Last week, I wrote a bit about maternal genes, specifically bicoid, and described how this gene was expressed in a gradient in the egg. Bicoid is both a transcription factor and a morphogen. The gene product regulates the activity of other genes, controlling their pattern of expression in the embryo. Today I thought I'd get more specific about the downstream targets of bicoid, the gap genes.
The gap genes are zygotic genes. Unlike maternal effect genes, which are transcribed from the mother's DNA, zygotic genes are activated in the fertilized embryo (the zygote) and are transcribed from the zygote's DNA. The gap genes in Drosophila get their name from the observation that mutations in these genes knock out, or cause a gap, in the body plan—lose the gap gene krüppel, for instance, and a chunk of the embryo's middle fails to develop. The genes themselves are expressed in restricted bands that correspond to the regions that are lost when they are mutated, as can be seen in this diagram and photographs of a few stained embryos.
The interesting developmental question here is how a gradient of one gene product, bicoid is translated into a pattern of stripes of expression of other genes. The answer is that there is a complex pattern of interactions that have been teased apart, one by one.
For example, one gene, hunchback, is relatively simple. It is directly regulated by bicoid, so that wherever the bicoid concentration is above a certain level, hunchback is turned on. As you can see from the diagram, hunchback is therefore turned on in cells in the anterior half of the embryo. We can play with this, increasing the concentration of bicoid and seeing the expression of hunchback turned on in more cells, or turning bicoid down and seeing hunchback similarly downregulated.
What about making stripes? Here, regulation gets more complicated. Krüppel, for instance, overlaps in its expression with hunchback, and is regulated by it. If there is no hunchback, there is no Krüppel expression. If there is a little bit of hunchback, Krüppel is turned on. If there is lots of hunchback, though, Krüppel is turned off again...so Krüppel is only turned on in a narrow band where the hunchback concentration is just right, where the fly has the perfect balance between repression and activation.
There is more to this story, of course. The gap genes regulate each other, so in addition to being controlled by hunchback, Krüppel is also inhibited by another gap gene, knirps. All of these genes are jostling one another, turning on some genes and turning off others, all triggered initially by the gradient of bicoid expression, to produce their final arrangement.
And it doesn't stop there! The gap genes are only the second step in the process. The gap genes in turn regulate another set of genes, the pair rule genes, which produce lovely alternating stripes of gene expression that correlate with every other segment, like this:
The pair rule genes are numerous, and also interact with one another, and will in turn regulate yet another level of the hierarchy, the segment polarity genes. The segment polarity genes are turned in in every segment, within specific subregions of the segment. Years of tracing these interactions now allows us to assemble diagrams of the regulatory cascade involved that look like this, where arrows indicate that a gene activates another, and bars indicate that it inhibits it:
Note that this is not a complete map of all of the regulatory interactions, and that there is much more that has to be filled in below it—with the segment polarity genes wingless (wg) and engrailed (en), we're still describing transcription factors that are going to go on to regulate yet another set of genes!
One last summary cartoon, since it illustrates these branching hierarchies of gene regulation simply and leaves me with a point that I'll want to follow up on later:
This shows how a localized deposition of maternal bicoid mRNA eventually leads to cells in different regions of the embryo expressing different genes. Note that I've only been discussing the effect of one gene, bicoid, and have said nothing so far about another interesting maternal gene product that is also localized in the egg, nanos (nos). Nanos is also expressed in a gradient, like bicoid, but unlike bicoid, it is high at the posterior end and low at the anterior end. In Drosophila, though, nanos has a peculiar but enlightening job to do, that I'll write about in a day or two.