Downstream targets of HOXB4 in a cell line model of primitive hematopoietic progenitor cells

Abstract

Enforced expression of the homeobox transcription factor HOXB4 has been shown to enhance hematopoietic stem cell self-renewal and expansion ex vivo and in vivo. To investigate the downstream targets of HOXB4 in hematopoietic progenitor cells, HOXB4 was constitutively overexpressed in the primitive hematopoietic progenitor cell line EML. Two genome-wide analytical techniques were used: RNA expression profiling using microarrays and chromatin immunoprecipitation (ChIP)-chip. RNA expression profiling revealed that 465 gene transcripts were differentially expressed in KLS (c-Kit(+), Lin(-), Sca-1(+))-EML cells that overexpressed HOXB4 (KLS-EML-HOXB4) compared with control KLS-EML cells that were transduced with vector alone. In particular, erythroid-specific gene transcripts were observed to be highly down-regulated in KLS-EML-HOXB4 cells. ChIP-chip analysis revealed that the promoter region for 1910 genes, such as CD34, Sox4, and B220, were occupied by HOXB4 in KLS-EML-HOXB4 cells. Side-by-side comparison of the ChIP-chip and RNA expression profiling datasets provided correlative information and identified Gp49a and Laptm4b as candidate "stemness-related" genes. Both genes were highly ranked in both dataset lists and have been previously shown to be preferentially expressed in hematopoietic stem cells and down-regulated in mature hematopoietic cells, thus making them attractive candidates for future functional studies in hematopoietic cells.

Figure 1

HOXB4 overexpression down-regulates the lymphoid marker B220 in EML cells. (A) Western blot analysis of total cell lysates fractionated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and probed with the I12 monoclonal anti-HOXB4 antibody revealed the presence of HOXB4 only in the lysate of EML cells transduced with retrovirus expressing HOXB4 (EML-HOXB4) and not in control cells. (B) By immunophenotyping of cell surface markers, the percentage of cells that express B220 decreases when HOXB4 is overexpressed in EML cells. All other lineage markers and stem/progenitor cell markers are expressed at similar levels in EML, EML-GFP, and EML-HOXB4 cells.

Figure 2

CD34 is expressed in a higher percentage of KLS-EML-HOXB4 cells than in KLS-EML-GFP cells. The percentage of cells that express CD34 is much higher in lineage-depleted EML-HOXB4 cells than in lineage-depleted EML-GFP cells, whereas the percentages of cells that express c-Kit and Sca-1 are similar. The cell surface profiles of the numbered lineage-depleted single-cell clones shown are representative of those of the other lineage-depleted single-cell clones that were used for microarray analyses.

Figure 3

Real-time PCR analysis of selected gene transcripts observed to be differentially expressed in RNA expression profiling experiments. Thirteen gene transcripts that were observed by microarray analysis to be differentially expressed at least 2-fold were chosen and tested. However, by microarray analysis, it was not possible to conclude which of the 2 isoforms of the HMBS gene (erythroid-specific and housekeeping isoforms) was differentially expressed because the probes for this gene on the array chip did not discriminate between the 2 isoforms. Therefore, real-time PCR was performed using primers specific for each of the 2 different isoforms of HMBS. Real-time PCR results confirmed that all 13 gene transcripts (including the 2 isoforms of HMBS) were indeed differentially expressed between KLS-EML-GFP and KLS-EML-HOXB4 cells, as observed in our RNA expression profiling experiments. Interestingly, the erythroid-specific isoform of HMBS was observed to be down-regulated approximately 28-fold more than the housekeeping isoform of HMBS (61.2-fold vs 2.2-fold). Values are mean plus or minus SD of triplicate PCR analyses of RNA from 3 different single-cell clones of each cell type.

Figure 4

GeneGO Pathway analysis of microarray results. GeneGo Pathway analysis of the RNA expression profiling results obtained by microarray indicates that the erythropoietin pathway in hematopoiesis development is the most highly affected pathway in KLS-EML cells that overexpress HOXB4.

Figure 5

Western blot analysis of samples from ChIP experiments, showing specificity of the IP obtained with the I12 anti-HOXB4 monoclonal antibody. (A) Formaldehyde cross-linked chromatin was prepared from clone 4 of KLS-EML-HOXB4 cells, as described in supplemental Methods (ChIP-chip), and incubated with the I12 antihuman HOXB4 rat monoclonal antibody that had been previously bound to magnetic beads (Dynabeads) conjugated to antirat IgG2a antibody. The IP lane represents the material that remained bound to the magnetic beads after repeated washings. Eluates 1 to 8 represent the supernatants from the serial washings of the magnetically bound beads. The HOXB4 protein was specifically immunoprecipitated using the I12 monoclonal anti-HOXB4 antibody, as shown by the presence of the HOXB4 protein and the absence of actin in the IP lane. Some residual HOXB4 protein was observed in the first eluate, indicating that binding was not 100% efficient. (B) Using an antirat IgG2a isotype control antibody, HOXB4 protein is not observed in the IP lane, further documenting the specificity of the I12 monoclonal anti-HOXB4 antibody used for ChIP. Similar results were obtained when chromatin extracts from clones 11 and 17 of KLS-EML-HOXB4 cells were used.

Figure 6

Results of ChIP-chip and their validation by quantitative PCR. (A) ChIP-chip analysis using the I12 anti-HOXB4 monoclonal antibody revealed that there was increased binding to the promoter regions of the B220, CD34, Sox4, Laptm4b, Gp49a, and HMBS genes by the anti-HOXB4 IP, compared with input. In contrast, the control IP obtained using antirat IgG2a antibody beads alone showed no comparable increased binding. By microarray analysis, the Khdrbs1 gene was not shown to be differentially expressed in KLS-EML-HOXB4 cells, and no binding of HOXB4 to its promoter region was detected by ChIP; it is shown as a negative control. *The promoter region for the erythroid-specific isoform of HMBS, which is located within the sequence region of the first intron of the housekeeping isoform of HMBS. Interestingly, there is a greater degree of increased binding of HOXB4 to the promoter region for the erythroid-specific isoform of HMBS than to the promoter region for the housekeeping isoform of HMBS. Arrows represent transcription start sites of the genes. (B) Primers were designed corresponding to the promoter regions of the negative control Khdrbs1 gene and of 17 genes that were observed by ChIP-chip to be enriched at least 2-fold in the IP obtained with the I12 anti-HOXB4 monoclonal antibody. Validation of the enrichment was carried out by quantitative real-time PCR. The PCR results confirmed that 15 of the 17 selected promoter regions were indeed enriched in these IPs compared with the input. Results using the IPs obtained using the control beads conjugated to antirat IgG2a antibody are shown for comparison. Values are mean plus or minus SD of triplicate PCR analyses of ChIP DNA from the 3 different single-cell clones of KLS-EML-HOXB4 cells. (C) Top-scoring DNA-binding motif identified, using the software program Weeder (Version 1.3), in an unbiased analysis of the top 1000 promoter sequences bound by HOXB4 in the ChIP-chip experiments. The known HOX protein binding consensus sequence, TAAT, is found prominently in this motif.

Figure 7

Model of HOXB4-mediated HSC self-renewal. HOXB4 may promote self-renewal in HSCs by down-regulating the expression of lineage-specific genes and up-regulating “stemness-related” genes. As HSCs encounter microenvironmental changes that are conducive to differentiation, the binding partners of HOXB4 may change, resulting in a reversal of HOXB4 function and/or binding specificity, such that lineage-specific genes that were once down-regulated by HOXB4 directly or indirectly would no longer be down-regulated. In addition, “stemness-related” genes that were once up-regulated by HOXB4 would now be down-regulated as the cells differentiate.