How cats get their stripes

[lpetrich]

How the Cat Gets Its Stripes: It’s Genetics, Not a Folk Tale - The New York Times

Although we have succeeded in solving some big biological mysteries, like metabolism and genetics, there is a big one that continues to elude our grasp: genes to shapes and patterns.

That has been the subject of a lot origin-story folklore, usually involving Lamarckian inheritance from something that some ancestor experienced.

But while genetics and metabolism have become very well-understood, development is still largely a mystery. We don't have much of a clue of how to get from genes to shapes, especially macroscopic shapes. The same is true of patterning, especially macroscopic patterning. It is generally agreed that development involve various morphogens in various ways, though they have been difficult to track down.

In 1952, Alan Turing proposed that patterns can be produced by a combination of diffusion of morphogens and chemical reactions between morphogens.  The Chemical Basis of Morphogenesis -  Turing pattern -  Reaction–diffusion system -  French flag model - 13.6: Reaction-Diffusion Systems - Mathematics LibreTexts

Some morphogens have indeed been discovered, like those in the Hox mechanism ( Hox gene) Hox genes are expressed in zones along the length of the animal, and the resulting proteins help specify identity along the nose-tail axis. There is a similar system that specifies identity along the dorsoventral axis ( Inversion (evolutionary biology)) and in plants, there is the  ABC model of flower development

Turning to patterns, this discovery was made back in 1977:
(PDF) A unity underlying the different zebra patterns

To elucidate the relationship between the complex striping patterns of the different species of zebras, a simple conceptual experiment has been performed. Using data from horse embryos, the normal growth of the zebra from early foetus to adult has been reversed to see what happens both to the spacing and to the orientation of the stripes. It turns out that for each species, there is a point in time when all the body stripes would have been perpendicular to the dorsal line and equally spaced. Moreover the spacing is roughly the same (0·4 mm) for the three main species of zebra at this time. This point is during the third week of development for E. burchelli, fourth week for E. zebra and fifth week for E. grevyi. As striping only appears at about the eighth month of foetal development, it seems that the pattern is determined a long time before the cells actually lay down pigment. Further analysis of the pattern so laid down on a rapidly-growing foetus shows how shadow and gridiron stripes can arise. The reason why leg stripes are orthogonal to body stripes cannot however be derived from this phenomenological approach. These results suggest that a single mechanism generating equi-spaced stripes of separation 0·4 mm could lay down the body stripes of zebras and that species differences arise from pattern formation occurring at different times in embryogenesis.

Just recently, a followup has been published on domestic cats.

 

[lpetrich]

Alan Turing: The Chemical Basis of Morphogenesis

How the Cat Gets Its Stripes: It’s Genetics, Not a Folk Tale - The New York Times
The article shows a picture of an American-shorthair cat with a mackerel-tabby pattern, gray and black stripes. This is the wild-type pattern, the pattern that domestic cats' wild ancestors have.

(Turing reaction-diffusion...)

... Dr. Barsh said the team’s research had confirmed this hypothesis.

Further, he said, the study shows for the first time that the gene Dkk4 and the protein it produces are central to the process. Dkk4 is the inhibitor in the process.

Then how the researchers got the cat embryos: from the wombs of spayed stray cats. Some of them were pregnant, and the embryos went out with the wombs.

From more than 200 prenatal litters, Dr. McGowan looked for patterns in the tissue at the different stages of growth in the embryos. She found a pattern of what she described as thick and thin areas of tissue in the top layer of the embryonic skin, never before reported. The regions, she said, “mimic what’s going on in the adult cat pigmentation patterns.” The same patterns that will appear in an adult cat’s coat as stripes or blotches appear first in the embryo before there is any hair or even hair follicles.

This is much like what happens in zebras -- the patterns are laid down long before the fur coloring emerges.

There were different amounts of Dkk4 in the thick and the thin tissue areas. The Dkk4 protein was inhibiting the genes that produce other signaling molecules known as Wnt proteins, Dr. Barsh said. Even more telling, when there was a mutation in the Dkk4 gene, the stripes became thinner, to the point that a plain pattern called Ticked emerged.

Like in zebras, where different species lay down their stripe patterns at different times.

The authors emphasize that the patterns they investigated are only a “fraction of the pattern diversity that exists among domestic cat breeds.”

In the future, Dr. Barsh said, one target for the team will be to uncover how the tissue pattern translates to color when hair follicles grow.

Dr. Hoekstra said the work highlighted the value of domestic animals to science. “Cats are a fantastic model — easier to study than zebras or leopards — that have developed a dazzling array of spots, stripes and everything in between.”

Developmental genetics of color pattern establishment in cats | Nature Communications

STRIPED, SPOTTED AND TICKED CATS - ticked cats are partially striped

Some earlier work on cat-coat genetics: Defining and Mapping Mammalian Coat Pattern Genes: Multiple Genomic Regions Implicated in Domestic Cat Stripes and Spots | Genetics

 

[steve_bank]

How about natural selection?

 

[lpetrich]

How about natural selection?

How does that figure in this? Natural selection can only work on what variations are available, and an important problem in biology is finding out what variations are available.

Returning to cats, Manx cats have stub tails or no tails at all. There is a bit of folklore that says that they are from an ancestor getting its tail stuck in Noah's Ark - a typical Lamarckian sort of explanation. But why do other cats continue to grow tails? A clue may be in the Manx mutation. If a cat has both copies of that mutated gene, it dies in the womb as an embryo. But if it has only one copy along with a copy of the original version, it can grow to adulthood and live to reproduce. This sort of thing can get in the way of the continued propagation of that gene, unless another mutation happens that makes it less lethal in the womb.


Horses, donkeys, and zebras diverged over the last few million years, and they are all placed within one genus:  Equus (genus) Their phylogeny is:

• Subgenus Equus, species Equus (Equus) ferus (wild horse, ancestor of domestic horse)
• -
  • Subgenus Asinus, species:
    • Equus (Asinus) africanus (African wild donkey, ancestor of domestic donkey)
    • Equus (Asinus) hemionus (Onager / hemione / Asian wild donkey)
    • Equus (Asinus) kiang (kiang)
  • Subgenus Hippotigris ("tiger horse"), species:
    • Equus (Hippotigris) quagga (burchelli) (Burchell's zebra / plains zebra, quagga) - stripes wk 3, coarse
    • Equus (Hippotigris) zebra (mountain zebra) - stripes wk 4, coarse on rear end, fine elsewhere
    • Equus (Hippotigris) grevyi (Grevy's zebra) - stripes wk 5, fine

Subgenera Equus and Asinus are solid-colored, though often with countershading (light-colored bellies). They often have  Primitive markings including stripes. This may mean that the ancestor of Equus was striped with some of its descendants losing stripes.

 

[lpetrich]

Domestic animals seem like good model systems for studying a lot of genetic variation, since they often have variations that are lacking in their wild ancestors, and because their recent ancestry is sometimes very well documented.

Consider dogs, which vary in size from the Chihuahua (2 kg) to the Great Dane (60 kg). Their wild ancestor, the gray wolf, weighs in at 50 kg.

 Dog coat genetics -  Cat coat genetics -  Equine coat color genetics

 Tabby cat - the color pattern

 

[steve_bank]

Natural selection, mutations and selection by the environment for survival.

I recently watched a PBS show about colors and animals.

The prey of a particular lion does not see orange well. The predator being a shade of orange has survival advantage.

Random mutations over long periods of time selected for survival. More lions with orange mutation are better fed, procreate more, and the gene becomes common.


There are bugs with bright colors that are poisonous and predators know to avoid. There are other bugs with similar coloring that are not poisonous but are avoided by predators. It is not s if the non poisonous bugs figured it out, it is just happenstance of evolution. No intent o conscious analysis and change.

If you accept evolution and natural section, then that can account for coloration. Not allmutaions and variations are based on selection.

 

[lpetrich]

That's all beside the point. One must ask what variations are available.

Consider that mammalian hair colors have a very limited color palette. Decomposing colors by hue, saturation, and lightness, they do vary over the complete range of saturation and lightness, but they vary very little in hue: orange to yellow. Human, domestic-dog, and domestic-cat hair colors cover the full range of possible hues.

Why not green hair or blue hair or purple hair? Birds have a greater range of colors in their feathers, sometimes on a single feather.

Human-eye irises vary more in color, going into blue, and some dogs and cats also have blue eyes.

Human skin is also rather limited in color, ranging from pinkish to yellowish to dark brown. Some monkeys, like mandrills have blue skin patches, I must note, and lizards, snakes, frogs, fish, insects, and spiders have a full range of colors.


This leads to an important question: what makes hair so limited? A larger color palette would be valuable for camouflage, and also for advertisement, like recognizing fellow species members, impressing rivals, impressing mates, and showing off that one is dangerous. A bird that tries to eat a monarch butterfly will refuse to eat any other ones.

 

[lpetrich]

Why do cats—and so many other animals—look like they’re wearing socks?

Pet lovers refer to this particular color pattern as an animal’s “socks,” “booties,” “mittens,” or “tuxedo” for obvious reasons. The phenomenon of pigment mixed with white splotches can occur in pigs, deer, horses, dogs, guinea pigs, birds, and, in rare cases, humans. But it’s particularly prominent in cats, as evidenced by the fact that Socks consistently ranks in the top names for felines. (Even former President Bill Clinton bestowed it upon his black-and-White House pet, who notoriously did not get along with the family’s monochromatic chocolate lab, Buddy.)

But scientists have another name for it: piebaldism. It’s the result of a mutation in the KIT gene, which causes an unusual distribution of melanocytes—the cells that give eyes, skin, and hair or fur pigment.

When a cat is still an embryo, all of its available melanocytes are bunched up toward its back, where its spinal column will eventually form. As the fetus develops into a mewling kitten, pigment cells spread throughout the developing body. If the melanocytes are evenly distributed, the cat could have a unicolor coat, like Sabrina the Teenage Witch‘s all-black cat, Salem, or the all-white Hello Kitty. But in many animals, the cells spread irregularly. That’s how you get a cat like Sylvester, who’s black from his back to his legs, but white down to his toes.

 Neural crest
The spinal cord forms from infolding skin on the back, making a tube, and the edges of that tube are the neural crest. Its cells then travel to different places in the embryo and becoming different things.

Neural crest cells are a temporary group of cells unique to vertebrates that arise from the embryonic ectoderm germ layer, and in turn give rise to a diverse cell lineage—including melanocytes, craniofacial cartilage and bone, smooth muscle, peripheral and enteric neurons and glia.

Reconciling diverse mammalian pigmentation patterns with a fundamental mathematical model | Nature Communications
Conclusion: those white spots are caused by melanocytes not dividing enough, thus leaving some parts of the skin unpigmented.

Piebaldism isn’t the only genetic quirk that can alter an animal’s fleece, according to the UC Davis Veterinary Genetics Lab. The tabby cat’s signature look is served up by the agouti gene, which determines the distribution of black pigment. The same gene gives rise to “bay” horses, which have ruddy brown bodies, but pitch black manes and tails. Norwegian Forest cats harbor two mutations of note: The aptly-named “Orange gene” on the X chromosome can produce a red coat in many cats, but an alteration on the MC1R gene appears specific to this breed. Born one color, these felines can mature into another golden or “amber” hue. And Siamese and Burmese cats have a form of selective albinism that allows them to suppress melanin production based on temperature. The activating enzyme tyrosinase explains the Siamese’s ombre appearance, with its sandy-colored abdomen (the warmest part of the body) that darkens around the extremities, including its ear tips and paws.

Why Cats' Color Points Are More Fascinating Than You'd Think
and
Tyrosinase mutations associated with Siamese and Burmese patterns in the domestic cat (Felis catus) - PubMed - these mutations make this enzyme not function well in relatively high temperatures, thus keeping melanin from being formed very much in warmer parts of the skin. That's what makes the Siamese-cat pattern -- dark face, ears, feet, and tail, and light elsewhere. Some domestic rabbits also have this pattern, likely due to the same mechanism.

 

[lpetrich]

I think that the greatest success in solving the genes-to-shapes problem has been in the specification of identity along the body axes.

The primary one is found in nearly all of the animal kingdom, and it is variously called the head-tail, nose-tail, anterior-posterior, rostral-caudal, cranial-caudal, and oral-aboral axis.

The secondary one is found in bilaterian animals, and it is variously called the back-belly and dorsal-ventral axis. It is orthogonal to the primary one.

The tertiary one is found in at least some bilaterian animals, and it is the left-right axis. It is orthogonal to the first two, and it is relatively weak.


The primary body axis, the head-tail one, is associated with Hox genes, and these genes are expressed in bands or stripes along that axis. The resulting proteins in turn control the expression of other genes. What's upstream of the Hox genes I couldn't find out very quickly. But fortunately, it's rather easy to find out where they are expressed and how they are related. Their expression zones typically overlap, and those regions also typically overlap several segments of segmented animals: arthropods, annelids, and vertebrates (vertebrate segmentation is on the inside, without the outside visibility of arthropod and annelid segmentation).

Hox genes are typically clustered on their genomes, arranged in head-to-tail order in their clusters, with their transcription being from tail-to-head in the cluster. But they have a complicated history of duplication and loss and in some cases cluster splitting and reversal of transcription direction to head-to-tail, so reconstructing Hox genes' ancestral state is not a trivial task. But it is evident that the ancestral bilaterian animal, a little worm that lived in the late Proterozoic, had one cluster with something like 8 Hox genes in it.

Hox clusters are sometimes duplicated, often as a result of genome duplication: polyploidy. There is evidence of several genome-duplication events in the history of our planet's biota:  Paleopolyploidy It often happened in plants, not so much in animals. But it happened twice in early vertebrates - the  2R hypothesis - and it happened again in the ancestors of teleost bony fish, for instance. After a polyploid event, members of duplicate sets of genes are often lost, sometimes enough to obscure evidence of that event.

Outside of Bilateria, cnidarians have Hox-gene clusters with three or four genes, and they are also expressed along the animals' primary body axis, the oral-aboral (mouth to opposite end) axis.


I've found several online illustrations of Hox-gene organization and expression zones, even if I haven't found one that I find very satisfactory. A lot of them are embedded in research papers that are often very arcane.

 

[lpetrich]

Before I continue, I wish to note a feature of animal embryonic development. When egg cells divide, they become a solid ball of cells, and then a hollow ball of cells or else a disk on a big ball of yolk ("germinal disc" or "blastodisk"). The ball itself can then form the animal, or else a disk on the ball does so, in the fashion of a disk on yolk. Disk-on-yolk development also typically involves making a yolk sac that grows around the yolk.

Reptiles, birds, and egg-laying mammals all use the disk-on-yolk-ball approach, while live-bearing mammals use a disk-from-hollow-ball approach. So all amniotes go through an embryonic-disk phase, with live-bearing mammals carrying it over rather than reverting to developing from the whole egg cell.

Fish also use the disk-on-yolk approach, and baby fish often hatch with their yolk sacs sticking out of their bellies.


That aside, the early embryo makes two or three "germ layers", thus being "diploblastic" or "triploblastic". Bilaterians have three, while other animals have two. From outside to inside, three-layer animals' germ layers are:

• Ectoderm: skin, nervous system with the brain and sense organs
• Mesoderm: blood, blood vessels with the heart, muscles, gonads, and vertebrate cartilage and bone
• Endoderm: gut (digestive system) and outpouchings of it like the vertebrate liver and land-vertebrate lungs

Two-layer animals have a gelatinous layer called "mesoglea" instead of a mesoderm.

Back to Hox genes, the Hox mechanism operates in the ectoderm. There is a similar mechanism in the endoderm, with the ParaHox genes, related to Hox ones, and another one in the mesoderm, with the NK genes, also related to Hox ones.

Hox genes are named from their "homeoboxes", the active part of Hox proteins. Fungi, plants, and a slime mold have proteins with homeoboxes in them, though they are not involved in Hox-style patterning. Homeoboxes bind to DNA sequences, as one would expect of proteins involved in gene regulation.

Homeoboxes got their name from genes with them having "homeotic" mutations, mutations that make body parts develop like other body parts. For instance, antennapedia in fruit flies is antennae developing like legs, proboscipedia is mouthparts developing like legs, and ultrabithorax is growing four wings instead of two wings and two "halteres", two small balls on stalks that flies and mosquitoes grow instead of hind wings.

Anatomically, arthropod antennae and mouthparts are all modified limbs, just like bird, bat, and pterosaur wings are all modified front limbs, and mutations like antennapedia and proboscipedia fit in very well.

 

[lpetrich]

I note that most invertebrates develop in whole-egg fashion, and that many aquatic invertebrates have the twist of a larval phase very different from the adult phase. Larvae known by such names as nauplius, zoea, trochophore, veliger, bipinnaria, brachiolaria, pluteus, and tornaria.

The nauplius and zoea are crustacean larvae, they have a small number of body segments, and they use their limbs for swimming. As they grow, they add segments on their rear ends, and they turn their frontmost limbs into antennae and mouthparts.

The others have much less resemblance. Trochophores grow up to be annelids and some mollusks, veligers to be bivalves and sea snails, bipinnarias grow into brachiolarias then and grow up to be starfish, pluteuses grow up to be sea urchins, and tornarias grow up to be acorn worms.

Larval stages have developed elsewhere, notably in insects and amphibians.

Four-stage insects go through egg, larva, pupa, and adult. The larva is a wormlike phase that is from continuing a late embryonic state of development. The pupa is a sort of second egg, and in that form, the insect then catches up on growing up.

Amphibians are tetrapods that have partially made it onto land, while spending the earlier part of their lives in water. Frogs are well-known for that, with tadpoles being aquatic and not looking much like adult frogs. But some species of frog skip the tadpole phase and hatch into miniature adult frogs. Much like how amniotes originated from early amphibians by hatching as land animals. Their embryos still have gill bars, an aquatic-vertebrate feature, though they never have functional gills. Their gill bars develop part of the way before being reused in various ways.

I also note that vertebrate jaws developed from frontmost gill bars.

 

[lpetrich]

As to segmentation, it is rather widely dispersed, in annelids, arthropods, and vertebrates. That raises the question of its origin. Was it invented once then lost several times? Or twice then also lost several times? Or three times, one in each group. It may be possible to resolve that issue by finding out the mechanisms involved in segment layout and growth. Are they too different to have the same origin? I haven't seen much that might resolve this issue.

Losing segmentation several times may not be as farfetched as it might at first seem. From molecular phylogeny, two groups of segmentless marine worms have been shown to descend from annelids: echiurans (spoon worms) and sipunculids (acorn worms). So it may happen elsewhere, and with improved resolution of animal phylogeny, we can get a better handle on how many losses are necessary for an early origin.


With some exceptions, segmentation grows in all three groups in the same way, but adding segments to the rear end.

The exceptions are some insects, like fruit flies. "Long germ band" insects like fruit flies lay out all their segments all at once as they grow. Most other insects lay out only the forward segments if they lay out any at once, being "short germ band" and "intermediate germ band" insects. The remaining segments are added on their rear ends.

This quirk of development illustrates a risk of model systems: they may contain very atypical features.

 

[lpetrich]

I'll move on from the primary to the secondary body axis, the back-belly or dorsoventral one. That one has a very interesting story to tell.  Inversion (evolutionary biology)

In 1822, biologist Étienne Geoffroy Saint-Hilaire dissected a crayfish, and he found that its internal anatomy was upside down by vertebrate standards.

• Arthropods: ventral, vertebrates: dorsal
• Central nervous system
• Gut
• Heart
• Arthropods: dorsal, vertebrates: ventral

What he found for that crayfish turned out to be true of not only arthropods, but most other invertebrates. Does that mean that some protovertebrate flipped over?

This hypothesis ran into problems very quickly. A big one was the architecture of the central nervous system. Vertebrates have a tube that is formed by infolding skin, while many invertebrates have a head-to-tail string of ganglia. At the head end, in arthropods and annelids, the CNS loops around the gut to reach the brain, which is on the dorsal side. Mollusks and various other invertebrates have a variation: the brain surrounds the gut.

So for a long time it was not accepted, though it was revived every now and then for over a century and a half.

But in recent decades, the discovery of growth factors / morphogens related to dorsoventral patterning has provided strong evidence for that hypothesis, strong enough to make it generally accepted. Ones expressed ventrally in arthropods and annelids are expressed dorsally in vertebrates, and vice versa.

"There is also evidence from left-right asymmetry. Vertebrates have a highly conserved Nodal signaling pathway that acts on the left side of the body, determining left-right asymmetries of internal organs. Sea urchins have the same signaling pathway, but it acts on the right side of the body."

Evidence from the remaining body axis, tertiary or left-right body axis.

 

[lpetrich]

How Humans Lost Their Tails - The New York Times
noting
The genetic basis of tail-loss evolution in humans and apes | bioRxiv

Looking over the entirety of Metazoa, most animals with a through gut have a mouth at one end of the body or near it, and an anus at the other end or near it. Lobster "tails" are really their abdomens, much like insect and arachnid abdomens.

Chordates are an exception. They have a tail that projects further rearward, a "post-anal tail".

This is sometimes well-developed, as in fish, and sometimes reduced. Some vertebrates have very short tails or none at all, like frogs and bears and us and our closest relatives, the apes. The Barbary ape is a macaque monkey that has lost its tail, and Manx cats are domestic cats with stumpy tails or no tails.

Let's look at phylogeny. Ancestors of simians branched off of prosimians about 50 million years ago, Old World and New World simians separated about 40 million years ago, and apes branched off of Old World simians some 20 - 30 million years ago. Great apes and lesser apes separated about 10 - 15 million years ago, and our ancestors separated from chimp ones about 5 - 10 million years ago. These dates are approximate from the patchiness of the fossil record.

 

[lpetrich]

Back to the research.

"This question — where’s my tail? — has been in my head since I was a kid,” said Bo Xia, a graduate student in stem cell biology at N.Y.U. Grossman School of Medicine.

A bad Uber ride in 2019, in which Mr. Xia injured his coccyx, brought it back to his mind with fresh urgency. “It took me a year to recover, and that really stimulated me to think about the tailbone,” he said.

...
Researchers have identified more than 30 genes involved in the development of tails in various species, from an iguana’s long whip to the stub on a Manx cat. All of these genes are active in other parts of the developing embryo as well. Scientists are still learning how their unique activity at the end of an embryo gives rise to a tail.

Mr. Xia reasoned that there must be some mutation that disabled the growing of a tail, some "Manx monkey" mutation.

In 1923, the Russian geneticist Nadezhda Dobrovolskaya-Zavadskaya X-rayed male mice and then allowed them to breed. She found that a few of them gained a mutation that caused some of their descendants to grow kinked or shortened tails. Subsequent experiments revealed that the mutation was on the TBXT gene.

So he looked for mutations in that gene, and he found an insert of some 300 genetic letters in the human and ape versions of TBXT and not in monkey ones, an insert in the same position in each species' gene.

So, he reasoned, some 20 million years ago, some monkey had a Manx mutation, a mutation that spread over some population of monkeys, a population that became the ancestors of apes -- and us.

Mr. Xia brought the finding to his supervisors, Itai Yanai and Jef Boeke, to see what they thought. “I nearly fell off my chair, because it is just a stunning result,” Dr. Yanai recalled.

To test the hypothesis that this TBXT mutation causes taillessness, they genetically engineered a mouse version and put it into some lab mice. The animals grew only short tails or no tails at all -- Manx mice.

The scientists said that the TBXT mutation is not the sole reason that we grow a coccyx instead of a tail. While the mice in their experiments produced a range of altered tails, our coccyx is almost always identical from person to person. There must be other genes that mutated later, helping to produce a uniform anatomy.

Not just in our species, but also in ape ones. They also are uniformly tailless.

So there must be some additional mutations that caused consistent taillessness. BTW, Manx cats vary in how much tail that they have, so their mutation is likely unaccompanied like these Manx-TBXT mice. A further issue with such mutations is that the a double copy of the cat Manx mutation is lethal as an embryo. Adult Manx cats have only a single copy along with an "normal" copy of that gene. So a population of Manx cats where taillessness is normal would need such additional mutations.

What might be the adaptive value of losing one's tail? I'm reluctant to speculate.

Nevertheless, if it is hard to stop growing a tail, that may explain why many mammals have vestigial-looking tails.

 

[lpetrich]

I've gotten into embryonic development quite a bit, and I must bring up a well-known unifying theory of embryology,  Recapitulation theory As Ernst Haeckel stated it "Ontogeny recapitulates phylogeny" - or in simpler language, growth reruns evolution. Since we have no way of observing the original, we are stuck with watching reruns, and according to the theory, we can watch reruns in embryonic development.

I recall from somewhere that EH coined both "ontogeny" and "phylogeny". While "phylogeny" has become widely used, "ontogeny" hasn't.

Rerun theory turns out to be grossly oversimplified, and often just plain wrong in a lot of details. That is especially true of EH's version, where embryonic phases resemble the *adult* phases of ancestors.

A colleague, Karl Ernst von Baer, proposed that it's *ancestral* stages that are shared. He came up with  von Baer's laws (embryology)

1. The more general characters of a large group appear earlier in the embryo than the more special characters.
2. From the most general forms the less general are developed, and so on, until finally the most special arises.
3. Every embryo of a given animal form, instead of passing through the other forms, rather becomes separated from them.
4. The embryo of a higher form never resembles any other form, but only its embryo.

But even those laws of development turn out to be oversimplified and just plain wrong in some details.

One gets closer with a  Phylotypic stage -- "In Embryology a phylotypic stage or phylotypic period is a particular developmental stage or developmental period during mid-embryogenesis where embryos of related species within a phylum express the highest degree of morphological and molecular resemblance."

But even there, it is not a clear picture. The von-Baer-Haeckel model may be called a funnel model, with the most resemblance earliest in development. An alternative is the hourglass model, of convergence on close resemblance, then divergence. That also has some support.

So we are stuck with lack of any great unifying theory of embryonic development.

 

[lpetrich]

1874 Ernst Haeckel Embryo drawings - Stock Image - C008/9505 - Science Photo Library
fish, salamander, turtle, chicken, pig, bovine, rabbit, and human embryos

EH was rather imprecise in identifying the first three, especially the fish.

Creationist foists “fraudulent” embryo picture on his readers | Playing Chess with Pigeons - discussing one creationist's discussion of Ernst Haeckel's drawings. Notes that EH's fish-embryo drawings were much more like amniote ones than is justified, using contemporary drawings and recent pictures. Fish embryos don't have stubby heads like amniote ones do, for instance.

The author presents some embryo pictures from a 1998 article: Michael Richardson embryo pictures:
fish (lamprey, dogfish, gar, salmon, lungfish), salamanders (axolotel, hellbender), snake, chicken, opossum, cat, bat, human

It has problems of its own, like including yolk sacs and other extra-embryonic material, as discussed in Michael Richardson's photographs | National Center for Science Education

-

Even within amniotes, where embryos look very similar when limbs start to grow, there are curious variations, like turtles, lizards, snakes, crocodilians, and birds having much bigger eyes than mammals.