DNA Damage: Biomarkers of DNA Damage

8-Hydroxy-2′-deoxyguanosine (8-OH-dG)

The DNA guanine base oxidation product 8-hydroxy-2′-deoxyguanosine (8-OH-dG) is a crucial and well-characterized biomarker of oxidative stress-induced DNA damage. Reactive oxygen species (ROS) and reactive nitrogen species (RNS) are key mediators of oxidative stress which is defined as an “imbalance towards pro-oxidative changes which leads to damage and dysfunction of cellular structures, cells and whole organisms” 376, 377. These pro-oxidative species are permanently formed during the entire metabolism or by exogenous stressors including numerous variants of radiation and noxious substances. Reactive oxygen species (ROS) play a predominant role in DNA oxidation as an excess of ROS is able to oxidize DNA at both, the desoxyribose moiety and the four nitrogenous bases. The HO• radical accounts for the majority of DNA damages by attacking the C5 or  the C atom from the methyl group (CH3) abundant in the pyrimidines, corresponding to the C8 in purines or the amino group of adenines 45. HO• also attacks the carbohydrate moiety of nucleosides in close vicinity to the DNA bases 83. Amongst all nucleic bases, guanine (G) is highly susceptible to ROS-mediated oxidation 84. It is well-documented that the interaction of reactive radicals with G or free 2′-deoxyguanosine (dG) leads to the formation of radical adducts. Not only HO• radicals but also HOO• radicals as well as singlet oxygen (1O2) have the capacity to attack the C8 of the imidazole ring of dG to form 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxo-dG). For a review see Giorgio et al. (2020) 85 and Urbaniak et al. (2020) 377. As given in Figure 10, the addition of HO• to the C8 of the guanine moiety in dG gives rise to the reducing 8-hydroxy-7,8-dihydro-2′-deoxyguanosyl radical which is subjected to competitive one-electron reduction in cellular DNA to generate 2,6-diamino-4-hydroxy-5-formamidopyrimidine (FapyG) 378. One-electron reduction of dG also yields guanosyl radical cations that are converted subsequently by hydration into the 8-hydroxy-7,8-dihydro-2′-deoxyguanosyl radical as well 379.  However, successive dehydration and nucleophilic attack of HO• at C4, C5 and C8 of guanine also forms guanine radicals that decay to oxazolone in the case of C4 and C5 oxidation, or to 8-oxo-dG after hydrolytic cleavage of the C8-OH 89. Keto-enol tautomerism then allows for the generation of 8-OH-dG through electron abstraction (Figure 10). Amongst the four putative tautomeric species, the diketone variant has been found as being the most favourable one 380. 8-OH-dG can be considered as the most prominent mutagenic modification of DNA. The pro-mutagenic potential of 8-OH-dG can be attributed to its capacity to block the pairing process during DNA replication 381, 382. Herein, defective 8-oxy-guanine gives rise to incorrect base pairs with adenine, and, if left unrepaired, induces the occurrence of GC → TA transitions in the following replication round that in turn may lead to malignant cell transformation.

Figure 10: Generation of 8-OH-dG and 8-oxo-dG. 2′-deoxyguanosine (dG); 4,8-endoperoxide-2′-deoxyguanosine (1); 4R/4S 4-Hydroxy-8-oxo-4,8-dihydro-2′-deoxyguanosine (2); 4R/4S spiroiminodihydantoin nucleoside (3); 8-Hydroperoxy-2′-deoxyguanosie (4); 8-oxo-7,8-dihydro-2′-deoxyguanosine (5); 8-Hydroxy-7,8-dihydro-2′-deoxyguanosyl radical (6); 7-Hydro-8-Hydroxy-2′-deoxyguanosine (7); 2,6-diamino-4-Hydroxy-5-formamidopyrimidine (FapyG) (8); 8-Hydroxy-2′-deoxyguanosine (9), oxazolone (10). Image was reproduced from Urbaniak et al. (2020) 377. License at https://creativecommons.org/licenses/by/4.0/.


5,6-Dihydro-5,6-dihydroxythymidine (thymidine glycol, Tg)

Thymidine glycol (5,6-dihydro-5,6-dihydroxythymidine, Tg) represents one of the most abundant oxidation products of thymidine 383, 384. Tg is formed, on the one hand, by exogenous stressors such as chemical oxidants or ionizing radiation, and, on the other band, endogenously in the course of the aerobic metabolism (Figure 11). Almost 400 Tg molecules are formed per cell and day by oxidation of the double-bond in thymidine 385, and the existence of Tg in DNA has been utilized as a biomarker for oxidative stress 385, 386. As Tg harbors two chiral C-atoms, four diastereomers exist: cis-(5R,6S), trans-(5R,6R), cis-(5S,6R), trans-(5S,6S) 387, amongst which cis-stereoisomers represent the predominant variants in nature 388. It has been shown previously that γ-irradiation produces equal amounts of the (5R, 6S) and (5S, 6R) cis-diastereomers 389, wherease oxidative modification leads to the preferential production of (5R, 6S) cis-diastereomers 390. Epimerization at C6 leads to interconversion between cis– and trans-diastereomers, even though the equilibrium favors the cis-diastereomers by the factor 3-5 387, 391. Although Tg is characterized by a poor mutagenic potential 392, it has been identified as being a potent inhibitor of DNA replication 393-395. Fixing of Tg occurs in conjunction with specific endonucleases such as prokaryotic Endo III and Endo VIII from E. coli as well as their eukaryotic counterparts 396. Similar to the mouse ortholog mNeil1, which is capable of excising both stereoisomers of Tg 397, the Endo III orthologs yNtg1 and yNtg2 from yeast have been shown to remove distinct Tg isomers 398. In addition, human thymine-DNA glycosylase (TDG), also termed G/T thymidine glycosylase, as well as methyl-CpG-binding protein 4 (MBD4) have been described to eliminate Tg from a G/Tg mismatch 399. More recent data raised by the group of Arthur P. Grollman revealed that the main DNA glycosylases in bacteria, yeast and mammalian cells stereoselectively remove the (5S, 6R) and (5R, 6S) isomers of Tg from duplex DNA, thereby implying the occurrence of these enzymes as being complementary pairs in their respective organisms 400. Several studies revealed that Tg is primarily removed by BER, although NER may also be involved in the repair of this respective DNA lesion 401-403. Since G/T pairs in DNA have been reported as being efficiently repaired by the MMR system 404, it has been proposed that the MMR system might also play a crucial role in the replacement of Tg-bearing DNA 405, 406. However, data raised in E. coli by Perevozchikova and collaborators indicate that Tg-DNA obviously does not function as a substrate for the DNA MMR system 407.

Figure 11: Generation of thymidine glycol through ROS-mediated oxidation. Reprinted with permission from Laverty and Greenberg (2017) 646. Copyright 2017 American Chemical Society.


Pyrimidine Dimers

The UV-induced formation of pyrimidine dimers in DNA is a main detrimental mechanism in both, eukaryotic and prokaryotic cells. UV irradiation is categorized into three classes on the basis of the range of wavelength: UV-A (320–400nm), UV-B (290–320 nm), and UV-C (190–290 nm). Previous studies reported on the prominent DNA-damaging potential of UV-C as being mediated by covalently linking two adjacent pyrimidines that belong to the same DNA strand or to different DNA strands. This dimerization process predominantly generates cyclobutane pyrimidine dimers (CPD) and pyrimidine (6 – 4) pyrimidone photoproducts [(6 – 4)PP; Figure 1] 2-8. In CPD, the two adjacent pyrimidines are covalently connected by a cyclobutane ring, while in (6 – 4)PP the C6 position of one pyrimidine is covalently attached to the C4 position of the neighboring pyrimidine. These bulky dimers lead to conformational rearrangements of the helix, thereby accounting for cell death and mutagenicity. UV-A and UV-B are also capable of inducing the formation of pyrimidine dimers, but to a much lesser extend 1, 4. Numerous investigations reveal that pyrimidine dimers induce various mutations such as CG→TA, TA→CG and typical tandem CC→TT transitions 408-410. An intriguing feature of (6 – 4)PP is represented by the fact that it can be converted by UV-B to a Dewar photoproduct, a valence isomer of (6 – 4)PP, which in turn has the capacity to regenerate the regular (6 – 4)PP isomer upon exposure to UV-C 6, 7. (6-4)PP and its Dewar photo-isomer are important cytotoxic determinants of UV light contributing to sunlight carcinogenesis 7. Unrepaired UV lesions, however, hinder DNA replication and block normal progression of replication forks.

Figure 1. UV-induced pyrimidine dimers in DNA. (Left) Cyclobutane pyrimidine (thymine-thymine) dimer; (Right) pyrimidine (6 – 4) pyrimidone photoproduct. Image was reproduced, with permission, from Nikitaki et al. (2015) 644. License at https://creativecommons.org/licenses/by/4.0/


Organisms in all kingdoms of life have developed multiple repair pathways aiming to resolve and circumvent UV-induced DNA lesions 411. The plainest pathway, also known as direct reversal, utilizes photolyases that disrupt the covalent bonds between the two pyrimidines by using visible blue-light in sunlight 3. CPD photolyases are divided into class I, II, III 412, 413 and ssDNA photolyases 414 that bear a flavin adenine dinucleotide (FAD) molecule as the key coenzyme 3, 415. As outlined by Zhang et al. (2017) 416, the mechanism of CPD repair can be initiated either by direct light absorption of reduced FADHor via energy transmission from the excited antenna chromophore. Afterwards, the excited FADH−* assigns an electron to the CPD substrate which spontaneously disbands into its individual monomers, two thymine bases. In the final step, the electron turns back to the coenzyme and thus restores the enzyme activity. An alternative mechanism can be found in prokaryotes, in which a DNA glycosylase (T4 endonuclease V) specifically dissects the 5′-connected CPD base and the DNA backbone via the process of β-elimination 417. In bacteria and fungi, UV damage endonuclease (Uve1p, UVED) has also been reported to cleave the 5’-connection 418. The CPD process is completed by BER thus culminating in the restoration of the DNA 419, 420.

The central repair pathway for UV lesions, nonetheless, is represented by the NER system. NER is found in all organisms and can be subdivided into two signaling pathways: i) global genomic NER (GG-NER), and ii) transcription-coupled NER (TC-NER). Both pathways utilize a large number of signaling proteins and differ remarkably in the first fundamental step, the detection of DNA lesions (reviewed by Kusakabe et al. (2019) 301. NER starts with the recognition of UV lesions followed by segregation and incision of the DNA strand on both sides of the lesion, proceeding to lesion repair and subsequent de novo-DNA synthesis and ligation (summarized in 411, 421). During DNA replication, bypass of UV lesions can be achieved by either homologous recombination (HR) 422 or translesion synthesis (TLS) triggered by specialized DNA polymerases 411, 423. TLS aims at exploiting benefits from transcription blockade mediated by the presence of DNA lesions and recognizes lesions on the transcribed strand. Various ATPases play crucial roles in the NER mechanism contributing to lesion recognition. In bacteria, lesion detection is facilitated by the ATPases UvrA and UvrB during GG-NER 421, while in eukaryotes this process is performed by damaged DNA-binding protein 2 (DDB2), also termed Xeroderma pigmentosum complementation group E (XPE), as part of the heterodimeric UV-DDB protein in conjunction with the XPC complex 424-426. ATP-dependent proofreading and lesion-mediated strand segregation in eukaryotes are obviously performed by XPB and XPD ATPases/helicases in the oligomeric transcription factor complex TFIIH 303-305. After successful verification, specific nucleases are recruited for DNA incision and lesion repair. Additional information is given in section Nucleotide Excision Repair (NER).

γ-H2AX (H2AX)

H2AX is a mammalian histone variant and a member of the H2A family. As the main protein constituent of chromatin, histones are subjected to various types of post-translational modifications, amongst which phosphorylation, ubiquitination, and sumoylation play critical roles in DNA double-strand break repair (DSBR). Once DSB are formed in DNA, H2AX becomes rapidly phosphorylated at the C-terminal S139. This phosphorylation site is unique amongst the H2A histone family members and designated as “γ-phosphorylation”. Accordingly, the label “γ-H2AX” illustrates the specific phosphorylation at S139 of H2AX. γ-H2AX represents a sensitive marker for DNA DSB. As summarized by Pinder and colleagues 292 (Figure 12), the induction of DSB in DNA leads to the recruitment of the MRN (Mre11-Rad50-Nbs1) complex to the damage site 363, 364. It has been shown previously that constitutive ubiquitination of Nbs1 by the E3 ubiquitine ligase RNF8 (E3 ring finger protein 8)-UbcH5 (ubiquitin-conjugating enzyme H5c) during early DSBR is a prerequisite for the targeted recruitment of the MRN complex to DSB 427. Moreover, BMI1-RNF2 is recruited to DSB after chromobox homolog 4 (CBX4)-mediated sumoylation of BMI1 at K88 428. The E3 ubiquitin ligase RNF2 (RING1b/RING2)-UbcH5 and its adapter molecule BMI1 (B lymphoma Mo-MLV insertion region 1 homolog) catalyze the mono-ubiquitination of H2AX at K118/119 as a prerequisite for the successive targeting of the ataxia telangiectasia mutant (ATM) protein kinase and valuable formation of γ-H2AX 429, 430. ATM is activated and recruited to DNA damage sites in complex with the histone acetyltransferase Tip60 359, 360. In the next step, H2AX is rapidly phosphorylated by ATM and various other protein kinases such as ATR (ATM- and RAD3-related), and DNA-dependent protein kinase (DNA-PK) 361, 362.

Figure 12: Role of g-H2AX and post-translational modifications in the early double-strand break repair (DSBR). For details see section Biomarkers of DNA Damage. Image reproduced with permission from Pinder et al. (2013) 292. License at https://creativecommons.org/licenses/by/3.0/.


Phosphorylated H2AX (γ-H2AX ) binds to the mediator of DNA damage checkpoint 1 (MDC1) which is phosphorylated by ATM and thus qualified to recruit RNF8 431-433. RNF8 is responsible for the recruitment of another E3 ubiquitin ligase, RNF168, both of which act together in ubiquitinating γ-H2AX at K13/15 434. RNF8 interacts with the E2 ubiquitin ligase Ubc13–Mms2 to form K63-linked ubiquitin chains 435-437. RNF8-mediated poly-ubiquitination of γ-H2AX has been found to depend on the presence of the HECT E3 ligase HERC2 438 which is inevitably sumoylated by SUMO E3 ligase PIAS4 (protein inhibitor of activated STAT protein 4) 439.

Poly-ubiquitinated γ-H2AX mediates the recruitment of 53BP1 to damage sites where it contributes to DSBR 440. K63-polyubiquitination of γ-H2AX is antagonized by RNF169 and the de-ubiquitinases BRCC36 (a BRCA1-A complex member) as well as the 19S proteasomal lid subunit POH1 441-443. Poly-ubiquitinated γ-H2AX also acts as an adapter for several chromatin-modifying complexes such as SWR1 and INO80 369, 370. The DSBR continues as described under the section Homologous Recombination. These data collectively demonstrate the heterogeneous and dynamic nature of the DSBR mechanism and underline its internal complexity.