HA binding by HPCs from mobilized blood, steady-state BM, or mobilized BM exhibit different HA receptor usage. HA binding was measured using three-color IF in the presence or absence of the indicated MoAbs as inhibitors of binding. Files were then gated for the HPCs and their HA binding was plotted as a histogram. (A) Treatment of mobilized blood HPCs. Anti-β1 integrin (JB1A) gave no detectable inhibition in 7 of 7 samples. Values are the mean percentage of HPCs binding HA ± SE of all 7 samples. For the 4 samples in which inhibition by MoAb CD44 was observed, the mean was 13% ± 6% of HPCs. For samples inhibited by anti-RHAMM, the mean was 10% ± 4% of HPCs. NS, mean inhibition was not significantly different from that of untreated samples. ^***P = .007 as compared with untreated or anti-β1 integrin-treated samples. (B) Treatment of steady-state BM HPCs. Treatment was as for (A) of 3 different steady-state BM HPC samples. Relative increase was calculated as the MFI of HA binding after pretreatment with anti-RHAMM or anti-CD44 divided by the MFI after anti-β1 integrin. The value of anti-β1 integrin-treated cells was set as 1.0 for each sample; anti-β1 integrin-treated and untreated cells had a similar intensity of HA binding. For all three samples, the pattern of MoAb modulation was the same. (C) Treatment of mobilized BM HPCs. HPCs from 3 different samples were treated as for (A). Relative decrease was calculated as the MFI of HA binding after pretreatment with anti-RHAMM or anti-CD44 divided by the MFI after pretreatment with anti-β1 integrin. The value of anti-β1 integrin-treated cells was set as 1.0 for each sample; anti-β1 integrin-treated and untreated cells had a similar intensity of HA binding. For all 3 samples the pattern of MoAb modulation was the same.