Models aren’t just pretty

Science Spotlight

Models aren’t just pretty

From the Schiffer lab, Vaccine and Infectious Disease Division

Sept. 17, 2018

Chronic Herpes simplex virus 2 (HSV-2) infection causes prolonged viral shedding in the genital track with episodes of reactivation leading to lesion formation. These local reactivation events vary in duration, severity and are controlled by tissue resident T cells (Trm). Trm are memory CD8+ T cells that are located in tissue near sites of infection. In the case of HSV-2, Trm are located in the mucosa where limited migration puts them in contact with a site of infection. Once Trm encounter infected cells, the Trm proliferate and produce INF-gamma and other cytokines to initiate the host immune response. In order to better understand this interaction in human tissue, researchers lead by Dr. Schiffer, at Fred Hutch (Vaccine and Infectious Disease Division) created a computational model of HSV-2 shedding and Trm activity. Their findings are published in the Journal of Immunology.

Researchers created a stochastic mathematical model to mimic genital mucosa. The mucosa is broken up into 200 regions that are immunologically independent and also allow for concurrent reactivation and HSV-2 release from neurons and infected cells. As the viral antigen concentration increases, the Trm cells expand and then contract upon viral control. This model has achieved good fit and reproducibility of natural HSV-2 shedding, kinetics and viral end titer. In general, the model predicts almost uninterrupted ganglia shedding in order to achieve the high level of observed shedding and spread seen in patients. It also predicts that the number of infected cells and peak viral load is determined by Trm density at the site of infection. However the model does have its limitations, which include the assumption that Trm are the sole providers of immune control. This excludes the effects from CD4+ T cells, NK cell, neutrophils and antibody-dependent cellular toxicity, however the modeled Trm effect probably captures impacts of other innate cells.

Once the model was set, the group looked at Trm number and behavior over a five-year simulation with daily sampling. This method produced results that showed increased Trm numbers during shedding events but a relatively fixed range over time. There is spatial heterogenicity throughout the tissue of Trm with density re-assortment over time contributing in the relative stable number of Trm seen across the five-year simulation. This suggests that not all regions are covered by Trm at all times, thus allowing new symptomatic events, as seen in patients. In order for Trm to expand after antigen encounter, two factors need to be true: there needs to be seeding by HSV-2 and a low starting Trm density. This allows for proliferation of Trm in some regions and not in others, thus contributing to the overall stable density of cells. Some conditions that affect these two factors include, random nature of HSV-2 seeding events and the time between local infections, which effects the starting concentration of Trm at the local location. To further characterize the role of Trm in the mucosa the group modeled the effect on viral outcome when Trm were allowed to traffic between sites of infection. This however led to a poorer fit of the model, suggesting that once Trm find antigens, the cells enter a less mobile stage.

Top picture is a stained human genital mucosa biopsy. Red staining represents CD8 T cells within the mucosal tissue. Bottom image shows the make-up of the tissues microregions with each color representing a single microregion (100 endothelial cells) and + representing the center of the region. The center image represents CD8 T cells (triangles) present in the tissue and color is matched to microregion below. This depiction shows the heterogeneity of the Trm populations within mucosal tissue.

Figure provided by Dr. Schiffer.

To further validate their model, the group looked at paired human biopsies from HSV-2 infected patients. The biopsies were analyzed by division into microregions and epithelial and Trm cells were counted (see figure). The biopsies show a clustering of high density CD8+ T cells along with regions of low density. This results in overall heterogeneity of Trm similar to the model. Lastly the group plugged treatment or vaccination into their model. They showed that treatment with antivirals, which reduced rate of viral release and replication, did in fact decrease replication but also decreased Trm proliferation. This allowed for viral breakthrough once treatment was stopped. Simulated vaccination, which increased trafficking of Trm to sites of viral replication did not curb infection, however increases in Trm proliferation due to vaccination did. Dr. Schiffer points out that, “It is clear that enough CD8+ T cells are elicited during chronic HSV-2 infection to eliminate each episode of shedding.  However, not enough CD8+ T cells are generated to protect all spatial regions against more severe, clinically relevant reactivations.  A therapeutic vaccine which aims to limit the severity of this lifelong infection, must increase the levels of CD8+ T cells in previous low-density regions.” In conclusion, Dr. Schiffer states, “In this paper, we demonstrated that HSV-2 induces a fixed spatial structure of tissue-resident CD8+ T cells across the entire genital tract during chronic infection. This spatial distribution is characterized by a high density of CD8+ T cells in certain regions, allowing elimination of HSV-2 before the virus spreads to more than a few cells. However, other regions have a very low density of CD8+ T cells allowing infection of thousands of infected cells, and development of painful ulcers, before all virally infected cells are eliminated.  We used a mathematical model to demonstrate how this spatial structure across the entire at-risk tissue is maintained over decades, despite the fact that individual micro-regions have massive increases the number of CD8+ T cells during local HSV-2 replication events.”


Schiffer JT, Swan DA, Roychoudhury P, Lund JM, Prlic M, Zhu J, Wald A, Corey L. 2018. A Fixed Spatial Structure of CD8(+) T Cells in Tissue during Chronic HSV-2 Infection. J Immunol. 201:1522-1535.

This work was supported by the National Institutes of Health.

 Fred Hutch/UW Cancer Consortium faculty members Jenny Lund (Fred Hutch), Martin Prlic (Fred Hutch), Larry Corey (Fred Hutch), Joshua Schiffer (Fred Hutch), Jia Zhu (UW) and Anna Wald (UW) contributed to this research.