Background: Poor graft function (PGF) is a serious complication after allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, the mechanisms underlying PGF remain to be elucidated, which increases the difficulty of managing PGF. In murine study, effective cross-talk between hematopoietic stem cells (HSCs) and bone marrow (BM) micro-environment plays an important role in hematopoiesis. HSCs occupy a hypoxic BM micro-environment to protect them from oxidative stress, whereas excessive reactive oxygen species (ROS) could inhibit HSCs self-renewal and induce HSCs exhaustion resulting in hematopoietic dysfunction. We recently reported that the impaired BM micro-environment may contribute to the occurrence of PGF post-HSCT using a prospective nested case-control study (Kong Y, et al. Biol Blood Marrow Transplant. 2013;19:1465-1473). Nevertheless, it is largely unknown whether the quantitatively and functionally impaired HSCs pre- and post-HSCT operate in the occurrence of PGF in allotransplants patients.

Aims: To investigate whether the quantitative and functional abnormalities of the donor BM CD34+ cells pre- and post-HSCT are involved in the pathogenesis of PGF.

Methods: The hematopoietic reconstitution activities of the CD34+ cells, sorted from the donors' BM of PGF and good graft function (GGF) patients, were evaluated in xenografted NOD-Prkdcscid IL2rgnull mice as an indicator of donors' HSCs function. To further investigate the effect of oxidative stress on normal hematopoiesis post-HSCT, the BM CD34+ cells of GGF allotransplant patients were treated of hydrogen peroxide (H2O2) with or without antioxidant N-acetyl-L-cysteine (NAC) in vitro. Subsequently, a prospective nested case-control study was performed enrolling 15 patients with PGF, 30 matched patients with GGF after allo-HSCT and their healthy donors. Quantification of the frequency, intracellular ROS levels, and cell cycle status of the BM CD34+ cells were analyzed by flow cytometry pre- and post-HSCT. Colony-forming capacity was investigated in CD34+ cells post-HSCT in vitro. The study was approved by the Ethics Committee of Peking University People's Hospital and written informed consent was obtained from all subjects.

Results: The hematopoietic reconstitution activity of the BM CD34+ cells in NOD-Prkdcscid IL2rgnull mice demonstrated no significant differences between the donors of PGF and GGF patients. In the subsequent in vitro study, increased ROS were found to play an important role in the exhaustion of the quiescent BM CD34+ cells of GGF patients, whereas treatment of ROS-abrogated CD34+ cells with the antioxidant NAC could partially, but significantly restore the exhaustion and colony-forming capacity of CD34+ cells. In the prospective nested case-control study, all patient- and therapy-related variables were similar between patients with PGF and GGF. Polymerase chain reaction DNA fingerprinting of the STRs confirmed 100% donor chimerism in these patients. The frequency, intracellular ROS levels and cell cycle status of the transplanted donor BM CD34+ cells showed no remarkable differences pre-HSCT. Nevertheless, the percentage of CD34+ cells post-HSCT and their colony-forming capacity, especially the quiescent CD34+ CD38low fraction, decreased remarkably in PGF patients when compared to that in GGF patients. Notably, significantly increased ROS levels were observed in CD34+ and CD34+ CD38low fractions of PGF patients post-HSCT.

Summary/Conclusion: Although the frequency and function of the transplanted donor BM CD34+ cells of PGF were demonstrated normal pre-HSCT, the increased levels of ROS and exhaustion of the quiescent CD34+ cells may operate in PGF post-HSCT. Our preliminary data indicate that an impaired BM micro-environment which may hamper the hematopoietic reconstitution of the donor HSCs in the recipients, rather than the defective donor HSCs, was involved in the occurrence of PGF. Therefore, novel therapeutic approaches, such as antioxidative therapy to maintain hypoxia BM micro-environment, promise to facilitate hematopoietic recovery in PGF.

Acknowledgment: Supported by the National Natural Science Foundation of China (grant nos. 81370638&81230013), and the Beijing Municipal Science and Technology Program (grant nos. Z141100000214011& Z151100004015164& Z151100001615020).


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.