Dear evoldir members, Thank you all for your suggestions on how to solve my problem. Most people so far have advised me to design internal primers to amplify fragments of 150 to 300 bp, and then assemble them to get the full gene. I also got some references to columns that allow one to remove smaller (< 200 bp or so) fragments from DNA extracts. Below i post the replies i got so far (in order of arrival). I think i got them all, but if i forgot some answer please let me know. Also I have ommited the email addresses (I think there is such a thing as email harvesters out there in the internet) so if you need to reply to someone, let me know and I will send you the emails. For now it seems like my extracts contain some inhibiting substance, but that it is possible to remove it using the Qiaquick spin columns. I did a couple of "spiking" PCRs. This is when you add to a PCR two different DNA extracts, one from one sample that you know that works and another from a sample that is not working. In the first such PCR, the amplification did not occur, meaning that the extract that doesn't work probably contains inhibitors. After running these extracts through Qiaquick columns I did a new "spiking" PCR, and now it works, which probably means that I no longer have inhibitors in my non-working extract. I also tried a PCR on the extracts that I run through Qiaquick, but no amplification occurred. Thus it now seems that the problem is fragmentation of my DNA and I will try to design primers to amplify shorter fragments. Once again thank you all for your help Best wishes to all of you, cheers Bruno --------- Hi Bruno, You may not be able to see your fragments on a gel (I suppose you're talking about 2-3% low-melting agarose gel), if the DNA is degraded. It is easier to design primers for shorter target fragments (100-400bp each), than to try to fish out longer fragments, if any are still there - again, not very likely, if the DNA is degraded. Once you amplified the short amplicons, build a contig and you will be able to assemble your target sequence of 900bp. Cheers, sergios --------- Sergios-Orestis Kolokotronis --------- Well, your problem looks rather strange. To have a good quality of extraction of your samples you must first be sure that the hole ethanol has evaporated before beginning. What I usually do is to cut the samples in little pieces, absorbe the ethanol on a paper, and after dry them in an autoclave (the time depend on your samples) and after it works fine. You also have to increase the quantity of proteinase to digest (2-3) times, and wait that the digestion is complete (shake gently in the autoclave). After you should maybe cheek if you have pcr inhibitors in your extract. Use the DNA and Primers of someone for how it works and add in the pcr the same amount of your sample. Make a positive control. After, try also to be sure that your primers works, what I suppose it's more the problem. Maybe you can try to amplify another fragment that is more conserved just to cheek if you have your mtDNA. Short fragment of dna to my knowledge don't inhibate the pcr, it's more enzymes and ions. You can also try in diluting your dna (sometimes it solve the problems of inhibitors). If you can, try to quantify the DNA you have in your extract. --------- 500bp may still be too long--I had good luck with 200-300bp in the control region with degraded DNA. Good luck! -- Rachel E. Simmons --------- Hi, Bruno, I guess, your target size is still too big...from my experience, up to 300 bp usually work fine in older samples. If your could design internal primers (from fresh sample sequences), it might work. Matthias ---------- Hi Bruno Several things. Firstly I would be surprised if the number of small frags alone was the problem, but here are some things you can try. 1) Put the extract through a silica based PCR column, like Qiagen's Qiaquick. Stuff under 70bp (which as DNA degrades in a weird exponential like fashion will probably be the majority of fragments) will wash through, leaving you the larger fragments at the end. 2) If you wan't to try and get rid of stuff from a larger cutoff, you can try Biorad's PCR Kleen Spin columns. They work on a different techology and remove stuff <200bp. The problem with the above is that you lost some of your large DNA too. Maybe you will get 80% back of what you put in. But it may help. However, I am not sure that inhibition from DNA frags is your problem. So, some more things. Maybe your size of assay is too big. DNA really degrades fast. I would suggest that you start aiming for something small, like a 100bp fragment of the D-loop. If it works then you can increase the size. There is pretty good evidence that as you reduce size of fragment, you get an exponential increase in numbers of template molecules. So small is good. Also, do you know that your extract actually has any amplifiable DNA in it? If you did not remove all the ethanol before extracting the tissue, maybe the extraction failed and you have no DNA. Also, I would suggest that you actually test your inhibition idea. Its simple to do. Take any extract you have (from anything) that you have done successful PCR on before (using any primers). Then set up the pcr again (same extract, same primers). It should work obviously. But, once set up, add 1ul (well, as much as you use in your PCRs that you have been trying before that are failing) of one of your new extracts to the reaction. If the reaction fails, you know the added DNA must be causing inhibition. This is known as spiking a PCR, and is a pretty useful technique. Lastly, what enzyme are you using for the PCR, and how many pcr cycles are you using. Ancient DNA people, who face similar issues, often use somewhat expensive enzymes that are more sensitive than others. Such as Amplitaq Gold, or Invitrogen's Platinum Taq or Platinum Taq Hifidelity. And we tend to use pretty high concentration primers...say 1uM final concentration of each. Anyway, if you have questions about the above, or want to talk more I am happy to help. I have done a lot of this kind of work so have a fair amount of ideas and experience, yours Tom Gilbert Hi Bruno Nice to hear from you so fast. Well... Thank you very much for your suggestions. Regarding your questions: i am sure there is DNA in the extracts, because they were analysed by espectofotometric (not sure this is the name in english). Ok. Although if the ratios are too low then maybe what you are measuring is not just DNA. I have tried some different extraction techniques (Quiagen easyDNA kit, CTAB, phenol-chloroform, salt-ethanol, and i even sent 3 samples to a company, BaseClear, for extraction) but they all give the same results, there is DNA in the extract but it is not possible to amplify it. Ok, well it sounds like inhibition or fragment size is the problem then. Also the concentration of DNA, although low, is not too low and it should be enough for PCR. I am using the red Taq polymerase from Sigma, and the primer concentrations i use is 2 µM. This is the concentration of the woking solution of primer, and because i use 2.5 µL of each primer in a total of 25 µL of PCR reaction, the final concentration should be 0.2 µM. As for the PCR conditions, i use the following: 94º ' 3 min; than 35 cycles of 94º 1 m, 50º 1 m, and 72º 1m30s; the final elongation step is at 72 º for 10 minutes. I have tried with 40 cycles, and also ive used different temperatures for the annealing (up to 55 º) but nothing seems to help. Ok, well I doubt it will make a difference, but I would suggest for now always using 40 cycles, and boost the primer concentration up to 0.5 µM. The only reason why you would want to drop it again is if you get lots of dimer problems. The reason I suggest using a lot is this...if you have not so much template, then the more primer you add, the more chance that primer will find the template (it is a random process after all...they have to bump into each other in the mixture). So the more chance you will have of amplification. Maybe it is also important to refer that the espectofotometric analysis shows that the ratio 260/280 is low and the ratio 260/230 is very very low. However, some of the samples that worked also have this low ratios (some of the samples that worked have lower ratios than some samples that didnt work...). Well, this really sounds to me like your fragment size is too long then, or that its inhibition. I am now going to try to "spike" my PCR, if that inhibits the PCR then i will try to get my hands on some of the columns you suggest. Good. Make sure you run a control 'unspiked' reaction in parallel. The extracts may not fully inhibit the reaction..sometimes the good DNA is so strong that it amplifies anyway, although less efficiently, so you may need to compare band brightnesses between an unspiked and spiked reaction. Anyway, let me know what you find, and I will see what I can do to help you trouble shoot! Also a quick question, how are you currently purifying your pcr products that work (what method)? Maybe you can use that on your samples so you don't have to go get new kits.... Yours Tom Hi i have already tried the 40 cycles on some PCRs, but it didnt change the outcome. I will however change my "default" to 40 cycles, maybe it can make a difference. My spiked PCR is almost done. I have used 3 samples that worked before, and i run each sample alone and with the extract from one of the samples that didnt work. i kept one negative control for each set (the negative control also contains the extract that doesnt work). I should know the results still today (will run an agarose gel in a few). How did it go? Very curious. Concerning the purification method i am using... well, when i did the espectofotometric analysis i used several samples at different points of different protocols (may sound confusing, but it isnt or well maybe it is). I was using the Amersham BioSciences GFX PCR purification kit, and i noticed in the espectofotometric results that the ratios 260/280 260/230 were becoming even lower after the purification with this kit. Ok, well supposedly this kit should only preserve fragments that are over 100bp in length, so very fragmented stuff should be lost during the purification. Maybe its not that efficient though. However, it would suggest that you are getting rid of your very small bands. I also sent some samples that i purified by ethanol precipitation. In these samples the ratios, for the same samples, were sligthy better. Thus i am now doing the following protocol: DNA extraction -> EtOH precipitation -> PCR -> EtOH precipitation -> Sequencing PCR -> Sequencing. If it sounds like a lot of work, thats because it is. Precipitation I think normally only preserves fragments over about 50bp (I think). So you should be losing the short fragments here too. The protocol i am using for EtOH precipitation: "x" µL of extract (or PCR product) "1/10 x" µL Sodium Acetate 3M (ph 5.2) "2 x" 100% cold (-20°) EtOH Incubate at -20° for 2 nigths Spin for 15 minutes at 14000 rpmm at 4 ° Remove ethanol Add 200 µL 70% cold EtOH Spin for 10 minutes at 14000 rpm at 4° Remove ethanol Dry in oven at 37° for ~ 20 minutes (or until tubes are dry) Add destiled water (the amount depends on the initial volume of extract or on "brigthness" of PCR band in agarose gel) Vortex and incubate at 37° for 5 minutes, and it is ready to use. Sounds pretty normal. If you want to reduce the volume of the purification you can switch the ethanol for 0.6-1 volume 100% isopropanol. Also you can try adding a carrier if you want - its useful if you have low amounts of DNA as it seems to help it precipitate, so you get higher recovery. Glycogen is a common carrier (invitrogen sells some special stuff that you just add 1 ug of to your precipitation). Also Ambion sells 'Glycoblue', which you add 1ul to your precipitation, and in the end you get a nice big blue, easy to see pellet. Although this may all not be necessary. Depends how much money you have! I will also boost the primer concentration in my next PCRs, that sounds like a nice idea. Hi Good results! It really looks like inhibition is the problem. However, as you suggest, you should probably try dropping the DNA amount to maybe 2.5ul of each, and even 1ul of each) just to make sure its not an artefact of the amount added. But if you used water to resuspend your DNA after the DNA precipitation this is probably not an issue. Still, worth checking. Just do it on one sample if you can\t be bothered to do all 3 again. As you can see there is something inhibiting the PCR. I have some questions about this: -Should i repeat this PCR and use only 2,5 µL of each DNA extract (working and not working)? Can it be that i have too much DNA in the PCR mixture? -If no, then how can i know what is inhibiting the PCR? How do i know it is the short fragments or something else? Now that is a difficult question I am afraid. Its often hard to know what it is, but some options are as follows 1) Use an anti inhibition additive in your PCR. One common one that often works is BSA (bovine serum albumin). You can either buy it ready made for PCR from various companies, or you can buy the powder, make it up to 10mg/ml, then add about 1ul of this to a 25ul reaction. The idea is that the BSA binds to inhibiting crap, getting them out of the way of the enzymes and primers and templates. 2) Try and repurify your problem samples using a PCR clean up column like a Qiaquick from Qiagen. However, these may lead to you losing a lot of the DNA (50% at a guess).....so try it on something not valuable first. The idea is that such columns normally have quite a wide mess silica membrane. The DNA binds to the silica. The other crap gets washed out. Then at the end you release the DNA from the silica and its now purer. 3) How sure are you dried all the ethanol off your extract before you resuspended the pellet in the precipitation? If you couldn\t see any ethanol at all, then that is good. But if enough was left it would inhibit the PCR reaction. Should i now try some separation column, or simply forget about it and go on to design primers ? (i think i should have enough good sequences to design primers, just need to get them aligned and run on some program ). Depends how much of a hurry you are in. And how much money you have. Obviously if you have inhibition designing small primers won\t help. But if you solve the inhibition you may need to do that anyway...depends on the fragmentation. Tom --------- Hi Bruno, You can try to use Microcone 100 spin columns (Millipore) to cut off DNA fragments shorter than 100-125 bp. You will also concentrate your DNA in the same time and remove some inhibitors, so there is a chance that it helps. I use it for my poor quality DNA and until some point it's great help, but if your DNA is really degraded nothing will help, I guess. Ola --------- Dear Bruno I had the same problem for amplifying a 16Smit DNA fragment of about 600 bp. I designed overlapping primers to cut the portion in 3 smaller fragments, and found that for most samples (80%) it worked very well amplifying using two smaller pieces (C:250bp and AB: 400bp) instead of the big one (ABC). Some samples were resistant but the improvement was striking. Just be careful to put more negative controls (without DNA) since smaller fragments amplify much more easily, you will get more often contaminations. I actually observed contaminations in some cases even though the no- DNA control produced no band (for a given sample: piece C was identical to an already known haplotype X, while piece AB was identical to an other already known haplotype Y). Good luck Anne --------- Hi Bruno, It seems very unlikely that the failure of your PCR:s would be due to the prescence of too many short DNA fragments. I have been working for many years with DNA from old museum samples as well as fossils, and have never had that problem. My advise to you would be to look into other sources of problems. Maybe your PCR isn't properly optimized? (a suboptimal PCR could still work on high quality DNA)... Have you tried to run an optimization using different annealing temperatures and concentrations of MgCl? The best advise I can give you, however, is to add BSA (Bovine serum albumin) to your PCR. The final concentration of BSA in your PCR should be 0.1 mg/ml. BSA sometimes works miracles in PCR's... Cheers, Love Dalén --------- Hi, Try using Microcon-100. It cuts off fragments below approx. 100bp of dsDNA so you'll get rid of the shortest fragments at least. However I don't like it as it is very risky while manipulation. You should avoid purifying too many samples at one time. Good luck Maciek --------- Bruno, Sephadex columns will remove small fragments and leave larger ones. We use this for cleanup of sequencing products, and it might help in your situation. There are different grades/types of Sephadex (G-50 removes fragments smaller than 50 bp, for example). It is made by Amersham Pharmacia and Sigma (USA) sells it as well. You can buy pre-filled columns or buy bulk Sephadex and make your own. Libby --------- BRUNO NEVADO PhD Student Vertebrate department Royal Belgian Institute of Natural Sciences Vautierstaat 29, 1000 Brussels, Belgium phone +32 2 627 42 86 (ext. 428) E-mail: bruno.nevado@naturalsciences.be Molecular Laboratory website: http://www.naturalsciences.be/institute/ structure/molelabo/vertebrates/ bruno.nevado@naturalsciences.be