Deleting a gene can turn an ovary into a testis in adult mammals

By Grant Jacobs 19/01/2010

You’d think that in adult mammals ovaries are ovaries, and that’s it. They’re committed to being what they are.

Or as geneticists would say, they’re terminally differentiated: they’ve reached the end of their differentiation pathway.

Well, it seems you’d think wrong. (This writer, too!)

In a stunning paper Henriette Uhlenhaut and 14 others show that if adult mice lose a Foxl2 gene, ovaries become testes.

These researchers raised mice in which they could delete the Foxl2 gene by treatment of tamoxifen, a compound that competes to block the estrogen receptor. (They used an inducible cre-recombinase system that responds to estrogen antagonist, tamoxifen. I’m summarising very lightly here for a general reader; biologists should start with the summary article.)

The surprising and unexpected result was that when adult mice were induced to lose their Foxl2 gene, their ovaries changed into testes! (I emphasis ‘adult’, as while embryos are still developing, adults are not.)

One up-shot is that this tells us that the ovary ’fights’ to maintain it’s status as an ovary throughout life. It’s not permanently committed. It is not a pathway that comes on ’by default’, either. Lose the Foxl2 gene and it changes.

Slightly more formally, this research shows that the ovary has to maintain constant suppression of the key testis development gene Sox9 by Foxl2; if not ovarian granulosa and theca cells change to become testicular Sertoli and Leydig cells, respectively.

Uhlenhaut and colleagues observe that the full set of genes associated with testis development becomes active and these XX (genetically female) mice produce similar amounts of the male sex hormone testosterone as XY (genetically male) mice.

My initial thoughts were that this work might be related to the well-known sex-reversal seen in some species of fish. The Australian scientists writing the summary article share this view (but coming from much more knowledge of the background than I have). They say it may also explain the female to male sex reversal of goats in polled intersex syndrome, where a region of their genome containing the Foxl2 gene has been lost. These workers also point to their own work, where ’knockdown’ of another gene, Dmrt1, required for testis development in chickens (and probably all other birds), results in feminization of the birds.

Taken together, it seems the view of the permanence of the status of the ovary in adult females in many species, not just mice, is under review.

It’s possible that this gene control pathway is also involved in human patients with premature ovarian failure or other disorders of sex development. I’m sure that these patients and their families will follow this emerging story closely.

Added after the article was written:

I’ve just realised that there is a video available explaining this work. I’m not sure if there is a way to “embed” this video, and I’m too short on time to experiment, so in the meantime at least you’ll have to visit there yourself.

It’s short (4 minutes, 18 seconds) and worth viewing, although you may want to replay it a couple of times if the terminology is a bit much for you.

It’s good to see Cell (the journal that published this work) making moves in this direction.

Update (23rd July 2011)

Ed Yong has written an account of a recent addition to this line of research, which shows the opposite situation to the masculinisation of ovaries described above – feminisation of testis cells through loss of the gene that represses feminisation of germ cells. This new work has parallels with the work I mentioned briefly that knockdown of Dmrt1 results in feminisation of chickens – it also acts through loss of Dmrt1. (I may attempt to write my own account of this work later, but I’m in two minds about trying to given Ed has already written on it. [He has the advantage of having access the journal embargoes so he can get these out on the day, before bloggers like me.])

References (both require a subscription, or a trip to a university library, sorry)

Summary article

Sinclair and Smith

Females Battle to Suppress Their Inner Male

Cell 139(6)1051 (2009)


Full paper

Uhlenhaut et al.

Somatic Sex Reprogramming of Adult Ovaries to Testes by FOXL2 Ablation

Cell 139(6) (2009)


Uhlenhaut NH, Jakob S, Anlag K, Eisenberger T, Sekido R, Kress J, Treier AC, Klugmann C, Klasen C, Holter NI, Riethmacher D, Schütz G, Cooney AJ, Lovell-Badge R, & Treier M (2009). Somatic sex reprogramming of adult ovaries to testes by FOXL2 ablation. Cell, 139 (6), 1130-42 PMID: 20005806

More medical/health/genetics-related posts on Code for life:

Monday potpourri: maps, malaria in the USA, cholera in Dunedin and vaccines

Minorities, disabilities and scientists

Neti pots now validated as sound science?

Genetic tests and personalised medicine

Donated to Science to screen

Book excerpt – Losing the faces of your wife and children

Autistic children and blood mercury levels

Metagenomics-finding organisms from their genomes

Monkey business, or is my uncle also my Dad?

0 Responses to “Deleting a gene can turn an ovary into a testis in adult mammals”

  • What I found particularly interesting about this was that during my studies in development (genetics), we were told that female was the default, as it were… This would seem to suggest that strange hermaphrodite may have been a more accurate description 🙂

    • Speaking as a strange hermaphrodite, or as we like to say in the non-rude camp, transgender person…

      …I believe the article author should revise the last statement to note that transgender folks will also be very interested in this technology advance when it reaches fruition (no pun intended).

  • Yeah, that’s what everyone gets told eh?! It’s a bit of a case of reality always being more complex, I suspect. FWIW, my own interest was if there was an epigenetics aspect to this. (Need to do more reading…)

    Couldn’t think of a decent graphic/image for the story.

  • It’s very unlikely… these effects have only been found in specially bred mice, and birds… but it’s just barely possible that this might account for a few of the cases of human dichogamous pseudohermaphroditism that have so far eluded other explanation. Most cases of human dichogamous pseudohermaphroditism are protogynous, and well understood – they’re due to either 5alpha-reductase-2 deficiency (5alpha-RD-2) or 17beta-hydroxysteroid dehydrogenase-3 deficiency (17beta-HSD-3) . These conditions can lead to a somewhat female appearance at birth slightly masculinising later, all the way to an apparent “natural sex change” from approximately female to almost normally male, or anywhere in between.

    But… there’s a few, a mere handful, of cases that are protandrous, going from male to female. They’re not well understood, and there appears to be many different etiologies. This could be one of them – or part of some of them.

  • Zoe,

    Thanks for your input. Just for other reader’s clarity, Zoe is referring to the final paragraph and it’s mention of the work possibly relating to human “premature ovarian failure or other disorders of sex development”.

    My final paragraph was just passing on (and condensing) what the authors of the summary paper thought. (I’m not a specialist in sexual development, my interest in this is if are any possible epigenetic aspects, i.e. the molecular genetics side of things.) They do in fact refer to the unexplained cases, which I dropped. I suspect with that in place you’ll be agreeing with them.

    Here is the relevant paragraph from their article (the final one in the article):

    Many disorders of sex development in
    humans remain unexplained. Similarly,
    the molecular mechanisms underlying
    premature ovarian failure in women are
    not fully understood. If the same phenomenon
    observed in this study applies
    to humans, then it may at least partly
    explain the etiology of these conditions.

    Nothing against what you’re saying but while you can look at the pathways involved and see possible routes from male to female, this particular research doesn’t directly address this to my reading. (Bear in mind that my focus here is the observation of transdifferentiation of ovaries to testes in adult XX mice, not the wider work in the field.)

    I’ve just realised that there is a video summary of the this article. It’s a pity I hadn’t seen this sooner. I’ll add a link to it in the article in a moment, but here it is in the meantime:

  • Hmm… blockquotes don’t work in comments. Sigh.

    (The quoted bit that is “ragged right” is “supposed” to be a block quote, but the software isn’t presenting it as one.)

  • Unfortunately I’m unable to be adequately objective on the issue. On the other hand, I make a great experimental subject, able to give full and informed consent.

    It would have been nice to make science journals for something I’ve done, rather than something I am, but I have to take what I can get. It will soon be 5 years since my metabolism went haywire. I really thought we would have found out the cause by now.

  • Unfortunately I’m unable to be adequately objective on the issue.

    Great to see someone who’s upfront and honest!

    I really thought we would have found out the cause by now.

    Unfortunately, science is often a very slow process, so for a lot of things (not all) it tends to produce treatments for the next generation rather than the current one.

    Looking at the photos you link, at least you can say you have a lot of company! 🙂

    (Sorry for the slow reply, I completely forgot to send it!)

  • Does anyone know if/when they plan to conduct any research on humans? In complete seriousness, I would love the opportunity.
    If the ovaries can be made to produce testosterone instead of estrogen – that would mean a lot for the female-to-male transgender community!

  • Andrew,

    I don’t know of any plans like that, not that I would as I don’t follow developments in that area.

    Aside from any ethical issues, you couldn’t use the method used in the research as that starts with one generation to create the loss of the Foxl2 gene in their offspring, whereas a treatment (for want of a better term) would want to work on the patient. (I’d explain how this method works, but seeing as I might have done that in the original article and that’s now over three years old perhaps I can be excused for leaving that for now.)

    I can conceptually see how someone might try shut down the Foxl2 gene, in the way you can speculate endlessly on these sorts of things. It’d come down to the details and I’m not familiar enough with these cells to talk about that! That said, there are a number of attempts to shut down specific genes in recent experimental gene-targeted therapies. I’ve meant to write about these, but one way or other haven’t. Perhaps I’ll put this right sometime. In simple terms, ‘designer’ DNA-binding proteins are made that aim to bind to just one region of a genome to encourage shutting down the gene targeted.* It’s very interesting stuff, but it’d be a long time coming if this were attempted – this area if very exploratory and faces a lot of issues, for just example how to target particular cells (e.g. within the ovary) and not other parts of the body.

    * A different variation of this general theme was tried for haemophilia, which I did write about a couple of years ago.