Bringing alloreactivity into focus with the lens of TCR cross-reactivity

SUMMARY
CD8+ T cells drive acute cellular rejection (ACR) of transplanted tissues via alloreactivity. Old paradigms about
the broad degeneracy of alloreactivity are giving way to the idea that many, if not most, alloreactive T cells
recognize unique peptide-allo-HLA complexes, reflecting what is referred to as allospecificity. The targets
recognized by alloreactive T cells, however, remain almost completely undefined. Of the few allo-ligands
identified, fewer have been validated in transplant patients. This presents a major problem as ACR remains the
number one cause of allograft loss, and widely used immunosuppressive therapies for ACR are associated with
serious risks and toxicities. Mimicking what has been observed in cancer, autoimmunity, and infectious disease,
knowledge of the targets recognized in ACR would enable advances in predicting, monitoring, and ultimately
treating ACR. Recent work has shed light on the challenge of finding allo-ligands: significant links are now known
to exist between alloreactivity and anti-viral immunity. Our recent mouse model of rejection demonstrated this
conclusively: compared to viral-naïve mice, viral-immune mice exhibit substantially accelerated allograft
rejection, characterized by massive infiltration and activation of virus-specific memory T cells. Strikingly, these
cells display substantial responsiveness to allogeneic but not syngeneic allografts. The connection between viral
infection and alloreactivity extends is not limited to mouse models. Several groups have begun using “off-the-
shelf” viral-specific T cells (VSTs) to treat transplant patients with active infection, and there is evidence that
HLA-mismatches with VST therapy can accelerate rejection. Altogether, these observations provide a conceptual
framework for re-envisioning alloreactivity through the dual lenses of allospecificity and viral cross-reactivity. We
propose to develop this framework, using viral reactivity as the “hook” for identifying allo-ligands. Our driving
hypothesis is that viral/allo cross-reactivity occurs predominantly through molecular mimicry at the level of
peptide/MHC structure, with the surfaces of specific allo-peptide/allo-MHC complexes on donor tissue mimicking
those of viral peptide/self-MHC complexes in recipients, facilitating cross-reactive binding of TCRs. Accordingly,
we have developed a novel approach for identifying mimics that incorporates state-of-the-art methods for
sequence analysis, structural modeling, and structural comparison, and leverages these with transcriptomic
databases and other immunoinformatics tools. Our preliminary data shows our approach can successfully
identify allo-mimics of viral-specific TCRs implicated in rejection. Building on this, we aim here to expand our
observations and refine our approach for translation to humans. Our two aims are 1) to define the allo antigens
recognized by viral-specific TCRs in mice and 2) to apply this knowledge to human cells. Our proposed work will
allow immunologists to (finally!) begin identifying allo-ligands recognized by CD8+ T cells in ACR, enabling
significant advances in understanding alloreactivity and ultimately predicting, monitoring, and treating rejection.

Grant Number: 1R21AI191060-01
NIH Institute/Center: NIH
Principal Investigator: Brian Baker