|Home | About | Journals | Submit | Contact Us | Français|
Accumulating evidence indicates that inflammatory signals required for maximizing effector T cell generation have opposing effects on the development of memory T cell precursors. Toll-Like receptor (TLR)2, and TLR9 significantly contribute to the inflammatory milieu and therefore in this study we examined whether the absence of TLR9 alone or the combined absence of TLR2 and TLR9 would affect vaccine-mediated immunity to Mtb. We found that TLR9KO and TLR2/9DKO mice vaccinated with a live Mtb auxotroph, akin to vaccinated WT mice, exhibited early control of Mtb growth in the lungs compared to their naïve counterparts. The granulomatous response, IFNγ production and cellular recruitment to the lungs were also similar in all the vaccinated groups of mice. These findings indicate that there is minimal contribution from TLR2 and TLR9 in generating memory immunity to Mtb with live vaccines. Defining the innate milieu that can drive maximal memory T cell generation with a tuberculosis vaccine needs further inquiry.
Mycobacterium bovis, Bacillus Calmette-Guérin (BCG) is the only vaccine currently used against tuberculosis (TB). Whereas, BCG is clearly effective in preventing disseminated TB in children, the protection conferred in adults has been variable, ranging from 0 to 80% in different studies . In the last decade, therefore, a great deal of research effort in TB was invested in generating new TB vaccines . This concerted effort from many TB investigators has produced several new vaccine candidates including live recombinant BCG, viral vector-based and subunit vaccines . In the pipe line are several live vaccine candidates in clinical trials [4, 5]. The recent failure of MVA85A booster immunization to provide better protection than BCG vaccination alone [6, 7], strongly indicates the need to continue efforts to understand the complexity of Mtb-specific memory T cell development, especially at the molecular level.
There is now accumulating evidence that inflammatory signals from the environment regulate the magnitude of effector and memory T cell generation. For example, high levels of Interleukin (IL)-12 leads to high T box transcription factor (T-bet) expression and induction of short-lived effector T cells while low expression favors the generation of memory precursor effector T cells . Transcription factors such as Eomesodermin (Eomes) and B cell lymphoma 6 favor the development of memory T cells [9, 10]. Cytokine levels dictate the ratio between T-bet and Eomes and this maintains the balance between effector and memory T cells . Consistent with the notion that curbing inflammatory response is conducive to memory T cell development, it was found that IL-21/IL-10/Signal transducer and activator of transcription 3 signaling axis is critical for the maturation of memory CD8+ T cells .
Mycobacterium tuberculosis (Mtb) expresses several ligands that stimulate TLRs on macrophages and dendritic cells (DCs) . In particular, Mtb has been shown to stimulate signaling through TLR2 and TLR9 [14–16]. We and others have previously shown that although TLR2 is not essential for protection against acute infection , it is required in maintaining the granulomatous response and in controlling inflammation during chronic Mtb infection [18, 19]. We have also reported the redundant role of TLR2 in mediating protective immunity to secondary Mtb infection . However, the requirement of TLR9 in the generation of protective immune response to Mtb has not been addressed. Based on the growing evidence that decreased IL-12 levels favor central memory T cell (TCM) generation [8, 11] and our finding that DCs lacking TLR9 have diminished IL-12 production , we posited that memory immune response against Mtb would be enhanced in TLR9KO mice. The main approach to improve BCG vaccine efficacy has been to modify the vaccine by inclusion of different Mtb antigens [21–23]. However, in this study, the approach was to modify the host innate environment and determine whether vaccine efficacy would improve. We therefore examined memory immunity against Mtb in the presence and absence of TLR9 signaling in a live auxotroph vaccination-challenge model.
Breeding pairs of TLR9KO and TLR2/9DKO were obtained from Alan Sher’s laboratory, NIH, Bethesda, MD [14, 24]. These mice were bred under pathogen free conditions at the animal facility at Rutgers-New Jersey Medical School, Newark, NJ. C57Bl/6 WT mice were purchased from National Cancer Institute. Mtb infected mice were housed in the BSL3 facility at Rutgers, New Jersey Medical School. Mice protocols used in this study have been approved by the Rutgers-Institutional Animal Care and Use Committee. ΔleuCΔpanCD double auxotroph mutant of H37Rv, mc26206 was obtained from the laboratory of William Jacobs at Albert Einstein College of Medicine, NY. The virulent strain of Mtb Erdman (Trudeau Institute, Saranac Lake, NY) was prepared as previously described .
Age matched 6–8 week old WT, TLR9KO and TLR2/9DKO mice were vaccinated subcutaneously with 4×106 colony forming units (CFU) of mc26206. After 9 weeks [25, 26], vaccinated mice of all genotypes and an equal number of genotype-matched unvaccinated (naïve) mice were challenged with a low dose of Mtb Erdman (50–100 CFU) by aerosol infection (Glas-Col) as previously described .
Superior lobe of lung or spleen tissue was homogenized in 1ml of PBS with 0.05% Tween-80. Serial dilutions of the tissue lysates were plated on 7H11 plates and CFU was enumerated after incubation at 37°C for 14–21 days.
The middle and left lobes of the lungs were perfused with 10ml sterile PBS, and digested with 2mg/ml collagenase D at 37°C for 30minutes. Reaction was stopped with the addition of 20µl of 5mM EDTA. Digested tissue was mashed through a 40µm nylon membrane filter to obtain single cell suspensions. Red blood cells were lysed using ACK lysis buffer. Single cell suspensions were surface stained with directly conjugated fluorochrome labeled anti-mouse CD4-V450 (clone RM4-5; BD Horizon), anti-mouse CD8-AF488 (clone 53-6.7; BD Pharmingen), anti-mouse CD44-AF700 (clone IM7; BD Pharmingen) and anti-mouse CD62L-PerCPCy5.5 (clone MEL-14; BD Pharmingen) antibodies. For tetramer staining, lung cells were first surface stained with ESAT6(1–20)-PE tetramer (NIH Tetramer Facility) for 1hr at 37°C, followed by surface staining for T cell markers. Following staining, cells were washed with FACS buffer, fixed in 4% paraformaldehyde and acquired on LSR-II (BD Biosciences). Frequency of specific cell types was calculated using Flow Jo software.
Mtb specific IFNγ producing cells in the lungs were determined by the enzyme linked immunospot assay (ELISPOT). 96 well MultiScreen HTS filter plates (Millipore) were coated overnight with 8µg/ml anti-IFNγ antibody (clone R4-6A2; BD Biosciences). Lung cells were seeded at a concentration of 0.025–0.1×106 per well. The cells were re-stimulated ex vivo with Mtb infected bone marrow derived dendritic cells (BMDC) overnight at a ratio of 1:2. Cultures were supplemented with 20U/ml IL-2 and incubated at 37°C for 2 days. Post incubation period, the plates were washed, treated sequentially with biotinylated secondary antibody (clone XMG1.2; BD Biosciences), streptavidin HRP and the substrate: 3-amino-9-ethyl-carbzole. Spot forming units were determined using an ELISPOT plate reader (Cellular Technology).
Post caval lobe of the lung from infected mice was homogenized in 1ml of TRIzol reagent (Invitrogen) and total RNA was extracted using RNeasy column (Qiagen). cDNA was prepared from the total RNA by reverse transcription using Superscript II (Invitrogen). Real time PCR was performed using TaqMan probes. Beta-actin was used as the endogenous control. (Anti-mouse IL-12p40 accession number: Mm00434174_m1; anti-mouse beta actin accession number: Mm00607939_s1; anti-mouse IL-12p35 accession number: Mm00434169_m1; anti-mouse IL-23p19 accession number: Mm01160011_g1; Life Technologies). Total RNA from uninfected naïve lung was used as calibrators. Relative gene expression was expressed as 2−ΔΔCt.
Inferior lobe of the lung was perfused with 10ml of PBS and fixed in 4% paraformaldehyde for 4 days. Paraffin embedded-tissue sections were cut and stained with haematoxylin (H) and eosin (E). Stereoscopic images of H&E stained sections of whole lungs were obtained using Act-1 software from Nikon.
Graph pad Prism software was used. One or Two way ANOVA with Bonferroni’s correction was used depending on the number of groups compared.
9 weeks following auxotroph immunization, the percentage of IFNγ producing cells in the lungs, spleen and draining lymph nodes from WT, TLR9KO and TLR2/9DKO mice were analyzed by ELISPOT. Interestingly, we observed an increased frequency of IFNγ producing cells in the lungs and spleen of TLR2/9DKO mice compared to WT mice (Fig 1A,B). Consistent with the increased IFNγ producing cells, flow cytometric analysis showed a higher percentage of ESAT6-tetramer+ CD4+ T effector memory cells (TEM: CD44hi CD62Llo) in the lungs of TLR2/9DKO mice when compared to WT mice (Fig 1E). A similar increase was not observed in the lungs of TLR9KO mice (Fig 1A,B,E). The total percentage of CD4+ T cells in the lungs was similar in the three groups of vaccinated mice (Fig 1D). Increased IFNγ producing cells were also present in the lymph nodes of the TLR2/9DKO mice (Fig 1C), although not statistically significant. These data indicate that auxotroph immunization induces a higher frequency of IFNγ producing cells when both TLR2 and TLR9 are absent.
Since there was a higher frequency of IFNγ producers and TEM in the lungs of TLR2/9DKO mice, we hypothesized that this would lead to enhanced protection in these mice upon challenge with Mtb. Consistent with an earlier study , WT mice vaccinated with mc26206 were able to control bacterial burden in the lungs at a significantly lower number than their naïve counterparts at 2 and 4 weeks following Mtb challenge (Fig. 2A,B). Similarly, significant reduction in Mtb load in the lungs was also observed in the vaccinated TLR9KO (Fig. 2A) and TLR2/9DKO mice (Fig. 2B) at both 2 and 4 weeks, in comparison to their naive counterparts. It is worthy of note that the presence of increased IFNγ producers in the lungs of TLR2/9DKO mice did not improve their ability to control Mtb infection. Bacterial burden in the spleens of all groups of vaccinated mice was also significantly lower compared to corresponding naïve mice (Fig. 2C,D). These data indicate that signaling through TLR2 and TLR9 is not required to generate memory immunity and concomitant enhanced restriction of Mtb growth in the lungs and spleens of vaccinated mice.
The enhanced control of bacterial burden in the lungs of vaccinated mice is typically associated with smaller granulomatous lesions . As expected, WT naïve mice exhibited extensive granulomatous inflammation in the lungs at 4 weeks post challenge. In contrast, the vaccinated WT mice displayed compact and localized granuloma with more clear alveolar spaces (Fig 3A,B). Vaccinated TLR9KO (Fig 3A) and TLR2/9DKO (Fig 3B) mice also presented with smaller granulomatous inflammation than their naïve counterparts. Also, the total granulomatous area in the lungs of vaccinated WT, TLR9KO and TLR2/9DKO mice was significantly lower in comparison to their naïve counterparts (Fig 3C,D).
As expected, in comparison to their naïve counterparts, vaccinated WT mice had a significantly higher recruitment of CD4+ T cells to the lungs 2 weeks following Mtb challenge (Fig. 4A–D). A similar increased recruitment was also observed in vaccinated TLR9KO (Fig. 4A,C) and TLR2/9DKO (Fig. 4B,D) mice when compared with the naïve controls. The recruitment of CD8+ T cells however did not significantly differ in the three groups of vaccinated mice (Fig. 4A–D). Consistent with the earlier recruitment of CD4+ T lymphocytes to the lungs, we observed a higher frequency of IFNγ producing cells in the lungs of vaccinated WT, TLR9KO and TLR2/9DKO mice when compared to their naïve counterparts (Fig. 4E,F). Interestingly, we found a lower frequency of IFNγ producers in the lungs of vaccinated TLR9KO mice when compared to vaccinated WT mice (Fig 4E), although there was no difference in bacterial burden between the two vaccinated groups of mice. We also found that the expression of IL-12p40 (Fig. 5A,B), IL-12p35 (Fig. 5C,D) and IL-23p19 (Fig. 5E,F) in the lungs following Mtb challenge was not affected either in TLR9KO or TLR2/9DKO mice. Taken together, these findings suggest that the early recruitment of IFNγ secreting CD4+ T cells to the lungs and in vivo expression of IL-12 upon Mtb challenge are not affected by the loss of TLR2 and TLR9 signaling.
In an earlier in vitro study we had found that IL-12 production from macrophages and DCs is reliant on the innate receptors, TLR2 and TLR9 . Based on that finding, we argued that the production of IL-12 in vivo would be significantly reduced in the absence of TLR2 and TLR9 and this would in turn favor an enhanced memory response, analogous to what has been reported with other pathogens [29, 30]. However, findings from the current study indicate that TLR2 and TLR9, are dispensable for a live vaccineinduced memory immune response against Mtb. Furthermore, examination of lung gene expression revealed that TLR9KO and TLR2/9DKO naïve and vaccinated mice were fully capable of expressing IL-12 in the lungs. This suggests that other innate receptors such as the C-type lectin receptors (CLRs) and Nucleotide-binding Oligomerization domain (NOD) like receptors (NLRs) could compensate for the lack of TLR2 and TLR9 in IL-12 induction in vivo during Mtb infection. Indeed, Mtb induced signaling through Dectin-1 (type II CLR) on splenic DCs has been shown to enhance IL-12 production in vitro . Despite the requirement for NLRs on macrophages for IL-12 production in vitro, NOD-2 deficient mice can control Mtb infection similar to WT mice. , suggesting that during Mtb infection in vivo, multiple signaling pathways induce IL-12.
Importance of IFNγ in protection against Mtb infection is clear since mice lacking the cytokine are highly susceptible to Mtb infection [33, 34]. However, recent studies indicate that CD4+ T effector cells can mediate protection against Mtb in an IFNγ independent manner [35, 36]. It is possible that in the absence of IFNγ, Th17 cells may confer protection early on during Mtb infection . In this study we found that either increase or decrease in IFNγ producing cells in the TLR-deficient mice, in comparison to WT mice, did not influence the disease outcome. These data support the growing consensus that IFNγ is not the critical effector cytokine mediating the control of Mtb infection.
TLR ligands have been widely used in various subunit and protein based vaccines. Using adjuvants such as Lipokel, a TLR2 ligand in a subunit vaccine preparation  and CpG-complexed liposomes with ESAT6 protein  have been demonstrated to mediate protection to Mtb challenge. These studies suggest a critical role for TLR stimulation at the time of vaccination. In our study, mice vaccinated with mc26206 were able to protect against Mtb infection even in the absence of TLR2 and TLR9 signaling. This indicates that live vaccines, unlike subunit or peptide/protein vaccines, can bypass the requirement for TLR in generating memory immunity to Mtb.
Based on the differential expression of a number of cell surface markers, antigen-experienced CD4+ and CD8+ T cells can be separated into TCM cells that chiefly migrate to lymphoid organs and TEM cells that largely migrate to extra-lymphoid organs [40, 41]. While both TCM and TEM provide protection, only TCMs are long-lived  and also provide better and long-term protection [43–45]. BCG vaccination induces predominantly TEMs, [46–48], which could partly explain its variable efficacy in protection. Consistent with this theory, BCG ΔureC::hly vaccine, a recombinant variant of BCG, induces an increased TCM response in the vaccinated mice and confers better protection against challenge Mtb infection in comparison to BCG which induces predominantly a TEM response . Mechanistically, the enhanced protection afforded by BCG ΔureC::hly vaccine was shown to be at the level of AIM2 inflammasome activation with a concomitant increase in production of IL-1β and IL-18 . These findings suggest that perhaps skewing the innate response towards enhanced IL-1β and IL-18 may promote TCM development and enhance vaccine efficacy. Clearly, future studies aimed at defining the innate milieu that can drive maximal memory T cell generation with a TB vaccine candidate are needed.
We conclude that vaccine mediated immunity to Mtb is not dependent on TLR2 and TLR9 signaling.
This work was supported by National Institutes of Health Grant AI084822. We thank Dr. Sher for TLR9KO and TLR2/9DKO mice, Dr. Stockinger for the Granulocyte colony stimulating factor-secreting cell line (X63) Dr. Jacobs for the Mtb auxotroph, mc26206 and John Altman, NIH Tetramer Facility for the ESAT6 tetramer.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1Abbreviations: TLR, Toll-like receptor; TB, tuberculosis; BCG, Bacillus Calmette-Guérin; IL, Interleukin; T-bet, T box transcription factor; Eomes, Eomesodermin; Mtb, Mycobacterium tuberculosis; DC, dendritic cells; T central memory, TCM; CFU, colony forming units; ELISPOT, enzyme linked immunospot assay; BMDC; bone marrow derived dendritic cell; H, hematoxylin; E, eosin; T effector memory, TEM; CLR, Ctype lectin recpetors; NOD, Nucleotide-binding Oligomerization domain; NLR, Nod-like receptors.