Photoreactivation (PHR), also known as “light repair,” is a process that utilizes visible light to reverse UV-induced lesions, either CPDs or (6-4)PPs, by directly rearranging bonds. A photolyase enzyme recognizes a lesion, binds to the site, and from there it is a single-step chemical process that uses blue to near-UV light energy to return the CPD or (6-4)PP to its original state (Sancar, 2000). The catalytic cycle of photolyases rely on a non-covalently bound cofactor, flavin adenine dinucleotide (FAD) (reviewed in Weber, 2005). Both the ground-state redox properties and the excited-state properties of the FAD cofactor are utilized. All photolyases are homologous across bacteria, archaea and eukaryotes, which suggests this mechanism developed early in evolution (Eisen and Hanawalt, 1999).

Photoreactivation genes, phr1 and phr2, that encode photolyase enzymes have been described in several studies on halophilic archaea (DasSarma et al., 2001; McCready and Marcello, 2003; Baliga et al., 2004; Capes et al., 2011) and the PHR process has been observed in some species and described (Hescox and Carlberg, 1972; Martin et al., 2000; McCready and Marcello, 2003). Interestingly, in gene knockout studies of phr1 and phr2, only phr2 was associated with PHR in H. salinarum (Baliga et al., 2004). The phr2 gene product did not display (6-4)PP repair activity, only efficient CPD repair (McCready and Marcello, 2003). There may also be species-specific regulation (Kish and DiRuggiero, 2012) since UV irradiation induced transcription of the ph2 gene in Halococcus hamelinensis (Leuko et al., 2011) but not H. salinarum (Baliga et al., 2004).

The function of phr1 is unclear. Kanai et al. (1997) suggested that the phr1 gene encodes a blue light receptor, descended from ancestral photolyase genes, and may function in circadian rhythms. A study on the evolution of photolyase genes also demonstrates that specificity for CPD vs. (6-4)PP lesions can change through time and across species (Eisen and Hanawalt, 1999).

Nucleotide excision repair (NER), or “dark repair,” is a universal and highly conserved system that allows cells to excise DNA lesions including CPDs, (6-4)PPs, and other bulky adducts (Sancar, 1996). Its machinery does not require light for the reactions to occur. There are several proteins involved that carry out this multi-step process involving recognition of the DNA damage (e.g., in bacteria, UvrA), single strand cutting on both the 5′ and 3′ sides of the lesion (UvrB and UvrC), and removal of the damaged strand by a helicase (UvrD). A DNA polymerase must then build a new strand complementary to the undamaged one, and finally, ligase seals the phosphodiester backbone. All halophilic archaea examined have the uvrABCD genes (Capes et al., 2011), the necessary DNA polymerases (Kish and DiRuggiero, 2012), and the ligases (e.g., Zhao et al., 2006).

Halophilic archaea species may have eukaryotic homolog NER genes as well as the bacterial UvrABCD system, as homologs from both the XP system (mammalian) and Rad system (yeast) have been described in the archaea domain (Eisen and Hanawalt, 1999). For example, H. salinarum has xpf and the rad genes (rad2, rad3, rad25) (DasSarma et al., 2001; Capes et al., 2011). Rouillon and White (2011) postulated that the XPF-like nuclease (which does 5′ cleavage of the damage site in mammals) may be involved in a different repair pathway, and not NER, since the archaeal XPFs studied have a broader specificity than the nuclease of mammalian cells. Despite the observation of eukaryotic repair genes, at least the lab model species H. salinarum appears to depend entirely on the UvrABCD system for NER (Crowley et al., 2006), but it is not clear if this is true for all other halophilic archaea. It has been theorized that other genes may be involved in repair-supportive processes such as addressing damage causing stalled replication forks (Boubriak et al., 2008).

An early investigation of H. salinarum suggested halophilic archaea do not have NER (Sharma et al., 1984); however, this was later corrected in the literature (McCready, 1996; McCready and Marcello, 2003). To date, a number of halophilic archaea species have been shown to use NER to repair photodamage, including H. volcanii (McCready, 1996), H. salinarum (McCready, 1996; McCready and Marcello, 2003; Baliga et al., 2004; Boubriak et al., 2008), and a Great Salt Lake Halorubrum species (Baxter et al., 2007). Verifying the importance of the UvrABCD system, H. volcanii mutants lacking uvrA are significantly more UV sensitive than their wild-type counterparts (Lestini et al., 2010). Furthermore, H. salinarum mutant studies knocking out the function of UvrA, C, or AC double mutants reduced the repair of CPDs and thus, the survival of these strains (Crowley et al., 2006).

Halophilic archaea are also capable of transcription-coupled repair (TCR), a subpathway of NER that functions in removing RNA-polymerase-arresting DNA lesions from the template strands of active genes (Savery, 2007). Stantial et al. (2016) demonstrated that H. salinarum and H. volcanii employ TCR to repair CPDs following UV irradiation. A uvrA dependence was observed in H. salinarum, but not H. volcanii. It was proposed that a unique mechanism for TCR exists in halophilic archaea in which NER proteins are recruited by arrested RNA polymerase complexes following lesion recognition by the RNA polymerase itself.

The base excision repair (BER) pathway removes damaged or modified bases in DNA, which can be caused by UV-induced oxidative damage or other intracellular metabolites that modify the DNA base structure (reviewed in Krokan and Bjørås, 2013). DNA glycosylases that are specific to the particular photooxidative damage cleave the N-glycosidic bond between the base and the deoxyribose ring. The DNA backbone is then cleaved by an abasic-site endonuclease and the deoxyribose sugar is removed. The opposite strand provides the template for a repair polymerase to replace the removed nucleotide, and ligase seals the backbone. ROS damage to bases is repaired predominantly by BER across all species studied (Eisen and Hanawalt, 1999; Krokan and Bjørås, 2013) and likely in halophilic archaea as well (Capes et al., 2011).

Base excision repair glycosylase genes include mutY (A/G-specific adenine glycosylases), alkA (alkyladenine glycosylase), and nth (endonuclease III) (Denver et al., 2003; Krokan and Bjørås, 2013). These are found across the halophilic archaea with some exceptions and variations (Capes et al., 2011). Notably, alkA is missing from Haloquadratum walsbyi, and the nthA gene has three variants in some species. Other genes involved in this repair pathway are also present, indicating halophilic archaea have a fully functional BER apparatus. Upon UV-irradiation, Baliga et al. (2004) observed the up-regulation of six genes involved in repair of photooxidative damage.

It is unclear how halophilic archaea handle UV-induced single strand breaks (SSBs). In bacteria, the majority of these are breaks in the backbone and are repaired by ligase, but damage that creates an apurinic or apyrimidinic site is repaired by BER (e.g., Peak and Peak, 1982).

Homologous recombination (HR) is also employed by cells to repair UV damaged DNA, in particular, double-strand breaks (DSBs), but to a lesser extent, lesions such as CPDs and (6-4)PPs that stall replication forks. Following this damage, there are several steps: DSB recognition, excision at broken ends to create recognition sites, recombinase binding, strand pairing/exchange, branch migration, and branch resolution (Cox, 1991). The RecA protein brings homologous molecules together and facilitates this strand exchange. Recombinational repair can result in mutation as it has the potential to cause genome rearrangements.

In bacteria (e.g., E. coli), HR is highly conserved, and there are at least four pathways for the initiation of recombination, all of which produce substrates used by the RecA protein to catalyze the pairing and exchange (Roca and Cox, 1997). Interestingly, despite much focus on NER and BER, HR may play a larger role than generally thought in addressing UV damage. Mutations in the recA gene are more sensitive to UV light than NER genes such as uvrA (Cox, 1991). The eukaryotic Rad51 family of proteins (e.g., Saccharomyces cerevisiae) is related to RecA in bacteria, and homologs are present in at least some species of archaea (Sandler et al., 1996). The archaeal RadA proteins have been shown to function similarly in recombinational repair to RecA/Rad51 (Seitz et al., 1998), and two distinct radA genes are found in sequenced halophilic archaea genomes (Capes et al., 2011). Also, halophilic archaea have homologs to the yeast proteins Mre11, an HR nuclease, and Rad50, an HR ATPase, suggesting that the archaeal systems are likely similar in complexity to the eukaryotic yeast model (Woods and Dyall-Smith, 1997).

Halophilic archaea do employ HR following UV assault if DSBs occur. When a radA mutant of H. volcanii was exposed to UV light, this strain demonstrated sensitivity, which underscores the significance of this repair system for UV damage (Woods and Dyall-Smith, 1997). In wild type H. salinarum cells, UV-B or UV-C exposure induced the radA1 as well as other genes implicated in HR (McCready et al., 2005; Boubriak et al., 2008). Also, in this strain, mutant studies show mre11 is likely involved in DSB end processing as in eukaryotes, but not rad50 (Kish and DiRuggiero, 2008), and double mutants of these genes in H. volcanii are sensitive to DSB accumulation (Delmas et al., 2009). Halophilic archaea are polyploid (Breuert et al., 2006), and this may create a disadvantage in HR, given that concatemers can form between circular chromosomes as resolution proceeds (Delmas et al., 2009). However, polyploidy may also give the cells more correct sequence templates from which to draw in repairing the damaged area (Kish and DiRuggiero, 2008; Kish and DiRuggiero, 2012).

The HR RecA/Rad51 protein families are also known to induce an “SOS response” to excessive DNA damage, especially when single strands are exposed (Radman, 1975; Janion, 2008). This global response arrests DNA replication and induces genes in repair, mutagenesis and other DNA metabolisms. When looking at UV-induced gene induction in H. salinarum, two independent studies noted an increase in radA1 transcription but not other genes expected for an SOS response (Baliga et al., 2004; Breuert et al., 2006). To date, the SOS response is thought to be absent in halophilic archaea.

The red-orange and pink colors characteristic of aquatic hypersaline ecosystems such as Great Salt Lake, Utah are attributed to the accumulation of carotenoid pigments within cell membranes of resident halophilic archaea (Figure 1). Though not the subject of this review, we should note that there are also halophilic, carotenoid-containing bacteria, such as the Salinbacter genus, present in lower abundance.

These compounds are comprised of long, conjugated hydrocarbon chains that generally possess oxygen-containing functional groups and symmetry about the central carbon (Figure 4). Halophilic archaea are distinguished by a unique set of carotenoids (Kelly et al., 1970; Kushwaha et al., 1974, 1975; Marshall et al., 2007), the predominating pigment being bacterioruberin (Kelly et al., 1970; Ronnekleiv, 1995; Lobasso et al., 2008; Mandelli et al., 2012; Jehlicka et al., 2013; Naziri et al., 2014; Yatsunami et al., 2014), a compound implicated in protecting from UV photodamage (Shahmohammadi et al., 1998; Asgarani et al., 1999).

The pathway of carotenoid biosynthesis in halophilic archaea (reviewed in Rodrigo-Baños et al., 2015) begins with the isoprenoid precursor, isopentenyl pyrophosphate, which is converted to geranylgeranyl pyrophosphate, the first carotenoid of the pathway. Two of these molecules are joined to form phytoene, which is subsequently converted to lycopene through stepwise desaturation (Kushwaha et al., 1976). Lycopene gives rise to two of the major carotenoids of halophilic archaea, bacterioruberin and β-carotene. β-carotene is a precursor to retinal. In H. salinarum, retinal is incorporated as a chromophore into bacteriorhodopsin, or “purple membrane” protein, which pumps protons out of the cell upon exposure to light (Oesterhelt and Stoeckenius, 1971) to power ATP synthase enzymes. Other retinal-containing, light-energy transducing proteins are found in H. salinarum, such as halorhodopsin (Mukohata et al., 1980; Mukohata and Kaji, 1981), sensory rhodopsin, and photorhodopsin (Mukohata et al., 1999).

Carotenoid biosynthesis in halophilic archaea is regulated by a variety of factors including salinity (D’Souza et al., 1997; Lobasso et al., 2008; Biswas et al., 2016), pH (Hamidi et al., 2014; Rodrigo-Baños et al., 2015), oxygen tension (Sumper et al., 1976; Ng et al., 2000; DasSarma et al., 2001), and, of note, light exposure. The pigmentation levels of halophilic archaea grown under bright light are visibly higher than those cultured in the dark (Figure 1d) (Baxter et al., 2007). A number of genes connected to the carotenoid biosynthetic pathway that are regulated in response to light (and O₂) have been identified in H. salinarum (reviewed in Ng et al., 2000; DasSarma et al., 2001). Several are organized in the purple membrane regulon (crtB1, blp, bat, brp, and bop). It has been shown that bacterioruberin synthesis, specifically, the bop gene cluster of this species, is induced by low oxygen tension and high light intensity (Shand and Betlach, 1991). It has also been shown that the conversion of lycopene to bacterioruberin (Dundas and Larsen, 1963; Shahmohammadi et al., 1998), as well as β-carotene to retinal (El-Sayed et al., 2002), are enhanced by light in H. salinarum. This underscores the physiology of halophilic archaea, which must rise to the surface of the water to utilize their proton pump, but in doing so may encounter photodamage.

How do carotenoids protect halophilic archaea from photodamage? The best-established mechanism is through their antioxidant activity, which prevents photooxidative damage through ROS scavenging (most notably, ¹O₂ and •OH quenching) and deactivating excited photosensitizers (Krinsky, 1979; Truscott, 1990; Miller et al., 1996; Saito et al., 1997; Young and Lowe, 2001; Stahl and Sies, 2003; Mandelli et al., 2012; Igielska-Kalwat et al., 2015; Islamian and Mehrali, 2015). The antioxidant capacity of carotenoids increases with the number of conjugated π-bonds as well as the length of the carbon chain. For example, the increased conjugation of bacterioruberin (13 π-bonds) by comparison to β-carotene (9 π-bonds) (Figure 4) affords it a higher efficacy of ROS scavenging (Saito et al., 1997). The mechanisms by which carotenoids prevent oxidative damage take place in a manner that leaves them intact (Stahl and Sies, 2003). ¹O₂ quenching takes place through a direct transfer of energy between molecules, after which the energy gained by the carotenoid dissipates into the solvent as heat. The quenching of free radicals leads to subsequent reactions; •OH scavenging in particular is thought to play an important role in preventing oxidative damage to membranes (Sies and Stahl, 1995).

Carotenoids then certainly provide antioxidant protection from photochemical damage not only to DNA, but also to membranes and other cell components. This notion is well demonstrated by the increased sensitivity of colorless mutant halophilic archaea to UV irradiation (Dundas and Larsen, 1963; Rodriguez-Valera et al., 1982; Shahmohammadi et al., 1998; Baxter et al., 2007). Dundas and Larsen (1963) were the first to demonstrate that non-pigmented H. salinarum cells are sensitive to the damaging effects of light when compared with pigmented cells, despite both cell types growing equally well with no light exposure. The consequence of pigment loss was described as extensive lysis of the irradiated cells. Rodriguez-Valera et al. (1982) also observed membrane lysis of colorless or pale halophilic archaea exposed to intense light. These findings point to the most significant ramifications of intense photooxidative damage occurring outside of DNA.

Carotenoids apparently offer protection from direct forms of DNA photodamage. The formation of CPDs is suppressed by the presence of carotenoids; Baxter et al. (2007) demonstrated that the relative levels of TT damage were decreased 3.5-fold in UV-irradiated Halorubrum cells that were rich in pigmentation due to full-spectrum light exposure, when compared to irradiated cells that had been grown in the dark and had reduced carotenoid pigmentation (Figure 1d). These findings are in agreement with in vitro studies of Asgarani et al. (1999), which demonstrate the formation of CPDs in plasmid DNA is reduced in the presence of bacterioruberin. The specific mechanism through which this form of photoprotection occurs remains unknown (see conclusion). Many studies do suggest direct absorption of UV (e.g., Shahmohammadi et al., 1998). However, carotenoid compounds absorb light in the range of 340–550 nm (Takaichi and Shimada, 1992), whereas the UV spectrum ranges from 200 to 400 nm. Most likely then, they do not afford photoprotection by acting as a complete optical filter (Cockell and Knowland, 1999).

Carotenoids also exhibit interplay with the PHR system. Sharma et al. (1984) examined the UV sensitivity of several pigmented and colorless strains of Halobacteria and saw the levels of photoreactivation were reduced in the colorless mutants. The authors suggested the interpretation that the pigments do not play a role in direct absorption of UV, but instead function by supplying energy to photolyase during repair of pyrimidine dimers. However, this does not explain the observation that carotenoids provide photoprotection from UV under photolyase-inhibiting (dark) conditions (Baxter and Zalar, in press). Interestingly, Shahmohammadi et al. (1998) noted the effects of bacterioruberin were more protective in the case of UV exposure in H. salinarum than when cells were exposed to ionizing radiation or H₂O₂. They suggest the same explanations offered above: absorbance of UV energy by the carotenoid and a supplying of energy to the photoreactivation system. Nevertheless, these explanations do not complete the picture of how carotenoids shield DNA from UV light, particularly in the absence of visible light.

In addition to carotenoids, a number of overlapping pathways for avoiding oxidative damage via ROS detoxification are seen in archaea (reviewed in Pedone et al., 2004). Of particular relevance to the present review are hydroperoxidases and superoxide dismutases. These enzymes work together to prevent oxidative damage through ROS scavenging ( and H₂O₂ in particular), and are found widely among aerobic and facultatively anaerobic organisms.

Hydroperoxidases are heme proteins that facilitate the elimination of H₂O₂ (Pedone et al., 2004). They are divided into two classes, catalases, which catalyze the decomposition of H₂O₂ into O₂ and H₂O, and peroxidases, which catalyze the oxidation of other organic compounds by H₂O₂. Active catalase and peroxidase enzymes have been reported for H. salinarum (Fukumori et al., 1985; Brown-Peterson and Salin, 1995). Bifunctional catalase-peroxidase enzymes have also been observed. That of H. salinarum was found to shift between catalase- and peroxidase-dominant activity in response to pH and NaCl concentration (Fukumori et al., 1985; Brown-Peterson and Salin, 1993), and was not induced by environmental stressors including H₂O₂ and intense light (Long and Salin, 2000). Additionally, a catalase-peroxidase enzyme was purified from H. marismortui (Cendrin et al., 1994).

Superoxide dismutases provide protection from oxidative damage by catalyzing the dismutation of to O₂ and H₂O₂ (Cannio et al., 2000; Imlay, 2003). The yielded H₂O₂ is not only less toxic than its precursor, but also may be subsequently scavenged by hydroperoxidases. The presence of superoxide dismutase has been verified in H. salinarum (May and Dennis, 1987) and H. volcanii (May et al., 1989). In H. salinarum, the encoding gene (sod) is positioned adjacent to that of photolyase (Takao et al., 1990). Superoxide dismutase activity has been shown to increase in response to elevated intracellular in the aforementioned organisms (May and Dennis, 1989; May et al., 1989; Brown-Peterson and Salin, 1993); however, activity in H. salinarum decreased with prolonged exposure, yet was sustained in H. volcanii.

The superoxide dismutase of H. salinarum is associated with cofactor Mn(II), as opposed to Fe(II) (May and Dennis, 1987; May et al., 1989). It has been shown that H. salinarum, as well as the highly radioresistant model bacterium Deinococcus radiodurans, have higher intracellular ratios of Mn to Fe than less radiation-resistant organisms (Daly et al., 2004; Kish et al., 2009). While intracellular Mn does not directly provide protection against DNA damage in H. salinarum (Robinson et al., 2011), it is hypothesized to play a role in protecting DNA repair proteins from oxidative damage via antioxidant activity (Daly et al., 2007). Indeed, Mn complexes (with phosphates and small organic molecules) have been shown to reduce protein carbonylation (Daly et al., 2010; Matallana-Surget and Wattiez, 2013), a recognized consequence of UV-induced oxidative stress, in H. salinarum (Robinson et al., 2011).

One strategy for maintaining osmotic balance with the extracellular environment employed by certain groups of halophilic archaea is to accumulate ions intracellularly, particularly K⁺ and Cl^- (da Costa et al., 1998; Oren et al., 2002; Oren, 2008). Concentrated Cl^- attenuates oxidative damage by transferring an electron to •OH, producing a hydroxyl anion and atomic chlorine (Cl •) (Shahmohammadi et al., 1998). The subsequent reaction of Cl • with Cl^- produces chloride radicals (Cl₂^∙-), which are less damaging to DNA than •OH (Ward and Kuo, 1968). Cl^- and Br^- have been shown to protect DNA from oxidative damage incurred by γ-radiation (Shahmohammadi et al., 1998; Asgarani et al., 1999; Daly et al., 2004; Kish et al., 2009). Kish et al. (2009) further demonstrated that H. salinarum accumulates fewer base oxidation products than the non-halophilic D. radiodurans when subjected to the same doses of γ-radiation. Potassium chloride also suppresses the formation of CPDs in H. salinarum, although it appears to play a larger role in protecting from γ-radiation (Asgarani et al., 1999).

Other common pathways for oxidative damage avoidance, such as thioredoxin/glutaredoxin systems and peroxiredoxins, have been observed in archaea, particularly methanogens (Pedone et al., 2004; Erkel et al., 2006), but remain poorly described for halophilic archaea. However, the presence of γ-glutamylcysteine, a known detoxifier of H₂O₂ and (Quintana-Cabrera et al., 2012), has been observed in millimolar concentrations in H. salinarum, H. volcanii, H. marismortui, and Halorubrum saccharovorum (Newton and Javor, 1985; Sundquist and Fahey, 1989).

Altogether, H. salinarum demonstrates a remarkable capacity to withstand H₂O₂ and . Kaur et al. (2010) observed fairly constant cell survival after 2 h of exogenous H₂O₂ exposure up to a threshold of approximately 30 mM H₂O₂, after which small increases in concentration induced significant loss. A similar effect was observed on cell growth. For comparison, cell survival of E. coli reached 10% after 20 min of exposure to 20 mM H₂O₂ (Asad et al., 1998). H. salinarum cell survival and growth in the face of decreases more gradually, with 20–30% loss of survival occurring at approximately 4 mM paraquat, a compound that generates during metabolism (Kaur et al., 2010). It is difficult to compare studies of paraquat toxicity among these microorganisms due to its sensitivity to growth conditions, especially NaCl concentration (Kitzler and Fridovich, 1986). Nevertheless, Korbashi et al. (1986) observed 90% cell loss of E. coli treated with 0.75 mM paraquat for 30 min, and Kitzler et al. (1990) observed significant loss after 2–4 h exposure to 2.5 mM.

DNA damage, if unrepaired and replicated, can lead to mutation. This underscores the paradigm that while intact DNA is critical to survival, mutation is critical to evolution (Friedberg, 2003). Much has been written about duplication of genes as an evolutionary strategy, since one functional copy allows other copies to change DNA sequence over time (reviewed in Zhang, 2003). However, little has been discussed about the use of polyploidy as a strategy for genome protection. In the case of halophilic archaea, which inhabit UV-intense, hypersaline environments, one mechanism for photoprotection might be simply gene duplication, or in this case, genome duplication.

Halophilic archaea have more than one copy of their genome, and some species have up to 25 copies during their fastest growth phase (Breuert et al., 2006). This polyploidy may provide redundancy of genetic information and can lead to gene conversion or back mutation (Soppa, 2011). Gene duplication has notably led to a variety of rhodopsins in archaea (Ihara et al., 1999). In addition to evolutionary potential, polyploidy provides a nutritional phosphate storage mechanism (Zerulla et al., 2014). With respect to photoprotection, polyploidy would give halophilic archaea more resistance to DNA damaging conditions (Kottemann et al., 2005; Soppa, 2011; Zerulla et al., 2014) such as UV-exposure or desiccation. Logically, increasing the number of copies of a given gene should reduce the probability of its function being lost to DNA damage globally.

A relatively slow rate of global genome repair of CPDs has been reported in polyploid halophilic archaea H. salinarum and H. volcanii by comparison to the monoploid archaeon Sulfolobus solfataricus (Dorazi et al., 2007; Romano et al., 2007; Stantial et al., 2016). Stantial et al. (2016) proposed that this may be attributed to the larger amount of DNA that must be scanned and repaired in polyploid organisms, suggesting a potential tradeoff to the advantage of genome duplication. Also, it should be noted that in yeast, polyploid (4–10 genome copies) cells show no advantage over diploid cells in resistance to ionizing radiation (Mortimer, 1958; Mable and Otto, 2001). To date, there are no UV survival studies probing the significance of ploidy in halophilic archaea.

Direct UV damage to DNA predominantly occurs through the cyclization of adjacent pyrimidine nucleotides, producing CPDs, or by the formation of covalent bonds that produce 6-4PPs (Figure 2) (Besaratinia et al., 2011). The photochemical susceptibility to lesion formation differs among the four bipyrimidine sequences, decreasing in the order of (5′ to 3′) TC > TT > CT > CC (Matallana-Surget et al., 2008). The more photoreactive bipyrimidines being T-containing, it has long been suggested that organisms with high G+C content, such as halophilic archaea (63.1% G+C on average) (Jones and Baxter, 2016), may be less susceptible to UV-photodamage (Haynes, 1964; Setlow and Carrier, 1966; Joux et al., 1999; Kennedy et al., 2001). Indeed, high G+C content is correlated with a photoprotective bipyrimidine signature (Figure 5) (Jones and Baxter, 2016).

While halophilic archaea do have lower genomic photoreactivity with respect to bipyrimidine signature (P_g) than most other microorganisms, it should be noted that they have higher P_g scores than others with comparable G+C content (Figure 5). Interestingly, halophilic archaea have significantly higher incidences of 5′-TC-3′ sites than the average bacterium, archaeon, or random DNA sequence of comparable G+C content (Jones and Baxter, 2016). It has been proposed that this feature is attributed to a demand for acidic amino acids (Zhou et al., 2007), an important adaptation to protein function in high salinity (Kennedy et al., 2001). Notwithstanding the high incidence of 5′-TC-3′ sequences in halophilic archaea genomes does increase susceptibility to bipyrimidine lesion formation, there is, paradoxically, a photoprotective benefit to such: the associated amino acid bias equips these microorganisms with fewer residues susceptible to ROS (Zhou et al., 2007).

The high G+C content of halophilic archaea also decreases their susceptibility to photooxidative DNA damage. Wei et al. (1998) observed a negative relationship between G+C content and the formation of 8-hydroxy-2′-deoxyguanosine (8-OHdG), a guanine oxidation product, in UV-irradiated DNA. These authors hypothesized that thymidine may serve as an intrinsic photosensitizer and therefore, its limitation reduces ¹O₂ generation.