Age-dependent expression of a novel protein in mouse liver immunologically and functionally homologous with dihydrofolate reductase

56

Biochimica et Biopt~vsica Acta. 993 (1989) 56-62

Elsevier BBAGEN 23196

Age-dependent expression of a novel protein in mouse liver immunologically and functionally homologous

with dihydrofolate reductase D a v i d K y n e r a n d S h e l d o n P. R o t h e n b e r g Department of Medicine. Veterans Hospital and ~ UN Y Health Science Center. Brooklyn. N Y (U.S.A.)

(Received 15 February 1989)

Key words: Dihydrofolatereductase; (Mouseliver) Dihydrofolate reductase ~ith a molecular weight of 22000 has been purified by salt precipitation and methotrexateSepharose affinity chromatography from mouse livers having a mean weight of 2.4 g each. When the same purification procedure was followed using livers with a mean weight of 1.4 g or less, a protein with a molecular weight of 27500 co-purified with dihydrofolate reductase. This 27.5 kDa species was recovered with dihydrofolate reductase following a second passage through the affinity column and it reacted by Western immunoblotting with a monospeci[ie polyclonal antiserum raised to the purified 22 kDa enzyme. The two proteins could not be separated in a native state to compare their functional activity, but the 27.5 kDa protein appeared to have catalytic reductase activity when assayed directly on the polyacrylamide gel following non-denaturing eleetrophoresis. The catalytic activity of the mixture of the purified proteins was stuichiometrically inhibited by a molar equivalent of methotrexate.

Introduction Dihydrofolatc reductase (DHFR), the enzyme which catalyzes the NADPH-dependent reduction of dihydrofolate (H2-PteGlu) to tetrahydrofolate (H4-PteGlu), has been isolated from both bacteria and mammalian cells and tissues [1~2] and characterized as an 18-22 kDa protein. It is the target enzyme for the folate antagonists, and following exposure of cells to these agents, the synthesis of the enzyme frequently increases [3,4]. Sometimes this 'induced" enzyme has a slightly greater molecular weight [5-7], a molecular weight of 42 000 [8], or a lower affinity for methotrexate (MTX) [9,10]. The functional properties of D H F R from normal mouse tissues has been the subject of several studies (11-15), but is has not yet been purified to homogeneity from mouse liver. In this study, we purified D H F R

Abbreviations: DHFR, dih)droft,late reductase(EC 1.5.1.3); PteGlu, pteroylglqtamic (folic)acid; H2.PteGlu, dibydrof,alatc: Ha-PteGlu, tetrahydrofolate; MTX, methotrexate; PAGE. polyacrylamide gel electrophoresis,SDS, sodium dodecylsulfate; NADPH, nicotinamide adenine dinucleotidepho~phale, reduced form; MTT-tetrazolium, 3(4.5-dimethyllhiazol-2)2.5-diphenyitetrazoliumbromide. Correspondence: S.P. Rothenberg, SUNY-Health Science Center. Brooklyn Veterans Hospital, 800 Poly Place, Brooklyn, NY 11209, U.S.A.

from the liver of normal mice and observed that a 27.5 kDa protein co-purified with the 22 kDa enzyme only from younger mice. This protein crossreacted immunologically with D H F R against a monospecific polyclonal antiserum raised to the purified 22 kDa enzyme from L1210 murine leukemia cells and appeared to have catalytic activity by an assay directly on the polyacrylamide gel. The catalytic activity of this mixture of both proteins was inhibited by MTX. Materials and Methods [3H]MTX (spec. act. > 5 C i / m m o l ) and [3H]PteGlu (24 Ci/mmol), were purchased from Amersham/Searle, Arlington Heights, 1L. Protamine sulfate, fast red violet, napthol AS-BI phosphoric acid, goat anti-rabbit IgG (whole molecule) alkaline phosphatase conjugate, trypsin coupled to agarose (insoluble trypsin), bovine serum albumin, MTT-tetrazolium, NADPH, PteGlu, and H2PteGlu were purchased from Sigma. St. Louis, MO. Acrylamide gel, silver stain kit, and protein assay kit were obtained from Bio-Rad, Richmond, CA. MTX was obtained through the courtesy of Lederle Laboratories, Pearl River, NY, and purified by the method of Galleli and Yokoyama [16]. BDF l mice (C57BIxDBA/2F t) were obtained from the National Institutes of Health. D H F R was purified f"om the livers of BDF t mice using the same procedure described previously for the

0304-4165/89/$03.50 ,3 1989 ElsevierScience PublishersB.V.(Biomedical Division)

57 purification of the enzyme from calf liver [17,18]. The steps essentially involve homogenization of the liver in 3 vols. of buffer followed by precipitation wit.~, protamine sulfate, zinc sulfate and ammonium sulfate (50-80%). Though the buffers used in the first purification of D H F R did not contain proteinase inhibitors, for subsequent purifications aprotinin (1000 K1U/I) and phenylmethylsulfonyl fluoride (3.5 mg/l) were added to the working buffers. The precipitate obtained at 80% ammonium sulfate saturation was dissolved in 0.1 M KPO 4 (pH 7.0) buffer and then centrifuged at 100000 × g for 60 min to sediment insoluble debris. N A D P H (50 ttM) was added to the supernatant fraction which was then applied to the column of Sepharose 4B to which was coupled methotrexate [19]. The column was then extensively washed with 50 mM KPO4 and 0.5 M KPO4 (pH 7.4) containing 50 ttM NADPH and the enzyme was then eluted with 0.5 M KPO4 buffer (pH 8.9) containing dihydrofolate (0.2 mg/ml) without any NADPH. Protein was measured using the Bio-Rad protein assay with bovine serum albumin as the standard. The catalytic activity of D H F R was determined by a radioenzymatic assay using [sH]PteGlu [20]. Since this assay is carried out at pH 4.8, the NADPH, which is unstable in ~cid solution, was prepared in 0.5 M sodium citrate (pH 7.2) and added to the assay buffer at the start of the reaction and in a 50000-fold excess of the folate substrate.

Polyac~.'lamide gel electrophoresis Electrophoresis was carried out in a 5-12.5% gradient polyacrylamide gel prepared in 25 mM Tris/192 mM glycine at pH 8.8 with the running buffer at pH 8.3. SDS-PAGE was carried out using a 12.5 or a 15% gel by the method of Laemmli [21]. The proteins were stained using the silver stain kit as instructed by the manufacturer. D H F R activity was detelmined directly on non-SDS gels using the MTT-tetrazolium-staining method of Mell et al. [22], with the polyacrylamide gel incubated at 37°C on top of 1.6% agarose containing 2.5 mg MTT-tetrazolium, 2.5 mg NADPH, and 1.5 mg H2-PteGlu in 8.5 ml of 100 mM Tris (pH 7.2).

Imrnunoblotting The SDS-polyacrylamide gel was incubated in transfer buffer (25 mM Tris/192 mM glycine (pH 8.3)/20% methanol) for 30 min. The protein was then transferred to a 0.2 ~ nitrocellulose membrane for 5 h at 200 mA and for 13 h at 70 rnA using a Bio-Rad Trans-Blot R Electrophoretic Transfer Cell. The membrane was then quenched for 2 h in Tris-saline buffer (10 mM Tris (pH 7.2)/150 mM NaCI) containing 3% gelatin, washed twice in Tris-saline buffer, and then i'acubated for 2 h at room temperature with the antiserum diluted 1:50 in this buffer containing 1% ovalbumln. The antiserum to D H F R was monospecific and polyclonal and was raised in New Zealand White rabbits to the 22 kDa enzyme

from L1210 murine leukemia cells purified to homogeneity by affinity chromatography [18]. The nitrocellulose membrane was then washed three times with Trissaline buffer containing 0.05% Tween 20 and twice with Tris-saline buffer, and then it was incubated for 2 h at room temperature with alkaline phosphatase conjugated goat anti-rabbit IgG diluted 1:1000 with Tris-saline buffer containing 1% gelatin. The membrane was then washed three times with Tris-saline buffer containing 0.05% Tween 20 and twice with Tris-saline buffer alone. The color of the bound conjugate was developed at 3 7 " C using 400 #g of Napthol AS-BI phosphoric acid and 200 ~g Fast Red Violet per ml of 100 mM Tris (pH 8.8)/1 mM MgCI~. Results

Purification of the enzyme The results of the typical purification of D H F R from 28 g of liver from 20 mice are summarized in Table I. The average weight of each liver was 1.4 g. The protamine sulfate, acidification with HC1 (pH 5.5), zinc sulfate precipitation, and 50-80% ammonium sulfate fractionation steps removed about 90% of the starting protein. Passage of the preparation through the MTXSepharose affinity column purified the enzyme 1000fold, yielding an overall purification of approx. 10000fold. The same procedure was followed for the purification of the livers from older and younger mice.

Polyacrylamide gel electrophoresis Electrophoresis of the final preparation on SDS-15% polyacrylamide gels separated two proteins with molecular weights of 22000 and 27500 (Fig. 1A, lane 2). There was no evidence of any other proteins on this gel. The relative concentration of each protein could not be determined from the stain intensity because the interaction of I':,: ~i:,.'cr ~t; in is not stoichiometrically related to the protein corcentration. Purified D H F R from LI210 marine leukemia cells [18] had a molecular weight of 22000 (Fig. 1A, lane 3) and this preparation lacked the 27.5 kDa protein. lmmunoblotting of these proteins using a monospecific polyclonal antiserum raised to the purified 22 kDa D H F R from the L1210 leukemia cells after transfer to a nitrocellulose membrane demonstrated two bands (Fig. IB, lane 1) corresponding to the 22 kDa D H F R and 27.5 kDa protein in the purified mouse liver preparation. Only a 22 kDa DHFR from the LI210 cells stained with the antiserum (Fig. 1B, lane 2). Having exhausted our supply of the purified proteins from the initial mouse livers for these studies, we purified another preparation using exactly the same procedure from livers having a mean weight of 2.4 g each. This time only the 22 kDa DHFR species of protein was obtained (Fig. 2A, lane 5). Fig. 2A, lane 3, shows

58 TABLE I Purifwation of mouse fiver DHFR ~

Purification step

Volume (ml)

Total protein b (mg)

Total DHFR protein ~ (mg)

Relative purity d

Crude cell extract

112

4525

0.30

6.6.10- 5

_

Precipitation with protamine sulfate

108

2290

0.26

1.1.10 a

1.7

87

Precipitation with acid and zinc sulfate

105

651

0.22

3.4-10- *

5.2

73

367

0.24

6.6-10 -,t

0.14

0.70

Precipitation with (NH4)2SO4 (50-80%)

4.7

MTX-Sepharoseeluate

6.0

0.20

Purification factor

Yield (%)

_

10

81

10606

46

The livers having an average weight of 1.4 g each were obtained from 20 mice. b The total protein is based on the Bio-Rad protein assay using bovine serum albumin as the standard. c DHFR protein was determined from the MTX-binding capacity of the fraction as previously described [18] and was converted to mg of DHFR by multiplying the mg of MTX bound by the ratio of 22000 (molecular weight of the predominant form of DHFR as determined ~,v SDS-PAGE) divided by 451 (molecular weight of MTX). a Ttfis is mg of DHFR divided by mg of total protein.

940006600045OO0-

30000-

L)I

20000_

14000-

~T I

2

3

1

2

A Fig. l. (A) SDS-15% PAGE of DHFR purified from mouse livers (1.4 g average weight) and LI210 leukemia cells. Lane 1. molecular weight markers (top to bottom: phosphorylase b, 94000; bovine serum albumin, 66000; ovalbumin, 45000; carbonic anhydrase, 30000: soybean trypsin ifihibitor, 20000; lysozyme, 14000); lane 2, mouse liver DHFR (200 ng); lane 3, LI210 DHFR (150 ng). The proteins were visualized by silver s~aining. (B) lmmunoblot analysis of DHFR transferred from the SDS-polyacrylamide gel (A) to nitrocellulose paper. Lanes 1 and 2 were exposed to antiserum raised in rabbit to DHFR purified from L1210 cells: Lane 1. mouse liver DHFR (400 ng); lane 2, LI210 DHFR (300 ng).

the two proteins purified from the livers having a mean weight of 1.,~ g. Since the only difference between the first and second preparations was the age of the mice from which the livers were obtained, we purified D H F R from the livers from 2-week-old mice having an average liver weight of 0.7 g and again both the 27.5 k D a protein and the 22 k D A species of D H F R were obtained (Fig. 2B, lane 2). Fig. 2B, lane 3, shows the single 22 k D a protein purified from livers weighing 2.4 g each. The same two proteins were purified from the livers of n e w b o r n mice (Fig. 2C, lane 3) having an average liver weight of 0.1 g. Fig. 2C, lane 5, again shows the 22 kDa protein purified from livers weighing 2.4 g each. Since a 27.5 k D a species of D H F R had not been previously reported, we believed initially that this protein was a c o n t a m i n a n t and therefore subjected the preparation to repurification through the MTX-Sepharose. Even with greater pre-elution washing of the column, we obtained exactly the same electrophoretic pattern by the S D S - P A G E . Boiling the purified enzyme with 2% SDS in the presence or absence of 5% 2-mercaptoethanol did not alter the migration of either protein band. We could not selectively elute either species of protein from the MTX-Sepharose column using 0.5 M K P O 4 ( p H 7.4) or 0.05 M KPO 4 ( p H 5.5) containing PteGlu or H2-PteGlu (0.2 m g / m l ) . Both proteins eluted together only when the p H of the eluting buffer was 8.9 and contained H2-PteGlu. A t t e m p t s to separate the two species of D H F R by gel-filtration, hydrophobic interaction chromatography, or anion- or cation-exchange chromatography of the ternary complex of protein-[ 3H ] M T X - N A D P H were unsuccessful. Although S D S - P A G E clea, ly separated the two proteins, we were unable to remove the SDS by

3000030000-

30000-

24ooo-

2,ooc-

20000-

20000-

1

3

5

.,

240O0-

' ~ i ~ . ~i

1

2o0oo-

23 1

A

B

3

5

C

Fig. 2. SDS-PAGE of D H F R purified from livers of mice of various ages. (A) Lane 1, molecular weight markers; lane 3, 200 ng D H F R from livers having an average weight of 1.4 g; lane 5. 200 ng D H F R from livers having an average weight of 2.4 g. (B) Lane 1, molecuhtr weight markers; lane 2, 200 ng D H F R from 2-week-old mousL; livers having an average weight of 0.7 g; lane 3, 200 ng D H F R from livers (2.4 g average weight). (C) Lane 1, molecular weight ,'narkers; lane 3, 200 ng D H F R from newborn mouse livers having an average weight of 0.1 g; lane 5, 200 ng D H F R from livers having an average weight of 2.4 g. The proteins were visualized by silver staining.

30000-

20000 . B

......~ii~i~~ii i~

N .

.

1

.

.

.

3

.

i~!~i!~i~:

5

7

Fig. 3. (A) NON-SDS 5%-12.5%-PAGE of DHFE from mouse livers having an average weight of 1.4 g. Lane 1, purified DHFR (] /~g) preincubated at room temp for 20 min in 50 mM Tris (pH 6.4); lane 2, purified DHFR (! /zg) preincubated at room temperature for 20 rain in 50 mM Tfis (pH 6.4), ]25 # M NADPH and 12 ng MTX. The gel was stained with MTT-tetrazolium as describ,~d in the Materials and Methods section. (B) SDS-12.5% PAGE of mouse liver DHFR following elution Irom a non-SDS-polyacrylamide gel. Lane 1, molecular weight markers (top to bottom: carbonic anhydrase, 30000; tr,jpsinogen, 24000; trypsin inhibitor, 20000); lane 3 contains the slowest migrating protein from the non-SDS-PAGE (first large band of Fig. 3A, lane 2); lane 5 contains proteins eiuted from the middle band of the non.SDS.polycrylamide gel (shown in Fig. 3A, lane 2); lane 7 contains the fastest moving area of DHFR from the non-SDS-polyacrylami¢;e gel (fastest h::,d cf Fig. 3A. lane 2). The proteins were visualized by ,',il',,.:~ s.,aini~o

precipitation as the potassium salt [8] following elution of the proteins from the gel. Accordingly, in order to determine whether each protein was catalytically active, n o n - S D S - P A G E was used, and enzyme activity was determined using an agarose overlay containing H~PteGlu, N A D P H and M T l ' - t e t r a z o h u m as described by Mell et al. [22]. O n a 5%-12.5% non-SDS gradient gel, :aly one wide b a n d was observed with the purified preparation alone (Fig. 3A, lane 1). However, when the preparation was incubated with N A D P H and MTX, three distinct protein bands were obtained, and each appeared to be catalytically active (Fig. 3A, lane 2). The amount of H2-PteGlu in the assay was in excess by 200000-fold over the amount of M T X a d d e d to the preparation before electrophoresis, and this was sufficient to displace the inhibitor. The tetrazolium reaction did not stain the gel if H2-PteGlu was c~mitted from the incubation mixture (data not shown). Silver staining of duplicate lanes of the polyacrylamide gel showed protein bands that matched exactly the catalytically active b a n d s (data not shown). To determine which of the catalytically active forms of D H F R identified by the non-denaturing P A G E corresponded to the proteins observed on the S D S - P A G E , one lane of the non-denaturing gel was stained for enzymatic activity and the adjacent gel was then cut in segments corresponding to the three areas of enzyme activity. The protein in each segment was then eluted in 0.15% SDS, concentrated, and run in the S D S - P A G E (Fig. 3B). The fastest moving catalytically active b a n d corresponded to the 27.5 kDa protein on the SDS-gel (Fig. 3B, lane 7); the catalytically active middle b a n d from the non-SDS-gel contained both the 27.5 and the 22 kDa proteins by S D S - P A G E (Fig. 3B, lane 5); and the slower moving catalytically active b a n d contained predominantly the 22 kDa enzyme and some of the 27.5 kDa species by S D S - P A G E (Fig. 3B, lane 3). Since the pure 22 kDa enzyme contained only the two slower nioving isoenxyme~ oa non-denaturing electrophoresis (data not shown), the contamination of the gel segments containing the two slower moving bands by the 27.5 kDa protein war likely due to dissociation of the highly negatively charged MTX from this protein during electrophoresis, resulting in slower migration as observed in Fig. 3A, lane 1. Treatment of the purified preparation containing both forms of the parified enzyme with agarose-coupledtrypsin followed by S D S - P A G E showed progressive digestion of both the 27.5 kDa protein and the 22 kDa D H F R (F-g. 4, lanes 8-10) as the incubation time of lhe reaction was increased. However, if N A D P H and M f X were added to the preparation before incubation with trypsin, the 22 kDa D H F R but not the 27.5 kDa protein was protected (Fig. 4, lanes 12-14). The disappearance of the 27.5 kDa protein by trypsin digestion did not yield any polypeptide smaller than 22 kDa.

30000-

2400020000-

1

3

5 6

8 9 10

12 13 14

Fig. 4. SDS-15% PAGE of the two proteins purified from mouse livers having an average weight of 1.4 g. Lane 1. molecular weight markers (top to bottom: carbonic anhydrase, 30000; trypsinogen, 24000; trypsin inhibitor, 20000; lactalbumin 14000); lane 3, purified mouse liver DHFR (200 ng); lane 5, DHFR incubated with 125 ,uM NADPH for 20 rain at room temperature; lane 6, DHFR incubated with 125 ,uM NADPH plus 14 ng MTX for 20 min at room temperature; lane 8, DHFR incubated with 125 ~M NADPH plus 0.07 U of trypsin for 5 min at room temperature; lane 9, DHFR incubated with 125 #M NADPH plus 0.07 U of trypsin for 5 min at 37°C; lane 10, DHFR incubated with 125/~M NADPH plus 0.07 U of trypsin for 45 min at 37°C; lane 12-14, the same as lanes 8-10, respectively,with addition of MTX (14 ng) to the reaction mixture. Lanes 2, 4, 7 and 11 served as controls and contained no protein. The proteins were visualized by silver staining.

Therefore, it is not possible to prove by this study that the 27.5 kDa protein was converted to the smaller form or whether it was completely digested by the trypsin.

Inhibition of catalytic activity by MTX The effect of M T X on the catalytic activity of the purified preparation containing both the 22 kDa D H F R and the 27.5 k D a protein is shown in Fig. 5. The

o

4

0 0.0

0.2

0.4

0.6

0.8

1.0

pmolea MTX added

Fig. 5. Inhibition by MTX of catalytic activity of the purified DHFR from. mouse livers having an average weight of 1.4 g. DHFR (0.7 pmol) was incubated in 50 mM citrate (pH 4.8) and 125 ,uM NADPH in the absence or presence of MTX for 20 min at room temperature. [3H]PteGlu was then added and after 20 rain at 37°C the reaction was stopped by the addition of unlabeled PteGtu and zinc sulfate [20].

titration of the inhibition of catalytic activity by MTX was linear (Fig. 5) and complete inhibition of enzymatic activity required 0.8 pmol of MTX. This value is in close agreement with the 0.7 pmol of total prot:iT, contained in the reaction mixture. The protein c.mcentration of the purified preparation was determined using the Bio-Rad protein assay, and the molar concentration was computed from the mean value of the molecular weights of the two proteins. The K m for the reduction of PteGlu by the mixture of both species of D H F R was 5 ~ M as computed from the Lineweaver-Burk double-reciprocal plot. With MTX in the assay reactions, a g i of 1 nM was obtained by plotting 1/v against the inhibitor concentration [23]. Discussion The results of these studies indicate that the liver from young mice contains a 27.5 kDa protein that reacts with a monospecific polyclonal antiserum raised to the 22 kDa D H F R enzyme. This 27.5 kDa protein was not found when livers having a mean weight of 2.4 g each were obtained from older mice for the purification of DHFR. Since treatment of the preparation containing both proteins with 2-mercaptoethanol before boiling with SDS did not affect their migration on SDS-PAGE, a disulfide bond does not appear to be involved in coupling an additional polypeptide domain to the 22 kDa species. The two proteins could not be separated in their native state in solution to permit effective characterization of the catalytic properties of the 27.5 kDa species. Three proteins, however, could be separated by non-SDS-PAGE and it appears that the protein in each band, one of which corresponded to the 27.5 kDa by SDS-PAGE, could reduce H2-PteGlu to H4-PteGlu directly on the gel. In addition, MTX on a mole for mole basis inhibited the catalytic activity of the final preparation containing both purified proteins. D H F R is an extremely well-studied enzyme because it is essential for maintaining the intracellular pool of reduced folate cofactors. For this reason it has been the target for inhibition by folate antagonists in the treatment of many forms of cancer and leukemia. Although a high-molecular-weight functional form of D H F R has been isolated from some trypanosomes [24] and a nonfunctional immunoreactive form of this protein has been identified in human leukocytes [18] and L1210 cells [25], the molecular weight of the functional enzyme from a variety of bacteria and mammalian cells [1,2] appears to range from 18000 to 22000. Even when an altered form of the enzyme with a lower affinity for MTX is expressed by cells following exposure to this inhibitor, the molecular weight of the enzyme usually remains unchanged [9,10], may be slightly increased [5-7] or, in one instance, was 42000 [8]. However, to the

best of our knowledge, there are no previous reports demonstrating two molecular species of D H F R isolated from normal mammalian tissue. Four distinct mRNA(s) from mouse liver containing 750 to 1600 nucleotides were found by Setzer et al. [26] to code for DHFR, but translated in vitro only a 21 kDa protein. The major difference in size of the RNAs was attributed to the length of the 3' untranslated region. Though several groups have studied D H F R in mouse liver (11-15), the size of the enzyme has not been previously established. Both the 27.5 kDa protein and the D H F R in the mouse liver bound tightly to the MTX-Sepharose affinity colunm so that they were not eluted with KPO 4 at pH 7.0, even with excess H2-PWGIu. Selective elution of either form with PteGlu at a lower pH was also unsuccessful. Both forms eluted from the affinity column with H2-PteGlu only when the pH of the buffer was raised to 8.9. The inhibition of catalytic activity by MTX also provides evidence that the 27.5 kDa protein has functional folate reductase activity. Complete inhibition of the catalytic activity of 0.7 pmol of protein in the purified preparation (using the average molecular weight of both proteins to compute this value) required 0.8 pmol of MTX. This type of stoichiometric inhibition is characteristic of the interaction of MTX with D H F R [27] and indicates that this inhibitor has substantially greater affinity than the [3H]PteGlu substrate for the catalytic sites of both species of protein. The 22 kDa DHFR, as a complex with MTX and NADPH, was resistant to digestion with trypsin, whereas the 27.5 kDa protein was completely digested in the presence of the inhibitor and cofactor. These findings suggest that the 27.5 kDa protein has a lower affinity than the 22 kDa DHFR for MTX a n d / o r N A D P H and that isolation of the higher-molecular-weight protein as a ternary complex with the inhibitor and cofactor would be difficult. Following digestion of the 27.5 kDa protein with trypsin it could not be determined from the comparative densities of the 22 kDa band in the digested and undigested control sample whether a greater amount of the 22 kDa species was present after digestion, a finding which would have indicated that the 27.5 kDa species was converted 1,; the sn~,~ller enzyme. A 5 kDa fragmeJ~: was not detected on the gel, bat this peptide might have been completcly digested by the trypsin. Under non-denaturing conditions, the purified forms of D H F R migrated as a wide band on PAGE, but these isoenzymes could be partially separated when the preparation was incubated with N A D P H and MTX before electrophoresis. By selectively eluting the separated proteins from the non-SDS-gel and subjecting them to SDS-PAGE, we could establish that the two slower moving and catalytically active bands of the nonSDS-PAGE corresponded to the 22 kDa pro!ein by SDS-PAGE. The fastest migrating band on tile non-

62 SDS-gel corresponded to the 27.5 kDa protein by SD.~PAGE. The change in migration of D H F R on nonS D S - P A G E in the presence of N A D P H and MTX appears to be a conformational change of the enzyme previously observed for D H F R by Dully et al. [28]. There have been reports that mouse liver contains two isoenzymes of D H F R with Pi values of 8.1 and 8.7 [1<151. A n o t h e r important finding concerning '.he 27.5 k D a protein in the mouse liver was that it was isolated only from the liver of y o u n g animals. Although the functional properties of D H F R in embryonic tissues have been reported to be similar to the enzyme in mature animals [29], the size of the protein in fetal tissue has not been established. The relativc concentration of the 27.5 kDa protein (vis-a-vis the 22 k D a enzyme) in the livers from n e w b o r n mice did not appear greater than the concentration of the protein from the liver of 2week-old mice, suggesting that, if the higher-molecularweight species is a 'fetal enzyme', its expression begins to decrease early in fetal development. The persistence of this higher-molecular-weight protein after birth may be a consequence of the slow turnover of liver cells. Acknowledgements This work has been supported by the Medical Research Service of the Veterans Administration znd .the Kresevich F o u n d a t i o n of New York. References 1 Blakeley, R.L. (1984) Folates and Pterins, Chemistry and Biochemistry Folates (Blakeley, R.L. and Benkovic, S.J., eds.), Vol. I, pp. 191-253, John Wiley, New York. 2 Freisheim, J.H. and Matthcws, D.A. (1984) Folate Alatagonists as Therapeutic Asents (Sirotnak, F.M., Btrchall, J.J., Ensminger, W.D. and Montogomery,J.A., eds.), Vol. 1. pp. 69-131, Academic Press, Orlando, FL.

3 Albrecht, Alberta M. and Biedler, June L. (1984) Folate Antagonists as Therapeutic Agents (Siromak. F.M., Burchall, J.J.. Ensminger, W.D. and Montgomery,J.A., eds.). Vol. 1. pp. 317-353, Academic Press, Orlando, FL. 4 Schimke, R.T. (1984) Cancer Res. 44. 1735-1742. 5 Dedhar, S. and Goldie, J.H. (1983) Cancer Res. 43, 4863-4871. 6 Lovesay, A.C. (1071) Biochem. Pharmacol. 20, 2033-2041. 7 Melera, P.W., Hession. C.A., Davide, J.P., Scotto. K.W., Biedler, J.L., Meyers, M.D. and Shanske. S. (1982) J. Biol. Chem. 257, 12939-12949. Dedhar, S. and Goldie, J.H. (1985) Biochem. Biophys. Res. Commun. 129, 536-545. Flintoff, W.F. and Essani, K. (1980) Biochemistry 19, 4321-4327. Haber, D.A., Beverley. S.M., Kiely, M i . and Schimke, R.T. (1981) J. Biol. Chem. 256, 9501-9510. Jarabek, L. and Bachur, N. (1971) Alch. Biochem. Biophys. 142, 417-425. McCullough, J.L. and Bertino, J.R. (1971) Biochem. Pharmacol. 20, 56~-574. Soucek, J. and Slavikova, V., Kuchar, M. and Kotynkova. E. (1972) Biochem. Pharmacol. 21, 1907-1919, Schalhorn, A. (1980) Enzyme, 25, 241-249. Soucek, J., Zizkovsky, V. and Fort. P. (1977) Acta Univ. Carol. Medical 79, 193-200. Galleli, J.F. and Yokoyama, G.J. (1967) Pharma. Sci. 56, 387-389. Kaufman, B.T, and Kemerer, V.F. (1976) Arch. Biochem. Biophys. 172, 289-300. Rothenberg, S.P. and lqbal. M.P. (1982) Proc. Natl. Acad. Sci. USA 79. 645-649. Kaufman. B.T. and Pierce, J.V. (1971) Biochem. Biopbys. Res. Commun. 44, 608-613. Rothenberg, S.P. (1966) Anal. Biochem. 16, 176-179. Laemmli, U.K., (1970) Nature (Lond.I 227~ 680-695 Me!l. G.P., Martctli, M., Kirchner, J. and Huer~r!eken,~, F.M. (1968"~ Ric,chem. Biophys. Res. Commun. 33, 74-79. Freudental, R. (1970) Anal. Biochem. 35, 346-350. Gutteridge, W.E., Jaffe. J.J. and McCormack, J.J. (1969) Biochim. Biophys. Acta 191,753-755. Rothenberg, S.P. and Iqbal, M.P. (1983)Cancer Res. 43, 529-535. Setzer, D.R., McGcu~;an, M., Nunbe'rg, J.H. and Schimke, R.T. (1980) Cell 22, 361-370. Werkheiser, W. (1961) J. Biol. Chem. 236, 888-893. Duffy, T.H., Beckman. S.B., Sato, J.K., Nagae, H., Vitols. K.S. and Huennekens, F.M. (1985) Adv. Enzym. Regul. 23, 3-12. Roberts, D. and Hall, T.C. (1965) Cancer Res. 25, 1894-1898