(376a) Mechanistic Insights into Early Endoderm Differentiation of Human Embryonic Stem Cells Using Systems Level Analysis of Signaling Interactions

Introduction: Human
Embryonic Stem Cells (hESCs) are an attractive raw material for regenerative
medicine due to their unique properties of self-renewal and lineage specific
differentiation. hESCs are directed to mature cell types using various chemical
cocktails that replicate signaling programs active during in vivo
development. These signaling pathways employ large number of mediators that are
controlled by several regulatory mechanisms. Current efforts to develop
chemical modulators for controlling differentiation are purely experimental and
they have successfully identified the key players. However, systems level
analysis of the dynamics of this process is still under-developed. We recently
demonstrated the strength of quantitative systems level approaches to uncover
the role of regulatory mechanisms in the self-renewal state of hESCs [1]. Here,
we present a complete systems level analysis of the dynamics of regulatory
interactions in the TGF-β/SMAD and PI3K/AKT pathways, directing the early
endoderm differentiation of hESCs, by integrating efficient computational tools
and targeted experimental perturbations.

Materials and Methods:

Experimental analysis: For self-renewal, H1 hESCs were maintained on
matrigel-coated plates in mTeSR1. Endoderm differentiation was performed using
100 ng/ml Activin A (to activate TGF-β/SMAD2,3) with or without modulation
of PI3K/AKT pathway using PI3K inhibitor, Wortmannin. The phosphorylation
dynamics of participating signaling molecules were measured using MagPix
Luminex xMAP technology. The initial selection of key molecules was based on
the study by Singh et al. [2]. Nucleo-cytoplasmic shuttling rates of
molecules were measured using Fluorescence Recovery After Photobleaching (FRAP)
analysis. Key transcription factors characterizing endoderm were measured using
qRT-PCR. Mathematical and computational tools described in the next section
were used to guide generation of data by in-house experiments.

Mathematical analysis: We first employed data-driven modeling tools like
Partial Least Squares Regression (PLSR) and Dynamic Bayesian Network Analysis
(DBN) to identify key molecules, hypothesize interactions and
identify most informative time points. Detailed mechanistic
Ordinary Differential Equation (ODE) model for the TGF-β/SMAD2,3 pathway
with crosstalk interactions of PI3K/AKT was developed for a systems level
analysis. The model was calibrated using Replica Exchange Ensemble Modeling (EM)
and sensitive reactions were identified using computationally efficient Global
Sensitivity Approach (GSA).

Results and Discussion: PLSR results indicated that endoderm markers of hESCs
correlate well with the early but not the late signaling events. Application of
DBN on the early signaling dynamics showed that the molecules, p-SMAD2 and
SMAD4 form the core interactions (± p-AKT) and, p-SMAD3 and p-ERK were only influenced
by the core interactions (Fig. 1A). Interestingly, our experiments showed that
the dynamics and fold-change of p-SMAD2 and p-SMAD3 diverge, a novel result
seen only in hESCs. The reason for the divergence was investigated using a
detailed ODE model with crosstalk interactions with AKT. We generated hESC
specific model with SMAD2 and SMAD3 interactions modeled separately. EM on the
detailed ODE model captured the parameter ensembles that explained the
differentiating hESC system (Fig. 1B). The nucleo-cytoplasmic shuttling rates
of SMAD2 and SMAD3 (separately) measured through FRAP analysis was used to
constrain the parameter ensembles. From among the various hypotheses, AKT was
found to primarily influence the phosphorylation rates of SMAD molecules. GSA
results indicated that the sensitive parameters for p-SMAD2 and p-SMAD3 were of
similar ranking but of varying strengths resulting in the divergent dynamics. Phosphorylation
and de-phosphorylation of SMADs were the most sensitive reactions, while SMAD
nucleo-cytoplasmic shuttling rates were associated with modulation of the
aforementioned reactions. Further, negative feedback via SMAD7 constrained the
propagation of parameter uncertainty as well as promoted a robust signaling
response.

Conclusions:
Our results show that early signaling dynamics of p-SMAD, p-AKT and p-ERK encode the long-term endoderm
differentiation response of hESCs. We provide mechanistic explanations for the
divergence of p-SMAD2 and p-SMAD3 dynamics in hESCs. Further, our results
present the importance of negative feedback via SMAD7 in controlling the
long-term signal propagation and population variability in hESCs. Application
of mechanistic information revealed by such an analysis will direct precise
perturbations through designed small molecules, hence offering an avenue to
remove xenogenic factors in current culture conditions.

References:

[1] Mathew et al. Bioinformatics 2014, Article
in Press, ID: bioinf-2013-2200.r1 (btu209).

[2] Singh et al. Cell Stem Cell 2012, 10(3): 312-26