- Email: [email protected]
Molecular Cloning and Functional Characterization of Rat Δ-6 Fatty Acid Desaturase
Biochemical and Biophysical Research Communications 255, 575–579 (1999) Article ID bbrc.1999.0235, available online at http://www.idealibrary.com on
Molecular Cloning and Functional Characterization of Rat D-6 Fatty Acid Desaturase Tsunehiro Aki, 1 Yayoi Shimada, Katsuya Inagaki, Hirofumi Higashimoto, Seiji Kawamoto, Seiko Shigeta, Kazuhisa Ono, and Osamu Suzuki Department of Molecular Biotechnology, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima 739-8527, Japan
Received January 6,1999
Mammalian cDNA fragments putatively encoding amino acid sequences characteristic of the fatty acid desaturase were obtained using expressed sequence tag (EST) sequence informations. These fragments were subsequently used to screen a rat liver cDNA library, yielding a 1573-bp clone. Expression of DNA fragment containing either of two possible open reading frames (nucleotide numbers 97-1431 and 148-1431) of the isolated clone in yeast led to the accumulation of g-linolenic acid in the presence of exogenous linoleic acid. In this system, the addition of a-linolenic acid also resulted in the accumulation of its D-6 desaturated product whereas dihomo-g-linolenic acid failed to be a substrate. These results indicate that the protein encoded by the rat cDNA is D-6 fatty acid desaturase, and the first 17 amino acids corresponding to the coding region 97-147 of the clone are not required to function in yeast. © 1999 Academic Press
D-6 Desaturase catalyzes the conversion of linoleic acid (LA, C18:2D-9, 12) to g-linolenic acid (GLA, C18: 3D-6, 9, 12) by inserting a double bond between carbon 6 and 7, in conjunction with cytochrome b 5-mediated electron transfer system in mammals. Since GLA and its elongation product, dihomo-g-linolenic acid (DGLA, C20:3D-8, 11, 14), are barely detectable in mammalian cells, it is generally accepted that the D-6 desaturation step is rate-limiting (1). In this context, the activity of the D-6 desaturase is considered to affect directly to the cellular content of arachidonic acid (AA, C20:4D-5, 8, 11, 14) which is a D-5 desaturated product of DGLA. It is feasible to extend this aspects on the n-6 pathway to another pathway (n-3) where a-linolenic acid (ALA, C18:3D-9, 12, 15) is converted to eicosapentaenoic acid (EPA, C20:5D-5, 8, 11, 14, 17) through D-6 desaturation. To whom correspondence should be addressed. Fax: 181-824-227191. E-mail: [email protected]. 1
AA is well-known as a precursor of a large family of eicosanoids that have multiple effects related to the regulation of e.g. blood pressure, inflammatory reactions, and platelet function (1–3). EPA exhibits antagonizing effect aganst AA metabolism, and vise versa (4, 5). Since the amounts and types of eicosanoids synthesized are partially determined by the availability of the fatty acid precursors, imbalance of these acids is suggested to contribute to numerous clinical symptoms. An early study indicated that affinity of the D-6 desaturase for ALA is greater than that for LA, implying that these fatty acids might not be metabolized in the same fashion (6). Therefore, the imbalance of the levels of fatty acid precursors could be due to the impaired activity of the D-6 desaturase on either of the two pathways. Indeed, depression of the D-6 desaturase activity, mainly reported on the n-6 pathway, is associated with various physiologic and pathophysiologic states including aging, diabetes, atopic dermatitis, cardiovascular disorders, and cancer (1, 7, 8). Also, the differences in nutritional and hormonal conditions influence the D-6 desaturase activity, resulting in the altered composition of intracellular fatty acids and membrane phospholipids (1, 9). Up to now, these observations have been led, in part, by tracing the activity of the enzyme, detected predominantly in the microsomal membrane fraction. However, molecular characterization of the membrane-bound desaturase protein especially in mammals has not been fully established. Recently, genes coding for D-6 desaturases from the borage Borago officinalis (10) and the nematode Caenorhabditis elegans (11) and D-5 desaturases from C. elegans (12) and the fungus Mortierella alpina (13, 14) were identified. Mutual comparisons of their deduced amino acid sequences revealed the presence of highly conserved heme-binding motif and histidine boxes, located in same order, which appeared to be common in all desaturases of bacteria and plants (15). Taking advantage of the sequence informations on the desaturases in eukaryotes, we newly identified a rat liver
575
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
cDNA encoding functional D-6 fatty acid desaturase, as reported here. MATERIALS AND METHODS General laboratory chemicals were purchased from Katayama Chemical (Osaka, Japan). Fatty acid standards were from Sigma Chemical Co. (St. Louis, MO). Reagents and enzymes for genetic manipulations were from Takara Shuzo (Kyoto, Japan), otherwise stated. Male BALB/c mice were obtained from Charles River Japan (Hiroshima, Japan). Messenger RNAs were extracted from mouse liver by guanidinium thiocyanate method using QuickPrep mRNA Purification Kit (Amersham Pharmacia Biotech, Uppsala, Sweden). A cDNA pool was prepared from the mouse mRNAs by TimeSaver cDNA Synthesis Kit (Amersham Pharmacia Biotech) according to the manufacturer’s instruction. Oligonucleotide primers of following sequences were synthesized for amplification of two apart regions of gene coding for desaturase-like protein by polymerase chain reaction (PCR); m3F, 59-GTCAGGGTGCTGGAGAGCCACTGG-39; m3R, 59-GTAGTGTAGGCCGTGCTTCGCGC-39; m5F, 59-GATGCTACGGATGCCTTCCGTGCC-39; m5R, 59-TTCATGTCCTCAGCAGTCTTCTTC-39. The cDNA from mouse liver was subjected to PCR reactions (LA PCR kit, Takara) with primer pairs, m3F and m3R, or m5F and m5R. Successfully amplified products, m3 and m5 respectively, were cloned on plasmid pGEM-T Easy Vector (Promega, Madison, WI), and the inserts were confirmed by DNA sequencing analyses. Rat liver cDNA library constructed on lZAP II (#937507, Stratagene, La Jolla, CA) was probed with alkaline phosphatase-labeled m3 or m5 fragment, by which labeling of the probes, hybridization, and detection of hybrids were performed using AlkPhos Direct System (Amersham Pharmacia Biotech). In this protocol, hybridization and washing steps were done at 55°C. Plaques positively detected by either of the two probes were picked up, and the accuracy of the first screening was reevaluated with purified plaques by another round of plating and hybridization. The plasmids containing positive cDNA were recovered from selected l clones by in vivo excision, and the insert was entirely sequenced on both strands of DNA. Since one of the positive clones, r24, seemed to contain full-length cDNA of interest, a plasmid derived from clone r24 was used as a template for PCR amplification of regions which were deduced as open reading frames. Because two ATG sequences could be considered as putative translation initiation codons, two forward primers, r24aF, 59-ACAAAGCTTATGGGGAAGGGAGGTAACCAG-39 (corresponding to the first ATG indicated by boldface type) and r24bF, 59-CAGAAGCTTATGCCCACCTTCCGCTGGGAG-39 (corresponding to the second ATG indicated by boldface type) were used to amplify the coding frames, r24a and r24b, respectively, and to generate Hin dIII site (underlined) adjucent to the ATG. A reverse primer r24R, 59-TCTTCTAGATCATTTGTGGAGGTAGGCATC-39 (annealing to the complement of the stop codon indicated by boldface type) was used for each PCR reaction, generating Xba I site (underlined). The PCR products treated with Hin dIII and Xba I were inserted respectively to the yeast vector pYES2 (Invitrogen, San Diego, CA). It was confirmed by DNA sequencing analyses that the entire and flanking sequences of the inserts were as we designed. Transfer of the constructs into Saccharomyces cerevisiae strain INVSc1 (Invitrogen) was done by the lithium acetate method, and recombinant yeast cells were selected on uracil-deficient medium. The yeast cells were cultivated in a medium containing 4% raffinose, 0.7% yeast nitrogen base without amino acid, 1% tergitol type NP-40, 20 mg/ml histidine, 60 mg/ml leucine, and 40 mg/ml tryptophan at 28°C, overnight. The culture broth was supplemented with fatty acid substrate so as to be a final concentration of 0.5 mM, followed by further cultivation until cell density reached at 5 3 10 6 cells/ml. The expression of the transgene was performed by the addition of galactose to 2% (w/v) and an additional cultivation for 10 hr.
Culture broths were harvested, and total intracellular lipids were extracted with a mixture of chloroform/methanol (2:1, v/v). The lipid fraction was subjected to methyl-esterification with 10% hydrochloride in methanol. Fatty acid methyl esters were applied on a gasliquid chromatograph (GC; model GC-17A, Shimadzu, Kyoto, Japan) equipped with a TC-70 capillary column (GL Science, Tokyo, Japan) and a flame ionization detector. The condition for GC analysis was as previously described (16). GC-mass spectrometry (GC-MS) analysis of the fatty acid methyl esters was performed using a MS-BU20 (JEOL, Tokyo, Japan) high-resolution mass spectrometer linked to a gas chromatograph (model MS-5890, Hewlett Packard) equipped with the TC-70 column as the sample inlet, and operated in the electron impact mode at 70 eV. Comparison of the mass spectra of authentic standards and interest peaks in total ion chromatogram was done by visual- and computer-based examinations.
RESULTS A nucleotide sequence corresponding to the highly conserved region (indicated by dotted line in Fig. 1) in D-6 desaturase from C. elegans (11) was used as a query to search databases for related sequences in mammals. When the database of mouse expressed sequence tag (EST) at DNA Data Bank of Japan was searched using both the BLAST and FASTA algorithms, several entries registering DNA sequences partially homologous to the query were retrieved. The nucleotide sequence in 59-region of one (GenBank accession number W53753) of the ESTs was then used to further search the database and found to be partially overlapped with another EST clone (AA512429). By similar sequential searchs toward 59-end of a putative desaturase gene in mouse, a clone (AA036321) overlapped with the AA512429 was found, and the AA036321 led us to an additional clone (AA250162). Nucleotide sequence of the AA250162 could be translated to the amino acid sequence bearing partial resemblance to that of N-terminal domain of previously characterized desaturases. Based on the sequence informations from ESTs W53753 and AA250162, we made two nonoverlapping DNA fragments, m3 (39-region) and m5 (59-region), respectively, by PCR with a mouse liver cDNA pool as a template. These fragments were used as hybridization probes for isolation of entire coding region of desaturase gene from rat liver cDNA library. We elected rat, instead of mouse, as a source of the target gene, because the desaturases had been best-characterized biochemically in rat, which included a report of partial purification of linoleoylCoA desaturase (17; see Discussion). The condition for hybridization was set at medium stringency making allowance for differences in animal species. As a result of screening, five individual clones were isolated as positives to the probe m3 and only one of them, termed r24, was hybridized also with m5 probe. Sequencing of these clones revealed that the clone r24 had a cDNA insert of 1573 basepairs (bp) in length (GenBank accession number AB021980), and
576
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Composite alignments of the amino acid sequence deduced from the rat cDNA clone r24 with D-6 desaturases from other sources. Borage D6d, B. officinalis D-6 desaturase (GenBank accession number U79010); C.elegans D6d, C. elegans D-6 desaturase (AF031477). Nucleotide sequence corresponding to the highly conserved region (indicated by dotted line above the sequences) in C. elegans D-6 desaturase was used as a query to search databases, as mentioned in the text. Identical residues are boxed, and the conserved heme-binding motif and three histidine boxes are marked with single and double underlines, respectively.
cDNAs in remaining four clones (all of them are less than 500 bp in length) were corresponded to 39region of r24 (data not shown), supporting the failure of hybridizing with probe m5. A putative protein encoded by the clone r24 seemed to be a rat homolog of the protein from the mouse EST AA250162 (96.5% identity in 491 nucleotides overlap). Thus, we chose the clone r24 for further characterization. Two ATG initiation codons (nucleotide numbers 97-99 and 148-150) were found in the sequence of 59terminal domain of the r24 cDNA and those were placed in-frame. According to the Kozak consensus sequence, AGXXATGG, that has been advocated as favored sequence for eukaryotic initiation sites (18), the first of the two initiation codons is credible. If this is the case, the cDNA contains a coding frame of 1335-bp long including a TGA termination codon (nucleotide numbers 1429-1431), which can be translated into 444 amino acid polypeptide. Comparisons of the deduced amino acid sequence of r24 with D-6 desaturases from C. elegans and B. officinalis showed homology scores of 27.9% and 26.4%, respectively (Fig. 1). It is noted in Fig. 1 that a typical heme-binding motif, HPGG (19), and three histidine boxes highly conserved within fatty acid desaturases (15) are present in the r24 sequence as well as the others. At the third histidine box, the first histidine residue in the conventional motif, HXXHH, was substituted with glutamine. This variance had occured in D-6 and D-5 desaturases from fungus, plant, and lower animal (10 –14).
For functional analysis of the clone r24, two possible coding regions, named r24a and r24b (nucleotide numbers 97-1431 and 148-1431, respectively) were amplified by PCR, and respective expression plasmids were constructed on the yeast vector pYES2. The PCR products were located at just downstream of the galactoseinducible GAL1 promoter on each construct. After obtaining yeast transformants carrying pYES2/r24a, pYES2/r24b, or pYES2 (control), cells were cultivated, supplemented by the addition of substrate LA (C18: 2D-9, 12), and induced in the presence of galactose. Aliquots of cells in the induced culture broth were taken for analyses of the intracellular fatty acid composition by GC, and the resultant chromatograms of fatty acid methyl esters were shown as Fig. 2. A novel peak, which was not apparent in the case of control (Fig. 2A), was detected in charts from both induced pYES2/r24a (peak 6 in Fig. 2B) and pYES2/r24b (data not shown). Similarly, when the substrate LA was replaced with ALA (C18:3D-9, 12, 15), a peak additional to the background level in the control case (Fig. 2C) was found in a GC profile obtained from the yeast transformed with either pYES2/r24a (peak 8 in Fig. 2D) or pYES2/r24b (data not shown). We confirmed that these and other additional peaks did not appear when the yeast carrying pYES2/r24a or pYES2/r24b was not induced by galactose or was supplemented with none of exogenous fatty acids. Comparisons of the retention times of the newly yielded peaks with those of authentic standards have anticipated that the fatty
577
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. GC analyses of methyl esterified fatty acids from the induced yeast cells containing pYES2 (A and C) or pYES2/r24a (B and D). Before the induction with galactose, LA (peak 5 in A and B) or ALA (peak 7 in C and D) was added to be incorporated to the cells. The peaks that found only in the case of pYES2/r24a were indicated by arrows (peak 6 in B and peak 8 in D). Identities of other peaks were determined by comparing their retention times with those of authentic standards. Peaks 1, C16:0; 2, C16:1D-9; 3, C18:0; 4, C18:1D-9.
acids giving the peaks 6 and 8 are GLA (C18:3D-6, 9, 12) and cis-3, 6, 9, 12-octadecatetraenoic acid (C18: 4D-6, 9, 12, 15), which are the D-6 desaturation products of LA and ALA, respectively. These prospects were positively supported by definitive assignments of the compounds in peaks 6 and 8 by GC-MS analyses (data not shown). In a separate experiment, when DGLA (C20:3D-8, 11, 14), a substrate of D-5 desaturase, was added to our expression system, no extra peak was observed in chromatograms from the r24a/r24b recombinants, compared to the negative control. Taken together, the recombinant yeast containing the inducible r24 cDNA had gained function of D-6, but not D-5, fatty acid desaturation. DISCUSSION Here we isolated a rat liver cDNA coding for the D-6 fatty acid desaturase. Although the cDNA, r24, was successfully expressed in yeast, we could not predict the actual ATG initiation codon corresponding to a methionine residue at the amino terminus of the native desaturase protein. This is because, in our study, no significant differences have not been detected between the two lines of expression analyses on r24a and r24b. This observation suggested no other than the needless of the first 17 amino acids in the protein expressed from r24a to function in yeast although this portion might be indispensable in rat. Another set of experiments including the purification of the native D-6 desaturase is essential to clarify this point and is being undertaken.
Okayasu, et al. (17) described a purification of rat liver linoleoyl-CoA desaturase that was capable of converting linoleoyl-CoA to g-linolenoyl-CoA in vitro. The apparent molecular weight of this enzyme (66 kD) obviously differs from either molecular weights calculated from the deduced amino acid sequence of r24a (52.4 kD) or r24b (50.7 kD). This inevitably raises a possibility of the presence of more than two types of the enzymes taking charge of the D-6 fatty acid desaturation. Sprecher and his colleagues have proposed a novel pathway, docosapentaenoic acid to docosahexaenoic acid via D-6 desaturation, for the biosynthesis of polyunsaturated fatty acids (20, 21). However, the putative involvement of a single cycle of peroxisomal b-oxidation in this pathway is under a critical reevaluation, excluding also the necessity of the proposed D-6 desaturation step (22). A metabolic study by Christiansen, et al. (23) suggested that liver microsomes might contain separate enzymes for desaturation of LA and ALA. Their observations, however, seem to be inconsistent with a result of competitive study using fatty acid tracers (24), implying that a single enzyme may govern desaturating fatty acids at D-6 position. To date, no clear conclusions have been made whether multiple forms of D-6 desaturase exist. In relation to these pending questions, we are attempting to isolate and characterize a full-length cDNA corresponding to the EST clones W53753, AA512429, and AA036321 since nucleotide sequences of these clones can be translated into amino acid sequences that are significantly homologous, but not
578
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
identical, to the sequence from our clone r24 (data not shown). This gene may encode an isoform of the D-6 desaturase, which is dominantly expressed in tissues other than liver or at the different developmental stages. This assumption does not contradict the facts that we were unable to isolate a liver cDNA whose sequence is matched with the probe m3 (from W53753), and these ESTs are derived from embryo and mammary gland. Otherwise, a protein encoded by this gene may be one of other desaturases, for example, D-5 desaturase which has not yet been identified in mammals. In either case, the cloning of the mammalian desaturase gene(s) will accelerate to elucidate the molecular mechanisms on the regulation of various cellular events by the enzyme possibly through the alteration of physical state of membrane lipids and of the level of pooled precursors for signal transducers. ACKNOWLEDGMENTS We thank Dr. Dai Hirata for his helpful advice on management of yeast. This work was supported in part by a Grant-in-Aid for Encouragement of Young Scientists (#10760052) from the Ministry of Education, Science and Culture, Japan.
REFERENCES 1. Cook, H. W. (1996) in Biochemistry of Lipids, Lipoproteins and Membranes (Vance, D. E., and Vance, J., Eds.), pp. 129 –152 Elsevier Science, Amsterdam. 2. Marx, J. L. (1982) Science 215, 1380 –1383. 3. Samuelsson, B., Goldyne, E., Granstro¨m, E., Hamberg, M., Hammarstro¨m, S., and Malmsten, C. (1978) Annu. Rev. Biochem. 47, 997–1029. 4. Li, B-. Y., Birdwell, C., and Whelan, J. (1994) J. Lipid Res. 35, 1869 –1877.
5. Whelan, J. (1996) J. Nutr. 126, 1086S–1091S. 6. Brenner, R. R. and Peluffo, R. O. (1966) J. Biol. Chem. 241, 5213–5219. 7. German, J. B., Dillard, C. J., and Whelan, J. (1995) J. Nutr. 126, 1076S–1080S. 8. Horrobin, D. F. (1992) Prog. Lipid Res. 31, 163–194. 9. Brenner, R. R. (1981) Prog. Lipid Res. 20, 41– 47. 10. Sayanova, O., Smith, M. A., Lapinskas, P., Stobart, A. K., Dobson, G., Christie, W. W., Shewry, P. R., and Napier, J. A. (1997) Proc. Natl. Acad. Sci. USA 94, 4211– 4216. 11. Napier, J. A., Hey, S. J., Lacey, D. J., and Shewry, P. R. (1998) Biochem. J. 330, 611– 614. 12. Michaelson, L. V., Napier, J. A., Lewis, M., Griffiths, G., Lazarus, C. M., and Stobart, A. K. (1998) FEBS Lett. 439, 215–218. 13. Michaelson, L. V., Lazarus, C. M., Griffiths, G., Napier, J. A., and Stobart, A. K. (1998) J. Biol. Chem. 273, 19055–19059. 14. Knutzon, D. S., Thurmond, J. M., Huang, Y. -S., Chaudhary, S., Bobik, E. G., Chan, G. M., Kirchner, S. J., and Mukerji, P. (1998) J. Biol. Chem. 273, 29360 –29366. 15. Shanklin, J., Whittle, E., and Fox, B. G. (1994) Biochemistry 33, 12787–12794. 16. Aki, T., Matsumoto, Y., Morinaga, T., Kawamoto, S., Shigeta, S., Ono, K., and Suzuki, O. (1998) J. Ferment. Bioeng. 86, 504 –507. 17. Okayasu, T., Nagao, M., Ishibashi, T., and Imai, Y. (1981) Arch. Biochem. Biophys. 206, 21–28. 18. Kozak, M. (1981) Nucleic Acids Res. 9, 5233–5262. 19. Napier, J. A., Sayanova, O., Stobart, A. K., and Shewry, P. R. (1997) Biochem. J. 328, 717–720. 20. Voss, A., Reinhart, M., Sankarappa, S., and Sprecher, H. (1991) J. Biol. Chem. 266, 19995–20000. 21. Sprecher, H., Luthria, D. L., Mohammed, B. S., and Baykousheva, S. P. (1995) J. Lipid Res. 36, 2471–2477. 22. Infante, J. P. and Huszagh, V. A. (1998) FEBS Lett. 431, 1– 6. 23. Christiansen, E. N., Lund, J. S., Rørtveit, T., and Rustan, A. (1991) Biochem. Biophys. Acta 1082, 57– 62. 24. Geiger, M., Mohammed, B. S., Sankarappa, S., and Sprecher, H. (1993) Biochem. Biophys. Acta 1170, 137–142.
579
Molecular Cloning and Functional Characterization of Rat D-6 Fatty Acid Desaturase Tsunehiro Aki, 1 Yayoi Shimada, Katsuya Inagaki, Hirofumi Higashimoto, Seiji Kawamoto, Seiko Shigeta, Kazuhisa Ono, and Osamu Suzuki Department of Molecular Biotechnology, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima 739-8527, Japan
Received January 6,1999
Mammalian cDNA fragments putatively encoding amino acid sequences characteristic of the fatty acid desaturase were obtained using expressed sequence tag (EST) sequence informations. These fragments were subsequently used to screen a rat liver cDNA library, yielding a 1573-bp clone. Expression of DNA fragment containing either of two possible open reading frames (nucleotide numbers 97-1431 and 148-1431) of the isolated clone in yeast led to the accumulation of g-linolenic acid in the presence of exogenous linoleic acid. In this system, the addition of a-linolenic acid also resulted in the accumulation of its D-6 desaturated product whereas dihomo-g-linolenic acid failed to be a substrate. These results indicate that the protein encoded by the rat cDNA is D-6 fatty acid desaturase, and the first 17 amino acids corresponding to the coding region 97-147 of the clone are not required to function in yeast. © 1999 Academic Press
D-6 Desaturase catalyzes the conversion of linoleic acid (LA, C18:2D-9, 12) to g-linolenic acid (GLA, C18: 3D-6, 9, 12) by inserting a double bond between carbon 6 and 7, in conjunction with cytochrome b 5-mediated electron transfer system in mammals. Since GLA and its elongation product, dihomo-g-linolenic acid (DGLA, C20:3D-8, 11, 14), are barely detectable in mammalian cells, it is generally accepted that the D-6 desaturation step is rate-limiting (1). In this context, the activity of the D-6 desaturase is considered to affect directly to the cellular content of arachidonic acid (AA, C20:4D-5, 8, 11, 14) which is a D-5 desaturated product of DGLA. It is feasible to extend this aspects on the n-6 pathway to another pathway (n-3) where a-linolenic acid (ALA, C18:3D-9, 12, 15) is converted to eicosapentaenoic acid (EPA, C20:5D-5, 8, 11, 14, 17) through D-6 desaturation. To whom correspondence should be addressed. Fax: 181-824-227191. E-mail: [email protected]. 1
AA is well-known as a precursor of a large family of eicosanoids that have multiple effects related to the regulation of e.g. blood pressure, inflammatory reactions, and platelet function (1–3). EPA exhibits antagonizing effect aganst AA metabolism, and vise versa (4, 5). Since the amounts and types of eicosanoids synthesized are partially determined by the availability of the fatty acid precursors, imbalance of these acids is suggested to contribute to numerous clinical symptoms. An early study indicated that affinity of the D-6 desaturase for ALA is greater than that for LA, implying that these fatty acids might not be metabolized in the same fashion (6). Therefore, the imbalance of the levels of fatty acid precursors could be due to the impaired activity of the D-6 desaturase on either of the two pathways. Indeed, depression of the D-6 desaturase activity, mainly reported on the n-6 pathway, is associated with various physiologic and pathophysiologic states including aging, diabetes, atopic dermatitis, cardiovascular disorders, and cancer (1, 7, 8). Also, the differences in nutritional and hormonal conditions influence the D-6 desaturase activity, resulting in the altered composition of intracellular fatty acids and membrane phospholipids (1, 9). Up to now, these observations have been led, in part, by tracing the activity of the enzyme, detected predominantly in the microsomal membrane fraction. However, molecular characterization of the membrane-bound desaturase protein especially in mammals has not been fully established. Recently, genes coding for D-6 desaturases from the borage Borago officinalis (10) and the nematode Caenorhabditis elegans (11) and D-5 desaturases from C. elegans (12) and the fungus Mortierella alpina (13, 14) were identified. Mutual comparisons of their deduced amino acid sequences revealed the presence of highly conserved heme-binding motif and histidine boxes, located in same order, which appeared to be common in all desaturases of bacteria and plants (15). Taking advantage of the sequence informations on the desaturases in eukaryotes, we newly identified a rat liver
575
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
cDNA encoding functional D-6 fatty acid desaturase, as reported here. MATERIALS AND METHODS General laboratory chemicals were purchased from Katayama Chemical (Osaka, Japan). Fatty acid standards were from Sigma Chemical Co. (St. Louis, MO). Reagents and enzymes for genetic manipulations were from Takara Shuzo (Kyoto, Japan), otherwise stated. Male BALB/c mice were obtained from Charles River Japan (Hiroshima, Japan). Messenger RNAs were extracted from mouse liver by guanidinium thiocyanate method using QuickPrep mRNA Purification Kit (Amersham Pharmacia Biotech, Uppsala, Sweden). A cDNA pool was prepared from the mouse mRNAs by TimeSaver cDNA Synthesis Kit (Amersham Pharmacia Biotech) according to the manufacturer’s instruction. Oligonucleotide primers of following sequences were synthesized for amplification of two apart regions of gene coding for desaturase-like protein by polymerase chain reaction (PCR); m3F, 59-GTCAGGGTGCTGGAGAGCCACTGG-39; m3R, 59-GTAGTGTAGGCCGTGCTTCGCGC-39; m5F, 59-GATGCTACGGATGCCTTCCGTGCC-39; m5R, 59-TTCATGTCCTCAGCAGTCTTCTTC-39. The cDNA from mouse liver was subjected to PCR reactions (LA PCR kit, Takara) with primer pairs, m3F and m3R, or m5F and m5R. Successfully amplified products, m3 and m5 respectively, were cloned on plasmid pGEM-T Easy Vector (Promega, Madison, WI), and the inserts were confirmed by DNA sequencing analyses. Rat liver cDNA library constructed on lZAP II (#937507, Stratagene, La Jolla, CA) was probed with alkaline phosphatase-labeled m3 or m5 fragment, by which labeling of the probes, hybridization, and detection of hybrids were performed using AlkPhos Direct System (Amersham Pharmacia Biotech). In this protocol, hybridization and washing steps were done at 55°C. Plaques positively detected by either of the two probes were picked up, and the accuracy of the first screening was reevaluated with purified plaques by another round of plating and hybridization. The plasmids containing positive cDNA were recovered from selected l clones by in vivo excision, and the insert was entirely sequenced on both strands of DNA. Since one of the positive clones, r24, seemed to contain full-length cDNA of interest, a plasmid derived from clone r24 was used as a template for PCR amplification of regions which were deduced as open reading frames. Because two ATG sequences could be considered as putative translation initiation codons, two forward primers, r24aF, 59-ACAAAGCTTATGGGGAAGGGAGGTAACCAG-39 (corresponding to the first ATG indicated by boldface type) and r24bF, 59-CAGAAGCTTATGCCCACCTTCCGCTGGGAG-39 (corresponding to the second ATG indicated by boldface type) were used to amplify the coding frames, r24a and r24b, respectively, and to generate Hin dIII site (underlined) adjucent to the ATG. A reverse primer r24R, 59-TCTTCTAGATCATTTGTGGAGGTAGGCATC-39 (annealing to the complement of the stop codon indicated by boldface type) was used for each PCR reaction, generating Xba I site (underlined). The PCR products treated with Hin dIII and Xba I were inserted respectively to the yeast vector pYES2 (Invitrogen, San Diego, CA). It was confirmed by DNA sequencing analyses that the entire and flanking sequences of the inserts were as we designed. Transfer of the constructs into Saccharomyces cerevisiae strain INVSc1 (Invitrogen) was done by the lithium acetate method, and recombinant yeast cells were selected on uracil-deficient medium. The yeast cells were cultivated in a medium containing 4% raffinose, 0.7% yeast nitrogen base without amino acid, 1% tergitol type NP-40, 20 mg/ml histidine, 60 mg/ml leucine, and 40 mg/ml tryptophan at 28°C, overnight. The culture broth was supplemented with fatty acid substrate so as to be a final concentration of 0.5 mM, followed by further cultivation until cell density reached at 5 3 10 6 cells/ml. The expression of the transgene was performed by the addition of galactose to 2% (w/v) and an additional cultivation for 10 hr.
Culture broths were harvested, and total intracellular lipids were extracted with a mixture of chloroform/methanol (2:1, v/v). The lipid fraction was subjected to methyl-esterification with 10% hydrochloride in methanol. Fatty acid methyl esters were applied on a gasliquid chromatograph (GC; model GC-17A, Shimadzu, Kyoto, Japan) equipped with a TC-70 capillary column (GL Science, Tokyo, Japan) and a flame ionization detector. The condition for GC analysis was as previously described (16). GC-mass spectrometry (GC-MS) analysis of the fatty acid methyl esters was performed using a MS-BU20 (JEOL, Tokyo, Japan) high-resolution mass spectrometer linked to a gas chromatograph (model MS-5890, Hewlett Packard) equipped with the TC-70 column as the sample inlet, and operated in the electron impact mode at 70 eV. Comparison of the mass spectra of authentic standards and interest peaks in total ion chromatogram was done by visual- and computer-based examinations.
RESULTS A nucleotide sequence corresponding to the highly conserved region (indicated by dotted line in Fig. 1) in D-6 desaturase from C. elegans (11) was used as a query to search databases for related sequences in mammals. When the database of mouse expressed sequence tag (EST) at DNA Data Bank of Japan was searched using both the BLAST and FASTA algorithms, several entries registering DNA sequences partially homologous to the query were retrieved. The nucleotide sequence in 59-region of one (GenBank accession number W53753) of the ESTs was then used to further search the database and found to be partially overlapped with another EST clone (AA512429). By similar sequential searchs toward 59-end of a putative desaturase gene in mouse, a clone (AA036321) overlapped with the AA512429 was found, and the AA036321 led us to an additional clone (AA250162). Nucleotide sequence of the AA250162 could be translated to the amino acid sequence bearing partial resemblance to that of N-terminal domain of previously characterized desaturases. Based on the sequence informations from ESTs W53753 and AA250162, we made two nonoverlapping DNA fragments, m3 (39-region) and m5 (59-region), respectively, by PCR with a mouse liver cDNA pool as a template. These fragments were used as hybridization probes for isolation of entire coding region of desaturase gene from rat liver cDNA library. We elected rat, instead of mouse, as a source of the target gene, because the desaturases had been best-characterized biochemically in rat, which included a report of partial purification of linoleoylCoA desaturase (17; see Discussion). The condition for hybridization was set at medium stringency making allowance for differences in animal species. As a result of screening, five individual clones were isolated as positives to the probe m3 and only one of them, termed r24, was hybridized also with m5 probe. Sequencing of these clones revealed that the clone r24 had a cDNA insert of 1573 basepairs (bp) in length (GenBank accession number AB021980), and
576
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Composite alignments of the amino acid sequence deduced from the rat cDNA clone r24 with D-6 desaturases from other sources. Borage D6d, B. officinalis D-6 desaturase (GenBank accession number U79010); C.elegans D6d, C. elegans D-6 desaturase (AF031477). Nucleotide sequence corresponding to the highly conserved region (indicated by dotted line above the sequences) in C. elegans D-6 desaturase was used as a query to search databases, as mentioned in the text. Identical residues are boxed, and the conserved heme-binding motif and three histidine boxes are marked with single and double underlines, respectively.
cDNAs in remaining four clones (all of them are less than 500 bp in length) were corresponded to 39region of r24 (data not shown), supporting the failure of hybridizing with probe m5. A putative protein encoded by the clone r24 seemed to be a rat homolog of the protein from the mouse EST AA250162 (96.5% identity in 491 nucleotides overlap). Thus, we chose the clone r24 for further characterization. Two ATG initiation codons (nucleotide numbers 97-99 and 148-150) were found in the sequence of 59terminal domain of the r24 cDNA and those were placed in-frame. According to the Kozak consensus sequence, AGXXATGG, that has been advocated as favored sequence for eukaryotic initiation sites (18), the first of the two initiation codons is credible. If this is the case, the cDNA contains a coding frame of 1335-bp long including a TGA termination codon (nucleotide numbers 1429-1431), which can be translated into 444 amino acid polypeptide. Comparisons of the deduced amino acid sequence of r24 with D-6 desaturases from C. elegans and B. officinalis showed homology scores of 27.9% and 26.4%, respectively (Fig. 1). It is noted in Fig. 1 that a typical heme-binding motif, HPGG (19), and three histidine boxes highly conserved within fatty acid desaturases (15) are present in the r24 sequence as well as the others. At the third histidine box, the first histidine residue in the conventional motif, HXXHH, was substituted with glutamine. This variance had occured in D-6 and D-5 desaturases from fungus, plant, and lower animal (10 –14).
For functional analysis of the clone r24, two possible coding regions, named r24a and r24b (nucleotide numbers 97-1431 and 148-1431, respectively) were amplified by PCR, and respective expression plasmids were constructed on the yeast vector pYES2. The PCR products were located at just downstream of the galactoseinducible GAL1 promoter on each construct. After obtaining yeast transformants carrying pYES2/r24a, pYES2/r24b, or pYES2 (control), cells were cultivated, supplemented by the addition of substrate LA (C18: 2D-9, 12), and induced in the presence of galactose. Aliquots of cells in the induced culture broth were taken for analyses of the intracellular fatty acid composition by GC, and the resultant chromatograms of fatty acid methyl esters were shown as Fig. 2. A novel peak, which was not apparent in the case of control (Fig. 2A), was detected in charts from both induced pYES2/r24a (peak 6 in Fig. 2B) and pYES2/r24b (data not shown). Similarly, when the substrate LA was replaced with ALA (C18:3D-9, 12, 15), a peak additional to the background level in the control case (Fig. 2C) was found in a GC profile obtained from the yeast transformed with either pYES2/r24a (peak 8 in Fig. 2D) or pYES2/r24b (data not shown). We confirmed that these and other additional peaks did not appear when the yeast carrying pYES2/r24a or pYES2/r24b was not induced by galactose or was supplemented with none of exogenous fatty acids. Comparisons of the retention times of the newly yielded peaks with those of authentic standards have anticipated that the fatty
577
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. GC analyses of methyl esterified fatty acids from the induced yeast cells containing pYES2 (A and C) or pYES2/r24a (B and D). Before the induction with galactose, LA (peak 5 in A and B) or ALA (peak 7 in C and D) was added to be incorporated to the cells. The peaks that found only in the case of pYES2/r24a were indicated by arrows (peak 6 in B and peak 8 in D). Identities of other peaks were determined by comparing their retention times with those of authentic standards. Peaks 1, C16:0; 2, C16:1D-9; 3, C18:0; 4, C18:1D-9.
acids giving the peaks 6 and 8 are GLA (C18:3D-6, 9, 12) and cis-3, 6, 9, 12-octadecatetraenoic acid (C18: 4D-6, 9, 12, 15), which are the D-6 desaturation products of LA and ALA, respectively. These prospects were positively supported by definitive assignments of the compounds in peaks 6 and 8 by GC-MS analyses (data not shown). In a separate experiment, when DGLA (C20:3D-8, 11, 14), a substrate of D-5 desaturase, was added to our expression system, no extra peak was observed in chromatograms from the r24a/r24b recombinants, compared to the negative control. Taken together, the recombinant yeast containing the inducible r24 cDNA had gained function of D-6, but not D-5, fatty acid desaturation. DISCUSSION Here we isolated a rat liver cDNA coding for the D-6 fatty acid desaturase. Although the cDNA, r24, was successfully expressed in yeast, we could not predict the actual ATG initiation codon corresponding to a methionine residue at the amino terminus of the native desaturase protein. This is because, in our study, no significant differences have not been detected between the two lines of expression analyses on r24a and r24b. This observation suggested no other than the needless of the first 17 amino acids in the protein expressed from r24a to function in yeast although this portion might be indispensable in rat. Another set of experiments including the purification of the native D-6 desaturase is essential to clarify this point and is being undertaken.
Okayasu, et al. (17) described a purification of rat liver linoleoyl-CoA desaturase that was capable of converting linoleoyl-CoA to g-linolenoyl-CoA in vitro. The apparent molecular weight of this enzyme (66 kD) obviously differs from either molecular weights calculated from the deduced amino acid sequence of r24a (52.4 kD) or r24b (50.7 kD). This inevitably raises a possibility of the presence of more than two types of the enzymes taking charge of the D-6 fatty acid desaturation. Sprecher and his colleagues have proposed a novel pathway, docosapentaenoic acid to docosahexaenoic acid via D-6 desaturation, for the biosynthesis of polyunsaturated fatty acids (20, 21). However, the putative involvement of a single cycle of peroxisomal b-oxidation in this pathway is under a critical reevaluation, excluding also the necessity of the proposed D-6 desaturation step (22). A metabolic study by Christiansen, et al. (23) suggested that liver microsomes might contain separate enzymes for desaturation of LA and ALA. Their observations, however, seem to be inconsistent with a result of competitive study using fatty acid tracers (24), implying that a single enzyme may govern desaturating fatty acids at D-6 position. To date, no clear conclusions have been made whether multiple forms of D-6 desaturase exist. In relation to these pending questions, we are attempting to isolate and characterize a full-length cDNA corresponding to the EST clones W53753, AA512429, and AA036321 since nucleotide sequences of these clones can be translated into amino acid sequences that are significantly homologous, but not
578
Vol. 255, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
identical, to the sequence from our clone r24 (data not shown). This gene may encode an isoform of the D-6 desaturase, which is dominantly expressed in tissues other than liver or at the different developmental stages. This assumption does not contradict the facts that we were unable to isolate a liver cDNA whose sequence is matched with the probe m3 (from W53753), and these ESTs are derived from embryo and mammary gland. Otherwise, a protein encoded by this gene may be one of other desaturases, for example, D-5 desaturase which has not yet been identified in mammals. In either case, the cloning of the mammalian desaturase gene(s) will accelerate to elucidate the molecular mechanisms on the regulation of various cellular events by the enzyme possibly through the alteration of physical state of membrane lipids and of the level of pooled precursors for signal transducers. ACKNOWLEDGMENTS We thank Dr. Dai Hirata for his helpful advice on management of yeast. This work was supported in part by a Grant-in-Aid for Encouragement of Young Scientists (#10760052) from the Ministry of Education, Science and Culture, Japan.
REFERENCES 1. Cook, H. W. (1996) in Biochemistry of Lipids, Lipoproteins and Membranes (Vance, D. E., and Vance, J., Eds.), pp. 129 –152 Elsevier Science, Amsterdam. 2. Marx, J. L. (1982) Science 215, 1380 –1383. 3. Samuelsson, B., Goldyne, E., Granstro¨m, E., Hamberg, M., Hammarstro¨m, S., and Malmsten, C. (1978) Annu. Rev. Biochem. 47, 997–1029. 4. Li, B-. Y., Birdwell, C., and Whelan, J. (1994) J. Lipid Res. 35, 1869 –1877.
5. Whelan, J. (1996) J. Nutr. 126, 1086S–1091S. 6. Brenner, R. R. and Peluffo, R. O. (1966) J. Biol. Chem. 241, 5213–5219. 7. German, J. B., Dillard, C. J., and Whelan, J. (1995) J. Nutr. 126, 1076S–1080S. 8. Horrobin, D. F. (1992) Prog. Lipid Res. 31, 163–194. 9. Brenner, R. R. (1981) Prog. Lipid Res. 20, 41– 47. 10. Sayanova, O., Smith, M. A., Lapinskas, P., Stobart, A. K., Dobson, G., Christie, W. W., Shewry, P. R., and Napier, J. A. (1997) Proc. Natl. Acad. Sci. USA 94, 4211– 4216. 11. Napier, J. A., Hey, S. J., Lacey, D. J., and Shewry, P. R. (1998) Biochem. J. 330, 611– 614. 12. Michaelson, L. V., Napier, J. A., Lewis, M., Griffiths, G., Lazarus, C. M., and Stobart, A. K. (1998) FEBS Lett. 439, 215–218. 13. Michaelson, L. V., Lazarus, C. M., Griffiths, G., Napier, J. A., and Stobart, A. K. (1998) J. Biol. Chem. 273, 19055–19059. 14. Knutzon, D. S., Thurmond, J. M., Huang, Y. -S., Chaudhary, S., Bobik, E. G., Chan, G. M., Kirchner, S. J., and Mukerji, P. (1998) J. Biol. Chem. 273, 29360 –29366. 15. Shanklin, J., Whittle, E., and Fox, B. G. (1994) Biochemistry 33, 12787–12794. 16. Aki, T., Matsumoto, Y., Morinaga, T., Kawamoto, S., Shigeta, S., Ono, K., and Suzuki, O. (1998) J. Ferment. Bioeng. 86, 504 –507. 17. Okayasu, T., Nagao, M., Ishibashi, T., and Imai, Y. (1981) Arch. Biochem. Biophys. 206, 21–28. 18. Kozak, M. (1981) Nucleic Acids Res. 9, 5233–5262. 19. Napier, J. A., Sayanova, O., Stobart, A. K., and Shewry, P. R. (1997) Biochem. J. 328, 717–720. 20. Voss, A., Reinhart, M., Sankarappa, S., and Sprecher, H. (1991) J. Biol. Chem. 266, 19995–20000. 21. Sprecher, H., Luthria, D. L., Mohammed, B. S., and Baykousheva, S. P. (1995) J. Lipid Res. 36, 2471–2477. 22. Infante, J. P. and Huszagh, V. A. (1998) FEBS Lett. 431, 1– 6. 23. Christiansen, E. N., Lund, J. S., Rørtveit, T., and Rustan, A. (1991) Biochem. Biophys. Acta 1082, 57– 62. 24. Geiger, M., Mohammed, B. S., Sankarappa, S., and Sprecher, H. (1993) Biochem. Biophys. Acta 1170, 137–142.
579