Epitope-specific CD4+, but not CD8+, T-cell responses induced by recombinant influenza A viruses protect against Mycobacterium tuberculosis infection. Manuela Flórido1, Roman Pillay1, Caitlin M. Gillis1, Yingju Xia2, Stephen J. Turner3, James A. Triccas4, John Stambas2 and Warwick J. Britton1,4,5

1

Tuberculosis Research Program, Centenary Institute, Newtown, NSW, Australia

2

School of Medicine, Deakin University, Geelong, Victoria, Australia

3

Department of Microbiology and Immunology, University of Melbourne, Parkville, Victoria

4

Department of Infectious Diseases and Immunology Sydney Medical School, University of

Sydney, NSW, Australia 5

Department of Medicine, Sydney Medical School, University of Sydney, NSW, Australia

Keywords:

recombinant influenza A virus, tuberculosis, lung, CD4+ T cells, interferon-

Corresponding author: Dr M. Flórido, Tuberculosis Research Program, Centenary Institute, Locked Bag No. 6, Newtown, NSW 2042. Australia. Fax: +61-2-9565 6110 Email: [email protected]

Abbreviations: Recombinant Influenza A virus, rIAV; Mycobacterium tuberculosis, M. tuberculosis; Spot Forming Cell, SFC.

Received: 19-Jun-2014; Revised: 29-Oct-2014; Accepted: 24-Nov-2014 This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/eji.201444954. This article is protected by copyright. All rights reserved.

1

Abstract Tuberculosis remains a global health problem, in part due to failure of the currently available vaccine, BCG, to protect adults against pulmonary forms of the disease. We explored the impact of pulmonary delivery of recombinant influenza A viruses (rIAV) on the induction of Mycobacterium tuberculosis-specific CD4+ and CD8+ T-cell responses and the resultant protection against M. tuberculosis infection in C57BL/6 mice. Intranasal infection with rIAVs expressing a CD4+ T cell epitope from the Ag85B protein (PR8.p25) or CD8+ T cell epitope from the TB10.4 protein (PR8.TB10.4) generated strong T cell responses to the M. tuberculosis-specific epitopes in the lung that persisted long after the rIAVs were cleared. Infection with PR8.p25 conferred protection against subsequent M. tuberculosis challenge in the lung, and this was associated with increased levels of poly-functional CD4+ T cells at the time of challenge. By contrast, infection with PR8.TB10.4 did not induce protection despite the presence of IFN-γ-producing M. tuberculosis-specific CD8+ T cells in the lung at the time of challenge and during infection. Therefore, the induction of pulmonary M. tuberculosis epitope-specific CD4+, but not CD8+ T cells, is essential for protection against acute M. tuberculosis infection in the lung.

This article is protected by copyright. All rights reserved.

2

Introduction Tuberculosis (TB) is still one of the major causes of death worldwide, and it is estimated that one third of the world’s population is infected by M. tuberculosis (WHO 2012, http://www.who.int/tb/publications/global_report). The only vaccine currently used, BCG, although effective at protecting young children, fails to protect adults from the pulmonary form of the disease with efficacies that range from 0-80% [1]. The most advanced vaccine in human studies is MVA85A, a non-replicative vaccinia viral vector expressing M. tuberculosis Ag85A, but the results of a phase II clinical trial showed no additional protective effect of this vaccine over BCG against clinical TB or the acquisition of M. tuberculosis infection in infants [2]. The development of more effective vaccines is therefore essential. The immune response against primary infection with M. tuberculosis in a naïve host involves mostly CD4+ T cells so unsurprisingly anti-TB vaccines were initially designed to induce a strong CD4+ T cell response. That was the case for BCG and subunit vaccines that are potent inducers of systemic anti- M. tuberculosis CD4+ T cell responses [3]. Emerging evidence that CD8+ T cells also played a role in both the immune response to M. tuberculosis infection and in the vaccine-induced anti-mycobacterial immunity [4, 5] prompted the more recent development of DNA and viral vector-based anti-TB vaccines that are able to induce potent CD8+ T cell responses [6-8]. In addition, several experimental vaccines, both subunit vaccines and viral vectors, were found to have increased protective efficacy when delivered by the respiratory track than when delivered by peripheral immunization, [9-12] engendering interest in developing vaccines capable of inducing local pulmonary CD4+ and CD8+ protective T cell responses against M. tuberculosis infection. Nanoparticles coated with M. tuberculosis specific antigens is one of the mucosal delivery platforms being developed. Ag85B-coated nanoparticles delivered intranasally into mice conferred protective immunity

This article is protected by copyright. All rights reserved.

3

against tuberculosis both when administered alone [13] and adjuvanted by CpG [14]. Similarly, M. tuberculosis-specific proteins adsorbed to Bacillus subtilis killed spores administered intranasally to mice induced humoral responses and multi-functional T cells and protected against M. tuberculosis challenge [15]. Another strategy is the use of spray-dried particles containing mycobacterial components for pulmonary delivery. Aerosol delivery of BCG using this methodology significantly protected guinea pigs from challenge with M. tuberculosis at higher levels than standard subcutaneous BCG [12]. Recently we have shown that pulmonary vaccination using TLR2-targeted M. tuberculosis secreted proteins in a dry powder formulation generated CD4+ and CD8+ T cell responses in the lung and protect mice against M. tuberculosis challenge [16]. The majority of studies evaluating respiratory mucosal vaccination use intranasal delivery of viral vectors, particularly replication-deficient recombinant adenovirus and recombinant vaccinia virus [6]. Recently, the delivery of the candidate vaccine MVA85A by aerosol to the lungs of macaques was found to be safe and highly immunogenic [17]. Despite the current high interest in these new pulmonary delivery platforms for anti-TB vaccination, the requirement for the M. tuberculosis-specific CD4+ and CD8+ T cell subsets in the lung and their role in the protection conferred by these mucosally delivered vaccines against M. tuberculosis has not been extensively studied. Respiratory infection with IAV establishes strong CD8+ T cell responses in the lung, which are essential for the control of the primary influenza infection. In the context of IAV infection, CD4+ T cell help is essential for the development of the memory CD8+ T cell response and both IAV-specific CD8+ and CD4+ T cells persist lifelong. In our previous work, infection of mice with a recombinant influenza A virus (rIAV) expressing the CD8+ epitope of the M. tuberculosis MPT64 protein (MPT64190-198) induced a strong pulmonary epitope-specific CD8+ T cell response that persisted in the lung long after viral clearance [18]. This suggests

This article is protected by copyright. All rights reserved.

4

that rIAVs may be useful tools to induce M. tuberculosis epitope-specific CD4+ and CD8+ T cell responses in the lung and to analyze their role in protection against M. tuberculosis. To investigate the impact of pulmonary M. tuberculosis-specific CD4+ and CD8+ T cell responses against M. tuberculosis infection, we generated recombinant H1N1 PR8 strain rIAV expressing dominant H-2b-restricted T cell epitopes recognized during acute murine M. tuberculosis infection, viz the major CD4+ T cell epitope of Ag85B protein (Rv1886c), Ag85B240-254 (PR8.p25), or the CD8+ T cell epitope of the TB10.4 protein (Rv0288 or ESXH), TB10.43-11 (PR8.TB10.4). We then characterized the M. tuberculosis-specific CD4+ and CD8+ T cell responses induced by infection with PR8.p25 and PR8.TB10.4 and determined their role in protection against M. tuberculosis challenge.

Results Generation of recombinant influenza A virus expressing M. tuberculosis-specific epitopes To develop viruses that stimulate high levels of pulmonary M. tuberculosis-specific CD4+ and CD8+ T cells, we generated rIAVs expressing immuno-dominant M. tuberculosis-specific epitopes using reverse genetics. Ag85B and TB10.4 are two of the most immunogenic proteins expressed during acute M. tuberculosis infection in both mice and humans. In C57Bl/6 mice, the anti-Ag85B CD4+ T cell response is directed at the H-2b-restricted epitope Ag85B240-254 (p25) and the anti-TB10.4 CD8+ T cell response is focused on the Kb-restricted epitope TB10.43-11. Therefore we engineered the H1N1 PR8 influenza A virus strain to express at position 43 of the neuraminidase stalk either the CD4+ T cell epitope of the M. tuberculosis Ag85B protein, Ag85B240-254, (PR8.p25) or the CD8+ T cell epitope of the M. tuberculosis TB10.4 protein, TB10.43-11, (PR8.TB10.4) (Fig 1A). The insertion of the correct

This article is protected by copyright. All rights reserved.

5

DNA sequence for each epitope (Fig 1B) was confirmed by PCR (Fig 1C and D) and sequencing of the PCR products.

rIAV PR8.TB10.4 induces TB10.43-11-specific CD8+ T-cell responses To determine if infection with PR8.TB10.4 was able to induce a TB10.4-specific CD8+ T cell response, mice were infected with either the wild-type PR8 strain or PR8.TB10.4. The TB10.43-11-specific response was quantified in the lung and spleen at 2 and 4 weeks after infection, using IFN-γ ELIspot assay and MHC-I tetramer analysis (Fig 2). High frequencies of IFN-γ-producing CD8+ T cells specific to TB10.43-11 peptide were found in the lungs (Fig 2A) and spleens (Fig 2B) of mice infected with PR8.TB10.4, and, as expected, these were absent in mice infected with the wild-type PR8 strain and in uninfected control mice. Remarkably, the frequencies of IFN-γ-producing CD8+ T cells in response to the exogenous M. tuberculosis CD8+ T cell epitope TB10.43-11 in the lungs and spleens of mice infected with PR8.TB10.4 reached similar levels as the frequency of IFN-γ producing CD8+ T cells specific to the immunodominant endogenous influenza A virus nucleoprotein peptide (NP366374).

Moreover, despite evident contraction of the T cells responses following clearance of the

virus at 4 weeks post infection, a population of TB10.43-11-specific CD8+ T cells remained detectable in the lungs (~0.1% of total cells) and spleens (~0.01% of total cells) of mice infected with PR8.TB10.4 (Fig 2A, B). Consistent with these findings, enumeration of TB10.4-specific CD8+ T cells using Kb TB10.44-11 MHC-I tetramer revealed that 2 weeks after infection 6-10% of the CD8+ T cells in the lung of mice infected with PR8.TB10.4 were specific for TB10.44-11 epitope. Despite contraction of the response, more than 3% of the CD8+ T cells present in the lungs of mice 4 weeks after infection with PR8.TB10.4 were specific for the TB10.44-11 epitope (Fig 2C).

This article is protected by copyright. All rights reserved.

6

Smaller, but significantly increased, frequencies of TB10.44-11-specific CD8+ T cells were detected in the spleens of the mice infected with PR8.TB10.4 at both time-points analyzed (Fig 2D). To determine if TB10.4-specific CD8+ T cells generated by infection with PR8.TB10.4 were cytolytic, we performed an in vivo cytotoxicity assay. Syngeneic CFSE-labelled splenocytes, which were pulsed with the endogenous IAV peptide NP366-374, exogenous M. tuberculosis peptide TB10.43-11 or irrelevant ovalbumin peptide SIINFKEL (OT-I), were transferred into mice infected with either IAV PR8 or with rIVA PR8.TB10.4. After contraction of the acute T cell response, at 4 weeks post-viral infection, a reduction in the transferred splenocytes loaded with the peptides NP366-274 relative to the cells loaded with the irrelevant peptide OT-I was observed in the mice infected with both strains of IAV, while reduction of splenocytes loaded with TB10.43-11 was only observed in mice infected with PR8.TB10.4 (Fig 3). Importantly, in mice infected with PR8.TB10.4 a significant increase in killing of TB10.43-11pulsed targets was observed both in the lung (Fig 3B) and the spleen (Fig 3C) compared to mice infected with PR8. These results demonstrate that the TB10.43-11-specific CD8+ T cells induced by infection with PR8.TB10.4 are cytotoxic. Thus, a single intranasal infection with PR8.TB10.4 induced a strong and functional TB10.43-11-specific CD8+ T cell response.

rIAV PR8.p25 induces Ag85B240-254 –specific CD4+ T-cell responses Next we examined whether infection with PR8.p25 induced Ag85B240-254-specific responses CD4+ T cell responses. To that end, the Ag85B240-254-specific responses were analyzed in the lungs and spleens from mice infected with PR8.p25, wild-type PR8 and uninfected controls using an IFN-γ ELIspot assay and by measuring the amount of IFN-γ released in culture supernatants following stimulation with Ag85B240-254 peptide.

This article is protected by copyright. All rights reserved.

7

Cells producing IFN-γ in response to recall with the Ag85B240-254 peptide were detected in the lungs (Fig 4A) and in the spleens (Fig 4B) of mice infected with PR8.p25 at 2 and 4 weeks post-infection. The frequency of Ag85B240-254 -specific cells was higher in the lung than in the spleen of mice infected with PR8.p25 at both time-points. As observed for the rIAV-induced TB10.43-11-specific CD8+ T cell response, despite evident contraction of the Ag85B240-254 – specific CD4+ T cell response from 2 to 4 weeks post infection, after clearance of the virus, IFN-γ producing Ag85B240-254 specific cells remained in the lung (~0,06%, Fig 2A) and in the spleen (~0.02%, Fig 2 B). Consistent with these results, both lung cells and splenocytes from PR8.p25 infected mice secreted significantly increased levels of IFN-γ following Ag85B240-254 stimulation (Fig 4C, D). The IFN-γ levels were reduced, but detectable, at 4 weeks post-infection. Therefore, infection with PR8.p25 induces a strong pulmonary Ag85B240-254 specific CD4+ T cell response.

Comparison of Ag85B240-254 and TB10.43-11-specific T-cell responses induced by rIAVs and BCG To address the potential use of rIAVs as vaccines we compared the Ag85B240-254 and TB10.43-11–specific T cell responses induced by i.n. immunization with PR8.p25 and PR8.TB10.4 with those induced after subcutaneous BCG immunization. At 4 weeks postimmunization the Ag85B240-254 and TB10.43-11–specific T cell responses were measured by IFN- ELISpot in the lung (Fig 5A) and in the spleen (Fig 5B). In both organs infection with PR8.p25 induced higher frequencies of IFN- producing Ag85B240-254–specific T cells than immunization with BCG. Similarly, infection with PR8.TB10.4 induced higher frequency of IFN--producing TB10.43-11–specific T cells than BCG immunization. Significant IFN-

This article is protected by copyright. All rights reserved.

8

response to BCG lysate was only found in mice immunized with BCG, and the response to influenza A virus endogenous epitope NP was restricted to rIAV-immunized mice and was similar following both rIAVs (Fig 5A and B). We next analyzed the cytokine secretion profile of the splenic T cells in mice infected with the rIAVs or immunized with BCG. Production of the cytokines IFN-γ, TNF and IL2 by T cells was measured 4 weeks after infection with the rIAVs or BCG-immunization by intracellular staining following ex-vivo re-stimulation of spleen cells with TB10.43-11 and Ag85B240-254 peptides. Both poly- and single-cytokine CD4+ T cell responses to stimulation with Ag85B were significantly higher in mice infected with PR8.p25 than in mice immunized with BCG, in the lung (Fig 5C) and in the spleen (Fig 5D). By contrast, the increased IFN- production in response to stimulation with TB10.4 observed in mice infected with PR8.TB10.4 compared to BCG was mainly due to increase in the frequency of IFN-γ single producing CD8+ T cells in the lung (Fig 5E) and in the spleen (Fig 5F). Curiously, in the lungs, BCG infected mice had higher frequency of CD8+ T cells producing IL2 alone than PR8.TB10.4 infected mice (Fig 5E), while in the spleen, mice infected with PR8.TB10.4 had higher frequencies of CD8+ T cells producing IL2 alone or IL2 and TNF than BCG immunized mice (Fig 5F). These results demonstrate that infection with the rIAVs induced higher levels of CD4+ and CD8+ T cell response to the epitopes Ag85B and TB10.4 than immunization with BCG and that the PR8.p25-induced CD4+ T cell response was characterized by single- and poly-cytokine production while PR8.TB10.4-induced CD8+ T cell response was mostly comprised of cells producing IFN-γ alone.

Long-term M. tuberculosis-specific T-cell responses induced by infection with rIAVs

This article is protected by copyright. All rights reserved.

9

To determine if the M. tuberculosis-specific responses induced by the rIAV persisted, we analyzed the Ag85B240-254 and TB10.43-11–specific T cell responses in the lung and spleen 8 weeks after infection with PR8.p25 and PR8.TB10.4 and compared these with those induced 12 weeks after immunization with BCG. This was at the time of challenge with M. tuberculosis to assess protective efficacy. IFN-γ-producing Ag85B240-254-specific CD4+ T cells were detected in the lungs (Fig 6A) and spleens (Fig 6B) of mice infected with PR8.p25 at 8 weeks post-infection. By contrast, in mice immunized with BCG there was no Ag85B240-254 CD4+ T cell response detected in the lungs (Fig 6A) or spleen (Fig 6B) compared to uninfected mice. Similarly, there was a significant frequency of IFN-γ-producing CD8+ T cells in response to TB10.43-11 peptide in the lungs (Fig 6C) and spleens (Fig 6D) of mice infected with PR8.TB10.4, and not in mice immunized with BCG. As expected, a strong response to recall with NP366-374 peptide was observed in the cells from mice infected with PR8.p25 (Fig 6A,B) and with PR8.TB10.4 (Fig 6C,D). There was a trend for higher IFN-γ T cell responses to BCG lysate in BCGimmunized mice, but this was only significant in the lung in 2 out of 4 experiments. Furthermore, the T cell responses induced by infection with PR8.p25 and PR8.TB10.4 were sustained 12 weeks after infection, with the same levels of IFN-γ producing Ag85B240-254specific CD4+ T cell response (390±65 SFC/106 lung cells) and IFN-γ producing TB10.43-11– specific CD8+ T cell response (291±89 SFC/106 lung cells) respectively. Using specific MHC I tetramers, TB10.43-11 -specific CD8+ T cells were detected in the lungs (Fig 6E) and spleens (Fig 6F) of mice immunized with PR8.TB10.4 and BCG. Consistent with our IFN-γ ELISpot results (Fig 6A-D) and previous comparison to BCG immunization (Fig 5), infection with PR8.p25 resulted in a higher frequency of IFN-γproducing CD4+ T cells in response to the Ag85B240-254 antigen (0.3%) than immunization

This article is protected by copyright. All rights reserved.

10

with BCG (0.1%). Interestingly, higher frequencies of TNF single-positive and polyfunctional CD4+ T cells were found in mice infected with PR8.p25 (Fig 7A). Infection with PR8.TB10.4 resulted in a higher percentage of CD8+ T cells producing IFN-γ alone in response to TB10.43-11 peptide than following BCG immunization (Fig 7B). As previously noted (Fig 5E,F), the majority of TB10.43-11-specific CD8+ T cells were producing IFN-γ alone, or to a lesser extent, TNF alone, with only minor populations of poly-functional TB10.43-11-specific CD8+ T cells (Fig 7B). Together these data demonstrate that infection with PR8.p25 and PR8.TB10.4 generate long-lasting T cell responses to Ag85B240-254 and TB10.43-11 antigens respectively, at higher levels and with enhanced cytokine production compared to the response generated by standard BCG immunization.

M. tuberculosis epitope-specific CD4+ but not CD8+ T cells induced by infection with rIAVs confer protection Finally, to determine if the presence of M. tuberculosis-specific CD4+ and CD8+ T cells in the lung contribute to protection against M. tuberculosis infection, mice previously infected i.n. with PR8.p25 or with PR8.TB10.4 were challenged by aerosol with 100 cfu of M. tuberculosis 8 weeks later, and the levels of protection were compared to that following subcutaneous BCG immunization. Prior infection with PR8.p25, but not with PR8.TB10.4, was protective in the lung against subsequent M. tuberculosis challenge, reducing bacterial load by 0.8 log10 (P

Epitope-specific CD4+, but not CD8+, T-cell responses induced by recombinant influenza A viruses protect against Mycobacterium tuberculosis infection.

Tuberculosis remains a global health problem, in part due to failure of the currently available vaccine, BCG, to protect adults against pulmonary form...
1MB Sizes 1 Downloads 8 Views