Context-specific regulatory genetic variation in MTOR dampens neutrophil-T cell crosstalk in sepsis, modulating disease

Context-specific regulatory genetic variation in MTOR dampens neutrophil-T cell crosstalk in sepsis, modulating disease | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Context-specific regulatory genetic variation in MTOR dampens neutrophil-T cell crosstalk in sepsis, modulating disease Ping Zhang, Patrick MacLean, Alicia Jia, Callum O'Neill, Alice Allcock, and 13 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6457289/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 25 Feb, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Sepsis is a heterogeneous clinical syndrome with a high mortality rate and personalised stratification strategies are proposed as essential to successful targeted therapeutics. Here, we characterise genetic variation that modulates MTOR , a critical regulator of metabolism and immune responses in sepsis. The effects are highly context specific, involving a regulatory element that affects MTOR expression in activated T cells with opposite direction of effect in neutrophils. The lead variant, rs4845987, significantly interacts with the known sepsis prognostic marker neutrophil-to-lymphocyte ratio, shows activity specific to sepsis endotype, and a pleiotropic effect on type 2 diabetes (T2D) risk. Using ex vivo models, we demonstrate that activated T cells promote immunosuppressive sepsis neutrophils through released cytokines, a process dampened by hypoxia and the mTOR inhibitor rapamycin. The G-allele of rs4845987, associated with decreased risk of T2D, is associated with reduced mTOR signaling in T cells and improved survival in sepsis patients due to pneumonia. We define a novel epigenetic mechanism that fine-tunes MTOR transcription and T cell activity via the variant-containing regulatory element, which exhibits an allelic effect upon vitamin C treatment. Our findings reveal how common genetic variation can interact with disease state/endotype to modulate immune cell-cell communication, providing a patient stratification strategy to inform more effective treatment of sepsis. Biological sciences/Genetics/Genomics/Personalized medicine Health sciences/Diseases/Infectious diseases/Bacterial infection Health sciences/Medical research/Outcomes research Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Sepsis is a heterogeneous clinical syndrome characterised by life-threatening organ dysfunction associated with a dysregulated host response to infection. This includes, in a patient and disease state-specific manner, decreased lymphopoiesis and adaptive immunity, sustained inflammation, and myeloid pathologies 1 . The mechanistic basis for these diverse features and individual risk, which impact not only acute illness but also poor long-term clinical outcomes in survivors, remains unresolved 1 , 2 . To progress the goal of tackling heterogeneity in sepsis and advance personalised stratification strategies for successful targeted therapeutics, subtypes of sepsis defined by clinical features and white blood cell gene expression have been reported 3 – 6 . For example, a specific transcriptomic sepsis response signature, SRS1, identifies a subset of patients with more severe disease involving T cell exhaustion, immunosuppressive neutrophils, hypoxia and glycolysis networks 3 , 7 , informing a more stratified approach to understanding specific maladaptive sepsis responses 8 – 10 . The dysregulated host responses in sepsis span innate and adaptive immunity, and the neutrophil-to-lymphocyte ratio (NLR) has been proposed as a biomarker that reflects this interplay. Dysregulation of NLR is observed in many specific infectious diseases, as well as autoimmune disorders and cancer, providing insights into susceptibility and pathogenesis 11 , 12 . A high neutrophil-to-T cell ratio is correlated with poor prognosis and increased mortality in sepsis 13 and COVID-19 patients 14 . As one of the critical modulators of adaptive immunity, neutrophils can regulate the adaptive immune response either by releasing antimicrobial peptides and cytokines, or by direct cell-surface interactions 15 , 16 . The complex role of neutrophil subsets in immunostimulation versus immunosuppression is not fully understood. Antigen-presenting aged neutrophils (APANs) drive CD4 + T cell proliferation, exacerbating inflammation in sepsis 17 , while immunosuppressive neutrophils inhibit the proliferation and activation of CD4 + T cells and are prone to detrimental NETosis during severe sepsis 8 . However, the reciprocal interaction between neutrophils and T cells, and the role of neutrophils as potential effectors in the context of sepsis, remain incompletely defined. Recent studies have identified human expression quantitative trait loci (eQTLs) as important genetic contributors to context-dependent phenotypic variation 18 – 20 . While most currently reported eQTLs exhibit consistent effects across various contexts, those associated with disease may display cell-type specificity and can be linked to cellular dysregulation 21 , 22 . The knowledge gained from emerging evidence revealing differences in chromatin accessibility, histone modifications and transcription factor binding across cell types can be harnessed to prioritise causal eQTLs and potential drug targets 23 , 24 . However, understanding the mechanisms by which eQTLs affect gene expression, and how the cellular environment interacts to create cell-type specific effects, remains a challenge. Additionally, eQTL can exhibit inverted effects, a given genetic variant being associated with opposing effects on gene expression dependent on cell type 25 – 27 , which further emphasises the complexity of gene regulation at eQTL, the potential to impact cell-to-cell interactions, and the development of highly heterogeneous diseases such as sepsis. Driven by the potential contribution of genetic factors to the observed cellular dysregulation in sepsis 3 , 28 , here we investigate the mechanistic basis of a context-specific eQTL involving MTOR . MTOR is a key metabolic regulator, controlling a signalling hub that integrates immune receptor and metabolic signals including the switch to pro-inflammatory glycolysis in sepsis 29 . We have previously reported that MTOR is downregulated in SRS1 patients, and that there was an eQTL at this locus whose effect was specific to patients with the SRS1 endotype 3 . We now map this association to a genetic variant that we find residing in a context-specific regulatory element, and show that this has an inverted effect on MTOR expression between activated T cells and neutrophils, decreasing and increasing expression respectively. We demonstrate how an eQTL can modulate cell-to-cell communication, here in sepsis through a negative feedback loop between activated T cells and detrimental neutrophils, mediated by mTOR activity and hypoxia. The G-allele of this variant is associated with reduced mTOR signalling and cytokine release in T cells, leading to improved survival for sepsis patients. Furthermore, we characterise the epigenetic mechanisms regulating MTOR transcription in resting and activated T cells, providing insights into the survival of memory T cells and the detrimental effects of T cell activation during sepsis. Results 1. A sepsis MTOR eQTL involves directional association between neutrophils and activated T cells We first sought to understand the cellular context and specificity of the MTOR eQTL showing association dependent on sepsis endotype 3 . We fine mapped the association in patients from the UK Genomic Advances in Sepsis (GAinS) cohort using total white blood cell bulk RNA-seq and SNP genotyping (n = 823 samples from 638 sepsis patients) 28 . This localised the eQTL to a haplotype block with 25 variants in strong linkage disequilibrium (LD) (r² > 0.95), tagged by rs4845987 (Fig. 1 a and Fig. S1a-b ). The minor G-allele of rs4845987 is associated with increased expression of MTOR in whole blood and this is magnified in SRS1 patients compared to non-SRS1 patients (Fig. 1 b). To explore how this observation of endotype specificity may relate to the cellular specificity of the MTOR eQTL, we performed cell-type deconvolution of the total white blood cell bulk RNA-seq. This showed differences specific to SRS endotype involving both innate and adaptive cell populations, specifically neutrophil and T cells subsets (Fig. 1 c). We therefore considered the relationship of the eQTL with NLR, noting that higher NLR based on hospital laboratory leukocyte differential counts from our previously published cohorts 30 , 31 shows association with more severe all cause sepsis, COVID-19 and post-operative pneumonia ( Fig. S1c-e ); and a positive correlation with the quantitative transcriptomic sepsis response score (SRSq) for the likelihood of SRS1 endotype 7 ( Fig. S1f ). We found that the MTOR eQTL tagged by rs4845987 shows interaction with NLR, with increased expression of MTOR for the G allele in patients with heightened NLR (compared to those with lower NLR) (Fig. 1 d). We then asked how often this relationship may be observed. From our whole-blood sepsis eQTL data 28 , we found that 565 out of 12,726 genome-wide independent eQTL effects in sepsis show interaction with NLR (FDR < 0.01, Table S1 ), and 65% (365 out of 565) of these also showed significant interaction (FDR < 0.01) with SRS endotype ( Fig. S1g ), with strong concordance between SRS and NLR interaction effects (Pearson's r = 0.95; Fig. S1h ). This demonstrates that there are likely to be multiple independent genetic modulators of immune-relevant phenotypes dependent on relative abundance of neutrophil and lymphocyte populations, increasing the potential for clinical relevance. To identify immune cell types and treatment settings linked to the MTOR association, we next performed colocalisation analysis using data curated by the eQTL Catalogue and GTEx consortium 32 , 33 . Among 127 datasets from 75 tissues/cell types and 14 treatments, 11 different cellular or tissue contexts demonstrated shared genetic associations for the lead eQTL with MTOR expression, including T cells stimulated with anti-CD3/CD28 beads and neutrophils (PP4 > 0.95; Fig. 1 e and Fig. 1 g; Table S2 ), but not resting T cells (Fig. 1 h). We found that the minor G-allele effect of rs4845987 on MTOR expression was inverted between activated T cells and neutrophils, decreasing and increasing expression respectively (Fig. 1 f). This aligns with our interaction analysis showing lower MTOR expression with possession of the G-allele in non-SRS1 and low NLR patients (Fig. 1 b and 1 d), where T cell function was less attenuated compared to SRS1 and high NLR patients (phenotypes associated with more severe disease). 2. The MTOR allele associated with reduced T cell activity is protective for survival during sepsis We next investigated whether there was evidence of association with outcome in sepsis for the MTOR eQTL lead SNP. In the UK GAinS cohort, we found that patients with a copy of the minor G-allele exhibited significantly reduced 28-day mortality compared to those with the major C-allele ( P = 0.0010, HR = 0.60 [95% CI 0.44–0.81]; Fig. 2 a; n = 737). This association was restricted to patients with sepsis due to community-acquired pneumonia (CAP) (Fig. 2 a), and not observed in non-CAP sepsis patients ( P = 0.45; Fig. S2a ; n = 384), which is consistent with the role of mTOR in hypoxia/oxidative stress and evidence linking mTOR inhibitors to pneumonitis 35 . In an independent Sepsis Immunomics (SI) cohort 8 , we genotyped the sepsis CAP patients for rs4845987 (see Methods ). Consistent with UK GAinS, the G allele was associated with improved sepsis survival ( P = 0.031, HR = 0.32 [95% CI 0.11–0.90]; Fig. 2 b; n = 102). A similar protective effect was observed in a further independent validation cohort from the UK Biobank for participants who had confirmed bacterial pneumonia ( P = 0.0041, HR = 0.68 [95% CI 0.52–0.88]; Fig. 2 c; n = 1125; see Methods ). To further investigate the interplay between cellular features and allelic differences in clinical outcomes, we stratified the GAinS cohort into groups based on SRS status (defined as SRS-latest [assigned based on samples from the latest available time point for each patient]; SRS1-ever [assigned if any SRS1 sample was detected across time points for a given patient]) and NLR levels. We found that in non-SRS1 patients, those with the G-allele have a significantly better survival using logistic regression compared to the SRS1 patient group (OR = 0.43, P = 0.0012 when considering SRS_latest; OR = 0.38, P = 5.5e-04 using SRS1_ever; Fig. 2 d). A stronger protective association of the G-allele was also observed in patient with low NLR compared to high NLR (OR 0.46 vs 0.51; Fig. 2 d). Consistent with the immune paresis nature of the SRS1 endotype, the protective effect of the G-allele in UK Biobank patients (Fig. 2 c) was observed only in those without an immunosuppressed status (IS) (Fig. 2 e; see Methods). Prolonged hypoxia significantly contributes to multi-organ failure, a hallmark of severe sepsis, and is relevant to a potential role for genetic variants of MTOR . We therefore analysed patients requiring a high fraction of inspired oxygen (FiO 2 ≥ 0.4) or with a low ratio of arterial oxygen tension (PaO 2 ) to FiO 2 (0 < PaO2/FiO2 ≤ 100) as indicators of hypoxaemia during sepsis 36 , 37 . As expected, we observed high FiO 2 and low PaO 2 /FiO 2 were associated with poor survival in the GAinS cohort (HR = 2.6 and 3.8 respectively; P < 0.0001; Fig. S2b ). Significant association for FiO 2 level, but not PaO 2 /FiO 2 , was also observed in another independent cohort GenOSept 38 ( Fig. S2b ). Consistently, in patients with low FiO 2 across both cohorts, we observed stronger allelic differences in survival (OR = 0.46, P = 0.033 for GAinS; OR = 0.24, P = 0.0071 for GenOSept; Fig. 2 d). We did not observe significant associations between the eQTL and survival in non-CAP patients with or without any stratification (Fig. 2 d yellow and cyan dots; Fig. S2a ). 3. Pleiotropic effects of genetic variant on T2D risk and sepsis survival through modulation of MTOR expression We next further investigated the relationship between the MTOR sepsis-associated eQTL and other disease-related traits using the GWAS Catalog 39 . We found that rs4845987 or SNPs in high LD (r 2 > 0.8) were significantly associated ( P ≤ 5e-08) with 18 GWAS traits, 11 of which were related to T2D or its associated metabolic traits such as body mass index and insulin resistance ( Table S3 ). We observed strong colocalisation between the MTOR eQTLs and risk variants for T2D in two GWAS studies 40 , 41 (PP4 = 0.75 and 0.97 respectively; Fig. S3a-b ). The minor G-allele of rs4845987, which is protective for sepsis survival, was also associated with a reduced risk of T2D ( Fig. S3a ). Individuals with T2D have an increased risk of developing sepsis, and sodium-glucose cotransporter-2 inhibitors used to treat T2D suppress MTOR and reduce pneumonia risk 42 , 43 . To test the potential causative effect of MTOR expression on T2D risk, we performed a summary data-based Mendelian randomisation (SMR) analysis 44 . We focused on the genes within a 2Mbp window around the GWAS risk loci of the MTOR locus and observed significant effects on increased T2D risk through higher expression of MTOR in activated T cells ( P SMR = 1.17e-06 and 1.92e-6 in CD4 + and CD8 + T cells respectively; Fig. 2 f-g; Table S4; see Methods ). In line with the eQTL associations (Fig. 1 f), we observed inverted effects in neutrophils and adipose tissue (Fig. 2 h and Fig. S3c-d ), both of which are known to interact with T cells in disease settings 15 , 45 , 46 . Given the disease and immunological associations, we hypothesised that the MTOR variant would have been under selection pressure. We observed higher frequency of the G-allele in Africans, compared to other populations, with allele frequencies of 0.91, 0.31, 0.21, 0.30, and 0.38 in Africans, Americans, East Asians, Europeans, and South Asians respectively ( Fig. S3e ), as well as high genetic divergence (measured by fixation index, F ST 47 ) ( Fig. S3f-g ). This would be consistent with potential functional consequences associated with survival fitness and disease modulating allele frequency following migrations and genetic adaptations within human populations 48 . 4. Activation of sepsis neutrophils by T cells is dampened by mTOR inhibition and hypoxia Our findings of a protective G-allele effect in patients with relatively reduced MTOR expression in activated T cells, and with higher expression in neutrophils, led us to hypothesise a potential role for T cells in driving sepsis-associated neutrophil phenotypes via the mTOR pathway (Fig. 2 i). To test this hypothesis, we established an ex vivo co-culture model of neutrophils and T cells in which we could compare effects of co-culture with either neutrophils from sepsis patients or healthy donors, with primary T cells that were resting or activated (Fig. 3 a and Fig. S4a ). We assayed three neutrophil surface markers, CD64, CD123, and PD-L1, previously associated with sepsis severity 51 , 52 , as indicators of neutrophil activity. All three markers exhibited upregulation in neutrophils isolated from sepsis patients compared to healthy controls, at both the protein (Fig. 3 b) and mRNA levels ( Fig. S4b ). The expression levels were further elevated in SRS1 compared to non-SRS1 patients ( Fig. S4b ). We found that co-culture with activated T cells, relative to resting T cells, resulted in a further increase in marker expression on sepsis neutrophils (Fig. 3 b-c). This upregulation was associated with enhanced NETosis ( Fig. S5a ). Consistent with the observed eQTL association, both CD4 + and CD8 + T cells demonstrated similar effects on neutrophil markers ( Fig. S4c ). However, the immortalised Jurkat T cell line did not affect these markers on sepsis neutrophils compared to primary T cells ( Fig. S4d ), potentially due to altered T cell function in the Jurkat line. Interestingly, co-culture with activated T cells did not significantly affect CD123 and CD64 expression on neutrophils from healthy donors or convalescent patients compared to resting T cells (Fig. 3 c), suggesting a specific priming of sepsis neutrophils for hyperactivation by activated T cells. T cell activation is known to be heavily reliant on mTOR and HIF-1α (hypoxia-inducible factor 1α) 29 , and mTOR inhibition and hypoxia attenuate T cell activity and cytokine release 53 , 54 . Consistent with these previous findings, in co-culture assays in the presence of rapamycin, a specific mTOR inhibitor, under hypoxia condition, or with rapamycin pre-treated T cells, we found a decreased effect on sepsis neutrophils (Fig. 3 d-e and Fig. S5b . This effect was specific to T cell-neutrophil interactions, as rapamycin treatment did not affect sepsis neutrophils cultured alone ( Fig. S5c ). Together, these results highlight the critical role of mTOR activity in T cell activation and the consequent deleterious effects on neutrophils in the context of sepsis. 5. Reciprocal interactions between activated T cells and sepsis neutrophils Previous studies have revealed that neutrophils regulate adaptive immune responses either by releasing secreted peptides and cytokines or by direct cell-surface interactions 15 , 16 , and neutrophils from SRS1 sepsis patients exhibit a stronger inhibitory effect on T cell proliferation and activation relative to non-SRS1 patients 8 . In our co-culture experiments, PD-1 and CD69 expression was significantly suppressed on T cells in the presence of sepsis neutrophils, an effect not observed when using neutrophil-conditioned media alone (Fig. 4 a), suggesting direct cell-surface contact is crucial Conversely, CD123 and CD64 expression was markedly upregulated on sepsis neutrophils when co-cultured with activated T cells or exposed to conditioned T cell media (Fig. 4 b). This upregulation was also observed in a sepsis whole blood culture model ( Fig. S6a-b ), where T cell activation enhanced neutrophil markers ( Fig. S6c-d ), and the increase in immunosuppressive neutrophils subsequently inhibited T cell activation (48h vs 24h; Fig. S6e ). Together, these results indicate a negative feedback loop between T cell activation and neutrophils in sepsis, mediated by both cytokine release and direct cell-surface interactions (Fig. 4 c). Disruption of this loop, due to excessive mTOR signalling in T cells, may contribute to poor clinical outcomes. To prioritise the potential cytokines mediating the cell-cell interactions, we generated RNA-seq datasets in primary CD4 + and CD8 + T cells following activation with anti-CD3/CD28 beads. As expected, TCR stimulation triggered substantial transcriptomic reprogramming (Fig. 4 d; Table S5 ), with 68 and 59 differentially expressed (DE) genes encoding annotated cytokines identified in CD4 + and CD8 + cells, respectively. Overall, 81.4% (48 out of 59) of DE cytokines in CD8 + T cells overlapped with those in CD4 + cells (Fig. 4 e) and displayed a strong correlation (Pearson’s r = 0.84; Fig. 4 f). Interestingly, 11 of the 48 commonly DE cytokines in both T cell subsets were also rapamycin-sensitive (Fig. 4 g), including IL-3 (ligand for its receptor CD123) and IFN-γ, a cytokine known to augment NETosis via ligand-receptor interaction, contributing to acute lung injury during sepsis 17 . Additionally, the transcription of the sepsis surface markers was also upregulated in response to T cell cytokines ( Fig. S5d ), suggesting a more intricate regulatory network. 6. Regulatory mechanism underlying the MTOR eQTL in activated T cells We proceeded to investigate the regulatory genomic landscape of the MTOR genetic association in order to prioritise and further characterise the MTOR eQTL. Studies have shown that causal eQTLs and GWAS risk SNPs are more likely to be found in open chromatin 23 , 32 . We investigated context specific epigenomic datasets and found that the MTOR eQTL variant rs4845987 is located at the centre of an ATAC-seq peak in intron 8 of MTOR , a peak identified exclusively in resting T cells (both CD4 + and CD8+), but not in other immune cells (Fig. 5 a and Fig. S7a ). This ATAC peak was completely silenced in activated T cells (Fig. 5 a-b), coinciding with a specific reduction in MTOR expression, while the expression of other nearby genes remained unaffected (Fig. 5 c-d and Fig. S8b ). We also observed a similar pattern of the ATAC peak at this location being present in central memory T cells, but not in naive cells, following TCR activation and CAR T cell production ( Fig. S7b-c ). Upon stimulation, this locus exhibited coordinated changes in epigenetic features, marked by the absence of enhancer markers H3K27ac and H3K4me1 and presence of hydroxymethylation in activated T cells (Fig. 5 e and Fig. S8c ). The importance of this specific genomic region was further supported by proximal CTCF-binding and its interaction with the MTOR promoter (Fig. 5 e and Fig. S8a ). We hypothesised that the enhancer plays a critical role in maintaining MTOR expression and cell survival in resting memory T cells while masking the eQTL effect within the enhancer itself, and that in activated T cells, the eQTL-associated genetic variant may exert its effect through hydroxymethylation, leading to an allelic effect observed in bulk T cells (Fig. 1 f) and single-cell T cell subsets ( Fig. S7d ). To experimentally confirm this, we manipulated hydroxymethylation using vitamin C (VC), a known activator of TET enzymes that catalyse demethylation and 5-hydroxymethylcytosine (5hmC) 56 . As predicted, we observed increased MTOR expression upon VC in activated T cells with the C/C genotype compared to G allele carriers (Fig. 5 f). We next utilised a CRISPR/dCas9-based epigenetic activation system delivered via lentivirus to manipulate this locus and demonstrate direct evidence for enhancer function (Fig. 5 g; see Methods). We observed a significant upregulation of MTOR expression in activated T cells with sgRNA targeting the variant-containing regions (sgRNA-e) compared with a non-targeting control sgRNAs, or sgRNAs targeting the flanking regions (Fig. 5 h). Together, these results demonstrate a unique epigenetic mechanism that ensures precise control of MTOR expression, which is critical for proper T cell function in immune responses (Fig. 5 i ) . Discussion In this study, we identify a genetic variant located within an enhancer element that regulates MTOR transcription in T cells. Our findings are consistent with this enhancer sustaining essential baseline mTOR activity in resting T cells while masking the MTOR eQTL effect. Upon T-cell receptor activation, mTOR is highly upregulated through post-translational modifications, with silencing of this enhancer providing a negative feedback loop to prevent excessive mTOR activity at the transcription level. In activated T cells, the eQTL effect is exposed through hydroxymethylation involving the variant which can be manipulated by VC via activation of TET enzymes. The C allele amplifies mTOR signalling and T cell activity during sepsis, ultimately leading to persistent neutrophil hyperactivation, NETosis, and increased mortality in patients with sepsis due to pneumonia. mTOR is a critical signalling hub that integrates immune receptor and metabolic signals to determine the fate of T cells 29 . In this study, we utilised epigenetic profiling to identify a regulatory element that controls MTOR transcription, distinguishing between resting and activated T cells. This enhancer is essential for memory T cell survival and functions as a negative feedback loop to suppress mTOR activation upon TCR stimulation. Further studies are needed to elucidate key extrinsic modulators by examining epigenetic remodelling of cytokine loci, with a focus on the mTOR pathway and those affecting sepsis-associated neutrophil dysfunction. The deleterious effects of cytokine release during sepsis are supported by recent observations of sustained cytokine signaling upregulation and the potential harmful effects of interferon gamma-1b therapy in pneumonia cases 61 , 62 . Understanding these modulators could also provide insights into whether these factors impact the efficacy of CAR-T cell therapy, which is often complicated by severe infection 63 . Pretreatment of CAR-T cells with rapamycin has shown promise in promoting bone marrow infiltration and enhancing therapeutic efficacy in acute myeloid leukaemia 64 . This raises the possibility of combining rapamycin with approaches targeting immunosuppressive neutrophils to improve outcomes in sepsis patients. Cytosine modifications, including 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC), are crucial epigenetic mechanisms that regulate chromatin structure and gene transcription. While 5mC is known to be associated with gene repression, 5hmC as an intermediate in demethylation is often linked to gene activation. Recent studies have highlighted the broader implications of 5hmC in cell differentiation and disease 65 . Our findings revealed a novel regulatory mechanism involving 5hmC at the MTOR locus, mediating T cell function and vitamin C response. Currently, the analysis of 5mC and 5hmC is primarily conducted at the bulk level, obscuring potential heterogeneity between different cell types and/or subsets. Future studies could leverage single-cell approaches to obtain insights into the distinct 5mC/5hmC dynamics during T cell differentiation and acute infections. Chronic inflammation and metabolic dysfunction in adipose tissue are key drivers of insulin resistance and T2D 66 . Both CD4 + and CD8 + T cells infiltrate adipose tissue, with increased interferon-γ-expressing T cells promoting inflammation 67 , 68 . The adipocyte-secreted hormone leptin directly regulates T cell proliferation, glycolytic metabolism and cytokine production 45 , 46 . Individuals with T2D have an increased risk of developing infections including sepsis and COVID-19 69 . However, the overall evidence linking T2D to sepsis outcomes remains unclear but appears more robust for pneumonia 70 , 71 . Interestingly, sodium-glucose cotransporter-2 inhibitors used to treat T2D suppress MTOR 43 and reduce pneumonia risk 42 . Our study demonstrated colocalisation of a genetic variant between adipose tissue, T cells for MTOR expression and T2D GWAS loci. This tissue-specific variant had opposite allelic effects on MTOR expression, with the G allele decreasing MTOR expression in activated T cells while increasing it in adipose tissue. These results suggest a potential role for mTOR signaling in the adipose-T cell axis with implications for the interplay between sepsis and T2D. Vitamin C (VC) supplementation has been proposed as a potential treatment for sepsis due to its anti-inflammatory and antioxidant properties. However, clinical trials have shown mixed results 72 – 75 , potentially due to different patient severities at enrolment. Our results revealed a detrimental effect of mTOR-mediated T cell activity on pneumonia risk. Furthermore, we showed that VC increases MTOR expression in activated T cells from risk C-allele carriers, providing insights into the potential mechanism of action of harmful VC therapy in sepsis 72 . In contrast to VC, which activates TET enzymes 56 , TET inhibition could potentially reduce T cell-mediated cytokine release, alleviate neutrophil dysfunction and improve patient survival. Indeed, treatment of TET inhibitors reduces 5hmC levels and attenuates LPS-induced pulmonary oedema and lung injury in a sepsis mouse model 76 . However, the specific role of TET inhibition in modulating T cell-neutrophil crosstalk during human sepsis remains to be seen. Materials and Methods eQTL interaction and colocalisation analysis eQTL interaction was determined using a linear mixed model as implemented in R package lmerTest 77 : p ~ g + i + g:i + (1|donor), where p is the gene expression corrected for population structure and PEER factors as described in 28 , g is the genotype vector, i is the term for SRS endotypes or log2 transformed NLR, g:i is the interaction between genotype and either SRS or log2 transformed NLR, and (1|donor) is a random effect accounting for variability between donors. Gene expression and genotype data were obtained from 823 RNA-seq samples derived from 638 sepsis patients, as described by Burnham et al 28 . Colocalisation analysis was performed in a ± 200-bp window surrounding the MTOR lead eQTL, using R package coloc 34 with default settings. eQTL summary statistics from healthy tissue and cell types were retrieved from eQTL Catalogue 32 . Type 2 diabetes GWAS summary statistics were downloaded from https://ftp.ncbi.nlm.nih.gov/dbgap/studies/phs001672/analyses/ 40 , and https://www.diagram-consortium.org/downloads.html 41 , respectively. Cell type deconvolution Aligned bam files derived from sepsis leukocytes were obtained from the GAinS study. RNA-seq count matrix was generated using featureCounts and normalised by DESeq2. As described previously 31 , the CIBERSORTx 78 absolute scores of each cell type in bulk samples were obtained using the count matrix for the bulk RNA-seq and the signature matrix derived from sepsis single cell RNA-seq 8 with S-mode batch correction and 100 permutations. Summary data–based Mendelian randomisation T2D GWAS in Europeans and eQTL summary data were obtained from Suzuki K. et al. 41 and eQTL Catalogue 32 , respectively. We performed the SMR analysis 44 using eQTLs as instrumental variables to identify genes whose expression is associated with T2D risk due to pleiotropy and/or causality. Genes were included in the analysis if they had at least one cis-eQTL ( P < 5e⁻⁸) in either activated CD4+, CD8 + T cells 20 , neutrophils 49 , or adipose/fat tissue 79 , 80 within a 2 Mbp window around GWAS loci, following the default settings of the SMR tool (v1.3.1). The HEIDI (heterogeneity in dependent instruments) test was applied to differentiate functional associations from linkage effects. LD correlation between SNPs was estimated using 1000 Genomes Project data for Europeans. Survival analysis Patients with sepsis were recruited from the Genomic Advances in Sepsis (GAinS) and Sepsis Immunomics (SI) studies. The inclusion and exclusion criteria for GAinS and SI studies were described previously 8 , 81 . Blood samples were obtained on Day 1, 3 and 5 of hospital admission, with follow-up samples taken an average of 3 months after hospital discharge. CAP was diagnosed as a febrile illness with cough, sputum production, breathlessness, leukocytosis and radiological evidence of pneumonia, acquired in the community or within 2 days of hospital admission. To assess the association between genetic variants and 28-day mortality, Cox proportional-hazards model and logistic regression adjusting for age, sex, and the first seven principal components were used. Multivariable Cox regression was performed to evaluate the predictive ability of hypoxia-related factors (FiO2, PaO2:FiO2 ratio) for 28-day mortality while controlling for age (> 65 vs ≤ 65 years) and sex. All statistical analyses were conducted using R (v4.2.1). UK Biobank (UKB) data curation and analysis The UKB recruitment and ethical approval process has been described previously 82 . Briefly, half a million men and women aged 40–69 years attended one of 22 UKB assessment centres located throughout England, Scotland and Wales between 2006 and 2010. All participants completed a touchscreen questionnaire, verbal interview and had a range of physical measurements and blood, urine and saliva samples taken for long-term storage. Genotype data was generated as described previously 83 . Withdrawn participants were removed from the available UKB dataset and only individuals who received a diagnosis of bacterial pneumonia based on International Classification of Diseases 10th Revision (ICD10) J15 in any instance of the UKB-specified Field 41270 were included in the analysis (n = 1,461). The corresponding date of diagnosis was extracted from the matching instance in Field 41280. For those with multiple diagnoses the sole or earliest date of diagnosis was used as the only reference point for impatient start date. Participant death was determined via the presence of a date in instance 0 of Field 40000. Individuals who died within 28 days of their earliest ever recorded diagnosis of the relevant ICD10 code were classified as a case with the remainder as controls. The time in days between the earliest date of diagnosis and date of death was used as the time to event metric. Age in years was calculated by subtracting the estimated date of birth (using ‘year_of_birth_f34_0_0’, ‘month_of_birth_f52_0_0’, and day of birth manually assigned as the 15th to avoid regression to the mean) from the date of the relevant ICD10 diagnosis and dividing by 52.143. Self-reported sex in instance 0 of field 31 was used to define males and females. Patients with a genetically determined ancestry of White British Subset (WBS) were included (n = 1,125), and the first seven principal components were used for survival analysis as described above. Immunosuppressed status (IS) was defined based on ICD-10 digital codes (Field 41270) as previously described 84 , covering chronic viral hepatitis, Human Immunodeficiency virus (HIV) or Human T-cell lymphotropic virus (HTLV) infection, organ transplantation, medication, autoimmunity, blood cancer, and congenital disease. SNP genotyping We genotyped the SNP rs4845987 for the Sepsis Immunomics (SI) cohort using samples derived from CAP patients with hospitalisation dates up to 04-12-2024. The inclusion and exclusion criteria for this study were detailed above. Briefly, genomic DNA was extracted from buffy coat samples (stored at -80°C) using the Monarch Genomic DNA Purification Kit (NEB #T3010) following the manufacturer’s instruction. gDNA was then quantified using the Qubit dsDNA HS Assay Kit (ThermoFisher). Sample genotypes were determined using TaqMan Genotyping Assays with the Universal PCR Master Mix and a predesigned probe for rs4845987 (C_2524855_10, ThermoFisher) on a CFX-96 C1000 platform (Bio-Rad). For healthy volunteers, purified cells were used for gDNA extraction. Isolation and culture of human neutrophils and T cells Peripheral blood mononuclear cells (PBMCs) from healthy donor leucocyte cones were isolated by Ficoll-Paque (Sigma) density gradient centrifugation. CD4 + and CD8 + T cells were then separated from PBMCs by positive selection with magnetic MicroBeads (Miltenyi Biotec) following the manufacturer’s instructions. Isolated T cells were frozen in FBS (Sigma, #F7524) with 10% DMSO at 2 x 10 7 cells/ml and stored in liquid nitrogen. Total T cells were maintained in RPMI-1640 (Gibco) medium supplemented with 5% FBS (Sigma, #F7524), 100 mM L-glutamine (Sigma), 1x penicillin-streptomycin (Sigma) and 500 IU/ml human recombinant IL-2 (Biolegend). T cells were activated using anti-CD3/CD28 Dynabeads (ThermoFisher, #11131D) at a 1:1 cell:bead ratio at 2 x 10 6 cells/ml. Neutrophils were extracted from 2–10 ml whole blood from sepsis patients or healthy donors using EasySep HLA Chimerism Whole Blood CD66b positive selection kit (STEMCELL) as per manufacturer’s instructions. Neutrophils were maintained in RPMI-1640 (Gibco) medium supplemented with 10% FBS, 100 mM L-glutamine and 1x penicillin-streptomycin. For co-culture, cryopreserved CD4 + T cells were thawed and cultured with CD66b + neutrophils in 24-well plates at a 1:2 T cells to neutrophils ratio in the presence or absence of anti-CD3/CD28 beads for 48 hours. Rapamycin (#A8167-APE) was obtained from Stratech Scientific. L-ascorbate (Vitamin C) was from Merck (#11140). Flow cytometry Cells were harvested by centrifugation at 300 g for 3 minutes, followed by live/dead cell staining (Fixable Green Dead Cell Stain Kit; ThermoFisher). Cells were stained using surface makers CD3-APC (Biolegend, #300412), PD-1-PE-Cy7 (Biolegend, #367414), CD69-PerCP-Cy5.5 (Biolegend, #310926), CD66b-AF700 (Biolegend, #305114), PD-L1-BV605 (Biolegend, #329724), CD123-PE (BD Biosciences, #554529) and CD64-BV421 (Biolegend, #305020) for 45 minutes at room temperature, and washed with 0.2% bovine serum albumin (BSA) in phosphate-buffered saline (PBS). Samples were then acquired on a LSRFortessa X-20 (BD Biosciences) flow cytometer and analysed using FlowJo software (v10.10). Whole blood samples from sepsis patients were collected in BD Vacutainer EDTA tubes and cultured in T cell media at a 1:4 ratio, with or without anti-CD3/CD38 Dynabeads at a 1:10 v/v ratio, for 48 hours. Cultures were then treated twice with RBC lysis buffer prior to antibody staining and flow cytometry. qRT-PCR and RNA-seq Total RNA was extracted from lysed cells using the Monarch Total RNA Miniprep Kit (NEB #T2010). cDNA synthesis was subsequently performed with LunaScript RT SuperMix Kit (NEB #E3010). Gene expression levels were quantified by qRT-PCR using SYBR Green Real-Time PCR Master Mix (Qiagen) on a CFX-96 C1000 platform (Bio-Rad), with β-actin as the normalisation control (see Table S8 for primer sequences). 1µg RNA was used for library preparation using the NEBNext Ultra II RNA Library Prep Kit for Illumina (#E7770S) following the manufacturer's protocol. Poly(A) mRNA enrichment was performed using the NEBNext Poly(A) mRNA Magnetic Isolation Module (#E7490). Library quality control, including size distribution and quantification, was assessed using TapeStation 4200 (Agilent) with High Sensitivity D1000 reagents and Qubit HS DNA kit (ThermoFisher), respectively. Final library molarity was determined using the KAPA Library Quantification Kit (Roche). Libraries were sequenced on the Illumina NextSeq 500 platform using a 150-cycle High Output Kit v2.5. RNA-seq analysis Raw RNA-seq reads were trimmed using Trim Galore (v0.6.2) and aligned to the human genome (hg38) using HISAT2 (v2.1.0). Transcript quantification was performed using featureCounts (v1.6.2) with GENCODE v31 annotations. The bigwig files normalised by RPKM (Reads Per Kilobase per Million mapped reads) were generated using the bamCoverage function of deepTools (version 3.3.1). Differential gene expression analysis was conducted using DESeq2 (v1.36.1) on raw read counts. Omni-ATAC-seq and analysis 50,000 cells were prepared by centrifugation and resuspended in 50 µl of lysis buffer (10 mM Tris-HCL pH 7.4, 10 mM NaCl, 3 mM MgCl2, 0.01% Digitonin, 0.1% Tween-20 and 0.1% Igepal CA-630) for nuclear isolation. Following transposition and DNA purification, library preparation was performed using standard protocols as described previously 85 , 86 . Final libraries were quality controlled and sequenced on the Illumina NextSeq 500 platform. Sequencing reads for ATAC-seq were aligned to the human genome (hg38) using Bowtie2 (v2.2.5). As described previously 23 , 86 , data were filtered for quality control using Picard (v2.0.1) and Samtools (v1.9) before peak calling with MACS2 (v2.1.0). Differential peak analysis was performed using DESeq2, considering peaks present in at least 30% of samples. Potential batch effects and/or technical variation were assessed through principal component analysis and incorporated as covariates in the DESeq2 design formula. Public hMeDIP-seq analysis Raw sequencing reads for hydroxymethylated DNA Immunoprecipitation Sequencing (hMeDIP) data were obtained from GSE74850 60 , trimmed using Trim Galore (v0.6.2), and aligned to human genome (hg38) using the BWA-mem alignment algorithm (v0.7.12) 87 . The binary alignment and map (BAM) files were filtered to remove reads with a mapping quality score less than 10 and duplicate reads using SAMtools (v1.9) and Picard (v2.21.1). The normalised fold enrichment tracks over the corresponding input controls were generated by using the callpeak function with the --SPMR flag, then passing the bedgraph outputs into the bdgcmp function of MACS2 and the bedGraphToBigWig tool. Lentivirus production Human embryonic kidney (HEK) 293FT cells were maintained in Opti-MEM I Reduced Serum Medium (OPTI-MEM) with GlutaMAX Supplement (ThermoFisher, #51985034) supplemented with 5% FBS (Sigma, #F7524), 1 mM sodium pyruvate (ThermoFisher), and 1× MEM nonessential amino acids (ThermoFisher) in T175 flasks. Cells were seeded per 150 mm dish (Corning, #430599) in 14 ml of medium overnight to achieve confluency about 90% at the time point of transfection. Cells were transfected with a plasmid mixture containing 5.7 µg psPAX2 (Addgene #12260), 3.2 µg pCMV-VSV-G (Addgene #8454), and either 4.6 µg of a sgRNA expression vector (Addgene #96923) or 7.0 µg of dCas9-VP64 (Addgene #180263) in equimolar ratios using jetPRIME reagent (Polyplus). After 6 h, the transfection medium was replaced with fresh medium supplemented with ViralBoost (Alstem Bio, #VB100). Lentiviral supernatant was harvested in 24 h post-transfection, filtered with a 0.45 µm membrane filter (Millipore), and concentrated using ultracentrifugation at 29,000 rpm for 2h at 4°C. The pellet was resuspended in PBS with 1.5% BSA, aliquoted and stored at -80°C. Lentivirus titers of dCas9-VP64 were determined by quantifying mCherry-positive cells via flow cytometry post-transduction. Viral volume that results in at least 50% transduction efficiency was used. CRISPR-dCas9 mediated epigenetic editing for primary T cells We designed and selected top ranked single guide RNA (sgRNA) based on the scoring metrics using FlashFry 88 . For the human U6 promoter-based transcription, a guanine base was added to the 5′ of the sgRNA when the 20bp guide sequence did not begin with G. The sgRNA sequences are listed in Table S8 . Primary human CD4 + T cells were activated using anti-CD3/CD28 Dynabeads (ThermoFisher). One million cells were transduced with sgRNA- and dCas9-VP64-containing lentiviruses in the presence of 8 µg/mL polybrene (Merck, #28728-55-4) for 24 hours. Transduced cells were maintained in media and assayed in 6 or 7 days. Declarations Ethics approval and consent to participate Peripheral blood samples were obtained from healthy volunteers following informed consent (Oxfordshire Research Ethics Committee approval REC reference 06/Q1605/55); and from sepsis patients in the Sepsis Immunomics (SI) study (South Central Oxford REC C, reference:19/SC/0296) and UK GAinS (REC approvals 05/MRE00/38, 08/H0505/78, and 06/Q1605/55) with ethics approval granted nationally and locally, and informed consent obtained from all patients or their legal representative. UK Biobank has obtained ethics approval from the North West Multi-centre Research Ethics Committee (approval number: 11/NW/0382) and had obtained informed consent from all participants. Declaration of interests JCK reports a grant to his institution from the Danaher Beacon Programme for work on RNA biomarker point-of-care test development in sepsis for endotype assignment which includes support for KC-G and JCK. All remaining authors declare that they have no competing interests. Acknowledgments This project was supported by the Medical Research Council (MR/V002503/1) (JCK), a Wellcome Trust Investigator Award [204969/Z/16/Z] [JCK], Chinese Academy of Medical Sciences (CAMS) Innovation 537 Fund for Medical Science [2018-I2M-2-002] [PZ, JCK], Wellcome Trust Grants [090532/Z/09/Z and 203141/Z/16/Z to core facilities Centre for Human Genetics, and the NIHR Oxford Biomedical Research Centre. Computation used the Oxford Biomedical Research Computing (BMRC) facility, a joint development between the Wellcome Centre for Human Genetics and the Big Data Institute supported by Health Data Research UK and the NIHR Oxford Biomedical Research Centre. The Wellcome Sanger Institute is funded by the Wellcome Trust [220540/Z/20/A]. CO’N. is supported by a Wellcome Trust Doctorate Award (228321/Z/23/Z). AJM received support from the Academy of Medical Sciences Starter Grant (SGL024∖1096) and the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC). Data and code availability GAinS gene expression and genotyping data were deposited at the European Genome-phenome Archive (EGA), under accession number EGAD00001008730 7 and EGAD00001015369 28 . RNA-seq and ATAC-seq raw FASTQ files for CD4+ and CD8+ T cells are available under accession number EGAS50000000894 (https://ega-archive.org/studies/EGAS50000000894). Processed data, including raw and normalised counts and bigWig files for genome-wide signal data can be accessed on Zenodo (https://zenodo.org/uploads/14907264). The raw ATAC-seq data for primary immune cells were obtained from GSE172116 (macrophages), EGAS00001007362 (monocytes), GSE150018 (neutrophils), GSE118189 (NK and dendritic cells) and GSE168882 (CAR T cells). RNA-seq data in rapamycin-treated CD4+ T cells were obtained from GSE129829. MeDIP-Seq for 5hmC in CD4+ T cells were obtained from GSE74850. The processed histone modification and CTCF ChIP-seq results were downloaded from the ENCODE project (https://www.encodeproject.org/; see also Table S7 ). All code used for data processing and analysis in this study is publicly available on GitHub (https://github.com/jknightlab/MTOR-Genetics-Project). For the purpose of Open Access, the authors have applied a CC BY public copyright licence to any Author Accepted Manuscript arising from this submission. References Shankar-Hari M et al (2024) Reframing sepsis immunobiology for translation: towards informative subtyping and targeted immunomodulatory therapies. Lancet Respir Med 12:323–336 Cajander S et al (2024) Profiling the dysregulated immune response in sepsis: overcoming challenges to achieve the goal of precision medicine. Lancet Respir Med 12:305–322 Davenport EE et al (2016) Genomic landscape of the individual host response and outcomes in sepsis: a prospective cohort study. Lancet Respir Med 4:259–271 Scicluna BP et al (2017) Classification of patients with sepsis according to blood genomic endotype: a prospective cohort study. Lancet Respir Med 5:816–826 Sweeney TE et al (2018) Unsupervised Analysis of Transcriptomics in Bacterial Sepsis Across Multiple Datasets Reveals Three Robust Clusters. Crit Care Med 46:915–925 Seymour CW et al (2019) Derivation, Validation, and Potential Treatment Implications of Novel Clinical Phenotypes for Sepsis. JAMA 321:2003–2017 Cano-Gamez E et al (2022) An immune dysfunction score for stratification of patients with acute infection based on whole-blood gene expression. Sci Transl Med 14:eabq4433 Kwok AJ et al (2023) Neutrophils and emergency granulopoiesis drive immune suppression and an extreme response endotype during sepsis. Nat Immunol 24:767–779 Boomer JS et al (2011) Immunosuppression in patients who die of sepsis and multiple organ failure. JAMA 306, 2594 – 605 Antcliffe DB et al (2019) Transcriptomic Signatures in Sepsis and a Differential Response to Steroids. From the VANISH Randomized Trial. Am J Respir Crit Care Med 199:980–986 Song M, Graubard BI, Rabkin CS, Engels EA (2021) Neutrophil-to-lymphocyte ratio and mortality in the United States general population. Sci Rep 11:464 Templeton AJ et al (2014) Prognostic role of neutrophil-to-lymphocyte ratio in solid tumors: a systematic review and meta-analysis. J Natl Cancer Inst 106:dju124 Huang Z, Fu Z, Huang W, Huang K (2020) Prognostic value of neutrophil-to-lymphocyte ratio in sepsis: A meta-analysis. Am J Emerg Med 38:641–647 Liu J et al (2020) Longitudinal characteristics of lymphocyte responses and cytokine profiles in the peripheral blood of SARS-CoV-2 infected patients. EBioMedicine 55:102763 Bert S, Nadkarni S, Perretti M (2023) Neutrophil-T cell crosstalk and the control of the host inflammatory response. Immunol Rev 314:36–49 Li Y et al (2019) The regulatory roles of neutrophils in adaptive immunity. Cell Commun Signal 17:147 Jin H et al (2023) Antigen-presenting aged neutrophils induce CD4 + T cells to exacerbate inflammation in sepsis. J Clin Invest 133 Fairfax BP et al (2014) Innate immune activity conditions the effect of regulatory variants upon monocyte gene expression. Science 343:1246949 Consortium GT (2020) The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science 369:1318–1330 Schmiedel BJ et al (2018) Impact of Genetic Polymorphisms on Human Immune Cell Gene Expression. Cell 175:1701–1715e16 Yazar S et al (2022) Single-cell eQTL mapping identifies cell type-specific genetic control of autoimmune disease. Science 376:eabf3041 Natri HM et al (2024) Cell-type-specific and disease-associated expression quantitative trait loci in the human lung. Nat Genet 56:595–604 Zhang P et al (2022) Epigenomic analysis reveals a dynamic and context-specific macrophage enhancer landscape associated with innate immune activation and tolerance. Genome Biol 23:136 Fang H et al (2019) A genetics-led approach defines the drug target landscape of 30 immune-related traits. Nat Genet 51:1082–1091 Mizuno A, Okada Y (2019) Biological characterization of expression quantitative trait loci (eQTLs) showing tissue-specific opposite directional effects. Eur J Hum Genet 27:1745–1756 Ota M et al (2021) Dynamic landscape of immune cell-specific gene regulation in immune-mediated diseases. Cell 184:3006–3021e17 de Klein N et al (2023) Brain expression quantitative trait locus and network analyses reveal downstream effects and putative drivers for brain-related diseases. Nat Genet 55:377–388 Burnham KL et al (2024) eQTLs identify regulatory networks and drivers of variation in the individual response to sepsis. Cell Genom, 100587 Chi H (2012) Regulation and function of mTOR signalling in T cell fate decisions. Nat Rev Immunol 12, 325 – 38 COVID-19 Multi-omics Blood ATlas Consortium (2022) A blood atlas of COVID-19 defines hallmarks of disease severity and specificity. Cell 185:916–938e58 Torrance HD et al (2023) A Transcriptomic Approach to Understand Patient Susceptibility to Pneumonia After Abdominal Surgery. Ann Surg Kerimov N et al (2021) A compendium of uniformly processed human gene expression and splicing quantitative trait loci. Nat Genet 53:1290–1299 Consortium GT et al (2017) Genetic effects on gene expression across human tissues. Nature 550:204–213 Giambartolomei C et al (2014) Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using Summary Statistics. PLoS Genet 10 Albiges L et al (2012) Incidence and management of mTOR inhibitor-associated pneumonitis in patients with metastatic renal cell carcinoma. Ann Oncol 23:1943–1953 Sweeney RM, McAuley DF (2016) Acute respiratory distress syndrome. Lancet 388:2416–2430 Villar J et al (2015) Assessment of PaO(2)/FiO(2) for stratification of patients with moderate and severe acute respiratory distress syndrome. BMJ Open 5:e006812 Rautanen A et al (2015) Genome-wide association study of survival from sepsis due to pneumonia: an observational cohort study. Lancet Respiratory Med 3:53–60 Cerezo M et al (2025) The NHGRI-EBI GWAS Catalog: standards for reusability, sustainability and diversity. Nucleic Acids Res 53:D998–D1005 Vujkovic M et al (2020) Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1.4 million participants in a multi-ancestry meta-analysis. Nat Genet 52:680–691 Suzuki K et al (2024) Genetic drivers of heterogeneity in type 2 diabetes pathophysiology. Nature 627:347–357 Henney AE et al (2024) Comparative estimate of glucose-lowering therapies on risk of incident pneumonia and severe sepsis: an analysis of real-world cohort data. Thorax 80:32–41 Fukushima K, Kitamura S, Tsuji K, Wada J (2021) Sodium-Glucose Cotransporter 2 Inhibitors Work as a Regulator of Autophagic Activity in Overnutrition Diseases. Front Pharmacol 12 Zhu Z et al (2016) Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet 48:481–487 Gerriets VA et al (2016) Leptin directly promotes T-cell glycolytic metabolism to drive effector T-cell differentiation in a mouse model of autoimmunity. Eur J Immunol 46:1970–1983 Lord GM et al (1998) Leptin modulates the T-cell immune response and reverses starvation-induced immunosuppression. Nature 394:897–901 Holsinger KE, Weir BS (2009) Genetics in geographically structured populations: defining, estimating and interpreting F(ST). Nat Rev Genet 10:639–650 Prohaska A et al (2019) Human Disease Variation in the Light of Population Genomics. Cell 177:115–131 Chen L et al (2016) Genetic Drivers of Epigenetic and Transcriptional Variation in Human Immune Cells. Cell 167:1398–1414e24 Goudet J (2005) HIERFSTAT, a package for R to compute and test hierarchical F-statistics. Mol Ecol Notes 5:184–186 Wang X et al (2015) Neutrophil CD64 expression as a diagnostic marker for sepsis in adult patients: a meta-analysis. Crit Care 19 Meghraoui-Kheddar A et al (2022) Two New Neutrophil Subsets Define a Discriminating Sepsis Signature. Am J Respir Crit Care Med 205:46–59 Shi LZ et al (2011) HIF1alpha-dependent glycolytic pathway orchestrates a metabolic checkpoint for the differentiation of TH17 and Treg cells. J Exp Med 208:1367–1376 Martin AR et al (2017) Rapamycin-mediated mTOR inhibition uncouples HIV-1 latency reversal from cytokine-associated toxicity. J Clin Invest 127:651–656 Hiemer S et al (2019) Integrated Metabolomic and Transcriptomic Profiling Reveals Novel Activation-Induced Metabolic Networks in Human T cells. bioRxiv, 635789 Blaschke K et al (2013) Vitamin C induces Tet-dependent DNA demethylation and a blastocyst-like state in ES cells. Nature 500:222– Consortium EP (2012) An integrated encyclopedia of DNA elements in the human genome. Nature 489:57–74 Gennert DG et al (2021) Dynamic chromatin regulatory landscape of human CAR T cell exhaustion. Proc Natl Acad Sci U S A 118 Soskic B et al (2022) Immune disease risk variants regulate gene expression dynamics during CD4(+) T cell activation. Nat Genet 54:817–826 Marina RJ et al (2016) TET-catalyzed oxidation of intragenic 5-methylcytosine regulates CTCF-dependent alternative splicing. EMBO J 35:335–355 de Brabander J et al (2025) The blood transcriptional response in patients developing intensive care unit-acquired pneumonia. Eur Respir J Roquilly A et al (2023) Interferon gamma-1b for the prevention of hospital-acquired pneumonia in critically ill patients: a phase 2, placebo-controlled randomized clinical trial. Intensive Care Med 49:530–544 Wudhikarn K, Perales MA (2022) Infectious complications, immune reconstitution, and infection prophylaxis after CD19 chimeric antigen receptor T-cell therapy. Bone Marrow Transpl 57:1477–1488 Nian Z et al (2021) Rapamycin Pretreatment Rescues the Bone Marrow AML Cell Elimination Capacity of CAR-T Cells. Clin Cancer Res 27:6026–6038 Parry A, Rulands S, Reik W (2021) Active turnover of DNA methylation during cell fate decisions. Nat Rev Genet 22:59–66 Chait A, den Hartigh LJ (2020) Adipose Tissue Distribution, Inflammation and Its Metabolic Consequences, Including Diabetes and Cardiovascular Disease. Front Cardiovasc Med 7:22 Winer S et al (2009) Normalization of obesity-associated insulin resistance through immunotherapy. Nat Med 15:921–929 Rocha VZ et al (2008) Interferon-gamma, a Th1 cytokine, regulates fat inflammation: a role for adaptive immunity in obesity. Circ Res 103:467–476 Zhou K, Lansang MC et al (2000) Diabetes Mellitus and Infection. in Endotext (eds. Feingold, K.R. (South Dartmouth (MA) Kornum JB et al (2007) Type 2 diabetes and pneumonia outcomes: a population-based cohort study. Diabetes Care 30:2251–2257 Angriman F et al (2024) Preexisting Diabetes Mellitus and All-Cause Mortality in Adult Patients With Sepsis: A Population-Based Cohort Study. Crit Care Explor 6:e1085 Lamontagne F et al (2022) Intravenous Vitamin C in Adults with Sepsis in the Intensive Care Unit. N Engl J Med 386:2387–2398 Wacker DA et al (2022) Evaluating Vitamin C in Septic Shock: A Randomized Controlled Trial of Vitamin C Monotherapy. Crit Care Med 50:e458–e467 Sato R et al (2021) Effect of IV High-Dose Vitamin C on Mortality in Patients With Sepsis: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. Crit Care Med 49:2121–2130 Patel JJ et al (2022) IV Vitamin C in Critically Ill Patients: A Systematic Review and Meta-Analysis. Crit Care Med 50:e304–e312 Chen LL et al (2022) Itaconate inhibits TET DNA dioxygenases to dampen inflammatory responses. Nat Cell Biol 24:353–363 Kuznetsova A, Brockhoff PB, Christensen RHB (2017) lmerTest Package: Tests in Linear Mixed Effects Models. J Stat Softw 82:1–26 Newman AM et al (2019) Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat Biotechnol 37:773–782 Taylor DL et al (2019) Integrative analysis of gene expression, DNA methylation, physiological traits, and genetic variation in human skeletal muscle. Proc Natl Acad Sci U S A 116:10883–10888 Buil A et al (2015) Gene-gene and gene-environment interactions detected by transcriptome sequence analysis in twins. Nat Genet 47:88–91 Burnham KL et al (2017) Shared and Distinct Aspects of the Sepsis Transcriptomic Response to Fecal Peritonitis and Pneumonia. Am J Respir Crit Care Med 196:328–339 Sudlow C et al (2015) UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 12:e1001779 Bycroft C et al (2018) The UK Biobank resource with deep phenotyping and genomic data. Nature 562:203–209 Karp-Tatham E, O'Neill CR, Knight JC, Mentzer AJ, Chong AY (2024) Letter to the Editor: Lack of association between classical HLA genes and asymptomatic SARS-CoV-2 infection. HGG Adv, 100382 Corces MR et al (2017) An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues. Nat Methods 14:959–962 Brown AC et al (2023) Comprehensive epigenomic profiling reveals the extent of disease-specific chromatin states and informs target discovery in ankylosing spondylitis. Cell Genom 3:100306 Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25:1754–1760 McKenna A, Shendure J (2018) FlashFry: a fast and flexible tool for large-scale CRISPR target design. BMC Biol 16:74 Additional Declarations Yes there is potential Competing Interest. JCK reports a grant to his institution from the Danaher Beacon Programme for work on RNA biomarker point-of-care test development in sepsis for endotype assignment which includes support for KC-G and JCK. All remaining authors declare that they have no competing interests. Supplementary Files SuppFig.docx Cite Share Download PDF Status: Published Journal Publication published 25 Feb, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6457289","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":448138516,"identity":"54b33bb2-e3f4-426f-85f9-1259fbc85d7b","order_by":0,"name":"Ping Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCklEQVRIie2RsUoDQRBA5zg4m9Vt9xD0FyYc3OUw4q+sBLS1THkQsNoPyEek2CqQ7mQKm0jag7MQhGAjbCUnXuFeAuk2uVJwXzEMzDxmhgHweP4i4TaWXWbDw3BfwB5KJG2f6KHAXmHYT8lOTukD2tfLjM+/hLEKFzIwDVDiUvLp2d0Q2GawnH0u4plVYlWGsQJKXQoSSxEEBbp6WdTMKvhcwDkAjQ4rSDe6Wm3qtlMIwp8jSvIGkm71WkU17KZE3RTnYvmUpSBLGusqyr4VCharp8dc4b3z/IyvEmNautZresdmMrrgYkxVM7kaFK7F7AuF7LJdBAYQFAcfaUuh2Wa8dHd5PB7P/+YXqEhTtwdFUHgAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0001-7063-7769","institution":"University of Oxford","correspondingAuthor":true,"prefix":"","firstName":"Ping","middleName":"","lastName":"Zhang","suffix":""},{"id":448138517,"identity":"4be23d0e-f90f-498c-b05e-a6c9a492b142","order_by":1,"name":"Patrick MacLean","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Patrick","middleName":"","lastName":"MacLean","suffix":""},{"id":448138518,"identity":"e3ad31fa-9eee-43e1-9652-46327cacdd42","order_by":2,"name":"Alicia Jia","email":"","orcid":"https://orcid.org/0000-0002-0619-7313","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Alicia","middleName":"","lastName":"Jia","suffix":""},{"id":448138519,"identity":"1544e580-a56b-4132-a262-c937c5321f1f","order_by":3,"name":"Callum O'Neill","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Callum","middleName":"","lastName":"O'Neill","suffix":""},{"id":448138520,"identity":"4d682bc8-a31e-4243-b7de-dfdc8904e9a2","order_by":4,"name":"Alice Allcock","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Alice","middleName":"","lastName":"Allcock","suffix":""},{"id":448138521,"identity":"8c2dc308-dd5b-4575-9cf1-fb64fce4929f","order_by":5,"name":"Ethan Prince","email":"","orcid":"https://orcid.org/0009-0009-8928-1622","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Ethan","middleName":"","lastName":"Prince","suffix":""},{"id":448138522,"identity":"9b0871ac-3d14-4ddf-8206-35fb42e382ac","order_by":6,"name":"Imogen Dyne","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Imogen","middleName":"","lastName":"Dyne","suffix":""},{"id":448138523,"identity":"8449b112-bfc8-4a7c-8489-d130f2fb2b64","order_by":7,"name":"Kiki Cano-Gamez","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Kiki","middleName":"","lastName":"Cano-Gamez","suffix":""},{"id":448138524,"identity":"90f11635-caa1-49eb-991d-829ede68e005","order_by":8,"name":"Hanyu Qin","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Hanyu","middleName":"","lastName":"Qin","suffix":""},{"id":448138525,"identity":"955ae45d-33f2-41eb-8180-b65a1cd11087","order_by":9,"name":"Chloe Wainwright","email":"","orcid":"https://orcid.org/0009-0003-4196-2433","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Chloe","middleName":"","lastName":"Wainwright","suffix":""},{"id":448138526,"identity":"c7f734fe-d05b-4f90-a41b-88b12e6a8bd6","order_by":10,"name":"Giuseppe Scozzafava","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Giuseppe","middleName":"","lastName":"Scozzafava","suffix":""},{"id":448138527,"identity":"262453b8-e6f0-4ee6-a5d9-4a380ba69416","order_by":11,"name":"Andrew Brown","email":"","orcid":"https://orcid.org/0000-0002-4951-3056","institution":"Centre for Human Genetics, University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"Brown","suffix":""},{"id":448138528,"identity":"4bbe91cd-d5ae-4015-9861-69611a525831","order_by":12,"name":"James Davies","email":"","orcid":"https://orcid.org/0000-0002-4108-4357","institution":"Weatherall Institute of Molecular Medicine","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"Davies","suffix":""},{"id":448138529,"identity":"7bc8ecad-4682-44d1-9fe7-63b00c28c345","order_by":13,"name":"Amanda Chong","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Amanda","middleName":"","lastName":"Chong","suffix":""},{"id":448138530,"identity":"ab646f3f-39d6-4915-a4e9-11de4b3836cb","order_by":14,"name":"Alexander Mentzer","email":"","orcid":"https://orcid.org/0000-0002-4502-2209","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Mentzer","suffix":""},{"id":448138531,"identity":"3369ef64-f706-4141-817c-a2e6320d54c6","order_by":15,"name":"Katie Burnham","email":"","orcid":"https://orcid.org/0000-0001-8680-2933","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Katie","middleName":"","lastName":"Burnham","suffix":""},{"id":448138532,"identity":"e17dde04-6bdb-4bc7-9202-fd75af6585be","order_by":16,"name":"Emma Davenport","email":"","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Emma","middleName":"","lastName":"Davenport","suffix":""},{"id":448138533,"identity":"f59f6ff0-6d0e-4f8a-8932-8232ff64e5d1","order_by":17,"name":"Julian Knight","email":"","orcid":"https://orcid.org/0000-0002-0377-5536","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Julian","middleName":"","lastName":"Knight","suffix":""}],"badges":[],"createdAt":"2025-04-15 18:15:50","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6457289/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6457289/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-026-69919-7","type":"published","date":"2026-02-25T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":83438478,"identity":"4b78353e-645e-415d-a5ed-a5da8e0b22db","added_by":"auto","created_at":"2025-05-26 09:00:52","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":609418,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe intronic \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eMTOR\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e eQTL interacts with SRS endotypes and shows inverted association between neutrophils and activated T cells. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Regional plot illustrating \u003cem\u003eMTOR\u003c/em\u003e lead eQTL variants (in red) identified in whole blood of sepsis patients. (\u003cstrong\u003eb\u003c/strong\u003e) The interaction effects of SRS endotypes on the \u003cem\u003eMTOR\u003c/em\u003e eQTL association in whole blood. (\u003cstrong\u003ec\u003c/strong\u003e) Volcano plot for the differential cell abundance in SRS1 sepsis patients relative to the non-SRS1 patients from the GainS cohort. The absolute proportion of each cell type (CIBERSORTx absolute scores) was obtained from deconvolution of bulk RNA-seq using a single-cell RNA-seq reference panel derived from sepsis patients \u003csup\u003e8\u003c/sup\u003e. \u003cem\u003eP\u003c/em\u003e value was calculated by linear regression adjusted for age and sex. The horizontal dashed line represents the Bonferroni-corrected \u003cem\u003eP\u003c/em\u003e value 0.01. (\u003cstrong\u003ed\u003c/strong\u003e) The interaction effects of NLR on the \u003cem\u003eMTOR\u003c/em\u003e eQTL association in whole blood. (\u003cstrong\u003ee\u003c/strong\u003e) Colocalisation of the \u003cem\u003eMTOR\u003c/em\u003e eQTL association from sepsis with 127 datasets derived from normal tissue/cell types with different treatment conditions. Summary statistics for the eQTL datasets were retrieved from \u003cem\u003eeQTL Catalogue\u003c/em\u003e\u003csup\u003e32\u003c/sup\u003e\u003cem\u003e.\u003c/em\u003e Posterior probabilities of shared (PP.H4) or distinct (PP.H3) causal variants were computed using R package \u003cem\u003ecoloc\u003c/em\u003e\u003csup\u003e34\u003c/sup\u003e. (\u003cstrong\u003ef\u003c/strong\u003e) Effect size and nominal p value for the lead eQTL across the top 11 colocalised datasets as highlighted in (e) in orange with PP4 ³0.95. (\u003cstrong\u003eg-h\u003c/strong\u003e) Regional association plots for the \u003cem\u003eMTOR\u003c/em\u003e eQTL in sepsis whole blood, neutrophils, activated T cells and naïve T cells derived from healthy donors. The lead eQTL variants identified in sepsis are highlighted in red. PP4 was calculated using \u003cem\u003ecoloc\u003c/em\u003e (see \u003cstrong\u003eMethods\u003c/strong\u003e). See also Fig. S1.\u003c/p\u003e","description":"","filename":"Picture1.png","url":"https://assets-eu.researchsquare.com/files/rs-6457289/v1/4c383bcc584520fb0b54bfd6.png"},{"id":83438479,"identity":"e23a8b29-8d14-408e-8fc0-7a23d25bd8d9","added_by":"auto","created_at":"2025-05-26 09:00:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":667650,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe G-allele associated with reduced T cell activity is protective for survival during sepsis and reduces type 2 diabetes risk. \u003c/strong\u003e(\u003cstrong\u003ea-c\u003c/strong\u003e) Kaplan–Meier curves for 28-day mortality in sepsis CAP patients of GAinS (a; n=737), SI (b; n=102) and UK Biobank (c; n=1,125) cohorts carrying different alleles of the \u003cem\u003eMTOR\u003c/em\u003e eQTL rs4845987. \u003cem\u003eP\u003c/em\u003e value was calculated using the cox regression adjusted for age and gender. (\u003cstrong\u003ed\u003c/strong\u003e) Forest plot showing 28-day mortality for GAinS or GenOSept cohort patients carrying the G alleles, stratified according to SRS status, NLR, or FiO2 levels. Beta coefficient and \u003cem\u003eP \u003c/em\u003evalue were calculated using additive logistic regression, adjusting for age, gender and the first seven genotype principal components from Europeans. (\u003cstrong\u003ee\u003c/strong\u003e) Forest plot showing 28-day mortality in the UKB pneumonia cohort, stratified by immunosuppressed status (IS). Beta coefficients and P values were calculated as described above. (\u003cstrong\u003ef-h\u003c/strong\u003e) Dot plots showing the effect sizes of SNPs used for the SMR/HEIDI analysis from T2D GWAS \u003csup\u003e41\u003c/sup\u003e on the y axis, against \u0026nbsp;eQTLs in activated CD8+ (f), activated CD4+ T cells (g) \u003csup\u003e20\u003c/sup\u003e and neutrophils (h) \u003csup\u003e49\u003c/sup\u003e on the x axis.\u0026nbsp; Error bars represent the standard errors of the SNP effects. (\u003cstrong\u003ei\u003c/strong\u003e) Summary plot illustrating the pleiotropic effects of rs4845987-G on reducing T2D risk and sepsis mortality via its allelic effect on \u003cem\u003eMTOR\u003c/em\u003e expression. See also Fig.S2-3.\u003c/p\u003e","description":"","filename":"Picture2.png","url":"https://assets-eu.researchsquare.com/files/rs-6457289/v1/b0c0e2f34e45dbf29bf281c3.png"},{"id":83438480,"identity":"cf86c3bd-8352-48fe-9d65-9309ec5593b6","added_by":"auto","created_at":"2025-05-26 09:00:52","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":359000,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNeutrophil activation induced by T cells was dampened under hypoxia and by mTOR inhibition.\u003c/strong\u003e (\u003cstrong\u003ea\u003c/strong\u003e) Schematic of ex vivo co-culture for CD66b+ neutrophil and allogeneic T cells from healthy donors in the absence or presence of anti-CD3/CD28 Dynabeads. (\u003cstrong\u003eb\u003c/strong\u003e) Flow cytometry analysis showing the expression of sepsis surface makers on neutrophils derived from sepsis patients and healthy volunteers (HV). (\u003cstrong\u003ec\u003c/strong\u003e) Box plots showing the quantification of proteins by flow cytometry. Each dot represents an individual (n=12 sepsis patients; n=3 Conv.; n=10 HVs). Conv.: convalescent. \u003cem\u003eP \u003c/em\u003evalue was calculated by two-tailed t-test. *P\u0026lt;0.05; **P\u0026lt;0.01;***P\u0026lt;0.001;****P\u0026lt;0.0001. (\u003cstrong\u003ed-e\u003c/strong\u003e) Box plots showing the expression of markers on sepsis neutrophils (n=4) co-cultured with activated T cells in the presence of rapamycin (d), or under hypoxia (1.5% O\u003csub\u003e2\u003c/sub\u003e) vs normoxia condition (e). \u003cem\u003eP\u003c/em\u003e value was calculated by two-tailed paired t-test. See also Fig. S4-5.\u003c/p\u003e","description":"","filename":"Picture3.png","url":"https://assets-eu.researchsquare.com/files/rs-6457289/v1/b06a5ebe3801e69dc052e526.png"},{"id":83439188,"identity":"502aa11c-3bdd-424d-9e86-6da67d11cbfd","added_by":"auto","created_at":"2025-05-26 09:08:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":588273,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNeutrophil activation induced by T cells was dampened under hypoxia and by mTOR inhibition. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) representative flow cytometry panels showing the inhibition of T cell activation (CD69 and PD-1 expression) in the presence of sepsis neutrophils, but not with conditioned neutrophil media cultured for 2 days. Bar plot (right panel) for the number of activated T cells co-cultured with sepsis neutrophils compared to conditioned neutrophil media. n=5. (\u003cstrong\u003eb\u003c/strong\u003e) Flow cytometry result showing the activation of sepsis neutrophils (CD123 and CD64 expression) by activated CD8+ T cells. Bar plot (right panel) for the number of activated neutrophils co-cultured with activated T cells compared to conditioned T cell media. n=5. (\u003cstrong\u003ec\u003c/strong\u003e) Interplay between sepsis neutrophils and allogeneic T cell upon TCR activation. (\u003cstrong\u003ed\u003c/strong\u003e) Volcano plot showing differentially expressed (DE) genes upon TCR activation in CD4+ T cells derived from three healthy volunteers (left panel). DE genes (FDR\u0026lt;0.05, Fold change \u0026gt;2) encoding cytokines annotated by UniProtKB (n=189: searched using Cytokine (KW-0202) for ones reviewed by Swiss-Prot) were highlighted in purple. (\u003cstrong\u003ee\u003c/strong\u003e) Venn diagram showing the overlap of DE cytokines between CD4+ and CD8+ T cells. (\u003cstrong\u003ef\u003c/strong\u003e) Correlation of log2 fold change of DE cytokines identified in CD4+ and CD8+ T cells upon activation. Pearson’s r and \u003cem\u003ep\u003c/em\u003e values are shown. (\u003cstrong\u003eg\u003c/strong\u003e) Volcano plot showing differentially expressed genes upon rapamycin (100nM) treatment in activated T cells with anti-CD3/CD28 beads. RNA-seq raw data was downloaded from GSE129829 \u003csup\u003e55\u003c/sup\u003e. See also Fig. S6.\u003c/p\u003e","description":"","filename":"Picture4.png","url":"https://assets-eu.researchsquare.com/files/rs-6457289/v1/57f1768b5c821b48c4a00f64.png"},{"id":83438485,"identity":"3ca6c7d5-a28a-4e8f-9484-d8b56f6ea0d7","added_by":"auto","created_at":"2025-05-26 09:00:53","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":692712,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRegulatory mechanisms underlying the \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eMTOR\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e eQTL in activated T cells.\u003c/strong\u003e (\u003cstrong\u003ea\u003c/strong\u003e) ATAC-seq showing an \u003cem\u003eMTOR\u003c/em\u003e eQTL resides in a differential open chromatin (in grey) of T cells following anti-CD3/CD28 stimulation. (\u003cstrong\u003eb-c\u003c/strong\u003e). Quantifications of the ATAC peak (\u003cstrong\u003eb\u003c/strong\u003e) (highlighted in a) and (\u003cstrong\u003ec\u003c/strong\u003e) the expression of its proximal gene \u003cem\u003eMTOR\u003c/em\u003e in activated T cells relative to resting T cells (n=6 from 3 healthy volunteers). Adjusted p values were calculated using \u003cem\u003eglm.nb \u003c/em\u003eand\u003cem\u003e Wald's test \u003c/em\u003eas implemented in \u003cem\u003eDESeq2\u003c/em\u003e. (\u003cstrong\u003ed\u003c/strong\u003e) \u003cem\u003eMTOR\u003c/em\u003e transcripts measured using qRT-PCR normalised to β-Actin in activated CD4+ T cells (anti-CD3/CD28 Dynabeads for 3 days) relative to resting T cells (2\u003csup\u003e–∆∆Ct\u003c/sup\u003e) extracted from healthy donors (n=8). (\u003cstrong\u003ee\u003c/strong\u003e)\u003cem\u003e \u003c/em\u003eEpigenetic status surrounding the \u003cem\u003eMTOR\u003c/em\u003e eQTL locus in T cells following anti-CD3/CD28 activation. ChIP-seq for histone modifications and CTCF binding were downloaded from the Encyclopedia of DNA Elements (ENCODE) project \u003csup\u003e57\u003c/sup\u003e (see \u003cstrong\u003eTable S7\u003c/strong\u003e). \u003cem\u003eMTOR\u003c/em\u003e lead eQTLs are represented by red bars (see also \u003cstrong\u003eFig. S1a-b\u003c/strong\u003e and \u003cstrong\u003eFig. S8a \u003c/strong\u003efor the full locus). (\u003cstrong\u003ef\u003c/strong\u003e) \u003cem\u003eMTOR\u003c/em\u003e transcripts in CD4+ T cells treated with vitamin C (200 uM for 48h) relative to the vehicle control. Each dot represents an individual healthy donor (n=8). (\u003cstrong\u003eg\u003c/strong\u003e) Schematic of CRISPR/dCas9-mediated epigenetic gene activation and the delivery strategy in primary T cells. (\u003cstrong\u003eh\u003c/strong\u003e) \u003cem\u003eMTOR\u003c/em\u003e transcripts in CRISPRa (dCas9-VP64) edited CD4+ T cells with an sgRNAs targeting the variant-containing region (sgRNA-e; highlighted in e), or sgRNAs targeting the surrounding regions (sgRNA-c1 and -c2) compared to a non-target sgRNA control (sgRNA-nc) (see \u003cstrong\u003eTable S8\u003c/strong\u003e for sgRNA sequences). Error bars represent SEM of 4 independent replicates from 2 healthy donors. P value was calculated by two-tailed t-test. (\u003cstrong\u003ei\u003c/strong\u003e) Model for the enhancer activity in resting T cells, and the allelic difference in \u003cem\u003eMTOR\u003c/em\u003e expression in activated T cells. See also \u003cstrong\u003eFig. S7-8\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"Picture5.png","url":"https://assets-eu.researchsquare.com/files/rs-6457289/v1/a08a6d0d49212804d0341c25.png"},{"id":106389899,"identity":"9daf6a83-1e3b-4f83-903a-646e07d887a7","added_by":"auto","created_at":"2026-04-08 07:07:53","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4747423,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6457289/v1/52cbc1cb-b3ff-4cd2-9001-889d4132e8a7.pdf"},{"id":83439191,"identity":"dbd0161e-8a80-4088-8101-21334bfad96d","added_by":"auto","created_at":"2025-05-26 09:08:53","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3969910,"visible":true,"origin":"","legend":"","description":"","filename":"SuppFig.docx","url":"https://assets-eu.researchsquare.com/files/rs-6457289/v1/45889f0193e384f8b2faa952.docx"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nJCK reports a grant to his institution from the Danaher Beacon Programme for work on RNA biomarker point-of-care test development in sepsis for endotype assignment which includes support for KC-G and JCK. All remaining authors declare that they have no competing interests.","formattedTitle":"Context-specific regulatory genetic variation in MTOR dampens neutrophil-T cell crosstalk in sepsis, modulating disease","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSepsis is a heterogeneous clinical syndrome characterised by life-threatening organ dysfunction associated with a dysregulated host response to infection. This includes, in a patient and disease state-specific manner, decreased lymphopoiesis and adaptive immunity, sustained inflammation, and myeloid pathologies\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. The mechanistic basis for these diverse features and individual risk, which impact not only acute illness but also poor long-term clinical outcomes in survivors, remains unresolved\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. To progress the goal of tackling heterogeneity in sepsis and advance personalised stratification strategies for successful targeted therapeutics, subtypes of sepsis defined by clinical features and white blood cell gene expression have been reported\u003csup\u003e\u003cspan additionalcitationids=\"CR4 CR5\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. For example, a specific transcriptomic sepsis response signature, SRS1, identifies a subset of patients with more severe disease involving T cell exhaustion, immunosuppressive neutrophils, hypoxia and glycolysis networks\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, informing a more stratified approach to understanding specific maladaptive sepsis responses\u003csup\u003e\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe dysregulated host responses in sepsis span innate and adaptive immunity, and the neutrophil-to-lymphocyte ratio (NLR) has been proposed as a biomarker that reflects this interplay. Dysregulation of NLR is observed in many specific infectious diseases, as well as autoimmune disorders and cancer, providing insights into susceptibility and pathogenesis\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. A high neutrophil-to-T cell ratio is correlated with poor prognosis and increased mortality in sepsis\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e and COVID-19 patients\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. As one of the critical modulators of adaptive immunity, neutrophils can regulate the adaptive immune response either by releasing antimicrobial peptides and cytokines, or by direct cell-surface interactions\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. The complex role of neutrophil subsets in immunostimulation versus immunosuppression is not fully understood. Antigen-presenting aged neutrophils (APANs) drive CD4\u0026thinsp;+\u0026thinsp;T cell proliferation, exacerbating inflammation in sepsis\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, while immunosuppressive neutrophils inhibit the proliferation and activation of CD4\u0026thinsp;+\u0026thinsp;T cells and are prone to detrimental NETosis during severe sepsis\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. However, the reciprocal interaction between neutrophils and T cells, and the role of neutrophils as potential effectors in the context of sepsis, remain incompletely defined.\u003c/p\u003e \u003cp\u003eRecent studies have identified human expression quantitative trait loci (eQTLs) as important genetic contributors to context-dependent phenotypic variation\u003csup\u003e\u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. While most currently reported eQTLs exhibit consistent effects across various contexts, those associated with disease may display cell-type specificity and can be linked to cellular dysregulation\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. The knowledge gained from emerging evidence revealing differences in chromatin accessibility, histone modifications and transcription factor binding across cell types can be harnessed to prioritise causal eQTLs and potential drug targets\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. However, understanding the mechanisms by which eQTLs affect gene expression, and how the cellular environment interacts to create cell-type specific effects, remains a challenge. Additionally, eQTL can exhibit inverted effects, a given genetic variant being associated with opposing effects on gene expression dependent on cell type\u003csup\u003e\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e, which further emphasises the complexity of gene regulation at eQTL, the potential to impact cell-to-cell interactions, and the development of highly heterogeneous diseases such as sepsis.\u003c/p\u003e \u003cp\u003eDriven by the potential contribution of genetic factors to the observed cellular dysregulation in sepsis\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, here we investigate the mechanistic basis of a context-specific eQTL involving \u003cem\u003eMTOR\u003c/em\u003e. MTOR is a key metabolic regulator, controlling a signalling hub that integrates immune receptor and metabolic signals including the switch to pro-inflammatory glycolysis in sepsis\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. We have previously reported that \u003cem\u003eMTOR\u003c/em\u003e is downregulated in SRS1 patients, and that there was an eQTL at this locus whose effect was specific to patients with the SRS1 endotype\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. We now map this association to a genetic variant that we find residing in a context-specific regulatory element, and show that this has an inverted effect on \u003cem\u003eMTOR\u003c/em\u003e expression between activated T cells and neutrophils, decreasing and increasing expression respectively. We demonstrate how an eQTL can modulate cell-to-cell communication, here in sepsis through a negative feedback loop between activated T cells and detrimental neutrophils, mediated by mTOR activity and hypoxia. The G-allele of this variant is associated with reduced mTOR signalling and cytokine release in T cells, leading to improved survival for sepsis patients. Furthermore, we characterise the epigenetic mechanisms regulating \u003cem\u003eMTOR\u003c/em\u003e transcription in resting and activated T cells, providing insights into the survival of memory T cells and the detrimental effects of T cell activation during sepsis.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003e1. A sepsis\u003c/strong\u003e \u003cstrong\u003eMTOR\u003c/strong\u003e \u003cstrong\u003eeQTL involves directional association between neutrophils and activated T cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe first sought to understand the cellular context and specificity of the \u003cem\u003eMTOR\u003c/em\u003e eQTL showing association dependent on sepsis endotype\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. We fine mapped the association in patients from the UK Genomic Advances in Sepsis (GAinS) cohort using total white blood cell bulk RNA-seq and SNP genotyping (n\u0026thinsp;=\u0026thinsp;823 samples from 638 sepsis patients)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. This localised the eQTL to a haplotype block with 25 variants in strong linkage disequilibrium (LD) (r\u0026sup2; \u0026gt; 0.95), tagged by rs4845987 (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ea and \u003cstrong\u003eFig. S1a-b\u003c/strong\u003e). The minor G-allele of rs4845987 is associated with increased expression of \u003cem\u003eMTOR\u003c/em\u003e in whole blood and this is magnified in SRS1 patients compared to non-SRS1 patients (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eb).\u003c/p\u003e\n\u003cp\u003eTo explore how this observation of endotype specificity may relate to the cellular specificity of the \u003cem\u003eMTOR\u003c/em\u003e eQTL, we performed cell-type deconvolution of the total white blood cell bulk RNA-seq. This showed differences specific to SRS endotype involving both innate and adaptive cell populations, specifically neutrophil and T cells subsets (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ec). We therefore considered the relationship of the eQTL with NLR, noting that higher NLR based on hospital laboratory leukocyte differential counts from our previously published cohorts\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e shows association with more severe all cause sepsis, COVID-19 and post-operative pneumonia (\u003cstrong\u003eFig. S1c-e\u003c/strong\u003e); and a positive correlation with the quantitative transcriptomic sepsis response score (SRSq) for the likelihood of SRS1 endotype\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e (\u003cstrong\u003eFig. S1f\u003c/strong\u003e). We found that the \u003cem\u003eMTOR\u003c/em\u003e eQTL tagged by rs4845987 shows interaction with NLR, with increased expression of \u003cem\u003eMTOR\u003c/em\u003e for the G allele in patients with heightened NLR (compared to those with lower NLR) (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ed). We then asked how often this relationship may be observed. From our whole-blood sepsis eQTL data\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, we found that 565 out of 12,726 genome-wide independent eQTL effects in sepsis show interaction with NLR (FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.01, \u003cstrong\u003eTable S1\u003c/strong\u003e), and 65% (365 out of 565) of these also showed significant interaction (FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.01) with SRS endotype (\u003cstrong\u003eFig. S1g\u003c/strong\u003e), with strong concordance between SRS and NLR interaction effects (Pearson\u0026apos;s r\u0026thinsp;=\u0026thinsp;0.95; \u003cstrong\u003eFig. S1h\u003c/strong\u003e). This demonstrates that there are likely to be multiple independent genetic modulators of immune-relevant phenotypes dependent on relative abundance of neutrophil and lymphocyte populations, increasing the potential for clinical relevance.\u003c/p\u003e\n\u003cp\u003eTo identify immune cell types and treatment settings linked to the \u003cem\u003eMTOR\u003c/em\u003e association, we next performed colocalisation analysis using data curated by the eQTL Catalogue and GTEx consortium\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Among 127 datasets from 75 tissues/cell types and 14 treatments, 11 different cellular or tissue contexts demonstrated shared genetic associations for the lead eQTL with \u003cem\u003eMTOR\u003c/em\u003e expression, including T cells stimulated with anti-CD3/CD28 beads and neutrophils (PP4\u0026thinsp;\u0026gt;\u0026thinsp;0.95; Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ee \u003cstrong\u003eand\u003c/strong\u003e Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eg; \u003cstrong\u003eTable S2\u003c/strong\u003e), but not resting T cells (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eh). We found that the minor G-allele effect of rs4845987 on \u003cem\u003eMTOR\u003c/em\u003e expression was inverted between activated T cells and neutrophils, decreasing and increasing expression respectively (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ef). This aligns with our interaction analysis showing lower \u003cem\u003eMTOR\u003c/em\u003e expression with possession of the G-allele in non-SRS1 and low NLR patients (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eb and \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ed), where T cell function was less attenuated compared to SRS1 and high NLR patients (phenotypes associated with more severe disease).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2. The\u003c/strong\u003e \u003cstrong\u003eMTOR\u003c/strong\u003e \u003cstrong\u003eallele associated with reduced T cell activity is protective for survival during sepsis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next investigated whether there was evidence of association with outcome in sepsis for the \u003cem\u003eMTOR\u003c/em\u003e eQTL lead SNP. In the UK GAinS cohort, we found that patients with a copy of the minor G-allele exhibited significantly reduced 28-day mortality compared to those with the major C-allele (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0010, HR\u0026thinsp;=\u0026thinsp;0.60 [95% CI 0.44\u0026ndash;0.81]; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea; n\u0026thinsp;=\u0026thinsp;737). This association was restricted to patients with sepsis due to community-acquired pneumonia (CAP) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea), and not observed in non-CAP sepsis patients (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.45; \u003cstrong\u003eFig. S2a\u003c/strong\u003e; n\u0026thinsp;=\u0026thinsp;384), which is consistent with the role of mTOR in hypoxia/oxidative stress and evidence linking mTOR inhibitors to pneumonitis\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn an independent Sepsis Immunomics (SI) cohort\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, we genotyped the sepsis CAP patients for rs4845987 (see \u003cstrong\u003eMethods\u003c/strong\u003e). Consistent with UK GAinS, the G allele was associated with improved sepsis survival (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.031, HR\u0026thinsp;=\u0026thinsp;0.32 [95% CI 0.11\u0026ndash;0.90]; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb; n\u0026thinsp;=\u0026thinsp;102). A similar protective effect was observed in a further independent validation cohort from the UK Biobank for participants who had confirmed bacterial pneumonia (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0041, HR\u0026thinsp;=\u0026thinsp;0.68 [95% CI 0.52\u0026ndash;0.88]; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ec; n\u0026thinsp;=\u0026thinsp;1125; see \u003cstrong\u003eMethods\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eTo further investigate the interplay between cellular features and allelic differences in clinical outcomes, we stratified the GAinS cohort into groups based on SRS status (defined as SRS-latest [assigned based on samples from the latest available time point for each patient]; SRS1-ever [assigned if any SRS1 sample was detected across time points for a given patient]) and NLR levels. We found that in non-SRS1 patients, those with the G-allele have a significantly better survival using logistic regression compared to the SRS1 patient group (OR\u0026thinsp;=\u0026thinsp;0.43, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0012 when considering SRS_latest; OR\u0026thinsp;=\u0026thinsp;0.38, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.5e-04 using SRS1_ever; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed). A stronger protective association of the G-allele was also observed in patient with low NLR compared to high NLR (OR 0.46 vs 0.51; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed). Consistent with the immune paresis nature of the SRS1 endotype, the protective effect of the G-allele in UK Biobank patients (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ec) was observed only in those without an immunosuppressed status (IS) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ee; see Methods).\u003c/p\u003e\n\u003cp\u003eProlonged hypoxia significantly contributes to multi-organ failure, a hallmark of severe sepsis, and is relevant to a potential role for genetic variants of \u003cem\u003eMTOR\u003c/em\u003e. We therefore analysed patients requiring a high fraction of inspired oxygen (FiO\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0.4) or with a low ratio of arterial oxygen tension (PaO\u003csub\u003e2\u003c/sub\u003e) to FiO\u003csub\u003e2\u003c/sub\u003e (0\u0026thinsp;\u0026lt;\u0026thinsp;PaO2/FiO2\u0026thinsp;\u0026le;\u0026thinsp;100) as indicators of hypoxaemia during sepsis\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e36\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. As expected, we observed high FiO\u003csub\u003e2\u003c/sub\u003e and low PaO\u003csub\u003e2\u003c/sub\u003e/FiO\u003csub\u003e2\u003c/sub\u003e were associated with poor survival in the GAinS cohort (HR\u0026thinsp;=\u0026thinsp;2.6 and 3.8 respectively; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; \u003cstrong\u003eFig. S2b\u003c/strong\u003e). Significant association for FiO\u003csub\u003e2\u003c/sub\u003e level, but not PaO\u003csub\u003e2\u003c/sub\u003e/FiO\u003csub\u003e2\u003c/sub\u003e, was also observed in another independent cohort GenOSept\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e (\u003cstrong\u003eFig. S2b\u003c/strong\u003e). Consistently, in patients with low FiO\u003csub\u003e2\u003c/sub\u003e across both cohorts, we observed stronger allelic differences in survival (OR\u0026thinsp;=\u0026thinsp;0.46, P\u0026thinsp;=\u0026thinsp;0.033 for GAinS; OR\u0026thinsp;=\u0026thinsp;0.24, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0071 for GenOSept; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed). We did not observe significant associations between the eQTL and survival in non-CAP patients with or without any stratification (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed yellow and cyan dots; \u003cstrong\u003eFig. S2a\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3. Pleiotropic effects of genetic variant on T2D risk and sepsis survival through modulation of\u003c/strong\u003e \u003cstrong\u003eMTOR\u003c/strong\u003e \u003cstrong\u003eexpression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next further investigated the relationship between the \u003cem\u003eMTOR\u003c/em\u003e sepsis-associated eQTL and other disease-related traits using the GWAS Catalog\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. We found that rs4845987 or SNPs in high LD (r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.8) were significantly associated (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026le;\u0026thinsp;5e-08) with 18 GWAS traits, 11 of which were related to T2D or its associated metabolic traits such as body mass index and insulin resistance (\u003cstrong\u003eTable S3\u003c/strong\u003e). We observed strong colocalisation between the \u003cem\u003eMTOR\u003c/em\u003e eQTLs and risk variants for T2D in two GWAS studies\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e (PP4\u0026thinsp;=\u0026thinsp;0.75 and 0.97 respectively; \u003cstrong\u003eFig. S3a-b\u003c/strong\u003e). The minor G-allele of rs4845987, which is protective for sepsis survival, was also associated with a reduced risk of T2D (\u003cstrong\u003eFig. S3a\u003c/strong\u003e). Individuals with T2D have an increased risk of developing sepsis, and sodium-glucose cotransporter-2 inhibitors used to treat T2D suppress \u003cem\u003eMTOR\u003c/em\u003e and reduce pneumonia risk\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eTo test the potential causative effect of \u003cem\u003eMTOR\u003c/em\u003e expression on T2D risk, we performed a summary data-based Mendelian randomisation (SMR) analysis\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. We focused on the genes within a 2Mbp window around the GWAS risk loci of the \u003cem\u003eMTOR\u003c/em\u003e locus and observed significant effects on increased T2D risk through higher expression of \u003cem\u003eMTOR\u003c/em\u003e in activated T cells (\u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003eSMR\u003c/em\u003e\u003c/sub\u003e = 1.17e-06 and 1.92e-6 in CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;T cells respectively; Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ef-g; \u003cstrong\u003eTable S4;\u003c/strong\u003e see \u003cstrong\u003eMethods\u003c/strong\u003e). In line with the eQTL associations (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ef), we observed inverted effects in neutrophils and adipose tissue (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eh and \u003cstrong\u003eFig. S3c-d\u003c/strong\u003e), both of which are known to interact with T cells in disease settings\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eGiven the disease and immunological associations, we hypothesised that the \u003cem\u003eMTOR\u003c/em\u003e variant would have been under selection pressure. We observed higher frequency of the G-allele in Africans, compared to other populations, with allele frequencies of 0.91, 0.31, 0.21, 0.30, and 0.38 in Africans, Americans, East Asians, Europeans, and South Asians respectively (\u003cstrong\u003eFig. S3e\u003c/strong\u003e), as well as high genetic divergence (measured by fixation index, F\u003csub\u003eST\u003c/sub\u003e\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e) (\u003cstrong\u003eFig. S3f-g\u003c/strong\u003e). This would be consistent with potential functional consequences associated with survival fitness and disease modulating allele frequency following migrations and genetic adaptations within human populations\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e4. Activation of sepsis neutrophils by T cells is dampened by mTOR inhibition and hypoxia\u003c/h2\u003e\n \u003cp\u003eOur findings of a protective G-allele effect in patients with relatively reduced \u003cem\u003eMTOR\u003c/em\u003e expression in activated T cells, and with higher expression in neutrophils, led us to hypothesise a potential role for T cells in driving sepsis-associated neutrophil phenotypes via the mTOR pathway (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ei). To test this hypothesis, we established an \u003cem\u003eex vivo\u003c/em\u003e co-culture model of neutrophils and T cells in which we could compare effects of co-culture with either neutrophils from sepsis patients or healthy donors, with primary T cells that were resting or activated (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea \u003cstrong\u003eand Fig. S4a\u003c/strong\u003e). We assayed three neutrophil surface markers, CD64, CD123, and PD-L1, previously associated with sepsis severity \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e51\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e, as indicators of neutrophil activity. All three markers exhibited upregulation in neutrophils isolated from sepsis patients compared to healthy controls, at both the protein (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb) and mRNA levels (\u003cstrong\u003eFig. S4b\u003c/strong\u003e). The expression levels were further elevated in SRS1 compared to non-SRS1 patients (\u003cstrong\u003eFig. S4b\u003c/strong\u003e). We found that co-culture with activated T cells, relative to resting T cells, resulted in a further increase in marker expression on sepsis neutrophils (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb-c). This upregulation was associated with enhanced NETosis (\u003cstrong\u003eFig. S5a\u003c/strong\u003e). Consistent with the observed eQTL association, both CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;T cells demonstrated similar effects on neutrophil markers (\u003cstrong\u003eFig. S4c\u003c/strong\u003e). However, the immortalised Jurkat T cell line did not affect these markers on sepsis neutrophils compared to primary T cells (\u003cstrong\u003eFig. S4d\u003c/strong\u003e), potentially due to altered T cell function in the Jurkat line. Interestingly, co-culture with activated T cells did not significantly affect CD123 and CD64 expression on neutrophils from healthy donors or convalescent patients compared to resting T cells (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ec), suggesting a specific priming of sepsis neutrophils for hyperactivation by activated T cells.\u003c/p\u003e\n \u003cp\u003eT cell activation is known to be heavily reliant on mTOR and HIF-1\u0026alpha; (hypoxia-inducible factor 1\u0026alpha;) \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, and mTOR inhibition and hypoxia attenuate T cell activity and cytokine release \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e53\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. Consistent with these previous findings, in co-culture assays in the presence of rapamycin, a specific mTOR inhibitor, under hypoxia condition, or with rapamycin pre-treated T cells, we found a decreased effect on sepsis neutrophils (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ed-e \u003cstrong\u003eand Fig. S5b\u003c/strong\u003e. This effect was specific to T cell-neutrophil interactions, as rapamycin treatment did not affect sepsis neutrophils cultured alone (\u003cstrong\u003eFig. S5c\u003c/strong\u003e). Together, these results highlight the critical role of mTOR activity in T cell activation and the consequent deleterious effects on neutrophils in the context of sepsis.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003e5. Reciprocal interactions between activated T cells and sepsis neutrophils\u003c/h3\u003e\n\u003cp\u003ePrevious studies have revealed that neutrophils regulate adaptive immune responses either by releasing secreted peptides and cytokines or by direct cell-surface interactions \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, and neutrophils from SRS1 sepsis patients exhibit a stronger inhibitory effect on T cell proliferation and activation relative to non-SRS1 patients \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. In our co-culture experiments, PD-1 and CD69 expression was significantly suppressed on T cells in the presence of sepsis neutrophils, an effect not observed when using neutrophil-conditioned media alone (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea), suggesting direct cell-surface contact is crucial Conversely, CD123 and CD64 expression was markedly upregulated on sepsis neutrophils when co-cultured with activated T cells or exposed to conditioned T cell media (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb). This upregulation was also observed in a sepsis whole blood culture model (\u003cstrong\u003eFig. S6a-b\u003c/strong\u003e), where T cell activation enhanced neutrophil markers (\u003cstrong\u003eFig. S6c-d\u003c/strong\u003e), and the increase in immunosuppressive neutrophils subsequently inhibited T cell activation (48h vs 24h; \u003cstrong\u003eFig. S6e\u003c/strong\u003e). Together, these results indicate a negative feedback loop between T cell activation and neutrophils in sepsis, mediated by both cytokine release and direct cell-surface interactions (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ec). Disruption of this loop, due to excessive mTOR signalling in T cells, may contribute to poor clinical outcomes.\u003c/p\u003e\n\u003cp\u003eTo prioritise the potential cytokines mediating the cell-cell interactions, we generated RNA-seq datasets in primary CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;T cells following activation with anti-CD3/CD28 beads. As expected, TCR stimulation triggered substantial transcriptomic reprogramming (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ed; \u003cstrong\u003eTable S5\u003c/strong\u003e), with 68 and 59 differentially expressed (DE) genes encoding annotated cytokines identified in CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;cells, respectively. Overall, 81.4% (48 out of 59) of DE cytokines in CD8\u0026thinsp;+\u0026thinsp;T cells overlapped with those in CD4\u0026thinsp;+\u0026thinsp;cells (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ee) and displayed a strong correlation (Pearson\u0026rsquo;s r\u0026thinsp;=\u0026thinsp;0.84; Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ef). Interestingly, 11 of the 48 commonly DE cytokines in both T cell subsets were also rapamycin-sensitive (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eg), including IL-3 (ligand for its receptor CD123) and IFN-\u0026gamma;, a cytokine known to augment NETosis via ligand-receptor interaction, contributing to acute lung injury during sepsis\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Additionally, the transcription of the sepsis surface markers was also upregulated in response to T cell cytokines (\u003cstrong\u003eFig. S5d\u003c/strong\u003e), suggesting a more intricate regulatory network.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6. Regulatory mechanism underlying the\u003c/strong\u003e \u003cstrong\u003eMTOR\u003c/strong\u003e \u003cstrong\u003eeQTL in activated T cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe proceeded to investigate the regulatory genomic landscape of the \u003cem\u003eMTOR\u003c/em\u003e genetic association in order to prioritise and further characterise the \u003cem\u003eMTOR\u003c/em\u003e eQTL. Studies have shown that causal eQTLs and GWAS risk SNPs are more likely to be found in open chromatin\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. We investigated context specific epigenomic datasets and found that the \u003cem\u003eMTOR\u003c/em\u003e eQTL variant rs4845987 is located at the centre of an ATAC-seq peak in intron 8 of \u003cem\u003eMTOR\u003c/em\u003e, a peak identified exclusively in resting T cells (both CD4\u0026thinsp;+\u0026thinsp;and CD8+), but not in other immune cells (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ea \u003cstrong\u003eand Fig. S7a\u003c/strong\u003e). This ATAC peak was completely silenced in activated T cells (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ea-b), coinciding with a specific reduction in \u003cem\u003eMTOR\u003c/em\u003e expression, while the expression of other nearby genes remained unaffected (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ec-d \u003cstrong\u003eand Fig. S8b\u003c/strong\u003e). We also observed a similar pattern of the ATAC peak at this location being present in central memory T cells, but not in naive cells, following TCR activation and CAR T cell production (\u003cstrong\u003eFig. S7b-c\u003c/strong\u003e). Upon stimulation, this locus exhibited coordinated changes in epigenetic features, marked by the absence of enhancer markers H3K27ac and H3K4me1 and presence of hydroxymethylation in activated T cells (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ee \u003cstrong\u003eand Fig. S8c\u003c/strong\u003e). The importance of this specific genomic region was further supported by proximal CTCF-binding and its interaction with the \u003cem\u003eMTOR\u003c/em\u003e promoter (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ee \u003cstrong\u003eand Fig. S8a\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eWe hypothesised that the enhancer plays a critical role in maintaining \u003cem\u003eMTOR\u003c/em\u003e expression and cell survival in resting memory T cells while masking the eQTL effect within the enhancer itself, and that in activated T cells, the eQTL-associated genetic variant may exert its effect through hydroxymethylation, leading to an allelic effect observed in bulk T cells (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ef) and single-cell T cell subsets (\u003cstrong\u003eFig. S7d\u003c/strong\u003e). To experimentally confirm this, we manipulated hydroxymethylation using vitamin C (VC), a known activator of TET enzymes that catalyse demethylation and 5-hydroxymethylcytosine (5hmC)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. As predicted, we observed increased \u003cem\u003eMTOR\u003c/em\u003e expression upon VC in activated T cells with the C/C genotype compared to G allele carriers (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ef).\u003c/p\u003e\n\u003cp\u003eWe next utilised a CRISPR/dCas9-based epigenetic activation system delivered via lentivirus to manipulate this locus and demonstrate direct evidence for enhancer function (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eg; see Methods). We observed a significant upregulation of \u003cem\u003eMTOR\u003c/em\u003e expression in activated T cells with sgRNA targeting the variant-containing regions (sgRNA-e) compared with a non-targeting control sgRNAs, or sgRNAs targeting the flanking regions (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eh). Together, these results demonstrate a unique epigenetic mechanism that ensures precise control of \u003cem\u003eMTOR\u003c/em\u003e expression, which is critical for proper T cell function in immune responses (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ei\u003cstrong\u003e)\u003c/strong\u003e.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we identify a genetic variant located within an enhancer element that regulates \u003cem\u003eMTOR\u003c/em\u003e transcription in T cells. Our findings are consistent with this enhancer sustaining essential baseline mTOR activity in resting T cells while masking the \u003cem\u003eMTOR\u003c/em\u003e eQTL effect. Upon T-cell receptor activation, mTOR is highly upregulated through post-translational modifications, with silencing of this enhancer providing a negative feedback loop to prevent excessive mTOR activity at the transcription level. In activated T cells, the eQTL effect is exposed through hydroxymethylation involving the variant which can be manipulated by VC via activation of TET enzymes. The C allele amplifies mTOR signalling and T cell activity during sepsis, ultimately leading to persistent neutrophil hyperactivation, NETosis, and increased mortality in patients with sepsis due to pneumonia.\u003c/p\u003e \u003cp\u003emTOR is a critical signalling hub that integrates immune receptor and metabolic signals to determine the fate of T cells\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. In this study, we utilised epigenetic profiling to identify a regulatory element that controls \u003cem\u003eMTOR\u003c/em\u003e transcription, distinguishing between resting and activated T cells. This enhancer is essential for memory T cell survival and functions as a negative feedback loop to suppress mTOR activation upon TCR stimulation. Further studies are needed to elucidate key extrinsic modulators by examining epigenetic remodelling of cytokine loci, with a focus on the mTOR pathway and those affecting sepsis-associated neutrophil dysfunction. The deleterious effects of cytokine release during sepsis are supported by recent observations of sustained cytokine signaling upregulation and the potential harmful effects of interferon gamma-1b therapy in pneumonia cases\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e,\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e. Understanding these modulators could also provide insights into whether these factors impact the efficacy of CAR-T cell therapy, which is often complicated by severe infection\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e. Pretreatment of CAR-T cells with rapamycin has shown promise in promoting bone marrow infiltration and enhancing therapeutic efficacy in acute myeloid leukaemia\u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e. This raises the possibility of combining rapamycin with approaches targeting immunosuppressive neutrophils to improve outcomes in sepsis patients.\u003c/p\u003e \u003cp\u003eCytosine modifications, including 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC), are crucial epigenetic mechanisms that regulate chromatin structure and gene transcription. While 5mC is known to be associated with gene repression, 5hmC as an intermediate in demethylation is often linked to gene activation. Recent studies have highlighted the broader implications of 5hmC in cell differentiation and disease\u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e. Our findings revealed a novel regulatory mechanism involving 5hmC at the \u003cem\u003eMTOR\u003c/em\u003e locus, mediating T cell function and vitamin C response. Currently, the analysis of 5mC and 5hmC is primarily conducted at the bulk level, obscuring potential heterogeneity between different cell types and/or subsets. Future studies could leverage single-cell approaches to obtain insights into the distinct 5mC/5hmC dynamics during T cell differentiation and acute infections.\u003c/p\u003e \u003cp\u003eChronic inflammation and metabolic dysfunction in adipose tissue are key drivers of insulin resistance and T2D\u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e. Both CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;T cells infiltrate adipose tissue, with increased interferon-γ-expressing T cells promoting inflammation\u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e,\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e. The adipocyte-secreted hormone leptin directly regulates T cell proliferation, glycolytic metabolism and cytokine production\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. Individuals with T2D have an increased risk of developing infections including sepsis and COVID-19\u003csup\u003e69\u003c/sup\u003e. However, the overall evidence linking T2D to sepsis outcomes remains unclear but appears more robust for pneumonia\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e,\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e. Interestingly, sodium-glucose cotransporter-2 inhibitors used to treat T2D suppress \u003cem\u003eMTOR\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e and reduce pneumonia risk\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Our study demonstrated colocalisation of a genetic variant between adipose tissue, T cells for \u003cem\u003eMTOR\u003c/em\u003e expression and T2D GWAS loci. This tissue-specific variant had opposite allelic effects on \u003cem\u003eMTOR\u003c/em\u003e expression, with the G allele decreasing \u003cem\u003eMTOR\u003c/em\u003e expression in activated T cells while increasing it in adipose tissue. These results suggest a potential role for mTOR signaling in the adipose-T cell axis with implications for the interplay between sepsis and T2D.\u003c/p\u003e \u003cp\u003eVitamin C (VC) supplementation has been proposed as a potential treatment for sepsis due to its anti-inflammatory and antioxidant properties. However, clinical trials have shown mixed results\u003csup\u003e\u003cspan additionalcitationids=\"CR73 CR74\" citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e, potentially due to different patient severities at enrolment. Our results revealed a detrimental effect of mTOR-mediated T cell activity on pneumonia risk. Furthermore, we showed that VC increases \u003cem\u003eMTOR\u003c/em\u003e expression in activated T cells from risk C-allele carriers, providing insights into the potential mechanism of action of harmful VC therapy in sepsis\u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. In contrast to VC, which activates TET enzymes\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e, TET inhibition could potentially reduce T cell-mediated cytokine release, alleviate neutrophil dysfunction and improve patient survival. Indeed, treatment of TET inhibitors reduces 5hmC levels and attenuates LPS-induced pulmonary oedema and lung injury in a sepsis mouse model\u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e. However, the specific role of TET inhibition in modulating T cell-neutrophil crosstalk during human sepsis remains to be seen.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eeQTL interaction and colocalisation analysis\u003c/h2\u003e \u003cp\u003eeQTL interaction was determined using a linear mixed model as implemented in R package lmerTest \u003csup\u003e\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e\u003c/sup\u003e: p\u0026thinsp;~\u0026thinsp;g\u0026thinsp;+\u0026thinsp;i\u0026thinsp;+\u0026thinsp;g:i + (1|donor), where p is the gene expression corrected for population structure and PEER factors as described in\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, g is the genotype vector, i is the term for SRS endotypes or log2 transformed NLR, g:i is the interaction between genotype and either SRS or log2 transformed NLR, and (1|donor) is a random effect accounting for variability between donors. Gene expression and genotype data were obtained from 823 RNA-seq samples derived from 638 sepsis patients, as described by Burnham \u003cem\u003eet al\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Colocalisation analysis was performed in a\u0026thinsp;\u0026plusmn;\u0026thinsp;200-bp window surrounding the \u003cem\u003eMTOR\u003c/em\u003e lead eQTL, using R package coloc \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e with default settings. eQTL summary statistics from healthy tissue and cell types were retrieved from eQTL Catalogue\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Type 2 diabetes GWAS summary statistics were downloaded from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ftp.ncbi.nlm.nih.gov/dbgap/studies/phs001672/analyses/\u003c/span\u003e\u003cspan address=\"https://ftp.ncbi.nlm.nih.gov/dbgap/studies/phs001672/analyses/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003csup\u003e40\u003c/sup\u003e, and \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.diagram-consortium.org/downloads.html\u003c/span\u003e\u003cspan address=\"https://www.diagram-consortium.org/downloads.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003csup\u003e41\u003c/sup\u003e, respectively.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eCell type deconvolution\u003c/h2\u003e \u003cp\u003eAligned bam files derived from sepsis leukocytes were obtained from the GAinS study. RNA-seq count matrix was generated using featureCounts and normalised by DESeq2. As described previously\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, the CIBERSORTx\u003csup\u003e\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e\u003c/sup\u003e absolute scores of each cell type in bulk samples were obtained using the count matrix for the bulk RNA-seq and the signature matrix derived from sepsis single cell RNA-seq\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e with S-mode batch correction and 100 permutations.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSummary data–based Mendelian randomisation\u003c/h3\u003e\n\u003cp\u003eT2D GWAS in Europeans and eQTL summary data were obtained from Suzuki K. \u003cem\u003eet al.\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e and eQTL Catalogue\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, respectively. We performed the SMR analysis\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e using eQTLs as instrumental variables to identify genes whose expression is associated with T2D risk due to pleiotropy and/or causality. Genes were included in the analysis if they had at least one cis-eQTL (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;5e⁻⁸) in either activated CD4+, CD8\u0026thinsp;+\u0026thinsp;T cells\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, neutrophils\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, or adipose/fat tissue\u003csup\u003e\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e,\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e\u003c/sup\u003e within a 2 Mbp window around GWAS loci, following the default settings of the SMR tool (v1.3.1). The HEIDI (heterogeneity in dependent instruments) test was applied to differentiate functional associations from linkage effects. LD correlation between SNPs was estimated using 1000 Genomes Project data for Europeans.\u003c/p\u003e\n\u003ch3\u003eSurvival analysis\u003c/h3\u003e\n\u003cp\u003ePatients with sepsis were recruited from the Genomic Advances in Sepsis (GAinS) and Sepsis Immunomics (SI) studies. The inclusion and exclusion criteria for GAinS and SI studies were described previously \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e\u003c/sup\u003e. Blood samples were obtained on Day 1, 3 and 5 of hospital admission, with follow-up samples taken an average of 3 months after hospital discharge. CAP was diagnosed as a febrile illness with cough, sputum production, breathlessness, leukocytosis and radiological evidence of pneumonia, acquired in the community or within 2 days of hospital admission. To assess the association between genetic variants and 28-day mortality, Cox proportional-hazards model and logistic regression adjusting for age, sex, and the first seven principal components were used. Multivariable Cox regression was performed to evaluate the predictive ability of hypoxia-related factors (FiO2, PaO2:FiO2 ratio) for 28-day mortality while controlling for age (\u0026gt;\u0026thinsp;65 vs\u0026thinsp;\u0026le;\u0026thinsp;65 years) and sex. All statistical analyses were conducted using R (v4.2.1).\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eUK Biobank (UKB) data curation and analysis\u003c/h2\u003e \u003cp\u003eThe UKB recruitment and ethical approval process has been described previously\u003csup\u003e\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e\u003c/sup\u003e. Briefly, half a million men and women aged 40\u0026ndash;69 years attended one of 22 UKB assessment centres located throughout England, Scotland and Wales between 2006 and 2010. All participants completed a touchscreen questionnaire, verbal interview and had a range of physical measurements and blood, urine and saliva samples taken for long-term storage. Genotype data was generated as described previously\u003csup\u003e\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e\u003c/sup\u003e. Withdrawn participants were removed from the available UKB dataset and only individuals who received a diagnosis of bacterial pneumonia based on International Classification of Diseases 10th Revision (ICD10) J15 in any instance of the UKB-specified Field 41270 were included in the analysis (n\u0026thinsp;=\u0026thinsp;1,461). The corresponding date of diagnosis was extracted from the matching instance in Field 41280. For those with multiple diagnoses the sole or earliest date of diagnosis was used as the only reference point for impatient start date. Participant death was determined via the presence of a date in instance 0 of Field 40000. Individuals who died within 28 days of their earliest ever recorded diagnosis of the relevant ICD10 code were classified as a case with the remainder as controls. The time in days between the earliest date of diagnosis and date of death was used as the time to event metric. Age in years was calculated by subtracting the estimated date of birth (using \u0026lsquo;year_of_birth_f34_0_0\u0026rsquo;, \u0026lsquo;month_of_birth_f52_0_0\u0026rsquo;, and day of birth manually assigned as the 15th to avoid regression to the mean) from the date of the relevant ICD10 diagnosis and dividing by 52.143. Self-reported sex in instance 0 of field 31 was used to define males and females. Patients with a genetically determined ancestry of White British Subset (WBS) were included (n\u0026thinsp;=\u0026thinsp;1,125), and the first seven principal components were used for survival analysis as described above. Immunosuppressed status (IS) was defined based on ICD-10 digital codes (Field 41270) as previously described \u003csup\u003e\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e\u003c/sup\u003e, covering chronic viral hepatitis, Human Immunodeficiency virus (HIV) or Human T-cell lymphotropic virus (HTLV) infection, organ transplantation, medication, autoimmunity, blood cancer, and congenital disease.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eSNP genotyping\u003c/h2\u003e \u003cp\u003eWe genotyped the SNP rs4845987 for the Sepsis Immunomics (SI) cohort using samples derived from CAP patients with hospitalisation dates up to 04-12-2024. The inclusion and exclusion criteria for this study were detailed above. Briefly, genomic DNA was extracted from buffy coat samples (stored at -80\u0026deg;C) using the Monarch Genomic DNA Purification Kit (NEB #T3010) following the manufacturer\u0026rsquo;s instruction. gDNA was then quantified using the Qubit dsDNA HS Assay Kit (ThermoFisher). Sample genotypes were determined using TaqMan Genotyping Assays with the Universal PCR Master Mix and a predesigned probe for rs4845987 (C_2524855_10, ThermoFisher) on a CFX-96 C1000 platform (Bio-Rad). For healthy volunteers, purified cells were used for gDNA extraction.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eIsolation and culture of human neutrophils and T cells\u003c/h2\u003e \u003cp\u003ePeripheral blood mononuclear cells (PBMCs) from healthy donor leucocyte cones were isolated by Ficoll-Paque (Sigma) density gradient centrifugation. CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;T cells were then separated from PBMCs by positive selection with magnetic MicroBeads (Miltenyi Biotec) following the manufacturer\u0026rsquo;s instructions. Isolated T cells were frozen in FBS (Sigma, #F7524) with 10% DMSO at 2 x 10\u003csup\u003e7\u003c/sup\u003e cells/ml and stored in liquid nitrogen. Total T cells were maintained in RPMI-1640 (Gibco) medium supplemented with 5% FBS (Sigma, #F7524), 100 mM L-glutamine (Sigma), 1x penicillin-streptomycin (Sigma) and 500 IU/ml human recombinant IL-2 (Biolegend). T cells were activated using anti-CD3/CD28 Dynabeads (ThermoFisher, #11131D) at a 1:1 cell:bead ratio at 2 x 10\u003csup\u003e6\u003c/sup\u003e cells/ml.\u003c/p\u003e \u003cp\u003eNeutrophils were extracted from 2\u0026ndash;10 ml whole blood from sepsis patients or healthy donors using EasySep HLA Chimerism Whole Blood CD66b positive selection kit (STEMCELL) as per manufacturer\u0026rsquo;s instructions. Neutrophils were maintained in RPMI-1640 (Gibco) medium supplemented with 10% FBS, 100 mM L-glutamine and 1x penicillin-streptomycin. For co-culture, cryopreserved CD4\u003csup\u003e+\u003c/sup\u003e T cells were thawed and cultured with CD66b\u003csup\u003e+\u003c/sup\u003e neutrophils in 24-well plates at a 1:2 T cells to neutrophils ratio in the presence or absence of anti-CD3/CD28 beads for 48 hours. Rapamycin (#A8167-APE) was obtained from Stratech Scientific. L-ascorbate (Vitamin C) was from Merck (#11140).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eFlow cytometry\u003c/h2\u003e \u003cp\u003eCells were harvested by centrifugation at 300 g for 3 minutes, followed by live/dead cell staining (Fixable Green Dead Cell Stain Kit; ThermoFisher). Cells were stained using surface makers CD3-APC (Biolegend, #300412), PD-1-PE-Cy7 (Biolegend, #367414), CD69-PerCP-Cy5.5 (Biolegend, #310926), CD66b-AF700 (Biolegend, #305114), PD-L1-BV605 (Biolegend, #329724), CD123-PE (BD Biosciences, #554529) and CD64-BV421 (Biolegend, #305020) for 45 minutes at room temperature, and washed with 0.2% bovine serum albumin (BSA) in phosphate-buffered saline (PBS). Samples were then acquired on a LSRFortessa X-20 (BD Biosciences) flow cytometer and analysed using FlowJo software (v10.10). Whole blood samples from sepsis patients were collected in BD Vacutainer EDTA tubes and cultured in T cell media at a 1:4 ratio, with or without anti-CD3/CD38 Dynabeads at a 1:10 v/v ratio, for 48 hours. Cultures were then treated twice with RBC lysis buffer prior to antibody staining and flow cytometry.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eqRT-PCR and RNA-seq\u003c/h2\u003e \u003cp\u003eTotal RNA was extracted from lysed cells using the Monarch Total RNA Miniprep Kit (NEB #T2010). cDNA synthesis was subsequently performed with LunaScript RT SuperMix Kit (NEB #E3010). Gene expression levels were quantified by qRT-PCR using SYBR Green Real-Time PCR Master Mix (Qiagen) on a CFX-96 C1000 platform (Bio-Rad), with β-actin as the normalisation control (see \u003cb\u003eTable S8\u003c/b\u003e for primer sequences).\u003c/p\u003e \u003cp\u003e1\u0026micro;g RNA was used for library preparation using the NEBNext Ultra II RNA Library Prep Kit for Illumina (#E7770S) following the manufacturer's protocol. Poly(A) mRNA enrichment was performed using the NEBNext Poly(A) mRNA Magnetic Isolation Module (#E7490). Library quality control, including size distribution and quantification, was assessed using TapeStation 4200 (Agilent) with High Sensitivity D1000 reagents and Qubit HS DNA kit (ThermoFisher), respectively. Final library molarity was determined using the KAPA Library Quantification Kit (Roche). Libraries were sequenced on the Illumina NextSeq 500 platform using a 150-cycle High Output Kit v2.5.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eRNA-seq analysis\u003c/h2\u003e \u003cp\u003eRaw RNA-seq reads were trimmed using Trim Galore (v0.6.2) and aligned to the human genome (hg38) using HISAT2 (v2.1.0). Transcript quantification was performed using featureCounts (v1.6.2) with GENCODE v31 annotations. The bigwig files normalised by RPKM (Reads Per Kilobase per Million mapped reads) were generated using the bamCoverage function of deepTools (version 3.3.1). Differential gene expression analysis was conducted using DESeq2 (v1.36.1) on raw read counts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eOmni-ATAC-seq and analysis\u003c/h2\u003e \u003cp\u003e50,000 cells were prepared by centrifugation and resuspended in 50 \u0026micro;l of lysis buffer (10 mM Tris-HCL pH 7.4, 10 mM NaCl, 3 mM MgCl2, 0.01% Digitonin, 0.1% Tween-20 and 0.1% Igepal CA-630) for nuclear isolation. Following transposition and DNA purification, library preparation was performed using standard protocols as described previously \u003csup\u003e\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e,\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e\u003c/sup\u003e. Final libraries were quality controlled and sequenced on the Illumina NextSeq 500 platform.\u003c/p\u003e \u003cp\u003eSequencing reads for ATAC-seq were aligned to the human genome (hg38) using Bowtie2 (v2.2.5). As described previously \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e\u003c/sup\u003e, data were filtered for quality control using Picard (v2.0.1) and Samtools (v1.9) before peak calling with MACS2 (v2.1.0). Differential peak analysis was performed using DESeq2, considering peaks present in at least 30% of samples. Potential batch effects and/or technical variation were assessed through principal component analysis and incorporated as covariates in the DESeq2 design formula.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003ePublic hMeDIP-seq analysis\u003c/h2\u003e \u003cp\u003eRaw sequencing reads for hydroxymethylated DNA Immunoprecipitation Sequencing (hMeDIP) data were obtained from GSE74850\u003csup\u003e60\u003c/sup\u003e, trimmed using Trim Galore (v0.6.2), and aligned to human genome (hg38) using the BWA-mem alignment algorithm (v0.7.12)\u003csup\u003e\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e\u003c/sup\u003e. The binary alignment and map (BAM) files were filtered to remove reads with a mapping quality score less than 10 and duplicate reads using SAMtools (v1.9) and Picard (v2.21.1). The normalised fold enrichment tracks over the corresponding input controls were generated by using the \u003cem\u003ecallpeak\u003c/em\u003e function with the \u003cem\u003e--SPMR\u003c/em\u003e flag, then passing the bedgraph outputs into the \u003cem\u003ebdgcmp\u003c/em\u003e function of MACS2 and the bedGraphToBigWig tool.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eLentivirus production\u003c/h2\u003e \u003cp\u003eHuman embryonic kidney (HEK) 293FT cells were maintained in Opti-MEM I Reduced Serum Medium (OPTI-MEM) with GlutaMAX Supplement (ThermoFisher, #51985034) supplemented with 5% FBS (Sigma, #F7524), 1 mM sodium pyruvate (ThermoFisher), and 1\u0026times; MEM nonessential amino acids (ThermoFisher) in T175 flasks. Cells were seeded per 150 mm dish (Corning, #430599) in 14 ml of medium overnight to achieve confluency about 90% at the time point of transfection. Cells were transfected with a plasmid mixture containing 5.7 \u0026micro;g psPAX2 (Addgene #12260), 3.2 \u0026micro;g pCMV-VSV-G (Addgene #8454), and either 4.6 \u0026micro;g of a sgRNA expression vector (Addgene #96923) or 7.0 \u0026micro;g of dCas9-VP64 (Addgene #180263) in equimolar ratios using jetPRIME reagent (Polyplus). After 6 h, the transfection medium was replaced with fresh medium supplemented with ViralBoost (Alstem Bio, #VB100). Lentiviral supernatant was harvested in 24 h post-transfection, filtered with a 0.45 \u0026micro;m membrane filter (Millipore), and concentrated using ultracentrifugation at 29,000 rpm for 2h at 4\u0026deg;C. The pellet was resuspended in PBS with 1.5% BSA, aliquoted and stored at -80\u0026deg;C. Lentivirus titers of dCas9-VP64 were determined by quantifying mCherry-positive cells via flow cytometry post-transduction. Viral volume that results in at least 50% transduction efficiency was used.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eCRISPR-dCas9 mediated epigenetic editing for primary T cells\u003c/h2\u003e \u003cp\u003eWe designed and selected top ranked single guide RNA (sgRNA) based on the scoring metrics using FlashFry\u003csup\u003e\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e\u003c/sup\u003e. For the human U6 promoter-based transcription, a guanine base was added to the 5\u0026prime; of the sgRNA when the 20bp guide sequence did not begin with G. The sgRNA sequences are listed in \u003cb\u003eTable S8\u003c/b\u003e. Primary human CD4\u0026thinsp;+\u0026thinsp;T cells were activated using anti-CD3/CD28 Dynabeads (ThermoFisher). One million cells were transduced with sgRNA- and dCas9-VP64-containing lentiviruses in the presence of 8 \u0026micro;g/mL polybrene (Merck, #28728-55-4) for 24 hours. Transduced cells were maintained in media and assayed in 6 or 7 days.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePeripheral blood samples were obtained from healthy volunteers following informed consent (Oxfordshire Research Ethics Committee approval REC reference 06/Q1605/55); and from sepsis patients in the Sepsis Immunomics (SI) study (South Central Oxford REC C, reference:19/SC/0296) and UK GAinS (REC approvals 05/MRE00/38, 08/H0505/78, and 06/Q1605/55) with ethics approval granted nationally and locally, and informed consent obtained from all patients or their legal representative. UK Biobank has obtained ethics approval from the North West Multi-centre Research Ethics Committee (approval number: 11/NW/0382) and had obtained informed consent from all participants.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJCK reports a grant to his institution from the Danaher Beacon Programme for work on RNA biomarker point-of-care test development in sepsis for endotype assignment which includes support for KC-G and JCK. All remaining authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis project was supported by the Medical Research Council (MR/V002503/1) (JCK), a Wellcome Trust Investigator Award [204969/Z/16/Z] [JCK], Chinese Academy of Medical Sciences (CAMS) Innovation 537 Fund for Medical Science [2018-I2M-2-002] [PZ, JCK], Wellcome Trust Grants [090532/Z/09/Z and 203141/Z/16/Z to core facilities Centre for Human Genetics, and the NIHR Oxford Biomedical Research Centre. Computation used the Oxford Biomedical Research Computing (BMRC) facility, a joint development between the Wellcome Centre for Human Genetics and the Big Data Institute supported by Health Data Research UK and the NIHR Oxford Biomedical Research Centre. The Wellcome Sanger Institute is funded by the Wellcome Trust [220540/Z/20/A]. CO’N. is supported by a Wellcome Trust Doctorate Award (228321/Z/23/Z). AJM received support from the Academy of Medical Sciences Starter Grant (SGL024∖1096) and the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData and code availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGAinS gene expression and genotyping data were deposited at the European Genome-phenome Archive (EGA), under accession number EGAD00001008730\u003csup\u003e7\u003c/sup\u003e and EGAD00001015369\u003csup\u003e28\u003c/sup\u003e.\u0026nbsp;RNA-seq and ATAC-seq raw FASTQ files for CD4+ and CD8+ T cells are available under accession number EGAS50000000894 (https://ega-archive.org/studies/EGAS50000000894). Processed data, including raw and normalised counts and bigWig files for genome-wide signal data can be accessed on Zenodo (https://zenodo.org/uploads/14907264). The raw ATAC-seq data for primary immune cells were obtained from GSE172116 (macrophages), EGAS00001007362 (monocytes), GSE150018 (neutrophils), GSE118189 (NK and dendritic cells) and GSE168882 (CAR T cells). RNA-seq data in rapamycin-treated CD4+ T cells were obtained from GSE129829. MeDIP-Seq for 5hmC in CD4+ T cells were obtained from GSE74850. The processed histone modification and CTCF ChIP-seq results were downloaded from the ENCODE project (https://www.encodeproject.org/; see also \u003cstrong\u003eTable S7\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAll code used for data processing and analysis in this study is publicly available on GitHub (https://github.com/jknightlab/MTOR-Genetics-Project).\u003c/p\u003e\n\u003cp\u003eFor the purpose of Open Access, the authors have applied a CC BY public copyright licence to any Author Accepted Manuscript arising from this submission.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eShankar-Hari M et al (2024) Reframing sepsis immunobiology for translation: towards informative subtyping and targeted immunomodulatory therapies. Lancet Respir Med 12:323\u0026ndash;336\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCajander S et al (2024) Profiling the dysregulated immune response in sepsis: overcoming challenges to achieve the goal of precision medicine. Lancet Respir Med 12:305\u0026ndash;322\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavenport EE et al (2016) Genomic landscape of the individual host response and outcomes in sepsis: a prospective cohort study. Lancet Respir Med 4:259\u0026ndash;271\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScicluna BP et al (2017) Classification of patients with sepsis according to blood genomic endotype: a prospective cohort study. Lancet Respir Med 5:816\u0026ndash;826\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSweeney TE et al (2018) Unsupervised Analysis of Transcriptomics in Bacterial Sepsis Across Multiple Datasets Reveals Three Robust Clusters. Crit Care Med 46:915\u0026ndash;925\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeymour CW et al (2019) Derivation, Validation, and Potential Treatment Implications of Novel Clinical Phenotypes for Sepsis. JAMA 321:2003\u0026ndash;2017\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCano-Gamez E et al (2022) An immune dysfunction score for stratification of patients with acute infection based on whole-blood gene expression. Sci Transl Med 14:eabq4433\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKwok AJ et al (2023) Neutrophils and emergency granulopoiesis drive immune suppression and an extreme response endotype during sepsis. Nat Immunol 24:767\u0026ndash;779\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoomer JS et al (2011) Immunosuppression in patients who die of sepsis and multiple organ failure. \u003cem\u003eJAMA\u003c/em\u003e 306, 2594\u0026thinsp;\u0026ndash;\u0026thinsp;605\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAntcliffe DB et al (2019) Transcriptomic Signatures in Sepsis and a Differential Response to Steroids. From the VANISH Randomized Trial. Am J Respir Crit Care Med 199:980\u0026ndash;986\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong M, Graubard BI, Rabkin CS, Engels EA (2021) Neutrophil-to-lymphocyte ratio and mortality in the United States general population. Sci Rep 11:464\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTempleton AJ et al (2014) Prognostic role of neutrophil-to-lymphocyte ratio in solid tumors: a systematic review and meta-analysis. J Natl Cancer Inst 106:dju124\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang Z, Fu Z, Huang W, Huang K (2020) Prognostic value of neutrophil-to-lymphocyte ratio in sepsis: A meta-analysis. Am J Emerg Med 38:641\u0026ndash;647\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu J et al (2020) Longitudinal characteristics of lymphocyte responses and cytokine profiles in the peripheral blood of SARS-CoV-2 infected patients. EBioMedicine 55:102763\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBert S, Nadkarni S, Perretti M (2023) Neutrophil-T cell crosstalk and the control of the host inflammatory response. Immunol Rev 314:36\u0026ndash;49\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Y et al (2019) The regulatory roles of neutrophils in adaptive immunity. Cell Commun Signal 17:147\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin H et al (2023) Antigen-presenting aged neutrophils induce CD4\u0026thinsp;+\u0026thinsp;T cells to exacerbate inflammation in sepsis. J Clin Invest 133\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFairfax BP et al (2014) Innate immune activity conditions the effect of regulatory variants upon monocyte gene expression. Science 343:1246949\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConsortium GT (2020) The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science 369:1318\u0026ndash;1330\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchmiedel BJ et al (2018) Impact of Genetic Polymorphisms on Human Immune Cell Gene Expression. Cell 175:1701\u0026ndash;1715e16\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYazar S et al (2022) Single-cell eQTL mapping identifies cell type-specific genetic control of autoimmune disease. Science 376:eabf3041\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNatri HM et al (2024) Cell-type-specific and disease-associated expression quantitative trait loci in the human lung. Nat Genet 56:595\u0026ndash;604\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang P et al (2022) Epigenomic analysis reveals a dynamic and context-specific macrophage enhancer landscape associated with innate immune activation and tolerance. Genome Biol 23:136\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFang H et al (2019) A genetics-led approach defines the drug target landscape of 30 immune-related traits. Nat Genet 51:1082\u0026ndash;1091\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMizuno A, Okada Y (2019) Biological characterization of expression quantitative trait loci (eQTLs) showing tissue-specific opposite directional effects. Eur J Hum Genet 27:1745\u0026ndash;1756\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOta M et al (2021) Dynamic landscape of immune cell-specific gene regulation in immune-mediated diseases. Cell 184:3006\u0026ndash;3021e17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Klein N et al (2023) Brain expression quantitative trait locus and network analyses reveal downstream effects and putative drivers for brain-related diseases. Nat Genet 55:377\u0026ndash;388\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurnham KL et al (2024) eQTLs identify regulatory networks and drivers of variation in the individual response to sepsis. Cell Genom, 100587\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChi H (2012) Regulation and function of mTOR signalling in T cell fate decisions. \u003cem\u003eNat Rev Immunol\u003c/em\u003e 12, 325\u0026thinsp;\u0026ndash;\u0026thinsp;38\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCOVID-19 Multi-omics Blood ATlas Consortium (2022) A blood atlas of COVID-19 defines hallmarks of disease severity and specificity. Cell 185:916\u0026ndash;938e58\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTorrance HD et al (2023) A Transcriptomic Approach to Understand Patient Susceptibility to Pneumonia After Abdominal Surgery. Ann Surg\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKerimov N et al (2021) A compendium of uniformly processed human gene expression and splicing quantitative trait loci. Nat Genet 53:1290\u0026ndash;1299\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConsortium GT et al (2017) Genetic effects on gene expression across human tissues. Nature 550:204\u0026ndash;213\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiambartolomei C et al (2014) Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using Summary Statistics. PLoS Genet 10\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlbiges L et al (2012) Incidence and management of mTOR inhibitor-associated pneumonitis in patients with metastatic renal cell carcinoma. Ann Oncol 23:1943\u0026ndash;1953\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSweeney RM, McAuley DF (2016) Acute respiratory distress syndrome. Lancet 388:2416\u0026ndash;2430\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVillar J et al (2015) Assessment of PaO(2)/FiO(2) for stratification of patients with moderate and severe acute respiratory distress syndrome. BMJ Open 5:e006812\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRautanen A et al (2015) Genome-wide association study of survival from sepsis due to pneumonia: an observational cohort study. Lancet Respiratory Med 3:53\u0026ndash;60\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCerezo M et al (2025) The NHGRI-EBI GWAS Catalog: standards for reusability, sustainability and diversity. Nucleic Acids Res 53:D998\u0026ndash;D1005\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVujkovic M et al (2020) Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1.4 million participants in a multi-ancestry meta-analysis. Nat Genet 52:680\u0026ndash;691\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuzuki K et al (2024) Genetic drivers of heterogeneity in type 2 diabetes pathophysiology. Nature 627:347\u0026ndash;357\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenney AE et al (2024) Comparative estimate of glucose-lowering therapies on risk of incident pneumonia and severe sepsis: an analysis of real-world cohort data. Thorax 80:32\u0026ndash;41\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFukushima K, Kitamura S, Tsuji K, Wada J (2021) Sodium-Glucose Cotransporter 2 Inhibitors Work as a Regulator of Autophagic Activity in Overnutrition Diseases. Front Pharmacol 12\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Z et al (2016) Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet 48:481\u0026ndash;487\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerriets VA et al (2016) Leptin directly promotes T-cell glycolytic metabolism to drive effector T-cell differentiation in a mouse model of autoimmunity. Eur J Immunol 46:1970\u0026ndash;1983\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLord GM et al (1998) Leptin modulates the T-cell immune response and reverses starvation-induced immunosuppression. Nature 394:897\u0026ndash;901\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolsinger KE, Weir BS (2009) Genetics in geographically structured populations: defining, estimating and interpreting F(ST). Nat Rev Genet 10:639\u0026ndash;650\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eProhaska A et al (2019) Human Disease Variation in the Light of Population Genomics. Cell 177:115\u0026ndash;131\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen L et al (2016) Genetic Drivers of Epigenetic and Transcriptional Variation in Human Immune Cells. Cell 167:1398\u0026ndash;1414e24\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoudet J (2005) HIERFSTAT, a package for R to compute and test hierarchical F-statistics. Mol Ecol Notes 5:184\u0026ndash;186\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang X et al (2015) Neutrophil CD64 expression as a diagnostic marker for sepsis in adult patients: a meta-analysis. Crit Care 19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeghraoui-Kheddar A et al (2022) Two New Neutrophil Subsets Define a Discriminating Sepsis Signature. Am J Respir Crit Care Med 205:46\u0026ndash;59\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi LZ et al (2011) HIF1alpha-dependent glycolytic pathway orchestrates a metabolic checkpoint for the differentiation of TH17 and Treg cells. J Exp Med 208:1367\u0026ndash;1376\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin AR et al (2017) Rapamycin-mediated mTOR inhibition uncouples HIV-1 latency reversal from cytokine-associated toxicity. J Clin Invest 127:651\u0026ndash;656\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHiemer S et al (2019) Integrated Metabolomic and Transcriptomic Profiling Reveals Novel Activation-Induced Metabolic Networks in Human T cells. bioRxiv, 635789\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlaschke K et al (2013) Vitamin C induces Tet-dependent DNA demethylation and a blastocyst-like state in ES cells. Nature 500:222\u0026ndash;\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConsortium EP (2012) An integrated encyclopedia of DNA elements in the human genome. Nature 489:57\u0026ndash;74\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGennert DG et al (2021) Dynamic chromatin regulatory landscape of human CAR T cell exhaustion. Proc Natl Acad Sci U S A 118\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoskic B et al (2022) Immune disease risk variants regulate gene expression dynamics during CD4(+) T cell activation. Nat Genet 54:817\u0026ndash;826\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarina RJ et al (2016) TET-catalyzed oxidation of intragenic 5-methylcytosine regulates CTCF-dependent alternative splicing. EMBO J 35:335\u0026ndash;355\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Brabander J et al (2025) The blood transcriptional response in patients developing intensive care unit-acquired pneumonia. Eur Respir J\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoquilly A et al (2023) Interferon gamma-1b for the prevention of hospital-acquired pneumonia in critically ill patients: a phase 2, placebo-controlled randomized clinical trial. Intensive Care Med 49:530\u0026ndash;544\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWudhikarn K, Perales MA (2022) Infectious complications, immune reconstitution, and infection prophylaxis after CD19 chimeric antigen receptor T-cell therapy. Bone Marrow Transpl 57:1477\u0026ndash;1488\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNian Z et al (2021) Rapamycin Pretreatment Rescues the Bone Marrow AML Cell Elimination Capacity of CAR-T Cells. Clin Cancer Res 27:6026\u0026ndash;6038\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParry A, Rulands S, Reik W (2021) Active turnover of DNA methylation during cell fate decisions. Nat Rev Genet 22:59\u0026ndash;66\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChait A, den Hartigh LJ (2020) Adipose Tissue Distribution, Inflammation and Its Metabolic Consequences, Including Diabetes and Cardiovascular Disease. Front Cardiovasc Med 7:22\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWiner S et al (2009) Normalization of obesity-associated insulin resistance through immunotherapy. Nat Med 15:921\u0026ndash;929\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRocha VZ et al (2008) Interferon-gamma, a Th1 cytokine, regulates fat inflammation: a role for adaptive immunity in obesity. Circ Res 103:467\u0026ndash;476\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou K, Lansang MC et al (2000) Diabetes Mellitus and Infection. in \u003cem\u003eEndotext\u003c/em\u003e (eds. Feingold, K.R. (South Dartmouth (MA)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKornum JB et al (2007) Type 2 diabetes and pneumonia outcomes: a population-based cohort study. Diabetes Care 30:2251\u0026ndash;2257\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAngriman F et al (2024) Preexisting Diabetes Mellitus and All-Cause Mortality in Adult Patients With Sepsis: A Population-Based Cohort Study. Crit Care Explor 6:e1085\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLamontagne F et al (2022) Intravenous Vitamin C in Adults with Sepsis in the Intensive Care Unit. N Engl J Med 386:2387\u0026ndash;2398\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWacker DA et al (2022) Evaluating Vitamin C in Septic Shock: A Randomized Controlled Trial of Vitamin C Monotherapy. Crit Care Med 50:e458\u0026ndash;e467\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSato R et al (2021) Effect of IV High-Dose Vitamin C on Mortality in Patients With Sepsis: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. Crit Care Med 49:2121\u0026ndash;2130\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatel JJ et al (2022) IV Vitamin C in Critically Ill Patients: A Systematic Review and Meta-Analysis. Crit Care Med 50:e304\u0026ndash;e312\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen LL et al (2022) Itaconate inhibits TET DNA dioxygenases to dampen inflammatory responses. Nat Cell Biol 24:353\u0026ndash;363\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuznetsova A, Brockhoff PB, Christensen RHB (2017) lmerTest Package: Tests in Linear Mixed Effects Models. J Stat Softw 82:1\u0026ndash;26\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman AM et al (2019) Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat Biotechnol 37:773\u0026ndash;782\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor DL et al (2019) Integrative analysis of gene expression, DNA methylation, physiological traits, and genetic variation in human skeletal muscle. Proc Natl Acad Sci U S A 116:10883\u0026ndash;10888\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuil A et al (2015) Gene-gene and gene-environment interactions detected by transcriptome sequence analysis in twins. Nat Genet 47:88\u0026ndash;91\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurnham KL et al (2017) Shared and Distinct Aspects of the Sepsis Transcriptomic Response to Fecal Peritonitis and Pneumonia. Am J Respir Crit Care Med 196:328\u0026ndash;339\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSudlow C et al (2015) UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 12:e1001779\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBycroft C et al (2018) The UK Biobank resource with deep phenotyping and genomic data. Nature 562:203\u0026ndash;209\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarp-Tatham E, O'Neill CR, Knight JC, Mentzer AJ, Chong AY (2024) Letter to the Editor: Lack of association between classical HLA genes and asymptomatic SARS-CoV-2 infection. HGG Adv, 100382\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCorces MR et al (2017) An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues. Nat Methods 14:959\u0026ndash;962\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown AC et al (2023) Comprehensive epigenomic profiling reveals the extent of disease-specific chromatin states and informs target discovery in ankylosing spondylitis. Cell Genom 3:100306\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25:1754\u0026ndash;1760\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcKenna A, Shendure J (2018) FlashFry: a fast and flexible tool for large-scale CRISPR target design. BMC Biol 16:74\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6457289/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6457289/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSepsis is a heterogeneous clinical syndrome with a high mortality rate and personalised stratification strategies are proposed as essential to successful targeted therapeutics. Here, we characterise genetic variation that modulates \u003cem\u003eMTOR\u003c/em\u003e, a critical regulator of metabolism and immune responses in sepsis. The effects are highly context specific, involving a regulatory element that affects \u003cem\u003eMTOR\u003c/em\u003e expression in activated T cells with opposite direction of effect in neutrophils. The lead variant, rs4845987, significantly interacts with the known sepsis prognostic marker neutrophil-to-lymphocyte ratio, shows activity specific to sepsis endotype, and a pleiotropic effect on type 2 diabetes (T2D) risk. Using \u003cem\u003eex vivo\u003c/em\u003e models, we demonstrate that activated T cells promote immunosuppressive sepsis neutrophils through released cytokines, a process dampened by hypoxia and the mTOR inhibitor rapamycin. The G-allele of rs4845987, associated with decreased risk of T2D, is associated with reduced mTOR signaling in T cells and improved survival in sepsis patients due to pneumonia. We define a novel epigenetic mechanism that fine-tunes \u003cem\u003eMTOR\u003c/em\u003e transcription and T cell activity via the variant-containing regulatory element, which exhibits an allelic effect upon vitamin C treatment. Our findings reveal how common genetic variation can interact with disease state/endotype to modulate immune cell-cell communication, providing a patient stratification strategy to inform more effective treatment of sepsis.\u003c/p\u003e","manuscriptTitle":"Context-specific regulatory genetic variation in MTOR dampens neutrophil-T cell crosstalk in sepsis, modulating disease","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-26 09:00:48","doi":"10.21203/rs.3.rs-6457289/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"234f6fd2-3cee-4ae7-8153-342a357b9ace","owner":[],"postedDate":"May 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47702134,"name":"Biological sciences/Genetics/Genomics/Personalized medicine"},{"id":47702135,"name":"Health sciences/Diseases/Infectious diseases/Bacterial infection"},{"id":47702136,"name":"Health sciences/Medical research/Outcomes research"}],"tags":[],"updatedAt":"2026-04-08T07:07:40+00:00","versionOfRecord":{"articleIdentity":"rs-6457289","link":"https://doi.org/10.1038/s41467-026-69919-7","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2026-02-25 05:00:00","publishedOnDateReadable":"February 25th, 2026"},"versionCreatedAt":"2025-05-26 09:00:48","video":"","vorDoi":"10.1038/s41467-026-69919-7","vorDoiUrl":"https://doi.org/10.1038/s41467-026-69919-7","workflowStages":[]},"version":"v1","identity":"rs-6457289","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6457289","identity":"rs-6457289","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}