Activation, differentiation, and survival of T cells are controlled by events that take place during direct cell-cell interactions with antigen-presenting cells (APC). Our laboratory is investigating the membrane bound molecules that are critical for these events (Croft, 2003, Nat Rev Immunol; Croft, 2003, Cytokine and Growth Factor Rev).

In order to elicit an effective immune response, T cells need to become activated and divide extensively, thereby allowing sufficient numbers to accumulate to overcome the particular infection. Upon encounter with the antigen, expressed as a peptide bound to MHC molecules, a number of events need to take place in order to stimulate the T cell efficiently to allow it to expand over the next 3-4 days and then survive for weeks and even months. During the past few years we have been working to understand which molecular interactions control different phases of the T cell response.

Costimulation and the Early T Cell Response

The response of T cells can be controlled by costimulatory molecules present on APC that interact with co-receptors present on the T cell. The ligand-receptor pairs which may determine the fate of the T cell include B7:CD28, ICAM-1:LFA-1, 4-1BB-L:4-1BB, OX40-L:OX40, and CD40:CD40-L (Croft and Dubey, 1997, Crit Rev Immunol). The initial activation of a naïve T cell is controlled by recognition of peptide/MHC complexes and does not appear to require any other interactions. This was shown in studies where activation of naïve T cells was analyzed with peptide presented on surrogate fibroblast APCs that either expressed B7 and/or ICAM, or were deficient in these molecules. Blastogenesis, entry into cell cycle, and upregulation of many surface molecules (e.g. CD25, CD69) occur equivalently regardless of the presence of costimulation, and expression of only two molecules, namely membrane-bound CD40L and secreted IL-2, critically require costimulatory signals (Jaiswal, 1996, Int Immunol; Croft, 1997, J Immunol). Functionally, T cells stimulated by antigen but without costimulatory signals, are able to proliferate and expand in numbers, but only transiently, with proliferation being very short-lived and few T cells being able to survive over time (Rogers, 1998, J Immunol). Additionally, the small number of T cells that are generated are unresponsive to antigen, entering into a state of tolerance (Croft, 1997, J Immunol). Our older data had shown that the molecules B7 and ICAM could costimulate naïve T cell activation when expressed in isolation, but an efficient response, including high levels of CD40L, IL-2, and short-term proliferation, resulted only when both B7 and ICAM were able to signal through CD28 and LFA-1 at the same time (Dubey, 1995, J Immunol; Jaiswal, 1996, Int Immunol; Rogers, 1998, J Immunol). Collectively, this demonstrates that the T cell membrane proteins CD28, LFA-1, and CD40L and the secreted protein, IL-2, are crucial molecules that may determine the short-term fate of T cells.

Costimulation and the Late T cell Response

The hallmark of an effective T cell response is the formation of a stable long-lived population of cells that mediate immune memory. Although the interactions of ICAM/LFA, CD40/CD40L, and B7/CD28 explain how the early T cell response is driven, these interactions are not sufficient for promoting effective memory. Our present studies are focusing on the molecular interactions which prevent excessive T cell death after the initial period of clonal expansion and which are essential for promoting late T cell division and providing long-term survival signals. It is likely that again these events are mediated by a number of different cell bound molecules. Candidates are several members of the tumor necrosis factor receptor (TNFR) family, and our laboratory is focusing on a number of these molecules.

OX40 (CD134)

The TNFR homologue OX40 is expressed on CD4 and CD8 T cells, but unlike CD28 and LFA-1, it is not constitutive with peak levels only seen 2-3 days after activation (Gramaglia, 1998, J Immunol). Its ligand, OX40L, has been visualized on the surface of activated APC such as dendritic cells and B cells and also peaks in expression 2-3 days after these cells are stimulated. This immediately suggests that OX40 action may be late in the T cell response, and our initial data using fibroblast APCs transfected with OX40L supported this with CD4 cells being induced to proliferate over a prolonged period after receiving OX40 signals (Gramaglia, 1998, J Immunol). Our laboratory has extended this work using both an agonist antibody to OX40 and in studies of OX40-deficient mice. By tracking antigen-specific CD4 T cells in vivo with flow cytometry, anti-OX40 was shown to be capable of promoting 5-15 fold more T cells to accumulate in a primary immune response and to result in a similar greater number surviving over several weeks as memory cells (Gramaglia, et al., 2000, J Immunol). In support of this, OX40-deficient mice generate 10-20 fold fewer T cells 5-7 days after primary immunization, and are greatly reduced in their ability to generate high numbers of memory cells over 5 weeks (Gramaglia, 2000, J Immunol). Using OX40 and CD28 knockout mice bred to TCR transgenic animals in which all CD4 cells are specific for a single antigen, we confirmed that CD28 controls early T cell expansion and proliferation with little participation from OX40. In contrast, OX40 is essential for allowing proliferation 4-6 days into the response and more significantly is essential for preventing the death of T cells at these late stages (Rogers et al, 2001, Immunity). We are focusing on defining the mechanism of action of OX40 in promoting long-term T cell survival. We demonstrated that OX40 signals can regulate the levels of the anti-apoptotic molecules such as Bcl-xL, Bfl-1, and Bcl-2, and that death of T cells brought about by a lack of OX40 signals can be reversed if these molecules are overexpressed (Rogers, 2001, Immunity). We have recently shown that this action is mediated by sustaining the activation of protein kinase B, also known as Akt (Song, 2004, Nat Immunology). We have also shown that TRAF5, an adaptor protein appears to antagonize OX40 signals, and the TRAF5 mice generate enhanced Th2 responses when OX40 is ligated (So, 2004, J Immunol). Our continuing studies are addressing whether OX40 targets other pro-survival molecules such as the inhibitor of apoptosis molecule survivin, and other intracellular signaling pathways that are involved in mediating OX40 effects on survival or Th2 differentiation such as those involving PKCθ, PI3k, NF-αB, and JNK.

4-1BB (CD137)

4-1BB is another TNFR family member that may function similarly to OX40. It can also be expressed on activated CD4 and CD8 T cells and is induced with similar kinetics, peaking 2-3 days after antigen recognition (Gramaglia, 2000, Eur J Immunol). Additionally, 4-1BBL is also only expressed on activated APC such as dendritic cells and B cells. Using surrogate fibroblast APCs expressing 4-1BBL, our studies have shown that it can synergize with B7 in promoting IL-2 production and proliferation of naïve CD4 cells and additionally that 4-1BB signals can suppress T cell death (Gramaglia, 2000, Eur J Immunol). Recent studies have confirmed that 4-1BB signals function similarly to OX40 and also regulate the anti-apoptotic members of the Bcl-2 family. Several reports have demonstrated that agonist antibodies to 4-1BB can promote CD8 T cell responses, and it has been suggested that 4-1BB may be more essential for CD8 cells than for CD4 cells. Our laboratory has to date found very similar effects of 4-1BB signals on CD8 T cells. Blocking 4-1BB strongly inhibits the generation of high numbers of CD8 cells in a primary immune response, and this is associated with increased T cell death, again suggesting that this molecule is highly analogous to OX40 (Cooper, 2002, Eur J Immunol). We are currently studying 4-1BB-deficient mice and generating TCR transgenic 4-1BB-deficient CD4 and CD8 T cells in order to more closely define the contribution of this molecule to the response of these T cells.

OX40 and 4-1BB and Tolerance

A number of costimulatory receptors have been targeted with great success in attempts to prevent the induction of T cell tolerance, including CD28 and CD40. However, being able to break established tolerance has proved more difficult. To date, our studies have focussed on OX40 using a system whereby TCR transgenic CD4 cells can be tracked through flow cytometry and assessed at the single cell level. We have shown that OX40 signals, provided at a time after unresponsiveness has been induced, are capable of reversing the tolerant state, allowing CD4 T cells to expand normally and to regain their functionality (Bansal et al, 2001, Nat Med). More recently, we have also shown that OX40 signals can also antagonize tolerance induction of CD8 cells and that an agonist 4-1BB antibody can also deliver similar signals (Bansal-Pakala, 2002, J Immunol). Our current studies are aimed at determining the mechanism of action of OX40 at the molecular level, and how expression of OX40 and OX40L may dictate the ability to overcome tolerance to tumors. In this regard, we have recently shown that OX40-deficient CD8 cells do not undergo efficient clonal expansion in vivo, due to extensive apoptosis, and that in a response to an injected thymoma, agonist anti-OX40 can completely protect mice from growth of this tumor (Bansal-Pakala, 2004, J Immunol).

OX40 and 4-1BB in Disease

Our studies of disease have focused on lung inflammation. We have shown that OX40-deficient animals develop only a mild inflammatory response in the lungs in an experimental model of asthma that involves Th2 responses (Jember et al, 2001, J Exp Med). More recently, our studies have defined a critical contribution of OX40 to the response of memory T cells that mediate this disease, and have shown that OX40 signals are essential for long-term survival of memory T cells, similar to their role on naïve T cells (Salek-Ardakani, 2003, J Exp Med). Other current work is focusing on a role for 4-1BB in these responses, and whether OX40 or 4-1BB are also crucial to lung inflammation that is caused by Th1 cytokines, either mediated through CD4 or CD8 cells.