Properties of two multifunctional plant fatty acid acetylenase/desaturase enzymes Anders S. Carlsson 1 , Stefan Thomaeus 1 , Mats Hamberg 2 and Sten Stymne 1 1 Department of Crop Science, Swedish University of Agricultural Sciences, Alnarp, Sweden; 2 Department of Medical Biochemistry and Biophysics, Division of Physiological Chemistry II, Karolinska Institutet, Stockholm, Sweden The properties of the D6 desaturase/acetyle nase from the moss Ceratodon purpureus and the D12 acetylenase from the dicot Crepis alpina were studied by e xpressing t he encoding genes in Arabidopsis thaliana and Saccharomyces cerevisiae. The acetylenase from C. alpina D12 desaturated both oleate and linoleate with about equal efficiency. The d esaturation of oleate gave rise to 9(Z),12(E)- and 9(Z),12(Z)-octadeca- dienoates in a ratio of approximately 3 : 1. Experiments using stereospecifically deuterated oleates showed that the pro-R hydrogen atoms were removed from C-12 and C-13 in the introduction of the 12(Z) double bond , whereas the pro-R and pro-S hydrogen atoms were removed from these carbons during th e formation of t he 12( E) double bond. The results suggested that the D12 acetylenase could accommo- date oleate having either a cisoid or transoid conformation of the C 12 -C 13 single bond, and t hat these conformers served as precursors of the 12(Z)and12(E) double bonds, respectively. However, only the 9(Z),12(Z)-octadecad ieno- ate isomer could be further d esaturated to 9(Z)-octadecen- 12-ynoate (crepenynate) by the enzyme. The evolutionarily closely related D12 epoxygenase from Crepis palaestina had only weak desaturase activity but could also produce 9(Z),12(E)-octadecadienoate from oleate. The D6 acetyle- nase/desaturase from C. purpureus, on the other hand, produced only the 6(Z) isomers using C16 and C18 ac yl groups possessing a D9 double bond as substrates. T he D6 double bond was efficiently further con verted to an acety- lenic bond by a second round of desaturation but only if the acyl substrate had a D12 double bond and that this was in the Z configuration. Keywords:acetylenase;Ceratodon purpureus; Crepis sp.; desaturase; Saccharomyces cerevisiae. Recently, a number of plant genes have been cloned that encode enzymes evolutionarily related to the D12 desatu- rase converting oleate to linoleate [9(Z),12(Z)-octadeca- dienoate]. These enzymes catalyze not only D12-cis desaturation but also hydroxylation [1,2], epoxidation [3], formation of acetylenic bonds [3,4] of conjugated double bonds [5–8] and of E double bonds [9]. The hydroxylases have been shown to also carry out D12(Z) desaturation [2] and conjugases to be able to efficiently D12 (E) desaturate oleate [7]. Here we report on the multifunctionality a nd stereochemistry o f another mem- ber of the D12 desaturase-like enzyme family, the Crepis alpina D12 acetylenase, and show that this enzyme, beside its ability to form triple bonds, is able to convert oleic acid into a mixture of 12(E)and12(Z) isomers of 18:2. The catalytic activity of the C. alpina acetylen ase enzyme is compared with that of the evolutionarily closely related Crepis palaestina epoxygenase and another, evolutionarily distantly related, bifunctional acetylenase/ desaturase from the moss Ceratodon purpureus. Materials and methods Plant material and growth conditions Nontransformed Arabidopsis tha liana L. (Heynh) plants of ecotype Columbia (wild type), its fad2 mutant defective in the endoplasmatic reticular oleoyl D12 desaturase, as well as transgenic A. thaliana, were grown in controlled growth chamber at a photosynthetic flux of 100–120 lEÆm )2 Æs )1 , 20 °C in a photoperiod of 16 h light/8 h dark. C. palaestina and C. alpina plants were grown under greenhouse condi- tions. Expression of D12 desaturase-like enzymes in A. thaliana The cloning of the C. alpina D12 acetylenase (CREP1) as well as the C. palaestina D12 epoxygenase ( CPAL2) genes has been described earlier [3]. Full-length cDNA of CREP1 and CPAL2 were inserted into the binary vector o f pBI121 under the transcriptional c ontrol of a truncated version (FP1) of the seed-specific napin promoter [10]. Transfor- mation and cultivation of the A. thaliana plants was as described earlier [11]. All fatty acid analyses of the transformants were performed on T 3 seeds. Correspondence to A. S. Carlsson, Department of Crop Science, Swedish University of Agricultural Sciences, PO Box 44, 230 53 Alnarp, Sweden. Fax: + 46 40 415519, Tel.: + 46 40 415561, E-mail: anders.carlsson@vv.slu.se Abbreviations: 18:1D9(Z), oleic acid; 18:2D9(Z),12(Z), linoleic acid; 18:2D9(Z),12a, crepenynic acid. Note: a web page is available at http://www.vv.slu.se (Received 2 April 2004, revised 21 May 2004, accepted 26 May 2004) Eur. J. Biochem. 271, 2991–2997 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04231.x Expression of D6- and D12-desaturase-related enzymes in Saccharomyces cerevisiae The C. alpina CREP1 was expressed in S. cerevisiae (w303– 1 A strain) with the plasmid p VT-CREP1 containing the constitutive alcohol dehydrogenase promoter [3]. The D6 acetylenase (CER1)fromC. purpureus [12] was cloned behind the galactose-inducible promoter GAL1 of the yeast expression vector pYES2 (Invitrogen) generating the plas- mids pYCER1 [12] and transformed in S. cerevisiae INVSc1 cells (Invitrogen) using the poly(ethylene glycol) method [13]. Cell growth Transformed yeast cells were grown in complete minimal dropout-uracil medium (CMdum), i.e. YNB complete medium (Sigma, St Louis, MO, USA) and complete supplement mixture without uracil (Bio101,inc) with 2% (v/v) glucose, a t 30 °C for 48 h. The cells were then diluted to A 600 ¼ 0.2 with fresh CMdum (total volume 10 mL). Fatty acids dissolved in 10% (v/v) Tween 40 (BDH Chemicals Ltd, Poole, UK) were then added to the cultures transformed with pYCER1 to a final concentration of 0.03% (w/v) fatty acids and 1% (w/v) Tween 40. These cultures were incubated further for 8 h after which expression of the CER1 gene was induced by adding galactose to a final concentration of 1.8% (w/v). The addition of fatty acids in Tween 40 (at the same concentrations as described above) was carried out after 8 h of incubation to th e cultures t ransformed with CREP1. All cell cultures were then grown for additional 72 h before harvesting. The cells were subsequently broken by a glass-bead shaker and their lipids extracted into chloroform according to the method devised b y Bligh and Dyer [14]. Fatty acid and lipid analysis Polar lipids were sep arated from neutral lipids by thin l ayer chromatography in hexane/diethyl ether/acetic acid (70 : 30 : 1, v/v/v) on precoated silica gel 60 plates (Merck). Gels from polar lipid areas (application spots) were scraped off and methylated in situ with 0.1 M sodium methoxide . Fatty acid methyl esters of total lipids from chloroform extracts of yeast cells were prepared by alkaline transme- thylation using 2 mL of 0.1 M sodium methoxide for 5 min at 90 °C. Preparation of methyl esters from the seeds were performed by treatment of whole seeds with 2 mL of 0.1 M sodium methoxide for 55 min at 90 °C. Fatty acid methyl esters were extracted with hexane and quantified by GC using a WCOT fused silica capillary column (50 m · 0.32 mm; film thickness, 0.25 lm), CP-wax 58 (FFAP)-CB (Chrompack, Middelburg, the Netherlands). GC was performed on a Shimadzu GC-17 A equipped with an autoinjector AOC-20i a nd an FID ( flame ionization detector) detector using helium (4.0 mLÆmin )1 )ascarrier gas. The injector and detector temperatures were at 230 and 270 °C, respectively. Oven temperature was held at 165 °C for 0.5 min and then raised to 25 0 °Cat4.0°CÆmin )1 . Heptadecanoic acid methyl ester was used as an internal standard. Fatty acid preparations 9(Z),12(E)-Octadecadienoic a cid w as prepared from cis - 12,13-epoxy-9(Z)-octadecenoic acid by hydrolysis into the threo-12,13-diol f ollowed by syn elimination of the diol using the Corey–Winter procedure [15]. The 9(Z),12(E)- octadecadienoate was obtained in 41% yield and in > 98% purity following preparative RP-HPLC. cis-12,13-Epoxy- 9(Z)-octadecenoic acid, 9(E),12(Z)-octadecadienoic acid and 9(E),12(E)-octadecadienoic acid were purchased from Larodan Fine Chemicals, Malmo ¨ , Sweden. [12(S)- 2 H]Oleic acid was prepared from 12(R)-hydroxy-9(Z)-octadecenoic acid (ricinoleic acid) using previously described m ethodo- logy [16] (Fig. 1). The isotope composition was 96.2% monodeuterated and 3.8% undeuterated molecules. [(+/–)-erythro-12,13– 2 H 2 ]Oleic acid was synthesized from the methyl ester of cis-9,10-epoxy-12(Z)-octadecenoic acid by catalytic deuteration using Wilkinson’s catalyst followed by elimination of the cis-epoxide function using triphenylphosphine selenide (Fig. 1). The material obtained was purified by RP-HPLC to afford [(+/–)-erythro- 12,13- 2 H 2 ]oleic acid in 55% yield. The isotope composition was 97.0% dideuterated, 1 .4% monodeuterated and 1.6% undeuterated molecules. It is well established that deuter- ations using Wilkinson’s catalyst take place by clean syn Fig. 1. Reaction used to prepare deuterated substrates. Reactionsusedtoprepare(A)[12(S)- 2 H]oleic acid and (B) [ erythro-12,13– 2 H 2 ]oleic acid. 2992 A. S. Carlsson et al. (Eur. J. Biochem. 271) Ó FEBS 2004 addition of deuterium [17], i.e. deuteration of a Z double bond will afford the erythro dideuterated product. Also the triphenylphosphine selenide-promoted elimination takes place stereospecifically, i.e. producing a Z alkene from a cis-epoxide [18]. GC-MS and GLC analysis GC-MS was carried out with a Hewlett-Packard model 5970B mass selective detector connected to a Hewlett- Packard model 5890 gas chromatograph equipped with a capillary column of Supelcowax (30 m ; film thickne ss 0.25 lm; carrier gas helium). Injections were made in the split mode using an initial column temperature of 150 °C. The temperature was raised at 3 °CÆmin )1 until 250 °C. GLC with flame ionization detection was carried out with a Hewlett-Packard model 5890 gas chromatograph under the same conditions as those used in GC-M S. Identification of 9( Z ),12( E )-octadecadienoic acid An unknown 9,12-octadecadienoate produced in th e pres- ence of recombinant C. alpina D12 acetylenase in A. thaliana and yeast and also present in seeds of C. alpina was isolated as its methyl ester by RP-HPLC using a solvent syst em of acetonitrile/water 75 : 25 (v/v). The ester cochromato- graphed w ith methyl 9(Z),12(E)- and 9 (E),12(Z)-octadeca- dienoates (effluent volume, 65 mL) but separated from methyl linoleate (59 mL) and methyl 9(E),12(E)-octadeca- dienoate (71 mL). The material obtained was analyzed by GLC. Its retention time was identical to that of methyl 9(Z),12(E)-octadecadienoate (20 min 30 s), but differed from those of m ethyl 9(E),12(Z)-octadecadienoate (20 min 50s),methyl9(Z),12(Z)-octadecadienoate (20 m in 16 s), and from those of a number of isomeric 9,11- and 10,12-octadecadienoates (22 min 58 s to 24 min 16 s). Analysis by GC-MS was in full agreement with the identification of the unknown octadecadienoate as methyl 9(Z),12(E)-octadecadienoate. Results The fatty acid composition of lipids from C. alpina and C. palaestina seeds were compared with that of seeds from wild type and the fad2 mutant as well as seeds from A. thaliana lines transformed w ith either the C. alpina D12 acetylenase gene (CREP1)ortheC. palaestina epoxygenase gene (CPAL2)(Table1).C. alpina seed lipids contained significant amount (3.3%) of 9(Z),12(E)-octadecadienoate (Materials and methods section), which have also been reported earlier in Crepis rubra [19], whereas the seed lipids from C. palaestina, wild type and the fad2 mutant did not contain detectable amounts of this un usual stereoisomer of 18:2. In order to investigate if the synthesis of this E isomer of 18:2 in C. alpina seed was a product of t he C. alpina D12 acetylenase activity, the acetylenase gene was expressed in both wild type and t he fad2 mutant. Both transgenic lines produced this fatty acid and the concentration was about 50% higher in the background of the fad2 mutant than in the wild type background (3.5% vs. 2.3%) indicating that oleate was the substrate for the formation of the D12E double bond. It should be noted that the level of 18:2D9Z,12E was nearly four times higher than the level of the acetylenic acid 18:1D9Z,12a (crepenynate) in t he wild type transformed with CREP1.Inthefad2 mutant trans- formed with CREP1, only trace amount of crepenynate was detected. There was a significant increase of 18:2D9Z,12Z in the fad2 line transformed with CREP1 compared with the nontransformed mutant, indicating that the acetylenase also could carry out D12(Z)desaturationofoleate. We also analyzed the acyl composition of seed lipids from cwild type and fad2 mutant transformed with the C. palaestina D12 epoxygenase gene (CPAL2), which is evolutionary closely related to the C. alpina acetylenase [3]. Small amounts of 18:2D9Z,12E were detected in both transgenic lines (0.2%), and this was both 30 and nine times less than the epoxy fatty acids formed by the epoxygenase in the t ransformed wild type and fad 2 lines, r espectively (Table 1). The wild type transformed with CPAL2 showed Table 1. Acyl c omposition o f seeds from vild species and transgenic plant lines. Acyl composition in seeds from C. alpina, C. palaestina and wild type and fad2 mutants of nontransforme d A. thaliana and transformed with either the C. alpina D12 acetylenase gene (CREP1)ortheC. palaestina epoxygenase gene (CPAL2). nd, Not detected. Fatty acids Acyl composititon (area percentage) C. alpina A. thaliana fad2 A. thaliana + CREPI fad2 + CREPI C. palaestina A. thaliana + CPAL2 fad2 + CPAL2 16:0 4.6 7.6 4.9 8.7 6.1 4.4 9.0 5.0 16:1 D9Z 0.1 0.3 0.3 0.3 0.4 0.1 0.4 0.5 18:0 2.0 4.6 2.6 3.2 2.7 2.5 4.3 3.0 18:1 D9Z 1.1 17.6 42.8 29.0 46.3 9.4 38.8 55.0 18:1 D11Z 0.9 1.2 1.7 2.3 2.2 0.5 1.8 3.0 18:2 D9Z,12Z 9.8 24.7 4.3 18.7 6.6 23.7 4.7 2.0 18:2 D9Z,12E 3.3 nd nd 2.3 3.5 nd 0.2 0.2 18:2 D9Z,15Z nd nd 0.7 nd 0.3 nd nd 0.4 18:3 D9Z,12Z,15Z nd 18.9 11.2 10.6 7.6 nd 6.7 4.0 20:1 D11Z nd 20.9 21.8 18.2 20.8 nd 21.8 18.0 18:1 D9Z, 12a 77.8 nd nd 0.6 Trace nd nd nd 12-Epoxy-18:1 D9Z nd nd nd nd nd 54.1 3.6 0.9 12-Epoxy-18:2 D9Z,15Z nd nd nd nd nd nd 2.1 0.9 Others 0.4 4.1 9.7 6.8 3.6 5.4 6.6 7.1 Ó FEBS 2004 Stereochemistry of acetylenase/desaturase enzymes (Eur. J. Biochem. 271) 2993 a drastic reduction in 18:2D9Z,12Z content compared with the untransformed plants (4.7% vs. 24.7%) and in the fad2 line transformed with CPAL2, the 18:2D9Z,12Z content was decreased from 4 % to 2%. It has p reviously been shown that epoxy fatty acid production in A. thaliana severely inhibits formation of linoleate and that the degree of inhibition is correlated with the levels of the epoxy fatty acids accumulated [11]. In order to further investigate t he catalytic activities of the C. alpina acetylenase, we transformed baker’s yeast (S . cerev isiae)withtheCREP1 gene. In contrast to A. thaliana, S. cerev isiae lacks D12 desaturase activities and can readily incorporate exogenous added fatty acids into their lipids. In absence of exogenous fatty acids, the yeast expressing the C. alpina acetylenase gene converted up to 0.65% of their fatty acids into 1 8:2 in a ratio of D12(E) to (Z) isomers of approximately 3 : 1 (Table 2). Similar conversion rates were seen with yeast cultures fed o n exogenously added oleate (Table 2). Only trace amounts (0.01–0.02%) of 18:1D9Z,12a were produced in these cultures. When linoleate was added t o the cultures, the formation of the E isomer of 18:2 was still evident, although its c oncentration was reduced by 80% (to 0.09%) and 0.53% of crepenynic acid was formed. When 18:2D9Z,12E was added to the yeast transformed with CREP1,no crepenynic acid was formed. As acyl groups linked to phospholipids are the s ubstrate for the D12 desaturase-like enzymes, we measured the acyl composition of the polar lipids in the yeast after incubation with the exogenous fatty acids. The l evels of t he two e xogenously s upplied 18:2 isomers were approximately the same in polar lip ids as in total lipids, accounting for 49 and 41% of the total acyl groups in the incubation with added (Z)and(E) isomer, respectively (data not shown). As it has previously been shown that a D6 acetylenase (CER1) from the moss C. purpureus also has D6desatu- rase activity [12], w e i nvestigated if that enzyme was capable of also carry out E desaturation. Yeast trans- formed with the CER1 was grown in presence of either 18:2D9Z,12Z or 18:2D9Z,12E (Table 3). The CER1 desaturated endogenous 16:1D9Z,18:1D9Z and exogenous 18:2D9Z,12Z to the corresponding D6( Z) derivatives and also efficiently further converted the formed 18:3D6Z,9Z,12Z (c-18:3) to 18:3D6a,9Z,12Z. Exogenous 18:2D9Z,12E wasconvertedto18:3D6Z,9Z,12E but not further to t he acetylenic compound. No D6(E) isomers were found in the yeast transformed with the D6acetyl- enase gene. It was of particular interest to investigate the stereochem- istry in the removal of hydrogen atoms in the formation of the two geometrical isomers of 18:2 produced by CREP1. Extracts containing fatty acid m ethyl esters obtained following feeding of [12(S)- 2 H]- and [erythro-12,13- 2 H 2 ]oleic acids to yeast expressing the CREP1 gene were subjected to GC-MS, and the isotope contents of methyl 9(Z),12(Z)- and 9(Z),12(E)-octadecadienoates were determined by selected ion monitoring. Additionally, the isotope contents of oleates recovered in the two experiments were determined. The oleate recovered a fter feeding o f [12(S)- 2 H]oleate was a mixture o f 39.7 deuterated and 60.3% undeuterated mole- cules showing that the a dded oleate was diluted approxi- mately 1.5-fold with unlabeled material during the incubation period. A similar dilution was noted for the oleate recovered after feeding of the [erythro-12,13- 2 H 2 ] oleate, i.e. 34.8% dideuterated, 0.6% monodeuterated, and 64.6% undeuterated molecules. As seen in Table 4, 9(Z),12(Z)- and 9(Z),12(E)-octadecadienoates both retained most of the deuterium label w hen formed f rom [12(S)- 2 H]oleate, and accordingly, the 12(R) hydrogen was lost from C-12 when the 12(Z)and12(E) double bonds were introduced. F eeding of [erythro-12,13- 2 H 2 ]oleates produced 9(Z),12(Z)-octadecadienoic acid, which was mainly either undeuterated or dideuterated, had an isotopic composition close to that expected for removal of the 12(R),13(R) or 12(S),13(S) deuteriums (Table 5). In contrast, the 9(Z),12(E)-octadecadienoate mainly consisted of either undeuterated or monodeuterated molecules in accordance with a removal of the 12(R),13(S)or12(S),13(R)deute- riums. By combining the results in Tables 4 and 5 it could be deduced that the 9(Z),12(Z)-octadecadienoate was formed by elimination of the pro-R hydrogen atoms at C-12 and C-13 whereas the 9(Z),12(E)-octadecadienoate was formed by elimination of the pro-R hydrogen at C-12 and the pro-S hydrogen at C-13. Table 2. Acyl composition in nontransformed (empty plasmid) yeast (S. cerevisiae ) and in yeast expressing D12 acetylenase gene (CREP1). Yeast was grown in the absence and presence of e xogenous fatty acids. nd, Not detected. Acyl group Acyl composition (area percentage) Empty plasmid Fatty acid added CREP1 Fatty acid added None 18:1D9z 18:2D9Z,12Z 18:2D9Z,12E None 18:1D9z 18:2D9Z,12Z 18:2D9Z,12E 16:0 22.3 23.1 23.1 23.0 20.4 23.4 25.4 24.1 16:1D9Z 37.2 29.2 8.1 18.3 25.9 21.1 5.3 10.6 18:0 10.8 9.4 10.4 9.2 14.9 12.6 13.2 13.1 18:1D9Z 28.9 37.3 6.3 12.6 37.2 41.8 5.7 10.7 18:1D11Z 0.7 0.9 0.01 0.2 0.9 0.06 0.08 0.1 18:2D9Z,12Z nd nd 51.9 nd 0.18 0.19 49.3 nd 18:2D9Z,12E nd nd nd 36.6 0.47 0.41 0.09 41.0 18:1D9Z,12a nd nd nd nd 0.02 0.01 0.53 nd Others 0.1 0.1 0.1 0.1 0.03 0.4 0.4 0.4 2994 A. S. Carlsson et al. (Eur. J. Biochem. 271) Ó FEBS 2004 Discussion We demonstrate that the C. alpina D12 acetylenase is a tri- functional enzyme that efficiently produces a mixture of D12 (Z)and(E) isomers of 18:2 from o leate as well as 18:1D9Z,12a (crepenynate) from l inoleate. The evolutionarily closely related D12 epoxygenase from C. palaestina [11] was, on the o ther hand, primarily an 18:2 epoxygenase utilizing linoleate as the substrate, although it had also a weak D12 (E) oleate desaturase activity when ex pressed in A. thaliana. The evolutionarily distantly r elated bifunctional D6 d esaturase/acetylenase from C. purpureus did not exhibit (E) desaturase a ctivity, demonstrating that such activity is not an inherent property of acetylenases. The tra ns desaturation of oleate to 18:2D9Z,12E has now been shown for three different D12 desaturase-like enzymes, the C. alpina acetylenase a s shown here, the conjugase from tung [7] and the desaturase from Dimorphoteca sinuata [9]. In D. sinuata seeds, the D12(E) desaturase produces the substrate for a second D12 desaturase-like enzyme converting 18:2D9Z,12E into 9-hydroxy-18:2D10E,12E, the dominating acyl group in D. sinuata seeds. However, in C. alpina seeds the 18:2D9Z,12E is an end product, as we show here that it cannot be converted further into crepenynic acid. Of all the fatty acid desaturases so far reported it is only t he C. alpina acetylenase that produces a m ixture of E and Z isomers. However, a plant sphingolipid D8 desaturase that converts 4-hydroxy-sphinganine into a mixture of 7 : 1 of E and Z isomers o f 4-hydroxy-8-sphingenine has been cloned and characterized [20]. A single fatty acid desatu- rase enzyme has been implicated in both E and Z D11 desaturation of myristoyl-CoA in insects [21]. It is of special interest that the two isomeric products of this desaturation in insects are precursors for two different pheromones with different biological activities. However, no gene encoding a D11 desaturase yielding both E and Z isomers has yet been cloned. The physiological significance of the formation of 18:2D9Z,12E by C. alpina acetylenase is unknown. The e nzyme serves to produce acetylenic fatty acids, wh ich are most probably i nvolved in pathogen Table 3. Acyl composition of nontransformed yeast (S. cerevisiae) and yeast expressing D6 acetylenase (CER1). Yeastweregrowninthepresenceof 18:2D9Z,12Z or 18:2:2D9Z,12E. nd, Not detected. Fatty acids Acyl composition (area percentage) Empty plasmid Fatty acid added CER1 Fatty acid added 18:2D9Z,12Z 18:2D9Z,12E 18:2D9Z,12Z 18:2D9Z,12E 16:0 32.8 30.4 28.5 20.3 16:1D9Z 5.5 17.7 5.5 22.6 16:2D6Z,9Z nd nd 0.15 0.64 16:2D6E,9Z nd nd nd nd 18:0 10.0 10.9 11.5 12.2 18:1D9Z 7.7 21.4 8.1 30.1 18:1D11Z 0.16 0.61 0.12 0.90 18:2D6Z,9Z nd nd 0.05 0.22 18:2D6E,9Z nd nd nd nd 18:2D9Z,12Z 43.5 nd 44.8 nd 18:2D9Z,12E nd 18.9 nd 12.3 18:3D6Z,9Z,12Z nd nd 0.69 nd 18:3D6E,9Z,12Z nd nd nd nd 18:3D6Z,9Z,12Z nd nd nd 0.69 18:2D6a,9Z,12Z nd nd 0.57 nd Others 0.34 0.09 0.02 0.05 Table 4. Isotope compositions of 9(Z),12(Z)- and 9(Z),12(E)-octa- decadienoates generated from [12(S)- 2 H]oleic acid. The oleate given to the cells had the following isotopic composition: 96.2% monodeuter- ated, 3.8% undeuterated molecules, an d the oleate recovered consisted of 39.7% deuterium-labeled and 60.3% unlabelled molecules due to dilution with endogenous material. d 0 (%) d 1 (%) 9(Z),12(Z)-octadecadienoate 36.4 63.6 9(Z),12(E)-octadecadienoate 34.4 65.6 Expected for R elimination 39.7 60.3 Expected for S elimination 0 100 Table 5. Isotope compositions of 9(Z),12(Z)- and 9(Z),12(E)-octa- decadienoates generated from [erythro-12,13- 2 H 2 ]oleic acid. The oleate given to the cells had the following isotopic composition: 97.0% dideuterated, 1.4% monodeuterated, 1.6% undeuterated molecules, and the oleate recovered consisted of 34.8% dideuterated, 0.6% monodeuterated, and 64.6% undeuterated molecules due to dilution with endogenous material. d 0 (%) d 1 (%) d 2 (%) 9(Z),12(Z)-octadecadienoic acid 18.5 0.3 81.2 9(Z),12(E)-octadecadienoic acid 2.0 30.6 67.4 Expected for R,R or S,S elimination 17.4 0.3 82.3 Expected for R,S or S,R elimination 0 35.1 64.9 Ó FEBS 2004 Stereochemistry of acetylenase/desaturase enzymes (Eur. J. Biochem. 271) 2995 defense mechanisms [4], however, other functions of the enzyme are conceivable. Stereochemical and mechanistic studies have been per- formed on the trans and cis desaturation by the myristoyl- CoA desaturase [21], the sphingolipid desaturase [20] and of the formation of the acetylenic bond by the C. alpina acetylenase [22]. Our studies using stereospecifically deuter- ated oleates revealed that the desaturations leading to the 12(Z) a nd 12(E) alkenes by the C. alpina acetylenase proceeded with distinct ste reochemistries. Thus, the Z double bond-forming reaction resulting in 9(Z),12(Z)- octadecadienoic acid took place by selective removal o f the pro-R hydrogen atoms from C-12 and C-13, a result that was in accord with previous studies of other Z-desaturases [21, 23, 24]. Formation of the E double bond of 9(Z),12(E)- octadecadienoic acid took place by selective removal o f the pro-R hydrogen from C-12 and the pro-S hydrogen from C-13. It is well established from studies of kinetic isotope effects [20–24] that en zymatic desaturations take place in a stepwise manner and involve an initial, slow hydrogen abstraction followed by rapid collapse of the intermediate thus formed and expulsion of the second hydrogen atom. Importantly, because of the short lifetime of the singly desaturated intermediate, it is unlikely that a conformational change can take place p rior to eliminatio n of t he second hyd rogen atom. It seems v ery likely that this mechanism is valid also for the Crepis alpina D 12 acetylenase/desaturase studied in the presen work. Although the three-dimensional structure of this enzyme is unknown, it is possible t o rationalize the stereochemical results in terms of the different conforma- tions the oleate substrate has to adopt t o generate e ither a 12(Z)ora12(E) double bond. Thus, it can be postulated that the C 11 –C 14 segment of the oleate carbon chain has to be in the cisoid conformation to produce the 9(Z),12(Z)- octadecadienoate and in the transoid conformation to produce the 9(Z),12(E)-octadecadienoate (Fig. 2). In the biosynthesis of the 9(Z),12(Z) isomer, the pro-R hydrogens at C 12 and C 13 will be in close contact w ith the hydrogen- abstracting groups on the enzyme surface and be eliminated. In order to produce the 9(Z),12(E) isomer, the C 12 –C 13 single bond has to rotate to produce the transoid confor- mation of the C 11 –C 14 segment. As a consequence of this rotation, the pro-R hydrogen at C-13 will move away from the hydrogen-abstracting group whereas the pro-S hydrogen will come in contact and be eliminated (Fig. 2). A comparison between the end products produced by the D6 acetylenase and the D12 acetylenase in y east expressing genes for these enzymes revealed an interesting difference. When linoleate was fe d to the y east expressing the D6 acetylenase, the products of the desaturation reactions, i.e. 18:3D6Z,9Z,12Z and 18:2D6a,9Z,12Z accounted for about 1.3% of all fatty acids in the cell and of which 40% was found as the acetylenic fatty acid and thus had undergone two rounds of desaturation catalyzed by the acetylenase. Although this efficient conversion of 18:3D6Z,9Z,12 Z to 18:2D6a,9Z,12Z. Thus the product of the first desaturation round was efficiently competing with the about 40 times higher concentration of linoleate in the membranes. Although a D6(Z) desaturase gene from C. purpureus has been cloned [12], it i s t hus plausible that much of the 18:3D6Z,9Z,12Z used for the production of the acetylenic fatty acid is generated by the acetylenase enzyme itself in this moss. T his would be consistent with the very low levels of 18:3D6Z,9Z,12Z compared with linoleate found in phos- phatidylcholine (the site for D6 desaturation and acetylena- tion) in C. purpureus [25]. The C. alpina acetylenase, on the other hand, does not have such a metabolic channeling. Of the s mall amount of linoleate produced by the acetylenase in yeast, only 5–10% is converted further to 18:1D9Z,12a. It is interesting to note that 18:2D9Z,12E, which cannot be converted further into 18:1D9Z,12a, amounts to 75% of the products of oleate desaturation catalyzed by the C. alpina acetylenase. The C. alpina seed lipids (mainly triacylglycerols) contain about 3% of 18:2D9Z,12E and nearly 80% of 18:1D9Z,12a. Thus, the acetylenase is responsib le for the production of about 1% of 18:2D9Z,12Z in C. alpina seedsandtherestofthe linoleate produced in these seeds, which is 86% of all acyl groups in the seed (% linoleate +% crepenynate –% linoleate formed by the acetylenase), must be a result of D12(Z) oleate desaturase activity catalyzed by a separate enzyme. Because the C. alpina acetylenase appears as efficient as a desaturase as an acetylenase, this n early total out competing of the acetylenase by the D12(Z)desaturase for oleate substrate in C. alpina seeds appears at fi rst sight puzzling. However, results presented here using cells having a different ratio of oleate to linoleate indicate that this ratio determines the ratio of utilization of the two substrates by the acetylenase. The ratio of o leate to linoleate in C. alpina seeds is 1 : 10 and under such conditions a d esaturation to acetylenation ratio of about 1 : 20 would be expected based on ou r results in transgenic organisms. This is close to the ratio that could be calculated on the basis of the amount of 18:2D9Z,12E and 18:1D6a,9Z found in these seeds. The conjugase from tung also has, in addition to its conjugase activity with linoleate, a high oleate D12(E) desaturase activity [7]. The tung seed lipids has about 3% of 18:2D9Z,12E and 80% of a-eleostearic acid (18:3D9Z, 11E,13E) and it is probable that the relative proportion of oleate and linoleate in these seeds also determines the relative rate of utilization of these two substrates by the conjugase [7]. However it is not known if 18:2D9Z,12E is an end product or if it could be used by the conjugase for the synthesis of a-eleostearic acid. Fig. 2. Conformations of oleic ac id bound to C. alpina D12 acetylenase. The conformation in (A) will prod uce 9(Z),12(Z)-octadecadienoic acid and the one in (B) will produce 9(Z),12( E)-o ctadecadienoic acid. 2996 A. S. Carlsson et al. (Eur. J. Biochem. 271) Ó FEBS 2004 The data presented here and by Sperling et al. [12] show that the D6 acetylenase can utilize a variety of D9 desaturated a cyl groups as substrates for the first desatu- ration reaction, such as 16:1D9Z, 18:1D9Z, 18:2 D9Z,12Z, 18:2D9Z,12Eand18:3D9Z,12Z,15Z. However, the forma- tion of acetylenic bonds is only occurring with substrates having a D12(Z) double bond. Thus in order f or the double bond to be orientated properly t owards the active site f or further hydrogen subtraction, a D12 double bond in (Z) configuration seems essential. The r esult p resented in this paper is an additional demonstration of the great flexibility in substrate acceptance and catalytic outcome of the fatty acid desaturase enzymes. In case of the D6 acetylenase from the moss C. purpureus, this implies that the enzyme not only produces D6acetylenic fatty acids in this species, but also provides its own D6 desaturated substrate f or this reaction. In contrast, the C. alpina D12 acetylenase produces mainly a dead-end substrate from oleate and thus must obtain its substrate for the a cetylenase r eaction of a separate D12 des aturase enzyme. Acknowledgement Financial su pport from the Swedish University of Agricultural Sciences strategic research grant ÔThe Biologica l F actory/Agr iFunGenÕ,Swedish Oil Growers Association (SSO), VL-stiftelsen, and the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (project number 2001–2553) are gratefully acknowledged. Arabidopsis fad2 seeds transformed with CREP1 and CPAL1 genes were kindly provided by Drs Allan Green and Surinder Singh at CSIRO, Canberra, Australia. References 1. Broun, P. & Somerville, C. (1997) Accumulation of ricinoleic, lesquerolic, and densipolic acids in s eeds of transgenic Arabidopsis plants that express a fatty acyl hydroxylase cDNA from castor bean. Plant Physiol. 113, 933–942. 2. Broun, P., Boddupalli, S. & Somerville, C. (1998) A bifunctional oleate 12-hydroxylase: desaturase from Lesquerella fendleri. Plant J. 13, 201–210. 3. Lee,M.,Lenman,M.,Banas,A.,Bafor,M.,Singh,S.,Schweizer, M., Nilsson, R., Liljenberg, C., Dahlqvist, A., Gummeson, P.O., Sjo ¨ dahl, S., Green, A. & S tymne, S. (1998) Identification of non- heme diiron proteins that catalyze triple bond and epoxy group formation. Science 280, 915–918. 4. Cahoon, E .B., Schnurr, J.A., Huffman, E.A. & Min to, R.E. (2003) Fungal responsive fatty acid acetylenases occur widely in evolu- tionarily distant plant families. Plant J. 34, 671–683. 5. Cahoon, E.B., Carlson, T.J., Ripp, K.G., Schweiger, B.J., Cook, G.A., Hall, S.E. & Kinney, A.J. (1999) Biosynthetic origin of conjugated double bonds: production of fatty acid components of high-value drying oils in transgenic soybean embryos. Proc. Natl Acad. Sci. USA 96, 12935–12940. 6. Qiu, X., Reed, D.W., Hong, H.P., MacKenzie, S.L. & Covello, P.S. (2001) Identification and analysis of a gene from Calendula officinalis encoding a fatty acid conjugase. Plant Physiol. 125, 847–855. 7. Dyer, J.M., Chapi tal, D .C., K uan , J.C.W., Mullen, R.T., Turner, C., McKeon, T.A. & Pepperman, A.B. (2002) Molecular analysis of a b ifunction al fatty acid c onjugase/de saturase from tung: implications for the evolution of plant fatty acid diversity. Plant Physiol. 130, 2027–2038. 8. Iwabuchi, M., Kohno-Murase, J. & Imamura, J. (2003) Delta 12- oleate desaturase-related enzymes associated with formation of conjugated trans-delta 11, cis-delta 13 double bonds. J. Biol. Chem. 278, 4603–4610. 9. Cahoon, E.B. & Kinney, A.J. (2004) Dimorphecolic acid is syn- thesized by the coordinate activities of two divergent {delta}12- oleic acid desaturases. J. Biol. Chem. 279, 12495–12502. 10. Stalberg, K., Ellerstrom, M., Josefsson, L.G. & Rask, L. (1993) Deletion analysis of a 2s seed storage protein promoter of b rassica- napus in transgenic tobacco. Plant Mol. Biol. 23, 671–683. 11. Singh, S., Thomaeus, S., Lee, M., Stymne, S. & Green, A. (2001) Transgenic expression of a delta 12-epoxygenase gene in Arabi- dopsis seeds inhibits accumulation of linoleic acid. Planta 212, 872–879. 12. Sperling,P.,Lee,M.,Thomas,G.,Zahringer,U.,Stymne,S.& Heinz, E . (2000) A bifunctional delta (6)-fatty acyl acetylen ase/ desaturase from the m oss Ceratodon purpureus – a new member o f the cytochrome b(5) superfamily. Eur. J. Biochem. 267, 3801–3811. 13. Dohmen, R.J., Strasser, A.W.M.,Honer,C.B.&Hollenberg,C.P. (1991) An efficient tran sforma tion procedure enabling long-term storage of competent cells of various yeast genera. Yeast 7, 691–692. 14. Bligh, E.G. & Dyer, W.J. (19 59) A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911–917. 15. A., Corey, E.J. & Winter, R.A.E. (1963) A new, stereospecific olefin synthesis from 1,2-diols. J. Am. Chem. Soc. 85, 2677–2678. 16. Hamberg, M. & Samuelsson, B. (1967) On the mechanism of the biosynthesis of prostaglandins E 1 and E 1a . J. Biol. Chem. 242, 5336–5343. 17. Tulloch, A.P. (1979) Syn thesis o f d eu terium an d c arb on-13 labelled lipids. Chem. Phys. Lipids 24, 391–406. 18. Hamberg, M. (1992) A method for determination of the absolute stereochemistry of alpha,b eta-epoxy alcohols d erived from fatty- acid hydroperoxides. Lipids 27, 1 042–1046. 19. Gunstone, F.D. (1986) Chapter 1. In The Lipid Handbook (Gunstone F., Harwood J. and Padley F., eds), pp. 4. Chapman & Hall, London, UK. 20. Beckmann, C., Rattke, J., Oldham, N.J., Sperling, P., He inz, E . & Boland, W. (2002) Characterization of a delta(8)-sphingolipid desaturase from higher pla nts: a stereochemical a nd mechanistic study on the o rigin of E,Z. isomers. Angew. Chem. Int. Ed. Engl. 41, 2298–2300. 21. Pinilla, A., Camps, F. & Fabrias, G. (1999) Cryptoregiochemistry of the delta (11)-myristoyl-CoA desatu rase involved in the bio- synthesis of Spodoptera littoralis sex pheromone. Bioc hem istry 38 , 15272–15277. 22. Reed, D.W., Polichuk, D.R., Buist, P.H., Ambrose, S.J., Sasata, R.J., Savile, C.K., Ross, A.R.S. & Covello, P.S. (2003) Mechan- istic study of an improbable reaction: alkene dehydrogenation by the delta 12 acetylenase of Crepis alpina. J. Am. Chem. Soc. 125, 10635–10640. 23. Schroepfer, G.J.J. & Bloch, K. (1965) The stereospecific conver- sion of stearic acid to oleic acid. J. Biol. Chem. 24 0, 54–63. 24. Morris, L.J. (1970) Mechanisms and stereochemistry in fatty acid metabolism. Biochem. J. 118, 681–693. 25. Kohn, G., Hartmann, E., Stymne, S. & Beutelmann, P. (1994) Biosynthesis o f acetylenic fatty-acid s in the moss ceratodon pur- pureus (hedw) brid. J. Plant Physiol. 144, 265–271. Ó FEBS 2004 Stereochemistry of acetylenase/desaturase enzymes (Eur. J. Biochem. 271) 2997 . Properties of two multifunctional plant fatty acid acetylenase/desaturase enzymes Anders S. Carlsson 1 , Stefan Thomaeus 1 ,. for about 1.3% of all fatty acids in the cell and of which 40% was found as the acetylenic fatty acid and thus had undergone two rounds of desaturation