Deletion of donor-reactive T cell clones after human liver transplant

Abstract

We recently developed a high throughput T cell receptor β chain (TCRβ) sequencing-based approach to identifying and tracking donor-reactive T cells. To address the role of clonal deletion in liver allograft tolerance, we applied this method in samples from a recent randomized study, ITN030ST, in which immunosuppression withdrawal was attempted within 2 years of liver transplantation. We identified donor-reactive T cell clones via TCRβ sequencing following a pre-transplant mixed lymphocyte reaction and tracked these clones in the circulation following transplantation in 3 tolerant and 5 non-tolerant subjects. All subjects showed a downward trend and significant reductions in donor-reactive TCRβ sequences were detected post-transplant in 6 of 8 subjects, including 2 tolerant and 4 non-tolerant recipients. Reductions in donor-reactive TCRβ sequences were greater than those of all other TCRβ sequences, including 3rd party-reactive sequences, in all 8 subjects, demonstrating an impact of the liver allograft after accounting for repertoire turnover. Although limited by patient number and heterogeneity, our results suggest that partial deletion of donor-reactive T cell clones may be a consequence of liver transplantation and does not correlate with success or failure of early immunosuppression withdrawal. These observations underscore the organ- and/or protocol-specific nature of tolerance mechanisms in humans.

Keywords: T cell biology; basic (laboratory) research/science; immunobiology; liver transplantation/hepatology; monitoring: immune; tolerance.

Fig. 1.. Study design.

(A) Schematic of subjects in ITN030ST study randomized to immunosuppression withdrawal. Tolerant subjects diagrammed on left and non-tolerant subjects on right. Key time points shown at top and sample collection at bottom. (B) Schematic of identification and tracking of the donor-reactive TCRβ repertoire via the pre-transplant CFSE-MLR.

Fig. 2.. Reduction in detection of donor-reactive TCRβ sequences after liver transplantation.

Counts of donor-reactive TCRβ sequences in pre- and post-transplant CD4 and CD8 samples with ≥90% detection power for (A) tolerant and (B) non-tolerant non-immune, non-viral subjects, and for (C) tolerant and (D) non-tolerant HCV subjects. n corresponds to the number of total donor-reactive CD4 or CD8 sequences. Fold change in detection of donor-reactive sequences among (E) non-immune, non-viral subjects and (F) HCV subjects. Fold change is defined as the ratio of the odds of detecting donor-reactive sequences post-transplant relative to pre-transplant. Notation following each subject in the key refers to p-value at each time point pre-randomization/post-randomization. *p<0.05 reduction or increase in odds of detection of donor-reactive sequences compared to pre-transplant (two-sided Fisher’s exact test); tabulated data in Table S4.

Fig. 3.. Repertoire turnover post-transplant.

Jensen-Shannon-Divergence (JSD) comparing the top 1000 TCRβ sequences at each post-transplant time point to the top 1000 TCRβ sequences pre-transplant for (A) non-immune, non-viral subjects and (B) HCV subjects. JSD ranges from 0 to 1, with 1 representing complete divergence and 0 representing identical samples.

Fig. 4.. Reduced detection of donor-reactive sequences after accounting for repertoire turnover.

Relative change in detection of donor-reactive TCRβ sequences for (A) non-immune, non-viral subjects and (B) HCV subjects, where black are tolerant and red are non-tolerant subjects. Relative change at a post-transplant time-point is the ratio of the odds of detecting pre-transplant-identified donor-reactive sequences to the odds of detecting any pre-transplant-identified (unstimulated) sequences at the same post-transplant timepoint. A value of 1 indicates that the proportion of pre-transplant-identified donor-reactive sequences detected at a given time point was equal to that for any pre-transplant-identified sequences. Values <1 and >1 indicate a lower rate and a greater rate, respectively, of detection of donor-reactive versus any pre-transplant sequences. Notation following each subject in the key refers to p-value for each time point pre-randomization/post-randomization. *p<0.05 reduction or increase in odds of post-transplant detection of pre-transplant-identified donor-reactive sequences compared to unstimulated pre-transplant sequences (two-sided Fisher’s exact test); tabulated data in Table S5.

Figure 5.. Reduced detection of donor-reactive relative to 3^rd party-reactive TCRβ sequences.

At pre-randomization and post-randomization time points, comparison of (A) fold change in detection of donor-reactive sequences to fold change in detection of 3^rd party-reactive sequences and (B) change in detection relative to pre-transplant unstimulated sequences of donor-reactive sequences compared to 3^rd party-reactive sequences. *p<0.05, **p<0.01, p-value calculated from two-sided paired t-test.