Re: What is the ratio of tautomers of the bases to their normal forms?

Jennifer,

First of all, let me apologize for the time it has taken me to respond to 
your question, but the difficulty of your question and my own work load 
have made it tough to get back to you any sooner than this.

Unfortunately, I don't quite have a direct answer to your question.  The 
problem of tautomer formation in nucleic acid base pairs is a difficult one 
to study experimentally.

As you may be aware, a hydrogen bond is an interaction between a donor and 
acceptor involving dipolar interactions between a proton and electrons.  
Donors are typically an imide or amide (N-H), or a hydroxyl group (O-H); an 
acceptor is usually a lone pair of electrons on either a nitrogen or an 
oxygen atom.  The hydrogen from the donor loosely associates with the 
electrons on the acceptor through dipole-dipole interactions, also 
resulting in some sort of sharing of the proton and/or electrons between 
the heavy atom donors and acceptors.  In a tautomerisation reaction, there 
is an exchange of the hydrogen and lone pair electrons; the donor becomes 
the acceptor and vice versa.

For example, in the following equilibrium:

	-N-H...N-  <=>  -N...H-N-

The hydrogen from the first nitrogen is transferred to the second (the lone 
pair electrons are effectively transferred from the second to the
first, though the actual transfer depends on other factors including pH).  
In base pairing interactions, there must be a transfer of a second proton 
in either A-T or G-C base pairs for the tautomer to be
stabilized.

Several lines of evidence from some key papers indicate the ration of the 
enolic and keto forms of monomer (non-base paired) thymine and guanine 
bases is very small, leading to the conclusion that the amount of unusual 
tautomers in base pairs is equally small.  NMR and mass spectrometry are 
the techniques most commonly used to investigate tautomerisation reactions, 
providing direct information on individual atoms involved, and the rate 
tautomer formation.

Some of the key papers in this field (see references below) indicate that 
the rate of single proton transfer is quite fast, with time scales on the 
order of 250 femtoseconds.  The rate of the second proton transfer is on 
the order of 2 picoseconds, about 10 times slower.  The energy barrier for 
the transfer of the first proton is small, allowing the backward reaction 
to occur rapidly, resulting in the original Watson-Crick base pairing 
interaction rather than a stable tautomer.  The same is true of the reverse 
reaction, occurring with similar rate
constants and equally small energy barriers.  The equilibrium constant for 
the reaction has been estimated around 0.01, though it may be approximately 
1 depending on the temperature.  The gist is tautomers can form quickly and 
easily, and may constitute 1% or as much as 50% of the total base pair 
conformations.

The second part of your question, can tautomers of base pairs result in 
mutations in the DNA sequence, is a function of the fidelity of the DNA 
polymerase.  Typical rates of base pair addition for DNA and RNA polymerase 
complexes are on the order of 100 to 10000 per second.  While this may seem 
fast, it is quite slow relative to the rate of tautomer formation.  
Therefore, a given base pair will undergo a complete tautomerisation and 
the reverse reaction several time on the time scale of base pair addition. 
 The other aspect of the equation is that a polymerase does not look at a 
base pair per se; the base pairs are broken (usually by a separate helicase 
protein, though bacterial and viral polymerases may contain their own 
helicase activity), prior to the interaction of a template base with the 
active site of the polymerase.  It is not clear to me if anyone has 
demonstrated any function of a helicase in discriminating between 
Watson-Crick base pairs and various tautomeric forms of base pairs.  
Generally, it is thought that a helicase does not "read" base pairs in a 
specific sense, rather they seem to chug along breaking hydrogen bonds as 
the go.  My guess is, the base pairs will be broken in their tautomeric 
form just as easily as in their canonical base paired configuration.

Addition of a new nucleotide on the growing chain involves base pairing 
with the template strand in the active site, formation of the 
phoshpodiester bond, and translocation of the enzyme; both during the 
reaction, and subsequent to polymerase translocation, the polymerase itself 
and proof-reading enzymes assay for possible mismatch additions.  The 
efficiency of this proof-reading function is variable among different 
organisms, but is usually quite high, on the order of 10^8 or 10^15 in 
humans, and as low as one mistake in 1000 for some viruses (where the 
ability to mutate is necessary for survival).  Unfortunately, the rate of 
mutation and the equilibrium of tautomer formation are not easily separable 
experimentally.  Tautomers are formed all the time, but polymerases seem to 
deal with it without any problems.


I hope that helps a little...for some primary data on tautomeric 
equilibria, you might take a look at the following papers, as well as 
references therein.  Most medical libraries should have them:

  1)  H. Rutterjans et al., 1982, Nucleic Acids Research vol.10 num.21, 
pp7027-7039.  "Evidence for tautomerism in nucleic acid base pairs. 1H NMR 
study of 15N labeled RNA".

  2)  B. P. Cho and F. E. Evans, 1991, Nucleic Acids Research vol.19 num.5, 
pp1041-1047. "Structure of oxidatively damaged nucleic acid adducts. 3. 
Tautomerism, ionization and protonation of 8-hydroxyadenosine studied by 
15N NMR spectroscopy".

  3)  A. Douhal et al., 1995, Nature vol.378, pp260-263.  "Femtosecond 
molecular dynamics of tautomerisation in model base pairs".


Regards,
Dr. Jim Kranz