Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mech Dev. Author manuscript; available in PMC 2011 January 1.
Published in final edited form as:
PMCID: PMC2818266

Delayed differentiation in embryonic stem cells and mesodermal progenitors in the absence of CtBP2


Mammalian embryonic stem cells (ESCs) are characterized by an ability to self-renew and give rise to each of the three germ layers. ESCs are a pluripotential source of numerous primitive progenitors and committed lineages and can make stoichiometric decisions leading to either asymmetric or symmetric cell division. Several genes have been identified as essential for maintenance of self-renewal, but few non-lineage specific genes have been identified as essential for differentiation. We selected the chromatin factor Ctbp2 from microarray data for its enriched expression in stem cells, in comparison to committed progenitors. RNA interference (RNAi) was used to knockdown gene expression in mouse ESCs and the potential for transduced cells to self-renew and differentiate was assessed in ESC and mesodermal assays. Here, we demonstrate an important role for Ctbp2 in stem cell maintenance and regulation of differentiation using an in vitro system. The knockdown of Ctbp2 increases the prevalence of ESCs in culture, delays differentiation induced by LIF withdrawal, and introduces developmental changes in mesodermal differentiation. A model is presented for the importance of Ctbp2 in maintaining a balance in decisions to self-renewal and differentiate.

Keywords: CtBP2, embryonic stem cells, hemangioblast, self-renewal, Oct4, mesoderm, RNA interference

1. Introduction

C-terminal binding protein (CtBP) was identified because of its interaction with the C-terminus of the adenovirus E1A oncoprotein (Boyd et al., 1993; Schaeper et al., 1995). In adenovirus, CtBP represses E1A transcriptional activation and oncogenesis (Sollerbrant et al., 1996). Detailed reviews of CtBP’s isolation, evolutionary conservation across eukaryotes, and biochemical activity have been published by Turner and Crossley (2001) and Chinnadurai (2002). Briefly, CtBP proteins are highly conserved across Drosophila, zebrafish, mice, and humans. Drosophila CtBP (dCtBP) interacts physically and genetically with the transcriptional repressors kruppel, knirps, and hairy in early development (Hildebrand and Soriano, 2002). In Drosophila, CtBP2 also interacts with snail, which is required for proper mesodermal development and cell fate decisions (Hemavathy et al., 2004).

The two CtBP proteins in mice - CtBP1 and CtBP2 - differ in expression, cellular localization, and regulatory role. Ctbp1 is expressed throughout embryogenesis and into adulthood. Ctbp2 is expressed in the forming placenta and is restricted to embryogenesis (Furusawa et al., 1999). CtBP1 is cytoplasmic and CtBP2 is almost exclusively found in the nucleus (Zhao et al., 2006). Both mammalian CtBPs associate with Polycomb proteins and complex with transcription factors to repress genes involved in mouse development. They have roles that somewhat overlap but they have differing degrees of importance. Ctbp1 nulls are born in expected Mendelian proportions. Ctbp2 nulls are embryonic lethal with developmental abnormalities appearing by E10.5 (Hildebrand and Soriano, 2002). The primary causes of lethality are defects in placental and yolk sac development. There is also an observed decrease in Brachyury expression. CtBP2 appears to be the more critical of the two CtBPs in embryonic development, as Ctbp1 nulls with a heterozygous Ctbp2 background are not viable and die around E15.5-18.5 (Hildebrand and Soriano, 2002).

CtBP2 interacts physically with key factors in the stem cell regulatory network. SMAD6 recruitment of CtBP2 to repress BMP-associated transcription has been demonstrated in vitro and in vivo (Lin et al., 2003; Hassel et al., 2004). CtBP2 can also bind SOX6 and repress transcription of Fgf-3 (Murakami et al., 2001). Other stem cell-related transcription factors that associate with CtBP2 to co-repress transcription include the AML/EVI1 oncoprotein, the Gata1-binding factor known as Friend of Gata (FOG), the mesodermal and ectodermal-specific transcription factor γ-EF1, and Krupple-like factor 3 (KLF3) (Turner and Crossley, 1998; Fox et al., 1999; Furusawa et al., 1999; Holmes et al., 1999; Turner et al., 2003). Although these protein-protein interactions have been documented biochemically, there are few studies that describe the functional importance of CtBP2 in mammalian embryonic development. Based on the in vivo description of embryonic lethality in Ctbp2 nulls and heterozygotes and based on decreased expression of Brachyury, we predict that Ctbp2 is important in early embryonic differentiation and proper mesodermal emergence. In this study, we use ESC and blast assays, as in vitro proxies for the blastocyst and mesoderm, respectively, with the objective of providing an in vitro parallel of what happens developmentally prior to E10.5 in a Ctbp2 null.

2. Results

2.1 Lentiviral-mediated RNA interference of CTBP2 is efficient

Lentiviral RNA interference vectors were designed to knockdown expression of Ctbp2 in mouse ESCs (mESCs). A lentiviral RNAi vector against Oct4, provided by Dr. Natalia Ivanova at Princeton University, was used to transduce mESCs in order to generate a self-renewal control and lentiviral vectors against Smarca4 were used to generate a cell-cycle control. RNAi was successful with all vectors, however, the vectors containing shRNA sequences designed against the more 3’ region of the open reading frame (ORF) were most efficient. The Ctbp2-A vector, which targets the more 5’ region of the Ctbp2 ORF, decreased gene expression by a broad range of 28–54% (Figure 1A). The Ctbp2-B vector, which targets the more 3’ region of the ORF, decreased gene expression by a narrower and more specific range of 85–98% (Figure 1A). The Oct4 RNAi control decreased gene expression consistently by 75%, which is sufficient to recapitulate the in vitro and in vivo phenotypes. The efficiency of CtBP2 knockdown was confirmed with Western blots of whole cell extracts from cells sorted into media at Days 12 and 15 post-transfection. Blots show that there is no observable change in CtBP2 with Ctbp2-A RNAi, however, there is a clear knockdown of CtBP2 observed with Ctbp2-B RNAi (Figure 1B). The Western blot data is consistent with the quantification of each vector’s ability to knockdown Ctbp2, as measured by RT-PCR.

Figure 1
(A) Real-time PCR (RT-PCR) confirmation of Ctbp2 RNAi, as compared to empty vector (EV) and Oct4 RNAi positive control. The Ctbp2-A RNAi construct targets a more 5’ region of the open reading frame (ORF) of Ctbp2 and the Ctbp2-B RNAi construct ...

2.2 Self-renewing ESCs are more prevalent in the absence of CtBP2

After confirming the efficiency of CtBP2 knockdown, transduced ESCs were placed in a competition assay to characterize self-renewal ability. The premise of the assay, as described in more detail in the Experimental Methods section, is to monitor the ratio of GFP−/GFP+ ESCs over time. The competition assay was read using FACS at 48-hour intervals that overlapped with culture passage. The raw percentage change in GFP−/GFP+ is depicted in Figure 2A and was normalized against the mock infected and empty vector cultures to adjust for an expected ~10% decrease in cells. Our findings are based on cumulative changes in GFP−/GFP+ from Day 2 through Day 14 of competition and indicate that a knockdown of CtBP2 causes an increase in the ratio of GFP−/GFP+ cells (Figure 2A). The extent of the phenotype is related to the degree of RNAi associated with the Ctbp2-A and Ctbp2-B vectors. In Ctbp2-A RNAi cultures, 45–80% of Ctbp2 is expressed but without change to protein levels. The absence of change in protein levels is associated with an insignificant increase in the GFP ratio (p>0.05) (Figure 2A). In contrast, Ctbp2-B RNAi reduces Ctbp2 expression to less than 15% and there is an observable loss of protein. This loss of protein is associated with a significant (p<0.05) 20% increase in the GFP ratio (Figure 2A).

Figure 2
ESC self-renewal and maintenance. (A) Percent change in RNAi GFP+ ESCs in competition with GFP− ESCs as measured by FACS. Percent change is cumulative and measured from Day 2 through Day 14 of a timecourse with empty vector (EV), Ctbp2-A RNAi, ...

Since Ctbp2-B RNAi cells experience an apparent gain in self-renewal ability and increased prevalence in culture, then the assumption is that either the progression through the cell cycle quickens or the balance between asymmetric and symmetric self-renewal is altered. To explore the former possibility, we performed PKH26 dye and Annexin V staining assays using Smarca4 RNAi cultures as a positive control. We did not observe significant cell cycle-related changes in Ctbp2 RNAi cultures although it was apparent that independent cultures of Ctbp2-B RNAi ESCs produce greater numbers of ESCs over time. In a proliferation assay with Ctbp2-B RNAi, we did not observe a significant (p>0.05) increase in cell division (Figure 2B) or change in STAT3 and AKT signaling (Supplementary Materials, Figure S1). In the absence of data in support of increased cell division in Ctbp2 RNAi ESCs, we turned to assays that would explore the possibility of a shift in self-renewal and differentiation.

We used three methods to characterize self-renewal and differentiation - AP assay, colony measurements, and LIF withdrawal. An increase in the self-renewal ability of ESCs should affect colony morphology and maintenance. Changes in colony morphology can be visualized in an alkaline phosphatase (AP). In the AP assay for Ctbp2 RNAi, we analyzed the morphology of Ctbp2-B RNAi ESCs in LIF+ (self renewal) and LIF− (differentiation) conditions. In the presence of LIF, we observed Ctbp2-B RNAi colonies that were larger in comparison to the controls and in comparison to Ctbp2-A RNAi (Figure 3A). We used IPLab software to trace colony boundaries and to select the pixels at colony boundaries to generate measurements of perimeter and area. Our findings of larger colony area and perimeter for Ctbp2-B RNAi cultures confirm the results of the AP assay (Figure 3B). In the AP assay in the absence of LIF, differentiation ensues as expected in all cultures except for Ctbp2-B RNAi. In the Ctbp2-B RNAi cultures, large round colonies are still visible in contrast to mock, empty vector (EV), Ctbp2-A RNAi and Oct4 RNAi cultures (Figure 3A). This difference in colony morphology of Ctbp2-B RNAi is consistent with the demonstrated difference in degree of gene expression knockdown and significance of competition ratios. The morphological observation in the LIF-panel of the AP assay is supported by analysis of Oct4 levels during a LIF withdrawal timecourse. In the AP assay when LIF is removed, Ctbp2-B RNAi cultures do not differentiate as quickly as control cultures (Figure 3A, compare LIF to No LIF). After 7 days without LIF, the EV cultures can be categorized as “differentiated” and the Ctbp2-B RNAi cultures resemble the “semi-differentiated” cultures pictured in Ivanova et al. (2006). The morphological delay in differentiation is confirmed by a corresponding delay in decrease of Oct4 expression after LIF withdrawal. In Figure 3C, we show that Oct4 levels decrease sharply and as expected in empty vector cultures by Day 4 of LIF withdrawal, whereas Oct4 levels in Ctbp2-B RNAi cultures do not decrease sharply and are not at comparably low levels until Day 6 of differentiation.

Figure 3
(A) Alkaline phosphatase (AP) assay of ESCs in self-renewal (LIF+, top) and differentiation (LIF−, bottom) conditions. Cells were sorted for GFP expression at Day 2 post-transduction (p.t), plated in their respective conditions at Day 4 p.t., ...

2.3 Blast colony formation is delayed in Ctbp2 RNAi cultures

Our findings in ESC self renewal assays demonstrate that a decrease in CtBP2 affects stem cell maintenance by increasing the prevalence of ESCs in culture and shifting cell fate decisions away from differentiation. We wanted to further characterize the impact of Ctbp2 RNAi on differentiation and chose mesodermal differentiation assays given reported in vivo changes in yolk sac development and Brachyury expression in the null (Hildebrand and Soriano, 2002).

The transferability and efficiency of RNAi in the in vitro mesodermal differentiation assays was verified using a proof-of principle experiment targeting Aml/Runx1. In vivo, Aml/Runx1 is required for the development of definitive, but not primitive, hematopoietic lineages from the hemangioblast (Okuda et al., 1996; Wang et al., 1996). Lentiviral-mediated RNAi against Aml/Runx1 recapitulates the in vivo phenotype and results in normal primitive and blocked definitive hematopoiesis (Supplementary Materials, Figure S2A). The RNAi vector is stable in the hemangioblast system and expression from the RNAi cassette is maintained throughout mesodermal induction and differentiation (Supplementary Materials, Figure S2B). We used the Ctbp2-B RNAi vector for the experiments in the hemangioblast system because this vector yielded the highest level of knockdown and the most significant results in the ESC assays.

Formation of mesodermal progenitors, known as hemangioblasts (in vivo) or blast colonies (in vitro) is delayed in Ctbp2-B RNAi cultures (Figure 4A). Upon differentiation and during the blast formation window of Days 2.5–3.5, C-KIT expression should decrease and FLK-1, a marker for blast colonies, should increase. At Day 3.5 when blast colony formation is expected to peak, C-KIT expression should rebound because c-kit is necessary for hematopoietic development. CD31 expression should decrease through Day 3.5. The pattern of marker expression at Day 3.5 is very specific and critical because it predicts the emergence of the in vitro hemangioblast (Lacaud et al., 2004). Blast colonies can only form properly if self-renewal has been sufficiently suppressed. A temporal comparison of the empty vector cultures (EV RNAi) and Ctbp2-B RNAi shows a clear half-day delay in kinetics, even though EV RNAi and Ctbp2-B RNAi cultures are similar at Day Zero (Figure 4B). This confirms that neither has entered differentiation prematurely. A snapshot of FACS kinetics at Day 3.5 shows that EV RNAi has the expected expression pattern of FLK-1 and C-KIT expression and that CD31 expression has decreased as expected (Figure 4A). In contrast, the C-KIT expression in Ctbp2-B RNAi has not rebounded and the expression of CD31 is lagging in its descent. Quadrant statistics on biological replicates of the FACS timecourse confirm that the differences in FLK-1 and C-KIT expression between EV RNAi and Ctbp2-B RNAi at Day 2.5, 3, and 3.5 are significant (p<0.05) with the largest difference at Day 3.5 (Figure 4B). CD31 expression is also significantly different between the two populations at Days 3 and 3.5 (Figure 4B, p<0.05).

Figure 4
Delay in hemangioblast formation. (A) FACS profile of embryoid body (EB) differentiation in empty vector (EV) and Ctbp2-B RNAi cultures at Day Zero (left) and Day 3.5 (right). (B) Percent differences between Ctbp2-B RNAi EBs and EV RNA EBs in FLK-1/C-KIT ...

The FACS kinetics for expression of FLK-1, C-KIT and CD31 define the window of blast colony formation, but the colony-forming assays confirm blast colony emergence during this window. The morphology of the blast colonies and blast precursors (secondary EB and transitional) do not differ between normal and RNAi cultures (Supplementary Materials, Figure S3). We found that there are significantly more (p<0.05) mesodermal progenitor colonies in Ctbp2-B RNAi cultures at Day 3.5 with the largest difference in the Flk+ sorted population (Figure 5A). It is important to note that the Flk+ population gives rise to the blast colony; however, when we characterized the colonies according to type we found that although Ctbp2-B RNAi has a larger total number of colonies, it has a much lower number of mature blast colonies. The bulk of the total colony count consists of immature transitional colonies and undifferentiated secondary EBs (Figure 5B).

Figure 5
(A) Total number of blast assay colonies from a FLK+ population isolated from embryoid bodies (EBs) at Day 3.5. Data is representative of cultures in triplicate with p < 0.05. (B) Distribution of type of blast assay colony, generated from a FLK+ ...

If we look at the distribution of the three colony types (blast, transitional and secondary EB) over time, we can characterize a difference in the pattern and distribution of blast colony emergence with a significant difference (p<0.05) in the number of blast colonies at Day 3.5 in Ctbp2-B RNAi (Figure 6). The EV RNAi cultures peak in blast formation at Day 3.5, as expected, but the Ctbp2-B RNAi cultures do not experience a peak in blast colony formation until Day 4. At the time of EV RNAi peak (D3.5), blast colonies represent roughly 60% of total colonies, but at the time of Ctbp2-B RNAi peak (D4), blast colonies are in the minority at approximately 30% of total colonies. The kinetics of Brachyury expression, conducted in quadruplicate, correlate well with the FACS profiles. Brachyury expression in EV RNAi peaks at Day 3 of mesodermal induction and precedes the point of hemangioblast formation at Day 3.5, as expected (Figure 7). Brachyury expression in Ctbp2-B RNAi is down regulated and expression does not reach its zenith until Day 3.5. This half-day delay in Brachyury expression goes hand-in-hand with the significant half-day delay in emergence of blast colonies.

Figure 6
Distribution of secondary embryoid body (EB), transitional and blast colonies across Days 2.5–6. The significant difference (p < 0.05) in blast colonies between empty vector (EV, top) and Ctbp2-B RNAi (bottom) at D3.5 noted by asterick ...
Figure 7
Expression of Brachyury in embryoid body (EB) differentiation. Quantified for empty vector (EV) and Ctbp2-B RNAi (Ctbp2) cultures using RT-PCR.

2.4 The distribution of colonies in early hematopoiesis is altered

In the hematopoietic “big mix” assay of Ctbp2-B RNAi cultures, there are more total colonies but there is no significant difference in the absolute number of hematopoietic colonies (primitive erythroid, macrophage, erythroid-macrophage, and mixed lineage). There is a clear difference in the distribution of colonies in Ctbp2-B RNAi, with the number of secondary EBs dominating the total colony distribution and dwarfing the number of hematopoietic colonies (Supplementary Materials, Figure S4B). In addition to a relative deficiency in hematopoietic colonies, which characterize in vitro hematopoiesis, we also characterized a deficiency in Scl expression in Ctbp2-B RNAi (Supplementary Materials, Figure S4A).

3. Discussion

In this study, our goal was to explore the effect of loss of CtBP2, in vitro, in order to better characterize the observed in vivo defects that arise between E6.5–E10.5, prior to embryonic lethality. ESC assay results described herein provide an in vitro view of null embryonic development and demonstrate that although the defect in null mice is not fully visible until E10.5, there are important and progressive changes in the developmental program from the blastocyst to gastrulation stages. There are also important changes to the development of the hemangioblast. The hemangioblast does not represent the mesoderm, per se, but does represent the potential of the mesoderm to progress through development and give rise to differentiated cells.

This study focused primarily on characterizing Ctbp2 knockdown culture development from ESC (blastocyst) to EB (gastrula) and then on the emergence and function of the blast colony (hemangioblast). A functional approach was used to characterize the effect of a loss of Ctbp2 in ESC maintenance and in the development of mesodermal progenitors. Findings indicate that a knockdown of Ctbp2 expression skews the balance between self-renewal and differentiation towards self-renewal even when differentiation is forced. A competition assay was first used to quantify an increase in prevalence in ESCs in the absence of Ctbp2 and then cell cycle assays verified that this change is not attributable to changes in rate of cell division. Since there are only two mechanisms by which a stem cell could increase in prevalence, these being to divide more rapidly or to experience a change in symmetry of fate decisions, it was hypothesized that the latter mechanism was responsible for our observations. The hypothesis was supported by results from the AP assay in LIF- conditions and the quantification of Oct4 expression in a differentiation timecourse. These results lead to the conclusion that Ctbp2 is important for normal asymmetric differentiation. A decrease in CtBP2 creates a noticeable delay in ESC differentiation, but it does not completely block differentiation as evidenced by our ability to generate embryoid bodies. However, the embryoid bodies with knocked down Ctbp2 expression are not normal. Emergence of blast colonies from Ctbp2-B RNAi EBs is delayed. In vitro, the expectation is to isolate the highest number of mesodermal progenitor colonies, or blast colonies, from EBs that are at Day 3.5 of differentiation but in the absence of Ctbp2, peak blast formation is delayed until Day 4. The timing of this isolation is critical as primitive hematopoiesis rides on the coat tails of blast formation and what might appear to be a small delay of 12 hours is significant for the rapidly developing embryo.

In addition to a delay in blast formation, there is a significant accumulation of immature blast colony precursors. If differentiation occurred normally, then the expectation is to isolate a larger number of differentiated blast colonies from Ctbp2-B RNAi. Transitionals, defined in Robertson et al. (2000) as colonies containing cell populations undergoing commitment from mesoderm to hematopoietic and endothelial lineages, should be in the minority by Day 3.5. Instead with Ctbp2-B RNAi, there are a proportionally smaller number of mature blast colonies and a larger number of immature transitional and secondary EB colonies. Additionally, if the blast colonies that emerged from Ctbp2-B RNAi EBs were normal, then in the hematopoietic assay the total number of primitive and definitive hematopoietic colonies should have greatly surpassed those of EV RNAi. Instead the numbers of hematopoietic colonies are comparable to EV RNAi but only make up 10% of the total colonies in Ctbp2-B RNAi.

Gene expression data and results of Western blots for signaling molecules can be used in conjunction with existing knowledge to build a mechanistic model for Ctbp2 in embryonic development. In addition to characterizing functional delays in differentiation, this study found changes in Oct4, Brachyury and Scl expression. Abnormal changes in protein levels and phosphorylation were not evident in STAT3 or AKT signaling or in levels of phosphorylated GSK3-β and β-catenin. STAT3 is phosphorylated in response to an activating signal induced by the LIF ligand. We see normal STAT3 availability and phosphorylation in LIF+ conditions and the expected loss of phosphorylation in LIF− conditions. Although cultures lacking CtBP2 lose STAT3 phosphorylation upon LIF withdrawal, they retain Oct4 expression well into induced differentiation. This retention of Oct4 expression is attributed as the mechanism behind the Ctbp2-B RNAi cultures’ self-renewal resiliency in the absence of LIF. Since CtBP2 is a transcriptional co-repressor, it would be worth exploring in future biochemical studies whether there is a physical binding of CtBP2 to the Oct4 promoter or if CtBP2 is repressing Oct4 by binding to other genetic elements or co-activators.

A recent study reports that STAT3 and AKT pathways make up a parallel circuit for maintaining self-renewal in ESCs in the absence of LIF (Niwa et al., 2009). We were able to exclude STAT3 signaling as a potential mechanism for the Ctbp2-B RNAi phenotype and in our analysis of AKT signaling in the presence of LIF, we do not observe a change in protein or phosphorylation levels. This study does not exclude the potential for an AKT-associated signaling mechanism in Ctbp2-B RNAi cultures in the absence of LIF (Supplementary Materials, Figure S1C). However, given our identification of both a gain of function phenotype in the presence of LIF and a downward trajectory of Oct4 expression after LIF withdrawal, the signal regulating CtBP2’s repressor activity should be apparent in both the presence and absence of LIF.

The phenotypes observed in the blast assays, namely a delay in blast colony emergence and an unusual abundance of transitional colonies, are attributed to a loss of Brachyury and Scl. These in vitro findings parallel those from in vivo embryos with defects in Brachyury expression and from Scl−/− EBs (Robertson et al., 2000). A decrease in Brachyury expression was noted in the Ctbp2 null and linked to Wnt signaling, since CtBP2 is reported to function with β-catenin as a transcriptional activator of Brachyury (Yamaguchi et al., 1999; Hildebrand and Soriano, 2002). Our study does not find a change in GSK3-β or β-catenin levels in the absence of CtBP2, therefore, we can confirm that CtBP2 functions alongside β-catenin instead of functioning upstream of β-catenin. We can conclude that CtBP2 is required for normal transcription of Brachyury and that Wnt signaling alone is necessary but not sufficient to activate normal transcription of Brachyury. There are no published findings with regard to CtBP2 regulation of Scl; however, defects in yolk sac vascularization are noted in the Ctbp2 null (Hildebrand and Soriano, 2002). Scl is recognized as a key regulator of hematopoietic specification and yolk sac angiogenesis, so the in vivo hematopoietic defects and the abundance of transitionals in vitro point toward a loss of Scl (Robertson et al., 2000; Lecuyer and Hoang, 2004). Gene expression analysis confirms a decrease in expression of Scl in Ctbp2-B RNAi and leads to the conclusion that Ctbp2 is important in embryonic development as a transcriptional activator of Scl. Protein-protein interactions and promoter studies are still needed to characterize CtBP2’s effect as a direct activator or as an indirect activator of Scl gene expression.

The decision to self-renew or differentiate is not as black and white as it might appear and this existence of a gray area enables ESCs to respond to the constantly changing demands of development. Pluripotency depends on an ability to balance fates and choose between symmetric self-renewal, asymmetric cell division, or commitment. The findings from this study demonstrate a clear role for Ctbp2 in embryonic development. The model proposed is one in which CtBP2 normally represses Oct4, either directly or via a yet to be determined signaling pathway, to permit ESC differentiation and maintain asymmetric self-renewal. The model also proposes that CtBP2 directly activates transcription of Brachyury and Scl to permit normal blast formation and induction of hematopoiesis.

The proposed role of CtBP2 is to establish a pattern of gene expression that favors pluripotency and asymmetric self-renewal. The loss of CtBP2 does not completely block differentiation in ESC or blast assays and does not completely block transcription of Brachyury or Scl, which suggests the presence of a secondary transcriptional activator with similar function. From a developmental perspective, the presence of a secondary regulator in the absence of CtBP2 will only rescue the developing embryo if the activity occurs in parallel or if normal expression of Brachyury and Scl is maintained. Given the observed in vivo defects which lead to embryonic lethality and the delays in the in vitro assays that lead to aberrant ESC differentiation and blast formation, we infer that the secondary regulatory does not function in parallel to CtBP2, but rather, is activated in response to a loss of CtBP2. In sum, the Ctbp2-B RNAi cultures do eventually differentiate but unfortunately, with regard to embryonic development, delays in developmental cues are often lethal. In addition to continued characterization of biochemical interactions between CtBP2 and stem cell gene promoters to further our knowledge of the developmental regulation in the in vivo null, the finding of an abundance of undifferentiated progenitors also suggests that future studies include an analysis of Ctbp2 in tumor models.

4. Experimental Procedures

4.1 Candidate Selection and RNA Interference Design

Ctbp2 was selected from microarray expression data sets based on differential expression (enrichment) in stem cells, in comparison to committed progenitors (Ivanova et al., 2002; Ramalho-Santos et al., 2002). The criteria used were identification as a chromatin factor and a greater than two-fold level of enrichment in stem cells. Ctbp2 satisfied both criteria sufficiently as a chromatin-related co-repressor of transcription with a level of expression in mouse embryonic stem cells five times that measured in committed progenitors (Ivanova et al., 2002).

Dharmacon’s RNAi design program ( was chosen based on specificity in the design of siRNA sequences and use of the National Institutes of Health (NIH) Blast algorithm to exclude sequences with any homology to Ctbp2. Blast of Ctbp1 and Ctbp2 shows 76% sequence homology. Blast of the siRNA sequence Ctbp2-A to the coding sequence of Ctbp2 shows pairing of all 19 nucleotides at position 1924–1942. There are 6 additional pairings possible but the largest of these is only 8/19 nucleotides. Blast of the siRNA sequence Ctbp2-A to the coding sequence of Ctbp1 shows the largest match as only 9/19 nucleotides, with an additional 7 pairings of no more than 8/19 nucleotides. Blast of the siRNA sequence of Ctbp2-B to the coding sequence of Ctbp2 shows pairing of all 19 nucleotides at position 2594–2612. There are an additional 4 pairings possible but the largest match is only 9/19 nucleotides. Blast of the siRNA sequence Ctbp2-B to the coding sequence of Ctbp1 shows that there is only one possible pairing of 7/19 nucleotides.

Some publications have suggested an effect of the anti-sense strand of the siRNA construct, so to be cautious, we blasted this sequence as well. Our findings show that the anti-sense sequence of Ctbp2 construct A with the Ctbp2 coding sequence yields one match of 7/19 while a blast of the same construct with the Ctbp1 coding sequence yields three pairings, all of 8/19 base pairings or less. The anti-sense sequence of Ctbp2 construct B blasted against the Ctbp2 coding sequence produces six matches of 8/19 base pairings or less, and a blast of the same construct with the Ctbp1 coding sequence yields only one match of 7/19 base pairings. The conventional minimum threshold for bioinformatics exclusion is less than 15 base pairs of homology. The homology of the additional pairings shown here is far below that threshold. Jackson et al. (2003) tested the thresholds at which siRNA becomes ineffective and their findings affirm 15 base pairs of homology as the threshold.

4.2 Lentiviral Vector Production and RT-PCR for Gene Expression

A third generation lentiviral vector FUGW (Lois et al., 2002) was modified as described in Ivanova et al. (2006) to express a short-hairpin RNA (shRNA) from an H1 promoter. Sequence-specific RNAi cassettes were designed in part using Dharmacon’s siRNA Design Center. The Dharmacon software was used to, first, identify a gene specific target within the open reading frame (ORF) and, second, to generate a 19 base, sequence-specific oligonucleotide. Two shRNA cassettes were generated per gene target. The cassette that targeted the more 5’ region of the ORF was labeled “A” and the cassette that targeted the more 3’ region of the ORF was labeled “B”. The oligonucleotides generated are as follows: Ctbp2-A (ATTCGACAGTCTGCCACAT) and Ctbp2-B (AGAATTCTTCGTGACTTCA). The shRNA cassettes were cloned directionally into the vector using XbaI and SmaI cloning sites. Cloning was verified by restriction digestion with PacI and sequencing of the fragment product with a primer for the HIV-1 flap (ACAGCAGAGATCCAGTTTG). Each sequence-verified shRNA lentiviral plasmid was co-transfected with pVSV-G and pCMVd8.9 packaging plasmids into 293FT cells. At 16–22 hours prior to co-transfection, the 293FT cells were plated on pre-gelatinized 15cm dishes at 18×106 cells per dish and cultured at 37°C. At transfection, 3–15cm dishes of cultured 293FT cells were used per lentiviral construct. Superfect was used to mediate transfection and the transfected 293FT cells were cultured at 37°C for 20–24 hours and moved to 32°C for an additional 40–44 hours of culture. The viral supernatant was harvested from the dishes, filtered through a 0.44µm low-protein binding filter, and centrifuged for 150 minutes at 50,000×g. The resulting pellet was resuspended in 300µL sterile PBS, aliquoted, and stored at −80°C until use. Viral titers were measured using serial dilutions of viral construct to transduce NIH 3T3 fibroblasts and then FACS analysis measured expression of GFP to confirm transduction efficiency. For quantification of RNAi efficiency in gene expression knockdown, all transduced ESCs were sorted by FACS, according to expression of GFP, into Trizol for RNA extraction and cDNA synthesis. Quantitative polymerase chain reaction (QPCR) was conducted on cDNA samples from cells sorted on Days 4, 8, 11 and 12 post-transduction (p.t).

4.3 Cell Culture

CCE murine ES cells (Advanced Cell Technologies) were maintained on gelatinized plates in self-renewal media (DMEM supplemented with 10% FCS, glutamine, non-essential amino acids, sodium pyruvate, BME, P/S, and LIF). J1 murine ES cells (a gift from Dr. Gordon Keller, Mt. Sinai School of Medicine) were cultured on irradiated mouse primary embryonic feeders as reported (D’Souza et al., 2005). Transductions were performed as follows: CCE ES cells were trypsinized and resuspended to a concentration of 400,000 cells/mL in CCE self-renewal media supplemented with polybrien. For each vector, 1mL of cells was mixed with the lentiviral RNAi construct at low multiplicity of infection (MOI of 5), plated in one well of a pre-gelatinized 6-well dish, and cultured at 37°C for 20–24 hours. After the first 20–24 hour period, 1mL of new CCE self-renewal media was added to each well of the dish and the dish was cultured for an additional 24 hours to maximize GFP expression from the lentiviral construct. J1 cells were transduced in a similar fashion except for the inclusion of a feeder depletion step, on an ungelatinized dish for 1–1.5 hours, prior to the addition of lentiviral RNAi construct.

4.4 Quantitative and Semi-Quantitative PCR

Cells were sorted into Trizol at a concentration of 100,000 cells/mL. Total RNA was extracted using a standard RNA isolation protocol. Total RNA was transcribed into cDNA with random hexamers using Powerscript. Semi-quantitative PCR and Q-PCR were carried out using gene-specific primers.


4.5 Immunochemistry

For Western blotting, whole cell proteins were isolated with RIPA buffer and nuclear proteins were isolated with TMSB buffer (25mM NaCl, 10mM Tris-Hcl at pH7.8, 1.5mM MgCl2, 0.25M sucrose, 10mM BME, 0.5mM PMSF) and F buffer (10 mM Tris, pH 7.05, 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM NaF, 5 µM ZnCl2, 0.1 mM Na3VO4, 10% glycerol, 0.1% Triton X-100, aprotinin, PMSF). After SDS-PAGE, the nitrocellulose was blotted with antibodies to CtBP2 (Santa Cruz #5967), Phospho-STAT3 (Cell Signaling Technologies #9138), STAT3 (Cell Signaling Technologies #9139), Actin (Sigma #A2066), SMARCA4 (Upstate #07-478), Phospho-AKT Pathway Sampler Kit (Cell Signaling Technologies #9916), MAPK/ERK-1/2 (Cell Signaling Technologies #9102), β-catenin (a gift from Dr. Kateri Moore, Princeton University), and ERK-1/2 (a gift from Dr. Jean Schwarzbauer, Princeton University).

4.6 Competition and Proliferation Assays

Transduced cells were sorted on a FACS Vantage according to GFP expression. GFP-positive cells were plated into competitive assays, in a 6-well dish at 37°C as described in detail in Ivanova et al., 2006. Cultures were analyzed for GFP expression by FACS at each 48-hour passage. Briefly, the ESC competition assay has been used in previously published studies to assess self-renewal ability (Ivanova et al., 2006; Galan-Caridad et al., 2007) and is based on the classic hematopoietic stem cell (HSC) competitive repopulating assay. Normal stem cells, GFP−, set the threshold for self-renewal and differentiation activity in culture. Mutant ESCs, GFP+, are placed in self-renewal competition with the normal ESCs. The ratio used for most competition assays is 50% normal ESCs and 50% mutant ESCs. If the mutation or knockdown in the mutant ESCs has no affect on self-renewal ability, then the ratio of 50/50, as assayed by FACS for GFP−/GFP+, will be maintained over the course of the culture period (+. If the mutation leads to a loss of ESC self-renewal ability, then the mutant ESCs will differentiate and will not be included in the GFP+ stem cell category. Differentiated ESCs are larger and can be separated out from undifferentiated ESCs according to FACS scatter profile. The resulting ratio will favor a higher percentage of normal stem cells (GFP−) as the dominant cell type in the culture. If the mutation leads to a gain of self renewal ability above what is normally observed in normal cultures, then the ratio of GFP−/GFP+ ESCs will shift towards the GFP+ cell population. We used internal controls to ensure that the difference in GFP−/GFP+ is not attributable to transfection reagents, handling (mock transfection) or to RNAi method (empty vector). We also used a positive control for loss of self-renewal (Oct4 RNAi).

To ascertain rate of proliferation, cultures were first analyzed using PKH26 dye retention (Sigma) and Annexin V staining (Roche) according to manufacturer’s instructions to exclude the likelihood of abnormal changes in cell cycle or rate of apoptosis. Afterwards, control and knockdown cells were grown in separate dishes with cell counts at passage, every 36-hours, for 4 passages. For each passage, we first counted the cell number in culture for the previous passage and then we replated a specific number of cells (identical number for both control and knockdown cultures) so that each passage began with the same cell number per culture dish.

4.7 Differentiation

For the alkaline phosphatase (AP) assay, ESCs were sorted at Day 2 post-transduction and plated in LIF+ or LIF− conditions in DMEM supplemented with 10% FCS (tested for ESCs), glutamine, non-essential amino acids, sodium pyruvate, BME, and P/S as reported (Ivanova et al., 2006). Cultures were stained and assayed using an AP kit (Sigma) at Day 9 post-transduction and colony size was measured with IP Lab (Scanalytics, Inc.) software.

For embryoid body (EB) differentiation, ESCs were differentiated under EB-generating conditions in DMEM, 10% FCS (tested for differentiation), glutamine, non-essential amino acids, sodium pyruvate, BME, and P/S. EBs were trypsinized and dissociated prior to analysis by FACS with antibodies against FLK-1 (a gift from Dr. Gordon Keller, Mt. Sinai), PE-conjugated C-KIT (BD Pharmingen #553355), and biotin-conjugated CD31 (BD Pharmingen #553371). Blast assay colonies and hematopoietic “big mix” colonies were isolated and scored using published methods (Choi et al., 1998; Fehling et al., 2003; Kubo and Shinozaki, 2003; D’Souza et al., 2005; Ivanova et al., 2006).

Supplementary Material






We would like to thank all members of the Lemischka Research Group at Princeton University and the Keller Research Group at Mt. Sinai School of Medicine. We would specifically like to thank Dr. Gordon Keller and Dr. Sunita D’Souza for their feedback with regard to this project. We would also like to thank Dr. Gordon Keller for the provision of materials as indicated. This research was supported by NIH 5T32 GM07388 (PI: IR Lemischka) and NIH 5R37 DK042989 (PI: IR Lemischka). During manuscript preparation and editorial review, Dr. Heather Tarleton was supported by NIH T32 CA09142 (PI: ZF Zhang).


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Boyd JM, Subramanian T, Schaeper U, La Regina M, Bayley S, Chinnadurai G. A region in the C-terminus of adenovirus 2/5 E1a protein is required for association with a cellular phosphoprotein and important for the negative modulation of T24-ras mediated transformation, tumorigenesis and metastasis. EMBO Journal. 1993;12:469–478. [PubMed]
  • Chinnadurai G. CtBP, an unconventional transcriptional corepressor in development and oncogenesis. Molecular and Cellular Biology. 2002;9:213–224. [PubMed]
  • Choi K, Kennedy M, Kazarov A, Papadimitriou JC, Keller G. A Common Precursor for Hematopoietic and Endothelial Cells. Development. 1998;125:725–732. [PubMed]
  • D’Souza SL, Elefanty AG, Keller G. SCL/Tal-1 is Essential for Hematopoietic Commitment of the Hemangioblast But Not For Its Development. Blood. 2005;105:3862–3870. [PubMed]
  • Fehling HJ, Lacaud G, Kubo A, Kennedy M, Robertson S, Keller G, Kouskoff V. Tracking Mesoderm Induction and Its Specification to the Hemangioblast During Embryonic Stem Cell Differentiation. Development. 2003;130:4217–4227. [PubMed]
  • Fox AH, Liew C, Holmes M, Kowalski K, Makay J, Crossley M. Transcriptional cofactors of the FOG family interact with GATA proteins by means of multiple zinc fingers. EMBO Journal. 1999;18:2812–2822. [PubMed]
  • Furusawa T, Moribe H, Kondoh H, Higashi Y. Identification of CtBP1 and CtBP2 as corepressors of zinc finger-homeodomain factor delta EF1. Molecular and Cellular Biology. 1999;19:8581–8590. [PMC free article] [PubMed]
  • Galan-Caridad JM, Harel S, Arenzana T, Hou ZE, Doetsch FK, Mirny LA, Reizis B. Zfx controls self-renewal of embryonic and hematopoietic stem cells. Cell. 2007;129:345–357. [PMC free article] [PubMed]
  • Hassel S, Eichner A, Yakymovyeh M, Hellman U, Knaus P, Souchelnytskyi S. Proteins associated with type II bone morphogenetic protein receptor (BMPR-II) and identified by two-dimensional gel electrophoresis and mass spectrometry. Proteomics. 2004;4:1346–1358. [PubMed]
  • Hemavathy K, Hu X, Ashraf SI, Small SJ, Ip YT. The repressor function of snail is required for Drosophila gastrulation and is not replaceable by Escargot or Worniu. Developmental Biology. 2004;269:411–420. [PubMed]
  • Hildebrand J, Soriano P. Overlapping and unique roles for C-terminal binding protein 1 (CtBP1) and CtBP2 during mouse development. Molecular and Cellular Biology. 2002;22:5296–5307. [PMC free article] [PubMed]
  • Holmes M, Turner J, Fox A, Chisholm O, Crossley M, Chong B. hFOG-2, a novel zinc finger protein, binds the co-repressor mCtBP2 and modulates GATA-mediated activation. Journal of Biological Chemistry. 1999;274:23491–23498. [PubMed]
  • Ivanova NB, Dimos JT, Schaniel C, Hackney JA, Moore KA, Lemischka IR. A stem cell molecular signature. Science. 2002;298:601–604. [PubMed]
  • Ivanova NB, Dobrin R, Lu R, Kotenko I, Levorse J, DeCoste C, Schafer X, Lun Y, Lemischka IR. Dissecting self-renewal in stem cells with RNA interference. Nature. 2006;442:533–538. [PubMed]
  • Jackson AL, Bartz SR, Schelter J, Kobayashi SV, Burchard J, Mao M, Li B, Cavet G, Linsley PS. Expression profiling reveals off-target gene regulation by RNAi. Nature Biotechnology. 2003;21:635–637. [PubMed]
  • Kubo A, Shinozaki K, Shannon JM, Kouskoff V, Kennedy M, Woo S, Fehling HJ, Keller G. Development of definitive endoderm from embryonic stem cells in culture. development. 2004;131:1651–1662. [PubMed]
  • Lacaud G, Keller G, Kouskoff V. Tracking mesoderm formation and specification to the hemangioblast in vitro. Trends Cardiovasc Med. 2004;14:314–317. [PubMed]
  • Lecuyer E, Hoang T. SCL: from the origin of hematopoiesis to stem cells and leukemia. Experimental Hematology. 2004;32:11–24. [PubMed]
  • Lin X, Liang YY, Sun B, Liang M, Shi Y, Brunicardi FC, Shi Y, Feng XH. Smad6 recruits transcription corepressor CtBP to repress bone morphogenetic protein-induced transcription. Molecular and Cellular Biology. 2003;23:9081–9093. [PMC free article] [PubMed]
  • Lois C, Hong EJ, Pease S, Brown EJ, Baltimore D. Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors. Science. 2002;295:868–872. [PubMed]
  • Murakami A, Ishida S, Thurlow J, Revest JM, Dickson C. SOX6 binds CtBP2 to repress transcription from the Fgf-3 promoter. Nucleic Acids Research. 2001;29:3347–3355. [PMC free article] [PubMed]
  • Niwa H, Ogawa K, Shimosato D, Adachi K. A parallel circuit of LIF signaling pathways maintains pluripotency of mouse ES cells. Nature. 2009;460:118–122. [PubMed]
  • Okuda T, van Deursen J, Hiebert SW, Grosveld G, Downing JR. AML1, the target of multiple chromosomal translocations in human leukemia, is essential for normal fetal liver hematopoiesis. Cell. 1996;84:321–330. [PubMed]
  • Ramalho-Santos M. Stemness: Transcriptional Profiling of Embryonic Adult Stem Cells. Science. 2002;298:597–600. [PubMed]
  • Robertson SM, Kennedy M, Shannon JM, Keller G. A transitional stage in the commitment of mesoderm to hematopoiesis requiring the transcription factor SCL/Tal-1. Development. 2000;127:2447–2459. [PubMed]
  • Schaeper U, Boyd JM, Verma S, Uhlmann E, Subramanian T, Chinnadurai G. Molecular cloning and characterization of a cellular phosphoprotein that interacts with a conserved C-terminal domain of adenovirus E1A involved in negative modulation of oncogenic transformation. PNAS. 1995;92:10467–10471. [PubMed]
  • Sollerbrant K, Chinnadurai G, Svensson C. The CtBP binding domain in the adenovirus E1A protein controls CR1-dependent transactivation. Nucleic Acids Research. 1996;24:2578–2584. [PMC free article] [PubMed]
  • Turner J, Crossley M. Cloning and characterization of mCtBP2, a co-repressor that associates with basic Kruppel-like factor and other mammalian transcriptional regulators. EMBO Journal. 1998;17:5129–5140. [PubMed]
  • Turner J, Crossley M. The CtBP family: enigmatic and enzymatic transcriptional co-repressors. Bioessays. 2001;23:683–690. [PubMed]
  • Turner J, Nicholas H, Bishop D, Matthews JM, Crossley M. The LIM protein FHL3 binds basic Kruppel-like factor/Kruppel-like factor 3 and its co-repressor C-terminal-binding protein 2. Journal of Biological Chemistry. 2003;278:12786–12795. [PubMed]
  • Wang Q, Stacy T, Binder M, Marin-Padilla M, Sharpe AH, Speck NA. Disruption of the Cbfa2 gene causes necrosis and hemorrhaging in the central nervous system and blocks definitive hematopoiesis. Proceedings of the National Academy of Sciences of the USA. 1996;93:3444–3449. [PubMed]
  • Yamaguchi TP, Takada S, Yoshikawa Y, Wu N, McMahon AP. T (Brachyury) is a direct target of Wnt3a during paraxial mesoderm specification. Genes Dev. 1999;13:3185–3190. [PubMed]
  • Zhao LJ, Subramanian T, Zhou Y, Chinnadurai G. Acetylation by p300 regulates nuclear localization and function of the transcriptional corepressor CtBP2. Journal of Biological Chemistry. 2006;281:4183–4189. [PubMed]