Multidimensional analyses reveal modulation of adaptive and innate immune subsets by tuberculosis vaccines

We characterize the breadth, function and phenotype of innate and adaptive cellular responses in a prevention of Mycobacterium tuberculosis infection trial. Responses are measured by whole blood intracellular cytokine staining at baseline and 70 days after vaccination with H4:IC31 (subunit vaccine containing Ag85B and TB10.4), Bacille Calmette-Guerin (BCG, a live attenuated vaccine) or placebo (n = ~30 per group). H4:IC31 vaccination induces Ag85B and TB10.4-specific CD4 T cells, and an unexpected NKTlike subset, that expresses IFN-γ, TNF and/or IL-2. BCG revaccination increases frequencies of CD4 T cell subsets that either express Th1 cytokines or IL-22, and modestly increases IFNγ-producing NK cells. In vitro BCG re-stimulation also triggers responses by donor-unrestricted T cells, which may contribute to host responses against mycobacteria. BCG, which demonstrated efficacy against sustained Mycobacterium tuberculosis infection, modulates multiple immune cell subsets, in particular conventional Th1 and Th22 cells, which should be investigated in discovery studies of correlates of protection.

The manuscript entitled "H4:IC31 and BCG induced immune responses in a prevention of M. tuberculosis infection efficacy trial" by Rozot et al, study immunological responses of the first 90 enrolled participants, 30 per arm, of the C-040-404 trial. Previously the results of C-040-404 trial, were published in NEJM by Nemes et al in 2018 in which, authors showed that BCG vaccination reduced the rate of sustained QFT conversion at 4 IU/mL (P=0.03), but not vaccination with H4:IC31.
The present manuscript study the range, function and phenotype of innate and adaptive cellular immune responses conferred by H4:IC31 vaccination or by BCG revaccination. Authors observed that additional functions to Th1 response such as cell subsets contributed to BCG revaccination significantly boosted response of Th22 T cells. The study proposes a new framework to analyse vaccine-induced immune responses in clinical trials to provide a view of the global response induced by vaccination. Authors propose that the diverse components of this complex "collaborative immunity" are likely to be modulated by vaccination in unconventional ways that can only be revealed by multidimensional analyses. And authors conclude that where H4:IC31-specific responses would be predominated by Th1 cytokine-expressing CD4 T cells, while BCG would induce a broader range of immune subsets and functions.
Manuscript would be much more attractive if the authors will focus results and discussion on the direct comparing the immunological differences between regimen of vaccination that did prevent Mtb infection, measured as less sustained QFT conversion (BCG revaccination), as compared by the control and H4: C31 vaccine, that did not prevent Mtb infection. After all the exhaustive immunological study of the entire manuscript at the end of the discussion L457, is said, "the mechanism of BCG-induced protection against TB remains unknow" which greatly reduces the reader's expectations, who is looking for an immunoplogical correlation between BCG revaccination and decreased QFT sustained conversion. To include a final discussion on the relevance of the results obtained for future immunological studies, with other live attenuated vaccines currently in clinical trials, could help to increase the relevance the presented results in the field of TB vaccines.

Some Comments
Title: should be focused on BCG revaccination immunological results since to include H4:IC31 that did not prevent Mtb infection, does not add anything.
Discussion, Line 341: should say "reduce " and no "protect" since BCG vaccine reduced the rate of sustained QFT conversion.
That H4:IC31 induced increases in polyfunctional CD4 Th1 responses to Ag85B and TB10.4, should be interpreted as vaccine taken since are the Ags contained in H4: C31 vaccine and not too much discussion is needed. Discussion should be concentrated on the immunological results of BCG revaccination that could be of interest to determine the correlates of prevention of infection results of the Clinical Trial. Introduction 1. L1 Would a vaccine that prevents Mtb infection also by default prevent disease? A vaccine that prevent disease does not necessarily prevent infection, but a vaccine that prevents infection must surely also prevent TB disease. This opening sentence does not flow well to the next sentence. 2. L82 explain that Ag85B and TB10.4 are antigen from Mtb. 3. L86 explain that BCG is live. 4. L94 explain how 4IU/ml differs from the manufacturer's threshold 5. L95 with neither meeting "significance by" standard ...? 6. L90-99 -have these data been published elsewhere? (I think so, in reference 2). If they have, please include the citation and less detail. If not, they would be better placed in the results rather than introduction and state in the methods how many participants were in each study arm to measure efficacy? 7. L108 -while CD4 T cell deficiency occurs during HIV infection, many other cell populations are depleted; therefore it lends only weak support to your argument. 8. L115-116 simply because CD4 T cells are important for protection in a naive individual, it does not mean that by default a protective memory immune response must be generated from CD4 T cells. There is a lot of evidence in the literature that suggests Th1 cytokines induced by vaccines do not correlate with protection. Results 1. L144 abbreviations PBMC and ICS 2. L147-152 -different IL-17 and IL-22 antibody clones were used between WB and PBMCs, this should be clearly stated. 3. L163/L178-179 -you should state clearly which lymphocytes you have eliminated from your analysis (eg ILCs, B cells etc) from PBMCs. Why were these 'dumps' excluded from the WB analysis? Did you measure the contribution of the excluded populations to cytokine production? Discussion 1. L348 -can you say the Th1 responses were dominant, as you have restricted your analyses to only Th1/Th17/Th22 responses? 2. It may be useful to comment on the prevalence on the non-conventional populations in blood and whether you necessarily would detect their activation/change in frequencies that may occur in other sites, including the lung.
Supplementary Data 1. In Supplementary Figure 1, where in the gating strategy do you gate our LD/CD19 and CD14 positive cells or was this WB and not PBMCs? It is unclear in L163.
Methods 1. L496 how many samples were excluded for poor viability? 2. L521-525 how many samples were excluded based on these criteria? 3. L553-560 how many samples were excluded based on these criteria?
General comment -this manuscript would be easier to read with better paragraph structure.

Reviewer #1:
The manuscript entitled "H4:IC31 and BCG induced immune responses in a prevention of M. tuberculosis infection efficacy trial" by Rozot et al, study immunological responses of the first 90 enrolled participants, 30 per arm, of the C-040-404 trial. Previously the results of C-040-404 trial, were published in NEJM by Nemes et al in 2018 in which, authors showed that BCG vaccination reduced the rate of sustained QFT conversion at 4 IU/mL (P=0.03), but not vaccination with H4:IC31.
Response: We thank the reviewer for this summary and would like to clarify few points regarding the published manuscript (Nemes.et.al.), which we think are important to interpret the work under consideration. BCG revaccination was associated with reduced rates of sustained QFT conversion above 0.35 IU/mL and QFT conversion above 4IU/mL (sustained QFT conversion above 4IU/mL was not evaluated), with efficacy of ~45% for both end-points, which was significant at 95% confidence interval (CI). Vaccination with H4:IC31 showed efficacy for the same end-points (30% and 35% efficacy, respectively), but these were only significant at the 80% CI, which was the protocol-defined criterion for efficacy. However, protection conferred by H4:IC31 was not signification at the more rigorous and standard 95% CI. As such, we cannot definitively rule out that H4:IC31 showed some efficacy.
The present manuscript study the range, function and phenotype of innate and adaptive cellular immune responses conferred by H4:IC31 vaccination or by BCG revaccination. Authors observed that additional functions to Th1 response such as cell subsets contributed to BCG revaccination significantly boosted response of Th22 T cells. The study proposes a new framework to analyse vaccine-induced immune responses in clinical trials to provide a view of the global response induced by vaccination. Authors propose that the diverse components of this complex "collaborative immunity" are likely to be modulated by vaccination in unconventional ways that can only be revealed by multidimensional analyses. And authors conclude that where H4:IC31-specific responses would be predominated by Th1 cytokine-expressing CD4 T cells, while BCG would induce a broader range of immune subsets and functions.
Manuscript would be much more attractive if the authors will focus results and discussion on the direct comparing the immunological differences between regimen of vaccination that did prevent Mtb infection, measured as less sustained QFT conversion (BCG revaccination), as compared by the control and H4: C31 vaccine, that did not prevent Mtb infection.
Response: Unfortunately, this trial was not designed (nor powered) for a head-to-head comparison between BCG and H4:IC31, but to compare each vaccine arm to a shared placebo arm. In keeping with the predefined statistical design of the trial, and to protect against biased reporting, we therefore feel that it is important to report in the main figures the results for each vaccine separately. Further, because the vaccines are very different, and because we cannot say definitively that H4:IC31 was not efficacious, we contend that showing immunogenicity results for both regimens is correct and of interest to the reader. However, we agree that a side-by-side comparison of immune responses induced by the two vaccines is of relevance for understanding the results and have now added the head-to head comparison for the 2

vaccine arms and the placebo arm for the responses measured in CD4 T cells, CD8 T cells, gd T cells, MAIT cells and NKT-like cells and NK cells as Supplementary figure 5.
After all the exhaustive immunological study of the entire manuscript at the end of the discussion L457, is said, "the mechanism of BCG-induced protection against TB remains unknow" which greatly reduces the reader's expectations, who is looking for an immunoplogical correlation between BCG revaccination and decreased QFT sustained conversion. To include a final discussion on the relevance of the results obtained for future immunological studies, with other live attenuated vaccines currently in clinical trials, could help to increase the relevance the presented results in the field of TB vaccines.
Response: We thank the reviewer for this comment. Unfortunately, it is not possible to infer correlates of protection from the results presented in this paper. This is because the immunogenicity analyses in this paper were performed in a subset of the trial participants. As already emphasized in the discussion of the manuscript 'We propose that analysis platforms such as the one presented here are useful to understand complex vaccine-induced interactions between cell subsets that could inform correlates of risk or protection and should be routinely evaluated in clinical studies of TB vaccines and in other research areas". We have now added the following statement to clarify how the results reported here are relevant to the current efforts aimed at discovering immune correlates of protection: "Such studies are currently ongoing and the results reported here partially informed the experimental approach to discover immune correlates of protection, which includes measurement of multiple immune cell subsets beyond conventional Th1 cells".

Some Comments
Title: should be focused on BCG revaccination immunological results since to include H4:IC31 that did not prevent Mtb infection, does not add anything.  Response: It is not definitely true that a vaccine that prevents infection must also prevent disease, unless the vaccine is 100% efficacious. For a partially protective vaccine, such as BCG, the individuals that are protected against infection may be the "special" ones that would not have progressed to TB disease anyway. As such, we cannot assume that protection against infection equals protection against disease. We have now modified the sentence by removing the mention of TB disease to make it less confusing: "A vaccine that can prevent infection with Mycobacterium tuberculosis (Mtb) could have a major impact on the tuberculosis (TB) epidemic"