Electron Donor Systems to Facilitate Development of Assays for Two Flavoproteins Involved in Tetrahydromethanopterin Biosynthesis

Methane production by archaea depends on tetrahydromethanopterin (H 4 MPT), a pterin-containing cofactor that carries one-carbon units. Two redox reactions within the nine steps of H 4 MPT side chain biosynthesis have been hypothesized. Biochemical assays have demonstrated that the archaeal iron-sulfur flavoprotein dihydromethanopterin reductase X (DmrX or MM1854) catalyzes the final reaction of the pathway, the reduction of dihydromethanopterin to H 4 MPT, using dithiothreitol (DTT) as an artificial electron donor. The crystal structure of DmrB, a bacterial DmrX homolog that lacks iron-sulfur clusters, has led to a proposed ping-pong mechanism of electron transfer between FMNH 2 and the FMN prosthetic group of DmrB. However, an enzymatic assay to test the hypothetical DmrB mechanism is lacking because a suitable electron donor has not previously been identified. Furthermore, a second uncharacterized archaeal flavoprotein (MM1853) has been hypothesized to function in H 4 MPT side chain biosynthesis. In this work, to facilitate the development of assays to elucidate the functions of DmrB and MM1853, we tested a variety of electron donors, including dithiothreitol, ferredoxin, and a system consisting of NADH and an NADH-dependent fla-vin-reducing enzyme sis. While NADH and NADPH were incapable of directly reducing the FMN cofactor of MM1853, DTT or NADH/Fre could eliminate the FMN peaks. These results establish the basis for new oxidoreductase assays that will facilitate testing several proposed DmrB mechanisms and defining the specific function of MM1853 in methanogen cofactor biosynthesis.


Introduction
Methane-producing microorganisms (methanogenic archaea) require strictly anaerobic conditions for growth and are considered extremophiles due to their inability to grow in the presence of O 2 [1]. Methane production by archaea depends on tetrahydromethanopterin (H 4 MPT), an unusual pterin-containing cofactor that carries one-carbon units [2]. Chemical compounds that block a committed step in H 4 MPT biosynthesis have been shown to inhibit the microbial production of methane, a greenhouse gas associated with climate change [3] [4] [5]. In addition, because methane-producing microbes in the intestinal tracts of rodents and humans have been correlated with obesity [6], newly discovered enzymes in the pathway of H 4 MPT biosynthesis could serve as potential targets for novel pharmaceutical drugs intended to treat morbid obesity.
Of the nine enzymatic steps in the pathway for synthesis of the H 4 MPT side chain, only the last step and the third step are proposed to be oxidation-reduction reactions [7] [8]. Genetic complementation studies [9] and biochemical assays [10] have together demonstrated that the final reaction of H 4 MPT side chain biosynthesis (step 9) is catalyzed by an archaeal flavoprotein A (AfpA) that is also called dihydromethanopterin reductase X (DmrX or MM1854) (Figure 1(a)). In addition to containing a binding site for FMN, the DmrX sequence contains two binding sites for four-iron four-sulfur (4Fe-4S) centers [11], which are associated with the ability of dithiothreitol (DTT) to serve as an artificial electron donor [10]. The third step of H 4 MPT side chain biosynthesis is also a putative oxidation-reduction step [7], but the enzyme catalyzing this reaction has not yet been discovered.
Although DmrX is a strong candidate as a target for inhibiting methane production by methanogenic archaea, detailed structural and mechanistic studies of DmrX have been hindered by the presence of the two 4Fe-4S clusters that make DmrX susceptible to inactivation by O 2 [10]. However, a DmrX homolog from the aerobic bacterium Burkholderia xenovorans (DmrB) has been identified that can serve as an oxygen-stable model system for flavin-dependent dihydromethanopterin reductases. The air-stable DmrB protein lacks 4Fe-4S cluster bind- ing sites and yet retains one FMN binding site per monomer [9]. Curiously, a crystal structure for DmrB has been reported that shows two FMN molecules per monomer rather than one [12]. This observation led the authors to propose a ping-pong mechanism of electron transfer (Figure 1(b)) in which a loosely bound FMNH 2 molecule (FMN-2 reduced ) transfers electrons to the tightly bound FMN prosthetic group of DmrB (FMN-1); this would be followed by the binding of the substrate dihydromethanopterin and reduction to H 4 MPT using electrons from reduced FMN-1 [12]. The model predicts that the loosely bound FMN-2 may be re-reduced by a flavin-reducing enzyme (Fre) with electrons from NAD(P)H. This hypothetical ping-pong mechanism for DmrB electron transfer has not yet been tested biochemically, and a reductant for DmrB has not been previously reported.
A second flavoprotein, the archaeal protein MM1853 and its bacterial homo-

Production and Purification of His6-DmrB
Burkholderia xenovorans DmrB with an N-terminal histidine tag (His 6 -DmrB) was produced in Escherichia coli BL21 (DE3) RIL cells (Stratagene, La Jolla, CA) containing pET41a:his 6 -DmrB, and cells were harvested by centrifugation, as described previously by McNamara et al. [12]. The cell pellet (about Advances in Microbiology Some His 6 -DmrB eluted upon washing twice with 2 ml of 50 mM Tris, pH 8.0, 200 mM NaCl, 100 mM imidazole, but more highly purified protein (~95% pure) eluted with two additions of elution buffer (2 ml) containing 250 mM imidazole. The protein was stored at 4˚C until use.

Production and Purification of His6-Fre from Salmonella enterica
A codon-optimized gene encoding an NAD(P)H-flavin reductase (Fre) from S. enterica (NP_462864.1) [14] with an N-terminal six-histidine tag (His 6 -Fre) was synthesized by GenScript (Piscataway, NJ), subcloned into the NdeI-HinDIII sites of pET41a, and transformed into E. coli BL21(DE3)-RIL cells. Cells were grown and harvested in the same manner as for DmrB, and protein was purified aerobically using the same protocol as DmrB, except that His 6 -Fre tended to elute in 100 mM imidazole. All centrifugation and incubation steps were carried out at 4˚C.
His 6 -Fre could be stored in its elution buffer and could remain stable for at least one month at 4˚C in that buffer. However, His 6 -Fre was found to be sensitive to temperature fluctuations when stored in a refrigerator that was opened and closed frequently. Therefore, the active fraction (Elution 1, 100 mM imidazole) was separated into 200-μl aliquots, sealed in 2-ml vials, and kept in a 4˚C cold room to minimize temperature fluctuations. For enzymatic studies, His 6 -Fre was maintained at 4˚C, either on ice or on Lab Armor cold beads (Fisher Scientific).

Dithionite and Dithiothreitol (DTT) Electron Transfer Assay
Electron transfer assays were conducted anaerobically in a chamber containing 98% N 2 and 2% H 2 (Coy Laboratories, Grass Lake, MI) or in anaerobic cuvettes

Development of Small-Scale Flavin Reductase (Fre) Assay
A ping-pong mechanism for DmrB has been proposed in which an enzyme (Fre) reduces a loosely bound FMN cofactor (FMN-2) of DmrB [12]. To begin to test this hypothesis, an assay using Fre from S. enterica [14] was developed by mon- For experiments to estimate the reaction rate, the protein amounts used above were too high, and thus the enzyme solution was diluted 50-fold with assay buffer. His 6 -Fre was kept on 4˚C beads during these assays since it would lose activity at higher temperatures throughout the day, as described above. For studies in the absence of exogenous FMN, the standard assay (200 μl) included 196 μl of enzyme and 2 μl of 10 mM NADH (100 μM final). The reaction was initiated by the addition of His 6 -Fre. The solution buffering the enzyme was used as assay buffer.

Breakage of M. thermophila Cells and Storage of Cell-Free Extracts for Carbon Monoxide Dehydrogenase (CODH)/Ferredoxin Studies
Methanogen Methanosarcina thermophila cells were previously grown on ace- lysed as described previously [17]. Cells (2.5 g) were mixed with breakage buffer (5 ml of 50 mM TES, pH 7.0, 10 mM MgCl 2 , 5% glycerol, and 10 mM 2-mercaptoethanol) and 2 μl DNase and broken anaerobically with one passage using a cold (4˚C) French Press (ThermoScientific). After anaerobic centrifugation at 32,000 × g (4˚C) for 45 min, the supernatant (cell-free extract) was flash-frozen and stored as pellets in liquid N 2 until use.

Electron Transfer to DmrB via CO and M. thermophila Extracts
It has previously been shown that CO and extracts of the sulfate-reducing archaeon Archaeoglobus fulgidus could be used to reduce archaeal flavoprotein A (AfpA), a DmrB homolog from A. fulgidus [11]. Electrons from CO presumably flow to a large carbon monoxide dehydrogenase (CODH) complex in the extracts, where they are transferred to a ferredoxin or a ferredoxin-like protein, and then to an archaeal flavoprotein AfpA [11] that is a homolog of DmrB. Us- ing similar reasoning, we tested the ability of CO and extracts from the methanogen M. thermophila to reduce purified His 6 -DmrB.
His 6 -DmrB was purified as described above and stored at 4˚C until use. Enzyme solutions were centrifuged before use and then desalted using prepacked

Production and Purification of His6-MM1853
To produce an expression plasmid for heterologous protein production, a   [11]. Unfortunately, dithionite is too strong for use in a DmrB assay because it can also nonenzymatically reduce the substrate dihydromethanopterin. We then tested DTT at concentrations from 30 to 100 mM, but DTT was unable to reduce the FMN prosthetic group of His 6 -DmrB over the period of one hour (Figure 2). This result was unexpected since DTT effectively reduces the archaeal homolog DmrX and has been used to develop a successful activity assay for that enzyme [10]. We

DmrB Reduction Using an NADH/Fre Electron Transfer System
Having failed to identify a chemical reductant suitable for a DmrB assay among our selected chemical candidates, we next examined the hypothesis of McNamara et al. (12) that an electron transport system consisting of exogenous FMN, To investigate these results in a more systematic way, we first purified transfer from NADH to Fre to DmrB; however, contrary to the prediction of the ping-pong mechanism [12], exogenous FMN did not appear to be a requirement for Fre-mediated electron transfer to DmrB.
To quantify the rate of His 6 -DmrB reduction in each time interval by the NADH/Fre system using the molar extinction coefficient for FMN at 460 nm

DmrB Reduction Using CO, CO Dehydrogenase (CODH), and Ferredoxin
Ding and Ferry [11] showed previously that archaeal flavoprotein (AfpA, an archaeal homolog of DmrB) could be reduced by CO in the presence of CODH and ferredoxin. Both AfpA and DmrB lack iron-sulfur clusters, raising the possibility that the archaeal CO/CODH/ferredoxin system may act as an artificial electron donor system for His 6

Test of DTT and NAD(P)H as Reductants for MM1853
DTT (

NADH/Fre Enzyme Reduction of MM1853
Because NADH alone was unable to reduce the FMN of His 6 -MM1853, we tested whether His 6 -Fre could facilitate electron transfer from NADH to His 6 -MM1853.
When the assay was initiated by adding His 6 -Fre, a decrease of 0.135 absorbance unit was observed compared to the original absorbance at 380 nm, while the 460-nm peak was decreased completely to baseline level (Figure 7, line 2). For the 460-nm peak, the reaction appeared to be complete within 5 min since the absorbance at this wavelength did not decrease substantially at later time points   CO/CODH/ferredoxin has been proposed as a physiological electron donor for both DmrX [10] and an archaeal flavoprotein homolog (AfpA) that lacks 4Fe-4S clusters [11]. The current work demonstrates that archaeal CO/CODH/ferredoxin effectively reduces bacterial DmrB (Figure 4). A future assay for DmrB could be developed by combining CO/CODH/ferredoxin with DmrB and the assay components for detecting the product (H 4 MPT), which is coupled to NADH production via methylene H 4 MPT dehydrogenase [11] [26]. However, a CODH-based assay would be complicated by the need to grow strictly anaerobic methanogens and use an O 2 -free chamber to purify CODH and ferredoxin, which are not commercially available.
An alternative electron donor system for DmrB has been proposed from a DmrB crystal structure in which two molecules of FMN (rather than one) were unexpectedly found per active site of DmrB [12]. The presence of an extra FMN in the structure led to a hypothetical ping-pong mechanism for electron transfer ( Figure 1(b)) in which a tightly bound DmrB-FMN (FMN-1) receives electrons from a loosely bound FMNH 2 molecule (reduced FMN-2), which could be re-reduced by a hypothetical NAD(P)H-dependent flavin-reducing enzyme (Fre). In the current work, we tested the feasibility of this mechanism using Fre from S. enterica. NADH in combination with Fre successfully reduced the prosthetic group of DmrB (Figure 3), providing the first biochemical evidence for the feasibility of the ping-pong mechanism. Complete reduction using the NADH/Fre system took about an hour, and the initial velocity (1.3 µM/min) was more than an order of magnitude lower than the rate with the CO/CODH/ferredoxin system ( Figure 4), showing that the conditions might be further optimized in the future.
Nevertheless, for reasons discussed above, if DmrB can be pre-reduced by incubating with NADH and Fre prior to mixing with the substrate, the creation of an NADH/Fre-based DmrB assay may be more convenient and less cumbersome than using CO/CODH/ferredoxin as the electron donor.
With regard to the proposed DmrB ping-pong mechanism of electron transfer [12], it was unexpected to find that DmrB could be reduced by NADH and Fre alone without the need for added FMN. Since FMN-1 not covalently bound (12) that could alter the redox properties of the FMN prosthetic group and contribute to its DTT-mediated reduction. Interestingly, the NADH/Fre system was also shown to fully reduce MM1853 within 5 min (Figure 7), which is faster than when DTT was used as the electron donor ( Figure 6). This result could facilitate the development of a future enzyme assay to discover the function of MM1853 and raises the possibility that a similar Fre-type system may function in methanogenic archaea. Although it is possible that the physiological electron donor in methanogens may be CODH/ferredoxin, F420, or another uncommon methanogen coenzyme, these materials are not commercially available. Thus, the NADH/Fre system currently offers an advantage over other potential electron donors for the development of a new assay to probe the function of MM1853.
Our current hypothesis is that MM1853 might contribute to the third reaction of H 4 MPT side chain biosynthesis, which was previously proposed by White [7] to involve the reductive opening of a ribose ring to form the linear ribitol of H 4 MPT, as illustrated in Figure 8.

Conclusion
In conclusion, this work describes the discovery of electron donors for two previously uncharacterized flavoproteins involved in H 4 MPT biosynthesis. DmrB could be reduced by a CO/CODH/ferredoxin and an NADH/Fre system, providing the first biochemical evidence supporting the proposed ping-pong mechanism Figure 8. The flavoprotein MM1853 might potentially be involved in the third step of H 4 MPT side chain biosynthesis. As an oxidoreductase, one possible role for MM1853 could be in the reductive opening of the ribose ring to form the ribitol moiety of H 4 MPT, as described by White [7].