Roles of Coenzyme F420-Reducing Hydrogenases and Hydrogen- and F420-Dependent Methylenetetrahydromethanopterin Dehydrogenases in Reduction of F420 and Production of Hydrogen during Methanogenesis

Reduced coenzyme F₄₂₀ (F₄₂₀H₂) is an essential intermediatein methanogenesis from CO₂. During methanogenesis from H₂ andCO₂, F₄₂₀H₂ is provided by the action of F₄₂₀-reducing hydrogenases.However, an alternative pathway has been proposed, where H₂-dependentmethylenetetrahydromethanopterin dehydrogenase (Hmd) and F₄₂₀H₂-dependentmethylenetetrahydromethanopterin dehydrogenase (Mtd) togetherreduce F₄₂₀ with H₂. Here we report the construction of mutantsof Methanococcus maripaludis that are defective in each putativepathway. Their analysis demonstrates that either pathway supportsgrowth on H₂ and CO₂. Furthermore, we show that during growthon formate instead of H₂, where F₄₂₀H₂ is a direct product offormate oxidation, H₂ production occurs. H₂ presumably arisesfrom the oxidation of F₄₂₀H₂, and the analysis of the mutantsduring growth on formate suggests that this too can occur byeither pathway. We designate the alternative pathway for theinterconversion of H₂ and F₄₂₀H₂ the Hmd-Mtd cycle.

INTRODUCTION

Top

INTRODUCTION

The methanogenic Archaea (methanogens) occupy a variety of anaerobichabitats, where they play essential roles in the conversionof hydrogen and other intermediates to methane (10). The hydrogenotrophicmethanogens use hydrogen to reduce CO₂ to methane. In addition,some hydrogenotrophs use formate, and a few substitute certainlow-molecular-weight alcohols for hydrogen.

The deazaflavin F₄₂₀ is an essential coenzyme of methanogenesis.The reduction of CO₂ to methane requires reduced F₄₂₀ (F₄₂₀H₂),since it is the sole electron donor for the step that reducesmethylenetetrahydromethanopterin (methylene-H₄MPT) (Mer in Fig.1). In addition, F₄₂₀H₂ is the electron donor for F₄₂₀H₂-dependentmethylenetetrahydromethanopterin dehydrogenase (Mtd), one oftwo enzymes that reduce methenyl-H₄MPT. The other enzyme, H₂-dependentmethylenetetrahydromethanopterin dehydrogenase (Hmd), uses H₂directly. mRNA abundance for mtd increased markedly under hydrogen-limitedgrowth conditions (4), suggesting that Mtd may be more importantwhen H₂ is limiting.

View larger version (15K):
[in this window]
[in a new window]

FIG. 1. The hydrogenotrophic methanogenic pathway. See reference 3 for a full description of methanogenesis. The Hmd-Mtd cycle is boxed. Abbreviations: CoB, coenzyme B; CoM, coenzyme M; F₄₂₀, coenzyme F₄₂₀; Fd, ferredoxin; Frc, cysteine-containing F₄₂₀-reducing hydrogenase; Fru, selenocysteine-containing F₄₂₀-reducing hydrogenase; Mer, methylenetetrahydromethanopterin reductase; MFR, methanofuran.

The F₄₂₀-reducing hydrogenases (Fru and Frc) reduce F₄₂₀ withH₂. However, an alternative route for this process has beenproposed. In Methanothermobacter marburgensis the specific activityof F₄₂₀-reducing hydrogenase, a Ni-Fe hydrogenase, decreased20-fold under nickel-limited growth conditions. In contrast,the specific activities of Hmd and Mtd, neither of which requiresnickel for activity, increased six- and fourfold, respectively(1). These observations led to the proposal that under nickel-limitedconditions, F₄₂₀ may be reduced by the concerted action of Hmdand Mtd, the former working in the forward direction (with respectto the methanogenic pathway) and the latter in the reverse direction(1, 2). This pathway is boxed in Fig. 1.

Here we report on the properties of mutants of Methanococcusmaripaludis that are deficient in Hmd, Mtd, or the F₄₂₀-reducinghydrogenases. The results demonstrate that neither Hmd nor Mtdis essential, confirming that either enzyme is sufficient formethenyl-H₄MPT reduction. The results also indicate that, invivo, Hmd and Mtd do indeed constitute an alternate pathwayfor the reduction of F₄₂₀ with H₂, which we designate the Hmd-Mtdcycle. Furthermore, we show that during growth on formate, H₂production occurs, evidently by reversal of either the F₄₂₀-reducinghydrogenase or the Hmd-Mtd cycle.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

Growth of strains and measurement of H₂. M. maripaludis was grown on H₂ and CO₂ by standard anaerobictechniques in McCas medium as described elsewhere (6). For growthon formate, McCas medium was modified to contain 200 mM sodiumformate and 200 mM MOPS (morpholinepropanesulfonic acid) buffer(pH 7.0), NaCl was decreased to 0.18 M, and the gas atmospherewas 80% N₂ and 20% CO₂ at a pressure of 15 lb/in². Cultures(5-ml volume) were inoculated with 0.25 to 0.5 ml of a cultureactively growing on formate. Growth was monitored by opticaldensity at 660 nm. The accumulation of H₂ in the headspace (20-mlvolume) was measured using a Hach CARLE Series 100 AGC gas chromatographequipped with a Supelco 60/80 mesh molecular sieve 5A column(6 ft by 1/8 in.) and a trace analytical RGD2 reduction gasdetector.

Construction of plasmids and strains. Primers are listed in Table 1, and strains and plasmids arelisted in Table 2. PCR products containing the genes hmd, mtd,frcA, and fruA and their flanking regions were generated usingthe primer pairs hmdcln5for and hmdcln5rev, mtdcln5for and mtdcln5rev,frcAfor2 and frcArev2, and fruAfor and fruArev, respectively.The products were cloned into pCR2.1topo to generate phmdtopo,pmtdtopo, pfrcAtopo, and pfruAtopo. An in-frame deletion ofhmd was produced by PCR of phmdtopo using primers hmddel1 andhmddel2, followed by digestion with AscI and ligation to producephmddeltopo. pmtddeltopo, pfrcAdeltopo, and pfruAdeltopol weregenerated in the same way using pmtdtopo and the primers mtddel1and mtddel3, pfrcAtopo and the primers frcdel1 and frcdel2,and pfruAtopo and the primers frudel1 and frudel2, respectively.The in-frame deletion of hmd was amplified from phmddeltopousing the primers hmddelamp1 and hmddelamp3; the resulting fragmentwas digested with BamHI and ligated into the vector pCRprtneoto produce pCRprt ${Delta}$ hmdneo. pCRprt ${Delta}$ mtdneo was produced in the sameway from pmtddeltopo using the primers mtddelamp1 and mtddelamp2and digesting with BamHI. pCRprt ${Delta}$ frcneo was produced from pfrcAdeltopousing frcdelamp5 and frcdelamp6 and digesting with XbaI, andpCRprt ${Delta}$ fruneo was produced from pfruAdeltopo using frudelamp5and frudelamp6 and digesting with XbaI.

View this table:
[in this window]
[in a new window]

TABLE 1. Primers

View this table:
[in this window]
[in a new window]

TABLE 2. Strains and plasmids

Strains containing markerless in-frame deletions of hmd, mtd,frcA, and fruA were constructed in strain Mm900 as describedelsewhere (6) using the plasmids pCRprt ${Delta}$ hmdneo, pCRprt ${Delta}$ mtdneo,pCRprt ${Delta}$ frcneo, and pCRprt ${Delta}$ fruneo, respectively, to produce strainsMm1097, Mm1020, Mm1183, and Mm1145, respectively. A double mutantof frcA and fruA was constructed by the same procedure fromthe frcA mutant strain Mm1183 by using pCRprt ${Delta}$ fruneo to produceMm1184. Deletions were confirmed by Southern analysis. For experimentstesting whether hmd deletion mutations could be made, pCRprt ${Delta}$ hmdneowas transformed into recipient strains. The resultant merodiploidswere streak purified, allowed to grow overnight without antibioticselection, and plated on counterselection plates containing8-azahypoxanthine. Colonies were analyzed by Southern blottingto distinguish strains containing deletions of the hmd genefrom those containing the wild-type hmd gene.

RESULTS AND DISCUSSION

Top

RESULTS AND DISCUSSION

F₄₂₀ reduction during growth on H₂. We used a genetic approach in M. maripaludis to test whetherF₄₂₀-reducing hydrogenase and the Hmd-Mtd cycle constitute twoalternative pathways for the reduction of F₄₂₀ in vivo. M. maripaludiscontains genes for Hmd and Mtd and two sets of genes for F₄₂₀-reducinghydrogenases, fruADGB and frcADGB (5). FruA contains selenocysteineresidues, while FrcA contains cysteine residues in correspondingpositions, and in the closely related Methanococcus voltae frcexpression is repressed in the presence of selenium in the medium(7, 8). We hypothesized that if Hmd and Mtd can provide an alternativepathway for the reduction of F₄₂₀, then mutants with deletionsin fru and frc should be viable in the presence of wild-typehmd and mtd. Conversely, mutants with mutations in either hmdor mtd should be viable in a fru⁺ frc⁺ background.

Using H₂ and CO₂ as growth substrates, we made the followingmutants, all containing markerless in-frame deletions: ${Delta}$ fruA, ${Delta}$ frcA, double mutant ${Delta}$ fruA ${Delta}$ frcA, ${Delta}$ mtd, and ${Delta}$ hmd strains. ${Delta}$ fruA ${Delta}$ frcA, ${Delta}$ mtd, and ${Delta}$ hmd strains each grew normally on H₂ and CO₂ (Fig.2A). Since F₄₂₀H₂ is essential for methanogenesis, each mutantmust retain a pathway for F₄₂₀ reduction using H₂. Hence, theresults imply that F₄₂₀-reducing hydrogenase and the Mtd-Hmdcycle are each sufficient for this function.

View larger version (12K):
[in this window]
[in a new window]

FIG. 2. Growth and H₂ production by wild-type and mutant strains of M. maripaludis. (A) Growth on H₂; (B) growth and H₂ production on formate. For the ${Delta}$ mtd mutant on formate, data from three separate growth experiments are plotted and are represented by a single line. OD₆₆₀, optical density at 660 nm.

As a formal possibility, a third, unknown pathway for the reductionof F₄₂₀, different from the F₄₂₀-reducing hydrogenase and theMtd-Hmd cycle, could exist. To test this possibility, we attemptedto construct an ${Delta}$ hmd mutation in a ${Delta}$ fruA ${Delta}$ frcA background. Followingour regular procedure for generation of markerless mutations(6), we introduced ${Delta}$ hmd (containing the N- and C-terminal flankingregions of hmd) on an integrative vector to produce merodiploidsof ${Delta}$ hmd and hmd⁺. We made such merodiploids in the ${Delta}$ fruA ${Delta}$ frcA, ${Delta}$ frcA, and fru⁺ frc⁺ backgrounds. We then selected for resolutionof the merodiploids via a second recombination event and analyzedthe resulting strains by Southern blotting. In principle a mixtureof wild-type and deletion strains should result, depending onwhere the second recombination event occurs. We counted thenumbers of resulting ${Delta}$ hmd and hmd⁺ strains in each background.In the fru⁺ frc⁺ background six out of eight strains testedcontained ${Delta}$ hmd and the remaining two contained hmd⁺. In the ${Delta}$ frcAbackground, which should express fru and therefore retain activeF₄₂₀-reducing hydrogenase, three strains contained ${Delta}$ hmd and fivecontained hmd⁺. In contrast, in the ${Delta}$ fruA ${Delta}$ frcA background all40 strains tested contained only hmd⁺. The results indicatethat while Hmd can be eliminated in a strain with active F₄₂₀-reducinghydrogenase, it is essential in a strain lacking F₄₂₀-reducinghydrogenase. Therefore, no evidence could be found for the existenceof a third pathway that would produce F₄₂₀H₂ from H₂.

H₂ production during growth on formate. Growth on formate differs from growth on H₂ and CO₂ becauseF₄₂₀H₂ is a direct product of formate oxidation (Fig. 1). Neitherthe F₄₂₀-reducing hydrogenase nor the Mtd-Hmd cycle should benecessary for the production of F₄₂₀H₂. However, the reversalof either pathway might result in H₂ production. We characterizedthe growth of the ${Delta}$ fruA ${Delta}$ frcA, ${Delta}$ mtd, and ${Delta}$ hmd mutants on formate.The ${Delta}$ fruA ${Delta}$ frcA and ${Delta}$ hmd mutants grew normally, while the ${Delta}$ mtd mutantgrew after a lag. For each strain, H₂ accumulated in the headspaceof the tubes as growth commenced and disappeared when growthended (Fig. 2B). This observation suggests that H₂ is producedfrom F₄₂₀H₂ and that either the F₄₂₀-reducing hydrogenase orthe Mtd-Hmd cycle can mediate this conversion. H₂ accumulatedto a substantially higher level in tubes containing culturesof the ${Delta}$ mtd mutant than in tubes containing any of the otherstrains. In the ${Delta}$ mtd strain, Hmd is the only enzyme for the reductionof methenyl-H₄MPT. Therefore, H₂ production, which would occurby the action of the F₄₂₀-reducing hydrogenase, should be essential.Due to the relatively low affinity of Hmd for H₂ (9), substantiallyhigher H₂ levels accumulate. In contrast, in the other strainsMtd is present and can use F₄₂₀H₂ for the reduction of methenyl-H₄MPT.These results indicate that H₂ production from F₄₂₀H₂ occursduring growth on formate and that either the F₄₂₀-reducing hydrogenaseor the Mtd-Hmd cycle can carry out this process.

Whether H₂ is a necessary intermediate during growth on formatecannot be determined from the present data. The generation ofa ${Delta}$ fruA ${Delta}$ frcA ${Delta}$ hmd triple mutant, which is expected to grow inthe presence of formate, could resolve this question. Growthof the mutant on formate alone without the addition or generationof H₂ would indicate that H₂ is not a required intermediate.A requirement for added H₂ would indicate that H₂ productionis required during growth on formate. Efforts to construct sucha mutant are under way.

Concluding remarks. The genetic approach taken here has shown that two alternativepathways, the F₄₂₀-reducing hydrogenase and the Hmd-Mtd cycle,can function in vivo for the reduction of F₄₂₀ with H₂. Furthermore,during growth on formate the same pathways function in reverseto produce H₂ from F₄₂₀H₂. The lack of growth differences betweenthe wild-type and mutant strains on H₂ and CO₂ (Fig. 2A) suggeststhat neither pathway for F₄₂₀ reduction was rate limiting. However,in nature the F₄₂₀-reducing hydrogenase may constitute the majorpathway when sufficient nickel is present, while the Hmd-Mtdcycle may be important when nickel is limiting (1, 2).

ACKNOWLEDGMENTS

This work was supported by Department of Energy Basic Researchfor the Hydrogen Fuel Initiative grant DE-FG02-05ER15709 andgrant NCC 2-1273 from the NASA Astrobiology Institute.

We thank William Whitman for helpful comments.

FOOTNOTES

* Corresponding author. Mailing address: University of Washington, Department of Microbiology, Box 357242, Seattle, WA 98195-7242. Phone: (206) 685-1390. Fax: (206) 543-8297. E-mail: leighj{at}u.washington.edu

Published ahead of print on 16 May 2008.

REFERENCES

Top