Abstract

Chronic graft-versus-host disease (cGVHD) is the most frequent long-term complication after allogeneic hematopoietic cell transplantation (HCT). While acute GVHD (aGVHD) is largely donor T cell (TC) driven in the setting of tissue damage, less is understood about the pathophysiology of cGVHD. There is clear evidence that besides TC, B cells (BC) contribute to this syndrome. cGVHD has not been widely studied in animal models. The existing cGVHD models use either MHC-mismatched strains, or the manifestations appear in a time frame more akin to aGVHD. To generate a comprehensive picture of cGVHD we studied 3 different MHC-matched, minor histocompatibility (miHA) mismatched strain combinations over an extended time period (day (d) 30, 60, 90, 120, 150, 300) for clinical signs, chimerism, IgG synthesis (donor/host), and histology: C57BL/6 (B6)→ BALB.B (H2Db), B10.D2→BALB/c (H2Dd), and AKR/J→BALB.K (H2Dk). Recipients were lethally irradiated and given FACS purified hematopoietic stem cells (HSC: c-Kit+Thy1.1loLin-Sca-1+). To induce GVHD titrated doses of splenocytes or purified CD4 or CD8 TC were co-transplanted. BALB.B recipients of B6 HSC + TC develop aGVHD. Morbidity as assessed by weight loss and mortality directly correlated with TC dose. Survivors of the acute phase stabilized by d50–70, with only subtle signs for months. However, they ultimately developed a full-blown burned-out picture of cGVHD, including erythrosquamous skin lesions, alopecia, cirrhotic liver changes, conjunctivitis and corneal clouding at 1 year post-HCT. Periportal liver infiltrates progressed over time to fibrosis. Intestines were primarily affected in the acute phase, whereas skin changes such as subcutaneous atrophy and infiltrative hair follicle damage were manifested late. Splenocytes or isolated CD4 TC (+HSC), but not CD8 TC, induced the full clinical and histological picture of GVHD. Despite a low barrier to engraftment, recipients of HSC only remained mixed chimeras. With addition of TC, conversion to full donor chimerism promptly occurred. However, delays in BC reconstitution correlated with the presence and degree of GVHD. Despite this lymphopenia high levels of IgG synthesis were observed post-HCT. BALB/c recipients of B10.D2 HSC grafts develop cutaneous GVHD (alopecia, erythrosquamous exanthema) between d20–50 post-HCT, which subsequently resolves. These mice had low mortality and morbidity and did not display clinical signs of systemic GVHD. Histological changes were subtle and engraftment occurred promptly. HCT of AKR/J→BALB.K required high numbers of HSC to achieve engraftment and chimerism remained mixed (beyond d100). Low numbers of co-transferred splenocytes or CD4 TC (but not CD8 TC) induced fulminant hyperacute GVHD with a very high mortality, approaching 80–100%. In contrast, CD8 cells appeared to facilitate engraftment of donor cells. Here, we characterized and compared miHA-mismatched HCT models with regard to cGVHD and engraftment. B6→BALB.B offers the most convincing histological signs of systemic cGVHD. Whereas the B10.D2→BALB/c developed an attenuated form, and AKR/J→BALB.K developed a devastating aGVHD. We believe mouse studies which allow the delineation of the GVHD-inducing (presumably CD4) TC subsets will not only yield a better understanding of the pathophysiology of GVHD, but will, together with the identification of engraftment facilitator cells (contained within the CD8 TC fraction) be the basis for the future of clinical graft engineering.

Author notes

Disclosure:Consultancy: JA Shizuru: Cellerant; Stem Cells, Inc. Research Funding: NIH.