Final Report Summary - BLOODY SIGNALS (Multidimensional analysis of signaling during normal hematopoietic differentiation and stress-induced regeneration)
Hematopoietic stem cells (HSCs) are the foundation of all the differentiated hematopoietic lineages in an organism. During situations like myeloablation, massive bleeding or disease, the organism needs to quickly adapt and produce hematopoietic cells in a regenerative or simply stress response. Steady state and stress hematopoiesis are both tightly regulated by intrinsic transcription factors (TFs) like lineage regulators and external signals interpreted by several signal-responsive TFs. One of the main goals of this proposal was to dissect the interplay between lineage regulators and signaling factors in hematopoietic differentiation.
We worked on two major hematopoietic differentiation branches: the erythroid and the myeloid differentiation. By in vitro differentiating human progenitor hematopoietic cells towards erythroblasts, we were able to dissect step-by-step human erythropoiesis. To understand the interplay between lineage and signaling regulators we used as exemplary lineage factors GATA2, that is mainly expressed in progenitor cells and GATA1, which is a master regulator of erythropoiesis. As an exemplary signaling factor, we study SMAD1, since SMADs and BMP signaling have been implicated in erythropoiesis and specifically in stress erythropoiesis. Using a panel of genome-wide techniques we managed to dissect the chromatin landscape of human erythroid differentiation. First, we performed chromatin immunoprecipitation followed by sequencing for these three TFs. We discovered that GATA2 binding is diminishing as cells progress towards erythroblasts. On the other hand, GATA1 binding occurred as the cells had committed to the erythroid fate. Interestingly, the binding of SMAD1 was dynamic across the whole differentiation time-course. By comparing the binding of lineage regulators with SMAD1 binding, we show that most of the SMAD bound regions correlated with GATA bound regions. We then wanted to investigate these co-bound regions. To this end, we first examined the genome-wide chromatin accessibility changes throughout the differentiation time course as well as the binding of H3K27ac, a marker of active enhancers. Indeed, we found that co-bound GATA-SMAD regions are preferentially located in open chromatin regions, enriched for H3K27ac, namely active enhancer regions. Chromatin immunoprecipitation analysis of the binding of other signaling TFs showed that they are also localized in the genomic regions bound by both GATA and SMAD. These data led us to hypothesize that the co-bound regions would be crucial for gene expression. To investigate this possibility we performed expression analysis at various time points during erythroid differentiation. We then assigned the co-bound and the GATA-only bound peaks to genes and compared them with the expression data. Indeed, co-bound regions were regulating stage specific genes that were important for erythroid differentiation in contrast to GATA-only bound regions. These results show that regions bound by lineage and signal responsive TFs are located in genomic “hotspots” that regulate the expression of stage specific genes. To understand the importance of signal responsive and lineage regulators in these “hotspots” we used CRISPR-Cas9 technology to mutate or delete specific TF motifs. This experiment clearly showed that deletion of a signal responsive factor in a respective enhancer greatly affects the expression of the corresponding gene. Collectively these data show that by identifying the genomic regions where a single lineage regulator and a single signaling TF bind, we can identify novel genes that are crucial for various stages of hematopoietic differentiation. These data will be a useful resource for the community and could be used to identify the cause of various genetic traits that do not lie in the coding regions of the genome.
Our next interest lied in the delineation of myeloid differentiation and especially the function of HLX (H2.0-Like Homeobox) TF in this process. HLX is a non-clustered homeobox transcription factor implicated in several developmental and physiological processes but also in disease, such as Acute Myeloid Leukemia (AML) where it is commonly overexpressed. Since developmental pathways are often reactivated in cancer, we hypothesized that HLX plays a critical role during developmental hematopoiesis in zebrafish. We showed that endothelial-specific overexpression of human HLX causes aberrant production of hematopoietic stem and precursor cells (HSPCs) and increased numbers of immature myeloid cells at 48 hpf (hours post fertilization), thereby mimicking a pre-leukemic phenotype. Conversely, embryos with hlx1 morpholino-mediated knockdown had reduced numbers of HSPCs. Expression analysis showed that HLX overexpression resulted in downregulation of mitochondrial electron transport chain (ETC) enzymes and upregulation of the peroxisome proliferator activated receptor family member δ (PPARδ), with concomitant changes in various metabolic parameters. These results suggest that HLX regulates the fate of hematopoietic cells by fine-tuning their metabolic status. Chromatin immunoprecipitation in human cell lines revealed that ETC genes and PPARδ are directly regulated by HLX. Importantly, metabolic and molecular changes observed in zebrafish were recapitulated in human cell lines and primary HSPCs (hematopoietic stem and progenitor cells). Moreover, AMPK signalling and autophagy were increased in HLX-overexpressing human cells. Modulation of PPARδ signalling rescued the hematopoietic phenotypes of both zebrafish and human HSPCs but had no effect on cancerous cell lines. However, inhibiting AMPK signalling induced death of AML cells with minimal effects in primary HSPCs. Thus, our findings show that HLX controls hematopoietic differentiation by modulating the cell metabolic status. This knowledge could be used for treatment of AML patient that HLX is highly expressed.
Finally, we also progressed our work on the interplay between inflammatory and developmental signalling factors in HSC development.
Taken together our research led to significant results that helped our understanding on the genomic binding of TFs and the cooperation between them that regulates fundamental hematopoietic processes. This knowledge is fundamental for the community. Additionally, it created a solid foundation for our current and future research.
Trompouki lab website: https://www.ie-freiburg.mpg.de/trompouki
We worked on two major hematopoietic differentiation branches: the erythroid and the myeloid differentiation. By in vitro differentiating human progenitor hematopoietic cells towards erythroblasts, we were able to dissect step-by-step human erythropoiesis. To understand the interplay between lineage and signaling regulators we used as exemplary lineage factors GATA2, that is mainly expressed in progenitor cells and GATA1, which is a master regulator of erythropoiesis. As an exemplary signaling factor, we study SMAD1, since SMADs and BMP signaling have been implicated in erythropoiesis and specifically in stress erythropoiesis. Using a panel of genome-wide techniques we managed to dissect the chromatin landscape of human erythroid differentiation. First, we performed chromatin immunoprecipitation followed by sequencing for these three TFs. We discovered that GATA2 binding is diminishing as cells progress towards erythroblasts. On the other hand, GATA1 binding occurred as the cells had committed to the erythroid fate. Interestingly, the binding of SMAD1 was dynamic across the whole differentiation time-course. By comparing the binding of lineage regulators with SMAD1 binding, we show that most of the SMAD bound regions correlated with GATA bound regions. We then wanted to investigate these co-bound regions. To this end, we first examined the genome-wide chromatin accessibility changes throughout the differentiation time course as well as the binding of H3K27ac, a marker of active enhancers. Indeed, we found that co-bound GATA-SMAD regions are preferentially located in open chromatin regions, enriched for H3K27ac, namely active enhancer regions. Chromatin immunoprecipitation analysis of the binding of other signaling TFs showed that they are also localized in the genomic regions bound by both GATA and SMAD. These data led us to hypothesize that the co-bound regions would be crucial for gene expression. To investigate this possibility we performed expression analysis at various time points during erythroid differentiation. We then assigned the co-bound and the GATA-only bound peaks to genes and compared them with the expression data. Indeed, co-bound regions were regulating stage specific genes that were important for erythroid differentiation in contrast to GATA-only bound regions. These results show that regions bound by lineage and signal responsive TFs are located in genomic “hotspots” that regulate the expression of stage specific genes. To understand the importance of signal responsive and lineage regulators in these “hotspots” we used CRISPR-Cas9 technology to mutate or delete specific TF motifs. This experiment clearly showed that deletion of a signal responsive factor in a respective enhancer greatly affects the expression of the corresponding gene. Collectively these data show that by identifying the genomic regions where a single lineage regulator and a single signaling TF bind, we can identify novel genes that are crucial for various stages of hematopoietic differentiation. These data will be a useful resource for the community and could be used to identify the cause of various genetic traits that do not lie in the coding regions of the genome.
Our next interest lied in the delineation of myeloid differentiation and especially the function of HLX (H2.0-Like Homeobox) TF in this process. HLX is a non-clustered homeobox transcription factor implicated in several developmental and physiological processes but also in disease, such as Acute Myeloid Leukemia (AML) where it is commonly overexpressed. Since developmental pathways are often reactivated in cancer, we hypothesized that HLX plays a critical role during developmental hematopoiesis in zebrafish. We showed that endothelial-specific overexpression of human HLX causes aberrant production of hematopoietic stem and precursor cells (HSPCs) and increased numbers of immature myeloid cells at 48 hpf (hours post fertilization), thereby mimicking a pre-leukemic phenotype. Conversely, embryos with hlx1 morpholino-mediated knockdown had reduced numbers of HSPCs. Expression analysis showed that HLX overexpression resulted in downregulation of mitochondrial electron transport chain (ETC) enzymes and upregulation of the peroxisome proliferator activated receptor family member δ (PPARδ), with concomitant changes in various metabolic parameters. These results suggest that HLX regulates the fate of hematopoietic cells by fine-tuning their metabolic status. Chromatin immunoprecipitation in human cell lines revealed that ETC genes and PPARδ are directly regulated by HLX. Importantly, metabolic and molecular changes observed in zebrafish were recapitulated in human cell lines and primary HSPCs (hematopoietic stem and progenitor cells). Moreover, AMPK signalling and autophagy were increased in HLX-overexpressing human cells. Modulation of PPARδ signalling rescued the hematopoietic phenotypes of both zebrafish and human HSPCs but had no effect on cancerous cell lines. However, inhibiting AMPK signalling induced death of AML cells with minimal effects in primary HSPCs. Thus, our findings show that HLX controls hematopoietic differentiation by modulating the cell metabolic status. This knowledge could be used for treatment of AML patient that HLX is highly expressed.
Finally, we also progressed our work on the interplay between inflammatory and developmental signalling factors in HSC development.
Taken together our research led to significant results that helped our understanding on the genomic binding of TFs and the cooperation between them that regulates fundamental hematopoietic processes. This knowledge is fundamental for the community. Additionally, it created a solid foundation for our current and future research.
Trompouki lab website: https://www.ie-freiburg.mpg.de/trompouki