Novel Roles for the Homeobox Gene, HoxA1, and Its Truncated Form, HoxA1-T, in the Regulation of Hematopoietic Stem Cell Self-Renewal.

There is increasing evidence that homeobox (Hox) genes play critical roles in the regulation of hematopoiesis. We found that both forms of the earliest HoxA gene, HoxA1 and its alternatively spliced transcript, HoxA1-T, which lacks the homeobox domain, are expressed in immature populations of hematopoietic stem cells (HSCs) and progenitor cells. In more mature bone marrow cell (BM) populations the levels of HoxA1-T increases relative to HoxA1 with both transcripts absent in mature peripheral blood (PB) cells. Roles for either of these Hox transcripts in hematopoiesis have not yet been described. We overexpressed either HoxA1 or HoxA1-T in BM using a modified GFP-containing MSCV vector (MXIE). Overexpression of either HoxA1 transcript did not affect expression levels of other Hox genes relative to control (empty vector). Overexpression of HoxA1-T significantly reduced the numbers of colony-forming cells (CFCs) produced by 500 GFP+ BM compared to control GFP+ BM (mean±SEM control 73±6.4; HoxA1 57±6.8; HoxA1-T 26.7±5.4; n=6; P<0.01 control vs. HoxA1-T). Interestingly, colonies generated by HoxA1 GFP+ BM contained ~3-fold more cells (mean±SEM: control 22,000±900 cells; HoxA1 61,700±12,000 cells; HoxA1-T 20,600±6,360 cells; n=6; P<0.005 HoxA1 vs. control and HoxA1-T). In each of 2 experiments 10,000 HoxA1 overexpressing BM grew in liquid suspension culture for up to 15 weeks, increasing by ~239-fold weekly. In contrast, control BM proliferated for only 3 weeks, with weekly ~53-fold increases, whereas HoxA1-T BM expired after 2 weeks of culture, with weekly cell increases of only ~7-fold (control v HoxA1 P<0.01; HoxA1 v HoxA1-T P<0.001). There was no difference in the number of day 12 colony-forming unit-spleen (CFU-S) formed from 2500 control or HoxA1 GFP+ BM (11.2±0.47 and 12.6±1.26 respectively). In contrast, 2500 HoxA1-T GFP+ BM produced significantly fewer CFU-S (8.9±0.61) compared to both control and HoxA1 BM (n=4, P<0.02). To assess HSC potential, lethally irradiated CD45.2+ recipients (n=6/group) were injected with 5x10⁶ congenic CD45.1+ BM immediately post-transduction without selection (all groups had similar transduction efficiencies). All recipients had >75% donor cells (CD45.1+, GFP+/−) in their PB at 3 months post-transplant. The average %GFP+ cells in recipients were similar for control (24.5±7.0%) and HoxA1 cells (20.8±2.5%). Strikingly, HSCs overexpressing HoxA1-T had markedly reduced repopulating ability (3.5±0.6% GFP+, P<0.05 HoxA1-T vs. control or HoxA1). Poisson statistics were used to quantitate HSC frequency in secondary transplant recipients. Mice were injected with 5x10³, 5x10⁴, 2x10⁵ or 1x10⁶ BM from control or HoxA1 primary recipients together with 2x10⁵ congenic CD45.2+ BM (n=10/group). The frequency of HSCs was markedly higher (~32-fold) in the HoxA1 cells (~1 HSC per 1.8x10⁵ BM) compared to control (~1 HSC per 5.8x10⁶ BM; P<0.001). Although secondary recipients injected with 1x10⁶ HoxA1-T BM showed donor (CD45.1+) contribution (15.3±1.8%), none of the 10 recipients had GFP+ donor cell reconstitution. These data therefore suggest that HoxA1 enhances HSC self-renewal whereas HoxA1-T rapidly promotes HSC differentiation. These novel findings highlight important roles for the two HoxA1 transcripts in the regulation of HSCs, the mechanisms of which are currently being assessed.