THE JOURNAL OF BIOLOGICAL CHEMISTRY (6 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Val. 266, No. 5, Issue of ’ February

Human NAD+-dependent Mitochondrial Malic Enzyme cDNA CLONING, PRIMARY STRUCTURE, AND EXPRESSION IN ESCHERICHIA COLI*

(Received for publication, August 27, 1990)

Gerhard LoeberS, Anthony A. Infante§, Ingrid Maurer-Fogy, Edeltraud Krystek, and Mark B. Dworkin From the Ernst Boehringer Institut, Dr. Boehringergasse 5-11, A-1121 Vienna, Austria

Mitochondrial NAD+-dependent malic enzyme (EC 1.1.1.40) is expressed in rapidly proliferating cells and tumor cells, where it is probably linked to the conver- sion of amino acid carbon to pyruvate. In this paper, we report the cDNA cloning, amino acid sequence, and expression in Escherichia coli of functional human NAD+-dependent mitochondrial malic enzyme. The cDNA is 1,923 base pairs long and contains an open reading frame coding for a 584-amino acid protein. The molecular mass is 65.4 kDa for the unprocessed precursor protein. Comparison of the amino acid se- quence of the human protein with the published NADP+-dependent mammalian cytosolic or plant chlo- roplast malic enzymes reveals highly conserved re- gions interrupted with long stretches of amino acids without significant homology. Expression of the proc- essed protein in E. coli yielded an enzyme with the same kinetic and allosteric properties as malic enzyme purified from human cells.

Malic enzyme (ME)’ catalyzes the oxidative decarboxyla- tion of malate to pyruvate, malate + NAD(P)+ --* pyruvate + CO, + NAD(P)H+, and can be found both in eukaryotic and prokaryotic cells. Three different isoforms of ME have been described in mammalian tissues: a strictly cytosolic NADP+- dependent enzyme, an NADP+-dependent mitochondrial iso- form, and a mitochondrial isoenzyme which can use both NAD’ and NADP+ but is more effective with NAD’ (Fraen- kel, 1975). The mammalian isoforms are about 62-64 kDa in size (Moreadith and Lehninger, 1984b; Magnusson et al., 1986; Bagchi et al., 1987). A native size of 240,000 Da suggests a tetrameric structure for the active enzyme (Fraenkel, 1975; Moreadith and Lehninger, 1984b).

The highest levels of the cytosolic ME activity are found in the liver and in adipose tissue, where this isoform is linked to the generation of cytosolic NADPH for de novo fatty acid synthesis. This isoenzyme is under dietary control and can be induced by a carbohydrate-rich diet or thyroid hormones (Fraenkel, 1975; Dozin et al., 1985). NADP+-dependent mi- tochondrial ME activity is found in many tissues, including brain, heart, and skeletal muscle and adrenals (Lin and Davis,

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence($ reported in thispaper has been submitted

M559Ot5. to the GenBank”/EMBL Data Bank with accession number(s)

$ To whom correspondence should be addressed. § On sabbatical leave from the Dept. of Molecular Biology and

’ The abbreviation used is: ME, malic enzyme. Biochemistry, Wesleyan University, Middletown, CT 06457.

1974; Fraenkel, 1975; Nagel et al., 1980), tissues which also express the soluble NADP+-dependent ME isozyme. This enzyme may be important for the cycling of NADPH into the mitochondria for biosynthetic reactions there (Simpson and Estabrook, 1969).

Mitochondrial NAD+-dependent ME activity can be found in tissues which undergo high rates of cell division, such as spleen, thymus, and the basal cells of the small intestinal mucosa (Sauer et al., 1979; Nagel et al., 1980). It is also expressed during the rapid cleavage stages of early Xenopus development (Dworkin and Dworkin-Rastl, 1990). Activity for this isoform is low or absent in brain, muscle, and normal and regenerating liver tissue from rat (Nagel et al., 1980) but has been reported in rat adrenal cortex (Sauer, 1973), pigeon and human skeletal muscle (Lin and Davis, 1974; Taroni et al., 1988), and in heart muscle of several species (Lin and Davis, 1974; Liguori et al., 1989). It is also expressed in mitochondria of all tumor cells investigated to date, including ascites tumors (Sauer and Dauchy, 1978), hepatoma cells (Sauer et al., 1980), and a variety of other tumors and trans- formed cell lines (Moreadith and Lehninger, 1984b).2 In the Morris hepatoma series, expression of the NAD+-dependent ME is progression linked (Sauer et al., 1980).

Many tumor cells in culture are able to use glutamine as well as glucose as their main respiratory fuel (Reitzer et al., 1979), and many malignant cell lines do not even have an absolute requirement for glucose per se (Wice et al., 1981). The NAD+-dependent mitochondrial ME activity, then, may be linked to the conversion of amino acid carbon to pyruvate (Sauer et al., 1980). There is some direct evidence supporting this role (Moreadith and Lehninger, 1984a). In this paper, we report the isolation and nucleotide sequence of a cloned cDNA of human mitochondrial NAD+-dependent ME and the de- duced amino acid sequence of the protein. We have expressed the functional protein in E. coli and have compared the allosteric and kinetic properties of the purified recombinant protein with natural human lymphocyte NAD+-dependent ME and endogenous bacterial ME.

EXPERIMENTAL PROCEDURES3

RESULTS AND DISCUSSION

Cloning and Nucleotide Sequence of Human Mitochondrial NAD+-dependent Malic Enzyme cDNA-Our approach to cloning NAD+-dependent ME was to identify a human cell line that expressed the isozyme and to purify it in order to

’ M. B. Dworkin and A. A. Infante, unpublished results. ’’ Portions of this paper (including “Experimental Procedures,” Table 1, and Figs. 1-4 and 6) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

3016

Human NAD+-dependent Mitochondrial Malic Enzyme 3017

get amino acid sequence information. We found that many transformed lymphocyte cell lines express this isoform of ME and purified it from one such line using ion exchange and affinity chromatography. The cloned sequence was identified from a human fibrosarcoma library using a polymerase chain reaction with degenerate oligonucleotides based on the amino acid sequence of isolated tryptic peptides (see “Experimental Procedures”). The complete nucleotide sequence and the de- duced amino acid sequence of the human mitochondrial ME are shown in Fig. 3. The cDNA clone has an overall length of 1,923 base pairs, contains an open reading frame coding for 584 amino acids between positions 90 (ATG) and 1844 (TAG), and ends with a poly(A) tail (Fig. 3). The 5”untranslated region is extremely G/C-rich (76.3% G/C; 61 out of the first 80 nucleotides are G/C). A high G/C content has also been found in the 5’-untranslated region of NADP+-dependent cytosolic ME genes from rat (Magnusson et al., 1986) and mouse (Bagchi et al., 1987) and an NADP+-dependent ME gene from maize chloroplasts (Rothermel and Nelson, 1989). In contrast to human NAD+-dependent ME, in the NADP+- dependent ME genes the G/C-rich region extends consider- ably into the coding region. The nucleotides surrounding the ATG start codon match reasonably well with the Kozak consensus sequence (Kozak, 1987). The 3’-untranslated re- gion is 79 base pairs long, followed by the poly(A) tract; 26 consecutive A residues were found in the isolated cDNA clone. The closest match (AGATAA) to the consensus polyadenyl- ation signal (AATAAA) (for review, see Birnstiel et al. (1985)) is located 13 base pairs upstream of the poly(A+) tract.

The predicted molecular weight deduced from the amino acid sequence is 65,400 Da (including the mitochondrial leader sequence), which is similar to the size of the processed mito- chondrial NAD+-dependent ME isolated from canine small intestine as seen in sodium dodecyl sulfate-polyacrylamide gels ( M , = 62,000 (Nagel and Sauer, 1982)). The first 20 amino acids of the predicted protein show the characteristics of a typical mitochondrial leader sequence: a possible amphi- philic CY helix in a region of considerable potential secondary structure, with nonpolar amino acids on one side of the helix and basic amino acids on the other side (von Heijne, 1986; Allison and Schatz, 1986). Of the protein’s 584 amino acids, 67 are acidic, 78 are basic, and 11 are cysteines. Although glycosylation has been reported in a cytosolic NADP+-de- pendent ME isolated from human heart muscle (Taroni and DiDonato, 1988), N-linked glycosylation consensus sequences are absent from the human NAD+-dependent ME. Glycosyl- ation of asparagine 421 (Asn-Pro-Thr) is not possible because of the proline residue (Mononen and Karjalainen, 1984).

Comparison of the Amino Acid Sequence with Other Known Malic Enzymes-Sequences of four other malic enzymes have been published to date: cytosolic NADP+-dependent ME from mouse (Bagchi et al., 1987) and rat (Magnuson et al., 1986), an NADP+-dependent ME from maize chloroplasts (Rother- me1 and Nelson, 1989), and a thermostable ME from Bacillus stearothermophilus (Kobayashi et al., 1989). Table 2 shows the percentage of amino acid identity and similarity (scores are from the “gap” alignment program of the UWGCG pro- gram package (Devereux et al., 1984)) of all known malic enzymes to each other plus a comparison of the different ME sequences with bean cinnamyl alcohol dehydrogenase (Walter et al., 1988), an NADP+-dependent enzyme involved in lignin synthesis with a high sequence similarity to the malic en- zymes.

The sequence of human mitochondrial NAD+-dependent ME is significantly different from that of rat and mouse cytosolic ME, the two rodent cytosolic enzymes are 95%

TABLE 2 Comparison of the amino acid sequences of human mitochondrial NAD+-dependent ME (human mt.ME), rat and mouse cytosolic

NADP-dependen t ME (rat and mouse cyt.ME, respectively), maize chloroplast M E (maize chLME), NAD+-dependent M E from B.

stearothermophilus (bac.st. ME), and bean cinnamyl alcohol dehydrogenase (cin.alc. dehydr)

The first number of a matching pair indicates the percentage of amino acid identity; the second number indicates the percentage of amino acid identity plus similarity. Scores are from the “gap” program of the University of Wisconsin Genetics Computer Group program package (Devereux et al., 1984).

Human Rat Mouse Maize Bac.st. mt.ME cyt.ME cyt.ME chi.ME ME

Rat cyt. ME 55.4/72.9 Mouse 54.2172.5 95.3197.2

Maize 472167.1 50.4166.5 50.2167.0

Bacst. ME 27.4/52.4 26.5/50.0 26.5/49.6 28.3/52.5 Cin.alc. 48.4/67.6 52.7/67.9 51.8/67.6 72.8/84.7 27.0/52.1

cyt.ME

chl.ME

dehydr.

identical with each other. The human NAD+-dependent mi- tochondrial enzyme shares only 54-56% identity with the rat and mouse ME, which is only slightly higher than the 47- 49% identity it shares with maize chloroplast ME or bean cinnamyl alcohol dehydrogenase. The eukaryotic ME se- quences have low but significant similarity (27% identity) to the thermostable ME from B. stearothermophilus. The ho- mology between cinnamyl alcohol dehydrogenase and the ME sequences (for example 72.8% identity with maize malic en- zyme) is unexpected, since the substrates for the two enzymes, cinnamyl alcohol/cinnamaldehyde and malate/pyruvate, re- spectively, and their catalytic activities, reduction, and reduc- tive decarboxylation, respectively, are quite different. Whether cinnamyl alcohol dehydrogenase has malic enzyme activity is unknown.

The similarity among the different ME sequences is not distributed evenly over the protein; rather, highly conserved stretches are interrupted with longer segments which are poorly conserved. Fig. 4 shows a direct linear alignment of the human mitochondrial ME with rat cytosolic ME and maize chloroplast ME and serves to highlight conserved do- mains in the protein. It is striking that several long stretches of amino acids are almost identical among the ME sequences from human, rat, and maize, whereas other regions display little similarity. Particularly conserved are regions from amino acids 160 to 189, 251 to 294, and 411 to 452 (the numbering refers to the human sequence).

The region between amino acids 111 and 119 has similarity to the proposed NADP+ binding sites of goose fatty acid synthetase and human glyceraldehyde 3-phosphatase (Pou- louse and Kolattukudy, 1983; Fig. 5 A ) . A sequence element that matches the consensus sequence for the “ADP binding pap fold,” as proposed by Wierenga et al. (1985), is located between amino acids 163 and 193 (Fig. 5B). These structural elements are thought to bind the ADP moieties of the dinu- cleotides and have a characteristic arrangement of glycines (indicated by a G in Fig. 5 B ) and nonpolar ( n ) and hydrophilic (h) residues at certain crucial positions on a compact pap fold. A segment between amino acids 311 and 343 includes a motif which also has some homology to a proposed nicotina- mide coenzyme binding site (Scrutton et al., 1990, Fig. 5 C ) . Scrutton et al. (1990) were able to alter the cofactor specificity of glutathione reductase from NADP+ to NAD+ by altering two alanines (indicated by arrows in Fig. 5 C ) to glycines and regions with some basic amino acids (which may establish

Human NAD+-dependent Mitochondrial Malic Enzyme 3018 A : ... .,.

htWE V Y T P T V G L A gFAS V P T - T V G S A hGAP V F T - T M E K A

1:9

B: 161 GERILGLGDLGVYGMGIPVGKLCLYPACAG?.

~".G.~"G".""n""---"n

: 9 1

C: hmtm

111 343

=ME GAGEAALGIANLIVMSMVENGLSEQEAQKKIWH GAGE~LGIARLIVHIUIEKEGLSKE~RQKIW~

CONS,NADP G . ~ " ~ " ~ ~ " " " G " " " n b - - - - - b - ~ CONS. N m G-G"G"-G"""G""..~"-lll;-

T t

FIG. 5. A , comparison of putative dinucleotide binding sites of human NAD+-dependent ME (hmtME), goose fatty acid synthetase ( # A S ) , and human glyceraldehyde 3-phosphatase (hGAP). Amino acids are given in the single letter code. Regions of amino acid identity are in boldface. B, putative ADP binding pap fold of human NAD+- dependent ME. Consensus sequences as identified by Wierenga et al. (1985) are shown above the ME sequence. h, hydrophilic amino acid; n, nonpolar amino acid; G, glycine in the consensus sequence; a dash indicates a position for any amino acid. C, comparison of the NAD(P)+ binding fold of NAD+-dependent human ME (hmtME) and NADP+- dependent rat ME ( r M E ) with the consensus sequences for NADP+ and NAD' binding as identified by Scrutton et al. (1990). Amino acids are shown in the single letter code. n, neutral; b, basic; a, acidic amino acid. Glycines crucial for the discrimination of the coenzymes are indicated by arrows. The two underlined regions contain ( a ) basic amino acid(s) required for NADP+ binding (see text).

contact with the phosphate group of the NADP', underlined in Fig. 5C) to nonpolar or acidic residues. In the case of malic enzyme, however, the coenzyme specificity cannot be easily deduced from the sequence. Both the rat NADP+-dependent ME and the human NAD+-dependent ME have alanines at the positions of the arrows, although in the first underlined region human NAD+-dependent ME has two glutamates and rodent NADP+-dependent ME has two lysines. These puta- tive dinucleotide binding domains are conserved among the different ME sequences described so far, as well as in cinna- my1 alcohol dehydrogenase. Three cysteines at positions 120, 198, and 428 are also conserved among all these different enzymes and are possible sites for disufide bridges.

Regions of the NAD+-dependent ME with little homology to NADP+-dependent ME are amino acids 1-90 (including the mitochondrial leader sequence), the carboxyl terminus between amino acids 500 and the end of the protein, and a region between amino acids 355 and 410. In this last region, NAD+-dependent human ME shares only 33-36% identity with the other eukaryotic NADP+-dependent ME sequences. Among the NADP+-dependent ME sequences from rodents and plants this region shows 47-67% amino acid identity. It is therefore possible that this region of the protein is respon- sible for some of the unique allosteric properties (for example, stimulation by fumarate, inhibition by ATP) of the human mitochondrial ME.

Expression of the Human Mitochondrial NAD+-dependent M E in Bacteria-The expression plasmid coding for amino acids 19-584 (complete coding region of ME minus the puta- tive mitochondrial leader sequence) resulted in high expres- sion of a novel protein of M, 64,000, the expected size of human NAD+-dependent mitochondrial ME. The human re- combinant ME was separated from bacterial ME by ion exchange chromatography (see "Experimental Procedures"), and the purified protein had a specific activity of 35 milli- units/bg, which is similar to the specific activity reported for NAD+-ME isolated from human muscle tissue (Taroni et al., 1988). A construct expressing the full-length coding region (including the putative mitochondrial leader) did not produce any active protein in the supernatant, although sodium do- decyl sulfate gel electrophoresis showed a significant induc- tion of a protein of the correct size. Whether the absence of

activity of the unprocessed precursor protein was due to complete insolubility or to interference of the leader peptide with enzyme activity is currently under investigation.

The bacterially produced human mitochondrial NAD+-de- pendent ME was characterized by its substrate and allosteric characteristics, which further enabled it to be distinguished from the endogenous bacterial ME activity and from other ME isoforms. A comparison of the activities of the recombi- nant human ME, the bacterial ME, and the enzyme purified from transformed human lymphocytes is shown in Fig. 6. The bacterial enzyme uses both NAD' and NADP+ effectively as the electron acceptor, whereas the recombinant human en- zyme is virtually inactive with NADP+. This strong preference for NAD' was also observed for the natural human enzyme isolated from lymphocytes (Fig. 6A). Fumarate, an activator of NAD+-dependent ME (Mandella and Sauer, 1975), acti- vated the recombinant and natural human ME similarly but had no effect on the endogenous bacterial enzyme (Fig. 6B). ATP inhibited the activity of both the NAD+-dependent natural human ME and recombinant ME but did not inhibit the endogenous bacterial ME (Fig. 6C).

The K,,, value of several preparations of the recombinant human ME for malate was 1.6 * 0.6 mM in the presence of fumarate. This matches well with the published K,,, values for human NAD+-dependent ME of 1.18 mM (Zolnierowicz et al., 1988) and 1.91 mM (Taroni et al., 1988). In the absence of fumarate, the K,,, for malate was quite high, approximately 10 mM. The K,,, for malate of the endogenous bacterial ME was 0.53 f 0.3 mM in the presence and absence of fumarate.

In sum, we have cloned, sequenced, and expressed in E. coli human mitochondrial NAD+-dependent ME. The enzyme expressed in bacteria has similar kinetic characteristics as the natural human NAD+-dependent ME, strongly suggesting that extensive posttranslational modifications are not neces- sary for protein activity. It further indicates that the active tetramer is composed of identical subunits. The cDNA clone can now be used to investigate the expression of this isozyme in cells in order to study its role in the intermediary metabo- lism of rapidly proliferating cells.

Acknowledgments-We wish to thank C. Stratowa for the human HS913 cDNA library, R. Hauptmann for the expression vector pRH 281, D. Infante and T. Gramanitsch for expert technical assistance, and E. Dworkin-Rastl for discussion and support. We are particularly grateful to Dr. L. A. Sauer for many discussions about NAD'- dependent ME and its possible role in tumor metabolism.

REFERENCES Allison, D. S., and Schatz, G . (1986) Proc. Natl. Acad. Sci. U. S. A.

Bagchi, S., Wise, L. S., Brown, M. L., Bregman, D., Sul, H. S., and

Benton, W. D., and Davis, R. W. (1977) Science 196, 180-185 Birnstiel, M. L., Busslinger, M., and Strub, K. (1985) Cell 41, 349-

Bodo, G. (1981) Methods Enzymol.78,69-75 Devereux, J., Haeberli, P., and Smithies, 0. (1984) Nucleic Acids Res.

Dozin, B., Magnusson, M. A., and Nikodem, V. M. (1985) Biochem-

Dworkin, M. B., and Dworkin-Rastl, E. (1990) Deu. Biol. 138, 177-

Dworkin-Rastl, E., Swetly, P., and Dworkin, M. B. (1983) Gene

Fraenkel, R. (1975) Curr. Top. Cell. Regul. 9, 157-181 Kobayashi, K., Doi, S., Negoro, S., Urabe, I., and Okada, H. (1989)

Kozak, M. (1987) Nucleic Acids Res. 15,8125-8148 Kyhse-Andersen, J. (1984) J. Biophys. Biochern. Methods 10, 203-

Laemmli, U. K. (1970) Nature 227, 680-685

83,9011-9015

Rubin, C. S. (1987) J. Biol. Chem. 262, 1558-1565

359

12,387-395

istry 24,5581-5586

187

(Amst . ) 21, 237-248

J. Biol. Chem. 264,3200-3205

209

Human NAD'-dependent Mitochondrial Malic Enzyme 3019

Liguori, M., Tessarolo, D., Ferro, M., and Giacanelli, M. (1989) Ado.

Lin, R. C., and Davis, E. J. (1974) J. Biol. Chem. 249, 3867-3875 Magnuson, M. A., Morioka, H., Tecce, M. F., and Nikodem, V. M.

Mandella, R. D., and Sauer, L. A. (1975) J. Biol. Chem. 250, 5877-

Mononen, I., and Karjalainen, E. (1984) Biochim. Biophys. Acta 788,

Moreadith, R. W., and Lehninger, A. L. (1984a) J . Biol. Chem. 259,

Moreadith, R. W., and Lehninger, A. L. (1984b) J. Biol. Chem. 259,

Nagel, W. O., and Sauer, L. A. (1982) J. Biol. Chem. 257, 12405-

Nagel, W. O., Dauchy, R. T., and Sauer, L. A. (1980) J. Biol. Chem.

Poulouse, A. J., and Kolattukudy, P. E. (1983) Arch. Biochem. Bio-

Reitzer, L. J., Wice, B. M., and Kennell, D. (1979) J. Biol. Chem.

Rothermel, B. A., and Nelson, T. (1989) J. Biol. Chem. 264, 19587-

Myochem. 2,203-204

(1986) J. Riol. Chem. 261, 1183-1186

5884

364-367

6215-6221

6222-6227

12411

255,3849-3854

phys. 220,652-656

254,2669-2676

19592

Sanger, F., Nicklen, D., and Coulson, A. R. (1977) Proc. Natl. Acad.

Sauer, L. (1973) Riochem. Biophys. Res. Commun. 50,524-531 Sauer, L. A., and Dauchy, R. T. (1978) Cancer Res. 38, 1751-1756 Sauer, L. A., Dauchy, R. T., and Nagel, W. 0. (1979) Biochem. J.

Sauer, L. A., Dauchy, R. T., Nagel, W. O., and Morris, H. P. (1980)

Scrutton, N. S., Berry, A., and Perham, R. N. (1990) Nature 343,

SimDson. E. R.. and Estabrook. R. W. (1969) Arch. Biochem. Biophys.

Sci. U. S. A. 74,5463-5467

184,185-188

J. Biol. Chem. 255,3844-3848

38-43

129,384-395 Taroni. F.. and DiDonato. S. (1988) Int. J. Biochem. 20.857-866 Taroni: F.; Gellera, C., and DiDonato, S. (1988) Biochlm. Biophys.

von Heijne, G . (1986) EMRO J. 5,1335-1342 Walter, M. H., Grima-Pettenati, J., Grand, C., Boudet, A. M., and

Lamb, C. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,5546-5550 Wice, B. M., Reitzer, L. J., and Kennell, D. (1981) J . Riol. Chem.

Wierenga, R. K., De Maeyer, M. C. H., and Hol, W. G. J. (1985)

Zolnierowicz, S., Swierczynski, J., and Zelewski, L. (1988) Biochem.

Acta 916,446-454

256,7812-7819

Biochemistry 24,1346-1357

Med. Metab. Biol. 39, 208-216

Human NAD*-depmdcn l M i lochondr ia l Ma l ic Enzyme: cDNA Cloning. Primary Slructure and Expression in E coli.

Gerhnrd Loeher. Anthony A. Infnnte. Ingrid Maurer-l'ogy. Edcltraud Kryrtek and Mark 0. Dworkin

b y

C : x p e r i n l e n I a l P r o c e d u r e s :

Fu r i f i r a l i on o f nli locllondrial NAD+.drpcndenI malic enzyme from a trans- formed human lympharyle cel l line. Mitachondrirl NAD+.dcpcndenl malic en. lyme (ME) was purified from lhe human ~tansformcd T-lymphoeyle cell line 1301 (G.

-2.5 I 1010 ~ymp~mcytes were c o ~ ~ e c ~ c d at 2.500 rpm for 5 min in B lleracus Clltirt Hdo. 19x1). following iu gcnerrl the procedure of Morcsdilh and Ishninger (1984h).

nminiluge 2. rinsed with a buffer conmining R2.5 m M NaCI. 2.5 m M KCI. I mM Na211P04. 5 m M IIEPES pl l 7.9, and homogcni7cd in 400 nd hulfcr A (20 m M Tcir.IICI pl l 7.4. I nth4 EDTA. 0.25 M rucrorel. Nuclei and dehris were removed by spinning lor IO Inin at 2.500 rpm i n a Bcckmatm JA2l rotor. hlmd,ondria wcrc then pellcled lor IS mi" at l(1.000 rpm i n lhe same rolor. The mitochondria were washed twicc with huflcr A. suspended i n 50 ml 5 mhf IIEPIIS. pl l 7.6. SO m M KCI. 0.1 m M EDTA. 0.5 mM

mmimum lcvd on kc . cooling 1.2 min hclwcetl sonicalion periods. l h e sonicalc was DIT. 0 . 2 % Luhrol PX (Sigma) and sonicated 6 x 30 IEC with a Branson sonicator at

lrroughl to IS 4. glycerol and spun at 40.000 rpm for 30 lmin wing a Dcckmann TiSU.2 m o r . Thc supernalanl was diluted with one volume of huller D (30 mM Tris-IICI, p l l 7.4. 1.5 I ~ M M ~ C I ~ . 0.5 ~ I M w r A . 0.2 SUM m-r, 2 0 % g~ycero~) :tnd theta applied IO a

wmhcd with 200 t n 1 huller I) and lhen eluted will1 a 40 . 120 m M KC1 gradient in 12 x 2.5 cm DliAI: CCIIUIOSC (Whrtman I)iiS2) ion exchsnge COIUIIIII . 'The column wa1

huller D. NAW.dependcst n d i c cnzymc clulrd a1 X0 . 90 niM KCI. l:raelions con~:tin. ing MII aclivity were pooled. diluted wit0 w e volu~nc 01 huller D. hroughl to I mM lumrrnte and i mM MnC12 and !hen loadcd onlo i( 5 ml ATP affinity column (Type II A'IP agarose. Pharmacia). ll ighiy purilicd ME was cluled from lhe ATP column will) bufler I) + I(x) mM KC1 + 4 mM NAD+. diluted will, one volume huller D and concen. lratcd hy hinding to a 2 ml DEAE column and eluting ~ 8 t h 1 0 0 mM KC1 in huller D without glyccrol. For sequenenng. the M E wits crlenrively di:tlyred against 2.5 mhl Tr i r p l l 7.5. lyophilized Io drynew. 2nd dissolved in SDS gel ;applicalion huller (Laemnlli. 1970). l h e first DIiAE column yielded 6.400 mU NAD+.dependent ME (I #nu = I m m l c NADl l formedl min :,I 33C). and the final yicld was 30 %. I:ig. I sllows gcI clecrrophorerir of lhc I ~ w m a n prolein alter lhe ion exchange column ( a ) and after lhc ATP agarose affinity column (h). Two pOiypCptidCs of 35 kDa wcre also prcqcn~ it, ;dl prcprr;nltons of ME from lymplnoeytcr.

Amino acid ~equen~ing of mitochondr ia l NAD+-dependenf mal ic enzyme. l'hc purified M E ( -50 pg) was eleetrophorcred on 3 I0 % SDS.polyacryiamid gel (I.rcmmli. 1970) and blotted onlo an Immohilon P mcmhranc (Millipore) using thc remi.dry.proecdure with a discontinuous bufler (300 m M I t i s . p l l 10.4; 25 niM Tris. pl l 10.4: 25 mM Tris. 40 m M r.aminocapronic acid. p l l 9.4: 911 hullers contained 20 % melhanol: Kyhse.Andcrsen. 1984). Since there was indication of N.tcrminal blockage. MI: was cleaved directly on thc membrane with trypsin. Memhnne pieces conlaining line 64 kDa protein (idenlified by Coomassic hlue staining of I neighhoring lane) werc suspended i n 200 ml 1 0 0 m M Trir. p l l X.2. 5 % acclonilrile. and 2 % wlw trypsin (Iloehringer Mmnheim. sequencing grade). incuhaled 6 11 at 37°C. and 11csh trypsin was addcd for an additional 18 h. Thc released tryptic pcptides wcrc scprraled hy reverse phase l lPLC urizlg a Waters Vcll:~ Pak CIX column (3.9 x IS0 mnl, 5 mm psr- t ick diameter) at 30°C and 0.1 70 trifluoroacelie reid in water (A) and ileetOnitrilC (I)) as mobile phrrer (Fig. 2). l h c gmdicnt was 0 . 55 0 B i n 55 min. The major lryptic pepfidcr were collected. lyophiiizcd. rnd lhe residues dissolved in 7.5 ml 70 0 formic :acid and directly applied to the cmridgc 01 a pulsed liquid phase scquenalor ( A D L 477A). For each sequencing run. one Fil-l.cyclc. one Ilegin.l.cycle and 20 10 40 Norm~l.l.eyeles wem used. The cycle conditions wcrc as recommended hy the nunut- fncturer. Thc phcnylthiohydantoin derivslives of clcaved amino acids wcrc automa. ticnlly lransferred to rev~rse phase III'LC (AB1 120A) and identified hy comparison of lheir retention limes to a SEI of standard amino acids. Tahle I A shows Ihe peptide ~cquences from the tryptic fragments 01 human ME.

A B (D

9 9

6 7

4 3

30

C kD

"66 Fig. I: SDS.polyrerylrmide gel

-45 eleetropharcris (Laemmli. 1970)

4 6 of the human NAD+.dcpendent mahc enzymc. A.0: Natural hu. man protein purified from lym-

=2g,24 phoeytcr after DEAE ccl lulox chromatography ( A ) and alter ATP agarose aflinily column (B). C: The recomhinnnl protcin puri. ficd from induced E.coli X L I hluc cells. Gclr were rtrincd using the Diorad silver staining kit.

Human NAD+-dependent Mitochondrial Malic Enzyme 3020

Human NAD+-dependent Mitochondrial Malic Enzyme

Natural Recornbmant Baclerlal

3021