Skip to main content

Role of transcription factor Hlf in hematopoietic stem cells

Final Report Summary - HLF IN HSC (Role of transcription factor Hlf in hematopoietic stem cells)

Report for CiG grant entitled “HLF in HSCs” #618647
Our body is not the same as it used to be. It generates new cells every day and regenerate following insults. This is possible thanks to special cells, namely adult stem cells, which retain the developmental potential to make new cells throughout life. Hematopoietic Stem Cells (HSCs) are the adult stem cells responsible for making all blood and immune cells. HSCs enable the lifesaving Bone-Marrow transplantation (also called HSC-transplantation, HSCT). Surprisingly, however, we do not yet understand how HSCs are keeping their potency. The various cell types in our body are all generated from one cell, and differ of each other by the genes that they express. With extensive information of gene expression in HSCs and in other blood and immune cells we had identified some genes that are most preferentially expressed in these adult stem cells and not in their progeny. One of these genes is called Hlf.
Hepatic Leukemia Factor (Hlf) is a PAR-bZIP transcription factor mostly known and studied in Acute B-cell leukemia with E2A-HLF fusion oncogene. However, there is not much study of its endogenous role, despite the clear interest. Hlf is profoundly expressed in HSCs and not in differentiated hematopoietic cells. It is also expressed in other organs such as the liver and brain, but not extensively studied there too. We had published of the findings of Hlf as one of the top HSC-specific transcription factors, and got interested in its specific study as it is regulating the expression of many other genes, thus may contribute to the identity of HSCs. Functional abilities of Hlf to induce extended colony-formation and partial HSC-like phenotype was published as part of our study of the ImmGen consortium database, in which I was fortunate to take part (Gazit R, Garrison B et al., 2013). This was not the first publication to suggest Hlf role in HSCs, as an earlier study by Shojaei and Batiha had already pointed the human HLF, together with HES1, as conadidate positive-regulators of human HSCs. Both studies suggested functional potency but no further details of how Hlf may act. Strikingly, we had also included Hlf as part of a short list of 36 factors for direct reprogramming of blood cells into HSCs. It turned to be one of the key 6 core factors that have the ability to turn committed progenitors directly into functional HSCs by their transient expression, as we published with colleagues (Riddel J, Gazit R et al., 2014). Importantly, this publication included extensive functional experiments but little mechanistic understanding for how few factors can turn one cell type into another. Attempts to translate our proof-of-concept for direct reprogramming into HSCs in human cells did not report robust success so far, suggesting that we better gain more information towards rational design of similar protocol in human.
We aimed to generate a specific mouse model for the study of Hlf in vivo, and this project had good start with construction of a targeting vector that includes disruption of the Hlf gene together with insertion of a reporter fluorescent gene (Knock-in knock-out strategy). However, in order to avoid possible problematic interpretation of data, we performed numerous controls, including specific assessment of the activity of the targeted-Hlf in cells. To our disappointment, we found that even the minute changes made of this gene in our construct had disabled its robust activity in primary hematopoietic cells. Thus we set this approach aside, to avoid possible abnormalities, and retreated for an alternative approach. Using the recently developed technique of genomic editing by CRISPR/Cas9 we had targeted the endogenous locus of the Hlf gene in the mouse. We obtained a positive hit, with a specific disruption on the second exon of Hlf, predicted to stop production of the full-length protein. We are now on the process of expanding the colony of mice and study the physiological role of Hlf in HSCs in this novel model that we generated.
In order to better understand Hlf in HSCs, and essentially any other gene of interest, we had in parallel advanced our research of gene expression for the new technology of RNA-sequencing. While our previous study of HSCs’ genes using microarray technologies is well established, the emerging RNA-Seq further enables novel discoveries. In-depth analysis of data enabled us to map the whole transcriptome splicing of mouse HSCs for the first time. Most genes, in mouse like human, do express by more than one variant- increasing the complexity of our transcriptome. This study, by Goldstein O et al., (2017) presented the broad information with possible resolution down to single-base, in a way accessible to all scientists to explore. Hlf was found in this study to consist of 2 main variants in HSCs. We had functionally tested both of these Hlf transcripts, and revealed that only the long-form has the robust activity in primary hematopoietic cells, while the short form does not. Intriguingly, in the abovementioned CRISPR-targeted mouse, we realize that the disruption is predicted to interfere with Hlf-long and possibly not with Hlf-short, making it all of the more interesting.
Our interest in direct reprogramming into HSCs is having Hlf as one of the key factors we study. Elaborated experiments with these factors in hematopoietic cells produced data for Hlf along with other factors. The robust functional activity of Hlf in cultured cells had been robustly reproducible in multiple experiments, confirming its extreme ability to endow cells with growth advantage in vitro. This is of great relevance for its possible role in malignancy, which is already known of its oncogenic translocations but less so for the endogenous gene itself. We had analyzed the transcriptome wide impact of Hlf, finding it to induce more genes than any of the other reprogramming factors, except MycN. The list of Hlf-targets consists of multiple interesting genes that we keep on studying. We had surprisingly identified that Hlf is also downregulating some well-known HSC genes, including cKit and Sca1, and validated their suppression at the surface-expression level. This may suggest the inability of Hlf to confer robust advantage in cells in vivo, as both cKit and Sca1 are major surface receptors of HSCs. Hence, Hlf on its own is inducing part of the HSC-program but at the same time it also suppresses another portion. We had progressed in demonstrating cooperative activity. We had also identified molecularly that Hlf is turning the cells to become less reliance on external stimulation of cytokines for their growth. Thus suggesting another hazard of tis potential oncogenic role, and highlighting its possible targeting in cancer. To complete this section, we examined the direct targets of Hlf on the genome using Chromatin-immunoprecipitation (ChIP). This revealed not many genes, but several of these are of known significance in HSCs; the binding MOTIF from our data is partially overlapping with predicted one for HLF, and presents precise targets in hematopoietic cells.
In summary, our research during this grant period yielded several findings on the specific role of Hlf in HSCs, highlighting molecular mechanisms by which this potent factor function within hematopoietic cells. It further imply for direct reprogramming into HSCs, and to blood cancers.