Dear all, Many thanks to all the contributors, it really helps. Many asked for the answers. They where quite varied, so I copied below the original messages, with slight editions. Cheers, Matthieu -------- 1. Rachel I have been doing exactly this over the past year or so.... and so I have a few pointers. Hopefully they're useful. 1. I went the silver staining route initially but had a hard time extracting the band from the polyacrylamide and found the whole process rather messy and time consuming... 2. After ditching the silver staining method I began using MetaPhor agarose (I assume this is like Nusieve...), and that's worked really well. I do isolate the band after whole amplification, and re-amplify, and then run them out again. I've been able to resolve bands of sizes from 75 to 400 bp with no trouble. 3. For the final extraction I ran a 100bp ladder next to 5uL of the PCR followed by an empty well and then the rest of my sample (~20uL). After running it overnight at about 34V I then cut out the ladder and 5uL PCR areas of the gel and stained them in EtBr - this then became my reference by which I could cut out the fragment of interest from the unstained 20uL lane. I did this so that I didn't have any EtBr knicks in my fragment of interest - maybe overly cautious, but I thought it was worth it. -------- 2. Erick (my summary + translation) You could use the same fluorescent PCRs as the ones used on the automatic sequencer, but loaded into a classical vertical electrophoresis (denaturing, 0.4 cm thick), visualise the bands with a FMBIO II scanner, and cut them at this very moment. Labour intensive and boring but efficient. -------- 3. Bob I can confidently tell you that separation on acrylamide followed by silver-staining is really not all that painful. I've done it hundreds of times (many hundreds of gels!) for microsatellites (though now I use fluor-labeled primers and run them on the 3730). Forget the silver-staining kits, by the way. They're complicated and expensive. If you're interested, I can send you the silver-staining technique we used that is cheap and fast and dependable. I still use a modified silver-staining method for SSCP analysis on tiny Bio-Rad Protean-3 minigels. I used Bio-Rad's 38x30cm rig called the Sequin-Gen GT. Great for microsatellites, but not sure you'll get the separation you'll need for AFLPs, depending on how many AFLP bands you have, and where in the spread they occur. But Bio-Rad makes a variety of formats, and I found their setup for pouring, etc., to be really sweet. I did most of this work while at postdoc at Duke, and I had several undergrads trained in the procedure, from loading gels all the way to silver-staining. Lots more "fun" than agarose gels! The nice thing about the 38x30 is that it's just less gel to handle, making the staining and handling much easier. We actually did quite a few AFLP runs on the 38x30, and got good enough results that we were able to score bands - and compare these to the results we were getting from the fragment analysis on the ABI instrument (this was a few years ago, pre-3730). I doubt Nusieve is going to work - but it might be worth a try - and I know nothing about spreadex. -------- 4. Phillip My student has the attached paper currently in press, in which she describes using a Dark Reader(r) Transilluminator (Clare Chemical Research, Dolores, CO, USA) to extract AFLP bands from Spreadex gels. [note: the paper is in Journal of Heredity: AFLP Fragment Isolation Technique as a Method to Produce Random Sequences for Single Nucleotide Polymorphism Discovery in the Green Turtle, Chelonia mydas, by Roden et al 2008/9] -------- 5. David If you have some funds for equipment purchase, the combination of a Li-Cor DNA analyzer (a slab polyacrylamide gel-based sequencing/genotyping system using infrared dyes) and Li-Cor's Odyssey fluorescent imager would probably be an option to consider. (I'm assuming they still market the Odyssey system.) The Odyssey provides the opportunity to visualize the bands and locate the spot where the bands you want to excise are. -------- 6. Adam You can easily do it on just agarose. We do it regularly. 1) Run the amplification again, adding one base to one of the primer sets. There are 4 reactions, and your band will show on one of them. You may need to do this using the regular ABI separation gel. 2) Do this again with one primer added to the other base. Unless you have a high band density, you can probably do this on agarose. 3) If necessary, add another base and do it again. 4) Your band will show up sufficiently separated from the others that you can get it out with a cut-off 1-ml micropipette tip, purify it and sequence it. 5) Then if necessary, make primers to get it out of the rest of your material using regular PCR. It works best if you have a family that you have already bred from, so that you can be even more sure you are getting the right band. But if it is already a band that is consistently diagnostic within one taxon, that probably isn't necessary. -------- 7. Kathrin I am planning a similiar project to detect sex specific sequences and a collegue experienced with AFLP told me that they clone the entire PCR, pick the clones, amplify them with the selective primers, run them on a gel, and just choose the one showing the corresponding length. I never tried it, but it seems a good way to circumvent cutting, eluting, purifying et al. -------- 8. Craig I did some band extractions from AFLP gels many years ago (for the original work see J.Evol.Biol. 14 611-619). I used 33P labelled Eco primers and overlaid the gel with the film, pierced the gel around the band with pins to locate the rectangle of gel containing the fragment of interest, cut this out with a scalpel and placed it in water to soak. I then amplified from this with preselective primers and cloned and sequenced. I realise that you say that you can use radiolabeled primers and would therefore say (without having used the method) that silver staining would be best, including some appropriate size standards to enable locating your band. How much work this will be will depend on how many primer combinations you have bands for but from my experience it necessitates having quite a few spare glass plates for your electrophoresis rig as these will be out of action whilst you undertake the extraction. I would warn you not to expect every clone to contain your band of interest and to undertake some checks. I cloned, and sequenced multiple clones and often obtained very different sequences. I homed in on the 'real' band by first of all obtaining the same sequence from 2 unrelated individuals (and in the case of this study from the 2 different morphological forms we were looking at). Of course, the insert had to be the correct size and I also confirmed that the presence of the particular insert matched that seen in the original AFLP gels by running a subset of samples out on a standard agarose gel, blotting and probing with the clone's insert. The presence/absence here had to match that seen previously. I consider this part to be vital to know you have the correct insert - I think you will get inserts that are not the real AFLP band and you have to do this step for certainty. If you need to know anything else please let me know. -------- 9. Sean I am also hoping to perform similar AFLP genome scans. The AFLP band extraction procedure can be quite long and I am told that is not as straight forward as it may seem. I would prefer to use an outside company that can perform the procedure rapidly and reliably. One such company is Baseclear (http://www.baseclear.com/index.php), I have attached the order form for AFLP extraction too. At the minute I am in the very early stages of my AFLP genome scans and have not used Baseclear, but it is something you might wish to look into. -------- 10. Nicolas I faced the exact same dilemma months ago and I went for the silver staining protocol. I have no idea about Nusieve agarose, and doubt seriously of the use of Precast Spreadex for cDNA AFLP. In return, I would add another technique based on Cy5 labeled priming. It uses a Licor type system and a powerful imager (Typhoon for instance). Afterwards you need to print out a transparency sheet to locate the region containing the bands of interest by aligning it to the gel. Please see corresponding article enclosed. Turning to the silver staining method, I forwarded you my protocol. Basically, we separated fragments using electrophoresis through 6% polyacrylamide gels run at 1000V for 2h using a 200 model S2 sequencing gel electrophoresis apparatus (Gibco BRL). That's an second-hand old school apparatus that I found unused in a lab. It was rather labor intensive but works out just fine. -------- 11. Lena I tried this for a little while at the end of my phd, but then gave up - so no success story, but I think that it would have worked with more effort. I tried a Licor sequencer with an odyssey scanner. This works well for strong bands, but I often found that bands of interest I could see perfectly well in the scan from the sequencer where all but invisible on the gel scanner. My reactions had way too many bands for the spreadex gels (which otherwise are great). If I were to try it again, I would give the protocols in this paper another shot: A new and versatile method for the successful conversion of AFLP markers into simple single locus markers. Brugmans B, van der Hulst RG, Visser RG, Lindhout P, van Eck HJ. Nucleic Acids Res. 2003 May 15;31(10):e55. -------- 12. Kym Another very long-winded way of doing it would be to use selective primers with increasing numbers of bases (e.g. +AGC, +AGCT, +AGCTA, +AGCTAC) until you end up with your fragment of choice plus only a small number of other bands. This would make it easier to cut out your band on a NuSieve (or whatever) gel, especially if the fragments were widely separated. Matthieu Boulesteix Postdoctoral Research Associate Imperial College London Silwood Park Campus Buckhurst Road, Ascot, Berkshire, SL5 7PY, UK Tel +44 (0)20 7594 2306 m.boulesteix AT imperial DOT ac DOT uk m.boulesteix@imperial.ac.uk