It is widely known that a functional placenta is vital for normal embryonic development, but how much it may contribute to embryonic lethality has never before been systematically studied. Our research, published in Nature, demonstrates for the first time a remarkable co-association between embryonic lethality and placental defects.

A healthy placenta is vital to sustain normal pregnancy, ensuring proper supply of nutrients and oxygen to the baby. Abnormalities in the placenta can therefore have serious repercussions on fetal development, even causing miscarriage. Despite this, remarkably little is known about the identity of genes essential for a normal, functioning placenta and even less about the extent to which placental abnormalities contribute to defects that can arise as the fetus develops.


We screened more than 100 mouse mutant lines in which affected embryos die before or immediately at birth. Almost 70% showed serious abnormalities in the placenta; in extreme cases this resulted in a placenta incapable of supporting embryo development beyond an early stage (Figure 1), in others, abnormalities in the developing embryo were accompanied by abnormalities in the placenta.

FIGURE 1 – mouse mid-gestation embryos and placentas shown at the same magnification

LEFT: a normal, wild-type (WT) genotype. RIGHT: Nubpl mutation (MUT) shows a growth-retarded and developmentally delayed embryo that will not survive until birth.

The placentas are stained for a marker of the exchange surface (MCT4, in green) across which nutrients are transported from the mother to the embryo. Note the complete absence of this cell type from the MUT placenta. Red staining is for a cell surface protein (CDH1) demarcating the cells underneath the MCT4-positive layer (arrows), which are greatly reduced in number in the MUT placenta.


Not only do these results identify a large number of genes essential for normal development of the placenta; in addition they show an intriguing link between placental defects and abnormalities affecting the brain , heart and vascular system of the embryo itself. The research, led by Dr Myriam Hemberger and her colleagues at the Babraham Institute demonstrates how common placental abnormalities are when embryos develop abnormally.


The team examined in detail three different genes that cause embryonic lethality, and showed that for two of them the loss of the gene affected proper differentiation of placental cell types. For one of these genes they were also able to show that embryo death was a direct result of gene loss in the placenta, by providing the mutant embryo with a genetically normal placenta, which prevented embryo death.

Although the DMDD study uses mice, the results are likely to be just as relevant for studying human pregnancy and the role the placenta may play in pregnancy complications and the origins of birth defects in newborn babies.


The Advance Online Publication on Nature, ‘Placentation defects are highly prevalent in embryonic lethal mouse mutants is available now .

All image and phenotype data gathered by the DMDD programme is freely available to the scientific community at The research described in this blog post was funded by the Wellcome Trust with support from the Francis Crick Institute.


Every developing heart is subtly different. Hearts and their blood vessels don’t always grow at the same rate, to exactly the same size or precise shape, and this can complicate things if we want to identify an abnormality. To be sure if any feature of a heart is abnormal, first we need to understand the range of differences that we might see in normal hearts as they grow.

Surprisingly, this is a much-understudied area – something that it has only recently been possible to determine using modern imaging techniques. A new article published in Journal of Anatomy [1] uses high resolution 3D imaging to study more than 200 genetically normal mouse embryos from the DMDD programme, identifying the typical range and occurrence of different variations in the heart’s development. The image below shows the hearts of two genetically normal mouse embryos that were determined to be at exactly the same stage of development. However, one feature of their hearts that is very different is the extent to which the ventricular septum has grown to separate what was initially a single cavity into the right and left ventricles. The heart on the left has only a very small gap left in the developing septum, while the heart on the right has a much larger gap. Without this sort of study we wouldn’t be able to tell whether the heart on the right is normal or whether it has a ventricular septal defect – the most common congenital heart defect in newborns.


Click to view larger image.
The hearts of two genetically normal mouse embryos at precisely the same developmental stage show significant differences in the development of the ventricular septum.


The new data will be a valuable reference when identifying phenotypes in the heart and vessels of mouse embryos around the 15th day of gestation.

[1] Morphology, topology and dimensions of the heart and arteries of genetically normal and mutant mouse embryos at stages S21-S23, S. H. Geyer et al., J. Anat (2017), doi: 10.1111/joa.12663




A large-scale study of DMDD data [1] has shown that inactivating the same gene in mouse embryos that are virtually genetically identical can result in a wide range and severity of physical abnormalities. This suggests that the relationship between gene mutation and embryo development is more complex than previously thought.

The article ‘Highly variable penetrance of abnormal phenotypes in embryonic lethal knockout mice‘ was published in Wellcome Open Research.


The study considered 220 mouse embryos, each with one of 42 different genes inactivated. These genes are part of a set known as ‘embryonic lethal’, because they are so crucial to development that an embryo missing any one of them can’t survive to birth. Studying these genes can help us understand how embryos develop, why some miscarry and why some mutations can lead to abnormalities.

Each embryo was scanned in minute detail, meaning that even the smallest differences in features could be identified – right down to the level of whether the structures of individual nerves, muscles and small blood vessels were abnormal. It was also possibe to see whether having the same, single missing gene affected all embryos in exactly the same way.

Clinicians commonly find that people with the same genetic disease can show different symptoms or be affected with differing severity. In part this is likely to be due to the fact that we all differ in our precise genetic makeup. However, the results of this study in mice shows that even when individuals have virtually identical genomes, the same mutation can lead to a variety of different outcomes amongst affected embryos.


Click to view larger image.
A comparison of two embryos that are both missing the embryonic lethal gene Coro1c. The embryo on the right has abnormal viscerocranium (facial skeleton) morphology, while the embryo on the left does not.


A total of 398 different abnormalities (known as abnormal phenotypes) were observed across the 220 embryos. Surprisingly, almost none of the phenotypes were found in every embryo with the same missing gene. A phenotype that does not occur for every embryo missing a particular gene is said to have ‘incomplete penetrance’ and this phenomenon was seen regardless of how profound the abnormality was. Incompete penetrance was observed for phenotypes ranging from severe heart malformations to relatively minor defects such as the abnormal positioning of nerves.

Dr Tim Mohun, who led the study at DMDD said:

“This is a striking result, coming as it does from such a large study in which embryos have been analysed in unprecedented detail. It shows us that even with an apparently simple and well-defined mutation, the precise outcome can be both complex and variable. We have a lot to learn about the roles of these lethal genes in embryonic development to understand why this happens.”

The result was put in context by Dr Andrew Chisholm, Head of Cellular and Developmental Sciences at the Wellcome Trust, who added that “the fundamental processes driving how we develop have been conserved through evolution, which makes studying animal models enormously helpful in increasing our understanding of why some babies develop birth defects. This study throws new light on what we thought was a fairly straightforward relationship between what’s coded in our genes and how we develop. Researchers need to appreciate this added layer of complexity, as well as endeavouring to unpick the intricate processes of genetic control at play.”

All image and phenotype data gathered by the DMDD programme is freely available to the scientific community at Dr David Adams, Group Leader at the Wellcome Trust Sanger Institute who contributed to the work, said “this study highlights the power of genetic analyses in mice and provides an unprecedented resource of data to inform clinical genetic studies in humans.”

The research described in this blog post was funded by the Wellcome Trust with support from the Francis Crick Institute.


[1] Wilson R, Geyer SH, Reissig L et al. Highly variable penetrance of abnormal phenotypes in embryonic lethal knockout mice, Wellcome Open Res 2016, 1:1 (doi: 10.12688/wellcomeopenres.9899.2)


This week sees the publication of a new technique to precisely find the developmental stage of a mouse embryo, simply by looking at its fingers. We talked to the lead author, Stefan Geyer, to find out more about the technique, and the impact it could have on the field.

Stefan Geyer

What’s the main question that you have been trying to answer during your research career?

I have been fascinated by developmental biology since I was a student, and during my diploma thesis I became really interested in imaging. I was using High Resolution Episcopic Microscopy (HREM) and it was such a powerful tool that we could do 3D analysis and measurements of tiny structures like blood vessels and arteries in mouse embryos. I’m especially interested in the pharyngeal arch arteries – transient structures that undergo complex re-modelling during development. Abnormal re-modelling of these arteries accounts for a range of congenital cardiovascular abnormalities, so it’s really important that they develop as normal.

I’m also trying to expand the potential applications of HREM imaging. It was originally developed to study mouse, chick and zebrafish embryos but I believe that it has incredibe potential to be used with adult tissue samples, or even to study non-biological materials.


You published a paper today in Journal of Anatomy – it describes a new way to very precisely find the developmental stage of a mouse embryo. Why is this such an important result?

In the DMDD programme we screen a lot of mouse embryos – both those that have had a single gene knocked out and wild-type controls (with no gene knocked out). We study their morphology in great detail and look for developmental abnormalities, as a way to gain understanding about the function of the different genes that have been knocked out. Very early on in the project we realised that even though we were looking at embryos all harvested at E14.5 (14.5 days gestation) there was a dramatic variation in their appearance. And crucially this applied to the wild-type controls, as well as the knockouts. Some looked as you might expect for an embryo harvested at E13.5, others looked older. Even embryos from the same litter didn’t all look the same.

Obviously this is a big problem if you want to detect malformations in mouse embryos. If the embryos that are supposed to be ‘normal’ are so diverse in appearance, how can you begin to tell what is ‘abnormal’? Our new staging technique splits ‘normal’ at E14.5 into 6 developmental stages. Once we have identified the precise developmental stage of a knockout embryo using these new descriptions, we can compare it to what is normal for that stage and truly know whether it has developed as it should.

The appearance of the forelimb at stages S21 and S23.
The forelimb has the appearance of a paddle at stage S21 (left) but by stage S23 (right) it has developed into a hand with separate fingers.


Can you describe your technique in a few sentences?

It’s actually quite simple. We defined the 6 new developmental stages based on the amount of webbing between the embryo’s fingers. Between E14 and E15 an embryo’s forelimbs develop from paddles to hands with individual fingers. We defined the stages S21, S22-, S22, S22+, S23- and S23 according to how much of this webbing remained.

I think the simplicity is one of the main advantages of this technique. We use it on 3D models produced from HREM data but it could also be used on data derived from a variety of other imaging techniques.


In the past, people have staged mouse embryos using the Theiler stages (a set of 26 stages covering the whole of embryonic development). Why do we need more detailed staging than this?

Because there are structures and organs that change rather rapidly in embryos around E14.5, even within a single Theiler stage. So on some occasions the Theiler stages don’t give enough granularity. The most striking example that we found was the appearance of the palatal shelves, which completely reorient themselves during this period of development. We found that the changes happen in a predictable order, but extremely quickly. In some embryos harvested at E14.5 the shelves are positioned laterally to the tongue, in others they have already elevated and are positioned above the tongue. In a few embryos, elevation is in progress and the shelves are arranged asymmetrically, one laterally to the tongue, the other above it. If we want to diagnose cleft palate with confidence in E14.5 embryos, exact staging is essential.

We also think that our technique simplifies the staging process, as it’s based only on one parameter – the amount of webbing between the fingers. Theiler staging is based on the appearance of several features, so sometimes you can get conflicting results.


Could your new technique also impact other areas of developmental biology?

It would be useful for any study that compares mutant and control embryos around E14.5. This isn’t limited to gene knockouts, it’s relevant to any mutation that is being studied to look for resulting abnormalities in the embryo. Relying on simple comparisons between mutant and control embryos harvested at the same time point might falsely diagnose a malformation. For example an embryo at stage S22- still shows an interventricular foramen in the heart (a temporary gap between the developing ventricles), which is normal for his developmental stage. In a embryo at stage S23 the interventricular foramen has disappeared. Therefore comparing a mutant at stage S22- to a control embryo at stage S23 is problematic. Only precise staging allows us to identify true malformations.


What is next on the horizon for you?

We are now able to tell the stage of an embryo very precisely. But we don’t know exactly how every organ develops during each of our new substages. To really understand how an embryo’s organs develop I want to understand at what a ‘normal’ mouse looks like and exactly what structures are changing between stages 21 and 23.  Our research group will do statistical analyses of different structures, to find what a normal range of variation looks like. We also plan to look at variations in different strains of mouse. I believe that phenotype screens would really benefit from such a comprehensive study of a ‘normal’ control mouse.


The original article described in this post is SH Geyer et al., ‘A staging system for correct phenotype interpretation of mouse embryos harvested on embryonic day 14 (E14.5)‘, J. Anat., 10.1111/joa.12590.

Stefan Geyer is Assistant Professor in the Division of Anatomy, Center for Anatomy and Cell Biology at the Medical University of Vienna.


A new paper published in Journal of Anatomy shows that measuring the amount of inter-digital webbing in mouse embryos between 14 and 15 days gestation is the best way to find out their exact stage of development.

So why is this important to a geneticist? If we want to discover a causal link between a gene mutation and a developmental abnormality in a mouse, we need to be really sure that any abnormalities we see aren’t just standard features of embryo development. The morphology of an embryo can change rapidly as it grows, and things that seem unusual at first glance might be completely typical for a given stage of development. Understanding and avoiding these potential pitfalls is hugely important when studying embryos for the effects of a gene mutation.

Using the new technique described by the authors, it’s now possible to find an embryo’s exact developmental stage simply by looking at its fingers.


To be sure that an embryo has developed abnormally, we need to compare it to one that we know has developed normally. But this can be a complicated comparison to make when an embryo’s ‘normal’ appearance is changing dramatically as it develops.

Cleft palate is a serious abnormality, as it can prevent a mouse pup from suckling and eventually lead to death. But the palate is also one of many tissues that changes its structure and orientation dramatically between E14 and E15 (days 14 and 15 of gestation). The image below shows three snapshots of a typical mouse embryo palate during this period.


During days 14 – 15 of embryo development, the palatine plates change their orientation rapidly. Initially, both plates are lateral to the tongue (left). They begin to elevate, but do so asymmetrically (centre), ending with both shelves above the tongue (right) at which point the plates fuse together.


Not only do the palatine plates completely change their orientation, they also move one at a time. Without knowing the precise developmental stage of an embryo, it would be impossible to tell whether the palate was developing normally, or whether a cleft palate had developed as a resust of a gene mutation.

Dramatic changes within a similarly short timeframe can also be seen in both the heart and the intestine. It’s vital to measure the developmental stage of the embryo as precisely as possible, so that it can be compared to what is normal at that exact stage.


In the past, the development of a mouse embryo has been described using a set of 26 Theiler stages. Days 14 – 15 of gestation are typically described by the stages 21, 22 and 23.

However, the authors found that for embryos of this age many organs and tissues developed rapidly even within a single Theiler stage. The stages weren’t granular enough to be sure that tissues like the palate, heart and intestines were really showing abnormal development. To address this the article defines 6 alternative stages, together with a novel technique to identify the stage solely by measuring the amount of webbing between the embryo’s fingers.


An image of each of the six newly defined stages, and the appearance of the forelimb at each stage.
Six individual stages were defined, each of which can be identified by measuring the amount of inter-digital webbing.


During this time, the forelimb develops from a paddle to a hand with separate fingers. The full set of new stages S21, S22-, S22, S22+, S23- and S23 allow the developmental stage of the embryo to be determined much more precisely, meaning that developmental abnormalities can be identified with confidence.

This blog post is based on the original article SH Geyer et al., ‘A staging system for correct phenotype interpretation of mouse embryos harvested on embryonic day 14 (E14.5)‘, J. Anat., doi:10.1111/joa.12590.


Today sees the launch of Wellcome Open Research, a new publication platform for Wellcome Trust funded researchers. A set of articles released to coincide with the launch includes ‘Highly variable penetrance of abnormal phenotypes in embryonic lethal knockout mice‘, a publication by the DMDD Programme. It explores the results of systematic efforts by the consortium to image and phenotype embryos from embryonic-lethal knockout mouse lines.


Papers submitted to WOR are published immediately as preprints, following a series of objective checks. They are then subject to a process of open peer review. WOR requires all supporting data to be made available, enabling other researchers to analyse and replicate published studies. By using services developed by the publisher F1000Research, research outputs can be made available faster and in ways that support reproducibility and transparency.

DMDD’s Tim Mohun on the reasons the team chose to publish in WOR:

“For us, the timing of the launch was perfect. We had reached a bit of a milestone in the DMDD project, having analysed enough different embryonic lethal genes that we could begin to consider the dataset as a whole and draw some initial conclusions. We wanted to share our findings with the research community as quickly as possible because we think the conclusions are interesting, important and, in part, puzzling.

Lots of research never sees the light of day. We only ever see the choice results and final conclusions in conventional scientific publications, but rarely all the accumulated data on which those studies rest. For screening studies like ours which are necessarily limited in depth but wide in scope, publications like Wellcome Open Research give the opportunity to make our data available for other scientists to use, hopefully helping to advance their work and avoiding duplication of effort”.

The DMDD paper ‘Highly variable penetrance of abnormal phenotypes in embryonic lethal knockout mice‘ is currently undergoing peer review and this process can be followed online.

Read a longer interview with Tim Mohun and Jim Smith on the WOR blog.


Around a third of mammalian genes are essential for life, and the recent Nature paper from the IMPC  ‘High-throughput discovery of novel developmental phenotypes‘ [1] describes some achievements from sytematic study of these genes in knockout mice.

Screens like those of the IMPC and DMDD are vital to understand gene function on a genome-wide scale and, based on the results recently published in Nature, here are some reasons why.


Lethal genes in the mouse are known to be enriched for human disease genes [2,3]. When additional data from the IMPC was included on the genes essential for survival of the embryo, this enrichment was increased even further. More than half of the human disease genes considered were essential for mouse embryo survival. The study also found a remarkable correlation between the core essential genes in humans and mice.

Systematic knockout mouse screens provide data that could not be derived from human patients. These new results further underline the importance of mouse models in the study of human disease, and their relevance in a clinical setting.


A suprising observation from knockout mouse screens is the incomplete penetrance of phenotypes for many lines.

One example of this is the sub-viability of lines. The IMPC has found that in around 11% of knockout lines some homozygous pups were observed, but fewer than the 1 in 4 pups predicted by Mendelian genetics. Some pups were able to survive with the homozygous gene knockout, but some weren’t.

Incomplete penetrance is a result also echoed in DMDD data. For example, in the seven Adamts3 knockout embryos studied, all display subcutaneous edema and absent lymph sac, while only two display a bifid ureter.

Click to view larger image.
Subcutaneous edema and bifid ureter (left side) observed in an Adamts3 mutant embryo. The red arrows highlight a single ureter on the right side, but two branches on the left side.

Data from systematic screens of knockout mice is showing, on an unprecedented scale, that even for a complete gene knockout, the observed phenotypes can vary from embryo to embryo. Given the standardised background and allele construction, this is a suprising result and could suggest an underlying stochastic process.


As part of its systematic screen, the IMPC has identified 22 essential mouse genes with human orthologs that are not known to be associated with any human disease. These are potential candidates for undiagnosed diseases and could shine new light on the causes of genetic disorders.

Efforts are continuing to study knockouts of every gene in the mouse genome. As these datasets grow in size, so too does the potential for them to help us understand gene expression and the genetic basis of human disease.

The DMDD database of embryonic-lethal mouse knockouts can be found at

The IMPC database of knockout mice can be found at


[1] The IMPC Collaboration (2016)
High-throughput discovery of novel developmental phenotypes
Nature  doi:10.1038/nature19356

[2] B. Georgi1, B. F. Voight1, M. Bućan1 (2013)
From mouse to human: evolutionary genomics analysis of human orthologs of essential genes
PLoS Genet 9(5): e1003484. doi: 10.1371/journal.pgen.1003484

1 Department of Genetics, Perelman School of Medicine, University of Pennsylvania, USA

[3] J. E. Dickerson 1, A. Zhu1, D. L. Robertson1 K. E. Hentges1 (2011)
Defining the role of essential genes in human disease
PLoS ONE, 6(11), e27368.

1 Faculty of Life Sciences, University of Manchester, UK


For the first time, a reference mouse embryo atlas has been created using HREM image data. For other embryo imaging methods such as micro-CT, a reference embryo atlas has previously been shown to be the basis of automated phenotyping (Wong et al, 2014). This new work (a collaboration between the DMDD programme and the Mouse Imaging Centre in Toronto) is a proof of principle for automated phenotyping of HREM data, which can now be tested in more detail.


In order to phenotype an embryo, we need to compare it to another embryo that we consider to be normal. With histology this can be very difficult, as it’s virtually impossible to take exactly the same slice through two embryos.

So this leads to the question ‘have I really observed a phenotype, or am I simply looking at a different part of the embryo?’

The ideal scenario for a phenotyper is to have exactly equivalent slices of anatomy to compare. They can then do a direct visual comparison, or an algorithm can be used to do a statistical comparison, therefore automating the process to some degree.


Using a technique previously employed with micro-CT data (Wong et al, 2012), the team have created a reference embryo atlas from HREM data. The atlas can be used to find these equivalent sections of anatomy.

21 wild-type embryo image stacks were merged together to give an unbiassed ‘average embryo’. Every possible pair of embryos was compared in terms of displacement. The embryos were then combined together so that the displacement for all embryos to join the atlas was a minimum. This technique avoids giving undue weighting to any one embryo in the average.

As HREM data is so large, the resolution was scaled back first to make the computation feasible.

For each individual embryo, a different deformation field is needed to bring the image into alignment with the average. The inverse of this field can then be used to propagate back from any plane in the reference embryo to a section of the original input embryo.

This means it’s possible to find equivalent embryo sections to compare, ensuring that any differences observed are much more likely to be genuine phenotypes.

Equivalent embryo sections of HREM data.
Bottom left, the merged reference embryo where the plane of interest is defined. Bottom right, the equivalent data from four individual mouse embryos. Top, the crumpled sections show the effective ‘cut’ plane through the original HREM data that is homologous to the plane of interest in the reference embryo. (Image recreated with the permission of Elsevier).

It’s important to note that these sections are never planes but crinkly slices through the embryo. This means that the sections couldn’t have been identified by manually scrolling through the image stacks.


Automated selection of equivalent embryo sections is the starting point for automated phenotyping. Given the time-consuming nature of phenotyping embryos, and the large task of phenotyping many mouse lines, automated phenotyping is an attractive prospect to many. It can act as a guide – a primary screen to highlight areas of interest to manual phenotypers.

The next step, then, is to compare other HREM embryo data with this reference atlas.

The results of automated primary screens are typically presented as a heatmap (Wong et al, 2014), highlighting areas of interest. The authors showed a proof of principle for an automated primary screen of HREM data by separating the embryos into male and female groups. After selecting equivalent sections through the gonads to compare between the two groups, they were able to automatically distinguish statistically significant differences in this region.


Click to see full-size image.
The coloured regions show statistically significant differences between males and females in the region of the gonads.

This is an exciting result for the future of automated phenotyping of HREM mouse embryo data, which will be tested much more extensively over the coming months.


R. Mark Henkelman1, Miriam Friedel1, Jason P. Lerch1, Robert Wilson2, Tim Mohun2 (2016)
Comparing homologous microscopic sections from multiple embryos using HREM
Developmental Biology [Epub ahead of print], DOI: 10.1016/j.ydbio.2016.05.011

1 Mouse Imaging Centre (MICe), Hospital for Sick Children, University of Toronto, Toronto, Canada
2 The Francis Crick Institute Mill Hill Laboratory, London, UK


M. D. Wong1, Y. Maezawa2, J. P. Lerch1, R. M. Henkelman1 (2014)
Automated pipeline for anatomical phenotyping of mouse embryos using micro-CT
Development, DOI: 10.1242/dev.107722

1 Mouse Imaging Centre (MICe), Hospital for Sick Children, University of Toronto, Toronto, Canada
2 The Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, Toronto, Canada


M. D. Wong1,2, A. E. Dorr1,3, J. R. Walls4, J. P. Lerch1,2, R. M. Henkelman1,2 (2012)
A novel 3D mouse embryo atlas based on micro-CT
Development, DOI: 10.1242/dev.082016

1 Department of Medical Biophysics, University of Toronto, Toronto, Canada
2 Mouse Imaging Centre (MICe), Hospital for Sick Children, University of Toronto, Toronto, Canada
3 Sunnybrook Hospital, Toronto, Canada
4 Regeneron Pharmaceutecals, Tarrytown, USA