Activation of Methionine Synthetase by a Reduced Triphosphopyridine Nucleotide-dependent Flavoprotein System*

SUMMARY Two flavoproteins, both of which are required for the TPNH-and adenosylmethionine-dependent activation of the Bll-containing methionine synthetase, have been purified to homogeneity from Escherichia coli K-12. The larger flavoprotein (mol wt 27,000) contains 1 mole of noncovalently bound FAD per mole of protein and has an atypical spectrum (X,,, at 400 and 456 nm). The smaller flavoprotein (mol wt 19,400) is acidic, contains 1 mole of noncovalently bound FMN per mole of protein, and has absorbance maxima at 369 and 465 nm. The methionine synthetase, which was also purified to homogeneity from E. coli K-12, contains 1 mole of B12 per mole of protein (mol wt 186,000) and has an absorbance maximum at 474 nm. In the presence of TPNH, both flavoproteins, and adenosylmethionine, the synthetase has a specific activity at 37” of 3.8 pmoles per rnin per mg of protein with respect to formation of methionine from 5-methyltetrahydrofolate and homocysteine. Methionine

From the Department of Biochemistry, Scripps Clinic and Research Foundation, La Jolla, California 92037 SUMMARY Two flavoproteins, both of which are required for the TPNH-and adenosylmethionine-dependent activation of the Bll-containing methionine synthetase, have been purified to homogeneity from Escherichia coli K-12.
The larger flavoprotein (mol wt 27,000) contains 1 mole of noncovalently bound FAD per mole of protein and has an atypical spectrum (X,,, at 400 and 456 nm).
The smaller flavoprotein (mol wt 19,400) is acidic, contains 1 mole of noncovalently bound FMN per mole of protein, and has absorbance maxima at 369 and 465 nm. The methionine synthetase, which was also purified to homogeneity from E. coli K-12, contains 1 mole of B12 per mole of protein (mol wt 186,000) and has an absorbance maximum at 474 nm. In the presence of TPNH, both flavoproteins, and adenosylmethionine, the synthetase has a specific activity at 37" of 3.8 pmoles per rnin per mg of protein with respect to formation of methionine from 5methyltetrahydrofolate and homocysteine.
Methionine synthetase (5-methyltetrahydrofolate-homocysteine cobalamin methyltransferase), which catalyzes Reaction 1, 5-Methyltetrahydrofolate + homocysteine + tetrahydrofolate + methionine (1) has been purified extensively from several strains of Escherichia coli (l-3), from porcine liver (4), and from bovine brain (5): As isolated, however, these enzymes are inactive unless the assay system is supplemented with a reducing system and a catalytic amount of S-adenosylmethionine; these accessory factors appear to be required for the reductive methylation of the protein-bound Blz moiety (&9).
In previous investigations involving the purification of methionine synthetases or studies of the mechanism of Reaction 1, chemical reductants (e.g. FMNH2 plus dithiols) or various DPNH-dependent flavoproteins have been employed to activate * This work was supported by grants from the National Cancer Institute, National Institutes of Health (CA 6522), the American Cancer Society (BC-62), and the American Heart Association (69-1082). the enzyme (2, 3, l&15).
Although these chemical and enzymatic reducing systems have been useful empirically, their nonspecificity and high concentrations have prompted us to search for a more efficient enzymatic system in cell lysates from which the synthetase is purified.
The present paper describes a new, TPNH-dependent system from E. coli K-12 which meets this criterion.
It consists of two flavoproteins (designated R and F components), both of which have been purified to homogeneity. The B12-containing methionine synthetase (designated M component), which has also been purified to homogeneity from E. coli K-12, catalyzes Reaction 1 at an optimal rate when supplemented with TPNH, AdoMet,' and the R and F components.
Prior to chromatography, each column was equilibrated with the starting buffer. E. coli K-12 cells, grown on B12 were obtained from Grain Processing Corp. B1ja and &L-5-methvltetrahvdrofolate were kindlv " supplied bi Drs. D. Jacobsen and j. Whiteley, from this Departmerit. 5-Methyltetrahydrofolate (5-"CH3) (60&i per pmole) was purchased from Amersham-Searle; stock solutions of this material (20 mM, 350 to 600 cpm per nmole) were stored frozen under argon in Thunberg tubes. Homocysteine was prepared from the 6746 thiolactone by the procedure of Hatch et al. (16). Solutions of homocysteine (0.1 M) and dithioerythritol (0.2 M) were stored at 4" under argon in Thunberg tubes.

Methods
Absorbance spectra were obtained with a Cary recording spectrophotometer, model 14. Electrophoresis of proteins on polyacrylamide gel w&s carried out according to the methods of Ornstein (17) and Davis (18). Staining was achieved with Amido black, followed by electrophoretic destaining using a Canalco rapid destainer.
Analysis of flavins was performed by (a) thin layer chromatography (19) on SIL S plates with 5yo NasHPOd as the solvent system or on CEL plates with butanol-l-water-acetic acid-methanol (7:7:0.5:3) or l-butyl alcohol-water (6:4) as the solvent systems; and (b) paper electrophoresis on Whatman No. 3 MM paper with a 0.05 M sodium acetate buffer, pH 5.1, as the solvent system (20). Flavins were located by their fluorescence under ultraviolet light.
Molecular weights of proteins were determined by gel filtration according to Whitaker's method (21) or by electrophoresis on sodium dodecyl sulfate polyacrylamide by means of the procedures of Weber and Osborn (22) and Dunker and Rueckert (23). In the gel filtration method, Bio-Gel P-200, Sephadex G-75, and Sephadex G-50 were used for the M, R, and F components, respectively. Pyruvate kinase, y-globulin, alcohol dehydrogenase, and dihydrolipoate dehydrogenase served as standard proteins for the Bio-Gel P-200 column, while cytochrome c, myoglobin, a-chymotrypsinogen, trypsin, and pepsin served as standards for both Sephadex columns. For all three columns, the void volume was determined with blue dextran 2000. In the sodium dodecyl sulfate polyacrylamide gel method, proteins were stained with Coomassie blue.
The M component was examined in a 5Q/o polyacrylamide gel prepared according to Dunker and Rueckert (23). For the It and F proteins, 10% polyacrylamide gels with 0.75 of the normal amount (22) of cross-linking were employed.

M. Component
Methionine formation from 5-methyltetrahydrofolate and homocysteine, catalyzed by the M component, was measured by the following methods which are based upon the procedure of Weissbach et al. (24).
Reactions were performed in conical tubes (1.2 X 11.5 cm) equipped with two side arms and sealed with a rubber stopper. To ensure uniform deoxygenation, the tubes for each experiment were connected in series with Tygon tubing and then evacuated and flushed with argon several times; during these manipulations the tubes were kept in an ice-water bath. The reactions were initiated by placing the tubes in a 37" bath, and, after 15 min, the reactions were terminated by returning the tubes to the ice bath. The tubes were then evacuated and flushed with air several times. Cold water (0.25 ml) was added to each tube, and the contents were applied to small (0.5 X 4.0 cm) columns of Bio-Resin AG l-X8.
The tubes were washed with two 0.5-ml portions of cold water, and the washings were also passed through the columns. Samples (0.5 ml) of the effluents were added to 10 ml of scintillation fluid (100 g of naphthalene and 5 g of 2,5diphenyloxazole (PPO) per liter of dioxane), and radioactivities were determined with the use of a liquid scintillation spectrometer, Beckman model LS233. Results were corrected for blanks in which the M protein was omitted.
In accordance with the procedure of Taylor and Weissbach (2), one unit of activity was defined as the amount of enzyme required for synthesis of 1 rmole of methionine per 15 min. Specific activity was expressed as units per mg of M protein.
Protein in crude preparations was measured by the biuret method (25) with bovine serum albumin as the standard; in more purified preparations, the Lowry method (26) was used.
Assay B: Chemical Reducing System-This was similar to the enzymatic system, except that TPNH and the It and F components were replaced by 6 rmoles of dithioerythritol.

R Component
The activity of the R component was determined either by its DPNH-cytochrome c reductase activity or its ability in the presence of TPNH (or DPNH) and the F component to support methionine synthesis catalyzed by the M component.
For measurement of the cvtochrome c reductase activity. the assay mixture contained, in 5.5 ml, 150 pg of cytochrome &.0.30 lrmole of DPNH. 25 Lamoles of KPB. and the indicated amount of 'R component; tie biank was identical except for the omission of enzyme. The absorbance change was monitored at 550 nm with the use of a Gilford multiple sample absorbance recorder at 37". Results are expressed as change in absorbance (AAsso per min) during the initial, linear portion of the curve (0.5 to 2 min). One unit of activity was defined as the amount of enzyme required to produce a ~~~~~ of 1.0 per min. Specific activity was expressed as units per mg of protein.
The activity of the R component in support of methionine synthetase was measured by means of Assay System A with fixed amounts of M and F and the R component as a limiting factor. Under those conditions, one unit of R activity was defined as the amount required for the synthesis of 1 *mole of methionine per 15 min. Specific activity was expressed as units per mg of R protein.

F Component
This component was also measured under conditions in which it was the limiting factor in Assay A. The volume of the assay mixture was 0.25 or 0.125 ml, but the concentrations of all reactants were unchanged.
Activity and specific activity were defined in a manner similar to that described above for the R component.

Purification of Proteins
In the following procedures, all operations were carried out at approximately 5". Centrifugations were performed using a Sorvail centrifuge with GSA head for 60 to 90 min at 14,600 X g. During chromatography, fractions were collected automatically. Solutions were concentrated by pressure dialysis under argon with the use of a Diaflo apparatus (Amicon) with indicated membranes.

M Component
Frozen cells (664 g, wet weight) of E. coli K-12 were suspended in 5 volumes of 0.05 M KPB and lysed with a Branson Sonifier, model S-125, at a setting of 5 for two 5-min periods.
During sonication the temperature was kept below 8". The homogenate was centrifuged, and to the supernatant (34.5 g of protein) was added dropwise, with stirring, 293 ml of 20/o protamine sulfate (adjusted to pH 7.5 with 1 N KOH).
After being stirred for 30 min at 4", the mixture was centrifuged and the residue discarded. Solid ammonium sulfate was added slowly, with stirring, to the supernatant (313 g per liter, 50yo saturation). The suspension was then stirred for an additional 30 min and centrifuged.
The supernatant was saved for isolat,ion of the 11 or F components, as described below. The precipitate, which contained the M component, was dissolved in 0.14 M KPB and dialyzed against three changes of the same buffer.
In the followine steos. AdoMet-chloride was added to all solutions at a final coicen&ation of 1 &M. The dialyzed solution from the 0 to 50% ammonium sulfate precipitate was applied to a column (4 X 52.5 cm) of DEAE-cellulose.
The column was washed with 750 ml of 0.14 M KPB, followed by gradient elution with 1 liter of 0.15 M KPB in the mixing chamber and 1 liter of 0.45 M KPB in the reservoir.
The flow rate was 34 ml per hour, and 16.2-ml fractions were collected.
Fractions containing the M component (tubes 49 to 72) were combined and concentrated by pressure dialysis (PM-10 membrane).
The resulting solution  System-In agreement with the results of previous investigations (l--5), the M component isolated by the present procedure was unable to catalyze Reaction 1 in the absence of a reducing system (Fig. 4). Addition of a chemical reducing system such as dithioerythritol plus B1za activated the methionine synthetase, and under these conditions product formation was proportional to the concentration of M. The specific activity of M, calculated from these data, was 29 pmoles of methionine synthesized per 15 min per mg of protein, or 1.9 pmoles per min per mg. The activity of M was increased considerably when the 6748 chemical system was replaced by the present TPNH-dependent The concentration dependence of R and F for activation of the enzymatic system (Fig. 4). Activity was again proportional to the concentration of M, and the specific activity of the latter was methionine synthetase is illustrated in Fig. 5. In each instance 57 pmoles per 15 min per mg of protein, or 3.8 pmoles per min per the amount of M was fixed, TPNH and one of the flavoproteins w.
were present in excess, and the remaining component was varied A component study for methionine synthesis supported by the as indicated.
Under the conditions of these experiments, maxienzymatic reducing system is summarized in Table I. The mum activity was obtained with approximately 0.2 and 0.1 /~g following points are apparent: (a) Omission of any of the oxida-of the R and F components, respectively. In the 0.25-ml assay tion-reduction components (R, F, or TPNH) caused the activity mixture, these amounts corresponded to FAD and FMN concentrations of 0.030 and 0.020 pM. of M to decrease markedly.
The residual activities were due, in Putijication and Properties of R Component-The ability of the part, to the presence in the reaction mixture of homocysteine and R and F components to support methionine synthesis (cf. Fig. 5) Bit,, which together provide a weak chemical reducing system. was used to monitor these proteins during their purification.
(b) Omission of other components of the assay system, such as Because this activation is a catalytic rather than stoichiometric adenosylmethionine and homocysteine, also produced a marked diminution of activity, verifying that the characteristics of methi-process, it was necessary to define arbitrarily the activity of the limiting component (R or F) in terms of methionine synthesized onine synthesis are not changed when M is activated by the by a given amount of $1 component. The cytochrome c reenzymatic, rather than a chemical, reducing system. (c) Replacement of TPNH by DPNH resulted in no loss of activity,  Table IV) and 0.8 mg of a partially purified preparation of F (pooled Fractions 30 to 50, from the Sephadex G-75 column described in Step 3 of this Table).
6 From 648 g of cells. d -, not determined.
ductase activity of the R component also provided an independent (and absolute) measure of its specific activity. The parallelism between the results of the two assays for R component activity during the various steps in the purification procedure is shown in Table II. Although TPNH is more efficient than DPNH in both the R-dependent reduction of cytochrome c and in the enzymatic activation of methionine synthetase, DPNH was used for routine assays.
The 50 to 85% ammonium sulfate fraction of the cell-free extract served as the starting material for purification of the R component.
A sequence of steps, described under "Experimental Procedures," separated R from other flavoproteins and led to a homogeneous preparation.
During purification, care had to be taken not to confuse the R component with other cytochrome c reductases which also appeared in the elution profiles (cf. Fig. 1). It should be noted that none of these other cytochrome c reductases can substitute for R in the activation of the M component.
The final chromatographic step in the purification of the R component was characterized by a somewhat asymmetric elution profile with respect to the catalytic activities and protein. The highest specific activities and homogeneous protein (see below) were found in Fractions 38 to 44. The data in Table II indicate that the enzyme in these fractions was purified about 300-to 360-fold as judged, respectively, by the cytochrome c and methionine synthetase assays. Additional enzyme having a slightly lower specific activity was found in Fractions 34 to 37 and'45 to 54. The over-all yield of R component in Fractions 34 to 54 was about 11 To. This reasonably good recovery is attributable to the stability of R during purification. It is, likewise, very stable to storage; full activity was retained after 10 months at -20" under argon. Homogeneity of the R component was demonstrated by electrophoresis on polyacrylamide (Fig. 6). Prior to protein staining, the material was visible as a green band at Rp = 0.62. The molecular weight of the R component was determined by two procedures. Gel filtration through a standardized column of Sephadex G-75 gave a value of 26,500, which was in very good agreement with that of 27,500 from sodium dodecyl sulfate polyacrylamide electrophoresis. Based upon these results, 27,000 was chosen as the best value for the molecular weight of the R component. The general procedure for separating and staining the proteins is given under "Experimental Procedures." A 20.rg quantity of each protein was applied to the gels.
The absorbance spectrum of the R component at neutral pH (Fig. 7) is unusual for a flavoprotein. Maxima occur at 274,400, and 456 nm, with a shoulder at 480 nm; millimolar extinction coefficients calculated from the data in Fig. 7 and a molecular weight of 27,000 are 48.5, 7.4, 7.1, and 6.0, respectively, for these peaks. Heat denaturation (6 min at 97") of R released the flavin, and the spectrum of this chromophore (shown also in Fig. 7) was normal, i.e. maxima at 375 and 450 nm. Thin layer and paper chromatography identified this flavin as FAD. From the data in Fig. 7, and using 11.3 as the millimolar extinction coefficient for FAD (27), the FAD content was calculated to be 0.89 mole per mole of protein.
Purijication and Properties of F Component-During purification, the F component could be assayed only by its ability to activate methionine synthetase. Purification of F from a 50 to 85% ammonium sulfate fraction of the cell-free extract was accomplished by the sequence of steps described under "Experimental Procedures." In the DEAE-cellulose chromatographic step, a brown ferredoxin band (measured by absorbance at 416 nm) preceded the F component and served as a convenient marker 6750  7 (Zejt). Absorbance spectrum of It component.
-, holoenzyme (0.685 mg per ml in 0.08 M phosphate buffer, pH 7.0). ---> chromophore released by heat denaturation of It (0.2 ml of the above solution was heated at 97" for 6 min. Denatured protein w&s removed by filtration, washed twice with water, and the combined filtrate and washings were lyophilized. The residue was redissolved in 0.2 ml of water). FIG. 8 (right). Absorbance spectrum of F component.
---, chromophore released by heat denaturation of F (procedure similar to that in Fig. 7, except that the filtration step was omitted owing to absence of coagulated protein).
on the column. Activity in the methionine synthetase system coincided with flsvoprotein absorbance measured at 465 run. The final step, chromatography on hydroxylapatite, resolved the F component into two peaks. The initial peak (Fractions 41 to 47) in Fig. 3 contained material of higher specific activity and most of the total activity. The data in Table III indicate that the enzyme in these fractions is about 250-fold purified and that the recovery is about 10%. If enzyme in the second peak (Fractions 48 to 54) is also included, the over-all recovery is about 14oJ,. Like the R component, the F component is quite stable during purification and storage (e.g. full activity is retained after 10 months at -20').
Homogeneity of the F component was also demonstrated by electrophoresis on polyacrylamide (Fig. 6). At pH 8.3, the F component migrated very rapidly and overlapped the marker dye (Rp = 1.0) ; this behavior is consistent with its acidic nature and low molecular weight. The F component could be detected visually on the gels as a brown band (the result of mixing the flavoprotein with the marker dye).
The molecular weight of the F component was determined by the procedures used for the R component. Gel filtration through a standard column of Sephadex G-50 and electrophoresis on sodium dodecyl sulfate polyacrylamide gave values of 18,900 and 19,500, respectively; 19,400 was taken as the best value for the molecular weight of the F component.
The absorbance spectrum of the F component at neutral pH (Fig. 8) is that of a typical flavoprotein. Millimolar extinction coefficients calculated from the data in Fig. 8 and a molecular weight of 19,400 are 53.2 at 274 nm, 7.17 at 369 nm, and 8.42 at 465 nm. Heat denaturation of F released the flavin chromophore whose spectrum is also shown in Fig. 8. Thin layer and paper chromatography identified the flavin as FMN. From the data in Fig. 8, and using 12.5 as the millimolar extinction coefficient for FMN at 445 nm (27), the FMN content was calculated to be 0.97 mole per mole of pro&in.
Purification and Properties of M Component-Activity of the M component was measured most conveniently during purification by Assay B (see under "Experimental Procedures") in which activation was achieved by dithioerythritol plus Bns. This chemical reducing system also aids in the removal of oxygen which may remain after flushing with argon. It should be noted that BUZZ is not being used to convert apoenzyme of the M component b From 664 g of cells.
to the holoenzyme since replacement of Blz, by FMNH:! leads to no reduction in methionine synthetase activity. Purification of the M component was achieved by the procedure described under "Experimental Procedures." In the final step, i.e. chromatography on DEAE-Sephadex, selected fractions (54 to 60) yielded a homogeneous preparation of M as judged by several criteria (see below). At this stage, the enzyme had been purified about IOO-fold and was recovered in an over-all yield of about 3oJ, (Table IV).
The loss of activity observed during purification of the M component was consistent with its instability, particularly when present in dilute solution, during storage. Maximum stability was obtained by freezing more concentrated solutions (4 to 10 mg per ml) of the protein and storing them at -20" under argon; under these conditions, full activity was retained even after 1 year. If deoxygenation was incomplete, however, a gradual loss of activity, paralleled by an increase in absorbance of the preparation at 357 nm, was observed. AdoMet, which had been shown previously to protect the enzyme during ultracentrifugation or gel filtration (28), was found to accomplish the same purpose during purification and was included (at 1 PM), therefore, in all solutions after Step 2 in the procedure. In Steps 3 to 6 of the procedure, dilute solutions of the enzyme were concentrated by pressure dialysis rather than by ammonium sulfate precipitation, since the latter was attended by an apparent oxidation of the BU chromophore. A similar lability, induced by exposure to ammonium sulfate, has been reported for a B12-containing protein from Clostridium thermoaceticum (29).

6751
Homogeneity of the M component., following Step 6 in the purification procedure, was demonstrated by electrophoresis on polyacrylamide (Fig. 6). Even when larger amounts of protein (about 50 pg) were applied to the gel, only the single band at Rp = 0.41 was seen. This was visible, prior to protein staining, by its brown-orange color.
The molecular weight of the M component was determined by three different methods: (a) filtration through Bio-Gel P-200, and comparison of its position in the elution profile against those of various standard proteins, gave a value of 183,000; (b) electrophoresis on sodium dodecyl sulfate polyacrylamide, and comparison of its relative mobility with those of standard proteins, yielded a very similar value, viz. 187,000; and (c) conversion of the chromophore to the dicyano derivative, assuming millimolar extinction coefficients of 10.3 at 581 nm and 31.8 at 367 nm for the latter (averages of three previously reported values (30)(31)(32)), gave a slightly lower value, viz. 175,000. In view of some uncertainties inherent in the latter procedure, the first two methods were considered to be more reliable, and 186,000 was taken as the best value for the molecular weight of the M component.
The absorbance spectrum of the M component at neutral pH is shown in Fig. 9. Principal maxima occur at 278 and 474 nm with a small peak at 404 nm and a shoulder at 312 nm. The following millimolar extinction coefficients were calculated from the spectrum of the holoenzyme (Fig. 9) using a molecular weight of 186,000:194 at 278 nm, 30.1 at 312 nm, and 11.0 at 474 nm (ratio 17.6:2.74:1.00). When the M component was treated with KCN (final concentration 0.75 M) at pH 12.5, even in the absence of light, the spectrum of the solution changed to that of the dicyano form of a corrinoid (X,,, at 367, 541, and 581 nm). When the chromophore was removed by treatment of the M component with 80% ethanol at 82" for 30 min in the dark, followed by phenol extraction of the mixture and back extraction into water, the resulting solution had a spectrum (X,,, at 357, 510, and 545 nm) resembling that of Bu, (33).  Chemical reducing systems (e.g. combinations of thiols with reduced flavins or B12 compounds) have proved to be convenient for the assay of methionine synthetase activity during purification of the enzyme. They are less satisfactory, however, in mechanistic studies because of the danger of side reactions or masking of absorbance changes in the Blz moiety during the catalytic cycle. In order to overcome these difficulties and to obtain further insight into the mechanism of methionine synthesis, several laboratories have searched for an enzymatic counterpart to these reducing systems. The following DPNH-dependent systems from various strains of E. coli have been used for this purpose: (a) FAD reductase (10); (b) a 60 to 90% ammonium sulfate fraction (11) ; (c) a flavoprotein (mol wt 126,000) purified from the 28 to 43% ammonium sulfate fraction (3, 15); (d) a flavoprotein with dihydrolipoate dehydrogenase activity plus a low molecular weight, heat-stable protein (12); (e) two protein fractions, one of which was heat-stable (13) ; cf) a diaphorase (mol wt 68,900) (14); and (g) a dihydrolipoamide dehydrogenase (mol wt 112,000) (14).
These diverse observations emphasize the ability of methionine synthetase to utilize a variety of enzymatic, as well as chemical, reducing systems. However, even in these enzymatic systems, relatively large amounts of the oxidation-reduction proteins were required, Maximum stimulation of methionine synthetase was achieved, for example, when the flavoproteins were present in amounts corresponding to flavin concentrations of 10 to 20 PM. This is not much lower than the concentration (50 to 60 PM) of free, reduced flavin that would have sufficed. The present investigation was undertaken, therefore, to search for a more specific and efficient enzymatic reducing system. Fractionation of E. coli K-12 extracts by various techniques disclosed the presence of an apparently new system consisting of two separate fiavoproteins which, when supplemented with TPNH (or higher concentrations of DPNH), supported methionine synthetase with a high degree of efficiency.
The R component has an unusual absorbance spectrum for a flavoprotein, viz. an anomalous band at 400 nm (Fig. 7). However, when the prosthetic group was released by heat denaturation, it showed a typical flavin spectrum (Fig. 7) and was identified chromatographically as FAD. The anomalous 400-nm band thus appears to be due to interaction of the FAD with some group or groups on the R protein.
The F component has the characteristics of a flavodoxin (34), viz. low molecular weight, FMN as the prosthetic group, and acidic nature as evidenced by tight binding to anion exchangers and rapid mobility during electrophoresis. Vetter and Knappe (35) havereported that E. coli contains a flavodoxin whose molecular weight was found to be 13,500 to 14,500 by sodium dodecyl sulfate polyacrylamide gel electrophoresis and amino acid analysis, and 19,000 to 21,000 by gel filtration and sucrose gradient centrifugation.
The molecular weight found for the F component in this investigation (19,400) compares with the higher value reported by Vetter and Knappe.
The strongly acidic properties of the F component caused some difficulties in the initial purification of this protein. When preparations were chromatographed on DEAE-cellulose, elution with KPB alone removed most of the protein but none of the F activity; it was found that the latter could be eluted by including 0.6 M KC1 in the buffer. The acidic nature of F is probably also responsible for its association with other proteins. Thus, in a trial experiment, passage of a 50 to 85% ammonium sulfate fraction