Dissecting the role of distinct OCT4-SOX2 heterodimer configurations in pluripotency

Abstract

The transcription factors OCT4 and SOX2 are required for generating induced pluripotent stem cells (iPSCs) and for maintaining embryonic stem cells (ESCs). OCT4 and SOX2 associate and bind to DNA in different configurations depending on the arrangement of their individual DNA binding elements. Here we have investigated the role of the different OCT4-SOX2-DNA assemblies in regulating and inducing pluripotency. To this end, we have generated SOX2 mutants that interfere with specific OCT4-SOX2 heterodimer configurations and assessed their ability to generate iPSCs and to rescue ESC self-renewal. Our results demonstrate that the OCT4-SOX2 configuration that dimerizes on a Hoxb1-like composite, a canonical element with juxtaposed individual binding sites, plays a more critical role in the induction and maintenance of pluripotency than any other OCT4-SOX2 configuration. Overall, the results of this study provide new insight into the protein interactions required to establish a de novo pluripotent network and to maintain a true pluripotent cell fate.

Figure 1. Modeling the OCT4-SOX2 interactive configurations.

The Utf1-like (A) or the Fgf4-like (B) composite is used as a DNA template and is represented in gray. The OCT4 POU_S, the OCT4 POU_HD, and the SOX2 HMG domains are depicted in green, blue, and orange, respectively. The amino acids mutated in this study are indicated in both configurations. To highlight the differences between both arrangements, the orientation of OCT4 is kept the same in both DNA templates.

Figure 2. The OCT4-SOX2 configuration plays an important role in reprogramming.

(A) Human SOX2 amino acid alignment indicating the position of the point mutations included in each construct. (B) iPSC reprogramming efficiency generated using each different SOX2 mutant and measured by the number of Oct4-GFP positive colonies. Error bars correspond to standard deviations between three independent biological replicates. (C) SOX2-transgene expression in MEFs measured by qRT-PCR three days after transduction. Values are normalized to the expression of C- that corresponds to non-transduced MEFs and that is considered as 1. The error bars correspond to the standard deviation between three technical replicates. (D) Images showing Oct4-GFP expression and AP staining of one independent iPSC clonal cell line generated from each SOX2 construct. (E) Expression of endogenous pluripotency markers was measured by qRT-PCR using two mESC lines and MEFs as positive and negative controls, respectively. Values are normalized to the expression of ESC line #1, which is considered as 1. The error bars correspond to standard deviations between three technical replicates.

Figure 3. A specific OCT4-SOX2 configuration is essential for preventing ESC differentiation.

(A) Self-renewal rescue experiment. Sox2-null ESCs were transfected with hygromycin-resistant PiggyBac plasmids coding for the different SOX2 constructs, the empty vector was used as a negative control. The transfected Sox2-null ESCs were then treated with doxycycline and hygromycin for 1 week prior to AP staining. The rescue index was calculated by dividing the number of AP-positive colonies for each construct by the number of AP-positive colonies in the empty PiggyBac vector. Error bars correspond to standard deviations between three independent biological replicates. (B) Western blot analysis showing that all SOX2 constructs express equal levels of SOX2 protein after overexpression in 293T cells. α-TUBULIN is used as loading control. (C) Images of clonal cell lines established for each construct expressing the constitutive dsRed protein transgene are shown. (D) Pairwise scatter plots of global gene expression profiles comparing the average of two S113 clonal cell lines or three S98/102 clonal cell lines versus two SWT clonal cell lines. Black lines indicate a two-fold change in gene expression level between the paired populations. Color bar on the right indicates scattering density. Genes up- and downregulated are shown by red and green circles, respectively. (E) Heat map from microarray data showing the expression of a subset of pluripotent marker genes in stable Sox2-null ESC lines transfected with the different constructs. Values correspond to the average of two SWT-, two S113- and three S98/102-clonal Sox2-null ESC lines. The color bar at the top indicates gene expression in log₂ scale. Blue and orange colors represent higher and lower gene expression levels, respectively.

Figure 4. The abundance of the bound composite motif determines the impact of each OCT4-SOX2 configuration on pluripotency.

(A) Distribution of the motif sequences 100 bp around the peak summit obtained from OCT4 and SOX2 co-pulling down CHIP-seq experiments is depicted in black. Gray curves show the distribution in background control sequences selected using homer (http://homer.salk.edu/homer/index.html). Logos of the OCT4-SOX2 degenerate consensus sequences generated using http://weblogo.berkeley.edu/ are shown above. (B) EMSA analysis to evaluate the heterodimerization and binding capacity of OCT4 with wild-type or mutant SOX2 constructs on Hoxb1 (left), Utf1 (center) and Fgf4 (right) motifs. OCT4/SOX2/DNA supershift band, OCT4/DNA band and free DNA are indicated as , and , respectively. Of note, full-length SOX2 binding alone cannot be observed in this assay. (C) Pairwise scatter plots exhibiting the expression level of genes co-bounded by OCT4 and SOX2 on Hoxb1-like motif (left), Utf1-like motif (center) and Fgf4-like motif (right). A red and a green line represent a 2-fold expression level increase or decrease in comparison with SWT. Genes up- and downregulated are shown by red and green dots, respectively. Expression levels below 5 were considered as background.