POLSKIE ■ TOWARZYSTWO BIOCHEMICZNE
Advances in Biochemistry
TOM 41, NR 5 supplement, 1995
Special issue dedicated to
Professor David Shugar
on the occasion of his 80th birthday
Profile — David S hugar
298
Tautomerism of thioguanine . . . 300
Mutations, DNA damage & rep air. 308
Proteins secreted by £
coli
. . . 3 1 3
Gene transfer s y s te m ...318
Scope and limitation of BNCT . . 321
DNA polym erases...326
BuPdGTP and DNA polymerase . 332
Therapy of HIV infection... 338
Herpesviruses-encoded TK . . . . 342
Antiviral phosphonates... 347
Symposium — P rogram ... 352
This edition of a supplement to
the quarterly Postępy B ioch em ii
w as made possible through the f i
nancial support of the Polish Bio
chemical Society and the C o m m it
tee fo r Scientific Research (K B N )
CONTENTS
Editor
POLSKIE TOWARZYSTWO BIOCHEM ICZNE
Polish Biochemical Society ul. L. Pasteura 3
02-093 Warszawa Poland
(2) 658-20-99 fax: (22) 22-45-52
D rukarnia N aukow o-Techniczna M ińska 65 03-828 W arszawa Editorial Board Editor-in-Chief Z O F IA ZIE L IŃ SK A tel. 31-24-03 Editors
GRAŻYNA PA LAM A RCZY K tel. 658-47-02 A N D R Z E J JE R Z M A N O W SK I tel. 659-70-72 ext. 1217 AN NA SZAKIEL tel. 23-20-46 JO LA N TA GRZYBOW SK A tel. 13-05-15
E ditor of this issue D A N IELA BARSZCZ tel. 658-44-99
ext. 3334
Address
REDAKCJA KWARTALNIKA „POSTĘPY BIOCHEM II” INSTYTUT BIOLOGII DOŚWIADCZALNEJ im. M. Nenckiego PAN ul. L. Pasteura 3 02-093 Warszawa tel (2) 659-85-71 ext. 332 fax: (22) 22-53-42
Profile — David S h u g a r ...298
Tautom erism of thioguanine. Experim ental m atrix isola tion and th eo retical quantum -m echanical studies
KRYSTYNA SZCZEPANIAK, W ILLIS B. PERSON, JERZY LESZ CZYŃSKI, JÓZEF S. K W IA T K O W S K I...300
Mechanism s of action of m ethyl m ethanesulfonate on
Escherichia coli: M utagenesis, D N A dam age and repair
CELINA J A N I O N ... 308
Secretion of proteins by Escherichia c o /i and its ap plication in production of recom binant proteins
M A G D A LE N A M. FIKUS, BOŻENNA R E M P O Ł A ... 313
S y n th etic virus-like tra n sfer system
STAN ISŁAW S Z A L A ...318
Scope and lim itatio n s o f boron neutron cap ture therapy as th e selective tool against cancer
H A N N A W O JTO W ICZ, H A LIN A TH I EL-PAW LICKA,
KRZYSZTOF G O L A N K IE W IC Z ... 321
M am m alian DN A polymerases
JA N U S Z A. SIEDLECKI, RAD O SŁAW A N O W A K ...326
Butylphenol-dG TP as a stru ctu re probe of B fam ily DN A polymerases
JA M E S M. STATTEL, GEORGE E. W R IG H T ...332
C hem otherapy o f human im m unodeficiency virus (H IV ) in fectio n based on chem o therapeutic intervention w ith early steps o f th e virus replicative cycle
ERIK DE C L E R C Q ... 338
Herpesvirus-encoded th ym id in e kinase in chem otherapy
TO M ASZ OSTROWSKI, JO LA N T A NOW OTNA, BOŻENNA
G O LANKIEW ICZ ... 342
The potential of acyclic nucleoside phosphonates as broad-spectrum antiviral agents
LIEVE NAESENS, JA N BALZAR IN I, ERIK DE C L E R C Q 347
Symposium on stru ctu re and biological functions of nucleic acid com ponents and th e ir analogues, and related topics — P r o g r a m ... 352
On behalf o f the Polish Biochemical Society the Editorial Board o f
Postępy Biochemii (Advances in Biochemistry) takes pleasure in dedica
ting this special issue o f the journal to Professor D a v i d S h u g a r on the
occasion o f his 80th birthday. It contains papers written by his coworkers
and friends.
On the same occasion a special Symposium was organized in the
Institute o f Biochemistry and Biophysics,
Polish Academy o f Sciences.
The program o f the Symposium is included in this issue o f Postępy
Biochemii.
Professor David Shugar
delivering his lecture during the Symposium Warsaw, September the 2nd, 1995
Profile — David Shugar
David Shugar was born in Poland on the 10th of September, 1915. W hen he was three years old, the family emigrated to Canada. There he grew up, was educated, and obtained his P hD in physics at M cGill University (M ontreal) in 1940.
His interest in the biological sciences arose during his P hD research when he worked, for financial support, on the construction of an ultracentrifuge for medical purposes at the University hospital. U ntil 1948 he worked in the field of physics and biophysics in scientific laboratories in Canada. The first occasion to start biological research opened up in postw ar Europe, when he received a fellowship at the Pasteur Institute, and subsequently at the Sorbonne. He worked there on kinetic of enzymes, protein denaturation and p h o to chemistry. These studies were continued at the U niver sity of Brussels.
In 1952 he published four papers with J. J. Fox on the UV absorption spectra and tautom erism of nucleic acid derivatives. In an elegant way they identified tautom eric forms of pyrimidines and proved N (l) as the position of attachm ent of a sugar moiety to the base in nucleosides. These findings became fundam ental inform ation on the structure of nucleic acids. The simple m ethod of absorption spectra m easurem ent was used for years in studies on the structure of natural and synthetic polynucleotides and their interactions with various agents.
In the early fifties D. Shugar came to Poland, initially for a one-year stay on the invitation of Professor Leopold Infeld and the M inistry of Higher Education. He was a young researcher with broad knowledge, brilliance of m ind and the rare ability of finding interesting and im portant problem s th at could 298
http://rcin.org.pl
POSTĘPY B IO C H E M II 4 1 ( 5 ) , 1995be solved under the available conditions. He held a basic belief th at in science, nearly in any circum stan ces, one can do something useful with hard work and a little creativity. He found here the opportunity to use both.
To start with, he held the post of head of a laboratory in the Biochemistry D epartm ent at the State Institute of Hygiene in W arsaw. Later, he headed the D ep art m ent till 1966 when he left the Institute.
In the Institute of Nuclear Research at Świerk he participated in the setting up of an analytical lab o rato ry for m easuring the purity of reactor materials.
From 1954 on, D. Shugar was involved in the organization of the Institute of Biochemistry and Biophysics of the Polish Academy of Sciences. At the beginning he organized a L aboratory of Biological Physical-Chemistry, which grew to become a D epart ment, after some time renam ed the Biophysics D ep art ment. After some years this was divided into two D epartm ents, with M olecular Biology headed by D. Shugar till his retirem ent in 1985. It is w orth m en tioning that, due to his organisational activity and influence on the scientific profile, the Institute was given the second part of its name.
D. Shugar organized in 1955, with the help of coworkers and the Faculty of Physics (University of Warsaw), the first course in Poland on isotope tech niques for biologists and medical personnel. A collec tion of the lectures in book form was very helpful for a long time. At the State Institute of Hygiene he designed, P oland’s first, cobalt-60 source for research in radiation chemistry and biology, and supervised its construction.
In 1966, at the request of the M inistry of Higher Education and the Faculty of Physics, University of W arsaw, D. Shugar began to organize a C hair of Biophysics, currently the D epartm ent of Biophysics, at the Faculty. He was the principal au th o r of the program s for teaching and research, and the head of the U nit for m any years. It is the only biophysics departm ent in Poland in a physics faculty; others were organized later usually as parts of biology dep art ments. Over a hundred physicists have already specia lized in m olecular biophysics in the D epartm ent.
D. Shugar served also for some time as a consultant in the D epartm ent of Radiobiology in the Institute of Oncology in W arsaw. Due to his cooperation and inspiration there is now a m odern laboratory of m olecular medicinal biology.
These huge organisational efforts ran sim ultaneous ly with experimental studies. D. Shugar started in Poland with a small group of people varying in educational background, shills and ability and age, having available only poor equipm ent and very limited finances. However, he m anaged in the first years to perform valuable studies on ribonuclease and lysozy- me establishing, for the latter, a unit of enzymatic activity, called the Shugar unit. Research was also
conducted on the mechanism of G ram staining in m icroorganism s and on histochemical localization of enzymes. The latter problem became a flourishing field of investigation in D. Shugar’s laboratories, involving synthesis of specific labeled substrates and studies on the properties and function of nucleolytic enzymes.
In 1957 he started research on the photochem ical transform ation of pyrimidines and polynucleotides in relation to mutagenesis and lethal effects, and after some time in relation to DN A repair as well. The synthesis of polynucleotides was introduced for studies on the structure of nucleic acids, their photochem istry, and also in relation to the genetic code and mechanism of protein biosynthesis. A part from these m ain sub jects, research was also pursued on the mechanism of bacterial transform ation, radiation chemistry of prote ins and nucleic acids in relation to mutagenesis and radiation damage.
In the seventies studies on antiviral and anticancer agents became the central problem. These involved new m ethods for synthesis of analogues of nuclesides and nucleotides modified in the sugar moiety, suppor ted by studies on their conform ation to establish correlations between modification, conform ation and biological activity. Later extensions of these inves tigations are studies on protein kinases and their inhibitors. O n these problems D. Shugar has been mainly engaged during recent years.
In his curriculum vitae he wrote: ”It is obvious that the im plem entation of such research program s was possible only because of collaboration with young, capable scientific research students, whose names appear on m ost of the publications. It is also obvious th at w ithout such capable coworkers, it would not have been possible for me to carry out such a p ro gram ”. Being deeply involved in scientific research he was able to work with everybody in whom he felt a spark of passion for research. The door to his room was always wide open enabling people to drop in at any time. He attracted people to his laboratories but was equally open for collaboration with other institutions. A list of D. Shugar publications includes over 450 positions (including books and reviews), the m ajority with one or several coworkers, m any from laboratories both inside and outside the country. Friendly and informal contacts with laboratories abroad always served to enrich the studies and possibilities in his laboratories.
The above does not embrace all the activities and achievements of Professor David Shugar. Despite his retirement, he continues to be an active mem ber of the research staffs at the Institute of Biochemistry and Biophysics, and at the Biophysics D epartm ent of the University. He attacks research problems with the same eagerness and sparkle in the eye as in preceding years. We hope Professor David Shugar will remain with us for further fruitful years.
Coworkers
ARTICLES
Tautomerism of thioguanine. Experimental matrix
isolation and theoretical quantum-mechanical studies
K R Y ST Y N A S Z C Z E P A N IA K
1
W ILLIS B. P E R S O N
1
JERZY L E SZ C Z Y Ń SK I
2
JÓ ZEF S. K W IA T K O W SK I
3
Contents: I. Introduction II. Methods II-1. Experimental II-2. Calculation III. Results and discussionIII-l. N1H, N7H, and NH2 stretching modes III-2. S-H stretching mode
III-3. Other vibrations III-4. Energies IV. Concluding remarks
Abbreviations used: B3LYP — Backe’s nonlocal three para meter exchange and correlation functional with Lee- -Yang-Parr correlation functional; DFT — density func tional theory; DZP — double zeta polarization; M P — Mollett-Plesset; ZPE — zero point energy.
I. Introduction
Stim ulated by fruitful and enjoyable discussions with Professor David Shugar, a group of physicists in W arsaw about 25 years ago began studies of tautom er ism of isolated pyrimidine and purine bases, first in the vapor phase and then in low tem perature inert m a trices (noble gases and nitrogen). This latter exciting technique had been brought to W arsaw University from the University of Florida, Gainesville, where one of the authors (KS) spent her postdoctoral study. M atrix isolation infrared spectroscopy appeared to be very useful for studies of vibrational spectra and tautom erism of nucleic acid bases and their analogs, particularly because it does not require the extensive heating of the sample th at is necessary in vapor studies. In this technique, the sample is prepared by sublim a tion of the solid at a tem perature m uch below its melting point into a stream of inert gas followed by rapid condensation to form the inert cold rigid m atrix [1]. The procedure avoids therm al decom position and
1 Department of Chemistry, University of Florida, Gaines ville, FL 32611, USA, 2 Department of Chemistry, Jackson State University, Jackson, MS 39217, USA, 3 Institute of Physics, N. Copernicus University, 87-100 Toruri, Poland
the isolation in the rigid m atrix prevents additional reaction and further tautom eric changes in these sensitive com pounds th at often present problem s for studies of their vapors.
In the late 70's the W arsaw group was the only one in the world doing m atrix isolation studies of the vibra tional spectra of isolated nucleic acid bases and their analogs [2-6]. These studies provided much needed experimental d ata on tautom erism of isolated py rimidine and purine bases, which are crucial for verification of the reliability of theoretical calculations. In view of this, it is not surprising that the experimental m atrix studies stim ulated m any new ab initio calcula tions of the stabilities and vibrational spectra of different tautom eric forms. By the 80's other labo ra tories also began experimental studies of vibrational spectra of these m atrix isolated pyrimidine and purine bases. At the present time several groups in different countries perform such studies for these molecules [7-12], with the W arsaw continuing to contribute to these investigations (e.g., see reference [13-17]).
O ur laboratory at the Chemistry D epartm ent of the University of Florida, joined several years ago by two members of the W arsaw group (KS and M arian Szczesniak), has collected a vast am ount of data for m any pyrimidine and purine bases and their deriva tives. The spectra of these molecules are being analyzed on the basis of better and better ab initio m olecular orbital calculations [18-31]. The progress in the com putational m ethods is so rapid that before we complete 300
http://rcin.org.pl
POSTĘPY B IO C H E M II 4 1 ( 5 ) , 1 9 9 5the analysis of our spectra using results from one calculation, newer and better calculations are com pleted, often requiring revision of the interpretation. This fact is partly responsible for the delay in publica tion of our results. But the m ain reason for this delay is the extreme difficulty of the interpretation of the complex vibrational spectra of these molecules, p a r ticularly when several tautom eric forms are present. The results discussed below illustrate some of the problem s involved.
In this paper we shall present briefly results from our recent experimental m atrix isolation studies of the vibrational spectra of tautom erism of thioguanine, together with results from some very recent calcula tions at the H artree-Fock and density functional theory levels for thioguanine tautom ers [32]. This molecule has special biological interest as a known m etabolic inhibitor with antitum or and antineoplastic activity, and is used in cancer treatm ent and research (for example, see reference [33-37], and references therein).
As m any other purine and pyrimidine bases do, thioguanine can potentially occur in several tautom eric forms. Figure 1 shows some of the tautom ers and ro tational isomers th at are possible for this molecule. At present, calculations of structures, dipole moments, energies and vibrational spectra have been performed only for the first four forms shown in figure 1.
It is known from X-ray crystallographic studies [38] th at thioguanine occurs in the crystal in the amino- thione tautom eric form with hydrogen attached to nitrogen atom N7 of the imidazole ring (structure 2 in Fig. 1), in sharp contrast with guanine, which occurs in the crystal and in solution as the am ino-oxo form with the hydrogen on the imidazole ring at N9 [39-41]. The occurrence of the (N 7H)am ino-thione form in the crystal also is contrary to the prediction of relative tautom eric stabilities by recent ab initio calculations at M P2/(full)/D ZP [42] and M P2/6-311 + G(d,p) [43] levels, which is th at the (N9H)am ino-thiol form (struc ture 3 in Fig. 1, or its rotam er with SH trans to N l) is the m ost stable.
Such differences between theoretical predictions of relative stabilities of tautom ers and the structure experimentally observed in the crystal m ay be a result of the effect of the interm olecular interactions, particu larly hydrogen bonding, in the crystal which are not considered in the calculations at this relatively high level of theory. This result is somehow puzzling since the N 7H -thione tautom er 2 found to occur in the crystal has a m uch smaller calculated dipole m om ent (1.78 D [42]) than does the N 9H -thione tautom er 1 (7.42 D [42]), thus contradicting the usual belief that the m ore polar tautom er will be stabilized in the more polar medium of the crystal. In view of these observa tions, the im portance of experimental verification of the relative stabilities of tautom ers of isolated thioguanine
(N 9H) AMIN O-THIONE (N7H)AMINO-THIONE
(N9H)AM INO-TfflOL (N7H)AMINO-THIOL
(N9H, N3H)AMINO-THIONE (N7H, N3H)AMINO-THIONE
(N9H)IMINO-THIONE (N9H, N 1 H)IMINO-THIOL Fig. 1. Some possible tautom ers for thioguanine. Structures 1-4 are
draw n for the optimized geometry from the D FT(B3LYP)/ /6-31G(d,p) calculations, which have predicted that the NH2 group is non planar. Structures 5-8 are draw n for arbitrary geometries, starting from the first four structures but with changed positions of the hydrogen atoms.
for com parison with the ab initio calculations m ade for the molecule in vacuum is quite obvious.
II. Methods
II-1. Experimental
The details of the experimental technique of m atrix isolation infrared spectroscopy have been described elsewhere [1]. In this study a m ixture of the vapor from thioguanine (purchased from Sigma) heated to about 227°C and the m atrix gas (argon) was condensed onto a cold (about 12 K) window held in the cryostat (Air Products M odel 202E Displex). The infrared spectra were m easured at a resolution of 1 c m -1 using a Fourier Transform Infrared (FTIR) Nicolet M odel 740 spectrometer. The final spectra shown below were draw n by the SpectraCalc program (Galactic) on a personal computer, after electronic transfer of the data from the Nicolet spectrom eter to the PC. This
program was also used to subtract the interference fringes caused by the thin m atrix film and to correct the background to the horizontal “zero” level.
II-2. Calculation
The calculations reported here were m ade using density functional theory (DFT) [44-45] with the Becke three param eter exchange functional and the gradient corrected functional by Lee, Yang and P arr (the DFT(B3LYP) m ethod) (see references [46-51], and references therein). The standard 6-31G(d,p) basis set was used. O ptim izations were carried out w ithout any constraints on the planarity of the species under study. All quantum m echanical calculations were performed using the G A USSIA N 92/D FT program [52] at the Mississippi Center for Supercom puting Research. Cal culations at this level have recently been the focus of a lot of attention (see reference [53], and references therein) because the calculated vibrational spectra are usually in good agreement with experiment; the agree m ent for frequencies is often within the uncertainty due to anharm onicity corrections.
O n the basis of knowledge of small molecules [54_57] it can be expected th at the correction for anharm onicity will be the largest for the X-H (where X is N l, N7, S, C or N for symmetric or asymmetric NH2) stretching frequencies, and relatively smaller corrections for other frequencies. Based on our experi ence with other molecules, we take these different anharm onicity corrections into account by scaling all N -H and S-H stretching frequencies by 0.955 (force constants by (0.955)2 or 0.912), all C-H stretching frequencies by 0.964 (force constants by (0.964)2 or 0.929), and all other frequencies by 0.973 (force con stants by (0.973)2 or 0.947) in the com parisons shown here. The use of these different scaling factors does not change the norm al coordinate mixing because it changes only force constants for stretching coordinates which do not mix with other coordinates. These scaling factors are not significantly different from those recom mended by Pulay [53] for spectra from DFT(B3LYP)/ /6-31G(d,p) calculations for a large num ber of m ol ecules. In general, the calculated frequencies after scaling agree with experimental frequencies within an uncertainty of about 10-20 cm - 1 .
The calculated spectra, structures, and atom dis placements in the norm al modes displayed here are draw n using the Animol 3.1 program (Innovative Software) installed under W indows in a PC, which reads its initial input directly from the GAUSSIAN output file.
III. Results and discussion
Usually the m ost straightforw ard identification of tautom ers can be m ade on the basis of the presence or absence in the infrared spectrum of characteristic
absorption bands for the functional groups involved in the tautom eric transitions. In the case of thioguanine these functional groups are N1H , N3H, N7H, N9H ,
asymmetric and symmetric NH 2, imino (C = N H ) and C = S stretching vibrations. F or example the m ost direct evidence of the presence of the thiol tautom er would be the appearance of the SH stretching absorp tion band near 2600 cm - 1 [17, 24]. U nfortunately the infrared intensity of this band is expected to be very low (about 1 km m o l- 1 ; see references [17, 32]) and its presence or absence is difficult to decide due to confusion with com binations or overtones which are also expected to absorb in this region.
It is equally difficult to decide about the presence or absence of the C = S stretching band (and thus about the presence of the thione tautom er) because, in contrast with the C = 0 (carbonyl) stretching mode, the C = S stretching is not characteristic. The C = S stretch ing coordinate contributes to several norm al modes and is mixed strongly in them with other coordinates (see below).
F o r these reasons the identification of the thiol and thione tautom ers can be done with m ore assurance by examining carefully the infrared absorption in other spectral regions. The analysis relies heavily on the interpretation based on the calculated spectra for the different tautom ers.
III-l. N1H, N7H, and NH2 stretching modes
One of the m ost useful hints about tautom ers present in the m atrix comes from examining the effect of short (15 minute) irradiation of the m atrix isolated sample by UY irradiation from a medium pressure m ercury lamp. As a result of this irradiation, some weak bands observed before irradiation disappear; these can be identified as absorption by am ino-thione tautom ers. Figure 2a shows a com parison of the experimental spectra in the N H stretching region from thioguanine isolated in an Ar m atrix before and after the short UV irradiation with the spectra calculated for the am ino-thiol tautom ers. As can be seen there, the strongest bands observed at 3574,3564,3489,3456 and3449 c m -1 in the spectrum (practically unchanged after irradiation) have frequencies and intensities con sistent with those calculated for the asymmetric N H 2 stretch, the N 9H (and/or the N7H ) stretch, and the symmetric N H 2 stretch, respectively, for both of the am ino-thiol tautom ers, but considerably different from the spectrum calculated for the am ino-thione forms shown in figure 2b. T hat figure shows the experimental difference spectrum obtained by subtracting the spec trum after UV irradiation from the original spectrum com pared with the spectra calculated for the two thione tautom ers shown.
The weak bands observed at 3533,3526,3463,3432, 3426, 3415, 3405, and 3399 c m -1 in the initial spec trum , and also shown in the difference spectrum, have
b
WAVENUMBER (C M -l)
W A V EN U M B ER ( C M - l )
Fig. 2. C om parison of the experimental infrared spectrum in the N -H stretching region of thioguanine isolated in an Ar m atrix with the infrared spectra from the D FT(B3LYP)/ /6-31G(d,p) calculations, for the tautom ers shown. All cal culated frequencies for the spectra shown here have been scaled by a constant factor of 0.955 to account for the anharm onicity corrections and for rem aining basis set error, a) The experimental spectrum before and after UV irradi ation com pared with calculated spectra for am ino-thiol tautom ers. b) The experimental difference spectrum ob tained by subtracting the spectrum after UV irradiation from the spectrum before the UV irradiation, com pared with the calculated spectra for am ino-thione tautom ers.
frequencies close to those calculated for the asym m et ric and symmetric N H 2 and the N 1H stretching modes in both of the am ino-thione tautom ers. However, the experimentally observed band at 3463 cm - 1 is located at lower frequency than calculated for the N 7H or N 9H stretches (but within the uncertainty of the calculation). It seems m ost likely th at the species responsible for these absorption bands are amino- thione tautomers.
O ur first interpretation of the doublet structure for the bands observed for the strong N H 2 stretching vibrations (3574, 3564; 3456 and 3449 cm -1 ) and for the weak bands 3533,3526; 3432,3426; and 3415,3405
(and/or 3399 cm -1 ) was th at it was caused by the presence in the m atrix of a m ixture of N 7H and N 9H am ino-thiol tautom ers (strong bands) with N 7H and N 9H am ino-thione tautom ers (weak bands). Argu m ents causing some reservation about this interpreta tion are: (i) the observed splitting is different th an the calculated frequency differences; and (ii) the relative intensities of the com ponents of the experimental doublets differ from one m atrix to another. F o r these reasons we rather prefer to interpret the splitting in the experimental spectrum as resulting from “m atrix ef fects” on the spectrum of one tautom er. However, it is surprising that this doublet pattern is practically insensitive to annealing, in contrast with similar obser
vations for adenine ([16] and references therein). These splittings clearly need further study.
The com parison of experimental and calculated spectra shown in figure 2 suggests th at the predominant
tautom er in the m atrix is an amino-thiol from, but it is not possible to distinguish whether it is the (N7H) or (N9H) tautom er since the calculated spectra shown in figure 2a for both forms are very similar. The weak bands listed above which disappear after UV irradi ation are m ost probably nearly all due to vibrations of an am ino-thione form. Again, the results shown in figure 2 do not allow us to determine with certainty whether it is the (N7H) or the (N9H) am ino-thione form.
III-2. S-H stretching mode
Figure 3 shows a com parison of the experimental and calculated spectra in this region (near 2600 cm -1 ). Six bands are observed in the experimental spectra, each with an absorbance about 30 to 50 times lower th at for the N H 2 asymmetric stretching mode, consist ent with the calculated ratio of intensities for these two modes (note the intensity scale). However, it is very likely th at some of the observed bands are com bination bands involving the lower frequency fundamentals, which may also be expected to have similar intensities.
POSTÇPY B IO C H E M II 4 1 ( 5 ) , 1 995 303
III-3. Other vibrations
. 0 0 8
-W A V E N U M B E R ( C M - 1 )
Fig. 3. C om parison of the experimental infrared spectrum in the S-H stretching region of thioguanine isolated in an Ar m atrix with the spectra calculated (DFT(B3LYP)/6-31G(d,p)) for the two am ino-thiol tautom ers. Some of the bands observed in the experimental spectrum may be due to com bination bands, such as: (1590+1026 = 2616), (1352+1263 = 2615), (1499 + + 1088 = 2587), (1620 + 961 = 2581), (1263 + 1314 = 2571), or (1620 + 956 = 2576) c m -1 .
Some possible com binations that could be responsible for some absorption in this region are listed in the caption of figure 3.
Because of the uncertainty about the best value of the frequency scaling factor th at should be used the S-H stretch with these D F T calculations, there is uncertainty in the value of the calculated frequency th at should be com pared in figure 3 with the experi m ental spectrum. If the errors are the same for both the N 7H and the N 9H tautom er, the calculation predicts a slightly lower S-H stretching frequency in the latter tautom er.
Because of the very low intensity of the S-H stretch ing mode, the only judgm ent concerning the presence of thiol tautom ers in the m atrix from the spectrum in this region can be support for conclusions based on inform ation from other parts of the spectrum.
As has been emphasized above, the com parison of the experimental and calculated spectra of thioguanine tautom ers shown for the N H stretching region in figure 2 does not perm it the final determ ination of whether it is the (N7H) or (N9H) form th at is actually present in the matrix. Some inform ation bearing on this question can be obtained by exam ination of the spectra in the lower frequency region (1700-400 cm -1 ). Figure 4 shows a com parison of the experimental m atrix spectra before and after UV irradiation and the difference spectrum with the spectra calculated for the four tautom ers shown there. This spectral region includes 37 of the 42 norm al modes of thioguanine, and is very difficult to analyze. Full interpretation of the experimental vibrational spectrum based on a detailed analysis of the calculated and experimental spectra will be given in reference [32]. Here we shall focus attention only on some of the most distinctive bands.
As can be seen in figure 4, the calculated spectra for the different tautom ers are not significantly different for much of this spectral region. Fortunately, even so, there are a few relatively strong bands in the spectral patterns of the (N7H) and (N9H) thiol tautom ers th at are calculated to be characteristically different. These include the bands (marked by solid triangles ▼) cal culated at 1392,1281 and, particularly, 512 cm “ 1 in the spectrum of the (N9H)amino-thiol tautom er related to the displacements of atom s for these modes, shown in figure 5, with its contribution from the N 9H group. These modes appear to correspond to the bands observed in the experimental spectra at 1395,1263 and
513 cm -1 , respectively. On the basis of this observa tion, we conclude th at the predominant amino-thiol
tautom er present in the m atrix is the (N9H) form and
not the (N7H) form. This conclusion is strongly supported by the failure to observe any absorption in the m atrix spectrum from the N 7H wagging mode calculated to absorb strongly at 408 cm -1 for the (N7H)am ino-thiol form.
It is m uch m ore difficult to distinguish between the (N7H) and (N9H) forms of the am ino-thione ta u to mers, because of the very low concentration of this form in the m atrix and consequently the very low absorbances of these bands in the experimental spectra. Some indication th at the thione tautom er in the m atrix is the (N7H) form comes from the observa tion in the experimental spectra in figure 4 of the bands at 1533, 1196, 971 and 578 cm -1 (marked by filled circles • ) which disappear after UV irradiation and are shown in the difference spectrum in figure 4b. These bands are close to the relatively intense modes at 1523, 1168, 959 and 569 c m -1 (also m arked by filled circles • in Fig. 4b) calculated for the (N7H)am ino-thione tautom er th at are related to the norm al mode atom displacements (shown in Fig. 6) with contributions 304
http://rcin.org.pl
POSTĘPY B IO C H E M II 4 1 ( 5 ) , 1 9 9 5WAVENUMBER (CM -1)
from the C = S stretch and the N 7H m otions, strongly mixed with other displacements.
In summary, our conclusion from the exam ination of figure 2 and 4 is th at the predominant tautom er present in the initial deposit of the m atrix from the vapor is the (N9H)amino-thiol form shown as structure 3 in figure 1 (or the rotam er with SH trans to N l).
A smaller concentration of the (N7H)am ino-thione tautom er 2 in figure 1 is also present.
III-4. Energies
The predom inance in the m atrix spectrum of ab sorption by tautom er 3 in figure 1 agrees with the relative stabilities predicted for the isolated molecules by the ab initio calculations [42, 43].
If we m ake the assum ption th at the analysis given above has correctly identified only two tautom eric forms present in the m atrix spectrum of thioguanine (namely the (N9H am ino-thiol predom inant form and the lesser concentration of the (N 7H)am ino-thione form), then we may use the relative intensities observed to determine the equilibrium constant for the tautom eric equilibrium at the tem perature of the gaseous m ixture as the equilibrium was trapped during the form ation of the matrix.
The ratio of the observed integrated intensity of the
Fig. 4. C om parison of the experimental infrared spectrum in the 1700-400 c m -1 region of thioguanine isolated in an Ar m atrix with the spectra calculated (DFT(B3LYP)/6- -31G(d,p)) for the tautom ers shown. All calculated frequen cies in this region have been scaled by 0.973 to account for anharm onicity and basis set error, a) The experimental spectrum for the sample before and after UV irradiation com pared with spectra calculated for two am ino-thiol tautomers. b) The experimental difference spectrum ob tained by subtracting the spectrum after UV irradiation from that before UV irradiation com pared with the calculated spectra for two am ino-thione tautom ers.
Fig. 5. Calculated displacements of the atom s, shown by the lengths of the arrows, in norm al modes for the (N9H)amino-thiol tautom er th at are used to distinguish between the (N9H)ami- no-thiol and (N 7H)amino-thiol tautom ers in the matrix spectrum. N ote th at all these norm al modes show large contributions from m otion both of the N 9H group and of the SH group (as well as contributions from m otion of several other atoms). The mode at 512 c m -1 is an out of plane “wagging” motion.
1523 CM-1 1168 CM-1
959 CM-1 569 CM-1
Fig. 6. Calculated displacem ents of atom s, shown by the lengths of arrows, in norm al modes for the (N 7H)am ino-thione that distinguish between the (N 7H)am ino-thione and (N9H)ami- no-thione forms. N ote that the N 7H bending contributes to the modes at 1523 and 959 c m “ 1, while C = S stretching and N 1H bending m otions contribute to the norm al modes at 1168 and 959 cm “ 1. However, all these modes have sizable contributions from m otion of other atom s and cannot be considered to be characteristic C = S stretches. The 569 cm “ 1 mode is mostly “out of plane N H 2 wagging”.
asymmetric N H 2 stretch at 3564 and 3574 cm - 1 in the am ino-thiol form to that of the amino thione form at 3526 and 3533 c m -1 , together with the corresponding calculated absolute integrated m olar absorption coeffi cients A (41 and 37 km m o l- 1 , respectively) [32], allows us to evaluate the relative concentration of the (N9H)amino-thiol tautom er to that of the (N7H)ami
no-thione form, trapped in the matrix. The results are th at the percentage concentration of the former is 85%, and th at of the latter is 15%. F o r an equilibrium established in the gas phase during sublim ation at 500 K (227°C) and subsequently trapped in the matrix, the difference in the standard Gibbs free energy of the m ore stable am ino-thiol form from th at for the am ino-thione form is thus A G500 = —7.1+ 1.5 kJ m o l- 1 .
The values of the change in the internal energy AE® at 0 K have been obtained from the ab initio calcula tions as the sum of the calculated electronic energy differences AEe, and zero point energy differences (AZPE). Both contributions are given in table 1 for tautom eric transitions defined there. According to the calculation, the change in the Z P E is largely respon sible for the stability of the thiol tautom er with respect to the thione.
In order to com pare the calculated energies with the experimental m easurem ent of the equilibrium constant and the associated value of A G500, it is appropriate to use the statistical therm odynam ic relations [58] to convert the latter to the experimental value of AE°, using the calculated rotational constants and frequen cies of the norm al modes. From this, AEj] = —5.6 kJ m o l- 1 .
This value is com pared in table 1 with the calculated changes in internal energy for several tautom eric reactions involving tautom ers considered here. We see there th at the experimental value of AE° ( — 5.6 kJ m o l-1 ) for the tautom eric reaction from the (N7H)am ino-thione form of structure 2 to the (N9H)am ino-thiol form of structure 3 is about 3.8 kJ m o l-1 m ore negative than calculated value ( —1.8 kJ m o l-1 ). Structure 3 of the (N9H)amino-thiol form (in which the S-H bond is cis to the N1 atom) has been assumed to be m ore stable than the rotam er with the
Table 1.
Energies in kJ m o l“ 1 for the tautom eric reaction of (N 7H)am ino-thione thioguanine tautom er to form other tautom ers, from calculations. F o r the reaction:
(N 7H)am ino-thione —► B.
Tautom er (B) Calculation Experim ental Calc.b AE el‘ A ZPE3 AEq3 AE°98b AEqc A G W A G298
(N9H)amino-thiol cis + 7.90 - 9 .7 - 1 .8 + 0.1 - 2 .5 - 5 . 6 + 1 - 7 .1 + 1
trans __d ( -2 .7r - 0 .8 - 3 . 4
(N7H)am ino-thionef 0 0 0 0 0 0 0
(N 9H)amino-thione + 11.55 - 0 .5 5 + 11.1 + 2.7 > + 5 > + 5 + 2.5 (N7H)am ino-thiol cis + 27.99 -1 0 .8 7 + 17.0 + 4.8 > + 5 > + 5
a) M P 2(full)/D Z P //H F /D Z P calculations given by L e s z c z y n s k i , reference [42].
b) Values reported to be AE298 (internal energies including Z PE and therm al energies) or A G°98 from MP2/6-311 + +G (d,p)//H F/6-31 G(d) calculations by A l h a m b r a e t a I., reference [43],
c) This work. See text. The values for (N9H)amino-thiol could be for the cis or trans rotamer. Failure to observe any trace of a tautom er indicates its equilibrium constant is less than 0.05, implying AGsoo and AEo values 5-10 kJ m o l“ 1 m ore positive than the energy of the (N7H)amino- -thione tautom er.
d) N o t calculated.
e) Estim ated from the calculated value of the energy of the cis tautom er, added to the difference calculated between the energies of cis and trans
tautom ers in b.
f> Reference molecule. All energies calculated relative to this tautom er.
SH bond trans to the N1 atom. However, recent calculations by A l h a m b r a e t a l . [12], of the energies of tautom ers of thioguanines have found that the trans form is about 0.9 kJ mol ~ 1 m ore stable than the cis form (structure 3).
If the observed m atrix spectrum is for the trans form rather than the cis form, then the observed value of AEq ( — 5.6 kJ mol ~ x) should be com pared with a calculated value for th at rotam er. This value can be estim ated for the present by correcting the value from L e s z - c z y n s k i [42] in table 1 ( —1.8 kJ m o l“ x) by adding this calculated cis-trans energy difference [43], to obtain AEq for the trans rotam er to be —2.7 kJ m o l- 1 . Such close quantitative agreement between calculated ( — 5.6) and experimental ( — 2.7 kJ m o l-1 ) energies is m uch better than can be expected, and encourages us to believe th at the conclusions about tautom erism in thioguanine from this study may be correct.
IV. Concluding remarks
The m ain conclusions from this study are sum m arized below.
(1) Thioguanine occurs in a non-polar (hydrophobic) environm ent predom inantly (about 85% for an equi librium established at 500 K) as the (N9H)amino-thiol tautom er. Only a small concentration (about 15% at 500 K) of the molecules occur as the (N7H)amino- -thione form. It is possible th at small concentrations of im ino-thione or imino-thiol tautom ers may be present in the matrix. F u rth er experimental and theoretical studies are needed to confirm (or contradict) this last suggestion.
(2) This finding for thioguanine is in sharp contrast with the exclusive occurrence of the (N7H)amino- -thione form in the polar environm ent of the crystal. It is also in contrast with the finding for guanine in the m atrix environm ent th at three tautom ers [(N7H)ami- no-oxo, (N9H)amino-oxo, and (N9H)amino-hydroxy] occur with similar concentrations [4, 59].
(3) The greater stability found in this study for the (N9H)amino-thiol tautom er in the hydrophobic envi ronm ent confirms the predictions from recent ab initio
calculations, and the experimental change in energy, AEq, for the tautom eric transition from the (N7H)am ino-thione form is in good agreement with the value from the calculation. Additional calculations at the M P2(full)DZP level of the energy of the trans
rotam er of the (N9H)am ino-thiol form, which may be the most stable tautom er of thioguanine, are in progress.
Acknowledgments
First of all, it is a pleasure to acknowledge the inspiration provided by Professor David S h u g a r , and to dedicate this report to him. Partial financial support (KS and WBP) from NIH Grant No. 32988 and from a Research Development Award from the Division of Sponsored Research of the University of Florida is gratefully acknowledged. We are grateful also to NIH Grant No. 332090, to contract (DAAL
03-89-0038) with the Army Research Office, and to the Mississippi Center for Supercomputing Research for support (JL) of the calculations. JSK wishes to acknowledge support by the Committee for Scientific Research within grant KBN 3T09A 026 08.
References
1. P e r s o n WB , S z c z e p a n i a k K, S z c z ę ś n i a k M, D e l B e n e J E (1993) In: Fausto R (ed) Recent Experim ental and Com putational Advances in M olecular Spectroscopy. N ATO ASI Series C: M athem atical and Physical Sciences, vol 406. K luw er Academic Publishers, The N etherlands, D ordrecht, pp 141-169 2. N o w a k MJ , S z c z e p a n i a k K, B a r s k i A, S h u g a r D (1978) Z. Naturforsch Teil C 33: 876-883 3. N o w a k MJ , S z c z e p a n i a k K, B a r s k i A, S h u g a r D (1980) J M ol Struct 63: 47-69 4. S h u g a r D, S z c z e p a n i a k K (1981) I nt J Quantum Chem 20: 573-585 5. S z c z ę ś n i a k M, N o w a k MJ , S z c z e p a n i a k K (1984) J M ol Struct 115: 221-224 6. S z c z e p a n i a k K, S z c z ę ś n i a k M (1986) J M ol Struct 156: 29-42 7. R a d c h e n k o E D , P l o c h o t n i c h e n k o A M, S h e i n a G G , B l a g o i Y u P (1983) Biophysics 28: 559-563 (English edn) 8. B a r n e s A J, S t u c k e y M A, L e G a l l L (1984) Spectrochim Acta 40A: 419-431 9. R a d c h e n k o E D , S h e i n a G G , S m o r y g o N A, B l a g o i Y u P (1984) J M ol Struct 116: 387-396 10. S t e p a n i a n S G, S h e i n a G G , R a d c h e n k o A D, B a 1 - g o i Y u P (1985) J M ol Struct 131: 333-346 11. S t e p a n i a n SG, R a d c h e n k o E D , S h e i n a G G , B l a g o i Y u P (1990) J M ol Struct 216: 77-90 12. G r a i n d o u r z e M, S m e t s Y, Z e e g e r s - H u y s k e n s Th , M a e s G (1990) J M ol Struct 222: 345-364 13. N o w a k M J (1989) J M ol Struct 193: 35-49 14. N o w a k MJ , Ł a p i ń s k i L, F u l a r a J (1989) Spectrochim Acta 45A: 229-242 15. F u l a r a J, N o w a k MJ , Ł a p i ń s k i L, L e s A, A d a m o w i c z L (1991) Spectrochim Acta 47A: 595-613
16. N o w a k MJ , R o s t k o w s k a H, Ł a p i ń s k i L, K w i a t k o w s k i JS, L e s z c z y ń s k i J (1994) J Phys Chem 98: 2813- 2816 17. N o w a k MJ , R o s t k o w s k a H, Ł a p i ń s k i L, L e s z c z y ń s k i J, K w i a t k o w s k i J S (1991) Spectrochim Acta 47A: 339-353 18. S z c z ę ś n i a k M, N o w a k MJ , R o s t k o w s k a H, S z c z e p a n i a k K, P e r s o n WB , S h u g a r D (1983) J Am Chem Soc 106: 5969-5976 19. C h i n S, S c o t t I, S z c z e p a n i a k K, P e r s o n W B (1984) J Am Chem Soc 106: 3415-3422 20. S z c z ę ś n i a k M, N o w a k MJ , S z c z e p a n i a k K, C h i n S, S c o t t I, P e r s o n W B (1985) Spectrochim Acta 41A: 223-235
21. K w i a t k o w s k i J S, P e r s o n WB , S z c z e p a n i a k K, S z c z ę ś n i a k M (1987) Acta Biochim Polon 34: 165-181 22. S z c z ę ś n i a k M, S z c z e p a n i a k K, K w i a t k o w s k i JS,
K u B u l a t K, P e r s o n W B (1988) J Am Chem Soc 110: 8319-8330
23. S z c z e p a n i a k K, S z c z ę ś n i a k M, P e r s o n W B (1988)
Chem Phys L ett 153: 39-44
24. R o s t k o w s k a H, B a r s k i A, S z c z e p a n i a k K, S z c z ę ś n i a k M, P e r s o n W B (1988) J M ol Struct 176: 137-147 25. P e r s o n WB, S z c z e p a n i a k K, S z c z ę ś n i a k M, K w i a t k o w s k i JS, H e r n a n d e z L, C z e r m i ń s k i R (1989) J M ol Struct 194: 239-258 26. W o j c i k MJ , R o s t k o w s k a H, S z c z e p a n i a k K, P e r s o n W B (1989) Spectrochim Acta 45A: 499-502 27. J a w o r s k i A, S z c z ę ś n i a k M, S z c z e p a n i a k K, K u B u l a t K, P e r s o n W B (1990) J M ol Struct 223: 63-92 28. R o s t k o w s k a H, S z c z e p a n i a k K, N o w a k MJ , L e s z c z y ń s k i J, K u B u l a t K, P e r s o n W B (1990) J Am Chem Soc 112: 2147-2160 POSTĘPY B IO C H E M II 4 1 ( 5 ) , 1995
http://rcin.org.pl
30729. S z c z e p a n i a k K. S z c z ę ś n i a k M, S z a j d a W, P e r s o n W B (1991) Can J Chem 69: 1705-1720
30. S z c z ę ś n i a k M, L e s z c z y ń s k i J, P e r s o n W B (1992)
J Am Chem Soc 114: 2731-2377
31. P e r s o n W B, S z c z e p a n i a k K (1993) In: D urig JR (ed) Vibrational Spectra and Structure, vol 20. Elsevier, The N ether lands, Amsterdam, pp 239-325
32. S z c z e p a n i a k K, K ä s e n d e O, P e r s o n W B, L e s z c z y ń s k i J, K w i a t k o w s k i J S (Full interpretation of the vibrational spectrum of thioguanine; m anuscript in preparation for J Am Chem Soc)
33. H u n t e r CJ , D e e n D F , M a r t o n L J (1989) In t J Cancer 44: 658-660 34. P i e t e r s R, H u i s m a n s D R , L o o n e n A H , H a h l e n K, V e e r m a n A J P (1991) Jpn J Cancer Res 82: 1051-1055 35. H u n t e r CJ , D e e n D F , M a r t o n L J (1989) Int J Cancer 44: 658-660 36. P h i l l i p s M D , N a s c i m b e n i B, T i c e R R, S h e l b y M D (1991) Environ M ol M utagen 18: 168-183 37. P i e t e r s R , H u i s m a n s D R , L o o n e n A H , P e t e r s GJ , H a h l e n K, V a n d e r D o e s - v a n d e n B e r g A, V a n W e r i n g E R , V e e r m a n A J P (1992) In t J Cancer 51: 213-217 38. B u g g C E , T h e w a l d U (1970) J Am Chem Soc 92: 7441-7745 39. T h e w a l d U, B u g g C E, M a r s h R E (1971) Acta Cryst B27: 2358-2363 40. M a j o u b e M (1984) J Chim Phys 81: 303-315
41. S h u g a r D, P s o d a A (1990) In: Nucleic Acids. Landolt- Bernstein New Series of Biophysics, P a rt I. Springer Verlag, Berlin
42. L e s z c z y ń s k i J (1994) J M ol Struct (Theochem) 311: 37-44 43. A l h a m b r a C, L u q u e F J , E s t e l r i c h J, O r o z c o
M (1995) J Org Chem 60: 969-976
44. P a r r R G , Y a n g W (1989) In: Density Functional Theory of Atoms and Molecules. Oxford University Press, New Y ork 45. L a b a n o w s k i J K , A n z e l m J W (eds) (1991) In: Density
Functional M ethods in Chemistry. Springer Verlag, New York 46. B e c k e A D (1992) J Chem Phys 96: 2155-2160 47. B e c k e A D (1992) J Chem Phys 97: 9173-9177 48. B e c k e A D (1993) J Chem Phys 98: 5648-5652 49. L e e C, Y a n g W, P a r r R G (1988) Phys Rev B41: 785-789 50. V o s k o S H, W i l k L, N u s a i r M (1980) Can J Phys 58: 1200-1211 51. Z i e g l e r T (1991) Chem Rev 91: 651-667 52. F r i s h MJ , T r u e s G W , S c h l e g e l H B, G i l l P M W, J o h n s o n BG, W o n g M W , F o r e s m a n J B, R o b b MA , H e a d - G o r d o n M, R e p l o g l e ES, G o m b e r t s R, A n d r e s J L, R a g h a v a c h a r i K, B i n k l e y J S, G o n z a l e s C, M a r t i n RL, F o x DJ , D e f r e e s D J, B a k e r J, S t e w a r t J J P , P o p l e J A (1993) In: G aussian 92/D FT Revision F.2, G aussian, Inc. Pittsburg, PA
53. R a u h u t G, P u l a y P (1995) J Phys Chem 99: 3093-3100 54. H e r z b e r g G (1950) In: M olecular Spectra and M olecular
Structure, vol 1. Van N ostrand Reinhold Co, New York, pp 502-580
55. M i l l s I M (1975) In: Specialist Periodical Reports of the Chemical Society, Theoretical Chemistry, vol 1 pp 110-159 56. G r e e n W H Jr , W i l l e t t s A, J a y a t i l a k a D, H a n d y
N C (1990) Chem Phys L ett 169: 127-137 57. D u n c a n J L (1991) Spectrochim Acta 47A: 1-27
58. A t k i n s P W (1994) Physical Chemistry, 5th edn. Freem an, New York, chap. 20, pp 691-716; particularly E quation 33 59. S z c z e p a n i a k K, S z c z e s n i a k M, S z a j d a W, P e r
s o n WB, L e s z c z y n s k i J, K w i a t k o w s k i J S (Tautom erism of guanine; m anuscript in preparation for J Am Chem Soc)
This work is dedicated to Professor David S h u g a r my teacher and best friend
Mechanisms of action of methyl methanesulfonate on
Escherichia coli: Mutagenesis, D N A damage and repair
C E L IN A JA NIO N *
Contents:
I. DNA repair processes and cancer induction
II. MMS-induced lesions and defence systems against alkylating agents
III. Mechanisms of the induction of the adaptive response IV. Specificity of MMS-induced mutations
V. Mutation frequency decline (MFD)
VI. Does MFD operate after MMS-treatment? VII. Conclusions
Abbrevations used: XP — xeroderma pigmentosum, FA — Fanconi’s anemia; BS — Bloom’s syndrome; TFH II — pro tein subunits of transcription factor in human, required for initiation of transcription of genes and nucleotide excision repair; HNPCC — hereditary nonpolyposis colorectal can cer; MMS — methyl methanesulfonate; M NNG — N
-methyl-A'-nitro-A-nitrosoguanidine; MNU —
iV-methyl-* Institute of Biochemistry and Biophysics, Polish Acad emy of Sciences, A. Pawinskiego 5 A, 02-106 Warsaw, Poland
A-nitrosourea; M FD — mutation frequency decline; TRCF — transcription repair coupling factor.
I. D N A repair processes and cancer induction
Recent studies have provided astonishing d ata on the role of D N A repair in avoidance of cancer. It has long been known that a deficiency, or disturbance, in 308http://rcin.org.pl
POSTĘPY B IO C H E M II 4 1 ( 5 ) , 1 9 9 5D N A repair may lead to genetic instability, and a predisposition to cancer, in patients with xeroderm a pigm entosum (XP), Fanconi’s anem ia (FA), Bloom’s syndrom e (BS); and to such hereditary neurological and neurom uscular diseases as H ungtinton’s or Cockayne’s syndrome. The origin of these diseases is a deficiency in some of the systems of D N A repair: nucleotide excision repair, DNA-ligase, preferential nucleotide excision repair (hum an gene ERC C6 corre sponds to mfd of E. coli), in T F H II subunit required for initiation of gene transcription and nucleotide excision repair [1-3]. Quite recently it was discovered that hereditary nonpolyposis colorectal cancer (H N PCC) in hum ans develops because of m utation in hM SH 2, h M L H l, h P M S H l, or hP M SH 2 genes and deficiency in D N A mismatch repair [4-8].
All these DN A repair systems were first discovered and well recognized in bacteria, then in yeast and/o r in hum an cells.
In the last decade, m uch attention was focused on the p53-human gene encoding the p53 protein. A m u ta ted form of this protein was found in about 50% of all hum an cancer tissues [9,10]. In the norm al tissues, p53 protein is undetectable.
Protein p53 is a transcription factor, anti-oncogene product, suppressor of cancer, and a genomic guardian of the hum an cells. The synthesis of p53 is much increased after damage to D N A (e.g. by ionizing radiation) and this stimulates synthesis of the other proteins involved in arrest of cell cycle, D N A synthesis and repair [11-13].
The cascade of reactions which occurs after DN A damage in hum an cells resembles th at operating in E. coli where it is know n as the SOS system. The hypothesis th at m alignancy develops because the cells cannot scope with D N A repair, in 80-90% of cancer cases, seems to be true [14].
Organisms are continuously exposed to DN A dam age from insults of endogenous and exogenous (envi ronm ental) mutagens. M ethylating agents are wide spread in the environm ent. In this paper mechanisms of damage and D N A repair, the systems defending E. coli cells against the killing and m utagenic effects of methyl m ethanesulfonate (MMS), will be briefly re viewed.
II. MMS-induced lesions and defence systems
against alkylating agents
M M S is a m onofunctional agent which methylates bases predom inantly at the nitrogens (SN2 type of reaction, in contrast to SNl-type acting reagents al kylating to greater extent the oxygens) in DNA. As a result, about 82% of 7meG (non-m utagenic), 10.8% of 3meA and 0.6% of 3meG (mutagenic and highly toxic) and 0.3% of 0 6m eG (highly mutagenic) is formed [15]. The presence of 7meG and O em eG in D N A does not arrest D N A synthesis (although
0 6meG, with methyl group in syn conform ations, may slow it down [16]); however, 0 6m eG (but not 7meG) may occasionally m ispair with T, and lead to m uta tions of G C -»A T transitions [17, 18]. 3meA and apurinic sites, which may be formed by excision of 3meA, 3meG or 7meG, stop synthesis of DN A and may, in a wmwDC-dependent m anner, induce AT->TA or G C -»T A transversions [19-22].
So far, two constitutively expressed proteins, defend ing the cells against the chemical m éthylation of DNA, are known: (i) ogt-cncoded 0 6m ethylguanine-D NA methyltransferase, able to remove methyl, or (with a much lower efficiency) ethyl group from the 0 6alkylG in D N A [23, 24], and (ii) fug-encoded 3-methyladenine DN A glycosylase I, excising 3- m ethyladenine (3meA) and leaving the apurinic sites in DN A [25]. O em eG and apurinic sites m ay be excised by UvrABC-excinuclease known to remove m any of the base adducts from DN A [26, 27]. The level of the UvrA and UvrB proteins much increases after SOS induction.
Alkylating agents may activate two inducible DN A repair systems: (i) the SOS system [28], including about 20 genes involved in DN A mutagenesis (umuDC) and repair, and (ii) the adaptive response, including four genes at least, triggered by low, weakly mutagenic, doses of m ethylating agents [29]. The cellular signal to induce the SOS system is damage to DN A and arrest of D N A synthesis [30, 31]. F or the alkylating agents, the prim ary signal is the presence of 3meA (or 3etA) in DNA. Ability to induce expression from the sfi::lacZ
fusion (sfi belongs to the SOS-regulated system) occurs at 5-50-fold lower doses of the alkylanes in bacteria devoid of both constitutive, tag A, and inducible
alkA-encoded, 3meA DN A glycosylases, than in the wild type cells [32].
The prim ary signal for inducing the adaptive re sponse is m éthylation of the Ada-protein at Cys-69. Ethylating agents are poor inducers of the adaptive response. Ada protein contains two (of the twelve) cysteines able to accept methyl groups: Cys-69, and Cys-321. Ada-me-Cys-69 (or Ada-me-Cys-69, me-Cys- -321) protein is a potent positive inducer of the ada
regulon, including its own and three other genes, at least: u//cA-encoding 3meA DN A glycosylase II, and
alkB- and aidC- encoding proteins with as yet un known functions [33-36]. The 39 kD a Ada protein itself, and 19 kD a carboxyl-term inal fragm ent resulting from Ada proteolysis (with O m pt protein [37]), has the activity of a O ealkyl-DNA alkyl transferase able to accept alkyl group from 0 6-alkylguanine (and O 4- methylthymine) in DN A on their Cys-321 residue. Ability of ada to demethylate 0 6m eG is 7-fold higher than that to deethylate of O eetG. Both, Ada protein, as well as its 19-kDa fragment, which structurally and functionally resembles of ogi-encoded protein, are not in sensu stricto enzymes, and after alkyl acceptation are not able to recover their activity again; i.e. they are
“suicide proteins”.
Constitutively expressed 19-kDa Ogt and inducible, derived of Ada, 19 kD a 0 6m eG -D N A m ethyltran- sferase, differ in theirs affinities. Both, ogt- and
ada-encoded alkyl transferases remove alkyl groups (with decreasing efficiency) from O emeG, 0 6etG and 0 4meT in DNA; but the ability of Ogt protein to remove ethyl group from 0 6etG is 145-fold greater, and to remove methyl from 0 4meT, is 7-fold greater, than for Ada alkyltransferase [38].
M am m alian tissues contain non-inducible 0 6meG- -DNA methyltransferase, with its activity resembling of that of the bacterial Ogt protein [39]. Expression of O gt protein in m am m alian cells following transform ation is frequently suppressed. H um an cell lines devoid of O em eG -D N A m ethyltransferase activity (M er- phenotype), are highly sensitive to killing effect by M N N G and M N U [40].
Regulated by m ethylated, or not m ethylated, Ada- protein [41] is a/L4-encoded 31-kDa 3meA DN A glycosylase II. This glycosylase, is m uch less specific than 21-kD a 3meA glycosylase I, and apart from 3meA, excises a wide spectrum of modified nitrogen bases: 3meG, N 7m eG [42], 0 2meT, 0 2meC [43], N 2, 3ethenoG [44], and possibly, 3,N4ethenoC [45]. Re cently, it has been found th at 3-alkyl D N A glycosylase II may excise the oxidative-damaged pyrimidines: 5-formylU with the same rate as 3meA, and 5-hydro- xymethylU with a rate of 1-3 order of m agnitude slower [46]. Therefore the alkA gene may function in the repair of oxidative-damaged DN A lesions, occur ring as a consequence of aerobic m etabolism of cells, which in hum ans may be a cause of aging and cancer [47].
There is an interesting relationship between the
tag-and al/c-encoded 3meA D N A glycosylases. After induc tion of the Aik enzyme, the level of Tag enzyme decreases in the cells. O verproduction of alk DN A glycosylase in the wild type cells rendered them (most probably because form ation of too m any apurinic sites) highly sensitive to M M S. However, overpro duction of tag DN A glycosylase, alm ost completely suppresses sensitivity to m ethylating agents of alk~
cells [48].
Recently, it has been found th at fag-encoded enzyme may, albeit with 70-fold lower efficiency than
alk--encoded enzyme, remove 3meG from DNA. This explains why an alkA m utation renders the cells much m ore sensitive to M M S then a tag m utation, and implies th at 3meG is a highly cytotoxic lesion [49].
III. Mechanisms of the induction of the
adaptive response
The mechanism of M M S-induced mutagenesis is closely related to the ability of the cells to repair prem utagenic lesions. Induction of the adaptive re sponse in E. coli appears to be a highly specific process.
Expression of the Ada-protein, and proteins involved in the ada regulon, begins when Cys-69 become m ethylated. The m ethylating substrate is only one of the two stereoisomers (Sp) of m ethylphosphotriesters in D N A [36]. However, Cys-69 in Ada may be directly m ethylated by the m ethylating agents, and the ad ap tive response may be triggered w ithout a contribution of m ethylated DN A [50]. Therefore one can guess th at the ways of triggering the adaptive response directly, or
via methyl phosphotriesters in DNA, should depend on the ability of the m ethylating agents to m ethylate phosphodiesters in D N A and /o r Cys-69 in the Ada protein.
It has been also shown th at O em eG -D N A methyl- transferases, both from m am m alian and E. coli cells, may be inactivated by a direct action of the m e thylating agents [50-52]. 0 6m eG -D N A m ethyltran- sferases may be directly inactivated by a much lower concentration of M M S, that by M M S-m ethylated DNA. Among the m ethylating agents are those which inactivate 0 6m eG -D N A m ethyltransferase directly at a lower (SN2 type reagent, including MM S) at a similar (M NNG), or at a m uch higher (M NU) concentration than by M M S-(M N N G or M N U ) m ethylated DN A [50, 51].
Thus M M S is a direct inducer of the adaptive response, and an agent inactivating (or a substrate for?) O em eG -D N A methyltransferase. One can conclude, therefore, that: (i) Cys-69 in Ada protein is a better acceptor for methyl group from M M S, than phospho- triester (and perhaps the nitrogen bases) in DN A and, (ii) the adaptive response in M M S-treated cells is induced earlier, before D N A can be damaged. That can explain the low m utagenic potency of M M S as com pared to other alkylating agents. The enzymes involved in repair of prem utagenic lesions, ada-encoded O em eG -D N A m ethyltransferase and a/L4-encoded 3meA-DNA glycosylase II, are induced before DN A becomes m ethylated, and the cells can remove M M S- induced lesions in D N A very efficiently. Accordingly to above, we did not observe an essential decrease in M M S-induced m utations, when cells were pretreated, before M M S m utagenic treatm ent, with the low doses of M N N G to induce the adaptive response (unpub lished results and [53]).
The different reactivity of a m ethylating agent with DN A and with Cys-69 of Ada protein m ust be reflected in its m utagenic specificity and killing potency. There fore, after M M S-treatm ent of E. coli, G C ->A T tra n sitions in DNA, due to efficient O emeG repair are happened rarely, and induction of Alk enzyme leads to a great increase in the apurinic sites which, when U m uD ' U m uC proteins are available, may result in AT->TA [21, 22, 54] or G C -»T A [19] transversions.
IV. Specificity of MMS-induced mutations
M M S is regarded as a low efficient, and largely 310