specifically RNAseq or Q RT PCR, NOT for RNA in-situ hybdrizations.
1- Gene expression in discs is really dynamic during the last 24 hours of larval development, so it pays to standardize by staging or timing to make your results more easily comparable. We often see consistent changes among our selected lineages in developmental timing. Onset of wandering is not a bad timing event (a bit variable, but easy). There are a few others that can also be used.
2 - Dissecting discs directly in RNAlater can be tricky (see comments below). We tried numerous times and abandoned it. Instead we use cold PBS made in RNase free water (and dissect on ice). RNase-away treat forceps before hand. Obviously the larvae have plenty of their own RNase, so we typically transfer discs in groups of 5-8 to the tube with RNAlater (typically 250-500 microL), with as small amount of PBS as possible (less than 10 microL). These tubes are chilled in an ice bucket. We then repeat this (but putting all discs for a given sample into the same tube with RNAlater). If your folks have a hard time and need to transfer more PBS, increase RNAlater (but not too much, see the protocol). Generally I try to get discs into RNAlater in less than 5 minutes (from live larvae), so depending on how fast your people are, they may need to do smaller numbers of discs to start. Remember to do this all on ice to slow all of the biochemistry down (i.e. to minimize changes in the gene expression profile due to the dissection itself).
3- The larvae tissues tend to "stick" on the inside and outside of standard yellow pipette tips. As such for each sample we always prepare/condition a pipette tip before using it to move the precious disks. We do this by taking the pool of larval carcasses from which the discs have been dissected, and pipette some of the material (never large pieces) up and down 4-5 times. We then rinse this same tip (moving liquid in and out 8-10 times) using just PBS, so that the only things that are stuck to the tip are really stuck. There should be no obvious physical tissue (i.e. under the scope the tip should look like an empty tip). Then this tip is conditioned for transfering discs and they will not stick to the inside or outside of the tip.
4 - Probably obvious, but if you see obvious damage to discs, don't use them. The stress response is quite substantial.
5 - We store discs in RNAlater at -20C. for large tissues most labs will keep them at 4C overnight and then transfer to -20C, but discs are so small that the RNAlater penetrates almost immediately in our experience. However, you could do the O.N. at 4C and then to -20C for long term storage.
6 - Comments for dissecting in RNAlater (provided by Baldur Kristjánsson from the Palsson lab at the U. Iceland). For dissecting in RNAlater, you can do an initial partial dissection of the larvae (i.e. just inverting larvae, or opening up pupal cases), let them absorb the RNAlater (all at 4C) for a few hours then store at -20C prior to doing the fine dissection. Dissecting in RNAlater is similar to dissecting in PBS e.g. you keep it on ice, however you do not have to worry as much about timing as you do with PBS. RNAlater feels kind of viscous when you are moving the larva in it but they also tend to float so it can be quite hard to manage them on the plate. This also applies to the discs when you have isolated them, they to float around and its easy to loose them. RNAlater changes larval tissue and they become sort of brittle and it is easier to collect the wing discs, remember to remove the trachea (also applies for PBS dissections). However, the wing discs can also fall off easier and get lost in the eppendorf. It is fairly easy to isolate wing discs from inverted larvae but if you intent to collect other tissues as well like brain or other imaginal discs it become trickier. Brains get very deformed in RNAlater and difficult to disentangle from the rests other imaginal discs are harder to find than in "live" larvae but maybe that is my lack of experience with other discs.
Another VERY import thing when dissecting in RNAlater is to clean your forceps/scissors every time they dry out. This is because when they dry the RNAlater crystallizes on them and this crystal float and expand ones they hit the RNAlater solution again, this can quickly cloud your plate making it impossible to see anything on it. Cleaning it with 70% ethanol is fine. Same applies for the tips, it is best to change tips between wing disc collections, this also prevents the discs from getting stuck on crystal formations within the pipette.
7 - After extraction and purification of the RNA, let the core facility who is doing the QC (probably via an agilent bioanalyzer) know that this is insect RNA, as some of the "standards" that they use to evaluate and produce RNA integrity Number (RIN) don't work well with insect RNA. The basics (quantity and purity) still are fine, just not extent of RNA degradation (although if you look at the traces a badly degraded sample is obvious). See this link (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3016993/). We use a computational approach after sequencing to generate RIN like numbers to include as covariates in the model.If you happen to do bioanalyzer by yourself then you can check the "hidden break" which causes insect RNA to look degraded because the 28S RNA breaks down and falls in the same peak as 18s RNA. If it is really degraded it will be obvious and look like any other degraded RNA.