Monday, March 14, 2016

A Mosquito ExAC?

Okay, there's a scheme for a crazy big genomics project has bitten me, infecting my brain.  It's definitely not something I'm in a position at all to execute on, but I throw it out as an idea in case anyone finds it useful.  And admittedly, it is pretty much stealing straight from the ExAC human exome aggregation project, which contains huge numbers of human exomes.  Behind all those is a lot of phenotype data.  Now, inspired by recently re-reading Laurie Garrett's The Coming Plague and also faced with daily news items on the Zika virus epidemic, I've had this question: what if the same approach were applied to key disease vectors?
Powered by huge databases of exomes, various groups (in particular Daniel MacArthur's at the Broad Institute) have been finding all sorts of amazing population genetics information for humans -- genes we thought were critical that perhaps can be lived without, alleles which may provide beneficial phenotypes, genes that appear intolerant of non-synonymous substitutions and so much more.  Search PubMed for "ExAC" and you'll already pull in about 3 dozen citations, and the list seems to be growing on a steady pace.  This treasure trove of human biology, revealing on a grand scale how human evolution and the genome have interacted.

Certain mosquitoes, such as from the genus Aedes and Anopheles, carry a witches' brew of nasty diseases which cause great misery and death across the world, particularly in tropical regions. Zika is in a long line of such diseases.  Zika is unusual among such diseases in threatening large swaths of the U.S..  It should be remembered that once "tropical diseases" were also endemic in the U.S. -- much of the southeast was a hotbed of malaria, and a yellow fever epidemic decimated Philadelphia (I grew up near there!) in 1793.

Now, for this project the phenotype of interest is the collection of diseases the individual mosquito is bearing.   So, this would be the general idea. Design the necessary capture reagents for each species of mosquito, with capture probes also to target the DNA-based disease organisms.  For RNA viruses (such as Zika and yellow fever), a separate reverse transcription would be needed, followed either by capture  or simply scoring by PCR.  Then sequence the captured DNA and identify polymorphisms. It would be essential to collect large numbers of mosquitoes from a diverse number of geographical regions. 

From this, a huge database of mosquito exomes with vector status would be assembled.  The hope would be to identify alleles that correlate (positively or negatively) with specific diseases being carried (or not carried).  The existence of such a database would also be a valuable resource for engaging and training researchers around the world, particularly in developing countries in which these diseases greatly tax the health systems.

I found one large mosquito exome study in PubMed, which was looking at resistance to carrying a nematode disease agent (Brugia malayi -- is there no end to the awful payloads for mosquitoes?).  And this study indeed found clues to interactions that may determine whether a mosquito will carray or kill Brugia. So something much like that -- but far more grandiose and with the potential to inform future disease or mosquito eradication efforts.


Daniel Neafsey said...

Great idea! See for an example of an initiative like this.

One challenge to finding alleles associated with disease vectorial capacity in wild caught mosquitoes is phenotyping, where the absence of a pathogen or parasite does not necessarily imply lack of vectorial capacity. For example, even in areas hyperendemic for malaria, perhaps 1 in 100 female mosquitoes will be carrying malaria parasites, even if the vast majority of the mosquito population is susceptible to parasite infection. Population studies combined with lab-based infection phenotypes and linkage studies may be a good way forward.

Keith Robison said...


Thanks for the insight! The question along those lines only just dawned on me this morning. The question would be to what degree is that simple transmission dynamics, and what degree genetics - and how many individuals would need to be included to determine that or see associations in spite of transmission dynamics noise

Jonathan Jacobs said...

Excellent idea Keith. We're setting up to do Chip/Seq and some epigenetic-seq studies on a number of wt and mutant mosq ... cell lines... soon for various alphaviruses -- but not mosq populations. Cell lines first to establish methods (or so they say...) Exome studies on populations would be super cool.