Keratinocyte Growth Factor Increases the CD4⁺Foxp3⁺ Regulatory T Cell-Pool by an Early Thymus-Independent and a Late Thymus-Dependent Mechanism.

Administration of keratinocyte growth factor (KGF) before and shortly after experimental allogeneic bone marrow transplantation (allo-BMT) has been shown to reduce the incidence and severity of graft versus host disease (GVHD) and to inhibit graft rejection. As naturally occurring CD4⁺Foxp3⁺ regulatory T cells (Treg) play an important role in the prevention of GVHD and graft rejection, we evaluated the effect of KGF on peripheral and thymic CD4⁺Foxp3⁺ regulatory T cell numbers in normal mice and BMT recipients. KGF (5mg/kg/day) was administered subcutaneously to adult B6 mice for 3 consecutive days. KGF enhanced the frequency of CD4⁺Foxp3⁺ Treg in peripheral blood and spleen twofold, to more than 15 % of CD4⁺ T cells, resulting in an increase in absolute numbers of CD4⁺Foxp3⁺ Treg within one week. From 2 weeks onwards, the frequency of CD4⁺Foxp3⁺ Treg gradually normalized, but the absolute numbers of Treg remained elevated (> 10 weeks) due to an increase of both CD4 and CD8 T cell numbers. In order to assess the relative contribution of thymic output versus peripherally expanded Treg, we next determined the frequency of T-cell receptor excision circles (TREC) in peripheral blood T cells. Analysis of sorted CD4⁺Foxp3⁺ Treg and CD4⁺Foxp3⁻ conventional T cells showed a decline in TREC frequency in both subsets 1 week after KGF administration, suggesting peripheral expansion. In contrast, the TREC frequency in peripheral T cells was significantly higher at 5 and 10 weeks after KGF, indicating a late thymic dependent effect. Normal B6 mice that were adoptively transferred with congenic T cells showed an enhanced, but transient, increase of the frequency of CD4+Foxp3+ Treg within one week, also suggesting that the early effect of KGF on peripheral Treg is independent of thymic output. In addition, we assessed the effects of KGF on Foxp3⁺ thymocytes. Thymic weight and thymic cellularity of all subsets, including CD4⁺Foxp3⁺ thymocytes, were 2 to 3-fold increased one week after KGF administration as compared to control mice. In addition, we found that KGF transiently affected the thymic architecture. The medullary epithelial compartment was virtually absent one week after KGF administration. Normal thymic architecture gradually reappeared after 2–3 weeks. To determine whether KGF-enhanced Treg numbers in BMT recipients would inhibit graft rejection, KGF was administered from day -3 until day -1 to Rag-1^−/− mice. Subsequently, the mice were 3Gy irradiated, supplied with 10⁵ B6 CD45.1 congenic T cells and transplanted with a MHC-matched minor-Ag mismatched T cell depleted 129Sv bone marrow graft. As in normal B6 mice, KGF-treatment of BMT-recipients enhanced peripheral CD45.1⁺CD4⁺Foxp3⁺ Treg numbers. This was associated with a reduced rate of bone marrow graft rejection (2 out of 8 KGF-treated mice as compared to 5 out of 7 PBS-treated mice). Taken together, our data show that KGF enhances the peripheral Treg pool by an early, transient and selective thymus-independent mechanism and thereafter by a non-selective thymus-dependent mechanism. The KGF-mediated expansion of the pool of peripheral Treg may contribute to the inhibitory effects of KGF on graft rejection.