DNA repair systems in the nuclear and mitochondrial genomes (nDNA and mtDNA) work in parallel, where nucleotidases (nucleotide “sanitisation” or “house-cleaning enzymes”) remove non-canonical ‘rogue’ nucleotides. Any defect in these nucleotidase genes may reduce the efficacy of DNA repair.
This thesis explores the hypothesis that polymorphic deficiency of the nucleotidase inosine triphosphate pyrophosphohydrolase (ITPase) – encoded by the gene ITPA – leads to increased risk of nDNA and mtDNA mutation. Deficiency of this enzyme has previously been regarded as ‘benign’. The overall aim of this thesis is to examine ITPase deficiency as a mechanism for producing cellular mutations, i.e. the accumulation of non-canonical nucleotides in the cell pool, by analyzing whether imbalances of cell nucleotide pools may lead to mis-incorporation of bases into nDNA and mtDNA, producing changes in cellular viability or transformation.
Four studies were established to examine the hypothesis:
The first study was aimed at developing methods to detect any small changes in nucleotide content in cell pools and in DNA. Two novel HPLC methods using mobile phases compatible with mass spectrometry were successfully established to analyse nucleotides: 1) direct cell nucleotide analysis using aqueous normal phase HPLC with a ‘Type-C’ silica hydride column; 2) enzymatic digestion of DNA coupled with reversed-phase HPLC and mass-spectrometry to analyse DNA for non-canonical nucleotide content. Together, these two methods were capable of providing accurate and sensitive detection of any small changes in nucleotide content in the cellular pool and detection of incorporation of non-canonical nucleotides into DNA. Further development of the direct analysis method using silica hydride HPLC with mass-spectrometry is warranted.
The second study was aimed at examine whether incubation of cells with non-canonical base/nucleotides induced cell injury or nDNA/mtDNA changes. Fibroblast cellular injury was induced in vitro by incubation with 3 purine non-canonical bases, 8-hydroxy-adenine (8-OH-Ade), 6-thioguanine (6-TG) and 8-hydroxy-guanine (8-OH-Gua), and its nucleoside 8-hydroxy-guanosine (8-OH-Guo).
In the first stage of the study, cell viability, apoptosis/necrosis assays, and hypoxanthine-guanine phosphoribosyltransferase (HPRT) sequencing (as a mutation ‘hot-spot’ nuclear gene) were used to examine cell injury and nDNA damage. This demonstrated: 1) incubation with 6-TG and 8-OH-Ade induced late-apoptosis and injury/death; 2) DNA integrity, assessed by HPRT exon sequencing, showed no defined mutations but a general increase in base mismatching during sequencing, indicating widespread mutations and mixed-cell populations; 3) whole cell DNA digestion provided evidence of non-canonical base/nucleoside incorporation of 6-TG, 8-OH-Gua and 8-OH-Guo, exacerbated by ITPase deficiency. It appeared that two mechanisms were acting on cells: incubation of cells with 8-OH-Ade and 6-TG caused metabolic stress leading to apoptosis. On the other hand, incorporation into DNA and induction of mutations was typical of 6-TG and 8-OH-Guo, exacerbated by ITPase deficiency even though the enzyme is not directly involved in the removal of these non-canonical purines.
In the second stage, mtDNA from normal, ITPase deficient or knockdown fibroblast cells were sequenced using the MitoChip v2.0 resequencing microarray for the examination of mtDNA damage. Two fibroblast lines carrying the mutant ITPA 94A allele were found to have high mtDNA mutation levels compared with wildtype ITPA cell lines. Overall there was no increase in the mtDNA mutation load in the presence of non-canonical purines except for 8-OH-Guo compared to wildtype and complete ITPase deficient cells. Heteroplasmic mutation of the mtDNA was particularly raised in the highly variable area of the D-loop.
The third stage aimed at examining whether chronic ITPase deficiency could affect mtDNA integrity in vivo. Peripheral leucocyte/bone marrow DNA samples from 85 patients suffering adult haematological malignancies (AHM) were examined by MitoChip. The ITPA 94C>A variant was associated with significantly increased mtDNA homoplasmic mutations, and a significant but lesser increase in heteroplasmic mutations. Importantly, this effect remained after removing all haplogroup variants. The results were consistent with reduced ITPase activity leading to an increase in mitochondrial abnormalities. A new paradigm for the evolution of AHM was invoked, where nucleotide imbalances produced by defects in a ‘house-cleaning’ nuclear gene may compromise cell integrity, predisposing the cell towards transformation.
The fourth stage, as part of the MitoChip studies, aimed at reducing ‘N-call’ designations when the GSEQ software fails to assign a base during mtDNA sequencing by MitoChip. Utilizing sPROFILER (free online software from Kothiyal et al., 2010) reduced the ‘N-calls’, and searching for mutations/indels was optimized by focusing on longer N-calls stretches which were then chosen for conventional Sanger sequencing. The analyses of 16 mtDNA samples from AHM patients found 7 insertions and 12 point-mutations not previously recorded. This demonstrated that N-calls may mask unique mutations.
In conclusion, the evidence supported the hypothesis that ITPase deficiency is not ‘benign’, but may lead to mutations in both nuclear and mitochondrial DNA, particularly in older people, i.e. age-onset disease. The research alters our understanding of the aetiology of AHM and suggests that nucleotidase defects should be screened in selected cohorts.