Abstract
Abstract 2423
Homeobox (Hox) genes have been shown to play critical roles in normal and aberrant hematopoiesis, however, little is known about the role of Hoxa1. Unlike most Hox genes, Hoxa1 is expressed as two different isoforms: full length Hoxa1 (FL-Hoxa1) and a truncated form, Hoxa1-T, resulting from alternative splicing within exon 1 of the gene. The two isoforms are expressed differentially in immature hematopoietic cells.
Overexpression of FL-Hoxa1 and Hoxa1-T in mouse bone marrow (BM) cells significantly increased, and significantly decreased, respectively, cell proliferation in vitro compared to control (MXIE) transduced BM cells. These findings suggested that Hoxa1-T may negatively regulate FL-Hoxa1. We therefore generated a mutant Hoxa1 (muHoxa1) that expresses FL-Hoxa1 but was no longer capable of generating Hoxa1-T by changing the AGC serine to a TCT serine at the splice site of Hoxa1-T.
Hoxa1 levels in muHoxa1-overexpressing BM were significantly higher than those in FL-Hoxa1 BM. Furthermore, BM cells overexpressing muHoxa1 had higher proliferative potential in vitro than FL-Hoxa1-overexpressing BM. Their in vivo potential was therefore assessed. At 12 weeks post-transplant, all primary transplant recipients had similar %GFP expression in the peripheral blood (PB), being approximately 10%. At this time point all recipients of muHoxa1-overexpressing BM cells displayed thrombocytopenia (mean ± SEM PB platelets (x 106/ml): MXIE= 1062 ± 75; FL-Hoxa1= 1053 ± 146; muHoxa1= 681 ± 71*, *P<0.05 vs MXIE).
Secondary transplants were performed into irradiated and non-irradiated recipients. Strikingly, the %GFP+ve cells were markedly increased in the PB of recipients of muHoxa1 BM (%GFP: MXIE: 0.29 ± 0.06; FL-Hoxa1: 3.33 ± 1.38; muHoxa1: 27.46 ± 8.35*; *P<0.05 vs MXIE and FL-Hoxa1). All secondary recipients of muHoxa1 BM developed myeloid neoplasias, resembling myelodysplastic syndrome (MDS). The thrombocytopenia persisted (PB platelets (x 106/ml): MXIE: 923 ± 42; FL-Hoxa1: 812 ± 38; muHoxa1: 388 ± 108*; *P<0.001 vs MXIE and FL-Hoxa1) and secondary recipients of muHoxa1 BM developed anemia (PB Hb (g/L): MXIE: 143 ± 2.6; FL-Hoxa1: 146 ± 2.6; muHoxa1: 117 ± 5.2*, *P<.0005 vs MXIE and FL-Hoxa1). The PB leukocyte counts in the majority of muHoxa1 recipients were unchanged compared to MXIE and FL-Hoxa1 recipients, however, muHoxa1 PB cells were predominantly granulocytes. These neoplasms also occurred in non-irradiated recipients of muHoxa1 BM, although they had a much longer latency. Interestingly, 40% of the non-irradiated recipients developed acute myeloid leukemia between 6 and 12 months post-transplant. Importantly, integration site and cytogenetic analysis demonstrated that the malignant phenotype was not due to co-operating insertional mutagenesis or chromosomal instability in the transduced cells.
The PB anemia was accompanied by a significant two-fold reduction in Ter119+ cells in BM of muHoxa1-overexpressing cells (GFP+ve= 16.3± 5.3%, GFP-ve= 34.8 ± 5.1%, P<0.05). This was due to an accumulation of the cells in early stages of erythroid differentiation, with increased proportions of proerythroblasts (GFP+ve: 36.6 ± 7.6%; GFP-ve: 8.6 ± 1.3% P<0.002) and basophilic erythroblasts (GFP+ve: 34.3 ± 3.4%; GFP-ve: 13.2 ± 3.5%, P<0.001) in muHoxa1 Ter119+ cells.
The block in erythroid differentiation was also accompanied by significant alterations in the proportions of immature progenitors in recipients of muHoxa1 BM. GFP+ve cells were detectable in HSC and myeloid progenitor cell subsets, however, there was a significant increase in the proportion of megakaryocyte erythroid progenitors (MEPs) within the lineage negative, c-kit+, Sca-1 negative progenitor cell fraction (GFP+ve: 42.9 ± 2.1%; GFP-ve: 8.2 ± 1.3%, P<0.01).
Significantly reduced GATA-1 expression was observed in both the GFP+ve proerythroblasts (400-fold) and MEPs (100-fold) compared to GFP-ve populations sorted from the same mice (P<0.001). Taken together, these results suggest that overexpression of FL-Hoxa1 in the absence of Hoxa1-T results in the development of MDS. This is partly due to an accumulation of MEPs and impaired differentiation of erythrocytes, both of which have significantly downregulated expression of GATA-1. Given the striking similarities in hematological phenotype to human patients with MDS, this novel mouse model will be invaluable in identifying the mechanisms contributing to this disease.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.