PowerPoint Presentation

[Audio] Good afternoon everyone, my name is Itziar and I am doing my PHD in Felipe´s lab with Borja Saez our main goal is to deciphering the cellular complexity of the mouse and human bone marrow microenvironment and its implication in aging and disease..

[Audio] It is accepted that neoplastic disorders are diseases of the tissue, so the study of their parenchymal composition and identification of the interactions established between the different cell types are necessary to fully comprehend the biology and pathogenesis of each disease.

[Audio] Hematopoietic stem cells reside in a complex specialized microenvironment which is known as the niche. The niche is a dynamic place that provides signals for the control of HSC maintenance and localization. Within the niche endothelial and mesenchymal Cells represent two well defined populations involved in the regulation of hematopoiesis, Due to the importance of the niche in the maintenance of hematopoietic homeostasis, we believe that the bone marrow microenvironment may represent an Achilles heel in the study and treatment hematologic malignancies. When we refer to the niche it's important to take in mind that this non hematopoietic cells represent a very low fraction of the whole bone marrow..

[Audio] To understand better the regulation of hematopoiesis, it is essential to generate a complete taxonomy of the BM microenvironment. So, our first objective was to describe the cellular heterogeneity and transcriptional regulation of the bone marrow microenvironment in both human and mouse in homeostasis. To achieve this goal, we have used single cell RNA sequencing because this technology allow us to identify new cellular subtypes and to delineate the gene regulatory networks that define each cell type in the bone marrow..

[Audio] Nevertheless, these single cell studies are limited by the number of cells sequenced, so for further study the niche we decided to integrate three scRNA-seq datasets. Two of them were public available and the other one was generated for us in our lab. We focus our study targeting these two niche populations: endothelial and mesenchymal cells. The integration of these data sets required the developing of this bioinformatics pipeline that was developed by pur collaborators in KAUST. As it was necessary to discriminate among similar cell coming from the same origin we customized the Louvain high resolution clustering approach and after grouping cell into clusters we applied a bootstrapping-based methodology to ensure the robustness and quality of each cluster..

[Audio] We characterize the different clusters based on the specific expression of markers and gene sets analysis, Thanks to this integration strategy we have been able to identify and characterize for the first time multiple cellular states that define different functions and trajectories of differentiation in the endothelial and mesenchymal compartments. These results demonstrate the plasticity and the heterogeneity within the bone marrow niche populations..

[Audio] Then, we used an independent single cell dataset to validate the performance of our integration method. To that end, we applied our clusters' signature to annotate Baccin dataset using SingleR. And despite the lower number of cells that are obtained in an individual dataset we were able to identify many cellular states described on our data. So, these results demonstrate the usefulness of our integration approach for characterizing functional clusters within the bone marrow populations..

[Audio] Regarding the human analysis, we sequenced four bone maroow samples from aspirates of healthy human donor. While we were able to identify niche populations, our data do not present enough resolution to describe the plasticity of the cells due to the high level of hematopoietic contamination. Therefore, we decided to use the knowledge that was generated in mice to characterize the human marrow. Thanks to this analysis We were able to identify the functional cell states that were previously described in mice..

[Audio] The conservation of these cellular states was also sustained by the identification of an enrichment of this clusters in the human dara and the expression of cluster-defining genes that were shared among species. And finally to further investigate the regulation of hematopoiesis we analyze the expression of known factors that are key regulators for hematopoietic stem cells that were also present in mice. As shown in the heatmap, the coexpression of these key factors suggests that the layers of microenvironmental regulation of hematopoiesis may also be shared between species..

[Audio] After characterizing the situation in homeostasis, our next objective is to describe the cellular microenvironment during aging due to the known association of aging and malignant transformation. Specifically we want to investigate if niche cells contribute to the defects and changes that are observed in hematopoiesis in old people. To this end we are sequencing samples obtained from healthy elderly people thanks to a collaboration that we have stablished with the hospital of Tudela. These samples comes from knee of hip orthopedic surgeries while the ones from the young donors are obtained from iliac crest aspirates..

[Audio] As shown in the slide, these samples present higher frequency of niche cells and less hematopoietic contamination in comparison with the ones coming from the aspirates. Furthermore, although these are only preliminary results, we have observed a different composition of the niche cellular cluster in elderly samples compared to the distribution observed in young samples. For examples Regarding the endothelium we observe more arterial cells, although these modifications are remarkably higher in mesenchymal cells where we observe a bias in the differentiation towards adipogenic lineage in detritment of the osteolineage Cells. But we need to analyze more samples and, in more depth.

[Audio] Finally, as I mentioned at the beginning of this talk our main goal wad deciphering the cellular and molecular landscape of the bone marrow microenvironment in order to discover its implication in neoplastic development. The importance of the niche cells in hematologic malignancies has been already demonstrated and this knowledge is growing in the last years. However, whether the niche provides the conditions for mutant clones to arise or whether the neoplastic cells are the ones that remodel the niche in their benefit is still unclear. In this case my project is focused in the study of the microenvironment in myelodysplastic syndromes. So, we had stablished another collaboration with the University of Pavia in order to describe the transcriptome of the microenvironment in a specific cohort of patients diagnosed with MDS splicing mutations..

[Audio] However Due to the impossibility of obtaining a large volume of bone marrow from these oncologic patients, we are going to sequence these samples by bulkRNAseq instead of single cell doing a prospective isolation of both endothelial and mesenchymal cells. Therefore, we must take into account a number of considerations before to get into the analysis..

[Audio] The most important aspect and disadvantage of our project is the low number of cells that the niche represents within the whole bone marrow, since we are talking about numbers below 0,1%. This limits both the efficiency isolation and the future sequencing of these samples. For this reason we have made a pilot study of different bulkRNA sequencing methods in order to determine which is the most suitable for our low cell input. On this regard, we isolated 1000 and 1500 cells from the same population in 3 samples that were sequenced by 3 different bulkRNA protocols: messenger takara, total takara and MARSseq..

[Audio] After performing different analyses for checking sequencing quality, correlation between samples, marker identification and power of differential expression analysis, Takara's full length messengerRNA protocol was the most successful in obtaining information when sequencing low number of cells, so it will be the one we will use to sequence these samples from patients with MDS..

[Audio] Finally As I mentioned during the talk, thanks to the single cells from human donors, we realized that in the bone marrow aspirate samples, despite of the positive isolation of niche cells, the percentage of hematopoietic contamination is still very high. This aspect is very important in the interpretation of the results because if it is not considered, the differences observed may not be associated with niche cells but with hematopoietic cells. This bias in the results could be solve by deconvolution methods. These computational tools, as CIBERSORTx, allow to infer the percentage of each of the populations that compose a heterogeneous sample thanks to a signature matrix that specifies the transcriptional identity of each those populations in that sample..

[Audio] To test the usefulness of this tool in our data, we applied the cell-specific profiles obtained in our single data into bulk- RNAseq samples. As can be seen in the slide, the isolation of mesenchymal cells is very pure with a very residual contamination whereas in the case of endothelial samples, the actual frequency of endothelial cells is around 10- 20% and we need to be very careful with these samples. But, the main adventage of this computational package is not only to know the real fraction, but also to infer the gene expression of each of the populations and compare it between samples. Therefore, in our case it will give insight into the differential molecular landscape of the niche cells in MDS..

[Audio] to sum up We have demonstrated the usefulness of the multi-dataset integration and the clustering approach used in our study to improve the resolution cellular complexity of the bone marrow microenvironment in both human and mouse. Our study dissects the intrinsic organization and the functional heterogeneity within the endothelial and mesenchymal cell populations governing the BM suggesting that microenvironmental regulation of hematopoiesis may be shared between species. Our preliminary results on elderly human samples show that aging of the hematopoietic stem cell niche is accompanied by a different composition of microenvironment cellular states. The Deconvolution analysis revealed a high percentage of contamination on bulk- RNAseq EC human samples compared to MSC samples. mRNA Takara full length protocol revealed the best performance with low cell numbers and will be used for sequencing niche cells in bulkRNAseq samples..

[Audio] Finally, I would like to thank to Felipe and Borja for giving me the opportunity of doing this work. Also, to my group colleagues Isabel, Ana Cris and Miguel and the rest of the people in the lab for their support and assistance. Last but not least, I would like to especially thank David and the rest of his group for making possible my internship at KAUST university. Thank you for your attention and now I would be glad to answer any of your questions..