Tuberculosis, caused by Mycobacterium tuberculosis (Mtb), remains one of the leading causes of death from an infectious agent with 1.6 million deaths reported worldwide in 2021 (1). Bacille Calmette Guérin (BCG), the only preventative measure to combat tuberculosis is one of the most utilized vaccines in the world, with approximately 4 billion doses estimated to have been given throughout history (2). BCG successfully protects infants and young children from pulmonary and disseminated disease. Unfortunately, long-term protection afforded by BCG is highly variable at preventing pulmonary disease in adolescence and adulthood, and this is indicated as a contributing factor to continued transmission in TB endemic regions (3, 4). Several hypotheses exist for why BCG does not provide adequate long-term protection. Among these, our interest is in the age of vaccination and subsequent immune landscape of the recipient, as BCG is one of only two vaccines routinely given in the neonatal period (5). BCG is most often administered at the time of birth in TB endemic areas and this is a time at which the neonatal immune system is described as “immune tolerant”. This immune profile is characterized by a reduced expression of stimulatory receptors, a bias toward Th2 responses, and a skew to anti-inflammatory cytokines over proinflammatory cytokines (6).

Interleukin (IL)-27 is a heterodimeric cytokine of the IL-12 family with immune suppressive activity and has been shown to be elevated in the neonatal period (7–9). Previous studies that investigated the involvement of IL-27 in protection against TB demonstrated an inhibitory role toward pathogen control, with mice unresponsive to IL-27 exhibiting improved Mtb clearance (10–13). However, it is important to note that improved pathogen control was accompanied by heightened inflammation and tissue damage in these studies and those addressing other chronic infections (11, 14). This further points to a primary suppressive role of IL-27 in response to infection or vaccination with live agents. IL-27 has also been shown to negatively regulate IL-17 responses in a variety of inflammatory contexts including TB (15–17). IL-17 production following vaccination has been shown to drive a chemokine response that directs a protective T cell response to the lungs (18, 19). Collectively, this work demonstrates that IL-27 activity has important implications in the immune response against Mtb. However, these reports as well as other vaccination studies, have utilized adult animals that do not accurately reproduce the environment in which BCG protection is developed nor when IL-27 is most abundant in circulation.

We previously reported on the influence of IL-27 on neonatal dendritic cells (DCs) (20). IL-27 impaired the ability of neonatal bone marrow-derived DCs (BMDCs) to clear internalized BCG, produce IL-12, and stimulate antigen-specific T cells. This has led us to hypothesize that IL-27 opposes development of protective immune responses following BCG vaccination in neonates that could impact protection against TB disease later in life. To explore this hypothesis, we developed a model of neonatal vaccination with an adult challenge in C57BL/6 (WT) and IL-27 receptor-α-deficient (KO) mice. The latter are unable to respond to IL-27. Here we have shown that WT neonates vaccinated with BCG exhibit variable ability to control Mtb, which we interpreted as recapitulating protection developed in humans. Meanwhile, in the absence of IL-27 signaling, vaccinated neonates exhibit improved control of Mtb burdens and more robust and protective cytokine responses in the lung. This included elevated expression of IFN-γ and IL-17, two canonically important cytokines in defense against TB. These results suggest that blocking IL-27 signaling during neonatal BCG vaccination promotes superior protection from pulmonary TB later in life.

In the present study, we sought to understand the role of IL-27 during neonatal BCG vaccination and its influence on protection developed against Mtb challenge. An emphasis of our study design was to model human vaccination and challenge as best possible from a timeline and developmental perspective. Although it is difficult to accurately represent every aspect of human disease in animal models, due to the more rapid development of mice, it is possible to mimic the human scenario. That is one in which vaccination occurs in the neonatal period and robust exposure to pulmonary TB occurs in adolescence to adulthood when the benefits of BCG vaccination have been shown to wane as consistent long-term protection is not observed (3). We readily acknowledge that this model does not incorporate broad genetic diversity. First, we investigated the existing immune response five-weeks after vaccination, the time at which we challenged mice. As we have previously reported (20) but also important to establish in the challenged cohorts, we saw elevated IL-27 in vaccinated WT animals. Since IL-27 continues to rise through 5 weeks post-vaccination, it is likely that it shapes the immune response to BCG.

In our prior work, we demonstrated that clearance of BCG is impaired in neonatal DCs in response to IL-27 (20). In addition, IL-12 production was reduced in response to BCG and the stimulation of IFNγ by T cells from BCG-vaccinated mice was limited compared with DCs from IL-27Rα-deficient mice (20). Extending those findings to the current in vivo model ( Figure 6 ), we did not observe significant increases in circulating IFNγ in either genotype five-weeks after vaccination, but we did see statistically greater IL-17 in the serum of our vaccinated IL-27Rα-KO animals compared to WT. Previously we showed that neonatally-vaccinated WT mice maintain significantly greater BCG in their peripheral tissues compared to KO animals (20). We hypothesized this may drive a continual effector CD4⁺ T cell immune response as opposed to developing long-term central memory populations, and with the observed elevation of IL-27, a less effective inflammatory response that would not provide protective immunity. Therefore, we evaluated how the improved clearance of BCG and increased IL-17 production in our KO-vaccinated animals would impact responses against Mtb challenge. To do so in a manner that did not count persistent BCG as Mtb post challenge, we needed to employ an approach that discriminated the bacteria. The gold standard IFN-γ release assay (IGRA) to diagnose TB exposure and infection incorporates ESAT6 and CFP-10 that are not expressed by BCG. As such, we adopted similar rationale by measuring ESAT6 gene expression as an indicator of Mtb viability. While we cannot exclude some differences in ESAT6 expression with various in vivo conditions, it is our expectation that the in vivo environment is not drastically different between WT and KO mice such that the absence of a single host gene would dictate dissimilar ESAT6 expression. In line with this, we performed separate experiments that validate a significant positive correlation of ESAT-6 expression in the left lung with standard plate counts on media supplemented with hygromycin to select for Mtb only ( Supplementary Figure 2 ). We observed an overall decrease in Mtb burden by detection of ESAT6 expression in the lungs of both vaccinated mice WT and IL-27Rα-KO genotypes compared to non-vaccinated animals ( Figure 2 ). Furthermore, while KO mice vaccinated as neonates consistently controlled Mtb, with very low bacterial burdens detected in their lungs 8 weeks after challenge, WT mice vaccinated as neonates exhibited significant heterogeneity in their ability to elicit a protective response ( Figure 2B ). This resembles what is observed in humans whereby BCG has 0-85% protective efficacy (3). Importantly our model involves inbred mice that do not recapitulate the genetic diversity in humans, however, as it has not been examined in an experimental model, heterogeneity in the immune response of neonates as a result of the immaturity of the immune system may translate to variability in protection. Furthermore, while there are countless studies in which mice and even NHPs are vaccinated with BCG as adults, most do not report this extensive variability in protection from Mtb challenge (22–25). This underscores the novelty of our model and the importance of neonatal vaccination to better recapitulate the extensive variability of protection by BCG in humans.

We observed that neonatal vaccination and absence of IL-27 signaling significantly contributed to Mtb clearance after challenge, but these two variables were independent, with the rate of improved protection observed between vaccinated WT and vaccinated KO mice comparable to their respective non-vaccinated counterparts ( Figure 2C ). Interestingly, though, the amount of circulating IL-27 did not correlate with Mtb burden. Additionally, despite significant differences in circulating IL-27 between vaccinated and non-vaccinated animals, lung gene expression of the two subunits was comparable between treatment groups. Considering the higher levels of IL-27 in the serum of non-vaccinated mice, it is unclear if the rate of gene expression peaked prior to 8 weeks post-challenge, EBI3 is the limiting factor in assembly of the secreted heterodimer, there is post-transcriptional regulation of IL-27 subunits that impacts protein production between the two groups, or differences in IL-27 expression at other peripheral sites contributes to the circulating cytokine differences. Paired with findings in Figure 2 , this would suggest that IL-27 is detrimental in the control of Mtb but the absolute level does not dictate the magnitude of dysregulated host protection. Overall, the absence of IL-27 signaling significantly improved response against Mtb, as did vaccination in either genotype, with KO animals consistently controlling bacterial burden more efficiently than WT mice. This follows trends previously observed of IL-27 unresponsive animals exhibiting an improved ability to control Mtb after infection (10, 11, 13, 17). However, we utilized a novel approach here in which we replicated the age of BCG vaccination in humans by immunizing mice in the neonatal period.

To further improve vaccination for protection against TB it is necessary to identify the constituents of protective immunity. IFNγ is canonically associated with protective immune responses to Mtb because of its role in activating macrophages. Previously IFNγ expression was shown to be decreased in Mtb-infected IL-27Rα KO mouse lungs (10). However, other studies suggest this deficiency is only apparent early during infection, as IFNγ responses later in infection are equal or even greater in IL-27Rα KO mice (11, 26). Here, we observed elevated circulating IFNγ in KO animals which was significantly greater in vaccinated KO animals than their non-vaccinated counterparts as well as significantly increased above vaccinated WT mice. However, we observed vaccinated animals produced less IFNγ compared to their non-vaccinated counterparts in both genotypes. This may be reflective of reduced Mtb burdens and an indication of a contracting immune response following improved bacterial control. It is also possible that the lower systemic cytokine level is reflective of a more appropriate level of inflammation. In the lung, KO animals had elevated IFNγ expression compared to WT animals in both the vaccinated and non-vaccinated groups ( Figure 4 ). Interestingly, the degree of gene expression within each genotype was approximately the same regardless of vaccination status. This may imply that IFNγ expression has the potential to be induced up to a certain point within each genotype in the response to Mtb in the lung. Although this finding is consistent with IL-27 regulation of inflammation, it does suggest IFNγ expression alone is not correlated with the improved bacterial clearance in vaccinated mice, as the level of gene expression is the same between vaccinated and non-vaccinated KO mice.

Along with the role of IL-27 as a regulator of inflammatory responses, IL-27 has been demonstrated to inhibit development of Th17 immune responses (15–17). Surprisingly, though we observed increased IL-17 in the serum from vaccinated IL-27Rα-KO five weeks after vaccination ( Figure 1 ), we did not observe elevated circulating IL-17 following eight weeks of Mtb challenge. IL-17 has an established role in recruitment of CD4⁺ T cells to the lung following vaccination and promotion of granuloma development (19, 27). This includes accumulation of multifunctional T cells in the lung (17). Thus, increased IL-17 levels observed in KO mice vaccinated as neonates at the time of Mtb challenge is likely to have directed a protective response to the lungs and subsequently receded in the periphery. While serum IL-17 levels are low overall, it is notable that they were higher in WT mice following challenge, despite the increased expression of this cytokine in the lungs of KO animals ( Figures 3 , 4 ). One explanation for this could be that this systemic IL-17 is produced by neutrophils, which have been shown to be contributors to IL-17 during Mtb infection (28). However, the gene expression of IL-17 in the lungs was higher in KO animals, and significantly greater in those which were previously vaccinated ( Figure 4 ). This suggests an overall improved local Mtb response in the absence of IL-27. Indeed, IL-17⁺IFNγ⁺ CD4⁺ T cells have previously been shown to preferentially localize to the lungs of infected non-human primates and correlate with reduced Mtb burden (29).

Although there are differences in IL-17 and IFNγ expression patterns between WT and KO mice vaccinated as neonates, neither cytokine independently correlated with bacterial burdens. However, when we evaluate the combined inflammatory gene profile and bacterial burdens by multivariate analysis, we see a trend emerge indicating improved control of bacteria with an overall greater inflammatory response in KO mice vaccinated as neonates ( Figure 5 ). Our data demonstrate that in a correlation curve of Th1 cytokines, WT mice cluster towards a different balance of IFN-g and TNF that is consistent with higher bacterial burdens. This imbalance may not promote effective antimycobacterial activity in macrophages. A similar trend was demonstrated in the correlation analysis of IFNγ and IL-17. This suggests to us that, while the absence of IL-27 signaling clearly improves Mtb clearance, these effects are not mediated through a change in any single inflammatory Th1 or Th17 cytokine alone but is rather a total shift towards a combined inflammatory response. This is in line with prior work underscoring the importance of IL-17, TNF, and IFNγ in the control of Mtb (19, 30, 31) as well as the importance of multi-functional T cell responses (13, 17, 32).

Our data supports that neonatal BCG vaccination in the absence of IL-27 signaling better primes the immune system for a more rapid response to Mtb. We show here that in addition to more consistent control of Mtb, IL-27Rα KO mice also produce a more robust and balanced inflammatory response, which correlation analyses indicate contributes to the improved bacterial control. We suspect an additional element of this improved immune response could involve a more diverse T cell response and this is an aspect of ongoing investigation. In summary, as significant strength of this work is the combined modeling of neonatal vaccination with adult pulmonary challenge. In addition, the genetic approach to enumerating the Mtb burden allowed for direct correlation of cytokine expression data in the same piece of tissue. The former has important translational implications as it is consistent with trends in human vaccination and the duration of BCG protection in endemic TB regions. Efforts to improve TB vaccination should consider vaccination during the neonatal period and the implications of IL-27 signaling in early life immunity.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was approved by West Virginia University Institutional Animal Care and Use Committee. The study was conducted in accordance with the local legislation and institutional requirements.

SDB performed and designed experiments, analyzed and interpreted data, and wrote the manuscript. AD performed and designed experiments. SB performed experiments. JP and KR analyzed and interpreted data and contributed to the manuscript. CR supervised all research, designed and performed experiments, analyzed and interpreted results, and wrote the manuscript. All authors contributed to the article and approved the submitted version.