Purification and characterization of dihydrofolate reductase from soybean seedlings

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 255, No. 2, June, pp. 279-289,1987

Purification and Characterization of Dihydrofolate Reductase from Soybean Seedlings’ SHOBHA

RATNAM, AND

Department

of Biochemistry,

Received November

TAVNER J. DELCAMP, JOHN B. HYNES,’ JAMES H. FREISHEIM3 Medical College of Ohio, C.S. 10,008, Toledo, Ohio 43699 4, 1986, and in revised form February

4,1987

Dihydrofolate reductase (DHFR; EC 1.5.1.3) was purified to homogeneity from soybean seedlings by affinity chromatography on methotrexate-aminohexyl Sepharose, gel filtration on Ultrogel AcA-54, and Blue Sepharose chromatography. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the enzyme gave a single protein band corresponding to a molecular weight of 22,000. The enzyme is not a 140,000 Da heteropolymer as reported by others. Amino acid sequence-specific antibodies to intact human DHFR and also antibodies to CNBr-generated fragments of human DHFR bound to the plant enzyme on Western blots and cross-reacted significantly in immunoassays, indicating the presence of sequence homology between the two enzymes. The plant and human enzymes migrated similarly on nondenaturing polyacrylamide electrophoretic gels as monitored by activity staining with a tetrazolium dye. The specific activity of the plant enzyme was 15 units/mg protein, with a pH optimum of 7.4. Km values of the enzyme for dihydrofolate and NADPH were 17 and 30 PM, respectively. Unlike other eukaryotic enzymes, the plant enzyme showed no activation with organic mercurials and was inhibited by urea and KCl. The affinity of the enzyme for folate was relatively low (& = 130 PM) while methotrexate bound very tightly (Ko < 10-l’ M). Binding of pyrimethamine to the plant enzyme was weaker, while trimethoprim binding was stronger than to vertebrate DHFR. Trimetrexate, a very potent inhibitor of the human and bacterial enzymes showed weak binding to the plant enzyme. However, certain 2,4-diaminoquinazoline derivatives were very potent inhibitors of the plant DHFR. Thus, the plant DHFR, while showing similarity to the vertebrate and bacterial enzymes in terms of molecular weight and immunological cross-reactivity, can be distinguished from them by its kinetic properties and interaction with organic mercurials, urea, KC1 and several 0 1987 Academic Press, Inc. antifolates. Dihydrofolate reductase (DHFR; tetrahydrofolate:NADP+ oxidoreductase, EC 1.5.1.3)4 catalyzes a key reaction in folate

coenzyme metabolism, i.e., the NADPHdependent reduction of dihydrofolate to tetrahydrofolate. Extensive physical and chemical studies of the enzyme obtained from various vertebrate, bacterial, and protozoa1 sources have been reported (re-

’ Research sponsored by the Science and Education Administration of the U.S. Department of Agriculture under Grant 86-CRCR-l-2188 from the Competitive Research Grants Office. ’ Present address: Department of Pharmaceutical Sciences, Medical University of South Carolina, 171 Ashley, Charleston, SC 29425. ’ To whom correspondence should be addressed. 4 Abbreviations used: DHFR, dihydrofolate reductase; MTT, 3-(4,5-dimethylthiazol-ByI)-2,5-diphe-

nyltetrazolium bromide; SDS, sodium dodecyl sulfate; PMSF, phenylmethylsulfonyl fluoride; pHMB, p(hydroxymercuri)benzoate; MeHgOH, methylmercuric hydroxide; MTX, methotrexate; ABTS, 2,2’-azinobis(3ethylbenzthiazolinesulfonic acid); ELISA, enzymelinked immunosorbent assay. 279

0003-9861/87 $3.00 Copyright Q 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.

280

RATNAM

viewed in (l-4)). DHFR isolated from vertebrates and bacteria is a monomeric protein of M, 18,000-22,000. Interestingly, higher molecular weight forms have been isolated from certain protozoans and shown to be bifunctional proteins with additional thymidylate synthase activity (1). Although the need for DHFR activity for cell growth in plants was demonstrated many years ago (5) very little is known about the properties of DHFR from plant sources, presumably because of the inherent low activity of the enzyme in plants. Partial purification has been reported for DHFR from pea (5), corn (6), and petunia (7) and from suspension cell cultures of carrot and rice (8). Recently, DHFR purified to homogenity from a carrot cell line was reported to be a homo-trimer with a subunit molecular weight of 58,000 (9). It was previously reported (10) that DHFR isolated from soybean seedlings was a heteropolymer comprising four different types of subunits. In this report we demonstrate, by using a more rigorous purification procedure, including gel filtration and Blue Sepharose chromatography, that the soybean DHFR has a basic molecular weight of ca. 22,000, similar to that of the vertebrate and bacterial enzymes. We further report the immunological, kinetic, and biochemical properties of the homogeneous plant enzyme including its interaction with a number of folate antagonists. EXPERIMENTAL

PROCEDURES

Mattic&. Soybean seeds were obtained commercially, Blue Sepharose CL-6B and AH-Sepharose 4B were obtained from Pharmacia. Ultrogel AcA-54 was obtained from LKB. Methotrexate (MTX) was a gift from Lederle laboratories. Dihydrofolate was prepared from commercial folic acid by dithionite reduction (11) and stored lyophilized at -20°C. Leupeptin and PMSF were from Boehringer-Mannheim Biochemicals. Protamine sulfate, 2-mercaptoethanol, NADPH, peroxidase-conjugated goat-anti rabbit IgG, and 4-chloronaphthol were from Sigma Chemical Co. MTX-aminohexyl Sepharose was prepared according to Kaufman and Pierce (12). Reagents for amino acid analyses were from Beckman Instruments. The substituted 4,6-diaminodihydro-s-triazines were generous gifts from Dr. John A. R. Mead, Division of Cancer Treatment, National Cancer Institute, NIH. The methods of preparation of 6-(pcarboxyanilino-

ET AL. methyl)-2,4-diaminoquinazoline and 5-chloro-5,8-dideazoaminopterin have been reported previously (13a, b). Assay procedure for DHFR. The enzyme was assayed spectrophotometrically according to the method of Kempton et al. (14). Incubation mixtures for the determination of DHFR activity contained 100 mM Tris-HCl buffer, pH 7.5, 10 mM 2-mercaptoethanol, 100 pM NADPH, and various amounts of the enzyme preparation, in a final volume of 1 ml at 45°C. The reaction was initiated by the addition of 100 pM dihydrofolate. Enzyme activity was monitored by the decrease in absorbance at 340 nm using a Cary Model 219 spectrophotometer at 45°C. In the enzyme inhibition studies, various amounts of inhibitors were preincubated for 2 min with the enzyme and NADPH prior to initiating the reaction. Blank values were obtained in the presence of all the assay components except dihydrofolate. One unit of enzyme activity is defined as the amount of enzyme required to convert 1 pmol of substrate to product per minute under the above assay conditions. Protein estimation. The concentrations were determined by the dye-binding assay of Bradford (19) with the reagent obtained from Bio-Rad and using bovine serum albumin as a standard. Amino acid analysis. Samples were hydrolyzed in constant-boiling HCI in evacuated, sealed tubes flushed with nitrogen at 110°C for a period of 24 h. Amino acid analyses were performed on a Beckman Model 6300 amino acid analyzer and all results are expressed as molar ratios. Values for cysteine and methionine were determined as cysteic acid and methionine sulfone, respectively, after performic acid oxidation according to Moore (20). Polyacrylamidegel electrophoresis. Slab gel electrophoresis was performed under nondenaturing conditions at 4°C using polyacrylamide (12%) in 0.1 M Tris-glycine buffer, pH 8.4. The purified plant enzyme and a sample of pure human DHFR were electrophoresed simultaneously for 3 h at 4°C. Activity staining of DHFR was performed according to Gunlack (15). Following electrophoresis the slab gel was incubated at 37°C in 0.05 M Tris-HCl, pH 7.5, containing, 3 X 10m4 M NADPH, 2 X 10e4 M dihydrofolate, and 1.6 X 10m4M MTT. The appearance of a blue color was correlated with DHFR activity. Denaturing slab gel electrophoresis was performed according to Laemmli (16) using polyacrylamide gels (12%)with 0.1% sodium dodecyl sulfate at 22°C and 25 mA for 3 h using 0.1 M Tris-glycine buffer, pH 8.4. Protein bands were stained either with Coomassie brilliant blue R 250 or by silver staining (17). The molecular weight of the purified plant DHFR was estimated using bovine serum albumin (M, 68,000). ovalbumin (M, 44,000). carbonic anhydrase (M 29,000), chicken liver DHFR (M, 21,000), and cytochrome c (Mr 12,400) as standards

SOYBEAN

DIHYDROFOLATE

Western blots. The purified soybean DHFR was electrophoresed on a 12% polyacrylamide gel containing 0.1% sodium dodecyl sulfate at 25 mA for 3 h. The gel was then transferred onto nitrocellulose paper buffer, pH 8.4, previously soaked in 0.1 M Tris-glycine without sodium dodecyl sulfate. The transfer was carried out at 12 V for 3 h in an EC electroblot apparatus. The nitrocellulose blot was soaked in 10 mM sodium phosphate, pH 7.5/0.15 M NaCl containing 5% nonfat dry milk (Buffer C) for 1 h. The blot was then incubated in Buffer C with a 200-fold dilution of crossreacting rabbit anti-human DHFR antiserum (or a control normal rabbit serum) for 4 h. The blot was washed extensively with Buffer C and incubated with a 200-fold dilution of peroxidase-conjugated goat antirabbit IgG for 1 h. The blot was thoroughly washed with 10 mM sodium phosphate, pH 7.5, containing 0.15 M NaCl and the color was developed by incubating in a solution containing 50 mM sodium acetate, pH 6, 0.05% 4-chloronaphthol, and 0.015% H202 at 22°C. Enzyme-linked immunosorbent assays. Plant or human DHFR was immobilized in microtiter dishes and incubated with various dilutions of the cross-reacting rabbit anti-human DHFR antiserum or with a control normal rabbit serum. The binding of antibodies was assayed using glucose oxidase-labeled goat anti-rabbit IgG and measuring the absorbance at 405 nm in the presence of horseradish peroxidase, O-D-ghCose and 2,2’-azinobis(3-ethylbenzthiazolinesulfonic acid) (ABTS), as previously described (18). Rabbit antibodies to human DHFR and rat antibodies to CNBr fragments of human DHFR were kindly provided by M. Ratnam. Pura&cation of dihydrofolate reductase. The purification procedure was carried out at 4°C. Soybeans were germinated in moist sand in the dark over 4 days. Washed seedlings were homogenized with 0.1 M imidazole buffer, pH 7.5/0.1 M KCl/5 mM EDTA/lO mM 2-mercaptoethanol/25% glycerol (Buffer A) containing 1 mM PMSF and 10 pM leupeptin, in a Waring blender. The homogenate was filtered through cheesecloth and centrifuged at 90009 for 30 min. Protamine sulfate (0.2%) was added to the supernatant and centrifuged at 9OOOgfor 30 min. Step 1: MTX-aminohexyl Sepharose chromatography. The supernatant, after protamine sulfate treatment, was loaded onto an MTX-aminohexyl Sepharose 4B column (1.5 X 15 cm) equilibrated with Buffer A, pH 7.2. The column was washed extensively with this buffer, containing 0.5 M KC1 until the Azso of the effluent was less than 0.1. The enzyme was eluted by the addition of 30 ml of Buffer A, pH 8.0, containing 2 mM dihydrofolate. The DHFR activity eluted as a broad peak just after the void volume. The fractions containing the enzyme activity were pooled. Step 2: Ultrogel AcA-54 chromatography. The pooled fractions from the MTX-Sepharose column were concentrated to approximately 5 ml using an Amicon ul-

281

REDUCTASE

trafiltration apparatus with a YM-10 membrane. The concentrated sample was then passed through an Ultrogel AcA-54 gel filtration column (2 X 90 cm) equilibrated with Buffer A, pH 7.2, at a flow rate of 10 ml/ h. The DHFR activity eluted as a sharp peak. Enzyme fractions were pooled and dialyzed extensively against 50 mM imidazole buffer, pH 7.2, containing 2-mercaptoethanol (10 mM) and 25% glycerol (Buffer B). Step 3: Blue Sepharose chrcmzatography. The dialyzed fractions were loaded onto a Blue Sepharose affinity column (2 X 10 cm) equilibrated with Buffer B. The column was then extensively washed with Buffer B until the A280of the effluent was negligible. The elution of DHFR activity was carried out by passing successively 60 ml of Buffer B containing 0.5 M KCl, pH 7.2, followed by 30 ml of Buffer B containing 0.5 M KC1 and 2 mM folic acid, pH 7.2. Fractions containing DHFR activity were pooled and dialyzed and the homogeneous preparation was stored indefinitely at 4°C. RESULTS

Purification

of Soybean Seedling DHFR

Dihydrofolate reductase was purified to homogeneity from soybean seedlings using protamine sulfate precipitation, MTX-Sepharose affinity chromatography, gel filtration on Ultrogel AcA-54, and Blue Sepharose affinity chromatography. These steps are described under Experimental Procedures and summarized in Table I. One critical step in our purification was the use of Blue Sepharose affinity chromatography since formylaminopterin-agarose alone, used previously (lo), yielded an enzyme preparation containing at least three major bands and several minor ones. A typical elution profile shows that most of the proteins in the pool from the Ultrogel column do not bind tightly to the Blue Sepharose matrix under the stated conditions of ionic strength and pH (Fig. 1). The DHFR activity emerges as a sharp peak after switching to an elution buffer that contains folic acid (Fig. 1). We have also obtained a homogeneous enzyme preparation by using Blue Sepharose chromatography as the first step, followed by gel filtration on Ultrogel AcA-54 and MTX-Sepharose affinity chromatography. This avoids destruction of the MTX-Sepharose by substances in the crude extract. However, the latter procedure results in a lower yield. The total activity in the crude extract represented in Table I is somewhat of an overestimate be-

RATNAM TABLE

ET AL. I

PURIFICATION OF DIHYDROFOLATE REDUCTASE FROM SOYBEAN SEEDLINGS

Step (1) (2) (3) (4)

Volume (ml)

Total protein (mid”

3000 40 10 4

22,000 65 2.5 0.053

Post protamine MTX-Sepharose Ultrogel AcA-54 Blue Sepharose

Total activity (rmol/min)*

Specific activity (rmol/min/mg)

00) 5 3 3

’ Protein is estimated by the Bio-Rad procedure using bovine serum albumin *One unit of activity is defined as that amount of enzyme which catalyzes dihydrofolate to tetrahydrofolate per minute at pH 7.5 and at 45°C.

cause of the presence of NADPH oxidase activity which results in a high blank value causing inaccuracy in the estimation of DHFR. The purified enzyme was stable at 4°C in 0.1 M imidazole, pH 7.2, containing 25% glycerol with or without folic acid, indefinitely. Freezing and thawing of the enzyme resulted in significant loss of activity even in the presence of glycerol. Use of the protease inhibitors PMSF and leupeptin during homogenization and of 25% glycerol in all buffers throughout the purification is essential for maintaining the structural and catalytic integrity of the enzyme. Kinetic Properties of So$ean DHFR The Km values for NADPH and dihydrofolate for the purified enzyme were determined in 100 mM Tris-HCl buffer containI

+0.5M 300

KCI

+ZmM

FA t

t

60

I :

0.0005 0.076 1.2 56.6 as a standard. the conversion

Recovery (o/o) 50 30 30

of 1 pmol of

ing 10 mM 2-mercaptoethanol, pH 7.5 at 45°C. In the determination of the Km for NADPH, the initial concentration of the cofactor was varied between 5 and 100 pM while the concentration of dihydrofolate was 100 PM in each assay. The reaction was initiated by the addition of the enzyme as described under Experimental Procedures. The average of two determinations of the initial velocity for each NADPH concentration was plotted vs NADPH concentration in a double reciprocal plot (Fig. 2B). Similarly, in the determination of Km for dihydrofolate, the initial concentration of cofactor was varied between 5 and 140 ~.LM while the concentration of NADPH was 100 PM for each assay. From the double reciprocal plot the Km value for NADPH was determined to be 30 ~.LM and the Km value for dihydrofolate was 17 PM (Fig. 2A). The specific activity of the purified soybean DHFR is 56 units/min/mg protein at 45’C. Assuming a molecular weight of 22,000this corresponds to a turnover number of 330 min-‘. The pH optimum for the plant enzyme was determined by assaying the enzyme in 0.1 M sodium acetate (pH 3.5-5.25), 0.1 M potassium phosphate (pH 5.0~8.5), and 0.1 M Tris-HCl (pH 7.0-9.0) buffers. The enzyme showed a single pH optimum at pH 7.4. Molecular Weight of Soybean Seedling DHFR

FRACTION

FIG. DHFR column. taining further

NUMBER

1. Elution profile of total protein (-) and activity (- * -) from the Blue Sepharose affinity The enzyme was eluted with Buffer B con0.5 M KC1 and 2 mM folic acid, pH 7.2. For details see Experimental Procedures.

The plant enzyme and the human enzyme migrated similarly on a nondenaturing slab gel (12%) when monitored by activity staining with a tetrazolium dye, MTT (data not shown) as evidenced by the appearance of a blue band at identical posi-

SOYBEAN

DIHYDROFOLATE

‘/FAH,

283

REDUCTASE

‘/NADPH

(~4)

(~h4)

FIG. 2. Double reciprocal plots of dihydrofolate (A) and NADPH (B) saturation curves for soybean DHFR. The assays were performed in 100 mM Tris-HCl buffer containing 10 mM 2-mercaptoethanol, pH 7.5, at 45°C. The average of two values for initial velocity for each substrate concentration was

tions in the two lanes. This result indicates similar charge/mass ratios for the plant and vertebrate DHFRs. Electrophoresis on a sodium dodecyl sulfate-polyacrylamide gel gave a single protein band corresponding to M, 22,000 (Fig. 3A). Immunological Cross-Reactivity Human and Plant DHFRs

between

Rabbit antibodies raised against the human DHFR cross-reacted significantly

with the plant enzyme on Western blots of sodium dodecyl sulfate-polyacrylamide slab gels as evidenced by the staining of the single M, 22,000 band (Fig. 3B). Solidphase immunoassays (ELISA) were performed, in which the plant and human enzymes were immobilized in microwells and various concentrations of antisera to the human enzyme were used as primary antibody (Fig. 4). The plant enzyme cross-reacted to a significant extent (ca. 70%)with

Mr [Kd]

68 44 -

29 22 10-5

10‘4 ANTIBODY

12 -

A

B

FIG. 3. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12%)of the purified soybean DHFR. Lane A represents the Coomassie blue-stained protein. An identical gel was transferred onto nitrocellulose and probed with antihuman DHFR (lane B).

10-3

10-Z

10-I

DILUTION

FIG. 4. Binding of antibodies to immobilized human and plant DHFRs. DHFR (1 pmol per well) from human WIL-2 cells (0) or soybean seedlings (0) was immobilized in polystyrene microtiter wells. Various dilutions of anti-human DHFR antibodies (50 ~1) were applied as primary antibody in quadruplicate wells. The wells were further processed as described (Experimental Procedures) and the amounts of specific antibody binding are proportional to the absorbance at 405 nm.

284

RATNAM

the antibodies amounting to a significant proportion of the total antibody binding to human DHFR (Fig. 4). This result indicates the presence of one or more similar epitopes in the plant and vertebrate enzymes. Antisera to various CNBr-generated fragments of human DHFR, corresponding to the sequences 15-52 (CB 2, 3), 53-111 (CB 4), 112-125 (CB 5), and 140-186 (CB 7), were used to probe denatured plant and human DHFRs immobilized in microtiter plates by ELISA. These results are given in Table II. Since it was not possible to determine the exact amount of each protein immobilized in the microtiter wells, the absorbance at 405 nm obtained in the solid phase assay using antiserum to whole human DHFR was taken as the reference value in each case and the absorbances at 405 nm for each individual antiserum are represented in Table II relative to the reference value. Thus Table II does not in any case represent the absorbance at 405 nm of the plant enzyme relative to the human enzyme but, rather, highlights the differences in the amount of sequence homology in the various parts of the primary sequence between plant and human DHFRs. Amino Acid Analysis The results from amino acid analyses of the acid hydrolyzed soybean DHFR are given in Table III. Numbers of residues per molecule of subunit are based on the asTABLE RELATIVE

Peptide CB2,3 CB4 CB5 CB7

II

BINDING OF ANTIBODIES AGAINST CNBr PEPTIDES OF HUMAN DHFR &d& (DHFR)” Human DHFR 0.43 0.41 0.05 0.60

Plant DHFR 0.17 0.37 0.24 0.24

a The values represent ratios of absorbance at 405 nm in ELISA assays due to binding of antipeptide antibodies to the value for antibodies raised against whole human DHFR. The values do not represent antibody binding to plant DHFR relative to human DHFR.

ET AL. TABLE

III

COMPARISON OF AMINO ACID COMPOSITION OF SOYBEAN DHFR AND HUMAN WIL-2/M4 DHFR

Residues” Amino acid Asx Thr” Ser” Glx Pro GUY Ala Cyd Val Metd Ile Leu Tyr Phe His LYS Arg Tw

Soybean*

Human

17.4 8.2 22.6 21.5 8.2 26.2 17.4 2.2 11.3 2.4 7.2 13.3 7.2 11.3 3.1 6.1 6.1 -

19.7 (20) 6.6 (7) 11.4 (11) 22.1 (22) 10.7 (11) 13.2 (13) 5.2 (5) 0.9 (1) 13.1 (13) 5.8 (6) 8.5 (9) 19.0 (19) 6.6 (7) 9.8 (10) 3.7 (4) 17.7 (18) 7.2 (7) 3.0 (3)

(17) (8) (23) (22) (8) (26) (17) (2) (11) (2) (7) (13) (7) (11) (3) (6) (6)

a Based on hydrolysis time of 24 h in constant-boiling HCl. The numbers in parentheses represent the nearest integer. * Based on molecular weight of 22,000. ‘Extrapolated to zero time of hydrolysis. d Determined as cysteic acid and methionine sulfone, respectively.

sumption of M, 22,000 for the protein. The enzyme contains two cysteine and two methionine residues among other residues. Repeated attempts to obtain the aminoterminal sequence of the denatured soybean DHFR were unsuccessful, perhaps due to the presence of a blocked amino terminus. Interaction of Soybean DHFR with Activators of Vertebrate DHFRs A number of reagents such as urea, thiourea, KCl, and organic mercurials have been found to activate vertebrate DHFRs severalfold, leading to the hypothesis that there could exist physiological activators of DHFR (21). Investigation of the effects of these reagents on the purified plant DHFR showed the absence of activation. Further, while p-hydroxymercuribenzoate

SOYBEAN

1

I

0.4

DIHYDROFOLATE

0.8

1.2

1.6

REDUCTASE

468

2

Urea, M

KCI,M

FIG. 5. Effect of KC1 (O-l.2 M) (A) and urea (O-8 M) (B) on the soybean DHFR. The enzyme was assayed in 100 ItIM Tris-HCl buffer containing 10 mM 2-mercaptoethanol, pH 7.5, with the indicated amounts of KC1 and urea. Each point represents activity remaining as a percentage of the control which is activity of the enzyme alone.

(pHMB) and methylmercuric hydroxide (MeHgOH) either in the presence or in the absence of NADPH did not have any effect on the enzyme activity at 0 or at 22”C, increasing concentrations of KC1 (O-l.2 M) inhibited the enzyme (Fig. 5A) as did urea (O-8 M) (Fig. 5B). Interaction

with Antifolates

Methotrexate inhibited soybean DHFR linearly and completely (Fig. 6A) indicat-

ing that at the given enzyme concentration the compound binds to the enzyme very tightly (i.e., & < 10-l’ M) and stoichiometrically. On this basis, the turnover number of the enzyme was calculated from the data in Fig. 6A and was found to correspond to the value obtained on the basis of protein estimation of the pure enzyme and assuming a molecular weight of 22,000. This further confirms that the catalytic unit in soybean DHFR is M, 22,000. Inter-

0.08

MTX, nM

0.16

0.24

0.32

Folic acid, mM

FIG. 6. Inhibition by MTX (O-2 nM) (A) and folic acid (O-O.32 mM) (B) of the soybean DHFR. The enzyme was assayed in 100 mM Tris-HCl buffer, pH 7.5, with the indicated amounts of MTX and folic acid. Each point represents activity remaining as a percentage of the control which is the activity of the enzyme alone.

RATNAM

286

TABLE

ET AL. IV

COMPARISONOF THE INHIBITION OF SOYBEAN SEEDLING, HUMAN, AND L. cusei DIHYDROFOLATE REDUCTASES BY VARIOUS FOLATE ANTAGONISTS

Id&cl Compound number”

Structure

Soybean

(MTX) Human

1

L cusei

1

CH I 2 COOH

II

33,300

900

8.4

III

46,700

300,000

0.5

1,300,000

26

0.36

10.6

0.83

1.35

1.0

1.14

2.52

&OH

VIII

53

1

56

3.5

8

SOYBEAN

DIHYDROFOLATE TABLE

287

REDUCTASE

IV-Continued

&o~ho Structure

Compound number”

Soybean

WW Human

2.4

8.6

XI

83

L cmei

32

2.1

3.2

1.3

0.9

a I, methotrexate; II, pyrimethamine; III, trimethoprim; IV, trimetrexate; V and VI, 2,4-diaminoquinazolines; VII-X, sulfonyl fluoride derivatives of 4,6-diaminodihydro-s-triazine; XI, BW30lu.

estingly, folic acid bound very poorly to the plant enzyme (& = 130 PM, Fig. 6B). The effectiveness of several antifolate compounds in inhibiting soybean DHFR were tested and the 2,4-diaminoquinazolines were found to be the most potent (Table IV). DISCUSSION

It is clearly established that similar to bacteria, protozoans, and vertebrates, plants are dependent on DHFR for cell growth (5,21,22). Elucidation of the properties of DHFRs from plant sources, apart from being of intrinsic interest, should provide us with model systems for understanding general structure-function relationships in DHFRs. Further, DHFR could be exploited as a target for antifolate compounds in the selective inhibition of growth of certain plants and plant pathogens in agriculture.

We have purified DHFR from soybean seedlings to apparent homogeneity as evidenced by a single protein band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Fig. 3), a single band in nondenaturing electrophoretic gels that is positive to enzyme activity staining, and cross-reactivity of the single protein band on Western Blots with antibodies to human DHFR (Fig. 3). The apparent molecular weight of the purified enzyme on denaturing polyacrylamide gels was calculated to be 22,000. Our results differ from a previous report (10) that the soybean DHFR is a heteropolymer of M, 140,000.We believe that this difference is due to the presence of impurities and/or aggregated states of the enzyme from the single step affinity purification (on formylaminopterin-agarose) reported previously (10). Our view is consistent with the observation that the specific activity of the purified enzyme re-

288

RATNAM

ported herein is about 400-fold higher than that reported previously (10). The soybean DHFR is also considerably different from the enzyme recently isolated from a carrot cell line (9). The carrot enzyme appears to be a homo-trimer with a subunit molecular weight of 58,000.The pH optimum of the soybean DHFR is 7.4 unlike that reported for carrot (pH 5.9). Further, unlike the carrot DHFR (9) the soybean enzyme shows significant immunoreactivity with antibodies to human DHFR. While this difference might simply reflect the result of using different antibody preparations for the soybean and carrot DHFRs it is tempting to speculate that the soybean DHFR might have a greater sequence homology with vertebrate DHFRs than the carrot enzyme, especially in view of the similarity of its molecular weight to vertebrate DHFRs. In addition, the carrot DHFR (9) has more than a lo-fold greater turnover number than the soybean enzyme, a 12-fold lower Km for NADPH, and a 4fold lower Km for dihydrofolate. The soybean and carrot (9) DHFRs both have high affinities for MTX and low affinities for folate. Despite the striking physical and immunological similarities between the soybean and vertebrate DHFRs the soybean enzyme is quite distinct kinetically and in its interaction with a number of reagents and folate analogs (Table V). While the molecular weights and specific activities of all the three enzymes are similar, the soybean DHFR exhibits markedly higher Km values for NADPH and dihydrofolate. The soybean enzyme also does not display the double pH optima of vertebrate DHFRs. A striking, albeit little understood phenomenon among vertebrate DHFRs and in sev-

ET AL.

eral cases among bacterial DHFRs is the severalfold activation of these enzymes by KCl, urea, thiourea, and organic mercurials (20, 23-25). It is, however, known that organic mercurials covalently modify the single cysteine residue in vertebrate DHFRs (20). Although the soybean DHFR contains cysteine residues (Table III) organic mercurials such as pHMB and MeHgOH failed to activate this enzyme. KC1 and urea inhibited the soybean enzyme at high concentrations (Fig. 5), whereas activation occurs with vertebrate DHFRs. Polyclonal antibodies to human DHFR, as well as antibodies to CNBr peptides of human DHFR corresponding to the sequences 15-52, 53-111, 112-125, and 140186 all cross-reacted with the plant enzyme (Figs. 3, 4, Table II), suggesting the presence of homology between the two enzymes covering most parts of the primary sequence. However, Table II shows that the various antipeptide antibodies cross-reacted with the plant enzyme to different extents, emphasizing the presence of large differences in the primary structures of the human and plant enzymes as well. This observation is borne out by a comparison of the amino acid compositions of the plant and human DHFRs (Table III) which shows significant differences in the amounts of several residues such as alanine, proline, and lysine. Similar to DHFRs from vertebrates and bacteria, MTX displayed a very high affinity for soybean DHFR (Fig. 6A). In contrast, folic acid, a strong inhibitor of vertebrate DHFRs (Iso = 0.5 PM), bound relatively poorly to the plant enzyme (Fig. 6B). However, folic acid, together with a high ionic strength, was required to elute the soybean enzyme from the Blue Se-

TABLE PROPERTIES

OF DIHYDROFOLATE

V

REDUCTASES

FROM VARIOUS

SOURCES

K, (M X 106) Source

w (x10-3)

FAHz

NADPH

Soybean L. casei Human

22 18.3 21

17 0.36 0.04

30 0.78 0.25

pH optimum

Specific activity

7.4 6.5 4; 7.3-8.3

15 (22°C) 12 (30°C) 16 (22OC)

SOYBEAN

DIHYDROFOLATE

pharose affinity matrix (Fig. 1). This indicates that Cibacron Blue 3GA, the ligand immobilized in Blue Sepharose, interacts with the folate binding domain in the soybean DHFR similar to its interaction with vertebrate DHFR, but different from its interaction with bacterial DHFR (26). Regarding the interaction of the soybean DHFR with antifolates, the enzyme displayed much lower affinities for pyrimethamine and trimetrexate than the vertebrate and bacterial enzymes and an intermediate value for trimethoprim (Table IV). The nonclassical 2,4-diaminoquinazoline (V in Table IV) inhibited the plant enzyme less effectively than the human and L. casei ennopterin (VI) was approximately threefold more inhibitory toward plant DHFR than MTX making it the most potent of the compounds studied. Sulfonyl fluoride derivatives of 4,6-diaminodihydro-s-triazine compounds (compounds VII-X) which are potent inhibitors of vertebrate and bacterial DHFRs displayed a uniformly lower affinity for the plant DHFR (Table IV) (27). The above observations indicate that there are differences in the mode of interaction of folate analogs at the active site of the plant DHFR as compared with the interactions with the vertebrate and bacterial enzymes. ACKNOWLEDGMENTS We thank Ms. Jean Nussbaum for technical assistance and Dr. Manohar Ratnam for valuable discussions. We thank Ms. Valerie Murphy for typing the manuscript. REFERENCES 1. BLAKLEY, R. L. (1984) in Folates and Pterins (Blakley, R. L., and Benkovic, S. J., Eds.), Vol. 1, pp. 191-253, Wiley, New York. 2. FREISHEIM, J. H., AND MATTHEWS, D. A. (1984) in Folate Antagonists as Therapeutic Agents (Sirotnak, F. M., Burchall, J. J., Ensminger, W. D., and Montgomery, J. A., Eds.), Vol. 1, pp. 69-131, Academic Press, Orlando, FL. 3. HITCHINGS, G. H., AND BACCANARI, D. P. (1984) in Folate Antagonists as Therapeutic Agents (Sirotnak, F. M., Burchall, J. J., Ensminger, W. D., and Montgomery, J. A., Eds.), Vol. 1, pp. 151-172, Academic Press, Orlando, FL. 4. MONTGOMERY, J. A., AND PIPER, J. R. (1984) in Folate Antagonists as Therapeutic Agents (Sir-

REDIJCTASE

289

otnak, F. M., Burchall, J. J., Ensminger, W. D., and Montgomery, J. A., Eds.), Vol. 1, pp. 219260, Academic Press, Orlando, FL. 5. SUZUKI, N., AND IWAI, K. (1970) Pkrnt CeUPhysioL 11,199-208. 6. CROTI, P. (1981) J. Exp. Bot. 32,717-723. 7. BARG, R., PELEG, N., PERL, M., AND BECKMANN, J. S. (1984) Plant MoL Biol. 3,303-311. 8. CELLA, R., CROSTI, P., NIELSEN, E., AND PARISI, B. (1983) J. Exp. Bot. 34,1189-1195. 9. ALBANI, D., PARISI, B., CARBONERA, D., AND CELLA, R. (1985) Plant Mol. BioL 5,363-372. 10. REDDY, V. A., AND RAO, N. A. (1976) Arch Biochem. Biophys. 174,675-683. 11. BLAKLEY, R. L. (1960) Nature (London) 188, 231232. 12. KAUFMAN, B. T., AND PIERCE, J. V. (1971) Bicchem Biophys. Res. Commun. 44,608-612. 13a. HYNES, J. B., BUCK, J. M., D’SOUZA, L., AND FREISHEIM, J. H. (1975) J. Med Chem. 18,11911194. 13b. HYNES, J. B., EASON, D. E., GARRET, C. M., COLVIN, JR., P. L., SHORES, K. E., AND FREISHEIM, J. H. (1977) J. Med. Chem. 20,588-591. 14. KEMPTON, R. J., BLACK, A. M., ANSTEAD, G. M., KUMAR, A. A., BLANKENSHIP, D. T., AND FREISHEIM, J. H. (1982) J. Mea! Chem. 25,475477. 15. GUNLACK, B. G., NAL, G. E., AND WILLIAMS, D. C. (1968) Biochem. PharmacoL 17,484-487. 16. LAEMMLI, U. K. (1970) Nature (Londmz) 227,680685. 17. TSAI, C-M., AND FRASCH, E. (1982) Anal. B&hem. 119,115-119. 18. HOCHSCHWENDER, S., LARGELBERG, L. K., SCHNEIDER, D. W., AND LINDSTROM, J. M. (1985) in Hybridomas in Biotechnology and Medicine (Springer, T., Ed.) pp. 223-238, Plenum, New York. 19. BRADFORD, M. M. (1976) Anal B&hem. 72,248254. 20. MOORE, S. (1963) J. BioL Chem. 238.235-237. 21. DELCAMP, T. J., SUSTEN, S. S., BLANKENSHIP, D. T., AND FREISHEIM, J. H. (1983) Biochemistry 22,633-639. 22. CELLA, R., CROSTI, P., NIELSEN, E., AND PARISI, B. (1983) J. Exp. Bot. 34(146), 1189-1195. 23. DOMIN, B., CHENG, Y-C., AND HAKALA, M. T. (1982) MoL Pharmacol 21,231-238. 24. JARABAK, J., AND BACHUR, N. R. (1971) Arch B&hem. Biophys. 142,417-425. 25. KAUFMAN, B. T., KUMAR, A. A., BLANKENSHIP, D. T., AND FREISHEIM, J. H. (1980) J BioL Chem 255,6542-6545. 26. SUBRAMANIAN, S., AND KAUFMAN, B. T. (1980) J. BioL Chem. 255,10587-10590. 27. KUMAR, A. A., MANGUM, J. H., BLANKENSHIP, D. T., AND FREISHEIM, J. H. (1981) J. BioL Chem 256,8970-8976.