This would seem to be an easy to answer question, but I was unable to find an answer (in g/L) for generic double-stranded DNA or plasmid neither on Google nor on BioNumbers. I would expect the solubility to vary slightly based on the sequence or length, but it probably wouldn't affect it very much due to the relative hydrophobicity of the bases.

I know I have made solutions of plasmid maxipreps of at least 1µg/µl, and the Nanodrop DNA quantifier datasheet has a maximum DNA detectable limit of 15µg/µl, but can anyone find a source for the maximum solubility level of generic DNA? For example, the solubility of a specific plasmid of approx. 5kb in length would be a good ballpark answer for this question.

  • 1
    $\begingroup$ I think that we need to dig around for maximum molar concentrations of DNA for this question. I know that you can't spec a sample above 100g/L (ug/uL), and that to test high concentration giga preps I've had to do serial dilution of the sample for accurate reads. I've consistently made giga preps that were concentrated to 10+g/L but those were in buffers, not water. $\endgroup$
    – Atl LED
    Commented May 26, 2015 at 14:34
  • $\begingroup$ @AtlLED do you have a protocol for the giga prep kit if you used one? $\endgroup$
    – March Ho
    Commented May 26, 2015 at 14:54
  • 1
    $\begingroup$ qiagen.com/us/resources/… and also mn-net.com/Portals/8/attachments/Redakteure_Bio/Protocols/… At high concentrations, the prep becomes much more like a syrup/has an increased viscosity. $\endgroup$
    – Atl LED
    Commented May 26, 2015 at 15:03
  • $\begingroup$ This is just a back-of-the-envelop thought experiment: A human cell has approximately 6 billion bases, meaning that there are app. the same number of - charged phosphate bonds in the backbones of the helices. If you were to say arbitrarily, for argument sake, that you needed 10 water molecules to solvate the negative charge of the phosphate bond, then you would only need 1.8 femtoliters of water to solvate the DNA from a single human cell. The human genome weighs about 3.6 picograms, so if this accurately represented the solubility of dsDNA then it is equivalent to 2g/mL. $\endgroup$
    – AMR
    Commented Aug 30, 2015 at 4:08
  • $\begingroup$ This makes sense, because when you think about how absolutely small a cell is and how much water its even smaller nucleus contains, even in this minuscule amount of water, DNA and associated proteins are still solvated, then you get an idea for just how much DNA can be solvated in water. We can easily see small fractions of a drop of 1 µL of water with our naked eye (as you mentioned a nanodrop machine), but we need a decently powerful microscope to see the nucleus of a human cell, and yet all of that DNA is still solvated. It will take a huge number of 5kb plasmids to saturate water. $\endgroup$
    – AMR
    Commented Aug 30, 2015 at 4:15

1 Answer 1


DNA is a bit more complicated than some molecules due to it's length and composition variability.

In water.

According to Integrated DNA technologies:

DNA oligos can be resuspended to a near maximum concentration of 10mMolar; to achieve such a high concentration will require a lot of vortexing and it may take up to a day for the oligo to go into the solution.

DNA is a polar molecule and as far as I have experienced dissolves in water very readily. However there doesn't seem to be an exact quantifiable number for DNA solubility in water. If I correctly understand it, this paper claims that a 200,000Da-8,000,000Da oligonucleotide is soluble in water, however they never quantify this. An answer on a related question pointed out that high molecular weight DNA oligomers, like genomic DNA, is enhanced in solutions with dilute monovalent cations.

Cleaver & Boyer (1971) have some data on this, however it is somewhat outdated, and they were only using this solubility to demonstrate dialysis in water was capable of desalting the oligonucleotides.

Oligonucleotides larger than 5 bases in length remained within the dialysis bag (Fig. 2) and losses of 10% and 40% occurred for tetra- and trinucleotides, respectively.

There is a lot of work done on optimal temperature and pH for solubility of DNA so perhaps a more specific query might throw up some results.

In ethanol.

Ethanol precipitates at least 10–20 % of all classes of oligonucleotides. Oligonucleotides longer than 15–16 nucleotides are precipitated to about 65%.

  • $\begingroup$ Can you explain why organic solvents are likely to be needed? It seems that DNA is less soluble in isopropanol and ethanol (which are used in DNA extraction) than water, so making the solvent less polar might appear to have the opposite effect. $\endgroup$
    – March Ho
    Commented Jun 23, 2015 at 6:25
  • $\begingroup$ Hmm... Not sure. It could be to do with the fact that recovering DNA from water could be difficult. I have opened another question for that: biology.stackexchange.com/questions/35601/… $\endgroup$
    – James
    Commented Jun 28, 2015 at 17:25
  • $\begingroup$ @243 pointed me to a site with the answer you were looking for. I've edited it in. $\endgroup$
    – James
    Commented Jun 29, 2015 at 12:52
  • $\begingroup$ The link you provide to genscript is for peptide sequences which can have different solubility properties depending on amino acid composition. Peptides sequences can potentially be far more hydrophobic than DNA. $\endgroup$ Commented Jun 29, 2015 at 12:55
  • 1
    $\begingroup$ There is a bit of a problem using the IDT number of 10mMolar to think about the solubility of DNA. First off Oligos come single-stranded, so that is a different molecule from a plasmid. Secondly, there is a large amount of salt in the dehydrated microtube they send you. So your resuspension ends up being a buffered solution and not just pure water. Both of these will effect the saturation point. $\endgroup$
    – AMR
    Commented Aug 30, 2015 at 4:53

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .