Genes often get their names from a single function, or from a disease or disorder they’ve originally been associated with (except for fruit fly genes, which get great, but usually totally uninformative names like sonic hedgehog or sevenless or bride of sevenless or cheap date – which are not only entertaining, but their being uninformative is probably better than the restrictive and potentially misleading names given to everyone else). In humans, we’ve got things like ‘Testis-expressed’ genes, or breast cancer genes (BRCA), or immune genes, and so on. A lot of information about what these genes do goes missing when we rely on this kind of naming, an indication of just how much we’re still prisoners of Mendel’s single-gene, single-function approach (he didn’t know of genes per se, of course).
Here’s a picture of the expression in a mid-gestation mouse of a gene called ‘CFH’, or complement factor H, a protein that’s secreted into the bloodstream and that is involved in the complement pathway of the immune system that reacts to the detection of pathogen-infected cells.
The picture is from a very useful database called GenePaint (GenePaint) that is in the process of cataloguing expression of more than 20,000 genes in the developing mouse. Zooming in on such pictures shows the specific cells that are expressing the gene – choose your favorite gene and check it out on GenePaint yourself. Here, we’ve labeled various structures (and it’s in fact a figure we’ve used in a paper of ours in the Feb 2009 issue of BioEssays (vol. 31: 198-208.)).
The picture is of a transverse section showing the results of an ‘in situ hybridization’ experiment looking for expression of the CFH gene in a 14-day old mouse embryo. Imagine you’re looking at a profile of the embryo, it’s sliced into many thin sections and the sections are put onto a microscope slide and then bathed in a solution that contains an RNA ‘probe’ that binds to CFH message in whatever cell the gene is expressed. The probe is treated so that it will turn purple wherever it binds, so the dark purple in this picture indicates where the CFH gene is expressed.
The interesting thing is that this ‘immune gene’ is expressed in lots of unexpected places in the developing mouse. It’s clearly more than an immune gene, and must have something to do with function that’s not yet known. The same can be said of most genes.
Here’s another picture, with in situ results again from GenePaint, but annotated by Sam Sholtis, lately of our lab, now a post-doc in Jim Noonan’s lab at Yale.
Panel A shows expression of Hap1, huntingtin associated protein, clearly not just about brain cell function, and panel B shows expression of Tex2, testis-expressed gene 2, clearly not restricted to the testes -- the insets show that they are strongly expressed in developing teeth, as well as elsewhere. Sam was interested in genes expressed in developing teeth, and if he had restricted himself by gene names or the literature about them, he would never have looked at the two genes in this figure.
Whatever these (and most) genes do, it is inaccurate at best to think of them as genes ‘for’ some specific thing. Their expression as seen in such figures as these could be incidental and without functional importance, but more commonly there is some function that we can identify if we but look for it. It is accurate to say that these genes contribute to the particular structures in which they’re expressed in these embryos, but it may not be accurate even to say the genes are ‘for’ this set of structures. That’s because this is just one instance of one embryo, at one stage of development, in one inbred (and hence rather artificial) strain of laboratory mice, under one set of circumstances. Who knows what other structures or molecular processes might depend on or involve these same genes?
Genes-for thinking works when a mutation has a strong effect, as in, say, many genetic diseases of childhood such as cystic fibrosis or Tay Sachs disease. But it fails us when we’re trying to understand complexity. It is first of all a major problem to overcome gene-for thinking, which has been entrenched by the history of discovery in genetics that goes all the way back to Mendel, and which is so appealing since we like to find simple causes of things. It is much more challenging to determine how to think of organisms if not in gene-for terms.