Oral typhoid vaccine Ty21a elicits antigen-specific resident memory CD4+ T cells in the human terminal ileum lamina propria and epithelial compartments

Background Salmonella enterica serovar Typhi (S. Typhi) is a highly invasive bacterium that infects the human intestinal mucosa and causes ~ 11.9–20.6 million infections and ~ 130,000–223,000 deaths annually worldwide. Oral typhoid vaccine Ty21a confers a moderate level of long-lived protection (5–7 years) in the field. New and improved vaccines against enteric pathogens are needed but their development is hindered by a lack of the immunological correlates of protection especially at the site of infection. Tissue resident memory T (TRM) cells provide immediate adaptive effector immune responsiveness at the infection site. However, the mechanism(s) by which S. Typhi induces TRM in the intestinal mucosa are unknown. Here, we focus on the induction of S. Typhi-specific CD4+TRM subsets by Ty21a in the human terminal ileum lamina propria and epithelial compartments. Methods Terminal ileum biopsies were obtained from consenting volunteers undergoing routine colonoscopy who were either immunized orally with 4 doses of Ty21a or not. Isolated lamina propria mononuclear cells (LPMC) and intraepithelial lymphocytes (IEL) CD4+TRM immune responses were determined using either S. Typhi-infected or non-infected autologous EBV-B cell lines as stimulator cells. T-CMI was assessed by the production of 4 cytokines [interferon (IFN)γ, interleukin (IL)-2, IL-17A and tumor necrosis factor (TNF)α] in 36 volunteers (18 vaccinees and 18 controls volunteers). Results Although the frequencies of LPMC CD103+ CD4+TRM were significant decreased, both CD103+ and CD103− CD4+TRM subsets spontaneously produced significantly higher levels of cytokines (IFNγ and IL-17A) following Ty21a-immunization. Importantly, we observed significant increases in S. Typhi-specific LPMC CD103+ CD4+TRM (IFNγ and IL-17A) and CD103− CD4+TRM (IL-2 and IL-17A) responses following Ty21a-immunization. Further, differences in S. Typhi-specific responses between these two CD4+TRM subsets were observed following multifunctional analysis. In addition, we determined the effect of Ty21a-immunization on IEL and observed significant changes in the frequencies of IEL CD103+ (decrease) and CD103− CD4+TRM (increase) following immunization. Finally, we observed that IEL CD103− CD4+TRM, but not CD103+ CD4+TRM, produced increased cytokines (IFNγ, TNFα and IL-17A) to S. Typhi-specific stimulation following Ty21a-immunization. Conclusions Oral Ty21a-immunization elicits distinct compartment specific immune responses in CD4+TRM (CD103+ and CD103−) subsets. This study provides novel insights in the generation of local vaccine-specific responses. Trial registration This study was approved by the Institutional Review Board and registered on ClinicalTrials.gov (identifier NCT03970304, Registered 29 May 2019—Retrospectively registered, http://www.ClinicalTrials.gov/NCT03970304)

Background CD4+ T cells are crucial for the generation of vaccine-mediated immune responses providing effective immunity against pathogens. Recent studies have demonstrated that both circulating T memory (T M ) cells and tissue resident memory T cells (T RM ) abundant in peripheral tissues are central to elicit protective immunity [1,2]. T RM represent a non-migratory population of T M that is phenotypically different from circulating T M subsets [e.g., T central memory (T CM ) and T effector memory (T EM )] and mediate rapid effector immune responses following antigen recall [2]. Human T RM are mainly characterized phenotypically by high expression of CD69, a key marker for distinguishing between circulating and resident T M [3]. Integrin αEβ7 (CD103), the ligand to E-cadherin, is also used to characterize T RM but CD103 expression is mostly confined to CD8+ T RM and a minor subset of CD4+T RM [3][4][5][6]. Recent studies have characterized CD4+ T RM subsets in multiple organs including lungs, liver, skin, intestines, vagina and lymphoid sites where they provide protective responses and direct the recruitment of immune cells [7][8][9][10]. In the human intestine, the majority of CD4+T RM are CD69+CD103− and a minority are CD69+CD103+ [3] but little information is available concerning their role in oral immunization or enteric infections. It is unclear whether live oral attenuated typhoid vaccine, Ty21a, could elicit and generate antigen-specific CD4+T RM responses in the human terminal ileum (TI) (site of infection for S. Typhi).
Salmonella enterica serovar Typhi (S. Typhi), the etiological agent of typhoid fever, is a human restricted pathogen, that infects around 21 million people globally resulting in approximately to 223,000 deaths yearly [11][12][13]. After actively infecting primarily the TI [14][15][16], S. Typhi disseminate to the systemic organs (e.g., liver, spleen and others) leading to systemic illness [17]. To our knowledge, no information is available on the induction of CD4+T RM responses to S. Typhi in the TI mucosa following wild-type S. Typhi infection or immunization with Ty21a. Since the gastrointestinal track is a major reservoir of CD4+T RM , it is important to understand their role in protective immunity against S. Typhi and other enteric pathogens at their preferred site of natural infection. This knowledge will provide novel insights for the development of tailored mucosal vaccines to enteric pathogens. Currently, there are two licensed typhoid vaccines namely, Ty21a and the parenteral Vi polysaccharide vaccine [18,19] and both have their limitations. Following Ty21 immunization, a modest level of long-lived protection (60-80%, 5-7 years) is obtained [20][21][22] while the Vi vaccine is moderately immunogenic but well tolerated [17,23]. Therefore, effective vaccines that offer durable, long-lasting protection are urgently needed for children and adults.
Likewise, we have characterized intraepithelial lymphocytes (IEL) CD8+T RM S. Typhi specific responses in the epithelium compartment following Ty21a-immunization [16]. Interestingly, no information is available on the role of the IEL CD4+T RM following oral Ty21a-vaccination. Functional studies with human intestinal IEL are challenging due to their low yield during the isolation process. Nonetheless, IEL are among the first line of defense and it is therefore important to understand their role in oral Ty21a-vaccination and S. Typhi infection.
In this study, we have characterized TI LPMC and IEL CD4+T RM subsets following Ty21a immunization. We then determined and compared CD4+T RM S.
Typhi-specific responses between the two groups. Finally, we assessed single and multifunctional S. Typhi specific responses in CD4+ T RM subsets by single cell analysis. These comparisons provide unique insights into the generation of S. Typhi specific responses in the human TI mucosa.

Volunteers, immunization and sample collection
Individuals (aged 50-74 years) undergoing routine colonoscopy were enrolled from the Baltimore-Washington metropolitan area and University of Maryland, Baltimore campus. Volunteers who have no previous history of typhoid fever and were not vaccinated with the attenuated Ty21a typhoid vaccine were assigned to each of two groups. Four recommended doses of Ty21a (Vivotif enteric-coated capsules; Crucell, Bern, Switzerland) were administrated to the first group (n = 18) but not to the control group (n = 18) (Additional file 1: Fig. S1). Blood samples were collected at least 21 days before colonoscopy (pre-immunization) and on colonoscopy day (day 0) together with TI biopsies using large capacity forceps (Additional file 1: Fig. S1). PBMC were isolated immediately after blood draws by density gradient centrifugation and cryopreserved in liquid nitrogen following standard techniques [32].

Ethics statement
Written informed consent was obtained from subjects and all procedures were approved by the University of Maryland, Baltimore Institutional Review Board (IRB) and registered on ClinicalTrials.gov (identifier NCT03970304). The study was conducted in accordance with the principles of the International Conference of Harmonization Good Clinical Practice guidelines.

Target cell preparation and S. Typhi infection
Autologous Epstein-Barr virus (EBV)-transformed lymphoblastoid cell line (EBV-B cells) were generated from each participant's pre-immunized PBMC (Additional file 1: Fig. S1) as previously described [25,32]. Target cells were then infected with wt-S. Typhi strain ISP1820 at a MOI of 7:1 as previously described [35]. Infected target cells were then gamma-irradiated (6000 rad) before used for ex vivo LPMC and IEL stimulation. To confirm S. Typhi infection, target cells were stained with anti-Salmonella common structural Ag (CSA-1, Kierkegaard and Perry, Gaithersburg, MD) and analyzed by flow cytometry as previously described [25,32].

Stimulation of terminal ileum LPMC and IEL
Freshly isolated TI-LPMC and IEL were used as effector cells as previously described [16,34,37]. Briefly, LPMC and IEL were co-cultured, respectively, with either uninfected or S. Typhi-infected EBV-B (MOI of 7:1). LPMC and IEL cultured with media only or in the presence of α-CD3/CD28 (Life technologies, Grand Island, NY) were used as negative and positive controls, respectively. After 2 h, 0.5 μL of Golgi Stop (Monensin, BD) and 0.5 μL Golgi Plug (Brefeldin A, BD) were added and cultures continued overnight at 37 °C in 5% CO 2 .

Statistical analysis
Data were analyzed using the statistical software Graph-Pad Prism ™ version 5.03 (Graphpad, San Diego, CA, USA). Statistical differences in median values between two groups were determined using Mann-Whitney tests. Wilcoxon matched pair tests were used to assess statistical differences between LPMC and IEL paired responses. Correlations (LPMC versus IEL S. Typhi-specific responses) were evaluated using Spearman correlation tests.

Data availability
The datasets supporting the findings of this study are available within the article and its additional information files.

Activation of terminal ileum LPMC CD4+T RM
CD4+T cells is the principal population in the TI lamina propria (LP) comprising most of the local CD4+T RM (Fig. 1a). The observation that oral Ty21a-immunization influences CD4+T RM frequencies ( Fig. 1b) suggests that TI-LPMC CD4+T RM subsets might respond differently in magnitude and characteristics following stimulation.
To test this hypothesis, we determined the cytokine pro- Therefore, we observed that both CD4+T RM subsets are responsive to S. Typhi-stimulation following oral Ty21a immunization. Interestingly, we noted that both CD4+T RM subsets exhibited differences in the background (unstimulated) and α-CD3/CD28 stimulation responses between Ty21avaccinated and unvaccinated volunteers. Therefore, we hypothesized that Ty21a-immunization might influence the induction of spontaneous cytokines (background responses) and the capacity of CD4+T RM to respond to stimulation. To address this hypothesis, we evaluated CD4+T RM (CD103+ and CD103−) cytokines producing cells when LPMC cells were cultured overnight either alone (unstimulated) or following stimulation with α-CD3/CD28 beads. Cumulative data from unstimulated LPMC show that there were significant (p < 0.05) increase in IFNγ-S (but not IFNγ-MF) producing cells in both CD4+T RM subsets (Fig. 2a). Interestingly, we observed that the baseline level of IFNγ producing cells by CD103+ CD4+T RM cells were significantly higher than those from CD103− CD4+T RM as shown by green lines (Fig. 2a). Next, we determined spontaneous IL-17A responses and observed that CD103+ CD4+T RM exhibited significantly higher levels of IL-17A-S following Ty21a-immunization (Fig. 2b). Furthermore, we noted that the level of IL-17A (S and MF) were significantly higher in CD103+CD4+T RM than in CD103− CD4+T RM (Fig. 2b). Similar assessments were made for IL-2 and TNFα for background responses (Fig. 2c, d). Interestingly, no differences in either CD4+T RM subset were noted for IL-2 and TNFα, except for CD103− CD4+T RM that showed a decreased trend in IL-2-S ( Fig. 2c) and an increased trend in TNFα-S (Fig. 2d) responses following Ty21a-immunization. Interestingly, we also observed significantly higher levels of IL-2-MF and TNFα-MF CD4+CD103+T RM than CD4+CD103− T RM in Ty21a-vaccinees (Fig. 2d).

LPMC-CD4+T RM subsets respond differently following Ty21a immunization
CD4+T RM respond rapidly following re-exposure to pathogens. However, no information is available on CD4+T RM vaccine-induced responses, particularly the characteristics of CD103+ and CD103− CD4+T RM following oral Ty21a-immunization in the human TI. Thus, we hypothesized that TI CD103+ and CD103− CD4+T RM would contribute differently in the CMI responses following Ty21a immunization.

Multifunctional LPMC CD4+T RM S. Typhi responses following oral Ty21a-immunization
In various diseases including typhoid fever, induction of antigen-specific multifunctional T cells have been shown to be associated with favorable disease outcome, higher effector function and higher protective efficacy after immunization [39][40][41][42][43] compared to monofunctional T cells. In addition, recent evidences show that there are molecular differences between multifunctional and monofunctional CD4 + T cells [44]. Thus, we deemed that examination of multifunctional CD4+T RM in the terminal ileum mucosa is important for assessing the quality of the responses following Ty21a immunization.
To further investigate the differences in S. Typhi-specific responses between LPMC CD103+ and CD103− CD4+T RM , we used the Winlist FCOM function to analyze multiple cytokines (i.e., IFNγ, IL-17A, IL-2, and TNFα) in individual S. Typhi-specific responding cells and classified them as either single cytokine producer (S) or multifunctional (Sum of double, triple, and quadruple cytokine producers) (MF).

Oral Ty21a-immunization elicits S. Typhi-specific terminal ileum IEL CD4+T RM
The intestinal mucosa is composed of a single layer of intestinal epithelial cells that is superimposed on the LP. Because LPMC CD103 +CD4+T RM expresses CD103, a ligand for E-cadherin on epithelial cells, it is reasonable to speculate that CD103+ CD4+T RM are poised to migrate and contribute to S. Typhi responses in the epithelial microenvironment. Thus, we hypothesized that oral Ty21a-immunization might influence the frequencies and baseline responses of intraepithelial lymphocytes (IEL) CD4+T RM in the human TI epithelial compartment. To address this hypothesis, we freshly isolated IEL and phenotypically characterized CD4+, CD8+ T cells and CD4+T RM subsets in Ty21a-vaccinated (n = 17) and unvaccinated (n = 17) volunteers. Representative examples of the gating strategy for a Ty21a vaccinated and an unvaccinated individual are shown in Fig. 6a. Interestingly, we observed that the frequencies of IEL-T cells were significantly altered following oral Ty21a immunization. Specifically, the frequencies of CD3+CD8+ IEL T cells were significantly (p < 0.005) decreased while CD3+CD4+ IEL T cell frequencies were significantly (p < 0.0005) increased following Ty21aimmunization (Fig. 6b). To confirm that there were no discrepancies in cell yields, the number of viable IEL per mg of TI tissues obtained from biopsies of Ty21avaccinated and unvaccinated volunteers were compared. No significant differences in cell yields were observed between the two groups (Additional file 6: Fig. S6). Similarly, almost identical cell yields were observed in LPMC between the two groups. However, significantly higher cell numbers were observed in LPMC as compared to IEL (Additional file 6: Fig. S6). Next, we observed that TI CD103 + and CD103− CD4+T RM were present in similar proportions in IEL as in LPMC (Fig. 6a). Interestingly, the frequencies of IEL CD103+ CD4+T RM were significant (p < 0.05) decreased following oral Ty21a-immunization (Fig. 6C). In contrast, IEL CD103− CD4+T RM frequencies were significantly (p < 0.05) increased in Ty21a vaccinated as compared to unvaccinated volunteers (Fig. 6c). Thus, these data suggest that the increase in IEL CD4+T cells result mainly from increases in the CD103− CD4+T RM subset. We next measured and compared the frequency of baseline S. Typhi single (S) and multifunctional (MF) producing cytokines (IFNγ, IL-17A, IL-2 and TNFα) in unstimulated IEL-CD4+T RM subsets obtained from Ty21a vaccinated (n = 17) and unvaccinated (n = 17) volunteers following an overnight incubation (Fig. 7). Exvivo unstimulated IEL CD4+T RM obtained from Ty21a immunized volunteers exhibited significantly higher level of IFNγ-S but not MF than controls (Fig. 7a). In contrast, no significant differences in the production of spontaneous IL-17A-S or MF, TNFα-S or MF and IL-2-S from IEL CD4+T RM subsets following Ty21a-immunization ( Fig. 7b-d). Of note, IEL CD103− CD4+T RM produced significantly (p < 0.05) less IL-2-MF in Ty21a vaccinated than in unvaccinated volunteers (Fig. 7c). We then studied multifunctional responses of IEL CD103+ and CD103− CD4+T RM in Ty21a vaccinated (n = 10) and unvaccinated (n = 8) volunteers following stimulation with α-CD3/CD28 beads. Remarkably, while α-CD3/ CD28 stimulated equally CD4+T RM subsets from both groups (Additional file 7: Fig. S7A-D), we observed trends to show significant increases in TNFα-S CD103+ CD4+T RM and in IFNγ-S CD103−CD4+T RM , as well as significant (p < 0.05) increases in TNFα-S responses by CD103− CD4+T RM in Ty21a vaccinees (Additional file 7: Fig. S7A-D). Above we showed that LPMC S. Typhi-specific CD4+T RM were elicited by oral Ty21a immunization. Thus, we hypothesized that oral Ty21a-immunization would induce distinct S. Typhi specific IEL CD4+T RM responses. To test this hypothesis, net S. Typhi responsive IEL CD4+T RM were evaluated following stimulation with autologous S. Typhi-infected targets in Ty21a vaccinated (n = 7) and unvaccinated (n = 6) volunteers. No significant differences were observed in cytokine production (IFNγ, IL-17A, IL-2 and TNFα) from IEL CD103+ CD4+T RM following Ty21a-immunization (Additional file 8: Fig. S8A). In contrast, the predominant IEL CD103− CD4+T RM exhibited a trend to show increased IFNγ and TNFα (p = 0.06) responses in Ty21a-vaccinated than in unvaccinated volunteers (Additional file 8: Fig.  S8B).
Intestinal T RM (CD4+ and CD8+) are located in both the lamina propria and epithelial compartments, which provides unique immunological niches. Thus, we hypothesized that the CD69 and CD103 expression profile of CD4+T RM subsets may differ between LP and epithelial compartments following Ty21a immunization. To address this hypothesis, we measured the mean fluorescence intensity (MFI) of CD69 and CD103 expressed on LPMC and IEL CD4+T RM subsets obtained simultaneously from Ty21a vaccinated (n = 8) and unvaccinated (n = 8) volunteers (Additional file 9: Fig. S9). As shown in cytograms from a representative unvaccinated volunteer, CD103 is highly expressed on both LPMC and IEL CD103+ CD4+T RM with similar intensities in IEL and LPMC (Additional file 9: Fig. S9A). Next, we compared the level of CD103 MFI between IEL and LPMC CD4+T RM subsets obtained from either Ty21a-vaccinated (n = 8) or unvaccinated (n = 8). Cumulative data showed no significant differences in the levels of CD103 expression between LPMC and IEL CD4+T RM subsets (Additional file 9: Fig. S9B).
Furthermore, we explored the relationship between the generation of S. Typhi specific immune responses between LPMC and IEL individually by performing Spearman correlation tests between LPMC and IEL CD4+T RM S and MF responses in both Ty21a vaccinated and unvaccinated volunteers. Interestingly, in unvaccinated volunteers, the frequencies of LPMC CD103+CD4+ T RM IL-2-S responses only were significantly positively correlated to their IEL counterparts while CD103−CD4+T RM subset has significant positive correlation only in IL-17A-S (Additional file 10: Table S1). However, following Ty21a-vaccination, the frequencies of LPMC CD103+CD4+ T RM S (IFNγ) were significantly positive correlated while their MF (IL-17A and IL-2) counterparts were significantly negatively correlated to their IEL counterparts (Additional file 10: Table S1). In contrast, the frequencies of LPMC CD103−CD4+T RM S (IFNγ and IL-17A-negative; TNFα-positive) but not MF were significantly correlated to their IEL counterparts (Additional file 10: Table S1).

Discussion
Vaccine-mediated protective immunity is generally enabled by the induction of both, antibodies and appropriate cellular immune responses, including those mediated by CD4+T cells. There is growing evidence that T RM (e.g., CD4+T RM ) cells play a crucial role in protective immunity following natural infection and their subsequent secondary exposure. However, there are no previous reports on the role of CD4+T RM following oral Ty21a-immunization in the human TI. Here, we examined the responses elicited by Ty21a-immunization on TI LPMC and IEL CD4+T RM subsets. We observed that Ty21a-immunization activated LPMC CD103+ and CD103− CD4+T RM resulting in increased spontaneous cytokines (IFNγ, TNFα and IL-17A) production. Importantly, LPMC CD4+T RM subsets contributed significantly to S. Typhi-specific IFNγ, IL-17A and IL-2 responses following stimulation with S. Typhi-infected targets. Moreover, these responses differed in magnitude and characteristics between CD103+ and CD103− CD4+T RM , suggesting a dichotomy in their contributions and possibly different roles in S. Typhi immunity. Finally, we demonstrated that Ty21a-immunization induces IEL CD4+T RM subsets in the epithelial compartment both spontaneously and following antigenspecific stimulation. Interestingly, IEL CD103− CD4+ T RM contributed significantly to S. Typhi specific IFNγ, IL-17A and TNFα as single producing cells while IEL CD103+CD4+T RM contributed to IL-2 production. Taken together, these results contribute major novel information of the effects of oral Ty21a-vaccination and the role of CD4+T RM in human TI mucosal responses.
The intestinal mucosa is a major site of antigenic exposure from various sources including the microbiome and enteric pathogens. Multiple specialized populations of immune cells (adaptive and innate) including CD4+T RM reside in the intestine and contribute to intestinal "homeostasis" and protective immunity against enteric pathogens. S. Typhi or the attenuated S. Typhi vaccine Ty21a strain can effectively invade the TI mucosa, preferred site of S. Typhi infection [45,46]. Then, it is reasonable to hypothesize that exposure to S. Typhi elicits both innate and adaptive responses including CD4+T RM in the lamina propria and the epithelial compartments of the TI which might play a significant role in protection against S. Typhi. The complexities involved in obtaining cells from the human terminal ileum biopsies (LPMC and IEL), are likely to have contributed to the lack of information available on CD4+T RM immune responses following oral vaccination. Using this unique model of oral live attenuated Ty21a vaccine in humans, we showed that oral Ty21a-immunization resulted in the spontaneous production of cytokine cells in the LP from CD103+ (IFNγ-S and IL-17A-S) and CD103− (IFNγ-S and TNFα-S) CD4+T RM and in the epithelium where IFNγ-S was produced spontaneously from both CD4+T RM subsets. Similar effects (e.g., increase in IFNγ-S and TNFα-S in both LP and IEL) elicited by Ty21a-immunization were observed following stimulation with α-CD3/CD28 stimulation. We hypothesize that there may be multiple cell subsets within each CD4+T RM subpopulation that differ in their requirements for activation and/or their recognition of specific cognate antigens [47][48][49]. Future studies should focus on fully understanding the heterogeneity and contribution of the resident cells in spontaneous cytokine production following Ty21a-immunization. These cells might play a role in protection, albeit likely to be of lesser importance than those mediated by antigenspecific responses. Overall, these unique data indicate that immunization with bacterial oral vaccines may have immunomodulatory effects beyond those that are specific for the vaccine being administered leading to a generalized level of activation.
While the cellular requirements and transcriptional basis for the induction of CD8+T RM has been studied in other systems, our understanding of CD4+T RM generation and maintenance is poor. CD4+T RM unlike CD8+T RM are less uniform in their expression of CD103 and consequently two distinct subsets are present in tissues. Here, TI CD103+ and CD103− CD4+T RM elicited immune responses were evaluated following Ty21a immunization. Our data provided the first evidence of distinct responses between these two subsets in terms of spontaneous secretion of cytokines, stimulation with anti-CD3/CD28, and S. Typhi-specific responses resulting from stimulation with S. Typhi-infected targets and Ty21a homogenate antigens. Specifically, our results showed striking differences in terms of the characteristics (IFNγ and S for CD103+; IL-2 and MF for CD103−) of these responses. Of note, a recent study reported differences in IFNγ responses between lung CD103+ and CD103− CD4+T RM following stimulation with α-CD3/ CD28 [7]. Taken together, these results suggest that CD103+CD4+T RM subset (Th1 and Th17) are distinct from CD103− CD4+T RM subset (Th1) and might play a different role in the mucosa following oral vaccination and/or infection.
We have previously characterized S. Typhi-specific LPMC CD4+T EM cells and observed increasing trends in IFNγ and IL-17A production following Ty21a immunization [50]. Here, we showed that the IFNγ and IL-17A are primarily produced by CD4+T RM (a subset of CD4+ T EM ) in significantly higher levels in the terminal ileum mucosa following Ty21a immunization. In addition, we observed that CD103+ CD4+T RM produced cytokines as monofunctional cells whereas CD103− CD4+T RM cells produced mostly multifunctional cytokines, similar to our observations in LPMC CD4+T EM [50]. These results suggest that examination of heterogeneous CD4+ T populations may mask the 'true' impact of oral Ty21a immunization due to the 'averaging effect' inherent to the analyses of whole populations/subsets which are composed of responding and non-responding cells. Thus, it is essential to focus on the fine granularity (e.g., S. Typhispecific CD103− or CD103+ T RM subsets, S vs MF) to better characterize the responses and properly study differences between cells in various immune compartments. Here, we demonstrated that oral Ty21a immunization elicited CD4+ T RM subsets in both, the LP and epithelial compartments, involving effector mechanisms (e.g., IL-17A) that might be well suited for protection against intracellular pathogens.
The type of antigen used for in vitro stimulation influence the responses of CD4+T M . Thus, we addressed this issue by using S. Typhi-infected targets (moderately efficient CD4+ T stimulation) and Ty21a homogenate antigens (an efficient CD4+ T stimulation) [51] to stimulate CD4+T RM subsets. Remarkably, our results indicate that following stimulation with Ty21a homogenate antigens, LPMC CD103+T RM largely produced significantly higher levels of S. Typhi-specific IFNγ-MF, IL-2-MF and TNFα-MF in Ty21a vaccinated participants than those in the unvaccinated group. These results were different from those observed following stimulation with S. Typhi-infected targets (IFNγ-S and IL-17A-S), which largely display S. Typhi antigens in the context of both MHC-I and II molecules. Taken together, these data suggest that LPMC CD4+T RM subsets are versatile in responding to S. Typhi antigens. Of note, we recently described that oral Ty21a-immunization elicits the induction of S. Typhi-specific CD8+T RM [16] and were able to respond following in vitro stimulation with soluble antigens and S. Typhi-infected targets, indicating that multiple effector T cell responses are likely to be concomitantly induced. Future studies focusing on the activation requirements of CD4 and CD8 populations in human tissues would be vital to better understand these phenomena to accelerate the development of mucosal vaccines targeting the induction of CMI.
Intestinal T cells are located diffusely throughout the epithelial compartment as intraepithelial lymphocytes (IEL). The predominant population of IEL-T cells are CD8+T with a minor population of CD4+T cells. IEL are part of the first line of defense and are deemed to be an important cell subset involved in immune responses at mucosal surfaces. Thus, in this study, we hypothesized that CD103+CD4+T RM (CD103 binds to E-cadherin on IEC) located in the LP are poised to migrate to the epithelium and play a consequential role in S. Typhi immunity. We describe, for the first time, that oral Ty21a-immunization influenced the frequencies of IEL-T cells with a significant shift in the proportion of CD4+T to CD8+T cells (significantly lower CD8+ and higher CD4+T cells). Surprisingly, we observed that this increase in frequency of IEL CD4+T cells was chiefly in the IEL CD103− CD4+T RM subset but not in the CD103+CD4+T RM subset. These data suggest that there may be other receptor-ligand pair(s) (apart from CD103/E-cadherin) involved in the migration of CD103− CD4+T RM from LP to the epithelial compartment. Alternatively, CD103− CD4+T RM cells may be more likely to move freely around the epithelium since it does not depend on their interaction with the E-cadherin on epithelial cells. Regarding S. Typhi-specific cytokine production, we found that oral Ty21a-immunization elicits significant IEL CD103+ CD4+ T RM mainly producing IL-2-S. In contrast, IEL CD103− CD4+T RM exhibited significant increases in IFNγS, IL-17A-S and TNFα-S following stimulation with S. Typhi-infected targets in Ty21a-vaccinated than unvaccinated volunteers. These antigen-specific responses showed diverse characteristics in the epithelium, suggesting that migrating and/or locally differentiated CD103− CD4+T RM are likely Th1 and Th17 which might play a significant role in S. Typhi protective immunity. In contrast, CD103+ CD4+T RM producing IL-2 might be important in supporting T cells proliferation and likely supporting T regulatory cells in this compartment. Taken together, these data suggest that following oral Ty21a immunization, both CD4+T RM effector subsets are recruited from the LP and activated/ differentiated uniquely in the epithelium, contributing distinct effector responses involved in effective S. Typhi immunity. This compartmentalization of T RM responses in the TI LPMC and IEL may provide important clues for the optimal design of future vaccines.

Conclusion
In summary, we provide the first evidence of the induction of increased spontaneous and S. Typhi-specific cytokine production by CD4 + T RM in the human TI LP and epithelial compartments following oral Ty21a-immunization. These results contribute novel insights in our understanding of the generation of gut local immunity in humans following immunization with oral attenuated bacteria and suggest that CD4 + T RM play a key role in protection following immunization and/or infection with S. Typhi.