Dynamic lipidome alterations associated with human health, disease and ageing

Table of Contents

Comprehensive lipid profiling of a longitudinal cohort

From a cohort of >100 participants with IR or insulin sensitivity (IS), we previously collected longitudinal molecular data comprising genome, transcriptome, proteome, metabolome and 16S microbiome data across different timepoints (~1,000 in total¹⁷). Within this cohort, we explored various molecular signatures in health and disease and identified hundreds of molecular pathways associated with metabolic, cardiovascular and oncologic pathophysiologies^17,18. Here, we investigate the dynamics of a largely unexplored molecular layer—the ‘plasma lipidome’—and extend the longitudinal duration by 2 years to obtain a total of 1,539 samples.

To investigate lipidome alterations associated with health, disease and lifestyle changes, plasma samples from 112 participants were profiled at a median of ten timepoints across 2–9 years (average 3.2 years; one participant was sampled 163 times across 9 years; Fig. 1a, Supplementary Data 1 and 2 and Supplementary Figs. 1 and 2). Samples were collected every 3 months when the participants were healthy and with an increased frequency of three to seven collections over 3 weeks during periods of illness (for example, RVI) or notable stress, as previously reported^17,18. In addition to lipid profiling, we collected 50 clinical laboratory measurements at each sampling timepoint along with medical records (Supplementary Data 2). Finally, because samples were collected during periods of stress and illness, we also profiled 62 cytokines, chemokines and growth factors in plasma at the same timepoints.

**Fig. 1: Longitudinal lipidomics profiling.**

The human lipidome was characterized using a high-throughput quantitative lipidomics pipeline (Lipidyzer) consisting of a triple-quadrupole mass spectrometer (Sciex QTRAP 5500) in conjunction with a differential mobility separation (DMS) device^15,16. This setup allows the identification and robust quantification (estimated concentrations) of >1,000 lipid species across 16 subclasses (free FA (FFA), TAG, DAG, CE, PC, lysophosphatidylcholine (LPC), PE, alkyl ether substituent containing PE (PE-O), alkenyl ether (Plasmalogen) substituent containing PE (PE-P), lysophosphatidylethanolamine (LPE), SM, PI, CER, hexosylceramide (HCER), lactosylceramide (LCER) and dihydroceramide (DCER); Fig. 1b). In addition, we observed the differential behaviour of smaller and larger TAGs, which comprise ≤48 and ≥49 carbons across all FAs, respectively, and evaluated these separately in most analyses. For accurate quantification and to control for variance introduced during lipid extraction, we included a mix of 54 deuterated spike-in standards for nine lipid subclasses at known concentrations. Lipid species that were not present as labelled spike-in standards were normalized against the spike-in standards based on structural similarity and signal correlation (described in Methods).

We randomized the samples separately for lipid extraction and MS data acquisition. After filtering (described in Methods), we quantified, on average, 778 lipids in each sample and 846 lipid species across >1,600 samples (including quality control (QC) samples). We found the highest number (373) of lipid species in the large TAG subclass and the smallest number (4) in the DCER subclass (Fig. 1c). Lipids comprise chemically heterogeneous molecules that exert a broad spectrum of biological functions ranging from bioenergetics to cellular signalling. This is partially visible in lipid subclass-specific abundance distributions. Figure 1d shows the abundance distributions across more than four orders of magnitude and for each lipid subclass, and depicts two distinct properties: (1) the median abundance of that subclass and (2) the abundance range across all interrogated plasma samples (including healthy and disease timepoints). SMs and FFAs were observed, on average, as the most abundant subclasses, but they spanned a relatively small dynamic range. Other lipid subclasses, including LPCs, CEs and TAGs, had a lower median abundance but a much wider dynamic range.

Our study demonstrated high technical reproducibility. As anticipated, the 104 QC samples clustered distinctly (Extended Data Fig. 1); the median coefficient of variation (CV) for the QC samples was low, with values between 6.5% (small TAGs) and 20.7% (DAGs). In contrast, CVs calculated across participants and sampling timepoints ranged from 19.9% (SMs) to 91.4% (small TAGs), indicating sufficient assay reproducibility to discern biological differences. To ensure the highest robustness in our analysis, we focused on 736 lipid species for which (1) QC CVs were <20% and (2) CVs in biosamples were larger than CVs in QC samples. Except for FFAs, intraparticipant variance was consistently lower than interparticipant variance, suggesting that individual lipid signatures are distinct and stable over time (Fig. 1e). Interestingly, both small and large TAGs and ester- and ether-linked PEs (PE versus PE-O and PE-P) exhibited significant differences within their respective subclasses in terms of variance (Fig. 1e) and abundance distribution (Fig. 1d). This implies the existence of unique physiological and participant-specific differences, which may provide new insights into biological processes.

Lipid signatures are highly individualized

We first sought to investigate lipid abundance differences across individuals by characterizing the lipidome in ‘healthy’ baseline samples, defined as samples from participants in the absence of any self-reported acute disease. This does not preclude latent, asymptomatic chronic conditions such as prediabetes or potential undiagnosed conditions. Overall, we analysed 802 healthy baseline samples derived from 96 participants from whom we collected samples at two or more timepoints. The number of baseline samples per participant is shown in Supplementary Fig. 3; most participants had approximately ten healthy visits, except one outlier with 52 healthy baseline samples.

In comparison with the transcriptome, proteome and general metabolome, lipid signatures can be highly personalized when assessed longitudinally¹⁹. To investigate the participant specificity of lipid profiles for healthy sampling timepoints at timescales of months to years, we examined which lipid subclasses show the largest interindividual differences and quantified how much of the variance observed for each lipid species can be attributed to interparticipant differences (Fig. 2a). Many lipids, in particular among TAGs, SMs, HCERs and CEs, showed a high degree of participant-specific variance, in some cases >50%. In contrast, FFAs were found to have relatively low participant-specific variation. To further illustrate participant specificity, we performed t-distributed stochastic neighbour embedding (t-SNE) on data from participants with >12 healthy visits, based on 100 lipids that we determined to be the most personalized (Fig. 2b). Most, but not all, samples clustered by individual participants (Fig. 2b,c), showing that some lipids can comprise personalized signatures even across years.

**Fig. 2: Interindividual differences in healthy baseline.**

Key lipids are associated with important clinical measures

Given the high variation in specific lipid classes among individuals, we next investigated the degree to which global lipidome profiles from healthy baseline samples are associated with clinical measures. We first grouped lipids into seven modules by applying weighted gene correlation network analysis (WGCNA) based on the similarity of lipid profiles and then associated these seven modules with 50 clinical measures while controlling for the covariates sex, age, ethnicity and body mass index (BMI; Fig. 2d–f and Supplementary Fig. 4). Controlling for these covariates allows investigation of the direct associations between the lipid modules and clinical measures by ruling out potentially confounding effects from sex, age and BMI. Modules M1 and M5, which were enriched for CER and PE, as well as small TAG (mainly M1) and large TAG (mainly M5), showed the strongest positive association with T2D measures, including glycated haemoglobin (A1C), fasting blood glucose and fasting insulin. Moreover, they showed a positive association with inflammatory markers, including high-sensitivity C-reactive protein (CRP) level and white blood cell count, and a negative association with high-density lipoprotein (HDL; ‘good cholesterol’) levels. Hence, lipids in M1 and M2 have negative health associations based on conventional clinical measures. In contrast, M7, which contained some FFAs and LPCs, correlated with lower CRP and A1C levels. M3, which was enriched for PE-P and PE-O, showed an association with higher levels of HDL and lower levels of fasting insulin, and, compared with the dominant T2D patterns in M1 and M5, demonstrated a signature that is generally considered healthier.

In addition, we investigated lipid–microbiome associations and observed mostly negatively correlated lipids, including several TAG species for the bacterial family Oscillospiraceae and (L)PE, PC and CE for Clostridiaceae (Supplementary Fig. 5 and Supplementary Note 1). These microorganisms are known to be abundant in the gut of participants with IS in this cohort²⁰, suggesting a potentially beneficial role of Clostridia in host lipid metabolism. Finally, an outlier analysis identified participants with abnormally high or low lipid signatures, some of which we could correlate with underlying medical conditions such as hepatic steatosis (Supplementary Fig. 6 and Supplementary Note 2). Overall, this global analysis suggests that many lipid subclasses are associated with and potentially have a role (for example, proinflammatory, anti-inflammatory or metabolic role) in clinical conditions, or may serve as biomarkers to stratify health states.

Global lipidome disruption in IR

As many clinical measures were associated with specific lipid subclasses, we next determined how the lipidome is influenced by the chronic metabolic disorder IR. IR commonly occurs in T2D and is a condition in which cells, mainly muscle cells and adipocytes, are unresponsive to insulin, leading to high glucose levels in the blood. IR is often associated with chronic inflammation as well as metabolic syndrome, including dyslipidaemia, and can lead to non-alcoholic fatty liver disease. Elucidating how the lipid network is perturbed in individuals with IR is important to better understand the molecular mechanisms and prognosis of metabolic disorders.

IR can be diagnosed by measuring the steady-state plasma glucose (SSPG) level after endogenous insulin secretion is suppressed and insulin and glucose are infused at fixed concentrations²¹. IR or IS (IR/IS) status was measured using SSPG assays in 69 participants, of whom 36 and 33 were classified as having IR (SSPG >150 mg dl⁻¹) and IS (SSPG ≤150 mg dl⁻¹), respectively. At the global level, we observed some capacity of lipid signatures to distinguish IR and IS (Fig. 3a). Using regression analyses that controlled for age, sex, ethnicity and baseline BMI, we resolved comprehensive differences between IR and IS across most lipid subclasses (Fig. 3b–d), such that more than half of the lipids (424) were significantly associated with SSPG levels. Lipids and lipid subclasses that had a significant positive correlation with SSPG included TAGs and DAGs, which is consistent with our observations (Fig. 2d) and previous reports of higher levels of these lipids in individuals with dyslipidaemia and metabolic syndrome^22,23. We also observed subsets of CERs to have increased abundance, contributing to the development of obesity-induced IR in mice and humans²⁴ (Fig. 3b,c), and making possible the lipid-based differentiation of IR and IS (Fig. 3a and Supplementary Fig. 7).

**Fig. 3: IR- and IS-associated lipid signatures.**

To investigate over- and under-representations of specific subgroups of lipids, we performed an enrichment analysis on positive and negative model coefficients. As TAGs comprise the largest subclass of lipids in our data and could dominate the results, we performed enrichment analyses separately for each lipid subclass and across all lipids (Fig. 3e). Enrichments were evaluated at the subclass level (Fig. 1b) and for FA composition (global saturation level and specific FAs). Importantly, to our knowledge, new associations were found, including an association of ether-linked PE (PE-P)—in contrast to PE in general—with lower SSPG levels. Ether-linked PEs are involved in cell signalling and can act as antioxidants²⁵. Together with increased levels of TAGs with higher SSPG levels, reduced PE-P levels suggest IR-associated inflammation and may indicate a PE-mediated link between oxidative stress, inflammation and IR.

In Fig. 2d, we demonstrated lipid modules that correlate with a variety of clinical measures. As it is well documented that IR affects both lipid regulation (dyslipidaemia) and clinical phenotypes, we investigated whether associations between lipids and clinical measures are affected by the IR/IS status, which would have important health implications for these participants (Fig. 3f and Supplementary Fig. 8). Intriguingly, we found many significant differences in both the effect sizes and the direction of the correlation of lipid signatures and clinical measures between participants with IR and those with IS. For instance, in IR unlike in IS, the low-density lipoprotein (LDL) to HDL ratio was positively associated with the ether-linked PE-P and PE-O, and negatively associated with LPE (Fig. 3f). Moreover, in participants with IR and IS, we observed opposite correlations of immune and blood cell measurements with lipid subclasses, including A1C–SM, SSPG–CER and SSPG–PI, as well as immunoglobulin M–PE-P/PE-O, monocyte–PE-P/PE-O, eosinophil–TAG and white blood cell–PI (Fig. 3f). Overall, these data indicate that, depending on the IR/IS status, lipid–clinical measure associations can vary significantly and the key lipids involved in energy regulation, cell signalling and immune homoeostasis exhibit broad dysregulation in IR.

Dynamic lipidome alterations during viral infections

In addition to their role in chronic inflammatory and metabolic conditions such as IR, complex lipids are key mediators of acute inflammatory responses, for example, by releasing arachidonic acid (FA(20:4)). Hence, complex lipids may be modified, released and activated during RVIs and possibly vaccinations while also having important roles in these processes in an IR-dependent manner.

Participants in this cohort were densely sampled during periods of RVI (72 distinct RVI episodes in 36 participants for a total of 390 samples) and vaccination (44 episodes in 24 participants for a total of 275 visits; Supplementary Fig. 9). For both RVI and vaccination, we classified longitudinally collected samples as early-phase (days 1–6), later-phase (days 7–14) and recovery-phase (weeks 3–5) samples (Fig. 4a). Using linear models, we identified 210 lipids that were significantly changed during RVI (false discovery rate (FDR) < 10%) across most subclasses (Fig. 4b), some of which have previously been implicated in acute inflammation. For instance, PEs have been reported to have a critical role in apoptotic cell clearance and the aetiology of various viruses²⁶. Another example is PIs, which bind to the respiratory syncytial virus with high affinity, preventing virus attachment to epithelial cells²⁷. LPCs, which we observed in increased abundance during inflammation, have been demonstrated to have therapeutic effects (after intraperitoneal administration in mice) in severe infections through immune cell recruitment and modulation²⁸.

To further investigate the lipid-associated processes that are involved in acute infection, we examined enriched lipids during infection (Fig. 4c). We observed significant changes in specific lipid subclasses, including ether-linked PEs and TAGs containing saturated FAs (SFAs) such as dodecanoic acid (FA(12:0)), following RVI. Dodecanoic acid and palmitic acid (FA(16:0)) are proinflammatory compounds that upregulate cyclooxygenase 2 (ref. ²⁹) and have key roles in the activation of inflammatory responses. Overall, this suggests that different key lipid subclasses may be important for various aspects of viral biology as well as the immune response, and undergo significant changes during RVI.

To explore the choreography of lipid dynamics over time, we examined their trajectory during the different phases of RVI. The 210 significantly changed lipids were mapped to four major clusters, using a hierarchical clustering approach based on the Euclidean distance between lipid species as the similarity measure (Fig. 4d), and these main clusters were linked with clinical measures (Fig. 4e). Except for the green cluster, which was significantly enriched for PC, all profiles showed decreased levels during infection. The blue cluster was significantly enriched for small TAG and showed sharply decreased lipid levels in early RVI, with a rapid recovery that correlated with clinical measures of total lipids, including cholesterol and LDL. This indicates a metabolic shift in early infection, potentially to support increased energy metabolism. The orange cluster, enriched for LPC, large TAG and ether-linked PE, showed a similar profile to the blue cluster but a delayed recovery to baseline levels. Lipids in this cluster were positively correlated with the clinical lipid panel and blood glucose levels but negatively correlated with CRP level and neutrophils. This suggests that early changes in energy metabolism (reduction in lipid and blood glucose levels) are coupled with increased inflammation (reduction in LPC and ether-linked PE levels, as well as an increase in the CRP level and neutrophils) followed by a slow attenuation of inflammation at later stages of RVI. The purple cluster, which was enriched for FFA, represents slowly decreasing lipid levels and reached the lowest point during the RVI recovery phase before reverting to the baseline levels. In particular, the late-stage correlation with immune-related parameters (CRP level, lymphocyte count) suggests that reduction in the levels of some lipids in this cluster may relate to a temporary strong immunosuppression to attenuate early- to mid-phase inflammation and promote a return to homoeostasis. Overall, our data suggest links between differential responses of lipids and specific biological roles, with rapid shifts in energy metabolism to support inflammation early in infection and possible attenuation in later stages. Reflecting important global shifts in cell signalling, metabolism and inflammation during RVI, these lipids may allow the assessment of disease severity and prognosis or offer an opportunity for therapeutic intervention.

We next investigated whether individuals with IR and IS respond differently to infections and vaccination (Fig. 4f and Supplementary Fig. 10). Through a longitudinal differential analysis, we found distinct longitudinal profiles for IR and IS. We observed a higher abundance of several FFAs during the early stage of RVI and greater elevation of PC levels in the middle to late stage in participants with IR than in participants with IS. In contrast, TAGs and some PEs were differentially elevated in IS compared with IR throughout the middle to later stages of infection. The IR/IS-specific FFA and TAG responses may reflect the altered energy metabolism in IR, whereas differences in PCs and other lipid classes may indicate changes in immune-associated signalling pathways. Importantly, we found that the patterns after vaccination were distinct from those during infection (Supplementary Fig. 10). For example, fewer TAG species showed elevated levels in IS, whereas a distinct population of LCERs were upregulated in IR after vaccination. As individuals with T2D associated with IR often exert a more compromised immune response to RVI¹⁷, such changes may be biologically significant.

Altered ageing of participants with IR

Ageing increases the risk of cardiovascular diseases and is accompanied by a variety of diseases including T2D^30,31 and chronic inflammation³². In our study, the participants spanned an age range of 20–79 years (healthy timepoints, median 57 years) and were longitudinally sampled on average over 3 years (Fig. 5a). Across the cohort, we observed an increase in BMI with higher age (Fig. 5b). We previously identified age-associated molecular signatures in a subset of this cohort, including inflammation (acute-phase proteins), blood glucose and lipid metabolism (A1C, apolipoprotein A-IV protein), but had not investigated age-associated lipidome changes³³. To identify lipids and pathways that change with ageing and may be associated with the development of age-related pathologies, such as chronic low-grade inflammation, we investigated longitudinal changes in the lipidome. In cross-sectional studies, lipid content can differ across participants with different ages, owing to biological ageing or the period during which the cohort aged, or other cohort effects. Periods and cohorts are social contexts affecting individuals and are inherently and mathematically confounded by the individuals’ age³⁴. These comprise environmental factors differently affecting young and old participants, due to them being born in different generations, and include generation-dependent exposures that may also affect lipidome compositions (for example, diet, lifestyle and/or diseases) rather than actual age. However, we note that the longitudinal nature of our data better enabled us to eliminate some biases and focus on the same individual across time³⁴. Furthermore, we previously did not observe major dietary changes in the cohort¹⁸. To identify lipid changes that occur with ageing in our longitudinal cohort, we used a linear model that estimates relative lipid changes as a function of the change in age (Δage model) while also controlling for sample storage length and BMI. With this model, we determined the ‘ageing’ effect (β coefficient) for each lipid subclass (Fig. 5c) and across lipid species (Fig. 5d).

**Fig. 5: Age-associated changes in the lipidome.**

We found that the levels of most lipid subclasses increased with ageing, most prominently CERs (LCER, HCER, DCER), SMs, LPCs and CEs, with some of the observed variance suggesting more complex lipid–ageing dependencies (Fig. 5c). A general increase in the levels of multiple lipid species and subclasses is consistent with previous observations^35,36. Intriguingly, the levels of TAGs generally increased over time (Supplementary Fig. 11), but this trend disappeared when controlling for BMI. We performed an enrichment analysis on the Δage model coefficients at the species level and observed a shift in the physicochemical properties of lipids associated with ageing, including increased levels of SFAs and monounsaturated FAs, whereas the levels of polyunsaturated FAs (PUFAs) were reduced (Fig. 5d). This pattern has been previously associated with dyslipidaemia and inflammation³⁷, underlining progressive deterioration of metabolic health during ageing. We also observed depleted levels of (beneficial) omega-3 FAs. In particular, the levels of docosahexaenoic acid (FA(22:6), TAG) and eicosapentaenoic acid (FA(20:5), PE) decreased with ageing. These omega-3 FAs have been indicated to have beneficial health effects by lowering plasma cholesterol levels and serving as precursors for mediators that resolve inflammation, such as resolvins, protectins and maresins^38,39. In addition, decreased levels of linoleic acid (FA(18:2)) have been reported in aged skin⁴⁰; our data show that this is also a significant ageing biomarker in blood plasma, suggesting a more systemic decrease. Through desaturation and elongation, linoleic acid is metabolized to arachidonic acid (FA(20:4)), which we found to increase in abundance with increasing age when we applied less stringent filtering (Supplementary Fig. 12), further substantiating a general shift towards inflammation with ageing. Furthermore, large and small TAGs showed distinct patterns, underlining the different functional roles along the TAG spectrum. Interestingly, the levels of LPCs, which have been implicated in cardiovascular diseases and neurodegeneration⁴¹ and some of which are anti-correlated with CRP (Fig. 2), increased with ageing, further underlining their pleiotropic role in human health. We also observed a strong sex dimorphism for multiple subclasses (Fig. 5e). Beyene et al. reported sex-associated differences in lyso- and ether-phospholipid metabolism⁴², which we confirmed in our study. In addition, we observed sex-associated differences for small TAGs as a prominent signature in ageing, with higher levels in men and lower levels in women.

We next investigated the extent to which IR alters molecular ageing signatures and observed that participants with IR had larger coefficients for multiple subclasses, including HCER, LCER, SM and CE, than participants with IS. Larger coefficients indicate that ageing-related changes may be accelerated in IR versus IS (controlling for storage length, sex and BMI; Fig. 5f). In contrast to previous reports that did not distinguish IR status³⁵, our study identified a negative association between DAGs and ageing in participants with IR. Intriguingly, higher DAG levels are commonly linked to dyslipidaemia and IR^37,38; however, similar to TAGs (see above), DAGs may have a stronger association with BMI, which was controlled for in the model. Moreover, PI and PE showed opposite ageing effects in participants with IR and IS, which suggests IR-specific changes in phospholipid metabolism with ageing. In sum, the composition of many lipid subclasses (that is, degree of unsaturation, omega-3 FAs, large TAGs, ether-linked PEs) changes significantly with ageing, a process that—for some lipid subclasses—differs between the sexes and is distinctly accelerated in the presence of IR.

Specific associations of lipids with cytokines and chemokines

Given the importance of cytokines, chemokines and growth factors in diverse biological processes, we characterized their relationship to lipids across homoeostasis and various pathophysiological disease processes in our longitudinal cohort. We investigated the degree to which the abundance of a particular lipid predicts the level of cytokines, chemokines or growth factors, controlling for BMI, sex, ethnicity and multiple measurements per participant as random effects across all samples and timepoints for which both measures were available (1,180 samples). Overall, we found 1,245 significant (FDR < 5%) positive and negative associations between a majority of lipids (580) and 40 cytokines, chemokines and growth factors (Fig. 6a).

**Fig. 6: Lipid–cytokine associations.**

The largest numbers of positive associations were between granulocyte–macrophage colony-stimulating factor (GM-CSF) and TAGs and between leptin and TAGs (Fig. 6a and Supplementary Fig. 13). The adipokine leptin regulates caloric intake and is commonly present in elevated levels in obesity, contributing to the associated inflammatory state⁴³. Its amount in the blood correlates with the amount of adipose tissue. Its receptor is expressed in the hypothalamus, hippocampus and many immune cells; thus, it also acts as a neuroregulator and an immunoregulator^43,44. The cytokine GM-CSF, originally defined as a haemopoietic growth factor, has other biological roles, including exerting proinflammatory effects^45,46,47. These signatures are consistent with the inflammatory effect of the high TAG levels that we observed and are also found as a consequence of a high-fat diet, obesity and hepatic adiposity^48,49. The pleiotropic cytokine interleukin-6 (IL-6), whose inflammatory and anti-inflammatory effects are context dependent⁵⁰, together with the anti-inflammatory cytokine IL-10 (ref. ⁵¹), showed negative associations with some TAGs and clustered distinctly from the positive TAG–leptin and TAG–GM-CSF associations, suggesting functional differences among different TAG species in immunoregulatory networks (Fig. 6a). TAGs showed the overall highest number of associations with leptin and GM-CSF, whereas lipids from other subclasses, such as PE, PC and DAG, were also positively associated (Fig. 6b,c and Supplementary Fig. 13). In contrast, lyso species of PE and PC (Fig. 6d,e) showed fewer associations with and less central roles for GM-CSF and leptin. Overall, these results suggest regulatory commonalities across lipid classes (for example, positive associations of TAGs, DAGs, PCs and PEs with leptin) and differences within subclasses for proinflammatory and immunoregulatory pathways.

To elucidate the extent to which specific subsets of lipid species are associated with cytokines and chemokines, we performed an enrichment analysis (Fig. 6f). Overall, we observed strong associations of FAs with distinct cytokines. For instance, positive leptin–TAG associations were significantly enriched for SFA, the polyunsaturated FA(18:3) and small TAGs. In contrast, large TAGs were negatively associated with IL-6 and IL-10. Moreover, we observed a hub of negative associations between TAGs containing FA(22:5) and multiple cytokines, including the anti-inflammatory IL-10 and the proinflammatory IL-23, as well as IL-6. Enrichment of TAG subclasses for positive and negative associations within both proinflammatory and immunoregulatory cytokines suggests that TAG subclasses (in terms of both the length and saturation of the acyl chain) have distinct roles in immunoregulation and signalling.

In Fig. 2, we found that some LPCs were associated with anti-inflammatory, hence healthier, signatures. Here, LPCs were positively associated with several growth factors, such as epidermal growth factor (EGF), vascular endothelial growth factor (VEGF) and brain-derived neurotrophic factor (BDNF), and resistin. VEGF is involved in promoting angiogenesis, whereas BDNF and EGF promote cell proliferation, with BDNF having a cardinal role in neurogenesis and plasticity⁵². In addition, LPCs were positively associated with the soluble CD40 ligand (sCD40L), which is secreted by activated T cells and platelets during inflammation, as well as with the inflammatory cytokine IL-1⍺ and the adipose tissue-specific secretory factor resistin, which induces other cytokines and has been suggested to contribute to a chronic proinflammatory cascade in T2D⁵³. Together, LPCs demonstrated contrasting associations, including some anti-inflammatory and tissue repair as well as proinflammatory signatures. These associations may highlight a difference between chronic inflammation (for example, mediated by factors such as resistin during T2D) and acute inflammation (for example, during an infection), which is strongly associated with high CRP levels. It may also reflect that both inflammatory and anti-inflammatory mediators are present in amounts that regulate a response so that it is effective but not excessive. Moreover, PCs containing linoleic acid (FA(18:2)) were negatively associated with the chemokines CXC motif ligand 9 (CXCL9; also known as MIG (monokine induced by interferon-γ (IFNγ))) and CXCL10 (also known as IP-10 (IFNγ-induced protein 10 kDa); Fig. 6b,f). CXCL9 and CXCL10 are induced by IFNγ to recruit cells to sites of inflammation; they bind to the same chemokine receptor, CXCR3. This association suggests that these lipids may affect immune cell migration during inflammation, in addition to their immune modulation role that we observed during RVI (Fig. 4). Overall, our multiomics data outline complex associations between cytokines and lipid subclasses as well as differential associations of lipids with specific FA compositions, suggesting distinct roles ranging from immune activation to immunosuppression.

Source link

Rate this post