I have most often used a PCR purification kit from Life Technologies for this application.  However, I have also used ethanol precipitation successfully for short DNA strands such as yours.  Our protocol is below, best of luck to you.

1.      Add 1/10 volume of 3 M NaOAC

2.      Add 1 volume isopropanol

3.      Vortex, spin at max speed for 15 min at room temperature

4.      Decant alcohol, add 500µl 70% ethanol, vortex

5.      Spin at max speed for 5 min at room temperature

6.      Decant alcohol, remove trace ethanol with a quick spin

7.      Dry pellet briefly and re-suspend the pellet in 25µl TE

The max speed on our table-top centrifuge is 21,130 rcf.

I have most often used a PCR purification kit from Life Technologies for this application.  However, I have also used ethanol precipitation successfully for short DNA strands such as yours.  Our protocol is below, best of luck to you.

1.      Add 1/10 volume of 3 M NaOAC

2.      Add 1 volume isopropanol

3.      Vortex, spin at max speed for 15 min at room temperature

4.      Decant alcohol, add 500µl 70% ethanol, vortex

5.      Spin at max speed for 5 min at room temperature

6.      Decant alcohol, remove trace ethanol with a quick spin

7.      Dry pellet briefly and re-suspend the pellet in 25µl TE

The max speed on our table-top centrifuge is 21,130 rcf.

Step 2: if you use ethanol instead of isopropanol, you need to add 2.5 volumes.

Hi there

yes. I too have routinely used Ambers standard protocol

Regarding use of ethanol I woukd also say that unlike isopropanol incubation at -20C will improve recovery and in particular incubation at -20C Over night will definitely facilitate full recovery of DNA. I would add 2 more points

1. For both isopropanol precipitation and ethanol precipitation addition of 1ul of 20mg/ml glycogen as a carrier will significantly improve recovery ( to the extent that with ethanol precipitation for 1 hour at -20C Plus glycogen is just as sufficient as over night incubation at -20C  + ethanol in the absence of carrier). It also gives you the advantage of a more opaque conspicuous pellet before and especially after air drying which facilitates washing with 70  % Ethanol after both isopropanol or ethanol precipitation) 

2. For fragments of 100bp or less which tend to be much less efficiently precipitated even with but especially without carrier addition of 0.1mM MgCl2 in addition to 3M NaAc pH 4-5 will greatly improve your situation. In the past for example I would routinely precipitate dye terminator labelled DNA fragments using this technique in order to be able to read sequence all the way back to the 5' primer. Invariably I could. In contrast even with isopropanol ppt or ethanol plus carrier short DNA labelled fragments proximal to the primer and less than 20bp were nearly always missing. This implies from my experience that very short DNA fragments are much more efficiently ppt by adding mgcl2 to the sodium acetate plus ethanol or isopropanol.

http://download.bioon.com/view/upload/month_0902/20090219_9a3666a7d331d8ca9f8144ZMtaLSlb0w.attach.pdf

Thank you very much for such comprehensive answers, I will try with MgCl2 and a carrier and leave it overnight as I have small fragments at low concentrations!

A follow up question, these products then have to be sent for NGS how do I remove the glycogen carrier, or will it be washed out during the washing stage? I am concerned with it interfering during sequencing.

Hi Camilla

a carrier by definition is inert. It doesn't interfere. Years ago when participating in a transcription if project I routinely ppt with glycogen for subsequent big dye fluorescent sequencing and everything was always fine. In any case as you suggest any residual glycogen is probably washed away by 70% ethanol. Moreover as glycogen makes the pellet opaque and therefore very visible I would wash by vortex img in order to detach the pellet. Soon for 5 min to repellent remove most of the ethanol with a tip. Spin for a further min to pellet residual ethanol on sides of the tube. Finally remove this with s 10ul tip and air dry pellet for about 15min. The pellet will then go completely clear when air dry but you will still be able to see when you hold your tube up to the light. Also remember that you add just 1ul at 20mg/ml although I seem to recall that literature stated that you only require a final concentration of 1ug/ml if you are worried ( although as stated I used 1ul of 20mg/ml stock irrespective of final vol

with no adverse effect).  Also remember to use molecular grade glycogen I.e certified as RNAse and DNAse free.  Roche is an excellent source 

Hi Leila

Certainly the addition of 0.1mM mgcl2 to the sodium acetate p!us ethanol will definitely precipitate the 90bp PCR product and also single stranded DNA.  Indeed I have single stranded DNA sequencing products smaller than 10bp that have precipitated. 

So yes go ahead and follow my method as stated in my previous answer

Hi Laurence

I am confused about the MgCl2 concentration mentioned in your protocol because PCR reaction itself contain 1.5 mM MgCl2. Then how the addition of 0.1 mM more improve the small DNA fragment precipitation? 

ERRATUM: Dominic

sincere apologies. I have just looked at my old short fragment prec lab procedure and in fact you add 0.01M to 0.1M MgCl(2) final conc. NOT 0.1mM MgCl(2) as I suggested

Obviously the 1.5mM in your PCR reaction is optimised for Taq so you must supplement during the precipitation of your short fragments

Once again my mistake. Thanks for bringing that to my attention !!

You are welcome Laurence! Thank you for the correction. 

Hi Laurence, we are working with aptamers and most of the time we deal with short (70-100bp) DNA and RNAs. I am very interested in your protocol. Have you ever measured the % recovery for those short DNA/RNAs? Also, when you mention about the 0.01M to 0.1M MgCl2, do you mean the final concentration (after mixing with sample) or it is add to 3M NaAc, so in the final ( with 1/10 volume added to sample before adding Ethanol) will be 10-1mM MgCl2? Thanks very much!