Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.


When An Arm is Really a Leg

The hands of a person with Liebenberg syndrome resemble feet; the arms resemble legs. (Credit: Dr. Malte Spielmann)

Flipping the X-ray showed Stefan Mundlos, MD, that his hunch was right – the patient’s arms were peculiar and stiff because the elbows were actually knees.

The recent report from Dr. Mundlos’ group at the Max Planck Institute for Molecular Genetics, complete with a genetic explanation for the condition, flew under the radar of  the press-release-driven science news aggregators. But I noticed it because I worked on this sort of thing in grad school – flies with legs growing out of their heads.

An ungoogleable researcher named F. Liebenberg described in 1973 a family with the condition that would take his name: “A pedigree with unusual anomalies of the elbows, wrists and hands in 5 generations.” In the white South African family, four males and six females had stiff elbows and wrists, and short fingers held in a way that made them look strangely out of place. The pedigree revealed classic autosomal dominant inheritance – each child of a weird-limbed person had a 50:50 shot at being so too.

A second family appeared in the pages of the Journal of Medical Genetics in 2000. Those researchers noted that when a patient stood in the “anatomical position” – palms forward – he or she couldn’t bend the arms at the elbow, and the restricted movement had been present since birth.

In X-rays of patients, the elbow joints looked too big. Those of a 6-month-old were the size of elbows of a 3½-year-old, with stubby fingers. Similarly, a 2½-year -old’s elbow was the size of an 8-year-old’s. And like an episode of Law and Order in which a suspect showing up in the first 15 minutes can’t possibly be the perp, those researchers implicated a gene on chromosome 17. That wasn’t it.

Then in 2010 a report appeared on identical twin girls with the curious stiff elbows and long arms of Liebenberg syndrome, in Plastic and Reconstructive Surgery. These researchers noted that the lengths and shapes of the fingers and toes were the same.

Noting that the muscles and tendons of the elbows, as well as the bones, weren’t quite right, Dr. Mundlos and colleagues, experts in the distinctions between vertebrate forelimbs and hindlimbs, realized that the stiff elbows were acting like knees.

The human elbow joint is a hinge in one plane and rotates the forearm in another. The knee is a bit tighter. It extends the lower leg but the kneecap stabilizes the lateral rotation, which is why I constantly injure it in zumba class.

The Max Planck researchers analyzed three unrelated families with Liebenberg syndrome. “The phenotype wasn’t easy to interpret at first glance. It looked like a malformation of the elbow joint and an abnormality of the wrist bones. But wrist bone abnormalities aren’t unusual, in particular fusion of bones,” Dr. Mundlos explained.

But the Liebenberg elbows had a peculiar enlargement, and that was a major clue.

The areas in the white dots are patella — knee — that are parts of the elbow joint. (Credit: Malte Spielmann)

“Normally the elbow joint consists of the humerus, which sits in a socket of an elongation of the elbow, the olecranon, with the radius forming a smaller part of this complex joint. In the patients the olecranon was missing and the joint had a flat appearance, unlike the normal hinge joint of the elbow,” Dr. Mundlos said.

It was when Dr. Mundlos examined a patient’s X-ray from a different perspective that the truth popped out. “I realized that the entire limb had the appearance of a leg. Normally you would look at the upper limb X-ray with the hand up whereas the lower limb is looked at foot down. If you turn the X-ray around, it looks just like a leg,” he recalled.

To anyone familiar with developmental biology, a body part in the wrong place evokes one word: HOMEOTIC. (Warning: spellcheck and autofill convert this to homoerotic.)

A homeotic mutation mixes up body parts, so that a fly grows a leg on its head or antennae on its mouth. Assignment of body parts begins in the early embryo, when cells look alike but are already fated to become what they will become, thanks to gradients of “morphogen” proteins that program a particular region to elaborate particular structures. Mix up the messages, and a leg becomes an antenna, or an elbow a knee.

Just months after I got my PhD in Thom Kaufman’s lab at Indiana University circa 1980, where I savagely murdered millions of fruit flies, post-doc Matt Scott and fellow grad student Amy Weiner discovered how homeotics happen. They identified the homeobox. This 180-base-sequence encodes a protein domain that binds other proteins that turn on sets of other genes – crafting an embryo, section by section. Researchers at the University of Basel found the homeobox at about the same time. (For a fly’s eye view of life with homeosis see The Making of a Mutant, A Fruit Fly Love Story, which I will re-post here for Valentine’s Day.)

Once developmental biologists knew what to look for, homeoboxes turned up in all manner of genomes, affecting the positions of petals, legs, and larval segments. Humans have four clusters of homeotic genes, plus controls.

Homeotic genes line up on their chromosomes in the precise order in which they’re deployed in development, like chapters in an instruction manual to build a body. And they’re ancient. A fruit fly given a mutant homeotic gene from a chicken sports an antennal leg, one species reading the DNA sequence of a very distant other. The homeotic mutants from the Kaufman lab even starred in an episode of The X-Files, about which I recall little except that it featured a Cher impersonator.

In an Antennapedia fly, the antennae develop as legs. (Credit: Rudi Turner)

I left bench science because I’d thought homeotic mutations were a quirk of only fruit flies, mice, and mosquitoes. But after the homeobox discovery, researchers quickly found them in people. In lymphomas, white blood cells detour onto the wrong lineage, and in DiGeorge syndrome, the abnormal ears, nose, mouth, and throat echo the abnormalities in Antennapedia, the legs-on-the-head fly in the photo. Extra and fused fingers and various bony alterations also stem from homeotic mutations.

But nothing could quite match, in a human anomaly, the dramatically mutant fruit flies — until I saw photos of children’s faces with lower jaws turned into upper jaws, in the May 2012 issue of the American Journal of Human Genetics). (See Scientific American blogs for that story.)

The partial arm-to-leg transformation in people, once you know what it is, is even more astonishing.


The homeotic hypothesis explained a lot about Liebenberg syndrome.

This wrist is more like an ankle. (Credit: Malte Spielmann)

“The transformation affects the bones, tendons and muscles of the elbows, wrists and hands. The olecranon of the elbow is completely missing in the patients and the bones of the wrist form a large structure similar to the bones of the ankle. In the 3D CT scans of the elbow you can see a structure similar to the patella of the knee that is fused to the head of the humerus. The bones of the hands are too long and look similar to bones of the feet,” explained Malte Spielmann, MD, lead author of the paper.

Although the researchers ultimately used techniques we are accustomed to these days – genome sequencing, and comparative genomic hybridization (CGH) to detect deletions and copy number variants – their search began as many genetic searches have since the 1950s: with abnormal chromosomes.

Two of the three Liebenberg syndrome families have deletions (missing DNA), and the third has a translocation (two chromosomes exchanging parts). The families share a glitch in the same general region of chromosome 5, providing a toehold in the genome.

Genome sequencing didn’t turn up any likely suspects in the translocation family, but CGH found a 134 kilobase deletion in their genomes. Apparently 134,000 DNA bases were lost when the translocation initially happened, like lopping off letters when cutting-and-pasting text.

The missing DNA in all three families corresponded to the same “gene desert,” a genome region festooned with so-called “dark matter” that doesn’t encode protein. But one candidate DNA sequence did emerge from the regulatory wasteland: a gene called PITX1.

The gene doesn’t encode protein but controls other genes that do. And not only is the gene “highly conserved” – in many species and therefore pretty important – but it controls limb development in mouse embryos.

The researchers had found their gene.

In the Liebenberg families, missing genetic material places an enhancer gene near PITX1, altering its expression in a way that mixes up developmental signals. And so the forming arm gets mixed up, and fashions part of a leg — at first glance barely noticeable, as in this doll. Fortunately the condition appears more an annoying oddity than a disease, and because the gangly arms don’t seem to disrupt everyday life too much, you won’t find Liebenberg syndrome in the rare disease databases like CheckOrphanNORD, or the global genes project.

I like the arm-to-leg story so much that I hardly know where to begin.

#1 I feel better at having spent four years trying to figure out how flies grew legs on their heads, yet worse for having left the field.

#2 I marvel anew at the elegant evolutionary tale that the homeotic mutations tell. When mutation derails development so similarly in such different species as a plant and a person, descent from a common ancestor is the most logical explanation.

#3 Looking at an image from an unusual perspective revealed what no one else had seen. With all the fuss over genome sequencing and nano-everything, we shouldn’t lose sight of the power of larger-scale observation in science.

#4 The homeotic mutations steer development to an alternate pathway. They symbolize, for me, my veering from the path to becoming a scientist shortly after getting my doctorate.

The late paleontologist and science writer Stephen Jay Gould helped me. In his essay “Hopeful Monsters” published in the October 1980 issue of Natural History magazine, reprinted in his 1983 book “Hen’s Teeth and Horse’s Toes,” he wrote about the work in the Kaufman lab, mentioning us lowly graduate students – Barbara Wakimoto, Tulle Hazelrigg, and me.

Thrilled, I wrote to him. And he wrote back, five hand-scrawled pages, encouraging me to follow my instincts to become a writer at a time when many were telling me not to stray from science.

Two decades later, I was to thank him again, unfortunately in an obituary. And now, a decade later, I do so yet again. Thank you, Steve, for telling an unsure graduate student that it’s okay to follow an unusual path.

And congrats to Drs. Mundlos and Spielmann and their co-workers for their insightful discovery.


  1. Thom told me about your aritcle this morning at the YMCA and sent me the reference line so I could download it. I much enjoyed the evidence for a homeotic mutation in humans, Bateson was first to coin the terms homeotic(displacement of parts) and meristic (duplication oif parts). There are lots of meristic examples in humans (duplicated toes and fingers, even boys with two penises) but nothing as dramatic as the case you describe. Thanks so much for writing it. Elof

  2. I discovered your weblog web site on google and check a number of of your early posts.
    Continue to keep up the superb operate. I just extra up your RSS feed to my MSN Information
    Reader. In search of forward to studying extra from you in a while!

  3. I enjoyed reading the story of this interesting discovery! But there is an error in this statement:

    “The gene doesn’t encode protein but controls other genes that do.”

    PITX1 surely does encode a protein; it’s a transcription factor (with its own homeobox, in fact)! You are right that, as a transcription factor, it controls the expression of other genes.

    The deletions/translocations in the genomes of the patients do not knock out the PITX1 protein coding region, but instead move some “enhancer” DNA into closer proximity to the PITX1 gene, which seems to cause the gene to be expressed differently than it normally is (in the developing arms in addition to the legs, where it is normally expressed).

    It is this “enhancer” DNA that does not encode a protein, but instead controls the expression of protein-coding genes like PITX1.

    The PITX1 gene is the subject of another engrossing evolutionary genetic tale, with references for instance here:

  4. Every once in a while, I google Leibenberg syndrome just see in anyone else has come up with more information about it. My curiosity stems from my wife’s curious phenotype in which her tiny hands and crooked elbows are quite conspicuous. When I first me her, she said she got the “Bell arms”, which derives from the paternal surname. Turns out that her father, and two of his sisters all had the same phenotype. Photos of dad, and the aunts clearly show that my wife inherited something from dad’s side of the family.

    They have been puzzling to physicians since birth, but have caused her some problems, particularly nueral, and she’s sought medical care to deal with some of the issues. A neurologist at Duke told her that her elbows were dislocated at birth and the could have easily fixed them at that time. He, however, could not fix or locate the nerves that were causing her problems and set her to to tears in cursing his inability to find anatomy that should have been in the right place.

    I have biology degree, and was a little curious about the familial history of mishapen arms and tiny fingers. The best descriptions of what I observed where from descriptions of Leibenberg Syndrome. There aren’t many and really, the root of the issues has more do with genetics than anyway.

    It would be interesting to have a radiologist look carefully at xrays of her elbows and wwrists. I can sure tell you that her poor old elbows look way more like knees, than elblows.

    Maybe we will get to the bottom of it. We are both having our genomes sequenced through 23andMe. We’ll see if any of the aforementioned deletions or translocations show up and provide and explanation.

    Who knows, we might get a paper out of it.

  5. Thanks so much for writing Mike. Please email me if you’d like to share your 23andMe experience. I am soon revising my human genetics textbook and am in search of narratives from people having genetic tests. I hope your wife’s health improves.

    1. I’ve had fixed bent arms since birth but never knew it had a name. I’m 55 now. My wrists and hands are normal but elbows are locked at apprx 30 degrees and it prevents my arms from fully straightening. Other than looking weird, it doesn’t hurt.

  6. Good day to all,

    I’m 1 of three children that actually knew Dr F Liebenberg. If you’d do the effort, search me on facebook. Send me a friend request and you could view some of the photo’s that were taken when we were still children.

Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top