Ecological solar UV radiation is conventionally divided into two regions: UVB (290-320 nm) and UVA (320-400 nm), which includes UVA1 (340-400 nm) and UVA2 (320-340 nm). UVB is filtered by stratospheric ozone, and its minor fraction (about 2%) enters the biosphere. UVA photons are not absorbed by the ozone layer; UVA fully reaches the Earth’s surface and is 20 times higher in intensity than UVB. High-energy UVB photons exert a strong damaging effect on biological systems. Their role in inducing carcinogenesis in the human skin and cell death in plants increases as stratospheric ozone is destroyed and the UVB intensity consequently increases in the biosphere [1,2]. DNA is one of the critical molecular targets in cell structures of living organisms exposed to UVB. UVB photons are directly absorbed by DNA bases and thus efficiently induce generation of two main lesions, Cyclobutane Pyrimidine Dimers (CPDs) and, with a lower yield, pyrimidine (6-4) pyrimidone photoproducts (6-4PPs). Both lesions are formed by pyrimidine bases that are adjacent in the same DNA strand. To produce CPD, the cyclobutane ring is formed within 1ps when the 5-6 double bonds are broken in the bases. In the case of 6-4PP, the formation of an ordinary bond is preceded by cyclization between the C5-C6 double bond of pyrimidine and the C4 carboxyl group of thymine or the imino group of cytosine; unstable cyclization products are then reorganized to 6-4PP. The mechanism of the process is more intricate and takes more time (4 ms) to complete. In addition, the 6-4PP quantum yield is approximately seven times lower than the CPD quantum yield. Various biological responses, including cytotoxicity, apoptosis, and carcinogenesis, may develop in response to these DNA lesions. The biological efficiency of the reactions of DNA damage by UVA is two to four orders of magnitude lower than that of UVB. However, UVA photons also contribute to the genotoxic and cytotoxic processes because the UVA wavelength range accounts for approximately 95% of solar UV radiation. Endogenous photosensitizers, including protoporphyrin IX, riboflavin, pterins, mostly mediate UVA-induced reactions [3,4]. Such reactions are strongly oxygen dependent and can be classed as photodynamic reactions. Oxygen either reacts with the electronically excited photosensitizer or is involved in secondary reactions with radicals generated by the photosensitizer or the substrate molecule during its photosensitized oxidation.
Two types, I and II, of photosensitized oxidation reactions are recognized based on the primary process. A type I mechanism is initiated by electron transfer between a photoexcited Sensitizer (S*) and a biological substrate Molecule (M). The reaction yields a pair of radicals, that is, the anion radical S•− and the cation radical M•+. In an alternative primary bimolecular reaction that can initiate the type I mechanism, a photoexcited sensitizer reduces O2 to produce the cation radical S•+ and the superoxide anion radical O2•−. After primary one-electron oxidation of the substrate molecule, the resulting two radicals are involved in several subsequent reactions. S•− reacts with O2 to regenerate the photosensitizer and to produce O2•−. This process is thought to provide a main source of O2•− in photosensitized reactions and is far more important than direct O2 reduction by the excited sensitizer. O2•− can undergo spontaneous or enzymatic dismutation to Hydrogen Peroxide (H2O2), which is another Reactive Oxygen Species (ROS). Like O2•−, H2O2 does not display high reactivity towards the majority of biomolecules. However, H2O2 can migrate throughout the cell and to cause the Fenton reaction in the presence of Fe2+:
H2O2 + Fe2+ → •OH + OH− + Fe3+
The highly reactive •OH radical, which is produced in the reaction, can interact in its generation site with biomolecules by addition to a double bond or abstraction of a hydrogen atom. Both of the reactions yield neutral molecule radicals (M•), which are probably precursors of peroxyl radicals (MOO•), which result from the interaction of M• with O2. The process results in oxidative damage to molecules. In contrast to the radical mechanism of type I reactions, a primary mechanism of type II reactions includes energy transfer from a sensitizer excited to a triplet state to dissolved oxygen, which occurs in its triplet ground state (3O2). This yields singlet molecular oxygen (1O2) as a result of a spin flip in one of the two unpaired electrons of 3O2. Molecular oxygen in its activated (singlet) state is far more reactive than in its ground state. At the same time, 1O2 is a more selective oxidant as compared with •OH [4,5].
Among the above mentioned photosensitizers, protoporphyrin IX (PPIX) is a classic sensitizer of type II photodynamic reactions [4]. The absorption spectrum of PPIX has a major maximum in a region of 400-410 nm and minor peaks in a region of 500-620 nm. At the same time, intense photon absorption by PPIX is observed throughout the UVA range, expanding the spectral range of its photosensitizing activity associated with 1O2 generation. Oxygen quenches the excited triplet state of PPIX to produce 1O2 with high quantum efficiency (≈ 0.6). On the other hand, an extremely low efficiency of electron abstraction reactions is characteristic of PPIX with the majority of biomolecules. Protoporphyrin is consequently almost not involved in photosensitized type I oxidation reactions in contrast to the other photosensitizers (pterins, riboflavin). Riboflavin acquires strong oxidative properties in an excited state and is consequently classed with photosensitizers of type I reactions. Type II reactions can also be photosensitized by riboflavin, which can generate 1O2 with a quantum yield of approximately 0.5. A system of conjugated double bonds of the isoalloxazine ring determines the photophysical, photochemical, and spectroscopic properties of riboflavin [4]. The absorption and fluorescence excitation spectra of oxidized riboflavin state have two maxima in a region of 300-500 nm, at 360 and 450 nm, and its fluorescence spectra have a single maximum at approximately 530 nm. A type I mechanism may underlie riboflavin-photosensitized oxidation of G bases in DNA. Generation of 8-oxo-dihydroguanine (8-oxodG) on exposure to UVA (365 nm) occurs via electron transfer to a triplet excited state of riboflavin with the formation of the riboflavin anion radical and dG•+ as intermediates. A mechanism based on dG•+ hydration was assumed to explain how 8-oxodG forms in cell DNA of cultured mammalian cells exposed to UVA in the presence of riboflavin. The results gave grounds to assume that riboflavin-photosensitized DNA damage potentially plays a role in skin carcinogenesis. Pterins (Ptrs) are heterocyclic compounds widespread in living systems and are a conjugated system of two, pyrimidine and pyrazine, rings. The Ptr family includes many derivatives, which differ in the nature of side-chain substituents bound to the pyrazine ring. When excited with UVA radiation (365 nm), these biomolecules fluoresce with a maximum at 440–450 nm; generate ROS (1O2 and O2•−); or undergo photooxidative degradation to yield various products, which can also exert sensitizing activity [6]. On exposure to sunlight, Ptrs accumulate in the human skin and generate ROS when excited, thus causing photooxidative stress. Photochemical generation of ROS determines the sensitizing properties of Ptrs; i.e., Ptrs are capable of inducing oxidative damage to biomolecules on exposure to UVA. The purine nucleotide dGMP was used as an oxidation substrate to study the mechanisms of the photosensitizing effect of Ptrs. A type I photosensitization mechanism was shown to predominate when dGMP oxidation is induced with Ptrs in their neutral forms. The mechanism is initiated by electron transfer from the nucleotide to the triplet excited state of Ptr, yielding Ptr•− and the dGMP•+ cation radical; deprotonation and hydration of the latter lead to oxidative degradation of the nucleotide. Guanine is known to have the lowest ionization potential among all nucleotide bases and is therefore the only base whose one-electron oxidation is induced by the majority of type I photosensitizers. It is of interest in view of this that Ptrs are capable of photosensitizing thymine oxidation in the pyrimidine nucleotide dTMP. UVA-initiated electron transfer from the nucleotide to the triplet Ptr state yields the Ptr•− anion radical and the dTMP•+ cation radical. The latter occurs in equilibrium with the deprotonated form (dTMP•), and Ptr•− is protonated to produce PtrH•. Subsequent reactions of the radicals with O2 include generation of O2•−, H2O2, and •OH and yield several ultimate dTMP degradation products. An additional process involves the above radicals in the absence of O2. The process, which was not observed with other photosensitizers, is based on the binding of the PtrH• and dTMP• radicals to produce the covalent adduct Ptr–dTMP. It is important to note that a Ptr adduct with a thymine base similarly forms in dsDNA on exposure to UVA in the absence of O2 [7]. Adducts of the kind presumably form in cell DNA as well. The assumption is based on the following facts. First, the O2 concentration can be extremely low in certain tissues. Second, Ptr freely penetrates through biomembranes and can therefore occur in the nucleus. Third, the Ptr concentration used in experiments with isolated DNA is comparable with the Ptr concentration in skin cells. Based on the above data, Ptrs are thought to act as endogenous photosensitizers that are capable of triggering genotoxic processes [7].
It is now exactly established that both UVB and UVA components of solar UV radiation are strongly implicated in the etiology of major human skin cancers, namely, basal cell carcinoma and melanoma, which originate from keratinocytes and melanocytes, respectively. Currently, skin cancer is the most common tumor diagnosed in many countries and the numbers of nonmelanoma and particularly those of melanoma skin cancers have increased sharply over the last decades. Melanomas represent about 10% of all skin cancers, nevertheless they account for the majority of skin cancer-related deaths because of the high metastatic potencial and resistance to therapy. The carcinogenic effect of UVB attracted predominant attention in early studies because UVB photons were reliably demonstrated to efficiently induce the formation of the photoproducts in DNA that provoke skin cancer. A relationship was revealed between UVB irradiation and nonmelanoma skin cancers carrying characteristic mutation signature, that is, C→T transitions at dipyrimidine DNA sites and CC→TT tandem base substitution in human tumor suppressor gene p53 [1]. The mutations are caused mostly by CPDs, which are repaired slower than 6-4PPs; faster 6-4PP repair is due to the fact that 6-4PPs structurally distort the DNA helix to a greater extent and are consequently more efficiently recognized and eliminated by Nucleotide Excision Repair (NER). Data accumulated in subsequent years to indicate that UVA plays a greater role in carcinogenesis than believed earlier. Epidemiological evidence is available showing that prolonged and repeated exposures of humans to artificial UVA light from sun lamps or tanning beds constitute a major risk factor for melanoma induction. Also, an increased long-term risk of melanoma is observed in patients treated with the combined action of the sensitizer psoralen and UVA irradiation for photochemotherapy of psoriasis (PUVA therapy). So, it became clear that UVA1 is not photochemically/biologically inactive, and contributes to the photosensitized generation of ROS, which are presumably involved in carcinogenesis in the skin. UVA radiation is far more efficient than UVB in causing oxidative damage to DNA bases in isolated cells and the human skin [3]. The UVA-induced formation of 8-oxodG is mostly due to selective guanine oxidation by singlet oxygen (1O2), which is generated via a type II photosensitization mechanism. A lower contribution to the process is made by the hydroxyl radical (•OH), which can form after initial sensitized generation of the superoxide anion radical (O2•−) via a type I photosensitization mechanism. Apart from the 8-oxodG as a major photoproduct of oxidation reactions in DNA, UVA induces the formation of oxidized pyrimidines and single-strand breaks as well as oxygen-independent formation of CPDs, which involve predominantly thymine bases [5]. According to this complex spectrum of primary lesions, the spectra of UVA-induced mutations in cellular DNA include G→A transitions, which most likely originate from CPDs, and G→T transversions, which most likely originate from 8-oxodG. As mentioned above, C→T transitions are characteristic of UVB. It is clear that the spectra of lesions and mutations in DNA depend on the photon wavelength, the cell type, and the efficiency of cellular repair systems in eliminating various lesions.
To prevent the mutagenic effect of UV radiation, several specific DNA repair mechanisms are developed to maintain genome integrity at damaged sites within the complexicity of genome structures. Two main strategies exist to repair UV-induced DNA lesions: a light-dependent process (phoreactivation) that removes lesion using Photolyase (PL) and a light independent process (“dark repair”) that excises the UV-damaged region followed by de novo synthesis of an intact DNA strand. According to phylogenetic analyses, 3.8 billion years ago living organisms possessed photolyase-like genes, making photoreactivation the oldest known DNA repair mechanism [8]. DNA photolyase genes evolved in all branches of life, including marsupials, however, PLs are not found in placental mammals. PLs can be classified as CPD- or (6-4)-photolyases, according to their exclusive substrate specificity for CPD or 6-4PP, respectively. Studies in several organisms allowed decoding the modes of these PLs action. All PLs bind the catalytic cofactor Flavin Adenine Dinucleotide (FAD) in the form FADH− and also additional chromophore, which functions as a light-harvesting photoantenna. The excitation energy transfer from antenna chromophore generates excited flavin cofactor *FADH−, which in CPD-photolyases donates an electron to the CPD to catalyze the repair reaction by cleaving the C5–C5 and C6–C6 bonds of the cyclobutane ring. The repair reaction of (6-4)-photolyases also uses *FADH− as an electron donor to form a transient oxetan-type residue followed by C6–C4 ordinary bond splitting. In both cases, the result is restoration of the native DNA sequence [9].
The “dark repair” mechanism, also called NER, promotes the repair of UV-induced lesions in DNA via two pathways: Global Genome Repair (GGR) and Transcription Coupled Repair (TCR) processing damage throughout the genome or along actively transcribed DNA strands, respectively. The TCR damage recognition step relies on the stalling of RNA Polymerase II (RNA Pol II) and, as a consequence, TCR predominantly repair lesions (CPDs) on the transcribed DNA strand. The RNA Pol II trans locates along DNA template strand, synthesizing the complementary RNA molecule. Breaks, gaps, and modified nucleotides can lead to stalling and arresting of the polymerase [10]. Interestingly, the decisive recognition step only occurs by interaction between RNA Pol II and the human CSB (Cockayne Syndrome protein B) [10]. Mutations in the CSB gene result in a rare genetic disease called “Cockayne syndrome”. In accordance with the proposed model, CSB binds stalling RNA Pol II and promotes its forward translocation, increasing the bypass efficiency at minor barriers [10]. While base alkylation, abasic sites and 8-oxodG can be bypassed by the RNA Pol II, pyrimidine dimers cause stalling and arrest [11,12]. In the example of (T<>T) CPD, the stalling occurs by the stacking above the bridge helix of Pol II, slow incorporation of an A in front of the first T involved in the dimer, and an even slower misincorporation of an U in front of the second T. This misincorporation finally leads to the arrest of transcription [12,13]. The above data show that TCP pathway acting in transcribed genomic regions displays an efficient recognition mechanism of UV-induced pyrimidine dimers. The process depends on the stalling of RNA polymerase and complexation with CSB at the damage site. This implies that transcriptional activation directly promotes the control of genome integrity [14]. In contrast to TCR, GGR acts in poorly transcribed/untranscribed genomic regions to efficiently repair UV-induced lesions (CPD and 6-4PP). In this NER pathway, the damage recognition is performed independently of RNA Pol II. In GGR, the central initiator is Xeroderma Pigmentosum group C (XPC)-RAD23 protein complex. It was identified in the GGR pathway because of the ability of XPC to bind DNA lesions. The XPC complex is stabilized at damage sites when associated with Centrin 2 in human. To adopt the bound conformation, the XPC complex needs to overcome a consequent energy barrier, which was described as a primary regulator for the recognition specifity. Indeed, DNA lesions induce structural changes of the DNA helix structure and weak base pairing, causing a decrease of the energy barrier that the recognition complex needs to overcome for efficient binding [15]. In other words, XPC may patrol along the DNA until encountering a disturbed helical structure with weak base pairing, allowing XPC to bind the lesion site. This process can explain how the XPC complex detects different DNA lesions and why the identification of particular lesion is more efficient. For example, 6-4PP causes a massive thermodynamic destabilization of the helical structure and, as a consequence, 6-4PP recognition by the XPC complex is preffered compared to CPD. GGR pathway primarily recognizes the lesion by the damage sensor complex UV–DDB (DNA damage binding protein), which in able to scan DNA in compacted chromatin. Binding to the lesion, UV–DDB recruits the Rad4/XPC complex for a second recognition step. The stalled RNA Pol II–CSB complex and the Rad4/XPC complex recruit the Transcription Factor IIH (TFIIH), and XPD proceeds to a damage validation step. Upon this final recognition step, the damaged DNA region is excised by a dual incising process, and the gap is filled by de novo DNA synthesis and nick ligation [16].
As already mentioned, UV-induced DNA damage plays a crucial role in the early development of melanoma and basal cell carcinoma. If photoproducts are not repaired in DNA or damaged cells (melanocytes and keratinocytes) are not eliminated via apoptosis, then certain DNA defects exert their mutagenic properties and activate oncogenes. As was demonstrated in recent extensive studies, UVB and UVA cause different types and different amounts of DNA lesions in melanocytes and keratinocytes, and this fact can explain the difference in mutation spectrum between melanomas and carcinomas. A NER system repairs dimeric photoproducts in DNA and plays an important role in preventing UV-induced skin cancer. NER defects are associated with several rare genetic disorders, such as Xeroderma Pigmentosum (XP). Cells of XP patients are hypersensitive to UV radiation. The patients are several orders of magnitude more likely to develop skin cancers of all types, including melanoma, the fact suggesting a contribution to melanoma and nonmelanoma skin cancers for pyrimidine dimers. On the other hand, base excision repair defects, which decrease repair of oxidative base lesions, are not associated with skin cancer in the majority of cases [1].
Basal cell carcinoma is a skin cancer that originates from the keratinocytes that occur in the basal layer of the epidermis and undergo uncontrolled proliferation, as characteristic of cancer cells. Cancer is known to originate from proliferating cells rather than from fully differentiated cells, which are incapable of proliferation. UV-induced DNA lesions in keratinocytes do not cause mutations if damaged DNA is repaired prior to cell division. Another important fact is that DNA lesions differ in mutagenic potential. This pertains especially to cytosine-containing dimeric photoproducts, which are far more mutagenic than thymine- containing CPDs and 6-4PPs. CPDs, which are characteristic of UVB, are also a main class of photoproducts that arise in DNA upon UVA exposure of cells and the human skin according to several studies. However, while UVB-induced dimers contain both thymine and cytosine, UVA induces mostly thymine-containing dimers [3]. Interesting data were obtained in a study of how generation of thymine-containing dimers depends on the penetration of UVB or UVA1 photons into various layers of the human skin epidermis. The amount of UVB-induced dimers was found to decrease with the increasing depth within the epidermis, while the amount of UVA1-induced dimers substantially increased with the increasing depth and reached its maximum in the basal layer of the epidermis. UVB activity decreases with the increasing depth within the epidermis because chromophores present in the skin absorb UVB photons. UVA1 photons are absorbed to a lesser extent in the upper epidermal layers and consequently penetrate deeper. A hypothesis was advanced that backward dermal scattering (for example, from collagen) and forward epidermal scattering are responsible for the higher sensitivity of the basal epidermal layer, which harbors keratinocytes (and melanocytes), to the genotoxic effect of UVA1 [17]. The main UVA-induced product 8-oxodG, which provides a biomarker of oxidative damage to DNA, is generated mostly via 1O2-mediated reactions in cells and the human skin. Other oxidative reactions that eventually produce 8-oxodG and minor amounts of oxidized pyrimidine bases and single-strand breaks involve •OH. The DNA repair enzyme 8-oxoguanine DNA glycosylase 1 (OGG1) specifically repairs 8-oxodG and thus prevents the GC→TA mutations originating from this photoproduct. OGG1 is expressed in the basal layer of the human epidermis to a lower level than in the upper epidermal layers. OGG1 expression is presumably regulated by the keratinocyte differentiation status, being higher in differentiated cells. The 8-oxodG photoproduct is consequently repaired slower in the basal layer compared with the upper layers of the epidermis, the fact being possibly responsible for the higher sensitivity of the basal layer to 8-oxodG accumulation. Moreover, thymine-containing dimers form in UVA-exposed keratinocytes with a yield three times higher than the 8-oxodG amount. Thymine-containing dimers give origin to GC→AT mutations, while GC→TA mutations most likely originate from 8-oxoG. Such mutations of the p53 and Brm genes, respectively, were detected in nonmelanoma human skin cancers. The fact that the basal layer of the epidermis is vulnerable to the accumulation of the above photoproducts and mutations makes it possible to assume that the human skin is more susceptible to UVA-induced carcinogenesis than believed earlier [18]. Recent reports show that delayed dark CPDs and oxidized photoproducts arise in keratinocyte DNA on exposure to UVA1. Vitamin E, which possesses antioxidant properties, inhibited the formation of these photoproducts and the respective light defects, indicating that photooxidation reactions underlie their origin. Dark CPDs were observed to arise in the human skin 2 h after exposure to radiation with a wavelength of 385 nm and to persist in the skin for 24 h [19,20]. Mechanisms of the light and subsequent dark steps of the formation of CPDs and other photoproducts in keratinocyte DNA are important to study in order to better understand the contribution of UVA to human skin carcinogenesis.
Melanoma is a malignant tumor that originates from epidermal melanocytes. Melanocytes differ from keratinocytes by being resistant to apoptosis; living longer in the skin; and having another spectrum of oncogene-activating mutations, which can arise from another class of UV-induced DNA lesions. A comparative quantitation of UVB- and UVA-induced DNA lesions in human melanocytes and keratinocytes showed that CPDs form with similar efficiencies in cells of both types on exposure to UVB as well as to UVA [21]. At the same time, the 8-oxodG amount in UVA-exposed melanocytes was 2.2 times greater than in UVA-exposed keratinocytes. The data indicate that the products of UVA-induced oxidation reactions make a greater contribution to DNA damage in melanocytes than in keratinocytes. The difference may be associated with photosensitizing activity of the pigment melanin, which is synthesized in melanocytes. An important role of melanin in melanoma induction was demonstrated in a study that used UV radiation of strictly determined wavelengths [22]. The study showed that UVA induces 8-oxodG, a mutagenic oxidative lesion that can give origin to GC→TA transversions, and that the induction requires melanin to be present in melanocytes. In contrast to UVA, UVB induces melanoma regardless of whether the pigment is present and is associated with direct CPD formation in DNA. As several studies showed, delayed CPD formation takes place in the dark several hours after direct photochemical CPD generation in melanin-containing human melanocytes exposed to UVB or UVA [23]. Dark CPDs do not arise in pigment-free melanocytes exposed to UVA, in contrast to melanin-containing melanocytes. A protective effect exerted by antioxidants gave grounds to assume that an oxidation process with the participation of melanin is involved in dark CPD generation [23,24]. It is known that and NO• are produced enzymatically after a time lag during the cell response to UV-induced stress and that peroxynitrite (ONOO−) results from their recombination. Peroxynitrite was assumed to react with melanin monomers to produce unstable dioxetanes near DNA, and their decay to triplet-excited carbonyls in the dark yields CPDs via triplet–triplet energy transfer from carbonyls to pyrimidine bases in DNA [23-25]. The putative chemical excitation mechanism possibly underlies the shift in CPD composition, that is, a substantial increase in cytosine- and thymine-containing dimers that are more mutagenic than UVA-induced thymine-containing dimers. A dual, protective and sensitizing, role in generation of light and dark CPDs in the epidermis was demonstrated for melanin in a recent study of Fizpatrick Skin Types (FSTs) I/II and VI upon exposure to radiation that mimicked solar radiation [26]. Maximum generation of dark CPDs was observed 1-2 h after irradiation and was probably mediated by melanin-photosensitized oxidation reactions. On the other hand, light CPDs did not form in the FST IV basal layer, possibly, because melanin acts as a filter to protect DNA from penetration of UV radiation. The molecular basis of photosensitizing activity of melanin and its role in chemical excitation of components of oxidation reactions that underlie the formation of dark CPDs are important to study in order to develop better means to prevent UV-induced carcinogenesis in the human skin.
Photoexcited states of endogenous sensitizers have been considered as promising molecular targets for chemoprevention of skin photodamage. Compounds capable of quenching photoexcited states by direct physical interaction are called Quenchers of Photoexcited States (QPES). Secondary amines of the proline alkylester type were identified as protective physical QPES compounds. Strong protection of cultured human skin cells and reconstructed full thickness human skin against photodamage was observed using the prototype QPES compounds L-proline and L-proline methylester, which deserve further experimental evaluation as chemopreventive agents for skin protection and for therapeutic applications. Such types of QPES compounds may be envisioned as agents used together with sunscreens to exert skin protection by quenching excited states not only of endogenous photosensitizers but also UV filters [27].
This work was supported by a state contract with Moscow University (no. 121032500058-7).
SignUp to our
Content alerts.
Are you the author of a recent Preprint? We invite you to submit your manuscript for peer-reviewed publication in our open access journal.
Benefit from fast review, global visibility, and exclusive APC discounts.