ELSEVIER Biochimica et Biophysica Acta 1301 (1996) 263-272 Biochi~ic~a et Biophysica A~ta Biosynthesis of docosahexaenoic acid in human cells" evidence that two different A6-desaturase activities may exist Isabel Marzo, Maria A. Alava, AndrOs Pifieiro, Javier Naval * Departamento de Bioquimica 3' Biologia Molecular, Facultad de Ciencias, Uniuersidad de Zaragoza, 50009 Zaragoza, Spain Received 9 February 1996; accepted 29 February 1996 Abstract It has been proposed that synthesis of docosahexaenoic acid (22:6(n- 3)) in rat hepatocytes occurs by a route independent of A4-desaturase, which involves A6-desaturation and retroconversion (Voss A., Reinhart M., Sankarappa S. and Sprecher H. (1991) J. Biol. Chem. 266, 19995-20000). However, most cells exhibit these enzymatic activities and nevertheless synthesize low to undectectable amounts of 22:6(n - 3). Moreover, there are few data on the occurrence of this pathway in human cells. In the present work, we have analysed the biosynthetic pathway of 22:6(n - 3) in human Y-79 retinoblastoma and Jurkat T-cells. Y-79 cells were supplemented with 18:3(n- 3) and 20:5(n- 3) or incubated with [1-J4C]18:3(n- 3) and [1-14C]20:5(n- 3) and lipids analysed by argentation TLC, reverse-phase TLC and GLC-mass spectrometry. Pulse-chase experiments revealed that synthesis of 22:6(n - 3) from 20:5(n - 3) in Y-79 cells occurred through two successive elongations, followed by a A6-desaturation of 24:5(n - 3) to 24:6(n - 3) and retroconversion to 22:6(n- 3). Incubation of Y-79 cells with [1-14C]18:3(n- 3) in medium containing 50 /xM trans-9,12-18:2, a potent inhibitor of A6-desaturase, caused a reduction of 22:6(n -3) synthesis mainly by interfering with the desaturation of 18:3(n- 3). However, when [1-]4C]20:5(n- 3) was used as precursor, synthesis of 22:6(n- 3) was depressed to a lesser extent and mainly by reduction of 24:6(n- 3) retroconversion. Neuronal differentiation of Y-79 cells caused a great increase in A6-desaturase activity on 18:3(n- 3), though the amount of 22:6(n - 3) synthesized did not change or diminish, suggesting the existence of a particular A6-desaturase involved in the synthesis of 22:6(n - 3). The existence of a distinctive A6-desaturase activity could also explain why Jurkat cells growing in serum-free medium showed a near 3-fold increase in the synthesis of pentaenes from 18:3(n - 3) and, at the same time, a large decrease in the synthesis of 22:6(n - 3). The verification of the involvement of two A6-desaturase activities in 22:6(n - 3) synthesis would have important implications for the formulation of the nutritional requirements of this fatty acid during development. Keywords: Polyunsaturated fatty acid; Docosahexaenoic acid; Retroconversion; A6-Desaturase; A4-Desaturase; Human retinoblastoma; Human leukemia 1. Introduction Biosynthesis of long-chain polyunsaturated fatty acids (PUFA) in mammalian cells occurs through a sequence of alternating desaturation and chain-elongation reactions act- ing from the essential fatty acids, linoleic, 18:2(n - 6), and a-linolenic,18:3(n- 3) [1,2]. Arachidonic acid, 20:4(n- 6), the major product of the n - 6 series, is generated from Abbreviations: DMA, dimethyl acetal; FCS, fetal calf serum; PUFA, polyunsaturated fatty acids; FAME, fatty acid methyl esters; SFM, serum-free medium; PBS, phosphate-buffered saline, pH 7.4; GLC, gas- liquid chromatography; dibutyryl cyclic AMP, db-cAMP; TLC, thin-layer chromatography. * Corresponding author. Fax: + 34 76 761159; e-mail: jnaval@msf.un- izar.es. 0005-2760/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PH S0005-2760(96)0005 1 - 3 18:2 by the sequential action of A6-desaturase, an elongase and AS-desaturase. The same pathway acting on 18:3(n- 3) yields 20:5(n - 3). However, the most abundant PUFA of the n - 3 series is docosahexaenoic acid, 22:6(n - 3), a key fatty acid in membrane phospholipids of tissues such as retina, heart, cerebral cortex and spermatozoa [2-4]. According to the commonly accepted pathway, synthesis of 22:6(n- 3) requires the elongation of 20:5(n- 3) to 22:5(n - 3) followed by a A4-desaturation [1,2]. Neverthe- less, a direct demonstration of the existence of A4-de - saturase has remained elusive. Recently, Voss et al. [5], working with rat liver microsomes and hepatocytes, have indicated that the apparent A4-desaturase activity on 22:5(n- 3) is due, instead, to the combined effect of a further elongation to 24:5(n-3) followed by a A6-de - saturation to 24:6(n - 3) and retroconversion to 22:6(n - 3). The proposed intermediates for this new pathway have