Identification of novel Mycobacterium tuberculosis CD4 T-cell antigens via high throughput proteome screening

Summary

Elicitation of CD4 IFN-gamma T cell responses to Mycobacterium tuberculosis (MTB) is a rational vaccine strategy to prevent clinical tuberculosis. Diagnosis of MTB infection is based on T-cell immune memory to MTB antigens. The MTB proteome contains over four thousand open reading frames (ORFs). We conducted a pilot antigen identification study using 164 MTB proteins and MTB-specific T-cells expanded in vitro from 12 persons with latent MTB infection. Enrichment of MTB-reactive T-cells from PBMC used cell sorting or an alternate system compatible with limited resources. MTB proteins were used as single antigens or combinatorial matrices in proliferation and cytokine secretion readouts. Overall, our study found that 44 MTB proteins were antigenic, including 27 not previously characterized as CD4 T-cell antigens. Antigen truncation, peptide, NTM homology, and HLA class II tetramer studies confirmed malate synthase G (encoded by gene Rv1837) as a CD4 T-cell antigen. This simple, scalable system has potential utility for the identification of candidate MTB vaccine and biomarker antigens.

Introduction

MTB causes 10⁷ active tuberculosis infections (ATBI) and kills about 1.3 × 10⁶ persons annually [27]. Globally, 1 in 3 people have latent tuberculosis infection (LTBI) [48]. The licensed BCG vaccine is poorly active against adult disease and there is a global effort to improve vaccines [37]. The immune response to MTB infection includes CD4 T-cells with rearranged T-cell receptor (TCR) alpha beta receptors that recognize peptides derived from MTB-encoded proteins bound to human leukocyte antigen (HLA) class II. Notwithstanding disappointing results from a recent vaccine clinical trial of one MTB antigen [74] the elicitation or boosting of CD4 T-cells remains a valid proposed mechanism of action for candidate vaccines [77]. It has been proposed that antigen and epitope choice may be important in the context of bacterial gene expression during different phases of MTB pathogenesis. The IFN-gamma axis is vital for host defense against MTB and CD4 T-cell decline in HIV infection is associated with severe MTB outcomes [18]. Memory T-cells also form the basis for tests for MTB infection. The tuberculosis skin test (TST) measures in vivo leukocyte infiltration in response to a filtrate of MTB cultures, while licensed interferon-gamma release assays (IGRA) measure IFN-gamma production in response to MTB peptides from three or fewer MTB open reading frames (ORFs) [7], [50]. Neither test discriminates between latent and active TB infection or predicts risk of progression from a latent to active state. There is an unmet need for biomarkers in this area.

MTB T-cell antigen discovery is thus relevant to vaccines and diagnostics. The complexity of the MTB proteome, encoded by 4000 annotated genes, has hindered systematic screening of potential antigens in MTB. Approaches, as recently reviewed [25] have included expression libraries of MTB DNA fragments [52], prediction of HLA-binding peptides MTB ORFs [81], and expression of targeted subsets of MTB ORFs based on criteria such as phase- or nutrient-dependent gene expression [10], [26] or sequence motifs associated with protein secretion [9].

Advances in high throughput in vivo recombination and E. coli lysate-based in vitro transcription/translation (IVTT) allow expression of essentially the entire translated proteome of large-genome pathogens [6], [17], [21], [34]. The proteins are useful for probing humoral responses [53]. Our lab adapted these protein collections for CD4 T-cell research for viruses encoding up to 240 proteins [32], [34]. IVTT proteins are suitable for CD4 T-cell studies because these immune cells typically detect microbial proteins after they are digested to linear peptides of 8 to about 20 amino acids. The peptides are not post-translationally modified, with recognized exceptions [55]. Here, we report a novel approach to MTB CD4 T-cell antigen discovery that uses a proteome set [45], [46] originally created for antibody studies to probe the reactivity of polyclonal MTB-specific CD4 T-cell lines. We further developed modifications of the workflow to adapt to a resource-constrained, MTB-endemic region, obtaining adequate assay performance to confirm and extend MTB antigen discovery.