Oleic acid triggers metabolic rewiring of T cells poising them for T helper 9 differentiation

Summary T cells are the most common immune cells in atherosclerotic plaques, and the function of T cells can be altered by fatty acids. Here, we show that pre-exposure of CD4+ T cells to oleic acid, an abundant fatty acid linked to cardiovascular events, upregulates core metabolic pathways and promotes differentiation into interleukin-9 (IL-9)-producing cells upon activation. RNA sequencing of non-activated T cells reveals that oleic acid upregulates genes encoding key enzymes responsible for cholesterol and fatty acid biosynthesis. Transcription footprint analysis links these expression changes to the differentiation toward TH9 cells, a pro-atherogenic subset. Spectral flow cytometry shows that pre-exposure to oleic acid results in a skew toward IL-9+-producing T cells upon activation. Importantly, pharmacological inhibition of either cholesterol or fatty acid biosynthesis abolishes this effect, suggesting a beneficial role for statins beyond cholesterol lowering. Taken together, oleic acid may affect inflammatory diseases like atherosclerosis by rewiring T cell metabolism.


Transcriptomic analysis of oleic acid-exposed non-activated CD4 + T cells
In order to identify the molecular features that define the effect of oleic acid exposure on non-activated CD4 + T cells in vitro, we exposed nonactivated CD4 + T cells to 30 mg/mL oleic acid for 0.5, 3, 24, 48, or 72 h (n = 9; Figure 1A).First, we measured CPT1A expression and found that its expression consistently increased over time indicating a robust response to oleic acid exposure across donors, while CPT1A expression did not change under control conditions (Figures 1B and S2B).Next, we analyzed the transcriptome of non-activated CD4 + T cells after oleic acid exposure using RNA-seq.Oleic acid induced differential expression of 544 genes (P FDR < 0.05) that clustered into 310 upregulated genes and 234 downregulated genes (Figures 1C and S3, and Tables S1A and S1B).There was no statistical evidence for further subdivisions of the two clusters, for example, in fast-and slow-responding genes.
We first examined the functions of the 310 genes that were upregulated in non-activated CD4 + T cells by oleic acid exposure.We inspected the top differentially expressed genes (Figure S3A and Table S1A).The top differentially expressed gene was CPT1A highlighting the involvement of b-fatty acid oxidation.In addition, we found an increased expression of HMGCR (3-hydroxy-3-methyl-glutaryl-coenzyme A [CoA] reductase), encoding the rate-limiting enzyme for cholesterol biosynthesis, and ACACA (acetyl-coenzyme A carboxylase 1), encoding the rate-limiting enzyme of fatty acid biosynthesis.Furthermore, transcripts of several aerobic glycolysis-related genes, such as TKT and PGD, were upregulated (Figure S3A and Table S1A).A formal analysis of enriched biological processes among all 310 upregulated genes confirmed the involvement of metabolism.In particular, cholesterol biosynthesis (P FDR < 0.001), homeostasis (P FDR < 0.001), and signaling of mTORC1 (P FDR < 0.001), a key complex of mechanistic target of rapamycin (mTOR) which aids in the switch toward aerobic glycolysis and fatty acid biosynthesis, were enriched (Figure 2A).Mapping the upregulated genes to canonical metabolic pathways further supported a specific metabolic rewiring of oleic acid-exposed non-activated CD4 + T cells (Figure 2B).First, oleic acid can first be catabolized through beta oxidation to produce acetyl-CoA, which can then be used as a starting point for cholesterol and fatty acid biosynthesis.In addition to CPT1A, we found 4 out of 15 enzymes in b-fatty acid oxidation (including SLC25A20, ACADVL, and ACAA2) and 2 out of 15 enzymes in the aerobic glycolysis pathway to be upregulated (TKT and PGD; Figures 2B, S4, and S5).Remarkably, on top of HMGCR, 15 out of 20 enzymes involved in cholesterol biosynthesis were upregulated in our gene set, including several key rate-limiting genes (such as HMGCS1, SQLE, MVD, and MVK).More specifically, 9/11 components of the mevalonate, 6/9 of the Bloch, and 6/9 of the Kandutsch-Russell pathway, together responsible for cholesterol biosynthesis, were upregulated (Figure 2B and S6).The upregulated gene set also included ACACA and FASN that encode the two enzymes that together are responsible for the 37 reactions making up fatty acid biosynthesis (Figures 2B and S7).Of note, the genes ACACA and FASN have been implicated in the differentiation toward T H 17 cells, a highly pro-inflammatory subset of CD4 + T cells. 46Furthermore, aerobic glycolysis and cholesterol and fatty acid biosynthesis are the hallmark metabolic processes of activated T cells and suggest that non-activated CD4 + T cells undergo a metabolic reprogramming upon oleic acid exposure that may poise the cells for a different response to activation.
We then examined the functions of the 234 genes that were downregulated in non-activated CD4 + T cells by oleic acid exposure.We first inspected the top differentially expressed genes (Figure S3B and Table S1B).Among the top downregulated genes, decreased expression of CXCR6 and CCR5, important chemokine receptors in the T cell immune response, was measured.Moreover, expression of TPM4, encoding actin-binding proteins involved in the cytoskeleton, and DMTN, encoding an actin-binding and bundling protein that stabilizes the actin cytoskeleton, was also downregulated.A formal analysis of the enriched biological processes among all 234 downregulated genes revealed a wide variety of different pathways.In line with the genes observed among the top downregulated genes, this included processes involved in immune response (CCR2, CCR8, HLA-DRA, SLC2A1) (P FDR < 0.001) and actin cytoskeleton organization (ACTB, RAC2, ARPC2, IQGAP1) (P FDR < 0.001) (Figure 2C).In addition, processes involved in chemotaxis (P FDR < 0.001), chemokine and cytokine signaling (P FDR < 0.001), and Rho GTPase regulation (P FDR < 0.001) were also downregulated (Figure 2C).Overall, these data point to a broad yet specific downregulation of genes in oleic acid-exposed non-activated CD4 + T cells, perhaps to cope with the influx of the fatty acid.
Next, we investigated whether specific transcription factors may underlie the differential expression observed by testing the enrichment of transcription factor binding motifs in upregulated vs. downregulated genes.The top motifs enriched among upregulated genes included key transcription factors PU.1, EGR1, BHLHE40, and SREBP1 (Figure 2D).Notably, PU.1 is the key transcription factor for the development of T H 9 cells.BHLHE40 has been linked to T H 17 development and pathogenicity in autoimmune encephalomyelitis suggesting an additional possible preference toward T H 17 differentiation post-activation. 48,49Furthermore, EGR1 and SREBP1 are involved in either the activation of Tbet or    fatty acid and cholesterol biosynthesis, respectively. 50,51These data further support the notion that oleic acid-exposed non-activated CD4 + T cells may be poised to differentiate toward T H 9 and T H 17 T cell subsets after activation.

Oleic acid induced CD4 + T cell phenotypes after activation
To determine the functional impact of the transcriptomic changes identified, we characterized the phenotypes of CD4 + T cells that were preexposed to oleic acid or control conditions and subsequently activated in the absence of oleic acid.To this end, non-activated CD4 + T cells of 8 out of 9 donors, for whom sufficient cells were available, were again exposed to 30 mg/mL oleic acid (Figure 3A).The effect of exposure was confirmed by an upregulation of CPT1A (Figure S8A); cell viability was high (>90%), and there was no difference in diameter between cells exposed to oleic acid and control (Figures S8B-S8E).First, we examined phenotypes after oleic acid exposure without activation (Figure S9 and Table S1C).We observed decreased frequencies of CD127 low CD25 hi FoxP3 + and CD27 + CD4 + T cells in response to oleic acid pre-exposure (P FDR < 0.05; Figure 3B).In non-activated cells, the CD127 low CD25 hi FoxP3 + population is representative of T reg cells, and thus the decreased frequencies in the non-activated cells are in line with the lower FOXP3 expression observed in the RNA-seq analysis.Increased frequencies of interleukin (IL)-5 + cells were also observed (P FDR < 0.05; Figure 3C) These data suggest that the oleic acid-induced changes in gene expression are reflected in consistent functional characteristics of the CD4 + T cells without activation.
Activation of the CD4 + T cells led to an increased cell size irrespective of pre-exposure to oleic acid (Figure 3D).In contrast, the expression of surface and intracellular markers was influenced by exposure to oleic acid prior to activation (Figure S10 and Table S1D).Pre-exposure to oleic acid resulted in a higher proportion of IL-9 + cells (P FDR < 0.01) as compared to the control (Figure 3E).Additional analysis showed that IL-9 was not co-expressed with other T H 2-associated cytokines (Table S1E).This aligns with our finding that a large percentage of upregulated genes mapped to a PU.1 motif (Figure 2D), the key transcription factor controlling T H 9 differentiation.Furthermore, increased frequencies of IL-17A + cells were observed after pre-exposure to oleic acid as compared with control conditions (P FDR < 0.05).As IL-17A is mainly produced by T H 17 cells, it was hypothesized that other T H 17-associated cytokines, such as IL-21, may also have been upregulated.Indeed, IL-21 + cells were increased in frequency (p < 0.05), but this effect was no longer significant after correction for multiple testing (P FDR < 0.08).This aligns with our finding that a large percentage of upregulated genes mapped to the BHLHE40 motif (Figure 2D) involved in T H 17 differentiation. 48,49ctivated CD4 + T cells showed increased frequencies of CD127 low CD25 hi FoxP3 + and GATA3 + and decreased frequencies of CD27 + and CD38 + cells in response to oleic acid pre-exposure (P FDR < 0.05; Figure 3F).However, FoxP3 can be expressed on activated conventional T cells without a suppressor function; 52 therefore, we are unable to differentiate whether the increased proportion of CD127 low CD25 hi FoxP3 + cells post-activation is due to increased differentiation toward T reg or an artifact of T cell activation.GATA3 is the key transcription factor involved in T H 2 differentiation, and, as such, frequencies of T H 2-related cytokines IL-5 + and IL-13 + were increased (P FDR < 0.05; Figure 3E).Finally, we observed that the effect of oleic acid on differentiation is not secondary to a differential proliferative capacity (p > 0.92; Figure S11).Together, these data indicate that the metabolic changes in non-activated CD4 + cells upon oleic acid exposure skew the cells toward producing more cytokines characteristic of T H 9, T H 17, and T H 2 subsets upon activation.
In order to reinforce our findings, we repeated the spectral cytometry analysis with 8 independent donors.The effect of oleic acid exposure was confirmed by an upregulation of CPT1A (Figure S12A).Cell viability was high (>78%), and there was no difference in diameter between cells exposed to oleic acid and control (Figure S12B-S12E).Without activation, the phenotypes of oleic acid-exposed CD4 + T cells showed increased frequencies of both IL-17A + (P FDR < 0.05) and TNFa + cells (P FDR < 0.05; Figures S13A, S13B, and S14; Table S1F).After activation, the phenotypes of oleic acid-exposed CD4 + T cells showed an increased frequency of IL-9 + (P FDR < 0.05) and GATA3 + (P FDR < 0.05) cells as well as decreased frequencies of CD38 + cells (P FDR < 0.05; Figures S13C, S13D, and S15, and Table S1G).These findings in non-activated and activated cells confirm results of our experiment and substantiate that oleic acid exposure in non-activated CD4 + cells poised the cells toward producing more cytokines representative of T H 9 cells post-activation.

Oleic acid induced CD4 + T cell phenotypes blocked by metabolic inhibitors
We next determined whether induction of this profile, reminiscent of an increase differentiation toward T H 9, T H 17, and T H 2 subsets, was dependent on an upregulation of cholesterol and fatty acid biosynthesis in line with our RNA-seq data.We inhibited cholesterol synthesis with atorvastatin, targeting 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase (HMGCR), and fatty acid synthesis with CP-640186, targeting both ACC1 and ACC2 (ACACA and ACACB).To this end, non-activated CD4 + T cells of 3 out of 8 donors, for whom sufficient cells were available, were again exposed to control conditions, oleic acid only, oleic acid +10 mM atorvastatin, oleic acid +20 mM CP-640186, or oleic acid and both atorvastatin and CP-640186 for 48 h.The effect of oleic acid exposure was confirmed by an upregulation of CPT1A (Figure S16A).Cell viability was high (>88%), and there was no difference in diameter between cells exposed to control, oleic acid, or oleic acid + inhibitors (Figure S16B-S16E).
Subsequently, both oleic acid and the inhibitors were washed away and the pre-exposed CD4 + T cells were activated.We evaluated the expression of one key marker for each subset: IL-9 for T H 9, IL-17A for T H 17, and IL-13 for T H 2 cells (Figures 4A and S17).Remarkably, the ability of oleic acid to increase frequencies of IL-9 + cells was inhibited by both atorvastatin and CP-640186 (Figure 4B and Table S1H).Although similar trends were observed for frequencies of IL-17A + and IL-13 + cells, these effects were not statistically significant (Figure 4B).These data indicate that oleic acid promotes the differentiation to in particular IL-9 + -producing T cells via upregulation of cholesterol and fatty acid biosynthesis.

DISCUSSION
T cells are known to respond to fatty acids. 2 Using an in vitro model, we show that sub-physiological concentrations of oleic acid can already influence CD4 + T cells when in a non-activated state by upregulating the expression of genes that encode enzymes involved in core metabolic pathways responsible for cholesterol biosynthesis, fatty acid biosynthesis, and aerobic glycolysis.These metabolic processes are hallmarks of activated T cells. 53Indeed, upon activation, CD4 + T cells pre-exposed to oleic acid are characterized by increased production of cytokines, including IL-9, IL-17A, IL-5, and IL-13, indicative of a preferential differentiation toward the pro-inflammatory T helper subsets T H 9 as well as T H 17 and T H 2, which can have both pro-and anti-inflammatory effects.Interestingly, this effect is abolished in particular for IL-9 + -producing cells by blocking the cholesterol or the fatty acid biosynthesis pathways during the initial exposure to oleic acid.Our findings imply that increased fatty acid levels in the circulation can rewire the metabolism of non-activated T cells and poise them to particularly differentiate toward T H 9 cells, for example, when the cells infiltrate diseased tissues, including atherosclerotic plaques, and become activated.
Our results showed that cholesterol biosynthesis was the primary transcriptionally upregulated pathway in oleic acid-exposed non-activated CD4 + T cells (15 out of 20 genes).This upregulation is of particular interest because of this pathway's role in producing the necessary metabolites required for T cell activation. 54Cholesterol biosynthesis is upregulated in activated T cells to support membrane production, cell signaling through the formation of lipid rafts, and prenylation of signaling proteins. 55Additionally, intracellular cholesterol sensing has also been found to play a role in T cell differentiation, particularly toward pro-inflammatory subsets.For example, sterols were found to bind the T H 17 transcription factor RORgt and could promote its activity. 56Thus, the upregulation of gene expression in the cholesterol biosynthesis pathway due to oleic acid exposure may be indicative of a metabolic reprogramming of the non-activated CD4 + T cells toward an activated state and may lead to the differentiation toward pro-inflammatory subsets post-activation.
Additionally, expression of the two genes comprising the de novo fatty acid biosynthesis pathway was upregulated (ACACA and FASN).Together, cholesterol and fatty acid biosynthesis comprise part of the process known as lipogenesis, the synthesis of novel lipids in a cell.Lipogenesis is induced by the activation of the transcription factor SREBP1, which was associated with the upregulated transcripts in our RNA-seq data.Enrichment analysis of our transcripts also revealed upregulated genes in mTORC1 signaling, which is known to induce the activation of SREBP1. 57Although this effect is usually insulin dependent, obesity and overfeeding have been shown to hyperactivate mTORC1. 58Thus, it is possible that oleic acid alone could induce the activation of mTORC1, which in turn activates SREBP1, leading to lipogenesis and expression of cholesterol and fatty acid biosynthesis-related genes.
Fatty acid biosynthesis has also been related to the development of T H 17 cells. 17,32Specifically, the mRNA expression of genes ACACA, encoding for acetyl-CoA carboxylase 1 (ACC1), and FASN, encoding fatty acid synthase, was increased in our dataset.These genes are key determinants in the development of the pro-inflammatory subset T H 17 cells over the anti-inflammatory subset T reg cells. 22,31,39,46,59Correspondingly, FOXP3, the key transcription factor of T reg cells, was downregulated in oleic acid-exposed non-activated CD4 + T cells.Upregulated transcripts were found to be associated with the transcription factor PU.1.PU.1 is the key transcription factor in the development of T H 9 cells.This subset is a highly pro-inflammatory subset related to T H 2 cells. 60This further supports the idea that oleic acid exposure leads to a cellular metabolic reprogramming that could promote the development of pro-inflammatory T cell subsets, specifically T H 9, and possibly also T H 17 and T H 2 cells.These results indicate that oleic acid-exposed non-activated CD4 + T cells were upregulating genes involved in metabolism to initiate/prepare for the selective differentiation into T H 9/T H 17/T H 2 cells post-activation.Moreover, the metabolic processes being enhanced due to oleic acid exposure hint that the cells may preferentially differentiate toward T H 9, T H 17, and T H 2 cells upon activation.
Importantly, we provide evidence that the oleic acid-induced metabolic rewiring underpins the observed enhanced T H 9, T H 17, and T H 2 differentiation as exposing non-activated CD4 + T cells to oleic acid in combination with cholesterol or fatty acid synthesis inhibitors decreased the frequencies of IL-9 + , IL-17A + , and IL-13 + cells.While the role of T H 17 and T H 2 cells in atherosclerosis has not been resolved, these cell types have been identified as pro-inflammatory in other diseases such as autoimmune encephalomyelitis and allergy, respectively. 35,613][64] Additionally, statins have been hypothesized to have protective effects independent of cholesterol reduction; 65 our study hints that effect of statins on T cell responses could contribute to this protective role.
Immune-lipid interactions occur in the circulation, which is a complex environment comprising many factors that can affect T cell function prior to their recruitment to disease site like the atherosclerotic plaque. 66Fatty acids are a significant component of this environment and have been found to exert their effect not only on atherosclerosis but also on T cell function. 2Our model was designed to determine the effect of oleic acid exposure on non-activated CD4 + T cells.Here, we focus solely on the interaction between oleic acid and CD4 + T cells and thus make no claim to what effects this fatty acid might have in relation to atherosclerotic cardiovascular disease as a component of more complex lipids, like olive oil.[39][40][41][42][43] Circulating levels of oleic acid have been found to be related to pro-atherogenic effects, 37,38 and oleic acid is one of the most abundant fatty acids in the human circulation. 36However, this does not preclude any effects in vivo or of other types of fatty acids on non-activated T cells.
Taken together, our results suggest that oleic acid can rewire the metabolism of non-activated CD4 + T cells, as they exist in the circulation.This metabolic rewiring induces a preferential differentiation in particular toward T H 9 cell types following activation.Since T H 9 cell have proatherogenic effects [62][63][64] and we show that the oleic acid-induced differentiation into T H 9 cells can be inhibited by statins, our study indicates a new route by which fatty acids can contribute to atherosclerosis through modifiable effects on the immune system.

Limitations of the study
Although our experiments show that non-activated CD4 + T cells exposed to oleic acid undergo distinct changes in the expression of genes encoding key enzymes constituting core metabolic pathways, and that subsequent activation of pre-exposed cells results in a differentiation that is skewed toward IL-9 + -producing T cells, our study used an in vitro model to establish these relationships and lacked an in-depth functional and mechanistic characterization of the metabolic changes involved.First, studies in vivo will be required to determine the relevance of our findings to the etiology of inflammatory diseases including atherosclerosis.Second, additional functional support for the occurrence of metabolic rewiring by oleic acid as implied by our results will be important.However, it will be challenging to assay functional effects.The T cells exposed to oleic acid were in a non-activated state and hence are unlikely to display functional differences in cell metabolism.Metabolic pathways are involved in the differentiation of CD4 + T cells into specific subsets, and functional metabolic differences in T cells generally emerge only post-activation.Cell-subtype-specific and single-cell approaches can be informative to overcome the limitations of the bulk sequencing and spectrometry experiments as we performed in this study, 18 including flow cytometry-based methods to functionally profile energy metabolism, 67 mass spectrometry, and proteomics.Nevertheless, pharmacological inhibition of fatty acid and cholesterol metabolism in non-activated T cells abolished the oleic acid-induced skew toward IL-9 + -producing T cells upon activation, supporting our overall interpretation that metabolism is mechanistically involved in the effects we observed.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:
Next, non-activated CD4 + T cells were cultured with or without oleic acid for 0.5, 3, 24, 48, or 72 hours at 37 C under 5% CO 2 .To this end, CD4 + T cells from each donor were plated in a 24 wells plate (density of 4*10 6 cells/well) in 2mL 5% FCS DMEM for each time point, one exposed to oleic acid, one to the solvent control, and one to the negative control.Cells were cultured in medium containing FCS to ensure cell viability during culture and to be more comparable to physiological conditions of the circulation where other lipids are also present.Oleic acid (Sigma-Aldrich, O1383) was dissolved in HPLC grade ethanol (Thermo Fisher Scientific, 64-17-5) to a final concentration of 30,000mg/mL and complexed to fatty acid-free (FAF) bovine serum albumin (BSA) (Sigma-Aldrich, A7030) in a 2% FAF BSA DMEM mixture (Dulbecco's Modified Eagle's Serum, 2% FAF BSA, 1% Pen-Strep, 1% GlutaMAX-1 (100x)) to a final concentration of 150mg/mL.Complexing oleic acid mimics physiological conditions as fatty acids are also bound to albumin in the human circulation. 68Oleic acid was further diluted to the final concentrations of 10, 20, 30, and 50mg/mL.[42][43][44][45][46] For the solvent control samples, HPLC grade ethanol was diluted in 2% FAF BSA DMEM in the same volume as to dilute oleic acid to 150mg/mL and added to the wells.For the negative control samples, 2% FAF BSA DMEM was added directly to the wells with no additional solvent.The amount of 2% FAF BSA DMEM added to the wells was equal for each condition to keep the volumes equivalent.To assess the additional oleic acid stimulus to the non-activated CD4 + T cells due to FCS in the culture medium, an FCS sample was measured via the Shotgun Lipidomics Assistant (SLA) method 69 to estimate the fraction of oleic acid in the sample.The sample was prepped as previously described 70 but with two modifications, a starting volume of 25mL FCS and 600mL MTBE was added instead of 575mL during the first extraction.After exposure, the cells were flash frozen in liquid nitrogen and stored at -80 C until further use.Cell viability was measured via trypan blue staining (Sigma-Aldrich, T8154).

Spectral cytometry cell prep and activation
To study the effect of oleic acid pre-exposure on CD4 + T cell subset development, cells from 8 out of 9 donors that were previously analyzed using RNA-seq were thawed from liquid nitrogen; 1 donor could not be studied further because too few cells were available.Cells were cultured overnight to allow the cells to return to a resting state after the stress of the thawing, in T75 flasks at a density of 2.5*10 6 cells/mL in 5% FCS DMEM medium supplemented with 50 IU/mL IL-2 at 37 C under 5% CO 2 .To keep the cells in a non-activated state, no additional stimulus was added.Following overnight incubation, the cells were divided into 2 conditions, oleic acid and solvent exposed, and plated in a 24 wells plate (density of 4*10 6 cells/well) in 2mL 5% FCS DMEM.The oleic acid and solvent solution were prepared as stated previously, with one modification.To ensure that there was no effect of the solvent on T cell differentiation, the HPLC grade EtOH was evaporated before dissolving the oleic acid in 2% FAF BSA DMEM medium.The HPLC grade EtOH was also evaporated before adding the 2% FAF BSA DMEM medium in the solvent exposed condition, rendering it essentially the same as the negative control.These solutions were each added to the respective wells, where the final concentration of the oleic acid exposed conditions equaled 30mg/mL.The CD4 + T cells were cultured for 48h at 37 C under 5% CO 2 .
To ensure that the effect on CD4 + T cell differentiation was due to oleic acid pre-exposure, all medium of each condition was replaced by 5% FCS medium after 48h of exposure, before initiating the activation.Cell viability and diameter were first measured by Via1-Cassetteä (Chemometec, 941-0012) on a NucleoCounterâ NC-200ä (Chemometec, 900-0200) and found to be > 90% for each condition.Then, 2 million cells were harvested by flash freezing in liquid nitrogen for in vitro model confirmation by RT-qPCR.The remaining cells were plated in a round bottom 96 wells plate (Corning Incorporated, 3799), at a density of 100,000 cells/well, and were activated for 72h using Dynabeadsä Human T-Activator CD3/CD28 for T Cell Expansion and Activation (Thermo Fisher Scientific, 11161D; RRID:AB_2916088) according to the manufacturer's instructions at 37 C under 5% CO 2 .Half the cells from each exposure were activated and the other half was left in the non-activated state.Subsequently, the cells from each pre-exposure and activation state were pooled in Eppendorf tubes and the beads were magnetically removed from the activated cells.Cell viability and diameter were measured by Via1-Cassetteä after 72h.Cells were then used for T cell subset identification described in more detail below.All centrifugation steps were performed at 1500 rpm at room temperature.

Inhibitor culture conditions and activation
To study whether the effect of oleic acid pre-exposure on CD4 + T cell subset development could be prevented by metabolic inhibitors, cells from 3 out of 8 donors that were previously analyzed for subset development were thawed from liquid nitrogen; 5 donors could not be studied further because too few cells were available.Cells were cultured overnight to allow the cells to return to a resting state after the stress of the thawing, in T75 flasks at a density of 2.5*10 6 cells/mL in 5% FCS DMEM medium supplemented with 50 IU/mL IL-2 at 37 C under 5% CO 2 .To keep the cells in a non-activated state, no additional stimulus was added.Following overnight incubation, the cells were divided into 5 conditions, solvent, oleic acid, oleic acid + atorvastatin (Sigma-Aldrich, PHR1422), oleic acid + CP-640186 (Sanbio, 17691-5), and oleic acid + atorvastatin + CP-640186 exposed, and plated in a 24 wells plate (density of 4*10 6 cells/well) in 2mL 5% FCS DMEM.The oleic acid and solvent solution were prepared as stated previously, with HPLC grade EtOH evaporation.These solutions were each added to the respective wells, where the final concentration of the oleic acid exposed conditions equaled 30mg/mL.Atorvastatin and CP-640186 were added to the respective wells at a concentration of 10mM and 20mM, respectively.The CD4 + T cells were cultured for 48h at 37 C under 5% CO 2 .
To ensure that the effect on CD4 + T cell differentiation was due to oleic acid and inhibitor pre-exposure, all medium of each condition was replaced by 5% FCS medium after 48h of exposure, before initiating the activation.Cell viability and diameter were first measured by Via1-Cassetteä on a NucleoCounterâ NC-200ä and found to be > 90% for each condition.Then, 0.5-1.5 million cells were harvested by flash freezing in liquid nitrogen for in vitro model confirmation by RT-qPCR.The remaining cells were plated in a round bottom 96 wells plate, at a density of 100,000 cells/well, and were activated for 72h using Dynabeadsä Human T-Activator CD3/CD28 for T Cell Expansion and Activation according to the manufacturer's instructions at 37 C under 5% CO 2 .Subsequently, the cells from each pre-exposure were pooled in Eppendorf tubes and the beads were magnetically removed.Cell viability and diameter were measured by Via1-Cassetteä after 72h.Cells were then used for T cell subset identification described in more detail below.All centrifugation steps were performed at 1500 rpm at room temperature.

RNA isolation
To isolate total RNA for RNA sequencing and RT-qPCR, RNA was extracted from the cell samples using the Zymo Quick-DNA/RNA Microprep Plus Kit (Zymo Research, D7005) according to manufacturer's instructions.The RNA was quantified using a Qubitâ 2.0 Fluorometer (Q32866) with the Qubitâ RNA BR Assay Kit (Thermo Fisher Scientific, Q10211) according to manufacturer's instructions.RNA integrity (RIN) values of the samples were on average 8.40 SE 0.14 as determined using an Agilent 2100 Bioanalyzer Instrument (G2939BA) with the Agilent RNA 6000 Nano Reagents (5067-1511).RNA was divided into two samples and stored at -80 C, 1mg for RNA sequencing and the rest for cDNA synthesis and RT-qPCR measurements.

Real time-Quantitative PCR
To measure the expression of CPT1A in all the cell samples, cDNA was synthesized with 200ng of the stored RNA using the Transcriptor First Strand cDNA Synthesis Kit (Roche, 04897030001) according to the manufacturer's instructions.Quantitative real time PCR's for CPT1A (Thermo Fisher Scientific, 4331182; Assay ID: Hs00912671_m1) were performed using the TaqManä Fast Advanced Master Mix (Thermo Fisher Scientific, 4444557) with 10ng cDNA per reaction on a QuantStudio 6 Real-Time PCR system (Applied Biosystems).All RT-qPCR reactions were performed in triplicate and outliers were removed if the Ct value measured differed more than 0.5% from the mean.Relative gene expression levels (-DCt) were calculated using the average of Ct values of RPL13A (Thermo Fisher Scientific, 4448892; Assay ID: Hs03043887_gH) and SDHA (Thermo Fisher Scientific, 4453320; Assay ID: Hs00188166_m1) as internal controls. 71The fold change was determined using the 2 -DDCt method, using the negative control as the reference.All statistical analyses were performed in R. Data are expressed as mean of the relative fold change and standard error.The reported P values were determined by applying a paired two-tailed student's T test.P values < 0.05 were considered to be statistically significant.

RNA sequencing
RNA sequencing (RNA-seq) was performed to determine the differences in the transcriptome of oleic acid versus solvent exposed non-activated CD4 + T cells across time.1mg of total RNA from each of the samples was sent for sequencing (Macrogen, Amsterdam, NL), each with a concentration above 20ng/mL in 50mL solution.RNA-seq libraries were prepared from 200ng RNA using the Illumina Truseq stranded mRNA library prep (Illumina, 20020595) after depletion of ribosomal RNA with Ribo Zero Gold (Illumina, 20037135).Both whole-transcriptome amplification and sequencing library preparations were performed in two 96-well plates with half the samples each, to reduce assay-to-assay variability.Quality control steps were included to determine total RNA quality and quantity, the optimal number of PCR preamplification cycles, and fragment size selection.No samples were eliminated from further downstream steps.Barcoded libraries were pooled and equally divided across two lanes to ensure an equal distribution of all the samples across the two lanes.Barcoded libraries were sequenced to a read depth of 30 million reads using the Novaseq 6000 (Illumina) to generate 100 base pair paired-end reads.
For acquisition, cells were resuspended in FACS buffer and acquired on a 5L-Cytek Aurora instrument at the Leiden University Medical Center Flow Cytometry Core Facility with the SpectroFloâ v2.2.0.3 software (Cytek Biosciences).Data was manually gated in OMIQ (Dotmatics, 2023).All statistical analyses were performed in R. Data are expressed as mean of the relative fold change and standard error.The reported P values were determined by applying a paired two-tailed student's T test.Differences with P FDR < 0.05 (Benjamini-Hochberg) were considered to be significant.

Figure 1 .
Figure 1.Oleic acid exposure in non-activated CD4 + T cells induces changes in transcriptomics (A) Experimental setup for RNA sequencing of oleic acid-exposed non-activated CD4 + T cells, n = 9. (B) Line plot showing the relative expression of CPT1A per donor across time as a confirmation of the in vitro model by RT-qPCR.Values are colored by donor across time.On average CPT1A was upregulated 1.03 SE 0.10-fold at 0.5 h, 5.73 SE 0.40-fold at 3 h, 8.08 SE 0.53-fold at 24 h, 8.39 SE 0.62-fold at 48 h, and 11.09 SE 1.16-fold at 72 h as compared to the solvent control, n = 9. (C) Differentially expressed genes (DEGs) in oleic acid-exposed non-activated CD4 + T cells across time as compared to the solvent control.Heatmap obtained from the DESeq2 analysis resulting in 544 DEGs (P FDR < 0.05).DEGs were plotted across time to show the genes expression as log2FoldChange at each time point.Unsupervised K-means clustering indicated 2 clusters.Cluster 1 contains 310 of the DEGs, which are generally upregulated and are represented in red, and cluster 2 contains 234 of the DEGs, which are generally downregulated and are represented in blue.Genes of interest are labeled, n = 9.

Figure 2 .Figure 3 .
Figure2.Continued (D) De novo motif analysis on promoters of up-versus down-regulated genes.Enrichment of transcription factor binding motifs was performed using HOMER.6 motifs are shown with supplementing information on p value, percentage of genes in upregulated gene set and percentage of genes in downregulated gene set, transcription factor name, -log(p value), and percentage in sequence.

Figure 4 .
Figure 4. Metabolic inhibitors prevent oleic acid pre-exposure-induced changes in expression of IL-9, IL-17A, and IL-13 (*) p < 0.05, n = 3. (A) Experimental setup for spectral cytometry measurements of oleic acid + inhibitor exposed non-activated CD4 + T cells for 48 h with activation for 72 h postexposure.(B) Bar plot of IL-9, IL-17A, and IL-13 expression in CD4 + T cell after 48 h of control, oleic acid, oleic acid + atorvastatin, oleic acid + CP-640186, or oleic acid + atorvastatin + CP-640186 exposure followed by 72 h of activation with CD3/CD28 activation beads and 4 h additional stimulus with PMA/ionomycin.Values are expressed as fold change and standard error relative to control.

TABLE
d RESOURCE AVAILABILITY B Lead contact B Materials availability B Data and code availability d EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS B CD4 + T cell isolation and culture conditions B Spectral cytometry cell prep and activation B Inhibitor culture conditions and activation d METHOD DETAILS B RNA isolation B Real time-Quantitative PCR B RNA sequencing B Spectral cytometry d QUANTIFICATION AND STATISTICAL ANALYSIS B Statistical analyses B RNA sequencing analysis