How to make a vulva


The vulva is one of my favorite organs. Not only is it pretty and fun to manipulate, but how it responds tells us so much about its owner. And it is just amazing how much we’re learning about it now.

Don’t worry about clicking to read more…this article is full of pictures, but it is entirely work safe because it’s all about science.

It also helps that all I’m going to talk about is worm vulvas. Anybody who has taken a developmental biology class in the last ten years knew exactly what to expect: when a developmental biologist starts getting all enthusiastic about vulvas, you know he’s thinking about nematodes. The C. elegans vulva has become one of the classic model systems for studying induction in the formation of organ systems.

They have a number of virtues. Vulval development is very well characterized, and the organ itself consists of a relatively small number of cells: 22 cells total, organized into 3 general classes, with 7 distinct patterns of gene expression. Here’s what it looks like; it’s a few cells bracketing a small opening in the ventral body wall of the worm.


This photo has a pedigree superimposed on it, and as we’ve come to expect in the worm, the formation of this structure is in part the result of a specific series of cleavages by a small number of cells, which can be observed and mapped over time. The cells we care about are ac, the anchor cell, which is within the worm gonad, and a string of cells named P3.p, P4.p, P5.p, P6.p, P7.p, and P8.p along the ventral body wall. The Pn.p cells form an equivalence group: as near as anyone can tell, they are all functionally equivalent to one another initially, and the process of forming a vulva is one of progressively differentiating this pool of cells into several distinct types.

Here’s a cartoon illustration of the process.


During development, cells fall into one of 3 broad categories. The 1° fate (the orange cells) is to form a stack of cells that make the actual vulval opening; the 2° fate (in blue) is to form supporting cells that, for instance, anchor muscles that open and close the vulva; and the 3° fate (gray) is to fuse with the hypodermis, or “skin” of the worm, and become something less glamorous than a vulva. These cells are not predetermined to follow these fates, however, and all initially have the potential to become any part of the vulva. We know this because there are mutants that shift the position of the anchor cell, and suggestively enough, where ever ac is, that is where the vulva will form. If ac is shifted anteriorly to lie next to P4.p, for instance, P4.p will adopt the 1° fate, P3.p and P5.p will be 2°, and P6.p, P7.p, and P8.p will be 3°.



One other useful feature of the vulva (to the experimentalist) is that the worm vulva is dispensable. Mutations that disrupt vulva development aren’t immediately lethal, and don’t even prevent the animal from reproducing. The worm is hermaphroditic, so if it lacks a vulva it can just self-fertilize its own eggs internally. There is the unfortunate problem that the eggs and embryos have no opening for getting out of Mom, but hey, the babies are tough and resilient, so they go ahead and grow internally, producing the lovely ‘bag of worms’ phenotype shown above, and eventually chew their way out.

There are a number of genes that specify the 1° fate, and they are called Vul (for vulvaless) genes because when they are mutated, the animal fails to make a vulva.



There are also complementary mutations that produce the multivulva phenotype. All of the Pn.p cells decide that they want to adopt the 1° fate, never mind where the anchor cell is, and they make vulvas all over the place.

There are also multiple genes necessary for suppressing 1° fates and promoting 2° and 3° fates; they are called Muv (for multivulva) genes because, when they are mutated, the normal constraints that allow only one vulva to form are lifted.

By knocking out genes that affect vulva development one by one and in combination, and by carefully analyzing the results, investigators have come up with models for how all the molecular/genetic pieces interact. The diagram below is one summary; these models are being constantly refined and made more detailed, but this one is sufficiently simple to give you a taste. It’s like a kind of circuit diagram, illustrating patterns of interactions.


Start with the anchor cell. As we know from the experiments that move the anchor cell around, it is the source of an inductive signal (in developmental biology, induction is a process in which one cell or tissue instructs another cell or tissue to adopt a particular fate, or express or repress a particular pattern of gene expression) that specifies the 1° cell fate. In the diagram, the thick downward pointing arrow represents a strong signal that activates the Vul signal transduction pathway, which in turn activates 1° genes in the nucleus, and turns off 3° genes. Note that there also Muv pathway genes trying to turn off Vul, but the inductive signal is strong enough that Muv can’t overcome Vul.

In addition, one effect of activation of the Vul pathway is to promote the production of another signal, the lateral signal (LS). The 1° cell is going to instruct its neighbors to not become 1°there can be only one. It’s also going to turn off its receptor for the lateral signal, the lin-12 gene product, so that its warning to its neighbors doesn’t also shut down its own set of 1° genes.

Now this is the short, sweet, and simple explanation; the details can get overwhelming. For instance, molecular analysis has revealed that the 1° and 2° categories also have subtypes within them. In the overview diagram below, you can see that P6.p, which normally gives rise to the 1° cells, generates two cell types, vulE and vulF (E and F for short), which can be distinguished by their pattern of gene activity. P5.p and P7.p produce the 2° cells, which consist of four more distinct types, vulA, vulB, vulC, and vulD. The whole organ is formed by a specific and mirror-symmetric array of these cell types, forming an ABCDEFFEDCBA configuration.

An overview of vulval development. Lineal origins of 22 vulval nuclei are indicated. “ABCDEFFEDCBA” refer to vulval cell types vulA, vulB1, vulB2, vulC, vulD, vulE, and vulF. vulB is the only case in which a single VPC granddaughter gives rise to two cell types. Vulval cell nuclei at each stage are positioned as indicated (left, anterior; right, posterior)

Oy, but it gets even more intricate. Cells have histories; being an “F” cell is not a static thing, but is also indicative of a temporal pattern of gene activity. The diagram below shows which genes are active in each of A-F at different stages of development.

summary of cell-type specificity and timing of expression in the wild type. Boxes indicate stages at which gene expression is activated. The vertical order of events within each time block is arbitrary. For egl-17, vulE/vulF expression begins in P6.p (early L3) and persists in their descendants (vulE and vulF) until turned off in the early L4 stage. This inactivation, which is regulated by lin-29 and lin-11, is indicated by the box marked “egl-17 OFF.” ceh-2 is expressed at a higher level in vulB1 compared with vulB2.

Each of these genes can be taken apart in detail and the organization of their regulatory elements examined. Here, for instance, is part of the regulatory region for egl-17, showing enhancer elements and the factors that can bind to them.

cis-regulatory elements of egl-17. A map of the egl-17 5′ regulatory region. Boxes indicate enhancer elements defined by Cui and Han and Kirouac and Sternberg. “AND” and “OR” logic gate symbols indicate sites and logic of information integration. Temporal (blue) and spatial (red) information is integrated as indicated by the logic circuit diagram to produce the complete egl-17 expression pattern.

I’m just skimming through the details here, but the impression I want to leave you all with is that this is a relatively simple organ consisting of only 22 cells in a very simple organism, C. elegans, and the complexity is stunning. Everything is precisely regulated and interlocking, and the outcomes of the process are so robustly determined, that if I were a creationist, I might look at it and say, “There’s no way that can evolve,” and then I’d throw up my hands and go back to the simplistic certainties of religion. And that’s precisely what many of them do.

That is not what scientists do, though. Instead, they dig a little deeper into the natural world and see if they can find real answers, and this is a case where they’ve done just that. Evolution, it turns out, is very good at generating complexity. By looking at other worms related to C. elegans, we can see that they are just as complex, but differently complex, and what’s even more interesting, that we can start to see the pathways that bridge them.

Inoue T, Wang M, Ririe TO, Fernandes JS, Sternberg PW (2005) Transcriptional network underlying Caenorhabditis elegans vulval development. PNAS 102(14):4972-4977.

Sundaram MV (2004) Vulval development: the battle between Ras and Notch. Current Biology 14:311-313.


  1. #1 André Esteves
    January 18, 2007

    Would it be possible to create a closed circuit ecosystem with C.Elegans and produce speeded spontaneous macro-evolution to a new species?

    As C.Elegans is such a well understood model (from what i know there is a complete pedigree for each and every cell of a adult worm), it seems a perfect choice.

    The microscopic size of the worm also would permit to build a managable experiment.

    At least, it would finaly shut up the creat’s about macro-evolution…

  2. #2 André Esteves
    January 18, 2007

    Azkyroth: Exactly. For creaticionists a metabolic or a signalling change don’t matter. They cannot see it, they will not believe it…

    Although those changes could in reality be even more dramatic than spouting legs or changing size…