Acute graft-versus-host disease (aGVHD) is caused by alloreactive effector T cells attacking the gastrointestinal tract, liver and skin after allogeneic hematopoietic cell transplantation (aHCT). The mechanism by which alloreactive T cells target these organs and not others remains elusive. Recently, we reported that different secondary lymphoid organs (SLOs), as alloreactive priming sites, can imprint distinct homing phenotypes on evolving alloreactive effector cells in vivo. However, preventing access to selected lymphoid organs (via the use of blocking antibodies or recipient mice lacking Peyer’s patches (PP), PP and lymph nodes (LN) or spleens) did not alter the aGVHD organ manifestation. These findings not only suggested a high redundancy of SLOs as induction sites of aGVHD, but also questioned whether homing instruction of alloreactive T cells by these sites can explain the mechanism of aGVHD target organ manifestation.

To test the homing instruction model we transplanted transgenic luciferase+ (luc+) FVB/N (H-2q, Thy1.1+) splenocytes into conditioned (2×400rad) Balb/c recipients (H-2d, Thy1.2+). On day+3 we isolated luc+ donor lymphocytes from peripheral LN, mesenteric LN, or spleens and transferred them into conditioned secondary allogeneic recipients. 16 hours later, bioluminescence imaging revealed that allogeneic luc+ T cells irrespective of their original priming site targeted the intestinal tract and liver. Subsequently, we compared aHCT of conditioned with non-conditioned secondary Balb/cRag−/− cγ-Chain−/− recipients.

Surprisingly, we found allogeneic luc+ T cells accumulating in SLOs in non-conditioned recipients in contrast to intestinal and hepatic tissues in conditioned recipients.

These in vivo findings establish that alloreactive effector cells migrate to aGVHD target tissues because of attraction to these sites rather than specific instruction by SLOs. Therefore, we propose a signal hierarchy model of alloreactive cell trafficking whereby inflammatory signal/ligand interactions dominate over organ-specific homing receptor/ligand interactions.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author