SINGLE MOLECULE IMAGING TECHNOLOGIES
Several independent real-time single molecule-imaging platforms have been assembled in the Fishel Lab that are capable of visualizing a wide range of interactions at low msec time resolution with or without the introduced mechanical force. Techniques and instrumentation include: 1.) single molecule Förster resonance energy transfer (smFRET; 1-10 nm protein-protein or protein-DNA interactions); 2.) prism-based single molecule total internal reflection fluorescence (smTIRF; 50-4000 nm tracking); 3.) single molecule magnetic tweezers (smMT; mechanical forces of 1-20 pN and/or up to 20 plectonemic supercoils); 4.) single molecule flow-stretching with objective TIRF (smFS-TIRF; combined force and component tracking); and 5.) single molecule cellular illumination (smCI; cellular tracking). A corollary to the development of these single molecule-imaging systems is that we have solved the technical challenges associated with tracking and quantifying 50-150 individual particles simultaneously. The power of these multiple systems is that they may be used to quantitatively compare predictions that arise from different real-time measures. An example of one of our most effective single molecule instrumentation platforms prism-based smTIRF is shown below.
Representative Example of Prism-Based smTIRF Microscopy. A) diagram of the pellin-broca prism showing incoming and totally reflected laser light. A DNA molecule is attached to a quartz flow-cell surface below the prism where a TIRF-generated evanescent field penetrates ∼100-200 nm exciting laser emission-specific fluorophores. B) After imaging the DNA is stained with a Sytox dye to locate its position and measure it length. C) Overlay of real-time images with the stained DNA reveals single molecules that bind and diffuse along the DNA molecule. In this case we are imaging the retroviral intasome of Prototype Foamy Virus searching for a site to integrate.
The development of new single molecule imaging platforms as well as improved computational tools for real-time single molecule tracking and quantitative analysis has been in collaboration with Dr. Jong-Bong Lee (POSTECH, Korea). Physics or biophysics graduate students and postdoctoral fellows interested in this project should contact Dr. Fishel directly.
MISMATCH REPAIR
Mismatched nucleotides in cellular DNA result from polymerase misincorporation errors, recombination between heteroallelic parental DNAs and from chemical or physical damage to nucleotides . Mismatch repair (MMR) resolves these anomalous nucleotides and has been conserved in most terrestrial organisms as one of several mechanisms to maintain genome integrity . In the absence of MMR, spontaneous mutation rates increase 100-1000 fold. This hypermutable state has several important biological consequences. MMR-deficient bacteria appear better able to rapidly adapt to environmental stresses, resulting in increased competitiveness and antibiotic resistance. In humans, MMR mutations are the cause of the most common cancer predisposition, Lynch syndrome or hereditary non-polyposis colorectal cancer (LS/HNPCC), as well as sporadic colorectal, endometrial, ovarian and upper urinary tract tumors. Cancer progression in MMR defective tumors results from a similar rapid cellular adaptation and increased competitiveness.
The core MutS homolog (MSH) and MutL homolog (MLH/PMS) components of MMR have been conserved across terrestrial biology (see Table above). In bacteria the progenitor MutS and MutL proteins function as a dimers, while in eukaryotes including human the MSH and MLH/PMS proteins function as heterodimers, each of which is coded by a separate gene. The vast majority of cancer-causing LS/HNPCC mutations have been found in the human MSH2 and MLH1 genes, with a small fraction found in their heterodimeric partners MSH6 and PMS2 that have been associated with atypical cancer-predisposition families.
The goal of the mismatch repair project is to fully understand the mechanics of MMR in both human and the bacteria Escherichia coli. The human studies should provide enough information to assess the affect of a growing number of missense mutations on the function of the MSH2-MSH6 and MH1-PMS2 heterodimers, and to what extent they might cause cancer. In addition, there are two fundamentally different mechanisms for ensuring the MMR strand break is located on the error-containing DNA strand. The majority of organisms (including humans) most likely utilize persistent strand breaks associated with DNA replication. E.coli is a relative to subset of γ-proteobacteria that has recently evolved DNA adenine methylation (Dam) and MutH to specifically introduce a DNA break onto the newly replicated strand. These include a number of human pathogens such as Haemophilus influenza, Yersinia pestis, Vibrio cholerae, Shigella dysenteriae, and Salmonella typhimurium. There are a number of synthetic-lethal mutations with MMR components including simultaneous mutation of Dam with recombination repair genes such as RecA that may provide a useful antibiotic strategy once the mechanisms are fully understood.
Previous work from our group showed that MSH and MLH/PMS proteins form a cascade of sliding clamps on DNA containing a mismatch. The MSH proteins contain a highly conserved Walker A/B-type ATPase domain and initially recognize a mismatch, which provokes ATP binding and the formation of the first sliding clamp. The MSH sliding clamp diffuses randomly along the DNA much like a donut on a string until it interacts with N-terminal MLH/PMS that contains a conserved GHKL ATPase. The formation of an MSH-MLH/PMS complex alters the diffusion properties from rotation-independent diffusion to rotation-coupled diffusion along the DNA, which facilitates wrapping of the remaining MLH/PMS peptides around the DNA forming the second sliding clamp. These are dynamic and redundant processes where diffusion away from the mismatch permits the loading of multiple MSH and MLH/PMS sliding clamps. In E.coli (see hypothetical mechanism below) the MutL sliding clamp binds the MutH endonuclease, significantly aiding in the search and incision at hemimethylated GATC sites. Separately, a MutL sliding clamp may capture a UvrD helicase protein at the hemimethylated GATC strand scission, substantially increasing its unwinding processivity. This dramatic increase in processivity is sufficient to remove the mismatch-containing strand between two nicked hemimethylated GATC sites. The ssDNA exonucleases (ExoI, ExoVII, ExoX and RecJ) are not required for strand excision, but support the removal of at least one nucleotide from the nicked hemimethylated GATC site to ensure it is not sealed by DNA ligase before an MMR event is completed. Cascading MSH and MLH/PMS sliding clamps are formed during human MMR, but the details of strand specific excision remain largely speculative.
A complete model for strand specific excision by E. coli (Ec) mismatch repair. a) Cascading EcMutS (blue) and EcMutL (brown) clamps recruit EcMutH (green) to generate multiple strand scissions (red lightning bolts) on a mismatched DNA at hemimethylated GATC sites. b) EcMutL captures EcUvrD (yellow) near an EcMutH strand scission tethering it to the mismatched DNA where it randomly unwinds (red arrow) and rezips (green arrow) the DNA that is alternately bound and released by EcSSB. c) When the EcMutL–EcUvrD unwinding reaches an adjacent EcMutH GATC incision site, the intervening fragment is released creating an EcSSB bound gap. d, e) While displacement of GATC→GATC may occur randomly between adjacent sites, MMR is not completed until the segment containing the mismatch is released. f) The replicative DNA polymerase and ligase complete MMR by resynthesizing the gaps generated by EcMutL–EcUvrD strand displacement.
RECOMBINATION REPAIR AND MEIOSIS CHROMOSOME PAIRING
Recombination repair (RR) fixes double strand breaks (DSBs) in DNA that are cause by chemical or physical damage. In the absence of repair, a single DSB can provoke cell death. Genome instability resulting from defects in RR have been linked to hereditary breast cancer (BRCA1/2) as well as hematopoietic and other solid tumors (Ataxia telangiectasia mutated, ATM; Nijmegen breakage syndrome, NBS; Fanconi anemia, FANC; Bloom syndrome, BLM) among others. RR engenders a complex cascade of responses that include cellular signaling integrated with the physical processes of DSB repair. RR can result in high fidelity homologous recombination repair (HRR) as well as lower fidelity non-homologous end joining (NHEJ). Our group focuses on HRR where the repair event uses homologous chromosomal sequences to bridge the DSB.
Chromosomes where HRR occurs are composed of protein-DNA composites (chromatin) that compact ~ 1 meter of DNA into a cellular nucleus that is less than 10 micrometer in diameter. The most obvious repeating units of chromatin are nucleosomes composed of ~147 bp of duplex DNA wrapped around an octamer core containing dimers of four histones (H2A, H2B, H3 and H4). HRR biochemical reactions must manage the disrupted chromatin on the broken donor DNA in order to search and pair with the assembled chromatin of a bridging homologous acceptor DNA. Deficiencies in any one of the multiple enzymatic steps will affect the outcome of RR and ultimately affect genome stability. Our group has focused on understanding the biophysical processes associated with the formation and activities of the HRR protein complex, particularly on well-defined chromatin substrates.
Remarkably, very little is known about the detailed biophysical events associated with HRR on chromatin. This is largely because the fundamental repair reactions have not been reconstituted with well-defined chromatin, nor has there been any attempt to accurately account for the disposition of individual nucleosomes that structure the chromatin of the donor and/or the acceptor DNAs during HRR. We have developed a number of single molecule imaging methods to dissect the mechanics of HRR within chromatin. One of those techniques is shown below and details of RAD51 unraveling the nucleosome DNA from the histone octamer examined on single DNA molecules by smFRET (from: Nucl.Acids Res. 45:685 2017). Projects involving HRR are focused on describing the detailed mechanics of homologous pairing in model chromatin in collaboration with Michael Sehorn (Clemson University).
Visualizing the nucleosomal DNA unwrapping by human RAD51 from the entry-exit region. (A) An illustration of the single molecule mononucleosome experimental setup to visualize human RAD51 (HsRAD51) catalyzed nucleosome remodeling relative to the entry-exit region. Increasing concentrations (250–2000 nM) of RAD51 with 0.5 mM ATP were introduced onto surface immobilized mononucleosomes. Very small changes in the distance between the Cy3 fluorophore “donor” and the Cy5 fluorophore “acceptor” will reduce the FRET between them over time. (B) Population analysis showing the percentages of static and dynamic molecules obtained at the indicated experimental conditions. The total number of molecules (N) analyzed for each experimeMMR Model nt is also shown at the bottom of the corresponding pie chart. (C–F) Representative single molecule dynamic FRET trajectories, post-synchronized FRET density plots and time averaged FRET histograms for all the RAD51 concentrations tested. RAD51 concentration and the number of molecules (n) analyzed are indicated in each panel. Real time infusion of RAD51 and ATP is marked by vertical dotted lines at 60 s. The intensity bar indicates the relative frequencies of FRET values in post-synchronized FRET density plots (low = yellow, high = red).
The pairing of homologous chromosomes during meiosis is a complex process fraught with many pitfalls. Chromosome pairing is initiated in Prophase I by the Spo11 gene product, which actively introduces DNA double stranded breaks (DSBs) into the sister chromatids. The landscape and number of meiotic DSBs appears dependent on localized chromatin structure and the ATM damage checkpoint process. The repair of these DSBs by the nearest sister is suppressed by the formation of meiosis-specific lateral elements between the chromatids. This leaves the homologous chromosome as the usual sequences exploited for DSB repair to restore the integrity of the genome. The DSBs are ressected by a 5’→3’ exonuclease. The resulting 3’ single-stranded DNA (ssDNA) end is then used in a search, pairing and strand-invasion reaction with the homologous chromosome that requires RAD51: a homolog to the E.coli recombination-initiation protein RecA. The ssDNA binding protein RPA is an essential RAD51 cofactor. Mutation of Spo11 or RAD51 result in a dramatic reduction of homologous chromosome pairing, a high frequency of meiosis I non-disjunction, and gamete inviability. Immunofluorescent protein localization (IPL) has been used to determine the timing and abundance of recombination protein(s) on meiotic chromosomes. These results suggest that upward of 400 DSB sites are formed that contain RAD51 and RPA beginning in leptotene.
A Model for MSH4-MSH5 and MLH1-MLH3 in Crossover Stabilization that Eventually become Chiasmata.
Our group is interested in the biophysical role of MSH4-MSH5 and MLH1-MLH3 in meiosis. Mutation of the yeast, worm and mouse MSH4 or MSH5 genes do not display a mutator phenotype, supporting the notion that they do not function in MMR where most MutS and MutL homologs operate. However, mutation of MSH4 or MSH5 does result in defective meiosis and infertility that results for an inability to form stable initial pairing structures that ultimately results in a high frequency of meiotic chromosome nondisjunction. IPL studies indicate that MSH4 and MSH5 proteins arrive later than Rad51/RPA; but at an equivalent abundance. However, unlike Rad51/RPA, some MSH4 and MSH5 foci persist and appear to be fundamental components of chiasmata formation. Chiasmata eventually contain the MutL homolog heterodimer MLH1-MLH3. Mutation of either MLH1 or MLH3 results in gametes that do not progress to chiasmata and display a high frequency of meiotic chromosome nondisjunction. Based on our studies of bacterial and human MSH and MLH/PMS, we have developed a hypothesis where the MSH4-MSH5 sliding clamp loaded onto nascent Holliday Junctions provides a platform for sequential loading of MLH1-MLH3 sliding clamps (see diagram to the right). We are actively developing single molecule imaging technologies to visualize this process in vitro and in vivo. The lifetime and disposition of MLH1-MLH3 sliding clamps and their relationship to meiosis I chromosome segregation remains a significant question?
RETROVIRAL INTEGRATION
HIV infection continues to be a worldwide concern. As a retrovirus, HIV is defined by its reverse transcriptase (RT) and integrase (IN) activities. Both of these enzymes are current targets of anti-retroviral therapies (ART). However, the highly mutagenic nature of the RT frequently results in ART resistance mutations underlining the need for novel drugs and drug targets. Stable integration of HIV into the genome of a host cell is essential for a productive infection. The precise mechanism of HIV integration in vivo remains elusive. We have used smTIRF and smMT to visualizing the kinetic processes associated with retroviral integration in real time. Purifed intasomes (IN protein in complex with model retroviral cDNA) of HIV, Prototype Foamy Virus (PFV) and murine mammary tumor virus (MMTV) are under scrutiny by single molecule imaging methods in collaboration with Kristine Yoder.
CANCER CHEMOPREVENTION
Non-steroidal anti-inflammatory drugs (NSAIDs) are a structurally diverse family of compounds that are effective in the prevention of several types of human cancer including colorectal cancer (CRC). Acetylsalicylic acid (ASA), commonly known as aspirin, is the archetype of the NSAID family. A number of epidemiological studies have demonstrated an inverse relationship between ASA use and the incidence of CRC. A recent clinical analysis reported that regular ASA use was associated with a lower risk of cancer-specific mortality in individuals already diagnosed with colorectal cancer. Moreover, a variety of animal models have confirmed that administration of various NSAIDs results in fewer tumors. While several explanations have been proffered over the years, a clear elucidation of the molecular mechanism(s) responsible for NSAIDs cancer chemoprevention efficacy has been elusive. The Colorectal Adenoma-Carcinoma Prevention Program (CAPP) has examined the potential of ASA to reduce colorectal neoplasia in LS/HNPCC carriers in a randomized trial. The end-point for these studies was the detection of LS/HNPCC cancers during yearly colonoscopy screens. Remarkably, the benefits for LS/HNPCC patients only begins after 5 yrs of ASA exposure and results in at leasta 2-fold decrease in the rate of tumor formation compared to the placebo-treated cohort.
A substantial NSAID chemopreventive effect on a common genetic cause of cancer that is recapitulated from the mouse model to the human disease provides a unique tool capable of dissecting the molecular mechanism(s) responsible for NSAIDs cancer chemoprevention efficacy. We use a mouse model of LS/HNPCC where exon 12 of Msh2 is flanked by LoxP recombination sites (Msh2flox/flox). The Msh2 deletion is then targeted to the intestinal epithelia by tissue-specific expression of Cre recombinase under the control of the villin promoter (VpC+/-). Tumorigenesis is confined to the intestines of these VpC-Msh2 mice and has all of the pathological hallmarks of HNPCC-like tumors including microsatellite instability (MSI). Moreover, like their human counterparts, the LS/HNPCC mice develop (and ultimately succumb to) 1-2 tumors that appear to progress rapidly. Because the tumor numbers are small even modest effects on tumor initiation and/or progression can have a significant effect on the survival of these LS/HNPCC mice.
Survival of VpC-MSH2 Mice Treated with NSAIDS. Average lifespan in days is shown in parenthesis. Left) Kaplan-Meier survival curves for untreated (grey; 362 days), aspirin treated (orange, 431 days), naproxen treated (green, 540 days) and wild type (black, 764 days) mice. Right) Kaplan-Meier survival curves for VpC-MSH2-Tgfβ-RII mice that were untreated (black, 388 days), aspirin treated (orange, 414 days) and naproxen treated (green, 425 days). Light background is KaplanMeier survival from the right panel for overlay comparison.
Treating the VpC-Msh2 mice with aspirin or naproxen significantly prolongs their lifespan (see above left; compare grey with orange and green). The prolonged survival of the aspirin treated mice appear qualitatively similar to CAPP studies with LS/HNPCC patients. Remarkably, this analysis suggests that naproxen treatment with LS/HNPCC patients may result in a significantly better clinical response; a prediction that has led to additional randomized clinical trials. Perhaps more interesting is that adding a Tgfβ-RII mutation that knocks-out Tgfβ signaling (VpC-Msh2-Tgfβ-RII), completely eliminates the NSAID effect on survival. Which aspects of the Tgfβ pathway are responsible for the NSAID chemoprevention effect are unknown and under active investigation. The goal of this project is to identify more specific clinical targets by understanding the mechanism(s) of tumor chemoprevention.