Purification of Soluble Cytochrome b 5 as a Component of the Reductive Activation of Porcine Methionine Synthase*

In mammals, methionine synthase plays a central role in the detoxification of the rogue metabolite homocysteine. It catalyzes a transmethylation reaction in which a methyl group is transferred from methyltetrahydrofolate to homocysteine to generate tetrahydrofolate and methionine. The vitamin B12cofactor cobalamin plays a direct role in this reaction by alternately accepting and donating the methyl group that is in transit from one substrate (methyltetrahydrofolate) to another (homocysteine). The reactivity of the cofactor intermediate cob(I)alamin renders the enzyme susceptible to oxidative damage. The oxidized enzyme may be returned to the catalytic turnover cycle via a reductive methylation reaction that requires S-adenosylmethionine as a methyl group donor, and a source of electrons. In this study, we have characterized an NADPH-dependent pathway for the reductive activation of porcine methionine synthase. Two proteins are required for the transfer of electrons from NADPH, one of which is microsomal and the other cytoplasmic. The cytoplasmic protein has been purified to homogeneity and is soluble cytochrome b 5. It supports methionine synthase activity in the presence of NADPH and the microsomal component in a saturable manner. In addition, purified microsomal cytochrome P450 reductase and soluble cytochromeb 5 reconstitute the activity of the porcine methionine synthase. Identification of soluble cytochromeb 5 as a member of the reductive activation system for methionine synthase describes a function for this protein in non-erythrocyte cells. In erythrocytes, soluble cytochromeb 5 functions in methemoglobin reduction. In addition, it identifies an additional locus in which genetic polymorphisms may play a role in the etiology of hyperhomocysteinemia, which is correlated with cardiovascular diseases.

In mammals, methionine synthase plays a central role in the detoxification of the rogue metabolite homocysteine. It catalyzes a transmethylation reaction in which a methyl group is transferred from methyltetrahydrofolate to homocysteine to generate tetrahydrofolate and methionine. The vitamin B 12 cofactor cobalamin plays a direct role in this reaction by alternately accepting and donating the methyl group that is in transit from one substrate (methyltetrahydrofolate) to another (homocysteine). The reactivity of the cofactor intermediate cob(I)alamin renders the enzyme susceptible to oxidative damage. The oxidized enzyme may be returned to the catalytic turnover cycle via a reductive methylation reaction that requires S-adenosylmethionine as a methyl group donor, and a source of electrons. In this study, we have characterized an NADPH-dependent pathway for the reductive activation of porcine methionine synthase. Two proteins are required for the transfer of electrons from NADPH, one of which is microsomal and the other cytoplasmic. The cytoplasmic protein has been purified to homogeneity and is soluble cytochrome b 5 . It supports methionine synthase activity in the presence of NADPH and the microsomal component in a saturable manner. In addition, purified microsomal cytochrome P450 reductase and soluble cytochrome b 5 reconstitute the activity of the porcine methionine synthase. Identification of soluble cytochrome b 5 as a member of the reductive activation system for methionine synthase describes a function for this protein in nonerythrocyte cells. In erythrocytes, soluble cytochrome b 5 functions in methemoglobin reduction. In addition, it identifies an additional locus in which genetic polymorphisms may play a role in the etiology of hyperhomocysteinemia, which is correlated with cardiovascular diseases.
Homocysteine is a rogue metabolite that is generated by the hydrolysis of S-adenosylhomocysteine, the spent form of the ubiquitous methyl group donor S-adenosylmethionine. Elevated levels of plasma homocysteine constitute a significant and independent risk factor for cardiovascular diseases (1)(2)(3)(4)(5)(6)(7). There are two major metabolic avenues for detoxifying homocysteine in mammalian cells. Transmethylation, catalyzed by methionine synthase or by betaine-homocysteine methyltransferase, salvages homocysteine as methionine, whereas transsulfuration, catalyzed by cystathionine ␤-synthase, commits it to degradation (Fig. 1). Of these three enzymes, betaine-homocysteine methyltransferase has a very limited tissue distribution and has been found only in the liver and kidney (8). Mutations in either methionine synthase (9,10) or cystathionine ␤-synthase (11) result in severe hyperhomocystinemia, with attendant early and aggressive occlusive arterial diseases. Noticing the correlation between prominent arterial damage and elevated homocysteine, McCully (12) boldly postulated in 1969 that the vascular alterations could be attributed to high homocysteine concentrations.
Despite the longevity of this hypothesis, little is known about how cells regulate homocysteine levels or about the etiology of homocysteine-dependent vascular changes. Whereas the activity of methionine synthase directly influences intracellular homocysteine concentration, its activity is itself dependent on auxiliary redox proteins that could influence homocysteine levels indirectly (Fig. 2). This is strongly supported by the existence of the cblE class of patients with an inborn error of cobalamin metabolism resulting in a functional methionine synthase deficiency, albeit the methionine synthase itself is apparently normal (13)(14)(15). Biochemical analysis of fibroblast cells from cblE patients indicates that the reductive activation system is compromised. However, the identities of the mammalian proteins that regulate the activity of methionine synthase have not been established.
Methionine synthase (EC 2.1.1.13) is a cobalamin (or vitamin B 12 )-dependent enzyme that catalyzes the transfer of a methyl group from CH 3 -H 4 folate 1 to homocysteine (Fig. 2). During this transmethylation reaction, the methyl group is transiently transferred to the cofactor cob(I)alamin, a supernucleophilic and oxidatively labile intermediate, to form methylcobalamin (16,17). The latter then transfers its methyl group to homocysteine to generate methionine. During catalysis, accidental oxidation of the reactive intermediate, cob(I)alamin, results in leakage of enzyme out of the turnover cycle and its accrual in an inactive cob(II)alamin state. Readmission to the catalytic cycle requires an electron source and S-adenosylmethionine as a methyl donor. Under in vitro assay conditions, artificial donors such as titanium citrate or dithiothreitol and hydroxycobalamin can serve as an electron source (18).
In Escherichia coli, the physiological reductive activation system uses NADPH as the preferred electron donor and requires two flavoproteins, NADPH-flavodoxin (or ferredoxin) oxidoreductase and flavodoxin, for electron transfer to methionine synthase (19,20). In porcine liver, both NADH and NADPH can serve as the ultimate source of electrons, and crude fractionation studies indicated that at least two proteins are required for reductive activation (15). In this study, we have identified the cellular locations of the two proteins: one is microsomal, whereas the second is cytoplasmic. The latter com-ponent has been purified to homogeneity and is soluble cytochrome b 5 , which suggests the identities of the microsomal NADH-and NADPH-dependent oxidases to be cytochrome b 5 reductase and cytochrome P450 reductase, respectively. Furthermore, purified microsomal cytochrome P450 reductase can reconstitute methionine synthase activity in the presence of soluble cytochrome b 5 . The role of soluble cytochrome b 5 in the reductive activation of methionine synthase is a novel addition to its so far limited repertoire of known functions. Cytochrome b 5 represents an additional locus that is a candidate for genetic polymorphisms that may be correlated with mild hyperhomocystinemia.

Preparation of Microsomes
Two procedures were employed interchangeably for purifying microsomes. The first protocol (21) represents a slightly larger scale purification in which 300 g of cubed pig liver was placed in 1 liter of 5 mM Tris (pH 7.2) containing 0.25 M sucrose, 1 mM MgCl 2 , 13 mg of trypsin inhibitor, 25 mg of phenylmethylsulfonyl fluoride, 3 mg of TLCK, and 1 ml of aprotinin. The liver pieces were homogenized in a Waring blender at 4°C for 3 ϫ 1 min at high speed with 1-min intervals. The homogenate was centrifuged at 12,000 ϫ g for 1 h, and the supernatant was filtered through two layers of Miracloth and centrifuged at 105,000 ϫ g for 1 h. The pellet was resuspended in 200 ml of 5 mM Tris (pH 7.2) containing 0.25 M sucrose and 1 mM MgCl 2 and centrifuged at 105,000 ϫ g for 1 h. The washed microsomes were resuspended in 50 mM potas-sium P i (pH 7.2) containing 20% glycerol to ϳ30 mg/ml protein concentration and stored at Ϫ80°C until further use.
Alternatively, to obtain microsomes of higher purity, a sucrose density gradient ultracentrifugation method was employed as described previously (22). For this, 100 g of cubed pig liver was placed in 500 ml of 5 mM Tris (pH 7.2) containing 0.25 M sucrose, 1 mM MgCl 2 , 7 mg of trypsin inhibitor, 13 mg of phenylmethylsulfonyl fluoride, 2 mg of TLCK, and 0.5 ml of aprotinin. The initial homogenization, centrifugation, and filtration steps were conducted as described above, and the filtrate was centrifuged at 105,000 ϫ g for 90 min. The pellet was suspended in 9 ml of 5 mM Tris (pH 7.2) containing 0.44 M sucrose and 1 mM MgCl 2 and layered over a discontinuous gradient. The latter consisted of 3 ml of 3 M sucrose and 15 mM CsCl overlaid with 1.0 ml of 0.6 M sucrose and 15 mM CsCl. The extract was centrifuged for 3 h at 105,000 ϫ g. The upper band containing the smooth microsomes sedimented above the interface, whereas the rough microsomal fraction sedimented at the bottom of the tube. The two microsome-containing bands were separately resuspended in 50 mM potassium P i (pH 7.2) containing 20% glycerol and stored at Ϫ80°C until further use.

Marker Assay for Microsomes
Microsomal localization of component I activity was verified by measuring the activity of the marker enzyme glucose-6-phosphatase, in which the formation of free phosphate was followed (23). Microsomal samples (1 ml) were dialyzed against two changes of a 2-liter solution containing 0.25 M sucrose and 1 mM EDTA to remove the phosphate prior to the glucose-6-phosphatase assay being performed.

Purification of Component II
The following steps were performed at 4°C.
Step 1: Preparation of Homogenate-One pig liver was cubed and placed in 2 liters of 100 mM potassium P i (pH 5.9) containing 25 mg of trypsin inhibitor, 50 mg of phenylmethylsulfonyl fluoride, 6 mg of TLCK, and 2 ml of aprotinin and was homogenized at high speed in a Waring blender for 3 ϫ 1-min bursts with 1-min intervals to prevent overheating. The homogenate was centrifuged at 12,000 ϫ g for 1 h, and the supernatant was filtered through two layers of Miracloth.
Step 2: DEAE-cellulose Chromatography-Approximately 500 ml of DEAE-cellulose slurry pre-equilibrated with 50 mM potassium P i (pH 7.2) was added to the homogenate, stirred for 40 min, and filtered. The resin was resuspended in 2 liters of 50 mM potassium P i (pH 7.2), stirred for 30 min, and filtered. Then, 500 ml of 50 mM potassium P i (pH 7.2) was added to the resin, and the slurry was poured into a 5 ϫ 15-cm column. The column was washed with 500 ml of 50 mM potassium P i (pH 7.2) and eluted with a 2-liter gradient ranging from 50 to 300 mM potassium P i (pH 7.2). Component II elutes at ϳ200 mM potassium P i . Active fractions were pooled and concentrated to ϳ30 ml.
Step 3: Hydroxylapatite Chromatography-Protein from the previous step was diluted 20 -30-fold with cold 0.2 M KCl and loaded onto a 5 ϫ 14-cm HTP column pre-equilibrated with 5 mM potassium P i (pH 7.2) containing 0.2 mM KCl. The column was washed with 200 ml of 5 mM potassium P i (pH 7.2) containing 0.2 M KCl and eluted with a 2-liter gradient ranging from 5 to 100 mM potassium P i (pH 7.2) containing 0.2 M KCl (in both buffers). Component II activity elutes at ϳ30 mM potassium P i . The active fractions were pooled and concentrated to ϳ20 ml.
Step 4: Phenyl-Sepharose Chromatography-The protein from the previous step was brought to 0.4 M ammonium sulfate and loaded onto a 2.5 ϫ 18-cm phenyl-Sepharose column. The column was washed with 200 ml of equilibration buffer containing 50 mM potassium P i (pH 7.2) and 0.4 M ammonium sulfate. Component II activity elutes in the wash. Active fractions were collected and concentrated to ϳ5 ml.
Step 5: Bio-Gel P-60 Chromatography-The concentrated solution from the previous step was loaded onto a 2.5 ϫ 90-cm Bio-Gel P-60 column. The column was eluted with 50 mM potassium P i (pH 7.2) containing 0.15 M KCl. The active fractions were pooled, concentrated to ϳ1 ml, and stored at 4°C.

Assay for Component I and II Activities
Component I and II activities were monitored indirectly by following the activity of methionine synthase. Methionine synthase was purified as described previously (24). The anaerobic NADPH-dependent assay (15,25) was employed, in which the reductive activation system for methionine synthase was reconstituted by adding fractions to be tested for component II activity to an assay mixture containing microsomes (0.6 mg of total protein) with component I activity and purified methionine synthase. Under these conditions, the activity of methionine synthase is limited by the presence of component II. During the puri-

Reconstitution of Methionine Synthase Activity with Purified Cytochrome b 5 and Either Microsomes or Cytochrome P450 Reductase
The activity of methionine synthase was reconstituted under in vitro conditions using full-length purified recombinant rat cytochrome P450 reductase, which was a generous gift from Thomas Shea and Bettie-Sue Masters (University of Texas Health Science Center, San Antonio, TX). The anaerobic assay mixture contained, in a final volume of 1 ml, 100 mM potassium P i (pH 7.  Fig. 7B. The assay was initiated and conducted as described previously (25).

Mass Spectroscopy
Electrospray mass spectrometry was performed on a PLATFORM spectrometer (Micromass) at the Department of Chemistry, University of Nebraska. The protein samples were prepared for mass spectroscopy by exchanging the buffer with water by repeated dilution and concentration of the sample in a Microcon-3 filter (Amicon, Inc.). The protein concentration of the final sample was ϳ1 mg/ml.

Peptide Sequencing
The protein band was excised from a 15% SDS-polyacrylamide gel and submitted to the University of Nebraska Protein Core Facility for N-terminal and internal peptide sequence analysis.

RESULTS
Purification and Identification of Component II-We have employed a combination of anion-exchange, hydrophobic, and size-exclusion chromatographic steps to purify component II (Table I). Enzyme activity eluted in a broad peak following anion-exchange chromatography on a DEAE-cellulose column (Fig. 3A). Chromatography on two other columns, hydroxylapatite and Bio-Gel P-60 gel filtration, resulted in elution of a single and relatively sharp band containing component II activity (Fig. 3, B and C). Hydrophobic chromatography on a phenyl-Sepharose column resulted in a convenient negative purification since component II bound weakly to the column under the conditions employed. The four-step purification of component II yielded a single protein with a molecular mass of ϳ14 kDa as estimated on a denaturing SDS-polyacrylamide gel (Fig. 4A). Separation on a calibrated size-exclusion column yielded a slightly larger estimate of 17 kDa for the native protein (data not shown). Hence, component II is a small monomeric protein.
The identity of component II was established unambiguously to be soluble cytochrome b 5 by four independent methods. First, component II comigrated with purified soluble human cytochrome b 5 and cross-reacted with antibodies raised against it (Fig. 4B). Second, the molecular mass of purified component II is 10,977 Ϯ 0.81 Da (Fig. 5), which corresponds exactly to the calculated mass for porcine cytochrome b 5 extending from residues 2 through 97 in which the initiator methionine is re-moved and the N-terminal residue is acetylated (26). The apparent discrepancy between the molecular mass estimated by chromatographic techniques and the actual mass has been noted previously for soluble cytochrome b 5 (27). Third, we obtained the following amino acid sequences for two high pressure liquid chromatography-purified peptides generated by partial tryptic digestion of component II: TFIIGELHPDDR and QAGGDAXENFE, which exactly match the published sequence for porcine cytochrome b 5 (26). Two attempts were made to obtain the N-terminal sequence of isolated component II prior to its identity being known. Both were unsuccessful, consistent with the N-terminal residue being acetylated.
Finally, following the HTP column step, an obvious correlation between red color and component II activity was consistently observed, indicating that it is a hemeprotein. As shown in Fig. 3 (B and C), the activity of component II coeluted with heme absorption monitored at 413 nm. Purified component II has a typical oxidized cytochrome b-like spectrum, with a Soret absorption maximum at 413 nm and a broad absorption band in the 550 nm region (Fig. 6). Reduction with dithionite shifts the Soret band to 423 nm and sharpens the ␣ and ␤ bands at 555 and 525 nm, respectively. The ␣ band is asymmetric with a shoulder at 560 nm (Fig. 6, inset), which is a characteristic of reduced cytochrome b 5 (27). The presence of a b-type heme was also independently confirmed by the pyridine hemochrome assay (Ref. 28 and data not shown).
Kinetics of Activation-The anaerobic assay employed to monitor component II (hereafter referred to as soluble cytochrome b 5 ) activity during its purification is limited by its concentration. Fig. 7A shows that the activity of methionine synthase was reconstituted by soluble cytochrome b 5 in a saturable manner. A Michaelis-Menten analysis of this data yielded a K act for cytochrome b 5 of 0.19 Ϯ 0.05 M and a V max of 6.34 Ϯ 0.44 nmol min Ϫ1 . Based on the concentration of the methionine synthase used in these experiments, we estimate that soluble cytochrome b 5 is in 20-fold excess at maximal activation. However, only a low proportion (ϳ30%) of the cytochrome b 5 that we isolated had bound heme. We presently do not know whether or not the heme is required for the functioning of cytochrome b 5 in this activation system, as discussed below.
Microsomal Localization of Component I-Initial attempts at purifying component I activity from the soluble fraction were unsuccessful. Two observations indicated that component I is membrane-associated. First, component I activity could be precipitated from the liver homogenate at 25% ammonium sulfate saturation, which is typical for many membrane-associated proteins. Second, fractions containing component I activity from pilot chromatography columns that were tested initially were turbid and whitish. The microsomal location of component I was confirmed by the co-purification of glucose-6-phosphatase and component I activities (data not shown).
Reconstitution of Methionine Synthase Activity with Either Microsomes or Purified Cytochrome P450 Reductase and Soluble Cytochrome b 5 -The activity of methionine synthase in the presence of soluble cytochrome b 5 was dependent on the pres- ence of isolated microsomes in a saturable manner (Fig. 7B). The microsomal fraction could be substituted for purified fulllength cytochrome P450 reductase (Table II). NADPH did not support methionine synthase activity in the absence of either soluble cytochrome b 5 or cytochrome P450 reductase. Addition of both components reconstituted methionine synthase activity fully based on the activity obtained in the standard in vitro assay conducted in the presence of dithiothreitol and hydroxycobalamin.

DISCUSSION
Mammalian methionine synthase has been the subject of intense interest since it is one of the two cellular enzymes that ponent II activity and heme absorbance on the HTP and Bio-Gel P-60 columns are indicated in B and C, where the open circles represent the absorption data, and the closed circles represent the activity data. MS, methionine synthase. control homocysteine metabolism in cells. It catalyzes two successive methyl transfer reactions from the substrate CH 3 -H 4 folate to the cofactor cob(I)alamin and from methylcobalamin to the second substrate, homocysteine (Fig. 2). The reactivity of the cob(I)alamin intermediate results in its inadvertent escape to an oxidized and inactive enzyme form approximately once every 100 -2000 turnovers depending on the assay conditions (20,29). The oxidized enzyme is rescued to the catalytic cycle in a reductive methylation reaction that employs S-adenosylmethionine and an electron source.
The physiological pathway for electron transfer in bacteria is well studied and involves two flavoproteins, NADPH-dependent flavodoxin reductase and the proximal reducing partner, flavodoxin (19,20). In mammals, which apparently lack flavodoxin, the nature of the physiological reducing system was unknown. Partially purified porcine methionine synthase was reported to have an associated thiol oxidase activity (30) that was postulated to function in the reductive activation of the mammalian enzyme. However, we found an inverse correlation between specific thiol oxidase and methionine synthase activities during purification of the porcine enzyme, ruling out the presence of an inherent thiol oxidase activity (15).
Early studies on reconstitution of the physiological activation system of porcine methionine synthase in our laboratory revealed that at least two components were required for its activation (15). The first step in the purification of porcine methionine synthase involves anion-exchange batch chromatography and results in the separation of two other fractions that are required to support its activity in the anaerobic NADPH assay.
In this study, we have employed a biochemical reconstitution strategy for elucidating the cellular localization of the two components and have purified one of them to homogeneity. A similar approach has been used previously for identifying activation system components, for instance, for the anaerobic ribonucleotide reductase (31). Chromatography of the crude porcine liver homogenate on DEAE-cellulose yielded two fractions, one of which was turbid and milky in appearance. The particulate nature of this fraction indicated that component I was membrane-associated. The co-purification of component I and glucose 6-phosphatase activities confirmed its microsomal localization.
The second component, cytochrome b 5 , is soluble and in fact co-purifies with methionine synthase during the first three steps of the published protocol for purifying methionine synthase (24). This accounts for the early observation that a cytochrome b 5 with an intense absorption at 413 nm is a "major contaminant of the transmethylase" (32). In the purification protocol reported here that was optimized for soluble cytochrome b 5 , methionine synthase activity was separated in the second step following chromatography on the HTP column. The isolated protein has been unambiguously identified as soluble cytochrome b 5 by several independent methods including precise molecular mass and peptide sequence determinations. Electrospray mass spectroscopy revealed a molecular mass of 10,977 Da, which corresponds to a soluble cytochrome b 5 extending from residue 2 through 97. It is well known that the N-terminal methionine residue of cytochrome b 5 is removed and that the second amino acid (alanine in porcine cytochrome b 5 ) is acetylated post-translationally. The last two C-terminal residues in the soluble porcine cytochrome b 5 are Glu and Ser (26). The molecular mass of the isolated protein reveals that in addition to the N-terminal methionine, the C-terminal serine is also removed in the protein that we have isolated.
Cytochrome b 5 is a well studied hemeprotein that exists in membrane-associated and soluble forms. The cDNA encoding the soluble form has a 24-base pair insert that contains an in-frame stop codon and leads to the translation of a 98-residue-long polypeptide (26). The soluble form of cytochrome b 5 functions in methemoglobin reduction in erythrocytes (33). The membrane-associated form has a C-terminal hydrophobic membrane anchor (34) and is a component of electron transfer chains leading from NADH-dependent cytochrome b 5 reductase (35) or NADPH-dependent P450 reductase (36). Our observations that (i) either NADPH or NADH can serve as the ultimate electron donor for activation of porcine methionine synthase and that (ii) one of the two components in the electron transfer pathway to methionine synthase is microsomal suggest that either the cytochrome P450 reductase or cytochrome b 5 reductase is the other redox partner of porcine methionine synthase.
To test this hypothesis, we have examined the dependence of methionine synthase activity on purified microsomes and on purified full-length cytochrome P450 reductase. Both support methionine synthase activity, which also requires the presence of soluble cytochrome b 5 for activity in the NADPH-driven assay (Table II). We have considered the possibility that the soluble cytochrome b 5 that we have isolated represents a proteolytically clipped form of the membrane-associated protein.
However, this seems to be highly unlikely based on our results. If the membrane-associated cytochrome b 5 was needed, then microsomes by themselves, containing this form of cytochrome b 5 , would be sufficient for reconstitution of methionine synthase activity. As shown in Table II, this is clearly not the case. We have confirmed the presence of membrane-associated cytochrome b 5 in the microsomes that we have isolated by Western analysis (data shown).
Interestingly, using a genetic homology-based polymerase chain reaction approach, a gene encoding a putative methionine-synthase reductase with 38% identity to the human cytochrome P450 reductase has been isolated recently (37). Two mutations in this gene have been identified in cblE cell lines. The open reading frame in the isolated cDNA is postulated to encode a soluble homolog of cytochrome P450 reductase, although the assignment of the initiator codon has not been confirmed. Thus, a leader sequence targeting methionine-synthase reductase to the membrane cannot be ruled out. If on the other hand, this gene product is indeed capable of reducing methionine synthase directly, it would represent an alternative and parallel route to the pathway that we have described here (Fig. 8). Multiplicity of reductive activation pathways for methionine synthase may explain why cell extracts from cblE cell lines always show methionine synthase activity in the NADPH assay (15), whereas some cblG cell lines, in which methionine synthase is directly affected, have undetectable enzyme activity. At least under in vitro assay conditions, methionine synthase activity is completely dependent on the presence of a reductive activation system. A meaningful comparison of methionine-synthase reductase and of the reductive activation pathway described here will have to await biochemical characterization of the putative methionine-synthase reductase and demonstration that it can indeed activate methionine synthase directly.
The choice of cytochrome P450 reductase for reductive activation of methionine synthase in mammals may not be accidental. The bacterial flavodoxin and flavodoxin reductases share homologies with the FMN-and FAD/NADPH-binding domains of cytochrome P450 reductase, respectively (38). Thus, during evolution, the two domains may have fused to generate a single protein (39). This was apparently accompanied by the relocation of the flavoprotein from the cytoplasmic to the microsomal compartment. Methionine synthase, on the other hand, is a soluble protein in both bacterial and mammalian cells. Hence, nature may have recruited soluble cytochrome b 5 as an adaptor protein to bring an electron donor anchored in the membrane together with an electron acceptor in the cytoplasm (Fig. 9).
It is interesting to note that the bacterial flavoproteins flavodoxin and NADPH-flavodoxin reductase are able to reduce cytochrome P450, the target of cytochrome P450 reductase (38). Cytochrome b 5 was found to stimulate 17␣-hydroxypregnenolone lyase while inhibiting cytochrome P450 c17 -progesterone 17␣-hydroxylase activity when the bacterial flavoproteins were employed (40). We have also found that the E. coli flavodoxin and flavodoxin oxidoreductase at high concentrations can activate porcine methionine synthase in an NADPH-de- b This assay represents the standard in vitro assay and was performed under aerobic conditions as described previously (25). NADPH was replaced by the artificial reductants dithiothreitol and hydroxycobalamin, and the redox proteins cytochrome P450 reductase and soluble cytochrome b 5 were omitted. pendent reaction. 2 The effect of cytochrome b 5 and the characteristics of this heterologous activation system will be further examined in the future.
Severe thermodynamic constraints limit the direct involvement of cytochrome b 5 in electron transfer in a pathway leading from NADPH to methionine synthase. The redox potential of cytochrome b 5 is positive, in the range of Ϫ20 mV to ϩ20 mV (41)(42)(43). The redox potential for the cob(II)alamin/cob(I)alamin couple has been estimated to be Ϫ490 mV (44) and Ϫ520 mV (45). Activation of oxidized methionine synthase is postulated to occur via a thermodynamically uphill reduction of cob(I-I)alamin to cob(I)alamin rendered favorable by the irreversible trapping of cob(I)alamin in an exergonic methylation reaction (Equation 1), where AdoMet is S-adenosylmethionine and MeCbl is methylcobalamin. Coupling of reduction to a highly favorable alkylation reaction explains why reductive methylation of methionine synthase is observed even at an ambient potential as high as Ϫ82 mV (45). Thus, at least in principle, cytochrome b 5 could serve as a proximal electron donor to the porcine methionine synthase. This would be facilitated if the redox potentials of either the heme or cobalamin (or both) are significantly perturbed in a methionine synthase-cytochrome b 5 complex. However, it is unlikely that the magnitude of the shift, if any, would span the almost 0.5 V difference between the two redox couples. Instead, soluble cytochrome b 5 may function to facilitate complex formation between the microsomal redox donor and its soluble redox acceptor, methionine synthase. A similar role for the membrane-associated cytochrome b 5 has recently been revealed in two related systems. Cytochrome b 5 was postulated to donate the second electron to cytochrome P450 during the catalytic cycle (46). A similar discrepancy between the redox potential of the putative donor (cytochrome b 5 ) and acceptor (e.g. cytochrome P4503A4, approximately Ϫ310 mV) pertains to this system (47). It has been demonstrated recently that apocytochrome b 5 can replace holocytochrome b 5 in augmenting the oxidation activity of cytochrome P4503A4 and the 17,20-lyase activity of cytochrome P450c17 (47,48). These studies demonstrate unambiguously that the role of cytochrome b 5 in these systems does not depend on its redox activity.
In summary, we have purified soluble cytochrome b 5 , which is one of two proteins in the two-component system for activation of porcine methionine synthase. The NADPH oxidoreductase in this redox pathway is associated with the microsome and could be the well studied cytochrome P450 reductase whose three-dimensional structure has recently been determined (49). This hypothesis is supported by the reconstitution of methionine synthase activity by purified cytochrome P450 reductase and soluble cytochrome b 5 . A number of catastrophic mutations in enzymes that either directly (viz. cystathionine ␤-synthase and methionine synthase) or indirectly (methylenetetrahydrofolate reductase) influence homocysteine homeostasis have been described. These are rare, however, and are inherited as autosomal recessive inborn errors of metabolism. Over the past few years, there has been an intense interest in identifying polymorphisms in genes that influence homocysteine levels and that either alone or in concert with the environment, viz. nutritional status, may constitute a risk for hyperhomocystinemia. A fairly prevalent polymorphism in methylenetetrahydrofolate reductase has been identified that may be correlated with mild hyperhomocystinemia (50). Soluble cytochrome b 5 represents an additional and newly identified target in the homocysteine metabolic pathway for assessing correlations between potential genetic polymorphisms and propensity for cardiovascular diseases. acknowledge the generous gift of purified full-length cytochrome P450 reductase from Thomas Shea and Bettie-Sue Masters.