In a key stage of development, individual cells form into organized tissues. Human embryonic stem cells allow us to recreate this early stage of embryo development in the lab. When grown in confined spaces, the cells organize into clusters that can then develop germ layers. Previous work using these clusters showed that a network of signaling proteins — including one called WNT — trigger human embryonic stem cells to form the initial clusters.
This briefly activated several genes that are known to help to form germ layers. However, this gene activity was not maintained for long enough to cause the stem cells to specialize and organize into layers. Yoney et al. The germ layer gene activity was maintained in this case, leading to the cells specializing and forming the inner two germ layers.
The next step is to understand how cells store the memory of the WNT signal. As well as aiding our understanding of development, it could also help us to understand situations where signaling goes wrong, such as cancer. The technique used here to follow signals in real time could also be used to study other biological signaling processes. In the early embryo, secreted morphogens regulating a limited number of signaling pathways carry the task of instructing dynamic and coordinated cell differentiation across the developing tissue.
In the day 6. To what extent this signaling cascade is involved in human gastrulation can now be investigated using in vitro models of early human embryos derived from human embryonic stem cells hESCs. We have previously shown that the first step of this signaling hierarchy is conserved in humans and that in response to BMP4, hESCs grown in geometrically confined colonies, self-organize to induce and pattern embryonic and extra-embryonic germ layers.
Ectoderm was specified at the center of circular colonies, extra-embryonic tissue at the edge and mesendoderm in between. We, therefore, termed this system a human gastruloid. Among them are: how can a single signaling pathway carry these two opposite functions before and after the onset of gastrulation? To what extent do the dynamics of SMAD signaling affect these readouts? Finally, to what extent do cells have a memory of past signaling? This induction, however, was not sustained and cells reverted back to pluripotency at later times. This implies an unexpected ability of human embryonic stem cells to record their signaling history without overt changes in fate.
The colonies were fixed and analyzed by immunofluorescence. All micropatterned culture experiments were performed on at least three separate occasions with similar results. Each line was also transfected with ePiggyBac transposable elements carrying a nuclear marker H2B-mCitrine or H2B-mCherry in order to analyze the response of individual cells Figure 2—figure supplement 2A. N-terminal SMAD fusion proteins were shown to function similarly to endogenous proteins in biochemical and cell-based assays Schmierer and Hill, Additionally, the SMAD response dynamics measured with our reporter lines, matched the behavior by of the endogenous proteins measured by immunofluorescence and western blotting as discussed below.
The response of cells at the colony edge B and near the colony center C as a function of time following BMP4 presentation. The intensity range was adjusted to the same minimum and maximum values in all images in both B and C. The intensity range was adjusted to the same minimum and maximum values in all images in both E and F. The single-cell nuclear mCitrine intensity was quantified and normalized to the single-cell cytoplasmic mCitrine signal. In our previous work we quantified the SMAD1 response detected by immunofluorescence as the number of positive nuclei within the colony Etoc et al.
Interestingly, the long-term SMAD2 nuclear level did not return completely to the pre-stimulus level. In order to study signaling dynamics at the single-cell level and to eliminate modifier influences on both SMAD branches coming from neighboring cells within the micropatterned colony, we performed the same experiment on dissociated cells grown under regular culture conditions Figure 3A.
The increase in the average nuclear signal as a function of time resulted from nearly all cells responding to the added BMP4, which is evident in the shift in the histogram of RFP-SMAD1 nuclear intensity Figure 3—figure supplement 1A. This is consistent with other data based on immunofluorescence Heemskerk et al. Although, the adaptive response is somewhat more pronounced in our experimental setup. This also demonstrates that modifying signals do not influence the dynamics of the response at the edge of the micropatterned colonies.
A Schematic outlining the single-cell experimental protocol. Images were acquired every 10 min. Similar results were obtained in two independent experiments. Data were collectively obtained from three independent experiments. We have previously shown that BMP4 signaling induces a sustained transcriptional response leading to gastruloid differentiation Warmflash et al. This is consistent with the stable nature of SMAD1 signaling presented above. They fell into three distinct groups. The first, which consisted of the majority of transcripts 2, , peaked at 2.
The second group, which consisted of transcripts, showed stable induction Figure 4A , orange box. Finally, the third group, which consisted of transcripts, represented genes that were stably or transiently down regulated upon ACTIVIN presentation and included genes that are involved in signaling pathways not previously associated with pluripotency or differentiation, such as insulin signaling and cAMP response Figure 4A , gray box and Figure 4—source data 3.
The gene z-score at each time point was calculated by subtracting the average and dividing by the standard deviation of the normalized read counts across all time points. An additional sample was collected that was left untreated for the 12 hr time course untr. Similar results were obtained in three independent experiments using the parental RUES2 line. Examination of the signaling hierarchy involved in gastruloid self-organization revealed the presence of feedback loops at all three levels.
Wnt signaling in development and disease
However, despite the induction of the expression of the ligands and inhibitors, the overall threshold of signaling was not sufficient to induce and maintain mesendodermal fates from either the BMP or the WNT pathway. First, we performed motif enrichment analysis on our gene groups.
- Wnt signaling pathway!
- Stochastic systems in merging phase space MVspa.
- Advances in Clinical Trial Biostatistics.
- Geographical information systems : trends and technologies!
- OECD Reviews of Tertiary Education: Croatia 2008!
- Related Information!
- The Way It Works?
In support of our hypothesis, the motifs were significantly enriched only in the transiently expressed genes of group one and not within group 2 or 3 Figure 4—figure supplement 1A and Figure 4—source data 4. This suggests that the gene regulatory network activated during the peak of SMAD2 signaling is associated with mesendodermal differentiation.
Although all groups contained some significant enrichment of genes from one or more of the mouse germ layers, group one displayed the most significant enrichment of endodermal genes Figure 4—figure supplement 1B and Figure 4—source data 5. However, the mesendodermal differentiation program is not maintained and cells return to pluripotency. However, more anterior mesendodermal lineages are completely eliminated and the embryos fail at gastrulation suggesting a critical role for SMAD2 in this process Vincent et al.
We next asked if the increase in the SMAD2 baseline after the peak response is required for the maintenance of pluripotency. Analysis of the parental RUES2 line confirmed the loss of pluripotency under the same experimental conditions Figure 5—figure supplement 1A. We conclude that the elevated baseline at the tail of the SMAD2 response is ligand dependent and responsible for maintaining the pluripotency program long-term.
Cells were fixed and analyzed by immunofluorescence IF. Even at the highest ligand concentrations pluripotency was maintained and no differentiation was observed. A Schematic outlining the 2 day experimental protocol. After the second day cells were fixed and analyzed by immunofluorescence IF. B Quantification of the final cell density in the experiment shown in A and in additional replicates. Following the tradition of experimental embryology, during the late s, Nieuwkoop began a series of explant and transplant experiments in the amphibian blastula that ultimately led to the discovery of ACTIVIN as a mesoderm inducer in the s Nieuwkoop, ; McDowell and Gurdon, When presented to isolated animal cap explants that normally give rise only to ectoderm, ACTIVIN was sufficient to induce different types of mesendodermal cells based on its concentration, and was thus qualified in principle as a morphogen Green and Smith, ; Wilson and Melton, Elegant genetic experiments also performed in the mouse placed NODAL signaling in a positive feedback loop that drives gastrulation.
BMP signaling subsequently induces WNT3 expression, which in turn induces high levels of NODAL in the proximal-posterior part of the embryo, which marks the site of primitive streak initiation Arnold and Robertson, Thus, a transient signaling response is compatible with a stable change in cell fate, as shown previously in a murine myoblast cell line Sorre et al.
Both pluripotency maintenance and mesendodermal fate acquisition are not affected by the loss of SMAD3. Our data provide evidence for a previously undetected level of signal integration that implies the presence of a cellular memory. The ability of embryonic cells to record WNT signals may be a broadly conserved and fundamental aspect of animal development, as shown by elegant experiments in Drosophilia where the authors demonstrate a morphogen effect for Dpp a BMP homologue after cells have lost contact with the source of WNT Alexandre et al. For example, cells of the Xenopus animal cap derived from the blastula stage embryo have been under maternal WNT influence as early as the two-cell stage, hours before ACTIVIN is presented to the explants.
These experiments demonstrate that both pathways are required for complete mesendodermal patterning without explicitly distinguishing their temporal relationship. Our study revises this point of view by bringing a temporal order. Controversies persist, since the embryo is rapidly developing, there are no live reporters of signaling, and fates are often assayed in early gastrulation rather than in the late blastula. In the mouse the earliest manifestation of the streak is the proximal-posterior expression of WNT, which extends distally.
Individual micropatterned coverslips CYTOO, Grenoble, France were washed one time with water for 5 min at room temperature RT to activate the surface according to the manufacturers recommendation. Cell seeding was performed as follows. The cell suspension was then placed over the coverslip in a 35 mm tissue culture dish. The sample was left unperturbed for 10 min at RT in order to achieve homogeneous seeding of the cells throughout the chip before being moved to the incubator.
The samples were incubated overnight. Live imaging was carried out in E7 imaging medium.
Western blotting was performed using standard techniques. Tiled image acquisition was used to acquire images of large areas of dishes or coverslips in four channels corresponding to DAPI and Alexa Fluor , , and Live imaging was performed on a spinning disk confocal microscope equipped with , , and nm lasers and an environmental chamber CellVoyager CV, Yokogawa.
After 4 days, blasticidin or puromycin was added for 7 days to select for cells that had been targeted. Colonies arising from a single cell were handpicked, expanded, and screened for correct targeting by PCR amplification of the genomic region and Sanger sequencing. Correctly targeted clones were subsequently transfected with ePiggyBac plasmids containing either H2B-mCitrine or H2B-mCherry cassettes to enable nuclear labeling for cell tracking Lacoste et al.
Individual clones were again isolated and controlled for normal karyotype G-banding and pluripotency maintenance. Two different sgRNAs were used to control for off-target effects. We modified the pX plasmid to also express a puromycin-2A-EGFP cassette to enrich for cells that had been successfully nucleofected. On the following day puromycin was added for 24 hr. Cells that survived selection were allowed to recover for several days. The resulting chromatograms were decomposed using the Tide web-based tool Brinkman et al.
Samples three pooled-wells were collected in 1 mL Trizol at 0, 2. Primer sequences and source are listed in Table 1. Samples three pooled-wells were collected in 1 mL Trizol at 1, 2. For images acquired from micropatterned cell culture experiments, stitching and colony detection were carried out as described previously using custom software written in MATLAB Etoc et al.
For analysis of the SMAD reporter lines on micropatterned colonies a single z-plane through the middle of the colony was analyzed at each time point. Background in each channel was removed by subtracting a minimum intensity image that was generated by taking the minimum value at each pixel over all images in that channel in the same z-plane. Vignetting was corrected by dividing by a flat-field image that was generated by normalizing the intensity of the minimum image to 1. This procedure corrects the intensity drop-off at the border of the images without further altering the average image intensity.
Digital images of individual blastocysts were acquired using AxioVision software Zeiss and a high-resolution black and white Zeiss AxioCam MRm digital camera.
For a negative control, IgG of the same species was used to replace primary antibody using the same concentration. Nuclear labeling, slide mounting, and image acquisition were performed as described above. Quantification of intensity of labeling in either the entire embryo or in the nuclei was performed using ImageJ software U. For labeling of the entire embryo, the area encompassing the entire embryo was selected and the mean intensity obtained using the Measure Analysis feature of ImageJ.
Background intensity was obtained from the area surrounding the embryo using the same technique and the value subtracted from embryo intensity. Labeling in the nucleus was determined using a similar technique except that the software was used to isolate nuclear regions within each embryo based on labeling with Hoechst Those cells with nuclei labeled with CDX2 were considered TE and the number of ICM cells was determined by subtracting number of TE cells from the total number of cells determined by counting number of nuclei labeled with Hoechst Blastocyst development was evaluated on Day 7 of development hpi.
The experiment was performed in 5 replicates using a total of COC and conventional semen from 10 different bulls. Both reagents were reconstituted as previously described [ 24 ]. Blastocyst development was evaluated on Day 7 of culture hpi , and a fraction of blastocysts with a clearly delineable blastocoel were randomly selected for immunolabeling of CTNNB1. The experiment was performed in seven replicates using a total of COC and conventional semen from 10 different bulls. Blastocyst development was assessed on Day 7 of development hpi. A fraction of blastocysts with a clearly delineable blastocoel was randomly selected and subjected to immunolabeling to determine the numbers of TE and ICM cells.
The experiment was performed in five replicates with a total of COC and conventional semen from 13 different bulls. Immunolabeling was evaluated for 91 blastocysts. The concentration of DKK1 was chosen because it was effective at blocking actions of WNT signaling agonist on development of bovine embryos to the blastocyst stage [ 24 ].
Morulae on Day 5 6 h after treatment , morulae on Day 6 24 h after treatment , and blastocysts with a clearly delineated blastocoel on Day 7 48 h after treatment were harvested and fixed for immunolocalization of YAP1 and CDX2. The experiment was performed in five replicates with a total of COC and conventional semen from seven bulls.
A total of individual embryos were assessed for immunofluorescence. Morulae were harvested 6 h after adding the treatment and immunolabeled for pJNK. The experiment was performed in four replicates with a total of COC and conventional semen from 11 different bulls. Actions of WNT7A on development were examined because the gene is highly expressed in the bovine endometrium [ 44 ] but not in the bovine embryo [ 24 , 45 ].
The concentration of WNT7A was chosen because it is the upper limit of the range suggested for biological activity of the product by the manufacturer, as determined by inhibition of Wnt-3a-induced alkaline phosphatase production in MC3T3-E1 cells. Blastocyst rate was assessed on Day 7 of development hpi and a fraction of blastocysts with clearly delineated blastocoel was randomly selected for further analyses.
This experiment was performed in six replicates with a total of COC and conventional semen from eight bulls. For experiments 9 and 10, embryos were produced in vitro. Treatments were added on Day 5 of development hpi. A total of 54 blastocysts were assessed by immunofluorescence. Treatments were added on Day 5 of development hpi , and blastocysts were harvested on Day 7 of culture for immunolabeling of pJNK as described above. A total of 42 individual blastocysts were assessed by immunofluorescence to quantify total pJNK. Effects of treatment on the percent of oocytes or cleaved embryos developing to the blastocyst stage were evaluated using Proc Glimmix of SAS for Windows, version 9.
The analysis was performed with the dependent variable considered as a binomial distribution, and treatments as fixed effects. Treatments were fixed effects and replicate was considered a random effect. Two experiments were conducted to determine whether activation of canonical WNT signaling would affect development of embryos to the blastocyst stage.
Effect of exposure of embryos to GSK3 inhibitor from Day 5 to 7 of development on the ability of embryos to develop to the blastocyst stage. Results of these experiments indicate that exogenous molecules that induce accumulation of CTNNB1 are detrimental for blastocyst formation. Neither treatment altered the proportion of oocytes or cleaved embryos developing to the blastocyst stage Table 3.
Effects of inhibition of endogenous WNT signaling from Day 5 to 7 of development with either Wnt-C59 or DKK1 on ability of embryos to develop to the blastocyst stage, and cell number of Day 7 blastocysts. These two experiments were designed to test whether DKK1 alters competence of embryos to become blastocysts and the number of TE and ICM cells in the blastocyst. There was no effect of addition of DKK1 from Day 5 to 7 of development on the proportion of oocytes or cleaved embryos that became blastocysts or on the numbers of ICM or TE cells in the resulting blastocysts.
This was true whether embryos were produced using conventional semen Table 3 or whether male and female embryos were tested separately after fertilization Table 4. Taken together, actions of DKK1 do not affect development of embryos to the blastocyst stage or number of cells in the blastocyst.
Effect of treatment of embryos with DKK1 from Day 5 to 7 of development on the ability of male and female embryos to develop to the blastocyst stage and cell number of Day 7 blastocysts.
By Day 7, both types of cells were confined to the TE. Quantification of nuclei that were positive for YAP1 and CDX2 indicated that the number of cells positive for both markers increased during development Figure 2B and F. Embryos treated with vehicle are represented by closed circles and solid lines while embryos treated with DKK1 are represented by broken lines and open circles.
Embryos were exposed to DKK1 at Day 5 of development. B Quantification of intensity of pJNK in whole embryonic area. C Quantification of intensity of pJNK in nuclear area. This WNT was chosen because WNT7A is highly expressed in the female reproductive tract [ 44 ] but is not expressed in the morula or blastocyst [ 24 , 45 ]. There was no effect of day of treatment or interaction of day with treatment. Effect of treatment of embryos with WNT7A from Day 1 to 7 or from Day 5 to 7 of development on the ability of embryos to develop to the blastocyst stage.
Effect of treatment of embryos with WNT7A from Day 5 to 7 of development on the ability of embryos to develop to the blastocyst stage and cell number of Day 7 blastocysts. D Quantification of intensity of pJNK in whole embryonic area. E Quantification of intensity of pJNK in nuclear area. Collectively, data suggest that embryo-derived WNTs are dispensable for blastocyst formation in bovine embryos but participate in regulation of ICM proliferation, likely through a mechanism independent of CTNNB1. There are two lines of evidence that endogenous WNT are not required for development to the blastocyst stage.
In earlier studies as well, there was no effect of DKK1 on development [ 24 , 30 ]. Similar to our findings, endogenous WNTs do not play a role in development to the blastocyst in mouse embryos. Blastocyst formation was not impaired in either Ctnnb1 or Porcn- deficient mice [ 27 , 28 ].
Also, there was no effect of Dkk1 on development to the blastocyst stage [ 30 ]. Although the cow parallels the mouse with respect to the dispensability of WNT signaling for development to the blastocyst stage, there may be divergence between species in role of embryo-derived WNT in formation of the ICM. The mechanism by which WNTs regulate number of cells in the ICM could potentially be mediated by the proportion of blastomeres in the morula that remain pluripotent after differentiation of the TE or by the degree of proliferation and apoptosis of cells in the ICM. Specific WNTs can promote differentiation [ 26 ] and apoptosis [ 53 ] and decrease proliferation [ 54 ].
The importance of the number of ICM and TE cells at the blastocyst stage for competence of the embryo to establish pregnancy has not been well defined in cattle. In the human, characteristics of the ICM appear to be a less important determinant of potential for development after transfer than are properties of the TE [ 57 — 60 ]. Additional evidence against a role for DKK1 in differentiation of the bovine blastocyst was the finding that DKK1 reduced accumulation of YAP1, a transcription factor important for TE formation in mouse [ 61 ] and did not affect amounts of the transcription factor CDX2 that is responsible for TE differentiation [ 62 ].
These findings stand in contrast to earlier studies in cattle [ 38 ] and pigs [ 31 ] that DKK1 increases the number of TE cells in the blastocyst. The reason for the discrepancy between current findings and earlier ones with respect to actions of DKK1 on TE numbers is not known. Embryo sex was one possible cause of variation in response of the embryo to DKK1 that was examined.
Indeed, changes in gene expression in the bovine morula induced by DKK1 were different in some cases for male embryos than for female embryos [ 63 ]. Sex has an even larger effect on response of the bovine embryo to colony stimulating factor 2 [ 64 , 65 ]. Despite these observations, the lack of effect of DKK1 on TE numbers was seen for both male and female embryos. Consequences of maternal WNT signaling are likely to depend on a complex array of factors including the abundance of specific WNT ligand, receptor and coreceptor availability, and presence of WNT regulatory molecules such as DKK1 and soluble FZD receptors, which are also expressed in the endometrium [ 44 ].
The fact that fertility in cows [ 34 ] and heifers [ 67 ] is associated with endometrial expression of DKK1 could reflect the importance of optimal WNT signaling for successful pregnancy. Not all maternally derived WNT are likely to inhibit development. WNT7A is not expressed in the bovine preimplantation embryo [ 24 , 45 ] but is highly expressed in bovine endometrium [ 44 ].
One of the objectives of the series of experiments documented here was to identify downstream pathways affected by DKK1. Bovine embryos treated with DKK1 were more likely to establish and maintain pregnancy after transfer to recipient cows than embryos not treated with DKK1 [ 38 ]. Also, expression of DKK1 in endometrium was lower for heifers diagnosed as infertile compared to heifers considered fertile [ 67 ] and was lower for endometrium of lactating cows than nonlactating cows [ 34 ].
To our knowledge, data presented here are first description of localization of YAP1 in the bovine embryo. Taken together, these data suggest that embryo-derived WNTs are dispensable for blastocyst formation in bovine embryos but do participate in the formation of the ICM. In contrast, exogenous WNTs can affect embryonic development in a positive or negative manner depending upon the nature of the WNT ligand and the downstream outcome.
Such a result implies that maternally derived WNT could play important roles in development of the preimplantation embryo. Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search. Article Navigation. Close mobile search navigation Article Navigation.
Volume Article Contents. Summary Sentence. Materials and methods. Supplementary data. Oxford Academic. Google Scholar.
Related Wnt Signaling in Embryonic Development
Copyright 2019 - All Right Reserved