Development, medicine, and evolution of the neck and shoulder


Neck anatomy has long terrified me. Way back when I was a grad student, my lab studied the organization and development of the hindbrain, which was relatively tidy and segmental; my research was studying the organization and development of the spinal cord, which was also tidy and segmental. The cervical region, though, was complicated territory. It’s a kind of transitional zone between two simple patterns, and all kinds of elaborate nuclei and new cell types and structural organizations flowered there. I drew a line at the fifth spinal segment and said I’m not even going to look further anteriorly…good thing, too, or I’d probably still be trying to finish my degree.

Fortunately, Matsuoka et al. were braver than I was and they have applied some new molecular techniques to sort out some of the details of how the neck and shoulder are assembled. This is a developmental study of how the muscles and bones of the shoulder girdle and neck are derived, and what they’ve identified is 1) a fairly simple rule for part of the organization, 2) an explanation for some human pathologies, and 3) some interesting observations about evolution. It is very cool to find a paper that ties together molecular genetics, development, paleontology, and medicine together so inseparably.

There are many anatomical details in this paper, and in the interest of keeping my summary conceptual, I’m going to skip over most of them. There are two general developmental dichotomies at stake here: neural crest vs. mesodermal derivation, and dermal vs. endochondral bone.

Muscle and bone comes from two sources. The expected source is for it to come from the mesodermal germ layer (remember, embryos after gastrulation have 3 layers: endoderm, which contributes to the gut, ectoderm, which forms the skin and nervous system, and mesoderm, which will make muscle and connective tissue like tendons and bone). We vertebrates have another source of cell populations that can make various things, including bone and muscle: the neural crest. These are cells of ectodermal origin that arise when the neural tube forms; they are kind of left over scrap tissue, a collection of cells that are set aside at the boundary between neural and skin tissue. These cells are wonderfully plastic, and migrate all over the place to fill in and generate many useful parts of your anatomy (you wouldn’t have much of a face without them, for instance, and your peripheral nerves wouldn’t be sheathed with myelin.) One question developmental biologists ask when they examine the formation of a bone or muscle is whether it is derived from neural crest or from mesoderm—and the neck is one of those transitional areas where it isn’t always obvious.

There are also two kinds of bones, from a developmental perspective. Most of your long bones and some of the bones of the skull are endochondral. That is, a cartilaginous scaffold formed first, and then the rubbery chondral or cartilaginous matrix was progressively replaced by bone from within, hence the name “endochondral”, or “within cartilage”. A second kind of bone, dermal bone, skips the cartilaginous phase altogether, and forms directly from cells of the dermis layer of the skin. Dermal bone is associated with sheets and plates of bone, such as the bones of the skull, while endochondral bone is characteristic of most of the bones of your body below the neck.

These two dichotomies have been related by an observation: mesoderm gives rise to endochondral bones, while the versatile neural crest can give rise to either dermal or endochondral bone. This has suggested a shortcut in determining the origin of a bone: if it is dermal, it must be from neural crest. That is, all you need to do is puzzle out the pattern of ossification of a bone, and you’ve got a clue to what population of cells, neural crest or mesodermal, generated it. Unfortunately, this has caused much confusion in our understanding of the neck and shoulder, because it turns out that bones that are dermal in one species may be endochondral in another.

Below is a highly schematized version of the shoulder girdle. The bones are all in grey, and the muscles in red. The bones are also drawn as simple boxes: (1) is the scapula or shoulder blade, to which all the bones of the limb (L) are attached; (2) is the coracoid, which is only a small process in mammals but can be quite large and distinct in other vertebrates; (3) is the clavicle or collar bone; and (4) is the sternum. That’s the series of bones we’re talking about here, an arc that swings from the shoulder blade on our back to the collar bone to the sternum on our fronts, on which our arms are suspended. The muscles in red hold this collar in place. For instance, (T) is the trapezius, the large muscle at the back of your neck that attaches to the skull and flares out to insert on the top of the shoulder blade, while (S) is the sternocleidomastoid, a straplike muscle that attaches to the skull just behind your ear and wraps down and around your neck to attach to the collarbone.

Highly conserved neck muscle scaffolds (red) attach (hatched areas) on a shoulder skeleton (boxes 1−4) that displays variable dermal (light grey) and endochondral (dark grey) ossification type. Attachment regions (hatching) of the gnathostome trapezius muscle (T) are endochondral in sharks, salamanders and all amniotes but are dermal in fish and frogs. A, limb muscles; Co1 and Co2, coraco-branchialis and coraco-hyoideus; L, limb skeleton; R, trunk muscles; S, sterno-cleido-mastoid. In the shoulder skeleton, box 1 is the dorsal cleithrum (dermal) in bony fish (Polyodon or Amia) and frog (Rana) and the scapular region (endochondral) in salamander, mouse and living amniotes; box 2 is the acromio-coracoid (endochondral); box 3 is the clavicular region (dermal/dermal + endochondral), although in sharks bone is absent and its space is taken by part of the scapulo-coracoid (stippled); and box 4 is the sternal region, comprising the sternum (endochondral) or connective tissue (bony fish).

One other important detail about the diagram: bones are either dark gray if they are endochondral, or light gray if they are dermal. If you look at just (1), the scapula, you’ll notice that it is endochondral in mammals and salamanders, but partially dermal in frogs and fish. If the rule of ossification were correct, and the dermal bones were neural crest derived while the endochondral bones are mesodermal, that would suggest that the shoulder blades of frogs and fish were not homologous to the shoulder blades of mice. That would be very peculiar. What Matsuoka et al. show, though, is that the rule of ossification is not reliable, and there is another, better rule.

Green areas in the diagram are neural-crest-derived connective tissues; blue areas are mesodermal (somitic) connective tissues and muscle fibres.

The experiment was straightforward, but not trivial: they worked with lines of transgenic mice that had reporter genes coupled to a neural crest gene (Sox10) and a mesodermal gene (Wnt1). Basically, this puts pretty colored markers in every cell that shows whether it is derived from neural crest or from mesoderm, and allows them to probe the developing neck and shoulder and determine where every cell is coming from. To the right is the simple summary of their results, with all the neural crest tissues colored in shades of green, and the mesodermal derivatives in shades of blue (the part that doesn’t correspond to the diagram above, the blue stuff on the left, are muscles and bones of of the occipital cranium (OC) and vertebrae).

I think you can see the pattern. All the muscles of the neck that insert on the shoulder girdle are neural crest derived, while all the muscles of the limb (L) are mesodermal…and the bones of the shoulder girdle themselves are of mixed derivation, no matter whether they are dermal or endochondral. The rule is that muscles and their attachment points on the bone always have common embryonic origins—distinct populations of cells build the framework of muscles, and interact cooperatively to assemble the bones upon which they insert. The confusing phylogenetic pattern of dermal and endochondral bones is a red herring: homology is defined by the embryonic source population, not the mechanical details of how the bone differentiates. Bones are secondary epiphenomena.

If bones are reasonably labile products of muscle interactions, that goes a long way towards explaining an evolutionary mystery: the disappearing cleithrum. The cleithrum is a bone you may never have heard of unless you’ve studied comparative anatomy, because we don’t have one. It’s a prominent part of the pectoral girdle in fish, and it pops up intermittently in various tetrapods in the fossil record as you can see in the diagrams below. It’s distinguished by its position and the set of muscles that insert on it, and so is readily recognizable—but it seems to have been lost as a distinct bone in salamanders, mammals, turtles, and diapsid reptiles. What happened to it?

The cleithrum (light grey box 1) has been lost (C – ) several times independently in evolution and ‘morphs’ into scapular spines. In all species with a cleithrum—be they fossil mammalian (Varanops), amphibian (Mastodonsaurus) or tetrapod (Ichthyostega) stem-group representatives or extant frogs—it is identical in position to the scapular spine of extant tetrapods with respect to the dual trapezius (T) and mesodermal (A) muscle connectivity and its dorsoventral position. Red zones on fossil forms indicate observed trapezius (T) complex attachment regions.

The paper reports that they have found the “cell population ghost” of the cleithrum in mammals, the conserved set of cells that handle those same muscle insertions in us. These cells form a spine or ridge running up the scapula. Embryonically, then, we still have a vestige of the cleithrum—it’s just that its fate is to fuse seamlessly with the scapula.

In addition to telling us something about evolution, the study also gives us hints to the root causes of some serious human pathologies. Having identified the derivatives of a specific class of cell populations, the post-otic neural crest or PONC, they went searching for congenital defects that correlate specifically with just those derivatives. They found a laundry list of syndromes that map almost perfectly unto just that set of cells.

Several hitherto poorly understood human syndromes precisely match the profile of a PONC syndrome and permit first insights into their common cellular aetiology. Klippel−Feil disease, Sprengel’s deformity, cleidocranial dysplasia, Arnold−Chiari I/II malformation and ‘cri-du-chat’ syndrome are all characterized by co-occurrence of pharyngeal/laryngeal, (sub-)occipital, cervical and shoulder dysmorphologies and swallowing problems. They also share a spina bifida occulta that is unusually confined to the cervical region normally occupied by the trapezius muscle. In Sprengel’s deformity a large fibrous, sometimes endochondral, so-called omo-vertebral bone replaces all dorsal neural-crest-derived endochondral elements of occipital region, cervical spinous processes, spina scapulae and trapezius inside the PONC trapezius territory. On this basis we identify Sprengel’s deformity, which is one of the phenotypic facets of Klippel−Feil syndrome, as primarily affecting PONC fate choices and not cervical segmentation as is currently thought. Moreover, the cervical hypomobility of Klippel−Feil patients can also be understood as caused by defects in PONC fate choices: ectopic ossifications of PONC (trapezius) connective tissues around the somitic mesodermal neck vertebrae and an ectopic ossification of the PONC pre-vertebral ligaments of pharyngeal muscles. Similarly, loss or dysplasia of PONC-derived basicranial (clivus) bone attachments for the internal pharynx and larynx constrictors and ensuing widening of the foramen magnum are the primary mechanical cause of the Arnold−Chiari I and II malformation, a serious human congenital malformation associated with swallowing problems and sudden infant (cot) death syndrome (SIDS).

As a developmental biologist myself, it’s gratifying to see my discipline’s power in action, providing unifying mechanisms that help resolve issues in evolutionary biology and suggest common mechanisms underlying human disease.

Matsuoka T, Ahlberg PE, Kessaris N, Iannarelli P, Dennehy U, Richardson WD, McMahon AP, Koentges G (2005) Neural crest origins of the neck and shoulder. Nature 436:347-355.


  1. Volvox says

    As advertised in the title “Development, medicine and evolution of the neck and shoulder” this article has it all.

    The evolutionary pathway is fascinating and the consequential connection to human medicine is well demonstrated.

    As an aside, I notice nobody claims Intelligent Design for any of the human maladies like those above.


  2. mndarwinist says

    Thank you for the article. It was amazing.
    Of course PZ, you gather a lot more comments for your other posts when you step on some toes. But I find articles like this one way more interesting. Just that I miss the creationists here, offering their rebuttals. Real science is way beyond them.

  3. says

    Yeah, funny isn’t it. The one place where you’d think creationists would be all over the place with their alternate, cogent, informative ideas would be posts like these. But the only place they show up is posts which (I hope you’ll excuse the expression, PZ) don’t matter that much: ones where bluster and arrogance are as useful as brains.

  4. says

    These are my favorites, too. And this one’s especially nice – I’ve got a familial shoulder problem that it just might be shedding some light on. Cool.

  5. Steve LaBonne says

    Ouch. Vertebrate development makes my head hurt. Flies were as much as I could deal with. I leave the vertebrates to my sister the frog neural crest expert.

    Seriously, the lucid summaries of recent developmental bio papers that PZ posts here are amazing (he must be a really superb teacher), and are a wonderfully painless way for me to keep up with some of the neat stuff that I otherwise would be totally out of touch with nowadays. Bravo.

  6. Xanthir says

    Damn, that’s quite cool. Thank you very much, PZ. The bit at the end linking it to human disease was simply amazing. “Evolution can’t do anything useful” my ass. This is straightforward application of evolutionary theory to medical science, hopefully to good effect in the future.

  7. says

    One of the reasons I traditionally found biology harder than physics or chemistry was the terminology always seemed harder to learn. This illustrates that in spades – I had to wade through it here. The gist is quite interesting, however. So bone is not related to shells and coral hard parts and that sort of thing?

  8. speedwell says

    I am talking to a chat friend right now, and she mentioned her daughter-in-law has Chiari syndrome, and I was able to talk intelligently and compassionately to her about it because I had read this post, and some further information about Chiari and the other malformations. Wow, this really helped. THANK YOU.

  9. David Marjanović says

    So bone is not related to shells and coral hard parts and that sort of thing?

    Not at all, no. It could have some relationship to the hard parts of echinoderms, but that’s all.

  10. David Marjanović says

    So bone is not related to shells and coral hard parts and that sort of thing?

    Not at all, no. It could have some relationship to the hard parts of echinoderms, but that’s all.

  11. Stephen says

    I’d just like to add a belated cheer for this stunning article.

    To explain such a difficult and specialist topic clearly enough that a layperson can (sort of) understand it takes both skill and effort – thank you!

  12. clinteas says

    I never saw this until today.Amazing.
    Used to love embryology,but Ive forgotten most of it….