Emerging role of γδ T cells in vaccine‐mediated protection from infectious diseases

Abstract γδ T cells are fascinating cells that bridge the innate and adaptive immune systems. They have long been known to proliferate rapidly following infection; however, the identity of the specific γδ T cell subsets proliferating and the role of this expansion in protection from disease have only been explored more recently. Several recent studies have investigated γδ T‐cell responses to vaccines targeting infections such as Mycobacterium, Plasmodium and influenza, and studies in animal models have provided further insight into the association of these responses with improved clinical outcomes. In this review, we examine the evidence for a role for γδ T cells in vaccine‐induced protection against various bacterial, protozoan and viral infections. We further discuss results suggesting potential mechanisms for protection, including cytokine‐mediated direct and indirect killing of infected cells, and highlight remaining open questions in the field. Finally, building on current efforts to integrate strategies targeting γδ T cells into immunotherapies for cancer, we discuss potential approaches to improve vaccines for infectious diseases by inducing γδ T‐cell activation and cytotoxicity.


INTRODUCTION
Although representing only a small percentage of T cells (generally 2-5% of peripheral blood T cells in healthy adults), cd T cells have increasingly been recognised for their unique roles in establishing and regulating the inflammatory response to infectious diseases. These unconventional T cells have antigen recognition capacity, tissue tropism and cytotoxic functions that are distinct from ab T cells. cd T cells are the first T cells to appear in the thymus during foetal thymic ontogeny and, following gene rearrangement, express different Tcell receptor (TCR) sequences. 1 TCR diversity is different across different animals, but in humans, subsets expressing different Vc and Vd regions localise to different tissues and have differing effector functions. For example, the most abundant subset in human adult peripheral blood is Vc9Vd2 cells (also referred to as Vc2Vd2) while Vd1 + cells are more common in mucosal tissues. 2 Existing only in primates, Vc9Vd2 cells recognise phosphoantigens induced by stress or pathogens in a process that is dependent on butyrophilin 3A1 (BTN3A1, CD277), a type I glycoprotein in the B7 family. 3 Other signalling pathways for human cd Tcell activation involve TCR interaction with ligands such as F1-ATPase or endothelial protein C receptor, or additional cell surface receptors such as natural killer group 2 member D (NKG2D) receptors or toll-like receptors (TLR). 4 Unlike ab T cells, all of these pathways are independent of the major histocompatibility complex (MHC). In some animals (e.g. cattle, sheep, chickens), cd T cells express highly diverse TCRs regardless of tissue localisation, while in others (e.g. mice), almost all cd T cells in the epidermal layer of the skin (called 'dendritic epidermal T cells') express identical cd TCRs. Interestingly, cd TCRs are structurally more similar to immunoglobulins than ab TCRs; the CDR3 lengths of TCR d chains are long and variable, whereas those of the TCR c chains are short and constrained. 1 The presence of TCR chains that use antibody-like V domains is widely distributed in vertebrates, suggesting a selective pressure for TCR chains that recognise antigen in ways similar to that of antibodies.
Several cd T-cell subsets have long been known to rapidly increase in number following systemic infections and to perform numerous roles, including direct anti-microbial roles, recruitment of innate immune cells and activation of adaptive immune cells. 4 In many situations, including most bacterial and parasitic infections in humans, it is the Vd2 + T-cell subset that proliferates, while in some viral infections, Vd1 + T cells expand and exert anti-microbial activities. Interestingly, cd T cells also appear to have some level of functional plasticity, enabling them to adapt their function at different points during infection based on TCR signalling and environmental cues. Animal models have further provided support that these cells are not simply biomarkers of infection, but can in fact mediate protection from disease and/or recurrent infection. Despite being known to have an important role in immunity to infectious diseases, cd T cells have, with the exception of the Bacillus Calmette-Gu erin (BCG) vaccine for tuberculosis, largely been ignored in vaccine development. Whether cd T cells are stimulated directly by the antigen component of the vaccine or indirectly with an appropriate adjuvant, there may be many opportunities to improve vaccine effectiveness by targeting cd T cells. In this article, we will review the evidence for the role of cd T cells in vaccineinduced protection to bacterial, protozoan and viral infections. Many of these diseases, particularly those responsible for the highest mortality and morbidity worldwidetuberculosis, malaria and HIVdo not yet have an effective vaccine because of rapid pathogen evolution and other biological and technical challenges. However, considering the functional roles of cd T cells and incorporating them into a vaccine strategy could be an important step towards reducing the devastating impact of these diseases.

MYCOBACTERIA AND OTHER BACTERIAL INFECTIONS
A number of studies have shown expansion of cd T-cell populations in response to various bacterial infections, both in humans and in animal models. In humans, cd T cells accumulate at mucosal epithelial tissues, including the lungs, 5 and have been shown to rapidly proliferate following infection with Mycobacterium tuberculosis (Mtb). 6,7 These responding cd T cells primarily express Vc9Vd2 8 and recognise Mtb phosphoantigen. 6,9 Studies testing whether cd T cells expand in response to the Mtb heat shock protein HSP65 have had somewhat conflicting results, but suggest that while some cd T-cell clones can recognise HSP65, the majority of cells respond to other antigens. 7,10,11 Several in vitro studies have suggested that Vc9Vd2 T cells may mediate protection from Mtb. These cells appear to be capable of directly killing extracellular Mtb via release of granulysin and intracellular Mtb via granulysin and perforin. 12 Mycobacteria-specific Vc9Vd2 T cells from individuals positive for the tuberculosis skin test also produce granzyme A, which indirectly leads to Mtb destruction by stimulating TNFa production by infected macrophages. 13 In the mouse model, although cd T cells seem to be less essential to immunity against Mtb, 14,15 GM-CSF production by cd T cells in the lungs seems to play a role in protection and an additive effect between GM-CSF and IFNc promoted macrophage control of intracellular bacterial replication. 16 Clearly, the Vc9Vd2 T-cell subset is important in the human immune response to Mtb, but further work is required to evaluate the role of various cytokines in protection from disease at different timepoints during infection.
cd T cells also seem to play a role in immunity induced by BCG, the only current vaccination against Mtb. Similarly to natural infection, cd Tcell populations expand and produce IFNc in response to BCG vaccination. [17][18][19] In fact, IFNc production by these cells was greater than that of CD4 + T cells. 19 In adults, Vd2 + cd T cells from BCGvaccinated individuals expanded more than cells from non-vaccinated individuals in response to in vitro Mtb restimulation; this memory-like phenotype could not solely be attributed to increased helper functions from mycobacteriaspecific memory CD4 + T cells. 20 Given that BCG contains lower levels of phosphorylated nonpeptidic antigens compared to Mtb, 21 it is unclear whether cd T cells responding to BCG are recognising the same or different antigens compared to natural infection. Further studies are needed to evaluate the functional role of cd T-cell expansion following BCG vaccination, including any role for memory-like subsets and whether expansion provides protection upon challenge or infection with Mtb. Considering the importance of granulysin, perforin and granzyme A in response to Mtb, it may also be useful to incorporate strategies that elicit these responses into vaccine design.
Studies in non-human primates further support an important role for cd T cells in responding to Mtb infection and BCG vaccination. These studies may additionally provide insight into mechanisms driving immunity induced by cd T-cell expansion. Non-human primates serve as a useful model as they also express the Vc9Vd2 T-cell subset, which recognise Mtb, unlike murine cd T cells which do not recognise phosphoantigen or microbial antigens. 15 Administration of an Mtb phosphoantigen analog combined with IL-2 expanded the Vc9Vd2 T-cell population during Mtb infection. 22 Expanded Vc9Vd2 T cells differentiated into effector subpopulations, expressed cytokines such as IFNc, perforin, granulysin and IL-12, and led to enhanced pulmonary responses of peptide-specific CD4 + / CD8 + T cells. 22 Importantly, diminished TB lesions and reduced Mtb proliferation were also observed, suggesting a role for expanded/ differentiated Vc9Vd2 T cells in resistance to Mtb infection. 22 In another approach, adoptive transfer of autologous Vc9Vd2 T cells 1 or 3 weeks after Mtb infection led to significant protection from Mtb, including a rapid recall-like increase in the pulmonary Vc9Vd2 T-cell subset, decreased Mtb infectious burdens (particularly in the lungs) and reduced pathology. 23 Following BCG vaccination, Vc9Vd2 T cells expanded as early as 4-6 days post-vaccination with peak levels at 3-5 weeks post-vaccination; this expansion further coincided with clearance of bacteraemia and immunity to fatal tuberculosis after challenge. 24 Finally, a prime-boost approach using phosphoantigen followed by fusion proteins led to expansion of cd T cells displaying effector memory surface markers and producing cytokines such as IL-2, IL-6, IFNc and TNFa following primary vaccination. 25 As these cells anergised following boosts whereas ab T cells expanded, 25 future studies could investigate whether anergy can be prevented and cd T-cell recall responses preserved.
Together, the described studies in macaques provide evidence that cd T cells confer protection from symptomatic Mtb infection and support targeting these cells in vaccination approaches to Mtb.
The cd T-cell ontogeny is quite different in other mammals compared to humans and nonhuman primates; however, studies in cattle and pigs showed similar responses to those found in humans and macaques. Cattle and other ruminants express large proportions of cd T cells that decline with age, but remain high relative to human levels. 26,27 In cattle, cd T cells rapidly proliferate following infection with Mycobacterium bovis [28][29][30] or BCG vaccination. 31,32 Similarly, in pigs, cd T cells proliferated following vaccination with BCG. 33 Other bacterial agents demonstrating cd T-cell expansion following infection and vaccination include Leptospira borgpetersenii, Salmonella enterica, Francisella tularensis and Listeria monocytogenes. Similarly to the described response to Mtb, human cd T-cell populations, in particular the Vc9Vd2 subset, expand following leptospirosis infection. 34,35 In leptospirosis vaccination studies in cattle, IFNc-producing cd T cells expressing the WC1 co-receptor expand postvaccination and upon in vitro restimulation. [36][37][38] cd T cells also expand following salmonella vaccination in chickens and macaques 39,40 or following salmonella infection in humans. 41 Furthermore, following salmonella or listeria vaccination in macaques, cd T cells displaying Vc9Vd2 were the major T-cell subset proliferating. 40,42 Following subclinical Listeria monocytogenes infection, Vc9Vd2 T cells expanded, trafficked to the lungs and intestinal mucosa and evolved into effector cells producing IFNc, TNFa, Il-4, Il-17 and/or perforin. 42 These cells could then lyse infected target cells and inhibit intracellular bacterial growth, demonstrating a potential role in protection from listeria. 42 Interestingly, cd T cells displaying Vc9Vd2 expanded in humans infected with F. tularensis, 43,44 but did not expand following vaccination, perhaps because of different phosphoantigens present. 43 In summary, a number of studies have not only demonstrated cd T-cell expansion in various bacterial infections, but also possible mechanisms of protection provided by this cell population, including both direct killing and recruitment of other cell types via production of proinflammatory cytokines. Although clear that cd T cells respond differently based on infectious agent, specific proliferation of the Vc9Vd2 subset in response to a number of bacterial pathogens correlates with protection from symptomatic disease. Consequently, upregulating activation and/or functional responses of this subset by vaccination may enhance protection against the agent targeted by immunisation. However, given the cd T-cell anergy observed in the described vaccine study combining phosphoantigen with a subunit anti-tuberculosis vaccine, 25 as well as prevalent examples of T-cell exhaustion in other contexts, further work is needed to assess potential mechanisms driving such processes. Timing of interventions could therefore be optimised to induce maximal cd T-cell recall responses and promote activation without causing exhaustion.

MALARIA INFECTION
In addition to long-standing evidence that cd T cells play a role in initial responses to parasitic infections, there is increasing evidence that cd T cells are important in vaccine-induced protection from malaria. Studies over the past few decades have shown that cd T cells (particularly the Vd2 + subset) rapidly expand following infection with the most virulent human malaria parasite, Plasmodium falciparum (Pf), in children, malaria-na€ ıve adults and malaria-experienced adults. [45][46][47][48] Frequencies of cd T-cell subsets, including Vd2 + , Vd2 À , activated CD11c + or CD16 + /Tim-3 + cd T cells, have all been associated with malaria exposure. [49][50][51][52][53][54][55][56] Higher frequencies and malaria-responsive cytokine production of Vd2 + T cells correlate with protection against subsequent infection in children living in endemic settings, 57,58 and in vitro, these cells perform cytotoxic, anti-parasitic functions. 59,60 Furthermore, these cells can also act as antigenpresenting cells, [61][62][63][64] which may further enhance the response to infection and/or vaccination. In malaria-na€ ıve volunteers exposed to Pf-infected mosquitoes, while under chloroquine prophylaxis, cd T cells expand after infection. 65 Elevated frequencies of cd T cells expressing effector memory surface markers and enhanced responsiveness to Pf stimulation persist for over 1 year following experimental infectious challenge. 65 A recent small study from the same group reported that vaccination with BCG changed the course of experimental malaria infection and that BCG vaccination was associated with altered innate immune activation (including cd, NK and monocytes) following malaria challenge. Interestingly, expression of the activation marker CD69 on both NK cells and cd T cells was associated with reduced parasitaemia. 66 Trends towards increased degranulation and granzyme B production among cd T cells from BCG-vaccinated volunteers compared to unvaccinated were also observed. 66 Together, these results suggest an important role for cd T cells in mediating protective immunity to malaria.
Although there is not yet an effective vaccine for malaria, preliminary studies testing whole parasite vaccines in humans and mice suggest an important role for cd T cells in protection from subsequent infection. The malaria vaccine that has advanced farthest to date is the RTS,S vaccine, which is based on the Pf circumsporozoite (CSP) protein and targets the sporozoite and liver stages of infection. Interestingly, RTS,S phase 3 trials in African children detected no significant change in cd T-cell frequencies following vaccination and minimal cytokine production by these cells in response to in vitro CSP stimulation. 67 However, as the authors examined total cd T cells rather than Vd2 + or other cd T-cell subsets, it will be important for future studies to determine whether specific subsets correlate with protection and if so, whether future RTS,S formulations can target these subsets. RTS,S trials in malaria-na€ ıve populations have generally focused on anti-CSP antibody studies and CD4 + / CD8 + T-cell responses without examining innate populations like cd T cells. One recent study utilising a systems approach identified natural killer (NK) cell signatures that correlated with and predicted protection, 68 suggesting that depending on the precise vaccine regimen, innate immune responses could be significant.
In contrast to RTS,S, vaccine formulations using sporozoites (the stage of the parasite injected by the mosquito into the human) have indicated a direct or indirect role for cd T cells in protection. In malaria-na€ ıve individuals immunised with the attenuated Pf sporozoite (PfSPZ) vaccine, Vd2 + T cells expanded in a dose-dependent fashion and frequencies of these cells correlated with protection more significantly than any other cellular immune responses. [69][70][71] Numbers of memory Vd2 + T cells also correlated with protection in a recent PfSPZ trial in a malariaendemic region in Mali. 72 Finally, when malaria-na€ ıve individuals were immunised with nonirradiated PfSPZ combined with chemoprophylaxis (PfSPZ-cVAC), the frequency of Vd2 + T cells increased in a dose-dependent manner and memory cd T cells specifically increased expression of IFNc and the activation marker CD38. 73 Additional work is needed to further elucidate the mechanism of Vd2 + T-cell-induced protection, as well as to determine whether frequencies of these cells could be used as a biomarker for protection in PfSPZ vaccinations in malariaendemic regions.
In the mouse model, results have depended somewhat on the parasite strain used, but generally support cd T cells as a correlate of natural and vaccine-induced protection. In the lethal Plasmodium berghei ANKA model, cd T cells were not required to prevent infection upon blood-stage challenge following sporozoite vaccination, but did contribute to pre-erythrocytic immunity by recruiting dendritic cells and CD8 + T cells. 72 These cells may also be important in modulating functional T follicular helper (Tfh) cell and germinal centre B-cell responses. 74 In contrast to these indirect roles in protection, cd T cells appear to act as important effector cells following vaccination with nonlethal Plasmodium yoelii sporozoites. 75 Results from mice lacking ab T cells further suggest that cd T-cell cytotoxicity may become more effective after interaction with CD4 + T cells. 75 Mice lacking cd T cells further reveal that these cells may be particularly important in immunity targeting the liver stages of the parasite (before it enters the bloodstream). 76 Clearly, it will be important to evaluate whether these differing results between murine parasite strains are solely because of differences in the type of immunity induced (i.e. P. berghei-irradiated sporozoite vaccination induces sterile immunity, while P. yoelii vaccination does not). Interestingly, a vaccine using whole lysate of the promastigote stage of a related parasite, Leishmania amazonensis, led to protection against subsequent infection that was dependent on the presence of cd T cells. 77 The mechanisms driving this protection and implications for malaria vaccines, however, are unknown.
In sum, results from vaccination studies targeting malaria (and potentially other parasitic infections such as leishmaniasis) strongly suggest that cd T cells play an important role in protection from future infection. However, future work is required to definitively show that cd T cells directly mediate protection rather than act as a biomarker of infection, as well as to determine the mechanism of protection and the role of Vd2 À subsets (if any). In particular, it will be important to assess whether protection is mediated via direct cd T-cell cytotoxicity and/or more indirect effects such as antigen presentation, recruitment of other cell types, or stimulation of functional Tfh cells and antibodies. Given that most malaria vaccines in trials, including the leading RTS,S vaccine, use specific antigens rather than whole sporozoites, vaccine effectiveness may be improved by the addition of an adjuvant or other vaccine component that stimulates cd T-cell responses. BCG vaccination may be a potential approach based on recent results of increased activation of innate cell populations following CHMI in BCGvaccinated individuals; 66 however, given that this response only occurred in half of the vaccinated volunteers and the sample size was small, further study is warranted.

VIRAL INFECTIONS
There is evidence that cd T cells may play a role in response to viral infections, including influenza virus, HIV and cytomegalovirus (CMV), and that they can directly kill virally infected cells. There is also evidence that these cells can expand in vivo in response to bisphosphonate stimulation and viral vaccination strategies and may contribute to improved outcomes, thereby raising the possibility that these cells could be targeted to play an important role in vaccine-mediated protection.
Regarding influenza, several studies have shown that phosphoantigen or pamidronate-activated cd T cells are capable of inhibiting virus replication by killing influenza-infected macrophages 78 and/ or lung alveolar epithelial cells. 79 Phosphoantigenactivated cells also have non-cytolytic activities in response to pandemic H1N1, producing IFNc and expressing inflammatory chemokines. 80 Relatedly, it was also recently shown that Vc9Vd2 T cells can promote CD4 + T follicular helper cell differentiation, B-cell class switching and influenza virus-specific antibody production in an in vitro co-culture assay, 81 suggesting that these cells may provide both a direct cytotoxic and potential synergistic role in the adaptive immune response to influenza.
Although both inactivated and live attenuated influenza vaccine reduce influenza illness and disease complications, live attenuated influenza vaccine has been shown to have superior efficacy in children. 82 Influenza-responsive cd T cells were found to expand following live attenuated, but not inactivated, influenza vaccination, 83,84 suggesting a potential immunologic correlate for this observation. Despite not proliferating after vaccination, cd T cells in elderly individuals receiving the inactivated vaccine did increase perforin production and, after in vitro restimulation, proliferated and produced IFNc and IL-4. 84 Similarly, the cd T-cell response in the nasal mucosa was attenuated in cigarette smokers relative to non-smokers, 85 suggesting these cells may represent a correlate for why smokers respond less well to influenza vaccination. In a murine model of influenza, cd T cells significantly expand in bronchial alveolar fluid following infection, 86 and in a humanised mouse model, pamidronate administration to mice reconstituted with human PBMC reduced disease severity and mortality following H1N1 and H5N1 influenza infection. However, pamidronate had no effect in mice reconstituted with Vd2 À depleted cells. 87 Together, these studies suggest that cd T cells may not only represent an immunologic correlate of protection from influenza infection and vaccination, but that they might also be a mediator of protection.
Regarding HIV, it has long been known that both the Vd1 + and Vd2 + subsets of cd T cells have cytotoxic capacity against HIV [88][89][90] and can inhibit viral replication in vitro. HIV-infected elite controllers have elevated levels of Vd2 + T cells compared with HIV-negative controls or HIVinfected individuals on antiretroviral therapy, 91,92 suggesting a potential role for these cells in inhibiting viral replication in vivo. cd T cells may also play a role in controlling viral infection at mucosal barriers. A recent study reported that higher levels of pro-inflammatory Vd1 + T cells correlated with lower gut-associated HIV viral load, 93 and another study in rhesus macaques found that levels of CD8 + Vd2 + T cells in the female reproductive tract correlated with lower SIV viral loads. 94 Vd1 + T cells expanding in HIVinfected individuals may also protect from other infections. For example, Vd1 + T cells producing IFNc and IL-17A responded to Candida albicans 95 and further expanded upon influenza vaccination combined with the MF59 adjuvant. 96 Individuals with chronic HIV infection have been found to have Vd2 + T-cell depletion and dysfunction in response to phosphoantigenic stimulation. 97 It is possible, however, that some of these cells are not dysfunctional but rather have different functions. For example, He et al. identified a population of CD16 + Vd2 + T cells that had decreased responses to phosphoantigens but increased capacity for antibody-dependent cellular cytotoxicity (ADCC). A decline in this population was associated with faster disease progression, while no decline was observed in individuals with controlled infection. 98 Administration of zoledronic acid with IL-2 in HIV-infected, antiretroviral na€ ıve patients was associated with Vd2 + T-cell expansion, dendritic cell activation and increased HIV-specific CD8 + T-cell responses. 99 It was also recently shown that cd T cells can be isolated from antiretroviral suppressed, HIVinfected individuals and that these cells can kill autologous HIV-infected CD4 + T cells. In addition, these cells could expand ex vivo following pamidronate stimulation and could significantly reduce viral replication, suggesting a potential role for these cells to clear HIV infection from latent reservoirs. 100 Even though HIV vaccine trials to date have not investigated any changes in cd T-cell populations, an intriguing study looked at canarypox as a vector for HIV antigens and, after in vitro expansion, identified a Vc9 + population (specific for canarypox, not HIV antigens) that produced IFNc. 101 These results suggest that in addition to adjuvants, vaccine vectors could be used to target cd T-cell responses.
Finally, in the context of CMV infection, oligoclonal cd (primarily Vd2 À ) T cells expand and differentiate into effector/memory cells. [102][103][104][105] Expansion of Vd2 À T cells is associated with viral clearance both in immunosuppressed 102,106,107 and in healthy populations. 102,107 These cells likely contribute to viral clearance via effector functions such as cytotoxicity and IFNc/TNFa production, 108 'antibody-dependent cell-mediated inhibition', 109 and enhanced cytotoxicity via sensing of IL-18 from virus-infected cells. 110 During secondary infection, cells proliferate and resolve infection Table 1.    faster, suggesting a memory-like phenotype. 102 Several studies in mice have shown that (1) cd T cells are capable of protecting ab T-cell-deficient mice against CMV-induced pathology and (2) adoptive transfer of CMV-induced cd T cells provides long-term protection in immunodeficient mice. 111,112 These results suggest that cd T cells are important mediators of protection against CMV and support approaches using adoptive transfer of effector/memory cd T cells or targeting cd T cells in future CMV vaccine trials. The possibility of inducing exhausted cd T cells would need to be considered, however, as CMV infection has both been shown to result in higher numbers of these cells. 113 In sum, results from in vitro and natural infection studies suggest an important role for cd T cells in controlling influenza, HIV and CMV viral replication. Targeting cd T cells through stimulation could provide an important adjuvanttype role in vaccination and/or cure-related strategies for viral infections.

CONCLUSIONS
Across the different bacterial, protozoan and viral infections examined (summarised in Table 1), there are clear patterns of cd T-cell expansion, particularly of the Vd2 + subset, in response to both infection and vaccination. In several contexts, including infection with Mtb, malaria, influenza and HIV and vaccination with BCG, PfSPZ and live attenuated influenza, cd T cells are associated with protection. Further, evidence so far supports a role for cd T cells in mediating protection via direct killing and other mechanisms. Studies in animal models, such as BCG vaccination in macaques and PfSPZ vaccination in mice, are beginning to shed light on direct mechanisms of protection vs. stimulation of other immune cells that mediate protection. Clearly, future work is needed to further elucidate these mechanisms, as well as the host and infection-mediated factors that influence responsivity of cd T cells and the relevant differences between responses to natural infection compared to response to vaccination. As new vaccine formulations targeting these diseases progress through development, the question of whether to induce cd T cells or cd T-cell subsets will become an important consideration. In fact, this approach is already being implemented in cancer, whether via administration of Vc9Vd2 T-cell agonists 114 or using BCG to stimulate Vc9Vd2 T cells as treatment for bladder cancer. 115,116 Approaches incorporating cd T cells into strategies targeting B-or T-cell responses have also been promising so far. For example, as previously mentioned, a study testing a subunit tuberculosis vaccine combined with phosphoantigen observed a robust cd T-cell response, including expression of effector memory markers, following primary vaccination. 25 Finally, another intriguing approach is to expand functional cd T cells ex vivo, as has been tested with effector cells capable of inhibiting HIV replication 100 and Mtb infection. 23 To maximise functional responses in future similar studies, it will be important to improve our understanding of the timing of cd T-cell vs. ab Tcell responses following vaccination, as well as any potential negative effects of overstimulation of cd T cells. As specific subsets of cd T cells that correlate with protection in different contexts are identified, optimisation of methods to specifically target these subsets will be beneficial. Especially given the hypothetical possibility of cd T-cell anergy/exhaustion, it will be essential to define responses that optimally stimulate and antigens/ agonists that best elicit that response. Altogether, as development of vaccines targeting infectious diseases that have long proved elusive becomes more of a reality, it will be important to broaden our perspective beyond targeting antibody-driven or T-cell responses and to intentionally target innate cells, such as cd T cells.