Tissue localization shapes antiviral T cell dynamics

From the Prlic and Lund labs, Vaccine and Infectious Disease Division

T cells comprise a critical arm of the immune system during infection and immunization. Tissue-resident memory T cells (Trm), which reside in specific tissue sites and do not recirculate into the blood and lymph, are indispensable for generating protection from infection after immunization. Trm are especially important in mucosal barrier sites: these tissues, such as the female cervicovaginal tract (CVT), are the site of entry for many pathogens and sexually transmitted infections (STIs) and must rapidly clear infections at first contact. Previous work has shown that Trm in the CVT are crucial for viral control of infections such as herpes simplex virus 2 (HSV-2); therefore, efforts to design a vaccine that would elicit protective Trm in the vaginal mucosa are ongoing. However, Trm dynamics are highly tissue-dependent, and much about the CD8 Trm dynamics in the CVT is currently not well understood.

“CD8 T cells are a type of immune cell that can provide meaningful immunity against sexually transmitted viruses such as HIV and herpes simplex virus,” explained Dr. Veronica Davé, a former graduate student in the Lund and Prlic labs (Vaccine and Infectious Disease Division) and lead author on a recent publication in the Journal of Immunology. “In this project, we asked how long CD8 T cells persist in the vagina after an immunization and whether they could protect against vaginal infection with herpes simplex virus. To answer this question, we used a CD8 T cell-targeting immunization to prime herpes simplex virus-specific CD8 responses in the vaginal tissue of mice,” Dr. Davé said.  

To begin the project, the authors profiled CD8 T cells derived from vaginal biopsies collected from healthy women. Using flow cytometry, they found that many CVT CD8 T cells expressed CD69 and CD103—surface markers associated with tissue residency— but did not express T cell factor 1 (TCF-1), a transcription factor associated with self-renewal in CD8 T cells. A large fraction of this population also expressed Granzyme B (Gzmb), a cytotoxic molecule used by CD8 T cells to kill infected cells. These findings suggested that the CVT hosts a population of tissue-resident cytotoxic CD8 T cells that have unique properties, prompting the authors to ask if this phenotype was a result of ongoing antigenic stimulation or instead reflected a Trm program driven by CVT tissue residency alone.

To interrogate the determinants of the vaginal CD8 phenotype, the project moved to a mouse model, where infection and exposure to antigen could be tightly controlled. To create a population of CD8 T cells comparable to those seen in human CVT, wildtype mice received an adoptive transfer of HSV-2-specific transgenic CD8 T cells, followed by a systemic immunization with Listeria monocytogenes expressing an HSV-2 immuno-dominant epitope gB (Lm-gB), which exclusively primes HSV-2-specific CD8+ T cells. One month post-infection, CD8 T cells specific for HSV-2 were found in distal sites such as the CVT and vaginal-draining lymph node, and their phenotype mirrored human CVT CD8 T cells. Notably, immunization with Lm-gB does not induce vaginal inflammation, demonstrating that CVT tissue residency itself, and not baseline antigenic stimulation, is sufficient to drive the unique phenotype seen in CVT CD8 Trm. Additionally, CD8 T cells in other regions of the reproductive tract lacked CD103 expression and were TCF+, suggesting that residency in the vagina, specifically, confers a distinct CD8 Trm phenotype.

Microscopy image of human vaginal tissue stained for CD8 T cells (red), blood vessels (green), and nuclei stained with DAPI (blue).
Microscopy image of human vaginal tissue stained for CD8 T cells (red), blood vessels (green), and nuclei stained with DAPI (blue). Image by Veronica Davé.

After finding that a large fraction of CVT CD8 T cells lacked TCF-1 expression, the authors hypothesized that this population lacked proliferative self-renewal capacity and would undergo gradual decay. Using a cell proliferation assay one month post-immunization with Lm-gB, they found that TCF- CD8 T cells proliferated less than TCF+ cells and that HSV-2-specific CD8+ T cells in the CVT declined three-fold over a five-month period post-immunization. In contrast, HSV-2-specific CD8 T cells remained stable in other tissue and lymphoid locations. The CVT cell loss occurred primarily in CD103- CD8+ T cells, suggesting that the CD69+ CD103+ Trm population was maintained either through proliferation or conversion of CD103- to CD103+. Additionally, Gzmb expression on CD8 T cells was not present one-month post-Lm-gB immunization, but expression of Gzmb accumulated over time on CD103+ cells in the CVT. Conversely, Gzmb expression, or lack thereof, was stable in other tissue sites and lymph nodes. This suggests that CVT CD8 Trm differentiate asynchronously from other tissue Trm and that CD8 Trm numbers decline in the CVT over time, but those that do remain gradually acquire an effector phenotype armed to respond to viral reencounter.

Given that Trm are known to mediate protection from viral infection, the authors next asked how changes in the CVT CD8 T cell compartment affect immunity to HSV-2. By immunizing mice with Lm-gB and subsequently challenging with lethal, wildtype HSV-2 at either one-month (early) or four-months (late) post-immunization, they tested how well the corresponding early and late CD8 Trm compartments could control HSV-2 infection. Although mice from both groups survived similarly, demonstrating that antigen-specific CD8 Trm are sufficient to confer partial protection from HSV-2, the mice challenged earlier (those with higher numbers of HSV-2-specific CD8 Trm) were better able to clear HSV-2 virus from the vagina. Based on this findings, the authors to hypothesized that the phenotype and number of CD8 Trm present during early and late challenge were associated with better or worse viral control, respectively. To further understand the relationship between CVT HSV-2-specific Trm dynamics and viral clearance, they next fit mathematical models to empirical antigen-specific CD8 Trm numbers in mouse CVT to characterize CD8 Trm kinetics in early and late challenge. The models predicted that the mice challenged early after infection, when CD8 Trm numbers were highest, would clear virus earlier than those immunized after partial CD8 Trm compartment decay. These modeling predictions, together with the empirical challenge data, suggest that a temporal decrease in antigen-specific CD8 Trm in the CVT is associated with a diminished ability to control HSV-2.

“Together, these results suggest that immune cells in the female reproductive tract may have unique requirements in order to maintain long-term immune responses,” said Dr. Davé. “We monitored how long these cells were maintained in the vagina and found that they declined in number and underwent other changes in the six months following vaccination. Surprisingly, these changes were unique to the vagina and did not occur in other parts of the body, such as the spleen, lymph nodes, or small intestine. Some of these same characteristics were also observed in vaginal-resident CD8 T cells from women who underwent vaginal biopsies,” confirming that the mouse model accurately reflects human CVT immunology, Dr, Davé added. Additionally, this works highlights that a blood draw—which is often used to quantify T cell or antibody titers after vaccination—may not be representative of the cells resident in the tissue site of interest. These findings have implications for vaccine design, as a protective HSV-2 vaccine and subsequent booster vaccinations must elicit a population of CD8 Trm that are maintained in the CVT to quickly respond to virus at the point of entry. Going forward, “future studies will help us further understand the requirements for vaginal-resident T cell longevity. Our hope is that this knowledge could inform the development of new T cell-based vaccines for sexually transmitted infection,” said Dr. Davé.

Davé VA, Cardozo-Ojeda F, Mair F, Erickson J, Woodward-Davis AS, Koehne A, Soerens A, Czartoski J, Teague C, Potchen N, Oberle S, Zehn D, Schiffer JT, Lund JM, Prlic M. Cervicovaginal Tissue Residence Confers a Distinct Differentiation Program upon Memory CD8 T Cells. J Immunol. 2021 Jun 4;ji2100166. doi: 10.4049/jimmunol.2100166. 

UW/Fred Hutch Cancer Consortium members Martin Prlic and Josh Schiffer contributed to this work.

This work was supported by National Institute of Allergy and Infectious Diseases, the Doug and Maggie Walker Fellowship, and the National Institutes of Health.