We used an in vitro model of Pluripotent Stem Cell (PSC) development

in mice to analyze dynamic changes in transcriptomes of hundreds

of individual cells which were undergoing an induced transition from

naïve mouse Embryonic Stem Cells (mESC) towards primed pluripotent

Epiblast Stem Cells (EpiSC). <br />

The differentiation of mESCs to EpiSC-like cells takes about five

days after induction. We collected cell samples in 24-hour intervals

for four days after induction as well as untreated mESCs and primed

state EpiSCs. Single-cell isolation and scRNA-seq library preparation for

each time point were done on the commercial Fluidigm C1 platform. In

addition, we sampled C1-Cap Analysis of Gene Expression (C1-CAGE)

libraries for the same set of time points to enable detection of non-

coding RNAs (ncRNAs) such as anti-sense RNAs or enhancer RNAs. <br />

This C1-CAGE protocol was new and still undergoing optimization

at the beginning of our experiments. C1-CAGE was first published

by Kouno *et al.* (2019) and the author of this thesis contributed as a

co-author. Throughout the work on this project a data management

platform called SCPortalen was developed to share all data among

project collaborators. SCPortalen’s publication was also co-authored

by the author of this thesis (Abugessaisa et al., 2018). <br />

The combination of transcriptome datasets from two different

protocols allowed the elucidation of expression dynamics of the naïve-

to-primed stem cell conversion. We independently identified two

subpopulations of cells during the transition process with both the

Fluidigm scRNA-seq and C1-CAGE dataset. Pseudotime analysis

revealed the developmental trajectory of cells and is a powerful tool to

reliably identify developmental stages of cells without prior knowledge

of their actual stage. Among these two transition phase subpopu-

lations, one showed wide-spread repression of gene expression. The

small nuclear RNA (snRNA) *Rn7sk* was identified as one potential

regulator of this population specific phenomenon. The second subpop-

ulation shared some characteristics with primed EpiSCs such as cell

morphology and the expression of known primed state marker genes, but

it could be shown that cells from this population were still undergoing

Epithelial-Mesenchymal Transition (EMT). That is a clear sign that

these cells have not yet fully transitioned to primed pluripotent stem

cells. Interestingly, the characteristics of this subpopulation largely

match a predicted third pluripotency state called “formative” (Smith,

2017). Therefore, we believe that our dataset not only contains naïve

and primed pluripotent stem cells, but also formative pluripotent stem

cells. Thus, our dataset represents a unique resource to compare and

study this proposed formative pluripotency state. Last but not least,

we found several marker gene candidates for all developmental stages

of the naïve-to-primed transition, which will facilitate classification of

cells in future experiments. For example, we propose *Cd59a* as a highly

specific marker gene for primed EpiSCs. <br />

The results of this thesis project have also been compiled into

a manuscript for publication in a peer reviewed journal and will be

submitted soon after the submission of this thesis.