Indian Journal of Biochemistry & BiophysicsVol. 42, December 2005, pp. 378-385

Energy barriers and rates of tautomeric transitions in DNA bases:Ab initio quantum chemical study

*Soumalee Basuls, Rabi Majumdar", Gourab K Das2 and Dhananjay Bhattacharyya '

IDepartment of Biophysics, Molecular Biology & Genetics, University of Calcutta, 92 Acharya Prafulla Chandra Road,Kolkata 700 009, India

2Department of Chemistry, Visva-Bharati, Santiniketan 731 235, India

3Biophysics Division, Saha Institute of Nuclear Physics, IIAF Bidhannagar, Kolkata 700 064, India

Received 12 May 2005; revised 12 August 2005

Tautomeric transitions of DNA bases are proton transfer reactions, which are important in biology. These reactions areinvolved in spontaneous point mutations of the genetic material. In"the present study, intrinsic reaction coordinates (IRC)analyses through ab initio quantum chemical calculations have been carried out for the individual D A bases A, T, G, Cand also A:T and G:C base pairs to estimate the kinetic and thermodynamic barriers using MP2/6-31 G** method fortautomeric transitions. Relatively higher values of kinetic barriers (about 50-60 kcal/mol) have been observed for the singlebases, indicating that tautomeric alterations of isolated single bases are quite unlikely. On the other hand, relatively lowervalues of the kinetic barriers (about 20-25 kcal/rnol) for the DNA base pairs A:T and G:C clearly suggest that the tautomericshifts are much more favorable in DNA base pairs than in isolated single bases. The unusual base pairing A':C, T':G, C':A orG':T in the daughter DNA molecule, resulting from a parent DNA molecule with tautomeric shifts, is found to be stableenough to result in a mutation. The transition rate constants for the single DNA bases in addition to the base pairs are alsocalculated by computing the free energy differences between the transition states and the reactants.

Keywords: Tautomerism, DNA, spontaneous mutation, quantum chemistry, transition rate, unusual base pairing, intrinsicreaction coordinate analysis

IPC Code: C07H21/04

Proton transfer reactions are of fundamental interestin biology'. Transfer of protons occurring in DNAbases leads to their unusual tautomeric forms, denotedby A', T, G' and C. These rare tautomeric bases haveimino or enol groups, replacing the usual amino orketo groups in normal bases. As a result, the primedbases are capable of choosing wrong pairs. It is easilyrealized, for example, that the base C' pairs with A,instead of G, the base T' pairs with G, instead of A,

*To whom correspondence should be addressedE-Mail: dhananjay.bhattacharyya@saha.ac.inFax: +91-33-2337-4637; Phone: +91-33-2237-0379sPresent address:Department of BiotechnologyWest Bengal University of TechnologyBF 142, Salt Lake, Kolkata 700064, India

Abbreviations: IRC, intrinsic reaction coordinate; BSSE, basis setsuperposition error; SCRF, self consistent reaction field; PCM,polarisable continuum model; SNP, single nucleotidepolymorphism.

and so on. Thus, the presence of anyone of theseprimed bases in DNA would eventually result in aspontaneous point mutation after a single round ofreplication. Although it has never been demonstratedexplicitly that the unusual tautomeric forms of basesare the ones responsi ble for spontaneous mutations,the existing evidences tend to confirm the correctnessof this assumption2

. An understanding of spontaneousmutations is extremely important as these areinvolved in characterizing biological features, fromevolution to single nucleotide polymorphism (SNP),leading to human genetic diseases:'.

Unusual tautomeric forms of bases have beenfound in damaged DNA duplex, indicating that thetransition to such altered forms is indeed feasible~.5Femtosecond dynamics of tautomeric transitions inmodel base pairs measured these as severalpicosecond processes". Various groups investigatedthe details of such transitions on the basis of differentcomputational approaches T", Tautomerism in DNA

,eaifd~ss,;sIS

-eIn'),

:n)eis

Inal~dntA

BASU et al.: RATES OF TAUTOMERIC TRANSITIONS IN DNA BASES 379

base pairs has also been studied recently" by usingLowdins concept of proton tunneling". It must berealized that the tautomeric shifts of the base pairsmay lead to mutation only if (i) the bases open outduring replication phase in their unusual tautomericcondition and (ii) the unusual tautomers form stablebase pair with isosteric Watson-Crick geometry withtheir wrong suite, i.e. A' pair strongly with C and soon.

Although the transition rates from normal to.unusual tautomeric conditions of the bases and basepairs have been studied extensively T", their effect inmutation by mis-pairing has not been addressed tothat extent. We, therefore, addressed these questionson the basis of detailed ab initio quantum chemicalcalculation, including electron correlation effects. Wealso determined the complete reaction profiles (CRPs)for tautomeric transitions in DNA bases as well asbase pairs. CRPs are calculated as the intrinsicreaction coordinates (IRC), which allow us to locatethe transition state or the saddle point" unambi-guously. Replacing the complete energy profilesobtained from IRC by the actual barrier potential, weestimated the transition rates and found these inagreement with recent studies":". We determined therelative stabilities of the wrong base pairs formedduring replication by an unusual tautomeric base,such as C:A, T:G, etc. The strong base pairingenergies of these base pairs indicate that the questionof stability of the wrong base pairs mentioned abovecan cause mutation through tautomeric shift.

MethodDetermination of kinetic and thermodynamic barriers

Similar to any chemical reaction, the transition to atautomeric form also goes through a reaction barrier,where the saddle point structure is located at the topof this barrier. We generated an approximate reactionpath by elongating the covalent N-H bonds (Fig. 1) todifferent values, and calculated the energies for all thegenerated structures. The structure with maximumenergy is refined to obtain the transition state (TS) orthe saddle structure through the calculation of Hessianusing GAMESS software20 with AM121 and HF/3-21G*22 basis sets_ The optimized saddle structures arefinally used to obtain the intrinsic reaction coordinates(IRC) in both forward and backward directions. Asexpected, the two end points of the IRC correspond tothe canonical and unusual tautomeric forms of a baseor a base pair. Single point energies of the reactants,products and saddle structures are finally calculated

using HF/6-31G(2p, 2d) and MP2/6-31G(2p, 2d)basis sets.

Determination of rate constantsTransition state theory assumes a statistical energy

distribution among all possible quantum states at allpoints along the reaction coordinates. Assuming thatthe molecules at the transition state are in equilibriumwith the reactant, the transition rate constantr' wouldbe

kaT so'k =-xexp(---)h RT

(1)

where, ka is the Boltzmann's constant h is the

Planck's constant, sc' is the Gibbs free energydifference between the transition state and thereactant, R is the universal gas constant, and T is theabsolute temperature. The effect of quantummechanical barrier penetration is represented byX (=4 keTNo , where Va is the kinetic barrier'"). Theterm X is equivalent to quantum mechanical tunnelingconstant for reactions with low thermodynamic

barrier. Free energy difference t....C# at 298 K iscalculated as

sc' = CTS - G,

where TS represents transition state and r the reactant.Therefore,

t...C#=t...H# -Tt...S# =HTS -Hr -T(STS -Sr)

,,_ (2)

We have calculated' the free energy terms Cr andCTS using GAMESS with HF/3-21G* basis set Thevalue of CTS also contains additional contributionfrom the electronic energy difference between thereactant and transition states as a part of its enthalpycounterpart, HTS. As the electronic components arevery sensitive to the choice of basis functiorr", we

have calculated these contributions t...H#1 (kinetic. e~barrier) from the more accurate MP2/6-31G(2p, 2d)computations. Geometry and vibrational modes forthe reactant as well as the TS for all the single basesand base pairs are determined by HF/3-21G* basisset. For all the reactions considered here, the terms

t....H !ans, t...H:at and ss !ans are zero, while !:1S:ot has a

negligible value due to change in geometry. The

t....H ~ib components are negative as the TS lacks one

vibrational mode.

••••

380 fNDIAN J. BIOCHEM. BIOPHYS., VOL. 42, DECEMBER 2005

Determination of stability of unusual base pairs and the basepair parameters .

Unusual base pairs like A':C, T':G, etc. may beformed when the template strand has the raretautomeric form during replication. Structures of theseunusual base pairs have been optimized using HF/6-31G(2p, 2d) basis set. Their stability values have beenestimated considering basis set superposition error(BSSE) correction by Morokuma method" ofGAMESS software using the same basis set. In orderto form the pairs, single bases may deform slightly.Energy difference between the structures of the basesin a base pair (complex) and their free optimizedstructures is the deformation energy EOeformationand isgiven by

EOeformation=Ecomplex(base I)-Efree (base 1)+ Ecomplex(base 2) - Efree(base 2)

The stabilization energy, ~E is then defined by

~E '= Ecomplex(base pair) - Ecomplex(base 1)-Ecomplex(base 2) + BSSE + EOeformation

... (3)

where Efree(base. 1) is the energy of the optimizedstructure of first base in the absence of paired base inthe surrounding, Efree(base 2) is the energy of theoptimized structure of the other base forming the base

pair, and Ecomplex(base 1) and Ecomplex(base 2) are theenergies of bases 1 and 2, respectively, after they haveformed the base pair complex.

To estimate the amount of deviation of the unusualbase pairs from a standard Watson-Crick structure, wehave calculated the stretch (or C8-C6 distance,Fig. 1), which is one of the base pair parametersaccording to methodology developed by Bansal andco-workers'f".

In order to assess the effect of solvent ontautomerization reactions, we have calculated theenergy of the bases and the base pairs using selfconsistent reaction field (SCRF) and polarizablecontinuum model (PCM) methods by HF/3-21G*basis set. The base pairs are considered to be inacavity of radius 9 A surrounded by solvent ofdielectric 80.0, while the bases are kept in cavities ofradius between 6 to 7 A, depending upon the size ofthe 'bases. The PCM energy calculations are carriedout using charge-derivative method with watersurrounding each system characterized by solventradius 1.4 A.

Results and DiscussionKinetic and thermodynamic barriers

Fig. 1 shows the geometry optimized (HF/6-3IG(2p,2d» structures of the normal (canonical) and

Fig. I-Geometry oPtimiz~d [HF/6-31 G(2p,2d) basis set] structures of canonical Watson-Crick base pairs [(a) A:T and (b) G:C and thebase pairs formed by the unusual tautomers of the bases (c) A':T' and (d) G"C]

theinvbetbascalbebetsmofbartaudYIen!mehakc,thethe

babedetioth,nopntluG-ahtriph

anevSI!

th:10'chbaall

Ire theV have

the unusual tautomeric forms of the DNA basesinvolved in base pairing. The energy differencesbetween the different tautomeric forms of bases orbase pairs obtained from semi-empirical and ab initiocalculations agree reasonably well (Table 1). It maybe noted that the kinetic barriers (energy differencebetween canonical and saddle structures) for all thesingle bases are quite large, having an average valueof about 56 kcallmol, although their thermodynamicbarriers (energy difference between canonical andtautomeric structures) vary significantly. The thermo-dynamic barrier signifies whether the reaction will beendothermic or exothermic, while the kinetic barriermeasures the speed of reaction. The bases A and Thave high thermodynamic barriers of about 14kcal/mol, while bases G and C have almost negligiblethermodynamic barriers, indicating higher stability ofthe tautomeric forms, especially in the latter case.

It may be mentioned that the low thermodynamicbarrier does not necessarily ensure rapid transitionbetween states, as the kinetic barrier acts as the ratedetermining factor. The extended basis set calcula-tions reveal that the enol form of G (G') is more stablethan normal G, and imino C (C') is almost as stable asnormal e. This observation is in conformity withprevious analysis of the transition of G:C-7G':C'through the formation of an intermediate ion-pairG-:C+ (ref. 28). The highly stable imino C (C') mayalso play an important role in forming H-DNA, thetriple helical DNA with C:G*C' (Fig. 2) base triplet atphysiological pH29.

For the base pairs too, the kinetic barriers for A:Tand G:C (Table 1) are comparable (-23 kcallmol),even though their thermodynamic barriers differsignificantly. Moreover, we would like to emphasizethat kinetic barriers for base pairs are considerablylower, as compared to those for free bases. Thisclearly indicates that tautomeric shifts of isolatedbases are energetically hindered, while similaralterations of the base pairs are relatively much easier.

BASU et al.: RATES OF TAUTOMERIC TRANSITIONS IN DNA BASES 381

nusualre, weuance,neters11 and

nt on:d theg self'izable-21G*e in :amt ofties of;ize of.arriedwater

.olvent

(HF/6-ll) and

l'

t:::J

This is due to mutual assistance of the individualbases in undergoing the tautomeric shift.

Structures of transition intermediatesConversion of a base pair from a normal tautomeric

form to its unusual tautomeric form is a reversibletransition through a saddle structure. Almost all theatomic coordinates undergo movement during thisprocess. Fig. 3 shows how the total energy of a G:Cbase pair changes with IRe. The thermodynamic andkinetic barriers are clearly characterized from the plot.As expected, the lengths of two out of three hydrogenbonds undergo significant alterations as shown inFig. 4(a). The hydrogen bond-lengths decrease in theintermediate structures, as compared to that in the

Fig. 2-Structure of a C:G.C' base triplet optimized by HF/3-21 Gbasis set, where the cytosine (C) and guanine (G) are engaged inWatson-Crick base pairing and the guanine is also involved inHoogsteen base pairing with another cytosine in the unusualtautomeric form (C')

r I Table I-Kinetic and thermodynamic barriers of single DNA bases and base pairs calculated using different levels of theory

~Reactant Product Kinetic barrier Vo (kcallmol) Thermodynamic barrier V J (kcallmol)

AMI 321G* 631G** MP2 AMI 321G* 631G** MP2

A A' 67.58 62.07 62.72 62.78 13.56 14.21 13.44 14.02T T' 61.47 62.16 57.88 57.94 12.89 17.80 14.34 14.59

d " G G' 54.66 50.20 49.12 48.43 3.33 3.33 -0.042 -0.59.C C' 6l.43 53.56 55.97 55.82 3.47 -1.48 0.605 0.49

: and the I A:T A':T' 33.31 2l.90 21.44 21.95 24.52 20.43 17.82 18.26G:C G':C' 35.32 21.70 25.43 24.92 14.81 9.95 .•10.08 9.55

382 INDIAN J. BIOCHEM. BIOPHYS., VOL. 42, DECEMBER 2005

reactant, and finally reduce to covalent bond lengthsin the product. Fig. 4(c) shows a similar phenomenonoccurring in the A:T base pair.

In this study, the constraint imposed on the bases,if any, due to their attachment to phosphate-sugar

ee.eA

00.0

4(1.0

30.0

211.0

10.0

0.0

-'0.0 .. -3 ·2 ·1 a

70.0

Beo.o

50.0

"~ 4(J.a

s~1! 30.0...

20.0

10.0

0.0 .s .,25

C

20

'5

backbone has not been taken into account. We havethus calculated a vital in-plane intra base pairparameter, the stretch or C8-C6 distance, for all thestructures along IRC (Fig. 4 (b) and (d) for G:C andA:T base pairs, respectively). For both base pairs,individual bases come close to one another at thesaddle point. To estimate the possibility of suchdeformations in a regular double helical DNAstructure, we have calculated the stretch values ofboth G:C and A:T base pairs for all the' high

G:C

a

res,obtFrebasstrenotbasrentypsinofthn

Rat

~(

a sIml(lCtramufeatra:givtaueqionlwoThDI'rervelmutau

or

10

I

oL-__~~ ~ ~ -L ~ _

~ ~ ~ 0 2Intmslc Reaction Coorclnates il Bohr SQrt(~)

Fig. 3-lntrinsic reaction coordinate (IRC) versus total energy of(a) cytosine (b) adenine (c) G:C base pair undergoing tautomerism[The structures of the reactant, product and saddle point are alsoshown]

< 2.£!! .---------~ 1.7 -- .«: <.:.::.::.::~...i1.4 --------->:/:.::.--D .. I8. ./ : ...~ 1.1 .1..... \ _

t0.8

·5.5

9.9

b<.s 9.8

aj...'0 9.7•••9B

9.6

·5.5

2<.£

·2.75 o 2.75

G:C

·2.75 o 2.75

A:T

!!~ 1.7

•"6'0

§ 1.4D8.i 1.1

....•.......... .-"------------~,'\ ...

X=.,,=,.~:.:="~':.=:.:.=:.L.~:===~:.~~:.c;-r

0.8 IL. --L L- -L .....l-s.a -6.5 ·2.8 0.0

A:T9.9 n------.-~--....::::;-----..------,

~ 9.8

!!j.•• 9.7'0

9B 9.6

d

9.5 ll... ......l.. 1- ...L. ...l

-e.a ·5.54 -2.78 oIrml1llic Reaction CoOtdina1e Boh' aq,,(amu)

Fig. 4-Plots showing hydrogen bond distances for differentintermediate structures against IRC for (a) G:C to G':C' transition;and (c) A:T to A':T' transition. Also shown are C8-C6 distancesfor the same transitions (b) G:C to G':C' and (d) A:T to A':T'

B

P

~-have: pairill theC andpairs,at the. suchDNA

ies of'high

BASU et al.: RATES OF TAUTOMERIC TRANSITIONS IN DNA BASES

resolution crystal structures of DNA double helicesobtained from nucleic acid database, July, 20023°.Frequency histograms of stretch values of the twobase pairs (Fig. 5) clearly indicate that the smallstretch values that we obtained at saddle structures arenot unusual. It may be worth mentioning that thebases of a pair of all the transition intermediatesremain planar, devoid of any propeller twist or buckletype of movements (data not shown). Earlier, asimilar observation of planar base pairs in the absenceof sugar-phosphate backbone was also revealedthrough quantum chemical analysis".

Rates of tautomeric transitions for single bases and base pairsThe rate constants k, calculated from the values of

/1GII using Eq. 1, confirm that the tautomeric shift ofa single base is much less likely (Table 2), perhapsimpossible, as the transition rates are extremely small(10-21

S·I for G ~ G' to 10.31 S-I for A ~ A'). Thetransition rates for the base pairs are comparativelymuch higher (_10.4 S·I) and are, therefore, quitefeasible. However, it should be kept in mind that thistransition rate is not the mutation rate, but instead itgives us the number of base pair that would undergotautomeric shift per sec. For example, in a dynamicalequilibrium between the canonical and unusual forms,only one base pair out of 104 canonical base pairswould be seen in its unusual tautomeric form per sec.This transition may cause spontaneous mutation inDNA only if the transformed base pair opens up forreplication at that moment, the probability of which isvery small. Therefore, the rate of spontaneousmutation has to be much smaller than the rate oftautomeric transition.

We further characterized the transition profiles ofG, C and G:C base pair using an extended basis setand the results are shown in Table 3. The saddlepoints have been determined and the IRCs areanalyzed using HF/6-31G(p,d) method, while theirfinal single point energies are calculated by evenmore accurate MP2/6- 31l+G(2p,2d) method. Acomparison of the transition rates with those obtainedfrom HF/3-21G* basis set indicates that the results aregenerally independent of the choice of basis set.

G:C'l0.8

0.4 IIl

~I ,~ I , ,i 0

A:T1''10.8

0.4

01,1 I ~I9.5 9.6 9.7

C8-C& dill1an09 in A9.6

Fig. 5- The observed frequency distribution of CS-C6 distancesin crystal structures of double helical DNA are shown forcomparison for (a) G:C base pairs; and (b) A:T base pairs

Table 2-Enthalpy (H), entropy (S) and free energy (G) for reactant and transition state of a single base and a base pair and, alsoquantum transmission coefficient (X) and transition rate constant for the reaction (k) are shown. For the transition state (TS), an

additional component Vo (MP2/6-31G (2p,2d) energies from Table I) has been included with the enthalpy

Base/base Structure H (kcallmol) S G /::,.C# X k

pair type at Vibrational component Kinetic barrier (callmol-K) (kcal/rnol) (kcallmol)

AReactant SI.5 0.0 SO.3 57.5 59.32 0.03 _ 10-31

TS 77.7 62.S 79.3 54.0

GReactant 85.5 0.0 85.8 59.9

45.05 0.04 _10-21TS 81.7 48.4 84.2 56.6

TReactant 83.3 0.0 81.1 59.1

53.83 0.03 _10-27

TS 79.1 57.9 80.7 55.0

CReactant 72.2 0.0 77.5 49.1 52.49 0.03 _10-26

TS 68.5 55.8 76.3 45.8

lifferent A:TReactant 167.2 0.0 125.9 129.6

18.82 0.09 _10-3

msition; TS 162.9 21.9 121.8 126.6

istances G:CReactant 160.3 0.0 119.7 124.7

19.84 0.08 _10-4

l' TS 154.7 24.9 117.9 119.6

383

J

384 INDIAN J. BIOCHEM. BIOPHYS., VOL. 42, DECEMBER 2005

_ Table 3-Results of more precise calculations i.e., optimization by HF/6-31 G(p,d) basis set and single point energy calculation byMP2/6-311 +G(2p,2d)

Base/base Barrier heights (HF/6-31G (p,d) Barrier heights (MP2/6-31I +G (2p,2d) kpair type //HF/6-31 G (p,d» //HF/6-31 G (p,d»

Kinetic Thermodynamic Kinetic Thermodynamic

Guanine 48.67 -0.23 49.63 -0.34 10-23

Cytosine 55.47 0.61 56.65 1.05 10-28

G:C 25.56 9.57 26.46 9.51 10-4

Table 4-Effect of solvent on kinetic and thermodynamic barrier for single bases and the base pairs[The third column in Vo and V I correspond to the values also reported in Table 1 calculated by same method]

Reactant Product v« (kcallmol) V I (kcallmol)

SCRF PCM Vacuum SCRF PCM Vacuum

A A' 62.08 62.92 62.07 14.00 11.83 14.21

T T 61.63 60.73 62.16 17.52 15.73 17.80

G G' 50.50 54.34 50.20 1.95 7.17 3.33

C C 53.91 57.38 53.56 -0.67 3.66 -1.48

A:T A':T 21.98 22.22 2UJO 20.39 20.48 20.43

G:C G':C 22.14 25.23 21.70 10.03 12.03 9.95

Table 5-Stabilization energies for normal base pairs and unusual base pairs considering BSSE and deformation corrections

Base type Stabilization energy BSSE correction Deformation correction Stabilization energy (t.E) Stretch of base pairs(kcallmol) (kcallmol) (kcallmol) with corrections (A)

(kcallmol)

A:T -12.02 2.77 0.73 -8.52 9.96G:C . -27.22 3.45 2.87 -20.91 9.96C:A -11.77 2.65 0.74 -8.47 10.01A':C -20.32 2.83 1.51 -15.98 9.94

T:G -29.39 3.58 4.03 -21.78 9.76

G':T -15.77 3.38 1.46 -10.92 9.88

We also analyzed (Table 4) the effect of watermolecules surrounding the baseslbase pairs bycalculating the kinetic and thermodynamic barriers forthe bases and the base pairs after solvating them bythe Kirkwood-Onsager spherical cavity model(SCRF)32,33 and the polarizable continuum model(PCMi4 using 3-21 G* basis set. It is evident from ourresults and also from a recent report using densityfunctional theory" that the effect of solvation on thebarriers is rather insignificant.

Stability of non- Watson-Crick base pairsAs indicated earlier that one of the necessary

conditions for mutation through tautomeric shift is theformation of stable and isosteric wrong base pairs, i.e.A':C, G':T, etc. We found that the optimizedgeometries of the wrong base pairs have near perfectWatson-Crick type geometry. These complexes havegeometries very similar to normal Watson-Crick base

pairs, as reflected by the stretch values (Table 5), andhence may not be recognized as non-Watson-Cricktype by the proofreading mechanism of DNApolymerase. We calculated the stabilization energy~E (with BSSE and deformation corrections) forthese unusual base pairs, by using Eq. 3. A largenegative value of this energy stabilizes a base pair.Results of our calculations are reported in Table 5,where it is found that the corrected stabilization'energies, involving the unusual bases' can be asfavorable as the normal base pairs. For example,stabilization energy of C':A is quite close to that ofA:T, both forming two hydrogen bonds. The largevalue of ~E for G:C, as compared to that of A:T is notunusual and a similar feature was reported earlier31

,3S.

Although both A':C and C':A have two hydrogenbonds and very similar total energies (data notshown), they show significantly different stabilizationenergies (Table 5). The high stability of A':C, as

CO]

bomeSirCO]

fOJ

urnba:dodoiopas10\

W:eviomabrercarda!

Ac

Tv.forunrNo

Re1

2

34

5

6

7

8

rs

BASU et al.: RATES OF TAUTOMERIC TRANSITIONS IN DNA BASES

compared to C:A or A:T, all having two hydrogenbonds, may be partly due to relative instability of theindividual base A', as compared to A (Table 1).Similarly, the relative instability of the base T', ascompared to T leads to a lesser stabilization energyfor G':T, as compared to T':G.

ConclusionThe rates of transitions of nucleotide bases to

unusual tautomeric forms, and the reverse, withinbase pairs are quite significant and these transitionsdo not cause any structural alteration to the DNAdouble helix. These bases, in unusual form, whenopen up for replication, may remain in the same form,as the rates of transitions of isolated bases are verylow. These unusual tautomers then can form stableWatson-Crick like base pairs with wrong suite, asevident from their high interaction energies. Thus,once a wrong pairing Occurs due to tautomeric shift ofa base pair and the bases open up for replication, theyremain stable enough and do not return to theircanonical form. These then can cause mutation in thedaughter cells.

AcknowledgementWe thank Prof. RK Moitra for useful discussions.

Two of us (RM and SB) are grateful to CSIR, Indiafor providing the facility and financial supportunder the CSIR Emeritus Scientist SchemeNo. 21 (0474)/OOIEMR-II.

and'rick>NAergyfor

argepair.Ie 5,ation

ReferencesI Watson J D & Crick F H C (1953) Symposium on

Quantitative Biology XVIII, 123-131, Cold Spring Harbor2 Szczepaniak K & Szczepaniak M (1987) J Mol Struc 156,

29-423 Brookes A J (1999) Gene 234, 177-1864 Robinson H, Gao Y G, Bauer C, Roberts C, Switzer C &

Wang A H-J (1998) Biochemistry 37, 10897-109055 Chatake T, Hikima T, Ono A, Ueno Y, Matsuda A &

Takenaka A (1999) J Mol Bioi 294, 1223-12306 Douhal A, Kim S K & Zewail A H (1995) Nature 378,

261-2637 Kwiatkowski J S & Leszczynski J (1990) J Mol Struct 208,

35-44Smedarchina Z, Fernandez-Ramos A, Zgierski M Z &Siebrand W (2000) DOITl.2, Natl Res Council, Canada

e asnple,at oflarges notr31,35

'ogennot

ation:, as

. , 385

9 Garb L, Podolyan Y, Fernandez-Ramos A & Smedarchina Z(2002) Biopolymer 61, 77-83

10 Katritsky A R & Waring A J (1962) J Chern Sac 2,1540-1544.

11 Niedzwiecka-Kornas A, Kierdaszuk B, Stolarski ~ & ShugarD (1998) Biophys Chem 71; 87-98 .

12 Suen W, Sipro T G, Sowers L C & Fresco J R (1999) ProcNatl Acad Sci (USA) 96, 4500-4505

13 Sponer J, Sabat M, Gorb L, Leszczynski J, Lippert B &Hobza P (2000) J Phys Chem B 104, 7535-7544

14 Zoete V & Meuwly M (2004) J Chem Phys 121,4377-438815 Kryachko E S (2002) Int J Quantum Chem 90, 910-92316 Garb L, Podolyan Y, Dziekonski P, Sokalski W A &

Leszczynski J (2004) JAm Chem. Sac 126, 1OU9-1OI2917 Golo V L & Volkov Y S (2003) Int J Modern Phys 14,

133-15618 Lowdin P (1963) Rev Mod Phys 35, 724-73319 Schmidt M W, Gordon M S & Dupuis M (1985) JAm Chem

Soc 107,2585-258920 Schmidt M W, Baldridge K K, Boatz J A, Elbert S T, Gordon

M S, Jensen J J, Koseki S, Matsunaga N, Nguyen K A; Su S,Windus T L, Dupuis M & Montgomery J A (1993) J ComputChem 14, 1347-1363

21 Dewar M J S, Zoebisch E G, Healy E F & Stewart J J P(1985) JAm Chem Soc 107,3902-3909

22 Binkley J S, Pople J A & Hehre W J (1980) JAm Chem Sac102,939-947

23 Schiff L I Quantum Mechanics, Me Graw-Hill KogakushaLtd., Tokyo _

24 Jensen F (1999) Introduction to Computational Chemistry,chapt 12, John Wiley & Sons

25 Morokuma K (1971) J Chem Phys 55, 1236-123926 Bansal M, Bhattacharya D & Ravi B (1295) CABlOS 11,

281-287

27 Olson W K, Bansal M, Burley S K, Dickerson R E, GersteinM, Harvey S C, Heinemann U, Lu X-J, Neidle S, Shakked Z,Sklenar H, Suzuki M, Tung C-S, Westhof E, Wolberger C &Berman H M (2001) J Mol Bioi 313, 229-237

28 Florian J & Leszczynski J (1996) J Am Chem Soc 118,3010-3017

29 Frank-Kamenetskii M D, Mirkin S M (1995) Ann RevBiochem 64, 65-95

30 Berman H M, Olson W K, Beveridge D L, Westbrook J,Gelbin A, Demeny T, Hsieh S-H, Srinivasan A R, SchneiderB (1992) Biophys J 63, 751-759

31 Gould I R & Kollman P A (1994) J Am Chem Sac 116,2493-2499

32 Kirkwood J G (1934) J Chem Phys 2,351-36133 Onsager L (1936) JAm Chem Sac 58,1486-149334 Miertus S, Scrocco E & Tomasi (1981) J Chem Phys 55, 17735 Kudritskaya Z G & Danilov V I (1976) J Theor Bioi 59,

303-318