The taste of heat

We've posted a few times recently about the ease of identifying genes, but the inherent difficulty of determining their function.  E.g., there are more than 2000 alleles linked with cystic fibrosis, most of them guilty by association rather than by demonstrated function.  Function is often assumed, and has traditionally been unquestioned based on the 'theory' that this is the gene 'for' the trait.  Genome mapping should make us wary, but it doesn't seem to do that as often as it should.  Of course, the theoretical assumption could be true--it's just very hard to tell experimentally or otherwise based on present knowledge.  Many of more than 100 genes associated with X-linked intellectual disability may in fact have nothing to do with the trait.

Now a paper in this week's Nature ("A gustatory receptor paralogue controls rapid warmth avoidance in Drosophila," Ni et al.) reports that response to a steep temperature gradient by Drosophila may be through a taste receptor.  Another legacy of the "gene for" approach to understanding genetics. 

Most organisms prefer a small range of external temperatures, because it helps in body temperature maintenance, but the molecular mechanisms for thermal preference had not been understood.  Now Ni et al. suggest one.

Here we [show] that thermal preference is not a singular response, but involves multiple systems relevant in different contexts. We found previously that the transient receptor potential channel TRPA1 acts internally to control the slowly developing preference response of flies exposed to a shallow thermal gradient. We now find that the rapid response of flies exposed to a steep warmth gradient does not require TRPA1; rather, the gustatory receptor GR28B(D) drives this behaviour through peripheral thermosensors.

That is, fruit flies respond to minor temperature fluctuation with one mechanism, the TRPA1, and to sharp fluctuations with another, a taste receptor. 

Previous work has found conflicting results with respect to thermal sensing in Drosophila -- the anterior cell neurons, inside the head, were thought to be one sensing mechanism, and the hot cell neurons, in the bristly arista outside the head, were thought to be another, among others.  The internal cells are mediated via TRPA1 but the external neurons are not, and Ni et al. sought to determine the molecular basis of thermal sensing by these cells.

Arista, Drosophila; Wikipedia

There are six neurons reside in the arista, three that respond to warmth and 3 to cold.  Ni et al. found three unidentified cells in the arista that resembled thermoreceptors, and were subsequently found to be gustatory receptors. "Gustatory receptors and odorant receptors have been studied extensively as chemoreceptors for sweet and bitter tastants, food odours, carbon dioxide and other chemicals, but have not previously been implicated in thermosensation."

By inhibiting the detection ability of the neurons in the arista one by one, and exposing the flies to rapid temperature changes and then allowing them to choose between two test tubes of different temperatures, Ni et al. determined that when the gustatory receptors were blocked, the flies did not prefer, as usual, the cooler tube. Expressing GR28B(D) ectopically transferred thermodetection to sites that normally don't detect temperature, fairly strong evidence that this gustatory receptor is involved in detecting temperature as well as chemicals.

Gene names and attributed functions can be both enlightening and constraining. Above is a section from an in situ experiment on a 14 day old mouse embryo, for example, taken from the extensive gene expression database, GenePaint.  The purpose of the experiment was to detect expression of a gene called Spata5.  The darker areas in the photo are evidence of that.  In this case, the strongest expression can be seen in the brain and the central nervous system generally, as well as the tongue, thymus, lungs, the incisors, and other tissues. And what does this gene, with such widespread expression, do?  As characterized by Gene Cards, Spata5 "may be involved in morphological and functional mitochondrial transformations during spermatogenesis."  That is, it's only presumed function has to do with the development of sperm.  

Our point is not to suggest that researchers get it wrong with respect to gene function and so forth, but that there's a lot that we don't yet know about genes.  Gene mapping studies, which aim to identify genes 'for' diseases or traits, necessarily rely heavily on previous knowledge about genes to whittle down what can be extensive lists of potential candidate genes.  This spermatogenesis gene apparently does a whole lot more than help make sperm, but if you were interested in genes involved in the developing lungs, or CNS, and you know nothing more about Spata5 than its putative function, you'd reject it as a candidate immediately.

Mapping studies also typically find that many genes and hence many different genotypes contribute to similar traits in different individuals.  Indeed, single-gene causation is usually due to rare strong-effect variants, but often other instances of the same trait don't involve the gene, and successful single-gene discoveries can seduce investigators into thinking that the whole world is a simple Mendelian garden.

Finding temperature sensing in a taste receptor is an unexpected finding, and in fact seems to have been a lucky artifact of the experiment, not something the investigators set out to test.  Good science involves not just doing experiments well, but seeing around corners, questioning everything, assuming there's a lot we still don't understand -- and that there always will be -- and a lot of dumb luck, and humility.