Genetic fingerprint defines hematopoietic stem cell pool size and function

Hematopoietic stem cells (HSC) are at the apex of the hematopoietic tree, with self-renewal and multilineage differentiation potential. On the one hand, HSC can replenish the mature blood cells by differentiating into lineage-committed progenitor cells in response to the shortage of blood cells under both homeostatic and stressed conditions, such as bleeding and infection. On the other hand, HSC replicate themselves to maintain their number. This differentiation and self-renewal needs to be strictly regulated by gene expression regulation in order to maintain life-long hematopoiesis.1 Gene expression is in general regulated by “cis- and trans-regulatory elements”.2 Trans-regulatory elements are defined as factors which regulate expression of distal genes (e.g. transcription factors), while cis-regulatory elements are defined as non-coding DNA sequences which regulate expression of proximal genes (e.g. promoter and enhancer regions). These cis-regulatory elements play important roles in evolution in which polymorphisms occur in cis-regulatory elements and contribute to phenotypic diversity of organisms within well-conserved genes.3 Epigenetics plays important roles in HSC regulation, as epigenetic dysregulation in HSC is a key driver for HSC aging and hematopoietic malignancies.4 Epigenetic regulation is also controlled by cis- and trans-regulatory elements. For example, histone modifications function as trans-regulatory elements, whereas DNA methylations function as cis-regulatory elements.

Phenotypic diversity is frequently caused by genetic variations such as single nucleotide polymorphism (SNP). It was reported that the size and function of the HSC pool vary between mice strains,95 which suggests genetic background, such as SNP and copy number variations, define HSC homeostasis. In 2007, Liang et al. identified latexin (Lxn) as an HSC regulatory gene whose expression level is inversely correlated with HSC number.10 Although a variation in Lxn gene expression and HSC number in different tested mouse strains was shown, the mechanism underlying regulation of Lxn gene expression and its variation between mice strains remained unknown.

In this issue of Haematologica, the same group who published the above-mentioned paper,10 Zhang et al. identified the promoter region of the Lxn gene that controls the level of Lxn gene expression via both HMGB2, a chromatin protein, and genetic variations in the promoter region.11 To study the transcriptional regulation of Lxn, the authors characterized the upstream region of the Lxn gene. Based on the natural variation of Lxn expression, they searched SNP with CpG island and identified a region with strong promoter activity in the upstream. Subsequently, DNA pulldown in combination with mass spectrometry analysis were performed to identify proteins bound to this region, and HMGB2 was found to bind to the promoter region and to suppress Lxn gene promoter activity. To further confirm the regulatory role of HMGB2 in Lxn gene expression, Zhang et al. performed a gene knockdown experiment with EML cells, which share some of the characteristics of HSC.12 This showed that HMGB2 knockdown suppresses EML cell growth and that additional Lxn gene knockdown could rescue this growth, suggesting that Lxn was one of the transcriptional targets of HMGB2. Consistent with the previous reports concerning the phenotype of cells overexpressing Lxn,1413 the HMGB2 knockdown cells showed enhanced apoptosis and cell cycle arrest, which could in part be rescued by Lxn gene knockdown. These data suggest HMGB2 positively regulates HSC survival and proliferation by suppressing expression of Lxn and Lxn-regulated apoptosis. Similar data were also shown in LinSca-1c-Kit (LSK) cells primarily isolated from mouse bone marrow that contain HSC. HMGB2 knockdown in LSK cells showed suppressed proliferation, enhanced apoptosis and cell cycle arrest in vitro. In transplantation experiments, HMGB2 knockdown in LSK cells showed decreased reconstitution of whole peripheral blood cells and bone marrow LSK cells and long-term HSC in transplant recipients, indicating that HMGB2 plays an important role in HSC function in vivo.

The previous finding that Lxn expression level is correlated with HSC numbers10 led the authors to hypothesize that the SNP in the promoter region of the Lxn gene may contribute to a variation in Lxn expression and HSC number. To test this, the authors introduced G to C mutation in the HMGB2 binding sequence in the Lxn gene promoter region and found that the G allele showed higher promoter activity for Lxn expression compared to the C allele. Furthermore, when Lxn gene expression and bone marrow HSC number were analyzed in different mouse strains carrying the G or C allele in this SNP region, the mice strain carrying the C allele showed relatively lower Lxn protein expression and higher HSC number compared to those carrying the G allele. These data suggest that a genetic variant in the Lxn gene promoter region defines the variation in Lxn gene expression level and HSC number.

Together, Zhang et al. revealed that the transcription of Lxn regulating HSC function, at least as far as apoptosis is concerned, was controlled by both trans-regulatory element, HMGB2, and cis-regulatory element, as genetic variation was observed in the Lxn gene promoter region (Figure 1). HMGB2 is known as a chromatin-associated protein which remodels chromatin structure and gene expression.15 Although the molecular mechanism by which HMGB2 regulates Lxn gene transcription remains unclear, all the data provided in this study suggest Lxn gene transcription may be regulated through epigenetic modification. HMGB2 is also known to regulate senescence-associated gene expression by orchestrating the chromatin landscape of the gene loci.16 As the authors discussed, it would be interesting to study the role of Lxn gene regulation by HMGB2 in the context of HSC senescence and aging. The SNP identified in this study was in the CG-rich region. Although direct evidence was not shown, these data imply a functional role of the genetic variation, such as SNP, in Lxn gene regulation via DNA methylation. Taken together, this report suggests epigenetic regulation of HSC via Lxn gene transcriptional regulation. The molecular mechanism of how epigenetic regulation by HMGB2 protein and genetic variation in the Lxn gene promoter region work together needs to be better understood.

Figure 1.Cis- and trans-regulation of Lxn gene transcription modulates hematopoietic stem cell function. In trans-regulatory mechanisms of hematopoietic stem cell (HSC) function and number, HMGB2 protein binding to the Lxn gene promoter region positively regulates HSC by suppressing Lxn gene transcription and blocks apoptosis. In the cis-regulatory mechanism of HSC, the G allele in the SNP found in Lxn gene promoter region is associated with higher Lxn gene transcription and lower HSC number, whereas the C allele is associated with suppressed Lxn gene transcription and higher HSC number.

This report is the first demonstration that genetic variation, especially SNP, is a determinant for the variations among different mouse strains in HSC pool size and function. The authors also discussed a similar observation made in humans,11 opening the possibility that the genetic variation in the Lxn gene promoter region may contribute to the pathogenesis of hematopoietic aplasia/neoplasia, such as bone marrow failure or leukemia. If epigenetic regulation of Lxn gene transcription is involved in these hematopoietic diseases, it could be a new therapeutic target for genetic correction/modification such as genome editing. Since other genes that regulate HSC function, besides the Lxn gene, could undergo genetic variation through cis-regulatory elements, it would be interesting to characterize SNP in the regulatory region of HSC regulatory genes. This would help to develop therapeutic strategies for personalized medicine.

Acknowledgment

This work was supported by KAKENHI from Japan Society for the Promotion of Science (JSPS) to TM (19K17833), and KAKENHI (17H05651, 18H02843), Princess Takamatsu Cancer Research Fund, MIRAI Research program and Center for Metabolic Regulation of Healthy Aging at Kumamoto University to HT.