Professional Documents
Culture Documents
Dna Repair and Mutagenesis
Dna Repair and Mutagenesis
DNA Repair
journal homepage: www.elsevier.com/locate/dnarepair
Historical Perspective
a b s t r a c t
This article reviews the early history of the discovery of an DNA repair pathway designated as base excision repair (BER), since in contrast to the
enzyme-catalyzed removal of damaged bases from DNA as nucleotides [called nucleotide excision repair (NER)], BER involves the removal of damaged or
inappropriate bases, such as the presence of uracil instead of thymine, from DNA as free bases.
© 2015 Published by Elsevier B.V.
1. Introduction
http://dx.doi.org/10.1016/j.dnarep.2015.12.003
1568-7864/© 2015 Published by Elsevier B.V.
A36 E.C. Friedberg / DNA Repair 37 (2016) A35–A39
containing DNA, which I tacitly assumed was a nuclease. Members Degradation of Uracil-containing Deoxyribonucleic acid by E. coli
of my laboratory were then fully engaged in other experiments. Deoxyribonucleic Acid Polymerase I [13]. These studies provide a
However, I was ultimately able to persuade one of them to put perfect example of the well-worn aphorism attributed to Arthur
aside his experiments for a while and address the first question. Kornberg: “don’t waste clean thoughts on dirty enzymes.”
It was soon revealed that very soon after infecting B. subtilis The discovery of uracil-DNA glycosylase opened the door to the
phage PBS2 expresses an inhibitor of the enzyme that attacks uracil- discovery of multiple other DNA glycosylases. By the mid-1990s
containing DNA. In later years my colleagues and I characterized close to a dozen such enzymes were known to exist. Surprisingly,
this inhibitor [10], which subsequently became a useful reagent for the catalogue of known DNA glycosylases includes the bacterio-
specifically inhibiting uracil-DNA glycosylase in vitro. While in the phage T4 pyrimidine dimer-specific enzyme mentioned above, as
early stages of experiments to address the second issue of interest – well as a very similar enzyme in the bacterium M. luteus, discov-
purifying and characterizing the putative nuclease in B. subtilis that ered earlier by the late Larry Grossman. I was originally confident
attacks uracil-containing DNA, I encountered Lindahl’s paper in the that my colleagues and I had discovered a new nucleotide excision
September 1974 issue of the PNAS, announcing the discovery of a repair pathway and that the phage T4-encoded enzyme would be
enzyme in E. coli that catalyzes the release of free uracil from DNA. shown to catalyze the endonucleolytic incision of DNA close to sites
I soon established that the properties of the B. subtilis enzyme were of pyrimidine dimers, thereby setting the stage for subsequent exci-
essentially identical to those of the uracil-DNA glycosylase discov- sion of pyrimidine dimers. However, to my surprise the enzyme was
ered by Lindahl [11]. Since PBS2 DNA contains U-A base pairs our shown to catalyze cleavage of the glycosyl bond of the 5 thymine in
experiments providentially revealed that the substrate specificity thymine dimers in DNA, generating an apyrimidinic site in the DNA,
of the E. coli enzyme discovered by Lindahl was not dependent on while the two thymine residues remained dimerized [14]. Identical
the presence of U-G mispairs. Lindahl later demonstrated that the results were obtained by Grossman and his colleagues with respect
enzyme even removes uracil from single-stranded DNA. to the enzyme from M. luteus.
History will of course never reveal whether or not absent Lin-
dahl’s publication in September 1974 my colleagues and I would 3. The discovery of apurinic-apyrimidinic endonucleases
have independently discovered uracil-DNA glycosylase, though I and lyases
would like to think that such an outcome would have emerged in
due course! But the critical reality is that focused as he was on a The catalytic action of a DNA glycosylase leaves sites of
search for an enzyme activity that removes uracil from DNA, Lin- base loss in DNA. Hence, this class of enzymes essentially
dahl astutely established an assay that included the possibility that trades one type of DNA damage for another. The completion
uracil might be released from DNA as the free base rather than as a of BER initiated by a DNA glycosylase requires incision of the
nucleotide or part of oligonucleotide fragments, as had been shown sugar-phosphate backbone at the apurinic/apyrimidinic (AP) site.
by Setlow to be the case with thymine dimers during NER. My Two classes of enzymes are known to achieve this hydroly-
own focus was on the extensive degradation of phage PBS2 uracil- sis, designated as apurinic/apyrimidinic (AP) endonucleases and
containing DNA when it infected B. subtilis, the products of which apurinic/apyrimidinic (AP) lyases.
would be readily observed by measuring radioactivity in the acid- The first AP endonuclease, designated endonuclease II of E. coli
soluble fraction of reaction mixtures that included DNA containing was unwittingly discovered when I was a post-doctoral fellow in
radiolabeled uracil. I had no immediate cause to consider the pos- the late David Goldthwait’s laboratory at Western Reserve Uni-
sibility that uracil might be present in the acid-soluble fraction as versity from 1966 to 1968 [15,16]. I use the word “unwittingly”
the free base. advisedly since when I joined Goldthwait’s laboratory the research
In a book entitled Correcting the Blueprint of Life – An Historical project that he assigned to me was to go in search of anticipated
Account of the Discovery of DNA Repair Mechanisms, published in endonucleases involved in genetic recombination.
1997 I recounted fragments of a discussion between Francis Crick As detailed in an autobiographical article published in this
and the science historian Horace Judson, author of The Eighth Day journal a few years ago [17] Goldthwait (Fig. 3) believed that
of Creation – Makers of the Revolution in Biology, on the nature of endonucleases involved in genetic recombination would cut DNA
discoveries. Judson offered the comment to Crick that “discovery, in a highly specific manner and was interested in establishing the
examined closely – seemed curiously difficult to pin to a moment molecular basis of this specificity. In a paper published in the Cold
or to an insight, or even a single person.” To which Crick responded: Spring Harbor Laboratory Symposia in Quantitative Biology in 1968,
he and I wrote:
“I think that’s the nature of discoveries, many times: that the
reason they’re difficult to make is that you’ve got to take a series A model – that has not been formally excluded requires that
of steps, three or four steps, which if you don’t make them you homologous pairing of two DNA molecules is a primary event
won’t get there, and if you go wrong in any one of them you in recombination. This could conceivably occur through spe-
won’t get there. It isn’t a matter of one jump – that would be cific protein protein–DNA interactions that produce localized
easy. You’ve got to make several successive jumps. And usually disruption of normal base pairing in the 2 molecules. Such inter-
the pennies drop one after another until it all clicks. Otherwise molecular association would result in an area of distortion of the
it would be too easy! [12].
Other investigators encountered what was likely uracil-DNA
glycosylase activity while pursuing unrelated questions. As early
as 1958 Arthur Kornberg and his colleagues had shown that the
DNA polymerase he discovered could utilize dUTP as a precursor for
DNA synthesis in vitro. When Huber Warner and Merle Wovcha at
the University of Minnesota synthesized DNA with “purified” E. coli
DNA polymerase in the presence of dUTP they observed nicking
of the DNA at sites of uracil incorporation and erroneously con-
cluded that the 5 → 3 exonuclease activity of DNA polymerase
I possessed an intrinsic uracil-specific endonuclease. They pub-
lished these results in a paper entitled: Synthesis and Nucleolytic Fig. 3.
A38 E.C. Friedberg / DNA Repair 37 (2016) A35–A39
normal secondary structure of the DNA molecules at the region pyrimidine dimer-DNA glycosylases and their associated AP lyase
of their pairing. An enzyme capable of recognizing this distor- activities generate incisions of the DNA that allows for subsequent
tion and restricted so that it could cleave only a single-strand of excision of oligonucleotides that include the pyrimidine dimers.
DNA oriented in one direction (either 3 → 5 or vice versa) could
make the appropriate single-strand nicks to facilitate some sort
5. Excision of the 5 deoxyribose residue
of intermolecular exchange – [18].
Goldthwait suggested that I attempt to generate localized con- Following endonucleolytic incision of DNA by an AP endonucle-
formational distortions of the DNA structure in vitro by treatment ase the lingering 5 deoxyribose-phosphate residue must be excised
of the DNA with the alkylating DNA methylmethane sulfonate. I to generate a single nucleotide gap. A survey of known excision
was then too biochemically naive to debate with my mentor that activities in E. coli surfaced a 3 → 5 exonuclease active on sin-
alkylation might not generate any sort of distortion of the DNA gle stranded DNA by Richard Kolodner and Susan Lovett that is
structure! Duly following Goldthwait’s advice I generated alkylated encoded by the recJ gene, that can legitimately be referred to as a
and differentially radiolabeled DNA molecules so that recombina- deoxyribophosphodiesterase [27].
tion intermediates could be detected by sedimentation in sucrose
gradients, and searched for an enzyme activity that preferen-
tially attacked this substrate. I never recovered recombinational 6. Reconstitution of BER in vitro
intermediates! However, I discovered a novel endonuclease that
preferentially attacked alkylated DNA, which we dubbed endonu- In 1994 Lindahl and Grigory Dianov reported the reconstitution
clease II (since it was the second E. coli endonuclease shown to of BER of uracil in DNA in vitro using uracil-DNA glycosylase, the
attack DNA). In subsequent studies Goldthwait and post-doctoral AP lyase activity of endonuclease II, the Rec J exonuclease men-
fellow Sheikh Mumtaz Hadi demonstrated that the preferred sub- tioned above, DNA polymerase I (to fill in the single nucleotide
strate for endonuclease II was in fact depurinated sites in DNA gap), and DNA ligase to restore covalent integrity of the repaired
presumably generated by enzymatic removal of alkylated bases DNA [28]. Dianov and Lindahl also established a second pathway for
[19]. Hence, endonuclease II of E. coli is historically credited with BER in vitro which is initiated in the same manner as that described
being the first AP endonuclease to be discovered. above, except that following AP endonucleolytic nicking adjacent
In 1976 Bernard Weiss, then at the Johns Hopkins University, to the AP site subsequent strand displacement and DNA resynthesis
demonstrated that the properties of a highly purified preparation transpires as in NER and generates a longer repair patch.
of a well-characterized enzyme in E. coli called exonuclease III were BER is now firmly established as a fundamental DNA repair pro-
identical to those of endonuclease II. Hence, endonucleases II and cess essential to the viability of all living cells and conceivably
exonuclease III of E. coli are almost certainly one and the same evolved earlier than NER. One would also anticipate that BER is
enzyme. Weiss postulated that a single catalytic site in the enzyme a prominent DNA repair mode in thermophiles, in which the fre-
is able to recognize interstrand spaces in DNA generated either quency of spontaneous DNA damage is presumably significantly
by depurination or by spontaneous terminal unwinding of double higher than in organisms that grow at considerably reduced tem-
stranded DNA [20]. peratures. This author finds it remarkable that the literature on DNA
repair in thermophilic organisms is sparse.
4. Apurinic/apyrimidinic lyases
7. Base insertion repair?
Most DNA glycosylases are devoid of other catalytic activities.
As mentioned above, following release of the damaged or inappro- By way of concluding comments, one might anticipate that the
priate base (such as uracil) incision of the DNA sugar-phosphate reverse reaction of that catalyzed by DNA glycosylases, i.e., the
backbone is effected by a physically distinct AP endonuclease. direct insertion of a correct base into an AP site might exist in nature.
However, the incubation of some DNA glycosylases (the phage T4 In 1979 Zvi Livneh, Dov Elad and Joseph Sperling reported the iden-
and M. luteus pyrimidine dimer DNA glycosylases being primary tification of an activity in extracts of E. coli that they claimed directly
examples) with substrate DNA is associated with the breakage of inserts adenine or guanine into the appropriate apurinic sites in
phosphodiester bonds in vitro. In these instances DNA incision has double-stranded DNA from the corresponding deoxyribonucleo-
been shown to occur by a  elimination reaction 3 to the AP site side triphosphates as the purine donor. [29] They designated this
and enzymes that catalyze such reactions are termed DNA lyases. putative DNA repair mode as “base insertion repair.”
The first published indication of AP lyase activity (in 1985) was When I spent a sabbatical period in Tomas Lindahl’s laboratory
by Germaine Grondal-Zocchi and Walter Verly [21], who suggested in 1979–80 it was with the specific intention of reproducing the
that an AP endonuclease from E. coli designated as endonuclease IV results reported by Livneh et al. – to no avail. Additionally, in 1982
attacks AP sites in DNA by a  elimination mechanism, the chemi- Kataoka and Sukiguchi published a paper in which they demon-
cal details of which are not described here. A few years later Yoke strated that the incorporation of purines into depurinated DNA can
Kow and Susan Wallace [22] and Veronique Bailly and Walter Verly be ascribed to the combined action of an AP endonuclease and DNA
[23] independently demonstrated that another AP endonuclease polymerase I. Their abstract states: “An enzyme activity to incor-
in E. coli called endonuclease III (which is a DNA glycosylase that porate labeled dATP or dGTP preferentially into depurinated DNA,
removes ring-saturated or fragmented pyrimidines from DNA) is found in an extract of E. coli, does not seem to represent ‘purine
also possessed with the ability to promote cutting of the sugar- insertase’, but may be ascribed to a combined action of DNA poly-
phosphate backbone by a  elimination mechanism. In late 1989 merase I and AP endonuclease(s) on the basis of the following
Bailly and Verly formally suggested that “AP endonucleases” that findings. (1) The activity was found in polA+ but not in polA− cells.
promote scission of the sugar-phosphate backbone of DNA by a  (2) The base moiety and the ␣-phosphate group of the nucleotide
elimination mechanism be designated as DNA lyases [24]. About were equally incorporated into the DNA” [30]. In 1984 Bernard Mar-
that time a DNA lyase activity of the phage T4 pyrimidine dimer tin and Nicole Sicard at the Centre de Recherche et de Geneticique
DNA glycosylase was reported in papers from the collaborative Cellullulaires in Toulouse, France also published a paper in which
laboratories of John Gerit, Lois Rabow, JoAnne Stubb and Philip they reported the failure to observe direct insertion of purines into
Bolton [25,26]. The sequential activity of the phage T4 (or M. luteus) AP sites [31].
E.C. Friedberg / DNA Repair 37 (2016) A35–A39 A39
Acknowledgments [14] E.H. Radany, E.C. Friedberg, A pyrimidine dimer-DNA glycosylase activity
associated with the V gene product of bacteriophage T4, Nature 286 (1980)
182–185.
I thank Sam Wilson and Susan Wallace for valuable feedback [15] E.C. Friedberg, D.A. Goldthwait, Endonuclease II of E. coli. I. Isolation and purifi-
and advice. cation, Proc. Natl. Acad. Sci. 62 (1969) 934–940.
[16] E.C. Friedberg, S.M. Hadi, D.A. Goldthwait, Endonuclease II of E. coli, II. Enzyme
properties and studies on the degradation of alkylated and native deoxyribonu-
References cleic acid, J. Biol. Chem. 244 (1969) 5879–5889.
[17] E.C. Friedberg, A life of fixing DNA – an autobiographical sketch, DNA Repair 12
[1] E.C. Friedberg, Nucleotide excision repair of DNA: the very early history, DNA (2013) 389–393.
Repair 10 (2011) 668–672. [18] E.C. Friedberg, D.A. Goldthwait, Endonuclease II of E. coli, Cold Spring Harbor
[2] J. Duncan, L. Hamilton, E.C. Friedberg, The enzymatic degradation of uracil- Symp. Quant. Biol. 33 (1968) 271–275.
containing DNA. II. Evidence for N-glycosidase and nuclease activities in [19] S.-M. Hadi, D.A. Goldthwait, Endonuclease II of Escherichia coli. Degradation of
unfractionated extracts of B. subtilis, J. Virol. 19 (1976) 338–345. partially depurinated DNA, Biochemistry 10 (1971) 4986–4994.
[3] T. Lindahl, Instability and decay of the primary structure of DNA, Nature 362 [20] B. Weiss, Endonuclease II of Escherichia coli is exonuclease III, J. Biol. Chem. 251
(1993) 709–715. (1976) 1896–1901.
[4] T. Lindahl, An N-glycosidase from Escherichia coli that releases free uracil from [21] G. Grondal-Zocchi, W.G. Verly, Deoxyribonuclease IV from rat liver chromatin
DNA containing deaminated cytosine residues, Proc. Natl. Acad. Sci. U. S. A. 71 and the excision of apurinic sites from depurinated DNA, Biochem. J. 225 (1985)
(1974) 3649–3653. 535–542.
[5] E.C. Friedberg, Correcting the Blueprint of Life – An Historical Account of the [22] Y.W. Kow, S.S. Wallace, Mechanism of action of Escherichia coli endonuclease
Discovery of DNA Repair Mechanisms, Cold Spring Harbor Laboratory Press, III, Biochemistry 26 (1987) 8200–8206.
1997, pp. 121–122. [23] V. Bailly, W.G. Verly, Escherichia coli endonuclease III is not an endonuclease
[6] E.C. Friedberg, J.J. King, Endonucleolytic cleavage of UV-irradiated DNA con- but a -elimination catalyst, Biochem. J. 242 (1987) 565–572.
trolled by the v+ gene in phage T4, Biochem. Biophys. Res. Commun. 37 (1969) [24] V. Bailly, W.G. Verly, AP endonucleases and AP lyases, Nucleic Acids Res. 17
646–651. (1989) 3617–3618.
[7] E.C. Friedberg, J.J. King, Dark repair of ultraviolet irradiated deoxyribonucleic [25] M. Manoharan, A. Mazmumder, S.C. Ransom, J.A. Gerit, P.H. Bolton, Mecha-
acid by bacteriophage T4. Purification and characterization of a dimer-specific nism of UV endonuclease V cleavage of abasic sites in DNA determined by 13 C
phage-induced endonuclease, J. Bacteriol. 106 (1971) 500–507. labeling, J. Am. Chem. Soc. 110 (1988) 2691–2692.
[8] E.C. Friedberg, D.A. Clayton, Electron microscopic studies on substrate speci- [26] A. Mazmumder, J.A. Gerit, L. Rabow, M.J. Absalon, J.A. Stubbe, P.H. Bolton, UV
ficity of T4 excision repair endonuclease, Nature 237 (1972) 99–103. endonuclease V from bacteriophage T4 catalyses DNA strand cleavage at alde-
[9] E.C. Friedberg, Studies on the substrate specificity of the T4 excision repair hydic abasic sites by a syn -elimination reaction, J. Am. Chem. Soc. 111 (1989)
endonuclease, Mutat. Res. 15 (1972) 113–123. 8029–8032.
[10] R. Cone, T. Bonura, E.C. Friedberg, Inhibitor of uracil-DNA glycosylase induced [27] S.T. Lovett, R.D. Kolodner, Identification and purification of a single-stranded
by bacteriophage PBS2, J. Biol. Chem. 255 (1980) 10354–10358. DNA-specific exonuclease encoded by the recJ gene of Escherichia coli, Proc.
[11] E.C. Friedberg, A. Ganesan, K. Minton, N-glycosidase activity in extracts of Bacil- Natl. Acad. Sci. 86 (1989) 2627–2631.
lus subtilis and its inhibition following infection with bacteriophage PBS2, J. [28] G. Dianov, T. Lindahl, Reconstitution of the DNA base excision-repair pathway,
Virol. 16 (1975) 315–321. Curr. Biol. 4 (1994) 1069–1076.
[12] E.C. Friedberg, Correcting the Blueprint of Life – An Historical Account of the [29] Z. Livneh, D. Elad, J. Sperling, Enzymatic insertion of purine bases into depuri-
Discovery of DNA Repair Mechanisms, Cold Spring Harbor Laboratory Press, nated DNA in vitro, Proc. Natl. Acad. Sci. 76 (1979) 1089–1093.
1997, pp. 123. [30] H. Kataoka, M. Sekiguchi, Are purine bases enzymatically inserted into depuri-
[13] M.G. Wovcha, H.R. Warner, Synthesis and nucleolytic degradation of uracil- nated DNA in Escherichia coli? J. Biochem. 92 (1982) 971–973.
containing deoxyribonucleic acid by Escherichia coli deoxyribonucleic acid [31] B. Martin, N. Sicard, Absence of pyrimidine insertase activity in E. coli extracts,
polymerase I, J. Biol. Chem. 248 (1973) 1746–1750. using plasmid DNA containing apyrimidinic sites, Mutat. Res. 132 (1984) 87–93.