Hello, all-- Several months ago I posted a question on microsatellite dropout with touchdown programs, and realized that I have yet to put the replies together and send them out to the list. Better late than never, I suppose! Thanks to all who replied. There didn't seem to be a consensus position: some people noticed dropout, some didn't, many seem to run touchdown programs with great success. No one could point me to a paper, so the phenomenon may not be published. Seems like it may vary considerably with DNA source and genotyping technology. My original question is below, followed by the replies: Initial question: Is there anything published on the frequency of allele dropout when using touchdown programs to amplify microsatellites? Years ago, I tested some microsats (for crabs, I believe) with a touchdown program and with a range of single-annealing temp programs. When I ran the PCRs out on a BaseStation, it was obvious that in a significant number of instances the touchdown program resulted in homozygotes, while the single-temp programs showed heterozygotes. It appeared that in heterozygotes, sometimes the smaller-length alleles were being preferentially amplified with the touchdown program, and the larger alleles were dropping out. Taking this as a cautionary tale, I've always been careful to run microsats with optimized single-temp programs. Other than that, I never really followed up on it. I note now that it is not uncommon to see papers that use touchdown PCR to amplify microsats, and am curious if a) there is anything published on the matter, and b) anyone else has had noticed dropout (or not!) in program comparisons. Carolyn Tepolt carolyn.tepolt@gmail.com Replies: Unfortunately I don't know of any publications that discuss this issue but thought I'd drop you a line since I have experienced the same thing myself. It's been exactly as you describe with the larger of the two alleles dropping out with a touchdown but not with its own single temp. It's a shame since it was a good way of getting lots of loci with varying Tm's going in a multiplex. I guess it makes a bit of sense though since with a touchdown, the amplification gets going right at the extreme limit of its Tm range. This would tend to favour more easily amplified fragments ie the shorter ones. It's amazing really that such a small advantage can make such a big difference but I guess since it happens early in a reaction, the advantage is exponential. As an aside, I got around this by having multiple Tm's in a reaction ie alternate cycles of two (or maybe three, I can't remember for sure) different temperatures, but with no touchdown between them. This helped a lot, presumably because at the upper temp, the other loci didn't amplify at all so the allele length advantage was removed. It was a bit of fiddling around but meant I could keep my multiplex together and not have to subsequently pool. I would give this a whirl if I found myself in this situation again. Good luck. N.Aitken niccy.aitken@canberra.edu.au ***** We didn't notice any significantly greater null-allele problem with an optimized touchdown program vs. optimized single-temp program; in fact we noticed the opposite, but some of this may be due to our study. For work described in a manuscript currently in revision, we used touchdown while amplifying loci in extracts from pericarps of Quercus acorns, for which the DNA may be somewhat degraded. We used undiluted extracts, and amplified three separate multiplex panels totalling 10 loci, using touchdown programs optimized for our species and tissue. To quantify error rates we regenotyped a number of pericarps once or twice and found that, on average, even in very good samples about 1.7% of loci would be heterozygotes on one run and homozygotes on another. For this paper, we were assigning these pericarps to source trees, so I wrote an R program to handle the potential impact of null alleles upon assignment probabilities within a likelihood context. It seems that the large majority of our null alleles were due to issues with pericarp DNA; for us the optimized touchdown programs (plus not diluting our extracts) gave better results than optimized single-temp programs. Doug Scofield dgscofield@ucla.edu ***** I almost always use touchdown, as it saves on months of optimisation. Overall, it has given larger peaks for the same samples. I have been using it for parentage, so I know when there is a null allele. Null alleles have not been a problem so far (except with 1/7 loci), despite 3 loci having large size ranges. I am also keen to know if Qiagen multiplexing kits have any influence on dropout rates. I'd imagine they should as they are designed to amplify different microsats of vastly different lengths in a single PCR. Patrick Hamilton p.b.hamilton@exeter.ac.uk ***** I have learn and use touch down prc protocol for amplification of genes allèles with unspecific primers prior to cloning and sequencing them to design specific primers for qpcr... Since this method is starting less specific TM and finished with highly stringent temperature it is logical that it amplified the ''large-allelle-drop-out'' phenomenon. Large alleles can't compete with small alleles and it result in a false deficit in heterozygosity. I do myself use microsatellites in my study on american eel. Each locus is very rich in allele, and large alleles are systematically difficult to amplified compared to the small... I visualised them with ABI3100 sequencer and systematically small alleles lock saturated and large ones are barely detectable sometimes... Caroline caroline.cote.14@ulaval.ca ***** I don't remember anything exactly like that published, but in general the pattern varies. I have seen many cases where the situation was opposite: touchdown gave heterozygotes with clear but lesser amplification of one allele that was not visible with optimized standard program. These were cases where it was quite clear that one of the binding sites was polymorphic. M.Tapio@cgiar.org ***** I had problems in the opposite direction, so now I do all my PCRs with Touchdown, I got much better and reliable results for heterozygotes than before. Suska suskaathome@gmx.de ***** We haven't done a systematic comparison across multiple loci but recently observed a similar effect for the orthologous Z & W loci used to type sex in birds: touchdown can result in one of the two loci not being sexed, and hence the heterogametic females (ZW) can appear to be male (Z_ looking like ZZ). I interpret this as the result of slight differences in the target binding sites, effectively making the annealing temperature different so that the primer will bind to the target site at one locus at a lower temperature than it does at the other. It follows, therefore, that any polymorphism in a primer binding site can result in this effect, and that nulls are therefore more expected to occur using touchdown than using a set temperature. It also follows that it should be possible to reduce the frequency of nulls by adjusting the temperature upwards, though I don't recall anyone doing this. In practice, these days we design at a single temperature, use Qiagen Q mix to increase the temperature range over which primers will work, and reject any loci that show evidence of nulls. Following this hypothesis I wouldn't expect the size of alleles to be a factor affecting dropout, though dropout in general can increase with product length - presumably due to something more generally suboptimal about the PCR. Terry Burke T.A.Burke@sheffield.ac.uk carolyn.tepolt@gmail.com