SALL4 is a key transcription regulator in normal human hematopoiesis

Abstract

Background: Stem cell factor SALL4 is a zinc finger transcription factor. It plays vital roles in the maintenance of embryonic stem cell properties, functions as an oncogene in leukemia, and has been recently proposed to use for cord blood expansion. The mechanism(s) by which SALL4 functions in normal human hematopoiesis, including identification of its target genes, still need to be explored.

Study design and methods: Chromatin immunoprecipitation followed by microarray hybridization (ChIP-chip) was used for mapping SALL4 global gene targets in normal primary CD34+ cells. The results were then correlated with SALL4 functional studies in the CD34+ cells.

Results: More than 1000 potential SALL4 downstream target genes have been identified, and validation of binding by ChIP-quantitative polymerase chain reaction was performed for 5% of potential targets. These include genes that are involving in hematopoietic differentiation and self-renewal, such as HOXA9, RUNX1, CD34, and PTEN. Down regulation of SALL4 expression using small-hairpin RNA in these cells led to decreased in vitro myeloid colony-forming abilities and impaired in vivo engraftment. Furthermore, HOXA9 was identified to be a major SALL4 target in normal human hematopoiesis and the loss of either SALL4 or HOXA9 expression in CD34+ cells shared a similar phenotype.

Conclusion: Taken together, SALL4 is a key regulator in normal human hematopoiesis and the mechanism of its function is at least in part through the HOXA9. Future study will determine whether modulating the SALL4/HOXA9 pathway can be used in cellular therapy such as cord blood expansion and/or myeloid engraftment.

Figure 1. SALL4 expression in normal human hematopoietic cells

(A) SALL4 protein expression in normal CD34+ and CD34− cells examined by confocal immunofluorescent staining. Normal bone marrow cells were sorted into CD34+ and CD34− fractions. SALL4 expression is in green, and the nuclear staining by TO-PRO-3 is in blue. Merged images showed that nuclear expression of SALL4 was only detected in the CD34+ HSC/HPCs, but not in the CD34− cells. (B) SALL4 RNA expression in normal CD34+ and CD34− cells examined by qRT-PCR. CD34+ cells were sorted into CD34+CD38−, CD34+CD38+, and CD34-fractions. Quantification of SALL4A and B mRNA levels normalized to GAPDH, and showed that both SALL4 isoforms were expressed preferentially in CD34+CD38− HSCs, but down-regulated in normal CD34+CD38+ HPCs, and not present in the CD34− cells (N=5 Error Bars: +/− SE). Y-axis: SALL4A or B relative expression.

Figure 2. Identification of potential SALL4 direct target genes in normal CD34+ cells by genome-wide chromatin immunoprecipitation (ChIP)-chip

(A) Validation of ChIP-Chip results. 16 genes known to play a role in hematopoiesis with various ChIP-chip score were chosen for validation, with CCNA1 as a negative control. Fold enrichment of above 1.5 compared to input (no immunoprecipitation added) after GAPDH normalization was considered positive. ChIP-qPCR was performed with two biological replicates with each biological sample being done as triplicates. The results shown were the average data. (B) Gene ontology categories of SALL4 bound targets in CD34+ cells. A total of 1160 genes were found to be bound by SALL4 in CD34+ HPC/HSCs through ChIP-chip genome-wide analysis. Protein Analysis Through Evolutionary Relationships (PANTHER), a web-based software, was used to analyze the dataset. (C) GO function categories as defined by Ingenuity. The two percentages behind the gene numbers represent: in green: genes in this particular category bound by SALL4 relative to the total number of SALL4-bound gene in CD34+ cells; in red, total number of genes defined by this category relative to total number of genes represented on the array.

Figure 3. comparison of SALL4 targets in normal CD34+ cells versus leukemic cells

(A) NB4 AML cell line and CD34+ cells have 258 SALL4-bound targets in common (CD34+ cells = 18.2%; leukemic cells = 9.8%). SALL4− bound genes in (B) P53 pathways, (C) NF-kB pathway, (D) Wnt/beta-catenin pathway, (E) FGF pathway and (F) JAK/STAT pathways in NB4 and CD34+ cells.

Figure 4. Down-regulation of SALL4 in normal CD34+ cells affects genes involved in differentiation and self-renewal/proliferation

(A) The abilities of the two shRNA constructs to knock down SALL4 in CD34+ cells were confirmed by qRT- PCR with 25–35% knockdown efficiency compared to the scrambled pRS control vectors (N=3, Error Bars: +/− SD). (B) Gene expression changes upon the down-regulation of SALL4 for HSC/HPC genes CD34, and RUNX1, HOXA9, GATA-1, TPO and PTEN. (C) Gene expression changes upon SALL4 downregulation for proliferation, anti- and pro-apoptosis genes ABL1, TRO, BCL2, FYN, ATF3, MAP3K12, p53, CUL3 and TNF. Fold enrichment was calculated after normalization of both control and SALL4 knock down samples to GAPDH (N>3, Error Bars: +/− SD). * P<0.05

Figure 5. Impaired colony forming abilities in vitro and engraftment in vivo upon SALL4 knock down

Colony forming unit (CFU) assays were performed on CD34+ cells infected with control vector and SALL4 shRNA retroviruses that were then plated on methylcellulose supplemented with myeloid-differentiation inducing cytokines for 7–14 days. (A)The total number of all colonies when CD34+ cells were infected with SALL4 shRNA, and the number and size of CFU-GM (B&C) in SALL4-reduced group when compared to those that were infected with the control vectors. 5 × 10⁵ of human CD34+ BM cells infected with either SALL4 shRNA or control shRNA retroviruses were transplanted into sublethally irradiated (1.0 Gy) NOD/SCID/IL2rγ-null mice by tail vein injection. Peripheral blood (PB) (D) and bone marrow (BM) (E) Chimerisms were monitored at 8 weeks post transplantation. CD34 cells with reduced SALL4 expression (n=10) had statistically lower engraftment than control cells (n=7). P-values are listed as indicated.

Figure 6. SALL4 and HOXA9 shares similar effects on myeloid differentiation in human primary CD34+ cells

Purified CD34+ cells were either infected with scrambled shRNA or shRNAs against SALL4, HOXA9, then cultured in methylcellulose medium supplemented with SCF, IL-3, GM-CSF and EPO, with 1ug/ml puromycin for selection. (A) Down-regulation of SALL4 or HOXA9 mRNA expression level to 35–40% of those treated with scrambled controls (N=3, Error Bars: +/− SD), evaluated by qRT-PCR. T (B) Total colony number and (C) myeloid colony number on CFU plates were scored on day 14. The numbers of colonies formed by scrambled viral infected cells is set at 100%. Data depict average and standard deviation of 3 independent experiments. The P value was obtained by comparing to the control using a paired two-tailed distribution t-test **p<0.01. (D) Flow cytometric analysis of myeloid differentiation of cells harvested from the CFU plates. The level of CD11b expression was plotted on histograms in comparison to the control. (E) Morphology of CFU colonies after treatment with scrambled control and SALL4 or HOA9 knock down (10x).