what exactly do u mean arbitary PCR, random primers? or degenrate?

as long you have you have pcr products you can always ligate in a suitable vector followed by transformation and colony picking and then sanger sequencing.

It would be good to share your experimental design, as it might help offer a much tailored solutions.

Cheers

Thanks for the advice. The experimental desig is as follows: We have mutagenized a bacterial strain with a mini-transposon to obtain conditional lethal (antibiotic dependent) mutants with transposon insertions upstream of essential genes. In the presence of the antibioticum the essential genes are over-expressed. The mutated cells were pooled together and exposed to a stress. The population of surviving cells should call carry the transposon insertions next to genes which confer resistance to the stress. To isolate the genes of interested two subsequent PCR reactions were performed. First with a transposon specific primer and an arbitrary primer, then the product was purified (with a Qiagen kit) and another PCR was performed with a transposon specific primer and a shorter version of the arbitrary primer. Normally (if the mutants were not pooled together) I could just send it for Sanger sequencing.

Alternatively you can send your pooled sequences to a Next-Gen sequencing service (e.g. GATC), which can provide you with the entire list of reads. I believe this is currently how most researchers track mutants in transposon libraries.

Usually you can localize mobile element insertions by inverse PCR.

http://en.wikipedia.org/wiki/Inverse_polymerase_chain_reaction

In Brief: you cut the genome into few kb pieces by restriction enzime , then you circularize fragments and pcr it.

Hope that helps.

Cheers,

Stepan

Inverse PCR will not be compatible with Sanger sequencing from a pool of insertion mutants, I think.

As Jay said, you could simply subclone the PCR products and sequence individual clones. However this is going to be a lot of work and more importantly, you might not have the coverage you really want if your insertion library is complex.

I agree with Christopher that a NGS approach will probably be the best bet. However you will want to think a bit about how to prepare your library for NGS since the reads are usually quite short. You may want to modify your PCR amplification protocol to amplify from the very 3' end of the transposon, and use primers that contain the desired sequencing adapters. My understanding of NGS is only superficial so you'll want to discuss this with someone who really understands NGS library preparation.

Also since the reads from NGS are usually short (depending on the platform, can be as short as 30-50 bp) you will probably need the whole genome sequence of your bacterial strain in order to be able to precisely map your insertion sites.

My view is that NGS might by too expensive. And usually it's hard to get boss approval for that if you do not have a big gun (NGS sequencer) just in your lab ;) . General approach to inverse PCR is that after PCR itself you cut all of the bands you see on the gel after reaction and then clone it into any vector you have (pBluscript or pGEM-T-Easy or so) and then sequence individual plasmids. Anna, how many different insertions do you expect to have?

fair enough--although NGS costs are falling quickly as many institutions are adding NGS capabilities at their sequencing core facilities. The key issue in determining whether to go with isolating single clones vs NGS is how complex you expect your library to be. If you really have no idea, you can subclone and sequence 20-50 clones to give you at least a rough sense of your library complexity.

I sincerely thank you all for your advice. Stepan, I am not sure how many insertions to expect but I am guessing that it will be less than 15. Regarding the cost of NGS I am hoping that there are other researchers at my lab that want to send samples for sequencing so that we can get a discount.

Anna, if you plan to get the sequencing data and want to call the presence/absence of specific Transposable element (TE) insertions in you pooled sample, there are now computational tools to do it! I designed one tool called T-lex (http://petrov.stanford.edu/cgi-bin/Tlex.html) to detect the presence or/and absence of known TE sequences in ngs data sequences as single or pool. The approach is based on the mapping of reads at the flanking regions of the known TEs. I validated it using PCR data and got accurate results. Do not hesitate if you have questions!

Anna,

Inverse PCR is the way to go. You may need to learn the idea and adapt your thinking a bit. Inverse PCR does not use ramdom primers, therefore the PCR specificity and efficiency are much higher. Combine the inverse PCR and NGS you will get unexpectedly better results than you think you want.

If you care let me explain briefly. First, you don't know how many desired insertions there may be in your sample. What you estimated is probably based on the number of potential target genes in the genome. But how do you know there are (could be) only one insertions for each potential gene? There could be any number of insertions for each potential gene, as long as they are shifted a few bases up or down. Tn10 transposition is TA biased, but still you could have many insertions for each potential gene if your insertion / transposition experiment is well done. You will have hundreds, even thousands insertion mutants in total. Now you see, this is a perfect NGS job. In the end you may have a collection that allows you to pick and choose one or a few most effective insertions for a particular target gene. What a luxury option!

If you only want to know some of the insertions, all as what you did before, except the last step, you TOPO TA cloning your nested PCR products and send the colonies for sequencing. Alternatively, if you want to get larger data set about the insertions, you can use NGS. In your nested PCR primers, include NGS platform primer and adaptor sequences.