Thomas C. Mitchell, Ph.D.

Thomas C. Mitchell, Ph.D.


My laboratory studies the molecular mechanisms by which T cells are signaled to die or to survive during protective immune responses. Survival of human activated T cells is a critical determinant of successful vaccination against infectious diseases and tumors, and of T cell-dependent autoimmune disorders such as Type I diabetes. We perform experiments designed to understand how the survival of activated T cells from mice is controlled, so that decisive factors can be targeted for eventual manipulation in humans. In the mouse model, injection of purified protein antigens induces "abortive" T cell activation characterized by rapid proliferation and then rapid death of the stimulated cells. In contrast, giving purified antigens along with immunological adjuvants (as is done in routine vaccinations) produces T cell responses that are "productive" in the sense that many more of the cells survive in order to participate in immune defense. My laboratory uses normal and genetically modified mice, model antigens and adjuvants, gene expression microarrays, and retroviral gene transfer to identify which genes are the most critical in a T cell's decision to die or survive. By understanding exactly how this process occurs in healthy immune responses, we can learn exactly how it is impaired in autoimmune diseases.



Little is known about the mechanism by which immunological adjuvants increase the survival and therefore numbers of immunologically useful T cells. Some studies have shown that cells die by caspase action after engagement of Fas or of the receptor for tumor necrosis factor (TNF). However T cell death in our experiments is only to a small degree dependent on Fas or TNF signaling, indicating that adjuvants do not act primarily by interfering with Fas or TNF-induced death. Members of the Bcl-2 family of proteins are often involved in the life and death of cells. The increased survival of activated T cells could be caused either by increased levels of anti-apoptotic proteins such as Bcl-2 and Bcl-xL or decreased levels of pro-apoptotic proteins such as Bim, Bad, and Bax. However, our published and preliminary experiments have shown that levels of these proteins are not affected by exposure to adjuvants.

With the failure of tests of these and other obvious candidates, we decided to compare gene expression in activated T cells that had or had not been exposed to adjuvants. In order to find broadly important factors, we searched for genes that were similarly affected by exposure to two different kinds of adjuvant, bacterial or viral.

Affymetrix gene microarray technology was used to survey thousands of genes in T cells that had been activated for an abortive vs. a productive response. Of several genes identified as having increased as a result of adjuvant exposure was Bcl-3, a member of the NF-kB/IkB family or transcription factors. Retroviral gene transfer experiments showed that Bcl-3 expression increased the survival of activated T cells both in vitro and in vivo. We concluded that adjuvants increase T cell life expectancy at least in part by raising levels of Bcl-3 in the cells, which may operate by altering the balance of NF-kB/Rel transcriptional activities during immune responses.

Our next steps are to learn how Bcl-3 promotes the survival of T cells. Bcl-3 is little studied, in comparison to other members of the NF-kB/IkB family, and has attributes of both the inhibitory members of the family and of the transcriptionally activating members of the family. Answering the deceptively simple question, "Does Bcl-3 protect T cells by inducing new gene expression, or by inhibiting gene expression" will be a major goal of the laboratory in the year to come.


Vaccination with recombinant proteins is the safest way to immunize large numbers of people against biological weapons and infectious disease. However, the low rates of inflammatory side effects that make subunit vaccines desirable for large-scale immunization may also be responsible for their relatively poor efficacies (when compared to ‘live-attenuated' vaccines). This is so because inflammatory effects are increasingly recognized as being beneficial for the establishment of immunity. In many cases inflammation is attributable to innate recognition of molecular patterns that are specific to infectious organisms, which leads to secondary signaling from antigen-presenting cells to guide adaptive T cell responses. We believe the beneficial aspects of microbial pattern recognition can be separated from pro-inflammatory effects, and that 'pattern recognition' adjuvants can be used safely in subunit vaccines. GlaxoSmithKline has developed several chemical mimetics of the biologically active portion of LPS, lipid A, some of which have low toxicity and promising immunogenic activities. It is expected that some of these compounds will become useful as adjuvants for vaccines against biologicial weapons.


The abortive response of T cells that follows activation with purified protein antigens appears to play an important role in preventing autoimmunity. If a T cell with reactivity for self-antigens escapes negative selection during its development in the thymus, and encounters the antigen for which it is specific, it undergoes rapid clonal expansion. However, this cellular proliferation is usually not harmful to the animal in which it occurs because this abortive response is associated with loss of the proliferating cells due to cell death. So, the self-reactive cells are not only prevented from attacking self-tissues in the short-term, they are also deleted permanently from the repertoire of T cells that are in circulation at any given time. Hence, the abortive responses that occur in healthy tissues by self-reactive T cells continually reinforce the tolerance of self that is needed to prevent autoimmunity.

In mice that are prone to autoimmune attack by self-reactive T cells, such as in NOD mice that develop type I diabetes, some aspect of the process of self-tolerance has gone wrong. My laboratory will test the hypothesis that activated T cells from NOD mice fail to undergo fully abortive (tolerogenic) responses in either, or both, of two ways. 1) Because their intrinsic mechanisms of death induction are impaired, or 2) because the adjuvant induced pathways responsible for productive T cell responses in normal mice are overactive. Having identified Bcl-3 as a molecule implicated in keeping normal T cells alive during productive T cell responses, gives us one candidate whose activity can be tested in NOD mice. Other candidates will be tested as they are identified as being required for normal immune responses to occur. Eventually, identification of molecules and signaling pathways that are required for autoimmunity to occur will give the pharmaceutical industry new targets for discovery of drugs that can be used to cure Type I diabetes and other autoimmune diseases.

This program is supported by funds provided by the NIH, Barnstable-Brown Endowment for Diabetes Research, the Commonwealth of Kentucky Research Challenge Trust Fund, and the Kentucky Lung Cancer Research Program.

Selected Publications

Mata-Haro V, Cekic C, Martin M, Chilton PM, Casella CR, and Mitchell TC.The Vaccine Adjuvant Monophosphoryl Lipid A as a Trif-biased Agonist of TLR4.Science, 2007; 316(5831):1628-1632. Perspective on this paper published in Science, 2007; 316(5831):1574-1576.

Sengupta S, Jayaraman P, Chilton PM, Casella CR and Mitchell TC. Unrestrained glycogen synthase kinase-3beta activity leads to activated T cell death and can be inhibited by natural adjuvant.The Journal of Immunology, 2007; 178(10):6083-91.8(10):6083-91, 2007.

Chilton PM and Mitchell TC.CD8 T cells require Bcl-3 for maximal gamma interferon production upon secondary exposure to antigen.Infection and Immunity, 2006; 74(7):4180-9.

Chilton PM, Jayaraman P, Cai M, Mitchell TC.Adjuvant-mediated effects on activated T cell survival and the divergent roles played by Bcl-3.Research Signpost 2006, Vella A.T., editor.

Sengupta S, Chilton PM, and Mitchell TC.Adjuvant-induced survival signaling in clonally expanded T cells is associated with transient increases in pAKT levels and sustained uptake of glucose.Immunobiology, 2005; 210(9):647-59.