Anomalies in T Cell Function Are Associated With Individuals at Risk of Mycobacterium abscessus Complex Infection

Abstract

The increasing global incidence and prevalence of non-tuberculous mycobacteria (NTM) infection is of growing concern. New evidence of person-to-person transmission of multidrug-resistant NTM adds to the global concern. The reason why certain individuals are at risk of NTM infections is unknown. Using high definition flow cytometry, we studied the immune profiles of two groups that are at risk of Mycobacterium abscessus complex infection and matched controls. The first group was cystic fibrosis (CF) patients and the second group was elderly individuals. CF individuals with active M. abscessus complex infection or a history of M. abscessus complex infection exhibited a unique surface T cell phenotype with a marked global deficiency in TNFα production during mitogen stimulation. Importantly, immune-based signatures were identified that appeared to predict at baseline the subset of CF individuals who were at risk of M. abscessus complex infection. In contrast, elderly individuals with M. abscessus complex infection exhibited a separate T cell phenotype underlined by the presence of exhaustion markers and dysregulation in type 1 cytokine release during mitogen stimulation. Collectively, these data suggest an association between T cell signatures and individuals at risk of M. abscessus complex infection, however, validation of these immune anomalies as robust biomarkers will require analysis on larger patient cohorts.

Keywords: T cells; cystic fibrosis; immunoprofiling; non-tuberculous mycobacteria; pulmonary non-tuberculous mycobacteria infection.

Figure 1

Divergent T cell activation and exhaustion profiles in cystic fibrosis (CF) patients based on non-tuberculous mycobacteria (NTM) infection status. (A) Flow cytometric analysis of ex vivo CD4⁺ T cells show significant differences in Treg percentages and marker expression between patient groups. Significantly more Tregs were seen in both CF^Act and CF^Past groups compared to CF^Contro1 (one-way ANOVA with post-hoc testing P = 0.013 and P = 0.042, respectively). Significantly higher CD25⁺ CD4⁺ T cells were observed in both CF^Act and CF^Past groups compared to CF^Contro1 group (P = 0.0056 and P = 0.037, respectively). CD25 cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) double-positive T cells were significantly higher in CF^Act than in CF^Control (P = 0.019). A reciprocal reduction in CD25 CTLA-4 double-negative CD4⁺ T cells were seen in CF^Act, CF^Past, and HC^A groups compared to CF^Control (P = 0.001, P = 0.027, and P = 0.008, respectively). CF^Act patient with active Mycobacterium avium complex (MAC) infection is shown as a blue circle in scatter plots. This data point was not included in the ANOVA analysis but is shown here to demonstrate activation and exhaustion profile of a patient with an active NTM infection that is not MABS (B) Immune marker profiling of CD4⁺ T cells by SPICE showed differences in ex vivo phenotype in the CF^Contro1 group compared to CF^Act (P = 0.0002), CF^Past (P = 0.0002), and HC^A (P = 0.005). PMA/I stimulation resulted in minor changes in maker profile with the CF^Control profile still being significantly different to CF^Act (P = 0.025) and CF^Past(P > = 0.018) though the difference with HC^Awas reduced (P = 0.057). (C) SPICE dot plots show expression levels of all combinations of markers CD25, CTLA-4, and programmed cell death protein 1 in CD4⁺ T cells both ex vivo and after PMA/I stimulation in CF patient and control groups. Groups with significantly different expression compared to HC^A (Wilcoxon rank test P < 0.05) are indicated with # symbol. Symbol color indicates significantly different group.

Figure 2

Divergent T cell cytokine profiles in cystic fibrosis (CF) patients based on non-tuberculous mycobacteria (NTM) infection status. (A) Flow cytometric analysis of ex vivo activated CD4⁺ T cells show significantly lower IFNγ⁻ TNFα⁺ (P = 0.026 and P = 0.030) and CD107a⁻ TNFα⁺ (P = 0.026 and P = 0.027) T cells in CF^Act and CF^Past groups, respectively compared to the CF^Contro1 group. Differences were not significant when comparing the HC^A group with ANOVA post-hoc testing. A similar pattern of increased TNFγ⁺ CD4⁺ T cells was seen in the HC^A group. Significantly more IFNγ⁺ TNFα⁻ CD4⁺ T cells were seen in CF^Act patients compared to CF^Contro1 patients (P = 0.0315) and significantly more IFNγ⁺ TNFα⁻ CD8⁺ T cells were seen in both CF^Act and CF^Past groups (P = 0.014 and P = 0.0047, respectively) compared to CF^Contro1 group. CF^Past had significantly more IFNγ⁺ TNFα⁻ CD8⁺ T cells compared to HC^A(P = 0.005). CF^Act patient with active Mycobacterium avium complex infection is shown as a blue circle in scatter plots. This data point was not included in the ANOVA analysis but is shown here to demonstrate cytokine profile of a patient with an active NTM infection that is not MABS (B) Polyfunctionality profiling of CD4⁺ T cells by SPICE showed differences in ex vivo functions in CF^Act and CF^Past groups compared to CF^Control (P = 0.056 and P = 0.041, respectively) and HC^A groups (P = 0.022 and P = 0.012, respectively). TNFα mono-expressing CD4⁺ T cells (blue arc) were significantly lower in the two NTM patient groups compared to the CF^Contro1 and HC^A groups. (C) SPICE dot plots show polyfunctionality profile of all combinations of cytokine expression in CD4⁺ T cells after PMA/I stimulation. Groups with significantly different expression compared to CF^Contro1 (Wilcoxon rank test P < 0.05) are indicated with # symbol. Symbol color indicates significantly different group.

Figure 3

Multivariate T cell analysis of mitogen-stimulated cytokine secretion profiles. (A) Unsupervised hierarchical clustering of mitogen-stimulated CD4⁺ and CD8⁺ T cell cytokine secretion profile show divergence based on non-tuberculous mycobacteria infection status. Each column represents an individual (patient/control) and each row represents a clustering variable (cytokine secretion pattern). Clustering patterns shown according to primary group [patient with cystic fibrosis (CF) black, or healthy control gray] and secondary group (CF^Act-light green, CF^Past-dark green, CF^Contro1-white, and HC^A-black). Clustering of CF^Ac and CF^Past patients with one exception and clustering of CF^Contro1 and HC^A with one exception. (B) Principle component analysis (PCA) and canonical co-variate analysis (CCA) show significant clustering (P = 0.02) of subjects by patient group. Marked overlap between CF^Act and CF^Past groups is seen in terms of global cytokine secretion profile indicating an overall similarity in secretion profile. CF^Contro1 group diverges further from CF^Act and CF^Past groups.

Figure 4

Top immune biomarker candidates for diagnosis of non-tuberculous mycobacteria (NTM) infection status in cystic fibrosis cohorts. (A) Flow cytometric biomarkers for NTM disease state from cytokine secretion profile data (Wilcoxon rank test). Graph shows log odds ratio for each biomarker candidate for diagnosis of NTM disease state (negative versus exposed state) with 95% CI. (P, probability; FDR, false discover rate; AUC, area under the curve). (B) Stepwise regression model (forward selection) for NTM infection status. Fitted model contains 18 variables shown in table with a final AIC of −30 and an AUC 1. Plots show contribution of each variable to model and fitted value for each patient/control (green dots) when model is applied showing an AUC = 1.

Figure 5

Evidence of CD4⁺ T cell activation and exhaustion in immunocompetent individuals with active non-tuberculous mycobacteria (NTM) infection. (A) Flow cytometric analysis of ex vivo CD4⁺ T cells showed significantly more Tregs in NTM^Act group compared to HC^B group (P = 0.028). Significantly more CD25⁺ CTLA4⁺ and CD25⁺ CTLA4⁻ CD4⁺ T cells were seen in NTM^Act group (P = 0.002 and P = 0.009, respectively) compared to HC^B. A reciprocal increase in CD25⁻ CTLA4⁻ CD4⁺ T cells was seen in the HC^B group (P = 0.001) compared to disease group. Higher numbers of PD1⁺ CD4⁺ T cells (P = 0.021) and CD25⁺ PD1⁺ CD4⁺ T cells (P = 0.011) were seen in the NTM^Act group. CD25⁺ PD1⁻ T cells were significantly higher in NTM^Act group (P < 0.001) and CD25⁻ and PD1⁻ CD4⁺ T cells were significantly higher in HC^B group (P < 0.001). (B) Phenotyping and polyfunctionality profiling of CD4⁺ T cells by SPICE showed differences in NTM^Act and HC^B groups directly ex vivo and post PMA/I stimulation. Phenotype profiles were significantly different between groups both ex vivo (P = 0.0013) and post stimulation (P = 0.022). Significantly more CD25⁺ CTLA4⁺ T cells were observed in NTM^Act groups compared to HC^B. (C) SPICE dot plots show phenotype profile of all combinations of markers CD25, CTLA4, and PD1 in CD4⁺ T cells both ex vivo and post PMA/I stimulation between NTM^Act and HC^B groups. Significantly different expression compared to HC^B (Wilcoxon rank test P < 0.05) is indicated with # symbol.

Figure 6

Divergence of CD8⁺ T cell cytokine profiles in immunocompetent individuals with active non-tuberculous mycobacteria (NTM) infection. (A) Flow cytometric analysis of ex vivo CD8⁺ T cells showed significantly more IFNγ⁻ TNFα⁺ cells (P = 0.01) and IL2⁻ TNFα⁺ cells (P = 0.048) in NTM^Act compared to HC^B. (B) Polyfunctionality profiling of CD8⁺ T cells shown in pie charts were not significantly different between the two groups. However, significantly higher TNFα mono-functional CD8⁺ T cells were observed in NTM patient group (P = 0.002). (C) SPICE dot plots show polyfunctionality profile of all combinations of cytokine expression in CD8⁺ T cells after PMA/I stimulation between NTM^Act and HC^B groups. Significantly different expression compared to HC^B (Wilcoxon rank test P < 0.05) is indicated with # symbol.