J. Biochem. 110, 909-914 (1991)

Presence of Two Transcribed Malate Synthase Genes in an n-Alkane-Utilizing Yeast, Candida tropicalis

Masaki Hikida,* Haruyuki Atomi,* Yuki Fukuda,* Akihisa Aoki,* Tadashi Hishida,** Yutaka Teranishi,** Mitsuyoshi Ueda,* and Atsuo Tanaka*,1 *Laboratory of I ndustrial Biochemistry, Department of Industrial Chemistry, Faculty of Engineering, Kyoto University, Yoshida, Sakyo-ku, Kyoto, Kyoto 606; and **Research Center, Mitsubishi Kasei Corporation, 1000 Kamoshida-cho, Midori-ku, Yokohama, Kanagawa 227

Received for publication, August 15, 1991

The presence of two genomic DNA regions encoding malate synthase (MS) was shown by Southern blot analysis of the genomic DNA from an n-alkane-assimilating yeast, Candida tropicalis, using a partial MS cDNA probe, in accordance with the fact that two types of partial MS cDNAs have previously been isolated. This was also confirmed by the restriction mapping of the two genes screened from the yeast kEMBL library. Nucleotide sequence analysis of the respective genomic DNAs, named MS-1 gene and MS-2 gene, revealed that both regions encoding MS had the same length of 1,653 base pairs, corresponding to 551 amino acids (molecular mass of MS-1, 62,448 Da; MS-2, 62,421 Da). Although 29 nucleotide pairs differed in the sequences of the coding regions, the number of amino acid replacements was only one: 159Asn (MS-1)•¨159Ser (MS-2). In the 5•L-flanking regions, there were replace ments of four nucleotide pairs, deletion of one pair, and insertion of four pairs. In spite of the fact that two genomic genes were present and transcribed, RNA blot analysis demon strated that only one band (about 2kb) was observable even when the carbon sources in the cultivation medium were changed. A comparison of the amino acid sequences was made with MSs of rape (Brassica napus L.), cucumber seed, pumpkin seed, , and Hansenula polymorpha. A high homology was observed among these , the results indicating that the protein structure was relatively well conserved through the evolution of the molecule. The replacement of only one amino acid residue seems to have little effect on the structure, even if there is some effect on the enzymatic activity. Since MS-1 and MS-2 genes were very similar and TATA-box sequences were also detected in the respective 5•L-flanking regions, the mechanism regulating their biosynthesis was suggested to be present in the respective 5•L-flanking regions.

Malate synthase (MS) mediates the condensation of glyox /3-hydroxyacyl-CoA epimerase)), 3-ketoacyl-CoAthiolase, ylate and acetyl-CoA to form L-malate. The is one and acetoacetyl-CoA ], catalase, carnitine acetyl- of the key enzymes of the glyoxylate cycle, which has an , and the key enzymes of the glyoxylate cycle

important function in gluconeogenesis (1). The glyoxylate (isocitrate and malate synthase), while mitochondria cycle is essential in bacterial cells assimilating gluconeo possess the tricarboxylic acid cycle enzymes and carnitine genic substrates (2). This cycle has been found in toad but not the ƒÀ-oxidation system urinary bladder (3) and liver of fetal guinea pig (4), (10, 11). From these results, the glyoxylate cycle in the indicating that it also has an important role in eukaryotic yeast is supposed to function through cooperation of cells. As for plants, the cycle is localized in of peroxisomes and mitochondria, involving the transport of germinating seeds and functions exclusively in utilization acetyl units via the "acetylcarnitine shuttle" (12, 13). The of storage (5). function of malate synthase in peroxisomes is to condense In yeasts, this cycle exhibits a characteristic localization. acetyl-CoA produced via fatty acid ƒÀ-oxidation and glyox

A yeast, Candida tropicalis, contains many peroxisomes ylate formed through the reaction to when grown on alkanes or higher fatty acids and the synthesize malate. organelles play an indispensable role in the metabolism of We have already purified and characterized malate these compounds (6, 7). Peroxisomes of this yeast contain synthases of glucose-, -, propionate-, and n-alkane- the enzymes of the fatty acid ƒÀ-oxidation system [acyl- grown C. tropicalis cells (14, 15). Although their proteinic CoA oxidase, bi- (8) or tri- (9) functional enzyme (enoyl- and immunochemical properties were indistinguishable, CoA hydratase/3-hydroxyacyl-CoA dehydrogenase (and the enzymes inducibly synthesized in the propionate- and n-alkane-grown cells showed higher Km values for acetyl- CoA than the enzymes of glucose- and acetate-grown cells ' To whom correspondence should be addressed . Abbreviations: MS, malate synthase; BN, Brassica napus L.; CS, (15). Such a difference was not observed among isocitrate cucumber seed; CT, Candida tropicalis; EC, Escherichia coli; HP, from cells grown on these carbon sources even Hansenula polymorpha; PS, pumpkin seed. though both malate synthase and isocitrate lyase are

Vol. 110, No. 6, 1991 909 910 M. Hikida et al.

involved in the glyoxylate cycle (15). In Escherichia coli, isolated from C. tropicalis was first digested with SalI, two types of MS (types A and G) have been found. The KpnI, EcoRV, and BamHI, respectively, whose restriction biosynthesis of the type A of MS is regulated by sites were absent in the probe cDNA, then electrophoresed repression-derepression and that of the type G of MS is on 1.0% agarose gel, and transferred to a nitrocellulose induced by glyoxylate (16, 17). These types differ in their filter. Hybridization with the cDNA probe gave two positive chromatographic behaviors and affinities to one of the bands (Fig. 2), suggesting the presence of two different MS substrates (glyoxylate). genes which seemed to correspond to MS cDNA-1 and MS We then isolated and analyzed the genomic DNA encod cDNA-2 or, interestingly, the presence of one allelic MS ing MS of f the yeast as the first step to investigate the gene from which mRNAs corresponding to MS cDNA-1 and difference at the molecular level. In this studdy, we have MS cDNA-2 were transcribed, respectively. found the presence of two transcribed MS genes in the Identification and Sequencing of Genomic DNA for library prepared from C. tropicalis genomic DNA and Malate Synthase-A AEMBL 3 genomic DNA library of C. analyzed their nucleotide sequences. tropicalis was screened for the genes encoding MS with the cDNA probe described above. Among 90,000 plaques, 4 MATERIALS AND METHODS positive signals were detected. Of these isolated clones, 3 clones were partially superimposed and one clone was Strain and Vector Used for DNA Sequencing-E. coli independent, judging from the maps of the restriction DH 5ƒ¿ was used as a host cell for cloning and pUC 19 as a enzyme sites. The clone having the longest insertion among plasmid for recombination. the superimposed clones was named ƒÉCT-MS-1 and the Preparation of Probes-cDNA (18) and genomic DNA independent clone, ƒÉCT-MS-2 (Fig. 3). SalI-SalI DNA fragments encoding a part of malate synthase (MS) were fragments in these clones were 3.1 and 2.9kbp, respective- labeled with biotin-11-dUTP (Bethesda Research Labora ly. This corresponded to the results that two positive bands tories Life Technologies, Gaithersburg, MD, U.S.A.) by of 3.1 and 2.9kbp appeared in the Southern blot analysis nick translation and used as probes. Detection was per- for the MS gene after Sali digestion (Fig. 2) and indicated formed by a color formation reƒÉCTion with nitroblue tetra that two kinds of MS genes (ƒÉCT-MS-1 and ƒÉCT-MS-2) zolium and 5-bromo-4-chloro-3-indolylphosphate as sub were present and isolated. As illustrated in Fig. 3, a strates, as described previously (19). comparison of the restriction maps of ƒÉCT-MS-1, ƒÉCT- Screening of Clones-From a ƒÉEMBL 3 genomic DNA MS-2 genes, MS cDNA-1, and MS cDNA-2 (see in Fig. 1) recombinant library of C. tropicalis cells (20), the clones revealed that MS cDNA- 1 was derived from the ƒÉCT-MS- were screened according to the method of Benton and Davis 1 gene and MS cDNA-2 from the ƒÉCT-MS-2 gene. (21) using a probe of partial cDNA encoding MS. Sail-SalI DNA fragments of both ƒÉCT-MS-1 and Restriction Mapping and DNA Sequencing-Restriction ƒÉCT-MS-2 genes were then sequenced (Fig. 4). We found endonucleases were purchased from Toyobo (Osaka) and only one open reading frame composed of 1,653 by in the Takara Shuzo (Kyoto). ReƒÉCTions were carried out under respective genes. Although the number of base pairs in the the conditions recommended by the vendors. After map- coding region was the same, there were replacements of 29 ping, the analysis of the respective DNA fragments sub- base-pairs between MS-1 gene and MS-2 gene, which cloned into pUC 19 was carried out by the dideoxy chain resulted in the change of only one amino acid residue among termination method (22). The universal 17-primer and 551 amino acid residues, from 159Asn of MS-1 to 159Ser of reverse 17-primer purchased from Pharmacia (Uppsala, MS-2 (molecular mass of MS-1, 62,448 Da; MS-2, 62,421 Sweden) and 16-primers synthesized by a 381A DNA Da). It was confirmed that the nucleotide sequences of Synthesizer (Applied Biosystems, Foster City, CA, U.S.A.) MS-1 and MS-2 genes contained those of MS cDNA-1 (701 were used for sequencing. bp) and MS cDNA-2 (933 bp), respectively (Figs. 3 and 4). Genomic DNA and RNA Blot Hybridization Analyses- From the results of the analysis of the secondary struc Total genomic DNA and RNA blot hybridization analyses tures of MS-1 and MS-2 according to the method of Chou were carried out as described previously (19, 23) using the and Fasman (24), no change of their structures was recog biotin-labeled probe DNA. nized. The C-terminal sequence was -Glu-Arg-Leu in both MSs, whereas a candidate transport signal of -Ser-Lys-Leu

RESULTS AND DISCUSSION for import into peroxisomes has been reported to be present at the C-terminal of some proteins (25, 26). The Southern Blot Analysis of Malate Synthase Gene-We isolated two types of partial malate synthase (MS) cDNAs from the ăgt11 cDNA expression library prepared from mRNAs of n-alkane-grown C. tropicalis cells by screening with anti-MS antibody (18). Restriction maps of the respective MS cDNAs are shown in Fig. 1. Both restriction maps were almost identical except for the presence of the PstI site in the MS cDNA-1. The MS cDNA-1 was then used as a probe to investigate the MS gene(s) in the yeast genome, because immunochemical and proteinic studies showed that the enzymes isolated from the cells were very similar, independent of the growth substrates (15), and the nucleotide sequences of the respective MS cDNAs were Fig. 1. The restriction maps of isolated MS cDNA-1 (A) and ME also almost identical (data not shown). The genomic DNA cDNA-2 (B).

J. Biochem. Two Malate Synthase Genes in Candida tropicalis 911

Fig. 2. Southern blot analysis of MS gene. MS cDNA-1 seen in Fig. 1 was used as a probe. The nitrocellulose filter showed positive bands. Lane 1, Sail digests of the yeast genomic DNA; lane 2, KpnI digests; lane 3, EcoRV digests; lane 4, BamHI digests. The markers of molecular size were HindIII digests of ;L-DNA. Arrowheads represent 3.1 and 2.9kbp bands of Sail digests.

Fig. 5. Northern blot analy sis of MS mRNAs with the genomic DNA probe (KpnI- HincII DNA fragment, 755 bp). Total RNAs (20ƒÊg) iso lated from C. tropicalis cultured on propionate (lane 1), n-alkane Fig. 3. Restriction map and sequencing strategy of clones (lane 2), glucose (lane 3), and ƒÉCT- MS-1 (A) and ƒÉCT-MS-2 (B) containing MS genes. The thick acetate (lane 4) as the sole car- arrows indicate the direction of transcription and the thin arrows show bon sources were separated by the direction and extent of sequence determination. The cloned electrophoresis and hybridized genomic DNAs of C. tropicalis are shown in the upper lines, while the with the DNA probe shown in Sail-Sail DNA fragments are enlarged in the bottom lines, respec Fig. 3. Calf and E. coli rRNAs tively. The thin bars in A and B (bottom lines) correspond to MS were used as the size markers. cDNA-1 and MS cDNA-2 portions (Fig. 1). The thick bar illustrated in A represents the genomic DNA probe used for Northern blot analysis (see Fig. 5).

above sequence could be regarded as resembling the sequences categorized into the -Ser-Lys-Leu group (20), and MS-2 genes in which the length from the TATA box to because the C-terminal amino acid is hydrophobic and the the poly(A)-addition signal site was almost the same. The second amino acid from the C-terminus is basic. However, ratio of the amount of MS-1 mRNA to that of MS-2 mRNA the third amino acid is different in its properties. in the cells remains unknown because these mRNAs were In the 5•L-flanking region, the replacement of four nu not discriminated with any probe tested (data not shown). cleotide pairs, deletion of one nucleotide pair, and insertion Figure 6 shows the amino acid sequences of C. tropicalis of four nucleotide pairs were observed in the MS-2 gene and (CTMS or MS-1, 551 residues), rape (Brassica napus L.) there was a TATA box at a similar position to that of the (BNMS, 561 residues) (27), cucumber seed (Cucumis MS-1 gene. In the 3•L-flanking regions, both genes have the sativas) (CSMS, 568 residues) (28), pumpkin seed (Cucur same sequence, in which a candidate poly(A)-addition bita sp. Kurokawa Amakuri Nankin) (PSMS, 566 residues) signal was located at position 1724 (Fig. 4). (29), E. coli A type (ECMS, 533 residues) (30), and RNA blot hybridization with the probe DNA of KpnI (at -utilizing Hansenula polymorpha malate synthases position 745)-HincII (position 1500) (755 bp) genomic (HPMS, 555 residues) (31). The enzymes could be aligned DNA fragment or the cDNA probe used for screening of the with high homology, 46.8% between CTMS and BNMS, genomic DNA library demonstrated that only one band 47.2% between CTMS and CSMS, 47.0% between CTMS (about 2kb) was present in the glucose-, acetate-, propio and PSMS, 48.0% between CTMS and ECMS, and 51.5% nate-, and alkane-grown cells (Fig. 5). This result coincided between CTMS and HPMS. There were several regions with the analysis of the nucleotide sequences of the MS-1 with especially high homology as shown by thick bars in Fig.

Vol. 110, No. 6, 1991 912 M. Hikida et al.

Fig. 4. Nucleotide sequences of the genes for MS-1 and MS-2, and their flanking regions. The nucleotide sequences are aligned in the first and third lines. The marks "-" indicate the same nucleotides. The amino acid sequences were deduced from the nucleotide sequences in the second and fourth lines. Identical residues in the fourth line are indicated with "+ ." Position 1 corresponds to the first nucleotide of the ATG initiation codon. The TATA box sequence is boxed on the MS-1 gene and a candidate poly(A) -addition signal is underlined beneath the nucleotide sequence on the same gene. Underlined amino acid residues represent the coincident sequences from the analysis of the purified enzyme. The arrow (t) indicates the change of the amino acid residue (N-.S).

J. Biochem. Two Malate Synthase Genes in Candida tropicalis 913

to Ser as seen in MS-1 and MS-2 seemed to have little effect on the structures of MS, because the amino acid residues at this position were not always preserved in these align ments. Statistical comparison revealed that codons which appear in highly expressed genes (33) were also utilized with strong preference in these MSs (data not shown) as

well as other peroxisomal proteins of C. tropicalis reported

previously (20, 34). However, the codon of Ser changed from Asn was AGC, which is not utilized preferentially. Different codons observed at other changed positions were well used among several peroxisomal enzymes of this

yeast. Accordingly, such codon usage may have an effect on the difference of biosynthesis of MS-1 and MS-2, although the comparative amount of tRNA for AGC codon is un- known. The upstream regions including TATA boxes, which

probably contain promoter sequences, presumably contrib ute to the different expression of the MS-1 and MS-2 genes, even if the two genes are on different loci or an allelic gene, because the difference in the size of SalI-SalI DNA frag ments was derived from the 5•L-upstream region. These regions should be further investigated for promoter activ ity. As described here, although both MS-1 and MS-2 genes from C. tropicalis are transcribed, we cannot discriminate which is expressed preferentially in the cells growing on different carbon sources. Since no MS gene other than the MS-1 and MS-2 genes was detected, the synthesis of MS-1 and MS-2 seemed to be regulated by the difference in the activation of the respective promoters, if the replacement of Asn to Ser has any metabolic significance.

REFERENCES

1. Cioni, M., Pinzaufi, G., & Vanni, P. (1981) Comp. Biochem. Physiol. 70B, 1-26 2. Maloy, S.R., Bohlander, M., & Nunn, W.D. (1980) J. Bacteriol. 143,720-725 3. Goodman, D.B.P., Davis, W.L., &Jones, R.G. (1980) Proc. Natl. Acad. Sci. U.S.A. 77,1521-1525 4. Jones, C.T. (1980) Biochem. Biophys. Res. Commun. 95,849-856 5. Beevers, H. (1979) Annu. Rev. Plant Physiol. 30, 159-193 6. Osumi, M., Miwa, N., Teranishi, Y., Tanaka, A., & Fukui, S. (1974) Arch. Microbiol. 99, 181-201 7. Osumi, M., Fukuzumi, F., Teranishi, Y., Tanaka, A., & Fukui, S. (1975) Arch. Microbiol. 103, 1-11 8. Ueda, M., Morikawa, T., Okada, H., & Tanaka, A. (1987) Agric. Biol. Chem. 51, 869-875 9. Moreno de la Garza, M., Shultz-Borchard, U., Crabb, J.W., & Kunau, W.-H. (1985) Eur. J. Biochem. 148, 285-291 10. Fukui, S. & Tanaka, A. (1979) J. Appl. Biochem. 1, 171-201 Fig. 6. Alignment of the amino acid sequences of CTMS (MS-1 11. Tanaka, A., Osumi, M., & Fukui, S. (1982) Ann. N. Y. Acad. Sci. of C. tropicalis), BNMS [MS of rape (Brassica napes L.)], 386,183-199 CSMS (MS of cucumber seed), PSMS (MS of pumpkin seed), 12. Ueda, M., Tanaka, A., & Fukui, S. (1982) Eur. J. Biochem. 124, ECMS (MS-A of E. coli), and HPMS (MS of H. polymorpha). 205-2101 Identical residues among CTMS and others are indicated by "+." 3. Ueda, M., Okada, H., Tanaka, A., Osumi, M., & Fukui, S. (1983) Gaps have been inserted to achieve maximum homology. The regions Arch. Microbiol. 136, 169-176 with high homology are shown with thick bars beneath the fourth line. 14. Okada, H., Ueda, M., & Tanaka, A. (1986) Arch. Microbiol. 144, The asterisk marks the position of the only change of amino acid 137-141 residue (N•¨S) on MS-2 of C. tropicalis. 15. Okada, H., Ueda, M., Uchida, M., & Tanaka, A. (1987) Agric. Biol. Chem. 51, 869-875 16. Falmagne, P., Vanderwinkel, E., & Wiame, J.M. (1965) Biochim. 6, and these may be essential for the enzyme activity. Biophys. Acta 99, 246-258 Although a sulfhydryl group was reported to be important 17. Vanderwinkel, E. & De Vlieghere, M. (1968) Eur. J. Biochem. 5, for MS activity (32), we have no information on the active 81-90 site residues, so we cannot determine which residues are 18. Ueda, M., Okada, H. Hishida, T., Teranishi, Y., & Tanaka, A. (1987) FEBS Lett. 220, 31-35 directly related to the activity. The replacement from Asn

Vol. 110, No. 6, 1991 914 M. Hikida et al.

19. Ueda, M., Hikida, M., Atomi, H., & Tanaka, A. (1990) Mem. Fac. 27. Comai, L., Baden, C.S., & Harada, J.J. (1989) J. Biol. Chem. Eng., Kyoto Univ. 52, 15-24 264,2778-2782 20. TAtomi, H., Ueda, M., Hikida, M., Hishida, T., Teranishi , Y., & 28. Graham, I.A., Smith, L.M., Brown, J.W.S., Leaver, C.J., & anaka, A. (1990) J. Biochem. 107, 262-266 Smith, S.M. (1989) Plant Mol. Biol. 13, 673-684 21. Benton, W.D. & Davis, R.W. (1977) Science 196, 180-182 29. Mori, H., Takeda-Yoshikawa, Y., Hara-Nishimura, I., & Nishi 22. Sanger, F., Nicklen, S., & Coulson, A.R. (1977) Proc. Natl. Acad. mura, M. (1991) Eur. J. Biochem. 197, 331-336 Sci. U.S.A. 74,5463-5467 30. Byrne, C., Stokes, H.W., & Ward, K.A. (1988) Nucleic Acids 23. Maniatis, T., Fritsch, E.F., & Sambrook , J. (1982) Molecular Res. 16, 9342 Cloning: A Laboratory Manual, pp. 382-389, Cold Spring 31. Bruinenberg, P. (1989) Ph.D. thesis, University of Groningen, Harbor Laboratory, Cold Spring Harbor, N.Y. NN Haren, the Netherlands 24. Chou, P.Y. & Fasman, G.D. (1978) Annu. Rev. Biochem. 47,251- 32. Durchschlag, H. & Zipper, P. (1988) FEBS Lett. 237, 208-212 276 33. Sharp, P.M., Touhy, T.M.F., & Mosurski, K.R. (1986) Nucleic 25. Gould, S.J., Keller, G. -A., & Subramani, S. (1988) J. Cell Biol. Acids Res. 14, 5125-5143 107,897-905 34. Okada, H., Ueda, M., Sugaya, T., Atomi, H., Mozaffar, S., 26. Miyazawa, S., Osumi, T., Hashimoto, T., Ohno, K., Miura, S., & Hishida, T., Teranishi, Y., Okazaki, K., Takechi, T., Kamiryo, Fujiki, Y. (1989) Mol. Cell. Biol. 9, 83-91 T., & Tanaka, A. (1987) Eur. J. Biochem. 170,105-110

J. Biochem.