Determining T Cell Responses To Ntm Specific Peptides
The closer relatedness of the species in the mycobacteria genus and the lack of NTM-specific and MTB-specific reagents make it difficult to distinguish NTM exposure from MTB exposure. With the recent identification of T cell epitopes unique to NTMs by Lindestam Arlehman et al.,(Arlehamn et al., 2015) we set out to contribute to solving of this paradox by determining T cell responses to these NTM specific peptides and to ESAT-6 in a Ugandan population whose NTM and MTB sensitisation status had been determined using both TST and TSPOT.TB assays.
Firstly we set out to determine the agreement between TST – T-SPOT.TB categorisation of NTM and MTB sensitisation. We then investigated in detail the T cell phenotypes and assessed their functionality using multi-parameter flow cytometry. We observed slight agreement between TST – T-SPOT.TB and ELISpot categorisation of NTM sensitisation. Following stimulation with NTM-specific peptide pools in the ELISpot assay, we expected almost perfect agreement especially from samples that had been obtained from study participants who had a negative T-SPOT.TB result and a TST wheal diameter between 1 to 9 millimetres. Existing literature states that NTMs induce smaller TST reactions compared to MTB (Jeanes, 1969; Mandalakas & Starke, 2005) and, a negative T-SPOT. TB.
Similarly, slight agreement was observed between TST – T-SPOT.TB and ELISpot categorisation of MTB sensitisation (Immunotec, 2013). Following stimulation with ESAT-6 in the ELISpot assay, we expected almost perfect agreement especially from samples that had been obtained from study participants with a positive T-SPOT.TB result and a TST wheal diameter greater than 10 millimetres, significant of MTB sensitisation based on existing guidelines for MTB diagnosis. (Ministry of Health Uganda, 2010; WHO, 2015a) The slight agreement observed between TST – T-SPOT.TB and ELISpot categorisation of NTM and MTB sensitisation might have been influenced by a systematic error arising from the use of cryopreserved rather than fresh PMBCs; antigens used in the ELISpot assay; and interpretation of the ELISpot assay results.
Cryopreservation of PBMCs for more than 1 year has been associated with loss of T cell function, alteration of T cell phenotype due to loss of characteristic markers, and apoptosis. A seminal report on lessons drawn from flow cytometry and intracellular cytokine staining by the Tuberculosis Vaccine Initiative (TBVI) TB Biomarker Working group United Kingdom highlighted the following setbacks with regards to use of cryopreserved rather than fresh PBMCs for immunophenotyping. There was a significant reduction in the total cytokine (IFNγ, IL2, TNF) expression following stimulation of thawed cells with PPD, ESAT-6 and heparin-binding haemaggluttinin (HBHA).(S. G. Smith et al., 2015).
Additionally, there was a significant reduction in the frequencies of single, dual and triple cytokine secreting cells following stimulation with HBHA. (S. G. Smith et al., 2015) Another study assessing suitability of long term cryopreserved PBMCs for HIV immune monitoring in HIV infected patients in USA highlighted the loss of CD4+ T cell IFNγ responses to HIV whole Gag protein, HIV peptides and to CMV lysate; increase in caspase-3 expression in CD4+ T cells thus the proclivity to apoptosis; and loss of naïve CD4+ and CD8+ T cells due to reduced expression of characteristic markers CD28 on CD4+ and CD45RA on CD8+ T cells.
(Rachel E. Owen, Elizabeth Sinclair, Brinda Emu, John W. Heitman, Dale F. Hirschkom, C. Lorrie Epling, Qi Xuan Tan, Brian Custer, Jeffery M. Harris, Mark A. Jacobson, Joseph M. McCune, Jeffery N. Martin, Frederick M. Hechet, 2007) Therefore, there is a possibility that long-term cryopreservation of PBMCs contributed to the low IFNγ responses to NTM and ESAT-6 peptide pools obtained in the ELISpot assay in the present study. Due to the low IFNγ responses to NTM and ESAT-6 peptide pools obtained in the ELISpot assay, we deviated from the standard protocol and re-defined the criterion for categorisation of NTM and MTB sensitisation.
The standard protocol, which had been optimised using fresh PBMCs, required a high spot count i.e. greater than 20 spots per well in the NTM and ESAT-6 or CFP-10 stimulated wells, for an NTM or MTB positive result respectively. For the new criterion, we calculated the mean spot count (2.53 spots) in the un-stimulated wells for all the 33 samples and any spot count above this mean in the NTM-specific peptide pool and ESAT-6 peptide pool stimulated wells was considered positive for NTM sensitisation and MTB sensitisation respectively.
However, still with this new assay cut off, some less reactive samples that could have been potentially positive for NTM or MTB were classified as negative. This discrepancy was evinced by the significant IFNγ responses (expressed as spot forming count per million PBMC) to NTM and to ESAT-6 peptide pools that were higher in the 7 NTM sensitised compared with 14 Un-sensitised controls and in the 12 MTB sensitised compared with 14 Un-sensitised controls respectively. This data shows that the ELISpot assay cut off used could have influenced the slight agreement observed between TST – T-SPOT.TB and ELISpot categorisation of NTM and MTB sensitisation.
The use of ESAT-6 only to detect MTB sensitisation could have also contributed to the slight agreement between TST – T-SPOT.TB and ELISpot categorisation of MTB sensitisation. T-SPOT.TB is a commercial kit that relies on use of both ESAT-6 and CFP-10 to detect MTB sensitisation. There is a possibility that the 9 out of 12 samples which were flagged MTB negative by ELISpot could have been high CFP-10 responders and less or none – ESAT-6 responders. However, the choice for use of ESAT-6 in the present study was based on findings from studies conducted in guinea pigs, cattle, cynomolgus macaques and humans in the Gambia, Denmark, Holland, Zambia, US and Japan showing that ESAT-6 had equivalent or higher response rates compared to CFP-10.
(Arend et al., 2000; Chapman et al., 2002; Ellis et al., 2017; Fox et al., 2007; Laurens, Ravn, Agger, Pollock, & Andersen, 2000; Mori et al., 2004) Secondly, we were also limited by the PBMC counts, thus the need to reduce on the panel of stimuli. Following analysis of PBMC samples with multi-parameter flow cytometry, we did not detect significant differences in the CD4+ and CD8+ T cell memory subset profiles between NTM sensitised and Un-sensitised controls. However, when we examined for polyfunctionality after stimulation with NTM-specific peptide pools, we identified significantly higher dual functional IFNγ+ IL2+ CD4+ and CD8+ central memory T cells in NTM sensitised category compared to the Un-sensitised controls.
Our study complements several other studies that have highlighted the role of IFNγ and IL2 in the control of infections due to NTMs. (Chan, Bai, Kartalija, Ian M. Orme, & Ordway, 2010; Lutzky et al., 2018; Mortaz, Moloudizargari, Varahram, & Movassaghi, 2018) We identified significantly higher triple non-functional IFNγ- IL2- TNF- CD4+ effector memory T cells in the NTM sensitised compared to the Un-sensitised controls, following ESAT-6 stimulation. This implies that either the study participants in the NTM sensitised category are not exposed to MTB or that the NTMs they are exposed to do not produce ESAT-6 homologs. We did not detect TNF secreting cells in the NTM sensitised study group at a level that was significantly higher than in the Un-sensitised controls.
The role of TNF in the control of NTM disease is still controversial. Least incidences of NTM lung disease were reported among patients treated with TNF agonists whereas fatal M. avium-intracellulare lung disease, fatal pulmonary M. xenopi infection, fatal pulmonary M. abscessus infection, M. chelonae endophthalmitis, M. xenopi spinal osteomyelitis and pulmonary M. szulgai infection were reported among patients treated with Etanercept TNF agonist. (Sexton & Harrison, 2008) In more recent studies carried out in Israel, Taiwan, Germany, Korea, US and Australia, no significant differences were detected in TNF production in patients with pulmonary disease due to NTMs or Cystic fibrosis patients with NTM infection, when compared with healthy controls.
(Bar-On et al., 2015; Greinert, Schlaak, Rũsch-Gerdes, Hans-Dieter, & Ernst, 2000; Kwon et al., 2007; Lutzky et al., 2018; Sampaio et al., 2008; Shu et al., 2017) Given the significant role of TNF in the inhibition of mycobacterial growth in the macrophages, (Balcewicz-Sablinska, Keane, Kornfeld, & Remold, 1998; Bermudez & Young, 1988; Jayaraman et al., 2013; Roach et al., 2002) the low TNF production reported in the latter studies and in our study could be a subversion strategy by the pathogen to persist in the host.
On the contrary, we identified significantly higher dual functional IL2+ TNF+ CD4+ Naïve T cells following stimulation with NTM-specific peptide pools, and dual functional IFNγ+ IL2+ CD4+ Naïve T cells following ESAT-6 stimulation (Figure 10), in Un-sensitised controls compared to the NTM sensitised study group. The production of antigen-specific cytokines by naïve CD4 T cells in the Un-sensitised controls may represent the ability of a healthy person to respond to infection. In a study utilising mouse peritoneal and human macrophages (obtained from a healthy individual), cells stimulated with TNF alone or in combination with IL-2 showed enhanced killing of Mycobacterium avium complex compared to cells injected with IFNγ.
In the MTB sensitised study group, there were significantly higher CD4+ Effector memory (CD45RA- CCR7-) and Terminal effector (CD45RA+ CCR7-) T cell subsets compared with Un-sensitised controls. The presence of more differentiated Effector memory and Terminal effector CD4+ T cell phenotype in the MTB sensitised study group in our study reflects the persistent expression of ESAT-6 by MTB and the persistent activation of the memory T cells by the ESAT-6 in the chronic nature of latent TB infection. There was no significant difference in the proportions of CD4+ Naïve and CM subsets in the MTB sensitised compared with Un-sensitised controls. However, the principal effector functional subset following ESAT-6 stimulation was the central memory (CD45RA- CCR7-).
We identified significantly higher triple IFNγ+ IL2+ TNF+ polyfunctional CD4+ central memory T cells in the MTB sensitised study group compared to the Un-sensitised controls. This finding is in agreement with a study investigating MTB recall responses to PPD and RD1 antigens (CFP-10, ESAT-6 and EspC) conducted between active TB and LTBI patients in London where they observed that CD4+ central memory T cells were the principal effector subsets in latent infection compared to CD4+ effector memory T cells in active TB infection.
Polyfunctional IFNγ+ IL2+ TNF+ CD4+ memory T cells have been detected by several studies of individuals with latent TB infection,(Day et al., 2011; Pollock et al., 2013) however their role in prevention of LTBI reactivation has not been well established (Pollock et al., 2013). In our study, the proportions of single or dual functional IFNγ+, IL2+, or TNF+ secreting CD4+ and CD8+ memory T cells were not significantly higher in the MTB sensitised group compared to the Un-sensitised controls, in response to ESAT-6 stimulation.
ESAT-6 specific CD8+ T cell responses have been found to be greater in individuals with pulmonary TB compared to individuals with latent TB infection in studies conducted in South Africa and London. (Day et al., 2011; Pollock et al., 2013) In general, we observed significantly lower CD4+ and CD8+ IFNγ, IL2 and TNF responses in the MTB sensitised group compared to Un-sensitised control, in response to SEB stimulation.
We did not however detect significant impairment of SEB-specific cytokine responses in the NTM sensitised study group. This reflects general T cell exhaustion and anergy in the chronic nature of latent TB infection. Impairment of SEB-specific T cell responses has been reported to be greater in an HIV/MTB co-infection compared to HIV infected controls. (Chetty et al., 2015) The notion that NTMs induce smaller TST reactions compared to MTB and the utilisation of this approach to establish a TST cut off for NTM and MTB sensitisation may need to be revised. In our study, 2 out of 12 study participants classified as MTB sensitised based on TST – T-SPOT.TB were also found to be NTM sensitised by ELISpot, following stimulation of their samples with NTM-specific peptide pools. These study participants had TST wheal diameter of 24 mm and 30.5 mm.
Additionally, in the 5 MTB sensitised samples (categorised by TST – T-SPOT.TB) that were stimulated with NTM-specific peptide pools and analysed using multi-parameter flow cytometry, we made the following observations. At an individual functional parameter level, there was a significant increase in the frequency of IFNγ secreting CD4+ CM and EM T cells and CD8+ Naïve and CM T cells. At a polyfunctional level, there was a significant increase in the IFNγ+ single expressing CD4+ CM and CD8+ CM; more IFNγ+ single expressing CD8+ Naïve T cells; and more dual functional IFNγ+ IL2+ CD8+ CM cells in the MTB sensitised. TST wheal diameter readings for these 5 MTB sensitised samples were: 11.5 mm, 14.5 mm, 19.5 mm, 24.0 mm, and 30 mm.
