Formiminotransferase-Cyclodeaminase Purification ELIZABETH
J. DRURY, Department
from of the Enzyme
E. MACKENZIE Y6. Canada
A simple procedure for the purification of the formiminotransferase-cyclodeaminase enzyme complex is described. The crystalline preparation is homogeneous by ultracentrifugation and electrophoresis and appears to he composed 01’ eight apparently identical subunits of about 6.4 x 10” daltons. Both enzyme activities migrate with the single protein hand on electrophoresis and it is proposed that the activities are probably associated with different sites on one type of polypeptide chain.
Histidine is degraded in mammalian tissues to produce formiminoglutamic acid which is then further metabolized by the sequential activities of formiminoglutamate: tetrahydrofolate formiminotransferase (EC 22.214.171.124) and formiminotetrahydrofolate cyclodeaminase (EC 126.96.36.199) to yield glutamic acid, ammonia and 5,10methenyl tetrahydrofolate. Metabolically the formiminotransferase and cyclodeaminase are involved in the genesis of one-carbon units, but in a more general context. they can be considered gluconeogenic activities since the overall pathway involves the degradation of histidine to glutamic acid. Recentlv it has been demonstrated that the activity of the formiminotransferase responds rapidly to hormones in uiuo in a fashion consistent with a gluconeogenie function. Both glucagon and adrenalin cause rapid increases in hepatic formiminotransferase activity. while insulin causes a decrease (1). In order to provide explanations at a molecular level for these observations, further characterization of the enzyme system is required. The transferase and cyclodeaminase copurify from acetone powders of hog liver, and some of the properties of these prepara1 This MA-4179
tions have previously been described (2, 3). In this paper we present a method for the routine purification of this enzyme complex from hog liver, without the use of acetone powder preparations. The enzyme complex has been crystallized. and is homogeneous by the electrophoretic and ultracentrifugal techniques used to examine its subunit structure. It seems highly possible that the transferase and cyclodeaminase activities exist in a single polypeptide chain. MATERIALS
Formiminoglutamate. hemibarium salt. was obtained from Sigma Chemical Co. and converted to the sodium salt by precipitation of the barium with sodium sulfate. dl-I.-Tetrah?-drofolic acid was prepared hy catalytic reduction of folic acid in neutral aqueous solution (4) using one atmosphere hydrogen and platinum as catalyst. The tetrahydrofolate was purified by chromatography on DEAF,-cellulose (!5) using 0.2 hl triet hanolamine hydrochloride 0.5 hf 2.mercaptoethanol pH7. for elution. 5.Formiminotetrahydrofolate was prepared enzymatically. The purified complex of formiminotransferase and cyclodeaminase was treated with chymotrypsin which destroys cyclodeaminase activity preferentially (2). This treated enzyme (5 mg) was incubated with 200 ml 01 10 rnM dl-I.-tetrahydrofolate and ,j rnM formiminoglutamate. The formiminotetrahydrofolate produced after :X1 min at rclom temperature was purified by column chromatography accordin:: to Rahinowitz (6). and obtained as a dry powder after
by operating grant Research Council of
662 Copyright 0 1975 by Academic Press, Inc. All rights of reproduction in any form reserved.
FIG. 1. Photomicrograph of crystals of the formiminotransferase-c?clociearninahe complex in 0.2 N sodium acetate pH 7.0. Exposures were made on Kodak PIUS X film at a magnification 01 160x using the Nomarski different ial interi’erence-c(,lltr~~~~ optic,: of a Zei+ \VI, microscope. TABLE PURIFICATION Fraction
Extract Polyethylene glycol Sodium acetate pH precipitation Crystallization u Specific
Volume (ml) 1990 645 18.2 5.9 10 for cyclodeaminase
OF FOKMIMINOTRANSFERASE/~TLODEAMINASE Protein (mg)
79.600 27.100 131 :38 19 in the purified
l?;ophilization. Trizma base was obtained from Sigma Chemical Co: specially puril’ied sodium dodecyl sulfate from Rritish Drug House: acrvlamide and methylene-his-acrylamide from Eastman Chemical Co.: polyethylene glycol 6000 was from Raker Chemical Co.; other chemicals were reagent grade. EnzJ,me (ISSQ~. Both formiminotransferase and cyclodeaminase were assayed by spectrophotometric measurement of 5,10-methylenetetrahydrofolate, the final product of the sequential reactions. The overall conversion of tetrahydrofolate to 5,10-methylenetetrahydrofolate measures transferase activity while the cyclodeaminase assay involves the conversion of added 5-formiminotetrahydrofolate to 5,10-methenyltetrahydrofolate. Formiminoglutamate-tetrahydrofolate formiminotransferase activity was measured in an assay mixture containing 0.1 M potassium phosphate pH7.3. 0.1 M 2-mercaptoethanol, 1 mM dl-L-tetrahy-
3930 ‘755 1430 1250 $06
Specific activity 0.05 0.10 11 33 37 (ll,Sl”
Percent yield 100 70 36 3” 18
complex drofolate. ,3 mM formimino-I.-glutamate and enzyme in a total volume of 1 ml. After incubation for 3 min at 30 ‘C. the reaction was stopped hy the addition of 1 ml of 0.:;6 M HCI or ‘7.5’~ trichloroacetic acid. To complete the conversion of 5.formiminotetrahydrofolate to .i.lO-methenyltetrahydrofolate the tubes were heated in a boiling water bath for 55 s (2) and cooled in ice. After centrifugation to remove precipitated protein (necessary only with crude fractions). the amount of product was determined by measuring the absorbance at 350 nm against a blank incubated in the same fashion and to which the enzyme was added after acidification. The amount of enzyme was adjusted to provide absorbance changes of 0.1-l.:! O.D. Formiminotetrahydrofolate cyclodeaminase was assayed as described by Tahor and Wyngarden (2). The incubation mixture contained 0.1 M potassium maleate pH 6.5. 0.1 mM formiminotetrahydrofolate
and 0.02 M 2.mercaptoethanol. The reaction was initiated by addition of enzyme and the rate of formation of 5,10-methenyltetrahydrofolate in an assay volume of 0.5 ml was monitored at 355 nm on a Gilford recording spectrophotometer at room temperature against a control which lacked enzyme. Enzyme activities are expressed as Mmol productimin. Protein determinations were by the method of Lowry (7) using bovine serum albumin as standard. Elwtrophoresis. Disc gel electrophoresis was performed according to Davis (8), or, with the addition of l’t Triton X-100 (9) and 8 M urea. Sodium dodecyl sulfate gel electrophoresis was carried out by the method of Weber and Osborne (10). Cr~lracentrifugation. Meniscus depletion sedimenTABLE EFFECT
tation equilibrium was used to determine molecular weights (11). and the determinations of partial specific volume (a) were obtained using H,O and D,O by the method of Edelstein and Schachman (12). Densities of solutions were measured with a 10 ml pycnometer. Enzvme purification. Fresh porcine liver was obtained at local meat packing plants and frozen at -20°C in 200-300 g portions. For a typical enzyme preparation 600 g of’ frozen liver was cut into pieces and homogenized for three minutes in a commercial Waring Blendor with three volumes of 0.1 M potassium phosphate buffer, pH 7.3, containing 0.05 M Z-mercaptoethanoI.The homogenate was centrifuged at 20,OOOg for 1 h and the supernatant solution was poured through glass wool to remove lipid particles. To this solution was added slowly, with stirring, 0.14 vol of a 50 percent (w/v) solution of polyethylene glycol-6,000. After 30 min, the suspension was centrifuged at 20,OOOg for 20 min, and the pellets discarded. An additional 0.j vol (based on the original supernatant solution) of 504’ polyethylene glycol was added and allowed to stand 30 min. Following centrifugation at 20.000~ for 20 min, the supernatant solution was discarded and the pellets dissolved in 500 ml 0.1 M potassium phosphate. pH 7.3. Ammonium sulfate (19.4 gYlO0 ml) was added slowly to this solution maintained at pH7 with concentrated ammonium hydroxide. After standing for one hour, the suspension was centrifuged at 20,OOOg for 60 min and the soluble fraction discarded. The pellet was resuspended with 10 ml 0.05 M mercaptoethanol. and centrifuged 30,OOOg for 1 h. The oily top layer was removed and a further 5 ml of water added and mixed well with the lower suspension. After centrifugation
ON FORMIMINOTFUNSFERASE ACTIVITV’
Potassium Tris hydrochloride Tris maleate Triethanolamine hydrochloride Triethanolamine sulfate Triethanolamine phosphate Imidazole hydrochloride Potassium maleate
100 5 7 100 115 115 9 40
” All buffers were 0.1 M in concentration at pH 7.3 and replaced phosphate in the assay medium. Aliquots enzyme (10 ~1) in triethanolamine sulfate were used to initiate the reactions.
0.8 :< * 0.6;
0 ; 0.4 z F
FIG. 2. Electrophoresis of enzyme complex (20 pg) on a 6% polyacrylamide gel in Tris-glycine. The gel was sliced longitudinally, and one half was stained for protein with coomassie blue and is shown at the top of the figure. The other half of the gel was sliced into sections which were minced with 0.5 ml 0.1 M phosphate pH 7.2 and dialyzed overnight to remove Tris. Aliquots were used to assay for formiminotransferase (solid line) and cyclodeaminase (dotted line) and the results are plotted as changes in absorbance.
acetate:acetic acid). The precipitate that formed was collected hy rentril’ugation and suspended in a small volume (5 ml) 0.1 M phosphate buffer, pH 7.2. After centrit’ugation. the clear solution was dialyzed overnight against 1 liter 0.2 M sodium acetate. pH 7.0. in order rl> crystallize the enzyme. The crystals were collected 1)~ centrilugation and dissolved in 10 ml 0.1 hl triethanolamine sulfate huft’er. pH 7.2. RESULTS
The purification procedure was designed to eliminate the use of acetone powder preparations of liver as the source of the enzymes. In preliminary studies it was noted that the yield of acti\.ity, based on the fresh weight of liver. was reduced to 2L50’~ by this delipidation procedure, and in addition it is possible that an enzyme isolated from such a preparation could have altered properties. However, omission of this step caused problems with fractionation by ammonium sulfate because of difficulties in forming compact pellets of the precipitate by centrifugation. This problem was overcome by first carrying out a fractionation of the crude extract with polyethylene glycol 6000. The material obtained in the resultant pellet. when redissolved. is fractionated easily by ammonium sulfate precipitation. The remainder
FIG. 3. Polyacrglamide gel electrophoresis using 20 and 40 pg of formiminotransferase. A. Samples were treated with 5 M guanidine-HCI and dialyzed against 8 M urea-l? Triton X-100, 0.0.5 hl 2-mercal,toethanol. Electrophoresis was carried out accordiny to Davis (5) ‘except for the inclusion of 8 r~ urea- I“; Triton X-100 in the 7’; pels. B. Sodium dodecyl sulfate gel electrophoresis on 9’; polyacrylamide gels. The molecular weight of the peptide chain is 6.2 : 0.2 x IO’ using glyceraldehyde-:l-phosphate dehydrogenase (36,000) catalase (60,000). hovine berum albumin (68.000), alcohol dehydrogenase (41.000). and phosphorylase a (94.000) as standards. for one hour at 30,OOOg the dark supernatant solution was carefully removed. The pellet was extrac,ted successively with 5 x ,5 ml volumes of 0.2 \I sodium acetate, pH 7.2, for 1 h, followed by centrilugation for 60 min at 30,OOOg. A final extraction with 5 ml ot’O.1 >f potassium phosphate, pH 7.2, was carried out to ensure complete extraction of the enzyme activity. The extracts containing activity were dialyzed against 1 liter of 0.2 M sodium acetate that had heen adjusted to pH 6 with 3 M acetate buffer (9:1, sodium
r? FIG. 4. Sedimentation equilibrium analysis ot’formiminotranst’erase-cyclodeaminase (100 pg/ml) at 1O’C. A. Native enzyme in 0.1 M potassium phosphate huf’fer. 0.05 h( 2-mercaptoethanol pH 7.2 ivas allowed to come to equilibrium for 24 h at 8070 rpm. B. enzyme denatured in 5 M guanidine HCI 0.05 M 2-mercaptoethanolO.1 M potassium phosphate pH 7.2 was allowed to equilibrate for 24 h at 30,340 rpm.
Sedimentation equilibrium Sedimentation equilibrium 5 M guanidine HCI Sedimentation velocity Dodecyl sulfate gel electrophoresis LIDetermined by the method of Edelstein b Protein concentration was 2 mg/ml.
6.58 * 0.23 x 10’ 6.2
of the procedure is derived largely from modifications of the purification by pH precipitation described by Tabor and Wyngarden (2). Precipitation of the enzyme extracted into sodium acetate by adjustment of the pH is a very powerful purification step and was modified by employing dialysis to effect the pH change more slowly, and thereby significantly increasing the yield of enzyme. The enzyme complex was crystallized by dialysis of a concentrated solution of the enzyme in phosphate buffer against 0.2 .LI sodium acetate, pH 7.0, in which it was observed to be much less soluble. The resultant small needlelike crystals tend to aggregate as shown in Fig. 1. The crystals could not be viewed readily by regular bright field microscopy due to their small size (2-10 pm) and lack of contrast. Although phase contrast resulted in somewhat better images, the crystals are best defined by the differential interference-contrast method which gives clearly visible outlines while focusing on a plane with a very limited depth of field. Thus crystals in a given plane are well resolved with little interference from structures above and below this plane. When collected by centrifugation, the crystals form a white gellike pellet. The use of DEAESephadex chromatography as a final purification step was occasionally useful to remove trace protein contaminants in our earlier preparations. but is generally found unnecessary to achieve homogeneity with this purification procedure. A summary of a typical purification is shown in Table I. The crystalline preparation of formiminotransferase contains high levrels of cyclodeaminase activity in agreement with the observation of Tabor and Wyngarden
properties 6 : 0.77”
s zo,w = 14.30
(2). However, our efforts to date have been directed primarily at the formiminotransferase activity. In checking buffer systems for potential use in purification procedures we have found that imidazole and Tris buffers are potent inhibitors while the use of multivalent anions as advocated by earlier investigators (2) slightlv stimulates the formiminotransferase activity, as illustrated in Table II. Although the native enzyme has a very high molecular weight. it nevrertheless could be electrophoresed in a 6”; polyacrylamide gel in Tris buffer which suggests that it may migrate as a dissociated species under these conditions. Both formiminotransferase and cyclodeaminase activities migrated with the single protein band, as shown in Fig. 2. The enzyme complex is homogeneous by several criteria, and it has not been possible to demonstrate the presence of nonidentical subunits. Figure 3 shows the results of gel electrophoresis in dodecyl sulfate and in the presence of 8 M urea-l% Triton X-100 where a single protein band was obtained in each case. These gels are relatively heavily loaded to demonstrate homogeneity; single bands are observed with 5-50 pg of protein. Analysis of both the native and denatured enzyme complex by velocity and equilibrium centrifugation indicated homogeneity. Plots of log x versus t for native enzyme were linear as were plots of log (Y, - Y,) vs r2 for enzyme in phosphate buffer or 5 M guanidine hydrochloride (Fig. 4). The pertinent molecular weight data, from both electrophoresis and ultracentrifugation is summarized in Table III. Good agreement was obtained for the subunit molecular weights by dodecyl sulfate gel electro-
(3). It is not clear what the green compophoresis and by sedimentation equilibrium nent is, but from the properties of our in 5 M guanidine HCl. From these values, colorless enzyme preparation it can be the best estimate of the subunit structure concluded that it is not required for activis that of an octomer of apparently identical ity and is most likely a contaminant. In subunits. The value of B obtained by ultraaddition, these authors used Tris buffer in centrifugation in H,O and D,O is a higher which we have value than normally found for proteins and their enzyme preparation, demonstrated to be highly inhibitory to the suggests the presence of a nonprotein formiminotransferase, and possibly causes component, possibly lipid or glycolipid. The uv spectrum of the protein is not un- dissociation of the enzyme. The reported for formiminoglutamate to usual with a maximum at 278 nm and a requirement stabilize the enzyme is probably due to the 280/260 ratio of 1.24 (R. Beaudet, unpublished observations). use of Tris, since such treatment is not found to be necessary in our preparation. The enzyme complex is very large (5X DISCUSSION x 10” daltons) and appears from our molecular weight estimates to contain eight Formiminoglutamate is a degradation identical subunits. Of course it product of histidine in both bacteria and apparently must be emphasized that with such a large mammalia but only in the latter does number of subunits and considering the further metabolism involve tetrahydrofoaccuracy of measurement of molecular late-dependent enzymes. Metabolically, weights, the actual number could readily both transferase and cyclodeaminase activities are of interest with respect to their role be 7, 8. or 9 subunits. Eight is the nearest and regulation in gluconeogenesis and the integer from our measurements, and is the associated responses to hormones in mammost likely number from the standpoint of malia. Investigation of the nature of the symmetry. physical association of this pair of enzymes The subunits are apparently identical is relevant to the general topic of protein because of demonstrated homogeneity by structure, and is necessary to understand sedimentation equilibrium in .i ~1 guanimore fully the molecular aspects of funcdine HCl, dodecyl sulfate gel electrophoretion and regulation. sis and electrophoresis in 8 M urea. These A prerequisite to such investigations is a data suggest therefore. that both activities reproducible method to obtain reasonable are the properties of a single polypeptide amounts of pure enzyme. The purification chain. Previous work (2) demonstrated procedure outlined yields minimally al- that different conditions could be used to tered homogeneous enzyme preparations preferentially destroy either the forwithout the requirement for column chromiminotransferase or cyclodeaminase matography in most instances. In this which suggests that these activities are not respect, the procedure is similar to that of the properties of a single active site. Thus. Tabor and Wyngarden (2) which yielded the current data indicate that two sites an enzyme preparation that showed a sinprobably exist on one polypeptide. The exgle symmetrical peak on sedimentation istence of more than one enzyme activity in velocity analysis in the ultracentrifuge. a single polypeptide chain has been demOur procedure differs in that the requireonstrated for several enzymes including ment for acetone powder preparations has DNA polymerase from E. coli which COIIbeen circumvented in an efficient manner. tains polymerase, 3’-5’ exonuclease and and the purified enzyme is crystalline and 5’-3’ exonuclease activrities (1Sl.i). The demonstrated to be homogeneous by sev- threonine sensitive aspartokinase-homoseeral techniques. A more recent preparation rine dehydrogenase of Escherichia coli has of the complex from liver acetone powders been shown to possess two activities but involves the use of apparent affinity chrocontains identical subunits (16-19). The matography and isoelectric focusing to transferase and cyclodeaminase could be achieve a preparation that is green in color another example of’ this type of enzyme
complex. Such a situation could arise by f&ion of’ the products of’ two genes. \vhich has heen demonstrated in mutants ot’ Salmonclln t?‘phimurium where the enzymes r.-histidinol dehydrogenase and imidazoyl phosphate: r.-glutamate aminotranst’erase. normally distinct species, have been incorporated into a single polypeptide chain. The fusion occurs with only small effects on the catalytic properties of’ the component enzymes (20). The value of’ u of the native enzyme measured by the method of’ Edelstein and Schachman (12) was O.$‘i, a value that is higher than normally f’ound f’or pure protein, which indicates the presence of’ a non-protein component. Currently it is not clear what this component is. hut it does’ not cnntrihute to the ultraviolet spectrum of the enzyme. It is possible that the enzyme may contain lipid or glycolipid. The indication of’ a nonprotein component underscores the prohahle desirability of avoiding delipidating procedures in the enzyme purification at least until this apsect ot’the enzyme structure is resolved.
1. SXIFEI., F. B., TAI NTON. 0. D., GHEESE:, H. L., LL:FKI\, E. G., HACIXH, L., ASI) HEHMAS, R. H. (19741 Riochim. Riophvs. Acta 354, 194-205. 2. TAHOH, H., .4x1) WYN(;AHI)E.\, L. (1959) J. Bioi. Chem. IK,
FoK.I.. P. (1974)
R. I,. (1957) Hiochc,m. J. 65, :!:I1 Z1-1”. C. E., D’ARI, L., AND RABINOWITZ, J. C. (1970) J. H/r,/. (‘hem. 24;, 511:,~5l”l. 6. RAHI~OUI~I.~, .J. C. (I%%) hfethods in Etrzym~)logv (Colowick, S. P.. and Kaplatl. N. 0.. eds.). Vol. VI, pp. 812-814, Academic Press, New York. 7. LOWRY, 0. H., ROSEBROUGH, N. T., FAHR, A. L., ,411) RAV)II I. R. .J. ClY:,ll J. Hiol. C’hcm. 193, 265-27,j. 8. DAVIS, H. J. (19641 Ann. ,‘Vcw Ym-h Acnd. Sc,i. 121, 4.
404-427. ,I.. AXI) W~~~EKSI~~~, A. R. (1971) J. Hiol. (‘hem. 246, :FI:V :ia41. WEHEK. K.. 4NI, OSHOHNE, M. (1971) J. f%ioi. C&m. 244, 4406-4412. Ywo IW. D. A. (1964) Hiochemistrx 3, 297 :lli. EDELSTEIN, S. J., AND SCHACHMAN, H. K. (1967) J. Biol. (‘hem. 242, 306-:111. KOKSHEH(;, A. (1969) Science 163, 1410~1418. Jo\ IN, T. M.. E\(.I.I \I), P. T.. AI,) Bwrwti, L. I,. (1969) J. Hioi. Chem. 244, 2996-3008. KEI.I.~, R. B., COZUHEI.I.I. X. R., DEI~TPTHEK. M. 0.. LEHhlA\, I. R.. 4x1) KOKNHEH(;. A. (1970) J. Hiol. (‘hem. 245, 339-45. COHES, (;. h-. 11969) Cur. ?‘op. CC/( RP~u/. 1, 18:3%2x). CLAHK, R. G.. .4x1) OGII.VIK, .J. W. (1972) Biochemisty 11, 1278~1282. STAKSE~, W. L.. M~xK, P.. MAI I., S. B.. CLsNt\(;t[.j\t. G. N.. Cos, D. ?J., 4x1) Stuw, W. (1972) Riochemistr>, 1 I, 677-687. F.~.coz-KEI.I.\, .J., .I.\NIN, .J., SAAKI, .J. C.. VEHO~, M., TRUFFA-BACH~, P., AND COHEN, G. N. (1972) Eur. J. Biochem. 28, 507-519. AI.~.J, S.. BHI\I, C. B., EoEr.tIoctI, H.. ANI) RE(,HIXK, M. M. (1973) J. Viol. (‘hem. 248, t5880-5886.
10. 11. 12. 13. 14. 15.