Human lens lipids differ markedly from those of commonly used experimental animals

Abstract

Electrospray ionisation tandem mass spectrometry has allowed the unambiguous identification and quantification of individual lens phospholipids in human and six animal models. Using this approach ca. 100 unique phospholipids have been characterised. Parallel analysis of the same lens extracts by a novel direct-insertion electron-ionization technique found the cholesterol content of human lenses to be significantly higher (ca. 6 times) than lenses from the other animals.

The most abundant phospholipids in all the lenses examined were choline-containing phospholipids. In rat, mouse, sheep, cow, pig and chicken, these were present largely as phosphatidylcholines, in contrast 66% of the total phospholipid in Homo sapiens was sphingomyelin, with the most abundant being dihydrosphingomyelins, in particular SM(d18:0/16:0) and SM(d18:0/24:1). The abundant glycerophospholipids within human lenses were found to be predominantly phosphatidylethanolamines and phosphatidylserines with surprisingly high concentrations of ether-linked alkyl chains identified in both classes. This study is the first to identify the phospholipid class (head-group) and assign the constituent fatty acid(s) for each lipid molecule and to quantify individual lens phospholipids using internal standards. These data clearly indicate marked differences in the membrane lipid composition of the human lens compared to commonly used animal models and thus predict a significant variation in the membrane properties of human lens fibre cells compared to those of other animals.

Introduction

The main body of the lens consists of concentric layers of slender, crescent shaped fibre cells, effectively creating a tightly packed arrangement of cell membranes. The main structural components of these cell membranes are integral membrane proteins, phospholipids and cholesterol. Phospholipids are a diverse range of molecules, consisting of a “head” group and one or two fatty acid “tails”. While the “head” denotes the class of phospholipid (e.g., phosphatidylethanolamine, phosphatidylcholine etc.) different combinations of fatty acid “tails” result in a myriad of possible molecules within each class.

It has previously been observed that the lipid composition of lens fibre cell membranes varies between animal species [1], [2], [3], [4] and it has been suggested that phospholipid content may be correlated with maximum lifespan [5]. Human and primate lenses have been found to be rich in dihydrosphingomyelins, with these phospholipids comprising approximately 50% of total phospholipid in humans [6], [7]. Dihydrosphingomyelins are present as only minor constituents in non-primate lenses with the major phospholipid classes in shorter-lived animals being mostly phosphatidylcholine, sphingomyelin and phosphatidylethanolamine [2], [3], [8].

Previous investigations of lens cell membrane composition typically utilized techniques such as ³¹Phosphorus nuclear magnetic resonance (³¹P NMR) spectroscopy [2], [9], [10], [11] to analyse the phospholipid “heads”. This technique resolves phospholipids based on their head-groups alone and provides little information on the fatty acid composition. Gas chromatography has been used to examine the fatty acid “tails” [4] and this involves initial hydrolysis of the lipids followed by derivatisation. Experiments using gas chromatography have shown that human lenses contain largely palmitic acid and a monounsaturated 24-carbon chain fatty acid [12], [13], whilst in bovine and rabbit lenses a monounsaturated 18-carbon chain fatty acid predominated [14]. Thin-layer chromatography has been used in conjunction with gas chromatography in order to identify the fatty acids associated with a particular class of phospholipids in rabbit lenses [8] but such studies do not characterize the molecular structure of individual lipids.

Matrix-assisted laser desorption ionization mass spectrometry (MALDI) has also been used for the characterisation of phospholipids in porcine [15], bovine [16] and human [17] lenses. Analysis of phospholipid mixtures is difficult using MALDI as there are significant differences in the ionisation efficiencies of different classes [18] and thus only relative quantitation has been achieved [15]. In the current study the use of internal standards for each phospholipid class enabled quantitation of all molecules detected by allowing for the differences in ionisation efficiency. Experiments using MALDI deduced that there are approximately 80 different phospholipids in porcine lenses [15], however they were not capable of distinguishing between isobaric phospholipids because identification was based solely upon the mass-to-charge ratio (m/z) of the parent phospholipid [15]. MALDI provided only the total acyl chain composition of each phospholipid (e.g., 36:2) and therefore complete characterisation of individual phospholipids at a molecular level was not possible (i.e., 36:2 may be 18:1/18:1 or 18:0/18:2) [15]. By contrast a triple-quadrupole mass spectrometer, capable of neutral loss, precursor ion and fragment ion scans, such as the one used in this study, can be used to unambiguously identify the molecular composition of glycerophospholipids and sphingolipids in a tissue extract.

“Shotgun lipidomics” is a contemporary, electrospray ionisation tandem mass spectrometry (ESI-MS/MS) based method that can be used for the complete characterization of phospholipids in lipid extracts [19]. This method is capable of not only providing quantitative information on lipid classes, but also precise information on the fatty acid composition of each individual phospholipid [20], [21]. In this study ESI-MS/MS was employed to provide quantitative data on the phospholipid composition of lenses from several animals. In addition, a novel direct-insertion electron-ionisation mass spectrometry (DI/EI-MS) method was developed and implemented for the rapid quantification of free cholesterol in the same lipid extracts. Electrospray ionisation of cholesterol is generally inefficient in crude lipid extracts due to the presence of more readily ionisable phospholipids. While this limitation can be overcome by fractionation, the method presented here can be used in parallel with the electrospray analysis and does not require time-consuming chromatographic separation of the extract.

Human lenses are difficult to source reliably, so it is understandable that animal tissues are often used as models for the study of human lens disorders. Rodent lenses are the most commonly used [22], [23], [24], [25], [26], [27], [28], [29]; however investigators have also used pig [30], [31], [32], [33], sheep [34], [35], cow [36], [37], [38] and chicken [39], [40], [41]. One purpose of this study was to characterise the membrane composition of lenses from humans as well as those from commonly used experimental animals in order to discover how close the membranes of animal lenses are to humans, and in this way provide information as to the most appropriate animal model for the human lens. The current study is the first to identify the fatty acids associated with individual phospholipids in the human lens and also the lenses of commonly used experimental animals. This novel information provides valuable new insight into the biophysical and biochemical properties of the lens fibre cell membranes in these animals.