Dev Bio SuperGroup Meeting

Dynamic Heterogeneity in Single Embryonic Stem Cells
John Yong, Elowitz Lab

In stem cell biology, single-cell gene expression data can potentially enable the
identification of distinct cellular states in mixed populations. However, it generally remains unclear
whether heterogeneous gene expression patterns reflect stochastic fluctuations ('noise') in
expression, or the co-existence of multiple cellular states within the population. To disentangle
these two sources of heterogeneity, we combined four-color single molecule mRNA-FISH
and quantitative time-lapse microscopy of individual mouse embryonic stem cells (ESCs). Analysis
of mRNA distributions suggests that cells populate two distinct, and transcriptionally coherent,
cellular states. Within those states, additional heterogeneity is generated through infrequent
transcriptional bursts of a specific subset of regulatory genes. Dynamically, cells transition
abruptly and infrequently between states. Perturbation using "2i" signaling inhibitors modulates
both the transition rates between states and the rate of stochastic bursts within a state, affecting
the overall dynamics and heterogeneity of regulatory genes in diverse ways. These results show
how distinct genes and timescales can be used to discriminate cell state from noise,
potentially enabling improved control and prediction of cell fate decision-making behaviors.

Single-cell transcriptome analysis of cellular reprogramming
Daniel Kim- Wold lab
Somatic cell reprogramming into induced pluripotent stem (iPS) cells involves widespread changes
in the coding and noncoding transcriptomes. Numerous transcription factors and chromatin
regulators, some of which associate with long noncoding RNAs (lncRNAs), facilitate
reprogramming of the somatic epigenome to pluripotency. Genome-wide studies of
reprogramming cell populations have been informative, yet are confounded by the stochastic and
asynchronous nature of this clonal process. We performed single-cell RNA sequencing to
investigate the identity, order, and coordination of changes triggered by ectopic Oct4, Sox2, Klf4,
and Myc. Induced cells faithfully activated known pluripotency transcription factors at early timepoints,
which coincided with a large global decrease in transcriptome variation. We found that
germ cell regulators are re-activated after the endogenous pluripotency network is established, as
confirmed by single-molecule FISH. Furthermore, we characterized the dynamic progression of
lncRNA activation in induced cells, including hundreds of lncRNAs that are associated with
chromatin regulatory proteins. These lncRNAs, which have emerging roles in epigenetic
regulation, may exert broad effects in reprogramming the epigenome to the pluripotent state.
Finally, our findings highlight a host of novel coding and noncoding gene sets, whose activity in
individual cells begins to parse divergent and shared responses to reprogramming stimuli.