Dear Juliana

what you describe is not uncommon:  having in the past ran a genome sequencing facility and also worked in a public consortium lab that contributed to the mouse genome sequence I was very familiar with this situation. In many cases sequencing through homopolymeric regions particularly poly A regions is the most difficult scenario of all as you face both slippage culminating in multiple out of register ladders beyond your poly A motif culminating in a jumbled unreadable mess; but also weaker sequence caused by the Taq stalling and then detaching; a so called strong stop  

My only advice as elaborated in the paper below us to try a Poly T anchored primer that spans your offending poly A motif. In addition occasionally extension at lower temp e.g 60C rather than 72C can help but unfortunately nothing is particularly robust in terms of a reliable method for the situation you describe. In addition most of these slippage/ stop phenomenon are exacerbated by template impurities so extra clean up of your DNA can sometimes contribute to a successful solution 

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2291785/

Either sequence from the reverse direction or if the poly a is well into the sequence re-amplify with forward primer nearer to the poly a and either clone the product and sequence or sequence with more sequencing mix than is normally used as some of the effect of poor polymer base sequencing is caused by nucleotide depletion in the sequencing mix as well as the steric effects. Poly A is likely to be polymorphic so if your samples have different numbers of As then all of your bases following the polya will be 2 supeimposed bases from the sequence and the sequence+/- 1 base so will all be half strength. Cloning will separate these sequences and sequencing clones should give you N as or n+1  A's  but not both which is what may be happening with your sequence. Post a .abi file of one of your sequences  please ( ideally one good and one messy one)

Thanks Paul, That's really helpful - our last resort is using a triplicate of the reverse sequence to confirm the sequences we obtain however my supervisor is keen to be able to reverse complement instead if we can manage it. When you say nearer to the poly a what would be sufficient? And thanks for the suggestion regarding the nucleotide depletion as it was a factor we'd identified but I hadn't uncovered any possible solutions yet. Unfortunately I don't have a 'good' forward sequence but I'll post forward and reverse. Many many thanks!

Many thanks Laurence, Again this is really helping me get a grip on how to approach the problem, the limitations there are as well as some ideas to start on. I'll give those ideas a go! Cheers, Juliana

hi Juliana.

I did not answer your query about Mg ions. dideoxy and normal nucleotides are poorly differentiated using Mg ions in the sequencing reaction so most sequencing kits use manganese instead of Mg as the divalent ion needed for the enzyme to work. ddNTP and ntp incorporate at similar rates with Mn so adjusting the concentration of ddNTP makes the kits more reproducible. 

Concerning your problem I might try sequencing in presence of 1M BETAINE additionally to help get through the polyA but desalting afterwards must be complete before loading the sample on a sequencer. I would be interested to see a poor sequence ( much more informative than a good one) and also in text the 6 bases either side of the poly A please

just to expand slightly on the MG issue. As Mg ions will cause more read through without termination it seems possible that adding Mg might allow reading of longer sequences past an area which terminates too easily. I attach a very old paper from when researchers were still making their own sequencing mixes which you may find interesting although something better must have been published since

Hi Paul,

Thanks for the explanations. It seems that my files might not have attached properly - here they are again... I also posted pdfs of the chromatograms. The file labelled CPV2655F at the end is the poor sequence and ending CPV3511R is the same sequence in reverse which is a good sequence.

This is the polyA with 6 bases either side:

TCTTGCAAAAAAAAAAAAAGCCGGT

This is the published sequence with the primer we're using underlined:

CCAGATCATCCATCAACATCAAGACCAACAAAACCAACTAAAAGAAGTAAACCACCACCTCATATTTTCATCAATCTTGCAAAAAAAAAAAAAGCCGGTGCAGGA

Huge thanks for your input!

Juliana

thank you for the pdfs Juliana .I will have a close look at them. One  more thing please. If ( and only if) you are using exo sap method to purify your pcr product can I have the 2 pcr primers sequence please

thank you Paul

Ok looking at the F sequence and ignoring the slightly bizarre base calling near 195 base a couple of thoughts. The base heights are good and the spacing fine so this is good sequencing. The poly A at around 55 bases is clean so there is no mixed population of amplimers in your sequencing template. Also there is no polymorphism in the poly A . I do not like the poly A superimposed on the sequence at around 85 bases. It would be good to see the sequence of the forward pcr and the forward sequencing primer and the purification method used in cleaning up your sequencing template. I am wondering if the forward primer is either partially degraded , contaminated or simply capable of annealing at 2 positions at the conditions used in the sequencing reaction. Often companies are very conservative regarding sequencing conditions and suggest , for instance annealing at 50c for the sequencing reaction even though the primer is used in a pcr and anneals at 60c in that reaction. If there is a possibility of the primer annealing in 2 places so producing 2 sequences then I would run the sequencing reaction ( forward) at the higher annealing temperature of the pcr primer in the hope that the primer will only anneal at one position and maybe extend at 64c not 60c as recommended for sequencing ( the latter not so important but it deals with some template conformation problems)

Regards Paul

Wow Paul, thank you for looking through that - I'm certainly learning a huge amount through the problems!

Previously the sequencing lab has done the purification so I'm unsure, but I recently got very similar results when I did the purification with a Qiagen kit. We did try a new batch of primer as well as an alternative primer 5bp further forward in the sequence but still got the same results. I have also tried annealing at 58'C as part of a gradient PCR but again messy sequence. I'm thinking of trying a touchdown PCR with higher initial annealing temperature and also increasing Mg2+ in another PCR. I have been extending at 72'C and just tried a PCR with a gradient for this starting at extension at 62'C up to 72'C... so maybe among these options something will improve the results!

I haven't used exosap for clean up so didn't bother you with this. So far your insights have been tremendous help, thank you!

Juliana

Dear Juliana,

All of those things are generally good for improving the quality of a product but I think that your product is already good as shown by the good sequence...if you had 2 sequences present then they would both sequence from the primers and the sequence would be a mess at some stage in both directions. I had wondered if the template was hairpin looping and a sequence generated from the double/single interface but the reverse sequence would have shown that so I still think that teh problem is caused by one primer only ( possibly annealing in 2 positions). Cloning the product and sequencing with plasmid based primers M13 or similar would provide an answer. Increasing Mg generally results in multiple or poorer pcr products as the primers can anneal in more positions on the genome. but good luck and get back if there are any aspects that you think the site can help with

Dear Paul, thanks so much for explaining again! It really helps to understand what is actually going on and how you tell rather than just guessing and I've found it really hard to nail down these details. I'm going to give your suggestions a go as well as trialling an alternative primer... so hopefully I won't need to bother you again but thanks for the offer! Best wishes, Juliana

It is really no trouble  Juliana and I am sure that both Laurence and myself will be keen to know what eventually works in this sequence. Good luck

Paul

Hi Paul & Laurence,

I just wanted to let you both know changing the extension temperature gave me good enough sequences where the reverse complement agreed with the problem forward sequence in the end! It made a big difference to my project & I learnt a lot from all your suggestions. So thank you!

Best wishes, Juliana

Hi Juliana

im glad you were able to navigate your poly A tail with assistance; particularly from Paul

Kind regards