Sequencing from Large-Insert Clones
Automated sequencing can work with a variety of very large DNA clones such as BAC, Lambda or Cosmid DNA. We've even had some excellent success with direct sequencing from a bacterial genomic template. Here's how to do it:
This page is divided into two sections. The first gives information for the client to help them prepare good quality template DNA and primers. The second section gives information for other Core facilities on how to process BAC DNA for best results.
Section I - template preparation for Core Clients
Because large clones like BACs carry with them huge amounts of non-template DNA, the sequencing reaction must be adjusted to help the primer find the correct priming site efficiently. Various other parameters must also be adjusted to compensate for these changes.
- The template concentration must be very high
- Only a small percentage of the clone will be used as sequencing template; to provide an adequate *effective* level of template, the total concentration of DNA must be increased. We have had the best success with 1 ug/ul for BAC DNA, 0.5 ug/ul for Lambda and cosmid DNA. Yes, this is a *lot* of expensive DNA, but in our experience, it is quite essential. BAC samples should be visibly viscous and difficult to pipette at this concentration. Please note that Core techs may need to shear your sample in order to pipette it. Don't expect to retrieve tubes of full-length BAC after we're done.
- Primer concentration must be high as well
- The excess DNA in the reaction tends to bind primer non-specifically, so you have to increase the primer concentration to compensate. We suggest 30 pMol/ul (picomoles per microliter - 10X higher than the concentration usually used for sequencing). If you request sequencing with standard primers (T7, T3, SP6, M13F, M13R), the Core automatically uses these higher concentrations.
- Stringently avoid primer dimers
- Because the primers are used at such a high concentration, they have a greatly increased tendancy to form dimers. Please use a computer to design your primers, and make sure they have very low probability of forming dimers - particularly 3' dimers.
- Avoid genomic DNA contamination
- According to our sources, BAC preps are routinely contaminated with 20%- 80% genomic DNA. Obviously this DNA decreases the effective concentration of the template, so genomic DNA must be avoided. Be extremely gentle when you break open the cells (alkaline SDS step) and when precipitating the protein/genomic components (potassium acetate step). Restrict your DNA and examine the bands on a gel. If you see any smears in the background, there's too much genomic DNA in the prep. Smears may not look like much DNA, but there's quite a bit there.
- DNA Purity is critical!
- Yes, DNA preparation is always critical for successful automated sequencing, but with large templates, it is even more so. The best sequence has been routinely generated from BACs prepared using Qiagen protocols (Qiatip 500, for example) and with an added phenol-chloroform extraction at the end. We don't know what this removes, but it has routinely helped save "bad" BAC preps.
- Submit the sample as 'Large DNA'
- When you log the sample(s) into our computer, select 'Large DNA' as the sample type. Please do not choose any of the plasmid selections. The Core personnel have specific protocols they must follow to get good sequence from large-insert templates.
Please note that we typically keep large-insert samples in the queue until we accumulate enough to bother processing. The delay is rarely more than a couple of days, but please be patient.
Some final comments for Core clients regarding sequencing BACs and other large DNA:
Templates that sequence well will reliably do so. A good BAC preparation will produce ~700 nt of data time and time again, given good primers. However, the same client might prepare a new batch of template with the same protocol, and it may fail miserably time and time again. We do not yet know what makes a BAC prep "good".
Due to these vagaries of DNA preparation, BAC sequencing is simply not as reliable as plasmid sequencing. Our lab is quite experienced with this protocol, yet our success rate is approximately 50%. The difference between success and failure, as always, is with the client's sample. Let's face it - operating a sequencer ain't rocket science. We can't take credit for the successes, but as long as we perform our normal quality control checks, don't be too hasty to blame us for the failures either.
Section II - for Core Labs
The following is a description of the procedures used in the University of Michigan DNA Sequencing Core for obtaining sequence from large DNA clones, such as BACs, P1s, Lambda and Cosmid clones. Most of our experience is with BAC DNA, so the procedures refer to 'BAC' DNA, but they probably apply equally to the other types of large DNA. See our Table of Template and Primer Concentrations for some info on other large-template types.
To compensate for the large amount of non-template DNA present, BAC sequencing depends on alterations to the reaction recipe and to the cycling conditions that enhance the signal.
Handling the template:
It may be necessary to shear the BAC DNA in order to pipette it. This may, in fact, improve the final sequence.
Reaction Recipe, BigDye V3.1 Chemistry: (Recipe updated 15-May-07)
- 2 ul Template DNA (1 ug/ul)
- 6 ul primer (30 pMol/ul)
- 1.0 ul DMSO
- 3.8 ul Betaine
- 2.2 ul 5X reaction buffer
- 6 ul BigDye v3.1
- 2 ul dGTP v3.0
Total reaction volume: 23 ul
We used to do modified cycling, but the techs hated to do it, and figured out that most of the time, the regular ABI-recommended cycling worked fine for BACs and bacterial genomes.
Before cycling starts, we heat the samples to 95 degrees for 5 minutes. We then do 25 cycles of:
- Denature 95 degrees, 10 seconds
- Anneal 50 degrees 5 seconds
- Extend 60 degrees, 4 minutes
On more difficult templates, 50 cycles may be necessary; some people use as many as 100 cycles. We rarely go above 25 cycles, and we get excellent results with *good* BAC preps. It has even worked surprisingly well for sequencing directly on bacterial (5MB) genomic DNA!