Molecular analysis of the influences of positive selection, tolerance induction, and antigen presentation on the T cell receptor repertoire

Exposing T Cell Secrets Inside and Outside the Thymus

I've had serious misgivings about writing this article, because from living the experience day by day, it's hard to believe my accomplishments merit the attention. To skirt this roadblock, I forced myself to pretend I was in a conversation with my trainees, trying to distill the central driving forces of my career in science. The below chronicles my evolution from would-be astronaut/ballerina to budding developmental biologist to devoted T cell immunologist. It traces my work from a focus on intrathymic events that mold developing T cells into self-major histocompatibility complex (MHC)-restricted lymphocytes to extrathymic events that fine-tune the T cell receptor (TCR) repertoire and impose the finishing touches on T cell maturation. It is a story of a few personal attributes multiplied by generous mentors, good luck, hard work, perseverance, and knowing when to step down. Expected final online publication date for the Annual Review of Immunology, Volume 40 is April 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Development of CD1d-restricted NKT cells in the mouse thymus

Using genetic and phenotypic analyses, we have analyzed the developmental pathway of mouse CD1d-restricted invariant NKT cells. We provide strong evidence that similar to conventional T cells, positive selection of NKT cells occurs during a CD4(+)CD8(+) stage. Later stages of NKT cell development involved the down-regulation of both TCR and CD4 levels and therefore diverge from conventional T cell development pathways. A unique and complete dependency for development on Fyn, a Src family kinase member, also distinguishes the NKT cell and conventional T cell populations.

Oligoclonal expansion of CD45RO+ T lymphocytes in Omenn syndrome

Omenn syndrome comprises a rare form of combined immunodeficiency with TH2-type features of eosinophilia and elevated IgE. Previous studies have led to reports of restricted heterogeneity in the T lymphocyte repertoire, and in vitro cloned T lymphocytes have been shown to produce IL-4 and IL-5. We hypothesized that (1) T cell receptor beta V(D)J DNA sequence analysis would confirm and further define the putative restricted heterogeneity, and (2) increased production of IL-4 and IL-5 should be found in nonstimulated T lymphocytes, if the molecular pathogenesis of Omenn syndrome is an uncontrolled TH2 state. We report the results of molecular analyses of T lymphocytes from an untreated 3-month-old patient. Oligoclonal T cell receptor beta variable gene usage was found. Sequence analysis revealed sets of identical V(D)J sequences, each in-frame, with apparently normal N-diversification and no obvious antigen combining site motif. From fresh, nonstimulated lymphocytes, proinflammatory TH1 cytokines could be detected, but TH2 cytokines could not, so that a simple TH1/TH2 paradigm cannot explain the eosinophilia and elevated IgE in Omenn syndrome. Our studies fully document for the first time at the molecular level that clonally expanded populations of T lymphocytes are present in Omenn syndrome.

Positive selection of T cells induced by viral delivery of neopeptides to the thymus

The relation between an antigenic peptide that can stimulate a mature T cell and the natural peptide that promoted selection of this cell in the thymus is still unknown. An experimental system was devised to address this issue in vivo-mice expressing neopeptides in thymic stromal cells after adenovirus-mediated delivery of invariant chain-peptide fusion proteins. In this system, selection of T cells capable of responding to a given antigenic peptide could be promoted by the peptide itself, by closely related analogs lacking agonist and antagonist activity, or by ostensibly unrelated peptides. However, the precise repertoire of T cells selected was dictated by the particular neopeptide expressed.

Flexibility of the T cell receptor repertoire

Alternative T cell receptor (TcR) gene usage between mice of different Mls alleles has been demonstrated in a number of T cell responses. A clear illustration of a flexible TcR V beta usage in the same strain of mice remains to be established. Using a model system in which I-Ek-restricted T cells recognizing lambda repressor cI protein (cI) 12-26 and pigeon cytochrome c (pcc) 81-104 predominantly use V beta 3 in B10.A and B10.BR mice, and V beta 1 in Mls-2a-bearing A/J and C3H mice, we have first demonstrated that the hierarchy of TcR V beta usage can not be inferred from one strain of mice to the other. The presumed flexibility of V beta 3 to V beta 1 did not exist in B10.BR mice in the given responses. Instead, a switch of dominant TcR from V beta 1/V beta 3 to V beta 8 was identified in C3H and B10.BR mice. In contrast, there was an absolute rigidity in TcR repertoire usage in some mouse strains such as A/J. The lack of flexibility was not due to slow generating kinetics of replacing T cells; since A/J mice treated with staphylococcal enterotoxin A from birth on still responded poorly to cI 12-26 and pcc 81-104. Therefore, whether TcR V beta usage in a T cell response would be flexible or rigid is highly dependent on each strain of mice. However, even the plasticity seen in B10.BR mice is very limited and further tolerance of the V beta 8+ population results in non-responsiveness toward the given antigens.

The ligand for positive selection of T lymphocytes in the thymus

T cells are spared from programmed cell death in the thymus after an appropriate interaction between the T-cell receptor and a self peptide/MHC complex; this step is referred to as positive selection. Recent work has focused on precise identification of the positively selecting ligand, and the cell that presents it. First, it was shown that bone marrow derived cells or fibroblasts can substitute for epithelial cells in providing the ligand for positive selection. Second, in a T-cell receptor transgenic system, variants of the antigenic peptides were found to induce positive selection. Peptides that served as antagonists or weak agonists for the T-cell receptor efficiently selected immature thymocytes for survival. It appears that the peptide ligands for positive selection of T cells are self peptides, which serve as mimics or look alikes for the universe of pathogen peptides. The challenge remains to identify a naturally occurring thymic self peptide that can cause positive selection and determine the range of reactivities to foreign peptides which it can select.

Evidence for a superantigen in human tuberculosis

T cells are not only required for resistance to tuberculosis, but they likely contribute to the tissue damage characteristic of the disease. To define better the T cell populations that contribute to the immunopathogenesis of human tuberculosis, we investigated the T cell receptor (TCR) beta chain repertoire expressed in patients with tuberculous pleuritis. Analysis by polymerase chain reaction and flow cytometry indicated an expansion of V beta 8+ T cells at the site of disease in some donors, suggesting the possibility that Mycobacterium tuberculosis contains a superantigen. M. tuberculosis induced strong T cell proliferative responses in tuberculin-negative healthy donors in vitro, with preferential expansion of V beta 8+ T cells, independent of the CDR3 region. T cell stimulation was MHC class II-dependent and did not require antigen processing by the antigen-presenting cells. These findings are consistent with the presence of a superantigen in M. tuberculosis, aspects of which may contribute to the immunopathology of tuberculosis and to the adjuvant properties of M. tuberculosis.

Extrathymic development of V alpha 14-positive T cells

It is known that rearrangement of the T cell antigen receptor (TCR) gene occurs in the thymus during T cell development and consequently results both in the deletion of DNA between the variable (V) and diversity/joining segments and in the formation of a circular DNA with recombination signal sequences. Here, we provide evidence that V alpha 14+ TCR gene rearrangements take place in extrathymic sites, such as bone marrow, liver, and intestine, but not in spleen, because we were able to detect frequent productive and nonproductive V alpha 14+ coding and signal sequences as a result of TCR rearrangements in extrathymic sites. Similar findings were also detected in athymic mice. Quantitative analysis shows that the relative amounts of V alpha 14 gene-mediated signal sequences in extrathymic tissues are higher than those in thymus. On the contrary, TCR rearrangements of V alpha 1.1 T cells, which are known to develop in the thymus, were mainly detected in the thymus, Peyer's patch, and spleen, but not in other extrathymic tissues, showing patterns distinct from V alpha 14 TCR rearrangements. These findings are evidence of extrathymic development of V alpha 14+ T cells. Differential characteristic TCR rearrangement patterns also indicate that distinct TCR repertoires are generated in different lymphoid tissues.

Quantitative titration of nucleic acids by enzymatic amplification reactions run to saturation

In vitro enzymatic amplification of nucleic acids by PCR or other techniques is a very sensitive method to detect rare DNA segments. We present here a protocol that allows the rapid, sensitive and precise quantification of DNA molecules using PCR amplification run to saturation. The DNA (or cDNA) to be assayed is co-amplified with known amounts of an internal standard DNA. We show that the latter must be almost identical to the assayed DNA, otherwise quantification at the plateau is unreliable. The read-out of the amplification involves one or two additional oligonucleotides. Using fluorescent oligonucleotides as primers in run-off reactions together with an automated DNA sequencer, we could measure the level of expression of several genes, like the murine MHC class I H-2Kd or a specific T cell receptor beta chain transcript in the course of an immunization. mRNA levels were normalized by measuring in a similar manner the number of transcripts encoding the housekeeping gene HPRT. Finally, our procedure might allow the rapid analysis of a large number of samples at the same time, as illustrated by the simultaneous analysis of the mRNAs encoding the CD4 and CD8 murine T cell markers.

Characterization of T-cell responses to the house dust mite allergen Der p II in mice. Evidence for major and cryptic epitopes

Major histocompatibility complex (MHC) congenic strains can be defined as high and low responders to the major house dust mite allergen Der p II on the basis of the ability to sensitize T cells for in vitro lymphokine release. Mice of the H-2b haplotype were high responders, H-2k were intermediate and H-2d low responders. Like responses to other proteins, only a limited number of epitopes could be located by the response of T cells from mice immunized with allergen to a series of overlapping peptides. The epitopes for H-2b mice were 11-35, 78-104 and 105-129, 36-50 and 78-104 for H-2k mice and 36-60 for H-2d. Immunization with the peptides however revealed that spleen-adherent cells were required for lymph node cells to recall responses to the whole protein and in addition that mice could be sensitized by cryptic epitopes defined by peptides 22-50 and 1-20 for H-2b mice. Peptides containing these cryptic epitopes did not normally induce responses in mice primed with the allergen, but when they were used for immunizing they could prime mice for responses to the peptide and the whole allergen. The results both help to define a model for studying the presentation of allergens and have significant implications for peptide-based immunotherapy.

Molecular detection and in vivo analysis of the specific T cell response to a protein antigen

We have analyzed in detail the repertoire of transcripts encoding the V beta chains of the T cell receptor and investigated the T cell response of B10.A mice to pigeon cytochrome c. We were thus able to follow the specific T cell response in vivo after immunization with this protein antigen. The response is first detectable in the draining lymph nodes, then in the spleen and in the blood. It is qualitatively similar in individual animals. It is dominated by a major category of specific T cells harboring a V beta 3-J beta 1.2 rearrangement, and a limited and well-defined set of nucleotide sequences, previously found in several specific T cell hybridomas and clones. This predominance is observed from the onset of the immune response strongly suggesting the notion that there is no variation and, therefore, no maturation of the T cell response in the course of immunization.

Evidence for somatic selection of natural autoantibodies

Natural autoantibodies are primarily immunoglobulin M (IgM) antibodies that bind to a variety of self-antigens, including self-IgG. Accounting for a large proportion of the early B cell repertoire, such polyspecific autoantibodies are speculated to contribute to the homeostasis and/or competence of the primary humoral immune system. Recent studies indicate that the leukemia cells from most patients with chronic lymphocytic leukemia (CLL) also express such IgM autoantibodies. Similarly, the leukemia cells from many CLL patients react with murine monoclonal antibodies (mAbs) specific for crossreactive idiotypes (CRIs) associated with human IgM autoantibodies. In particular, leukemic cells frequently react with G6, a mAb specific for an Ig heavy chain (H chain)-associated CRI, and/or with 17.109, a mAb that defines a kappa light chain (L chain)-associated CRI. Generated against IgM rheumatoid factor (RF) paraproteins, G6 and 17.109 each recognize a major CRI that is present in many IgM RF paraproteins. Furthermore, over 90% of the IgM paraproteins found to bear both H and L chain-associated CRIs also are found to have RF activity. Molecular characterization of these CRIs demonstrates that each is a serologic marker for expression of a highly conserved Ig V gene. As such, the frequent production of IgM polyspecific autoantibodies in CLL simply may reflect the frequent use of such highly conserved autoantibody-encoding Ig V genes with little or no somatic mutation. To test this hypothesis, we generated murine transfectomas to pair the 17.109-reactive kappa L chain of SMI, a 17.109/G6-reactive CLL population, with the Ig H chain of SMI or other G6-reactive leukemia cells or tonsillar lymphocytes. Cotransfection of vectors encoding the Ig H and L chains of SMI generated transfectomas that produce IgM kappa RF autoantibodies reactive with human IgG1 and IgG4. In contrast to G6/17.109-reactive IgM kappa RF Waldenstrom's paraproteins, the SMI IgM kappa also reacts with several other self-antigens, including myoglobin, actin, and ssDNA. However, cotransfection of the SMI L chain with a vector encoding any one of 10 different G6-reactive Ig H chains generated transfectomas that produce IgM kappa antibodies without detectable polyspecific autoantibody activity. These results indicate that polyspecific antiself-reactivity of G6/17.019-reactive Ig is dependent on the somatically generated Ig third complementarity determining region. Collectively, these studies imply that selection may be responsible for the frequent expression of polyspecific autoantibodies in CLL and early B cell ontogeny.

Towards an integrated view of thymopoiesis

One prediction from the complex series of steps in intrathymic T-cell differentiation is that to regulate it the stroma controlling the process must be equally complex: the attraction of precursors, commitment to the T-cell lineage, induction of T-cell receptor (TCR) gene rearrangement, accessory molecule expression, repertoire expansion, major histocompatibility complex (MHC) molecule-based selection (positive and negative), acquisition of functional maturity and migratory capacity must all be controlled. In this review, Richard Boyd and Patrice Hugo combine knowledge of T-cell differentiation with thymic stromal cell heterogeneity to offer an integrated view of thymopoiesis within the thymic microenvironment.