The Mirkin laboratory uses a variety of models to understand DNA function and integrity in normal physiology and in several disease states.
Uncontrollable expansions of trinucleotide repeats lead to more than two dozens human hereditary neurological disorders, including Fragile X mental retardation, Huntington's disease, myotonic dystrophy, Friedreich's ataxia, etc. The molecular mechanisms of repeat expansions have, therefore, attracted a very broad attention.
Our lab got interested in this problem after we discovered, using two-dimensional electrophoretic analysis of the replication intermediates, that replication fork stalled within expandable repeats in vivo. This repeat-mediated fork stalling was observed in bacterial, yeast and mammalian cells. There was a good agreement between the repeat lengths, causing replication blockage in our systems, and their expansion thresholds in human pedigrees. Altogether, these results led us to hypothesize that abnormal replication of expandable repeats could be the cause of their instability.
Subsequently, we have developed a yeast experimental system, which allowed us to analyze large-scale repeat expansions similar to that observed in human pedigrees. This system uncovered several principal features of the repeat expansion process. First, the rate of repeat expansions increased exponentially with their lengths. Second, the median expansion step appeared to correspond to the median size of an Okazaki fragment. Third, vast majority of genes involved in repeat expansions, which came out from unbiased genetic screens, encode protein components of the replication fork. Finally, while every repeat studied so far has a propensity to expand, expansion rates are much higher for structure-prone DNA sequences.
Figure 1. Proposed mechanism for large-scale repeat expansions. Complementary strands of an expandable repeat are shown in red and green. Blue hexameric ring depict the replicative DNA helicase. Purple star designates fork-pausing complex. Yellow circles symbolize leading and lagging DNA polymerases. Gray square shows template switching protein(s).
Based on these observations, as well as data from many other labs, we proposed a model for the large-scale repeat expansions based on the template-switching during DNA replication. It hypothesizes that during replication of a repetitive DNA run (Fig 1A), leading strand DNA polymerase can accidentally (~10–5 per replication) switch its template to continue DNA synthesis along the nascent lagging strand (Fig 1B). Notably, in a long repetitive run, each sequence in the nascent lagging strand sequence is repeated multiple times in the leading strand template. This could make the template switch more feasible, compared to the unique DNA sequences, as an unwound portion of the repetitive leading strand can pair with multiple points along the repetitive lagging strand. After reaching the end of the Okazaki fragment (Fig 1C), the polymerase has to switch back to its primary leading strand template in order for replication to continue. This switch results in the appearance of an expanded repetitive run within the leading DNA strand (Fig 1D).
We are currently continuing to substantiate this model in a yeast experimental system, as well as are investigating its applicability to mammalian cells. In the long run, yeast and mammalian experimental systems for repeat expansions could help searching for drugs that affect the rates of expansions or contractions. These drugs would be invaluable for treatment of the debilitating disorders, caused by expandable repeats.
Since transcription and replication share the same template, occasional collisions between the two machineries are inevitable and can interfere with both processes. We have recently found that the head-on collisions with elongating RNA polymerase is much more detrimental for the replication fork progression in vivo than the co-directional collisions. Furthermore, we have proven that these collisions are caused by the direct physical interaction of the two machineries, rather than the long-range alterations of the DNA template. These results, combined with the data on the preferred co-directional alignment of transcription units with the direction of replication in prokaryotes, have led us to suggest that the main disadvantage of the head-on collisions could be in their inhibitory effect on DNA replication.
Figure 2. Collisions between replication and transcription in bacteria. In case of head-on collisions, replicative DNA helicase, depicted by the green hexagon, collides with the front end of RNA polymerase (golden convex). In case of co-directional collisions, the leading strand DNA polymerase (red oval) collides with the rear end of the RNA polymerase.
Besides collisions with elongating RNA polymerases, we study the effects of the transcription initiation or termination complexes on the replication fork progression. This could be even more important, since most genes are not actively transcribed during DNA replication. We have recently found that the steadfast transcription initiation complexes inhibit the replication fork progression in an orientation-dependent manner, during head-on collisions. Transcription terminators also appeared to attenuate DNA replication, but in the opposite, co-directional orientation. Notably in both instances, the replication fork is stalled immediately after passing the coding region. Transcription regulatory signals, thus, serve as “punctuation marks” for DNA replication in vivo by attenuating the replication fork progression, as it has traversed the coding areas. This attenuation could provide an extra time for the repair or recombination machineries to clear the coding areas off the newly acquired mutations.
This project is now developing in several directions. First, we are expanding our collision studies from the E. coli into yeast S cerevisiae and, eventually, mammals. Second, we plan to experimentally determine mutation rates in the transcribed areas that are replicated head-on or co-directionally. This study will be carried out in yeast, using selectable genes driven by the S-phase-specific promoters. Finally, we are starting a major bioinformatics project, aimed at estimating the sequence divergence between genes in numerous bacterial genomes depending on their positioning relative to the direction of the replication.
More than two decades ago, we have characterized an unusual three-stranded DNA structure - H-DNA, or triplex - formed by homopurine-homopyrimidine mirror repeats.
Figure 3. The structure of the triple stranded H-DNA. The two complementary strands of a homopurinehomopyrimidine repeat are colored in red and gray, while flanking DNA is colored green. The structure is called H-y when the red strand is homopyrimide, and H-r if when it is homopurine. One can see that the red and green strands in this structure are not linked, i.e. formation of H-DNA is topologically equivalent to an unwinding of the entire homopurine-homopyrimidine repeat.
Little did we know at a time that one of those repeats, (GAA)n/(TTC)n, will be eventually implicated in the development of a hereditary human disorder - Friedreich’s ataxia. We have since found that formation of unusual DNA structures by H motifs during the DNA synthesis in vitro could block various DNA polymerases. Remarkably, the polymerase itself triggered the formation of an unusual DNA structure that subsequently inhibited it. Simple DNA repeats including, but not limited to H motifs were thus called “suicidal sequences” for the DNA polymerization. It has now become apparent that various DNA repeats could serve as suicidal motifs for the RNA polymerase, as well. Considerable efforts are currently being devoted to the detection of DNA triplexes and other unusual DNA structures inside living cells and elucidating their biological roles in norm and disease.