Multimodal T cell profiling can enable more precise characterization of elusive cell states underlying disease. Here, we integrated single-cell RNA and surface protein data from 500,089 memory T cells to define 31 cell states from 259 individuals in a Peruvian tuberculosis (TB) progression cohort. At immune steady state >4 years after infection and disease resolution, we found that, after accounting for significant effects of age, sex, season, and genetic ancestry on T cell composition, a polyfunctional Th17-like effector state was reduced in abundance and function in individuals who previously progressed from Mycobacterium tuberculosis (M.tb) infection to active TB disease. These cells are capable of responding to M.tb peptides. Deconvoluting this state—uniquely identifiable with multimodal analysis—from public data demonstrated that its depletion may precede and persist beyond active disease. Our study demonstrates the power of integrative multimodal single-cell profiling to define cell states relevant to disease and other traits.

Background: The accuracy of different laboratory tests for diagnosis of diabetes mellitus (DM) and prediabetes (preDM) in populations exposed to tuberculosis (TB) remains poorly understood. Here, we examined the prevalence of DM and preDM in TB affected people in Lima, Peru. Methods: A prospective cohort study of patients affected TB and their household contacts (HHC), was conducted between February and November 2017 in Lima, Peru. Fasting plasma glucose (FPG), HbA1c and oral glucose tolerance test (OGTT) were used to detect DM and preDM in a prospective cohort of TB patients (n = 136) and household contacts (n = 138). Diagnostic performance of the laboratory tests was analyzed. Potential effects of sociodemographic and clinical factors on detection of dysglycemia were analyzed. Results: In TB patients, prevalence of DM and preDM was 13.97 and 30.88% respectively. Lower prevalence of both DM (6.52%) and preDM (28.99%) were observed in contacts. FPG, HbA1c and OGTT had poor agreement in detection of preDM in either TB cases or contacts. TB-DM patients had substantially lower hemoglobin levels, which resulted in low accuracy of HbA1c-based diagnosis. Classic sociodemographic and clinical characteristics were not different between TB patients with or without dysglycemia. Conclusion: High prevalence of DM and preDM was found in both TB patients and contacts in Lima. Anemia was strongly associated with TB-DM, which directly affected the diagnostic performance of HbA1c in such population.

The non-polymorphic nature of CD1 proteins creates a situation in which T cells with invariant T cell receptors (TCRs), like CD1d-specific NKT cells, are present in all humans. CD1b is an abundant protein on human dendritic cells that presents M. tuberculosis (Mtb) lipid antigens to T cells. Analysis of T cell clones suggested that semi-invariant TCRs exist in the CD1b system, but their prevalence in humans is not known. Here we used CD1b tetramers loaded with mycolic acid or glucose monomycolate to study polyclonal T cells from 150 Peruvian subjects. We found that CD1b tetramers loaded with mycolic acid or glucose monomycolate antigens stained TRAV1-2 + GEM T cells or TRBV4-1 + LDN5-like T cells in the majority of subjects tested, at rates ∼10-fold lower than NKT cells. Thus, GEM T cells and LDN5-like T cells are a normal part of the human immune system. Unlike prior studies measuring MHC-or CD1b-mediated activation, this large-scale tetramer study found no significant differences in rates of CD1b tetramer-mycobacterial lipid staining of T cells among subjects with Mtb exposure, latent Mtb infection or active tuberculosis (TB) disease. In all subjects, including "uninfected" subjects, CD1b tetramer + T cells expressed memory markers at high levels. However, among controls with lower mycobacterial antigen exposure in Boston, we found significantly lower frequencies of T cells staining with CD1b tetramers loaded with mycobacterial lipids. These data link CD1b-specific T cell detection to mycobacterial exposure, but not TB disease status, which potentially explains differences in outcomes among CD1-based clinical studies, which used control subjects with low Mtb exposure.