Technical problems related to creation of useful mosquito genetic control strains are a bit of a moving target. For example, early efforts to develop markers for transgenesis shifted quickly from the Drosophila standby, eye-color genes, to the fluorescent proteins as their specificity and ubiquitous application eclipsed the value of a lot of previous work. At this point in time, it’s difficult to expect anything better, but like all innovation, it often arrives unanticipated and will likely be replaced by something that we didn’t see coming. Such changes make the relative importance of research efforts dynamic.
None-the-less, basic problems must be solved, regardless of whether the approach is transgenic or conventional. For GM and classical SIT mosquitoes, here are the ones I consider essential.
One has to be able to identify the released insect and often its progeny
Not only must it be identifiable, but the mark must be durable and simple to distinguish. Why? When one is surveying target populations, your mosquito must be distinguished from wild type. How else can you know if it is present and its abundance? Current options are visible marker, fluorescent marker, DNA marker and other more exotic approaches like antigens. Sure, one can do PCR and DNA is indelible – at least for a while - but the economy and directness of visual identification is hard to beat in the field. Knowing what is out there gives the program, and regulators, a lot of reassurance that they know what is happening and where your mosquito has gone. For some situations where emergency control measures will be triggered if wild-type mosquitoes are found to be present (exclusion zones), a faint marker can mean unnecessary control expenses and consternation initiated simply by a poorly marked individual.
One has to be able to efficiently produce them
Small releases can be performed with manual methods, but these are out-grown quickly if the method requires large numbers to be effective over the area expected to be treated. This means that any strains based on elaborate or labor-intensive crossing schemes will be - can I say “self-limiting” - small efforts where cost is not a major consideration such as proof-of-principle. In theory, the more effective the releases are to accomplish the desired effect, the less important this is. Same comment if the strain has excessive semi-sterility, though this can be acceptable if it’s sufficiently stable.
It must have a predictable genotype and phenotype
Every early paper on transgenic mosquitoes had the word “stable” in the title. (While I argued that the evidence was inadequate, most probably were.) Even if it’s a classical chromosome rearrangement strain for SIT, the strain can’t have an inconsistent phenotype that sometimes does not have the desired effect. If it doesn’t have sufficient stability to allow amplification for two or three generations from a purified starter strain, kill it. Put it out of it’s suffering.
You must be able to eliminate females
Have we beat this horse to death? Apparently not given that there are so few ways of doing it. Hopefully next-generation sequencing will identify some methods for killing off females or converting them to males as has been done in medfly, but when you consider the options for how this can be done now, classical male-sex-determining locus pseudo-linkage doesn’t look so bad. One can do individual selection based on sex-specific fluorescence in some cases, but similar methods had a very limited life in medfly (with arguably better markers and sorters) so it’s hard to see that this will take us far down the road.
So there you go. Might be obvious to some of you, but without these characteristics, IMHO, genetic strains are not ready for prime time. I hope you’ll feel free to argue with me in the comments. It’s an important way to refine our thoughts - and our mosquitoes.