Salmonella lipopolysaccharide stimulates uptake of long-chain fatty acids in the small intestine Authors: Eugene Koo, Gabriella Quinn, Cassie L. Behrendt, Brian Hassell, Katie N. Kang, Kelly A. Ruhn, Gonçalo Vale, Jeffrey G. McDonald, Wenhan Zhu, Sebastian E. Winter, Lora V. Hooper Published: 2026 ABSTRACT SignificanceThe intestinal lining absorbs nutrients while also detecting bacteria that enter with food, but how these two functions intersect during infection is poorly understood. Salmonella is a food-borne bacterial pathogen that infects humans. By studying mice, we found that Salmonella infection rapidly enhances absorption of dietary lipids by intestinal epithelial cells. This response was triggered when immune cells beneath the gut lining detected lipopolysaccharide, a molecule in the outer coat of many Gram-negative bacteria. In mice with defective lipid absorption, more Salmonella spread beyond the intestine, suggesting that lipid uptake contributes to host defense. These findings reveal a previously unappreciated link between microbial sensing, nutrient absorption, and defense against intestinal infection. The gut microbiota plays a central role in shaping host physiology, with growing evidence indicating its profound influence on nutrient metabolism and energy balance. Among its best-characterized functions is the capacity to enhance digestion of complex carbohydrates, enabling the liberation and absorption of monosaccharides and short-chain fatty acids (1). In parallel, microbial signals strongly influence intestinal lipid absorption and systemic lipid metabolism (2–4). Previous studies have shown that the microbiota engages host-intrinsic pathways in epithelial cells to promote the transcription of lipid metabolism genes, thereby facilitating the uptake and processing of dietary lipids (5–7). These microbiota-driven effects on lipid handling have been linked to circadian regulation (5), histone modification (7), and long noncoding RNA activity (6), and require innate immune signaling via myeloid cells and group 3 innate lymphoid cells (5–7). Chronic exposure to microbial signals thus programs epithelial lipid uptake in a manner that contributes to long-term energy storage and the development of obesity. By contrast, less is known about how enteric pathogens influence host lipid metabolism. While intestinal pathogens such as Salmonella enterica serovar Typhimurium (S. Typhimurium) have a well-recognized ability to trigger inflammation and disrupt barrier function (8), their impact on host nutrient handling has not been fully explored. Although enteric pathogens are known to alter carbohydrate availability and utilization during infection (9), it remains unclear whether and how they influence lipid absorption and utilization. Given that infection induces a robust innate immune response and alters epithelial cell gene expression and function (10–12), we reasoned that acute exposure to a pathogen might impact host lipid metabolism. Here, we show that infection with S. Typhimurium rapidly enhances long-chain fatty acid absorption in the small intestine, coincident with the induction of epithelial transcriptional programs involved in lipid uptake and transport. This response is driven by bacterial lipopolysaccharide (LPS), depends on the acylation state of LPS, and requires Toll-like receptor 4 (TLR4) signaling in subepithelial myeloid cells. We further find that the related commensal bacterium Escherichia coli similarly promotes lipid absorption through an LPS-dependent mechanism. Together, these findings identify bacterial LPS as an acute regulator of host lipid metabolism and indicate that both pathogens and commensals can engage shared innate immune pathways to influence nutrient handling at the intestinal barrier. Results S. Typhimurium Infection Promotes Lipid Uptake and Metabolic Reprogramming in Small Intestinal Epithelial Cells (IECs). To investigate whether S. Typhimurium infection alters intestinal lipid metabolism, we reanalyzed a published single-cell RNA-seq dataset of IECs from infected and uninfected mice (12). This analysis revealed increased expression of genes involved in lipid metabolism and transport in IECs following infection (Fig. 1 A–C). Gene set enrichment analysis of differentially expressed transcripts showed that lipid metabolism pathways were prominently represented among genes upregulated during infection (Fig. 1B), accounting for more than 10% of enriched pathways (37 of 313). Notably, genes involved in multiple stages of lipid trafficking, including lipid import, intracellular transport, and export, were increased in IECs from infected mice (Fig. 1C). Salmonella Typhimurium infection promotes lipid uptake and metabolic reprogramming in small IECs. (A) Volcano plot from reanalysis of published scRNA-seq data of small IECs isolated from streptomycin-pretreated mice 48 h after S. Typhimurium SL1344 infection or from uninfected controls (12). Lipid metabolism–associated genes enriched in infected IECs are highlighted in red; Reg3 genes are shown in blue. (B) Gene ontology (GO) pathway analysis of transcripts enriched in IECs from S. Typhimurium–infected mice (12). Shown are pathways related to fatty acid metabolism. (C) Heatmap showing relative expression (Z-score) of lipid metabolism and transport genes in IECs from uninfected and infected mice (12). Schematic depicts long-chain fatty acid uptake and intracellular trafficking pathways in IECs. (D) qPCR analysis of Cd36, Fabp1, and Apoa4 expression in small intestine from conventional (Conv) mice or antibiotic-treated (Abx-treated) mice 24 h after infection with S. Typhimurium SL1344 (n = 5 to 12 mice per group). Significance was determined by one-way ANOVA. (E) Immunofluorescence microscopy of small intestines from antibiotic-treated uninfected or S. Typhimurium–infected mice, showing neutral lipid accumulation (LipidTOX, red) and nuclei (DAPI, blue). Representative images and zoomed region shown. (Scale bar, 100 µm.) (F) Triglyceride quantification in isolated small intestinal IECs from conventional or antibiotic-treated mice 24 h after S. Typhimurium infection (n = 3 to 5 mice per group). Significance was determined by one-way ANOVA. (G) Mass spectrometry–based lipidomic analysis of IECs from conventional or antibiotic-treated mice following S. Typhimurium infection. Heatmap depicts relative abundance (Z-score) of lipid classes, averaged across 3 to 4 mice per group. Lipid classes and their abbreviations are defined in SI Appendix, Table S1. (H) Fatty acid composition of glycerolipids was determined by mass spectrometry for the same IEC samples shown in (G). Heatmap depicts relative abundance (Z-score) of individual fatty acids, averaged across 3 to 5 mice per group. (I) In vivo uptake of alkyne-labeled long-chain fatty acids in IECs from antibiotic-treated mice infected with S. Typhimurium SL1344. Alkyne-labeled fatty acids were administered by intragastric gavage, followed by click-chemistry detection after IEC isolation. Fluorescence was quantified by flow cytometry as median fluorescence intensity (MFI; n = 5 to 7 mice per group). Significance was determined by Student’s t test. (J) In vivo uptake of the fluorescent palmitate analog C16:0 BODIPY in IECs from antibiotic-treated mice infected with S. Typhimurium SL1344. Mice received C16:0 BODIPY by intragastric gavage, and fluorescence was quantified by flow cytometry (MFI; n = 12 to 13 mice per group). Representative histograms and quantification are shown. Significance was determined by Student’s t test. scRNA-seq, single-cell RNA sequencing; S. Tm, Salmonella Typhimurium; GO, gene ontology; Sm. int., small intestine; qPCR, quantitative real-time PCR; Abx-treated, antibiotic-treated; Conv, conventional; Uninf, uninfected; IEC, intestinal epithelial cell; LP, lamina propria. Data shown in panels D–F, I, and J are representative of 2 to 3 independent experiments. Each bar graph data point represents one mouse. To independently validate these findings while minimizing infection-associated tissue pathology, we analyzed intestinal gene expression early during infection. Mice were treated with antibiotics to deplete the microbiota, were orally infected with S. Typhimurium, and intestinal gene expression was measured 24 h after infection. At this time point, infected mice showed minimal intestinal pathology and similar numbers of viable IECs compared with uninfected controls, indicating that infection did not cause substantial IEC loss (SI Appendix, Fig. S1 A–D). Nevertheless, intestinal Il22 expression was robustly induced (SI Appendix, Fig. S1E), consistent with the established early interleukin 22 (IL-22) response to S. Typhimurium infection and indicating productive host recognition of the pathogen (13). Infected mice exhibited increased expression of genes involved in lipid transport, including Cd36, which mediates long-chain fatty acid uptake (14); Fabp1, which facilitates intracellular lipid trafficking (15); and Apoa4, which enables lipid export from enterocytes (16) (Fig. 1D). Similar results were observed in germ-free mice, indicating that these transcriptional changes occur independently of the resident microbiota (SI Appendix, Fig. S2A). Conversely, conventional mice that were not infected with S. Typhimurium did not show increased expression of these genes, indicating that the microbiota alone is insufficient to drive the response (Fig. 1D). Together, these findings indicate that S. Typhimurium infection induces a transcriptional program consistent with enhanced lipid uptake and trafficking in IECs. The increased expression of Cd36, encoding a long-chain fatty acid transporter, suggested that infection might enhance lipid uptake by IECs. To determine whether lipid abundance was altered during infection, we first visualized neutral lipids in intestinal tissue using LipidTOX staining. IECs from infected mice showed marked accumulation of neutral lipids compared with uninfected controls (Fig. 1E). Consistent with this observation, direct biochemical measurement revealed increased accumulation of neutral lipid triacylglycerols (TAGs)—the primary storage form of absorbed dietary fatty acids—in IECs from S. Typhimurium–infected mice (Fig. 1F). Unbiased mass spectrometry–based lipidomic analysis further confirmed increased abundance of several lipid classes following infection, including TAGs (Fig. 1G). Infection was also associated with increased abundance of multiple long-chain fatty acid species (Fig. 1H), consistent with increased uptake of dietary fatty acids that are subsequently incorporated into intracellular TAG pools. Together, these data indicate that S. Typhimurium infection promotes lipid accumulation in IECs. Increased intracellular lipid levels can arise from either enhanced uptake of dietary lipids or increased endogenous fatty acid synthesis. To distinguish between these possibilities, we examined expression of genes encoding key enzymes involved in de novo lipogenesis. These included Fasn, encoding fatty acid synthase, which synthesizes palmitate (17); Scd1, encoding stearoyl-CoA desaturase-1, which generates monounsaturated fatty acids required for cellular lipid biosynthesis (18); and members of the Elovl gene family, which elongate fatty acyl-CoAs, producing very-long-chain fatty acids (19). Levels of these transcripts were not appreciably altered in S. Typhimurium–infected mice (SI Appendix, Fig. S2B), arguing that increased de novo lipogenesis is unlikely to account for the observed lipid accumulation and suggesting that infection instead enhances uptake of dietary fatty acids. We next tested directly whether S. Typhimurium infection modulates the uptake of dietary fatty acids. We focused on three long-chain fatty acids—palmitate (16:0), stearate (18:0), and oleate (18:1)—that were elevated in IECs during infection (Fig. 1H). To quantify uptake, we administered alkyne-labeled fatty acid analogs by intragastric gavage and measured IEC fluorescence by flow cytometry following click-chemistry labeling (Fig. 1I). IECs from infected mice exhibited increased uptake of alkyne-palmitate and alkyne-stearate, whereas alkyne-oleate uptake was unchanged (Fig. 1I). These findings indicate that S. Typhimurium infection selectively enhances uptake of specific dietary long-chain fatty acids by the intestinal epithelium. Because palmitate is a well-characterized substrate for CD36, which preferentially mediates uptake of long-chain fatty acids (14), we next asked whether infection-induced uptake exhibits chain-length specificity. Antibiotic-treated mice received the fluorescent fatty acid analogs C16:0 BODIPY or C12:0 BODIPY by intragastric gavage, and IEC fluorescence was quantified by flow cytometry. S. Typhimurium infection increased uptake of C16:0 BODIPY but not C12:0 BODIPY (Fig. 1J and SI Appendix, Fig. S2C). C16:0 BODIPY fluorescence remained elevated in IECs from infected mice over 8 h, although signal gradually declined over time, consistent with enterocyte lipid processing and chylomicron-mediated lipid export (20) (SI Appendix, Fig. S2D). Together, these findings indicate that S. Typhimurium infection selectively enhances epithelial uptake of dietary long-chain fatty acids in a manner consistent with CD36-mediated transport. S. Typhimurium LPS Promotes Epithelial Long-Chain Fatty Acid Uptake. To identify bacterial signals responsible for inducing epithelial cell lipid uptake and metabolism, we first examined whether Salmonella virulence mechanisms were required for this response. S. Typhimurium invades epithelial cells and survives intracellularly through Salmonella pathogenicity island (SPI)-encoded type III secretion systems (21). However, both heat-killed bacteria and a ΔinvAΔspiB mutant (22) lacking functional type III secretion retained the ability to induce lipid metabolism genes (SI Appendix, Fig. S3 A–D). These results suggest that bacterial invasion and intracellular survival mechanisms are not required for epithelial lipid metabolic responses. To determine whether defined bacterial components could recapitulate this phenotype, we tested representative pathogen-associated molecular patterns (PAMPs). These PAMPs included LPS, a Gram-negative outer membrane glycolipid; flagellin, the protein subunit of bacterial flagella; and lipoteichoic acid (LTA), a Gram-positive cell wall component. Antibiotic-treated mice were injected intraperitoneally with ultra-purified LPS, flagellin, or LTA, and intestinal gene expression and lipid uptake were assessed. Whereas intragastric administration of LPS failed to induce lipid metabolism genes in the small intestine (SI Appendix, Fig. S4A), systemic administration of LPS induced expression of lipid transport genes and increased C16:0 BODIPY uptake by IECs (Fig. 2 A and B). By contrast, systemic administration of flagellin or LTA had no effect. Consistent with the differential effects of oral and systemic LPS on lipid metabolism genes, oral administration of LPS elicited only a modest increase in intestinal Il22 expression, whereas systemic administration induced a robust Il22 response (SI Appendix, Fig. S4B). Thus, LPS is sufficient to promote epithelial lipid metabolic gene expression and fatty acid uptake. S. Typhimurium LPS promotes epithelial long-chain fatty acid uptake. (A) qPCR analysis of lipid metabolism mRNAs (Cd36, Fabp1, and Apoa4) in small intestines of antibiotic-treated mice injected intraperitoneally with 1 µg of PAMPs: LTA, flagellin, or ultra-pure LPS (n = 6 mice per group). (B) In vivo uptake of C16:0 BODIPY in small IECs from antibiotic-treated mice injected intraperitoneally with the PAMPs shown in (A). Mice were injected daily for 3 d. Mice were gavaged with C16:0 BODIPY 12 h after the final treatment, and IEC fluorescence was quantified by flow cytometry. Analysis was performed on live EpCAM+ cells. Representative histograms and quantification of MFI are shown (n = 7 to 8 mice per group). (C) Schematic of S. Typhimurium LPS biosynthesis and mutant strains used in (D–F). The LPS biosynthesis pathway generates penta-acylated LPS, which is converted to hexa-acylated LPS by the acyltransferase MsbB (LpxM); the ΔmsbB mutant retains penta-acylated LPS. (D) qPCR analysis of lipid metabolism mRNAs (Cd36, Fabp1, and Apoa4) in small intestines of germ-free or antibiotic-treated mice infected by intragastric gavage with parental wild-type (WT) or mutant S. Typhimurium IR715 strains shown in (C) (n = 5 to 8 mice per group). (E) Mass spectrometry–based lipidomic analysis of IECs from uninfected or WT or mutant S. Typhimurium IR715–infected mice is shown as bar plots. Relative abundance of 14:0, 16:0, and 18:0 fatty acid–containing glycerolipids is shown (n = 3 to 4 mice per group). (F) In vivo uptake of C16:0 BODIPY. Antibiotic-treated mice were infected with parental WT or mutant S. Typhimurium IR715 strains and orally gavaged with C16:0 BODIPY. IEC fluorescence was quantified by flow cytometry. Representative histograms and MFI quantification are shown (n = 6 to 11 mice per group). LPS, lipopolysaccharide; PAMPs, pathogen-associated molecular patterns; i.p., intraperitoneal; Abx-treated, antibiotic-treated; qPCR, quantitative real-time PCR; sm. int., small intestine; LTA, lipoteichoic acid; IEC, intestinal epithelial cell; Veh, vehicle; S. Tm, Salmonella Typhimurium; Uninf., uninfected; WT, wild-type. MFI, median fluorescence intensity. Data shown in panels A, B, D, and F are representative of 2 to 3 independent experiments. Each bar graph data point represents one mouse. Significance was determined by one-way ANOVA. We next asked whether LPS promotes lipid uptake during live S. Typhimurium infection. LPS is essential for bacterial viability, thus precluding generation of a bacterial strain completely lacking LPS. Instead, we used an isogenic S. Typhimurium ΔmsbB mutant, which lacks the LpxM acyltransferase that adds a sixth acyl chain to lipid A (23) (Fig. 2C and SI Appendix, Fig. S5A). As a result, the ΔmsbB mutant produces penta-acylated LPS with reduced innate immune stimulatory activity (24, 25) (Fig. 2C and SI Appendix, Fig. S5 B and C). Although the ΔmsbB mutant colonized the intestine at levels comparable to the WT and complemented strains (SI Appendix, Fig. S5D), it failed to induce lipid transport genes, increase long-chain fatty acid abundance in IECs, or enhance uptake of C16:0 BODIPY (Fig. 2 D–F). Thus, S. Typhimurium LPS promotes IEC lipid metabolism and fatty acid uptake during infection in an acylation-dependent manner. S. Typhimurium-Induced Epithelial Lipid Uptake Requires TLR4 Signaling in Dendritic Cells. We next sought to identify the host sensing pathways that mediate lipid uptake in response to S. Typhimurium LPS. LPS can be detected by multiple host pathways. These include TLR4, which signals through the adaptor MYD88 (26, 27), and cytosolic sensing pathways that activate Caspase-11 and downstream inflammasome signaling, leading to maturation and release of IL-1 family cytokines (28, 29). Because both TLR and IL-1 receptor family signaling converge on MYD88, we first asked whether the lipid metabolic response to S. Typhimurium requires MYD88-dependent signaling and, if so, whether this reflects TLR-mediated sensing or inflammasome-driven cytokine signaling. To address this question, we infected Myd88−/− mice and mice lacking functional Caspase-11 (129S1) with WT S. Typhimurium and assessed intestinal lipid metabolism. Mice lacking MYD88 exhibited reduced expression of lipid metabolism genes and decreased uptake of C16:0 BODIPY compared to WT controls (Fig. 3 A–C), indicating that this response requires MYD88-dependent signaling. In contrast, mice lacking functional Caspase-11 showed no defect in lipid metabolic gene expression or C16:0 BODIPY uptake, demonstrating that Caspase-11–dependent inflammasome activation is dispensable for this phenotype (SI Appendix, Fig. S6 A–C). These findings indicate that the MYD88 dependence of the response reflects TLR signaling rather than inflammasome-driven IL-1 family signaling. S. Typhimurium-induced epithelial lipid uptake requires TLR4 signaling in dendritic cells. (A) Mouse models and experimental overview for panels B–H. (B) qPCR analysis of lipid metabolism transcripts (Cd36, Fabp1, and Apoa4) in small intestines of antibiotic-treated Myd88+/+ or Myd88−/− littermates 24 h after oral S. Typhimurium SL1344 infection (n = 4 to 7 mice per group). (C) C16:0 BODIPY uptake by IECs from antibiotic-treated Myd88+/+ or Myd88−/− littermates 24 h after S. Typhimurium SL1344 infection. MFI was quantified by flow cytometry; representative histograms and quantification are shown (n = 5 to 6 mice per group). (D) qPCR analysis of lipid metabolism transcripts in small intestines of antibiotic-treated WT, Tlr4−/−, or Tlr5−/− mice 24 h after S. Typhimurium SL1344 infection (n = 6 to 9 mice per group). (E) C16:0 BODIPY uptake by IECs from antibiotic-treated WT, Tlr4−/−, or Tlr5−/− mice 24 h after S. Typhimurium SL1344 infection. MFI was quantified by flow cytometry; representative histograms and quantification are shown (n = 5 to 8 mice per group). (F) qPCR analysis of lipid metabolism transcripts in small intestines of antibiotic-treated Tlr4fl/fl, Tlr4ΔIEC, Tlr4ΔMφ, or Tlr4ΔDC mice 24 h after S. Typhimurium SL1344 infection (n = 4 to 10 mice per group). Only statistically significant comparisons are shown; all other within-genotype comparisons were not statistically significant. (G) C16:0 BODIPY uptake by IECs from antibiotic-treated Tlr4fl/fl, Tlr4ΔIEC, Tlr4ΔMφ, or Tlr4ΔDC mice 24 h after S. Typhimurium SL1344 infection. Fluorescence was quantified by flow cytometry; representative histograms and MFI quantification are shown (n = 3 to 8 mice per group). (H) Triglyceride quantification in small intestinal IECs isolated from antibiotic-treated Tlr4fl/fl or Tlr4ΔDC mice 24 h after S. Typhimurium SL1344 infection (n = 3 to 6 mice per group). LPS, lipopolysaccharide; qPCR, quantitative real-time PCR; S. Tm, Salmonella Typhimurium; Abx-treated, antibiotic-treated; IEC, intestinal epithelial cell; uninf., uninfected; MFI, median fluorescence intensity. Data shown are representative of 2 to 3 independent experiments. Each bar graph data point represents one mouse. Significance was determined by two-way ANOVA. We next investigated the specific TLR responsible for sensing S. Typhimurium. Because LPS was sufficient to stimulate epithelial lipid uptake (Fig. 2 A and B), and S. Typhimurium also expresses flagellin, we tested the roles of the corresponding receptors TLR4 and TLR5 by infecting Tlr4−/− and Tlr5−/− mice (Fig. 3A). Lipid metabolic gene expression and C16:0 BODIPY uptake were reduced in Tlr4−/− mice but not in Tlr5−/− mice (Fig. 3 D and E), identifying TLR4 as the receptor that mediates this response to S. Typhimurium. TLR4 is expressed by multiple cell types in the intestine, including epithelial cells and various myeloid populations. To identify the relevant cellular compartment, we bred Tlr4fl/fl mice to mice harboring cell-specific Cre transgenes to generate mice with selective deletions of Tlr4 in IECs (Villin-Cre; Tlr4ΔIEC), CD11c+ dendritic cells (Cd11c-Cre; Tlr4ΔDC), and macrophages (Lyz2-Cre; Tlr4ΔMΦ) (Fig. 3A). Following infection with WT S. Typhimurium, only Tlr4ΔDC mice exhibited reduced lipid metabolic gene expression and decreased uptake of C16:0 BODIPY, whereas Tlr4ΔIEC and Tlr4ΔMΦ mice were comparable to Tlr4fl/fl controls (Fig. 3 F and G). Consistent with these findings, triglyceride accumulation in IECs was reduced in Tlr4ΔDC mice during infection (Fig. 3H). Likewise, induction of the infection-responsive cytokine gene Il22 was impaired in S. Typhimurium–infected Tlr4−/− and Tlr4ΔDC mice (SI Appendix, Fig. S7), indicating that TLR4 signaling in CD11c+ cells is also required for the intestinal IL-22 response to S. Typhimurium. Together, these results show that S. Typhimurium LPS is sensed by TLR4 in CD11c+ dendritic cells to drive epithelial lipid uptake and metabolic reprogramming. The Commensal Bacterium E. coli Promotes Epithelial Long-Chain Fatty Acid Uptake via LPS. We next assessed whether a noninvasive commensal bacterium could similarly promote epithelial lipid uptake. We mono-associated germ-free mice with the probiotic commensal E. coli Nissle (30, 31), and assessed intestinal lipid metabolism transcripts and epithelial fatty acid uptake. E. coli colonized the intestine at levels comparable to S. Typhimurium (Fig. 4 A and B). This resulted in increased epithelial lipid metabolism gene expression together with modest enhancement of C16:0 BODIPY uptake relative to germ-free controls (Fig. 4 A–D). These findings indicate that commensal E. coli can stimulate epithelial long-chain fatty acid uptake, although less robustly than S. Typhimurium. Commensal E. coli promotes epithelial long-chain fatty acid uptake via LPS. (A) Experimental overview for panels B and C. (B) Comparison of bacterial burdens in the small intestines of germ-free mice after oral infection with S. Typhimurium SL1344 or E. coli Nissle. (C) qPCR analysis of lipid metabolism transcripts (Cd36, Apoa4, and Fabp1) in small intestines of germ-free or E. coli mono-associated mice (n = 12 to 13 mice per group). (D) C16:0 BODIPY uptake by IECs from germ-free or E. coli mono-associated mice. Mice were gavaged with tracer and IEC fluorescence was quantified by flow cytometry. Representative histograms and quantification of MFI are shown (n = 10 to 11 mice per group). (E) qPCR analysis of lipid metabolism transcripts in small intestines of antibiotic-treated mice injected intraperitoneally with 107 CFU of heat-killed bacteria, including E. coli (WT or ΔmsbB), S. Typhimurium SL1344, Enterococcus faecalis, or Bacteroides thetaiotaomicron (n = 5 to 6 mice per group). (F) C16:0 BODIPY uptake by IECs from antibiotic-treated mice injected with heat-killed bacteria as in (E). Representative histograms and quantification are shown (n = 5 to 6 mice per group). (G) Experimental schematic for panels H and I. Antibiotic-treated mice were associated with the indicated bacterial strains and fed a high-fat diet prior to analysis. (H) Bacterial burdens in small intestines of mice associated with the indicated bacterial strains (n = 5 mice per group). (I) Triglyceride abundance in isolated IECs from mice associated with the indicated bacterial strains (n = 3 to 5 mice per group). qPCR, quantitative real-time PCR; sm. int., small intestine; uninf., uninfected; IECs, intestinal epithelial cells; i.p., intraperitoneal; Abx-treated, antibiotic-treated; MFI, median fluorescence intensity. Data shown in panels C–F are representative of 2 to 3 independent experiments. Each bar graph data point represents one mouse. Significance was determined by Student’s t test for panels B–D and by one-way ANOVA for panels E, F, H, and I. We next asked why E. coli induces a weaker lipid metabolic response despite producing structurally similar hexa-acylated LPS (23). Because S. Typhimurium invades host tissues whereas E. coli remains largely luminal, we hypothesized that E. coli-derived LPS has reduced access to the subepithelial myeloid cells that mediate TLR4-dependent signaling. Consistent with this idea, oral administration of purified LPS was insufficient to trigger lipid metabolic responses (SI Appendix, Fig. S4 A and B), suggesting that luminal exposure alone does not effectively engage this pathway. To test this model directly, we bypassed differences in bacterial localization by systemic administration of heat-killed bacteria. Under these conditions, heat-killed WT E. coli induced lipid metabolic gene expression and epithelial C16:0 BODIPY uptake to a degree comparable to S. Typhimurium (Fig. 4 E and F). By contrast, an isogenic E. coli ΔmsbB mutant failed to induce these responses (Fig. 4 E and F). Similarly, other commensal species, including the Gram-positive species E. faecalis and Bacteroides thetaiotaomicron, which produces weakly immunostimulatory penta-acylated LPS (32, 33), also failed to induce these responses (Fig. 4 E and F). Together, these findings suggest that bacterial localization and LPS structure both influence the magnitude of epithelial lipid metabolic responses. Finally, we addressed whether increased dietary lipid availability amplifies bacterially induced epithelial lipid accumulation. Antibiotic-treated mice were colonized with individual bacterial strains, maintained on a high-fat diet, and IEC triglyceride abundance was measured. Under these conditions, mice colonized with either E. coli or S. Typhimurium exhibited increased epithelial triglyceride accumulation compared to mice colonized with other commensals or the E. coli ΔmsbB mutant, despite similar intestinal colonization levels (Fig. 4 G–I). These findings indicate that commensal bacterial LPS can promote epithelial lipid accumulation, and that the magnitude of this response is shaped by both host exposure and dietary context. CD36 Contributes to Epithelial Lipid Uptake and Limits Extraintestinal S. Typhimurium Burden. Since both S. Typhimurium and E. coli induce Cd36 expression in an LPS-dependent manner, we asked whether CD36 functionally contributes to epithelial lipid uptake during infection. To address this, we generated mice with an IEC-specific deletion of Cd36 (Cd36ΔIEC) by crossing Cd36fl/fl mice with Villin-Cre transgenic mice. Under steady-state conditions, IECs from Cd36ΔIEC mice exhibited reduced intestinal expression of lipid metabolism genes, including Fabp1 and Apoa4 (Fig. 5A), indicating that CD36 contributes to basal epithelial lipid metabolic programming. CD36 contributes to epithelial lipid uptake and limits extraintestinal S. Typhimurium burden. (A) qPCR analysis of lipid metabolism genes (Apoa4 and Fabp1) in IECs recovered by laser capture microdissection from Cd36ΔIEC and Cd36fl/fl littermates (n = 8 to 9 mice per group). (B) In vivo uptake of C16:0 BODIPY by IECs from antibiotic-treated Cd36ΔIEC and Cd36fl/fl littermates 24 h after S. Typhimurium infection. Mice were gavaged with tracer, and IEC fluorescence was quantified by flow cytometry. Representative histograms and MFI quantification are shown (n = 6 to 10 mice per group). (C) Triglyceride abundance in IECs isolated from antibiotic-treated Cd36ΔIEC and Cd36fl/fl littermates 24 h after S. Typhimurium infection (n = 3 to 5 mice per group). (D) S. Typhimurium SL1344 burdens in the small intestine lumen (sm. int.), MLN, spleen, and liver 24 h after oral infection of antibiotic-treated Cd36ΔIEC and Cd36fl/fl littermates (n = 8 to 20 mice per group). (E) Model: S. Typhimurium–derived hexa-acylated LPS activates dendritic cell TLR4 to induce epithelial lipid uptake programs, promoting long-chain fatty acid uptake. CD36-dependent lipid uptake limits extraintestinal dissemination through an as-yet undefined mechanism. Sm. int., small intestine; IECs, intestinal epithelial cells; LCM, laser capture microdissection; S. Tm, Salmonella Typhimurium; Abx-treated, antibiotic-treated; Uninf., uninfected; CFU, colony forming units; MFI, median fluorescence intensity. Data shown are representative of 2 to 3 independent experiments. Each bar graph data point represents one mouse. Significance was determined by Student’s ttest for A, B, and D, and by two-way ANOVA for C. We next assessed fatty acid uptake during S. Typhimurium infection. Cd36ΔIEC mice showed reduced C16:0 BODIPY uptake in IECs under both uninfected conditions and following S. Typhimurium infection (Fig. 5B), indicating that CD36 is required for efficient epithelial long-chain fatty acid uptake. Consistent with this finding, S. Typhimurium-induced accumulation of triglycerides in IECs was diminished in Cd36ΔIEC mice compared to Cd36fl/fl controls (Fig. 5C). Together, these findings establish CD36 as a key mediator of epithelial lipid uptake and accumulation during infection. We next addressed whether CD36-dependent lipid uptake influences host control of S. Typhimurium infection. Following oral infection, bacterial burdens in the distal small intestine were comparable between Cd36ΔIEC mice and Cd36fl/fl littermate controls. However, Cd36ΔIEC mice exhibited increased bacterial burdens in the mesenteric lymph nodes (MLNs) and a trend toward higher bacterial numbers in the liver (Fig. 5D), suggesting enhanced extraintestinal dissemination. By contrast, systemic bacterial burdens were unchanged following intravenous infection, which bypasses the intestinal epithelium (SI Appendix, Fig. S8). These findings indicate that epithelial CD36 contributes to host defense during S. Typhimurium infection and support a model in which bacteria-induced lipid uptake helps to limit dissemination (Fig. 5E). Discussion In this study, we identify a mechanism by which an enteric bacterial pathogen acutely reprograms host intestinal lipid metabolism. We show that infection with S. Typhimurium induces rapid uptake of dietary lipids by IECs. Lipid uptake is driven by LPS and depends on its acylation state. This response requires sensing of LPS through the TLR4–MYD88 pathway in intestinal dendritic cells and is accompanied by coordinated upregulation of epithelial lipid metabolism genes, including the long-chain fatty acid transporter CD36, which functionally contributes to limiting bacterial dissemination. These results establish that LPS is a potent regulator of epithelial lipid absorption and reveal a link between innate immune sensing and host nutrient uptake during infection. These findings expand our understanding of how microbes shape host metabolism by showing that bacterial regulation of lipid absorption depends on the context of host exposure. Previous work revealed that the commensal microbiota promotes lipid uptake through chronic interactions with the host, particularly in the setting of a high-fat diet (5–7). In contrast, our results show that acute exposure to S. Typhimurium rapidly enhances epithelial lipid uptake under basal dietary conditions. The observation that commensal E. coli induces a similar but attenuated effect suggests that shared microbial signals engage common host pathways, while differences in host exposure determine the magnitude of the response. Our data suggest that subepithelial sensing of LPS is a key determinant of this difference. Whereas luminal LPS exposure was insufficient to induce epithelial lipid uptake, bacterial invasion or systemic delivery allowed LPS to access subepithelial dendritic cells and stimulate this response. Consistent with this interpretation, luminal LPS elicited only a modest increase in Il22 expression, whereas systemic administration induced robust Il22 expression together with epithelial lipid uptake. These results suggest that spatial compartmentalization of microbial sensing is a critical determinant of metabolic outcomes, with invasive pathogens uniquely positioned to activate this pathway by breaching the epithelial barrier. Our findings further suggest that this pathway may overlap with a previously described microbiota-responsive circuit. In earlier work, we showed that microbiota-induced epithelial lipid metabolism requires IL-22 signaling. Consistent with this model, Il22 induction during S. Typhimurium infection was impaired in both global and CD11c+ cell-specific Tlr4-deficient mice. Although further studies are needed to determine whether IL-22 is the downstream mediator of S. Typhimurium-driven lipid absorption, these findings suggest that pathogen-derived LPS engages components of the same immune signaling circuit. A central question raised by our findings is the functional significance of increased epithelial lipid absorption during infection. Lipids can fuel bioenergetic demands, remodel membranes to enhance resilience (34), and act as signaling molecules (35). Consistent with a functional role for this pathway, loss of CD36 increased bacterial dissemination during infection. How CD36-dependent lipid uptake contributes to host defense remains unclear but may involve effects on epithelial barrier function, immune cell function, or the availability and fate of bioactive lipid species. An unexpected feature of the lipid uptake response was its selectivity for specific long-chain fatty acids. S. Typhimurium infection preferentially enhanced uptake of saturated long-chain fatty acids such as palmitate and stearate, whereas uptake of oleate, a monounsaturated long-chain fatty acid, was not altered under the conditions tested. This substrate selectivity may reflect differences in transporter preference, intracellular trafficking, or metabolic utilization. Whether infection also affects uptake of very long-chain fatty acids remains to be determined. More broadly, our findings suggest that Gram-negative bacteria with immunostimulatory LPS, particularly members of the Enterobacteriaceae, may be uniquely positioned to influence host lipid metabolism. Expansion of Enterobacteriaceae is associated with metabolic disorders, including obesity, and with conditions characterized by increased intestinal permeability (36–38). Prior work from our group revealed that the microbiota promotes intestinal epithelial lipid accumulation and CD36 expression during high-fat feeding (5–7), suggesting that these observations reflect related mechanisms. Consistent with this idea, repeated intragastric exposure to heat-killed S. Typhimurium modestly enhanced weight gain in mice maintained on a high-fat diet (SI Appendix, Fig. S9). Although the underlying mechanisms remain undefined, these results suggest that chronic exposure to immunostimulatory bacterial products may contribute to microbiota-associated metabolic phenotypes. Similarly, environmental factors such as dietary emulsifiers, which increase microbial encroachment and low-grade inflammation (39), may enhance host exposure to bacterial LPS and engage related pathways. Together, these findings suggest that bacteria-driven lipid absorption may link intestinal innate immune sensing to metabolic disease. Materials and Methods Additional methods are presented in SI Appendix, Extended Materials and Methods. Mice. C57BL/6J, 129SV1, Myd88−/−, Tlr4−/−, Tlr5−/−, Tlr4fl/fl, Tlr4ΔIEC, Tlr4ΔDC, Tlr4ΔMΦ, Cd36fl/fl, Cd36ΔIEC, Villin-Cre, Lyz2-Cre, and Cd11c-Cre mice were bred and maintained under specific pathogen–free (SPF) conditions at the University of Texas Southwestern Medical Center. Tlr4−/−, Tlr4fl/fl (40), Tlr5−/− (11, 41), Cd36fl/fl (42), Villin-Cre (43), Lyz2-Cre (LysM-Cre) (44), and Cd11c-Cre (45) mice were from Jackson Laboratories. All floxed strains were maintained as homozygotes; all Cre driver lines were maintained as heterozygotes. Tlr4ΔIEC, Tlr4ΔDC, and Tlr4ΔMΦ mice were generated by crossing Tlr4fl/fl mice with Villin-Cre mice, Cd11c-Cre mice, and Lyz2-Cre mice respectively. Cd36ΔIEC mice were generated by crossing Cd36fl/fl mice to Villin-Cre mice. Germ-free C57BL/6J mice were housed and bred in the gnotobiotic mouse facility at the University of Texas Southwestern Medical Center. All experiments used male and female 10- to 18-wk–old mice. All mice were given access to chow and water ad libitum and kept in 12-h day/night cycles corresponding to 6 AM/6 PM. Mice were killed humanely by isoflurane overdose, followed by cervical dislocation. All experiments were performed using protocols approved by the Institutional Animal Care and Use Committees (IACUC) of the University of Texas Southwestern Medical Center. Mouse Diet. Antibiotic-treated and germ-free mice were fed Rodent Diet 2918 (Inotiv) that was vacuum sealed and irradiated. Chow was tested for sterility by bacterial culture testing and fecal pellets from germ-free mice fed 2918 chow were tested for bacterial growth. For high fat diet studies, mice were fed TD.160495 (Inotiv). Mice were given ad libitum access to all chow. Antibiotic Treatment of Mice. Antibiotics were sourced from Research Products International. Antibiotic cocktail water was made by dissolving 1 g/L neomycin (cat. # N20040), 1 g/L metronidazole (cat. #50-213-513), 1 g/L streptomycin (cat. #62000-100), and 0.5 g/L vancomycin (cat. # V06500) in deionized water followed by sterile filtration with a 0.22 µM mesh filter. Mice were given ad libitum access to antibiotic cocktail water for at least 7 d and were treated by intragastric gavage with 100 µL of antibiotic cocktail water at the start of antibiotic treatment. Microbiota depletion was verified by aerobic and anaerobic culture of fecal pellets. PAMP Treatment of Mice. Lyophilized PAMPs were reconstituted in sterile, endotoxin-free water and administered via intraperitoneal (i.p.) injection or intragastric (i.g.) gavage. For i.p. administration, mice received 40 ng/g body weight of ultrapure flagellin (InvivoGen, cat. #tlrl-epstfla), LPS (InvivoGen, cat. #tlrl-pb5lps), or LTA (InvivoGen, cat. #tlrl-pslta). For i.g. gavage, mice were treated with 50 µg/g body weight of LPS or LTA, or 1 µg/g body weight of flagellin. In experiments involving lipid administration, lipids were given 12 h after the final PAMP treatment. Bacterial Strains. Unless noted otherwise, S. Typhimurium and E. coli strains were cultured aerobically in Lysogeny Broth (LB; Sigma-Aldrich, cat. #L7658) or on LB plates (Sigma-Aldrich, cat. #L7533) at 37 °C. Antibiotics were added to LB and LB plates at the following final concentrations: nalidixic acid (Research Products International, cat. #3374-05-8), 50 mg/L; kanamycin (Kan; Sigma-Aldrich, cat. #K1377), 100 mg/L; carbenicillin (Carb; Goldbio cat. #C-103-5), 100 mg/L, and tetracycline (Tet;Research Products International, cat. #T17000), 15 mg/L. S. enterica serovar Typhimurium strains IR715 and IR715 msbB::KanR (ΔmsbB), were from previous studies (46, 47). E. coli Nissle 1917 (WT strain O6:K5:H1; EcN) was from a previous study (30, 31). Bacteroides thetaiotaomicron cultures (strain VPI-5482) were grown in freshly prepared liquid Tryptone–Yeast extract–Glucose medium. E. faecalis cultures (strain V583) were grown in Brain Heart Infusion broth (Sigma-Aldrich, cat. #53286). Construction of mutant bacterial strains is described in SI Appendix, Extended Materials and Methods. Treatment of Mice with Alkyne– and BODIPY–Conjugated Fatty Acids. Lyophilized alkynated lipids were from Avanti Research: oleic acid (17-yne), cat. #900412E; palmitic acid (15-yne), cat. #900400P; stearic acid (17-yne), cat. #700383P). Fatty acids were reconstituted to 1 mg/mL DMSO. Mice were given 800 nM of lipid in a 100 µL bolus by intragastric gavage. Click-chemistry attachment of CalFluor 488 azide (Click Chemistry Tools; cat. #1369-1) to alkyne-lipids was performed with the Click-iT Cell Reaction Buffer Kit (Thermo Scientific; cat. #C10269) following the manufacturer’s recommended protocol with the following changes. Briefly, following preparation of IECs for flow cytometry analysis (see below), cells were fixed and permeabilized with 4% PFA and 0.02% saponin for 15 min at room temperature, protected from light. Cells were washed once in 1% BSA/PBS solution then incubated with Click-iT Reaction Cocktail containing 50 µM CalFluor 488 azide for 1 h at room temperature protected from light. Cells were washed once in 1% BSA/PBS solution then resuspended in FACS buffer [PBS supplemented with 3% fetal bovine serum (FBS) and 1 mM EDTA] for flow cytometry analysis. BODIPY-conjugated fatty acids were from Thermo Scientific: BODIPY™ FL C16 (C16:0 BODIPY; cat. #D3821) or BODIPY™ FL C12 (C12:0 BODIPY; cat. #D3822) were reconstituted in DMSO to 1 mg/mL. Mice were given 4 µg of BODIPY-conjugated fatty acid per gram body mass by intragastric gavage and were sacrificed 4 h later. Mice that were infected with S. Typhimurium and treated with BODIPY-conjugated fatty acids were excluded from analysis if their stomachs showed visible signs of gastroparesis. Isolation of IECs. Mouse small intestines were harvested immediately following euthanasia by excising tissue at the stomach and cecum. Mesenteric adipose tissue was removed and the intestinal lumen was flushed with cold wash buffer [PBS supplemented with 3% FBS] to remove luminal contents. Intestines were then opened longitudinally, and tissues were washed twice with gentle agitation in a 100 mm dish containing 10 mL of wash buffer. Tissues were cut into 2 to 3 cm segments, transferred to 50 mL conical tubes, and washed twice by vortexing in 10 mL of wash buffer. Epithelial cells were subsequently dissociated by incubation in 10 mL of IEC separation buffer (3% FBS and 5 mM EDTA in Hank’s balanced salt solution) with shaking at 37 °C (225 RPM, 45° angle) for 20 min. The supernatant containing IECs was collected and passed through a 70 µm cell strainer to remove residual tissue. Remaining tissue fragments were washed twice with cold wash buffer to remove residual EDTA and recover additional IECs. For lipidomics analysis, isolated IECs were pelleted at 4 °C, and washed once by resuspension in 1 mL cold PBS. Cell numbers were determined, and aliquots of 1 × 106 cells were pelleted at 4 °C and flash-frozen in liquid nitrogen. Flow Cytometry of IECs. IECs were isolated as described above, and 1% of the total cell suspension (200 µL of a 20 mL suspension) was used for flow cytometry analysis. Cells were washed once with PBS and stained with a viability dye (Ghost Dye 710, Tonbo Biosciences, cat. #13-0871), followed by Fc receptor blockade (anti-CD16/32; BD Bioscience cat. #553142) in PBS. After a 20-min incubation at room temperature, cells were washed once with PBS and stained with anti-EpCAM (PE/Cyanine-7 anti-mouse CD326, BioLegend, cat. #118216) in FACS buffer for 20 min at room temperature. Cells were then washed once in FACS buffer and resuspended in FACS buffer for analysis. Flow cytometry was performed using a NovoCyte instrument and data were analyzed with FlowJo software. Triglyceride Assay. IECs were isolated as described above. Cells were pelleted, and supernatants were removed prior to flash freezing. Cell pellets were processed according to the manufacturer’s protocol using a Triglyceride Assay Kit (Abcam, cat. #ab178780). Triglyceride levels were normalized to total protein concentration as determined by Bradford Assay. Supplementary Material Appendix 01 (PDF)