Identification of factors involved in the maintenance of embryonic stem cell self renewal and pluripotency 2

An RNAi screen in mouse ES cells E14 using OCT4 and Nanog promoter activity as readouts was then performed in joint effort with my colleagues to further narrow down the choice of candid

Chapter2

Results and Discussion

2.1 ES cell secretion of LEFTY2 maintains pluripotency

2.1.1 Introduction

The molecular basis of self-renewal and the maintenance of pluripotency in ES cells have yet to be fully elucidated Here, I identify additional factors and delineate additional pathways important for “stemness” using a candidate gene approach was adopted ES cells are derived from the ICM of blastocysts and are similar (although not completely identical) to the pluripotent cells of the ICM Hence, to shortlist candidates in

an informed and intelligent manner, extensive perusal of literature, particularly pertaining

to genomic studies on blastocysts, preimplantation embryos, the inner cell mass and ES cells were carried out

Some published literature that were exploited to shortlist candidates include

MPSS data by Wei et al (2005), ChIP-PET data by Loh et al (2006) performed on

mouse ES cells, ChIP-chip data by Boyer et al (2005) performed on human ES cells and other genomic data such as gene expression profile changes during mouse preimplantation development (Hamatani et al., 2004)

Several criteria were used in the selection of genes The first criterion was the expression level of the candidate genes in pluripotent cell types of both mouse and human,

such as in ES cells in the in-vitro scenario, primitive ectoderm, epiblast and the inner cell mass etc when we talk about relevance to in vivo physiology Genes that are highly

expressed in a particular cell type may play a critical role in the regulation of cellular

processes specific to that cell The in depth transcriptome profiling by Wei et al (2005) using MPSS also highlighted a list of genes with differential expression patterns between

ES cells and differentiated cells and these may be prime candidates to examine further for a role in ES cell self renewal Of primary interest are genes whose expressions are restricted to the embryonic stages, as pluripotency is characteristic of embryo derived stem cells

The similarity in co-expression pattern between the candidate genes and Oct4,

Nanog or Sox2 was also used as a selection criterion, as the co-expression suggests that

the genes can potentially regulate or are regulated by Oct4, Sox2 and Nanog Little is currently known about what is upstream of Oct4, Sox2 and Nanog The studies by Loh et

al and Boyer et al have however revealed the putative downstream targets of OCT4, SOX2 and NANOG These targets were of particular interest as it is hypothesized that these genes form key nodes in the OCT4/SOX2/NANOG network and plays major functional roles in ES cells, possibly as gatekeepers against differentiation by repressing

differentiation inducing genes such as Gata6 and Lhx1

Genes linked to the WNT pathway such as Zfx were also included This is because

WNT is a highly critical conserved pathway that may be critical in maintaining ES cell self renewal The use of a synthetic pharmacological inhibitor of GSK3 in ES cells has been reported by Sato et al to sustain the short term pluripotency and self-renewal via activation of the Wnt pathway (Sato et al., 2004)

Using these criteria, 70 candidate genes were shortlisted An RNAi screen in

mouse ES cells (E14) using OCT4 and Nanog promoter activity as readouts was then

performed in joint effort with my colleagues to further narrow down the choice of candidate genes This thesis documents only the 9 candidate genes for which I was solely responsible for the cloning of their shRNAs and measurements of OCT4 / Nanog promoter activity in their knocked down cells (See Appendix) The rationale behind the

use of the luciferase assay is as such: In wild type ES cells, OCT4 and Nanog promoters

are active because the factors that regulate it are present However, when the ES cells

differentiate, there is a consequential shutdown of the OCT4 and Nanog promoters This

would lead to a reduction in measured luciferase activity If any gene is important for the

maintenance of the undifferentiated state of the ES cells or regulation of Oct4, Sox2 or

Nanog, ablating it by RNAi would reduce the luciferase activity Co-transfection of the OCT4 or Nanog promoter luciferase construct with short-hairpin RNA (shRNA) over-

expression constructs can hence enable us to rapidly and effectively screen for candidate genes that have a biological effect in maintaining the undifferentiated cell state (Figure 2)

One interesting factor identified from the screen is Lefty2, a secreted protein well

known for its role in regulating Nodal signaling during embryo development Data that

are a validation to the functional importance of Lefty2 in ES cell biology are presented in this chapter The RNAi screen also identified Lefty1, the homolog of Lefty2 to be a potential candidate, since its depletion by RNAi also perturbed Oct4 and Nanog promoter activities Here, data that show a role for Lefty1 in ES cell self renewal is presented The

mechanism via which Lefty 1 and 2 operates to maintain ES cell self renewal is also

addressed

Figure 2a OCT4-promoter luciferase reporter comprising 3 kb of the Human OCT4

promoter upstream of the transcription start site (TSS) for firefly luciferase The SOX2 binding site is indicated (DE- distal enhancer; PE- proximal enhancer; PP- proximal promoter; CR- human-mouse conserved region, ATG- translation start codon; TGA- translation stop codon; PolyA- polyadenylation signal)

OCT4-Figure 2b Nanog-promoter luciferase reporter comprising mouse Nanog sequences (from

positions -289 to +117 relative to the transcription start site) upstream of firefly luciferase

reporter Position of OCT4-SOX2 binding site within Nanog promoter is indicated

Nanog transcription start site is defined by the position of the furthest 5' EST in public

database (NCBI) (ATG- translation start codon; TGA- translation stop codon; PolyA- polyadenylation signal)

Human OCT4 promoter TSS TGA

Figure 2c A schematic representation of the pSuper.puro vector which was used for the overexpression of shRNA The vector consists of a H1 RNA polymerase III promoter that drives the endogenous production of short hairpin RNA (shRNA) Oligo duplexes were inserted between the BglII and HindIII restriction sites The vector consists of a PGK-pur cassette that enables for positive selection with puromycin (Pur- puromycin)

2.1.2 Results

2.1.2.1 Identification of Lefty2 as a candidate for self renewal and pluripotency from

a high throughput RNAi screen

To ensure reliability and specificity of the silencing effect, 2-4 shRNA constructs were designed for each gene to be knocked down The RNAi experiment was performed

on feeder-free E14 mouse ES cell cultured in serum based ES cell growth medium supplemented with LIF Puromycin selection was used as a means to select for cells transfected with the shRNAs (Figure 3a) The endogenous levels of the targeted genes could be reduced to less than 50% in most cases by third day of puromycin selection It was found that at least 50% of the candidate genes selected for knockdown were able to

reduce the Oct4 luciferase activity, by 25% or more Most of the genes which perturbed

Oct4 promoter activity such as Rex2, Lefty1 and Lefty2 were also able to do the same for

some were transcription factors such as Zfx and Rex2 and some were signaling molecules

like the Lefty proteins which are involved in the Nodal pathway This chapter first

describes the functional dissection of Lefty2, followed by that of its homolog, Lefty1

Figure 3a The application of shRNA coupled with puromycin drug selection in E14 mouse ES cells was able to reduce the endogenous levels of targeted gene candidates to less than 50% in most cases, as evidenced by quantitative PCR analysis of mRNA level

of candidates Mean values ± standard error of the mean (SEM) are plotted as percentages relative to scrambled shRNA control (100%) for qPCR measurement of gene candidates Transfection experiments were performed in technical triplicates for each biological sample and in biological duplicates (n=2)

Figure 3b Knockdown of factors such as Lefty2, Lefty1, Zfx and Rex2 perturbed Oct4 promoter activity The strategy of RNAi knockdown of ES cell-specific genes coupled with an Oct4 promoter luciferase reporter was employed to test candidate genes for

Figure 3c Genes such as Lefty2, Lefty1 and Rex2 that perturbed OCT4 promoter activities also affected Nanog promoter activity The strategy of RNAi knockdown of ES cell-specific genes coupled with a Nanog promoter luciferase reporter was employed to test candidate genes which can perturb Nanog promoter activity when depleted Oct4 and Nanog RNAi served as positive controls

2.1.2.2 Lefty2 shRNAs reduce Lefty2 expression level at both the transcript and

protein level

ES cells will differentiate, or perhaps undergo cell death when deprived of signals that

are required for their self renewal To ascertain if Lefty2 plays a role maintaining ES cells, shRNAs were used to deplete Lefty2 in E14 cells and the Lefty2 knocked down cells were

observed for signs of differentiation This experiment was carried out under normal mouse ES cell growth condition; that is in the presence of 15% serum and LIF To ensure the reproducibility, reliability and specificity of the effects of RNAi, two short hairpin RNA (shRNA) constructs targeting different regions of the transcripts were used shRNA1 and shRNA3 target different region of the 3’ untranslated region (3’UTR) of the

Lefty2 transcript Both constructs were very effective in reducing the endogenous level of Lefty2 transcripts qPCR analysis shows that shRNA1 resulted in a 74% reduction

relative to the non targeting control while the reduction achieved with shRNA3 was 80%

(Figure 4a) The relative transcript levels between Lefty2 knockdown samples and scrambled control were tabulated using the Ct method (See Appendix) -Actin was used as the housekeeping gene for normalization purpose The observed decrease in Lefty protein level relative to the non targeting control reflects the changes in gene expression level induced by RNAi (Figure 4b)

Figure 4a Lefty2 RNAi in E14 mouse ES cells effectively reduced Lefty2 transcript

levels by more than 75% Mean values± standard error of the mean (SEM) are plotted as percentages relative to the scrambled control (100%) Transfection experiments were performed in technical triplicates for each biological sample and in biological duplicates

(n=2) The two shRNAs used target different regions of Lefty2 mRNA target different region of the 3’ untranslated region (3’UTR) of the Lefty2 transcript

Figure 4b Lefty2 RNAi caused a decrease in endogenous Lefty protein levels, as evidenced by western blot analysis Lefty bands represent the processed form, ie the 34kDa form of total Lefty proteins, ie both Lefty1 and Lefty2 The intensities of the bands were compared to the loading control -Actin

2.1.2.3 RNAi knockdown of Lefty2 caused mES cell differentiation, even in presence

of LIF

Various parameters were investigated to assess the differentiation status of the ES cell upon Lefty2 RNAi knockdown compared to the control As a start, constructs

containing either the Oct4 promoter or Nanog promoter that contain OCT4/SOX2

binding sites driving the transcription of a luciferase reporter gene were used to measure the transcription activity of these pluripotency gene promoters (Figure 2) This provided a rapid and sensitive readout to detect any loss of pluripotency and onset of differentiation

in Lefty2 RNAi knockdown samples Two independent shRNAs against Lefty2 caused significant reductions of Nanog promoter activity in ES cells cultured in the presence of serum and LIF Nanog luciferase activity for Lefty2 RNAi knockdown cells dropped by

40% compared to control cells for both shRNAs Transient overexpression of both the

shRNAs in E14 cells also reduced Oct4 promoter activity shRNA1 induced a weaker attenuation of Oct4 promoter activity by 18% while shRNA3 caused a more dramatic

reduction of 52% (Figure 5)

The phenotypic effect of Lefty2 RNAi on ES cells was then investigated By the third day of puromycin selection, the cells underwent striking morphological changes

indicative of differentiation upon Lefty2 knockdown even in the presence of LIF The

cells lost their compacted growth morphology Lefty2 RNAi led to flattening and disintegration of the ES colonies, with some cells assuming the appearance of fibroblast Cells transfected with an equivalent amount of non targeting shRNA however did not

undergo any morphological changes They were still ES like and maintained tight colony structures (Figure 6)

Figure 5 Lefty2 RNAi caused a reduction of both Oct4 and Nanog promoter activity

in E14 mES cells To correct for transfection efficiency, firefly luciferase activity was

normalized to Renilla luciferase activity, and expressed as relative percentage change

to that of the scrambled control Experiments were performed in biological triplicates (n=3), where error bars denote standard error (SE)

Pluripotent ES colonies stain positively for alkaline phosphatase Fewer cells stained for alkaline phosphatase on the fourth day of drug selection when Lefty2 RNAi was applied when compared to the non-silencing control samples This again suggests that Lefty2 RNAi induces ES cell differentiation and this data is consistent with the morphological observation (Figure 6)

Figure 6 Lefty2 is required for mouse ES cell self renewal Assessment of morphlogy

indicated E14 mouse ES cell differentiation upon Lefty2 RNAi, even in presence of LIF

Lefty2 knockdown led to loss of ES cell characteristics such as clustered growth morphology and alkaline phosphatase staining An undifferentiated phenotype was

however maintained when a scrambled control shRNA was used Puromycin (1 g/ml) selection was performed 24 h post-transfection for three days Experiments were carried

out in duplicates Magnification used was 100x Two Lefty2 shRNA constructs were used

to ensure reproducibility and specificity of the silencing effect

The changes in expression of key pluripotency genes were also assayed to assess

the loss of pluripotency after Lefty2 knockdown Quantitative PCR on total RNA extracted from Lefty2 RNAi transfected samples subjected to four days of puromycin selection was performed A reduction in Oct4 and Nanog transcript level were detected This is consistent with the changes observed with Nanog and Oct4 promoter activity upon RNAi The levels of other key pluripotency genes such as Sox2 and Utf1 were also downregulated after Lefty2 knockdown Both shRNAs yielded consistent results

(Figure7) Partial loss of Lefty2 also led to reduction of Nanog protein expression in LIF

supplemented cultures, as observed with immunofluorescence microscopy (Figure 8)

Figure 7 Lefty2 knockdown elicited the downregulation of pluripotency gene expression such as Oct4, Nanog, Sox2 and Utf1at the transcript level, as observed with

quantitative PCR analysis Experiments were performed in biological duplicates and duplicate Ct values were obtained for each of the biological replicates (n=2) Error bars denote standard error of the mean (SEM) Mean values ± SEM are plotted as percentages relative to the scrambled shRNA control (100%)

To determine what cell types the Lefty2 RNAi knockdown cells were

differentiating into, I assayed for the expression of lineage specific gene markers In culture conditions containing serum and LIF, elevated mRNA levels of the early

primitive ectoderm marker (Fgf5) and mesendodermal markers (Sox7, Cerberus1 and

Pitx2) were observed for Lefty2 RNAi knockdown cells relative to non targeting control

(Figure 9) This suggests that the cells with reduced Lefty2 were differentiating into epiblast derived cell types upon Lefty2 knockdown, possibly into the mesendodermal cell fate To elucidate in a more definitive manner the lineage the Lefty2 RNAi transfected cells were differentiating into, LIF was withdrawn from Lefty2 RNAi cultures as it may

Figure 8 Partial loss of Lefty2 led to reduction of Nanog protein expression in LIF

supplemented E14 mouse ES cell cultures, as observed with immunofluorescence microscopy Cell nuclei were stained with DAPI Magnification used was100x

analysis was performed on day6 post transfection samples Directed reduction of Lefty2 led to a significant upregulation of more mesendodemal markers: Brachyury (T), Mixl1,

Pitx2 and Claudin 6 for both shRNAs shRNA3 induced a strong upregulation of 12 fold

for Brachyury, 6 fold for Mixl1, 4 fold for Pitx2 and 2 fold for Claudin6 shRNA1 also

induced a similar panel of genes as that for shRNA3, albeit at a lower level In addition, the two-fold upregulation of Cadherin11, another mesendodermal marker, was also

induced by shRNA1 The expression of ectodermal markers such as Nestin, Pax6 and

Neurod1, and endodermal markers like Hnf4, Sox17 and Gata4 also prominently showed

no significant difference between the Lefty2 RNAi samples and the control (Figure 10a)

These data suggests that Lefty2 RNAi knockdown induces mesendodermal cell fates specifically and Lefty2 may be required to maintain ES cell self renewal by blocking

mesendodermal differentiation

Figure 9 Lefty2 knockdown in presence of serum and LIF led to upregulation of differentiation markers in E14 ES cells

Quantitative PCR analysis of germ layer markers in Lefty2 knockdown cells cultured with LIF indicated an upregulated expression

level of markers like Sox7 over controls Error bar represent standard error of the mean (SEM) of 2 biological replicates The

0 200 400 600 800 1000 1200 1400 1600 1800

Figure 10a Lefty2 knockdown using RNAi3 in the absence of LIF led to differentiation of E14 ES cell to mesendoderm Quantitative PCR analysis of germ layer markers in Lefty2 knockdown cells cultured under LIF free condition indicates an elevated

Ectoderm Endoderm

Mesendoderm Mesoderm

Figure 10b Photomicrographs of E14 cells cultured in serum based, LIF free ES cell growth medium show that by day4 of Lefty2 RNAi, cells were already morphologically different from control Controls remained as homogenous ES-like clusters while population transfected with Lefty2 shRNA contained differentiated cells Magnification used was 200x

Secondary replating assays were also used to compare the extent of differentiation between the RNAi samples and the control Withdrawal of LIF was used in this assay to

allow Lefty2 shRNA knockdown to differentiate the ES cells The cells were then

dissociated with trypsin and replated at single cell density to allow the ES cells to expand into colonies It was interesting to note that the number of secondary ES colonies that emerged from the differentiated cells transfected with the scrambled control replated on

MEFs was higher than that obtained for cells transfected with both the Lefty2 shRNAs

The number of colonies observed for the control was almost twice that relative to shRNA3 at two different seeding densities of 300 and 500 cells per 56cm2 (surface area

of each well of a 6 well dish is 56cm2) When 300 cells were reseeded, the number of secondary colonies formed for cells transfected with shRNA1 was only 75% of that

obtained with the non targeting control cells This data suggests that Lefty2 RNAi

knockdown also affects ES cell colony forming potential due to differentiation (Figure 11)

0 10 20 30 40 50 60 70

500 cells replated

0 5 10 15 20 25 30 35 40 45 50

Figure 11 ES cells with reduced Lefty2 had significantly reduced capacity to form

colonies when replated onto feeders, indicating greater extent of differentiation as compared to control cells Secondary replating assay was used to compare the extent

of differentiation between Lefty2 RNAi cells and control cells a) Two seeding densities (300 and 500 cells) were performed for one of the shRNAs while b) 300 cells were replated after RNAi treatment for both the shRNAs Error bars correspond to

four biological replicates, and P <0.01 as determined with non paired t-test

11a

11b

2.1.2.4 ES cell differentiation induced by Lefty2 RNAi is specific to loss of Lefty2

shRNAs achieve their effects through a complex array of molecular interaction, including, but not necessarily limited to those that confer the desired nucleotide sequence-based specificity-that is, they may cause off target effects (Moffat and Sabatini,

2006) Up to this point, consistent results obtained using two Lefty2 shRNAs targeting different regions of the Lefty2 transcript have given me the confidence that the

phenotypes observed are gene specific and are not mere experimental artifacts or reagent

specific For further proof of specificity, rescue experiments for the Lefty2 knockdown

experiments were also carried out

The first set of rescue experiment involved the cotransfection of a vector

overexpressing Lefty2 RNAi resistant transcripts with Lefty2 shRNA Dual drug selection

of gentamycin (G418) and puromycin was used to select for cells that had incorporated both vectors For the rescue experiment, shRNA3 was used as it gave the best knockdown

A construct overexpressing the open reading frame of Lefty2 (Lefty2 ORF construct) was

engineered (Figure 12a) This strategy allows the production of an abundance of Lefty2 transcripts that is resistant to RNAi targeting and degradation by shRNA3 since shRNA3 targets the 3’ untranslated region (3’UTR) and does not affect the coding region To

facilitate detection, the Lefty2 cDNA was tagged in frame with V5 at its carboxyl

terminus Biologically active forms of Lefty proteins require proper processing and secretion Hence, it was important to check that the ability to produce functional LEFTY2 was not compromised by the V5 tag Western blot analyses against the V5 tag detected

28kDa) in the medium conditioned by the transfected E14 cells (Figure 12b) This indicates proper processing and secretion and verified that the RNAi resistant overxpression construct could effectively produce LEFTY2

Lefty2 open reading frame with V5 tag

BclI NsiI

Figure 12a Vector map of pCAG_Lefty2V5 plasmid The Lefty2 gene is shown in blue The open reading frame coding for Lefty2 is inserted between the BclI and NsiI

restriction sites, upstream of the IRES-EGFP cassette A V5 tag has been cloned at the

3’ end of Lefty2 cDNA The vector consists of a CAG promoter that drives the endogenous production of Lefty2

First, the optimal ratio of overexpression construct to shRNA was determined

using Nanog level measurement as the readout It was found that cotransfection of 3ug and 4 ug of the Lefty2 ORF construct with 2ug shRNA3 could reverse the drop in Nanog level induced by shRNA3 Nanog level was remained unchanged when a ratio of 2ug

Lefty2 ORF construct to 1ug shRNA was utilized At this ratio, the co-transfection of

RNAi resistant overexpression construct with shRNA was also able to fully block the

reductions in expression level of pluripotency genes like Oct4 and Sox2 otherwise

Figure12b The precursor (42kDa) and processed forms of V5 tagged LEFTY2 (the

34kDa and 28kDa form) were detected in the medium conditioned by the Lefty2

overexpressing cells by Western blotting using anti-V5 antibody No signal was detected in the control The level of Lefty2 proteins in duplicate preparations of

medium conditioned by E14 cells transfected with the V5 tagged Lefty2

overexpression construct was analyzed using western blot Equivalent volume of medium conditioned by cells transfected with empty vector in exactly the same manner as the overexpression construct were analyzed in parallel

42kDa 34kDa

28kDa

LEFTY2 conditioned medium

CAG vector control conditioned medium

Figure 13a Co-transfection of 4µg RNAi-resistant overexpression construct with 2 µg

shRNA blocked reduction in pluripotency genes expression levels induced by Lefty2

knockdown This rescue experiment further confirms the specificity of Lefty2 RNAi Real-time PCR analysis was conducted 3 days after puromycin selection -Actin was used as an internal control for normalization Biological duplicates were used for real-time PCR analysis The measurements were done in duplicates and the average of the normalized ratio of target gene/ -actin was calculated and presented with standard error of the mean (SEM) (CAG-empty CAG vector; Non sil-non silencing RNAi; ORF-CAG vector overexpressing Lefty2 open reading frame)

Figure 13b Co-transfection of two different ratios of RNAi-resistant overexpression

construct and shRNA rescued the reduction in Nanog gene expression level induced by

Lefty2 knockdown Transfection of 4µg of overexpression plasmids was able to

completely block the effect of Lefty2 RNAi on Nanog gene expression This rescue experiment demonstrates the specificity of Lefty2 RNAi Real-time PCR analysis was conducted 3 days after selection -Actin was used as an internal control for

normalization Biological duplicates were used for real-time PCR analysis Measurements were done in duplicates and the average of the normalized ratio of target

gene/ -Actin was calculated and presented with standard error of the mean (SEM) (CAG

- empty CAG vector; Control RNAi-scrambled RNAi control; V5Lefty2-CAG vector

overexpressing V5 tagged Lefty2 open reading frame; Lefty2 RNAi- Lefty2 knockdown with Lefty2 shRNA3)

The second set of rescue experiment was performed by supplementing the RNAi cultures with LEFTY2 conditioned medium HEK293T was transfected with the V5

tagged Lefty2 overexpresssion construct and the conditioned medium was harvested from

these transfectants It was verified that the active form of Lefty2 proteins, ie the 34kDa product could be detected in the medium prior to using them for the rescue experiment

020406080100

120

Lefty2 RNAi+CAG vector

Scrambled RNAi+CAG vector

Lefty2 RNAi+V5Lefty2

or Lefty2shRNA3) + 4ug CAG

vector(empty or V5Lefty2)

(Figure 14a) To control for factors other than LEFTY2 that could have been released into the medium by 293T cells, conditioned medium were also harvested from HEK293Tcells transfected with the empty CAG vector In the presence of Lefty2

conditioned medium, Lefty2 RNAi knockdown cells showed greater staining for alkaline

phosphatase as compared to those that were conditioned with the control medium, and alkaline phosphatase staining was equivalent to that observed with the scrambled control

cells Morphological changes induced by Lefty2 RNAi also were eliminated in presence

of LEFTY2 conditioned medium (Figure 14b)

Figure 14a V5 tagged LEFTY2 could be detected in the conditioned medium used for the rescue experiment, as shown by western blot analysis To produce Lefty2 conditioned medium for the rescue experiment, 293T cells were transfected with pCAG empty vector

(Control) or with the open reading frame Lefty2 with a V5 tag in the carboxyl-terminal domain (pCAG-Lefty2-V5) Secreted and processed Lefty (the 34kDa form) could be

detected in the conditioned media by Western blotting using anti-V5 antibody, while no protein was detected in control (CM-conditioned medium)

a

Figure 14b Addition of LEFTY2 conditioned medium to Lefty2 RNAi cultures blocked differentiation induced by RNAi and maintained alkaline phosphatase positive

ES cells in RNAi treated cells Magnification used was 100x This rescue experiment

conformed the specificity of the Lefty2 shRNAs (Control- non silencing shRNA;

LEFTY2 CM-LEFTY2 conditioned medium; 293T CM-Medium used to control for factors other than LEFTY2 that could have been released into the medium by 293T cells during conditioned medium production process)

Put together, the data from the rescue experiments further strengthen our

conclusion that Lefty2 plays a role in regulating the ES cell state and that phenotypic effects that resulted from loss of Lefty2 function are specific and not off target effects of

RNAi knockdown

2.1.2.5 Lefty1, a homolog of Lefty2 is also required to maintain the ES cell state

There are two Lefty homologues in mouse, one of which is Lefty2 and the other is

Lefty1 The mature regions of Lefty1 and Lefty2 proteins are highly conserved, with a

homology of 97%, excluding the four carboxyl terminal amino acid residues which are variable In comparison, the proprotein region is less conserved between the two Lefty proteins, with an identity of 70% It has been suggested that the Lefty proteins evolved independently from duplication of a single Lefty gene (Tabibzadeh and Hemmati-Brivanlou, 2006)

Like Lefty2, Lefty1 is abundant in both human and murine embryonic stem cells and blastocysts Lefty1 is also regulated by key ES transcription factors Nakatake et al has recently reported that Lefty1 is a direct target of OCT3/4 They showed that OCT3/4 was able to activate an enhancer upstream of the Lefty1 promoter in cooperation with

SOX2 and Kruppel like factor 4 (KLF4) KLF4 is a transcription factor that has been shown to contribute to ES cell self renewal In the same report, it was stated that KLF4 could support LIF-independent self renewal (Nakatake et al., 2006) Li et al has also

reported that overexpression of Klf4 in mouse ES cells prevented differentiation of

embryoid bodies (Li et al., 2005) Ng et al’s chIP location analysis has also revealed that

Lefty1 is a common target bound by OCT4 and NANOG in mouse ES cells Put together,

all these evidence suggest that the regulation of Lefty1 by OCT4, KLF4 and NANOG

may be part of the ES cell self renewal mechanism

The role of Lefty1 in regulation of ES cell self renewal was hence dissected using loss of function approach First, the effectiveness of the shRNAs was verified Lefty1 shRNA1could induce a 31% decrease for the level of Lefty1 transcript whereas the

reduction that could be achieved with transfection of shRNA4 was 55% (Figure 15) Similar to the effects of Lefty2 RNAi, application of Lefty1 RNAi abrogated the ability

of ES cell to self-renew Two independent shRNAs against Lefty1 caused significant reductions of Nanog promoter activity in ES cells cultured in presence of serum and LIF

Nanog luciferase activity for Lefty2 RNAi cells dropped by ~6 folds (normalized to

control cells) for both shRNAs Transient overexpression of both the shRNAs in E14

cells also depressed Oct4 promoter activity shRNA1 induced a substantial lowering of

Oct4 promoter activity by 74% while shRNA4 caused a more subtle reduction of 37%

(Figure 16) Partial loss of Lefty1 function caused changes in ES cell morphology and the

ability to stain for alkaline phosphatase (Figure 17) Also, there was a downregulation of the endogenous expression of pluripotency genes for Lefty1 RNAi samples relative to non silencing controls This is indicative of differentiation RNAi with shRNA1 led to

reduction of Nanog and Sox2 gene expression while RNAi with shRNA4 led to the detectable downregulation of Oct4 as well (Figure 15) It was observed that the extent of differentiation was more effective in samples transfected with shRNA4 as compared to

that for shRNA1 This correlates well with the observation that knockdown of Lefty1 was

more dramatic in samples transfected with shRNA4 as compared to that for shRNA1

Figure 15 Lefty1 shRNAs1 and 4 were effective in reducing endogenous Lefty1 transcript

level shRNA 4 was the most effective of the 4 shRNAs designed and could elicit the

downregulation of Lefty1 expression by more than 50% RNAi with shRNA1 led to reduction of Nanog and Sox2 gene expression while RNAi with shRNA4 led to the downregulation of Oct4 as well Experiments were performed in biological duplicates and

duplicate Ct values were obtained for each of the replicates (n=2) Error bars denote standard error of the mean (SEM) Mean values±SEM are plotted as percentages relative

to the non-targeting shRNA control (100%)

0 20 40 60 80 100

Oct4 RNAi RNAi 1 RNAi 4 Control Vector

Oct4 RNAi Nanog RNAi Lefty1 RNAi1 Lefty1 RNAi4 Control Vector

Figure 16 Lefty1 RNAi caused a reduction of both Oct4 and Nanog promoter activity in

E14 mES cells To correct for transfection efficiency, firefly luciferase activity was

normalized to Renilla luciferase activity, and expressed as relative percentage change to

that of the scrambled control Experiments were performed in biological triplicates (n=3) The error bars denote standard error

Figure 17 Lefty1 knockdown with shRNA1and shRNA4 led to abnormal colony

morphology (100x magnification) and reduced alkaline phsphatase staining (200x magnification) Non-targeting shRNA construct was used as negative control and the cells showed normal ES cell morphology and stained positively for alkaline phosphatase after RNAi

Figure 18 Lefty1 mRNA level increased by ~4 folds upon Lefty2 knockdown with both

shRNAs Experiments were performed in duplicates and duplicate Ct values were obtained for each of the replicates (n=2) Error bars denote standard error of the mean (SEM) Mean values±SEM are plotted as percentages relative to the non-targeting

RNAi4

0 50

Interestingly, there was an upregulation of Lefty1 transcript level by 3.9 and 4.2 folds upon Lefty2 depletion with Lefty2 shRNA1 and shRNA3 respectively This led to the speculation that Lefty1 was compensating for the loss of Lefty2, and that Lefty1 and

phenotypic rescue of Lefty2 RNAi with recombinant Lefty1 protein was carried out

When10-50ng/ml Lefty1 was added to the Lefty2 RNAi cultures, the cells did not

differentiate and did not lose their ability to stain for alkaline phosphatase, as did the knock down cultures with no exogenous Lefty supplemented This piece of data shows

that presence of Lefty1 abolishes the effect of Lefty2 RNAi, thereby suggesting that

Summing our data thus far, it can be surmised that Lefty1 and Lefty2 play a role in

blocking ES cell differentiation

Lefty2 RNAi

Lefty1 rnai1

50ng/ml

LEFTY1

Concentration

Figure 19 Supplementing Lefty2 RNAi cultures with recombinant Lefty1 protein

was able to block reduction in level of alkaline phosphatase staining induced by

Lefty2 RNAi Alkalline phosphatase staining was carried out 5 day post

transfection of Lefty2 shRNA3 The rescue effect could be observed for a range of concentration of Lefty1 protein (10-50ng/ml) added

2.1.2.6 Knockdown of LEFTYA induced human ES cell differentiation

Loss of function experiment using siRNA was also carried out in human ES

cells(Hues9) cultured under feeder free condition When the level of LEFTYA mRNA

was lowered by 46% using siRNA, the human ES cells underwent dramatic

differentiation Upon LEFTYA RNAi, there was a reduction in cell number, flattening of cell morphology, and a marked reduction of OCT4 to 29%, SOX2 to 52%, NANOG to 31% and ZFP42 to 25% relative to scarmbled siRNA was noted There was also an upregulation of the differentiation marker, GATA6 From these, it can be concluded that

LEFTYA is also important for ES cell self renewal in human This highlights the

importance of Lefty2 as a factor important for regulating ES cell fate across species

Figure 20a Knockdown of LeftyA in human ES cells by 46% using siRNA

downregulates OCT4, NANOG, SOX2 and ZFP42 expression, while up-regulating

GATA6 expression Values are relative to that obtained with scrambled siRNA control

qPCR was carried out on samples harvested day5 post transfection Experiments were performed in biological duplicates and triplicate Ct values were obtained for each of the

0 50

2.1.2.7 ES cells secrete LEFTY 1 and LEFTY2

The molecular mechanism via which Lefty2 operates to maintain ES cell self renewal was investigated in mouse ES cells Lefty is known to be a secreted protein We

hence wanted to first confirm that its site of action is extracellular in the ES cell context The secretion of Lefty proteins into the supernatant was determined Western blot

Figure 20b Photomicrograhs of day5 post transfection LeftyA RNAi cells shows reduced cell number and flattened morphology in samples transfected with 100nM LeftyA siRNA The extent of differentiation was comparable to cells transfected with Oct4 siRNA Hues9 cells transfected with scrambled siRNA grew with normal human

ES cells morphology RNAi transfections were carried out under feeder free conditions Magnification used was 100x

cells This western blot is an indication of the total Lefty protein level as the antibody used cannot distinguish specifically between Lefty1 and Lefty2 The precursor(42kDa) and proteolytically processed forms (34kDa and 28kDa) of total Lefty proteins were detected even in unconcentrated conditioned supernatant, indicative of strong secretion(Figure 21) This confirms its extracellular deposition

There is evidence that biologically active forms of Lefty proteins require processing of the precursor by members of the convertases family of enzymes (Ulloa et al., 2001) It must be noted however that the cleavage sites of Lefty1 and Lefty 2 are not entirely the same and that processing of Lefty1 and Lefty2 have slightly different requirement of convertases Lefty proteins have two RXXR cleavage sites The RGKR site is processed to produce the long form of polypeptide for both Lefty1 and Lefty2 The 28kDa form of Lefty1 results from cleavage of the RHGR site while that the short form

of Lefty2 results from cleavage of the RFER site Lefty1 precursor protein is processed

Figure 21 Lefty proteins are expressed in, processed and secreted by mouse ES cells

as shown by western blot analysis A high abundance of Lefty protein could be detected in medium conditioned by E14 cells The conditioned medium harvested were subjected to 2, 4 and 8 times concentration with Millipore Amicon® ultra-15 centrifugal filter units prior to analysis All three forms of Lefty proteins (the precursor, as well as the cleaved forms) could be detected (1x- Unconcentrated; 2x, 4x and 8x – two, four and eight times concentrated respectively)

1x 2x 4x 8x

by SPC1, SPC4,SPC6A(PC5A) and SPC6B at both sites while Lefty2 precursor is cleaved to the larger 34kDa product by SPC1 and SPC6A (Sakuma et al., 2002) The enzyme(s) that cleave the RFER site of Lefty2 precursor is however unknown currently The western blot analysis mentioned above however does not have the resolution power required to tease apart Lefty1 processing from Lefty2 processing Hence to specifically check that Lefty2, which is the subject of this thesis, is indeed cleaved and secreted by ES cells, more experiments were performed There is no commercially available antibody that can probe specifically for Lefty2 Hence to facilitate investigation, a V5 tagged

Lefty2 overexpression construct was used Like most other TGF- s, the Lefty precursor

proteins are cleaved post translation to release the C-terminus monomeric active protein Hence to enable the detection of the precursor and mature forms, the V5 tag has to be incorporated at the C terminus and not the N terminus There is also the possibility that tagging at the N-terminus will also disrupt secretion signals

Transfection with C terminus V5 tagged Lefty2 overexpression construct resulted

in expression of a 42kDa precursor, which could be proteolytically cleaved to release two polypeptides of 34kDa and 28kDa All three forms of polypeptides could also be detected

in the supernatant conditioned for 24 hours by the transfected cells (Figure 12b) This data tells us specifically that the enzymes and cellular machinery required to process Lefty2 are present in ES cells and assures us that Lefty2 processing is indeed active in ES cells and that mature Lefty2 proteins are secreted into and deposited in the ES cell microenvironment