The first aim of the project was to develop a powerful platform for gene editing in primary human dermal fibroblasts (HDFs). I used at least 3 different donors of HDF, contemplating both female and male, and an age range between 22-55 years old. This was an important experimental decision to account for donor diversity and to increase the robustness of the results. I established HDFs constitutively expressing the Cas9 protein, as well as a resistance cassette for the antibiotic blasticidin for the selection of Cas9-only cells. I assessed the appropriate Cas9 dose by nock-out (KO) activity in the different donors, ensuring the as well the cell viability. As for the CRISPR/Cas9 sgRNA library, I designed a custom-made, pooled sgRNA library targeting several candidate chromatin remodelers, as well as RNA modifiers and splicing factors (SF), called KORE library (Knock-Out RNA and Epigenetic regulators). The list of genes comprises 1,367 genes, including controls, selected from published databases and from collaborators. For each gene, 5 targeting sgRNAs were retrieved and non-targeting guides were also included as controls to evaluate noise and screening success. In total, the KORE library comprises a total of 7,400 guides. These were cloned as a pool into a lentiviral vector that contains a chimeric guide RNA backbone and a GFP cassette as a reporter. To ensure that each cell receives only one sgRNA construct, I tittered the lentivirus particles to achieve a functional MOI of approximately 0.3-0.4. Cas9-HDFs from all donors were transduced with the KORE library and later selected for GFP-positive cells by fluorescence-activated cytometry sorting (FACS) to enrich the cell population to use in future experiments.
After the incorporation of the KORE library and gene KO, I performed cell reprogramming experiments to generate representative cells with different degrees of plasticity from the engineered HDF cells. I followed specific protocols to obtain iDC1, iHSPC and iPSC using polycistronic vectors, all factors inducing a certain cell fate in a single vector. By ectopic expression of the human TFs PU.1 IRF8, BATF3 (PIB), I achieved direct conversion to iDC1, an example of unipotent cells. In the case of iHSPC, the expression of GATA2, GFI1B, and FOS is sufficient to obtain multipotent cells committed to the hemogenic fate. Moreover, the enforced expression of OCT4, SOX2, KLF4, and c-MYC TFs instructs reprogramming to iPSC. Once reprogramming was completed for each protocol (iDC1: day 9, iHSPC: day 15, iPSC: day 20), the endpoint bulk cell population was classified into different subpopulations of non-reprogrammed or reprogrammed cells by immunostaining of validated cell surface markers (iDC1: CD45, HLA-DR; iHSPC: CD9, CD49f; iPSC: CD13, EpCAM, TRA-1-60, SSEA4) and cell sorting. A day 0 sample (D0) was also collected as baseline, for unreprogrammed cells that were edited since they contain the Cas9 and the KORE library. Once all samples were collected, I continued with mapping the gene perturbations by next-generation sequencing (NGS). In this sense, genomic DNA was extracted and gRNA cassette amplified, and finally sequenced the library to evaluate sgRNA enrichment.
Later on, I focused on the data exploration and identification of candidate genes by the analysis of hits retrieved from performing deep sequencing of the samples collected after the reprogramming experiments. I classified different cell populations regarding their reprogramming efficiency and identified groups of genes. On one hand, genes depleted in the reprogrammed cells could act as negative regulators or barriers to the reprogramming process. In this scenario, silencing of these genes would hypothetically benefit the reprogramming outcome. On the opposite, another list containing positive regulators or conversion facilitators, namely genes depleted in non-reprogrammed cells. Here, the lack of expression would be detrimental for reprogramming, and cells will not interconvert even in a favourable scenario. These barriers and facilitators were proposed as molecular targets that could be functionally important in plasticity regulation.
For the bioinformatic analysis, I applied the MAGeCKFlute pipeline for the identification of the plasticity-associated genes. First, I normalized the endpoint hits with the baseline sample D0 for every sorted population, to then calculate the sgRNA abundance-based rank difference between reprogrammed and non-reprogrammed cells to retrieve lists of genes. To hone the list of top candidates, I checked the correlation between the replicated screens, looking across donors. Taking into consideration a threshold of 1.5 standard deviation from the median, I collected the lists of top barriers and facilitators from the 3 reprogramming systems. Since I used the same sgRNA library in all the reprogramming systems, I carried out a comparative analysis to explore the relationship between sets of regulators previously sorted and examine genes shared in the intersections of unipotency, multipotency, and pluripotency. Through this assessment, I anticipated uncovering a list of chromatin and RNA modifiers that play a significant role in cellular plasticity. The genes identified in all the reprogramming systems were described as ‘universal’ regulators of reprogramming and represent master regulators of cell identity modification. Additionally, genes shared among pluripotent and multipotent cells were classified as regulators for increased plasticity given their possible role as molecular drivers that allow or preclude the generation of multiple cell types. Lineage-commitment regulators were identified at the intersection between multipotency and unipotency. In this case, involved in hematopoiesis specification, or inversely, in preserving the multipotent state, a context frequently associated with blood-related diseases.
The results obtained validated the experimental pipeline decided on the first aim and provided evidence of the feasibility of the approach. For iDC1 reprogramming, I listed 113 barriers and 72 candidate genes as facilitators. In the case of iHSPC, there were 95 barriers and 111 facilitators identified. Finally, for iPSC, the candidate genes came to 82 barriers and 125 facilitators. Regarding the barriers in the overlapping analysis, when doing 2 at a time comparisons, I identified 24 genes shared among iPSC-iHSPC, 33 genes among iHSPC-iDC, and 27 genes between iDC-iPSC. Finally, I was able to uncover 4 shared hits that could be described as possible universal barriers of cell plasticity.
Lastly, I proposed the validation of the top hits from the ranked lists of candidate genes by secondary screens. I focused on some of the interesting candidate genes including the ones present in the intersections from the integrative analysis, but also significant hits from the iDC1 and the iHSPCs screenings since they have a clear translational impact on cancer-related cell therapies. After shortlisting the genes, I selected the 2 sgRNAs that performed best for each candidate gene and validated them for the previously observed reprogramming outcome from the screening with the pooled KORE library. These guides were simultaneously cloned in a multiplex vector containing Cas9 expression as well, for lentiviral delivery. To start with, I selected barrier and facilitator candidates from the iDC1 screening and repeated the reprogramming experiments and flow cytometry analysis for a comprehensive phenotyping exploration at the respective endpoint. In addition, I analyzed genomic DNA disruption by Sanger sequencing with custom primers and ICE analysis for efficient gene KO.
Robust hits from the intersections of plasticity will be further studied to delve into their molecular mechanism and functionality. I plan to perform RNA-sequencing experiments to study the transcriptional changes that these KO have at a global scale on the cells. I will also perform ATAC-seq and ChIP-seq experiments to investigate the chromatin state and possible targets in the genome for the studied proteins.
