Mobilization of different peripheral blood cells in mice
Through our extensive research efforts over many years, we have developed a new class of small molecule agents that are potent antagonists of CXCR4. On the basis of our representative compound HF50731 (29), we have developed a new highly potent small molecule analog named HF51116, which features an unsymmetrical polyamine (Figure 1A). HF51116 binds strongly to CXCR4 with the IC50 of 12 nM in competitive binding with 12G5 (26). We examined the compositions and dynamics of different PB cells in mice following subcutaneous injection of HF51116. The PB showed time-dependent changes in WBCs and neutrophils in response to HF51116. At 5 mg/kg, total WBC numbers in PB achieved a maximum number (18.83 K/L) at 60 min post HF51116 injection (Figure 1B). Increases in neutrophil numbers occurred faster and lasted longer when compared to the vehicle control (0.50 K/L), with increases of approximately 9-fold occurring from 30 min to 2 h after HF51116 treatment and 3.6-fold increases observed at 4 h (Figure 1C). The lymphocyte numbers (Figure 1D) started to increase at 30 to 60 min, followed by a dramatic decrease at 1 to 4 h. No changes were observed in platelet numbers in the PB in response to HF51116 injection, when compared to the vehicle control (Figure 1E, 816.83 K/L), suggesting that HF51116 specifically mobilized WBCs.
Rapid mobilization of hematopoietic progenitor cells (HPCs) in mice
We demonstrated that the HPC mobilization induced by HF51116 was dose- and time-dependent (Figure 2A-D). At 1 h post injection of HF51116, the colony numbers reached a plateau at 5 mg/kg, with no further increase at 10 mg/kg and 20 mg/kg (Figure 2A). The plateau level was about 9.36-fold higher than the baseline circulating level (negative control group, 185 CFUs/mL). The mobilizing efficacy of HF51116 was comparable to AMD3100 (30) at dose of 5 mg/kg. In comparison with the negative control group, HF51116 induced 8.73-fold increases in CFU-GM (1035/mL), 11.01-fold increases in BFU-E (698/mL), and 9.75-fold increases in CFU-GEMM (33/mL) numbers in PB (Figure 2B).
CFUs was also rapidly increased at 15 min after HF51116 treatment and reached a pick level at 30 min; 9.57-fold higher than the level in PB collected before HF51116 injection (198 CFUs/mL). The mobilization efficacy of HF51116 was comparable to that of AMD3100 from 15 min to 1 h post injection (Figure 2C). The PB also showed time-dependent changes in CFU-GMs, BFU-Es, and CFU-GEMMs in response to HF51116 (Figure 2D).
Inter-individual variability in the mobilizations of HPCs in patients means that approximately 15% patients are insensitive to G-CSF (19). This phenomenon also exists in different mouse strains (9). We used C57BL/6, C3H/HEJ, and DBA/2 mice to test the variability in HF51116 response and sensitivity to HF51116 in different mouse strains (Figure 2E). At 30 min post injection of 5 mg/kg HF51116, C57BL/6 and C3H/HEJ strains showed comparable sensitivities, but DBA/2 strain exhibited the better sensitivity than C57BL/6 strain.
Synergistic mobilization by HF51116 and G-CSF
After confirming the optimal dose and time for HPC mobilization, we tested the potential for synergistic effects of co-administration efficiency of G-CSF+HF51116.
G-CSF, G-CSF+HF51116, or G-CSF+AMD3100 were subcutaneously injected into mice. The mobilization efficacy in G-CSF+HF51116 group (24963 CFUs/mL) was 5.50-fold higher than in G-CSF group (4538 CFUs/mL) and 1.35-fold higher than in G-CSF+AMD3100 group (18512 CFUs/mL) (Figure 3A). We simultaneously examined the absolute number of hematopoietic stem and progenitor cells (HSPCs, Lineage- Sca-1+ c-Kit+: LSK) and HSCs (CD150+ CD48– lineage- Sca-1+ c-Kit+: SLAM LSK) in PB mobilized by these three treatments (31). The absolute number of LSK cells in G-CSF+HF51116 group (8612/mL) was 4.25-fold higher than in G-CSF group (2024/mL) (Figure 3B, C). In addition, in comparison with G-CSF group (214/mL), G-CSF+HF51116 treated group (715/mL) showed approximately 4.7-fold increases in SLAM-LSK cells (Figure 3D, E).
Long-term repopulating and self-renewing capability of HSPCs and HSCs mobilized by G-CSF and HF51116
We also evaluated the long-term repopulating and self-renewing ability of HSPCs and HSCs post injection of G-CSF+HF51116. The recovery of neutrophils and platelets was reflected in the engraftment kinetics, as the HSPCs and HSCs mobilized by G-CSF+HF51116 treatment showed timely and early engraftment (Figure 4A-C). Lethally irradiated CD45.1+ recipients receiving light-density mononuclear cells (LDMNCs) obtained from PB mobilized by G-CSF, G-CSF+AMD3100, and G-CSF+HF51116 (Figure 4A) showed similar engraftment kinetics: neutrophils recovered to the baseline level at around 18 days (Figure 4B) and platelets recovered at around 35 days (Figure 4C).
We employed CD45 congenic mice to demonstrate the long-term repopulating ability of HSPCs and HSCs (Figure 4D). The percentage of CD45.2+ cells in G-CSF+HF51116 (73.50%) group was 2.0-fold higher than in G-CSF group (35.44%) and 1.3-fold higher than in G-CSF+AMD3100 (54.60%) group post 6 months transplantation (Figure 4E). We also collected CD45.1/CD45.2 mice BM cells and tested self-renewal of the long-term repopulated cells in a non-competitive pattern of secondary transplantation (Figure 4D). No significant different of the percentage of CD45.2+ cells between G-CSF+HF51116 (79.11%) and G-CSF+AMD3100 (70.98%) groups; the percentage of CD45.2+ cells was still 2.0-fold higher in G-CSF+HF51116 group than in G-CSF group (41.76%) after 6 months post transplantation (Figure 4F). These data confirmed that the HSPCs and HSCs mobilized by G-CSF+HF51116 not only produce timely and early engraftment but they also retain a long-term repopulating and self-renewing capability.
Mobilization of different peripheral blood cells in monkeys
We addressed mobilization activity of HF51116 in monkeys (32). HF51116 was subcutaneously injected into rhesus monkeys at 10 or 1 mg/kg. Kinetics of WBCs, neutrophils, and lymphocytes were in time-dependent manners (Figure 5A-C). A maximum number of WBCs was achieved at 2 h (Figure 5A). HF51116 induced a 3.73-fold (10 mg/kg) change in WBC numbers when compared to 0 minute (average 8.87 K/L). Neutrophil numbers reached highest level at 4 h post injection of 10 mg/kg and 1 mg/kg of HF51116 (Figure 5B). Lymphocytes showed the maximum increases in number at 2 h for both doses (Figure 5C), similar to the WBC response. However, lymphocyte numbers decreased quickly from 2 to 8 h and had reached 70% of the 0-minute value by 24 h. HF51116 did not induce changes in platelets at either dose, similar to the change in mouse (Figure 5D and Figure 1D).
Mobilization of HSCs in monkeys
We examined the CD34+ cell counts (33, 34) in the PB in monkeys (Figure 6A) (35), and determined that the HF51116-induced HPC mobilization was time dependent (Figure 6B). 2 h post injection of 10 mg/kg HF51116, there were 17 CD34+ cells /L PB. The area under the curve (AUC) were 38.47 for dose 1 mg/kg and 61.50 for dose 10 mg/kg (Figure 6A). At 2 h post injection, 10 (5900 CFUs/mL) or 1 mg/kg (4373 CFUs/mL) HF51116 produced maximum HPC mobilization effects at the same time point when WBCs and CD34+ cells reached their maximum numbers. 1 mg/kg HF51116 induced an approximately 8.5-fold increase when compared to 0 minute (510 CFUs/mL) (Figure 6B).
At 10 mg/kg, the highest concentration of HF51116 (6305.89 ng/mL) in the PB plasma occurred at 15 min post injection (Figure 6C), and 100.54 ng/mL remained in the plasma at 24 h. At 1 mg/kg, the highest concentration (704.641 ng/mL) of HF51116 appeared at 30 min and no HF51116 remained in the plasma 8 h later.