Cross-reactive public TCR sequences undergo positive selection in the human thymic repertoire
Mohsen Khosravi-Maharlooei, … , Todd M. Brusko, Megan Sykes
Mohsen Khosravi-Maharlooei, … , Todd M. Brusko, Megan Sykes
Published March 28, 2019
Citation Information: J Clin Invest. 2019;129(6):2446-2462. https://doi.org/10.1172/JCI124358.
View: Text | PDF
Research Article Autoimmunity Immunology

Cross-reactive public TCR sequences undergo positive selection in the human thymic repertoire

Abstract

We studied human T cell repertoire formation using high-throughput T cell receptor β (TCRβ) complementarity-determining region 3 (CDR3) sequencing in immunodeficient mice receiving human hematopoietic stem cells (HSCs) and human thymus grafts. Replicate humanized mice generated diverse and highly divergent repertoires. We observed repertoire narrowing and increased CDR3β sharing during thymocyte selection. Whereas hydrophobicity analysis implicated self-peptides in positive selection of the overall repertoire, positive selection favored shorter shared sequences that had reduced hydrophobicity at positions 6 and 7 of CDR3βs, suggesting weaker interactions with self-peptides than were observed with unshared sequences, possibly allowing escape from negative selection. Sharing was similar between autologous and allogeneic thymi and occurred between different cell subsets. Shared sequences were enriched for allo–cross-reactive CDR3βs and for type 1 diabetes–associated autoreactive CDR3βs. Single-cell TCR sequencing showed increased sharing of CDR3αs compared with CDR3βs between mice. Our data collectively implicate preferential positive selection for shared human CDR3βs that are highly cross-reactive. Although previous studies suggested a role for recombination bias in producing “public” sequences in mice, our study is the first to our knowledge to demonstrate a role for thymic selection. Our results implicate positive selection for promiscuous TCRβ sequences that probably evade negative selection, given their low affinity for self-ligands, in the abundance of “public” human TCRβ sequences.

Authors

Mohsen Khosravi-Maharlooei, Aleksandar Obradovic, Aditya Misra, Keshav Motwani, Markus Holzl, Howard R. Seay, Susan DeWolf, Grace Nauman, Nichole Danzl, Haowei Li, Siu-hong Ho, Robert Winchester, Yufeng Shen, Todd M. Brusko, Megan Sykes

×

Figure 3

Proportion of shared CDR3βs between animals and experiments.

Options: View larger image (or click on image) Download as PowerPoint
Proportion of shared CDR3βs between animals and experiments. (A–C) Box-a...
(A–C) Box-and-whisker plots (dot plot for experiment 3 because of the smaller sample size) comparing proportions of shared CDR3βs between each asymmetric mouse pair in experiment 1 (n = 6 comparisons), experiment 2 (n = 30 comparisons), and experiment 3 (n = 2 comparisons) for each cell population at the nucleotide nonproductive, nucleotide productive, and aa levels. (D) Box-and-whisker plot distributions of the proportion of shared CDR3βs comparing all versus the top 100 sequences by frequency in experiment 2. (E) Comparisons in both directions between each pair of mice in experiment 2, depending on whether the mice received the same (autologous thymus, n = 12 comparisons) or a different thymus (allogeneic thymus, n = 18 comparisons). (F) Distributions of the proportion of shared CDR3βs between each pair of mice within and across experiments. Exp, experiment. (G) Ratio of unique CDR3β nucleotide sequences per aa sequence (Nt/aa ratio) in shared versus unshared sequences for each pair of mice in experiment 2. Supplemental Table 5 shows the mean Nt/aa ratio for each subset and P values comparing different subsets. Box-and-whisker plots show the median, range, interquartile range, and outliers. ***P < 0.001, by paired t test with Bonferroni’s multiple testing correction.