APOL proteins tune gut immunity via commensal lipid recognition

In a recent paper published in Nature, Yang and colleagues identified a novel lipid–protein interaction that broadly regulates the behavior of dominant gut bacteria from the Bacteroides taxa. This interaction, in turn, tunes the expression of immune-regulating molecules in intestinal epithelial cells and promotes the expansion of specialized immune cells important for gut health.

Host–microbe mutualism relies on a delicate balance. The host immune system must control commensal microbe population while promoting “beneficial” function to the host.¹ Meanwhile, microbes compete for space and nutrients with each other and evolve strategies to persist within the host.² Although factors such as short-chain fatty acids promote mutualism, the full range of host- and microbe-derived factors that promote mutualism remains poorly defined.³ Uncovering what these factors are and how they interact could lead to fundamental insights into the nature of symbiosis and its evolutionary origins. This line of inquiry also holds significant potential for identifying novel microbiome-targeted therapies, underscoring its biomedical relevance.

To identify host proteins that interact with the microbiota, Yang and colleagues performed a proteomic screen comparing ileal mucus from germ-free (GF) mice and conventionally raised (CR) mice, and microbiota-enriched protein fractions from CR mice.⁴ This analysis identified apolipoprotein L9a and L9b (APOL9a/b) as the most enriched proteins in both the total mucus and microbiota-bound fractions of CR mice, but absent in ileal mucus of GF mice. APOL9a/b, belonging to the APOL family, are known to be interferon-inducible and are also reported to be expressed in tissues such as the liver and intestine.^5,6 Importantly, they have not been associated with interactions with the commensal bacteria. To determine which microbes these proteins target, the authors developed “APOL9-seq”, a modified version of IgA-seq, in which recombinant APOL9 proteins were used to label bacteria from CR mice before flow cytometry and 16S rRNA profiling.⁷ This approach revealed that APOL9a/b selectively bound members of genus Bacteroides, a finding confirmed by co-culture experiments with individual strains.

Reasoning that APOL9 may recognize unique lipids in Bacteroides, the authors turned to Bacteroides thetaiotaomicron (B. theta), a genetically tractable gut symbiont. Genetic deletion of enzymes required for ceramide-1-phosphate (Cer1P) synthesis in B. theta abolished APOL9 binding. This effect was observed across both mouse and human APOL orthologs, with APOL2 identified as the functional human counterpart. Direct lipid-binding assays further confirmed that APOL9a and APOL2 bound specifically to Cer1P, but not to unphosphorylated ceramides. These results establish Cer1P as the microbial ligand recognized by host APOL proteins.

Interestingly, while APOL9 binding did not affect B. theta growth under in vitro conditions, it induced membrane perturbation and enhanced outer membrane vesicle (OMV) release, as shown by electron microscopy. OMV induction was absent when Cer1P-deficient B. theta strains were used, confirming that APOL–Cer1P interactions trigger this response. Transcriptomic analysis of bacteria exposed to APOL9 or APOL2 showed upregulation of stress and membrane biogenesis genes, suggesting that OMV release reflects a stress adaptation rather than a direct antimicrobial effect.

To determine whether APOL-induced OMVs impact host immunity, the authors examined epithelial and immune responses in Apol9a/b-knockout mice. Bulk RNA-seq of ileal epithelial cells showed reduced expression of major histocompatibility complex class II (MHC-II) and interferon-γ (IFNγ) genes in knockout animals, which was restored by treatment with B. theta-derived OMVs both in vitro and in vivo. Using a co-culture system, the authors demonstrated that OMVs stimulate dendritic cells, in turn promoting epithelial MHC-II expression in a TLR2-MyD88-dependent manner.

Single-cell RNA-seq of intraepithelial CD45⁺ immune cells revealed that Apol9a/b-deficient mice had a marked reduction in CD4⁺CD8αα⁺TCRαβ⁺ intraepithelial lymphocytes, a population previously implicated in maintaining mucosal tolerance and resistance to infection. Loss of these cells correlated with impaired control of oral Salmonella typhimurium infection, leading to increased pathogen burden, systemic dissemination, and mortality. Notably, these phenotypes were rescued by oral administration of OMVs from B. theta. This study reveals an unexpected function for APOL proteins in promoting intestinal immune homeostasis through selective recognition of commensal lipids (Fig. 1). The discovery that a “non-immune” host protein can detect a “non-self” microbial lipid and initiate such a cascade challenges conventional boundaries between innate immunity, microbial sensing, and metabolic regulation. It suggests that host surveillance of the microbiota may involve a broader repertoire of molecular sensors, including metabolic proteins not traditionally associated with immune function. As the authors mention in their discussion, it is possible that commensal bacteria such as B. theta may have evolved lipid molecules like Cer1P not only for structural or metabolic purposes, but also to engage with stable, less inflammatory host pathways.⁸ Indeed, such adaptations could confer competitive advantages by promoting tolerance and long-term persistence in the gut environment. Alternatively, it is also conceivable that strains that did not adopt Cer1P synthesis were outcompeted. In this scenario, Cer1P synthesis-negative strains would have had to adopt alternative pathways to maintain their fitness in the presence of Cer1P-synthesizing strains; however, such questions would need to be addressed in a future follow-up study. Conversely, host recognition of these microbial lipids through non-immune proteins like APOL9 may reflect an evolutionary strategy to co-opt metabolic signaling for immune modulation, enabling fine-tuned responses that preserve mutualism without triggering full immune activation.

Fig. 1: APOL-mediated modulation of gut immunity through commensal lipid sensing.

Apolipoprotein L (APOL) proteins are secreted by intestinal epithelial cells into the gut lumen, where they bind specifically to B. theta via recognition of Cer1P lipids on the bacterial surface. This interaction induces the release of OMVs from B. theta, which in turn stimulate dendritic cells. IFNγ signaling promotes MHC-II expression on epithelial cells, supporting the expansion of CD4⁺CD8αα⁺TCRαβ⁺ intraepithelial lymphocytes. Illustration made using NIAID NIH BIOART: source https://bioart.niaid.nih.gov/.