Scott Keeney scottk at
Thu Jul 16 05:47:54 EST 1992

In article <1992Jul16.085212.23335 at>,
 PHILLIPSA at UK.AC.AFRC.LARS (Andy Phillips) writes:

|> Mike Poidinger asks about nucleic acid precipitation and salt.  I always
|> believed that the purpose of adding salts was to shield/counteract the 
|> negative charges of the phosphate groups on the DNA/RNA backbone, so that 
|> the strands can come closer together and precipitate as the ethanol removes
|> the water. 

Yes, the purpose of adding salts is to neutralize the charge on the
sugar-phosphate backbone of the DNA, but ethanol's task is a little
more complex than "removing" the water. For a precipitation, you're 
interested in forming ion pairs between the polyanion (DNA) and the
cation (Na+, Mg++, spermidine, protamine, etc). In dilute aqueous
solution, DNA and counterions like Na+ and Mg++ are more or less in
the free ion form rather than the ion pair form (that is, they are
surrounded by one or more layers of water molecules).

Water has a high dielectric constant (e), which from Coulomb's Law
tells us that the electrostatic force (F) between two ions of opposite
charge is very low in water:

		     Q1 * Q2
		F = ----------
		     e * r^2

where Q is the charge on each ion and r is the distance between them.
Adding organic solvent *decreases* the dielectric constant of the 
solution. As e goes down, F goes up and *BANG*, anion and cation
form an ion pair and promptly swoon out of solution.

|> On the same theme, a student asked my yesterday why NaCl increases
|> the stability of DNA duplexes, although you might expect salts to
|> intefere with hydrogen bonds, rather than strengthen them. Any
|> answers?

Same reason you gave above for adding Na+ to the ethanol precipitation:
it neutralizes the charge. Each strand of DNA has an enormous charge
density (charge per unit volume), so the two strands tend to
push each other apart. Cations added to the solution form a "cloud"
of positive charges around the DNA. This cloud of counterions lowers the
effective charge density and relieves the repulsion between the strands.

As for the effect of salt on hydrogen bonds, you have to realize that
the hydrogen bonds formed between bases in duplex DNA contribute little
to the stability of the duplex. For an interaction to stabilize the 
duplex, the interaction between bases must be stronger than the
interaction of the bases with water (if bases are not paired with
one another in a duplex, then they are surrounded by water). Hydrogen
bonding between the amines, carbonyl oxygens, etc. of G-C or A-T
is of the same energy (sometimes even less) than the hydrogen bonds
these same groups would form with water if the DNA were single-
stranded. (The H-bonds do contribute *something*: GC base pairs with 
three H-bonds are harder to melt than AT pairs with two.)

So, what drives DNA strands together? Entropy and enthalpy, of
course. Entropy in the form of "hydrophobic" interactions between
the bases (those big aromatic rings are quite hydrophobic, you
know). Enthalpy in the form of favorable, stabilizing interactions
between the pi electrons of the aromatic rings of bases as they
stack on top of one another.

So, what do hydrogen bonds do, if they don't stabilize the duplex?
They enforce the *specificity* of base pairing. Correct base
pairing is nice, but doesn't add much. Incorrect base pairing,
on the other hand, takes a lot away. Forcing unpaired H-bond
donors and acceptors (i.e. hydrophilic groups) into a hydro-
phobic environment makes everybody unhappy.

I've simplified a number of points for the sake of brevity (I
came close, but I didn't want to write a textbook). I'm sorry
if I glossed over your favorite factlet on DNA physical
chemistry ;-)

Excercise left to the student: why is "hydrophobic interaction,"
in the context of entropically-driven processes, a misnomer?

       ____         Scott Keeney, DNA repair-queer         ____
       \  /           scottk at           \  /
        \/        Biochemistry and Molecular Biology        \/ 
                            U.C. Berkeley

More information about the Methods mailing list