Plasticity at the cellular level

At the cellular level we use embryonic stem cell models (ESCs) to investigate the regulation of developmental potential during early stages of ESC differentiation. ESCs are highly plastic cells, capable of self-renewal and differentiation towards all cellular lineages.

Owing to these unique capabilities they have been the focus of extensive studies, yielding invaluable mechanistic insights on molecular mechanisms underlying the maintenance of pluripotency and the differentiation to particular lineages.

Still, a number of key questions remained incompletely understood, some of which are investigated in our lab. Specifically, we study:

  1. how the initial commitment to differentiation is affected by cell cycle phase,
  2. what are the intermediate stages of differentiation, and
  3. how these early events are regulated by microRNAs.


Interactions between cell cycle phase and differentiation in hESCs

hESCs are known to have unique cell cycle properties and were initially thought to avoid cell cycle check point signals and automatically commit to proliferation.
More recent evidence, however, suggests that hESCs may have active checkpoint and that the regulation of cell cycle and differentiation in hESCs are coupled.

To investigate these questions we established a centrifugal elutriation-based strategy enabling live isolation of hESCs in specific phases of their cell cycle. This includes fractionation of the G1 compartment into subsets of G1 cells that are committed to S-phase, and cells that have not yet made this commitment.
This system enables direct investigation of causal relations between cell cycle phase and differentiation in hESCs. Using this technique, we found that the fraction of non-committed G1 cells have a larger propensity to differentiate compared with cells in all other phases, and that, upon isolation, the pre-committed G1 cells differentiate at low cell densities even under self-renewing conditions.

We further showed that only the differentiation-prone fraction expresses high levels of hypo-phosphorylated retinoblastoma and that the differentiation of these cells is preceded by spontaneous cell cycle arrest. Notably, the differentiation of the G1 cells is partially prevented in dense cultures and completely abrogated in co-culture with S and G2 cells. This rescue from differentiation appears to be mediated by membrane-bound factors on neighboring cells. Read more in Sela et al., Stem Cells2012.




Uncovering intermediate stages of hESCs differentiation

Understanding the regulation of developmental potential during differentiation of ESCs could be greatly advanced by identifying and isolating cells undergoing initial phases of lineage commitment. In collaboration with the Weissman lab (Stanford Univ.), we initiated a large screen to identify such populations at a very early stage of committment.
We screened self-renewing and differentiating cultures of hESCs with over 400 antibodies recognizing cell surface antigens.

Based on gene expression signatures in sorted putative progenitor populations, comparison to mouse knowledgebase and functional assays, we discovered surface markers specific for 4 types of progenitors exhibiting, respectively, primitive endoderm, mesoderm, vascular endothelial, andtrophoblast characteristics.

We showed that these four progenitor groups also emerge from human induced pluripotent stem cells (hiPSCs), indicating the applicability of the markers for isolation of progenitors from any source of hPSCs. These markers and progenitors provide a novel platform for studying the commitment of pluripotent stem cells to lineage progenitors and purifying human tissue-regenerating progenitors.  Read more in Drukker et al., Nature Biotech2012.     


To further advance the ability to resolve specific subtypes of lineage committed cells we established high throughput proteomics platform in collaboration with the Walker lab (Weizmann Institute). The approach, designated differential cell-capture antibody array, is based on highly parallel, comparative screening of live cell populations using hundreds of antibodies directed against cell-surface antigens. We used this platform to fractionate the hitherto unresolved early endoderm compartment of hES-derived CXCR4+ cells and identify several endoderm (CD61+, CD63+) and non-endoderm (CD271+, CD49F+, CD44+, B2M+) subpopulations. We provide evidence that one of these subpopulations, CD61+, is directly derived from CXCR4+ cells, displays characteristic kinetics of emergence, and exhibits a distinct gene expression profile. This allows us to investigate the early stages of endoderm commitment at unprecedented resolution and shows that our platform provides efficient approach for resolving complex developmental programs. Read more in Sharivkin et al., Mol. Cell. Proteomics. 2012.


Gene regulation by microRNAs in hESCs

By analyzing the regulation of translation in hESCs, we uncovered microRNA-specific tendencies to reside in the translational pool and we deciphered the mechanism behind this interesting feature.microRNAs are known to repress the expression of genes by destabilizing target mRNAs or by repressing translation from these mRNAs.  

However, it is not clear what determines or affects the balance between mRNA destabilization and translation repression. One possibility is that the balance is regulated by the degree of association of microRNAs with polysomes (Polysomeoccupancy of microRNAs).
As a step towards testing this hypothesis, we established a novel high throughput method for quantifying polysome occupancies for hundreds of microRNAs (never been measured before).

Analysis in hESCs and human foreskin fibroblasts (hFF) revealed that the degree of association with polysomes is microRNA-specific. Bioinformatics and functional analysis based on over-expression of endogenous and chimeric microRNAs showed that the polysome occupancy of microRNAs is specified by their mature sequence and depends on the choice of seed.
Further analysis of microNRA•mRNA pairs revealed that the differential occupancy of microRNAs reflects their interactions with their mRNA targets: microNRA•mRNA pairs involving high occupancy microRNAs exhibit significantly higher binding energy compared to pairs with low occupancy microRNAs. Since the mRNAs reside primarily in polysomes (independently of the choice of microRNA), strong microNRA•mRNA interactions lead to high association of microRNAs withpolysomes and vice versa for weak interactions. 

These findings may contribute to understandingthe balance between mRNA destabilization and translation repression and might provide additional perspective for microRNAs’ mechanisms of repression. à See more in Molotski andSoen, RNA 2012.