Increase, DNA Damage leads to Increase, Mutations

High. DNA damage in the form of nucleotide damage, single strand and double strand breaks, and complex damage can generate mutations, particularly when a damaged cell undergoes replication.

Nucleotide damage

Damage to single nucleotides can generate mutations. Oxidative damage and ionizing radiation can induce a range of base lesions, but guanine is particularly vulnerable because of its low redox potential (David, O'Shea et al. 2007). Repair is generally accurate, but generates single strand breaks. Repair processes can also insert an incorrect nucleotide where a lesion has been excised. If not corrected by mismatch repair processes before a replication cycle, an incorrect base is matched with its pair and is made permanent (Tubbs and Nussenzweig 2017). In a cell undergoing replication, the replication fork typically stalls at a lesion until repairs are complete, but translesion synthesis allows the replication fork to proceed at the cost of increased errors, or mutations (Abbotts and Wilson 2017). Different lesions vary in their frequency and ability to escape repair or be replicated and incorrectly paired by DNA polymerase, making some lesions more mutagenic than others. For example, guanine lesion 8-oxoguanine is very common, so although it is efficiently repaired it contributes to guanine mutations. Other guanine lesions including Fapy and hydantoins are less common but very mutagenic, so likely also contribute to guanine mutations (Neeley and Essigmann 2006; David, O'Shea et al. 2007). Thymine glycol is another common oxidative lesion formed from thymine that can also generate mutations.

Single strand breaks

Single strand breaks are generally repaired efficiently through a variant of the base excision repair pathway. However, replication fork collapse can occur when the replisome encounters an unrepaired single strand break, resulting in a double strand break (Kuzminov 2001).

Double strand breaks

Double strand breaks can generate mutations ranging from point mutations to inversions, deletions, duplications, and chromosomal gaps, breaks, and micronuclei. Double strand breaks can be repaired via two to three major pathways depending on damage type and cell stage among other conditions, and the mutation type and frequency depends on the repair mechanism employed.

Double strand breaks generated from the stalling or collapse of replication forks around lesions or single strand break are processed using homologous recombination (HR) (Rothkamm and Lobrich 2003; Ceccaldi, Rondinelli et al. 2016). Because these breaks happen during replication, an identical sister chromatid may be present to use as a template and repair can restore the original sequence. However, HR can also occur using a non-sister chromatid or in the case of repeated regions can use another stretch of DNA as a template, resulting in loss of homozygosity, inversions, deletions, and duplications (Saleh-Gohari, Bryant et al. 2005; Shrivastav, De Haro et al. 2008). HR may also increase point mutations (Shrivastav, De Haro et al. 2008).

Double strand breaks occurring in all parts of the cell cycle may be processed by non-homologous end joining (NHEJ) (Rothkamm and Lobrich 2003), which can alter the nucleotide sequence of the two broken ends to achieve a fusible template leading to point mutations, deletions, and insertions (Ceccaldi, Rondinelli et al. 2016). NHEJ can fuse incorrect ends within or between chromosomes, resulting in major changes including translocations, deletions, inversions, and duplications. Compared with HR, NHEJ is considered to be more likely to generate mutations, particular the resection dependent classical or alternative end joining pathways (Ceccaldi, Rondinelli et al. 2016).

Complex damage

Complex damage delays repair and increases double strand breaks, increasing the likelihood of mutation. Clustered lesions or single strand breaks are processed more slowly than non-clustered lesions, increasing the number of lesions that will undergo replication and potentially generate mutations in daughter cells (Dianov, Timchenko et al. 1991; Eccles, O'Neill et al. 2011). Clustered damage made of closely opposed lesions and/or single strand breaks can also create double strand breaks (Chaudhry and Weinfeld 1997; Vispe and Satoh 2000; Yang, Galick et al. 2004; Schipler and Iliakis 2013; Sharma, Collins et al. 2016; Shiraishi, Shikazono et al. 2017). Complex damage involving double strand breaks is also repaired more slowly (Stenerlow, Hoglund et al. 2000; Schipler and Iliakis 2013; Lorat, Timm et al. 2016), and undergoes a form of NHEJ with excision that leads to increased translocations and deletions (Eccles, O'Neill et al. 2011; Sharma, Collins et al. 2016; Watts 2016).

Genomic Instability/Long term effects

Genomic instability is the prolonged appearance of DNA damage, chromosomal damage, and mutations. It is sometimes seen following agents that induce DNA damage including ionizing radiation, RONS, and NMU (Goepfert, Moreno-Smith et al. 2007; Kadhim, Salomaa et al. 2013; Stanicka, Russell et al. 2015). DNA damage occurring during genomic instability is associated with the appearance of mutations including deletions, inversions, and duplications (Murnane 2012; Kadhim, Salomaa et al. 2013; Sishc, Nelson et al. 2015).

High. It is generally accepted that DNA damage leads to mutations. Empirical support comes in part from the observation that agents which increase DNA damage also cause mutations, that DNA damage precedes the appearance of mutations, and that interventions that reduce DNA damage also reduce mutations. None of the identified studies measure both outcomes over the same range of time points. This constitutes a readily addressable data gap.

Ionizing radiation (IR) and reactive oxygen and nitrogen species (RONS) are both capable of causing DNA damage including lesions, single and double strand breaks, and these agents also cause mutations (Schiestl, Khogali et al. 1994; Kuhne, Rothkamm et al. 2000; Rydberg, Cooper et al. 2005; Dayal, Martin et al. 2008; Seager, Shah et al. 2012), and chromosomal aberrations (Choi, Kang et al. 2007; Jones, Riggs et al. 2007; Du, Gao et al. 2009; Buonanno, de Toledo et al. 2011; Patil, Rao et al. 2014; Fetisova, Antoschina et al. 2015; Padula, Ponzinibbio et al. 2016). DNA damage observed after IR and RONS precedes the appearance of mutations (Denissova, Nasello et al. 2012; Sharma, Collins et al. 2016) and precedes or is concordant with the appearance of chromosomal aberrations (Yang, Asaad et al. 2005; Du, Gao et al. 2009; Suzuki, Kashino et al. 2009; Patil, Rao et al. 2014; Padula, Ponzinibbio et al. 2016). Further evidence that mutations arise from DNA damage comes from agents that affect DNA repair: decreasing repair after IR maintains damage while decreasing translocations (mutations introduced by DNA repair) (Biehs, Steinlage et al. 2017).

Despite the generally accepted relationship between DNA damage and mutations, few studies uncovered in the literature for RONS or ionizing radiation measure both DNA damage and mutations in the same study (Denissova, Nasello et al. 2012; Sharma, Collins et al. 2016; Biehs, Steinlage et al. 2017) and none measure both key events at the same time points.

Mutations generally increase linearly with dose of DNA damaging agents (Sandhu and Birnboim 1997; Sharma, Collins et al. 2016), but multiple factors including DNA repair, bystander effects, and genomic instability can affect the shape of the dose-response. IR promotion of DNA repair mechanisms decrease major mutations (lethal recessive changes) at lower IR doses/dose rates in flies (0.2 Gy at 0.05 Gy/min gamma) (Koana and Tsujimura 2010). In contrast, non-targeted effects of IR contribute to supralinear responses at lower doses (Sandhu and Birnboim 1997; Hall and Hei 2003; Yang, Anzenberg et al. 2007). At higher doses (10-80 Gy) rearrangements from misrejoining (joining together of non-sequential DNA) increase linearly with dose for high LET IR, but supralinearly for low LET IR, attributed to the increase in the concentration and complexity of double strand breaks with LET  (Rydberg, Cooper et al. 2005).