1
Electrospray ionisation mass spectrometry analysis of differential turnover of phosphatidylcholine by human blood leukocytes

2
Synthesis and turnover of membrane phospholipids is essential for cell growth and function, and hydrolysis of membrane phospholipid is central to many intracellular signalling mechanisms.

3
Hydrolysis of phosphatidylcholine (PC) is a major signalling mechanism of neutrophils, leukocytes that phagocytose and kill bacteria as part of the innate immune response, generating phosphatidic acid, diacylglycerol and arachidonate-derived lipid second messengers.

4
We describe here the application of tandem MS/MS electrospray ionisation mass spectrometry to the analysis of molecular patterns of PC synthesis by blood neutrophil and lymphocyte cells from healthy volunteers.

5
This technique combined incorporation of the headgroup choline, methyl-labelled with deuterium (methyl-d9-choline), with precursor scans of diagnostic labelled and native fragment ions.

6
The technique was very sensitive, permitting detection of d9 enrichment <0.01%.

7
Results showed that the two different cell types maintained distinct molecular species compositions of PC, even though they were exposed to the same nutrient supply in blood.

8
Moreover, while the pattern of lymphocyte PC synthesis directly mirrored composition, the fractional synthesis of arachidonoyl (C20∶4,n-6)-containing PC species in neutrophils was greatly enhanced compared with composition.

9
This increased turnover of arachidonoyl species in neutrophils may be related to the active synthesis of eicosanoids and other arachidonoyl-derived mediators in this cell type.

Introduction

10
Enzymatic pathways for synthesis and turnover of membrane phospholipids are well characterised, but mechanisms regulating their molecular specificity remain unclear.

11
The combinations of fatty acyl moieties esterified to the glycerol backbone define the molecular species composition of glycerophospholipids and are major determinants of their biological functions in membranes.

12
These functions include providing the physico–chemical environment for optimal membrane protein function and acting as substrates for a wide range of intracellular signalling phospholipases.1–3

13
Traditional analysis of total esterified fatty acids in such lipids by gas chromatography provides little information about such specificities of phospholipid structure and function.

14
In contrast, the application of electrospray ionisation mass spectrometry (ESI-MS) has had a dramatic impact on studies characterising membrane lipid compositions, due to its exquisite sensitivity and the detailed nature of the information provided.4,5

15
In this paper, we describe the further application of ESI-MS/MS to quantify absolute rates of phosphatidylcholine (PC) synthesis in intact human white blood cells.

16
Studies of PC synthesis by livers6,7 and lungs8 of experimental animals, using HPLC to follow incorporations of radioactive substrates, have shown that newly synthesised PC is acyl remodelled to its final membrane composition by sequential actions of phospholipases and acyltransferase enzymes.

17
Little is known about the operation of comparable synthetic mechanisms in other tissues and cell type, especially from human subjects.

18
We have previously used ESI-MS/MS to document the incorporation of (methyl-d9-choline) into endonuclear PC of cultured cancer cell lines,9 but this provides no insight into the relative synthetic rates of quiescent primary human cell types.

19
White blood cells represent a readily accessible population of primary human cells, and sub-population of lymphocyte and neutrophils cells can be obtained in reasonable yield.

20
This ability is important, as these two cell types have very different functions in vivo.

21
Lymphocytes orchestrate the humoral immune response, while neutrophils are central to the immediate innate host defence responses to microbial pathogens.

22
Hydrolysis of PC by phospholipase D is a major mechanism regulating the activation of neutrophils NADPH oxidase,10 the enzyme complex responsible for generation of bacteriocidal oxygen radicals, and maintenance of an optimal PC composition is thought to be central to this process.

23
In addition, neutrophils are major sources for the generation of prostaglandins and leukotrienes derived from arachidonic acid (C20∶4,n-6),11 which in turn is released from membrane phospholipids by phospholipase A2 activity.

24
Consequently, this study provided an opportunity to compare mechanisms of PC synthesis by two cell types derived from the same environment which can then be linked with biological function.

Experimental

Incubation of blood cells with deuteriated choline

25
Venous blood samples (10 ml) were taken from healthy volunteers using lithium heparin as anti-coagulant, in accordance with permission from the Local Research Ethics Committee.

26
Whole blood was diluted 1∶1 (v/v) with phosphate buffered saline (PBS) and incubated with deuteriated choline (methyl-d9-choline, 100 µg ml–1) (Pointe-Claire, Quebec, Canada) for 3 h at 37 °C in an atmosphere of 5% (v/v) CO2.

27
Lymphocyte and neutrophil cell fractions were prepared from diluted blood (8 ml) by sequential centrifugation steps initially over 4 ml of polymorphprep (Robbins Scientific, Solihull, UK) and then over lymphozyten separation medium (PAA laboratories, Yeovil, UK).

28
Purified cell populations were suspended finally at 2 × 107 cells ml–1 in PBS.

29
Cells from 1 ml of each fraction were centrifuged at 13 000 rpm in a microcentrifuge (Sanyo MSE Micro Centaur) for 5 s, resuspended in 100 µl PBS and fixed in 900 µl of ice-cold methanol.

Phospholipid purification

30
Dimyristoyl-phosphatidylcholine (PC 14∶0/14∶0; 15 nmol/107 cells) dissolved in 50 µl chloroform was added before extraction as an internal recovery standard.

31
Phospholipids from cell fractions were extracted with acidified chloroform and methanol.12

32
The PC fraction was then purified from the chloroform phase on 100 mg BondElut aminopropyl solid phase extraction cartridges (Phenomenex, Macclesfield, UK) and finally dissolved in methanol∶chloroform∶water∶NH4OH (7∶2∶0.2∶0.2, v∶v) at a concentration of 1 µM.

Mass spectrometry

33
ESI-MS was performed on a Micromass Quatro Ultima triple quadrupole mass spectrometer (Micromass, Wythenshaw, UK) equipped with an electrospray ionisation interface.

34
PC samples were introduced to the mass spectrometer by direct injection at a flow rate of 5 µl min–1, and cumulative spectra were obtained for 10 scans each of 12 s.

35
Capillary voltage was set at +3.5 keV and cone voltage at 100 V. Following collision-induced fragmentation with argon gas, endogenous PC molecules produced a diagnostic fragment with m/z = +184, corresponding to the protonated phosphocholine headgroup.

36
Newly synthesised PC, which contained 9 deuterons, produced an analogous fragment at m/z = +193.

37
Sequential precursor scans of the m/z 184 and 193 moieties permitted the determination of endogenous and newly synthesised PC species.9,13

38
Mass spectra were quantified after conversion to centroid format and export to an Excel spreadsheet.

39
Ion intensities for selected species were corrected first for the 13C isotope effect and second for the decreasing efficiency of MS/MS fragmentation with increasing molecular mass.

40
Because of the approximately 1% natural abundance of 13C in biological compounds, all PC species were detected as a series of ion peaks of increasing unit mass and decreasing intensity.

41
The fractional abundance of the [M + H]+ molecular ion therefore decreased with increasing molecular mass, and [M + 2H]+ ions overlapped with PC species two mass units higher.

42
For each individual PC species, correction for these effects was achieved by subtracting the calculated [M + 2H]+ ion intensity of the preceding PC species and then dividing by the fractional [M + H]+ ion intensity calculated as a proportion of all ion peaks.

43
The formula for reduced response factor with increasing m/z was calculated experimentally as follows: a = 2 × 1013b–3.9873, where a = reduced response factor relative to a value of 1.00 for PC14∶0/14∶0 and b = m/z value.

44
Identities of the major PC molecular species in each ion peak were established from fragmentation patterns of [M-15] in negative ionisation, which represents loss of one methyl group from the molecular ion.

45
This was feasible as the BondElut sample purification removed any phosphatidylethanolamine and acidic phospholipid that would have interfered with such analysis in the total lipid extract under conditions of negative ionisation.

46
Product ions due to fatty acyl substitutes were major components of these fragmentation spectra and provided a direct assessment of their combinations for each PC species.

47
The relative intensities of each product fatty acid also enabled their assignment to either the sn-1 or sn-2 positions.14

48
Consequently, the molecular species of PC identified in this paper represent in each case the major component in each respective ion peak.

49
The majority of PC molecules detected were diacyl species where both fatty acyl moeties are esterified to the glycerol backbone of the molecule.

50
A minor population of sn-1-alkyl-2-acyl PC (plasmanyl) species was also detected and characterised by exhibiting molecular ions 14 m/z units lower than the corresponding diacyl species combined with fragmentation in negative ionisation to produce solely the acyl group from the sn-2 position.

51
The resistance of these ether PC species to acid eliminated the possibility that they were sn-1-alkenyl-2-acyl species, as the vinyl bond in such plasmenyl molecules is acid labile14.

Results and discussion

52
Due to the quaternary nitrogen in the choline headgroup, PC species are readily detected by ESI-MS in positive ionisation and, as PC ions carry only a single charge, mass∶charge (m/z) is equivalent to molecular mass.

53
Any methyl-d9-choline incorporated during PC synthesis will produce an ion peak 9 m/z units higher than the corresponding endogenous PC species.

54
However, given the relatively low rate of incorporation of label over the three hour incubation, it was not possible to distinguish any incorporated material in the full positive scan from background endogenous material (Fig. 1A).

55
This figure illustrates a portion of the spectrum with PC16∶0/18∶1 (m/z 760) as the predominant PC species present, and clearly demonstrates that any newly-synthesised PC16∶0/18∶1, expected at m/z 769, could not be resolved from minor ion peaks derived from endogenous PC species.

56
Fragmentation of the ion at m/z 769 in positive ionisation generated a major diagnostic product at m/z 184, corresponding to the phosphocholine headgroup, together with a less abundant but well resolved ion at m/z 193 derived from PC containing incorporated methyl-d9-choline (Fig. 1B, upper panel).

57
Comparable fragmentation of the ion at m/z 760 produced only the single product at m/z 184 (Fig. 1B, lower panel).

58
This result indicates that while m/z 760 was comprised solely of endogenous PC, the ion at m/z 193 contained both endogenous and a smaller amount of newly-synthesised PC.

59
The ability to distinguish these product ions formed the basis of the technique for quantifying PC synthesis by comparing ion counts of precursor scans of m/z 193 with those of comparable precursor scans of m/z 184 (Fig. 1C).

60
The upper panel of Fig. 1C represents newly synthesised PC species that fragmented to give phosphoryl-(methyl-d9)-choline (m/z 193) while the lower panel indicates endogenous PC species that generated solely the m/z 184 fragment ion.

61
Although d9-enriched material was only a minor product of the ion peak at m/z 769, the pattern of PC species with incorporated methyl-d9-choline was clearly distinguishable from the endogenous material.

62
Intensities of these two spectra are normalised to the major ion peak displayed in each case as the total ion count for the precursor scan of m/z 193 was less than 1% that of m/z 184 and would consequently be negligible if displayed on the same scale.

63
Enrichment of PC with methyl-d9-choline was calculated from the ratio of the sums of the corrected ion counts of identified species in precursor scans of m/z 193 compared with m/z 184.

64
This enrichment is a reasonable estimate of the absolute rate of PC synthesis in these cells, as methyl-d9-choline was added to the medium in at least a ten-fold excess over unlabelled choline in the cells.

65
Comparison of typical PC spectra, determined as precursor scans of m/z 184, demonstrate considerable differences between purified lymphocyte and neutrophil preparations (Fig. 2).

66
While monounsaturated species were the major components for both cell types (PC16∶0/18∶1, m/z 760), lymphocyte PC was enriched in species containing arachidonate (PC16∶0/20∶4, m/z 782; PC18∶1/20∶4, m/z 808; PC18∶0/20∶4, m/z 810) and, paradoxically, in the disaturated species PC16∶0/16∶0 (m/z 734) (Fig. 2A).

67
By comparison, neutrophil PC was relatively depleted in arachidonoyl-containing species but contained increased amounts of 1-alkyl-2-acyl species characterised by an ether rather than an ester bond at the sn-1 position (e.g. PC16∶0alk/18∶1, m/z 746) (Fig. 2B).

68
Comparable analyses of PC synthesis from precursor scans of m/z 193 are illustrated in Fig. 3.

69
Fractional PC synthesis rates over the experimental period were 1.49 ± 0.16% for lymphocytes and 0.59 ± 0.21% for neutrophils (mean ± sd, n = 5).

70
Despite this low abundance compared with endogenous material, the spectra in Fig. 3 demonstrate clearly the ready quantification of such scans, with typical signal∶noise ratios > 200.

71
Comparison of these spectra suggest low rates of incorporation of methyl-d9-choline into the 1-alkyl-2-acyl plasmanylcholine in both lymphocytes (Fig. 3A) and neutrophils (Fig. 3B), but synthesis of all other PC species was well resolved at 9 m/z units higher.

72
Analysis of the combined data suggested that negligible fatty acyl remodelling mechanisms operated in PC synthesis by human blood lymphocytes (Fig. 4A), as the pattern of methyl-d9-choline was essentially the same as that of endogenous PC.

73
For neutrophils by comparison, the fractional incorporation of methyl-d9-choline into arachidonoyl species of PC (especially PC18∶0/20∶4) was considerably greater than the relative abundance of the endogenous molecular species (Fig. 4B).

74
The simplest interpretation of this observation is a rapid turnover rate of the arachidonoyl moiety in neutrophils PC.

75
This in turn may be related to the high metabolic requirement in neutrophils for arachidonate, generated by the action of phospholipase A2 on membrane phospholipid,15 for generation of eicosanoid inflammatory mediators, especially leukotriene B4.

76
The mechanisms responsible for this increased turnover of arachidonate in neutrophil PC require further investigation, but the results of this study provide a good illustration how the application of a physical technique such as ESI-MS can provide considerably greater detail and sensitivity of information than previous approaches following the incorporations of radiolabelled precursors.

77
Moreover, they also demonstrate that previous concepts of acyl remodelling in PC synthesis, derived from studies in liver and lungs,6–8 are not applicable to human blood leukocytes.

78
These studies suggested that PC was synthesised in both tissues enriched in diunsaturated species, predominately PC16∶0/18∶2, and was remodelled by subsequent actions of phospholipases and acyltransferases to, respectively, polyunsaturated or disaturated species.

79
As both liver and lungs are organs characterised by secretion of phospholipid, as lipoprotein or lung surfactant, it is possible that this pattern of acyl remodelling is characteristic of PC destined for secretion rather than for structural membrane synthesis.