Drosophila Smell and Taste

Blogging on Peer-Reviewed Research

Sensing and reacting to one’s environment is necessary for survival. Different species have different expertise in regards to how they sense their environment. Humans, for example, have reduced olfactory abilities relative to other mammals, but excellent color vision. Cats have good night vision, but poor vision during day light. These proficiencies and deficiencies in sensory abilities hold for non-mammalian taxa as well.

Olfaction and taste have been well studied in a variety of taxa. Amongst the invertebrates, the genes responsible for olfaction and taste in Drosophila are one of the best studied systems. Much of the recent work mapping genes to functions in the olfactory system has been done in John Carlson’s lab, where they have figured out which genes are expressed in which neurons and which odors those neurons detect. Carlson’s work was done in D. melanogaster (the common work-horse in molecular biology and genetics), but with the availability of completely sequenced genomes from 11 other Drosophila species, researchers have begun to study the evolution of olfactory and gustatory genes within the entire genus.

As the genomes of multiple Drosophila species became available, multiple labs mined the sequences for olfactory receptor (Or) and gustatory receptor (Gr, or taste) genes. Two papers from two different research groups were published earlier this year reporting on the evolution of Drosophila Or genes (here and here). Additionally, at least two other groups have performed their own computational analysis of Or genes in Drosophila genomes (based on titles I’ve seen for talks and posters that will be presented at the SMBE meeting). The Drosophila Or repertoires evolved independently from the Or and Gr genes found in mammals. These studies found that the sizes of the gene family has remained relatively constant, despite frequent duplication and loss of Or genes. The evolution of Or genes appears to fit a birth-and-death model for gene family evolution.

Mining genomes for your favorite gene family is expected as genomes become available. While these studies are important in that they contribute necessary annotations to these genomes, the science is fairly mundane. Also, there’s something a bit questionable about publishing a paper in which all of the research reported is based on data that are publicly available (the genome assemblies), but the people who generated the data have not yet published their paper. In a way, they violated an unwritten moratorium in the genomics community. The labs who published papers on the Or genes in the 12 Drosophila genomes should have waited for the genome papers to be published before publishing their own analyses of those sequences. Or, better yet, they should have joined the consortium of labs analyzing the 12 genomes and published their papers along with the genome papers from the consortium.

There is also some really interesting research being done on Drosophila Or and Gr genes. These studies explicitly address questions that get at the biology of the organisms being studied. In one study, Lindy McBride compared the Or and Gr repertoires of the sister species D. simulans and D. sechellia. D. simulans is a generalist when it comes to host plant preference, whereas D. sechellia only feeds on the fruits of Morinda citrifolia — a plant that produces toxins that all other drosophilids find objectionable. McBride found that D. sechellia is losing Or and Gr genes much faster than D. simulans, and the rate of protein evolution in these genes is higher in D. sechellia than D. simulans.

The changes in rates of evolution along the D. sechellia lineage could be the result of positive selection or relaxed selective constraint. Ideally, one would use population genetic data to differentiate between the two, but D. sechellia has very few polymorphisms segregating in natural populations, limiting the power of the applicable analyses. Instead of using polymorphism data, McBride analyze where the substitutions were occurring within the genes. She found that D. sechellia Or genes have changes in the same parts of their proteins as D. simulans and D. melanogaster (the outgroup species) Or genes, but D. sechellia Gr genes do not. She uses this observation to argue that adaptive evolution has driven the accelerated evolution of D. sechellia Or genes (they are still functionally constrained because the changes can’t happen anywhere in the protein), whereas relaxed selective constraint is responsible for accelerated evolution of the Gr genes.

And if you’re more interested in the function of the genes responsible for the host shift in D. sechellia, this paper might tickle your fancy. Takashi Matsuo and colleagues used D. melanogaster lines with knocked-out odorant binding protein genes (which are different from Or genes) to identify genes involved in the tolerance to and preference for the Morinda plant. They showed that one gene is differentially expressed in D. sechellia compared to its close relatives, and knock-outs of that gene (along with another gene) allow D. melanogaster to tolerate some of the toxic volatiles produced by the Morinda fruit.

Interestingly, D. sechellia diverged from D. simulans approximately 1/2 million years ago, but the Morinda plant that D. sechellia feeds on was introduced to the Seychelles islands (home to D. sechellia) by humans (within the last few thousand years). That means the speciation event between D. simulans and D. sechellia was not driven by host adaptation of D. sechellia to the Morinda plant on the Seychelles. I can think of three scenarios for the host shift observed in D. sechellia:

  1. D. sechellia was adapted to Morinda (or some other plant that produces similar toxins) on mainland Africa (home to the native range of D. simulans and presumably the native range of the ancestor of the two species) and then colonized the Seychelles upon the introduction to the plant.
  2. D. sechellia and D. simulans diverged in the absence of a host-shift on mainland Africa, and D. sechellia shifted hosts upon arriving in the Seychelles.
  3. D. sechellia colonized the Seychelles and speciated from D. simulans in allopatry, with the host-shift coming subsequently upon the introduction of the Morinda plant.

Whatever the historical series of events giving rise to the adaptation of D. sechellia to Morinda citrifolia, it’s still interesting to figure out how the species came to tolerate and prefer this plant. These two studies shed light on whether the host-shift was a matter of adaptive evolution or lack of selective constraint, and which genes were responsible for the host shift. The study by Matsuo et al also addresses the issue of whether coding sequence or regulatory regions are involved in adaptive evolution — a major debate in the evolutionary genetics literature.

Guo S and Kim J. 2007. Molecular evolution of Drosophila odorant receptor genes. Mol Biol Evol 24:1198-1207 doi:10.1093/molbev/msm038

Matsuo T, Sugaya S, Yasukawa J, Aigaki T, and Fuyama Y. 2007. Odorant-binding proteins OBP57d and OBP57e affect taste perception and host-plant preference in Drosophila sechellia. PLoS Biol 5:e118 doi:10.1371/journal.pbio.0050118

McBride C. 2007. Rapid evolution of smell and taste receptor genes during host specialization in Drosophila sechellia. Proc Natl Acad Sci USA 104:4996-5001. doi:10.1073/pnas.0608424104

Nozawa M and Nei M. 2007. Evolutionary dynamics of olfactory receptor genes in Drosophila species. Proc Natl Acad Sci USA 104:7122-7127 doi:10.1073/pnas.0702133104


  1. #1 Martin R
    June 23, 2007

    With that headline, I was expecting to learn what Drosophila smell and taste like. Can I have mayonnaise with mine?

  2. #2 Peter Ellis
    June 23, 2007

    Also, there’s something a bit questionable about publishing a paper in which all of the research reported is based on data that are publicly available (the genome assemblies), but the people who generated the data have not yet published their paper.

    I do understand where this is coming from, but where does the line get drawn? What if the data’s been in the public domain for several years, but there’s no sign of an upcoming paper? How long can the sequencing consortium effectively control a whole field of research?

    Disclaimer: I’m not a fly person and certainly don’t know what the timing parameters look like in this specific case.

    Second question: how much non-genomic data would you have to have in the paper before it’s acceptable to make use of the genomic data?

    To take a concrete example: say you discover via transcriptional profiling that a particular oligonucleotide probe is regulated in some experimental model of choice with an interesting phenotype. You compare that probe to public cDNA sequences and discover that it’s an uncharacterised gene that has some full-length sequence available from one of the massive cDNA library sequencing projects.

    You compare the full-length sequence to the public genome data and determine the intron/exon structure. You do some data mining on the genome data to find orthologues and put together the evolutionary history of the gene – which turns out to be deeply weird, and directly driven by the gene function.

    In parallel, you do a more in-depth investigation of the transcriptional complexity of the gene. So you have your own data on variant splicing, expression in different cell types and developmental stages, and so on. Tying it all together leads to solid conclusions about how the function of the gene has affected its evolution and that of the chromosome it sits on.

    How much of that body of work should you hold up, waiting for the paper to come out from the sequencing consortium covering your genomic area of choice? Sure, a lot of it is dependent on someone else’s unpublished(?) work, but a lot of it isn’t. And the parts of it don’t make sense independently.

    Put more succinctly: what is the point of releasing genomic data if the unwritten rules say you can’t use it?

    (Disclaimer: the above is of course a real life example. I’m working on novel Y chromosome genes. Pretty much nothing on the Y makes sense except in terms of the structure and evolution of the chromosome. The mouse Y sequencing consortium hasn’t published their work yet.)

  3. #3 Peter Ellis
    June 23, 2007

    The mouse Y sequencing consortium hasn’t published their work yet

    Rereading the above, I should stress that I’m in no way saying they’re holding things up – it’s just that it’s an incredibly complex chromosome and will take a lot of time to do right. I don’t envy them the job!

  4. #4 RPM
    June 23, 2007

    Thanks for the comment Peter. I agree that there is no clear cut distinction. I’d recommend contacting the genome sequencing consortium and telling them you’re planning to publish your findings using their data. Technically, once the data are made available, it’s a free-for-all, but it’s only polite to let them know what you’re doing. I know people who have done that, and gotten no beef from anyone, while one of the groups that studied Or genes in Drosophila got some emails saying they jumped the gun.