The value of transcriptomics in advancing knowledge of the immune response and diagnosis in tuberculosis

Abstract

Blood transcriptomics analysis of tuberculosis has revealed an interferon-inducible gene signature that diminishes in expression after successful treatment; this promises improved diagnostics and treatment monitoring, which are essential for the eradication of tuberculosis. Sensitive radiography revealing lung abnormalities and blood transcriptomics have demonstrated heterogeneity in patients with active tuberculosis and exposed asymptomatic people with latent tuberculosis, suggestive of a continuum of infection and immune states. Here we describe the immune response to infection with Mycobacterium tuberculosis revealed through the use of transcriptomics, as well as differences among clinical phenotypes of infection that might provide information on temporal changes in host immunity associated with evolving infection. We also review the diverse blood transcriptional signatures, composed of small sets of genes, that have been proposed for the diagnosis of tuberculosis and the identification of at-risk asymptomatic people and suggest novel approaches for the development of such biomarkers for clinical use.

Figure 1. Heterogeneity in outcomes upon exposure to M. tuberculosis.

Upon contact with an active TB patient (red), an individual with recent (white) exposure to M. tuberculosis can manifest a range of infectious states. The majority of the exposed individuals will remain asymptomatic with the possible scenarios: remain uninfected or eliminate the bacteria (purple); become infected but control the bacteria either by innate immune responses (purple) or by M. tuberculosis antigen-specific T cell response as detected by the IGRA test (gradation from purple to black); develop subclinical TB and show pulmonary abnormalities by advanced radiographic approaches, and a transient blood signature (black). A small proportion of exposed individuals will progress to active TB (red) and further represent a spectrum of infection states based on the M. tuberculosis load as measured in sputum by a smear test (indicative of high bacterial load); M. tuberculosis culture or nucleic acid amplification test (GeneXpert®); or if negative in sputum, measured in BAL, when possible (indicative of lower bacterial load) and may manifest different degrees of symptoms (different degrees of red). Adapted from Pai et al., 2016 (Ref. 3)

Figure 2. The immune response to M. tuberculosis infection.

The immune response generated in the host upon exposure to M. tuberculosis is complex and remains incompletely understood, with limited information about host factors that determine control versus progression. The cytokines IL-12, IL-1 and TNF, produced by innate immune cells, as well as IFN-γ produced by T cells, are protective against TB. Upon infection with M. tuberculosis, resident lung alveolar macrophages can become infected. (a) Early and low levels of type I IFN from macrophages, inflammatory monocytes and myeloid dendritic cells (DCs) and other innate immune cells at low mycobacterial loads can induce IL-1, IL-12 and TNF. (b) High and sustained levels of type I IFN from the macrophage and other sources (e.g. paracrine type I IFN produced by DCs upon infection with virus), can be harmful and lead to the production of the suppressive cytokine IL-10 leading to the inhibition of the production of IL-1, IL-12 and TNF by macrophages and DC, and inhibition of their activation by IFN-γ. Thus in the context of low mycobacterial loads type I IFN may be protective, whereas high mycobacterial loads and increased and sustained levels of type I IFN may result in disease progression.

Figure 3. Modular host gene signatures in tuberculosis and in other infections and diseases.

Modular approaches can be utilized to tease out subtle differences between TB and other diseases and infections, by profiling blood from patients using transcriptomics approaches, such as RNA-sequencing, to capture the entire transcriptome. Each gene within the transcriptome is expressed at a particular level across each individual sample, and genes involved in similar biological pathways are co-ordinately expressed. These groups of co-ordinately expressed genes constitute individual modules that represent discrete biological pathways and can be identified using unbiased approaches such as weighted gene co-expression network analysis (WGCNA). Perturbation as a response to infection with M. tuberculosis or other pathogens, can be measured within each module of co-expressed genes, compared to healthy controls. Using such an approach, modular signatures can be identified for TB and other infections and diseases, to inform on the immune response, and this information can also be utilized to develop reduced gene signatures that are more specific to TB to develop biomarkers for diagnosis.