Dynamic lipidomics using ESI–MS (tandem electrospray ionization mass spectrometry) of 9-deuterated choline (choline-d9) incorporation into mammalian cell PtdCho (phosphatidylcholine) permits assessment of the molecular specificity of synthesis. Bulk cell PtdCho synthesis occurs in spatially distinct locations, using separate CPTs (1,2 diacylglycerol CDP:choline cholinephosphotransferases). We assessed whether in vitro molecular selectivity of DAG (diacylglycerol) incorporation between CPTs is manifest in situ, by monitoring choline-d9 incorporation into PtdCho and lyso-PtdCho molecular species over 3 h in control cells and in CHO-K1 cells overexpressing hCEPT1. Compared with controls, the basal molecular species composition of hCEPT1 overexpressors was significantly enriched in arachidonate. This was not due to net accretion of cellular PtdCho arguing against effects of inadequate unsaturated PtdCho degradation or remodelling. Rather, time-course analyses of PtdCho and lyso-PtdCho pools showed that both arachidonate-containing DAG incorporation and turnover of PtdCho is increased in hCEPT1 overexpressors. Increased choline-d9 incorporation into arachidonyl lyso-PtdCho shows that both phospholipase A1- and A2-mediated turnover is involved. Spatially distinct molecular specificity of DAG incorporation into cellular PtdCho at the level of hCEPT1 exists in situ.
A molecular diversity of the CDP:choline pathway is mandated by duplication of each relevant enzyme of PtdCho (phosphatidylcholine) biosynthesis; CK (choline kinase) , CCT (CTP:choline phosphate cytidylyltransferase)  and CPT (1,2 diacylglycerol CDP:choline cholinephosphotransferase) , on the mammalian genome. For instance, CK exists in at least three active forms derived from the ck-α and ck-β genes . Similarly, CCT has two distinct gene products, designated α and β; whereas the former has no published splice variants, the latter has three . CPT activity resides on both the CPT1 and CEPT1 gene products .
Aside from CK, where subcellular locations are as yet unassigned, discrete CCT and CPT isoenzyme distributions [3–5] distinguish multiple compartments of CDP:choline pathway PtdCho biosynthesis [3,6,7]. Thus bulk membrane PtdCho is assembled in both the nuclear periphery and Golgi apparatus , with the two pathways utilizing a terminal step catalysed by either the CEPT1 gene product  for perinuclear synthesis  or the CPT1 gene product  for Golgi synthesis . Moreover, CCTα is characteristically targeted to, and found associated with, the nucleus , whereas CCTβ forms are typically excluded from the nuclear environments . In addition, quantitatively minor but spatially distinct PtdCho synthesis occurs inside the nucleus  and distal intracellular locations, such as neurone body extensions, may also support autonomous PtdCho synthesis .
Compartment-specific acyl substitution patterns are apparent between whole cell and endonuclear PtdCho synthesis [1,6,10]. However, it is unclear to what extent further distinct molecular specificity may arise from spatial segregation of bulk membrane PtdCho synthetic pathways , despite compelling evidence in vitro that the human isoforms hCEPT1 and hCPT1 possess the capacity to discriminate between DAG (diacylglycerol) molecular species .
The use of dynamic lipidomics to track the flux of deuterated lipid precursors through biosynthetic and degradative products in situ [6,10,11] can permit metabolic analyses that are impractical with conventional radiolabelling. Choline-d9 (9-deuterated choline) labelling coupled with ESI–MS (electrospray ionization mass spectrometry) to demonstrate distinct molecular specificities of PtdCho synthesis between compartments [6,10,11] has depended on a capacity to isolate intact subcellular fractions. Unlike nuclei, successful separation of sufficiently pure Golgi and perinuclear components of bulk membrane PtdCho synthesis is more problematic. Direct manipulation of individual pathway flux in situ should offer an alternative, albeit indirect, access route to the assessment of molecular selectivity within the same compartment. Accordingly, in a previously characterized Chinese-hamster ovary cell line, CHO-K1, which overexpresses hCEPT1 20-fold under control of a TET-ON promoter , we have determined changes in PtdCho and lysoPtdCho synthesis/turnover/acyl chain remodelling over 3 h.
Materials and methods
CHO-K1 TET-ON control cells and hCEPT1 transfectants between passages 14 and 22 were cultured in Dulbecco's modified Eagle's medium, containing 5% (v/v) foetal bovine serum, 200 μg/ml hygromycin B, 34 μg/ml proline and 350 μg/ml G418 (Invitrogen) at 37°C in 5% (v/v) CO2. Choline-d9, phospholipid standards, doxycyclin and other laboratory chemicals were obtained from Sigma–Aldrich. hCEPT1 overexpression was achieved by cell incubation with 2 μg/ml doxycyclin for 24 h  before choline-d9 labelling for up to 3 h .
After subsequent lipid extraction and phospholipid class purification , ESI–MS/MS of endogenous and newly synthesized PtdCho was undertaken on a Micromass Ultima Quatro ESI tandem mass spectrometer by direct injection employing precursor scans of m/z 184+ and m/z 193+  and lysoPtdCho species were incorporated into the analyses by extending the scans to encompass the mass range m/z 460–840.
Results and discussion
Analyses of the endogenous compositions of PtdCho from control and overexpressor hCEPT1 transfectants using precursors of m/z 184+ fragments showed that, 24 h after doxycyclin induction, hCEPT1-transfected CHO-K1 cell PtdCho increased by 3.76-fold the proportion with arachidonate, C20:4, n−6, present. In six arachidonate-containing molecular species their representation increased from 1.7 to 6.4% of total PtdCho (P values for each species ranged between 0.0001 and 0.05, n=36 in each group). There was no net accretion of PtdCho compared with controls, suggesting that any increase in flux through the biosynthetic pathway was compensated for by degradative increases. PtdCho arachidonate enrichment was remarkable under these conditions, not least because ESI–MS analysis of all the major phospholipids pools (results not shown) indicated that CHO-K1 cells retain little arachidonate. This was true even in the conventionally unsaturated phosphatidylethanolamine and phosphatidylinositol pools, which in CHO-K1 cells contained predominantly monounsaturated molecular species. Indeed, diversion of limited arachidonate-containing DAG pools towards PtdCho biosynthesis may be even more marked against a background of higher arachidonate content found in phospholipids of other mammalian cell types in vivo.
Incorporation of choline-d9 into control and hCEPT1 overexpressor CHO-K1 cell PtdCho
Precursors of m/z 193+ fragments after 2 h 2H labelling (Figure 1) revealed that incorporation of the phosphocholine headgroup into arachidonate and docosahexaenoate (C22:6, n−3) containing acyl species was significantly increased in the hCEPT1 overexpressors and this difference was also apparent at 30, 60 and 180 min after labelling (results not shown), suggesting a selectivity for polyunsaturated DAG incorporation conferred either by a binding preference of this enzyme or by perinuclear membrane-specific compartmentalization of polyunsaturated DAG.
ESI–MS analyses of newly synthesized PtdCho in control and hCEPT1-transfected CHO-K1 cells after 2 h incubation with choline-d9
Incorporation of choline-d9 into control and hCEPT1 overexpressor CHO-K1 cell lysoPtdCho
The lack of cellular accretion of PtdCho in hCEPT1 overexpressor cells, 24 h after induction, suggests that enhancement of cellular enzyme activity  is balanced by rapid PtdCho turnover. Moreover, differential rates of 2H incorporation into hCEPT1 overexpressor lysoPtdCho and PtdCho between 30 min and 3 h are consistent with this idea (results not shown). It can be seen from Figure 2 that the patterns of endogenous and newly synthesized lysoPtdCho at 2 h for both control and hCEPT1 overexpressing CHO-K1 cells have different characteristics, recapitulated at each time point (results not shown). In particular, the hCEPT1-transfected cells show evidence of polyunsaturated lysoPtdCho species in addition to saturated and monounsaturated, implying that the cellular response to excess PtdCho generated in the perinuclear region involves degradation/recycling by both PLA1 (phospholipase A1) and PLA2 activities.
ESI–MS analyses of endogenous and newly synthesized lysoPtdCho in control and hCEPT1-transfected CHO-K1 cells after 2 h incubation with choline-d9
Accordingly, from thorough analyses of choline-d9 incorporation patterns over time, we conclude that (i) synthesis of PtdCho de novo through hCEPT1 increases the proportion of polyunsaturated DAG incorporated; (ii) increased flux through CDP:choline pathways is balanced by enhanced degradation of excess PtdCho; and (iii) enrichment and rapid flux characteristics of 2H label through lysoPtdCho species is consistent with enhanced activity of both PLA1 and PLA2.
Research Colloquia: Research Colloquia at BioScience2004, held at SECC Glasgow, U.K., 18–22 July 2004. Edited by M. Bouvier (Montreal, Canada), G. Milligan (Glasgow, U.K.), V. O'Donnell (Cardiff, U.K.), M. Brand (MRC-Dunn Human Nutrition Unit, Cambridge, U.K.), M. Schweizer (Heriot-Watt University, Edinburgh, U.K.), R. Insall (Birmingham, U.K.), A. Ridley (Ludwig Institute for Cancer Research, London, U.K.) and M. Sutcliffe (Leicester, U.K.). The first eight papers featured in this Section were presented as a part of the GPCR Regulation and Signalling Research Colloquium, incorporating the GPCR–Ion Channel Interactions Pfizer-Sponsored Research Colloquium.