T-cell progenitors migrate from bone marrow (BM) into the thymus. After maturation they are released as recent thymic emigrants (RTE) into the periphery ensuring the diversification of the T-cell repertoire. Both the kinetics with which RTE migrate through the periphery and the surface molecules they express are still unclear. In 1-and 18-month-old Lewis rats CD4 1 RTE were identified in blood, spleen, lymph node, and thoracic duct lymph by flow cytometry (CD45RC 2 and CD90 1 ), were differentiated from CD4 1 naive (CD45RC 1) and memory T cells (CD45RC 2 CD90 2 ), and were characterized regarding the expression of surface molecules. Both in 1-and 18-month-old animals the percentage of RTE among the CD4 1 population in blood was comparable to that in all other compartments. Surprisingly, RTE expressed a 4 -integrin, LFA-1, and interleukin (IL)-2 receptor at a significantly higher level than naive T cells and more comparable to memory T cells. Within lymphoid tissues RTE, naive, and memory T cells significantly upregulated the expression of CD44 and ICAM-1, and downregulated the expression of l-selectin. These changes were reversed before the cells reentered the blood. Thus, our data indicate that CD4 1 RTE travel through the periphery of young and old rats like mature T cells, continuously modulating their surface molecule expression.

Many lymphocytes enter tissues such as peripheral lymph nodes, and Peyer's patches through high endothelial venules (HEV). It is known that HEV differ in the expression of adhesion molecules as lymphocyte subsets do. Through the interaction of these molecules B and T lymphocyte subsets are thought to be preferentially directed into lymphoid organs. However, it is unclear which role these mechanisms play in vivo, since there are no studies demonstrating that blood lymphocyte subsets preferentially interact with different types of HEV in vivo. Therefore, in the present study the frequency of B, T, CD4+ and CD8+ lymphocytes in the wall of the HEV of rat peripheral lymph nodes and Peyer's patches was analyzed by immunohistology. In addition, the expression of CD44, VLA-4, LFA-1, ICAM-1, CD2 and L-selectin on B and T lymphocyte subsets of the blood was determined by flow cytometry. Although B and T lymphocytes showed significantly different levels of expression for each adhesion molecule investigated, the relation of B and T lymphocytes within the HEV of peripheral lymph nodes and Peyer's patches was strikingly comparable (38.0 +/- 5.2% vs. 40.6 +/- 5.7% and 62.0 +/- 5.2% vs. 59.4 +/- 5.7%, respectively). The same was true for CD4+ and CD8+ cells. Thus, although HEV and the blood lymphocyte subsets differ markedly in their expression pattern of adhesion molecules, the existing levels are sufficient to mediate comparable entrance of B and T lymphocyte subsets into both types of HEV.

After liver transplantation, the release of donor leukocytes into the host and the uptake of host leukocytes by the graft is one of the earliest immunologic interactions between donor and host. Using three-color flow cytometry, these interactions were investigated in eight patients from 5 min-24 h after receiving HLA unmatched liver grafts. Five minutes after reperfusion, 5.0 YO * 1.4 YO of all blood leukocytes in the host were of donor origin, decreasing to 1.1 % 0.8% after 24 h. Donor granulocytes preferentially disappeared from the host circulation, whereas no differences were found between NK-cells and various B-and T cell subpopulations. Furthermore, host granulocytes were preferentially retained in the donor liver. Thus, despite extensive pre-operative perfusion, more than lo9 donor leukocytes quickly leave the liver graft while host granulocytes preferentially accumulate there. A better understanding of the molecular mechanisms mediating these early interactions might help to develop new strategies for diagnosis and therapy of liver graft rejection.

The subset composition of the migrating lymphocyte pool is largely unknown. In order to determine the number of B, T, CD8+, CD4+ and CD4+ 'naive' (CD45RC+) and 'memory' (CD45RC-) lymphocytes in this pool, the thoracic duct lymph of the rat was drained for 7 days. The effect of lymphocyte depletion on the number of blood lymphocytes was also monitored. In addition, the influence of continuously applied interferon-gamma (IFN-gamma) on the mobilization of the migrating lymphocyte pool was investigated. Within 1 week 2 x 10(9) thoracic duct lymphocytes (TDL) were collected, which represents about 50% of the total lymphocyte pool of an adult rat. Among the migrating lymphocytes an early and a late mobilized population could be differentiated. In the former the CD4+ 'naive' (CD45RC+) T lymphocytes constituted the largest population, whereas in the latter it was the B lymphocytes. Continuous infusion of IFN-gamma did not affect the number of lymphocytes in the blood. In contrast, in the thoracic duct IFN-gamma reduced the appearance of all lymphocyte subsets. However, the pattern of reduction over time differed markedly depending on the population (early or late mobilized) and the phenotype (B- or T-lymphocyte subsets). Thus, the migrating lymphocyte pool of the rat is very heterogeneous regarding its populations and shows complex changes in the mobilization pattern after IFN-gamma stimulation. Future studies should focus on how the size and the composition of the migrating lymphocyte pool is regulated.

Although several distinct adhesion pathways are now well characterized, it is not clear whether analysis of adhesion molecule expression on leucocytes is sufficient to predict their interaction with endothelium in vivo. Therefore, in the present study this question was addressed by investigating the interaction between blood leucocyte subsets and high endothelial venules (HEV). The expression of different types of adhesion molecule (CD44, alpha 4-integrins, LFA-1, ICAM-1, CD2 and L-selectin) on lymphocytes, NK cells, monocytes and granulocytes of rat blood was determined by flow cytometry. In the same animals the numbers of blood leucocyte subsets present in the HEV of axillary lymph nodes and Peyer's patches were analysed using immunohistology. In the HEV of both axillary lymph nodes and of Peyer's patches lymphocytes (greater than 10,000 per mm2), as well as small numbers of NK cells and monocytes (less than 500 per mm2), were found. In contrast, granulocytes were not detected here. Lymphocytes, NK cells, monocytes and granulocytes each expressed CD44, alpha 4-integrins, LFA-1, ICAM-1, CD2 and L-selectin in a pattern characteristic to cell type, but this did not correlate with the different ability of the leucocyte subsets to interact with the two types of HEV. In conclusion, determining the expression of CD44, alpha 4-integrins, LFA-1, ICAM-1, CD2 and L-selectin on blood leucocytes alone is not sufficient to predict leucocyte/endothelium interaction in vivo.