The In Vivo Response of Naive CD4⁺ T Cells

It is customary for the president of the American Association of Immunologists to publish a paper in the Journal of Immunology related to one’s Presidential Address at the annual meeting. One purpose of an article like this may be to give readers a sense of the scientific career path taken by an American Association of Immunologists president. So, at the risk of being presumptuous, here is my story in brief.

I have spent my whole career trying to understand how CD4⁺ T cells, the ones that use α–β TCRs to recognize peptides bound to MHC class II (MHCII) molecules (1), respond during the immune response. As a graduate student with Stephen Miller at Northwestern University, I participated in a study showing that CD4⁺ T cells play a critical role in the delayed-type hypersensitivity reaction to protein Ags (2). During this experience, I realized that ear-swelling assays were probably only going to take us so far, so I looked for a postdoctoral position in a laboratory doing more molecular immunology.

I landed a position with Ronald Schwartz in the Laboratory of Immunology at the National Institutes of Health. Ron was doing pioneering work on the fine specificity of peptide recognition by CD4⁺ T cells using cytochrome c as a model Ag (3). It had been shown that protein Ag-pulsed splenocytes treated with the chemical 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (ECDI) induced Ag-specific immune tolerance when injected before immunization (4). My project in the Schwartz laboratory ended up focused on how ECDI inhibited the activation of pigeon cytochrome c peptide–specific Th1 clones in vitro. I found that peptide-pulsed, ECDI-treated splenocytes did not cause Th1 clones to proliferate (5). In fact, the Th1 cells became anergic, as evidenced by an inability to produce IL-2 or proliferate maximally in response to peptide Ag and viable splenocytes (5, 6). As described in Jeffrey Bluestone’s Pillars of Immunology article (7), this work, along with a similar study by Helen Quill in the Schwartz laboratory with purified peptide:MHCII complexes in planar membranes (8), provided clear experimental support for two early signal models (9, 10) positing that AgR signaling was necessary for productive lymphocyte activation but only in the context of a costimulatory signal. It may have also set the stage for later work, some from my laboratory at the University of Minnesota (11), showing that the costimulatory signal was transduced by CD28 (12–16). It is satisfying to see that CD28 blockade is now used to treat rheumatoid arthritis (17).

Although my postdoctoral experience taught me the power of reductionist in vitro studies, in my own laboratory at the University of Minnesota, I was anxious to develop an in vivo approach for the study of peptide:MHCII-specific CD4⁺ T cells to bolster the physiological relevance of the work. It seemed important for the approach to have the sensitivity to track the T cells in their naive state and after they became effector and memory cells during an immune response to the relevant Ag. TCR-transgenic mice had great potential in this regard because they contained an abundant supply of T cells with a known TCR specificity and could, in theory, get around the limitation that naive CD4⁺ T cells specific for any given peptide:MHCII epitope were likely to be exceedingly rare in a normal polyclonal repertoire. Surprisingly, we found that very few of the OVA peptide:I-A^d–specific CD4⁺ T cells in DO11.10 TCR-transgenic mice (18) showed signs of activation after injection of the OVA peptide (19). After scratching our heads for a long time about this odd phenomenon, we considered the idea that CD4⁺ T cells with the identical TCR were so abundant in the intact DO11.10 mouse that they experienced inefficient activation because of competition for OVA peptide:I-A^d complexes. To test this idea, we reduced the frequency of naive DO11.10 T cells by injection into normal BALB/c mice, such that ∼10⁵ of the transferred cells established residence in the secondary lymphoid organs of the recipient. The transferred T cells could easily be distinguished from the recipient’s T cells by flow cytometry using an anti-clonotypic Ab specific for the DO11.10 TCR produced by Marrack, Kappler, and colleagues (20). Remarkably, injection of the OVA peptide into the adoptive recipients caused the transferred naive DO11.10 T cells to undergo a marked clonal expansion and convert to the memory cell phenotype (19). We used this simple system to publish a series of “visualization” papers documenting the effects of adjuvants on T cell activation and differentiation (19, 21, 22), early interactions between CD4⁺ T cells and dendritic cells (23, 24), later interactions between CD4⁺ T cells and Ag-specific B cells (19, 25), and migration of effector cells to nonlymphoid organs at the whole-body level (26). This system is still used widely today to study the in vivo T cell response, as described in Jonathan Sprent’s Pillars of Immunology article about this work (27).

However, the TCR-transgenic adoptive-transfer system still suffered from the clonal competition problem, just to a lesser degree than intact TCR-transgenic mice. Lefrançois’ group (28) and then my group (29) and Harty’s group (30) showed that adoptively transferred TCR-transgenic T cells differentiated inefficiently when exposed to the relevant epitope in vivo. This problem was related to the frequency of the transferred cells, because reducing it in the adoptive recipients increased the efficiency of activation. Thus, although adoptive transfer of TCR-transgenic T cells is a useful tool for the study of in vivo T cell activation, the aforementioned studies taught us that the fewest possible cells should be transferred for the best results.

The issues related to the high-dose TCR-transgenic T cell adoptive-transfer system spurred us to search for a way to finally identify peptide:MHCII-specific T cells in normal polyclonal repertoires. We did so by building on the monumental discovery that peptide:MHC tetramers bind T cells with specific TCRs (31) and reports that these reagents could be used with magnetic bead–based cell techniques and flow cytometry to enrich rare specific T cells (32–38). Using peptide:MHCII tetramers in cell-enrichment mode, we detected naive peptide:MHCII-specific CD4⁺ T cells in normal mice (39) and studied the naive to effector to memory cell transition within polyclonal repertoires (40, 41). It is gratifying to see that tetramer-based cell enrichment is now being used to study the human preimmune CD4⁺ T cell repertoire (42, 43).

In a recent chapter of this continuing story, the sensitivity of the tetramer-based cell-enrichment method brought us back to the adoptive-transfer approach to study the naive to effector cell transition but this time by transferring single naive CD4⁺ T cells (44). I never dreamed as I was measuring swollen mouse ears and culturing T cells in 96-well plates that someday my colleagues would be tracking the fates of single epitope-specific T cells during an immune response. What a thrill.