Abstract

Chronic graft-versus-host disease (cGVHD) is the most common late complication after allogeneic hematopoietic cell transplantation (HCT). Its pathophysiology is poorly understood, a problem hampered by the scarcity of mouse models that accurately reproduce the human scenario. While acute (aGVHD) is characterized by donor T cell (doTC) infiltration damaging target tissues, cGVHD displays a wider range of alterations, often resulting in fibrosis with disrupted tissue architecture. Whether or not doTC drive this process versus (vs) other cell types has not yet been clarified. Indirect disruption of normal immunity caused by destruction of lymphoid tissues could contribute to the autoimmune-like syndrome. Here, we describe the progression of lymphocyte infiltrations of GVH targets post-HCT. Two MHC-matched, minor antigen-mismatched strain combinations were studied. Myeloablated BALB. B (H2b) and BALB/c (H2d) recipients were given FACS purified hematopoietic stem cells (HSC: c-Kit+Thy1.1loLin-Sca-1+) + 10^7 splenocytes (SP) from C57BL/6 (B6) (H2Db) or B10.D2 (H2d) donors, respectively. Mononuclear cells (MNC) from bone marrow (BM), spleen, lymph nodes (LN), thymus, peripheral blood, and liver were FACS analyzed for cell (sub-) type and chimerism 14, 28, 50 and 150 days (d) post-HCT. BALB.B recipients of B6 HSC + SP developed aGVHD with severe weight loss (>30%) and a mortality of 22%. Survivors never fully recovered baseline weight. Histologic changes shifted from defined infiltrates to disseminated disruptions of tissue structures. In contrast, BALB/c recipients showed sclerodermatous skin changes, but no mortality or other signs of systemic disease. In both strains the BM was markedly infiltrated with doTC during the acute phase: 14d post-HCT BALB.B BM consisted of ~50% mature doTC, and the majority (2/3) of these were CD8+ with an effector memory (EM) phenotype (<2% naïve CD8+). Tetramer staining revealed reactivity against the minor antigen H60 in 8–24% of CD8+ cells. The fraction of doTC decreased to ~20% of BM MNC on d28, and <5% by d50. CD4 and CD8 TC achieved a balanced ratio. While SP-derived doTC persisted long-term (d150) in the BM they were not H60 positive. B cells (BC) comprised only <3% of BM early post-HCT vs ~70% in recipients of pure HSCs. BC recovery was severely delayed (0.5–23% d50) and varied depending on the degree of GVHD. B10.D2 doTC infiltrates in the BM of BALB/c recipients (<42%) showed no preponderance of CD8 over CD4 TC. BC reconstitution was less impaired (up to 50% d28). By d50 TC numbers in the BM had normalized to <3%, and ~30% of cells were BC. Long-term follow up revealed no further abnormalities. Detectable doTC displayed equal proportions of naïve, central memory (CM) and EM phenotypes. The liver is a main target of B6 grafts in BALB.B recipients: 14d post-HCT Ficoll-Paqueisolated MNC contained 60–80% doTC, which were mostly EM CD8+ (80%), reactive to H60 in up to 30%. These subsets decreased over time, however, doTC infiltrates persisted beyond d150 (~30% of MNC). Infiltrating BCs were <2% early post-HCT vs ~15% in recipients of pure HSC. Of note, livers of the latter contained mainly Mac1Gr1+ cells and only modest amounts of doTC (<5%). BALB/c had less liver infiltrates on d14 and 28 (30–50%), and no prominence of CD8+. Instead, Mac1/Gr1+ cells predominated (30– 60%). 150 days post-HCT no difference was noted in lymphocyte distribution between normal controls, HSC recipients and those receiving SP grafts. Regarding other potential target organs, persisting mature doTC were notable in the thymuses of some B6 into BALB.B hosts. Generally, their proportion of immature double positive (DB) CD4/8 TC appeared decreased. Thymuses of B10.D2 into BALB/c recipients normalized completely by d150. Likewise their LN and spleen fully recovered by d50, with CD4>CD8 TC, naïve TC phenotypes exceeding EM, and ~40% BC. In conclusion, doTC infiltration of host organs is not inevitably associated with impairment of organ function and clinical disease. Nonspecific invasion of doTC likely occurs throughout the organism post-allogeneic HCT, and specific stimuli, such as minor antigens or cytokines are required to activate and perpetuate pathology. These data further suggest that characterizing the composition in the infiltrate in the acute setting may permit prediction of whether or not the initial infiltrates will lead to the fibrotic and sclerotic changes, the hallmarks of cGVHD.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author