Evolution of a complex T cell receptor repertoire during primary and recall bacterial infection

Abstract

The mechanisms underlying the genesis and maintenance of T cell memory remain unclear. In this study, we examined the evolution of a complex, antigen-specific T cell population during the transition from primary effector to memory T cells after Listeria monocytogenes infection. T cell populations specific for listeriolysin O (LLO)91-99, the immunodominant epitope recognized by H2-Kd-restricted T lymphocytes, were directly identified in immune spleens using tetrameric H2-Kd-epitope complexes. The T cell receptor (TCR) Vbeta repertoire of specific T cells was determined by direct, ex vivo staining with a panel of mAbs. We demonstrate that LLO91-99-specific, primary effector T cell populations have a diverse TCR Vbeta repertoire. Analyses of memory T cell populations demonstrated similar TCR diversity. Furthermore, experiments with individual mice demonstrated that primary effector and memory T cells have indistinguishable TCR repertoires. Remarkably, after reinfection with L. monocytogenes, LLO91-99-specific T cells have a narrower TCR repertoire than do primary effector or memory T cells. Thus, our studies show that the TCR repertoire of primary effector T lymphocytes is uniformly transmitted to memory T cells, whereas expansion of memory T cells is selective.

Figure 1

Folding and biotinylation of soluble H2-K^d and generation of tetramers. (A) The cDNA for murine H2-K^d was mutagenized by PCR to delete the leader sequence (LS), the transmembrane (TM), and the cytosolic domain (CD), and to extend the COOH terminus with a biotinylation sequence (indicated in single letter amino acid code) recognized by the E. coli BirA enzyme. The biotinylated lysine residue is enclosed in a box. (B) Recombinant H2-K^d and human β₂m were overexpressed in E. coli and inclusion bodies were purified, resolubilized, and folded with LLO_91–99 peptide as indicated in Materials and Methods. A typical fast protein liquid chromatography (FPLC) gel filtration absorbance profile demonstrates a large peak consisting of folded H2-K^d, β₂m (seen in the gel inset), and peptide. HC indicates a small peak of aggregated, unfolded H2-K^d heavy chains. (C) Refolded, FPLC-purified, and biotinylated H2-K^d complexes were either directly subjected to PAGE (first labeled lane) or precipitated with conformation-dependent anti–H2-K^d-specific antibody SF1-1.1.1 (anti-K^d), control mouse IgG (mIgG) or streptavidin-agarose beads (SA). After SA precipitation, only a very small amount of folded H2-K^d could be precipitated with SF1-1.1.1 (anti-K^d, right lane). (D) Biotinylated H2-K^d complexes were mixed with streptavidin and again subjected to FPLC gel chromatography. The absorbance profile demonstrates a high molecular weight complex consisting of tetramerized H2-K^d–β₂m–peptide complexes (gel inset shows H2-K^d heavy chain, β₂m, and a faint band of streptavidin). The large peak consists of free streptavidin and carrier BSA (BSA/SA).

Figure 2

Costaining with LLO_91–99–H2-K^d tetramers and TCR Vβ mAbs. A LLO_91–99-specific CTL line was generated from an L. monocytogenes–immunized BALB/c mouse by in vitro peptide restimulation. (A) P815 (H2^d) target cells were labeled with ⁵¹Cr and incubated in the presence (open circles) and absence (closed circles) of 10⁻⁶ M LLO_91–99 and decreasing numbers of LLO_91–99-specific CTL. The percentage of specific lysis and the E/T ratio are indicated. (B) The CTL line was stained for CD8 (anti-CD8α Cy-Chrome) and LLO_91–99 tetramers (PE-conjugated). Gating for CD8⁺ blasts revealed that nearly all T cell blasts stained with LLO_91–99 tetramers. (C) Lymphoblasts were stained with a panel of FITC-conjugated, Vβ-specific antibodies in the presence (white bars) and absence (black bars) of LLO_91–99 tetramers.

Figure 3

Direct ex vivo TCR staining of LLO_91–99-specific T cells. CD8⁺ T cells from the spleen of a BALB/c mouse immunized 7 d previously with a sublethal dose of L. monocytogenes were stained with LLO_91–99 tetramers (PE-conjugated) and FITC-conjugated antibody specific for TCR-α/β (left) and the TCR Vβ8 chain (right).

Figure 4

TCR Vβ staining reveals multiple subpopulations of LLO_{91– 99}-specific T cells. Immune BALB/c CD8⁺ splenocytes obtained 7 d after L. monocytogenes infection were stained with LLO_91–99-specific tetramers and a panel of 14 different FITC-conjugated, Vβ-specific mAbs. These histograms demonstrate the proportion of cells that are stained with each of the antibodies.

Figure 6

Primary and memory, LLO_91–99-specific T cells from individual mice have indistinguishable ratios of TCR Vβ chains. (top) Three BALB/c mice were immunized with a sublethal dose of L. monocytogenes and 7 d later peripheral blood lymphocytes were restimulated with LLO_91–99-coated splenocytes. 10 d later these T cell lines were stained with LLO_91–99 tetramers and the panel of TCR Vβ–specific antibodies (white bars; primary effector T cells). (bottom) 35 d after infection, these three BALB/c mice were killed and CD8⁺ T cells were isolated from spleens and stained with LLO_91–99 tetramers and the panel of TCR Vβ–specific antibodies (hatched bars; memory T cells). The percentage of cells stained with each of the TCR antibodies is indicated.

Figure 7

After a recall response, LLO_91–99-specific T cells express a more limited TCR repertoire than do primary effector T cells. Four BALB/c mice were infected with a sublethal dose of L. monocytogenes and 7 d later peripheral blood lymphocytes were used to generate LLO_91–99-specific T cell lines, as described for Fig. 6. These T cell lines were stained with LLO_91–99 tetramers and the panel of TCR Vβ-specific antibodies (white bars; primary effector T cells). 35 d after primary infection, these four BALB/c mice were reinfected with a 50-fold higher dose (100,000 bacteria) and 5 d later CD8⁺ splenocytes were isolated and stained with LLO_91–99 tetramers and the TCR Vβ panel (hatched bars; recall effector T cells). The percentage of cells stained with each of the TCR antibodies is indicated.

Figure 5

LLO_91–99-specific primary and memory T cell repertoires closely reflect the general TCR repertoire of BALB/c CD8⁺ T cells. Six BALB/c mice were infected with a sublethal dose of L. monocytogenes, and CD8⁺ T cells from three mice were stained for TCR Vβ expression 7 d after infection (top, primary effector T cells) and from the remaining three mice 35 d after infection (bottom, memory T cells). White bars indicate the percentage of LLO_91–99 tetramer-positive cells that stain with the individual TCR Vβ specific antibodies. Black bars indicate the percentage of overall CD8⁺ T cells that stain with the TCR Vβ-specific antibodies. Minimum number of gated CD8⁺ and tetramer-positive T cells for each TCR Vβ staining was 2,000 for primary effector T cells and 1,000 for memory T cells, respectively. n.d. = not done.

Figure 8

Model for TCR repertoire evolution during primary and recall infection with L. monocytogenes. The diversity of a pathogen-specific T cell population expanded during primary infection (arrows indicate time points of infection) is maintained in the memory pool. After rechallenge with the pathogen, the recall TCR repertoire is more restricted compared with the primary effector and memory T cell populations. These differences might be due to different in vivo expansion rates of T cells within the epitope-specific population.