Characterization of Hydrogen Metabolism in the Multicellular Green Alga Volvox carteri

Hydrogen gas functions as a key component in the metabolism of a wide variety of microorganisms, often acting as either a fermentative end-product or an energy source. The number of organisms reported to utilize hydrogen continues to grow, contributing to and expanding our knowledge of biological hydrogen processes. Here we demonstrate that Volvox carteri f. nagariensis, a multicellular green alga with differentiated cells, evolves H2 both when supplied with an abiotic electron donor and under physiological conditions. The genome of Volvox carteri contains two genes encoding putative [FeFe]-hydrogenases (HYDA1 and HYDA2), and the transcripts for these genes accumulate under anaerobic conditions. The HYDA1 and HYDA2 gene products were cloned, expressed, and purified, and both are functional [FeFe]-hydrogenases. Additionally, within the genome the HYDA1 and HYDA2 genes cluster with two putative genes which encode hydrogenase maturation proteins. This gene cluster resembles operon-like structures found within bacterial genomes and may provide further insight into evolutionary relationships between bacterial and algal [FeFe]-hydrogenase genes.


Introduction
Hydrogen is an essential component in the metabolism of a variety of microorganisms [1,2,3]. In biology, the production of H 2 is predominantly catalyzed by two classes of enzymes, hydrogenases and nitrogenases [1,4], with hydrogenases also contributing to H 2 uptake [2]. Microbes are able to use these enzymes to catalyze the oxidation of H 2 and/or the reduction of protons during fermentation [2,5,6,7]. In addition, recent reports suggest that H 2 may also play a role in deacidification, enhancing the cell viability of certain microbes in harsh environments [8,9,10]. The complex role of H 2 in the life-cycle of autotrophic microorganisms, such as green algae, is linked to both photosynthetic and fermentative processes [11,12], and studies of H 2 were incubated at 22°C in standard Volvox medium (SVM), a minimal medium modified from Kirk et al. [37] and Yamano et al. [38] and primarily composed of metal salts [0. 5  For physiological H 2 production assays, V. carteri cultures were incubated with 20 mM sodium bicarbonate for 72 hours prior to anaerobic acclimation to increase CO 2 levels for improved carbon fixation.

Anaerobic Acclimation
To acclimate V. carteri cells to anaerobiosis, a culture of light-adapted cells was centrifuged at 1,000× g for 5 min. The resulting cell pellet was washed three times in an anaerobic chamber (Coy Laboratory Products) with anaerobic media (degassed SVM supplemented with 0.16 mM MgCl 2 rather than 0.16 mM MgSO 4 to exclude sulfate from the medium). Cell shading is often sufficient to significantly reduce oxygenic photosynthesis in thick cultures of unicellular green alga. However, due to non-homogenous mixing of V. carteri cultures and the low densities of individual cells within spheroids, the vials were wrapped in aluminum foil to establish dark conditions. To prevent the possibility of residual oxygenic photosynthetic activity from stray light, sulfate-depleted medium was also employed to decrease photosystem II activity, a strategy successfully utilized previously in experiments with C. reinhardtii [39,40,41]. The cells were resuspended in degassed media to a final concentration of 75 μg chlorophyll/mL, and 2 mL of this mixture were sealed in a 10 mL air-tight serum vial (Wheaton). The resuspension was shaken for 4 h before further experimentation.

Hydrogen Production Measurements
H 2 evolution using an artificial electron donor was measured by incubating 0.1 mL of either aerobically-or anaerobically-acclimated cells (75 μg chlorophyll/mL) with 1.9 mL of H 2 evolution assay solution [degassed media, 100 mM sodium dithionite, 10 mM methyl viologen dichloride (MV)] in a sealed 10 mL serum vial at 22°C in the dark with continuous shaking. At various time points, 20 μL of headspace gas were injected into a TRACE GC Ultra Gas Chromatograph (Thermo Scientific) using a 100 μL syringe. The peak at 1.4 min corresponded to H 2 , and the absolute value of H 2 gas was determined by comparison to a standard curve. The resulting values were plotted versus time to monitor the accumulation of H 2 in the headspace, and the rates were determined per μg of chlorophyll.
Selected previous reports on H 2 production in algae have utilized detergents to rupture cells and allow the hydrogenase proteins access to the abiotic electron donors. Initial experiments demonstrated no significant difference upon addition of detergent (data not shown), suggesting that detergent is not essential under our assay conditions.

Cell Separation
All steps for the cell separation were performed anaerobically using a protocol adapted from Nematollahi et al. [42]. Briefly, an anaerobically-acclimated 1 L culture was centrifuged at 1,800× g in air-tight 250 mL centrifuge bottles (Nalgene) for 10 min. The supernatant was discarded and the pellet was resuspended in 30 mL of anaerobic media. The suspension was passed through a 100 μm filter (Genesee Scientific) and the flow-through discarded. The cells were washed off of the filter in 10 mL of anaerobic media and the cells were liberated from the extracellular matrix using a 50 mL glass Dounce homogenizer (Sigma-Aldrich), moving the pestle up and down 10 times. An additional 20 mL of anaerobic media was added to the suspension and the gonidia were allowed to settle for 10 min. The top 20 mL of the suspension enriched in somatic cells was passed successively through 30 μm and 10 μm meshes (Genesee Scientific and Membrane Solutions, respectively). Following passage through the 10 μm mesh, the flow-through contained pure somatic cells as determined by microscopy.
The settled cells enriched in gonidia were resuspended in 10 mL of anaerobic media and passed through a 100 μm mesh. The flow-through was then passed through a 30 μm mesh, collecting the cells on the filter. The purified gonidia were washed off of the filter and resuspended in 2 mL of anaerobic media. The homogeneity of the gonidia samples was confirmed by microscopy.

RNA Isolation
To isolate total RNA from V. carteri, cell cultures were centrifuged at 10,000× g for 10 min. The resulting pellets were resuspended in 0.5 mL of TRIzol (Invitrogen) and frozen in liquid nitrogen. The frozen TRIzol/cell mass mixture was crushed using a mortar and pestle, thawed, and resuspended with an additional 0.5 mL of TRIzol. To this suspension, 0.2 mL of chloroform was added, and the mixture was vortexed at top speed at room temperature for 15 min. To pellet cell debris, the suspension was centrifuged at 10,000× g for 15 min at 10°C. The aqueous phase was transferred to a 1.7 mL tube (Denville) containing 0.25 mL of 0.8 M sodium citrate and 1.0 M sodium chloride, and the suspension was gently mixed by inversion. Following the addition of 0.25 mL of isopropanol, the mixture was centrifuged at 10,000× g for 25 min at 10°C. The supernatant was discarded and the RNA pellet washed twice with ice-cold 75% ethanol. The pellet was allowed to dry and was then resuspended in 20 μL of diethylpyrocarbonatetreated water.

RT-PCR
Reverse transcription of mRNA was performed using M-MLV Reverse Transcriptase (Invitrogen), and the cDNA was amplified for 30 cycles via PCR using GoTaq Green Master Mix (Promega) and the primer sets listed in Table 1. Genomic and cDNA sequences of HYDA1, HYDA2, HYDEF, and HYDG can be accessed from the DDBJ/EMBL/GenBank database with the accession numbers XM_002948441, XM_002948437, XM_002948568, and XM_002948439, respectively.

Cloning of HYDA1 and HYDA2
The HYDA1 and HYDA2 genes were PCR amplified with PfuTurbo DNA polymerase (Stratagene) from reverse-transcribed cDNA using the primers listed in Table 1. The amplified products were ligated into the SacI/HindIII site of pAC-BAD, a pBAD/D-TOPO (Invitrogen) expression vector (contains a kanamycin-resistant cassette and an L-arabinose inducible promoter) that was modified to remove the N-terminal thioredoxin-tag [43]. To achieve optimal yields of active protein (using a heterologous expression system previously described [43,44]), the constructs were transformed into S. oneidensis MR-1 ΔhydA/ΔhyaB [45] electrocompetent cells as detailed by Ozawa [46] and selected for antibiotic resistance on 50 μg/mL kanamycin sulfate LB plates.

Overexpression, Protein Purification, and Hydrogen Evolution Assay
Cells harboring the pAC-BAD_HYDA1 and pAC-BAD_HYDA2 vectors were induced for gene overexpression and protein synthesis as previously described [43]. Enzymes were purified from S. oneidensis cultures in an anaerobic chamber and assayed for H 2 evolution activity as described by Cornish et al. [43].

In vivo and in vitro H 2 Evolution
V. carteri is closely related to C. reinhardtii, a green alga with well-described hydrogen metabolism [28,29]. Recent sequencing and annotation of the V. carteri genome [31] uncovered two genes with sequence similarity to algal [FeFe]-hydrogenase genes (HYDA1, HYDA2) (Fig 1) as well as two genes predicted to be necessary for [FeFe]-hydrogenase maturation (HYDEF, HYDG).
To assess V. carteri for in vivo and in vitro H 2 production, a culture of cells was acclimated to anaerobiosis for 4 h in degassed media. Aliquots of anaerobically-acclimated cells were transferred to anaerobic assay vials either containing 2 mL of anaerobic media alone or supplemented with an abiotic electron donor (Na 2 S 2 O 4 ) and an electron mediator (MV), and the headspace was tested for H 2 accumulation over time (Fig 2A). H 2 production could be measured within the first 45 min of incubation with MV/Na 2 S 2 O 4 , eventually achieving an overall integrated production rate of 30.8 nmol H 2 /mg chlorophyll/s after 3 h. Cells that were not supplemented with MV/Na 2 S 2 O 4 did not accumulate a significant amount of H 2 in the headspace even 48 h after initiating the assay. When oxygen-exposed cells (i.e. aerobic cells) were assayed under similar conditions, no appreciable H 2 accumulation was observed over the course of the assay (Fig 2A). These data indicate that, under these conditions, V. carteri requires both anaerobic acclimation and sufficient reducing equivalents to evolve H 2 .
In nature, green algae are able to ferment endogenous carbon stores to drive in vivo H 2 production. However, initially we did not observe appreciable H 2 accumulation in V. carteri cells that were anaerobically-acclimated in the absence of Na 2 SO 4 /MV and assayed for H 2 production. We hypothesized that, under our original growth conditions, fixed carbon levels might be too low to support fermentative H 2 production. To allow for sufficient carbon fixation to occur, cultures of V. carteri were supplemented with 20 mM sodium bicarbonate and incubated for 72 hours in the light. The cells were then acclimated to anaerobiosis for four hours and tested for H 2 formation using unsupplemented SVM as previously described. During the course of the assay, the cells produced appreciable amounts of H 2 under anaerobic, dark conditions doi:10.1371/journal.pone.0125324.g002 (Fig 3). These results demonstrate that the green alga V. carteri is competent for H 2 production under physiological conditions.

Gonidia and Somatic Cell H 2 Production
Previous work revealed distinct metabolic differences between the two cell types of V. carteri [42,47]. To investigate the potential different roles in H 2 production between the cell types, gonidia and somatic cells were isolated from an anaerobically-acclimated V. carteri culture and assayed for H 2 accumulation over a three hour incubation period using the MV/Na 2 S 2 O 4 electron donor system. H 2 evolution rates were roughly similar for the two cell types (Fig 2B) nd were~30% of the rate observed from whole cell extracts on a per chlorophyll basis. The observed difference in the rates may be due to cellular and plastidic damage from the separation procedure. Overall, no major differences in H 2 evolution rates were observed between the two cell types when supplied with an abiotic electron donor.

HYDA Transcript Analysis
We predicted that the putative HYDA genes in V. carteri play a role in H 2 production. To test for HYDA expression during anaerobiosis, RNA was collected from both aerobically-and anaerobically-acclimated cells. The RNA was reverse-transcribed and PCR amplified with primers specific to the putative [FeFe]-hydrogenase genes, HYDA1 and HYDA2. Amplification of mRNA for both genes was observed in anaerobic samples, while very little or no transcript could be observed in the aerobic samples when compared to a housekeeping gene, ACTA (accession number XP_002955536) (Fig 4). These results are consistent with the notion that anaerobic expression of HYDA1 and HYDA2 correlates with the H 2 production observed in V. carteri.

Hydrogenase Cloning and Characterization
The two putative HYDA genes have high sequence similarity to known green algal hydrogenase genes (Fig 1), suggesting that they code for active H 2 -producing enzymes. To test this hypothesis, the two genes were amplified from cDNA, cloned into pAC-BAD, and transformed into S. oneidensis ΔhydAΔhyaB cells. Gene expression was induced by the addition of L-arabinose. The N-terminally His-tagged proteins were purified under strict anaerobiosis by nickel-affinity chromatography and the proteins were analyzed by SDS-PAGE. The HYDA1 and HYDA2 proteins were active when supplied with MV/Na 2 S 2 O 4 , and both had similar K m values when measured for MV (2.76 mM and 2.82 mM, respectively). The apparent specific activity for HYDA1 was almost 70-fold greater than for HYDA2 (1,790 nmol H 2 /mg/s and 26.7 nmol H 2 / mg/s, respectively). This differs from previously reported values for recombinant C. reinhardtii HYDA1 and HYDA2 specific activities, (2,500 nmol H 2 /mg/s and 1,900 nmol H 2 /mg/s, respectively) [48], but is slightly more consistent with the greater than 3.5-fold difference in activity for CpI and CpII of Clostridium pasteurianum (2,060 nmol H 2 /mg/s and 7,310 nmol/H 2 /mg/s, respectively) reported by Adams et al. [49].

Gene Cluster Analysis
The two hydrogenase genes and the two assembly factor genes in the V. carteri genome (HYDA1, HYDA2, HYDEF, and HYDG) were found to be in relatively close proximity to one another. The distance between the furthest genes (HYDA1 and HYDA2) is~57 kb, and the main cluster of genes is within a~23 kb span (Fig 5). Although green algae are theorized to have obtained their hydrogenase genes via horizontal gene transfer [24,25,26,27]-likely from an operon or operon-like structure-this level of [FeFe]-hydrogenase gene clustering has not been previously observed in the sequenced genomes of other H 2 -evolving green algae (C. reinhardtii and Chlorella variabilis). [24,25]. Based on these analyses, this clade of hydrogenase genes demonstrated a clear relationship to bacterial genes, strongly indicating an ancestral horizontal gene transfer event. The presence of the hydrogenase gene cluster in V. carteri could be a remnant of horizontal gene transfer. However, no clustering of [FeFe]-hydrogenase-related genes is observed in the bacterial genomes noted in the clades adjacent to the green algal clade, as observed in the Meuser phylogram [25]. Expanding the search into thermophilic anaerobes, we identified several bacterial genomes which contain gene clusters composed of hydrogenase and hydrogenase maturation genes. In particular, the genomes of Fervidobacterium nodosum, Fervidobacterium pennivorans, and Thermotoga thermarum contain operon-like clusters composed of five hydrogenase genes that cover a <22 kb span (Fig 6, Table 2). It is worth nothing that these clusters demonstrate roughly similar patterns in gene distribution and orientation to each other and the V. carteri gene cluster.

Discussion
A variety of green algae have been characterized for their ability to couple energy derived from photosynthesis to the production of H 2 , especially the model organism C. reinhardtii [15,30,50]. Previous work has demonstrated that a species of Volvox is capable of H 2 production [36] to a similar extent as C. reinhardtii, an alga separated V. carteri by~220 million years of evolution. The recently sequenced V. carteri genome identified two putative [FeFe]-hydrogenase genes [31], prompting us to investigate further the H 2 metabolism of this organism.
To test the H 2 metabolism of V. carteri, algal cell cultures were acclimated to anaerobiosis, while the serum vials were wrapped with aluminum foil, thus eliminating physiological oxygen production and allowing residual oxygen to be removed via respiration. The anaerobic cells were then provided with an electron donor system and H 2 accumulation was measured. Under [FeFe]-hydrogenase operon-like gene cluster in V. carteri. Four genes with sequence similarity to HYDA1, HYDA2, HYDEF, and HYDG are arranged within 60 kb of one another. The gene orientation corresponds to the orientation of each shape. Three additional putative genes are within this region: ACKA (XP_002948613), a phosphate acetyltransferase (XP_002948484), and a predicted gene, Q85 (XP_002948486). These three predicted gene products are not hypothesized to catalyze H 2 production or assist in the assembly of the [FeFe]-hydrogenase active site, and these genes are not found within investigated bacterial [FeFe]-hydrogenase gene clusters.
doi:10.1371/journal.pone.0125324.g005 the conditions tested, V. carteri cultures evolved appreciable amounts of H 2 when supplied with an exogenous electron source (MV/Na 2 S 2 O 4 ) to drive H 2 production.
Previous work has detailed differences in basic metabolism between somatic and gonidia cells, which may have implications related to anaerobic H 2 production [42,47]. To test for differences in H 2 metabolism between the two cell types, we separately measured H 2 evolution activities of isolated gonidia and somatic cells in the presence of an abiotic electron donor. Although the rates of H 2 evolution in each cell type appeared to be similar (Fig 2B), the H 2 production rates were normalized based on chlorophyll content. This complicates any direct comparison between gonidia and somatic cells because the amount of photosynthetic machinery does not directly correlate to cell number between the different cell types. Nevertheless, the ability of the isolated cells to generate hydrogen indicates that sufficient hydrogenase machinery is present in both cell types.
Initially, V. carteri did not demonstrate physiological H 2 production under the conditions tested. Based on the rationale that cellular carbon sequestration was not sufficient to produce significant quantities of H 2 during fermentation, algal cultures were supplemented with sodium bicarbonate for 72 hours prior to anaerobic acclimation. In contrast to unsupplemented culture, the sodium-bicarbonate cultures produced demonstrable levels of H 2 , proving that V. carteri is capable of in vivo H 2 metabolism. The measured H 2 production reached a maximum of~3 μmol H 2 /mg Chl at 10 hours, at which point accumulation ceased in the headspace. This is likely due to exhaustion of the fixed carbon storage and represents the maximum yield of H 2 under these experimental fermentative conditions. Under anaerobic conditions, transcript for the hydrogenase homologs HYDA1 and HYDA2 accumulated in V. carteri, and these homologs are likely responsible for the measured in vivo H 2 production. In support of this theory, the genes were heterologously-expressed in S. oneidensis MR-1 and the purified proteins demonstrated quantifiable in vitro H 2 evolution activity when supplied with an artificial electron donor. The purified proteins have similar K m values to one another, matching previous literature [48], although the specific activities measured for each enzyme were dissimilar. This discrepancy may be due to incomplete maturation of HYDA2 in the S. oneidensis MR-1 expression system or may arise from natural differences in activity, as previously observed by Adams et al. [49]. In addition to similar kinetic values, HYDA1 and HYDA2 also share common structural features with other characterized green algal [FeFe]-hydrogenases. For example, both proteins contain two short sequences not observed in non-algal hydrogenases and also lack the canonical iron-sulfur cluster-containing Fdomains [14].
While examining the hydrogenase genes within the V. carteri genome, it was noted that the HYDA1 and HYDA2 genes are within close proximity to genes encoding maturation proteins, HYDEF and HYDG (Fig 5). This level of gene clustering is unique to V. carteri among the three sequenced green algal genomes, although HYDEF and HYDG are proximal to each other in the genomes of both C. reinhardtii and C. variabilis (S1 Fig). Entire gene clusters can be transferred between organisms during DNA transfer events [51,52,53], and these events can also result in gene fusion, reminiscent of the green algal HYDEF fusion [54,55].
[FeFe]-hydrogenase gene clusters, similar to the V. carteri cluster, were also recently reported in the genomes of the photosynthetic heterokonts Nannochloropsis oceanica CCMP1779 [56] and Nannochloropsis gaditana CCMP526 [57]. Thus, the acquisition of [FeFe]-hydrogenase genes in heterokonts and green algae may share an evolutionary history, although the exact origin of the HYD genes is not known [27,58,59]. Together, the presence of these clusters may provide additional tools to investigate the evolutionary history of [FeFe]-hydrogenases.
Using the phylogenetic analysis established by Meuser et al. [49] as a starting point, we analyzed the genomes of several bacterial species which contained [FeFe]-hydrogenase genes closely related to the V. carteri genes. Finding that these bacterial genomes did not contain hydrogenase gene clusters, we expanded our search to bacterial genomes not represented in the phylogram by Meuser et al. This search yielded three closely related thermophillic bacterial species (Fervidobacterium nodosum, Fervidobacterium pennivorans, and Thermotoga thermarum) that contained operon-like gene clusters (>22 kb span) within their genomes. It must be noted, however, that distinct differences exist between the V. carteri and bacterial gene clusters, such as the orientation of the genes and the fusion of the HYDE and HYDF genes in V. carteri. It is therefore intriguing that the relative positions of the genes are identical between V. carteri and bacterial clusters, although the bacterial genes, lacking introns, are shorter in length ( Table 2 and Fig 6). Despite these differences, the preserved order of the genes between the V. carteri and the bacterial genomes suggests that this clustering may be a remnant of horizontal gene transfer rather than a result of directed gene rearrangements for co-regulation.
Examining the operon-like gene clusters between the thermophiles and V. carteri-and assuming direct gene transfer from bacteria to green algae-it is tempting to suggest that both HYDA genes in green algae were obtained during a single horizontal gene transfer event. In contrast, previous evidence from Meuser et al. would instead indicate that the presence of two HYDA genes in V. cateri result from gene duplication, as these genes are reciprocally monophyletic . The presence of the gene cluster in V. carteri-with both putatively duplicate hydrogenase genes in close proximity to one another-is puzzling, especially as clustering of these genes is not observed in either C. reinhardtii or C. variabilis. It is interesting to note, however, that genes encoding phosphate acetyltransferase and acetate kinase are in close association with HYDA2 in both the V. carteri and C. reinhardtii genomes, albeit in different orders in the two species (S1 Fig). This clustering of related genes (acetate assimilation proteins are implicated in fermentative H 2 production [60]) observed in both species is intriguing, and additional green algal genomes containing hydrogenase genes need to be sequenced to explore the evolutionary history further.
In conclusion, herein we provide the first evidence that V. carteri, a complex multicellular eukaryote, is capable of producing H 2 under physiological fermentative conditions. We demonstrate that the two [FeFe]-hydrogenase gene products catalyze H 2 evolution in vitro, suggesting a role for these two enzymes in V. carteri H 2 metabolism. A previously unreported gene cluster within the V. carteri genome encodes these hydrogenases, as well as the essential maturation machinery. This cluster may share a relationship to similar gene clusters found in thermophillic bacteria, providing new avenues in studying the evolutionary origins of [FeFe]hydrogenases.
Supporting Information S1 Fig. Representation of HYD genes in the Volvox carteri, Chlamydomonas reinhardtii, and Chlorella variabilis sequenced genomes. The gene orientation corresponds to the orientation of each shape. The HYDEF and HYDG genes are within close proximity with one another in all three genomes, while the HYDA1 and HYDA2 genes are only associated with other HYD genes in the V. carteri genome. (PDF)