The aim of the project was the development and application of novel bioinformatic and statistical methods to analyze different sources of experimental data, produced with massive technologies, with a systems biology approach. At the starting phase of the project we have integrated our computational infrastructure with new components and built up a new and powerful computational facility for storage and analysis of big data sets. Several young researchers were recruited to work in the project.
The research has been focused on the characterization of the role of coding and non-coding RNAs on chronic myeloproliferative neoplasms (MPNs) and in particular in primary myelofibrosis (PMF) using microarray and RNA-seq data. The main goal was the discovery and characterization of small non-coding RNAs associated with MPNs pathogenesis and the reconstruction of regulatory networks in MPN, using matched mRNA and miRNA data.

In the first line of research:
The pipeline miR&moRe, for the discovery, characterization and quantification of small RNAs (Bortoluzzi et al. 2012), has been extended to the case of multiple samples. We obtained new small RNA-seq data of 6 CD34+ cells, including 3 samples collected from 3 pools of bone marrow CD34+ cells of healthy subjects (CTR), and 3 samples of circulating CD34+ cells of patients affected by primary myelofibrosis (PMF), that were deposited in Gene Expression Omnibus (GEO Series GSE69089). Small RNA expression in CD34+ cells of patients with PMF was characterized. We detected a total of 917 expressed sRNAs, including 784 know miRNAs and 133 new sRNAs. We discovered 34 new miRNAs produced from known hairpins; two of them, miR-2110* and miR-548ag- 2*, resulted highly expressed. We studied expressed isomiRs, since expressed miRNAs with unique sequence are very few (21%) and weakly expressed. A few mismatch isomiRs expressed in PMF samples and not in CTR samples, possibly due to genetic mutations, and of differentially expressed isomiRs were identified but resulted mostly weakly expressed and presented expression profiles highly correlated with those of miRNAs. We selected the most promising miRNAs for re-sequencing analysis of in one hundred samples of PMF-granulocytes, which did not show any mutations. Anyway, we considered isomiR counts for miRNA expression calculations and for target prediction.
We discovered 99 expressed microRNA-offset RNAs (moRNAs), probably produced by non-canonical processing of the hairpin precursors Of them 28 moRNAs were highly expressed, and in specific cases more expressed than the flanking miRNA. Notably, correlation values between expression profiles of moRNAs and miRNAs from the same hairpin arm indicated that moRNA expression is largely independent from that of the close miRNA.
sRNA differentially expressed in PMF vs CTR were identified and validated. Cluster analysis and heatmaps of pairwise sample correlations showed that sRNAs expression well separated CTR and PMF and highlighted a characteristic miRNA and moRNA expression profile in PMF. Indeed, we recognized 37 sRNAs significantly differentially expressed (DE), mostly upregulated, in PMF patients. RT-PCR experiments were conducted in PMF and CTR CD34+. miR-10b- 5p and moR-128- 2 were confirmed to be downregulated in PMF CD34+ cells, whereas miR-19b- 3p, miR-379- 5p and miR-543 resulted upregulated (Norfo et al. 2014, Guglielmelli et al. 2015). Additional data were obtained in granulocytes collected from an independent and sizeable cohort of normal controls (N = 10) and PMF (N = 50), PV (N = 30) or ET (N = 30) patients.
We studied more deeply 3’-moR- 128-2, a newly annotated small RNA that resulted expressed in normal CD34+ and downregulated in PMF. We investigated in detail sequence, structure, expression and possible functions of the newly discovered 3'-moR- 128-2, also by comparison with the cognate miR-128- 3p, showing that the miRNA and moRNA expression profiles are poorly correlated. At the functional level, assuming that moRNAs may act as miRNAs, we noticed that 3'-moR- 128-2 and miR-128- 3p potential target genes and pathways are markedly different. Interestingly, 3'-moR- 128-2, that was expressed in normal CD34+ cells and resulted absent in PMF cells, may target pathways related to the control of cell growth and proliferation and, strikingly, target several genes involved in microRNA biogenesis or in miRNA-mediated silencing.
A preliminary functional characterization of the possible biological role of sRNAs DE in PMF, was condicted by a double strategy. First we investigated possible target genes and pathways of the group of validated DE sRNAs, considered as a whole. Then, we focused on one of the most novel elements emerged by our results, 3’-moR- 128-2, to get specific insights on its possible functions in CD34+ and, in turn, in PMF disease. Predicted targets of differentially expressed miRNAs are enriched in many remarkable pathways involved in tumor development and progression, as “signaling by FGFR”, “DAP12 signaling” and “Oncogene Induced Senescence”. We then obtained a validation of 18 potential target genes for selected small RNA (moR-128- 2, miR-379- 5p, miR10b-5p, miR-19b- 3p, miR-29a- 3p and miR- 543) in an independent cohort of CD34+ cells from 20 PMF patients and 10 healthy subjects, using RT- PCR. According to miRNAs expression levels, and considering only significantly differentially expressed genes, an opposite behavior of expression variation was found for miR-379- 5p and MME, TCF4 and SYS1; miR-19b- 3p and miR-543 with TRPS1. Moreover, miR-10b- 5p and FSTL1 showed opposite behavior and 5 tested genes (AGO1, RAN, MECOM, MEIS1, and CAV1) putative targets of moR-128- 2 showed a trend toward increase in PMF, opposite to the moRNA.

In the second line of research, grounding on miRNA and gene expression data (GSE41812 and GSE53482) obtained by analyzing CD34+ cells of 42 patients with a diagnosis of PMF and 16 peripheral blood (CTR PB) and 15 bone marrow (CTR BM) samples from normal donors, integrative analysis of miRNA and gene expression patterns was used to identify and characterize regulatory networks involving miRNAs deregulated in PMF.
In a first study (Norfo et al., 2014) the differential expression of 50 and 34 miRNAs in PMF vs CTR CD34+ samples was detected and validated, and several biomarkers and putative molecular targets such as FGR, LCN2, and OLFM4 were identified. By means of miRNA-gene expression integrative analysis, we found different regulatory networks involved in the dysregulation of transcriptional control and chromatin remodeling. In particular, a network gathering several miRNAs with oncogenic potential (eg, miR-155- 5p) and targeted genes whose abnormal function has been previously associated with myeloid neoplasms, including JARID2, NR4A3, CDC42, and HMGB3 was highlighted. Because the validation of miRNA-target interactions unveiled JARID2/miR-155- 5p as the strongest relationship in the network, the function of this axis in normal and PMF CD34 +  cells was further investigated: JARID2 downregulation mediated by miR-155- 5p overexpression leads to increased in vitro formation of CD41 +  MK precursors. These findings suggested that overexpression of miR-155- 5p and the resulting downregulation of JARID2 may contribute to MK hyperplasia in PMF.
In a second study (Calura et al., under evaluation), we obtained a data-driven network model of primary myelofibrosis highlighting transcriptional and post-transcriptional alterations in CD34+ cells. By both knowledge-based and ab initio approaches for comparative analysis of CD34+ cells of PMF patients and healthy controls we identified the deregulated pathways involving miRNAs and genes and new transcriptional and post-transcriptional regulatory circuits in PMF cells. These converge in a unique and integrated cellular process, in which the role of specific miRNAs is to wire, co-regulate and allow a fine crosstalk between the involved processes. Even if the multistep approach we used for the PMF network reconstruction was specifically designed to go beyond the differential expression of single genes or miRNAs, giving a more informative picture of the impact of miRNAs deregulated in PMF to pathways and to mixed connected regulatory circuits it is worth note that we obtained confirmation of the differential expression in PMF samples compared to controls for about one half of the miRNAs included in the PMF network. In addition to the confirmation of differential expression in CD34+ previously obtained by array and RT-PCR (Norfo et al., 2014) and RNA-seq (Guglielmelli et al., 2015), we conducted also new experiments that confirmed a significant deregulation (p-value<0.05) of specific miRNAs in PMF patients granulocytes (50 PMF patients and 10 healthy controls) and plasma (25 PMF patients and 6 controls). This overview provides a global understanding of transcriptional and post- transcriptional deregulations in PMF, and, unifying consolidated and predicted data, could be helpful to identify new combinatorial therapeutic strategy.

Moreover, in order to clarify the contribution of miRNAs to the pathogenesis of JAK2V617F-positive MPNs, we analysed the expression of 365 miRNAs in erythroid (TER119+) and myeloid (GR1+) cells purified from BM of JAK2V617F knock-in (KI) mouse model (2 KI and 2 WT). We identified specific differentially expressed miRNAs shared by both erythroid and myeloid cells, suggesting a direct relationship between these miRNAs and JAK2V617F. Target prediction analysis identified pathways having functional relevance for MPN.

During the project the participants met in occasion of several progress report meetings in Padova and in Florence, participated to numerous national and International congresses. A Project Closing Meeeting was organized (February 3, 2016, Aula magna - Vallisneri Building) with invited speakers (Dr. Francesco Ferrari, IFOM, Milan “Chromatin architecture and transcription dynamics”, Oncoematology lab. Univ. of Padova “Transcription modules and genomic alterations associated to disease course in Juvenile Myelomonocytic Leukemia”, Dr. Paola Guglielmelli, Univ. of Florence “Non coding RNAs in Ph- negative chronic myeloproliferative neoplasms” and Dr. Manuela Ferracin, Univ. of Bologna “Applications of microRNAs in the diagnosis and prognosis of cancer”).

FINAL EVALUATION AND OUTLOOK

Overall, the project provided novel data, knowledge and methods.
New data were obtained regarding the expression profile of small RNA expressed in PMF CD34+ as well as in healthy control cells, by NGS analysis and their possible impact on specific genes and pathways; in addition, we described for the first time new moRNAs as possible contributors to disease pathogenesis. This information may represent the basis for further studies aimed at a deeper knowledge of the role on miRNAs and moRNAs in normal and pathological hematopoiesis. The integrative analysis of miRNA and gene expression patterns was used to identify and characterize regulatory networks involving miRNAs deregulated in PMF. We provided a significant increase of data on miRNAs deregulated in PMF and particularly of our understanding of their participation to specific regulatory axes, their impact on pathways deregulation and their involvement in mixed transcriptional and post-transcriptional regulatory circuits. Moreover, in relation to project aims, several important new methods, tools and services have been developed.
The project has been really important to carry out the planned research and to open novel, also international, collaborations, to have access to new funding opportunities and to start innovative lines of research related to the project, thus extending the project scope. Moreover, it has permitted the formation and qualification of young scientists that are now contributing, more and more in an independent way, to research in the field. We are continuing as well the methodological development in the field of transcriptome characterization by RNA-seq data analysis and of systems biology and integrative genomics methods applied to understand the impact on long and short regulatory RNAs in gene networks and pathways.