Hypothesis: Local dNTP depletion as the cause of microsatellite repeat instability during replication (Comment on DOI 10.1002/bies.201200128)

Abstract

Nucleotide depletion in vivo is correlated with global genomic instability and tumorigenesis 1-3. In contrast, localized instability of low complexity microsatellite sequences causes multiple neurodegenerative diseases (e.g. Huntington's disease, myotonic dystrophy, spinocerebellar ataxias), depending on the host gene. CNG microsatellites are particularly prone to the formation of imperfectly base paired hairpin structures in vitro 4 and in vivo 5, and are sites of genome instability exacerbated by replication fork slowing. Current models of microsatellite instability invoke the formation of hairpin structures during replication, repair or transcription. In replication or repair models, the microsatellite sequence or structure directly inhibits DNA polymerase movement. Contractions arise via two mechanisms: (i) slowing of the polymerase and consequent increase in hairpin formation in extended leading or lagging strand template ssDNA at the replication fork; (ii) threading of the leading strand template through the MCM helicase without concurrent polymerization. Expansions are provoked by nascent strand hairpins, because the repeat sequence directly causes polymerase ‘stuttering’ and primer slippage. In contrast, the new model presented in the current issue of BioEssays proposes that polymerase slowing and nascent strand hairpins are not direct effects of the repeat on the polymerase; rather they are indirect consequences of local nucleotide depletion, which results from a local stoichiometric imbalance in nucleotide incorporation into the nascent DNA (repeat sequences manifest such imbalanced stoichiometries) 6. The hypothesis predicts that nucleotide diffusion through the cell should be slower than the rate of precursor consumption at an individual replication fork. Although nucleotide diffusion rates have not been determined experimentally, based on the turnover of nucleotide pools in vivo, it is argued that the rate of nucleotide incorporation may be equal to or greater than the rate of diffusion. Additionally, replication is thought to occur in congregated replication forks, or factories: this raises the concern that if several forks simultaneously draw nucleotides from a common pool, imbalanced incorporation by a single fork would not have significant impact. However, the author presents evidence from ultra-high resolution microscopy that individual replication factories may contain only two forks, hence isolating the effects of nucleotide depletion. Trinucleotide repeat expansion or contraction in bacteria or yeast have shown significantly greater instability for CNG repeats than GNC repeats, possibly due to differences in the structure, stability or downstream metabolism of the respective hairpins. Dr. Kuzminov suggests in vitro tests of the nucleotide depletion model by analysis of replication in the response to differing nucleotide doses. Since nucleotide depletion by replication of CTG or GTC repeats should be similar, an alternative in vivo test of whether nucleotide depletion is the immediate cause of fork slowing would be to compare the distribution of replication intermediates by 2D gel electrophoresis during replication of plasmids containing CTG and GTC isomeric repeat sequences. Verification of the hypothesis that local nucleotide depletion potentiates microsatellite instability could lead to novel approaches to prognosis, prevention or therapy of neurodegenerative disorders.