RNA-sequencing analysis of naive and primed hPSCs in static and stirred suspension conditions

We have pioneered human pluripotent stem cell (hPSC) manufacturing in stirred suspension bioreactors. Cell therapies require large numbers of quality-controlled hPSCs yet technologies are limited in their ability to efficiently grow and scale clinically-viable hPSCs. We report here that naive hPSCs exhibit superior growth in suspension bioreactors compared to their primed counterpart. Naive hPSCs exhibited a shorter lag phase, and grew into more uniform, homogenous aggregates. Compared to static culture, gene expression analyses revealed that the bioreactor environment promoted the upregulation of naïve- and downregulation of primed-associated transcripts in both primed and naive hPSCs. Bioreactor-cultured naive hPSCs similarly showed more hypomethylated DNA and less primed hPSC-associated surface protein marker compared to statically-cultured naive hPSCs. Gene expression, epigenetic, and cell surface protein marker analyses all suggest that the bioreactor environment promotes the transition from primed-to-naive pluripotent state. Our research shows that reprogramming conventional hPSCs to the naive pluripotent state enhances hPSC manufacturing. Overall design: RNA-seq experiment was performed on naïve and primed human pluripotent stem cells from static culture and stirred suspension bioreactors. The aliquots of static-cultured cells for each naïve (P4) and primed hPSC sample were collected for RNA-seq analysis before inoculating them into stirred suspension bioreactors. Those cells were counted as “static-cultured donor cells”. The aggregates of day four post-inoculation for each bioreactor-cultured naïve and primed hPSC were collected for RNAseq analysis. We performed a pairwise comparison for transcriptomic analysis: 1) bioreactor-cultured naïve hPSCs and their statically-cultured naïve counterpart, 2) bioreactor-cultured primed hPSCs and their statically-cultured primed cognate. This study was exploratory (n =1), no p-values are given and therefore the simple normalized expression (normalization) value comparison was done as Reads per Kilobase per Million (RPKM). Expression levels for each gene were determined by RPKM in each sample. The RPKM values for each gene within each sample were compared to generate fold change and log2 fold changes values between the samples in a pair-wise comparison (a python script was used to compare RPKM values and to generate the foldchange tables). For the pairwise comparison between static- and bioreactor-cultured naïve hPSCs, the fold change > 1.95 was measured relative to static-cultured naïve hPSCs to derive differentially expressed genes among those samples. For the pairwise comparison between static- and bioreactor-cultured primed hPSCs, the fold change > 1.95 was measured relative to static-cultured primed hPSCs to derive differentially expressed genes among those samples. The differentially expressed genes (fold change > 1.95) among the pair-wise comparisons were used for enriched gene ontology (GO) terms (Molecular Function (MF), Biological Process (BP), and Biological Pathway (rea)) through g:Profiler.