Re: Why are RNA duplexes more stable than DNA duplexes in vitro?

Hi Sulzhan,

This is an interesting question you pose, and is one that has been widely
studied over the last 50 years.  The structural differences between RNA and
DNA are well understood, though there is still some debate surrounding the
dominant energetics underlying these nucleic acid interactions.  This is a
complicated issue, and in the interest of brevity I’m assuming you have
some general knowledge that I’ll skip over.  To address the differences in
thermostability, we have to understand the chemical and structural
differences of DNA versus RNA then the energetic consequences of these
differences.

First, let us address the problem of helix formation; both RNA and DNA have
comparable folding mechanisms, that is the formation of base pairing
interactions with a second strand (or same-strand in the case of hairpin
formation), which (1) involves hydrogen-bond formation between opposing
strands, (2) stacking of base pairs on top of one another, (3) reducing
conformational freedom of the phosphodiester backbone.

The first component, base-pairing through hydrogen-bonding interactions may
not be an important factor in comparing DNA versus RNA.  In terms of the
individual purines and pyrimidines, the only difference is found in
comparing uracil with thymine (which bears a 5’ methyl group lacking in
uracil).  This is known to contribute only a small fraction of the total
energy of base-pairing, adding slightly MORE energy to a DNA dA:dT base
pair compared to an RNA A:U base pair.  We can ignore this as the primary
source of the high relative thermostability of RNA.

The second and third components are interlinked, since it’s the
conformation of the phosphodiester backbone that ultimately determines the
relative orientation of one plane of paired nucleotide bases relative to
the nearest neighbor base pairs.  For a given nucleotide in either a DNA or
RNA strand, there are no fewer than 6 degrees of freedom (rotatable bonds),
counting two for each phosphate-oxygen, one to the C5’ carbon, one between
C5’ and C4’, one between the C3’ and the oxygen on the phosphate of the
next nucleotide (that’s five), and finally rotation of the purine/pyridine
base relative to the C1’ of the ribose or deoxyribose sugar.  In the case
of non-base paired, single-stranded RNA or DNA, all six of these bonds are
freely rotatable which makes these polymers extremely flexible.  In order
to become double-stranded, every nucleotide must adopt a single “preferred”
conformation, which requires that all six of these rotatable bonds be fixed
into a single orientation.  This is a VERY unfavorable process in terms of
the energetics of forming a double-stranded DNA or RNA, but is largely the
same process for both molecules.

The only other major difference between RNA and DNA is the detailed shape
of the double-helix, A-form for RNA and predominantly B-form for DNA
(please refer to your textbooks or to any of the references below for
additional detail).  RNA has never been observed to take on a B
double-helix; the presence of that 2’-OH almost exclusively locks the
ribose into a 3’-endo chair conformation, eliminating the possibility of a
stable B-helix.  However, the deoxyribose sugar may alternate between
2’-endo and 3’-endo conformations, allowing DNA to switch between B-form
and A-form under the right circumstances.  Note that hybrids of DNA:RNA
(one strand of each in a double-helix) adopt an A-form conformation.  (To
better understand the differences in allowable sugar puckers, you might
wish to return to your organic chemistry ball-and-stick models).

The B-form of DNA (in the presence of physiological Na+ or K+) is found at
high relative humidity; large numbers of water molecules are tightly bound
(to the tune of almost 1:1 water/nucleotide).  By comparison, it has been
shown that A-form RNA and A-form DNA both are dehydrated somewhat;
measurements of 75% the number of tightly bound water molecules compared to
B-form DNA are commonly cited.  There is a distinct difference between
tightly bound water and bulk solvent that will have profound energetic
consequences.  In fact, by putting DNA into a dehydrating medium (such as
low salt and high concentrations of ethanol), one can drive the
interconversion of B-form DNA into A-form.  Curiously, high salt (>2.5 M
NaCl) and high concentrations of ethanol will drive B-form to Z-form (a
left-handed helix) for DNA, and at elevated temperatures for RNA as well. 
The source of these effects are largely Coulombic (charge-charge
interactions) in nature, having to do with the unfavorable interactions
between adjacent phosphates on the backbone and the ability of solvent
composition to diminish (high salt/high humidity) or maximize (low salt/low
humidity) these unfavorable interactions, the details of which are
unapproachably complicated for our discussion.

There are important structural differences between A-form and B-form
helices that we must consider, notably in the diameter of the duplex, the
number of base pairs per turn, the tilt of paired bases relative to the
helical axis, and the solvent accessibility of major and minor grooves.  Of
these factors, it is the relative orientation and overlap of
nearest-neighbor base pairing interactions that, though only subtly
different, have contribute to the observed differences in thermostability
of RNA and DNA.

I’ve touched on a few important driving forces governing the transition
between duplex and single-stranded nucleic acids, and some of the potential
STRUCTURAL differences in these interactions between RNA and DNA.  In terms
of the relevant energetic contributions, the stacking of base pairs, one
above the other, plus the hydrogen bonds between bases provide the
stabilizing enthalpy of the helix, adding substantial energy stabilizing
the duplex when summed over the length of the DNA/RNA.  Both cross-strand
and same-strand van der Waals interactions among bases are important; the
magnitude of these favorable interactions are slightly different for RNA
(more stable) than for DNA; these small differences become large when
summed over many base pairs.  The charged phosphate groups repel one
another by Coulomb’s law of repulsion between like charges, an
enthalpically unfavorable interaction.  As mentioned above, the formation
of a double-helix results in a significant reduction in the conformational
degrees of freedom, which is entropically unfavorable in an equally big
way, and are also subtly different in A-form versus B-form.

In total, the single-strand to double-strand transition for both DNA and
RNA is enthalpically favors the helix and entropically favors
single-stranded conformation.  For RNA, deltaH ~ 40 kJ mol-1/base pair and
deltaS ~ 105 J K-1 mol-1/base pair (note the entropy is a function of
temperature).  For DNA, deltaH ~ 35 kJ mol-1/base pair and deltaS ~ 90 J
K-1 mol-1/base pair.  These are VERY large and OPPOSITE driving forces.

In terms of the free energy, the balance of these interactions, we observed
a higher melting temperature of RNA relative to the same sequence in DNA
under normal conditions.  The dominant source of this slightly higher
energy for RNA is generally attributed to modestly better base-stacking
energy in the A-form conformation.  The precise nature of the molecular
driving forces remain an active area of research.  Experimentally, one
observes very little difference in thermostability between RNA:DNA
double-helix compared to an all RNA double-helix, consistent with the
theory that the source of thermostability is due largely the result of
A-form versus B-form conformational differences, not strictly differences
in ribose versus deoxyribose chemistry.

For additional reading, I suggest the following textbooks, or refer to any 
introductory level biochemistry or molecular biology textbook:

     "Biochemistry" 4th ed., Lubert Stryer. New York: W.H. Freeman and Co.

     "Genes VII", Benjamin Lewin. New York: Oxford University Press,

     “Physical Chemistry”, Tinoco, Sauer, Wang, Publisi. New Jersey: 
           Princeton Hall, Inc.

     “Biophysical Chemistry”, Cantor and Schimmel.  New York: W. H. 
           Freeman and Co.

Thanks for the tough question.

Regards,
Dr. James Kranz