Featured Commentary

Issue No. 7; 2011

Lipid measurement is integral to global cardiovascular risk assessment. Emerging data show that profiling of lipid classes, and more recently, individual lipid species – lipidomics – offers further scope for understanding the complexity and dynamics of lipid changes, especially the case for high-risk individuals with cardiometabolic disease.

There is a logical rationale to support the use of lipidomics. Cells and plasma contain thousands of different lipid species, many of which play an integral role in modulating biological functions. As changes in individual lipid species are closer to the atherosclerotic process, it is expected that detailed lipidomic analysis will reveal more relevant information than that provided by traditional lipid measurement. Until recently, wider application of lipidomics has been hampered by cost and its time consuming nature. However, progress in mass spectrometry (MS)-based techniques and informatics has changed this, allowing researchers to profile vast numbers of lipid species in biological samples.

The LIPID MAPS Consortium has been at the forefront of these developments. In a recent study, the group has reported on the structural diversity of the plasma lipidome, based on fasting samples obtained from healthy individuals.¹ Six main categories of lipids were identified: fatty acyls, glycerolipids, glycerophospholipids, sphingolipids, sterols and prenols. Acetyl CoA is the building block of fatty acids and in turn may be transferred to form part of more complex lipids (sphingolipids, glycerolipids, glycerophospholipids and sterols), or converted to eicosanoids or isopentenyl pyrophosphate, the building block for prenol and sterol lipids. (Table 1).¹

Table 1. Main lipid species within each category for human plasma

Lipid category	Main species
Fatty acyls	Fatty acids, eicosanoids
Glycerolipids	Triacylglycerols, diacylglycerols
Glycerophospholipids	Glycerophosphocholine, glycerophosphoethanolamine
Sphingolipids	Sphingomyelins
Sterol lipids	Cholesterol, cholesteryl esters
Prenol lipids	Dolichols, coenzyme Q

Lipidomics has also been used by the same group to investigate the dynamics between different lipid categories in response to an inflammatory challenge, or statin inhibition of cholesterol synthesis. In a mouse macrophage model (RAW264.7),² activation of inflammatory mediators following challenge with ligand for toll-like receptor 4 (mimicking bacterial infection) led to differing temporal responses lipid species. Lipid responses were either immediate (as for changes in fatty acid metabolism leading to increased eicosanoid synthesis) or delayed (as for sphinolipid and sterol synthesis) (Fig. 1).


Fig 1. Temporal response in mouse macrophage lipidome after an inflammatory challenge with KLA (inflammatory polysaccharide specific for toll-like receptor 4). Adapted from Dennis EA et al (2010).²

Some of the largest changes in lipid levels were in cholesteryl ester containing saturated or monosaturated fatty acyl groups, suggesting potential for accumulation in vivo. Statin challenge also resulted in changes in lipid species as a result of lipid modelling and cross-talk between sterol and eicosanoid biosynthetic pathways. The physiological significance of these changes warrants further investigation.

The use of lipidomics may also add to our understanding of the different functionalities of HDL, a heterogeneous population of particles, varying in structure, intravascular metabolism and biological activity, building on data from proteomic analyses.³ Indeed, studies have shown that some lipid components are preferentially associated with specific types of HDL particles.^4,5 For example, prebeta-1 HDL and small, dense HDL3 are important for the efflux of cellular cholesterol and antioxidative activity; reduced content of phospholipids, depletion of sphingomylelin and a distinct apoliprotein A-I conformation preferentially favour these functions at least in HDL3.³

A lipid signature for atherosclerotic plaque?

Recent studies have investigated application of lipidomics to provide insight into differentiating atherosclerotic plaque. In one study,⁶ lipidomic profiling of human atherosclerotic plaques was used to evaluate differences between carotid endarterectomy samples obtained from symptomatic or asymptomatic patients. Lipids in these samples were extracted and MS analysis of individual lipid species was performed using shotgun lipidomics. Findings were compared with results obtained from non-diseased control samples.

Atherosclerotic plaques were characterised by:

• Enrichment with cholesteryl esters, predominantly polyunsaturated cholesteryl esters with long-chain fatty acids, relative to plasma. However, the relative content of these cholesteryl esters was lower in unstable versus stable plaque.
• Selective enrichment of certain sphingomyelins, suggesting either retention or de novo synthesis of these components in the plaque.
• An increase in phosphatidylcholines.

It is likely that changes in the profile of lipid species may reflect changes in substrate availability for mediators of inflammation, which underlie the atherosclerotic process. From a future practical perspective, systemic analysis of lipid species in atherosclerotic plaque may help in defining a lipid signature for the vulnerable atherosclerotic plaque, with the aim of a personalised approach to diagnosis and therapy.

Differentiating stable vs. vulnerable plaque?

Lipidomics may also have potential for risk stratification in unstable coronary syndromes (ACS). This is a particularly high risk patient group, with recurrent events occurring in ~10% of patients during the first year.⁷ However, available tests for identifying those at risk of events are less than ideal, costly and fail to differentiate stable vs. unstable disease. Screening tests based on key lipid species could offer additional predictive power in this setting.

In this study,⁸ electrospray-ionisation tandem MS lipidomics was used to identify individual lipid species in plasma samples from 220 subjects, either controls (n=80) or patients with stable angina (n=60) or ACS (n=80).

Different lipid species profiles were identified for stable vs. unstable coronary artery disease. One key difference was in the concentrations of certain alkyl- and alkenylphospholipids. For example the content of alkylphosphatidylcholine was 40% lower (p<0.01), and phosphatidylcholine plasmalogen, ~20% lower in samples from ACS patients than those with stable angina. As these lipids are known to be susceptible to oxidation,⁹ it is possible that this may reflect the higher oxidative stress in vulnerable versus stable plaque.

Another difference was in total phosphatidylinositol (PI) content, which was 14% lower in samples from ACS versus stable angina patients (p<0.01). PI is the primary source of arachidonic acid, required for biosynthesis of eicosanoids, including prostaglandins, which play a key role in monocyte activation and matrix metalloproteinase production, a key feature of plaque instability.¹⁰ Thus, the decrease in PI content in plasma samples from ACS patients suggests that production of arachidonic acid is linked with progression from a stable to unstable plaque.

Many of the lipids identified with unstable coronary artery disease were also associated with an increase in coronary artery disease severity.

Predictive models incorporating lipids and traditional risk factors improved the ability to differentiate stable versus unstable coronary artery disease (Fig 2).


Fig. 2 Accuracy for differentiating unstable versus stable coronary artery disease based on logistic regression analysis. Adapted from Meikle et al (2011).⁸

Overall, the results of this study suggest a possible role for plasma lipid profiling using lipidomics for earlier identification of unstable coronary artery disease.

Potential for predicting drug response?

Lastly, a recent study has highlighted the possibility of using lipidomics for identifying those patients most likely to benefit from certain types of drug therapy.

The Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study¹¹ was a major prospective study investigating fenofibrate in type 2 diabetes patients. While the study did not show a significant benefit in terms of the primary outcome (major coronary events), there was significant reduction in total cardiovascular events, largely attributable to reduction in revascularisation and nonfatal MI. Interestingly, changes in HDL cholesterol and apolipoprotein A-I were less than anticipated (increases of ~2% and ~4% versus placebo). Substudies showed that fenofibrate induced changes in the distribution of HDL subspecies,¹² as well as an increase in homocysteine.¹³ The clinical relevance of the latter effect remains uncertain.

In a subsequent study, lipidomic analysis was used to investigate the effect of fenofibrate on molecular lipid profiles of HDL particles in simulation models, and whether this differed in individuals with a high versus low homocysteine response.¹⁴

The analysis showed that fenofibrate treatment was associated with a number of compositional changes in HDL. These included an increase in apolipoprotein II, upregulation of sphingomyelins and a decrease in lysophosphatidylcholine content, the latter especially in patients in whom fenofibrate induced high homocysteine levels.

It is interesting that there was an increase in levels of ethanolamine plasmalogen and cholesterol ester in fenofibrate-treated patients with low homocysteine levels but not in those with high homocysteine levels. As plasmalogen species function as antioxidants,¹⁵ thereby preventing oxidation of cholesterol and phospholipids, it is possible that in fenofibrate-treated patients with high homocysteine levels, HDL may have reduced antioxidant ability.

While it is acknowledged that simulation models may not necessarily represent the structure of HDL in vivo, these results suggest that molecular profiling of HDL using lipidomic approaches might offer new surrogates for predicting treatment response to fenofibrate.

Ultimately such approaches may have application in personalised approaches to dyslipidemia management.