DNA Extraction – CTAB (Dan E.)

This is a protocol for extracting good yields of high quality DNA from sunflower leaf tissue.

The appealing thing about this protocol is the high yield of high quality DNA that it can produce from an easily achievable amount of starting material. The downside is that it is labour intensive and involves dangerous chemicals. See below for elaboration.

Why you might want to use this protocol – or not

If you need a lot of high quality DNA, or simply a lot of DNA, this protocol might be of interest. There are assays that require relatively large amounts of DNA Рmost NGS assays, including whole genome shotgun sequencing (WGS sequencing) and Genotyping By Sequencing (GBS or RAD Sequencing), require, at the first step, amounts of DNA  in the order of micrograms. This protocol is ideal for such applications.

By comparison with popular column-based DNA extraction methods this protocol is slow, labour intensive and dangerous (there are toxic chemicals involved). Column-based DNA extraction methods are ideal for quickly getting DNA from leaf tissue for the purposes of PCR-based assays – i.e. assays where you need a little bit of not too terribly impure DNA. They are quick and don’t involve any dangerous materials.

If you are doing some PCR-based assays, like microsats or Sanger sequencing of a few loci, you may well be better served going for a column-based DNA extraction. There is a standard column-based DNA extraction protocol in the lab (based on the very popular Qiagen kit).

If you want lots of DNA – much more than is required for a few PCRs – or high quality DNA – for restriction enzyme based assays (GBS/RAD sequencing) for example – then the column-based methods are not an option and this protocol may be suitable.


This protocol is a modified version of one we inherited from collaborators at the Burke lab at UGA in 2010. That protocol was authored by someone called Shunxue Tang from the Steve Knapp lab. You can find the original here.

I (Dan E.) modified the protocol slightly – I changed the NaCl concentration in the Extraction buffer and I switched the spooling step to a spin.

DNA Yeild and Quality

This protocol yields highly variable but generally good yields of DNA from moderate amounts of starting leaf tissue. There is however, a failure rate, (which I can’t explain – I haven’t done the experiments to figure out what might be behind total failure).

As of October 2011 I have used this protocol to extract DNA for several projects – for a Burke lab project for which I have had no feedback, although I have reason to believe that I would have heard about it had the DNA been a problem, for Rose’s RAD sequencing project, which produced the first RAD sequences in the lab, via Floragenex, just fine, and for Greg Baute’s and my own projects, which are not going smoothly but not due to known DNA factors.

This is a photo of an agarose gel where I’ve run 2ul (of 200ul) of DNA samples extracted using this protocol alongside some concentration standards. You can see good yields of high molecular weight sunflower DNA.

This pdf doc shows the yield and spectrophotometric (Nanodrop) quality stats for the batch of DNA samples I extracted for Rose in 2010. Note that I think I probably deleted some of the failures from these stats, so these figs under-represent the failure rate. But, also note, these samples include the 120 odd samples that were sent to Florogenex for RAD sequencing (successfully).

This pdf doc shows the spectrophotometric (Nanodrop) stats for the batch of 327 DNA extractions that Greg Baute and I did in August and September 2011. Note that the frequency distributions shown here include the all DNA extractions we did – including all failures. On that note: if you consider DNA extractions yielding less than 2ug to be failures (most of these really are total failures – hardly any DNA), then our failure rate was 5%. If you consider those extractions yielding less than 10ug to be failures then we had a 10% failure rate. These failure rates aren’t bad but you really do need to be aware of it and, if at all possible, you should have back up tissue to do repeat extractions from.

Things that I am (almost) certain affect yield and quality:

  • leaf tissue – better to start with young soft leaves.
  • amount of tissue – more is NOT better. It is easy to go beyond the capacity of the small volumes of buffers and solvents and get very poor yields or very poor quality DNA. The ideal amount is around 75mg. I aim for between 50mg and 85mg.
  • preservation of tissue – tissue should be frozen as soon as possible after collection or dried. See separate post about collecting tissue for DNA.
  • thawing – once the tissue has been frozen it needs to stay frozen until it hits the Extraction buffer. Not taking this seriously is an easy mistake to make – do whatever it takes to keep the tissue frozen.
  • grinding – the quality of the grinding is critical. The tissue must be ground to a homogeneous very fine powder.
  • aqueous layer only – at the step where you decant the aqueous layer off the chloroform layer you need to be slow and careful and watch what you are pipetting very closely.

Some basic tips

  1. Standardize your tissue collection. This will stop you from over doing it and minimizes one obvious source of sample to sample variation. The gold standard here is to weigh every sample. This is difficult if your collections are frozen but its not impossible (I’ve done it). The next best thing, and what I recommend if you are freezing your tissue immediately, is to use a hole punch to collect a standard area of tissue (the lab has a couple of sets of cork punches – red or yellow rubber handled sharpened metal tubes). For Greg B. and my projects we collected three #4 hole punches from nice soft young leaves and froze that immediately in the 2ml tubes we use for grinding (with the grinding beads).
  2. Do everything in small batches. At any step where you need to handle tubes but keep them frozen, do it in small batches – really small, do two tubes at a time if that is what it takes. I do extractions with this protocol in batches of 48 but I do most of the steps in batches of 12 and I do the frozen tissue handling in batches of two or four.
  3. Always handle the tubes in numerical or alphabetic or some sort of order. There are several steps where your labeling can be dissolved and if you don’t know what order the tubes were in you may as well throw them out. Also, if you know the order you handled tubes in you can sometimes track back an error or a quality problem.
  4. Have back up tissue.
  5. Keep an eye on spills and leaks. The tubes we use are not perfect and there are several steps, when you open tubes of frozen ground tissue for example, where it is very easy to get some of the contents of a tube on your fingers. This is an obvious contamination risk. Have a paper towel on hand so you can wipe your fingers between tubes.
  6. Clean up after yourself.


Liquid nitrogen is dangerous for obvious reasons. Beta-mercaptoethanol is toxic by contact, ingestion and inhalation. Chloroform is toxic too. If you are not familiar with these chemicals do some Googling and, at least, have a look at their MSDSs. There is no reason to be anxious and there is no reason this protocol can’t be done very safely. Wear gloves and long sleeves and, most importantly, eye protection when handling liquid N, especially if you are handling the big dewars. Wear gloves and work in the fume hood for the steps involving the beta-mercapto and chloroform.


If you publish somewhere where you might be expected to give a citation for your DNA extraction method I’m not sure what you should do. Somebody should look for publications out of the Knapp and Burke labs in recent years and see what they do.

Citations for DNA extraction methods for plant tissue have been a real annoyance for years because everybody does some version of a method developed by the Doyles that was never actually published anywhere that is remotely accessible. People tend to say they used “a modified CTAB method” then cite a Doyle and Doyle paper that is probably not even the right one.

[Note: See comment below! I am adding a modified version of this protocol, as I use it.–heather]


Dan Ebert, Oct 2011.

2 thoughts on “DNA Extraction – CTAB (Dan E.)

  1. Please note that there was an error in the protocol as originally uploaded to RLR.

    I have fixed it but anybody who downloaded the protocol before Nov 24 2011 should be aware that the volume of RNase I added per ml of 1M NaCl is 5ul not 50ul.

    Dan E.

  2. I have created a version of this protocol with my comments, since I do things a little differently–not necessarily better, but it might be worth checking out if you are having problems. The major differences are that I use a “full-salt” (1.4M NaCl) lysis buffer and an increased volume of CTAB precipitation buffer downstream. Minor differences: I add a potassium acetate precipitation step after the lysis; I like to use the heatblock instead of the water bath; I like a number of my centrifugations to be cold.

Comments are closed.