The transcription factor hepatic leukemia factor (HLF) is strongly expressed in hematopoietic stem cells (HSCs) and is thought to influence both HSC self-renewal and leukemogenesis. However, the physiological role of HLF in hematopoiesis and HSC function is unclear. Here, we report that mice lacking Hlf are viable with essentially normal hematopoietic parameters, including an intact HSC pool during steady-state hematopoiesis. In contrast, when challenged through transplantation, Hlf-deficient HSCs showed an impaired ability to reconstitute hematopoiesis and became gradually exhausted upon serial transplantation. Transcriptional profiling of Hlf-deficient HSCs revealed changes associated with enhanced cellular activation, and cell-cycle analysis demonstrated a significant reduction of quiescent HSCs. Accordingly, toxic insults targeting dividing cells completely eradicated the HSC pool in Hlf-deficient mice. In summary, our findings point to HLF as a critical regulator of HSC quiescence and as an essential factor for maintaining the HSC pool during regeneration.

Key Points RNA interference screen targeted to primary human HSPCs identified CYTH1 as a crucial mediator of cell adhesion. CYTH1 is required for homing and proper bone marrow localization of HSPCs following transplantation.

No abstract

The transcription factor Hepatic Leukemia Factor (HLF) was originally identified in a chromosomal translocation with the gene E2A causing a subset of childhood B-lineage acute lymphoid leukemia. Moreover, HLF has been described as a regulator of circadian rhythm and recent findings have implicated HLF as a candidate “stemness” gene in both normal and malignant stem cells. Accordingly, overexpression of HLF in human hematopoietic stem cells (HSC) results in an enhanced reconstitution capability in NOD-SCID mice. However, little is known about HLF’s physiological role in hematopoiesis and HSC regulation. Using quantitative PCR, we found that HLF is highly expressed in mouse (C57Bl/6) HSC and is downregulated upon differentiation (HSC 3.2 (±0.95) fold (p<0.001), LSK 1.9 (±0.47) fold (p<0.05), CMP, GMP MEP all less then 0.1 fold, all values are compared to HPRT). This encouraged us to further investigate HSC function in the absence of HLF. The conventional HLF knockout (KO) mice (C57bl/6 background) were viable, born at normal Mendelian ratios and showed normal hematopoietic parameters (bone marrow cellularity: WT 2.7x107 (±5.4 x106), KO 3.3x107 (±6.4 x106), p>0.2 n=9). In addition, the HLF KO mice demonstrated normal lineage distribution of both mature cells in the peripheral blood and bone marrow as well as the frequency of immunophenotypic HSC (Lin-Sca1+ckit+CD34-Flt3-: WT 0.0005 (±0.5x10-4)%, KO 0.0005 (±0.1x10-3)%; n>10). However, in a serial competitive transplantation assay using whole bone marrow (200 000 cells 1:1 ratio), HLF KO cells demonstrated a significant reduction in reconstitution capacity in primary recipients (WT 56 (±15)%, KO 40.2 (±16)%, p=0.028, n>10), which was further increased in the secondary recipients (WT 87.2 (±26)%, KO 8.7 (±5.8)%, p<0.001, n>10). Almost no engraftment was detected from the HLF KO cells in tertiary recipients. To further evaluate stem cell activity in the absence of HLF, we next enumerated the number of competitive repopulating units (CRU) by limiting dilution assay, which revealed a 2.6 fold reduction, of CRU in the HLF KO mice compared to WT controls (WT 1.6 (±0.4)/105 bone marrow cells, KO 0.6 (±0.2)/105 bone marrow cells). Similarly, transplantation of sorted HSC (Lin-Sca1+ckit+CD34-Flt3-) also showed a 2.4 fold (WT 47.3 (±24)%, KO 19.4 (±25)%, p=0.16, n=9) reduced engraftment of total cells but with enhanced T cell frequency in peripheral blood (WT 19.5 (±6.2)%, KO 40.8 (±7.4)%, p=0.01, n=9). Since we also found that HLF was highly expressed in fetal liver derived HSC, we transplanted fetal liver HLF KO cells from E14.5 in a competitive repopulation setting. In line with the phenotype seen in the adult HLF KO mice, the fetal liver HLF KO cells demonstrated impaired reconstitution ability (WT 52.8 (±16)%, KO 0.9 (±1.4)%, n>10). Intriguingly, the phenotype was stronger than in the adult HLF KO HSC, indicating that HLF is particularly important during the expansion phase of HSC in embryonic development. The underlying mechanism of the reduced HSC activity is still unclear, but preliminary findings show that HLF KO HSC have enhanced ROS levels (WT 337 (±33), KO 510 (±55), p<0.05, n=3) and increased cycling HSC (G0: WT 66.5 (±6.4)%, KO 58.5 (±4.7)%; G1/S/G2/M: WT 33.6 (±6.6)%, KO 41.7 (±4.9)%, n=3). We are currently performing global gene expression analysis to further understand the mechanism of HLF in HSC regulation. Interestingly, we also found that HLF appears to regulate the identity of HSC by modulating the expression of the SLAM code on the cell surface of the HLF KO HSC. In contrast to the normal frequency of LSK Flt3-CD34- cells, the HLF KO mice displayed a 3.5 fold reduction in the frequency of LSK CD150+CD48- cells (WT 1.94x10-4 (±4.4x10-5)%, KO 0.56x10-4 (±1.5x10-5)%, p<0.001 n>10). Strikingly, transplantation of as many as 150 LSK CD150+CD48-HLF KO cells showed a complete lack of repopulating capacity in vivo. This did not correlate to the number of functional HSC seen when transplanting whole bone marrow and indicates that HLF affects the identity of HSC by modulating the expression of the SLAM markers. Taken together, we show here for the first time that HLF has a fundamental role in HSC biology during both fetal and adult hematopoiesis by regulating HSC activity and identity. Disclosures: No relevant conflicts of interest to declare.

No abstract