Fig 1.
The esxH gene is highly polymorphic among clinical Mtb isolates.
(A) Reported TB10.4 epitopes recognized by human T cells from the IEDB database (www.iedb.org). Y axis represents the length of the 96 amino-acid TB10.4 protein. Each horizontal red bar spans an individual reported epitope. Grey shaded areas represent mouse TB10.4-specific T cell epitopes: H2-Kb-restricted TB10.44−11, H2-Kd-restricted TB10.420−28, and I-Ad-restricted TB10.474−88. (B) The locations of EsxH polymorphisms are represented along the linear sequence of the protein (green), with the number of isolates containing each polymorphism based on the whole genome sequences from Holt et al and Walker et al [25, 26]. Regions of the protein with a high frequency of polymorphisms are shaded grey. The bar color corresponds to the different Mtb lineages: lineage 1 (L1), purple; L2, blue; L3, white; and, L4, red. The table shows the total numbers of polymorphisms in each lineage, and the χ2 and p values compared to L1. (C) Whole-genome SNP-based phylogenies represent the polymorphisms across all clinical isolates within lineage 1. Each circumference line signifies a distinct polymorphism indicated in the grey box labeled “Lineage 1.” Each red dot cluster represents each time the indicated polymorphism evolved independently. Red stars designate the A10V polymorphism which evolve separately three times. The table shows the acquisition rate of esxH variations based on the identification of genome-wide SNP events within each phylogeny and the number of events which were within esxH.
Fig 2.
The hierarchy of immunodominant CD8 T cell responses after Erdman vs. 667 infection.
Lung T cell responses were evaluated five weeks after infection of C57BL/6J mice with ~100 aerosolized Erdman or 667 bacilli. (A) Representative flow cytometry plots of MTB32a309-318 and TB10.44−11 tetramer staining of pulmonary CD8 T cells from Erdman- or 667-infected C57BL/6J mice. The frequency (B) or absolute number (C) of TB10.4- or MTB32a-specific CD8, or ESAT6-specific CD4 T cells in the lungs of infected mice after Erdman or 667 infection. (D) The frequency (left) and total numbers (right) of Mtb-specific CD4 or CD8 T cells, from nine independent paired Erdman (closed symbols) and 667 (open symbols) infections. Each point is the average of 4–8 mice and lines connect the paired infections. (E) The number of TB10.4-specific CD8 T cells in the lungs of mice infected with Erdman (closed symbols) or 667 (open symbols) was quantified by staining the different cell populations using the TB10.44−11/Kb tetramer (i.e., loaded with IMYNYPAM) or the A10TTB10.44−11/Kb tetramer (i.e., loaded with IMYNYPTM) separately using individual mice from 4 independent paired infections, 38 mice in total. The data was pooled and analyzed. The diagonal line is the line of unity for this analysis. (F) Absolute numbers of TB10.4-specific CD8 T cells in the lungs of mice infected with Erdman or 667. The number of cells was determined by staining with the TB10.44−11/Kb tetramer (i.e., loaded with IMYNYPAM) or the A10TTB10.44−11/Kb tetramer (i.e., loaded with IMYNYPTM). Nine independent experiments were performed and analyzed 4–6 weeks post infection, with a total of 96 mice. Statistical testing by a two-tailed, unpaired Student’s T test. **, p<0.01; ***, p<0.005; and ****, p<0.0001.
Fig 3.
Infection with 667 is less virulent than Erdman.
The bacterial burden is measured by CFU from lung (A) or spleen (B) homogenates from Erdman (filled) or 667 (open) infected C57BL/6J at different timepoints post infections. CFU data were compiled from 13 (lung) or 6 (spleen) independent experiments, from 4 to 30 weeks post infections. (C) The survival of C57BL/6 mice after Erdman (solid) or 667 (dashed) infection, which is one of two independent results with similar results. In this experiment, the d1 CFU was 158 (667) or 55 (Erdman). (D-E) A barcoded pool of clinical Mtb isolates was administered intravenously to C57BL/6J (black) or RAG1 KO (red) mice, and 4–8 mice of each strain were harvested for lungs and spleens to recover bacteria at 1 (circle), 14 (triangle), and 21 (square) days post infection. (D) The pseudo CFU of 667 and Erdman in spleen were determined by the relative abundance of the respective Mtb strains multiplied by total CFU in the spleen. (E) CFU fold-change (versus d1 CFU) of 667 or Erdman abundance in the spleen was compared between C57BL/6J (black) and RAG1 KO (red) mice. (F) CFU fold-change (versus d1 CFU) of 667 (open) vs. Erdman (closed) in the spleens of C57BL/6J (black) and RAG1 KO (red) mice. Statistical significance of survival curves (C) was determined by log-rank (Mantel-Cox) test; p value is shown. Statistical significance of CFU fold-change (D, E, F) were analyzed by a two-way ANOVA with Tukey (D) or Sidak’s (E, F) multiple comparison test. p values are indicated by asterisks: **, p<0.01, ***, p<0.001, ****, p<0.0001. Not all comparisons are shown for clarity.
Fig 4.
The A10T polymorphism leads to a change in immunodominance of the CD8 T cell response after 667 infection.
C57BL/6J mice were infected with Erd.EsxHA10T or Erd.EsxHWT by the aerosol route and the lung T cell response was analyzed five weeks later by (A-B) tetramer staining and (C) the bacterial burden. (A) Representative flow cytometry plots of the frequency of MTB32a309-318- and TB10.44−11-specific CD8 T cells (left), or ESAT6-specific CD4 T cells (right), after Erd.EsxHWT or Erd.EsxHA10T infection in the lungs of C57BL/6J mice. (B) The frequency and absolute number of antigen-specific T cells elicited by five independent paired infections of isogenic Erd.EsxHWT (open) and Erd.EsxHA10T (closed) strains determined by tetramer staining. The TB10-specific CD8 T cell responses elicited by the isogenic strains Erd.EsxHWT and Erd.EsxHA10T, was determined using the specific tetramers WTTB10.44−11/Kb and A10TTB10.44−11/Kb tetramers, respectively. Each point is the average of 5 mice and lines connect the paired infections. (C) The bacterial burden was measured by CFU from lung (left) or spleen (right) homogenates from Erd.EsxHWT or Erd.EsxHA10T infected C57BL/6J at different timepoints post infections. CFU data were compiled from 4 independent experiments, from 5 to 19 weeks post infections. Statistical significance was determined by a two-tailed, unpaired Student’s T test. p values are indicated by asterisks: *, <0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.
Fig 5.
T cell recognition of infected macrophages.
CD4 (A, C) or CD8 T cells (B, D) purified from the lungs of Erdman- or 667-infected mice were cultured with H37Rv-infected (A, B) or 667-infected (C, D) macrophages and the MIM-ICS assay was performed. MIM-ICS assay of isogenic strains elicited (E) CD4 T cells and (F) CD8 T cells in recognizing H37Rv-infected macrophages. Isogenic strains elicited CD8 T cells were also measured for their recognition ability of (G) Erd.EsxHWT-infected and (H) Erd.EsxHA10T-infected macrophages. The X axis is the actual multiplicity of infection (MOI) as determined by CFU plating. Statistical significance was determined by multiple t testing, and p values <0.05 are shown above the bars. (I) Erd.EsxHWT- or Erd.EsxHA10T-elicited CD8 T cells were cultured with uninfected macrophages (UI) or macrophages plus the indicated peptide epitopes or anti-CD3 mAb. Statistical significance was determined by one-way Anova, and p values <0.05 are shown above the bars. Results are representative of two (A-H) or three (I) different independent experiments.
Fig 6.
Naturally occurring variants of the TB10.44−11 epitopes bind Kb and stimulate TB10.44−11-specific CD8 T cells.
(A) RMA-S Kb stabilization assay. Variant peptides were titrated, added to RMA-S cells, and the relative H-2 Kb surface expression (MFI) was determined by flow cytometry. The following peptides were used: SIINFEKL (positive control), IMANAPAM (negative control), “WT” TB10.44−11, and the following variants of TB10.44−11: A10T, A10V, P9S, and M11I. (B) The proliferation of the TB10.44−11-specific CD8 T cell line TB10RgR was measured by eFluor450 dilution 48 hours after co-culture with macrophages pulsed with titrated amounts of the indicated peptides. (C) Proliferative response of the TB10RgP, TB10Rg3, TB10RgR, and TB10RgL CD8 T cell lines 48 hours after co-culture with macrophages pulsed with titrated amounts of WT TB10 peptide (left) or A10T peptide (right). Each assay was repeated 4 times.
Fig 7.
Different peptide degradation patterns of TB10.41−34 with alanine or threonine at position 10 in macrophage lysosomes.
(A) Amino acid sequence of 34-mer peptides (i.e., TB10.41−34) containing alanine (WT) or threonine (A10T) at residue 10 (B) Experimental scheme of the in vitro peptide degradation assay in macrophage lysosomal extracts and WT and A10T TB10.4-1-34-mer sequences. (C) N- (left) and C- (right) terminus cleavage sites determined at 60 minutes. The relative amount of peptides starting (left panel) or ending (right panel) at each residue was quantified during the degradation of WT (black bars) and A10T (open bars) TB10.4-1-34-mer peptides. N = 6 experiments; *, p<0.05. (D) Production of TB10.44−11 epitope (left panel) and N-extended TB10.44−11 (right panel) from WT (black circles) or A10T 34-mer (open circles) at 60 minutes. One representative experiment. (E) Production of TB10.44−11 epitope from WT (black circles) and A10T (open circles) TB10.4-1-34-mer peptides was determined at 10, 30, 60 minutes in six independent experiments. P values calculated with Wilcoxon matched-pairs signed rank test.