Publication Date




Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Biochemistry and Molecular Biology (Medicine)

Date of Defense


First Committee Member

Zafar Nawaz

Second Committee Member

Sylvia Daunert

Third Committee Member

Sapna Deo

Fourth Committee Member

Michal Toborek

Fifth Committee Member

Thomas Malek


CD8 T cell exhaustion is the progressive, step-wise loss of T cell function due to persistent antigen stimulation. Given the importance of this phenomenon in sustaining chronic infections and preventing clearance of tumors, as well as potentially influencing autoimmune disorders, novel biomarkers must be established to monitor CD8 T cell exhaustion and the factors that influence its development (e.g. presence of CD4 helper cytokines, like IL-21). In this study, a six-day in vitro model of CD8 T cell exhaustion was characterized to evaluate the reduced functional capacity of these cells. Cells derived in this model possessed sustained inhibitory receptor expression (13.8% PD-1+ and 36.9% Tim-3+ at day 6), a reduced expression of effector cytokines (IL-2, TNFa and IFNg reduced at day 6 compared to day 3, p = 0.0117, p = 0.0247 and p = 0.0108, respectively), less secretion of the cytotoxic protein GZMB compared to day 3 (p = 0.0114), a decreased metabolic potential (increase in mitochondrial depolarization at day 6 compared to day 3, p = 0.0401), and a inhibited capacity to proliferate upon restimulation (46.1% of cells proliferating at day 4 compared to 10.5% proliferating at day 4 following restimulation, p = 0.0294). Further, statistical analysis of the differentially expressed genes in day 6 PD-1+ cells revealed a significant alignment with published gene sets for CD8 T cell exhaustion and tumor infiltrating lymphocytes in cancer (q values of 7.80e-12 and 1.83e-11, respectively). This data indicated that cells derived from this model reflected canonical CD8 T cell exhaustion, particularly within the PD-1+ subset at day 6. The interrogation of exhausted CD8 T cells from this model, allowed for the identification of 161 potential exhaustion-specific genes suitable for biomarker development that were elevated in the day 6 PD-1+ cells, but not at day 3, when the cultured cells were activated. IL-21, a CD4 helper cytokine speculated to help prevent CD8 T cell exhaustion, was added to the culture and shown to not only prevent the development of metabolic and proliferative dysfunction (p < 0.001 increase in basal OCR and ATP production at day 6), but also lower the expression of candidate exhaustion genes. Using this cytokine as a tool, 10 genes were identified by their elevated expression in exhausted day 6 PD-1+ cells that also modulated significantly down when exhaustion was prevented. This signature was confirmed by both a significant increase in the geomean of their relative expression values in Bat3 knockdown cells, which demonstrates enhanced exhaustion through engagement of the Tim-3 pathway (p = 0.0197 three days after gene knockdown), and by being significantly increased in renal cell carcinoma and colorectal patient TILs compared to whole blood CD8 T cells (p = 0.0157). Lastly, since IL-21 was shown to be able to prevent the development of exhaustion in vitro, new, ultrasensitive assays were established to detect accurate levels in autoimmune patients. Analysis of current ELISA methods to quantitate IL-21 were determined to be inaccurate, therefore two novel assays, a Quanterix Single Molecule Array (SiMoA) and S-PLEX® assay from Mesoscale Development (MSD) were created, characterized and compared. Following method optimization, each method was able to detect sub-pg/mL levels of IL-21 (predicted Limit of Detection [LOD] of approximately 1.0 fg/mL for both the Quanterix SiMoA and MSD S-PLEX® platforms) which is 200-500 times lower than current commercial assays. Additionally, it was demonstrated that rheumatoid factor did not interfere with measuring IL-21 in the Quanterix SiMoA assay. Results obtained with the two new ultrasensitive assays showed a strong correlation (r = 0.9428; p < 0.0001). Additionally, IL-21 levels were significantly increased in samples from patients with Systemic Lupus Erythematosus (mean+/- SD: n = 14, 202.64 +/- 111.47 fg/mL, p = 0.0001 for Quanterix SiMoA and 275.4 +/- 174.66 fg/mL p = 0.0001 for MSD S-PLEX®) as well as in samples from patients with Sjögren’s Syndrome (mean+/- SD: n = 11, 122.18 +/- 84.50 fg/mL , p = 0.0029 for Quanterix SiMoA and 183.64 +/- 153.00 fg/mL, p = 0.0082 for MSD S-PLEX®) when compared to healthy donors (mean+/- SD: n = 11, 38.1 +/- 27.8 fg/mL for Quanterix SiMoA and 58.1 +/- 30.7 fg/mL for MSD S-PLEX®). The outcome of this thesis provides two novel biomarkers: a gene signature for CD8 T cell exhaustion and two ultrasensitive assays for IL-21. Though more work must be conducted to evaluate any connections between CD8 T cell exhaustion in autoimmune or cancer patients with circulating IL-21 values, these tools now exist to complete the analysis. Eventually, these biomarkers may provide value for translational medicine by identifying patients who will benefit from checkpoint inhibitor therapies in oncology or IL-21 modulating therapies in autoimmunity, both helpful in targeting the right patients at the right time.


CD8 T cell exhaustion; IL-21; biomarkers

Available for download on Saturday, November 20, 2021