To the Editor
We read with great interest the findings described by Shi et al. They detected high levels of donor-derived CD56+, CD3+, and CD14+ T cells in the first explanted liver grafts in both the short and long term after liver transplantation (LT). They concluded that these long-lived intragraft leukocytes in LT patients could cause long-term hematopoietic chimerism, challenging our conclusion that long-term blood chimerism is derived from hematopoietic stem/progenitor cells (HSPCs) in the liver,1 although they also suggest the possibility that it may result from hematopoiesis of relocated donor HSPCs.
The critical issue here is whether the mature leukocytes in the intragraft are really long-lived. Natural killer (NK) cells (CD56+) are generally considered short-lived effector cells. Although it is still controversial, recent studies have demonstrated that NK cells have the ability to become long-lived memory cells and contribute to secondary immune responses.2 The mechanisms for NK cell longevity are unclear, but these self-renewing capable NK cells are not just simple mature NK cell proliferation. In liver tissue, CD14+ cells are mostly Kupffer cells with a lifespan of 3.8 days, which are generated and maintained by a high monocyte influx rate or local progenitor proliferation.3 However, what the local progenitor cells are and where they are derived from are yet to be clarified.
Based on the understanding as outlined above, we think high levels of donor-derived T cells in recipient liver are unlikely to result from a simple mature T-cell proliferation to survive for 2 years after LT. Instead, it may generate from an existent small population of donor HSPCs1 or blood HSPCs in liver grafts (Shi et al. appears to agree with the latter point), which are able to self-renew and differentiation to maintain the mature leucocyte pool. In this sense, the findings by Shi et al. prove our hypothesis.1 Regarding the origin of HSPCs, both consistent presences in liver or from blood HSPCs are possible. We prefer the former hypothesis, because the population was maintained in liver grafts after extensive perfusion, but we did not exclude the probability of blood HSPCs in liver grafts.1 Finally, in 5 of the explanted liver grafts, Shi et al. could not detect donor-derived Lin−CD34+ HSPCs, which does not mean the HSPCs do not exist in healthy liver grafts. That Lin−CD34+ or Lin−CD45+ liver cells isolated from over 30 healthy liver grafts are able to form a hematopoietic colony and engraft in immunodeficient mice are the evidence of the presence of HSPCs in liver grafts.1 Moreover, Lin−CD34+CD38−CD90+ HSPCs only account for 0.03% of total liver cells,1 whereas the Lin−CD34+ population described by Shi et al. was based on different human leukocyte antigen markers, which might further complicate the measurement.