About

Chaitan Khosla is a professor in the Department of Biochemistry at Stanford University. His research interests lie at the interface of chemistry and medicine, with a focus on the catalytic mechanisms of modular megasynthases such as polyketides synthases. His work aims to harness the programmable chemistry of these enzymes for the development of new antibiotics. More recently, his research has expanded to investigate the pathogenesis of celiac sprue, an autoimmune disease of the small intestine associated with HLA-DQ2. Professor Khosla has been recognized for his contributions to the enzymology of polyketide synthesis and Celiac Disease, including receiving the 2026 Abeles and Jencks Award for the Chemistry of Biological Processes.

Research topics

• Medicine
• Immunology
• Pathology
• Biology
• Computer Science
• Virology
• Internal medicine
• Business
• Bioinformatics
• Risk analysis (engineering)
• Computational biology
• Chemistry
• Biotechnology
• Pharmacology
• Data science

Selected publications

• Mechanistic Analysis of Programmed Iteration by Module 5 of the Nocardiosis-Associated Polyketide (NOCAP) Synthase

  Biochemistry · 2026-03-09

  articleSenior authorCorresponding

  Assembly line polyketide synthases (PKSs) possess multimodular architectures in which each module harbors the requisite protein domains to catalyze a single round of polyketide chain elongation and postelongation modifications. Exceptions to this paradigm are modules that catalyze multiple elongation cycles, a phenomenon referred to as “programmed iteration”. The molecular mechanism that allows PKS modules to iterate remains poorly understood. For example, Module 5 of the nocardiosis-associated polyketide (NOCAP) synthase catalyzes three elongation cycles during the biosynthesis of its undecaketide product, although in the absence of downstream modules, it has been shown to catalyze five elongation cycles. To understand the context-dependent control of its iterative capacity, we combined in vitro analysis of purified Module 5 of the NOCAP synthase with in vivo studies in Escherichia coli. Our findings reveal that, while the ability to iterate is an inherent property of Module 5, protein–protein interactions with its downstream module (Module 6) are key determinants of the number of elongation cycles catalyzed by Module 5 within the context of the complete assembly line. We also show that the intrinsic ability of Module 5 to iterate can be strongly influenced by the identity of its substrate. Our findings highlight the potential of Module 5 of the NOCAP synthase to reveal fundamentally new insights into the mechanistic differences between iterative and assembly line PKSs.

  PublisherDOI

• Association of Rare <i>APOE</i> Missense Variants R189C and W276C With Risk of Alzheimer Disease

  medRxiv · 2026-05-12

  article

  Abstract Rare APOE missense variants in admixed populations can clarify mechanisms of Alzheimer disease (AD) risk beyond ε2 and ε4. We analyzed APOE coding variation in the Alzheimer disease Sequencing Project and AllofUs and meta-analyzed ancestry-adjusted case-control associations. Two ε3-linked variants enriched in Native American ancestry among Admixed American individuals were associated with AD: R189C increased risk (odds ratio (OR), 3.35; 95% confidence interval (CI), 1.25–9.00; P = 0.016), whereas W276C was protective (OR, 0.28; 95% CI, 0.10–0.82; P = 0.020). In a nascent ApoE secretion assay that resolves high molecular weight lipid-bound species from lipid-poor species, R189C showed an APOE- ε4-like shift toward the lipid-bound fraction. In an orthogonal self-association assay, W276C reduced ApoE self-association to the protective APOE -Jacksonville variant level. These findings expand the spectrum of APOE missense variants that are associated with AD and implicate C-terminal ApoE conformation and lipidation states as tractable mechanisms for pathogenesis and therapeutic targeting.

  PublisherDOI

• Self-Resistance Guided Discovery of a Hybrid Polyketide-Peptide Antibiotic from <i>Vibrio ruber</i>

  Journal of the American Chemical Society · 2025-06-25 · 1 citations

  articleSenior authorCorresponding

  DSM 16370 based on its colocalization with a gene encoding an aminoacyl-tRNA synthetase homologue. Biological assays confirmed that the biosynthetic pathway produced an antibacterial metabolite while also verifying assignment of the aminoacyl-tRNA synthetase as a self-resistance mechanism. The structures of five closely related compounds of mixed polyketide─nonribosomal peptide origin were elucidated. Sequence analysis, mutagenesis and stable isotope feeding studies led to decoding of a plausible biosynthetic pathway for these natural products. The primary products, a trio of interconverting isomers designated vibriomycins A, B, and B', were found to be unusual allosteric inhibitors of bacterial threonyl-tRNA synthetases. Our study highlights the feasibility of discovering new anti-infective chemotypes via genome mining.

  PublisherDOI

• Peyer’s Patch B Cells Sample Transglutaminase-Gluten Complexes and Drive Celiac Disease Autoimmunity

  Gastroenterology · 2025-07-02 · 6 citations

  articleOpen access

  BACKGROUND & AIMS: T cells in gut-associated lymphoid tissues (GALT) via the formation of transient enzyme-substrate complexes formed between TG2 and gluten. Where in the body enzymatically active TG2 encounters gluten peptides remains unknown. METHODS: A model to study the celiac disease-relevant T-cell-B-cell interactions in GALT has been developed. Mice expressing HLA-DQ2.5 received TG2-specific B cells and gluten-specific T cells by adoptive transfer and were subsequently orally immunized with a model antigen containing the the B-cell and T-cell epitopes. RESULTS: Orally immunized HLA-DQ2.5 knockin mice developed TG2-specific gut immunoglobulin (Ig)A and serum IgG responses. Activated TG2-specific B cells were present in Peyer's patches and in gut-draining mesenteric lymph nodes, resembling what is seen in human celiac disease. We demonstrate that TG2-specific B cells in Peyer's patches sample TG2 when the protein is perfused into the gut lumen. CONCLUSION: The model supports a mechanism in which TG2-gluten complexes formed in the gut lumen are taken up by TG2-specific B cells in GALT. We propose that this pathway plays an important role in driving the anti-TG2 IgA autoantibody response in patients with celiac disease. The model provides a platform to explore novel approaches for celiac disease therapies.

  PublisherDOI

• A dendritic cell population responsible for transglutaminase 2-mediated gluten antigen presentation in celiac disease

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-06-05

  preprintOpen accessSenior authorCorresponding

  Abstract In celiac disease (CeD), a gluten-dependent autoimmune disorder, transglutaminase 2 (TG2) deamidates selected glutamine residues in gluten peptides, while HLA-DQ2 presents deamidated antigens to inflammatory T cells. The cellular sources of pathogenic TG2 and DQ2 are unclear. Using chemical biology tools, we show that intestinal CD103 + dendritic cells (DCs) couple cell-surface TG2 to the endocytic LRP1 receptor to simultaneously deamidate gluten antigens and concentrate them in lysosomes. In DQ2-transgenic mice, CD103 + DCs loaded with deamidated antigens migrate from intestinal lamina propria and Peyer’s patches into mesenteric lymph nodes, where they engage T cells. In turn, gluten antigen presentation upregulates intestinal TG2 activity. The tool (HB-230) used to establish a role of CD103 + DCs in gluten antigen presentation and TG2 activation in mice also revealed that the TG2/LRP1 pathway is active in human CD14 + monocytes. Within this population of circulating monocytes, a DC subset with the gut-homing β7-integrin marker is elevated in CeD patients with active disease compared to non-celiac controls or patients on a gluten-free diet. Our findings not only inform the cellular basis for gluten toxicity in CeD but they also highlight the immunologic role of an enigmatic protein of growing therapeutic relevance in CeD and other immune disorders.

  PublisherOA PDFDOI

• ApoE is Secreted as a Lipid Nanoparticle by Mammalian Cells: Implications for Alzheimer’s Disease Pathogenesis

  Biochemistry · 2025-10-24 · 1 citations

  articleSenior authorCorresponding

  The brain is the most cholesterol-rich organ in the body, and ApoE is the main lipid carrier protein in the brain. Although very little, if any, ApoE exists in its apoprotein form in physiological fluids, recombinant ApoE is typically prepared in a lipid-free state to study its physiological functions. We describe a lipid nanoparticle (LNP) form of ApoE as a primary extracellular product of the eukaryotic protein export system. Whereas the apoprotein is the dominant secreted product when the APOE gene is overexpressed in mammalian cells, an LNP form of ApoE is also observed. The LNP form is, however, the major secreted product from unmodified CCF-STTG1 astrocytoma cells. The C-terminal domain of ApoE plays a key role in LNP biosynthesis as the ApoE3 W210* truncation mutant is secreted without lipidation. Secreted ApoE LNPs are markedly better substrates than the apoprotein itself for further growth via the action of ATP-dependent lipid pumps. Compared to ApoE3 or the Alzheimer’s disease-protective ApoE2 variant, the recovered yield of the LNP form of the disease-predisposing ApoE4 variant is higher. Intriguingly, the LNP yield of the rare disease-protective R251G variant of ApoE4 is comparable to that of ApoE3 and ApoE2. Analogous to the well-documented intracellular biosynthesis of ApoB-containing LNPs, the biogenesis and pathophysiological relevance of the LNP form of ApoE warrant further investigation.

  PublisherDOI

• A dendritic cell population responsible for transglutaminase 2–mediated gluten antigen presentation in celiac disease

  JCI Insight · 2025-08-28 · 2 citations

  articleOpen accessSenior author

  In celiac disease (CeD), a gluten-dependent autoimmune disorder, transglutaminase 2 (TG2), deamidates selected glutamine residues in gluten peptides, while HLA-DQ2 presents deamidated antigens to inflammatory T cells. The cellular sources of pathogenic TG2 and DQ2 are unclear. Using chemical biology tools, we show that intestinal CD103+ dendritic cells (DCs) couple cell-surface TG2 to the endocytic LRP1 receptor to simultaneously deamidate gluten antigens and concentrate them in lysosomes. In DQ2-transgenic mice, CD103+ DCs loaded with deamidated antigens migrate from intestinal lamina propria and Peyer's patches into mesenteric lymph nodes, where they engage T cells. In turn, gluten antigen presentation upregulates intestinal TG2 activity. The tool (HB-230) used to establish a role of CD103+ DCs in gluten antigen presentation and TG2 activation in mice also revealed that the TG2/LRP1 pathway is active in human CD14+ monocytes. Within this population of circulating monocytes, a DC subset with the gut-homing β7-integrin marker is elevated in patients with CeD with active disease compared with nonceliac controls or patients on a gluten-free diet. Our findings not only inform the cellular basis for gluten toxicity in CeD, but they also highlight the immunologic role of an enigmatic protein of growing therapeutic relevance in CeD and other immune disorders.

  PublisherOA PDFDOI

• Individual Patient Data Meta-Analysis evaluating Camostat Mesilate to Treat COVID-19 in Community Settings

  medRxiv · 2024-05-17

  preprintOpen access

  Abstract Background In the COVID-19 pandemic, a number of phase II and III randomized trials were launched to evaluate the effectiveness of camostat, an orally administered TMPRSS2 inhibitor previously approved for other indications, for treating SARS-CoV-2 infections. Owing to the rapidly changing landscape during the pandemic, many of these trials were unable to reach completion. Further, methods for synthesizing data for trials that were launched and not completed were critical. Methods This study aimed to consolidate global evidence by identifying placebo-controlled, randomized trials of camostat and analyzing their collective clinical and virologic impact on SARS-CoV-2 through an individual participant data meta-analysis. We harmonized data from the included studies and utilized Bayesian statistical models to assess virologic outcomes (measured by the rate of change in viral shedding) and clinical outcomes (based on the time to the first of two consecutive symptom-free days), adjusting for age and sex. Findings The meta-analysis incorporated data from six countries, totaling 431 patients across the studies; 118 patients contributed data for the primary virologic outcome and 240 for the clinical symptom outcome. Camostat did not improve the rate of change in viral load (difference in rate of change = 0.11 Ct value/day higher, 95% credible interval 2.04 lower to 2.23 higher) or time to symptom resolution (hazard ratio = 0.87, 95% credible interval 0.51, 1.55) when compared to placebo. Interpretation In a meta-analysis prompted by a fast-changing landscape during the pandemic, we jointly synthesized evidence across multiple trials that did not meet their original recruitment goals. Despite its theoretically promising mode of action, camostat did not demonstrate a statistically significant virologic or clinical benefit in treating COVID-19, highlighting the complexity of drug repurposing in emergency health situations. Funding This work was partially supported by The Lundbeck Foundation, LifeArc, Assistance Publique Hôpitaux de Paris, anonymous donors, and awards from the National Institutes of Health. Research in context Evidence before this study Camostat mesilate, a therapy widely used in Japan for over two decades to treat pancreatitis and reflux esophagitis, showed promise against SARS-CoV-2 in early laboratory and animal studies. Numerous studies evaluating camostat as a treatment for COVID-19 were launched by autumn of 2020, but later stalled due to emerging treatments that altered the equipoise for placebo-controlled trials. Among the trials that reached publication, findings were mixed. Added value of this study Our research brings a fresh perspective by comprehensively analyzing both published and previously unseen data from randomized clinical trials on camostat. By pooling data across studies, our analysis provides a more robust assessment of the effectiveness of camostat against viral and clinical outcomes than any single study could offer. Novel analytic approaches, data sharing efforts, and international collaboration during the global health emergency are additionally described. Implications of all the available evidence After thorough analysis, our study concludes that, when considering all available data, camostat does not confer a virologic or clinical advantage in the treatment of COVID-19. This conclusion underscores the importance of pooling global research efforts to build a clearer understanding of potential treatments during health emergencies.

  PublisherOA PDFDOI

• Abstract 2119 Structural and Mechanistic Analysis of Ca2+ Dependent Regulation of Transglutaminase 2 Activity

  Journal of Biological Chemistry · 2024-03-01

  articleOpen accessSenior author

  Mammalian transglutaminases, a family of Ca2+ dependent proteins, are implicated in a variety of diseases. For example, celiac disease is an autoimmune disorder whose pathogenesis requires Transglutaminase 2 (TG2) to deamidate selected glutamine residues in diet-derived gluten peptides. Deamidation involves the formation of transient ϒ-glutamylthioester intermediates. Recent studies have revealed that, in addition to deamidated gluten peptides themselves, the corresponding thioester intermediates are also pathogenically relevant, although a mechanistic understanding of this relevance is hindered by the absence of any structure of Ca2+-bound TG2. We report the X-ray crystallographic structure of human TG2 bound to an inhibitory gluten peptidomimetic and two calcium ions in sites previously designated as S1 and S3. Site-directed mutagenesis confirmed that S1 regulates the formation of an inhibitory disulfide bond while also establishing that S3 is solely responsible for forming and resolving the ϒ-glutamylthioester intermediate. Further structure-based mutagenesis also demonstrated that two interacting residues, H305 and E363, play a critical role in resolving the thioester intermediate into an isopeptide bond (transamidation) but not thioester hydrolysis (deamidation). Together, our findings lead us to conclude that, although TG2 can bind to Ca2+ ions at least five allosteric sites, only one of them (S3) is necessary for both halves of the catalytic cycle.

  PublisherOA PDFDOI

• Discovery and Characterization of the Fully Decorated Nocardiosis-Associated Polyketide Natural Product

  Journal of the American Chemical Society · 2024-01-31 · 6 citations

  articleOpen accessSenior authorCorresponding

  The genomes of 40 strains of Nocardia, most of which were associated with life-threatening human infections, encode a highly conserved assembly line polyketide synthase designated as the NOCAP (NOCardiosis-Associated Polyketide) synthase, whose product structure has been previously described. Here we report the structure and inferred biosynthetic pathway of the fully decorated glycolipid natural product. Its structure reveals a fully substituted benzaldehyde headgroup harboring an unusual polyfunctional tail and an O-linked disaccharide comprising a 3-α-epimycarose and 2-O-methyl-α-rhamnose whose installation requires flavin monooxygenase-dependent hydroxylation of the polyketide product. Production of the fully decorated glycolipid was verified in cultures of two patient-derived Nocardia species. In both E. coli and Nocardia spp., the glycolipid was only detected in culture supernatants, consistent with data from genetic knockout experiments implicating roles for two dedicated proteins in installing the second sugar substituent only after the monoglycosyl intermediate is exported across the bacterial cell membrane. With the NOCAP product in hand, the stage is set for investigating the evolutionary benefit of this polyketide biosynthetic pathway for Nocardia strains capable of infecting human hosts.

  PublisherOA PDFDOI

Recent grants

• NIH Grant R01GM067937

  NIH · $489k · 2006

• Mechanisms and Evolution of Assembly-Line Polyketide Synthases

  NIH · $2.4M · 2021–2026

• NIH Grant R21DK065965

  NIH · $290k · 2005

• Localizing Pathogenically Relevant Transglutaminase 2 in Celiac Disease

  NIH · $1.8M · 2021–2026

• Structure, Mechanism, and Engineering of Assembly Line Polyketide Synthases

  NIH · $4.9M · 1995–2022

Frequent coauthors

• David E. Cane

  495 shared

• Alice Y. Chen

  69 shared

• Christopher T. Walsh

  Met Office

  62 shared

• John R. Jacobsen

  57 shared

• Rajesh S. Gokhale

  Indian Institute of Science Education and Research Pune

  55 shared

• Shiou‐Chuan Tsai

  University of California, Irvine

  53 shared

• Ludvig M. Sollid

  University of Oslo

  52 shared

• Chiara R. Valenzano

  50 shared

Labs

• Khosla LabPI

  Chemical Biology @ Stanford

Awards & honors

• 2026 Abeles and Jencks Award for the Chemistry of Biological…