environment, highlighting the advances and implications of bioremediation and phytoremediation as well as their utilization in cleaning-up toxic pollutants from contaminated environments.",
"conclusion": "Conclusion and Future Aspects Phytoremediation is an eco-friendly ‘green-clean’ technology that has tremendous potential to be utilized in the cleaning up of heavy metals and organic pollutants. For organic pollutants, plants, and rhizospheric bacteria have demonstrated the ability to detoxify and mineralize the former to harmless products that can be removed without causing accumulation. There are also a few reports of utilizing phytoremediation to successfully remove TCE and other organic compounds ( Dhankher et al., 2011 ). However, in the case of toxic metals, plants can uptake, detoxify, translocate, and accumulate them in the aboveground biomass, which has to be then harvested for metal recovery. Despite tremendous potential for the application of phytoremediation in the cleaning up of contaminated soil, sediment, and water, it has not been commercialized and used extensively on a large scale. There are many reports of heavy metal/metalloid uptake, detoxification, and accumulation but most of these were described at the laboratory scale in model plants ( Dhankher et al., 2011 ; Hossain et al., 2012 ; Ovečka and Takáč, 2014 ). According to our knowledge, none of these studies have been applied in the field for heavy metal detoxification and phytoremediation thus far. Furthermore, progress toward commercializing the phytoremediation of heavy metals and metalloids has been hampered due to a lack of complete understanding of the metal uptake process from soil to roots, translocation from roots to shoots and accumulation in the biomass tissues. Several recent studies have attempted to unravel the mechanism of heavy metal and metalloid transport and accumulation in plants using transcriptomic and proteomics approaches ( Cvjetko et al., 2014 ). Additionally, metabolomic analysis can help to identify the metabolites associated with heavy metal and metalloid stresses, which can be further mapped to its metabolic pathways to identify the related candidate genes ( Kumar A et al., 2014 ). One intriguing approach to enhance our knowledge about heavy metal and metalloid metabolism in plants is to develop suitable techniques for imaging. Efforts have been made to employ Laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), Matrix Assisted Laser Desorption Ionization (MALDI) and Fourier Transform Ion Cyclotron Resonance Mass Spectrometry (FT-ICR-MS) toward this aim ( Jones et al., 2015 ). However, more efforts are needed to enable imaging visualization and determination of metal and metalloid localization and distribution in plant tissues. Despite recent progresses in biotechnological applications and the availability of complete genome sequences of several plants species, the potential of phytoremediation has still not been fully exploited for the successful application of this technology on a commercial scale for the cleaning of contaminated soil and water. Another major factor for the lack of progress in this area is inadequate funding for phytoremediation research. Next generation sequencing was used to study the whole genomes and transcriptomes of several heavy metal-tolerant organisms ( Hu et al., 2005 ; He et al., 2011 ; Peña-Montenegro and Dussán, 2013 ). Mass spectrometry-based proteomics is extensively used to study heavy metal and other forms of stresses in candidate organisms including plants ( Hossain and Komatsu, 2012 ; Cvjetko et al., 2014 ), bacteria ( Zakeri et al., 2012 ), and marine organisms ( Muralidharan et al., 2012 ). Furthermore, proteogenomics, the alliance between proteomics and genomics ( Helmy et al., 2012 ), is being used to study the genomic and proteomic properties of microorganisms that tolerate high concentrations of contaminants and high levels of stress ( de Groot et al., 2009 ; Delmotte et al., 2009 ; Rubiano-Labrador et al., 2014 ). Collectively, these efforts promise an upcoming generation of tailored organisms with higher bio/phytoremediation efficiencies and lower costs ( Figure 3 ). FIGURE 3 Integration of “Omics” tools for developing plants for phytoremediation. Genomics, transcriptomics, proteomics, metabolomics, and phenomics could help on identifying the candidate genes which can be used for developing plants for phytoremediation through different approaches including transgenic, cisgenic, gene- stacking, metabolic engineering, and genome editing. In future, efforts should be made to develop strategies to improve the tolerance, uptake, and hyperaccumulation of heavy metals/metalloids using genomic and metabolic engineering approaches. Pathways that control the uptake, detoxification, transport from root to shoot tissues and translocation and hyperaccumulation in the aboveground storage tissues can be engineered using gene-stacking approaches ( Figure 3 ). Additionally, genome editing strategies can be designed using TALENs (transcription activator like effectors nucleases) technology or the powerful CRISPR-Cas9 (clustered regularly interspaced short palindromic repeats) system to produce microbes/plants for bio/phytoremediation purposes ( Figure 3 ). Recently, an efficient and successful CRISPR/Cas9-mediated targeted mutagenesis has been reported in Populus plants ( Fan et al., 2013 ). This is a particularly interesting finding since Populus plants are known to be ideal plants for the phytoremediation of several toxic pollutants. Additionally, efforts should be made to develop breeding programs to improve the biomass and growth habits of natural hyperaccumulators and breed those traits into non-food, high biomass, fast growing plants for commercial phytoremediation of heavy metals and metalloids. Furthermore, efforts should be made to combine the phytoremediation approach with bioenergy through the dual use of plants for phytoremediation and biofuel production on contaminated lands. This approach would be useful to phytoremediate contaminated sites and simultaneously produce renewable energy that can offset the costs of applying these type of methodologies.",
"introduction": "Introduction Bioremediation is the use of natural and recombinant microorganisms for the cleanup of environmental toxic pollutants. It is considered a cost-effective and environmentally friendly approach. It relies on improved detoxification and degradation of toxic pollutants either through intracellular accumulation or via enzymatic transformation to lesser or completely non-toxic compounds ( Brar et al., 2006 ). Many naturally or genetically modified microorganisms possess the ability to degrade, transform, or chelate various toxic chemicals and hence provide better strategies to combat environmental pollution. On a regular basis, scientists deploy either natural or modified microbes to remove contaminants, viz., heavy metals, metalloids, radioactive waste, and oil products from polluted sites ( Dixit et al., 2015 ). Plants possess the necessary genetic, biochemical, and physiological characteristics to establish themselves as the ultimate choice for soil and water pollutant remediation. Phytoremediation refers to a diverse collection of plant-based technologies that use either naturally occurring or genetically engineered plants to clean contaminated environments ( Salt et al., 1995 , 1998 ; Flathman and Lanza, 2010 ). Phytoremediation is a cost effective, green-clean technology with long-term applicability for the cleaning up of contaminated sites. However, the required time frame to clean-up contaminants from soil prevents its use on an industrial scale. It involves the cleaning up of contaminated soil and water by either root colonizing microbes or by the plants themselves and is best applied at sites with shallow contamination of organic and inorganic pollutants ( Pilon-Smits, 2005 ). Due to this shortcoming, the utilization of biotechnological approaches involving high biomass fast growing crops for remediation purposes combined with biofuel production has gained momentum in recent years ( Oh et al., 2013 ; Pidlisnyuk et al., 2014 ). The development of new genetic tools and a better understanding of microbe and plant gene structures and functions have accelerated advancements in pathway-engineering techniques (referred to as designer microbes and plants) for improved hazardous waste removal. This review focuses on the accomplishments of biotechnological applications and strategies for environmental protection, detoxification, and the removal of heavy metals and metalloids. The current review article also examines recent developments and future prospects for the bio/phytoremediation of toxic pollutants from contaminated soil and water."
} | 2,485 |
37437031 | PMC10337918 | pmc | 5,918 | {
"abstract": "Manipulating the microbiome of cropland soils has the potential to accelerate soil carbon sequestration, but strategies to do so need to be carefully vetted. Here, we highlight the general steps required to develop, implement, and validate such microbe-based strategies."
} | 67 |
35224313 | null | s2 | 5,919 | {
"abstract": "In multicellular organisms, metabolism is compartmentalized at many levels, including tissues and organs, different cell types, and subcellular compartments. Compartmentalization creates a coordinated homeostatic system where each compartment contributes to the production of energy and biomolecules the organism needs to carrying out specific metabolic tasks. Experimentally studying metabolic compartmentalization and metabolic interactions between cells and tissues in multicellular organisms is challenging at a systems level. However, recent progress in computational modeling provides an alternative approach to this problem. Here we discuss how integrating metabolic network modeling with omics data offers an opportunity to reveal metabolic states at the level of organs, tissues and, ultimately, individual cells. We review the current status of genome-scale metabolic network models in multicellular organisms, methods to study metabolic compartmentalization "
} | 242 |
38742878 | PMC11257099 | pmc | 5,920 | {
"abstract": "ABSTRACT Bacterial endosymbionts of eukaryotic hosts typically experience massive genome reduction, but the underlying evolutionary processes are often obscured by the lack of free-living relatives. Endomicrobia, a family-level lineage of host-associated bacteria in the phylum Elusimicrobiota that comprises both free-living representatives and endosymbionts of termite gut flagellates, are an excellent model to study evolution of intracellular symbionts. We reconstructed 67 metagenome-assembled genomes (MAGs) of Endomicrobiaceae among more than 1,700 MAGs from the gut microbiota of a wide range of termites. Phylogenomic analysis confirmed a sister position of representatives from termites and ruminants, and allowed to propose eight new genera in the radiation of Endomicrobiaceae . Comparative genome analysis documented progressive genome erosion in the new genus Endomicrobiellum , which comprises all flagellate endosymbionts characterized to date. Massive gene losses were accompanied by the acquisition of new functions by horizontal gene transfer, which led to a shift from a glucose-based energy metabolism to one based on sugar phosphates. The breakdown of glycolysis and many anabolic pathways for amino acids and cofactors in several subgroups was compensated by the independent acquisition of new uptake systems, including an ATP/ADP antiporter, from other gut microbiota. The putative donors are mostly flagellate endosymbionts from other bacterial phyla, including several, hitherto unknown lineages of uncultured Alphaproteobacteria , documenting the importance of horizontal gene transfer in the convergent evolution of these intracellular symbioses. The loss of almost all biosynthetic capacities in some lineages of Endomicrobiellum suggests that their originally mutualistic relationship with flagellates is on its decline. IMPORTANCE Unicellular eukaryotes are frequently colonized by bacterial and archaeal symbionts. A prominent example are the cellulolytic gut flagellates of termites, which harbor diverse but host-specific bacterial symbionts that occur exclusively in termite guts. One of these lineages, the so-called Endomicrobia, comprises both free-living and endosymbiotic representatives, which offers the unique opportunity to study the evolutionary processes underpinning the transition from a free-living to an intracellular lifestyle. Our results revealed a progressive gene loss in energy metabolism and biosynthetic pathways, compensated by the acquisition of new functions via horizontal gene transfer from other gut bacteria, and suggest the eventual breakdown of an initially mutualistic symbiosis. Evidence for convergent evolution of unrelated endosymbionts reflects adaptations to the intracellular environment of termite gut flagellates.",
"conclusion": "Conclusions Our comparative genomic analysis of Endomicrobiaceae provided new insights into the evolutionary processes underlying symbiogenesis in intracellular symbionts of termite gut flagellates. The transition from a free-living to an intracellular lifestyle not only restricts gene flow but also affects the ability to take up substrates and other nutrients from the environment. Members of the genus Endomicrobiellum possess functions that either represent adaptations to their intracellular niche or serve to compensate for gene losses during ongoing genome erosion. Some functions represent predispositions to an intracellular lifestyle that were present already in the common ancestor of Endomicrobiaceae , whereas others were apparently acquired by horizontal gene transfer. The acquisition of similar traits also by other established or putative flagellate symbionts underscores the importance of horizontal gene transfer from other gut microbiota in the convergent evolution of endosymbiotic lineages. The lack of biosynthetic functions common to all members of the genus Endomicrobiellum suggests that their original role in the symbiosis became compromised during the progressive genome reduction. It is possible that certain lineages now just represent a metabolic burden to their host. The acquisition of an NTT may indicate the transition from a mutualistic to a parasitic relationship, but it may also be the last straw to ensure survival for a beneficial endosymbiont whose energy metabolism became compromised owing to genome erosion. It is likely that dysfunctional members of Endomicrobiellum will be eventually replaced by secondary symbionts that co-occur in the same flagellates, as shown for other primary endosymbionts of eukaryotic hosts that have experienced severe gene losses ( 1 , 82 ).",
"introduction": "INTRODUCTION Bacterial endosymbionts of eukaryotes typically experience progressive genome erosion ( 1 – 4 ). Apart from several well-characterized endosymbionts in insect tissues, the absence of close but free-living relatives often impedes our understanding of the evolutionary processes and mechanisms underlying the transition from a host-associated to an intracellular lifestyle. A notable exception are the so-called Endomicrobia, which represent a family-level clade in the phylum Elusimicrobiota and have been detected almost exclusively in the intestinal tract of insects ( 5 – 7 ) and ruminants ( 7 – 9 ). The family Endomicrobiaceae comprises both free-living gut commensals of their respective hosts and closely related lineages of endosymbionts that abundantly colonize the cytoplasm of termite gut flagellates ( 10 , 11 ). In the context of this study, we use the term “free-living” to distinguish host-associated bacteria that occur freely within the gut from those in an endosymbiotic association with protists. All evolutionary “lower” termites (all families except Termitidae or “higher” termites) harbor diverse assemblages of flagellate protists that are essential for the digestion of lignocellulose ( 12 , 13 ). The flagellates themselves are generally associated with bacterial symbionts that colonize the exterior surface, cytoplasm, or nucleus of their hosts ( 14 – 18 ). Although the symbionts belong to different bacterial phyla, they share conspicuous similarities in the metabolic capacities encoded by their respective genomes. Despite a substantial reduction in genome size, the biosynthetic pathways of the symbionts are mostly conserved, which led to the consensus that the symbionts either provide their flagellate host with fixed nitrogen and/or essential amino acids and cofactors ( 16 , 19 – 21 ) or contribute metabolic capacities that facilitate cellulose degradation, such as reductive acetogenesis ( 19 , 22 ). The shared presence of such traits suggests a parallel evolution of unrelated lineages of endosymbionts. Although the diversity of Endomicrobiaceae in termite guts has been investigated in great detail and numerous associations with different termite gut flagellates have been documented ( 5 – 7 , 11 , 23 , 24 ), it is unclear whether all intracellular symbionts were derived from a single endosymbiotic event or were independently acquired from different lineages of free-living ancestors. This is mostly caused by the poorly resolved phylogeny of Endomicrobiaceae in 16S rRNA-based studies ( 10 , 11 ). Moreover, the family comprises only a single isolate, Endomicrobium proavitum ( 25 ), and a few endosymbionts with sequenced genomes ( 26 – 28 ). While the free-living Endomicrobium proavitum and the endosymbiotic Candidatus Endomicrobium trichonymphae share a purely fermentative metabolism, they differ in genome size and guanine-cytosine (GC) content and in their metabolic capacities, such as the uptake and activation mechanisms of their major energy substrates, suggesting that the substantial gene loss in the endosymbiont was accompanied by acquisition of novel functions ( 29 ). In the absence of genomic information on other representatives of the family, it remains open whether endosymbiosis in Endomicrobiaceae arose more than once and how free-living lineage(s) adapted to the intracellular lifestyle. To address these questions, we reconstructed more than 1,700 bacterial genomes from gut metagenomes of 48 termite species, representing all major families, including all subfamilies of the flagellate-free Termitidae. This effort yielded 67 novel metagenome-assembled genomes (MAGs) of Endomicrobiaceae . Together with previously published genomes, including numerous MAGs from termites ( 30 ), ruminants ( 31 – 34 ), and anaerobic bioreactors ( 35 ) that had not been analyzed to date, this provided a total of 106 genomes. We used this comprehensive data set to analyze the phylogeny of Endomicrobiaceae and the metabolic potential of the individual lineages, with the aims of identifying the distribution of putative endosymbionts and tracing the loss and gain of relevant functions during the evolutionary history of Endomicrobiaceae .",
"discussion": "DISCUSSION The discovery of the first sequences of the “Termite group 1” ( 54 , 55 ) and their association with termite gut flagellates ( 5 , 7 ), the phylogenetic origin of Endomicrobia, and the relationship between free-living and endosymbiotic lineages remained unresolved. Our study revealed that host-associated Endomicrobiaceae originated among aquatic ancestors that subsequently colonized the intestinal tract of insects and ruminants. Although most lineages represent free-living gut commensals, a common ancestor of the genus Endomicrobiellum engaged in an intracellular symbiosis with termite gut flagellates. Progressive genome reduction led to a breakdown of glycolysis and many biosynthetic pathways, which was compensated by the acquisition of new functions via horizontal gene transfer. In some subgroups, the apparent loss of all biosynthetic capacities that could potentially benefit the flagellate host indicates the decline of an originally mutualistic relationship. Phylogenetic diversity of Endomicrobiaceae The presence of two lineages from anaerobic digestors ( Proruminimicrobium and Praeruminimicrobium ) in a position ancestral to the host-associated members of the ruminant clade ( Ruminimicrobium and Ruminimicrobiellum ) corroborates the origin of the family Endomicrobiaceae among lineages of aquatic bacteria in the order Endomicrobiales ( 49 ). Although the environmental origin of the ruminant clade is supported also by 16S rRNA-based diversity data, the origin of the termite-associated clade remains unresolved (Fig. S3). Previous 16S-rRNA-based studies detected members of Endomicrobiaceae also in other insects, such as cockroaches ( 7 , 10 , 11 ) and a scarab beetle larva ( 56 ), but, with the exception of lower termites, always in low abundance ( 11 , 57 , 58 ). This explains the absence of Endomicrobiaceae among MAGs from cockroaches ( 59 , 60 ) and the low recovery from higher termites ( Fig. 2 ). Notably, the sister position of the 16S rRNA genes of Endomicrobiaceae from cockroaches to the genus Endomicrobium is only poorly supported (Fig. S3) and remains to be verified with phylogenomic data. The 16S rRNA gene tree documents an affinity of Ca . Endomicrobium superficiale, the ectosymbionts of Trichonympha magna from Porotermes adamsoni , and spirotrichosomid flagellates from the related Stolotermes victoriensis ( 24 ), to the genus Ectomicrobium , but in the absence of a genome sequence, this assignment must remain tentative. The genus Endomicrobiellum comprises all endosymbiotic representatives that were localized in previous studies ( 5 – 7 , 23 , 61 ). Compared to other genera of Endomicrobiaceae , their genomes are strongly reduced, which is in agreement with the endosymbiotic nature of the entire genus. Notably, also members of the genus Ectomicrobium , which are associated with gut flagellates ( 24 ), have slightly smaller genomes than most other family members, including their sister genus, Parendomicrobium (Table S4). The genera Ruminimicrobium and Ruminimicrobiellum comprise MAGs recovered from large-scale studies of ruminal microbiota of sheep, cows, deer, and goats ( 31 – 34 , 62 ). It has been speculated that Endomicrobiaceae present in the rumen are associated with ciliates ( 9 ). This is supported by a significantly smaller genome size in members of the genus Ruminimicrobiellum compared to most other lineages of the family (Table S4), and by our observation that 16S rRNA genes of Endomicrobiaceae recovered from capillary-picked suspension of rumen ciliates ( Isotricha spp.) fall into the radiation of this genus (U. S. Mies and A. Brune, unpublished data; see Fig. S3). Therefore, it is possible that an association with protists has evolved independently in three lineages of Endomicrobiaceae . Genome erosion and acquisition of new functions in Endomicrobiellum The fermentative energy metabolism of Endomicrobium proavitum ( 29 ) is shared by all members of Endomicrobiaceae . Unlike several other lineages of the class Elusimicrobia that encode respiratory chains for oxygen and other inorganic electron acceptors ( 63 ), all members of Endomicrobiaceae are strict anaerobes that activate glucose to Glc6P with glucokinase and/or a PTS, and ferment Glc6P via the EMP pathway to ethanol, lactate, acetate, hydrogen, and CO 2 . All members of the endosymbiotic genus Endomicrobiellum , including the previously characterized Em. trichonymphae ( 26 , 27 , 29 ), lack the ABC transporter and often also the PTS for the uptake of glucose. These gene losses are compensated by two organophosphate:phosphate antiporters (OPA1 and OPA2) that were independently acquired by different subclades and allow the direct uptake of sugar phosphates from the cytoplasm of their respective hosts. Although annotated as Glc6P and Gly3P transporters, both OPA1 and OPA2 fall into phylogenetic clades without biochemically characterized representatives ( Fig. 5 ); therefore, their exact substrates remain uncertain. Nevertheless, either of these substrates would serve to link the energy metabolism of endosymbiont and host, as observed in other bacteria with an intracellular lifestyle (e.g., Listeria and Chlamydia [ 64 , 65 ]). Here, the switch to sugar phosphates is considered an adaptation to the low concentrations of free glucose in eukaryotic cells that provides an additional, energetic advantage because no ATP is needed for substrate activation. It is not clear whether the acquisition of OPAs by members of the genus Endomicrobiellum was a predisposition for an intracellular lifestyle, as in the ancestors of Chlamydia ( 65 ), or a response to the conditions in their intracellular habitat. In any case, these transporters may be the only means to obtain an energy substrate for those members of the genus Endomicrobiellum that lack both glucokinase and PTS (lineages H and I). The absence of enolase in several basal lineages with an otherwise complete glycolytic pathway is rather perplexing. The same situation has been reported for a variety of host-associated bacteria, including Treponematales and Bacteroidales symbionts of termite gut flagellates ( 22 , 66 ) and several Saccharibacteria ( 67 , 68 ) and Clostridiales from mammals ( 69 ). Because a methylglyoxylate shunt, which allows circumvention of the enolase function in the rumen bacterium Butyrivibrio proteoclasticus ( 70 ), is absent from all Endomicrobiaceae , it remains unclear how members of these lineages generate the phosphoenolpyruvate required, e.g., by PTS. The fermentative metabolism of glucose and other energy substrates requires regeneration of the reduced cofactors formed in the oxidative part of the pathway. Although most members of Endomicrobiaceae encode homologs of Ldh and/or AdhE and at least one hydrogenase, the lineages of Endomicrobiellum without a functional glycolysis (lineages H and I) also lack the enzymes that reoxidize NADH and reduced ferredoxin, indicating that they lost the capacity for sugar fermentation. It is possible that they conserve energy by oxidative decarboxylation of amino-acid-derived 2-oxoacids to the corresponding acyl-CoA esters, as in Elusimicrobium minutum ( 71 ) and in hyperthermophilic archaea ( 72 , 73 ). An interesting case is presented by the MAGs that possess an ATP/ADP antiporter, which enables them to import ATP directly from the host. This NTT was independently acquired from uncultured alphaproteobacteria (orders UBA3830 and RUG11792) by Endomicrobiellum lineage D and several MAGs from other lineages. ATP/ADP antiporters and other NTTs are regarded as a hallmark of parasitism and are common in intracellular parasites of amoebae, such as Chlamydia and Rickettsia , where they are used by the parasites to obtain energy and nucleotide triphosphates for their growth and replication from their respective hosts ( 65 , 74 ). However, NTTs are found also in free-living, non-parasitic bacteria ( 75 ); therefore, it remains open to speculation whether members of Endomicrobiellum have evolved into energy parasites or still provide a function that is beneficial for their flagellate hosts. Host–symbiont interactions The intracellular location provides endosymbionts with a protective barrier against environmental forces. At the same time, an intact cell envelope may also hinder horizontal gene transfer and the exchange of metabolites between symbiont and host, which explains why the pathways for the synthesis of peptidoglycan and an outer membrane are frequently lost during genome erosion ( 76 ). Members of the genera Endomicrobiellum retained the capacity to synthesize a murein sacculus, LPS, and several outer membrane proteins (OMPs), which suggests that their cell envelope is still intact. However, the absence of the bifunctional transglycosylase/transpeptidase (MrcB) and d -alanine: d -alanine ligase (Ddl) in several subclades of Endomicrobiellum may affect the integrity of their cell wall ( 77 ), which may facilitate metabolite exchange between endosymbionts and host. By default, an intracellular bacterium must synthesize all amino acids and vitamins that it cannot acquire from its host. Therefore, any deleterious mutations in its biosynthetic machinery must be compensated by an uptake mechanism that was either present already in the free-living ancestors of the endosymbiont or was later acquired by horizontal gene transfer. These two scenarios are exemplified by the ancestral presence of an uptake system for a proline transporter (ProT) and the acquisition from other bacteria of transporters for serine (SdaC), glutamate (GltS), and aromatic amino acid (AroP). These latter transporters are present in almost all lineages of Endomicrobiellum and are most closely related to homologs from unrelated bacteria present in termite guts. The acquisition of amino acid transporters and conservation of many biosynthetic pathways in the genus Endomicrobiellum are in agreement with the long-standing hypothesis that flagellate endosymbionts provide essential amino acids to their respective hosts. Hongoh et al. ( 26 ) found that despite the severe genome erosion of Em. trichonymphae , pathways for the biosynthesis of aromatic amino acids are mostly conserved and several key genes are duplicated, which suggests that the endosymbionts provide the flagellate with aromatic amino acids. However, the capacity to synthesize tryptophane, tyrosine, and phenylalanine is not conserved in all lineages of the genus Endomicrobiellum . The same is true for biosynthetic pathways for proline and histidine. All members of the genus lack glutamine synthetase (GlnA), indicating a general dependence of the endosymbionts on the provision of glutamine by the host cell. It is unclear how this is accomplished because all members of Endomicrobiellum lack a glutamine transporter (GlnT). Lineage I lost the capacity to synthesize almost all amino acids, suggesting that they have lost their original function and may no longer play an essential role in the symbiosis. This applies also to a putative role of the endosymbionts in the provision of specific vitamins because the distribution of the corresponding biosynthetic pathways among different lineages is not conclusive ( Fig. 3 ). A function that is frequently encountered among bacterial symbionts of eukaryotic hosts is dinitrogen fixation. Examples among termite gut flagellates are the endosymbiotic Azobacteroides pseudotrichonymphae ( 78 ) and flagellate-associated bacteria in the gut of dry-wood termites ( 45 , 79 ). Unlike the homolog of nitrogenase group VI, which occurs in other lineages of the phylum Elusimicrobiota and is potentially involved in tetrapyrrole modification ( 63 ), the group IV homologs encountered in members of the genera Ectomicrobium and Endomicrobium were shown to encode a functional nitrogenase ( 25 ). All other lineages of Endomicrobiaceae , including the genus Endomicrobiellum , lack nitrogenase and cannot contribute to dinitrogen fixation. Horizontal gene transfer among flagellate symbionts drives convergent evolution There is abundant evidence supporting a convergent evolution among the endosymbionts of termite gut flagellates. Significant genome reduction has been reported for endosymbiotic representatives of Acutalibacteraceae ( 53 ), “ Ancillulaceae ” ( 21 ), Azobacteroidaceae ( 78 ), “ Adiutricaceae ” ( 19 ), and Treponemataceae ( 22 ). The same trend seems to be present also among the putatively endosymbiotic members of Rickettsiales and related Alphaproteobacteria (orders RUG11792 and UBA3830), Mycoplasmatales , and Opitutales (families LL51 and UBA9783) ( Fig. 7 ). Notably, endosymbiotic representatives of these lineages not only co-occur in the same termites but also often co-localize in the same flagellate species, which provides ample opportunity for horizontal gene transfer. Horizontal gene transfer among flagellate symbionts is most evident in the case of sugar phosphate transporters (OPAs), which are present in almost all lineages of established and putative flagellate endosymbionts except intracellular treponemes (Table S7). One homolog (OPA2) of Endomicrobiellum clusters among a clade of OPAs from termite-associated MAGs; the other homolog (OPA1) has close relatives among distantly related Elusimicrobiaceae associated with arthropods and ruminants but is absent from all other lineages of the Elusimicrobiota phylum. Their placement among the radiation of homologs from termite gut MAGs of other phyla suggests that both Endomicrobiaceae and Elusimicrobiaceae acquired their OPAs independently of each other from the same donor. Another case of convergent evolution involves the nucleotide antiporters (NTTs), which were independently acquired by Endomicrobiellum from putative flagellate endosymbionts in the alphaproteobacterial orders UBA3830 and RUG11792 (see \"Genome erosion and acquisition of new functions in Endomicrobiellum \"). The transporter of RUG11792 clusters also with a homolog from endosymbionts of the family Acutalibacteraceae ( Oscillospirales ) ( 53 ). All these lineages of endosymbionts have been localized in flagellates of the genus Trichonympha , suggesting that the genes encoding NTTs may be frequently transferred among the endosymbionts of termite gut flagellates. By contrast, the NTTs of the putatively endosymbiotic Opitutales , which comprise the endonuclear Candidatus Nucleococcus from a termite gut flagellate ( 17 ) and several ciliate endosymbionts ( 80 ), are embedded among lineages from other environments, suggesting an ancestral presence in Opitutales ( Fig. 6 ). A hexuronate pathway occurs in the endosymbiotic genera Endomicrobiellum , Ca . Ancillula, and Azobacteroides ( 19 , 21 , 78 ) but also in other members of the family Azobacteroidaceae . Although Endomicrobiaceae apparently acquired the pathway from a clade of Lachnospiraceae associated with termites and ruminants, “ Ancillulaceae ” acquired it from Lactobacillaceae (Fig. S16). The origin of the hexuronate pathway in several putatively free-living Endomicrobiaceae differs from that in the genus Endomicrobiellum (Fig. S16). The capacity of these endosymbionts to degrade hexuronates may represent a convergent adaptation to their intracellular niche, because the hydrolysis of hemicelluloses in the digestive vacuoles of their flagellate hosts should yield large amounts of hexuronates ( 81 ). The loss of pathways for the biosynthesis of amino acids in the genus Endomicrobiellum was compensated by the acquisition of several amino acid transporters absent from the ancestral, free-living lineages ( Fig. 4 ). In the case of AroP, the homolog from Endomicrobiellum originates among homologs of “ Adiutricaceae ” (Fig. S17), whereas those of other Endomicrobiaceae cluster with homologs from putative endosymbionts in the order UBA3830. Again, the respective lineages co-colonize the same flagellates ( Trichonympha spp.; [ 17 ]), providing further evidence that the convergent evolution of endosymbionts of termite gut flagellates is driven by horizontal gene transfer. Conclusions Our comparative genomic analysis of Endomicrobiaceae provided new insights into the evolutionary processes underlying symbiogenesis in intracellular symbionts of termite gut flagellates. The transition from a free-living to an intracellular lifestyle not only restricts gene flow but also affects the ability to take up substrates and other nutrients from the environment. Members of the genus Endomicrobiellum possess functions that either represent adaptations to their intracellular niche or serve to compensate for gene losses during ongoing genome erosion. Some functions represent predispositions to an intracellular lifestyle that were present already in the common ancestor of Endomicrobiaceae , whereas others were apparently acquired by horizontal gene transfer. The acquisition of similar traits also by other established or putative flagellate symbionts underscores the importance of horizontal gene transfer from other gut microbiota in the convergent evolution of endosymbiotic lineages. The lack of biosynthetic functions common to all members of the genus Endomicrobiellum suggests that their original role in the symbiosis became compromised during the progressive genome reduction. It is possible that certain lineages now just represent a metabolic burden to their host. The acquisition of an NTT may indicate the transition from a mutualistic to a parasitic relationship, but it may also be the last straw to ensure survival for a beneficial endosymbiont whose energy metabolism became compromised owing to genome erosion. It is likely that dysfunctional members of Endomicrobiellum will be eventually replaced by secondary symbionts that co-occur in the same flagellates, as shown for other primary endosymbionts of eukaryotic hosts that have experienced severe gene losses ( 1 , 82 ). Taxonomy Most members of the family Endomicrobiaceae belong to genus-level lineages that are either unclassified or require reclassification. The presence of both high-quality genomes and 16S rRNA gene sequences for most lineages allow the proposal of new taxa under the Code of Nomenclature of Prokaryotes Described from Sequence Data (SeqCode) ( 83 ). The new names and new combinations, together with the designated type material, are listed in Table 1 . The authors of previously proposed Candidatus names were assigned as descriptors of the corresponding new taxa. The full protologs including etymology, description, and other information are given in the supplemental material (Text S1). TABLE 1 New genera in Endomicrobiaceae proposed under SeqCode, their designated type species, and other species assigned to the respective genus a Genus Type species Other species in the genus Endomicrobium ( 25 ) Endomicrobium proavitum ( 25 ) Endomicrobium embiratermitis sp. nov., E. labiotermitis sp. nov., E. macrotermitis sp. nov., E. neocapritermitis sp. nov., E. procryptotermitis sp. nov. Endomicrobiellum gen. nov. Endomicrobiellum trichonymphae sp. nov. Endomicrobiellum africanum sp. nov., Em. agilis sp. nov., Em. basalitermitum sp. nov., Em. calcaritermitis sp. nov., Em. calonymphae sp. nov., Em. cryptotermitis sp. nov., Em. cubanum sp. nov., Em. dinenymphae sp. nov., Em. glyptotermitis sp. nov., Em. guadaloupense sp. nov., Em. incisitermitis sp. nov., Em. mastotermitis sp. nov., Em. meruensis sp. nov., Em. neotermitis sp. nov., Em. porotermitis sp. nov., Em. pyrsonymphae sp. nov., Em. roisinitermitis sp. nov., Em. siamense sp. nov. Ectomicrobium gen. nov. Ectomicrobium neotermitis sp. nov. Parendomicrobium gen. nov. Parendomicrobium reticulitermitis sp. nov. Parendomicrobium porotermitis sp. nov. Proendomicrobium gen. nov. Proendomicrobium guianensium sp. nov. Ruminimicrobium gen. nov. Ruminimicrobium bovinum sp. nov. Ruminimicrobiellum gen. nov. Ruminimicrobiellum bubulum sp. nov. Ruminimicrobiellum caprinum sp. nov., R. ovillum sp. nov., R. tauri sp. nov. Praeruminimicrobium gen. nov. Praeruminimicrobium purgamenti sp. nov. Proruminimicrobium gen. nov. Proruminimicrobium quisquiliarum sp. nov. \n \n a \n \n The nomenclatural types (genomes) of all new species and their full protologs can be found in the supplemental material (Text S1)."
} | 7,349 |
36905370 | PMC10128138 | pmc | 5,921 | {
"abstract": "Abstract \n Corynebacterium glutamicum experiences a transient iron limitation during growth in minimal medium, which can be compensated by the external supplementation of protocatechuic acid (PCA). Although C. glutamicum is genetically equipped to form PCA from the intermediate 3‐dehydroshikimate catalysed by 3‐dehydroshikimate dehydratase (encoded by qsuB ), PCA synthesis is not part of the native iron‐responsive regulon. To obtain a strain with improved iron availability even in the absence of the expensive supplement PCA, we re‐wired the transcriptional regulation of the qsuB gene and modified PCA biosynthesis and degradation. Therefore, we ushered qsuB expression into the iron‐responsive DtxR regulon by replacing the native promoter of the qsuB gene by the promoter P ripA and introduced a second copy of the P ripA ‐ qsuB cassette into the genome of C. glutamicum . Reduction of the degradation was achieved by mitigating expression of the pcaG and pcaH genes through a start codon exchange. The final strain C. glutamicum IRON+ showed in the absence of PCA a significantly increased intracellular Fe 2+ availability, exhibited improved growth properties on glucose and acetate, retained a wild type‐like biomass yield but did not accumulate PCA in the supernatant. For the cultivation in minimal medium C. glutamicum IRON+ represents a useful platform strain that reveals beneficial growth properties on different carbon sources without affecting the biomass yield and overcomes the need of PCA supplementation.",
"conclusion": "CONCLUSION This study demonstrates the successful re‐organization of transcriptional control by applying a naturally inspired approach. C. glutamicum might not encounter iron restricted conditions in its natural habitat, but does so in monoseptical cultivations. Consequently, by controlling the expression of the endogenous PCA synthesis gene qsuB in response to the iron availability and by mitigating its degradation, we ended up with a strain that exhibits a significantly higher intracellular Fe 2+ concentration, superior growth properties on different substrates and retains an identically high biomass yield as the WT. Hence, C. glutamicum IRON+ represents an interesting platform for further engineering approaches in an academic as well as industrial environment, because it overcomes the need of undesirable PCA supplementation, which is expensive on the one hand and provides another carbon source on the other.",
"introduction": "INTRODUCTION \n Corynebacterium glutamicum is a non‐pathogenic and facultatively anaerobic soil bacterium, which was originally isolated in 1957 as a natural l ‐glutamate producer (Kinoshita et al., 1957 ). Nowadays it is an established workhorse in industrial biotechnology for the large‐scale production of several amino acids such as l ‐lysine and l ‐glutamate in million ton scale per year (Becker et al., 2018 ; Ikeda & Takeno, 2013 ; Wendisch, 2020 ). The high robustness, the versatile metabolism and the GRAS status are beneficial characteristics of this Gram‐positive bacterium for the application in an industrial environment. A reliable genetic engineering toolbox to tailor the metabolism is available and has been applied to expand the product spectrum of C. glutamicum towards various amino acids, alcohols, aldehydes, diamines, organic acids, aromatic compounds and others (Becker et al., 2018 ; Kogure & Inui, 2018 ; Wang et al., 2020 ; Wendisch et al., 2018 ; Wieschalka et al., 2013 ). The wild type (WT) of C. glutamicum is not prototrophic and growth in minimal medium essentially relies on supplementation of the vitamin biotin (Peters‐Wendisch et al., 2014 ). Although not of vital importance, the addition of small amounts of iron chelators such as protocatechuic acid (PCA) or catechol haven been shown to significantly improve the growth properties of C. glutamicum in minimal medium, that is reduction of the lag phase and constant exponential growth (Liebl et al., 1989 ). Consequently, PCA became a standard ingredient of the widely used CgXII minimal medium (Keilhauer et al., 1993 ; Unthan et al., 2014 ). PCA (3,4‐dihydroxybenzoic acid) is a constitutional isomer of the more frequently employed siderophore precursor 2,3‐dihydroxybenzoic acid (2,3‐DHBA). Likewise, it can provide the functional moiety of such iron binding molecules (e.g. petrobactin) (Barbeau et al., 2002 ). A common response of many bacteria to iron limited conditions is the secretion of high affinity siderophores in order to make poorly soluble ferric iron (Fe 3+ ) available. But also the catecholate compounds 2,3‐DHBA and PCA themselves were detected under iron limitation in the culture supernatants of Bacillus subtilis , Paracoccus denitrificans and Bacillus anthracis , respectively (Neilands, 1981 ; Peters & Warren, 1968 ; Tait, 1975 ), where they can increase the iron availability by chelation of Fe 3+ and the chemical reduction to Fe 2+ (Müller et al., 2020 ). The iron homeostasis of C. glutamicum was extensively studied with the focus on the transcriptional regulation, iron storage and mobilization, as well as on the utilization of alternative iron sources (Blombach et al., 2013 ; Brune et al., 2006 ; Follmann et al., 2009 ; Frunzke et al., 2011 ; Keppel et al., 2019 ; Küberl et al., 2016 , 2020 ; Müller et al., 2020 ; Wennerhold et al., 2005 ; Wennerhold & Bott, 2006 ). The master regulator of iron homeostasis in C. glutamicum is DtxR which in response to the intracellular Fe 2+ concentration controls the transcription of genes encoding proteins for iron acquisition, storage and mobilization as well as proteins responsible for iron–sulphur cluster assembly. Moreover, DtxR represses under iron excess the transcription of the ripA gene coding for the transcriptional regulator of iron proteins A (RipA). Under iron limitation, RipA represses the transcription of genes encoding enzymes such as aconitase or succinate dehydrogenase. So far, it is not clear, how iron is initially transported into the cells when C. glutamicum grows as monoculture, because siderophore biosynthetic genes were not identified in the genome although a large repertoire of siderophore uptake systems is available (Brune et al., 2006 ; Frunzke & Bott, 2008 ; Wennerhold & Bott, 2006 ). Interestingly, C. glutamicum is genetically equipped to form PCA, but unlike in other bacteria, PCA biosynthesis nor its degradation is regulated in response to the iron availability (Brune et al., 2006 ; Kalinowski et al., 2003 ; Shen et al., 2012 ; Wennerhold & Bott, 2006 ). During growth on glucose PCA is synthesized from 3‐dehydroshikimate, an intermediate of the shikimate pathway, catalysed by the 3‐dehydroshikimate dehydratase (QsuB) encoded by qsuB (Figure 1A ). C. glutamicum also features the β‐ketoadipate pathway to utilize several aromatic compounds such as PCA as sole carbon and energy source (Kubota et al., 2014 ; Merkens et al., 2005 ; Shen & Liu, 2005 ; Teramoto et al., 2009 ; Zhao et al., 2010 ). PCA degradation is initiated by the PCA 3,4‐dioxygenase (PcaGH), which consists of two subunits encoded by pcaG and pcaH located in the pcaHGBC operon (Figure 1 ) (Zhao et al., 2010 ). Recently, we investigated biomass formation of the C. glutamicum WT in PCA‐deficient CgXII medium and observed a biphasic growth behaviour caused by a transient iron limitation (Müller et al., 2020 ). This growth retardation could be compensated by either the supplementation of PCA to the CgXII medium or by aeration of the bioreactor with an increased proportion of CO 2 in the inlet air. It turned out that the presence of CO 2 /HCO 3 \n − accelerates the chemical reduction of poorly soluble ferric iron (Fe 3+ ) to biologically active ferrous iron (Fe 2+ ) through phenolic acids and catechols including PCA (Müller et al., 2020 ). FIGURE 1 The central carbon metabolism of C. glutamicum including PCA biosynthesis via the shikimate pathway and its degradation via the β‐ketoadipate pathway (A). Metabolic engineering strategies to place the PCA synthesis under control of the master regulator of the iron homeostasis, DtxR (B): (i) replacement of the native qsuB expression control with the ripA promoter (P ripA ), which is repressed by DtxR at Fe 2+ excess, (ii) integration of an additional P ripA ‐ qsuB copy in the CgLP4 locus (Lange et al., 2017 ), (iii) increase of the PCA pool by mitigating expression of the pcaG and pcaH genes through the start codon exchange ATG → GTG. Abbreviations: 3‐DHS, 3‐dehydroshikimate; 6PG, 6‐phosphogluconate; AC‐CoA, acetyl‐CoA; CHO, chorismate; CMA, β‐carboxy‐ cis, cis ‐muconate; DAHP, 3‐desoxyarabinoheptulosanat‐7‐phosphate; E4P, erythrose‐4‐phosphate; GLC, glucose; GLC‐6P, glucose‐6‐phosphate; PCA, protocatechuic acid; PcaGH, PCA 3,4‐dioxygenase; PEP, phosphoenolpyruvate; PYR, pyruvate; QsuB, 3‐dehydroshikimate dehydratase; SUCC‐CoA, succinyl‐CoA. To improve the growth properties of C. glutamicum in minimal medium without PCA supplementation, we installed an iron‐responsive PCA biosynthesis by exchange of the native qsuB promoter with the ripA promoter (P ripA ), which is repressed by DtxR at Fe 2+ excess, integrated of a second copy of P ripA ‐ qsuB into the genome and increased the PCA pool by mitigating expression of the pcaG and pcaH genes (Figure 1B ).",
"discussion": "DISCUSSION Although genetically equipped for the biosynthesis of PCA, C. glutamicum does not regulate the expression of the responsible genes in response to the intracellular iron availability (Brune et al., 2006 ; Wennerhold & Bott, 2006 ). We have recently shown, that C. glutamicum WT experiences a transient iron limitation in CgXII medium lacking PCA, which results in a growth delay during the initial phase of the cultivation (Müller et al., 2020 ). The external supplementation of PCA in CgXII minimal medium is a commonly employed strategy (Keilhauer et al., 1993 ), but might be undesired because of being expensive and providing an additional carbon source (Graf et al., 2019 ; Shen et al., 2012 ). In this study we artificially ushered the control of qsuB expression into the iron‐responsive DtxR regulon. This approach was inspired by nature, reflecting that other organisms regulate their biosynthesis of PCA or the structural analogue 2,3‐DHBA in response to the intracellular iron availability in turn (Garner et al., 2004 ; Neilands, 1981 ; Peters & Warren, 1968 ; Tait, 1975 ). We replaced the native promoter of the qsuB gene by P ripA and introduced a second copy of the P ripA ‐ qsuB cassette into the genome of C. glutamicum , which indeed improved the growth properties but not to the level of PCA‐supplemented cultures. It was necessary to additionally reduce the degradation of PCA in order to enhance the growth phenotype further. This step is plausible since C. glutamicum possesses a functional β‐ketoadipate pathway and can efficiently utilize several aromatic compounds such as PCA (Shen et al., 2012 ). Accordingly, Okai et al. ( 2017 ) and Kallscheuer and Marienhagen ( 2018 ) engineered C. glutamicum for the production of hydroxybenzoic acids and showed the necessity of an inactive degradation of PCA for efficient overproduction of this aromatic compound. With our previously designed iron reporter strain C. glutamicum FEM3 (Müller et al., 2020 ) we show, that the improved growth properties in minimal medium caused by the introduced genetic modifications correlates with an increased intracellular Fe 2+ availability. Interestingly, the beneficial growth properties of C. glutamicum IRON+ could also be exploited when acetate was used as sole carbon and energy source. Recently, Graf et al. ( 2019 ) evolved a fast‐growing variant of C. glutamicum WT (designated as EVO5), which proliferates independently of PCA. Genome re‐sequencing of C. glutamicum EVO5 identified overall 10 mutations with three mutations located in the genes coding for the transcriptional regulators DtxR, RipA and RamA. Re‐engineering of the ramA mutation in C. glutamicum WT improved the growth rate in PCA‐deficient minimal medium on glucose, however, led to significantly impaired growth on acetate (Graf et al., 2019 ). In contrast, C. glutamicum IRON+ showed also in CgXII medium with acetate improved growth properties. We could not simply compensate the transient iron limitation during the initial growth phase on acetate with C. glutamicum IRON+ but the growth rate was 42% higher than that of the WT during the entire cultivation. A reduction of the growth rate has been reported previously (Wendisch et al., 2000 ) and might be due to an uncoupled membrane potential during growth on weak acids (Axe & Bailey, 1995 ; Baronofsky et al., 1984 ; Kiefer et al., 2020 ). In this context it is not clear, whether PCA can alleviate the growth limiting stress conditions through its inherent redox chemistry (i.e. acting as an electron shuttle, Perron & Brumaghim, 2009 ) or whether the enzymatic stress‐response benefits from the increased Fe 2+ availability of the strain. Given the fact, that acetate gains increasing importance as an alternative carbon and energy source for microbial production processes (Kiefer et al., 2020 ; Merkel et al., 2022 ; Schmollack et al., 2023 ), future research needs to address this effect as well as the performance of C. glutamicum IRON+ when exposed to different environmental stresses. Although the qsuB gene is monocistronically transcribed in ATCC 13032 (Pfeifer‐Sancar et al., 2013 ) it is located in a gene cluster with qsuC and qsuD encoding dehydroquinate dehydratase and quinate/shikimate dehydrogenase, respectively. Consequently, the replacement of the native promoter by P ripA might have increased qsuCD expression, too, which could additionally increase carbon flux towards the QsuB substrate 3‐dehydroshikimate (Kubota et al., 2014 ; Teramoto et al., 2009 ). Notably, not only PCA has a growth promoting effect. Also (di‐)phenolic compounds, such as catechol (Liebl et al., 1989 ), ferulic acid and vanillin facilitate growth of C. glutamicum (Siebert et al., 2021 ). We showed that functionalized aromatic compounds, with a mix of amino and hydroxyl groups or two adjacent hydroxyl groups chelate iron and/or reduce Fe 3+ , which improves the overall intracellular iron availability (Müller et al., 2020 ). And even indole was found to reduce extracellular Fe 3+ (Walter et al., 2020 ). Therefore, when feedstocks such as lignocellulosic hydrolysates, which contain diphenolic compounds, are supplemented to the minimal medium or strains, which overproduce such molecules are utilized (Kallscheuer & Marienhagen, 2018 ; Kim et al., 2022 ; Okai et al., 2016 , 2017 ), the application of C. glutamicum IRON+ might not be required. The fact that the soil bacterium C. glutamicum naturally encounters this substrate mixture might also explain the greater iron availability in its natural habitat, and why qsuB expression is not evolutionarily placed under control of the iron homeostasis regulators. However, the iron acquisition mode of C. glutamicum is not known, yet. An interesting future research objective is to differentiate whether the entire Fe 3+ ‐chelates are internalized by C. glutamicum prior to the release of Fe 2+ (i.e. via one of the siderophore‐specific transport systems) or whether the chemical reduction takes place spontaneously in the extracellular environment and iron is then taken up via Fe 2+ ‐specific transport proteins that are also annotated in the genome (Frunzke & Bott, 2008 ). The final strain C. glutamicum IRON+ performs well at different cultivation scales, does not accumulate quantifiable levels of PCA in the culture supernatant and maintains an equal Y X/S as the WT. By that, C. glutamicum IRON+ features an interesting genetic basis that could be further engineered for production purposes and represents a neat host strain for laborious screening approaches, as well as large‐scale fermentations. Moreover, the introduced genetic modifications might be combined with the engineered biotin prototrophic C. glutamicum strains (Ikeda et al., 2013 ; Peters‐Wendisch et al., 2014 ) to obtain a prototrophic platform strain, which provides reliable growth even at low initial biomass concentrations."
} | 4,103 |
29046540 | PMC5649373 | pmc | 5,923 | {
"abstract": "Horizontal gene transfer mediated by broad-host-range plasmids is an important mechanism of antibiotic resistance spread. While not all bacteria maintain plasmids equally well, plasmid persistence can improve over time, yet no general evolutionary mechanisms have emerged. Our goal was to identify these mechanisms, and to assess if adaptation to one plasmid affects the permissiveness to others. We experimentally evolved Pseudomonas sp. H2 containing multi-drug resistance plasmid RP4, determined plasmid persistence and cost using a joint experimental-modeling approach, resequenced evolved clones, and reconstructed key mutations. Plasmid persistence improved in fewer than 600 generations because the fitness cost turned into a benefit. Improved retention of naive plasmids indicated that the host evolved towards increased plasmid permissiveness. Key chromosomal mutations affected two accessory helicases and the RNA polymerase β-subunit. Our and other findings suggest that poor plasmid persistence can be caused by a high cost involving helicase-plasmid interactions that can be rapidly ameliorated.",
"discussion": "DISCUSSION The role of helicases in plasmid stabilization Strikingly, our and at least three other studies that evolved different host-plasmid pairs now suggest that maladapted interactions between plasmids and host-encoded helicases adversely affect plasmid cost and persistence. Moreover, these interactions can often be improved by single mutations, suggesting we are zooming in on a potential general mechanism of bacterial adaptation to plasmids. First, in P. aeruginosa PAO1, loss-of-function mutations in a putative accessory helicase with a UvrD-like helicase C-terminal domain ameliorated the cost of a small non-mobilizable plasmid ( 16 ). This initial cost was due to the plasmid’s replication initiation (Rep) protein triggering an SOS response in the ancestral host, and the helicase knockout mutation reduced Rep expression ( 11 ). Second, experimental evolution of an IncP-1β mini-replicon in Shewanella oneidensis MR-1 improved plasmid cost and persistence through loss of a helicase (DnaB) binding domain in the plasmid’s Rep protein, reducing the protein’s affinity for DnaB ( 14 ). This likely avoided an SOS response that may explain the high cost of the ancestral plasmid ( 20 ). Third, when we evolved that same plasmid in another host, it stabilized in two clones due to a SNP in either the dnaB promoter or a uvrD gene ( 18 ). Finally, in the present study SNPs affecting two accessory helicases again compensated for the cost of RP4, and improved the persistence of this and three other BHR plasmids. Helicases are involved in many aspects of DNA and RNA metabolism, such as replication (replicative helicases), and DNA repair, recombination, translocation, transcription, translation, and resolution of replication-transcription conflicts (accessory helicases) ( 31 , 32 , 33 ). Accessory helicases such as Xpd/Rad3 and UvrD, generally have variable C- and N-terminal accessory domains which determine their physiological specificity ( 34 ). Interestingly, UvrD and the Xpd/Rad3-like helicase DinG, have been shown to be upregulated as part of the SOS response induced by plasmid entrance and replication in a naïve host ( 35 , 36 , 37 , 38 ). In E. coli , the UvrD homologue Rep helicase has been shown to interact with DnaB, acting as a second motor that improves replication fork movement on the chromosome ( 39 ). UvrD helicases have also been shown to ‘backtrack’ the RNA polymerase complex to slow down transcription, thus preventing the complex from colliding with the replication fork and causing dsDNA breaks ( 32 , 38 ). In our study, host adaptation to plasmid carriage was facilitated by two different mutations; one that likely changed the Xpd/Rad3-like structure, and one that likely affected UvrD abundance. It was interesting to see that our plasmid decreased uvrD transcript levels in the ancestor where it imposed a high cost, but not in the evolved strains, where it had become beneficial. This suggests that higher UvrD levels are needed for plasmid persistence. The mechanisms by which these mutations affected plasmid cost and persistence are currently not understood but the topic of future work. A simple explanation like a change in plasmid copy number can probably be excluded based on very similar plasmid sequence coverage for the ancestral and evolved genomes (data not shown). We postulate three not necessarily mutually exclusive models: the accessory helicases interact with DnaB (i) or the plasmid replication initiation protein TrfA (ii) to modulate plasmid replication efficiency, or (iii) mutations in the helicases ameliorate the fitness cost of plasmid RP4 through their regulatory function. Whatever the mechanism, further research should confirm that accessory helicases are involved in plasmid persistence across pathogens. The two accessory helicase genes in our Pseudomonas strain were likely acquired by HGT, consistent with previous findings for P. aeruginosa PAO1 ( 11 ), where it was proposed they caused genetic conflict with the plasmid. We intend to test whether these helicases hamper or improve persistence of various resident MDR plasmids in other strains, as this could aid the development of strategies aimed at slowing down the spread of antibiotic resistance in bacterial pathogens. Potential epistasis between helicase and RpoB mutations Stabilization of plasmid RP4 required not only mutations in loci selected in the presence of the plasmid, but also at least one mutation that seemed adaptive to the growth environment. The gene rpoB , which encodes the β-subunit of RNA polymerase (RNAP), was mutated across all sequenced plasmid-containing and control clones. It is thus the most likely candidate for epistatic interactions with the accessory helicases. The RNAP holoenzyme, consisting of five subunits, αI, αII, β, β′ and ω, together with the σ factor, is responsible for transcription ( 40 ). The β-subunit specifically, in addition to DNA binding, is involved in the modulation of transcription through interaction with σ factors ( 41 ) and DNA helicases ( 38 ). The rpoB mutation in question was secondary after the initial rpoB mutation that resulted in Rif resistance (RifR) in our ancestral strain. In the absence of the helicase mutations, these rpoB mutations did not improve plasmid persistence at all ( Fig. 2 ). Thus, they most likely affect it only through epistatic interaction with the helicase mutations. We propose two possible mechanisms of epistasis between the rpoB and helicase mutations that are not mutually exclusive. First, based on their location, the SNPs in rpoB likely compensated for the cost imposed by the initial RifR rpoB mutation ( Table S5 ) by ameliorating transcription efficiency ( 42 , 43 ). It is possible that without this compensatory mutation the helicase mutations were unable to significantly improve plasmid cost and persistence. Second, it is striking that accessory helicases can bind to the RNAP complex, in particular RpoB, to slow down transcription and regulate backtracking ( 32 , 38 ). Was there a need for mutations in RpoB to modify this physical interaction? Our secondary RpoB mutations are closer to the active site than to the helicase-binding residues, suggesting they may not affect helicase binding. They are also not close to the rpoB mutations that were shown to rescue viability of strains without accessory helicases ( 34 ). Future studies are needed to determine the mechanism by which the helicase and possibly rpoB mutations can transform a plasmid cost into a benefit for its bacterial host. In conclusion, to combat the spread and persistence of plasmid-mediated antibiotic resistance, novel therapeutic approaches are needed that target mechanisms that affect stable retention of MDR plasmids ( 24 ). To do so we need to understand which chromosomal gene products stabilize or destabilize MDR plasmids across bacterial species and how. Our study led to at least three important conclusions that may impact the way we tackle MDR plasmid spread: (i) bacteria can adapt to conjugative MDR plasmids by changing plasmid cost into benefit, resulting in greatly improved plasmid persistence; (ii) this can be due to mutations affecting helicases that initially impaired plasmid persistence, a recurring evolutionary pattern that may lead to new antimicrobial therapies; and (iii) bacterial adaptation to one plasmid can lead to generally improved plasmid permissiveness, enabling future retention of MDR plasmids. So far as we know, this is the first time that antibiotic exposure is shown to select for bacterial mutants with increased general permissiveness toward transmissible drug resistance plasmids. These mutations may threaten the efficacy of traditional antibiotic treatments even more than single drug resistance mutations, as adaptation of a pathogen to one plasmid may result in improved retention of other plasmid-encoded antibiotic resistance determinants."
} | 2,268 |
36619370 | PMC9813533 | pmc | 5,925 | {
"abstract": "Hydrogen is generally considered as an ideal non-polluting future energy carrier because it releases energy and water as a byproduct on combustion. Besides, hydrogen possesses the highest energy density on mass basis compared to any other fuel. However, hydrogen production in a sustainable and environmentally friendly way still remains a challenge. Recently, biohydrogen production from green microalgae has gained significant attention due to availability of the feedstock, which are environmentally friendly and renewable. Biohydrogen production from photosynthetic microalgae is attractive, however in the current context, it has a low yield, and an optimization of the affecting parameters including algae concentration, light intensity, culture medium, etc. is critical. In this study, biohydrogen was produced in laboratory from Euglena acus microalgae as it was locally available in Bangladesh. • The effect of two different culture mediums (i.e. sulfur-rich and sulfur-deprived TAP mediums) for microalgae cultivation and biohydrogen yield were studied. • Depending on the concentration of microalgae (50% and 75% by weight) in the medium solution ∼3 ml to 5 ml biohydrogen was obtained.",
"discussion": "Discussion Biohydrogen production from microalgae is influenced by a number of factors including illumination quality and intensity, pH of the culture, ambient temperature, chemical composition of the medium (for cultivation and H 2 production), substrate type and concentration etc. Furthermore, hydrogen yield can vary greatly for different microalgae species. Since biophotolysis is linked to photosynthesis reactions, adequate illumination by either sunlight or artificial light source is crucial for biohydrogen production. In general, an increase in light intensity increases H 2 yield. However, at higher light intensities also enhances O 2 production rate, which subsequently impedes H 2 yield. The correlation between H 2 yield and light intensity further relies on the culture age, gas phase and density of culture. As the culture grew old H 2 production rate declined while the maximum rate of photo-production was recorded at the starting of the stationary phase [27] . However, in the current study, two concentration levels of the microalgae in the medium were investigated while the other parameters remained fixed. The higher concentration provided a higher H 2 yield in this study. The effect of light intensity variation and culture pH on H 2 yield is planned to be investigated in the future work."
} | 630 |
25392084 | PMC4229684 | pmc | 5,926 | {
"abstract": "Driven by its importance in nature and technology, droplet impact on solid surfaces has been studied for decades. To date, research on control of droplet impact outcome has focused on optimizing pre-impact parameters, e.g. , droplet size and velocity. Here we follow a different, post-impact , surface engineering approach yielding controlled vectoring and morphing of droplets during and after impact. Surfaces with patterned domains of extreme wettability (high or low) are fabricated and implemented for controlling the impact process during and even after rebound —a previously neglected aspect of impact studies on non-wetting surfaces. For non-rebound cases, droplets can be morphed from spheres to complex shapes —without unwanted loss of liquid. The procedure relies on competition between surface tension and fluid inertial forces, and harnesses the naturally occurring contact-line pinning mechanisms at sharp wettability changes to create viable dry regions in the spread liquid volume. Utilizing the same forces central to morphing, we demonstrate the ability to rebound orthogonally-impacting droplets with an additional non-orthogonal velocity component. We theoretically analyze this capability and derive a We −.25 dependence of the lateral restitution coefficient. This study offers wettability-engineered surfaces as a new approach to manipulate impacting droplet microvolumes, with ramifications for surface microfluidics and fluid-assisted templating applications.",
"conclusion": "Conclusion This study presented an investigation into morphing and vectoring liquid micro-volumes through energetic droplet impact on wettability-engineered surfaces, with an emphasis on controlling post-impact behavior —a promising approach to tuning impact outcome ( e.g. , deposition or bounce). Both procedures rely on the competition between surface tension and fluid inertial forces, and have important ramifications for surface microfluidics and soft templating applications. The work demonstrated highly repeatable events wherein droplets were converted into tori without loss of liquid, as well as splitting and high-rate sampling. The results indicate that it is possible to morph and fractionalize droplet volumes using simple patterning procedures and commercially available materials. Droplet vectoring using wettability-patterned surfaces was also studied, with results demonstrating the capability to manipulate vectoring behavior by controlling We and asymmetric pattern construction. Notably, achieved lateral velocities were comparable to, if not greater than those previously reported in the literature for surface microfluidic devices. Particular emphasis was placed on the efficiency and limits of the procedure for designing practical wettability patterns.",
"discussion": "Results and Discussion For droplets impacting onto solid surfaces, two outcomes are possible: deposition or rebound (total or partial). For surfaces with isotropic wettability, the outcome can be predicted by the receding contact angle ( ) 9 . Figure 2 illustrates this with a plot of contact line position ( x ) vs. time ( t ) for droplets impacting onto surfaces with isotropic wettability, i.e. , superhydrophobic ( ) or hydrophilic ( ); the contact line in the former case advances and then recedes (recoil), while the contact line in the latter case gets pinned after the droplet reaches its maximum lateral spread (no recoil). For cases with anisotropic wettability, predicting the outcome is less clear (see SI Section S4). Figure 2 also shows the evolution of contact line position for a droplet impacting onto a hydrophilic arc on a superhydrophobic surface (see inset in (a) for a schematic; see (b) for image sequence as seen from the side). The contact line on the hydrophilic arc behaves as if it were in contact with an isotropic hydrophilic surface; on the superhydrophobic region, the contact line behaves as if it is in contact with a superhydrophobic surface, until x = 0. In the isotropic case, this is the point when rebound occurs (the droplet separates from the surface). In the anisotropic case, the contact line continues to recede until it reaches the hydrophilic arc, at which point rebound occurs; it is important to note that in this case the droplet is launched in a direction that is not orthogonal to the surface. In the anisotropic case, since there is a significant difference in apparent contact angles from the hydrophilic to superhydrophobic regions ( i.e. , hysteresis), a net surface tension force develops during the receding phase of droplet impact —in the positive x -direction— potentially allowing lateral droplet rebound ( vectoring ) to occur. As shown in Figure 2 , contact angle hysteresis occurs only after the droplet has reached its maximum spread (when the liquid gets pinned on the arc), i.e. , only during the receding stage. The magnitude of this hysteresis force ( F σ ) can be defined as a function of time according to where t r is the time after the droplet begins to recede ( t r = 0), and the receding contact angles on the hydrophilic arc and superhydrophobic region, respectively, and τ r the total receding time on the surface (see SI Section S5 for a detailed derivation). The above equation accounts for the fact that during the receding phase, the droplet is continuously de-wetting (peels off) the hydrophilic arc up to the instant of detachment (rebound). By integrating Equation 1 in time, one obtains the change in momentum (Δ P ) due to the contact angle hysteresis force as where m is the mass of the droplet, and U x the velocity of the droplet in the x direction (parallel to the surface). The subscripts 1 and 2 indicate when the receding phase begins and ends, respectively. We assume that: (i) (receding contact angle does not vary on the superhydrophobic surface 10 ; inertial regime of drop impact, i.e. , We >> 1; K is a prefactor constant which depends on , see SI Section S5); and (ii) r o ≈ D 0 We 0.25 /2 (inviscid droplet 11 ). The latter condition expresses the design requirement to have the arc coincide with the area under the droplet close to its outer periphery at the instant of maximum lateral spread. To satisfy (ii), the radius of the hydrophilic arc must be adjusted depending on the value of We employed. Under these conditions, Equations 1 and 2 can be combined and a horizontal restitution coefficient ( ε x ) can be defined as where C is a prefactor defined as . From Table 2 , we see that (the subscripts r,phi and r,pho represent the receding contact angles on the philic and phobic regions); K = 3.1 as found experimentally (see SI Section S5); therefore, C ≈ 2.7. Equation 3 suggests that for drop impact events on hydrophilic arcs with diameter ≈ 2 r o , ε x has a We −0.25 dependence. This is supported by Figure 3a , which shows a plot of ε x vs. We along with a best-curve fit for a function with C 1 We −0.25 (prefactor C 1 = 0.8). Figure 3b shows a semi-log plot for the same variables along with a best-fit linear curve. The slope is −0.25, supporting the power-law behavior shown in Equation 3 . The main point is that for orthogonal impact on a surface with asymmetric wettability, a predictable launch direction and velocity can be obtained, and the total outcome of the drop impact event can be predicted a priori . Figure 4 shows an image sequence demonstrating the droplet vectoring process. One drawback of the current system is that at higher values of We , smaller satellite droplets are formed during the impact process ( Figure 4 , 10.7 ms); under the same impact conditions on isotropic surfaces, no satellite droplets form. It is also instructive to consider the relative horizontal velocity of the droplets in the context of other surface microfluidic systems. For a 1 μ L water droplet impacting with We = 20, we see from Figure 3 that U x,2 = 0.28 m s −1 . If we compare U x,2 with the meniscus velocities of other fast transport surface microfluidic systems (0.4 m s −1, 12 0.3 m s −1, 13 0.031 m s −1, 14 0.005 m s −1, 15 ), we see that the present velocities are either comparable or even higher than existing systems. In terms of miniaturization, the dimensionless groups that govern the scalability of the process are the Reynolds number ( Re = D 0 U y ,0 / ν ; ratio of inertial and viscous forces; ν is kinematic viscosity of water) and We (ratio of inertial and surface tension forces). For this vectoring process to operate effectively, inertia should dominate. This is to ensure that droplet deformation occurs ( We >> 1) and that viscous forces do not dissipate all of the energy ( Re >> 1). For the cases presented in Figure 3 , the ranges were We ≈ 10–70 and Re ≈ 1200–3300. If we consider a typical impact velocity from an inkjet printing process ( U y ,0 = 3.0 m s −1 ) and we fix We = 10, then the minimum droplet size is D 0 = 80 μ m (typical D 0 = 50–70 μ m 16 ). For the same impact velocity and D 0 we calculate Re = 240. In this case, it appears that viscous forces will be the limiting factor in vectoring droplets. The previous discussion emphasized controlling the outcome of droplets that strike surfaces with specially-designed wettability patterns; however, other designs of anisotropic surfaces may be useful for controlling the outcome of droplets that strike and stick to the surface, especially as this outcome relates to morphing the liquid volume. One of the major hurdles for droplet shaping on non-wetting surfaces is uncontrollable liquid ejection that can be caused by two primary mechanisms: 1) The final wetting state of the droplet on the wettability pattern is thermodynamically unfavorable ( e.g. , high free surface area of droplet); 2) the droplet does not remain pinned to the hydrophilic region during the vigorous drop impact process. The former problem can be dealt with by designing wettability patterns such that the final configuration of the droplet is stable, and is relatively trivial (see SI Section S6 for a detailed analysis of designing a stable annulus). The latter problem is more difficult to remedy, since the liquid regions near the contact line are very thin and are therefore the most susceptible to separation, provided that air bubbles exist below them 4 . Let us consider the case of a wettability pattern with two concentric hydrophilic circles; the final configuration of the liquid droplet after impact is an annulus (see Figure 5 ). In this case, r o and r i are predetermined from pre-impact parameters and the final equilibrium configuration of the liquid droplet; in other words, we modify w i and w o to ensure that the droplet remains pinned to the wettability pattern during the impact process. Both of these parameters were determined experimentally, and then validated theoretically; for a given value of We , w o and w i were incrementally increased until the pattern was capable of consistently stabilizing a liquid lens during vigorous droplet impact. In practice, w o was set to ≈500 μ m, and w i was set to the minimum value achievable with the present patterning process (≈100 μ m); any set width less than that did not produce a continuous line pattern. It should also be noted that liquid lens detachment generally did not occur from the inner ring, so a theoretical estimate of w i was not necessary. It has been previously shown that liquid lens detachment from wettability patterns can arise due to hole formation near the hydrophobic-to-hydrophilic terrain transition, as a result of air-bubble entrainment due to the fast moving contact line crossing from a Cassie-Baxter 17 to a Wenzel 18 wetting state 4 . By increasing w o , the spatial locations of the entrained air bubbles at the wettability transition are moved towards the center of the liquid lens; since the thickness of the liquid also increases closer to the center, it is hypothesized that hole formation in liquid lenses at the wettability transition area becomes less energetically favorable. In fact, for air-bubbles that are capable of forming penetrating holes of radius ~ 90 μ m at the wettability transition when D 0 = 2.1 mm and r o = 2.5 mm, 4 once w o exceeds 150 μ m, hole formation is no longer energetically favorable in the liquid lens. See detailed analysis in SI Section S7. For the relatively high present values of We (≈80), it was observed that stable penetrating holes appeared very early in the impact process. If we consider t = 0 ms to be the moment of initial liquid/solid contact, then the first stable penetrating hole in the liquid lens appears at t ≈ 5 ms ( cf. \n Figure 5 ). The process of hole formation at We ≈80 is difficult to interpret due to the highly dynamic character of the impact event. However, when the value of We was reduced to ≈60, and r o was reduced to 2.34 mm, a penetrating hole did not appear until much later, t ≈ 15 ms (see Figure 6 ), allowing a larger portion of the kinetic energy from the droplet impact to be dissipated and the penetrating hole formation process to be more visually clear. Such penetrating holes (air cavities) —which shape the liquid into a well-defined torus— have been observed before during droplet impact on superhydrophobic surfaces, are commonly referred to as “dry-out” events, and may lead to bubble entrainment 19 ; viscous effects are small and the free surface state is controlled by the competition between surface tension and inertia forces 19 20 . It is known that during droplet impact on surfaces, it is possible for droplets to entrain macroscopic air bubbles near the impact zone during the early stages 19 21 22 23 24 , and thus it is possible to imagine future platforms to exploit this fact, i.e. , capillary waves converging onto an entrained air bubble to produce a penetrating hole 21 . In any case, the fact that the penetrating hole is axisymmetric with respect to the liquid droplet, and forms later in the drop impact process, means that there is less of a tendency for liquid to be ejected, and is a preferable method for forming liquid annuli. Using the same technique resulting in the formation of the liquid ring ( Figure 6 ), it is possible to create droplet patterns which are not radially symmetric. To demonstrate this, Figure 7 presents droplet impact with We = 100 on a cross wettability pattern. The hydrophilic square has an outer dimension of 4.5 mm and the centered inner hydrophobic cross has dimensions 2.7 mm × 0.87 mm. A penetrating hole forms in the liquid film —at the wettability transition— which grows to take up the shape of the hydrophobic cross, as previously shown for ring patterns. This impact sequence is very repeatable and consistent; even when the droplet impacted slightly off center, a hole still formed with no ejected satellites. One possible application of this process is rapid soft (water) templating for the purposes of forming films of complex geometries. To demonstrate this technique, a low-temperature sintering silver nanoparticle ink was added to the water cross pattern formed in Figure 7 . The process is shown in Figure 8 , where the silver nanoparticle ink was added on the water template by a pipette; see Figure 8c . The sample was placed in a dry air convection oven at 120°C for 180 s to convert the ink to silver; the manufacturer states that if conditions are optimized, then this type of silver can sinter in seconds. Taking a close look at one of the corners of the silver pattern ( Figure 8e ), it can be seen that the silver was restrained onto the hydrophilic domain and did not extend onto the hydrophobic surface. While the above discussion emphasized complex shaping —in a well controlled and highly repeatable manner—of liquid volumes striking wettability-patterned surfaces, the fact that advancing contact line behavior is inertia-dominated and receding contact line motion is wettability-dominated can be used to also split liquid micro-volumes. Figure 9 shows a sequence of snapshots of a water droplet impacting with We = 60 centrally on a superhydrophobic line of width 0.5 mm. The surface on both sides of the superhydrophobic line is hydrophilic. As seen previously, the drop reaches its maximum spreading radius at t = 2.33 ms. The shape of the contact line is slightly irregular due to the differences in wettability of the surface, which affect symmetry. From the maximum spreading radius, the contact line on the hydrophobic strip recedes quickly, while the rest of the contact line is pinned on the hydrophilic parts of the surface, thus receding only slightly during the same period. As a result, a liquid bridge between two main separating parts of the drop forms; finally, the bridge collapses, leading to separation of the drop —in under 18 ms from initial contact— into two drops of equal size. Clearly, this wettability pattern acts as a liquid “scalpel” 25 . Experiments at lower values of We (<30) show that splitting is not possible with this same surface (see SI Section S8), suggesting the critical role inertia plays in the splitting process. Additional experiments revealed that as the superhydrophobic line became wider, the droplet required a higher maximum lateral spread ( i.e. , likewise We ) in order to reach the hydrophilic regions, leading to an increased risk of partial rebound and “sling shot” action, resulting in partial loss of the liquid volume (see SI Section S8). This loss is not desirable in surface microfluidic devices. While the above effort was focused on splitting a droplet in two, such wettability engineered surfaces are capable of a much higher rate of droplet sampling. Proper wettability patterns were also shown to be able to rapidly sample a large number (24 droplets) of small volumes in ~0.01 s (see SI Section S8)."
} | 4,475 |
23457582 | PMC3574064 | pmc | 5,927 | {
"abstract": "The animal gastrointestinal tract contains a complex community of microbes, whose composition ultimately reflects the co-evolution of microorganisms with their animal host. An analysis of 78,619 pyrosequencing reads generated from pygmy loris fecal DNA extracts was performed to help better understand the microbial diversity and functional capacity of the pygmy loris gut microbiome. The taxonomic analysis of the metagenomic reads indicated that pygmy loris fecal microbiomes were dominated by Bacteroidetes and Proteobacteria phyla. The hierarchical clustering of several gastrointestinal metagenomes demonstrated the similarities of the microbial community structures of pygmy loris and mouse gut systems despite their differences in functional capacity. The comparative analysis of function classification revealed that the metagenome of the pygmy loris was characterized by an overrepresentation of those sequences involved in aromatic compound metabolism compared with humans and other animals. The key enzymes related to the benzoate degradation pathway were identified based on the Kyoto Encyclopedia of Genes and Genomes pathway assignment. These results would contribute to the limited body of primate metagenome studies and provide a framework for comparative metagenomic analysis between human and non-human primates, as well as a comparative understanding of the evolution of humans and their microbiome. However, future studies on the metagenome sequencing of pygmy loris and other prosimians regarding the effects of age, genetics, and environment on the composition and activity of the metagenomes are required.",
"conclusion": "Conclusions We presented for the first-time the application of the shotgun metagenomic pyrosequencing approach to study the fecal microbiome of the pygmy loris. The overall goal of this study was to characterize the species composition and the functional capacity of the pygmy loris fecal microbiome. Taxonomic analysis of the metagenomic reads showed similarities among the gut microbiomes of the pygmy loris, humans, and other animals. Four phyla dominated the microbiomes, namely, Bacteroidetes, Proteobacteria, Actinobacteria, and Firmicutes. However, the relative proportion of the phyla was different; most of the less abundant phyla such as Proteobacteria and Actinobacteria were more prevalent, and most of the more abundant phyla such as Firmicutes were fewer in the pygmy loris fecal microbiome than in humans and other animals. At the genus-level taxonomic resolution, Bacteroides species were the most abundant, most of which were represented by B. fragilis . The organisms belonging to the said genus also represent one of the most abundant microbial taxa in the human intestinal microbiota [11] , [28] . The pygmy loris faecal samples contained more bacteria belonging to the phylum Verrucomicrobia, most of which were represented by the mucin-degrading bacterium A. muciniphilia . The high amount of A. muciniphilia present in the pygmy loris feces indicates a high turnover of mucins in these prosimians. Archaea, fungi, and viruses are minor constituents of the pygmy loris fecal microbiome. All archaea are members of Crenarchaeota and Euryarchaeota, with methanogens being the most abundant and diverse. Three fungi phylotypes were present in the pygmy loris fecal microbiome, namely, Ascomycota, Basidiomycota, and Microsporidia. Only about 0.1% of sequences were of viral origin, and all sequences were classified as bacteriophages. The hierarchical clustering of the gut metagenomic data from pygmy loris, humans, dogs, mice, chickens, and cows demonstrated the similarity of the microbial community structures of the pygmy loris and mouse gut systems despite the differences in functional capacity. The comparison of the fecal microbiota of NHPs with the microbiota of humans and other animals obtained in previous studies revealed that the gut microbiota of the pygmy loris are distinct and reflect host phylogeny. The comparative metagenomic analyses identified unique and/or overabundant taxonomic and functional elements in the pygmy loris distal gut microbiomes. Relatively abundant and diverse metabolic subsystems of aromatic compounds were found in the pygmy loris metagenome compared with all the other gut metagenomes. These results contribute to the limited body of primate metagenome studies and provide a framework for the comparative metagenomic analysis of the human and NHP metagenome, as well as a comparative understanding of the evolution of humans and their microbiome. More studies involving the deeper sequencing of metagenomes are required to fully characterize the gastrointestinal microbiome of the pygmy loris and other prosimians in healthy and diseased states, of varying ages or genetic backgrounds, and in the wild or in captivity.",
"introduction": "Introduction The gastrointestinal tract of animals harbors a complex microbial community, and the composition of this community ultimately reflects the co-evolution of microorganisms with their animal host and the diet adopted by the host [1] . As a result of the issues related to health and disease, the structure and function of the gut microbial community of humans has received significant attention from researchers. Previous studies have proven that the microbiomes of non-human primates (NHPs) exhibit a much higher similarity with those of primates than with other animals [1] . Therefore, the study of the microbiota from these NHPs provides important insights into the reflection of their features in humans. However, only a few reported culture-independent studies on fecal microbiota of non-human primates [2] – [9] are available, leading to limited comparative data on the intestinal microbiota of primates, either in captivity or in the wild. More extensive surveys of primate gastrointestinal microbiomes, particularly prosimian primates, about which little research work has been done [7] , combined with comparative analyses of their microbiomes with those of humans are necessary to better understand the evolution of humans and their microbiome. The pygmy loris ( Nycticebus pygmaeus ) is a small rare nocturnal prosimian primate found mainly in Vietnam, Laos, and China. Being nocturnal, the prosimians are less known than other primates, but are nonetheless important. Given that previous culture-independent 16S rRNA gene-based analyses have revealed impressive microbial diversity in the pygmy loris feces [7] , these analyses offer limited information on the physiological role of microbial consortia within a given gut environment. Random sequencing of the metagenomes has allowed scientists to reveal significant differences in metabolic potential within different environments [10] . Recently, next-generation sequencing technologies have been used to characterize the microbial diversity and functional capacity of a range of microbial communities in the gastrointestinal tracts of humans [11] – [14] as well as in several animal species [15] – [18] . The most important advantages of this cloning-independent approach are the avoidance of cloning bias and the bias introduced by PCR amplification. To best of our knowledge, this study was the first to apply a random sample pyrosequencing approach to analyze the metagenome of the pygmy loris to better understand how microbiomes relate to NHPs ecological and evolutionary diversity.",
"discussion": "Results and Discussion The analysis of the reads yielded a high percentage of species identification in complex metagenomes and even higher in less complex samples. Long sequence reads from 454 GS FLX Titanium pyrosequencing provided the high specificity needed to compare the sequenced reads with the DNA or protein databases and allowed the unambiguous assignment of closely related species. The initial pyrosequencing runs yielded 78,619 reads containing 34,473,384 bases of sequence, with an average read length of 438 bp. Prior to further processing, the raw read data were subjected to the MG-RAST v.3.0 online server quality control pipeline [19] to remove duplicate and low quality reads ( Table S1 ). The filtering step removed 22.1% of reads in the sample. The unique sequence reads that passed the QC filtering step were then subjected to further analysis focusing on biodiversity and functional annotation. All reads were deposited in the National Center for Biotechnology Information (NCBI) and can be accessed in the Short Read Archive (SRA) under the accession number SRX160437. Phylogenetic Analysis of Pygmy Loris Fecal Bacteria, Eukaryota, Archaea, and Viruses The overview of the phylogenetic computations provided 95.54% bacteria, 3.8% eukaryota, 0.39% archaea, and 0.12% viruses. In the pygmy loris intestinal metagenome, Bacteroidetes was the most predominant phylum (∼41%), followed by Proteobacteria (∼30%), Actinobacteria (∼11%), and Firmicutes (∼9%) ( Figure 1 ). Compared with the previous 16S rRNA gene-based data [7] , significantly lower percentages of Firmicutes and higher percentages of Bacteroidetes in the pygmy loris intestinal metagenome were observed. This discrepancy may have been caused by the biases associated with the primers, PCR reaction conditions, or selection of clones [25] . 10.1371/journal.pone.0056565.g001 Figure 1 Bacterial phylum profiles of the pygmy loris microbiome. The percentage of the pygmy loris fecal metagenomic sequences assigned to M5NR database is shown. Through the “Organism Abundance” tool in MG-RAST, the pygmy loris fecal sequencing runs were determined from the M5NR database with the BLASTx algorithm. The e-value cutoff for the metagenomic sequence matches to the M5NR database was 1×10 −5 , with a minimum alignment length of 30 bp. The relative paucity of the Firmicutes sequences is in conflict with data from the studies of humans [1] , [11] , [12] , [26] and other higher primates [2] , [3] , [4] , [6] , [8] , [9] . The reasons for the variation are difficult to identify because of the biases involved in the fecal lysis and DNA extraction methods [27] ; inter-individual variability may also contribute to this divergence. Given that 70% of the phylotypes existing in the human gastro-intestinal microbiome are subject-specific and no phylotype is present at more than 0.5% abundance in all subjects [13] , the gastro-intestinal microbiota of each individual has been shown to consist of a subject specific complement of hundreds of genera and thousands of species. As a prosimian, the pygmy loris is less like a primate than others with the same intestinal microbiome composition. However, this claim needs to be proved by further research on fecal samples of more pygmy loris and prosimians. Within the Bacteroidetes group, Bacteroidales were the most predominant, among which Bacteroides , Prevotella , and Parabacteroides were consistently overrepresented ( Table S2 ). Organisms belonging to the genus Bacteroides represent one of the most abundant microbial taxa in human intestinal microbiota [11] , [28] . Bacteroides fragilis comprises about 5.8% of the reads analyzed; therefore, it is considered the predominant species in the pygmy loris metagenome. B. fragilis is a ubiquitous Gram-negative anaerobic bacterium that inhabits the lower GI tract of most mammals [1] . Recent findings have revealed that this organism possesses the ability to direct the cellular and physical maturation of the host immune system and protect its host from experimental colitis [29] – [31] . Natural habitats of Prevotella sp. include the rumen and hindgut of cattle, sheep, and humans, where they help break down protein [32] and carbohydrate [33] . However, some species of this genus are known to be opportunistic pathogens to humans [34] . Proteobacteria were the second predominant phylum in the pygmy loris gastrointestinal tract with Pseudomonadales as the primary contributor to the Proteobacteria populations, followed by Enterobacteriales and Burkholderiales. The major genus in the Proteobacteria phylum is Pseudomonas , which is consistent with previous 16S rRNA gene-based data [7] . Several microbes belonging to the genus of Pseudomonas have a very diverse metabolism, including the ability to degrade organic solvents such as toluene [35] and phenol [36] , [37] . This ability may benefit the pygmy loris, given that they consume several toxic and pungent insects. Pseudomonas fluorescens was the predominant species among the Pseudomonas in the pygmy loris metagenome. P. fluorescens is a common Gram-negative bacterium that can be found in the low section of the human digestive tract [38] . Similarly, Clostridia and Bacilli are the primary contributors to the Firmicutes populations. However, various genera were found in the pygmy loris metagenome than in the 16S rRNA gene [7] ( Table S2 ). Clostridiales is the dominant order in Clostridia, which includes well-known gut bacteria, Faecalibacterium prausnitzii . F. prausnitzii is the most important n-butyrate producing gut bacterium with well-known effects on host energy metabolism and mucosal integrity [39] . A distinctive feature of the pygmy loris metagenome is the abundance of phylum Verrucomicrobia, particularly the members of the genus Akkermansia ; this abundance was unexpected and far greater than in humans ( Table S2 , Figure 2 ). The dominant species in the Verrucomicrobia phylum was Akkermansia muciniphila , which are common members of the human gut microbiota evident in human infants [40] . These mucin-degrading bacteria are related to normal mucosa development. Moreover, the Akkermansia species may have a role in maintaining intestinal integrity. 10.1371/journal.pone.0056565.g002 Figure 2 Phylogenetic clustering of pygmy loris, human, mouse, canine, cow, and chicken gastrointestinal metagenomes. A double hierarchical dendrogram was established through weight-pair group clustering methods based on the non-scaling Manhattan distance. The dendrogram shows the phylogenetic distribution of the microorganisms among the ten metagenomes from the six different hosts, including pygmy loris (WFH), human (HSM and F1S), mouse (LMC and OMC), dog (K9C and K9BP), cow (CRP), and chicken (CCA and CCB). The linkages of the dendrogram do not show the phylogenetic relationship of the bacterial phylum and are based on the relative abundance of taxonomic profiles. The heat map depicts the relative percentage of each phylum of microorganism (variables clustering on the y axis) in each sample (x axis clustering). The heat map color represents the relative percentage of the microbial descriptions in each sample, with the legend indicated at the upper left corner. Branch length indicates the Manhattan distances of the samples along the x axis (scale at the upper right corner) and of the microbial phyla along the y axis (scale at the lower left corner). Eukaryota were a minor constituent (∼4.0%) in the pygmy loris metagenome. Species of Blastocystis were also represented in small quantities (<0.01%) in the pygmy loris metagenome. These species have been reported as the most commonly occurring micro-eukaryote in human feces [41] , [42] . In addition, the presence of Blastocystis has been linked to a number of gut-related diseases. Some of these diseases could be the outcome of the predation of beneficial bacteria by Blastocystis in light of the similar observations in ruminant cattle and their communalistic protozoa [43] . Fungi have very low abundance sequences (0.5%), with Eukaryota with Dikarya being the primary contributor, followed by Microsporidia ( Table S3 ). Fungi in the intestinal ecosystem of NHPs have not yet been studied extensively. Through culture-independent methods, Scanlan and Marchesi revealed that more diverse fungi species are found in the human distal gut, including Saccharomyces , Gloeotinia , Penicillium , Candida , and Galactomyces \n [41] . Similar phylotypes of fungi ( Candida , Cladosporium , Penicillium , and Saccharomyces ) have also been identified in stool samples from patients with human inflammatory bowel disease and from healthy control subjects [44] . Compared with humans, more diverse fungi species belonging to 29 different genera exist in the pygmy loris metagenome ( Table S3 ). The most abundant fungi genera in the pygmy loris metagenome were Aspergillus (0.04%), Neurospora (0.03%), Gibberella (0.03%), and Neosartorya (0.03%). Three fungi species ( Gibberella zeae , Neurospora crassa , and Saccharomyces cerevisiae ) have been identified in the pygmy loris metagenome, which were also identified in feline [18] , canine (K9C and K9BP) [16] , and mouse metagenomes (OMC) [45] . \n Enterocytozoon bieneusi is the only species found in Microsporidia. Microsporidia are enteric protozoan parasites recognized as important pathogens in immunocompromised persons and malnourished children. E. bieneusi is the most common microsporidian species in humans, infecting enterocytes and other epithelial cells [46] . Interestingly, our study demonstrated a low percentage of fungi with high diversity among the fecal microflora. Future studies on the fungal diversity in the pygmy loris gastrointestinal tract would benefit from this next-generation sequencing. Archaea are a minor component of the pygmy loris metagenome, comprising ∼0.4% of the total sequencing reads. Archaea consist of two phyla, Euryarchaeota and Crenarchaeota, which diverged into seven classes and eight orders ( Table S4 ). Among the groups of archaea, methanogenic archaea is the most predominant and diverse group. Methanogenic archaea is also widespread in a variety of vertebrates, such as felines, dogs, humans, mice, and chicken [15] , [16] , [18] , [45] , [47] . In the pygmy loris fecal metagenome, Methanocorpusculum labreanum is the major component of archaea, having a percentage of 0.03% in all the analytic sequences ( Table S4 ). Although archaea have been observed in NHPs previously [5] , the diversity of this domain has not been elucidated yet. Archaea are considered commensals; however, they contribute to pathogeny in humans because of mutual interactions with other microorganisms [48] . For instance, methanogens consume hydrogen and create an environment that enhances the growth of polysaccharide fermenting bacteria, leading to higher energy utilization. Higher numbers of methanogenic archaea have been observed in obese humans [49] . However, the prevalence and medical importance of archaea in NHPs need to be determined. Despite the ability to carry methanogens in the intestinal tracts of animals linked to phylogeny [50] , the gut archaea in the pygmy loris metagenome are not similar to human ones than to those of other species. Only ∼0.1% of the total reads have viral origin, with only the order Caudovirales being identified. Two families were observed (Myoviridae and Siphoviridae) within the Caudovirales order, and all sequences were classified as bacteriophages ( Table S5 ). Bacteriophages influence food digestion by regulating microbial communities in the human GI tract through lytic and lysogenic replication [51] . Bacteriophages also contribute to human health by controlling invading pathogens [52] . Recent metagenomic analyses of the DNA viruses from human feces have revealed that the majority of DNA viruses in human feces are novel, and most of the recognizable sequences also belong to bacteriophages [53] . The close phylogenetic relationship between humans and NHPs, coupled with the exponential expansion of human populations and human activities within the primate habitats, has resulted in the exceptionally high possibility of pathogen exchange [54] . Therefore, studies on the viral community of NHPs and the potential for cross-transmission between humans and NHPs are needed. Given the type of methodology (shotgun DNA pyrosequencing approach) that we utilized, our study could only determine the dsDNA virus. Future studies need to provide a richer understanding of both RNA and dsDNA viruses to complete human knowledge of the viral intestinal ecosystem. Studies on cow [55] , cats [56] , mice [57] , rats [58] as well as humans [59] and NHPs [5] revealed a correlation between host diet and microbial community composition. However, the characterization of the dietary-induced changes in NHP microbiomes through high-throughput sequencing technologies has not been performed thus far. Hence, more attention should be given to it in future experiments. Metabolic Profiles of the Pygmy Loris Metagenome Carbohydrate metabolism is the most abundant functional category, representing 11.45% of the pygmy loris fecal metagenomes ( Figure 3 ). Genes associated with amino acids and derivatives, protein metabolism, cofactors (vitamins, prosthetic groups, pigments), membrane transport, cell wall and capsule. RNA metabolism and DNA metabolism are also very abundant in the pygmy loris metagenomes. Approximately 14.44% of the annotated reads from the pygmy loris fecal metagenomes were categorized within the clustering-based subsystems, most of which have unknown or putative functions. 10.1371/journal.pone.0056565.g003 Figure 3 Functional composition of the pygmy loris microbiome. The percentage of the pygmy loris fecal metagenomic sequences assigned to the general SEED subsystems is shown. Through the “Functional Abundance” tool in MG-RAST, the pygmy loris fecal sequencing runs were determined from the SEED database with the BLASTx algorithm. The e-value cutoff for the metagenomic sequence matches to the SEED subsystem database was 1×10 −5 with a minimum alignment length of 30 bp. We subjected our samples to the carbohydrate-active enzymes database (CAZy; http://www.cazy.org ), as described by Cantarel et al. [23] , to obtain a more in-depth view of the carbohydrate enzymes present in our data set. The comparison of the 61,281 metagenome reads post-QC processing based on the CAZy database provided 2260 hits at an E value restriction of 1×10 −6 . Candidate sequences that belong to the glycosyl transferase families GT2 (148) are the most abundant, followed by members of the glycoside hydrolase (GH) families GH3 (142) and GH2 (134) ( Figure S1 and Table S6 ). GHs are a prominent group of enzymes that hydrolyze the glycosidic bond among the carbohydrate molecules. The most frequently occurring GH families in the pygmy loris metagenome were GH3 (142; 8.96% of the total GH matches), 2 (134; 8.45%), and 43 (103; 6.5%) ( Figure S1 and Table S6 ). The most common activities of GH3 include β-D-glucosidases, α-L-arabinofuranosidases, β-D-xylopyranosidases, and N-acetyl-β-D-glucosaminidases [60] . In several cases, the enzymes have dual or broad substrate specificities with respect to monosaccharide residue, linkage position, and chain length of the substrate, such as α-L-arabinofuranosidase and β-D-xylopyranosidase [61] . GH2 components are β-D-galactosidases, β-glucuronidases, β-D-mannosidases, and exo-β-glucosaminidases. GH43 shows β-xylosidase, β-1,3-xylosidase, α-L-arabinofuranosidase, arabinanase, xylanase, and galactan 1,3-β-galactosidase activity ( www.cazy.org ). Glycosyl transferases are ubiquitous enzymes that catalyze the attachment of sugars to a glycone [62] . Candidate genes that belong to the glycosyl transferase families GT2 (148; 35.24% of the total GT matches) and GT4 (96; 22.86%) are the most abundant ( Figure S1 ). Comparative Metagenomic Analysis Despite the extensive variation among individuals, the gut microbiota of members of the same species are often more similar to one another compared with those of other species. Both humans and the pygmy loris are primates; however, the latter are prosimians and are different from humans in terms of primate evolution. The human gut is a natural habitat for various communities of microorganisms that have co-evolved with humans. Thus, it is important to provide a comparison between the gastrointestinal microbiomes of primates and those of other animals. The results of this study were compared with data sets from different animals and even humans in the MG-RAST database. Paired data from other studies were chosen, such as lean (LMC) and obese (OMC) mouse cecal metagenomes [45] , two chicken cecal metagenomes (CCA, CCB) [15] , two canine intestinal metagenomes (K9C, K9BP) [16] , and two human fecal metagenomes (F1S, HSM). F1S was a healthy human fecal metagenome [47] , whereas HSM was defined as human feces from a malnourished subject. A single cow rumen metagenome (4441682.3) was also utilized for comparison. The comprehensive overview of the ten data sets is shown in Table S7 . Clustering the metagenomes was carried out with unscaled Manhattan variance distances and presented through a double hierarchical dendrogram. The clustering-based comparisons were demonstrated at the phylogenetic level ( Figure 2 ) and the metabolic level ( Figure 4 ). In the phylogenetic comparison, the pygmy loris samples clustered with the two mouse metagenomes (OMC and LMC) and separated from those of the other animals and humans. The results do not correspond with previous studies, indicating that microbial community composition is more similar among primates than in other animals [1] . Although pygmy loris belong to primates,they are prosimians, or pre-monkeys and are different from humans in terms of primate evolution. Therefore, the gut microbiomes of the pygmy loris may show an obvious difference compared with human gastrointestinal microbiomes. Meanwhile, those of mice have a more exact similarity with the human genome; more than 90% of the mouse genome is similar to the human genome [63] . The microorganism composition of the animal gastrointestinal tract reflects the constant co-evolution of the animal with its host [28] . The clustering of the pygmy loris metagenome with that of the mouse metagenome may be a result of similar bacterial diversity influenced by co-evolution with the host. Similar clustering of mouse and human data from the IMG/M ER database was performed. Figure S2 demonstrates that the pygmy loris samples clustered with the mouse gut samples. 10.1371/journal.pone.0056565.g004 Figure 4 Metabolic clustering of pygmy loris, human, mouse, canine, cow, and chicken gastrointestinal metagenomes. A double hierarchical dendrogram was established through a weight-pair group clustering method based on the non-scaling Manhattan distance. The dendrogram shows the distribution of the functional categories among the ten metagenomes from the six different hosts, including pygmy loris (WFH), humans (HSM and F1S), murine (LMC and OMC), canine (K9C and K9BP), cow (CRP), and chicken (CCA and CCB). The linkages of the dendrogram are based on the relative abundance of metabolic profiles. The heat map depicts the relative percentage of each category of function (variables clustering on the y axis) in each sample (x axis clustering). The heat map color represents the relative percentage of functional categories in each sample, with the legend indicated at the upper left corner. Branch length indicates the Manhattan distances of the samples along the x axis (scale at the upper right corner) and of the microbial classes along the y axis (scale at the lower left corner). In all the samples, the Bacteroidetes, Firmicutes, Actinobacteria, and Proteobacteria were the most abundant. The pygmy loris metagenome was most distinguished by the greater prevalence of Verrucomicrobia compared with mice and other animals, as shown in Figure 2 . The heat map also demonstrates that the pygmy loris fecal metagenome contains lower Fibrobacteres, an important phylum of cellulose-degrading bacteria. Metabolism-based hierarchical clustering demonstrates that the pygmy loris, human, chicken, and dog samples clustered together. The mice and cow samples were the least similar samples to the pygmy loris ( Figure 4 ). The similarity of function among pygmy loris, humans, chicken, and dogs is not surprising, considering the fact that they are all omnivores with similar digestive tract structures and functions. Similar clustering of mouse and human data from the IMG/M ER database was performed. Figure S3 demonstrates that the pygmy loris samples clustered with the human gut samples. Interestingly, the two mouse samples that clustered together were most similar to the cow rumen samples, that is, those of an herbivore. As expected, all the gut metagenomes were dominated by carbohydrate metabolism subsystems with amino acids, protein, and cell wall and capsule; the DNA and RNA subsystems were represented in relatively high abundance as well. An interesting result was observed in terms of the metabolism of aromatic compounds, which accounted for a higher number of reads in the pygmy loris fecal metagenome than in other animals ( Figure 4 ). The profile of the metabolism of aromatic compounds in the pygmy loris fecal metagenomes was dominated by proteins annotated as subsystems of peripheral pathways for catabolism of aromatic compounds, metabolism of central aromatic intermediates, and anaerobic degradation of aromatic compounds ( Figure 5 ). 10.1371/journal.pone.0056565.g005 Figure 5 Percentage of sequences associated with the metabolism of aromatic compounds in the pygmy loris microbiome. Through the “Metabolic Analysis” tool in MG-RAST, the pygmy loris fecal sequencing runs were determined from the SEED database with the BLASTx algorithm. The e-value cutoff for the metagenomic sequence matches to the SEED subsystem database was 1×10 −5 with a minimum alignment length of 30 bp. KEGG Pathway Assignment Pathway assignment was performed based on the Kyoto Encyclopedia of Genes and Genomes (KEGG). First, the 78,619 reads were compared using BLASTX with the default parameters from the KEGG database. A total of 18,410 reads with corresponding enzyme commission (EC) numbers were assigned to the metabolic pathways. Given that the sequences related to the metabolism of aromatic compounds were more abundant in the pygmy loris fecal metagenomes compared with other animals in terms of subsystem, we focused our attention on xenobiotic biodegradation and metabolism. A high number of sequences in the benzoate degradation pathway was observed, which is coherent with the fact that benzoate is a central intermediary compound in the anaerobic and aerobic metabolism of various aromatic compounds, such as toluence, xylene, fluorine, carbazole, and biphenyl [64] . The key enzymes involved in benzoate degradation via hydroxylation, such as catechol 1,2-dioxygenase (EC 1.13.11.1), and protocatechuate 3,4-dioxygenase (EC 1.13.11.3) were identified in the pygmy loris fecal metagenomes ( Figure 6a ). The two usual methods of aerobic benzoate metabolism are dioxygenation to form catechol, utilized by some bacteria such as Pseudomonas putida and Acinetobacter calcoaceticus \n [65] , and monooxygenation to form protocatechuate, mostly by Aspergillus niger \n [66] . Almost all the enzymes involved in the two methods of aerobic benzoate metabolism in the KEGG pathway ( Figure 6a ). The main organisms, P. putida , A. calcoaceticus , and A. niger , involved in the course of metabolism were all represented in the pygmy loris fecal microbiome ( Table S2 and S3 ). 10.1371/journal.pone.0056565.g006 Figure 6 Reference pathway of benzoate degradation. a. Benzoate degradation by the hydroxylation reference pathway. b. Benzoate degradation by the CoA ligation reference pathway. The boxes colored red represent the enzymes identified from the pygmy loris fecal metagenomes based on KEGG. Although several key enzymes such as benzoyl-CoA reductase (EC 1.3.99.15) were missing in the method of anaerobic benzoate metabolism via CoA ligation, partial enzymes were identified ( Figure 6b ). This particular result may be due to the fact that the pathway of anaerobic benzoate metabolism in the pygmy loris was a little different. These results suggest that the fecal microbiota of the pygmy loris under study have a potential to degrade phenol and derivatives by the aerobic and anaerobic pathway. Moreover, these pathways may interchange because of the cross-regulation between the anaerobic and aerobic pathways for the catabolism of aromatic compounds, which may reflect a biological strategy to increase cell fitness in organisms that survive in environments subject to changing oxygen concentrations [67] . Aromatic compounds comprise one-quarter of the Earth's biomass and are the second most widely distributed class of organic compounds in nature, next to carbohydrates. Benzoate is produced from the microbial degradation of dietary aromatic compounds in the intestine [68] and is also naturally present in most berries, fruits, and fermented dairy products [69] . Normal patterns of personal dietary intake alters the phenolic substrates supplied to the intestinal bacteria and the aromatic metabolites formed, resulting in possible fluctuations in the microflora population [70] . The pygmy loris is an insectivorous species that also eats fruits, birds' eggs, chicks, geckos, and arboreal small mammals [71] . Such diet rich in aromatic compounds could result in the relative abundance of sequence-encoding enzymes and microbiota involved in benzoate degradation. Therefore, more attention should be given to the potential new genes and pathways generated by the metabolism of aromatic compounds in the pygmy loris fecal microbiome."
} | 8,379 |
22433663 | PMC3366867 | pmc | 5,929 | {
"abstract": "Background Direct conversion of solar energy and carbon dioxide to drop in fuel molecules in a single biological system can be achieved from fatty acid-based biofuels such as fatty alcohols and alkanes. These molecules have similar properties to fossil fuels but can be produced by photosynthetic cyanobacteria. Results Synechocystis sp. PCC6803 mutant strains containing either overexpression or deletion of the slr1609 gene, which encodes an acyl-ACP synthetase (AAS), have been constructed. The complete segregation and deletion in all mutant strains was confirmed by PCR analysis. Blocking fatty acid activation by deleting slr1609 gene in wild-type Synechocystis sp. PCC6803 led to a doubling of the amount of free fatty acids and a decrease of alkane production by up to 90 percent. Overexpression of slr1609 gene in the wild-type Synechocystis sp. PCC6803 had no effect on the production of either free fatty acids or alkanes. Overexpression or deletion of slr1609 gene in the Synechocystis sp. PCC6803 mutant strain with the capability of making fatty alcohols by genetically introducing fatty acyl-CoA reductase respectively enhanced or reduced fatty alcohol production by 60 percent. Conclusions Fatty acid activation functionalized by the slr1609 gene is metabolically crucial for biosynthesis of fatty acid derivatives in Synechocystis sp. PCC6803. It is necessary but not sufficient for efficient production of alkanes. Fatty alcohol production can be significantly improved by the overexpression of slr1609 gene.",
"conclusion": "Conclusions In this study the effects of fatty acid activation functionalized by a fatty acyl-ACP synthase on the production of fatty acid-based biofuels including fatty alcohols and alkanes in a photosynthetic cyanobacterium were evaluated and analyzed. We found fatty acid activation to be essential for efficient production of alkanes and plays a key role in manipulating fatty alcohol production. The results here provide promising clues for metabolically engineering cyanobacteria to improve photosynthetic production of fatty acid-based biofuels.",
"discussion": "Results and discussion Construction of Synechocystis sp. PCC6803 mutants with either overexpression or deletion of slr1609 gene To investigate the impact of AAS on production of free fatty acids and fatty acid derivatives, we constructed two plasmids pGQ11 (Figure 2B ) and pGQ49 (Figure 2D ) for over-expressing slr1609 gene, driven by a strong constitutive promoter Prbc or PpsbA2 and integrated into slr0168 [ 9 ] or psbA2 site, respectively. Two plasmids pGQ53 (Figure 2A ) and pGQ17 (Figure 2C ) were constructed for disruption of slr1609 gene with erythromycin or kanamycin resistance cassettes, respectively. The plasmid pGQ11 or pGQ53 was transformed into Synechocystis sp. PCC6803 generating GQ3 and GQ8 strains respectively for analysis of fatty acid and alkane production. The plasmid pGQ49 or pGQ17 was transformed into fatty-alcohol-producing strain Syn-XT14 generating GQ5 and GQ6 respectively for analysis of fatty alcohol production. Overexpressed AAS protein with C-terminal His-tag in GQ3 and GQ5 mutant were detected by western blotting as shown in Additional file 1 : Figure 1. Figure 2 Maps of the plasmids (A) pGQ53, (B) pGQ11, (C) pGQ17 and (D) pGQ49 . Due to cyanobacterial cells containing multiple copies of chromosomes [ 10 ], the complete replacement of wild type alleles must be established and confirmed by PCR (Figure 3 and 4 ). Primers used in this study were listed in Additional file 1 : Table 1. Because the whole inserted fragment is too long to amplify from genomic DNA, the over-expressed genes were checked by two reactions for successful insertion and correct orientation of the slr1609 or FAR gene and complete replacement of wild-type alleles (Figure 3B and 4 ). The first reaction with the primer (0168-2 or Pd1-3) of insert site and primer (1609NdeI, 1609R or far-1) for the inserted gene verified the genes were inserted in the correct orientation. The second reaction with primers (0168-1/0168-2 or pD1-1/p pD1-2d-2) of inserted site verified the wild-type was replaced completely (Figure 3B and 4 ). The disrupted genes were checked with internal primers (Figure 3A and 4B ) or primers that flanked the insertion site to prove the wild-type allele was replaced completely. The results of the PCR indicated the correct mutants were constructed. Figure 3 PCR analysis of the genotype of Synechocystis sp. PCC 6803 mutant strains . (A) lane 1: DNA marker (200 bp DNA Ladder Marker), lane 2: genomic DNA of GQ8 was amplified by primers kus and kdas outside the inserted gene segment, lane 3: plasmid pGQ53 was amplified by primers kvF and kvR located inside slr1609 gene (control), lane 4: genomic DNA of GQ8 was amplified by the same primers as lane 3, lane 5: genomic DNA of wild-type was amplified by the same primers as lane 3 (control). (B) lane1: DNA marker (1 kb DNA Ladder Marker), lane2: genomic DNA of wild type was amplified by primers 0168-2 and 1609NdeI (control), lane3: genomic DNA of GQ3 was amplified by the same primers as lane2, lane4: plasmid pGQ11 was amplified by the same primer as lane2 (control), lane5: genomic DNA of wild-type was amplified by primers 0168-1 and 0168-2(control), lane6 : genomic DNA of GQ3 was amplified by the same primers as lane5, lane7 : H 2 O was used as template, and the same primer as lane5 were used in PCR reaction(control). Figure 4 PCR analysis of the genotype of Synechocystis sp. PCC 6803 mutant strains . (A) lane1: DNA marker (1 kb DNA Ladder Marker), lane2: genomic DNA of wild-type was amplified by primers 0168-1 and far-1 (control), lane3: genomic DNA of GQ5 was amplified by the same primers as lane2, lane4: plasmid pXT14 was amplified by the same primers as lane2, lane5: genomic of wild-type was amplified by primers 0168-1 and 0168-2, lane6: genomic DNA of GQ5 was amplified by the same primers as lane5, lane7: plasmid pXT14 was amplified by the same primer as lane5 (control), lane8: genomic DNA of wild-type was amplified by primers pD1-3 and 1609R (control), lane9: genomic DNA of GQ5 was amplified by the same primers as lane8, lane10: plasmid pGQ49 was amplified by the same primers as lane8 (control), lane11: genomic DNA of wild-type was amplified by primers pD1-3 and pD1-2d-2 (control), lane12: genomic DNA of GQ5 was amplified by the same primers as lane11, lane13: plasmid pGQ49 was amplified by the same primers as lane11 (control). (B) lane1: DNA marker (1 kb DNA Ladder Marker), lane2: genomic DNA of wild-type was amplified by the primers 0168-1 and far-1 (control), lane3: genomic DNA of GQ6 was amplified by the same primers as lane2, lane4: plasmid pXT14 was amplified by the same primers as lane2 (control), lane5: genomic DNA of wild-type was amplified by primers 0168-1 and 0168-2 (control), lane6: genomic DNA of GQ6 was amplified by the same primers as lane5, lane7: plasmid pXT14 was amplified by the same primer as lane5 (control), lane8: genomic DNA of wild-type was amplified by primer 1609NdeI and 1609R (control, the size of target DNA fragment should be 2091 bp), lane9: genomic DNA of GQ6 was amplified by the same primers as lane8 (the size of target DNA fragment should be 1810 bp), lane10: plasmid pGQ17 was amplified by the same primers as lane8 (control, the size of target DNA fragment should be 1810 bp). The amount of free fatty acids can be doubled in the Synechocystis mutant strain with slr1609 knockout The total free fatty acids of the wild-type strain and the mutant strain GQ8 with slr1609 deletion were extracted using the methods described in the Experimental Procedures. The wild-type and the mutant strain displayed similar growth behaviors (Figure 5A ). However, the content of free fatty acids showed substantial differences between two strains (Figure 5B ). In the slr1609 deletion mutant, the concentration of total free fatty acids was 6.7 ± 0.2 μg/mL/OD, while that of the wild type was 3.5 ± 0.25 μg/mL/OD. The deletion of slr1609 increased free fatty acid accumulation close to two folds. It indicates that the dysfunction of fatty acid activation caused by the deletion of slr1609 results in an increase of free fatty acid accumulation. Figure 5 Comparison of cell growth (A) and production of free fatty acids (B) between Synechocystis sp. PCC 6803 wild-type strain labeled as 6803yu and the mutant strain GQ8 with the deletion of slr1609 gene . As to the contents of the pool of free fatty acids, the amount of unsaturated fatty acids with carbon chain length of C16 and C18 was significantly higher in the slr1609 knockout mutant strain compared to the wild-type strain. Double bonds can only be introduced into free fatty acid coupled to the glycerol backbone of membrane lipids by acyl-lipid-type desaturases. Indicating that unsaturated free fatty acids being released from membrane lipids of senescent or damaged cells, while unsaturated free fatty acids in AAS deletion mutant can not be recycled and incorporated to membrane lipids. In the mutant strain GQ3 with slr1609 over-expression, there is no significant change to the production of free fatty acids compared to the wild-type strain (data not shown). It has been confirmed that free fatty acids are released from membrane lipids in Synechocystis sp. PCC6803 [ 10 ]. Indicating free fatty acid production is not only determined by the fatty acyl-ACP pool size, but also by the biosynthesis of membranes and hydrolysis of membrane lipids which are physiologically regulated. The production of alkanes was significantly reduced in the slr1609 deletion mutant strain Alkanes are the predominant constituents of gasoline, diesel, and jet fuels. Production of alkanes has been reported in a diversity of cyanobacteria, with heptadecane and heptadecene being the most abundant hydrocarbons found in Synechocystis sp. PCC6803. In this pathway fatty acyl-ACP is reduced to a fatty aldehyde by a fatty acyl-ACP reductase (AAR) and then the fatty aldehyde decarbonylase (ADC) is able to convert the aldehyde into an alkane. Besides the fatty acyl-ACP produced by de novo fatty acid synthesis from acetyl-CoA, acyl-ACP synthetase (AAS) is essential for recycling fatty acids into fatty acyl-ACP. The results showed that the production of hydrocarbons was significantly reduced by around 90% in the mutant strain GQ8 with an slr1609 deletion (0.047 ± 0.01 μg/mL/OD) compared with the wild-type strain (0.38 ± 0.07 μg/mL/OD) (Figure 6B ), and this indicates that AAS plays an essential role in alkane biosynthesis. AAS can enhance the total amount of fatty acyl-ACP available for alkane production, and the acyl-ACP formed by AAS activity may be more accessible to the acyl-ACP formed by de novo fatty acid synthesis. Figure 6 Comparison of cell growth (A) and production of fatty alkanes (B) in Synechocystis sp. PCC 6803 wild-type strain labeled as 6803yu, the mutant strain GQ8 with the deletion of slr1609 gene and GQ3 with the overexpression of slr1609 gene . The production of alkanes was not enhanced by the over-expressing slr1609 gene alone in the GQ3 strain (0.39 ± 0.03 μg/mL/OD)(Figure 6B ). Due to the activities of downstream enzymes of the alkane producing pathway, AAR and ADC, are rather low and fatty acyl-ACPs might not be efficiently converted to alkanes [ 11 ]. Fatty acyl-ACPs are also a supplier of fatty acyl groups for biosynthesis of lipid A [ 12 ], phospholipids [ 13 ], and membrane-derived lipo-polysaccharides [ 14 ]. Synechocystis AAS plays an important role in fatty alcohol production Fatty alcohols possess carbon chain length which range from C8 to C22, and can be used as detergents, precursors for synthesis of other chemicals or fuels. We have constructed a fatty-alcohol-producing strain Syn-XT14, by the introduction of a jojoba acyl-CoA reductase gene into wide-type strain in previous work [ 4 ], and the effect of Synechocystis AAS on fatty alcohol production were examined by over-expressing (GQ5) or deleting the slr1609 gene (GQ6) in Syn-XT14. The results showed that fatty alcohol production was enhanced by about 60% in GQ5 (19.8 ± 2.3 μg/L/OD) or decreased by about 60% in GQ6 (4.9 ± 0.1 μg/L/OD) compared with Syn-XT14 (12.5 ± 2.0 μg/L/OD), respectively (Figure 7 ). The data indicates that Synechocystis AAS plays an important role in fatty alcohol production. Figure 7 Comparison of cell growth (A) and production of fatty alcohols (B) in engineered fatty alcohol-producing strain Syn-XT14, GQ5 with the overexpression of slr1609 gene and GQ6 with the deletion of slr1609 gene . Although the native jojoba FAR has a preference for very-long-chain acyl-CoA substrate (C20, C22 and C24), assays of jojoba extracts indicated that it is capable of reducing C16:0-ACP and C18:0-ACP [ 7 ]. It's a reductase with broad substrate specificity. It may be possible that the acyl-ACP produced by AAS can also be accepted as substrate in addition to acyl-CoA by jojoba FAR in engineered Synechocystis strains. It is also possible that the acyl-ACPs, which are synthesized by Synechocystis AAS, could be in turn transacylated to acyl-CoAs by a reverse catalysis of acetyl-CoA-ACP-transacylase (EC 2.3.1.38) type reaction."
} | 3,337 |
36072330 | PMC9441947 | pmc | 5,930 | {
"abstract": "Arbuscular mycorrhizal (AM) symbiosis in soil may be directly or indirectly involved in the reproductive process of sexually reproducing plants (seed plants), and affect their reproductive fitness. However, it is not clear how underground AM symbiosis affects plant reproductive function. Here, we reviewed the studies on the effects of AM symbiosis on plant reproductive fitness including both male function (pollen) and female function (seed). AM symbiosis regulates the development and function of plant sexual organs by affecting the nutrient using strategy and participating in the formation of hormone networks and secondary compounds in host plants. The nutrient supply (especially phosphorus supply) of AM symbiosis may be the main factor affecting plant's reproductive function. Moreover, the changes in hormone levels and secondary metabolite content induced by AM symbiosis can also affect host plants reproductive fitness. These effects can occur in pollen formation and transport, pollen tube growth and seed production, and seedling performance. Finally, we discuss other possible effects of AM symbiosis on the male and female functional fitness, and suggest several additional factors that may be involved in the influence of AM symbiosis on the reproductive fitness of host plants. We believe that it is necessary to accurately identify and verify the mechanisms driving the changes of reproductive fitness of host plant in symbiotic networks in the future. A more thorough understanding of the mechanism of AM symbiosis on reproductive function will help to improve our understanding of AM fungus ecological roles and may provide references for improving the productivity of natural and agricultural ecosystems.",
"conclusion": "Conclusion In conclusion, we reviewed the possible mechanisms and pathways of arbuscular mycorrhizal symbiosis' effects on plant reproductive fitness from the two aspects: the direct or indirect effects of arbuscular mycorrhizal symbiosis on male function (pollen) and female function (seed) of host plants. With the development of genomics and transcriptomics, the future application of various components analysis and protein-hormone interaction methods in plants will help us better understand the ecological functions of arbuscular mycorrhizal symbiosis on plant reproduction. At the same time, combining molecular biological and ecological methods to explore the ecological effects of arbuscular mycorrhizal fungus will allow us to better understand the interaction mechanisms between plants and arbuscular mycorrhizal fungus. The combined application of above-mentioned research methods will make us clearer how arbuscular mycorrhizal symbiosis manipulate the changes in plant reproductive fitness.",
"introduction": "Introduction The arbuscular mycorrhizal (AM) symbiosis formed between plants and AM fungus is a mutually-beneficial symbiosis prevalent in nature which emerged about 400 million years ago (Selosse et al., 2015 ). Nearly 80% of vascular plants in terrestrial ecosystems are able to form and maintain such symbiotic relationships with AM fungus (Bhantana et al., 2021 ). In the symbiotic system formed by plants and AM fungus, plants need to provide the AM symbiosis with carbohydrates produced by photosynthesis, and in return, AM symbiosis has a positive impact on the growth and reproduction of host plants by improving their ability to acquire mineral nutrients (Bhantana et al., 2021 ). Thus, the cooperative relationship can be maintained stably between AM fungus and plants through mutual help (Bhantana et al., 2021 ). Although AM symbiosis occurs underground, the regulation of underground AM symbiosis on aboveground growth and development of host plants cannot be ignored. Currently, there are abundant evidences indicated that underground AM symbiosis has directly positive effects on growth and reproductive traits of host plants (Derelle et al., 2015 ; Bennett and Meek, 2020 ; Vosnjak et al., 2021 ; Chen et al., 2022 ). As a biological factor in soil, AM symbiosis may affect the entire life history of flowering plants involving seed germination, vegetative growth, and sexual reproduction (flowering, pollination, fertilization, fruit set, and seed development) ( Figure 1A ). Sexual reproduction is an important stage of the plant life history, AM fungus may indirectly affect plant reproductive function through the formation of AM symbiosis with host plants and consequently influence plant population dynamics (Bennett and Meek, 2020 ). In particular, the effect of AM symbiosis on sexual reproduction fitness of flowering plants should be paid more attention. It is well-known that the sexual reproduction function of plants is manifested in the male and female functions of plants. The individual plants achieve their sexual reproduction fitness through male function, female function, or both depending on the sexual reproductive system of the plants (hermaphrodite/dioecious) (Varga, 2010 ). Male functional fitness is the ability of pollen production, pollen transfer, pollen germination, and pollen tube growth to fertilize ovules of plants, while female functional fitness is the ability of plants to product mature seeds and the subsequent performance of these seeds (Varga, 2010 ). The investment of flowering plants in sexual reproduction is influenced by individual nutritional status and environmental factors, especially the presence of AM symbiosis is an important factor regulating the process of sexual reproduction. It has been suggested that the male and female functions of flowering plants may be independently affected by AM symbiosis (Koide and Dickie, 2002 ) ( Figure 1B ). At present, it has been widely reported that AM symbiosis are positive relate to reproductive fitness of host plants, and the mechanism that how AM symbiosis affect plant fitness is relatively well-understood. AM symbiosis can assist host plants to successfully achieve reproductive fitness by providing nutrient, regulating hormone balance, and other secondary product production (Stanley et al., 1993 ; Varga and Kytöviita, 2014 ; Bennett and Meek, 2020 ). Figure 1 Effects of AM symbiosis on the reproductive function of host plants. (A) Represents the way in which AM symbiosis may participate in the realization of male and female functions during plant reproduction, and I represents the pollen production; II represents that changes in floral characteristics may potentially affect pollinators pollination behavior; III represents the pollen germination and pollen tube growth on stigma; IV represents the fruit formation and seed development after successful fertilization; V represents the seeds production; VI represents the seeds quality; VII represents the seeds germination; VIII represents the successful establishment of seedlings after germination; IX represents the growth of offspring seedlings. (B) Simplifies the AM symbiosis effects on plant reproduction fitness. AM symbiosis may directly and indirectly affect the male and female functions of plants through nutrient supply and hormone regulation, thus affecting their reproductive fitness. The Mutual adaptation and coordinate between male and female function may also be affected by the AM symbiosis, thus affecting the reproductive fitness of mycorrhizal plants. (C) Indicates that AM symbiosis may directly or indirectly regulate the reproductive function of plants through nutrient supply (phosphorus and other essential elements), regulation of hormone network and synthesis of secondary compounds (such as amino acids, proteins, terpenoids, and flavonoids) during the realization of male and female functional fitness of plants. It is generally believed that underground AM symbiosis may have direct and indirect effects on male and female functional fitness of host plants. The direct effect is that AM symbiosis promotes the host plant's ability to acquire nutrients and directly drives the host plant to increase resource investment in sexual reproduction. For example, AM symbiosis can improve the uptake and accumulation of major elements (nitrogen, phosphorus, and potassium) and trace elements (zinc, sulfur, copper, iron, calcium, and manganese) in soil by host plants (Chen et al., 2017 ; Turrini et al., 2018 ; Bhantana et al., 2021 ). This positive effect on mineral nutrient uptake directly promotes the growth and development of host plants, and changes the resource acquisition and allocation strategies of host plants, which will make host plants likely to invest more resources in reproductive functions, thereby improve male and female function fitness. For example, increased phosphorus content may have positive effects on flower bud formation and development, flower number, pollen size, and seed production, as it has been shown in the interaction between the non-mycorrhizal root endophytic fungus Piriformospora indica and Cyclamen persicum (Ghanem et al., 2014 ). The indirect effects of AM symbiosis on the male and female functions of plants may be that it regulates the synthesis and distribution of secondary compounds in plants by affecting related metabolic pathways and gene expression in host plants (Zouari et al., 2014 ; Bennett and Meek, 2020 ). Some studies have shown that AM symbiosis can regulate gene expression and indirectly participate in various metabolic processes of host plants. For example, the functions of photosynthesis, nutrient transport, amino acid synthesis, and terpenoid metabolism were enhanced after mycorrhizal colonization, which undoubtedly affected the growth and development of host plants (Zouari et al., 2014 ). In particular, AM symbiosis changes the levels of some endogenous hormones (e.g., auxin, gibberellin, etc.) in host plants, which regulate the formation and function of sexual organs (Nuortila et al., 2004 ; Foo et al., 2013 ). Thus, AM symbiosis can influence all components of plant sexual reproduction including pollen delivery, pollen germination, pollen tube growth, fruit and seed production, seed germination, etc. by regulating hormones, phenolic compounds, and secondary metabolites production and epigenetic modifications (Varga and Soulsbury, 2017 ; Cui et al., 2019 ; Bennett and Meek, 2020 ; Pons et al., 2020 ; Ran et al., 2022 ; Rashidi et al., 2022 ) ( Figure 1C ). These results are encouraging us to furtherly confirm that AM symbiosis can have a profound impact on plant reproductive fitness. However, we still lack a deeper understanding of how these two pathways work together. Although it is true that increased nutrient supply or altered hormone levels can affect plant reproductive function from one pathway, but both pathways may coexist in the plant-AM symbiosis system. When multiple regulatory pathways (nutrients, hormones, other gene products, etc.) exist together, does one pathway get overridden by the other or do these effects regulate different aspects individually or do they all work together? That's still unclear to us. Most flowering plants can produce offspring through sexual reproduction (Hiscock, 2011 ). AM symbiosis may be involved in various stages of the realization of reproductive fitness of host plants. The changes of male and female functional traits of host plant in the presence of soil AM symbiosis were focused on in this paper. Understanding the trade-off strategy between vegetative growth and sexual reproduction fitness in plants will be a meaningful reference for future efforts to use AM symbiosis to enhance plant productivity. Therefore, we reviewed the relevant researches on how soil AM symbiosis affect male and female fitness during sexual reproduction of host plants. We aimed to understand whether the promotion of host plant traits by AM symbiosis leads to changes in male and female functional fitness. Finally, we propose future research directions that will help to expand existing research area and enable us to more fully understand the feedback mechanism of plant sexual reproduction fitness to belowground AM symbiosis."
} | 3,011 |
30038374 | PMC6056493 | pmc | 5,931 | {
"abstract": "Geothermal springs are model ecosystems to investigate microbial biogeography as they represent discrete, relatively homogenous habitats, are distributed across multiple geographical scales, span broad geochemical gradients, and have reduced metazoan interactions. Here, we report the largest known consolidated study of geothermal ecosystems to determine factors that influence biogeographical patterns. We measured bacterial and archaeal community composition, 46 physicochemical parameters, and metadata from 925 geothermal springs across New Zealand (13.9–100.6 °C and pH < 1–9.7). We determined that diversity is primarily influenced by pH at temperatures <70 °C; with temperature only having a significant effect for values >70 °C. Further, community dissimilarity increases with geographic distance, with niche selection driving assembly at a localised scale. Surprisingly, two genera ( Venenivibrio and Acidithiobacillus ) dominated in both average relative abundance (11.2% and 11.1%, respectively) and prevalence (74.2% and 62.9%, respectively). These findings provide an unprecedented insight into ecological behaviour in geothermal springs, and a foundation to improve the characterisation of microbial biogeographical processes.",
"conclusion": "Conclusion This study presents data on both niche and neutral drivers of microbial biogeography in 925 geothermal springs at a near-national scale. Our comprehensive data set, with sufficient sampling density and standardised methodology, is the first of its kind to enable a robust spatio-chemical statistical analysis of microbial communities at the regional level across broad physicochemical gradients. Unequivocally, pH drives diversity and community complexity structures within geothermal springs. This effect, however, was only significant at temperatures < 70 °C. We also identified specific taxa associations and finally demonstrated that geochemical signatures can be indicative of community composition. Although a distance-decay pattern across the entire geographic region indicated dispersal limitation, the finding that 293 adjacent community pairs exhibited up to 100% dissimilarity suggests niche selection drives microbial community composition at a localised scale (e.g. within geothermal fields). This research provides a comprehensive dataset that should be used as a foundation for future studies (e.g. diversification and drift elucidation on targeted spring taxa). It complements the recently published Earth Microbiome Project 44 by expanding our knowledge of the biogeographical constraints on aquatic ecosystems using standardised quantification of broad physicochemical spectrums. There is also potential to use the two studies to compare geothermal ecosystems on a global scale. Finally, our research provides a springboard to assess the cultural, recreational, and resource development value of the microbial component of geothermal springs, both in New Zealand and globally. Many of the features included in this study occur on culturally-important and protected land for Māori, therefore this or follow-on future projects may provide an avenue for exploration of indigenous knowledge, while assisting in conservation efforts and/or development.",
"introduction": "Introduction Biogeography identifies patterns of diversity across defined spatial or temporal scales in an attempt to describe the factors which influence these distributions. Recent studies have shown that microbial community diversity is shaped across time and space 1 , 2 via a combination of environmental selection, stochastic drift, diversification, and dispersal limitation 3 , 4 . The relative impact of these ecological drivers on diversity is the subject of ongoing debate, with differential findings reported across terrestrial, marine, and human ecosystems 5 – 8 . Geothermally-heated springs are ideal systems to investigate microbial biogeography, because, in comparison to terrestrial environments, they represent discrete, aquatic habitats with broad geochemical and physical gradients distributed across proximal and distal geographic distances. The relatively constrained microbial community structures, typical of geothermal springs compared to soils and sediments, also allow for the robust identification of diversity trends. Separate studies have each implicated temperature 9 – 11 , pH 12 , and seasonality 13 as the primary drivers of bacterial and archaeal communities in these ecosystems; with niche specialisation observed within both local and regional populations 14 , 15 . Other geochemical variables, particularly hydrogen 16 , 17 and nitrogen 18 , 19 , may also contribute to community structure. Concurrently, the stochastic action of microbial dispersal is thought to be a significant driver behind the distribution of microorganisms 20 , with endemism and allopatric speciation reported in intercontinental geothermal springs 21 , 22 . It is important to note that significant differences have been found between aqueous and soil/sediment samples from the same springs 10 , 12 , 23 ; emphasising that the increased relative homogeneity of aqueous samples make geothermal water columns excellent candidate environments for investigating large-scale taxa–geochemical associations. Despite these findings, a lack of sampling quantity/density and physicochemical gradient scales, uniformity in sampling methodology, and a concurrent assessment of geographic distance, within a single study, has hindered a holistic view of microbial biogeography from developing. The Taupō Volcanic Zone (TVZ) is a region rich in geothermal springs and broad physicochemical gradients spanning 8000 km 2 in New Zealand’s North Island (Fig. 1 ), making it a tractable model system for studying microbial biogeography. The area is a rifting arc associated with subduction at the Pacific-Australian tectonic plate boundary, resulting in a locus of intense magmatism 24 . The variable combination of thick, permeable volcanic deposits, high heat flux, and an active extensional (crustal thinning) setting favours the deep convection of groundwater and exsolved magmatic volatiles that are expressed as physicochemically-heterogeneous surface features in geographically distinct geothermal fields 25 . Previous microbiological studies across the region have hinted at novel diversity and function present within some of these features 26 – 30 , however investigations into the biogeographical drivers within the TVZ are sparse and have focused predominantly on soil/sediments or individual hotsprings 9 , 11 , 31 . Fig. 1 Map of the Taupō Volcanic Zone (TVZ), New Zealand. The geothermal fields from which samples were collected are presented in yellow. All sampled geothermal springs ( n = 1019) are marked by red circles. The panel insert highlights the location of the TVZ in the central north island of New Zealand. The topographic layers for this map were obtained from Land Information New Zealand (LINZ; CC-BY-4.0) and the TVZ boundary defined using data from Wilson et al. 24 Here, we report the diversity and biogeography of microbial communities found in geothermal springs, collected as part of the 1000 Springs Project. This project aimed to catalogue the microbial biodiversity and physicochemistry of New Zealand geothermal springs to serve as a conservation, scientific, and indigenous cultural knowledge repository; a publicly accessible database of all springs surveyed is available online ( http://1000Springs.org.nz ). In order to determine the influence of biogeographical processes on bacterial and archaeal diversity and community structure within geothermal springs, we collected water-columns and metadata from 1019 spring samples within the TVZ over a 21-month period using rigorously standardised methodologies. We then performed community analysis of the bacterial and archaeal population (16S rRNA gene amplicon sequencing) and quantified 46 physicochemical parameters for each sample. This work represents the largest known microbial ecology study on geothermal aquatic habitats at a regional scale and complements a parallel study on protist diversity in New Zealand geothermal springs 32 . Our results demonstrate both the relative influence of physicochemical parameters (e.g. pH) and the effect of geographic isolation on the assemblage of communities in these extreme ecosystems. Collectively, these findings expand our knowledge of the constraints that govern universal microbial biogeographical processes.",
"discussion": "Results and Discussion Geothermal feature sampling Recent biogeography research has demonstrated that microbial diversity patterns are detectable and are influenced by both deterministic 6 and stochastic processes 7 . A lack of consensus on the relative impact of these factors, however, has been exacerbated by the absence of data across broad physicochemical gradients, and sampling scales and density across both geographic distance and habitat type. The inherent heterogeneity of terrestrial soil microbial ecosystems 33 further confounds attempts to distinguish true taxa–geochemical associations. To provide greater resolution to the factors driving microbial biogeography processes, we collected 1019 geothermal water-column samples from across the TVZ and assessed physicochemical and microbial community composition (Fig. 1 ). Samples included representatives of both extreme pH (< 0–9.7) and temperature (13.9–100.6 °C) (Supplementary Fig. 1 ). The filtering of low-quality and temporal samples yielded a final data set of 925 individual geothermal springs for spatial-statistical analysis (details can be found in Supplementary Methods). From these 925 springs, a total of 28,381 operational taxonomic units (OTUs; 97% similarity) were generated for diversity studies. Microbial diversity is principally driven by pH, not temperature Reduced microbial diversity in geothermal springs is often attributed to the extreme environmental conditions common to these areas. Temperature and pH are reported to be the predominant drivers of microbial diversity 9 , 34 , but their influence relative to other parameters has not been investigated over large geographic and physicochemical scales with appropriate sample density. Our analysis of microbial richness and diversity showed significant variation spanning pH, temperature, and geographical gradients within the TVZ (richness: 49–2997 OTUs, diversity: 1.1–7.3 Shannon index; Supplementary Figs. 2 and 3 ). As anticipated, average OTU richness (386 OTUs; Supplementary Fig. 4 ) was reduced in comparison to studies of non-geothermal temperate terrestrial 35 and aquatic 1 environments. Further, OTU richness was maximal at the geothermally-moderate temperature of 21.5 °C and at circumneutral pH 6.4. This is consistent with the hypothesis that polyextreme habitats prohibit the growth of most microbial taxa, a trend reported in both geothermal and non-geothermal environments alike 5 , 9 . A comparison of 46 individual physicochemical parameters (Supplementary Table 1 ) confirmed pH as the most significant factor influencing diversity (16.4%, linear regression: P < 0.001; Supplementary Fig. 3 ), with diversity increasing from acidic to alkaline pH. Further, multiple regression analysis showed \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{NO}}_3^ -$$\\end{document} NO 3 - , turbidity (TURB), oxidation–reduction potential (ORP), dissolved oxygen (dO), \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{NO}}_2^ -$$\\end{document} NO 2 - , Si, and Cd also had meaningful contributions (Supplementary Table 2 ). Cumulatively, along with pH, these factors accounted for 26.6% of the observed variation in Shannon diversity. Correlation of pH with Shannon index (Pearson’s coefficient: | r | = 0.41, P < 0.001) and significance testing between samples binned by pH increments (Kruskal–Wallis: H = 179.4, P < 0.001; Supplementary Fig. 1 ) further confirmed pH as a major driver of variation in alpha diversity. This finding is consistent with reports of pH as the primary environmental predictor of microbial diversity in several ecosystems, both in New Zealand and globally (e.g. soil 5 , 36 , water 32 , 37 , alpine 38 , 39 ). It has been previously hypothesised that pH has significant influence on microbial community composition because changes in proton gradients will drastically alter nutrient availability, metal solubility, or organic carbon characteristics 5 . Similarly, acidic pH will also reduce the number of taxa observed due to the decreased number that can physiologically tolerate these conditions 40 compared to non-acidic habitats. Here, we demonstrate that pH had the most significant effect on diversity across all springs measured, but due to our high sampling frequency, we see this influence diminishes at temperatures > 70 °C (Fig. 2 ). Inversely, the effect of temperature on diversity was lessened in springs where pH was < 4 (Supplementary Fig. 5 ). There is some evidence that suggests thermophily predates acid tolerance 40 , 41 , thus it is possible the added stress of an extreme proton gradient across cell membranes has constrained the diversification of the thermophilic chemolithoautotrophic organisms common to these areas 42 . Indeed, a recent investigation of thermoacidophily in Archaea suggests hyperacidophily (growth < pH 3.0) may have only arisen as little as ~0.8 Ga 40 , thereby limiting the opportunity for microbial diversification; an observation highlighted by the paucity of these microorganisms in extremely acidic geothermal ecosystems 11 , 40 . It is also important to note that salinity has previously been suggested as an important driver of microbial community diversity 43 , 44 . The quantitative data in this study showed only minimal influence of salinity (proxy as conductivity (COND)) on diversity (linear regression: R 2 = 0.001, P = 0.2720; Supplementary Table 1 ), bearing in mind that the majority of the geothermal spring samples in this study had salinities substantially less than that of seawater. Fig. 2 Alpha and beta diversity as a function of pH and temperature. a pH against alpha diversity via Shannon index of all individual springs ( n = 925) in 10 °C increments, with linear regression applied to each increment. b Non-metric multidimensional scaling (NMDS) plot of beta diversity (via Bray–Curtis dissimilarities) between all individual microbial community structures sampled ( n = 925) The relationship between temperature and alpha diversity reported in this research starkly contrasts a previous intercontinental study comparing microbial community diversity in soil/sediments from 165 geothermal springs 9 , which showed that a strong relationship ( R 2 = 0.40–0.44) existed. In contrast, our data across the entire suite of samples, revealed that temperature had no significant influence on observed community diversity ( R 2 = 0.002, P = 0.201; Supplementary Fig. 3 , Supplementary Table 1 ). This result increased marginally for archaeal-only diversity ( R 2 = 0.013, P = 0.0005), suggesting that temperature has a more profound effect on this domain than it does on bacteria. However, the primers used in this study are known to be unfavourable towards some archaeal clades 45 , therefore it is likely that extensive archaeal diversity remains undetected in this study. The lack of influence of temperature on whole community diversity was further substantiated via multiple linear modelling (Supplementary Table 2 ), and significance and correlation testing (Kruskal–Wallis: H = 16.2, P = 0.039; Pearson’s coefficient: | r | = 0.04, P = 0.201). When samples were split into pH increments, like Sharp et al. 9 , we observed increasing temperature only significantly constrained diversity above moderately acidic conditions (pH > 4; Supplementary Fig. 5 ). However, the magnitude of this effect was, in general, far less than previously reported and is likely a consequence of the sample type (e.g. soil/sediments versus aqueous) and density of samples processed 12 . Many samples from geothermal environments are recalcitrant to traditional DNA extraction protocols and research in these areas has therefore focused on those with greater biomass abundance 9 , 34 (i.e. soils, sediments, streamers, or biomats). Whereas aqueous samples typically exhibit a more homogenous chemistry and community structure, the heterogeneity of terrestrial samples is known to affect microbial population (e.g. particle size, depth, nutrient composition) 33 . Our deliberate use of aqueous samples extends the results of previous small-scale work 10 , 31 and also permits the robust identification of genuine taxa–geochemical relationships in these environments. Microbial communities are influenced by pH, temperature, and source fluid Throughout the TVZ, beta diversity correlated more strongly with pH (Mantel: ρ = 0.54, P < 0.001) than with temperature (Mantel: ρ = 0.19, P < 0.001; Fig. 2 , Supplementary Table 3 ). This trend was consistent in pH- and temperature-binned samples (ANOSIM: | R | = 0.46 and 0.18, respectively, P < 0.001; Supplementary Fig. 6 ); further confirming pH, more so than temperature, accounted for observed variations in beta diversity. Congruent with our finding that pH influences alpha diversity at lower temperatures (< 70 °C), the effect of temperature reducing beta diversity had greater significance above 80 °C (Wilcox: P < 0.001; Supplementary Fig. 6 ). The extent of measured physicochemical properties across 925 individual habitats, however, allowed us to explore the environmental impact on community structures beyond just pH and temperature. Permutational multivariate analysis of variance in spring community assemblages showed that pH (12.4%) and temperature (3.9%) had the greatest contribution towards beta diversity, followed by ORP (1.4%), \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{SO}}_4^{2 - }$$\\end{document} SO 4 2 - (0.8%), TURB (0.8%), and As (0.7%) ( P < 0.001; Supplementary Table 4 ). Interestingly, constrained correspondence analysis of the 15 most significant, non-collinear, and variable parameters (Supplementary Tables 4 and 5 ; pH, temperature, TURB, ORP, \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{SO}}_4^{2 - }$$\\end{document} SO 4 2 - , \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{NO}}_3^ -$$\\end{document} NO 3 - , As, \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{NH}}_4^ +$$\\end{document} NH 4 + , \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{HCO}}_3^ -$$\\end{document} HCO 3 - , H 2 S, COND, Li, Al, Si, and \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{PO}}_4^{3 - }$$\\end{document} PO 4 3 - ), along with geothermal field locations, only explained 10% of variation in beta diversity (Fig. 3 ), indicating physicochemistry, or at least the 46 parameters measured, were not the sole drivers of community composition. Fig. 3 Constrained correspondence analysis (CCA) of beta diversity with significant physicochemistry. a A scatter plot of spring community dissimilarities ( n = 923), with letters corresponding to centroids from the model for geothermal fields (A–Q; White Island, Taheke, Tikitere, Rotorua, Waimangu, Waikite, Waiotapu, Te Kopia, Reporoa, Orakei Korako, Whangairorohea, Ohaaki, Ngatamariki, Rotokawa, Wairakei-Tauhara, Tokaanu, Misc). Coloured communities are from fields represented in the subpanel. Constraining variables are plotted as arrows (COND: conductivity, TURB: turbidity), with length and direction indicating scale and area of influence each variable had on the model. b The top panel represents a subset of the full CCA model, with select geothermal fields shown in colour (including 95% confidence intervals). The bottom panel shows their respective geochemical signatures as a ratio of chloride (Cl − ), sulfate \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\left( {{\\mathrm{SO}}_4^{2 - }} \\right)$$\\end{document} SO 4 2 - , and bicarbonate \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\left( {{\\mathrm{HCO}}_3^ - } \\right)$$\\end{document} HCO 3 - Geothermal fields are known to express chemical signatures characteristic of their respective source fluids 46 , implying autocorrelation could occur between location and geochemistry. We therefore investigated whether typical geochemical conditions exist for springs within the same geothermal field and whether specific microbial community assemblages could be predicted. Springs are usually classified according to these source fluids; alkaline-chloride or acid-sulfate. High-chloride features are typically sourced from magmatic waters and have little interaction with groundwater aquifers. At depth, water–rock interactions can result in elevated bicarbonate concentrations and, consequently, neutral to alkaline pH in surface features. Acid-sulfate springs (< pH 3), in contrast, form as steam-heated groundwater couples with the eventual oxidation of hydrogen sulfide into sulfate (and protons). Rarely, a combination of the two processes can occur; leading to intermediate pH values 47 . It is unknown, however, whether these source fluid characteristics are predictive of their associated microbial ecosystems. Bray–Curtis dissimilarities confirmed that, like alpha diversity (Kruskal–Wallis: H = 240.7, P < 0.001; Fig. 4 ), community structures were significantly different between geothermal fields (ANOSIM: | R | = 0.26, P < 0.001; Supplementary Fig. 7 ). Gradient analysis comparing significant geochemical variables and geography further identified meaningful intra-geothermal field clustering of microbial communities (95% CI; Fig. 3 and Supplementary Fig. 8 ). Further, characteristic geochemical signatures from these fields were identified and analysis suggests they could be predictive of community composition. For example, the Rotokawa and Waikite geothermal fields (approx. 35 km apart) (Fig. 3 position N and F) display opposing ratios of \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{HCO}}_3^ -$$\\end{document} HCO 3 - , \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{SO}}_4^{2 - }$$\\end{document} SO 4 2 - , and Cl − , with corresponding microbial communities for these sites clustering independently in ordination space. Despite this association, intra-field variation in both alpha and beta diversity also occurred at other geothermal sites where geochemical signatures were not uniform across local springs (e.g. Rotorua, Fig. 3 position D), demonstrating that correlation does not necessarily always occur between locational proximity and physicochemistry. Fig. 4 Alpha and beta diversity as a function of geographic distance. a Alpha diversity scales (via Shannon index) across individual springs ( n = 925), split by geothermal fields. Fields are ordered from north to south ( H : Kruskal–Wallis test). b A distance-decay pattern of beta diversity (via Bray–Curtis dissimilarities of 925 springs) against pairwise geographic distance in metres, with linear regression applied ( m = slope). Geographic distance is split into bins to aid visualisation of the spread Aquificae and Proteobacteria are abundant and widespread To determine whether individual microbial taxa favoured particular environmental conditions and locations, we first assessed the distribution of genera across all individual springs. Within 17 geothermal fields and 925 geothermal features, 21 phyla were detected with an average relative abundance > 0.1% (Fig. 5 ). We found that two phyla and associated genera, Proteobacteria ( Acidithiobacillus spp.) and Aquificae ( Venenivibrio, Hydrogenobaculum , Aquifex spp.), dominated these ecosystems (65.2% total average relative abundance across all springs), composing nine of the 15 most abundant genera > 1% average relative abundance (Table 1 ). Considering the broad spectrum of geothermal environmental conditions sampled in this study (we assessed microbial communities in springs across a pH gradient of nine orders of magnitude and a temperature range of ~87 °C) and the prevalence of these taxa across the region, this result was surprising. Proteobacteria was the most abundant phylum across all samples (34.2% of total average relative abundance; Table 1 ), found predominantly at temperatures less than 50 °C (Supplementary Fig. 9 ). Of the 19 most abundant proteobacterial genera (average relative abundance > 0.1%), the majority are characterised as aerobic chemolithoautotrophs, utilising either sulfur species and/or hydrogen for metabolism. Accordingly, the most abundant (11.1%) and prevalent (62.9%) proteobacterial genus identified was Acidithiobacillus 48 , a mesophilic-moderately thermophilic, acidophilic autotroph that utilises reduced sulfur compounds, iron, and/or hydrogen as energy for growth. Fig. 5 Taxonomic association with location and physicochemistry. The heat map displays positive (red) and negative (blue) association of genus-level taxa (> 0.1% average relative abundance) with each geothermal field and significant environmental variables, based on Z -scores of abundance log ratios. Each taxon is colour-coded to corresponding phylum on the approximately maximum-likelihood phylogenetic tree Table 1 Average relative abundance and prevalence of phyla and genera Phylum Genus Abundance SD Max Prevalence Aquificae \n Venenivibrio \n 0.112 0.231 0.968 0.742 Proteobacteria \n Acidithiobacillus \n 0.111 0.242 0.994 0.629 Aquificae \n Hydrogenobaculum \n 0.100 0.235 0.999 0.608 Aquificae \n Aquifex \n 0.086 0.212 0.971 0.497 Deinococcus-Thermus \n Thermus \n 0.025 0.071 0.732 0.552 Proteobacteria \n Thiomonas \n 0.024 0.101 0.941 0.396 Proteobacteria \n Desulfurella \n 0.022 0.067 0.758 0.497 Crenarchaeota Sulfolobaceae (f) 0.020 0.091 0.951 0.416 Euryarchaeota Thermoplasmatales (o) 0.019 0.059 0.495 0.539 Proteobacteria \n Thiovirga \n 0.015 0.077 0.816 0.374 Proteobacteria Hydrogenophilaceae (f) 0.015 0.072 0.704 0.406 Thermodesulfobacteria \n Caldimicrobium \n 0.015 0.052 0.651 0.519 Proteobacteria \n Hydrogenophilus \n 0.013 0.045 0.432 0.484 Thermotogae \n Mesoaciditoga \n 0.011 0.033 0.286 0.410 Parcubacteria Parcubacteria (p) 0.010 0.024 0.193 0.608 Only taxa above a 1% average compositional threshold are shown. Maximum abundance of each taxon within individual features and standard deviation across all 925 springs. Where taxonomy assignment failed to classify to genus level, the closest assigned taxonomy is shown SD = standard deviation, f = family, o = order, p = phylum Aquificae (order Aquificales) was the second most abundant phylum overall (31% average relative abundance across 925 springs) and included three of the four most abundant genera; Venenivibrio , Hydrogenobaculum , and Aquifex (11.2%, 10.0%, and 8.6%, respectively; Table 1 ). As Aquificae are thermophilic ( T opt 65–85 °C) 49 , they were much more abundant in warmer springs (> 50 °C; Supplementary Fig. 9 ). The minimal growth temperature reported for characterised Aquificales species ( Sulfurihydrogenibium subterraneum and Sulfurihydrogenibium kristjanssonii ) 49 is 40 °C and may explain the low Aquificae abundance found in springs less than 50 °C. Terrestrial Aquificae are predominately microaerophilic chemolithoautotrophs that oxidise hydrogen or reduced sulfur compounds; heterotrophy is also observed in a few representatives 49 . Of the 14 currently described genera within the Aquificae, six genera were relatively abundant in our dataset (average relative abundance > 0.1%; Fig. 5 ): Aquifex , Hydrogenobacter , Hydrogenobaculum and Thermocrinis (family Aquificaceae), and Sulfurihydrogenibium and Venenivibrio (family Hydrogenothermaceae). No signatures of the Desulfurobacteriaceae were detected and is consistent with reports that all current representatives from this family are associated with deep-sea or coastal thermal vents 49 . Venenivibrio (OTUs; n = 111) was also the most prevalent and abundant genus across all communities (Table 1 ). This taxon, found in 74.2% ( n = 686) of individual springs sampled, has only one cultured representative, Venenivibrio stagnispumantis (CP.B2 T ), which was isolated from the Waiotapu geothermal field in the TVZ 29 . The broad distribution of this genus across such a large number of habitats was surprising, as growth of the type strain is only supported by a narrow set of conditions (pH 4.8–5.8, 45–75 °C). Considering this, and the number of Venenivibrio OTUs detected, we interpret this result as evidence there is substantial undiscovered phylogenetic and physiological diversity within the genus. The ubiquity of Venenivibrio suggests that either the metabolic capabilities of this genus extend substantially beyond those described for the type strain, and/or that many of the divergent taxa could be persisting and not growing under conditions detected in this study 30 , 50 . Geochemical and geographical associations exist at the genus level The two most abundant phyla, Proteobacteria and Aquificae, were found to occupy a characteristic ecological niche (< 50 and > 50 °C, respectively, Supplementary Fig. 9 ). To investigate specific taxa–geochemical associations beyond just temperature and pH, we applied a multivariate linear model to determine enrichment of taxa in association with geothermal fields and other environmental data (Fig. 5 ). The strongest associations between taxa and chemistry ( Z -score > 4) were between Nitrospira –nitrate \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\left( {{\\mathrm{NO}}_3^ - } \\right)$$\\end{document} NO 3 - and Nitratiruptor –phosphate \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\left( {{\\mathrm{PO}}_4^{3 - }} \\right)$$\\end{document} PO 4 3 - . Nitrospira oxidises nitrite to nitrate and therefore differential high abundance of this taxon in nitrate-rich environments is expected. Further, the positive Nitratiruptor – \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{PO}}_4^{3 - }$$\\end{document} PO 4 3 - relationship suggests phosphate is a preferred nutritional requirement for this chemolithoautotroph 51 and informs future efforts to isolate members of this genus would benefit from additional phosphate or the presence of reduced phosphorous compounds in the culture medium 52 , 53 . Thermus and Hydrogenobaculum were the only bacterial taxa to differentially associate (compared to other taxa) positively and negatively with pH respectively. This is consistent with the lack of acidophily phenotype (pH < 4) reported in Thermus spp. 54 and conversely, the preferred acidic ecological niche of Hydrogenobaculum 55 . Aquifex was the only genus to display above average association with temperature, confirming abundance of this genera is significantly enhanced by hyperthermophily 56 . Similar to the chemical–taxa associations discussed above, differential abundance relationships were calculated with respect to individual geothermal fields (Fig. 5 ). A geothermal field, which contains springs across the pH scale (i.e. Rotorua), was closely associated with the highly abundant and prevalent Acidithiobacillus and Venenivibrio . On the other hand, a predominantly acidic geothermal system (i.e. Te Kopia), produced the only positive associations with “ Methylacidiphilum” (Verrucomicrobia), Acidimicrobium (Actinobacteria), Terrimonas (Bacteroidetes), and Halothiobacillus (Proteobacteria). These relationships are likely describing sub-community requirements that are otherwise not captured by conventional spatial-statistical analysis, therefore providing insight into previously unrecognised microbe–niche interactions. Distance-decay patterns differ at local and regional scales Environmental selection, ecological drift, diversification, and dispersal limitation all contribute to distance-decay patterns 4 . While several studies have shown microbial dispersal limitations and distance-decay patterns exist in diverse geothermal and non-geothermal environments (e.g. hotsprings 21 , freshwater streams 1 , salt marshes), the point of inflection between dispersal limitation and selection, at regional and local geographic scales, remains under-studied. We identified a positive distance-decay trend with increasing geographic distance between 925 geothermal spring communities across the TVZ region (linear regression: m = 0.031, P < 0.001; Fig. 4 ). This finding strongly suggests that dispersal limitation exists between individual geothermal fields. Increasing the resolution to within individual fields, distance-decay patterns are negligible compared to the regional scale (Supplementary Table 6 ). Interestingly, the greatest pairwise difference ( y = 1) between Bray–Curtis dissimilarities was also observed in springs classified as geographically-adjacent (< 1.4 m). In the 293 geothermal spring pairs separated by < 1.4 m, temperature had a greater correlation with beta diversity than pH (Spearman’s coefficient: ρ = 0.44 and 0.30, respectively, P < 0.001). This result illustrates the stark spatial heterogeneity and selective processes that can exist within individual geothermal fields. Congruently, each OTU was detected in an average of only 13 springs (Supplementary Fig. 4 ). We propose that physical dispersal within geothermal fields is therefore not limiting, but the physicochemical diversity of geothermal springs acts as a barrier to the colonisation of immigrating taxa. However, even between some neighbouring springs with similar (95% CI) geochemical signatures, we did note some dissimilar communities were observed (for example, Waimangu geothermal field; Fig. 3 position E). These differing observations can be explained by either one of three ways: Firstly, the defining parameter driving community structure was not one of the 46 physicochemical variables measured in this study (e.g. dissolved organic carbon or hydrogen); secondly, through the process of dispersal, the differential viability of some extremophilic taxa restricts gene flow and contributes to population genetic drift within geothermal fields 57 . We often consider “extremophilic” microorganisms living in these geothermal environments as the epitome of hardy and robust. In doing so, we overlook that their proximal surroundings (i.e. immediately outside the host spring) may not be conducive to growth and survival 58 and therefore the divergence of populations in neighbouring, chemically-homogenous spring ecosystems is plausible. Thirdly, aeolian immigration 20 from the non-geothermal surrounding environment could alter the perceived composition of a community, even when immigrants are not competing to survive. Future work could be undertaken to understand individual population responses to community-wide selective pressures and the temporal nature of ecosystem functioning."
} | 9,552 |
38498725 | PMC10998617 | pmc | 5,933 | {
"abstract": "Significance Many natural functional materials comprise hierarchical micro- and nanostructures that are integral parts of biological surfaces. Distinctively, leafhoppers excrete brochosomes and actively use them as deployable materials on their body surfaces. Brochosomes are hollow, buckyball-shaped nanoscopic spheroids with through-holes on their surfaces. Since their discovery in the 1950s, understanding the functional significance of brochosomal geometry has remained elusive. Here, we demonstrate that the geometry and the through-hole design of brochosomes effectively reduce light reflection. Furthermore, brochosomes are a biological example exhibiting short-wavelength, low-pass filter functionality. The unique geometry of brochosomes provides a distinct approach for bioinspired optical manipulation. This represents a development distinct from the antireflective moth-eye effect (1973) and offers insight for engineering deployable optical materials.",
"conclusion": "Conclusion In summary, we have fabricated high-fidelity synthetic brochosomes using the two-photon polymerization 3D printing method and demonstrated that the hierarchical structures of brochosomes can effectively reduce light reflection through both Mie scattering and through-hole absorption effects. Brochosomes are biological structures that exhibit both short-wavelength, low-pass optical absorption and broadband antireflection functions. Inspired by the intricate 3D architecture of brochosomes, we anticipate that synthetic brochosomes could lead to the development of a class of bioinspired optical materials capable of interacting with a broad range of electromagnetic spectrum. Potential applications include omnidirectional antireflective coatings, light-absorbing materials, optical encryption, and multispectral camouflage.",
"discussion": "Discussion Geometrical Consistency of Natural Brochosomes. Our observations reveal that both the diameters of the brochosomes and their through-holes consistently fall within the range of hundreds of nanometers across different leafhopper species, showing weak dependence on leafhopper body length, as shown in Fig. 4 . Importantly, we found that the ratio between the diameter of natural brochosomes (~300 nm to ~700 nm) and the wavelength of UV–visible light (~300 nm to ~700 nm), ranges from ~ 0.4 to ~2.3. This places them within the effective Mie scattering regime for UV–visible light, as depicted by Eq. 1 ( Fig. 4 A and SI Appendix , Fig. S5 ). Furthermore, we observed that the through-hole sizes (~100 nm to ~280 nm) fall within the range that can effectively reduce UV light reflection, as depicted by Eq. 3 ( Fig. 4 B ). These findings suggest that the diameters of natural brochosomes and their corresponding through-hole sizes may have evolved to effectively reduce light reflection in both the visible (~400 nm < λ < ~700 nm) and UV (~300 nm < λ < ~400 nm) range. Fig. 4. Characteristic lengths of natural brochosomes and the corresponding optical regimes. ( A ) A scatter plot showing the diameters of brochosomes (closed red circles), and ( B ) the diameters of through-hole (open blue circles) plotted against leafhopper body length across different leafhopper species. These two characteristic lengths of natural brochosomes remain relatively consistent, with brochosome diameters ranging from approximately 300 nm to 700 nm, and through-hole diameters ranging from about 100 nm to 280 nm, regardless of the leafhopper body length (ranging from approximately 3 mm to 9 mm). The red area highlights the region of Mie scattering of visible light, as predicted by Eq. 1 , while the blue area indicates UV absorption via the through-hole effect, as predicted by Eq. 3 . The color bar represents the visual spectra of some of the leafhoppers’ predators, including birds and lizards ( 42 , 43 ). The characteristic lengths of brochosomes were obtained from our experimental measurements and the literature ( 18 ). To validate these analyses on biological samples, we measured the specular reflection on leafhopper wings, both with and without brochosomes ( SI Appendix , Figs. S11 and S12 ), to gain qualitative insights into the antireflection performance. It is important to recognize that directly validating the geometry-induced antireflection mechanisms on biological samples could be complicated by the light-absorbing properties of the materials comprising the leafhopper wing and brochosomes, particularly when the wavelength of light is less than 300 nm ( 44 ). Our measurements show that the specular reflection on a leafhopper wing coated with brochosomes can be reduced by ~28 to ~86% in the UV range (i.e., ~300 nm < λ < ~400 nm), and by ~28 to ~68% in the visible light range (~400 nm < λ < ~700 nm), compared to a bare leafhopper wing without brochosomes ( SI Appendix , Fig. S12 ). This degree of reflection reduction is comparable to that observed on moth wings with nano-pillars, which show an average reduction of approximately 69% compared to a bare moth wing without nano-pillars ( SI Appendix , Fig. S12 ). In the case of the leafhopper wing, the observed reduction in broadband UV–visible reflection is consistent with our results derived from synthetic brochosomes, where the combined through-hole effect and Mie scattering significantly reduce UV reflection, and Mie scattering alone contributes to broadband anti-reflection in the visible light range. In addition, natural brochosomes are typically found densely packed in a disordered arrangement ( Fig. 1 C ). To understand how the packing density and disorderliness of brochosomes affect light reflectivity, we conducted further study using synthetic brochosomes. We found that the through-hole effect becomes more pronounced when the packing density exceeds the critical packing density η c , which is approximately 58% ( SI Appendix , Fig. S13 ). This effect is particularly noticeable when the gap distance between individual particles is smaller than the through-hole diameter. Additionally, we have also performed a disorder analysis to examine the robustness of the through-hole effect. We fabricated synthetic brochosomes, both with and without through-holes, and arranged them in a disordered array ( SI Appendix , Fig. S14 ). Our experimentally measured specular reflectance reveals that the through-hole effect remains evident in a disordered arrangement ( SI Appendix , Fig. S14 ). Biological Implications. How do antireflective brochosomes benefit leafhoppers? Antireflective surfaces are commonly found across the insect kingdom, with some of the well-known examples being the moth’s eye ( 2 ) and insect wings ( 45 ). It has long been hypothesized that these antireflective surfaces help insects by reducing mirror-like reflections, thus providing camouflage or decreasing their detectability to predators ( 46 ). Similarly, we hypothesize that leafhoppers consistently apply a dense coating of brochosomes on their wings, potentially maximizing the surface anti-reflectivity to avoid attracting predators with mirror-like reflected light. To maintain a high packing density, leafhoppers frequently secrete and distribute brochosomes across their body surfaces. They engage in anointing and grooming behaviors every few hours, ensuring a multilayered, dense, and evenly distributed brochosome coating on their bodies ( 47 ). These behaviors, observed in both nymphs and adult leafhoppers, persist throughout their lifespan ( 47 ). The predators of leafhoppers, such as birds and lizards, possess tetrachromatic color vision, which enables them to perceive an extended range of colors, including UV colors ( 42 , 48 , 49 ). Plant leaves, which are a common habitat of leafhoppers, contain pigments that can absorb UV light, resulting in reduced UV reflection ( 50 ). Our results suggest that leafhoppers potentially utilize the through-holes on brochosomes to further decrease UV reflectance, thereby mimicking the low UV reflection of plant leaves. The combined effects of UV absorption by the through-holes and the scattering of visible light by the primary structures of brochosomes could synergistically aid leafhoppers in reducing their observability to predators. While this hypothesis remains to be tested in field studies, our results provide a physical basis for understanding why the size of brochosomes from various leafhopper species may have evolved within a size range on the order of a few hundreds of nanometers. These findings potentially suggest an evolutionary design strategy by leafhoppers, employing highly engineered, deployable optical materials as a means to evade their predators."
} | 2,158 |
21854462 | null | s2 | 5,935 | {
"abstract": "The marine bacterium Vibrio fischeri uses a biofilm to promote colonization of its eukaryotic host Euprymna scolopes. This biofilm depends on the symbiosis polysaccharide (syp) locus, which is transcriptionally regulated by the RscS-SypG two-component regulatory system. An additional response regulator (RR), SypE, exerts both positive and negative control over biofilm formation. SypE is a novel RR protein, with its three putative domains arranged in a unique configuration: a central phosphorylation receiver (REC) domain flanked by two effector domains with putative enzymatic activities (serine kinase and serine phosphatase). To determine how SypE regulates biofilm formation and host colonization, we generated a library of SypE domain mutants. Our results indicate that the N-terminus inhibits biofilm formation, while the C-terminus plays a positive role. The phosphorylation state of SypE appears to regulate these opposing activities, as disruption of the putative site of phosphorylation results in a protein that constitutively inhibits biofilm formation. Furthermore, SypE restricts host colonization: (i) sypE mutants with constitutive inhibitory activity fail to efficiently initiate host colonization and (ii) loss of sypE partially alleviates the colonization defect of an rscS mutant. We conclude that SypE must be inactivated to promote symbiotic colonization by V. fischeri."
} | 349 |
33960013 | PMC11469123 | pmc | 5,936 | {
"abstract": "Abstract Despite important breakthroughs in bottom‐up synthetic biology, a major challenge still remains the construction of free‐standing, macroscopic, and robust materials from protocell building blocks that are stable in water and capable of emergent behaviors. Herein, a new floating mold technique for the fabrication of millimeter‐ to centimeter‐sized protocellular materials (PCMs) of any shape that overcomes most of the current challenges in prototissue engineering is reported. Significantly, this technique also allows for the generation of 2D periodic arrays of PCMs that display an emergent non‐equilibrium spatiotemporal sensing behavior. These arrays are capable of collectively translating the information provided by the external environment and are encoded in the form of propagating reaction–diffusion fronts into a readable dynamic signal output. Overall, the methodology opens up a route to the fabrication of macroscopic and robust tissue‐like materials with emergent behaviors, providing a new paradigm of bottom‐up synthetic biology and biomimetic materials science.",
"conclusion": "3 Conclusions Working toward fully autonomous synthetic tissues, we used bio‐orthogonal chemistry for the programmed assembly of synthetic protocells into centimeter‐sized tissue‐like materials that are stable in water media, can communicate internally and with the external environment, and are capable of emergent non‐equilibrium biochemical sensing. This was achieved by packing millions of bio‐orthogonally reactive proteinosomes in oil at the water‐air interface inside a PTFE mold floating on a 5 wt% solution of polysorbate 80. Robust PCMs were generated from the synergistic effect of surfactant‐mediated oil removal, Marangoni flow, and interfacial bio‐orthogonal ligation of the protocell building blocks. We then showed that PCMs with complex 3D architectures could be easily constructed using this novel floating mold technique and we successfully generated patterns of different protocell phenotypes and stratified PCMs. It should now be possible to advance this methodology to generate even more complex 3D architectures where protocell populations with different specialized functions are patterned into individual layers of different thicknesses that can then be assembled into stratified PCMs. This would open up a way to the generation of PCMs with internalized complex communication pathways or to the microscale engineering of soft machines and devices that comprise localized components to carry out specific biosynthetic tasks. The communication properties of the PCMs were then investigated by assembling PCMs capable of an internalized GOx/HRP enzyme cascade. These PCMs were capable of sensing the external environment and triggering a coordinated internalized biochemical response using an endogenously produced signaling molecule (H 2 O 2 ). The unique communication properties of these materials were then employed to construct for the first time arrays of synthetic tissues that were capable of dynamically extracting encoded information provided by the external environment in the form of unidirectional diffusing fronts of chemical species. Our results open up a route from the synthetic construction of different protocell building blocks with adhesion capabilities to their programmed assembly and spatial integration into cm‐sized tissue‐like materials with precise architectures and geometries. These PCMs are stable in water and are capable of combining the specialization of individual protocell types with the emergent spatiotemporal biochemical response of the ensemble. From a more general perspective, the programmed assembly of non‐equilibrium materials capable of emerging bioinspired functions from protocell building blocks addresses important challenges of bottom‐up synthetic biology and biomimetic materials science and is expected to open new avenues toward novel organized platforms for tissue engineering, personalized therapy, pharmacokinetics, micro‐bioreactor technologies, and soft robotics. For example, we envision the possibility of engineering PCMs capable of integrating with living cells and tissues, interacting both through chemical and mechanical stimuli to influence cell growth, proliferation and differentiation or to provide targeted therapies. Our floating mold technique could be used to program the assembly of specific protocell phenotypes into arrays of biomimetic organoids that could be used to study the spatiotemporal diffusion and distribution of drugs. PCM arrays could also be chemically programmed to perform continuous biocatalytic synthetic tasks and deliver (bio)molecules of interest. We also envision the possibility of utilizing the techniques outlined in this work to engineer soft robots (e.g., swimmers and walkers) and soft robotic components (e.g., sensors and valves) from specialized protocells capable of chemo‐mechanical transduction. To conclude, the possibilities that our paradigm shifting approach to prototissue engineering opens up are many. Most importantly, the PCMs described here provide a highly modular platform to both start tackling important fundamental scientific challenges (e.g., understanding of the physicochemical basis of collective and emergent behaviors of living tissues) and to facilitate the development of new protocell applications through their spatial integration into tissue‐like materials endowed with higher‐order coordinated functions.",
"introduction": "1 Introduction Living tissues comprise complex 3D architectures of interconnected cell consortia that communicate and display emergent behaviors. Mimicking the structure of living tissues and understanding the physical–chemical basis of their emergent properties are two of the major goals of bottom‐up synthetic biology. Their achievement will lead to important technological advancements in tissue engineering, pharmacokinetics, personalized therapy, micro‐bioreactor technologies, and soft robotics. [ \n \n 1 \n \n ] \n In recent years, while working toward these goals, researchers in the field of bottom‐up synthetic biology started to develop methodologies to assemble different models of synthetic protocells [ \n \n 2 \n \n ] into interconnected 3D networks, termed prototissues, that communicate and display rudimental emergent behaviors. [ \n \n 1f \n \n ] For example, Bayley and co‐workers developed a 3D printing technique to pattern water‐in‐oil microdroplets connected by interface bilayers (DIBs) into synthetic tissues. They then demonstrated that DIBs are capable of membrane protein‐mediated electrical communication, macroscopic deformation, and light‐induced gene expression. [ \n \n 3 \n \n ] Li et al. used magnetic fields to manipulate diamagnetic giant unilamellar lipid vesicles (GUVs) into various spatially coded configurations of a few hundred micrometers in size. [ \n \n 4 \n \n ] Wang et al. showed that microarrays of hemifused GUVs could be patterned using acoustic standing waves, thus making progress toward the fabrication of prototissues with controlled geometries and lattice dimensions. [ \n \n 5 \n \n ] Instead of patterning using 3D printing or by applying magnetic fields or acoustic standing waves, our group has recently developed a synthetic approach to the programmed assembly of prototissue spheroids based on the interfacial bio‐orthogonal adhesion of two populations of reactive protein–polymer protocells, termed proteinosomes. [ \n \n 6 \n \n ] Proteinosomes are a well‐established protocell model and are generated using the Pickering emulsion technique. They comprise a semipermeable and elastic membrane which consists of a closely packed monolayer of conjugated bovine serum albumin/poly( N ‐isopropylacrylamide) (BSA/PNIPAM) amphiphilic nanoparticles, and because of this they can be classified as organic colloidosomes. The BSA/PNIPAM membrane is then chemically cross‐linked with poly(ethylene glycol)‐bis( N ‐succinimidyl succinate) (PEG‐diNHS) and the proteinosomes can be transferred into water media. Most importantly, proteinosomes can be engineered to display protocellular properties such as guest molecule encapsulation, selective permeability, gene‐directed protein synthesis, membrane‐gated internalized enzyme catalysis, predatory behaviors, and reversible contractility. [ \n \n 7 \n \n ] To assemble proteinosomes into prototissue spheroids, we first synthesized a new BSA/poly( N ‐isopropylacrylamide)‐ co ‐methacrylic acid (BSA/PNIPAM‐ co ‐MAA) nanoconjugate and functionalized it with either pendent azide or bicyclononyne (BCN) moieties. The amphiphilic bio‐orthogonally reactive protein–polymer nanoconjugates were then used to prepare two separate populations of azide‐ or BCN‐functionalized proteinosomes as water‐in‐oil (w/o) droplets using the Pickering emulsion technique. The proteinosome structures were stabilized via chemical crosslinking with PEG‐diNHS, which was pre‐dissolved in the water phase. Binary populations of the azide‐ and BCN‐functionalized proteinosomes were then spatially confined using a water‐in‐oil‐in‐water (w/o/w) Pickering emulsion procedure and structurally interlinked in situ via an interfacial strain‐promoted alkyne–azide cycloaddition (I‐SPAAC) reaction to afford prototissue spheroids 75–200 µm in diameter upon removal of the inner oil phase. [ \n \n 6 \n \n ] \n While all these different approaches provided important breakthroughs in prototissue design and synthetic construction, they are not without their drawbacks. The w/o/w Pickering emulsion method does not provide spatial control over the protocell organization and is currently limited to the generation of prototissue spheroids with micrometer‐scale dimensions; [ \n \n 6 \n \n ] acoustic patterning requires the standing waves to be constantly applied to avoid a rapid re‐dispersal of the GUVs into the bulk solution; [ \n \n 5 \n \n ] the diamagnetic GUVs require an aqueous media containing high levels of MnCl 2 and a constant magnetic field to maintain the patterns; [ \n \n 4 \n \n ] and the 3D‐printing of DIBs requires the presence of an external bulk oil phase and the resulting prototissues present a very short shelf life. [ \n \n 8 \n \n ] As a consequence, the possibility of using protocells as building blocks to assemble macroscopic materials that are robust, free‐standing, characterized by complex internal 3D architectures, capable of communicating both internally and with the external environment, and displaying emergent behaviors that generate from the synergistic interaction of their constituent parts still remains a considerable challenge. The development of such protocellular materials (PCMs) would open up new avenues in bottom‐up synthetic biology and bioinspired engineering and facilitate the transition of protocell research from fundamental to applied science. As a step toward this ambitious goal, herein, we describe the first bottom‐up methodology for the fabrication of PCMs that overcomes most of the current challenges in prototissue engineering. This methodology is based on a floating poly(tetrafluoroethylene) (PTFE) mold, which can be used for the programmed assembly of millions of bio‐orthogonally reactive synthetic protocells into centimeter‐sized free‐standing tissue‐like materials of any size and shape. These PCMs are stable in water media and are capable of communicating both internally and with the external environment. Significantly, this novel floating mold technique could also be used to generate for the first time 2D periodic arrays of PCMs, which were capable of an emergent non‐equilibrium spatiotemporal sensing behavior. These arrays were capable of dynamically translating the information provided by the external environment and encoded in the form of propagating reaction–diffusion gradients into a readable signal output. In general, our work moves beyond the engineering of a strategy to generate protocell–protocell adhesions. It aims to spearhead the programmed assembly and spatial integration of different protocell phenotypes into centimeter sized free‐standing PCMs with precise architectures and geometries. Thanks to these unique characteristics, the PCMs can then combine the specialization of individual protocell types with the emergent spatiotemporal biochemical response of the ensemble, thus providing a new paradigm of bottom‐up synthetic biology and biomimetic materials science.",
"discussion": "2 Results and Discussion 2.1 Programmed Assembly of Protocellular Materials (PCMs) Protein–polymer PCMs were generated from a binary population of bio‐orthogonally reactive proteinosomes in oil. First, rhodamine B isothiocyanate (RITC)‐labeled azide‐functionalized BSA/PNIPAM‐ co ‐MAA nanoconjugates (red fluorescence) and fluorescein isothiocyanate (FITC)‐labeled BCN‐functionalized BSA/PNIPAM‐ co ‐MAA nanoconjugates (green fluorescence) were synthesized using our previously established procedure. [ \n \n 6 \n \n ] Subsequently, samples of RITC‐labeled azide‐ and FITC‐labeled BCN‐functionalized proteinosomes in oil (mean diameter ≈ 25 µm; mean volume ≈ 8 pL) were prepared using the Pickering emulsion technique and internally cross‐linked with PEG‐diNHS ( Scheme \n 1 a ; Figure S1 , Section S1.2, Supporting Information). [ \n \n 6 \n , \n 7 \n \n ] Subsequently, the two populations of cross‐linked RITC‐labeled azide‐ and FITC‐labeled BCN‐functionalized proteinosomes in oil were mixed in a 1:1 ratio and drop‐cast inside a circular PTFE mold 2 mm in diameter floating on an aqueous solution of polysorbate 80 (5 wt%) to obtain a final emulsion volume of 0.64 µL mm −2 . The binary emulsion was then allowed to transfer to the water media for ≈2 h with an associated progressive color change from white to transparent (Scheme 1b ; Section S1.3, Supporting Information). Scheme 1 Generation of protocellular materials (PCMs). a) Scheme showing the preparation of a binary population of azide‐ (red shapes) and BCN‐functionalized (green shapes) proteinosomes in oil starting from the corresponding bio‐orthogonally reactive BSA/PNIPAM‐ co ‐MAA nanoconjugates. Step (1) involves the generation of 2 separate populations of bio‐orthogonally reactive proteinosomes in separate vials as w/o microdroplets using the Pickering emulsion technique. Step (2) involves mixing of the two populations in 1:1 ratio. b) Scheme illustrating the PCM programmed assembly process. Initially, a 1:1 binary population of azide‐ (red shapes) and BCN‐proteinosomes (green shapes) in oil prepared as described in (a) is cast inside a PTFE mold floating on a solution of polysorbate 80 in water (5 wt%). In this system the Pickering emulsion is subject to: 1) buoyancy, which keeps the emulsion inside the PTFE mold; 2) gravity, which acts to sediment the proteinosomes to the bottom of the oil droplet; and 3) Marangoni flow from the center of the PTFE mold to the sides and into the bulk solution as highlighted by the curved black arrows. With time the effect of polysorbate 80 and Marangoni flow extracts the oil from the emulsion and brings the proteinosomes in contact allowing them to react via an interfacial strain‐promoted alkyne–azide cycloaddition (I‐SPAAC) reaction and assemble the PCM. The photographs on top show the oil removal and PCM programmed assembly process on a 2 mm wide circular PTFE mold on a black background highlighting the associated opacity change from white to transparent; the appearance of black color is due to the background color. c) Scheme highlighting the I‐SPAAC reaction occurring upon oil removal. Time‐dependent fluorescence microscopy imaging showed that PCMs formed via a progressive oil removal process with concomitant bio‐orthogonal ligation of the binary proteinosome population (Scheme 1c ; Video S1 , Supporting Information). This resulted in membrane‐bounded PCMs with a spatially integrated tissue‐like structure that remained attached to the PTFE mold, as shown by fluorescence microscopy imaging ( Figure \n 1 a ). Confocal fluorescence microscopy showed that the PCMs prepared with an emulsion volume of 0.64 µL mm −2 had a homogeneous thickness of ≈180 µm (Figure 1b–d ). Significantly, the connection between the PCM and the PTFE mold was strong enough to allow for the mold to be lifted from the aqueous solution and for the sample to be handled in air (Figure 1e ). The PCM could also be easily detached from the mold, resulting in free‐standing PCMs in water media that could be manipulated with tweezers. This allowed for the preparation of samples for scanning electron microscopy (SEM) imaging, which showed tightly interconnected protein–polymer cell‐like structures resembling plant tissue or squamous epithelium tissue, highlighting the free‐standing nature of the material Figure 1f ; Figure S2 , Supporting Information). Figure 1 PCM characterization. a) Tiled epifluorescence microscopy image of a circular PCM 2 mm in diameter attached to a PTFE mold and immersed in an aqueous solution of polysorbate 80 (5 wt%). The PCM comprises an interlinked 1:1 binary population of FITC‐labeled (green fluorescence) BCN‐functionalized and RITC‐labeled (red fluorescence) azide‐functionalized proteinosomes internally cross‐linked with PEG‐diNHS. The PCM was prepared by adding an emulsion volume of 0.64 µL mm −2 . b) XY confocal fluorescence microscopy image showing a zoomed in area of the PCM in (a). c) XZ confocal fluorescence microscopy scan showing a zoomed in vertical section of the PCM in (a). d) 3D confocal image of the PCM in (a). The image shows that the PCM has a homogeneous thickness of ≈180 µm. Images in (b), (c), and (d) highlight the formation of a spatially interlinked network of closely packed bio‐orthogonally ligated proteinosomes. e) Photograph demonstrating the robustness and ease of lifting the PTFE mold with attached PCM from the aqueous solution of polysorbate 80 (5 wt%). In this image a water meniscus crossing the circular mold can be noted, highlighting the presence of the PCM inside. f) Scanning electron microscopy (SEM) image of a freeze‐dried free‐standing PCM showing the details of the spatially interlinked network of closely packed bio‐orthogonally ligated proteinosomes. g) Graph showing changes in the PCM thickness as a function of the emulsion volume used to assemble them. Data obtained from the analysis of Figure S3 , Supporting Information. Error bars: standard deviation. h) Graph showing onset of transfer time (blue plot) and final transfer time (orange plot) as a function of the emulsion volume used to assemble the PCMs, ranging between 0.16 and 0.64 µL mm −2 . Data obtained from the analysis of Figure S4 , Supporting Information. Error bars: standard deviation. The thickness of the PCMs could be controlled by varying the volume of the 1:1 binary emulsion. In a series of systematic experiments, we generated circular PCMs 2 mm in diameter by progressively increasing the emulsion volume from 0.08 (minimum volume we could inject) to 1.10 µL mm −2 (maximum volume that could reproducibly be contained in the mold) and used confocal fluorescence microscopy to characterize the thickness of the different PCMs. The PCM thickness could be varied from a minimum of 40 ± 9 µm (i.e., a mono/bilayer of proteinosomes) to a maximum of 190 ± 20 µm (Figure 1g ; Figure S3 , Supporting Information). The thickness was found to increase linearly with the emulsion volume between 0.08 and 0.64 µL mm −2 and reached a plateau at higher volumes due to the overflow of the emulsion from the bottom of the mold. Moreover, PCMs rich in voids formed using emulsion volumes < 0.32 µL mm −2 . 0.32 µL mm −2 was the minimum emulsion volume needed to obtain a continuous, non‐defective PCM 80 ± 10 µm thick (Figure S3d , Supporting Information). Due to the color change from white to transparent, the oil removal and PCM programmed assembly process could be monitored using a digital camera to obtain PCM transfer curves as a function of the emulsion volume used to assemble the PCMs, ranging between 0.16 and 0.64 µL mm −2 (Section S1.4, Figure S4 , Video S2 , Supporting Information). The PCM transfer curves displayed a sigmoidal shape, and the onset time of the curve was found to increase linearly with the emulsion volume used ( 1h , blue plot). In contrast, the final PCM transfer time, defined as the intersection of the slope of the sigmoidal curve and the plateau region, displayed a logarithmic growth (Figure 1h , orange plot). These observations seem to indicate that the onset of the oil removal process depends linearly on the volume of 2‐ethyl‐1‐hexanol present in the sample, but the rate of diffusion of the oil into the bulk solution tends to reach a threshold value at high emulsion densities. Moreover, the importance of polysorbate 80 in the oil removal and PCM programmed assembly process was also highlighted by control experiments carried out in the absence of the surfactant. Under this condition a strong osmotic pressure across the PCM caused the growth of large water bubbles on top of the PCM with concomitant PCM deformation and rupture when the bubbles reached a critical size (Video S3 , Supporting Information). This still allowed for a slow transfer of the prototissue into water (≈8 h, Figure S5 , Supporting Information), but the resultant PCM proved more fragile and inhomogeneous. In comparison, in the presence of polysorbate 80, the transfer of the PCM into water occurred in ≈1 h (Figure S5 , Supporting Information), highlighting the key role of the surfactant in the PCM formation. Significantly, control experiments carried out using non‐bio‐orthogonally reactive proteinosomes highlighted the critical role of bio‐orthogonal chemistry in the PCM generation. Video S4 , Supporting Information, compares the oil removal and PCM programmed assembly process of two experiments performed in parallel using normal (non‐bio‐orthogonally reactive) proteinosomes (left) and bio‐orthogonally reactive proteinosome (right). In the absence of bio‐orthogonal ligation the Marangoni flow pushed proteinosomes to the edge of the mold and dragged them into the bulk solution (Scheme 1 ), resulting at best in the formation of a thin and defective PCM (Video S4 , Supporting Information, left). In contrast, in the presence of bio‐orthogonal ligation, as soon as the oil was removed and azide‐ and BCN‐functionalized proteinosome entered in contact, they promptly reacted via the I‐SPAAC reaction and formed a spatially integrated tissue‐like structure (Video S4 , Supporting Information, right). Moreover, attempts to generate PCMs in the absence of the PTFE mold were unsuccessful as once the emulsion touched the aqueous solution it readily spread into individual proteinosomes under the action of the Marangoni flow. The individual proteinosomes then transferred to the aqueous phase without assembling into PCMs. Taken together, these observations seem to indicate that the PCM assembly process involves a synergistic effect of the mold (holding in place the bio‐orthogonally reactive proteinosomes in oil), the surfactant‐mediated oil removal, the Marangoni flow, and the bio‐orthogonal ligation. Most importantly, our new floating mold technique allows us to assemble proteinosome building blocks together into macroscopic and free‐standing PCMs with controllable thickness that are mechanically robust and stable in water media for months. 2.2 Generation of PCMs with Complex 3D Architectures Having established that the floating mold method can be successfully used to generate macroscopic and free‐standing tissue‐like materials from a binary community of bio‐orthogonally reactive proteinosomes, we next explored its versatility for the generation of PCMs with complex 3D architectures. First, we explored the possibility of generating PCMs of different shapes and sizes. As a step toward this goal, we built a PTFE mold in the shape of an equilateral triangle with 1.0 cm sides and a PTFE mold in the shape of a square with 0.5 cm sides and used them to generate PCMs at a 0.64 µL mm −2 emulsion volume. Epifluorescence microscopy images showed the successful programmed assembly of defect‐free PCMs in the shape of a triangle ( Figure \n \n 2 a ) and of a square (Figure 2b ) of the desired dimensions. Neither of the PCMs presented cracks or defects, both fully transferred to the water phase and remained attached to the PTFE mold. Despite the large area (43.3 and 25.0 mm 2 , respectively), both PCMs could be easily lifted from the surfactant solution using the mold and transferred to a different solution without breaking. Significantly, since the PCM shape is defined by the PTFE mold and this can be laser‐cut using a high‐precision CNC machine, there is virtually no limitation to the complexity of PCM shapes that can be fabricated with this method. To demonstrate this, we constructed different PTFE molds with our research group's logo (size: 6.5 × 2.2 cm) using 3 different text fonts and used them to assemble PCMs with complex shapes. Epifluorescence microscopy imaging showed the successful programmed assembly of millions of RITC‐ and FITC‐labeled bio‐orthogonally reactive proteinosomes into the Gobbo Group's logo (Figure 2c ; Figures S6 – S8 , Supporting Information). Figure 2 Generation of PCMs with complex 3D architectures. a) Tiled epifluorescence microscopy image of a PCM in the shape of an equilateral triangle with 1.0 cm sides. The PCMs comprised an interlinked 1:1 binary population of RITC‐labeled (red fluorescence) azide‐ and FITC‐labeled (green fluorescence) BCN‐functionalized proteinosomes internally cross‐linked with PEG‐diNHS. b) Tiled epifluorescence microscopy image of a PCM in the shape of a square with 5 mm sides with the same composition as the PCM in (a). c) Tiled epifluorescence microscopy images showing PCMs with the same composition as the PCM in (a) and composing the “Gobbo Group” logo. d) Tiled epifluorescence microscopy image of a patterned squared PCM with sides of 5 mm comprising interlinked 1:1 binary populations of non‐tagged azide‐ and differently tagged BCN‐functionalized proteinosomes internally cross‐linked with PEG‐diNHS. Blue fluorescence: Dylight405; green fluorescence: FITC; and red fluorescence: RITC. The patterns were manually generated using a mechanical pipette. e) 3D confocal fluorescence image of a 3‐tiered stratified PCM ≈270 µm thick. All layers are composed of an interlinked 1:1 binary population of BCN‐ and azide‐functionalized proteinosomes internally cross‐linked with PEG‐diNHS. Blue fluorescence: Dylight405; green fluorescence: FITC; and red fluorescence: RITC. The 3 PCM layers were perfectly attached to each other and no delamination was observed; see also Figure S11 , Supporting Information. Next, we explored the possibility of generating patterns of different proteinosome populations within the same PCM. To achieve this, we synthesized different populations of BCN‐ and azide‐functionalized proteinosomes in oil: 1) non‐labeled azide‐functionalized proteinosomes, 2) Dylight405‐labeled BCN‐functionalized proteinosomes, 3) FITC‐labeled BCN‐functionalized proteinosomes, and 4) RITC‐labeled BCN‐functionalized proteinosomes. These populations were then used to generate 3 binary populations of proteinosomes by mixing the non‐labeled azide‐functionalized proteinosomes in a 1:1 ratio with the Dylight405‐, FITC‐, or RITC‐labeled BCN‐functionalized proteinosomes. Patterned PCMs with circular and concentric green and red fluorescent proteinosome populations on a background of blue fluorescent proteinosomes were then generated by manual patterning of the 3 differently labeled binary populations of proteinosomes in oil inside a 0.5 cm wide square PTFE mold using a mechanical pipette. The patterned emulsions were then allowed to transfer to the aqueous phase and assemble into the patterned PCM (Section S1.5, Supporting Information). Epifluorescence microscopy imaging showed successful formation of the desired PCM with circular concentric patterns of proteinosome consortia Figure 2d ; Figure S9 , Supporting Information). Most importantly, no noticeable differences were observed in the pattern when the PCM was flipped upside‐down and imaged (Figure S9 , Supporting Information). This indicated that the technique produced patterns that were homogenous through the PCM thickness, that is, the pattern remained in the xy plane and no stacking of the proteinosome populations was observed. Patterned PCMs with 2 × 2 and 3 × 3 arrays of red fluorescent proteinosome populations on a background of green fluorescent proteinosomes were also generated. The arrays were achieved by manual patterning of a binary population of RITC‐labeled azide‐ and BCN‐functionalized proteinosomes in oil on a background emulsion comprising a binary population of FITC‐labeled azide‐ and BCN‐functionalized proteinosomes in oil. The patterned emulsions were then allowed to transfer to the aqueous phase and assemble into the patterned PCM (Section S1.5, Supporting Information). Epifluorescence microscopy imaging showed successful formation of the desired PCMs with 2 × 2 and 3 × 3 arrays of proteinosome consortia (Figure S10 , Supporting Information). Stratified PCMs could then be generated using a layer‐by‐layer technique. First, we cast a binary population of Dylight405‐labeled bio‐orthogonally reactive proteinosomes in oil and allowed them to transfer into water and assemble into a first PCM layer ≈90 µm thick. The further 2 proteinosome layers could then be cast on top this first layer simply by repeating the same protocol using different fluorescently labeled binary populations of bio‐orthogonally reactive proteinosomes in oil. Upon transfer, each proteinosome layer adhered to the layer underneath via an inter‐layer I‐SPAAC reaction, resulting, overall, in a stratified prototissue ≈ 270 µm thick. Confocal fluorescence microscopy imaging showed that all layers were homogenous in thickness and no layer delamination was observed (Figure 2e ; Figure S11 , Supporting Information). Furthermore, we did not observe any difference in the oil removal process when preparing single‐layered or stratified PCMs. We ascribed this to the high permeability of the proteinosome membrane to small molecules, which made the oil extraction process very effective even in the presence of transferred PCM layers. Overall, these results demonstrate the high versatility of our floating mold technique. Our novel approach is extremely promising and pioneers a route to the design and synthetic construction of PCMs of large size and of any shape that are stable in water media, and comprise patterns and layers of different protocell consortia. 2.3 Non‐Equilibrium Biochemical Sensing in PCMs Inspired by the above observations, we extended our methodology to construct the first PCMs capable of supporting a collective and coordinated spatiotemporal biochemical response via internally derived molecule‐based signaling. As a first step toward this goal, we investigated whether the PCMs were capable of sensing the external environment and triggering a coordinated internalized cascade of chemical signals via enzyme catalysis. To achieve this, we prepared a circular PCM 2 mm in diameter from a binary population of FITC‐labeled BCN‐functionalized and non‐labeled azide‐functionalized proteinosomes that were preloaded with glucose oxidase (GOx) and horseradish peroxidase (HRP), respectively (Sections S1.7 and S1.8, Supporting Information). Once transferred to the water media, the PCM was moved to a Petri dish containing an aqueous solution of glucose (Glc, 20 × 10 −3 \n m ) and Amplex Red (0.5 × 10 −3 \n m ). This initiated a spatially coupled GOx/HRP enzyme cascade reaction under diffusional equilibrium conditions between the 2 bio‐orthogonally interlinked protocell populations. The GOx containing protocells converted Glc to d ‐glucono‐1,5‐lactone (GDL) and internally produced the signaling molecule H 2 O 2 , which was used to communicate to the HRP‐containing protocells to oxidize Amplex Red to resorufin and produce a fluorescent signal ( Figure \n \n 3 a ). Epifluorescence microscopy was used to determine the location of the GOx‐containing FITC‐labeled BCN‐functionalized proteinosomes within the PCM and monitor the onset and in situ development of red fluorescence due to the endogenous production of resorufin (Figure 3b ; Video S5 , Supporting Information). Typically, the onset of red fluorescence in the HRP‐containing protocells occurred within the first minute, followed by diffusion into the neighboring GOx‐containing protocells and external environment. By contrast, control experiments involving PCMs lacking either GOx or HRP showed no fluorescence increase due to the inability of these PCMs to sense glucose in the surrounding environment or synthesize resorufin, respectively (Figure 3c ). Figure 3 Communication properties of PCMs. a) Scheme representing the GOx/HRP enzyme cascade reaction in a PCM (enclosed by the 2 blue dashed lines) consisting of GOx‐containing BCN‐functionalized proteinosomes (green shapes) and HRP‐containing azide‐functionalized proteinosomes (grey shapes). The substrates glucose (Glc) and Amplex red or o ‐phenylenediamine ( o ‐PD) freely diffuses through the PCM. The GOx‐containing protocells oxidize Glc to d ‐glucono‐1,5‐lactone (GDL) and H 2 O 2 . This initiates radial diffusion of H 2 O 2 from the GOx‐containing protocells, which is then used by HRP‐containing protocells to oxidize the non‐fluorescent molecules, Amplex red or o ‐PD to red fluorescent resorufin or green fluorescent 2,3‐diaminophenazine (2,3‐DAP), respectively. H 2 O 2 can therefore be considered as a signaling molecule between the two interlinked protocell communities. b) Time‐dependent epifluorescence microscopy images of a circular PCM 2 mm in diameter prepared as described in (a) and in the presence of glucose and Amplex Red (20 and 0.5 × 10 −3 \n m in PBS 10 × 10 −3 \n m , pH 6.8, respectively) at 25 °C. Green fluorescence, GOx‐containing FITC‐labeled BCN‐functionalized proteinosomes; red fluorescence, resorufin production. c) Graph showing the time‐dependent generation of resorufin from a circular PCM 2 mm in diameter and structured as described in (a) (red curve), in the absence of GOx (control experiment, blue curve), and in the absence of HRP (control experiment, green curve). Experiment repeated in triplicate; error bars: standard deviation. Given the PCM's ability to sense and respond to chemical changes in the environment, we next displayed the potential of our floating mold technique by assembling, for the first time, 2D arrays of spatially encoded PCMs that could detect and visualize advancing concentration fronts of chemical gradients under non‐equilibrium conditions. To achieve this, we first constructed a new PTFE mold featuring a 4 × 4 array of circular wells for PCM production and an injection point placed on the left‐hand side of the array ( Figure \n 4 a ). This PTFE mold was utilized to assemble a 4 × 4 array of enzymatically active circular PCMs 2 mm in diameter (Section S1.9, Supporting Information). Each PCM comprised an interlinked 1:1 binary population of non‐labeled GOx‐containing BCN‐functionalized proteinosomes and HRP‐containing azide‐functionalized proteinosomes. Subsequently, the PTFE mold loaded with the array of PCMs was allowed to float on 0.980 mL of a phosphate buffer solution (PBS, 10 × 10 −3 \n m , pH 6.8). The PCM array was then exposed to a left‐to‐right unidirectional reaction–diffusion gradient by injecting 20 µL of an aqueous solution of glucose (Glc, 100 × 10 −3 \n m ) and o ‐phenylenediamine ( o ‐PD, 50 × 10 −3 \n m ) through the injection point on the left‐hand side of the PCM array. This induced spatiotemporal oxidase/peroxidase responses to the co‐diffusion of Glc and o ‐PD substrates across the periodically ordered enzymatically active PCM array. The array's spatiotemporal responses could be followed using time‐dependent fluorescence microscopy by monitoring the development of green fluorescence associated with the HRP‐mediated oxidation of o ‐PD to 2,3‐diaminophenazine (2,3‐DAP) in the azide‐functionalized proteinosomes. In general, a wave of 2,3‐DAP production unidirectionally moved across the PCM array from left to right. We associated this with the progressive co‐diffusion of Glc and o ‐PD substrates (Figure 4b , Video S6 , Supporting Information). Time‐dependent mean fluorescence intensity analysis showed that the onset time of fluorescence (OT) increased quadratically across the rows of the array oriented parallel to the substrate diffusion front (Figure 4d ), whereas it showed minimal difference across the columns oriented perpendicularly to the substrate diffusion front (Table S2 , Figure S12 , Supporting Information). A similar trend was found for the initial rates of 2,3‐DAP production, which gradually diminished across the rows placed parallel to the direction of the substrate diffusion front and showed comparable rates across the columns placed perpendicular to the substrate diffusion front. The fluorescence associated with each individual PCM was found to increase to a steady state and then slowly decrease in intensity. We attributed this to the consumption of the substrates and to the diffusion of the 2,3‐DAP into the bulk solution. Taken together, these observations indicate that the substrates were progressively depleted as the reaction–diffusion front advanced through the PCM array from left to right. We also noticed that PCMs x \n 3 \n y \n 1 , x \n 4 \n y \n 1 , and x \n 3 \n y \n 4 , x \n 4 \n y \n 4 always developed a higher mean fluorescence intensity compared to PCMs x \n 3 \n y \n 2 , x \n 4 \n y \n 2 , and x \n 3 \n y \n 3 , x \n 4 \n y \n 3 (Video S6 , Figure S12 blue and green plots, Supporting Information). We attributed this to the PCMs in rows y \n 1 and y \n 4 being exposed to additional Glc diffusing along the top and the bottom of the field of view. By contrast, a control experiment carried out under diffusional equilibrium conditions where the enzymatically active 4 × 4 PCM array was placed on a PBS solution (10 × 10 −3 \n m , pH 6.8) preloaded with both Glc and o ‐PD (final concentrations 1.0 and 0.5 × 10 −3 \n m , respectively) showed a nearly immediate homogeneous fluorescence turn‐on through the entire PCM array (Figure 4c ; Video S7 , Figure S13 , Supporting Information). The onset time of fluorescence of all 16 PCMs took place during the first 2–3 min of the experiment, and, as expected, it was independent of the spatial position of the PCMs (Figure 4d ; Table S3 , Supporting Information). The time‐dependent mean fluorescence intensity curves reached a maximum after ≈60 min, which was followed by a progressive decrease of the signal due to the depletion of the substrates and diffusion of the 2,3‐DAP product into the bulk solution (Figure S13 , Supporting Information). These observations are coherent with our initial hypothesis that 2D arrays of PCMs can be biochemically programmed to collectively detect and visualize advancing concentration gradients of substrates of interest under non‐equilibrium conditions. Figure 4 Non‐equilibrium biochemical sensing in 4 × 4 arrays of enzymatically active PCMs. a) Scheme showing the circular PTFE mold used for the non‐equilibrium biochemical sensing experiments. The scheme highlights the injection point, the x \n 1–4 \n y \n 1–4 wells used for the assembly of the 4 × 4 array of enzymatically active PCMs, and the direction of the unidirectional diffusion front of chemical substrates (orange arrow). b) Sequence of false color epifluorescence microscopy images showing spatiotemporal response of a 4 × 4 PCM array of enzymatically active PCMs, which was exposed to a co‐diffusing mixture of Glc and o ‐PD substrates. The images show a consecutive fluorescence turn‐on of columns x \n 1–4 associated with the in situ production of 2,3‐DAP. See Section S1.9, Supporting Information, for experimental details. c) Sequence of false color time‐dependent epifluorescence microscopy images showing the control experiment performed on a 4 × 4 PCM array of enzymatically active PCMs under diffusional equilibrium conditions. The images show a simultaneous fluorescence turn‐on of all PCMs in the array. See Section S1.9, Supporting Information, for experimental details. d) Plot showing the trend of average onset times (OTs) of 2,3‐DAP fluorescence for each x \n 1–4 column of the 4 × 4 PCM array as a function of the distance from the injection point obtained for the experiments in (b) (orange plot) and (c) (blue plot). The orange plot highlights a quadratic relationship between the average OTs and the distance from the injection point, which is typical for diffusing chemical species. In contrast, the blue plot shows that the average OTs is independent of the spatial position of the PCMs. In order to improve our understanding of the transient spatiotemporal response of our enzymatically active 4 × 4 array of PCM, we performed additional experiments by single‐component diffusion of either o ‐PD or Glc into a solution preloaded with the second substrate, that is, Glc or o ‐PD, respectively. As was previously the case, in both experiments time‐dependent mean fluorescence intensity analysis in general showed a propagating fluorescence wave that moved through the PCM array from left to right (Videos S8 and S9 , Supporting Information). This was associated with the progressive diffusion of the o ‐PD or Glc and with the HRP‐mediated production of 2,3‐DAP. However, by comparing these two different experiments, we also noticed some important differences in the spatiotemporal response of the 4 × 4 PCM array. When we preloaded Glc and diffused o ‐PD through the PCM array we observed a slow sequential fluorescence turn‐on of columns x \n 1 and x \n 2 , whereas columns x \n 3 and x \n 4 did not turn on in the timeframe of the experiment (Figure S14 , Supporting Information). Moreover, columns x \n 1 and x \n 2 continued to produce 2,3‐DAP during the timeframe of the experiment and they did not turn off as in the previous experiment where we diffused both Glc and o ‐PD. These observations are consistent with the rate limiting step of the overall process being the production of H 2 O 2 , rather than the diffusion of o ‐PD through the array. This was attributed to a low concentration of preloaded Glc in the system and to the slower catalytic reactivity of the GOx‐containing protocells compared to those containing HRP. The latter was ascribed to a lower molar loading of GOx in the protocells compared to HRP. However, when we diffused Glc in a bulk solution pre‐loaded in o ‐PD we observed a fast and sequential fluorescence turn‐on of the entire array from columns x \n 1 to x \n 4 , followed by a similarly fast and sequential decrease of the fluorescence signal in each PCM (Figure S15 ). We attributed this behavior to a fast local production of H 2 O 2 when the concentrated diffusion front of Glc reached the PCM columns, followed by a fast local depletion of the o ‐PD substrate due to the high activity of the HRP‐containing protocells that composed the PCMs. Since in this instance the rate limiting step of the overall process was the diffusion of Glc, this allowed for the estimation of a rate of diffusion for Glc of 9.1 (±0.2) × 10 −8 m 2 s −1 under these experimental conditions (Figure S16 , Supporting Information). In this experiment we also observed that PCMs x \n 3 \n y \n 1 , x \n 4 \n y \n 1 , and x \n 3 \n y \n 4 , x \n 4 \n y \n 4 had a higher mean fluorescence intensity compared to PCMs x \n 3 \n y \n 2 , x \n 4 \n y \n 2 and x \n 3 \n y \n 3 , x \n 4 \n y \n 3 (Video S9 , Supporting Information). This was consistent with what was observed in the previous experiment where we co‐diffused Glc and o ‐PD and was due to the PCMs in rows y \n 1 and y \n 4 being exposed to additional Glc diffusing along the top and the bottom of the field of view. Overall, these experiments demonstrate that PCM arrays provide a novel chemically programmable framework in which to systematically study information encoded in propagating reaction–diffusion gradients of chemicals, such as direction of the diffusing front, spatiotemporal changes in chemical concentrations, estimation of the diffusion rates of chemical species, and identification of rate‐limiting steps of the PCM bioreactivity. These results therefore provide a first important example of spatially organized prototissues that can sense the external environment, trigger an endogenous coordinated response, and operate under non‐equilibrium conditions, providing a new paradigm of prototissue engineering."
} | 11,181 |
29446165 | null | s2 | 5,937 | {
"abstract": "Actin networks are adaptive materials enabling dynamic and static functions of living cells. A central element for tuning their underlying structural and mechanical properties is the ability to reversibly connect, i.e., transiently crosslink, filaments within the networks. Natural crosslinkers, however, vary across many parameters. Therefore, systematically studying the impact of their fundamental properties like size and binding strength is unfeasible since their structural parameters cannot be independently tuned. Herein, this problem is circumvented by employing a modular strategy to construct purely synthetic actin crosslinkers from DNA and peptides. These crosslinkers mimic both intuitive and noncanonical mechanical properties of their natural counterparts. By isolating binding affinity as the primary control parameter, effects on structural and dynamic behaviors of actin networks are characterized. A concentration-dependent triphasic behavior arises from both strong and weak crosslinkers due to emergent structural polymorphism. Beyond a certain threshold, strong binding leads to a nonmonotonic elastic pulse, which is a consequence of self-destruction of the mechanical structure of the underlying network. The modular design also facilitates an orthogonal regulatory mechanism based on enzymatic cleaving. This approach can be used to guide the rational design of further biomimetic components for programmable modulation of the properties of biomaterials and cells."
} | 372 |
34253787 | PMC8275744 | pmc | 5,938 | {
"abstract": "Process engineering of biotechnological productions can benefit greatly from comprehensive analysis of microbial physiology and metabolism. Ralstonia eutropha (syn. Cupriavidus necator ) is one of the best studied organisms for the synthesis of biodegradable polyhydroxyalkanoate (PHA). A comprehensive metabolomic study during bioreactor cultivations with the wild-type (H16) and an engineered (Re2058/pCB113) R. eutropha strain for short - and or medium-chain-length PHA synthesis has been carried out. PHA production from plant oil was triggered through nitrogen limitation. Sample quenching allowed to conserve the metabolic states of the cells for subsequent untargeted metabolomic analysis, which consisted of GC–MS and LC–MS analysis. Multivariate data analysis resulted in identification of significant changes in concentrations of oxidative stress-related metabolites and a subsequent accumulation of antioxidative compounds. Moreover, metabolites involved in the de novo synthesis of GDP- l -fucose as well as the fucose salvage pathway were identified. The related formation of fucose-containing exopolysaccharides potentially supports the emulsion-based growth of R. eutropha on plant oils.",
"introduction": "Introduction Ralstonia eutropha (also known as Cupriavidus necator ) is one of the most studied organisms for polyhydroxyalkanoate (PHA) homeostasis. PHAs have been shown to be biodegradable in soil and aqueous environments, which makes them promising “green” alternatives to conventional plastic 1 – 3 . After the genome of R. eutropha strain H16 was completely sequenced 4 , multiple “big data” studies were carried out, ranging from transcriptomics 5 – 9 and proteomics 9 – 12 to metabolomics 13 – 15 . In general, there are two main types of metabolomic studies: targeted metabolomics, where only specific, known metabolites are quantified and untargeted metabolomics, where compounds that were hitherto unknown or unidentified in an organism can be identified. Untargeted metabolic profiling has the potential for identifying novel pathways or biomarkers 16 – 19 . For example, Fukui et al. performed metabolomic analysis on R. eutropha using labeled glucose and observed the presence of ribulose 1,5-bisphosphate, suggesting that the Calvin-Benson-Bassham cycle is active in R. eutropha even during heterotrophic growth 13 . Alagesan et al. confirmed this by examining metabolic flux of R. eutropha on fructose and glycerol as carbon sources and have shown that the Calvin-Benson-Bassham cycle is active under hetrotrophic conditions 15 . The wild-type strain R. eutropha H16 synthesizes polymers containing solely short chain length ( scl) monomers (i.e., monomers with less than five carbon atoms). To synthesize a more flexible thermoplastic polymer, many strains have been engineered to integrate medium chain length ( mcl , 5 < C < 15) monomers into the polymers. Particularly, strain construction efforts of several research groups have focused on the production of poly(hydroxybutyrate- co -hydroxyhexanoate) [P(HB- co -HHx)] 20 – 27 . This kind of copolymer is more flexible, tougher, less crystalline and has lower melting temperatures compared to scl -PHA, which facilitates a wide range of applications 28 . A strain that has been studied extensively for the production of P(HB- co -HHx) from oleaginous feedstocks is the engineered R. eutropha strain Re2058/pCB113 20 , 29 – 35 . R. eutropha possesses the metabolic capabilities to metabolize oleaginous feedstocks, such as refined plant oils like palm oil (PO), soybean oil or canola oil, as well as waste oils and animal fats. These feedstocks have proven to be excellent substrates for efficient PHA production 21 , 23 , 32 , 36 – 38 . This effective production is facilitated in R. eutropha cultures by the secretion of lipases, which mediate the hydrolyzation of the triacylglycerols into free fatty acids, monoacylglycerols, diacylglycerols and glycerol and consequently allow the formation of a natural emulsion 39 . In this study, the metabolic characteristics of R. eutropha H16, a scl -PHA producer, and Re2058/pCB113, a scl-co-mcl -PHA producer, were examined during bioreactor cultivations under growth and nitrogen limitation (PHA synthesis) conditions using plant oil as the main carbon source. Both strains were evaluated by non-targeted metabolite profiling. Since the two strains produce different PHA polymer products, it was assumed that the metabolite profiles of the strains would exhibit significant differences, which were observed in this study. In addition, the metabolite profiling yielded the identification of the de novo and salvage pathway synthesis of GDP- l -fucose, a component of exopolysaccharide/lipopolysaccharide (EPS/LPS), as well as oxidative stress-related metabolite changes.",
"discussion": "Discussion Fucose is a 6-deoxyhexose, which is found in LPS or EPS. Fucose-containing LPS or EPS are synthesized from guanosine diphosphate fucose (GDP- l -fucose), which can be synthesized via two pathways: The de novo pathway produces GDP- l -fucose from GDP-mannose or the salvage pathway, where GDP- l -fucose is synthesized from LPS or EPS 41 . The latter has been identified apart from the presence in eukaryotes in only one prokaryote so far, the gut commensal bacterium Bacterioides fragilis 9343 42 . A successful integration of the B. fragilis pathway into the Escherichia coli was recently described 44 . We identified for the first time the presence of GDP- l -fucose and metabolites in the GDP- l -fucose de novo synthesis and fucose salvage pathway in R. eutropha . Fucose-containing EPS have been reported to be surface active compounds, which have emulsion stabilizing characteristics 45 – 47 . Even though several studies have shown the efficient PHA production from lipids with R. eutropha 21 , 23 , 32 , 36 – 38 , the presence of natural EPS emulsifiers was never described before. A large fraction of fucose in the EPS of R. eutropha IPT 027 was identified, but the emulsification behavior was not tested 48 . In this context, fucose-containing EPS potentially support the lipase-mediated emulsification process 39 , which would explain the excellent growth of R. eutropha on oleaginous substrates. A similar mechanism was described in Pseudomonas aeruginosa , which produces extracellular rhamnolipids from PO distillate with emulsifying properties 49 . The identification of this new metabolic pathway may facilitate the production of novel engineered strains with an increased fucose-EPS production for an enhanced emulsion formation. This could lead to strains with a reduced lag phase on plant oil or hard to emulsify solid waste animal fats, as it was shown for cultures with a chemical pre-emulsified PO or strains overexpressing an extracellular lipase 39 , 50 . The fucose-related metabolites in our study were identified by the annotation and subsequent statistical analysis of 656 metabolites, which are at least sixfold more annotated metabolites than in other comprehensive metabolomic studies with R. eutropha 13 – 15 . Genes for synthesis of GDP- l -fucose via both pathways have not been discovered or characterized in R. eutropha so far. The decrease of GDP- d -mannose and GDP- l -fucose throughout our cultivations suggests, that the de novo synthesis pathway is highly active in the beginning to build up EPS (Fig. 4 ). In contrast, the significant increase of l -fucose throughout the cultivations indicates, that the salvage pathway is more active during the later phase of the cultivation. The fucose salvage pathway was considered to be only present in eukaryotes until a bifunctional l -fucokinase/fucose-1-phosphate guanylyltransferase enzyme in the mammalian gut commensal bacterium B. fragilis was identified 42 . A BLAST search for a similar enzyme in R. eutropha did not result in a positive match (data not shown). High levels of methionine sulfoxide were detected in the beginning of the cultivations and these levels significantly decreased towards later time points (Fig. 5 ). Methionine sulfoxide is a product of methionine oxidation due to the presence of reactive oxidative intermediates, which are typically present in an aerobic bioprocess 43 , 51 . Although R. eutropha cells, which are exposed to oxidative stress, typically produce protective proteins, DNA damage occurs 10 , 11 . The increase in ethenodeoxyadenosine in our cultures can also be interpreted as an indication of such DNA damage (Fig. 5 ). Additionally, oxidative stress in R. eutropha is counteracted by antioxidative molecules 52 . In this study, we detected a significant increase of octyl gallate and two tocopherols throughout the cultivations (Fig. 5 ). The latter are hydrophobic antioxidants, which are present in high concentrations in PO 53 . It is most likely that R. eutropha is able to absorb these compounds, from the PO containing media, to cope with oxidative stress. The analysis of metabolites in the PHA homeostasis pathway revealed for both strains that the main precursor metabolite acetyl-CoA is available at higher levels prior to nitrogen limitation and the concentrations drastically decrease after the onset of nitrogen limitation (Fig. 3 ). It is known that β-ketothiolases (PhaAs) are highly selective for acetyl-CoA and an enhanced expression of the phaA gene can yield in enhanced PHA accumulation 54 – 56 . However, phaA was found to be constitutively expressed in R. eutropha H16, which would not explain the lower levels of acetyl-CoA after the nitrogen depletion 8 , 57 . Multiple PHA mobilizing enzymes, namely PHA depolymerases PhaZ1–7 and oligomer hydrolases PhaY1–2, are responsible for the degradation of the intracellular PHA and finally supplying acetyl-CoA 58 – 60 . The high activity of the PHA mobilizing enzymes under growth conditions and downregulation under nutrient depleted conditions was previously shown, and can help to explain the acetyl-CoA pool size during growth in our experiments 61 . Another interesting finding about PHA metabolism is the absence of any HHx-CoA pools in the engineered strain Re2058/pCB113, but it is known that very high contents of HHx in the PHA are present during the growth phase 20 , 32 . Together with the fact that overall the HB-CoA levels are lower in the engineered strain compared to the wild-type strain, a very high activity of the PhaC Ra can be assumed. A further engineering of the pathways supplying the acyl-CoA precursors for the PhaC Ra could therefor potentially enhance the PHA productivity. The high polymerase activity could also explain why the engineered strain produces PHA with a significantly lower average molecular weight compared to the wild-type strain or a similar engineered strain harboring a low active PHA synthase, such as PhaC BP-M-CPF4 27 . The comparison of the two strains resulted in the detection of a shortage of many amino acids in the engineered strain (Table S6 ). Due to the constant expression of the complete PHA operon, which is located on the plasmid pCB113, the engineered strain produces quantitatively more enzymes compared to the wild-type strain, which results in a PHA accumulation in the growth phase and could be the reason for the overall lower amino acid levels in the engineered strain. In conclusion, our comprehensive analysis of data from untargeted metabolomics of R. eutropha cultivations facilitated the identification of the fucose salvage pathway, which was previously identified in just one other bacterium. Even though R. eutropha is a well characterized organism, the presented approach allowed to add more understanding to the complexity of the microbial metabolic network. In detail, the findings presented here elucidate a potential additional emulsification mechanism for the efficient growth on oleaginous feedstocks. It can also be emphasized that diversity of fucose metabolism pathways in bacteria is poorly understood, suggesting that unknown enzymes involved in the fucose salvage pathway can be identified in future studies."
} | 3,021 |
33184268 | PMC7665198 | pmc | 5,940 | {
"abstract": "Previous studies have shown that copolymer compositions can significantly impact self-healing properties. This was accomplished by enhancement of van der Waals (vdW) forces which facilitate self-healing in relatively narrow copolymer compositional range. In this work we report the acceleration of self-healing in alternating/random hydrophobic acrylic-based copolymers in the presence of confined water molecules. Under these conditions competing vdW interactions do not allow H 2 O-diester H-bonding, thus forcing nBA side groups to adapt L-shape conformations, generating stronger dipole-dipole interactions resulting in shorter inter-chain distances compared to ‘key-and-lock’ associations without water. The perturbation of vdW forces upon mechanical damage in the presence of controllable amount of confined water is energetically unfavorable leading the enhancement of self-healing efficiency of hydrophobic copolymers by a factor of three. The concept may be applicable to other self-healing mechanisms involving reversible covalent bonding, supramolecular chemistry, or polymers with phase-separated morphologies.",
"introduction": "Introduction Placing monomer units in an orderly fashion into a macromolecule may facilitate self-healing because upon mechanical damage, neighboring polymer chains return to their original conformations due to enhanced van der Waals interactions 1 . This approach is advantageous because it eliminates chemical and physical alterations and enables multiple recovery of thermoplastic polymers upon mechanical damage, thus expanding their functionality and sustainability. Obtaining materials with a longer life span also requires consideration of external environments to which polymers are exposed, for example, water. Hydrophobic nature of the majority of polymers though suggests that the presence of hydrophilic water should not impact self-healing properties. For that reason, to achieve water-induced self-healing, multilayered polyelectrolytes 2 and redox-switchable supramolecular 3 were proposed or sugar moieties 4 incorporated into polymer networks. Considering that the hydrophobic effect is critical in many diverse phenomena, from the cleaning of laundry to emulsion synthesis or the assembly of proteins into functional complexes, theoretical studies 5 have taught us that this typically multifaced effect depends on whether hydrophobic molecules are individually isolated or reassembled into larger hydrophobic structures. For example, water molecules can readily participate in four H bonds with a single methane molecule, but in larger hydrophobic aggregates, such as polymers, hydration of water is significantly diminished 6 . Here, we show that if a polymer is physically damaged resulting in a chain separation, water molecules may disrupt vdW interactions and participate in self-H-bonding, thus affecting self-healing. When mechanical load is removed, unfavorable polymer–water interactions within hydrophobic domains will lead to the expulsion of water from the system and rapid regeneration of polymer–polymer interactions due to enhanced interchain cohesive energies, thus leading to potentially faster self-repair."
} | 791 |
19382535 | null | s2 | 5,945 | {
"abstract": "Mimicking nature's approach in creating devices with similar functional complexity is one of the ultimate goals of scientists and engineers. The remarkable elegance of these naturally evolved structures originates from bottom-up self-assembly processes. The seamless integration of top-down fabrication and bottom-up synthesis is the challenge for achieving intricate artificial systems. In this paper, technologies necessary for guided bottom-up assembly such as molecular manipulation, molecular binding, and the self assembling of molecules will be reviewed. In addition, the current progress of synthesizing mechanical devices through top-down and bottom-up approaches will be discussed."
} | 172 |
37061785 | PMC10114076 | pmc | 5,946 | {
"abstract": "Abstract Fatty acids are important molecules in bioenergetics and also in industry. The phylum cyanobacteria consists of a group of prokaryotes that typically carry out oxygenic photosynthesis with water as an electron donor and use carbon dioxide as a carbon source to generate a range of biomolecules, including fatty acids. They are also able to import exogenous free fatty acids and direct them to biosynthetic pathways. Here, we review current knowledge on mechanisms and regulation of free fatty acid transport into cyanobacterial cells, their subsequent activation and use in the synthesis of fatty acid-containing biomolecules such as glycolipids and alka(e)nes, as well as recycling of free fatty acids derived from such molecules. This review also covers efforts in the engineering of such cyanobacterial fatty acid-associated pathways en route to optimized biofuel production.",
"conclusion": "Future perspectives and concluding remarks FA metabolism in cyanobacteria shows many particularities, likely a reflection of photoautotrophy. These include a number of specific FA residue-containing small molecules, of unique enzymes acting on FA-derived moieties and of unique metabolic pathways associated with FAs, which could be useful to several industrial sectors (biofuel, pharmaceutics, materials, cosmetics, and agriculture). A comprehensive understanding of FAs metabolism in bacteria, including metabolite flow and kinetics is lacking, even more so in cyanobacteria. To this aim, isotopic labeling is the most widely used method to trace lipid metabolism by mass spectrometry lipidomics; however, recent development of non-invasive methods based on click and bioorthogonal chemistry, such as azides reporters, alkyne lipids and analogs, lipid head group labeling as well as their detection procedures, are opening new avenues both for tracing lipid metabolism by mass spectrometry lipidomics as well as tracking their localization by optical and electron microscopy (Gerbersdorf 2006 , Kurmayer et al. 2020 , Morón-Asensio et al. 2021 , Kuerschner and Thiele 2022 ). Such new methodologies have led to considerable progress in understanding the evolution and development of early life forms (Bhattacharya et al. 2019 ). These have also allowed the study of lipid post-translational modifications and have inspired the development of novel therapeutic strategies (Knittel and Devaraj 2022 , Wang and Chen 2022 ). Finally, a better understanding of cyanobacterial FA metabolism-associated pathways and their associated enzymes would equally pave the way toward the discovery of new biocatalysts and bioactive secondary metabolites of potential biotechnological interest. FAs metabolism is often studied as an internal cellular process. However, the global abundance of FAs and the fact that they are prone to be taken up or released by microorganisms implies a biological FA cycle (and not only a biological hydrocarbon cycle). Further insight into the FA intracellular metabolism, into the mechanisms of FFA incorporation originating from extracellular sources, their secretion pathways, will help gauge the importance of such a cycle at the ecosystem level. Likewise, the evolution and distribution of particular FA-associated traits in microbial communities will enable a better appreciation of the role of FFAs in nature.",
"introduction": "Introduction Fatty acids (FAs) are among the most important type of biomolecules, as they play key roles as cellular structural components and in energy flow. FAs share the same core structure, with a carboxylate terminus and an aliphatic chain that can be either saturated or unsaturated. These molecules can be classified according to the length of their aliphatic chain: small-chain fatty acids hold up to 6 carbons, medium-chain fatty acids feature from 6 to 10 carbons, and long-chain fatty acids (LCFAs) hold more than 12 carbons (Schönfeld and Wojtczak 2016 , Wu et al. 2017 ). FAs are most commonly linear and composed of an even number of carbon atoms, but can sometimes be branched. They are mostly insoluble in water and organize spontaneously in an aqueous solution to form micelles. FAs represent an important source of metabolic energy to all organisms. They play a key role in carbon and energy storage in the form of neutral lipids or lipid droplets (Wältermann and Steinbüchel 2005 ). It has also been known for decades that FAs can be found covalently bound to hundreds of different proteins, i.e. fatty acylated proteins. Protein acylation has been implicated in the regulation of intracellular trafficking and signaling, subcellular localization, targeting, as well as in protein–protein and protein–lipid interactions (Resh 1999 , 2016 , Frank et al. 2015 ). Recently, Monson et al. ( 2021 ) reported an increasing appreciation of lipid droplets, their FA components and other lipids in viral infection defense mechanisms. Additionally, a large number of secondary metabolites incorporate acyl groups derived from FAs (Morstein et al. 2022 ). Nevertheless, the bulk of FAs is utilized for the synthesis of complex lipids as essential elements of cell membranes. Because FAs are the major components of lipids, their structural properties and specific arrangement into lipids can be connected to the phylogeny of the producing organisms, to environmental conditions, and even to microbial community structure (Villanueva et al. 2017 , Ding et al. 2021 ). De novo production of FAs is an essential process across Eukaryota and Eubacteria. This is an energetically costly process and, to offset this, most microorganisms have the capacity to take up and incorporate exogenous free fatty acids (eFFAs) directly from their environment (Dippold and Kuzyakov 2016 , Yao and Rock 2017 ). Some types of lipids, such as acyclic isoprenoids, sterols, bacteriohopanepolyols, glycerol dialkyl glycerol tetraethers, and carotenoids are ubiquitous in the environment and are preserved in geological material for long periods. Their non-hydrolysable hydrocarbon cores (from simple hydrocarbons to FA chain esters) provide information related to the phylogeny, physiology, and environmental conditions of the microbial community where they originated from. Thus, those lipids are commonly used as microbial biomarkers in paleobiology (Summons et al. 2022 ). Lipid hydrolysis, or deacylation, results in the release of free fatty acids (FFAs). Even though no information is available regarding the global amounts of FFAs accessible to microorganisms in soil and aquatic environments, FAs remain valuable tools to measure carbon input, cycling, and transfer of materials, for instance in food webs (Ruess and Chamberlain 2010 , Parrish 2013 , De Carvalho and Caramujo 2014 ). Soil FFAs were first thought to originate mainly from plant material (Moucawi et al. 1981 ), as lipids are an important constituent of plant biomass. In fact, around 3%–10% of aboveground and 0.5%–5% of belowground plant biomass is comprised of lipids, which thus represent an essential component of the plant-derived carbon input in soil (Dippold and Kuzyakov 2016 ). However, it eventually became clear that soil biodiversity and biomass are strongly dominated by microorganisms, bacteria, and fungi, as opposed to plants (Bahram et al. 2018 , Delgado-Baquerizo et al. 2018 ). Worldwide, one billion bacterial cells distributed between 10 2 and 10 6 bacterial phylotypes reside per gram of soil (Roesch et al. 2007 , Bickel and Or 2020 ). As lipids represent around 10% of the microbial biomass, mainly in cell membranes and cell walls (Dippold and Kuzyakov 2016 ), they significantly contribute to the lipidic soil organic matter pool as well. In the marine environment, lipids originate mainly from phytoplankton, which includes the photoautotrophic cyanobacteria, fecal pellets, and, to a much lesser extent, heterotrophic bacteria (Frka et al. 2011 , Gašparović et al. 2013 ). The levels of lipids (and among them FFAs) in aquatic environments are estimated to be usually higher in coastal regions than in open waters (Marić et al. 2013 ). In the aquatic environment, FFAs are often reported as the second most abundant class of lipids, their main source being dissolution from the particulate fraction (Testerman 1972 , Frka et al. 2011 , Triesch et al. 2021 ). Despite their abundance, it is hard to estimate how often microbial cells are exposed to FFAs in nature, and how microbial cells balance de novo synthesis of FAs with import of eFFAs. Microbial incorporation of eFFAs has been previously reviewed (Yao and Rock 2017 ), but did not cover autotrophic bacteria such as cyanobacteria, which have a unique primary metabolism, and appear to have a differentiated FA metabolism (Beld et al. 2016 , Broddrick et al. 2016 ). These prokaryotes can sustain themselves from water, carbon dioxide, inorganic substances, and sunlight. It is widely accepted that their metabolism is supported by oxygenic photosynthesis, from which they can convert light energy into chemical energy. They synthesize numerous types of biomolecules (including FAs) through the fixation of carbon dioxide, and can thus be independent from the incorporation of organic carbon sources. Cyanobacteria are the only prokaryotes that have evolved oxygenic photosynthesis (Sánchez-Baracaldo et al. 2022 ). These organisms arose on Earth ∼3.5 billion years ago and contributed significantly to shaping the Earth’s present atmosphere, since they were responsible for the first great oxygenation event (Huisman et al. 2018 ). After such a long evolutionary history, they possess today diverse lifestyles, morphologies, and metabolic capabilities (Tomitani et al. 2006 , Beck et al. 2012 , Willis and Woodhouse 2020 ). This can be illustrated by the multiple habitats in which cyanobacteria can be found, including aquatic (freshwater and marine systems) and terrestrial (e.g. in microbial soil crusts) environments, as well as in extreme environmental settings, such as hot springs, hot and cold deserts, and even in highly contaminated areas (e.g. with metals) (Tomitani et al. 2006 , Singh et al. 2016 ). Despite their autotrophic lifestyle, cyanobacterial display metabolic plasticity as evidenced, e.g. by the globally distributed mixotrophic marine picocyanobacterial strains of the genera Prochlorococcus and Synechococcus (Muñoz-Marín et al. 2020 , Ford et al. 2021 ). One of the earliest reports of photo-assimilation of an organic carbon source dates back to 1967, when a small number of phylogenetically diverse cyanobacteria were found to take up acetate and further incorporate it into mainly lipids as well as free soluble glutamate and carboxylic acids, under light- and CO 2 -dependent conditions (Hoare and Moore 1965 ). In addition to these sources of organic carbon, the incorporation and assimilation of eFFAs in several cyanobacteria are well-established. Among the various studies revolving around eFFAs assimilation, it has been noted several times that cyanobacteria, unlike other microorganisms, seem to lack a β-oxidation pathway (von Berlepsch et al. 2012 , Figueiredo et al. 2021 , Kawahara and Hihara 2021 ). This pathway, considered ubiquitous across the tree of life (Pavoncello et al. 2022 ) generates energy and allows for recycling of cellular FAs by breaking them down in what can be regarded as the reverse pathway of bacterial fatty acid synthesis II (FASII). Thus, in cyanobacteria, recycling of FAs likely occurs through a different route. Particularly interesting is the high amount of FA-derived secondary metabolites that have been detected in cyanobacteria, which is in accordance with the existence of multiple FA-utilizing pathways (Figueiredo et al. 2021 ). Secondary metabolites derived from FAs were reported to come either directly from primary FA metabolism (Leão et al. 2015 ) or from FFAs that are loaded onto a secondary-metabolite biosynthetic pathway-specific acyl carrier protein (ACP) by fatty acyl AMP ligases (FAALs; Kleigrewe et al. 2016 , Fewer et al. 2021 , Martins et al. 2022 ), or even directly from FFAs, as reported in one study so far (Reis et al. 2020 ). Cellular functions associated with FAs in cyanobacteria are generally in line with those described for other bacteria (Fig. 1 ). Here, we review current knowledge on the incorporation, fate, and recycling of FFAs in cyanobacteria (Fig. 2 ). We first look at how eFFAs are incorporated into cyanobacterial cells, namely their transport and activation. We then follow their path as they are integrated into a number of different lipid families, most of which are specific to cyanobacteria. We finally examine their possible fates, including their re-conversion into FFAs and potential secretion routes from the cyanobacterial cell. Figure 1. General overview of the various cellular processes in cyanobacteria that utilize fatty acids (FAs). FAs are key elements of cyanobacterial membranes (PG—phosphatidylglycerol, MGDG—monogalactosyldiacylglycerol, SQDG—sulfoquinovosyldiacylglycerol, and DGDG—digalactosyldiacylglycerol), are used as energy source, are important for post-translational modifications of specific proteins, are involved in trafficking, signaling, and targeting, can be stored in the form of lipid droplets, and are components of several types of secondary metabolites. Figure 2. Schematic representation of the fate and turnover pathways of exogenous free fatty acids (eFFAs) upon import by cyanobacteria. After entering the cyanobacterial cell, eFFAs are initially incorporated (A) by the acyl–(acyl-carrier-protein) synthetase (Aas), which catalyzes their activation to fatty acyl-AMP, and then to fatty acyl-ACP. At this point, activated FFAs can be directed to lipid (B) or alka(e)ne (C, D) biosynthesis pathways, or be further elongated through the fatty acids biosynthesis II (FASII) pathway. In addition, FFAs can be released either by lipases directly from lipids (E) or by the sequential activity of aldehyde-deformylating oxygenase (Ado) and aldehyde dehydrogenase (ALDH) (F). FFAs can later be reactivated through the Aas (G) or be released to the extracellular medium. In the latter case, cell lysis is a major contributor to the release of FFAs (as well as other fatty acid-containing molecules, such as lipids and alka(e)nes) into the environment (I), but FFAs can also be actively secreted through dedicated transporters (H) or extracellular vesicles (J). eFA, exogenous fatty acid; FFA, free fatty acid; Aas, acyl–acyl carrier protein synthetase; FASII, fatty acid synthesis II; Aar, acyl–acyl carrier protein reductase; Ado, aldehyde deformylating oxygenase; Ols, olefin synthase; Lip, Lipase ALDH, aldehyde dehydrogenase; Rnd, resistance nodulation division; ACP, acyl–acyl carrier protein; AMP, adenosine monophosphate; SQDG, sulfoquinovocyldiacylglycerol; DGDG, digalactosyl diacylglycerol MGDG, monogalactosyl diacylglycerol; PG, phosphatidyl glycerol. The localization of most of the proteins and of their products remains unclear. Cyanobacteria appear to possess either the Aar/Ado pathway or the Ols one (not both)."
} | 3,779 |
28927381 | PMC5604501 | pmc | 5,947 | {
"abstract": "Background Recently, important discoveries regarding the archaeon that functioned as the “host” in the merger with a bacterium that led to the eukaryotes, its “complex” nature, and its phylogenetic relationship to eukaryotes, have been reported. Based on these new insights proposals have been put forward to get rid of the three-domain Model of life, and replace it with a two-domain model. Results We present arguments (both regarding timing, complexity, and chemical nature of specific evolutionary processes, as well as regarding genetic structure) to resist such proposals. The three-domain Model represents an accurate description of the differences at the most fundamental level of living organisms, as the eukaryotic lineage that arose from this unique merging event is distinct from both Archaea and Bacteria in a myriad of crucial ways. Conclusions We maintain that “a natural system of organisms”, as proposed when the three-domain Model of life was introduced, should not be revised when considering the recent discoveries, however exciting they may be.",
"conclusion": "Conclusion: The three-domain model of life best reflects biological reality and should thus be retained Because of the considerations brought forward here, we propose to retain the three-domain model, with the understanding that the primary taxonomic division among cellular organisms ought to reflect that 3 basic cell types exist. First, an archaeal cell type, with isoprenoid units attached to glycerol-1-phosphate by ether linkages forming their membranes and a codon usage not shaped by inosine. Then, another prokaryote, the bacterial cell type with ester-linked lipids (fatty acids linked to glycerol-3-phosphate by ester linkages) and a codon usage mainly moulded by a bacterial tRNA dependent uridine methyltransferase. Finally, the complex cell type resulting from eukaryogenesis, the type representing the “Eucarya” of Woese, Kandler and Wheelis. Here codon usage reflects the massive use of inosine in anticodons, membranes are of the bacterial type, two types of ribosomes - both different from prokaryotic ones- are present, and a completely reconfigured metabolism resulting from the interactions between both prokaryotes involved, can be found. Later on, some eukaryotes even acquired further ribosomes (e.g. in chloroplasts). In conclusion, our proposal does not challenge the primacy of the phylogenetic approach, and allows the first taxonomic division to be a functional division, reflecting biological reality. Three phylogenetic trees can now be build, separately, for each of the domains. We have to start with a tiny Darwinian forest before we can tend to its massive trees."
} | 667 |
34102056 | PMC8289248 | pmc | 5,948 | {
"abstract": "Polymer composites\nhave attracted increasing interest as thermal\nmanagement materials for use in devices owing to their ease of processing\nand potential lower costs. However, most polymer composites have only\nmodest thermal conductivities, even at high concentrations of additives,\nresulting in high costs and reduced mechanical properties, which limit\ntheir applications. To achieve high thermally conductive polymer materials\nwith a low concentration of additives, anisotropic, solid-state drawn\ncomposite films were prepared using water-soluble polyvinyl alcohol\n(PVA) and dispersible graphene oxide (GO). A co-additive (sodium dodecyl\nbenzenesulfonate) was used to improve both the dispersion of GO and\nconsequently the thermal conductivity. The hydrogen bonding between\nGO and PVA and the simultaneous alignment of GO and PVA in drawn composite\nfilms contribute to an improved thermal conductivity (∼25 W\nm –1 K –1 ), which is higher than\nmost reported polymer composites and an approximately 50-fold enhancement\nover isotropic PVA (0.3–0.5 W m –1 K –1 ). This work provides a new method for preparing water-processable,\ndrawn polymer composite films with high thermal conductivity, which\nmay be useful for thermal management applications.",
"conclusion": "Conclusions and Outlook This work provides a new method\nfor producing water-based polymer\ndrawn films with high thermal conductivity, which are potentially\nuseful for thermal management in electrical devices, like foldable\nvideo screens and flexible solar cells. Oriented PVA/GO composite\nfilms were fabricated through water evaporation\nand solid-state stretching. SEM images and WAXS results reveal the\nimproved dispersion of GO and the high orientation in drawn PVA/GO\nfilms when the co-additive (SDBS) is used. FTIR spectra of composite\nfilms demonstrated the presence of hydrogen bonding between PVA and\nGO. These results contributed to the high thermal conductivity of\ndrawn PVA-5 composite films in the drawing direction, which is higher\nthan most composite films and approximately a 50-fold enhancement\nin comparison with isotropic PVA. It is tempting to speculate\nfurther on the applications of these\nfilms in, for instance, flexible solar cells and foldable video screens.\nFor instance, solar cells usually have an efficiency below 25% and\nthe residual absorbed energy is transferred into heat. The heating-up\nof the devices can be quite significant (100 °C), which actually\nreduces their efficiency and lifetime enormously. To satisfy the need\nin the foldable and flexible devices, the light density and flexible\nthermally conductive polymer composites were presented in this work.",
"introduction": "Introduction Due to their ease of\nprocessing, high corrosion and electrical\nresistances, and relatively low costs and weight, polymers are widely\nused in daily life. 1 − 3 However, bulk polymers generally have relatively\nlow thermal conductivities (usually <1 W m –1 K –1 ), which limit their application as thermal management\nmaterials for heat exchangers, electronic devices, and solar cells,\nfor example. 3 − 7 The thermal conductivity of polymers can be enhanced by adding\nfillers with high thermal conductivity. 3 − 10 Typically, thermally conductive nanocarbon materials including graphite\nnanoplatelets, carbon nanotubes, graphene, and their derivatives are\nblended as additives into the polymer matrix to achieve an increased\nthermal conductivity. 4 , 6 , 8 Still,\nthe thermal conductivity of isotropic polymer composites with a high\nconcentration of additives is <10 W m –1 K –1 due to the poor compatibility and interaction between\nthe polymer matrix and additives, resulting in poor dispersion of\nthe additives and serious phonon scattering 3 − 6 as well as the deterioration of\nthe mechanical properties of the composites. Recently, high\nthermal conductivities in anisotropic polymers,\nincluding polyethylene (PE) microfilms and micro/nanofibers, 10 − 14 polyvinyl alcohol (PVA) microfilms, 15 − 17 polyamide nanofibers, 18 and their composite films, have been demonstrated\nvia high degrees of chain orientation, chain extension, and crystallinity.\nFor instance, stretched polyethylene containing graphene nanoplatelets\n(draw ratio of ∼5) showed a thermal conductivity of ∼6\nW m –1 K –1 with a weak van der\nWaals interaction at the interfaces between the polyethylene and graphene\nnanoplatelets. 7 Combined high thermal conductivity\n(∼75 W m –1 K –1 ) and visible\ntransparency were obtained using polyethylene with 0.1 wt % graphene\nand a low melting point compatibilizer (2-(2 H -benzotriazol-2-yl)-4,6-ditertpentylphenol,\nBZT) at a draw ratio of 100. 9 As\npreviously reported, enhanced intermolecular interactions between\nthe polymer chains and the additives are an advantage for obtaining\nhigh thermal conductivity in drawn composite films. 9 , 19 − 21 While anisotropic polymer composites with high thermal\nconductivities have been reported for stretched polyethylene, to date\nstudies of more polar anisotropic polymer composites, like PVA films\nemploying green processes (that is, avoiding organic solvents), have\nbeen less prevalent. PVA is an atactic, water-soluble polymer with\nan orthorhombic unit cell and planar zigzag configuration similar\nto polyethylene, resulting in dense packing in the crystal lattice\nwith a high theoretical modulus, strength, and thermal conductivity. 16 , 22 , 23 Here, we report high thermal\nconductivities in anisotropic PVA/graphene\noxide (GO) composite films containing a second additive (SDBS) produced\nvia casting and solid-state drawing. We demonstrate that the combination\nof hydrogen bonding between GO and aligned PVA contributes to an improved\nthermal conductivity (∼25 W m –1 K –1 ) in PVA/GO composite films, greater than that of most polymer composites.",
"discussion": "Results and Discussion To obtain\nhomogeneous PVA/GO composite films, PVA, GO, and SDBS\nwere employed ( Figure 1 ). SDBS, which has a high melting point (∼200 °C), was\nincluded as a compatibilizer, while GO was used instead of graphene\nto enhance the dispersion in PVA. The resulting solution/dispersion\nwas poured into polystyrene molds and dried at 60 °C for 2 days.\nThe composite films were then stretched at 130 °C to a draw ratio\nof ∼5. The effects of GO (0–5 wt %) and SDBS (0–5\nwt %) concentrations and draw ratios were systematically studied in\nthe drawn polymer films (see Table 1 : for this paper, the nomenclature PVA- n generally represents a drawn PVA composite film containing n wt % GO, variable amounts of SDBS, and is drawn to a ratio\nof 5). Higher concentrations of GO (>5 wt %) were not studied due\nto the generally poor dispersion and stretchability of the composite\nfilms. Figure 1 (a) Fabrication process for drawn PVA composite films. (b–d)\nChemical structures of graphene oxide (GO, simplified chemical structure),\nsodium dodecyl benzenesulfonate (SDBS), and PVA. The thermal conductivities of the drawn PVA/GO composite films\nwere measured as a function of the concentration of GO ( Figure 2 a), revealing that the thermal\nconductivity increases with increasing concentration of GO. The highest\nthermal conductivity, ∼25 W m –1 K –1 , was obtained with drawn PVA-5, exhibiting approximately 3 times\nthe thermal conductivity of drawn neat PVA-0 films and an approximately\n50-fold enhancement in thermal conductivity in comparison to pure,\nisotropic PVA-0 ( Figure 2 d and Table S1 ). Figure 2 (a) Thermal conductivity\nof drawn pure PVA-0 without SDBS (red\nblock) and thermal conductivities of PVA-0(2), PVA-1, PVA-2, and PVA-5\nfilms, each containing 1 wt % SDBS with an increasing GO content.\n(b) Thermal conductivities of drawn PVA-5(2), PVA-5, and PVA-5(3)\nfilms containing variable concentrations of SDBS with 5 wt % of GO.\n(c) Photographs of undrawn (i) and drawn (ii) PVA-5 films and the\nOM image (iii) of drawn PVA-5 films. (d) Thermal conductivities reported\nfor different films in the literature. 9 , 10 , 13 , 15 , 16 , 23 − 30 The x axis represents the wt % contents of the\nthermally conductive additives. The red symbols represent drawn polymers\nor composites. Here, GN, RGO, GT, CNT, and PVDF represent graphene,\nreduced graphene oxide, graphite, carbon nanotube, and polyvinylidene\nfluoride, respectively. The role of the SDBS\nsurfactant was also studied ( Figure 2 a,b). The thermal conductivity\nof drawn PVA-0(2) decreases with the addition of SDBS compared to\nPVA-0 ( Figure 2 a),\nwhich may be expected since the addition of low molecular weight additives\ncommonly reduces thermal conductivity. 9 In contrast, adding SDBS (up to 1 wt %) increases the thermal conductivity\nof drawn PVA-5 (compared to PVA-5(2)): further increasing the SDBS\ncontent to 5 wt % (sample PVA-5(3)) results in the thermal conductivity\ndecreasing again ( Figure 2 b). Apparently, adding up to 1 wt % of the surfactant likely\nenhances the dispersion of GO ( Figure S1 ), but the poor thermal conductivity of SDBS dominates at higher\ncontents, resulting in an overall lower conductivity. The experimental\ndata in Figure 2 also\nshow that adding GO without SDBS increases the thermal conductivity\nfrom ∼8 to ∼16 W m –1 K –1 , which illustrates that the increase in thermal conductivity (from\n∼8 to ∼25 W m –1 K –1 ; Figure 2 a) originates\npartly from both additives. The mechanical effects of adding SDBS\nwere examined using DMA ( Figure S2 and Table S2 ). The results indicate that a high content of SDBS decreases the\nstorage modulus and increases tan δ. In other words, SDBS behaves\nas a mechanical plasticizer, especially at high SDBS contents. Please\nnote that a very high SDBS content was used here (5 wt %), which is\nfar greater than the optimum SDBS content (1 wt %). In addition, Young’s\nmodulus decreases with the addition of SDBS, while the maximum draw\nratio increases ( Table S2 ), probably further\nresulting from the addition of SDBS as a plasticizer. Photographs\nof undrawn and drawn PVA-5 films are shown in Figure 2 c, i and ii. The\nOM image of the drawn PVA-5 film indicates that the addition of both\nadditives (GO and SDBS) improves the dispersion (compare Figure S3 and Figure 2 c, iii), although there is aggregation due\nto the excessive content of GO in PVA films. In Figure 2 d, the thermal conductivity of a wide variety\nof undrawn and drawn films described in the literature is shown as\na function of the content of thermally conductive additives, including\ngraphene, carbon nanotubes, graphene oxide, and mixtures ( Table S1 ), 9 , 10 , 13 , 15 , 16 , 23 − 33 revealing the impressive thermal conductivity in drawn PVA-5 films\nwith low additive contents. Drawn PE and PE/GN films showed greater\nthermal conductivity than drawn PVA/GO films in this work, probably\ndue to the high orientation and crystallinity induced by ultrahigh\ndraw ratios in drawn PE and PE/GN films, which could not be obtained\nin PVA/GO films. 10 , 11 WAXS was performed to further\nanalyze the drawn PVA/GO films ( Figure 3 a,b and Figure S4a,d ). Undrawn\nPVA-0 films without additives\nshow two scattering rings of lattice planes of (101) at 2⊖\n≈ 19.5° and (100) at 2⊖ ≈ 11.6°, corresponding\nto the PVA crystalline domains ( Figure S4a ), while these scattering rings coalesce into scattering dots in\ndrawn PVA films ( Figure 3 a), indicating some alignment of PVA crystalline domains. A significant\namorphous halo is also observed in both films, indicating little orientation,\nto be expected at these low draw ratios. Figure 3 WAXS patterns of drawn\nPVA-0 (a) and drawn PVA-5 (b) composite\nfilms. The insets are the 1D curves of X-ray scattering. Here, the\nplane of measured films is perpendicular to the incident X-ray ( Figure S5 ). SEM images of the cross section of\ndrawn PVA-0 (c) and drawn PVA-5 (d) composite films. (e) Non-polarized\nFTIR spectra of drawn PVA-0, PVA-0(2), and PVA-5. In the case of the polymer composite films, the scattering rings\nof PVA (at 11.6 and 19.5°) ( Figure S4c ) also transform into scattering dots upon stretching ( Figure 3 b), revealing the alignment\nof PVA and suggesting a low through-plane alignment of GO, although\nthe scattering rings of GO (lattice plane: (002), 2⊖: ∼25.5°)\ntransform into weak scattering arcs ( Figure 3 b). Herman’s orientation function\ncalculated by the full width at half-maximum (FWHM) 24 , 25 reveals that both the drawn PVA films with and without graphene\n(PVA-0 and PVA-5) have a high degree of orientation (∼0.9)\nof the crystalline domains of PVA, indicating that there is no obvious\neffect of adding GO and SDBS. Anisotropy of SDBS in drawn PVA-5 films\nwas also observed in the WAXS patterns due to the fact that the scattering\nrings of SBDS in undrawn films ( Figure S5c ) transfer into dots in drawn films ( Figure S5d ). Figure S5c reveals anisotropic GO in\nthe plane of drawn PVA-5 films. These results suggest drawing-induced,\nsimultaneous alignment of PVA and GO, similar to drawn PE and graphene\nfilms in the literature. 8 , 16 In the cross section\nof SEM ( Figure 3 c,d),\nthe drawn PVA-0 and PVA-5 films show more aligned and fibrillar structures\nthan undrawn PVA-0 and PVA-5 films ( Figure S4e,g ). Homogeneous dispersion of GO in PVA-5 films is observed in Figure S6a,b ; the morphology of GO is characterized\nin Figure S6c,d , indicating that the size\nof GO is smaller than 2 μm, which is smaller than that in Figure 2 , probably due to\naggregation or the interlayer splitting induced by stretching. 16 The interaction between PVA and the additives\nwas investigated\nusing infrared spectroscopy. FTIR spectra of drawn PVA-0 films with\nand without 1 wt % SDBS ( Figure 3 e and Figure S7a ) show vibration\npeaks at ∼3274 and 3297 cm –1 , respectively,\nattributed to the −OH group of PVA. In contrast, the drawn\nPVA films with 5 wt % GO and 1 wt % SDBS show a red-shifted (to lower\nenergies/wavenumbers) peak at ∼3260 cm –1 ,\nindicating the hydrogen bonding interaction between GO and PVA ( Figure S7a ). 16 , 34 , 35 Although WAXS results indicate the orientation of\nthe PVA chains, there is no obvious orientation of hydrogen bonding\nin the polarized FTIR spectra ( Figure S7b,c ), consistent with a previous literature report. 17 There is a drawing-induced shift to higher energies/wavenumbers\nof the absorption peak of −OH in both PVA-0 and PVA-5 films\nin comparison with the undrawn PVA-0 and PVA-5 films, respectively\n( Figure S7a ), as reported as undrawn and\ndrawn polyacrylonitrile and PVA in the literature. 36 , 37 FTIR spectra of pure SDBS powder and the undrawn PVA-0(2) film are\nshown in Figure S8 as control experiments,\nand the Raman spectrum confirms the oxidization group-induced defects\nin the GO powder ( Figure S7d ). Simple\nsetups were created to do a basic thermal analysis of representative\ncomposite films and to explore their potential application as heat\ntransport and heat release elements ( Figure 4 ). First, the ends of three films are fixed\nto the right end of a copper plate, and the left end of the copper\nplate was exposed to 194 F (80 °C) at time = 0 min ( Figure 4 b). After 2 min,\nthe drawn PVA-5 film shows the highest temperature, while the undrawn\nPVA-5 film exhibits the lowest temperature at the same position ( Figure 4 c), indicating a\nhigher thermal transport and thermal conductivity in the drawn PVA-5\nfilm than both drawn PVA-0 and undrawn PVA-5 films. After heating,\na cooling experiment was conducted to help clarify the heat release\nprocess of the composite films ( Figure 4 d,f). For the cooling measurement, the ends of two\ndrawn films, one PVA-0 and the other PVA-5, were fixed atop stainless\nsteel pillars, as shown in Figure 4 d, and exposed to 194 F (80 °C) for 10 min, before\nremoving the entire setup to ambient conditions, and the decrease\nin sample temperature was recorded ( Figure 4 e). The drawn PVA-5 film exhibited a greater\ntemperature decrease and a lower final temperature after 2 min than\nthe drawn PVA-0 film ( Figure 4 f), demonstrating that the drawn PVA-5 films more effectively\nrelease heat, which makes them potentially useful in devices as thermal\nmanagement materials. Figure 4 Schematic pictures of the thermal analysis module in the\nheating\n(a) and cooling (d) processes. (b, c) Thermal analysis of undrawn\nPVA-5, drawn PVA-0, and drawn PVA-5 films of similar sizes (highlighted\nby the dotted red boxes) during the heating process. (e, f) Thermal\nanalysis of the drawn PVA-0 and PVA-5 films (highlighted by the dotted\nwhite boxes) during the cooling process. F is Fahrenheit."
} | 4,117 |
21952217 | PMC3195254 | pmc | 5,950 | {
"abstract": "Rising petroleum costs, trade imbalances and environmental concerns have stimulated efforts to advance the microbial production of fuels from lignocellulosic biomass. Here we identify a novel biosynthetic alternative to D2 diesel fuel, bisabolane, and engineer microbial platforms for the production of its immediate precursor, bisabolene. First, we identify bisabolane as an alternative to D2 diesel by measuring the fuel properties of chemically hydrogenated commercial bisabolene. Then, via a combination of enzyme screening and metabolic engineering, we obtain a more than tenfold increase in bisabolene titers in Escherichia coli to >900 mg l −1 . We produce bisabolene in Saccharomyces cerevisiae (>900 mg l −1 ), a widely used platform for the production of ethanol. Finally, we chemically hydrogenate biosynthetic bisabolene into bisabolane. This work presents a framework for the identification of novel terpene-based advanced biofuels and the rapid engineering of microbial farnesyl diphosphate-overproducing platforms for the production of biofuels.",
"discussion": "Discussion Here we present a framework for the rapid identification and development of a terpene-based advanced biofuels that sets precedent for the discovery of other advanced biofuels. In the absence of a commercial source of bisabolane, chemical hydrogenation of commercial bisabolene led to the identification of bisabolane as a biosynthetic alternative to D2 diesel. The flexibility of the E. coli and S. cerevisiae FPP-overproducing platforms allowed us to rapidly switch from the overproduction of amorphadiene to that of bisabolene, the immediate precursor to bisabolane. In E. coli , bisabolene synthase screening and codon-optimization followed by metabolic engineering of the heterologous mevalonate pathway led to a tenfold increase in bisabolene titers. In S. cerevisiae , screening of the bisabolene synthases led to bisabolene titers >900 mg l −1 , the highest sesquiterpene titer in this organism to date. Further, we showed that bisabolene toxicity is not a limit in the microbial production of bisabolene in either S. cerevisiae or E. coli . Finally, we demonstrated chemical conversion of biosynthetic bisabolene into bisabolane using chemical hydrogenation. To our knowledge, this is the first report of a reduced monocyclic sesquiterpene, bisabolane, as a biosynthetic alternative to D2 diesel, and the first microbial overproduction of bisabolene at >900 mg l −1 . It is currently difficult to obtain large quantities of biosynthetic bisabolene to hydrogenate and test its fuel properties. The shake flask production experiments reported in this manuscript have been performed with an organic overlay for analysis purposes. Although both the E. coli and S. cerevisiae strains produce high levels of the sesquiterpene, purification of bisabolene is currently challenging due to co-evaporation of the product and the organic overlay. Further improvements in bisabolene production or an alternative separation technology will be required to obtain the quantities of highly pure bisabolene needed for fuel properties testing. Through multiple rounds of large-scale preparation in shake flasks, we have prepared ~20 ml of biosynthetic bisabolene, hydrogenated it, and performed preliminary fuel properties testing. Though the results are preliminary due to an insufficient sample amount for complete testing, the preliminary cetane number of the hydrogenated biosynthetic bisabolene is 52.6. Once the complete fuel properties of hydrogenated biosynthetic bisabolene can be obtained, we will be able to estimate the impact of byproducts present in the hydrogenated commercial bisabolene, such as farnesane and aromatized bisabolene ( Fig. 3b ). Given the similar ratios of potentially beneficial (farnesane) and detrimental (aromatized and partially reduced bisabolenes) byproducts in the hydrogenated commercial bisabolene, we estimate that the measured physicochemical properties reported in this work can be largely attributed to the major product, bisabolane. While bisabolane has physicochemical properties similar to D2 diesel, it must be produced at high yields from the sugar source using a relatively simple process to be economically viable. Here we produced bisabolene in E. coli at ~36% of apparent pathway-dependent theoretical yield, and at 4% of pathway-dependent theoretical yield in minimal medium. We are currently optimizing the fermentation conditions, pathway genes and the organism to improve the productivity in minimal medium and also to produce larger quantities of biosynthetic bisabolene for complete fuel property analysis. Further, the low toxicity of bisabolene to E. coli and S. cerevisiae should allow production of bisabolene at very high titers. An economic analysis on the production of bisabolene takes into consideration many variables including the cost and type of feedstock, biomass depolymerization method, and the microbial yield of biofuel. Assuming a break-even price of sugar at the mill to be close to US $0.10/lb, which is lower than the current volatile spot price but closer to the long-term nominal price of the commodity, we can perform a rough calculation of the theoretical cost of bisabolene. On the basis of that number, the raw material cost of bisabolene production would be, ignoring non-sugar costs, approximately $0.88 per kg of bisabolene. Assuming raw material costs to be only ~50% of the final cost, this would imply a final cost of $1.76 per kg, or $5.73 per gal of bisabolene on the basis of data for ethanol production 24 . (We assumed the raw material cost for bisabolene as 50% of total cost which is lower than that of ethanol (66%) on the assumption that the bisabolene production process will be more involved and, thus, more costly than the production of ethanol). At an estimated ~$6 per gal, bisabolene is currently more expensive than current D2 diesel. However, it is still promising to investigate this emerging alternative biofuel when considering its superior properties and renewable nature. Finally, in this work we have resorted to chemical hydrogenation of bisabolene into the final product bisabolane. While industrially feasible, the ultimate goal is the complete microbial production of bisabolane. This will require the reduction of terpenes in vivo using designer reductases and, potentially, balancing cellular reducing equivalents."
} | 1,605 |
28322307 | PMC5359605 | pmc | 5,951 | {
"abstract": "Structure-dependent colour is caused by the interaction of light with photonic crystal structures rather than pigments. The elytra of longhorn beetles Tmesisternus isabellae appear to be iridescent green in a dry state and turn to red when exposed to humidity. Based on the hygroscopic colouration of the longhorn beetle, we have developed centimeter-scale colorimetric opal films using a novel self-assembly method. The micro-channel assisted assembly technique adopts both natural evaporation and rotational forced drying, enhancing the surface binding of silica particles and the packing density by reducing the lattice constant and structural defects. The fabricated large-scale photonic film changes its structural colour from green to red when exposed to water vapour, similarly to the colorimetric feature of the longhorn beetle. The humidity-dependent colour change of the opal film is shown to be reversible and durable over five-hundred cycles of wetting and drying.",
"conclusion": "Conclusion We were able to design and fabricate a centimeter-scale opal film over a using a rotational self-assembled technique, inspired by tunable structural colour of the longhorn beetle T. isabellae . The photonic film changed in colour from green to red when exposed to water vapour, which is similar to the hygrochromic feature of the longhorn beetle. This proposed opal film showed reversible and durable colorimetric detections over five-hundred cyclic experiments, making it a low-cost structural coloration and sensing mechanism for quantifying environmental humidity without electricity.",
"discussion": "Results and Discussions Photonic structures of the longhorn beetle The structural colours of the dry and wet states of the longhorn beetle T. isabellae are shown in Fig. 1a,b . In the scales of the beetle, a multilayer photonic structure yields a colour switch from golden green to red exposed to humidity. As the relative humidity is increased, the golden coloured region turns red within a few minutes. The black band coloration of the elytra remains unchanged. The red coloured region recovers to the golden colour upon return to the dry state ( Supplementary video 1 ). Thus, the humidity-induced colour change is reversible. The magnified views of the elytral scales in the dry state were characterized by scanning electron microscopy (SEM, genesis-1000, EmCrafts) in Fig. 1c,d . The approximate length and thickness of a scale are 150 μm and 4 μm, respectively. The visible optical response to humidity is caused by multilayer structures in the interior of the scales. The transverse cross-sectional view of one scale is shown in Fig. 1e . The colorimetric photonic structure of the beetle scale consists of two alternating layers ( Fig. 1f ). One is a homogenous melanoprotein layer having a thickness of 100 ~ 110 nm. The other is a mixed layer consisting of melanoprotein nanoparticles and air voids. The mixed layer has a thickness of 70 ~ 80 nm with a relatively high lattice constant (the spacing between particles). The structural coloured region has a contact angle of 29.2°, facilitating water infiltration into the hydrophilic multilayers 21 . Although both melanoprotein and mixed layers absorb water, only melanoprotein layers swell noticeably after water absorption in the wet state ( Fig. 1g ). The air voids in the mixed layers are infiltrated with water without swelling. Therefore, hygroscopic colour change of the beetle scales occurs due to both the refractive index shift by water absorption of the mixed layers and the dimensional change by swelling of the melanoprotein layers 21 . When a longhorn beetle T. isabellae is initially in a dry state and exposed to water vapour, the reflection spectra of the coloured region of elytra can be measured using a spectrometer analyser (CCS100, Thorlabs, Newtown, NJ, USA). The elytra of the beetle in the dry state show a golden green colour with the reflectance peak at 578.3 nm ( Fig. 1h ). The reflection peak shifts from 578.3 nm to 632.6 nm with a peak increment of 54.3 nm, causing a dramatic change in the structural colour. The measured reflection peaks are consistent with colour observation by the naked eye. The reflection spectrum band is relatively broad because the beetle scale has multilayer structures consisting of alternating high and low refractive-index layers 21 . Design of bio-inspired colorimetric opal films In order to fabricate colloidal opal films to have a similar humidity-dependent colour change (green to red shift) to the longhorn beetle T. isabellae , we calculated the size of colloidal particles to fill in three-dimensional photonic crystals. The theoretical reflection wavelength of opal photonic crystals composed of spherical particles was approximately predicted using a modified Bragg’s equation for normal incidence, . Here λ is the peak wavelength of reflected light, d = 0.816 D is the interlayer spacing in the [1 1 1] direction, D represents the mean diameter of colloidal particles, and is the effective refractive index of the sample. The effective refractive index of a two-phase structure can be estimated as 28 34 , . Here f = 0.74 is the fraction of the colloidal particles for an ideal face-centred cubic package. n sphere and n void are the refractive indices of the colloidal particles and air void, respectively. In this study, we use silica particles with a refractive index of n sphere = 1.55. In the dry state, n void = n air = 1. In the wet state, water infiltrates into the air void and the refractive index becomes n void = n water = 1.333. Using the given parameters, we calculated the diameter of silica particles for the bio-inspired humidity sensor using the Bragg’s equation. At D = 252 nm, the theoretical reflection wavelength of opal photonic crystals is λ = 578 nm (green colour) in the dry state. When water penetrates into the air voids in the wet state, the theoretical reflection peak shifts to λ = 615 nm (red colour). The theoretical photonic band-gap shift due to only water penetration into the opal structures without swelling can be calculated asΔ λ = 615–578 = 37 nm. Therefore, we expected that opal photonic structures using silica particles with a diameter of 252 nm would generate bio-inspired hygrochromic coloration (from green to red), similarly to tunable color of the longhorn beetle. Optical and mechanical features of self-assembled opal films A novel self-assembled process for fabricating colloidal crystals using a micro-channel assisted dip-coating and a rotational drying method was proposed by the authors 35 . The rotating self-assembly technique for producing the colorimetric opal film is summarized in Fig. 2 . Besides predominant point and line defects within self-assembled colloidal crystals, macroscopic cracks inevitably evolve during the drying of the colloidal particles on rigid substrates due to compressive stresses perpendicular and tensile stresses parallel to the plane of the support material 36 37 . In order to reduce defects and cracks in opal films, the proposed self-assembly adopted a two-stage drying process of natural evaporation ( Fig. 2b ) and rotation-induced forced drying ( Fig. 2c ). Figure 3a shows the fabricated colorimetric opal film using the self-assembled silica particles with a diameter of 252 nm. The fabricated opal film was 15 mm × 40 mm × 100 μm in size. The structural green colour was clearly displayed in the dry state. When the film was exposed to water vapour, its colour changed to red in a few minutes. Figure 3b shows the structural colour change of the film according to the exposure time to water vapour. The entire film became completely red after 90 s. The red colour changed to the original green colour when the photonic film was exposed to air. On the other hand, the rotational drying process does not only enhance the structural uniformity and durability of the opal film, but it also reduces dramatically the time required for colloidal crystallization, compared to conventional evaporation methods. An average of approximately 150 min is required to fabricate opal structures with the size of a standard glass substrate using this method. The manufacturing time of the photonic crystal structures depends mainly on the thickness of the micro-channel and increases with decreasing thickness. The spectrum wavelengths of the fabricated film in the dry and wet states are shown in Fig. 3c . After perpendicularly generating a white light source to the surface of the film, the reflecting light was collected to analyse optical spectra. The reflection peak in the dry state was observed at a wavelength of 579.8 nm, which shifted to 626.1 nm in the wet state. The measured reflection shift was Δ λ = 626.1–579.8 = 46.3 nm. The experimental reflection peaks using silica particles with a diameter of approximately 250 nm agree with the theoretical values (578 nm and 615 nm) in the dry and wet states and a peak shift of 37 nm. The reflection spectra of the fabricated film in the dry and wet states are similar to those of the longhorn beetle T. isabellae . The opal film, consisting of monosized silica spheres, has much narrower reflection peaks compared to complex multilayer structures of the longhorn beetle. On the other hand, a higher reflection peak shift (Δ λ = 54 nm) of the longhorn beetle ( Fig. 1h ) was observed. It is because the colour change of the beetle elytra is due to the combined effect of the refractive index shift by water infiltration and swelling of the melanoprotein layers 21 . Using self-assembly with hard spheres, the particle size dispersity and the defect density deteriorate the optical properties of monodispersed photonic crystals 38 . The reduced defect density is directly reflected in remarkably sharp and tunable Bragg peaks in the optical absorption. Therefore, relatively sharp reflection peaks of the opal film ( Fig. 3c ) also indicate the low size dispersity of colloidal particles and low defects in photonic crystals. Figure 4a shows a magnified photograph of the fabricated opal film near an interface between the barrier-seed region and the rotation-induced crystallization. There existed many discontinuous layers in the barrier-seed region, which was constructed by natural evaporation and gravitational sedimentation of colloidal particles. This structural non-uniformity was mainly due to inconsistent convection and evaporation, induced by environmental variations such as temperature and humidity during the relatively long-time natural-drying process. Figure 4b,d display scanning electron microscope (SEM, genesis-1000, EmCrafts) images in two magnifications of the barrier-seed region. The natural evaporation process constructed loose-packed crystallization with lack of structural densification and uniformity. The crystal structure included many line defects on discontinuous interfaces of the loose-packed colloidal assembly, as well as point defects. The barrier-seed region, which was constructed at the channel tip during natural drying, prevented the leakage of suspension during rotational drying. Moreover, it acted as seed layers by providing a lattice frame for the self-assembly of colloidal spheres during rotational drying. Figures 4c,e show SEM images under two magnification factors of the highly ordered region, which was constructed by rotational drying. Since the surface binding of colloidal particles was enhanced, close-packed structures were assembled with high densification. There were no line defects or cracks in the opal film except for point defects. Some dot-like defects were present because of the size dispersity or geometric shape of the silica particles. The local defects that formed in the fabrication process grow extensively in disturbance or repeated use, and therefore structural durability of photonic crystals could be degraded. Since colloidal particles pack together without chemical bonds, the loose-packing of silica particles by natural drying include could be easily disassembled due to defect growth and crack propagation by exposure to repeatable humidfication. However, the rotational drying technique intensified colloidal aggregation and densification, enhancing structural durability to external disturbances. The quality of the nano-sized spheres used is also crucial in order to obtain crystalline order in colloidal crystals. Two main limiting factors for the fabrication of colloidal crystals are the presence of spheres smaller than the average diameter and nonspherical particle formation 39 . It is noted that the rotational drying method can dramatically reduce defect density in crystallization of colloidal particles, compared to the naturally dried photonic crystals with the same colloidal size dispersity. The transverse cross-sectional SEM images of the close-packed region display three-dimensional face-centered cubic (FCC) structures with approximately 400 layers in the 100-μm-thick film ( Fig. 4f,g ). The cross-sectional views confirmed that the photonic film consisted of FCC structures rather than hexagonal close-packed (HCP) crystallization with the same packing density (0.74). To directly compare the longhorn beetle and bio-inspired colorimetric film, partial pieces of the film and beetle elytra were exposed to cyclic humidity conditions. Figures 5a,b show the colour variations of both samples over a cycle of successive wetting and drying. Both samples showed very similar colour changes according to the exposure time of water vapour, and became red when they were completely wet in the water after 80 s. Moreover, the samples recovered to their original colours after drying, demonstrating that the hygroscopic colour-change of the film sensor is reversible ( Supplementary video 2 ). Colour-change images of the opal film and elytra were display after 200 and 500 cyclic wetting and drying experiments using a humidifier and an electric drier ( Fig. 5c ). The film sample showed good durability without additional defects or cracks for five-hundred cyclic wetting and drying experiments. Notably, the colorimetric opal film showed reversible and repeatable performance in humidity detection, as well as excellent mechanical stability. Most of colloidal crystals are very fragile because colloidal particles pack together without chemical bonds and the ordered packing of the particles can be easily disassembled in water or solvents 40 . Repeatable wetting-drying disturbances on self-assembled colloidal crystals can cause defect growth and crack propagation, leading to abrupt structural failure. However, SEM image of the opal film after 500 cycles of wetting and drying confirmed that the close-packed colloidal particles had not been disassembled, having structural durability to repetitive humidification ( Fig. 5d ). The experimental results of the fabricated opal film proved that the proposed drying technique dramatically reduced structural defects or cracks and it also intensified colloidal aggregation without additional crosslinking agents 41 42 . For the practical applications of tunable photonic crystals, it is required to manufacture large-scale photonic crystals with low defects in a highly efficient and reproducible method. Several self-assembly techniques have been proposed for large-area photonic crystals using spin coating 43 , doctor blade coating 44 , spray coating 45 , and a roll-to-roll Langmuir-Blodgett (LB) 46 methods. However, it is still difficult to fabricate highly scalable, defect-free colloidal crystals of non-crosslinked spheres. In addition to the requirements for high scalability and low defect density, structural durability to repeatable humidification is also necessary for the hygroscopic photonic films. This proposed fabrication method, based on the micro-channel deposition and the two-stage evaporation process, could be one of practical self-assembly techniques for the colorimetric application of tunable photonic crystals."
} | 3,998 |
32318116 | PMC7158082 | pmc | 5,952 | {
"abstract": "Industrial crops are grown to produce goods for manufacturing. Rather than food and feed, they supply raw materials for making biofuels, pharmaceuticals, and specialty chemicals, as well as feedstocks for fabricating fiber, biopolymer, and construction materials. Therefore, such crops offer the potential to reduce our dependency on petrochemicals that currently serve as building blocks for manufacturing the majority of our industrial and consumer products. In this review, we are providing examples of metabolites synthesized in plants that can be used as bio-based platform chemicals for partial replacement of their petroleum-derived counterparts. Plant metabolic engineering approaches aiming at increasing the content of these metabolites in biomass are presented. In particular, we emphasize on recent advances in the manipulation of the shikimate and isoprenoid biosynthetic pathways, both of which being the source of multiple valuable compounds. Implementing and optimizing engineered metabolic pathways for accumulation of coproducts in bioenergy crops may represent a valuable option for enhancing the commercial value of biomass and attaining sustainable lignocellulosic biorefineries.",
"conclusion": "Conclusions and future perspectives Plant metabolic engineering towards the production of chemicals has emerged as a promising approach to enhance crop value. With the advancement of biotechnological tools in both synthetic biology and plant transformation techniques, the understanding and implementation of metabolic pathways in plants became feasible [ 316 , 317 ]. These technical improvements dramatically accelerate the Design–Build-Test–Learn (DBTL) cycles of plant metabolic engineering, and important increases of target biochemicals in crops are now achieved at a faster pace. One of the challenges is the testing of engineered crops under field conditions to asses stress resilience and possible yield penalty. As exemplified by the case study of crops engineered for production polyhydroxyalkanoates, subcellular compartmentalization represents an effective strategy to alleviate toxicity associated with high titers of bioproducts [ 318 ]. Similarly, promising strategies have been developed for introducing organelles that allow bioproduct sequestration and accumulation in engineered plant tissues [ 319 ]. In several cases, extraction and purification of biochemicals from plant biomass represent other important challenges to overcome for rendering biorefineries economically attractive [ 8 ]. In addition to the work conducted by plant metabolic engineers to increase titers of specific chemicals in crops, which in some instances has resulted in remarkable increases by more than two orders of magnitude after several years of research [ 320 ], an emphasis should also be given to the development of isolation and purification processes to render plant-derived chemicals economically competitive compared to their petroleum-derived equivalents. Moreover, engineering approaches and biomass processing should be evaluated with life cycle analyses to assess the environmental impacts of a specific bioproduct and inform on the types of crops to improve. Next generation of holistic biorefineries that include upstream biomass extraction step(s) prior to hydrolysis and conversion of lignocellulose are expected to beneficiate from these value-added coproduct traits implemented in bioenergy crops. The development of solvents and extraction methods compatible with existing biorefineries should enable the integration of novel streams that generate valuable coproducts while reducing recovery costs [ 321 , 322 ]. Furthermore, in a concept of one-pot biomass conversion process, the release from engineered lignocellulosic feedstocks of either target bioproducts or their immediate metabolic precursors during biomass pretreatment and saccharification offers a potential for increasing final bioproduct yields, but this approach will necessitate the development of microbial strains that are tolerant to inhibitors found in lignocellulosic hydrolysates [ 323 , 324 ]. For our future bioenergy crops, exploiting diverse metabolic pathways inherent to plants such as the shikimate and isoprenoid pathways will certainly contribute to the supply of several valuable biochemicals that find multiple industrial applications. Such endeavor is intended to reduce the production of fossil fuel-derived chemicals and our dependence on petroleum."
} | 1,114 |
21963157 | null | s2 | 5,953 | {
"abstract": "Resilin is an elastomeric protein found in specialized regions of the cuticle of most insects, providing outstanding material properties including high resilience and fatigue lifetime for insect flight and jumping needs. Two exons (1 and 3) from the resilin gene in Drosophila melanogaster were cloned and the encoded proteins expressed as soluble products in Escherichia coli. A heat and salt precipitation method was used for efficient purification of the recombinant proteins. The proteins were solution cast from water and formed into rubber-like biomaterials via horseradish peroxidase-mediated cross-linking. Comparative studies of the two proteins expressed from the two different exons were investigated by Fourier Transform Infrared Spectroscopy (FTIR) and Circular Dichrosim (CD) for structural features. Little structural organization was found, suggesting structural order was not induced by the enzyme-mediated di-tyrosine cross-links. Atomic Force Microscopy (AFM) was used to study the elastomeric properties of the uncross-linked and cross-linked proteins. The protein from exon 1 exhibited 90% resilience in comparison to 63% for the protein from exon 3, and therefore may be the more critical domain for functional materials to mimic native resilin. Further, the cross-linking of the recombinant exon 1 via the citrate-modified photo-Fenton reaction was explored as an alternative di-tyrosine mediated polymerization method and resulted in both highly elastic and adhesive materials. The citrate-modified photo-Fenton system may be suitable for in vivo applications of resilin biomaterials."
} | 402 |
35198904 | PMC8851274 | pmc | 5,954 | {
"abstract": "Summary Extracellular electron transfer (EET) from microorganisms to inorganic electrodes is a unique ability of electrochemically active bacteria. Despite rigorous genetic and biochemical screening of the c -type cytochromes that make up the EET network, the individual electron transfer steps over the cell membrane remain mostly unresolved. As such, attempts to transplant entire EET chains from native into non-native exoelectrogens have resulted in inferior electron transfer rates. In this study we investigate how nanostructured electrodes can interface with Shewanella oneidensis to establish an alternative EET pathway. Improved biocompatibility was observed for densely packed nanostructured surfaces with a low cell-nanowire load distribution during applied external forces. External gravitational forces were needed to establish a bioelectrochemical cell-nanorod interface. Bioelectrochemical analysis showed evidence of nanorod penetration beyond the outer cell membrane of a deletion mutant lacking all outer membrane cytochrome encoding genes that was only electroactive on a nanostructured surface and under external force.",
"conclusion": "Conclusions Silver nanorod arrays were synthesized using template-assisted electrodeposition with periodic (AAO) and aperiodic (PC) commercial nanoporous templates. The resulting nanostructured electrodes obtained varied nanorod dimension, geometry, and distribution. Densely packed Ag nanorod arrays were obtained with the AAO templates, whereas the PC template yielded a sparser distribution of high-aspect-ratio nanorods. Using centrifugation as external force, S. oneidensis cells were attached to the nanostructured surface. This process led to decreased cell viability compared with flat Ag surfaces, where PC showed the most antimicrobial properties. Comparatively, the AAO nanostructured surface maintained a stable biocompatibility with about 40% of cells surviving the treatment. SEM imaging revealed a close interaction between cells and the nanostructured surface with evidence of cells being pierced by PC nanorods, while situated on top of the AAO nanorods. A feature that is represented in the low number of nanorods connected to each cell for PC, which caused an increased localized load distribution on each cell for the PC nanostructured surface compared with the AAO. The superior cell viability on AAO nanostructures is likely caused by the large number of nanorods connected to each cell leading to a low localized cell load during centrifugation. Bioelectrochemical analysis of the nanostructured electrodes showed a distinct oxidation current coming from S. oneidensis WT on the AAO electrode, similar to the response on 2D Ag electrodes. Effective cell-nanorod connections were observed in the MET potential region with dramatic current increase in the DET potential region. Analogous with the poor cell viability, no significant bioelectrochemical activity was observed on the PC nanostructured surface. Thus, there seems to be a strong correlation between cell viability during cell-nanorod interfacing and its bioelectrochemical activity. Studies on the S. oneidensis deletion mutant ΔOMC revealed a selective bioelectrochemical response for the AAO electrode with activation taking place at slightly higher potentials compared with the WT strain (i.e., 150 mV). The centrifugation step was also found to be crucial for establishing a strong cell-nanorod interaction as background level currents were observed without centrifugation prior to the bioelectrochemical analysis. In conclusion, interfacing electroactive bacteria with conductive nanostructured electrodes can be achieved using centrifugation as external force. A strong correlation between the nanostructured geometry and cell viability was observed with improved biocompatibility for highly ordered and densely packed nanorod arrays. Evidence of nanowire penetration beyond the outer cell membrane of S. oneidensis was found for an outer membrane cytochrome deletion mutant that was only electroactive on AAO nanostructured surfaces and under external force. This early development of intracellular nanostructured probes for electroactive bacteria could provides an important tool for the ongoing research into transplanting entire EET chains into non-native exoelectrogens and the overall understanding of the EET pathway. In this pursuit, further method development will be needed to actively connect as many living cells as possible to the nanostructured electrode, which can be achieved by, e.g., reducing the nanostructure dimensions and increasing the nanorods/cell coverage to create a more disperse load distribution over the cell membrane. The application of this new method is especially interesting for microbial bioelectrosynthesis technologies as we can now basically connect any microorganism to a nanostructured electrode and feed it electric current, the cheapest and most sustainable electron donor, for production of various important biofuels and platform chemicals. Limitations of the study This study takes the first step to integrating intracellular electron detection in exoelectrogenic bacteria using nanostructured electrodes. Although the Ag nanowired arrays based on AAO templates showed superior biocompatibility and electrochemical response, the use of 220-nm-wide nanorods was not ideal given the similar dimensions of S. oneidensis . Future studies will focus on developing nanowired arrays with sub-100-nm diameters, while maintaining a high areal density, to further improve the signal-to-noise ratio. The synthesis could thereby be aided by the use of templates with a metallic thin film sputtered on one side for electrical contact and more homogeneous growth rate during electrodeposition.",
"introduction": "Introduction Electrochemically active bacteria have sparked intense scientific interest recently, mainly due to their ability of converting various different organic substrates to higher oxidized compounds of biotechnological interest while producing electrical energy as a by-product in bioelectrochemical systems such as microbial fuel or electrolysis cells ( Logan 2009 ; Logan et al., 2019 ; Richter et al., 2012 ). To balance their redox metabolism, respiratory microorganisms require a pathway to transfer electrons to an external terminal electron acceptor. Certain bacteria have evolved a respiratory metabolism using non-soluble metals in oxygen-deprived environments. Bacteria capable of reducing iron- and manganese-based minerals can often achieve this by direct electrical contact between the cell and metal surface ( Kumar et al., 2017 ; Logan 2009 ; Logan et al., 2019 ; Richter et al., 2012 ). Direct electron transfer is made possible by a conductive multiprotein electron transport chain that extends through the cell membrane allowing electron flow outside of the cell. This rather unique ability is most notably expressed in model organisms for dissimilatory iron reduction such as Shewanella oneidensis and Geobacter sulfurreducens .( Kumar et al., 2017 ; Logan 2009 ; Richter et al., 2012 ). S. oneidensis is a particularly versatile organism as it can utilize a wide variety of terminal electron acceptors such as metallic- or carbon-based electrodes as well as various different metal ions (e.g., Fe 3+ , Mn 4+ , U 6+ , and Cr 6+ ).( Fredrickson et al., 2008 ; Gralnick et al., 2006 ; Hunt et al., 2010 ; Richter et al., 2012 ). Extracellular electron transfer (EET) in organisms like S. oneidensis stems from a dynamic network of c -type cytochromes most of which contain multiple electrochemically active heme groups. The complexity of the EET network is illustrated by the 41 genes that code for c -type cytochromes in S. oneidensis ( Richter et al., 2012 ). Through rigorous genetic and genomic investigations, a few key proteins have been identified to be crucial for extracellular electron transfers within the organism. CymA is located in the inner membrane and transfers electrons from the menaquinone pool into the periplasm. The periplasm of S. oneidensis is too wide (approximately 23.5 nm) for direct electron transfer between membrane-bound cytochromes (i.e., CymA to MtrA) ( Beblawy et al., 2018 ). Electrons are instead transported via soluble periplasmic proteins (e.g., FccA and STC) through the periplasm to the outer membrane. Three proteins (MtrA, MtrB, and MtrC) build an outer membrane-spanning complex that transports electrons through the outer membrane outside the cell where outer membrane cytochromes (e.g., MtrC and OmcA) function as terminal reductases and facilitate the final step of extracellular electron transfer to the electron acceptor. Despite intensive research, a complete picture of the electron transfer network remains elusive ( Kumar et al., 2017 ; Richter et al., 2012 ). This is mainly due to (1) the multitude of c -type cytochromes that are simultaneously expressed and have partially overlapping functions, (2) the lack of specificity of electron transfer reactions between the involved EET proteins, and (3) the commonly used experimental setups that fail in detecting individual electron transfer steps through cytoplasmic membrane, periplasm, and outer membrane. Moreover, experiments aiming at transplanting entire electron transport chains from S. oneidensis to E. coli revealed that electron transfer can only be partially resolved and the reason for this remains enigmatic ( Jensen et al., 2010 ). The recent development in nanotechnology has opened up possibilities of integrating nanostructures in biological or medical research, mainly due to their similarity in length scales with cellular components. This new interdisciplinary research field of bionanoelectronics utilizes nanostructured cell interfaces to establish bidirectional communication between cells and electronics ( Noy 2011 ). Trans-membrane nanostructures mainly based on 1D nanomaterials (e.g., nanorods and nanotubes) are introduced into living cells to study internal electrical and biochemical responses to stimuli ( Higgins et al., 2020 ; Robinson et al., 2012 ; Stewart et al., 2018 ; VanDersarl et al., 2012 ; Xie et al., 2012 ). Spontaneous cell membrane penetration events are rare and a topic of discussion in the field ( Higgins et al., 2020 ; Lee et al., 2014 ; Robinson et al., 2012 ; Qing et al., 2014 ; Xie et al., 2015 ). Recent modeling reports have shown that cells naturally settling on nanostructured surfaces with dimensions larger than 10 nm in diameter are unlikely to achieve membrane penetration ( Xie et al., 2013 ). The need for external forces to achieve cell-nanorod penetration has led to development of nanostructured intracellular penetration strategies based on well-established membrane disruption techniques such as electroporation, mechanoporation, and optoporation ( Higgins et al., 2020 ; Stewart et al., 2018 ). Although the majority of reports focus on eukaryotic cells, there is growing interest in the interaction between prokaryotic cells and high-aspect-ratio nanostructures. This trend is largely driven by the antimicrobial properties ( Lin et al., 2018 ; Tripathy et al., 2017 ) that nanostructured surfaces have, and there are only a handful of reports on interfacing exoelectrogenic bacteria with nanostructured surfaces for mechanistic studies ( Jeong et al., 2013 ; Jiang et al. 2010 , 2013 ). Bacteria such as S. oneidensis and Sporumosa ovata have been found to recognize and self-assemble around nanostructured surfaces ( Jeong et al., 2013 ; Sakimoto et al., 2014 ). Nanostructured electrodes, designed to selectively control the cell-electrode contact, have also been used to study the EET pathway of S. oneidensis and G. sulfurreducens ( Jiang et al. 2010 , 2013 ). Nevertheless, in order to study individual electron transfer steps in situ , we need nanomaterials that can interface directly with the microorganism by penetrating the outer membrane. These materials could also allow us to bridge at least the outer membrane of any Gram-negative organism thereby enabling an outer membrane conduit-independent electron transfer pathway. So far, no such studies have been reported to the best of the authors’ knowledge. Nanostructured platforms for intracellular probing are typically based on Si and C nanostructures that require delicate and expensive synthesis (e.g., lithography) ( Higgins et al., 2020 ). Metallic nanostructures can be synthesized using electrodeposition that offers a cost-effective and non-toxic synthesis route with a high degree of scalability in terms of nanostructure dimensions and chemical composition ( Edström et al., 2011 ; Feng et al., 2016 ; Rehnlund et al., 2015 ; Walter et al., 2003 ). Free-standing metal nanostructures can be achieved by template-assisted electrodeposition where the deposition is directed into the pores of nano- or microporous templates. This bottom-up technique utilizes periodic or non-periodic commercial templates based on anodic aluminum oxide and track-etched polymers, respectively ( Edström et al., 2011 ; Hulteen and Martin 1997 ). The metal nanostructure dimensions are dictated by the template pore dimensions and the coulombic charge of the deposition. Fine control over the deposition rate can easily be achieved through the applied current density. The electrochemical synthesis method has been used to prepare free-standing metallic arrays of Al, Cu, and Ag nanorods ( Feng et al., 2016 ; Oltean et al., 2011 ; Rehnlund et al., 2015 ). This well-established synthesis route is highly suited for electrochemically active bacteria since they have a naturally high tolerance for metals, even known antimicrobial noble metals such as Ag and Au ( Baudler et al., 2015 ), making it possible to utilize metal-based nanostructures for intracellular electron detection. In the present study, we set out to investigate if metallic nanostructured electrodes can be interfaced with the exoelectrogenic bacteria to bridge at least the outer membrane of the organisms. Two different kinds of silver nanostructured electrodes were prepared using template-assisted electrodeposition. The choice of periodic and non-periodic templates (i.e., AAO and PC) allows us to investigate the effect that the nanowire density and areal distribution plays on the bacteria-nanowire interface. The two silver nanorod arrays were investigated and compared based on their biocompatibility and functionality in achieving a stable intracellular interface with native and genetically engineered S. oneidensis using centrifugation as the membrane disruption method. This study therefore takes the first steps toward transferring the recent progress in intracellular electron detection used in eukaryotes to the study of electrochemically active bacteria vital for microbial bioelectrochemical systems.",
"discussion": "Results and discussion Synthesis of nanostructured electrodes The first step toward designing a conductive nanostructured surface suitable for intracellular interrogation of exoelectrogenic microorganisms is to define the electrode and media composition. Copper and silver were selected as the electrode material based on their low electric resistivity and high chemical stability. Both metals were characterized using chronoamperometry at 0 V (versus SHE) in LB and M4 minimal media. As shown in Figures 1 A–1D, the copper substrate produced a steady reductive current of about 1.3 μA/cm 2 in the M4 media, whereas an oxidative current of about 11 μA/cm 2 was seen in the LB media. The silver substrate showed considerably less electrochemical activity in both media. Besides an initial fluctuation of a few microamperes for the M4 media, the silver electrode showed quite similar results in both media with an average current density of about 100 nA/cm 2 . With the goal of minimizing abiotic electrochemical activity, silver is clearly the best choice of material for developing nanostructured electrodes suitable for nanorod-cell interfacing. The choice of media is less obvious as silver seems quite stable in both chemical environments. However, the yeast extract in the LB medium is known to contain flavins that could function as redox active shuttles and contribute to an increased mediated electron transfer (MET) between the organisms and the nanostructured electrode ( Logan et al., 2019 ). As the aim of the study was to investigate the intracellular direct electron transfer (DET) pathway, it would be best to minimize the effect of MET and therefore the flavin-free medium M4 was selected. Figure 1 Electrochemical synthesis of nanostructured metal electrodes (A–D) Abiotic chronoamperometry measurements of Ag and Cu planar electrodes in (A and B) M4 minimal media and (C and D) LB media. (B and D) Magnified views of the Ag planar electrode response in both media. Mean current densities are shown with standard deviation error bars. (E–H) Scanning electron microscope (SEM) analysis of pristine silver nanostructured electrodes prepared by template-assisted electrodeposition using nanoporous (E and F) anodic aluminum oxide (AAO) and (G and H) polycarbonate (PC) membranes. Highlighted regions in (e and g) are magnified in (f and h) to show the individual nanorod dimensions and areal distribution. Silver nanostructured electrodes were prepared using template-assisted electrodeposition. Nanoporous templates based on anodic aluminum oxide (AAO) and track-etched polycarbonate (PC) membranes were used to prepare arrays of Ag nanorods. The two different templates were chosen based on their periodic (AAO) and non-periodic (PC) pore distribution and the difference in pore density. Similar pore diameters were, however, selected to discard its effect on the biocompatibility. The resulting nanostructured electrodes were compared based on nanorod dimensions, areal distribution, and interrod distances. The electrodeposition was composed of an initial rapid potentiostatic nucleation step followed by pulsed galvanostatic deposition. This protocol is known to improve the deposition coverage and yield a more homogeneous nanorod growth rate in the narrow template channels ( Oltean et al., 2011 ; Rehnlund et al., 2015 ). The resulting nanostructured surfaces are shown in Figures 1 E–1H. With the AAO template, a densely packed array of well-defined Ag nanorods were observed with an average diameter of 220 nm. The dense distribution of nanorods is also seen in the narrow spacing between nanorods (i.e., 40–70 nm) as well as in the high areal loading of nanorods (i.e., 9.8 nanorods/μm 2 ). The PC template produced a more diverse distribution of well-defined cylindrical nanorods with an average diameter of 140 nm and a nanorod density of 1.7 nanorods/μm 2 (see Table 1 ). Measurement of interrod distances was complicated by the fact that the PC template does not have perpendicular pores, which causes the resulting nanorods to extend at an angle (<90°) from the surface. This also caused considerable clusters of leaning nanorods to form, a feature likewise observed for the AAO template, albeit to a lesser degree. Nanorod agglomeration is a well-known phenomenon for high-aspect-ratio nanostructures, which is caused by a competition between the nanostructure length, flexibility, and interrod adhesion forces ( Pokroy et al., 2009 ; Rehnlund et al., 2015 ). An important difference is that underneath the nanorod clusters lies a nanorod-covered surface for AAO, whereas the PC deposition leaves a flat Ag surface underneath the protruding nanorods. The AAO deposition therefore produced a more even nanostructure coverage. A comprehensive summery of the nanorod dimensions and electroactive area of the nanostructured electrode is shown in Table 1 . The electroactive area was calculated based on the nanostructured surface area and normalized to 1 cm 2 . Table 1 Dimensions of Ag nanostructures as analyzed from SEM micrographs with Fiji image processing software. Sample Nanorod area/μm 2 Diameter/nm Nanorods/μm 2 (pristine) Nanorods/cell (real) Nanorods/cell (theoretical) Electroactive area/cm 2 AAO 0.037 ± 0.005 220 ± 17 9.8 10 26 10 PC 0.015 ± 0.003 140 ± 15 1.7 1.8 65 8 S. oneidensis 0.97 ± 0.18 Biocompatibility of Ag nanostructured electrodes with S. oneidensis The biocompatibility of the Ag nanostructured surfaces in regards to S. oneidensis cells was evaluated with fluorescence microscopy using commercial live/dead dyes. S. oneidensis was cultivated anoxically and applied to the silver electrodes using centrifugation in customized well plates. The similar dimensions of the Ag nanostructures and bacteria such as S. oneidensi makes it unlikely that spontaneous membrane-poration would occur. Without external applied force (i.e., without centrifugation) we could not reliably detect cells attached to any of the electrode surfaces (i.e., 2D, PC, or AAO) with fluorescence microscopy. Analysis of the electrode biocompatibility was therefore focused on cells attached with external force. Mechanoporation, using centrifugation as the applying force, was selected as the method of achieving an intracellular bacteria-nanorod interface. In order to minimize cell rupture, a mild treatment of 1,000 × g for 5 min was applied to load the cells on the Ag nanostructured surfaces. The centrifugation treatment yielded a stable cell attachment to the nanostructured surfaces as confirmed visually by the nanostructured electrodes becoming pink after the centrifugation step, an indication of cytochrome-rich S. oneidensis cells coated on the electrode surface. Also, SEM analysis showed electrodes fully covered by cells, despite the rigorous sample preparation performed prior to electron microscopy analysis. After centrifugation, the cells were stained with the live/dead dye and analyzed by fluorescence microscopy. Figures 2 A–2C show the resulting live/dead ratio of S. oneidensis WT on both AAO and PC nanostructured surfaces as well as a reference planar (2D) Ag electrode. It is immediately clear that the centrifugation step was not harmful for the organism as most cells survived the treatment on planar Ag surfaces (see Figure 2 A). However, the cell viability dropped drastically on the nanostructured surfaces, with the least amount of damage produced by the AAO nanorods. Quantitative analysis of the percentage of living cells revealed that 73.2 ± 7.7%; 39.6 ± 8.7%, and 8.2 ± 0.6% of cells survived on 2D, AAO, and PC surfaces, respectively. This suggests that the AAO nanostructured surface is more biocompatible than its PC equivalent. As seen in the fluorescence image, there seems to be very few cells intact on the PC sample. To help understand the reason for this seemingly antimicrobial property of the PC nanostructures, electron microscopy analysis of the cell-nanorod interface was performed. S. oneidensis WT cells were here loaded on AAO and PC nanostructured surfaces (using centrifugation) and then prepared for electron microscopy (see experimental section for EM sample preparation). On the AAO sample, cells are seen to mostly lie evenly on top of the nanorods with some cells squeezed in between nanorods of varying lengths. No evidence of nanorods protruding through the cells could be found. On the PC nanostructures, cells were found to be attached to single and multiple nanorods with evidence that the nanorods are embedded deep within the cell. This fact can be seen in Figure 2 E where a highlighted cell is attached to several nanorods and held there above the electrode surface. It therefore seems that S. oneidensis cells are more likely to be pierced by PC nanorods than AAO and that this interface leads to the cell membrane puncturing. By comparing SEM micrographs before and after cell loading (i.e., Figures 1 E–1H and Figures 2 D–2E) we observed that approximately 10 nanorods could connect to one cell for the AAO nanostructures, whereas only approximately 1.8 PC nanorods connected to one cell. Moreover, there was evidence that the PC nanostructures were damaged and disconnected during the centrifugation treatment. In contrast, the AAO nanostructures seemed unaffected by the centrifugation treatment. It is possible that some of the dislodged PC nanostructures caused critical cell damage in the solution phase during the centrifugation, which could explain the low living cell count seen in Figure 2 C. Another possibility is that the centripetal force localized on the nanostructured surface exceeded the Young’s modulus of the cell membrane causing it to rupture. The local load distribution of one cell was calculated assuming that the entire cell density was evenly distributed over the electrode surface during the centrifugation step. Loads of 2.4 and 33 kPa were estimated on each cell situated on the AAO and PC nanostructured surfaces, respectively (see supplemental information for further details). The elasticity of S. oneidensis biofilms have been found to range between 33 and 38 kPa, as measured by atomic force microscopy in the liquid phase ( Kim et al., 2017 ). Based on the biofilm properties of S. oneidensis it would seem plausible that cell membrane rupture could occur on the PC nanostructured surface, whereas it is unlikely on the AAO nanostructured surface. This difference is mainly due to the significantly higher nanorod density found for AAO, leading to cell loads distributed over a larger area. Figure 2 Effect of electrode architecture on the cell viability of wild-type S. oneidensis (A–E) Fluorescence microscopy analysis of S. oneidensis WT propelled against planar Ag electrodes using centrifugation (i.e., 1,000 × g ). Live/dead staining was implemented to show the ratio of live (green) to dead (red) cells due to severe cell membrane disruption. The live/dead analysis was performed on (A) 2D planar, (B) AAO, and (C) PC silver electrodes. SEM analysis of (D) AAO and (E) PC nanostructured electrodes after centrifugation with S. oneidensis WT. The cell morphology is highlighted in each electron micrograph for clarity. Bioelectrochemical analysis of S. oneidensis on Ag nanostructured electrodes Next, we investigated the bioelectrochemical activity of S. oneidensis WT and a deletion mutant in all five genes encoding outer membrane cytochromes (ΔOMC) on both AAO and PC nanostructured as well as planar Ag (2D). Wild-type S. oneidensis can here be regarded as a positive control and is expected to show bioelectrochemical activity on all electrode types. The deletion mutant is a genetically engineered strain where all outer membrane cytochromes are removed leaving it without the ability for extracellular DET ( Bücking et al., 2012 ). First linear sweep voltammetry was implemented to investigate the potential region where microbial oxidation takes place. Both organisms were investigated on the three electrode types with and without centrifugation applied prior to bioelectrochemical analysis (see Figure 3 ). With centrifugation, the wild-type strain showed a gradual increased oxidative current starting at about −0.45 V that culminated at −0.2 V. The strain showed a direct activation at −0.2 V on 2D Ag electrodes, indicative of mainly DET. In contrast, no significant bioelectrochemical response was observed for the PC nanostructures. Without centrifugation, no significant oxidation took place on either nanostructured electrode, whereas a sharp activation was observed for the 2D electrode. This result is unexpected as the wild-type strain should be able to connect and transfer its metabolic electron pool even without centrifugation, Thus indicating that the centrifugation step can achieve a stable cell loading on the electrode, a feature that takes considerable time without external forces. The ΔOMC strain showed a more selective behavior where only the AAO nanostructured electrode enabled a bioelectrochemical response in the analyzed voltage window. Here an oxidative current response was observed from −0.3 V and steadily increased toward higher potentials. No significant bioelectrochemical signal was observed on the 2D and PC electrodes. Likewise, without applied centrifugation no bioelectrochemical response was observed for either electrode type. Although a lack of bioelectrochemical activity is expected for 2D electrodes with the ΔOMC strain, the results indicate that the AAO nanostructures are uniquely capable of interfacing with the strain to provide a solid electron transfer route that can bypass the lack of outer membrane cytochromes. Figure 3 Effect of electrode architecture on the activation potential for S. oneidensis strains Linear sweep voltammetry analysis of S. oneidensis (A and B) WT and (C and D) ΔOMC deletion mutant (A and C) with and (B and D) without the application of centrifugation prior to the bioelectrochemical analysis. Potential regions where mediative (MET) and direct electron transfer (DET) occur are annotated in each voltammogram. Current densities were normalized to the individual electroactive electrode area. After establishing that S. oneidensis can be interfaced with the nanostructured electrodes we wanted to further investigate electron transfer of the WT and ΔOMC strain. Chronoamperometry was therefore implemented and set to probe the current response at -0.2 V (versus Ag/AgCl), which corresponds to 0 V (versus SHE). For the WT strain after centrifugation, an initial spike followed by a gradual current density decline was observed on the AAO and 2D electrodes with an average current density of 77.7 ± 5.1 and 17.2 ± 6.5 μA/cm 2 , respectively (see Figures 4 A, S2 A, and S2B). The PC nanostructured electrode showed comparatively no current response throughout the measurement with an average current density of 0.012 ± 0.001 μA/cm 2 , in agreement with the linear sweep voltammetry results. As expected, without centrifugation no significant current response was observed for either electrode type. The ΔOMC strain showed a similar current response to the WT strain after centrifugation on the AAO nanostructured electrode, albeit with considerably reduced current density (3.8 ± 1.8 μA/cm 2 ) (see Figures 4 B, S2 C, and S2D). In contrast, essentially no current response was observed on the PC and 2D electrode with centrifugation. Without centrifugation, all electrodes showed background level current densities, in agreement with the WT strain results. It can therefore be concluded that an external force (i.e., centrifugation) is needed to achieve a stable cell-electrode connection for DET. The current response observed for the WT strain on 2D and AAO as well as the ΔOMC strain on AAO is typically observed for exoelectrogenic microorganisms in bioelectrochemical systems (see Figure S2 ). The organisms start fully charged with a fully reduced quinone and cytochrome pool. As the potential of DET is supplied, the organisms can quickly discharge leading to a rapid current response that decays with time as either new organisms are attracted to the electrode surface or existing organisms connected to the electrode produce a steady current through anaerobic respiration. The latter is more likely in the case of forced cell-nanorod integration as shown in Figure 3 C. In conclusion, of the two nanostructured surfaces investigated, only the AAO platform resulted in a stable cell-nanorod connection for both the WT and ΔOMC strain. These results are in good agreement with the live/dead fluorescence microscopy results presented in Figure 2 . Figure 4 Electrochemical performance of S. oneidensis strains (A and B) Chronoamperometry analysis of S. oneidensis (A) WT and (B) ΔOMC deletion mutant was performed at −0.2 V (versus Ag/AgCl) with and without the application of centrifugation prior to the bioelectrochemical analysis. The current response was recorded for both strains using 2D (gray), AAO (orange), and PC (blue) Ag electrodes. Statistical analysis of all experiments shows the mean current density (with SD error bars). Current densities were normalized to the individual electroactive electrode area."
} | 8,023 |
23996084 | null | s2 | 5,955 | {
"abstract": "Electrochemical impedance spectroscopy has received significant attention recently as a method to measure electrochemical parameters of Geobacter sulfurreducens biofilms. Here, we use electrochemical impedance spectroscopy to demonstrate the effect of mass transfer processes on electron transfer by G. sulfurreducens biofilms grown in situ on an electrode that was subsequently rotated. By rotating the biofilms up to 530 rpm, we could control the microscale gradients formed inside G. sulfurreducens biofilms. A 24% increase above a baseline of 82 µA could be achieved with a rotation rate of 530 rpm. By comparison, we observed a 340% increase using a soluble redox mediator (ferrocyanide) limited by mass transfer. Control of mass transfer processes was also used to quantify the change in biofilm impedance during the transition from turnover to non-turnover. We found that only one element of the biofilm impedance, the interfacial resistance, changed significantly from 900 to 4,200 Ω under turnover and non-turnover conditions, respectively. We ascribed this change to the electron transfer resistance overcome by the biofilm metabolism and estimate this value as 3,300 Ω. Additionally, under non-turnover, the biofilm impedance developed pseudocapacitive behavior indicative of bound redox mediators. Pseudocapacitance of the biofilm was estimated at 740 µF and was unresponsive to rotation of the electrode. The increase in electron transfer resistance and pseudocapacitive behavior under non-turnover could be used as indicators of acetate limitations inside G. sulfurreducens biofilms."
} | 399 |
37050822 | PMC10098871 | pmc | 5,957 | {
"abstract": "Robotic manipulation challenges, such as grasping and object manipulation, have been tackled successfully with the help of deep reinforcement learning systems. We give an overview of the recent advances in deep reinforcement learning algorithms for robotic manipulation tasks in this review. We begin by outlining the fundamental ideas of reinforcement learning and the parts of a reinforcement learning system. The many deep reinforcement learning algorithms, such as value-based methods, policy-based methods, and actor–critic approaches, that have been suggested for robotic manipulation tasks are then covered. We also examine the numerous issues that have arisen when applying these algorithms to robotics tasks, as well as the various solutions that have been put forth to deal with these issues. Finally, we highlight several unsolved research issues and talk about possible future directions for the subject.",
"conclusion": "7. Conclusions In this survey, we have provided an overview of deep reinforcement learning algorithms for robotic manipulation. We have discussed the various approaches that have been taken to address the challenges of learning manipulation tasks, including sim-to-real, reward engineering, value-based, and policy-based approaches. We have also highlighted the key challenges that remain in this field, including improving sample efficiency, developing transfer learning capabilities, achieving real-time control, enabling safe exploration, and integrating with other learning paradigms. One of the key strengths of the survey is its comprehensive coverage of the current state-of-the-art in DRL and exploring a wide range of techniques and applications for robotic manipulation. The survey does not, however, go into great depth on all strategies due to space considerations. Nevertheless, the survey can greatly benefit researchers and practitioners in the field of robotics and reinforcement learning by providing insights into the advantages and limitations of various algorithms and guiding the development of new approaches. Overall, this survey serves as a valuable resource for understanding the current landscape of deep reinforcement learning in robotic manipulation and can inspire further research to advance the field. Future research in this field should concentrate more on overcoming these difficulties and figuring out how to improve the performance and efficiency of deep reinforcement learning algorithms for robotic manipulation tasks. By achieving this, we can get one step closer to developing intelligent robots that can adjust to different surroundings and learn novel abilities on their own. It is likely that as the area of RL for robotic manipulations develops, we will witness the creation of increasingly more sophisticated algorithms and methodologies that will allow robots to perform increasingly difficult manipulation tasks.",
"introduction": "1. Introduction Industry 4.0’s embrace of artificial intelligence (AI) has drastically changed the industrial sector by allowing machines to operate autonomously, increasing productivity, lowering costs, and improving product quality [ 1 ]. In order to make decisions quickly and efficiently during the manufacturing process, artificial intelligence (AI) technology analyzes enormous amounts of data produced by machines and processes. In the next five years, its total market value is expected to triple [ 2 ]. Robotics is one of the markets that is anticipated to expand at the fastest rates. With the concept of robot manipulation proposed in 1962 [ 3 ], the idea of a robot is to mimic human behavior and tackle complex tasks. A branch of robotics called robotic manipulation is focused on developing robots that can manipulate items in their environment. Robotic manipulation seeks to develop robotic systems that can carry out a variety of activities that call for the manipulation of things, such as putting together goods in a factory, picking and placing items in a warehouse, or performing surgery in a hospital. Robotic manipulation systems typically consist of a robot arm, a gripper or end-effector, and a control system that coordinates the movement of the robot arm and gripper. The gripper is in charge of grabbing and manipulating items, while the robot arm is in charge of transferring them to the target area in the environment. Figure 1 shows a classic robotic manipulation workflow. Moreover, Matt Mason provided a thorough and in-depth description of manipulation in the introduction of his 2018 review paper [ 4 ]. Robotic manipulation is used in various fields such as manufacturing, agriculture, healthcare, logistics, space exploration, education, and research. It entails using robots to carry out activities including assembling, planting, harvesting, operating, managing goods, and performing experiments. In a variety of tasks, robotic manipulation can boost productivity, cut human costs, increase accuracy, and enhance safety. In the upcoming years, its application is anticipated to increase across a range of industries. In the manufacturing sector, robots are used for tasks including component assembly, welding, painting, and packaging. They have the ability to work carefully and diligently, increasing productivity and decreasing costs. In the healthcare sector, robots may assist with tasks including surgery, rehabilitation, and geriatric care. They can support healthcare personnel by letting them focus on more challenging tasks while the robot handles the basic tasks. In space exploration, robots are used to complete activities including sample gathering, structure construction, and equipment repair. They can operate in environments that are too dangerous or challenging for people, such as deep space or the ocean floor. However, the interaction between robotic manipulator arms and objects designed for humans remains a challenge. Up to now, no robots can easily achieve intelligent operations such as handwashing dishes, peeling a pineapple, or rearranging furniture. A subfield of artificial intelligence called deep reinforcement learning combines reinforcement learning and deep learning, a technique for training artificial neural networks [ 5 ]. Robots are programmed with a set of rules and instructions that specify how they should interact with items in their surroundings in conventional methods of robotic manipulation. This approach is effective for simple activities, but it becomes more challenging as the difficulty of the tasks increases. Robots may manipulate objects in their surroundings using a process called deep reinforcement learning, which allows them to make mistakes and learn from them. Deep reinforcement learning provides a more flexible and adaptable method, allowing robots to learn from experience and change their behavior [ 6 , 7 ]. For instance, the robot receives positive reinforcement if it successfully picks up and moves an object to the desired location. If it drops the object or fails to transfer it to the desired location, a negative reward is provided. Because it has the capacity to correlate some activities with good results and other actions with undesirable outcomes, over time, the robot develops a strategy for accomplishing the task at hand [ 8 ]. Robotic manipulation using deep reinforcement learning has the potential to change a variety of industries, including healthcare and manufacturing. Allowing robots to learn from experience and adjust to changing situations enables them to perform tasks that are too difficult or dangerous for humans to complete. As research in this area advances, we can expect to see more capable and advanced robots that can manipulate objects more precisely and effectively. RL has been successfully applied to a wide range of problems in various fields, including game-playing, recommendation systems, and finance. One successful use of RL is in the game of Go [ 9 ], where DeepMind’s AlphaGo program defeated the world champion Lee Sedol in 2016 to attain unparalleled success. In addition to its application in finance to improve trading tactics and portfolio management, RL has been used to create recommendation systems that can learn to give individualized suggestions depending on user behavior [ 10 ]. Moreover, RL has been used for complicated issues in which conventional rule-based systems or supervised learning approaches fall short, including those in the domains of natural language processing [ 11 ], drug discovery [ 12 ], and autonomous driving [ 13 ], among others. Creating algorithms that can precisely detect and distinguish objects in images or video streams is one of the primary issues in computer vision. By giving agents performance-based feedback in the form of rewards or penalties, RL can be utilized to teach agents to recognize objects [ 6 ]. As RL algorithms are capable of learning from experience and adapting to shifting settings, they provide a promising solution to a wide range of difficult issues in several industries. In this survey, we will examine the key concepts and algorithms that have been developed for DRL in the context of robotic manipulation. This will include a review of techniques for reward engineering, such as imitation learning and curriculum learning, as well as approaches to hierarchical reinforcement learning. We will also discuss the various network architectures that have been used in DRL for robotic manipulation and the challenges associated with transferring learned policies from simulation to the real world. Finally, we will review both value-based and policy-based DRL algorithms and their relative strengths and limitations for robotic manipulation tasks. The contributions of the paper are: A tutorial of the current RL algorithms and reward engineering methods used in robotic manipulation. An analysis of the current status and application of RL in robotic manipulation in the past seven years. An overview of the current main trends and directions in the use of RL in robotic manipulation tasks. The rest of the paper is organized as follows. The methodology used to find and choose relevant publications is described in Section 2 . In Section 3 , we introduce the key RL concepts and state-of-the-art algorithms. Next, Section 4 continues by describing the learning methods for DRL. In Section 5 , we discuss the current neural network architectures in RL. In Section 6 , we take a deep dive into the applications and implementations of robotic manipulation. Then, we describe the existing challenges and future directions with respect to previous work. The final paragraph of Section 7 provides a summary of the knowledge obtained."
} | 2,659 |
35540661 | PMC9076062 | pmc | 5,958 | {
"abstract": "The oxidation of a carbon anode has been reported to enhance electricity recovery in a microbial fuel cell (MFC). This study investigates the applicability of electrochemically oxidized graphite felt (EOGF) as the anode for the recovery of electricity from sewage wastewater when polarized at 0.2 V during MFC operation. EOGFs were prepared by polarizing graphite felt (GF) at 2 V in 1% sulfuric acid or nitric acid. The nitric acid-treated EOGF inoculated with an sewage sludge produced a maximum current of 110 μA cm −3 , which exceeds that produced by the original GF (91 μA cm −3 ) under electrochemical cultivation at 0.2 V vs. Ag/AgCl. This outcome is attributed to a decrease in charge-transfer resistance and an increase in the capacitance of the anode. In contrast, electrochemical oxidation did not affect the chemical oxygen demand (COD) removal rate or the microbial community structure of the anode. The MFC equipped with the EOGF delivered 340–560 mW m −3 -MFC of electricity during operation in the drainage water channel of a primary sedimentation tank, which corresponds to 11–15 μA cm −3 of current density. The lower current produced in the MFC compared to that observed during electrochemical cultivation indicates that factors other than the anode material restrict current production in the MFC. Even with the small amount of generated electricity, when operated for more than three days, the MFC provides a positive net energy balance when integrated with post-aeration treatment.",
"conclusion": "4. Conclusions This study evaluated the effect of the electrochemical oxidation of a GF anode on electricity recovery from sewage wastewater under two conditions: polarization using a potentiostat, and with the MFC running. The electrochemical oxidation of GF was found to enhance current production at a stable potential of +0.2 V vs. Ag/AgCl, although the current in the MFC was less than one-tenth of this current. To increase electricity recovery by the MFC, the entire system requires improvement rather than the anode material alone.",
"introduction": "1. Introduction Microbial fuel cells (MFCs) have received much attention for applications in sewage wastewater treatment systems. 1–3 Characteristically, an MFC has a single chamber containing an anode in wastewater and an oxygen-reducing cathode that is exposed to the atmosphere. 3–5 In general, microbes in the anodic chamber oxidize organic matter and negatively charge the anode through extracellular electron transfer. 6 The anode in an MFC functions as a collector of the electrons transferred from microbial cells and also as a carrier that holds the microbial cells. Hence, the anode requires good microbe affinity and a large specific surface area or 3D-structure that facilitates the adhesion of a large number of microbial cells. The surface chemistry of the anode is important for the formation of a microbial biofilm on the electrode, which is triggered by the adhesion of microbial cells through hydrogen bonding, electrostatic, and van der Waals forces. 7 Carbon-based anodes are becoming increasingly popular and are advantageous in terms of their commercial availability in various forms, such as felt, brush, cloth, and granules, 8,9 although non-carbon anodes are still being optimized for practical applications. 10 Focusing on the affinity of the electrode toward microbes, a charged and hydrophilic carbon-electrode surface enhances the adhesion of electrochemically active microbes, such as Geobacter species, which can be achieved by chemical oxidation. 7 Graphene oxide (GO), the oxidized form of graphene and the ultimate unit of single carbon sheets exfoliated by the chemical oxidation of graphite, has been demonstrated to exhibit considerably higher and more-stable energy production than graphite. 11–16 This superior performance is attributable to the selective growth of electrochemically active bacteria, 13 better biofilm growth, greater capacity, and a much smaller charge-transfer resistance. 16 However, it requires several weeks for the microbial reduction of GO and its subsequent use as an anode; hence, the preparation process needs to be improved for practical applications. A more practical procedure involves oxidizing the surface of a carbon electrode by heating, 17 acid soaking, 17 or electrochemical means. 18–22 In this study, we prepared and evaluated the performance of electrochemically oxidized graphite felt (EOGF), an alternative 3D-carbon anode prepared from GO. Firstly, the prepared EOGFs were evaluated as anodes for the recovery of electricity from sewage wastewater under polarization using a potentiostat, after which they were polarized in an operating MFC. The EOGFs were first evaluated during current recovery from sewage wastewater at a constant voltage by comparing them to the original graphite felt (GF) devoid of treatment. The GF and EOGF acclimated with sewage sludge were analyzed by cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS). Furthermore, electricity production was evaluated in an MFC floating in the drainage channel of a primary sedimentation tank of a sewage wastewater treatment plant. The MFC was also run as a batch reactor to evaluate organic-matter reduction.",
"discussion": "3 Results and discussion 3.1 Electrochemical oxidation of graphite felt \n Fig. 2A shows current density as a function of time during the electrochemical oxidation of GF at +2.0 V vs. Ag/AgCl in 1% nitric acid; approximately 30–35 mA cm −3 of current density was produced in 1% nitric acid, while almost no current was produced in water (data not shown). Polarization for 30 min produced about 59 000 C cm −3 of accumulated charge density. The GF was visually different following electrochemical oxidation, as shown in Fig. 2B and C . The texture of the EOGF is rough and it appears to have lost its surface gloss. However, the SEM images have no apparent changes in the surface morphologies of the GF and EOGFs (Fig. S1 † ). Fig. 2 Electrochemical oxidation of GF in 1% nitric acid. (A) Current produced during oxidation. XPS spectra of the GF (B) before and (C) after electrical oxidation in 1% nitric acid. The insets show photographic images of the GF surfaces. The XPS spectrum of the EOGF is also different to that of the original GF ( Fig. 2B and C ). The C 1s spectrum of the original GF exhibits a prominent CC or CH peak (285 eV), while EOGF clearly shows a peak ascribable to CCO or OCO (288 eV). The wide-range scan also shows an increase in the intensity of the O 1s peak following electrochemical cultivation (Fig. S2 and S3 † ). The electrochemical oxidation of GF in sulfuric acid showed similar current behavior (Fig. S4 † ), while the XPS data 3 reveal that less carbon was oxidized in 1% sulfuric acid (Fig. S5 † ). The GF immersed for 30 min in sulfuric acid or nitric acid in the absence of polarization was also examined by XPS and showed only single CC or C–H peaks (data not shown). The GF was also polarized at 2.0 V vs. Ag/AgCl in water, although the current was very low and no change in the XPS spectrum was observed (data not shown). 3.2 Electrochemical cultivation using EOGF The original GF and the two EOGFs were polarized at +0.2 V vs. Ag/AgCl in sewage wastewater to evaluate the effect of the anode on electricity production. Regardless of the GF type, all three anodes immediately produced current after 10 d of pre-incubation with sewage sludge ( Fig. 3 ). A current density of 60–150 μA cm −3 was observed within 5 d, but gradually decreased until the sewage wastewater was refreshed. At that time, all of anodes had abundant biomass on the surface (Fig. S1 † ). The current production immediately recovered when fresh wastewater was introduced into each bottle, which suggests that the decline in current is caused by a shortage of organic compounds. This decrease/recovery trend in current production was observed repeatedly for the GF as well as the EOGFs. On the whole, the EOGFs produced higher currents than the original GF over 45 d with three wastewater replacements, and the currents produced by the two EOGFs were equivalent. Fig. 4A shows peak current densities observed during the four cultivation terms according to wastewater replacement. The EOGFs produced peak current densities that were higher than those produced by the GF, while no significant differences between the two EOGFs were observed. These results indicate that the acid used during electrochemical oxidation does not affect the production of current. The COD removal rates were calculated from the initial and final CODs determined immediately after and before wastewater replacement. CODs in the 1.0–2.0 mg d −1 cm −3 range were determined for all bottles, and no significant differences were observed among the cultures incubated with the three types of anode ( Fig. 4B ). The coulombic efficiencies in the EOGF cultures were in the 14–38% range and are higher than those of the untreated GF (9–35%) ( Fig. 4C ). These differences appear to reflect the higher current produced in the EOGF cultures. Fig. 3 Current recovery from sewage wastewater using the original GF and EOGFs polarized using a potentiostat. The symbols indicate the average and the bars are standard deviations from triplicate experiments performed using three bottles in parallel. Fig. 4 Performance of the GF and EOGFs when polarized with a potentiostat. (A) Peak current, (B) COD removal, and (C) coulombic efficiency observed during electrochemical cultivation. The data shown are averages of three independent experiments performed in parallel. The error bars are standard deviations ( n = 3).* indicates a significant difference ( p < 0.05). Compared to graphene oxide, the enhancement in electricity production was limited. Specifically, the maximum peak current density obtained using the EOGF was about 150 μA cm −3 , while the complex formed between microbially reduced GO and microbes (the rGO complex) was reported previously to deliver a maximum peak current density of 310 μA cm −3 . 13 The current density produced using the original GF is similar to that obtained previously (89 μA cm −3 on average) and in this study (73 μA cm −3 on average), even when real sewage wastewater sampled in different years was used. These results suggest that electrochemical oxidation is unable to provide an anode material as capable of enhancing electricity production as the rGO complex. Anode modification by electrochemical oxidation for practical applications also requires large amounts of acid and produces acidic wastewater during the washing process, which is another obstacle to scale-up. 25 In addition, the washing process can potentially peel the oxidized carbon from the anode surface. Exposure of the anode surface to plasma is an alternative approach, but this method requires a large vacuum chamber. Atmospheric plasma or burning of the anode-surface can be more practically applied to MFCs used in wastewater treatment. 3.3 Cyclic voltammetry and electrochemical impedance spectroscopy of EOGF Fig. S6 † shows the cyclic voltammograms acquired in cultures using the three anodes after polarization for 40 d or more. Catalytic current was produced from −400 mV vs. Ag/AgCl and increased with increasing voltage in all cultures. The original GF exhibited 90–100 μA cm −3 at 200 mV vs. Ag/AgCl, while the two EOGFs produced higher current densities of 170–220 μA cm −3 . Comparing the cyclic voltammograms reveals that the EOGFs exhibit larger closed areas than the untreated GF, which indicates that the EOGFs retain more charge in the anode-sludge complex. The charge-transfer resistances ( R ct ) of the two EOGFs were determined to be <10 Ω cm −3 (Fig. S7 † ) from the diameters of the semicircles in their respective Nyquist plots. In contrast, the R ct of the untreated GF was >50 Ω cm −3 . In ideal electrochemical kinetics reactions, R ct and the angular frequency ( ω max ) at the top of semicircle are inversely proportional to capacitance ( C ), as expressed by the relationship: ω max CR ct = 1. According to Fig. S7 † and this formula, the capacitance of the EOGF must be much higher than that of the untreated GF. 3.4 Analyzing the enriched microbial communities on the anodes The microbial communities were analyzed using the 16S rRNA gene sequence, which revealed the predominance of Geobacter species, which are well-known bacteria capable of transferring electrons from cells to the electrode in all three anodes (Table S1 † ). The relative populations of Geobacter -OTUs were 23–31% on average. In addition to these Geobacter species, phylotypes belong to the Desulfovivrio and Desulfomicrobium genera, the Bacteroidales order, and the Holophagaceae family were also commonly observed in the 1.2–5.1% range. The comparative ratios of these OTUs do not significantly differ among the three types of anode. Additionally, the three microbial communities are similar to those observed for the GF-sludge complex polarized at 0.2 V in a past result, rather than those of the GO-sludge complex. 13 These results suggest that the electrochemically active microbial communities in the anode with sludge and sewage wastewater are highly stable and have no impact on the electrochemical oxidation of the anode surface. 3.5 Electricity recovery by the EOGF in the three MFCs GF sheets were next electrochemically oxidized in 1% nitric acid for use as anodes in the MFC (Fig. S8A † ). The current density produced on the GF-sheets was much lower than previously observed, which is ascribable to the larger anode size. Accordingly, the time required for electrochemical oxidation was extended to 17 h, which resulted in an accumulated charge density of approximately 60% of that produced by a small GF block. While XPS revealed the oxidation of carbon in the GF sheets, the CCO or OCO peak was less intense than that observed for the GF block (Fig. S8B † ). The EOGF sheets were further used as MFC anodes. \n Fig. 5 shows the electricity produced in the three MFC units equipped with the EOGF sheets; i.e. , the top MFC, middle MFC, and bottom MFC positioned at 0–13 cm, 13–26 cm, and 26–39 cm below the surface of the water, respectively. Electricity production increased dramatically and stabilized within 3 d from floating in the water channel of wastewater. Over 33 d of polarization, the top-, middle-, and bottom MFCs produced average power densities of 440, 560, and 340 mW m −3 -MFC volume, respectively, which correspond to 54, 69, and 42 mW m −2 -cathode area, respectively. The highest amount of electricity was produced by the middle MFC, which suggests that water depth affects electricity production. Specifically, oxygen dissolved from the water surface, as a competitive electron acceptor, can inhibit electricity recovery to the anode in the top-MFC. In contrast, there is possibly insufficient oxygen in the cathodic chamber of the bottom MFC. The large declines in electricity produced over days 4–5, 13–14, and 29–30 are probably associated with drops in the water level, which exposed the MFCs to the atmosphere. The electricity produced in the three MFCs exhibited small fluctuations (<20%) in amplitude during their daily cycles, which suggest that these fluctuations are the result of daily changes in organic content and the temperature of the sewage wastewater in the channel. Those trends, the highest electricity in the middle MFC and the dairy fluctuation were also observed in the previous study. 26 Fig. 5 Electricity recovery using the EOGF in a tuber MFC floating in a sewage wastewater channel. The current density produced in the MFC was about 13 μA cm −3 -anode volume on average, and much less than that observed in the EOGF polarized using a potentiostat. This suggests that factors other than the anode material restricted electricity production in the MFC. Possible rate-limiting factors are the separator, cathode, and accessibility of the substrate. The MFC used in this study produced 450 ± 110 mW m −3 -MFC (which corresponds to 54 ± 14 mW m −2 -cathode area) in the water channel. The value is somewhat lower than the electricity produced (82–170 mW m −2 -cathode area) in other MFCs that treat sewage wastewater containing 118–3300 mg L −1 of COD. 26 3.6 Batch evaluation of COD removal rate by a tuber MFC COD removal by the MFC-unit was evaluated using a batch reactor filled with sewage wastewater, the results of which are shown in Fig. 6 . Through 13 days of incubation, the top-MFC, middle-MFC and the bottom-MFC produced 73, 330, and 91 mW m −3 of electricity on average, respectively ( Fig. 6A ). Electricity generation by the MFC in the batch reactor was stable over 13 days, but the currents were less than that observed in the water channel. This is possibly the result of limited substrate availability in the anode biofilm due to interruptions in the continuous flow surrounding the anode. The change in COD concentration in the batch culture revealed a 73% reduction in COD after 13 days of incubation, and fitted a first-order rate formula well ( R 2 = 0.94) ( Fig. 6B ). Based on the daily COD levels ( Fig. 6C ) and electricity production, coulombic efficiency and electricity generation efficiency (EGE) were calculated to be 13% and 2.0 kW h kg-COD −1 on average ( Fig. 6D and E ). The fluctuations observed for coulombic efficiency and EGE reflect COD removal rather than electricity production. Fig. 6 Performance of the MFC in a batch culture over time. (A) Electricity production, (B) COD concentration in the sewage wastewater of the reactor, (C) COD removal, (D) coulombic efficiency, and (E) electricity-generation efficiency (EGE). Energy balance was calculated for wastewater treatment by combining batch MFC treatment with post-aeration for different MFC-operating times (0–15 d) ( Fig. 7 ). The values of energy recovery and COD removal by the MFC refer to the batch experimental data displayed in Fig. 6 . Energy consumption for the removal of COD by aeration was calculated to be −0.6 kW h kg-COD −1 . 27,28 The total energy balance calculation indicates that more than three days of MFC operation are required for the net energy balance to become positive when combined with post-aeration. Fig. 7 Calculating the net energy balance during MFC wastewater treatment and post-aeration for various MFC operating times. (A) Energy balances and (B) COD concentrations at the end of the MFC and post-aeration periods."
} | 4,621 |
28789484 | PMC5545993 | pmc | 5,959 | {
"abstract": "Anaerobic fungi reside in the gut of herbivore and synergize with associated methanogenic archaea to decompose ingested plant biomass. Despite their potential for use in bioconversion industry, only a few natural fungus–methanogen co-cultures have been isolated and characterized. In this study we identified three co-cultures of Piromyces with Methanobrevibacter ruminantium from the rumen of yaks grazing on the Qinghai Tibetan Plateau. The representative co-culture, namely ( Piromyces + M. ruminantium ) Yak-G18, showed remarkable polysaccharide hydrolase production, especially xylanase. Consequently, it was able to degrade various lignocellulose substrates with a biodegrading capability superior to most previously identified fungus or fungus–methanogen co-culture isolates. End-product profiling analysis validated the beneficial metabolic impact of associated methanogen on fungus as revealed by high-yield production of methane and acetate and sustained growth on lignocellulose. Together, our data demonstrated a great potential of ( Piromyces + M. ruminantium ) Yak-G18 co-culture for use in industrial bioconversion of lignocellulosic biomass.",
"introduction": "Introduction With energy consumption continuing to rise and fossil fuels inevitably trending toward limitation, humanity is urged to find alternative energy resources. One most recent emerging focus of energy generation has been the use of biofuels which can be generated from sustainable biomass feedstocks (Tilman et al. 2009 ). On the top of the list of renewable biomass resources that are suitable for biofuel production is crop straw, which is ranked as the fourth largest energy resources after coal, oil and natural gas. The major structural component of crop straw is lignocellulose, a heterogeneous complex mainly consisting of two carbohydrate polymers (cellulose, hemicellulose), and an aromatic polymer (lignin) (Bayer et al. 2004 ). Currently, the bioconversion strategy is regarded as the most common approach for industrial utilization of lignocellulosic biomass, which explores natural microbial colonizers of lignocellulose or their lignocellulose-degrading enzymes to decompose the recalcitrant structural polymers to easily metabolizable monosaccharides which are subsequently converted to products (Balan 2014 ). Improving the bioconversion efficiency of lignocellulosic biomass has received increased attention from researchers in recent years. Anaerobic fungi isolated from rumen of ruminants are well known for their plant biomass-degrading capabilities (Theodorou et al. 1996 ). Despite only accounting for a very small percentage (<8%) of the rumen microbial community, anaerobic fungi are largely responsible for degrading lignin tissue (Gruninger et al. 2014 ). Their strong lignocellulose degradation capacity stems from both physical disruption by penetrating their rhizoids into the plant cell wall and enzymatic digestion by producing a wide range of fibrolytic enzymes to effectively break down lignocelluloses to monosaccharides (Boxma et al. 2004 ). Subsequently, the resulting monosaccharides are taken up by the fungi and utilized via two alternative metabolic pathways, the cytoplasmic pathway to produce formate, ethanol, lactate and succinate as end products, or the higher energy-yielding pathway in hydrogenosome to produce formate, acetate, CO 2 , H 2 and ATP as end products (Muller 1993 ). The biodegradation capability of anaerobic fungi might be regulated by other members of rumen microbial community, as exampled by methanogens. Syntrophic relationship could form between anaerobic fungi and methanogens: methanogens can feed on some of the fungal metabolites including H 2 , CO 2 , formate to produce methane (CH 4 ) (Bauchop and Mountfort 1981 ). Consequently, anaerobic fungi are often found to be physically associated with methanogens when isolated from herbivore rumen (Bauchop and Mountfort 1981 ; Jin et al. 2011 ; Leis et al. 2014 ). It was also demonstrated that artificial consortiums of rumen fungi and methanogens generally degrade fiber more effectively than the fungus mono-cultures (Bauchop and Mountfort 1981 ). Such enhanced biodegradation activity could be attributed to the change of fungal metabolism in the presence of methanogen: hydrogenosome metabolism is favored over cytoplasmic metabolism, and consequently more energy is produced for fungi to grow, resulting in accelerated lignocellulose fermentation kinetics (Marvin-Sikkema et al. 1993 ; Williams et al. 1994 ) Thus, the rumen-derived natural anaerobic fungus–methanogen consortiums hold substantial promise for application in industrial bioconversion of lignocellulosic biomass. The domestic yak ( Bos grunniens ) is a large ruminant of the bovine family grazed mainly on the Qinghai Tibetan Plateau at an elevation higher than 3000 m. Given the facts that the yaks feed on wild grass instead of grain in order to thrive in the harsh environment and ruminal microbial community is critical for plant biomass digestion, we conceived the hypothesis that the natural fungus–methanogen co-cultures derived from the rumen of grazing yaks should evolve to acquire superior lignocellulose degradation ability as the result of long-term natural selection during the adaptation of yaks to the Qinghai-Tibetan Plateau. Our most recent study provided the first line of evidence supporting this hypothesis by isolating a total of 20 fungus–methanogen co-cultures from the rumen of the grazing yaks (Wei et al. 2016b ). Among the three fungal genera ( Orpinomyces , Neocallimastix and Piromyces ) identified in these natural co-cultures, Neocallimastix and Piromyces genera are of particular interest as their monoculture isolates were previously demonstrated to produce high levels of fibrolytic enzymes (Paul et al. 2010 ). In the present study, we further investigated three Piromyces with M. ruminantium co-cultures isolated from grazing yaks. Moreover, we focused on studying the representative ( Piromyces + M. ruminantium ) Yak-G18 co-culture and provided a comprehensive, quantitative characterization of its degradation capability of various lignocellulosic materials and the associated end-products profiles.",
"discussion": "Discussion Anaerobic fungi reside in the rumen of herbivore and synergize with associated methanogenic archaea to decompose ingested plant biomass. Given the relative short retention time of lignin tissues in the rumen, researchers have long speculated that rumen fungus–methanogen consortiums have acquired potent lignocellulose degradation capability attributable to natural selection. Thus, natural fungus–methanogen co-cultures may have the potential to be directly used in industrial bioconversion of renewable lignocellulosic biomass. Of equal importance, studies on these co-cultures may provide new insights into mechanistic aspects of fungal lignocellulolytic machinery, therefore facilitating the development of lignocellulolytic enzyme mixtures or complexes with improved biodegradation efficiency. Despite their potential importance in bioconversion industry, natural fungus–methanogen co-cultures are still poorly understood, largely attributable to the fact that only a few of them were isolated. A total of only ten natural fungus–methanogen co-cultures had been previously isolated from mule, camel, buffalo, goat, sheep and Alpine ibex (Bauchop and Mountfort 1981 ; Jin et al. 2011 ; Leis et al. 2014 ). Our search for highly active natural fungus–methanogen co-cultures has been focusing on yaks grazed on the Qinghai-Tibetan Plateau at 3000 m elevation and above. The underlying rationale can be ascribed to the supposition that the adaptation of yak to extreme environment drives the evolution of its rumen microbial community to be superior in decomposing recalcitrant lignocellulosic polymers. In this study, we investigated three Piromyces with methanogen co-cultures isolated from the rumen liquid of grazing yaks. Like most previously reported natural fungus–methanogen co-cultures from the rumen or feces of other herbivores (Bauchop and Mountfort 1981 ; Jin et al. 2011 ), all the three co-culture isolates showed the one fungus–one methanogen pattern. All the methanogens associated with the three isolates were identified to be Methanobrevibacter ruminantium , consistent with previous finding that Methanobrevibacter ruminantium is one of the most abundant H 2 - and CO 2 -consuming rumen methanogen species (Carberry et al. 2014 ; Janssen and Kirs 2008 ). Electron microscopy analysis of the representative co-culture isolate, ( Piromyces + M. ruminantium ) Yak-G18 co-culture, unveiled physical interactions between the methanogen and rhizoids but not the sporangia surface of the fungus, reminiscent of the majority of previously reported natural fungus–methanogen co-cultures (Bauchop and Mountfort 1981 ; Jin et al. 2011 ). Such close proximity between fungus and methanogen might facilitate effective interspecies hydrogen transfer (Leschine 1995 ; Thareja et al. 2006 ). Our comprehensive, quantitative analysis of the fermentation activities of ( Piromyces + M. ruminantium ) Yak-G18 co-culture on various substrates support our original hypothesis that Yak-derived ruminal fungus–methanogen consortiums have evolved to possess high efficiency to degrade plant lignocellulose. Given that the IVDMD values are often fluctuated at the late time points of the fermentation due to the outgrowth of the fungi, we also employed ADFD and NDFD values for accurately measuring the biodegrading activities. ( Piromyces + M. ruminantium ) Yak-G18 co-culture showed higher fiber digestion capability than most, if not all, of the previously reported anaerobic fungus isolates from other herbivores (Jin et al. 2011 ; Paul et al. 2010 ; Thareja et al. 2006 ). Specifically, the wheat straw IVDMD values of ( Piromyces + M. ruminantium ) Yak-G18 co-culture are significantly higher than those of many natural anaerobic fungus cultures isolated from various herbivores (Paul et al. 2010 ; Thareja et al. 2006 ). The 4-day wheat straw IVDMD and rice straw IVDMD values of ( Piromyces + M. ruminantium ) Yak-G18 co-culture are also slightly higher than those of the two previously reported natural fungus–methanogen co-cultures also with Piromyces as the fungal constituent (Jin et al. 2011 ). Consistent with its high IDVMD values, ( Piromyces + M. ruminantium ) Yak-G18 co-culture also exhibits generally high ADFD and NDFD values of all five substrates tested, further confirming its great lignocellulose degradation capability. Strong fiber degradation capability is normally associated with the production of highly active fibrolytic enzymes. As expected, ( Piromyces + M. ruminantium ) Yak-G18 co-culture showed high yield of a number of fibrolytic enzyme activities, especially xylanase activity, during the incubation with all the five lignocellulosic substrates assayed. Notably, it showed significantly higher xylanase activity (2500–5023 mU maximal value) than numerous fungus isolates of various herbivore origins (Cao and Yang 2011 ; Sirohi et al. 2013 ; Thareja et al. 2006 ; Tripathi et al. 2007 ). In comparison to the xylanase activity, the FPase and AE activities of the ( Piromyces + M. ruminantium ) Yak-G18 co-culture are relatively lower while the feruloyl esterase activities were the least significant, irrespective of the substrates. This result is corroborated by a most recent study that a gut Piromyces isolate exhibits approximately a fourfold increase of lignocellulose-induced xylanase activity relative to commercial Aspergillus enzyme preparations (Solomon et al. 2016 ). Thus, ( Piromyces + M. ruminantium ) Yak-G18 co-culture can be classified as a new practical agent for industrial xylanase production. There is accumulated evidence that the production of fibrolytic enzymes by anaerobic fungus is highly affected by growth conditions (Glass 2016 ; Solomon et al. 2016 ). Consistent with this notion, we found that different lignocellulosic materials induced different levels of the fibrolytic enzymes produced by ( Piromyces + M. ruminantium ) Yak-G18 co-culture. Among the five assayed substrates, medicago sativa which possesses the least NDF (Table 1 ) was identified with the highest IVDMD value. In contrast, compared with other substrates, medicago sativa showed least efficacy in inducing xylanase as well as AE activities throughout the incubation period. The induction of FPase by medicago sativa was also lower than other substrates except for Chinese wildrye at late incubation phase. The actual relationship between the fibrolytic enzyme production of ( Piromyces + M. ruminantium ) Yak-G18 co-culture and the fiber composition of fermented material requires future investigations. The advantage of anaerobic fungus–methanogen co-culture over fungus culture in degrading lignocellulose stems from the ability of methanogen to utilize H 2 and CO 2 for methane production, thus eliminating the inhibition of these end-products on the fungal growth and consequently the associated enzyme activities (Marvin-Sikkema et al. 1993 ; Williams et al. 1994 ). Indeed, for all the assayed substrates, the ( Piromyces + M. ruminantium ) Yak-G18 co-culture displayed a continuous accumulation of CH 4 , the levels of which were significantly higher than those of the natural fungus–methanogen co-cultures reported by Jin et al. ( 2011 ). One well-known metabolic impact of methanogen association on anaerobic fungus is the enhanced hydrogenosome pathway leading to more production of acetate and ATP (Williams et al. 1994 ). Consistent with this notion, ( Piromyces + M. ruminantium ) Yak-G18 co-culture produced high levels of acetate on all five assayed substrates with maximal value of 41.3 mM with Chinese wildrye on day 7. With regard to the acetate yield, a direct comparison between ( Piromyces + M. ruminantium ) Yak-G18 co-culture and natural fungus–methanogen co-cultures reported by Jin et al. ( 2011 ) is not valid because of different substrates being used for the determinations. End-products profiling also revealed a unique metabolism feature of the ( Piromyces + M. ruminantium ) Yak-G18 co-culture as the high-yield production of lactate. Significant lactate production has only been associated with some natural fungus–methanogen co-cultures isolated so far (Jin et al. 2011 ). Compared with these lactate-producing co-cultures previously reported (Jin et al. 2011 ), ( Piromyces + M. ruminantium ) Yak-G18 co-culture yields approximately four to sevenfold more lactate with l -lactate being the predominant isoform. It is worth mentioning that, since lactate cannot be metabolized by methanogen, the association with methanogen could impact the fungal lactate production either positively or negatively. It is as yet unclear why ( Piromyces + M. ruminantium ) Yak-G18 co-culture produces much more l -lactate than d -lactate and the underlying mechanism awaits further investigation. Finally, the identification of ( Piromyces + M. ruminantium ) Yak-G18 co-culture extends the repertoire of Yak-derived ruminal fungus–methanogen co-cultures applicable to the biodegradation industry. Previously, we isolated and identified two other types of fungus–methanogen co-cultures from yaks grazing on the Qinghai-Tibetan Plateau, namely ( Orpinomyces joyonii + M. ruminantium ) type, and the ( Neocallimastix frontalis + M. ruminantium ) type (Wei et al. 2016b ). Among the three types of co-cultures, based on the data presented here and our previous studies (Wei et al. 2016a , b ), we concluded that ( Piromyces + M. ruminantium ) Yak-G18 co-culture and the representative member of ( N. frontalis + M. ruminantium ) type, ( N. frontalis + M. ruminantium ) Yaktz1 co-culture, have an advantage over the nine members of ( Orpinomyces joyonii + M. ruminantium ) type in degrading lignocellulosic materials, as indicated by higher IVDMD values along with higher fibrolytic enzyme activities. Besides, the two co-cultures also exhibits higher yield of CH 4 and acetate, demonstrating a stronger metabolic fungus–methanogen syntrophy that is manifested by the fact that ( Piromyces + M. ruminantium ) Yak-G18 co-culture grew very stably in anaerobic medium with wheat straw and showed constant production of end-products even after 2 years’ continuous sub-culturing. It should be emphasized here that Piromyces and Neocallimastix have been shown to have growth advantages on plant lignocellulose over other anaerobic fungal genera (Paul et al. 2010 ). Most recently, Solomon et al. ( 2016 ) characterized three gut anaerobic fungi including Piromyces and Neocallimastix by transcriptome and proteome approaches. The analysis revealed that these gut fungi, particularly Piromyces , possess a wide array of cellulolytic carbohydrate-active enzymes (CAZy) with diverse functions in deconstructing cellulose and hemicellulose. The expansion of CAZy gene family, especially the xylan-degrading enzymes, provides a plausible explanation why Piromyces secretion showed limited substrate preference as compared to the commercial mixtures produced by Trichoderma and Aspergillus . This elegant study further endorses the candidacy of ( Piromyces + M. ruminantium ) Yak-G18 co-culture as a new model platform for developing more efficient fibrolytic enzyme formula in lignocelluloses degradation industry. In conclusion, three Piromyces with M. ruminantium co-cultures were obtained from the rumen contents of the yaks grazing on the Qinghai-Tibetan Plateau. The representative co-culture, ( Piromyces + M. ruminantium ) Yak-G18 co-culture, displayed robust lignocellulose degradation capability on various substrates by exhibiting high IVDMD, NDFD, and ADFD, along with the production of high levels of xylanase and other fibrolytic enzyme activities. The end-products profiling of ( Piromyces + M. ruminantium ) Yak-G18 co-culture validated the fungus–methanogen syntrophy, as is demonstrated by the high-yield production of methane and acetate and sustained growth on lignocellulose. These results, corroborated by recent advances in transcriptome and proteome analysis of anaerobic gut fungi, strongly support that ( Piromyces + M. ruminantium ) Yak-G18 co-culture and its future exploration might be of great use to lignocellulose biotechnology."
} | 4,615 |
33865676 | null | s2 | 5,962 | {
"abstract": "Interactions between microorganisms in multispecies communities are thought to have substantial consequences for the community. Identifying the molecules and genetic pathways that contribute to such interplay is thus crucial to understand as well as modulate community dynamics. Here I focus on recent studies that utilize experimental systems biology techniques to study these phenomena in simplified model microbial communities. These unbiased biochemical and genomic approaches have identified novel interactions and described the underlying genetic and molecular mechanisms. I discuss the insights provided by these studies, describe innovative strategies used to investigate less tractable organisms and environments, and highlight the utility of integrating these and more targeted methods to comprehensively characterize interactions between species in microbial communities."
} | 220 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.