Microtubule Stress Modifies Intra-Nuclear Location of Msh2 in Mouse ...

[Cell Cycle 3:5, 662-671; May 2004]; ©2004 Landes Bioscience

Microtubule Stress Modifies Intra-Nuclear Location of Msh2 in Mouse Embryonic Fibroblasts

Zealand

of Biological Sciences; Victoria University of Wellington; Wellington, New

Received 03/01/04; Accepted 03/15/04

No tD ist r

Bio

Previously published online as a Cell Cycle E-publication: http://www.landesbioscience.com/journals/cc/abstract.php?id=855

sci

*Correspondence to: Paul J. Smith; Department of Pathology; University of Wales College of Medicine; Cardiff, CF14 4XN, UK; Tel.: +44.0.2920742730; Email: [email protected]

Do

School of Biosciences; Cardiff University; Cardiff, UK

ce.

4School

The maintenance of genomic stability in mitotic and meiotic cycles through mismatch repair (MMR) demands the coordination of MMR functions with multiple processes including cell cycle traverse, linked changes in microtubule dynamics, protein translocation at chromatin sites and checkpoint activation. We have studied changes in the intracellular location of the MMR protein Msh2 in response to mitosis, microtubule disruption by colcemid and DNA damage induction by cis-platin in mouse embryonic fibroblasts (MEFs). Image analysis indicated that MEFs have a normally high nuclear retention of Msh2 during interphase with a precipitous dispersal of protein from chromatin sites into the cytoplasm at mitosis. Dispersal was also observed in cisplatin- and colcemid-treated interphase MEFs without any change in the overall Msh2 levels throughout the cell cycle. There was no evidence of co-localization of the punctate cytoplasmic Msh2 foci with any microtubule structures and knockout of Msh2 altered neither the extent of microtubule disruption nor the functional activation of the spindle assembly checkpoint by colcemid. Critically, extranuclear relocation of protein did not alter the ability to mount an Msh2-dependent G2 checkpoint delay in response to cisplatin-induced DNA damage. Depletion of the nuclear pool of Msh2 protein in cells undergoing dispersal was found to involve a rapid relocation of protein from AT-rich chromatin sites as defined by coassociation studies exploiting a newly-characterized base-pair preference of the fluorescent DNA binding probe DRAQ5. The study reveals the unexpected mobility of MMR protein pools during the MEF cell cycle and in response to different stress-inducing agents. The results link for the first time microtubule-integrity with intra-nuclear Msh2 protein dynamics. The high nuclear retention of Msh2 in interphase MEFs is in contrast to human tumor cells while the observations on protein dispersal suggest that only low levels of nuclear-located Msh2 are needed for G2 checkpoint activation by DNA damage.

en

Departments of Pathology1 and Medical Biochemistry4; University of Wales College of Medicine; Heath Park; Cardiff, UK

ibu

ABSTRACT

Nuria Marquez1 Sally C. Chappell1 Owen J. Sansom2 Alan R. Clarke2 Paul Teesdale-Spittle3 Rachel J. Errington4 Paul J. Smith1,* 3Cardiff

te

Report

KEY WORDS

La

ABBREVIATIONS

nd

es

Msh2, cell cycle, mouse embryonic fibroblasts, time-lapse microscopy, co-localization, colcemid, mismatch repair

20

04

MMR mismatch repair MEFs mouse embryonic fibroblasts APC anaphase promoting complex PBS phosphate-buffered saline

ACKNOWLEDGEMENTS

©

This work was partly supported by the UK MRC, AICR, Amersham Biosciences Plc, UK Joint Research Councils, BBSRC and Kinetic Imaging Ltd. P.J.S. is a director of Biostatus Ltd.

662

INTRODUCTION

The DNA damage response machinery and chromatin assembly/condensation mechanisms appear to share common molecular pathways.1-2 Some of these pathways must integrate their activity with spindle dynamics and the complex intracellular signaling network that operates to inhibit the multi-subunit ubiquitin ligase APC (anaphase promoting complex) during mitotic checkpoint activation.3-4 Coordination of the nuclear translocation of activated proteins and their assembly at sites of action is also an emerging feature of the early phases of response to DNA damage or microtubule stress. For example, integrating proteins such as 53BP1 appear to link stress pathways and microtubule function in the cell cycle.5 This integrating protein carries a kinetochorebinding domain becoming hyperphosphorylated in mitotic cell and undergoes an even higher level of phosphorylation in the presence of DNA lesions in response to spindle disruption with colcemid.5 Mismatch repair (MMR) proteins help to maintain the stability of eukaryotic genomes through the repair of errors made during DNA replication and the monitoring of homologous recombination. Accordingly we have hypothesized that MMR protein dynamics could be affected by both microtubule and DNA damage-originating stress. In humans the MMR damage recognition process is enabled through the hMutS alpha (hMsh2/hMsh6) or the hMutS beta (hMsh2/hMsh3) complexes that assemble at mismatch sites and subsequently act to recruit other repair functions highlighting the potential for complexes to assemble at discrete chromatin sites.6 Defective repair of mismatched bases can lead to an increased mutation rate and microsatellite instability resulting in complex contributions to the oncogenic process when checkpoint surveillance fails.7,8 Coprecipitation experiments in human cells have suggested that the MutSalpha complex is cytoplasmically preformed and subject to Cell Cycle

2004; Vol. 3 Issue 5

MSH2 AND COLCEMID

nuclear translocation dependent on alkylation damage and protein phosphorylation.9 Here we have specifically addressed the intracellular location of Msh2 protein foci in mitotic and interphase mouse embryonic fibroblasts under microtubule and DNA damage stress with a specific exploration of changes in the intranuclear associations of Msh2 with DNA/chromatin. Microtubule-associated pathways exist for nuclear translocation of important stress-response proteins such as p53.10 Such enhanced nuclear targeting of p53 and subsequent activation of p53-downstream target genes10 is associated with localization of p53 to microtubules and subsequent transport in response to DNA damage via the motor protein dynein.10 Furthermore, in human and rat fibroblasts there is evidence for the triggering of p53 activation and the arrest of cells at G1/S phase by colcemid-induced disruption of microtubules via an integrin-Raf-MAP kinase cascade. This cascade is associated with focal adhesion formation.11 Interestingly colcemid-treated MEFs are less likely to activate p53-driven pathways apparently due to the transience of integrin-induced Erk upregulation.11 Thus the MEF system permits an examination of the effects of agents such as colcemid on cell cycle checkpoint modulation and changes in the intranuclear dynamics of MMR proteins without changes being driven by collateral effects through p53 activation per se. We have studied the effects of microtubule disruption, sufficient to activate the mitotic spindle checkpoint, together with the sub-cellular localization of Msh2 protein using confocal microscopy and a novel labeling approach for the nuclear compartment. The development of fluorescent DNA-binding probes with base-pair/sequence preferences offers the opportunity to map intranuclear sites for protein location. A recently described novel cell-permeant anthraquinone (DRAQ5) has been shown to have a high degree of DNA selectivity in both live and fixed cell preparations as evidenced by its exclusively nuclear location.12 The far red fluorescence signal of DRAQ5 maintains spectral separation from the visible range fluors (e.g., fluorescent antibody labels with high quantum yield such as AlexaTM dyes) for use in mapping protein location in situ. Here we have studied the base pair preference of DRAQ5 to characterize intranuclear translocation events for Msh-2 protein. The results indicate that MEFs are highly geared for nuclear retention of Msh2 and that dynamic changes occur under microtubule and DNA damage stress.

MATERIALS AND METHODS Cells and Chemicals. Wt and Msh2-/- MEFs were derived from 13 day-old embryos as previously described (Marquez N, et al.; 2003). After 3 days in culture the MEFs were frozen down in 10% DMSO/FCS dilution. When in culture, cells were maintained in BHK-21 medium, containing 10% foetal bovine serum supplemented with penicillin and streptomycin, and grown in an atmosphere containing 5% CO2. Cells were plated at a density of 3.104 cells/mL (total volume of 1mL in 24 well plates, and 4 mL in 6 well plates) in advance of experiments. Colcemid® treatment was added at concentration of 0.5 µg/mL for designated times as described. Cisplatin (Faulding Pharmaceuticals Plc, UK) was kept at room temperature at a concentration of 1 mg/mL. 10 µM of cisplatin was added to an appropriate well and incubated for 1 h at 37˚C and washed twice with fresh media. Calf thymus DNA, polydeoxyadenylic acid-polythymidylic acid (Poly[dA]. Poly[dT]) and, polydeoxyguanylic acid-polydeoxycytidylic acid (Poly[dG]. Poly[dC]) were purchased from Sigma Chemical Co and were used without further purification. Trizma base (Tris[hydroxymethyl] aminomethane) and NaCl were supplied from Sigma and used for buffer preparation without further purification. Msh2 and Microtubule Immuno-detection. Cells were grown on glass coverslips to 70-80% confluency. After a specific drug treatment cells were fixed for 30 min in 4% paraformaldehyde/PBS. Preparations were quenched

www.landesbioscience.com

for 10 min in 100 mM glycine and permeabilized in PBS containing 0.2% Triton X-100 for 30 min, all at room temperature. Cells were then subjected to a block in PBS with 1% BSA over-night at 4˚C. Double labeling with mouse anti–Msh2 (Oncogene) and rabbit anti-β-tubulin (Oncogene) was followed by an incubation with AlexaTM 488-conjugated anti-mouse IgG and AlexaTM 568-conjugated anti-rabbit IgG (Merk Biosciences Ltd; Nottingham UK) for 45 min. Coverslips were mounted in PBS containing 20 µM DRAQ5, a far-red fluorescent DNA dye (Biostatus Ltd.; Leicester, UK). Single optical section, dual channel images were acquired using confocal laser scanning microscopy (1024 MP; BioRad Microscience; Hemel Hempstead, UK) at wavelength selections of: 488 nm excitation with 520/32 nm emission for AlexaTM 488; 568 nm excitation with 595/35 nm emission for AlexaTM 568 and 633 nm excitation with 680/32 nm emission detection for DRAQ5. Image Analysis to Extract Msh2 and DNA Co-localization. To analyse the degree of Msh2 localization within the nuclear compartment a co-localization coefficient was extracted using BioRad software following Manders’ method13 this was defined as the degree of Msh2 signal overlapping with an objectively segmented nuclear compartment derived from the DRAQ5DNA signal. Control cell data yielded a coefficient ≥0.85 corresponding to cells showing a significant degree of nuclear localization for Msh2. Therefore cells given a coefficient value of 4n

Cells were continuously exposed to 0.5 mg/mL colcemid only, or also pre-treated with 10 mM of cisplatin for 1 h. At 24 h post-treatment, the cells were trypsinised, permeabilized and ethidium bromide stained. The preparations were analysed using a Becton-Dickinson FACscalibur flow cytometer and the results extracted to obtain cell cycle phase quantification.

Microtubule filaments were clearly visible demonstrating an intricate network, fine structures were detected within the lammelipodia with thicker filaments often running the entire length of single cells (Fig. 2A). Upon treatment with colcemid this distinct network collapsed around the nucleus, with short microtubule fragments and tubulin aggregates present throughout the cell; furthermore the Msh2 protein relocalized to the cytoplasm compartment (Fig. 2B). In neither condition was the Msh2 colocated with the microtubule structures. Subcellular Relocalization and Cell Population Expression of Msh2 in Wild-type MEFs Exposed to Colcemid. Continuous exposure of MEFs to colcemid generated a dispersal of the Msh2 signal after only a 1 h exposure (Fig. 3A). Extensive image analysis showed that the co-localization coefficient attributed to each treatment condition varied significantly. In control conditions all cells exhibited a coefficient ≥0.85 to 1.0 (i.e., ≥ 85 % of the cellular Msh2 signal was localized to the nuclear compartment in each cell). Upon treatment with colcemid the colocalization coefficient ranged from 0.38 to 1.0, with 52% of interphase cells exhibiting a coefficient value of less than 0.8 despite the continued presence of a nuclear membrane (Fig. 3B). Therefore a high proportion of interphase cells showed detectable levels of cytoplasmic Msh2 foci. This was greater than observed for 10 µM cisplatin treatment alone with 17% of interphase cells showing an increase, beyond the normal threshold, for Msh2 levels in the cytoplasm (Fig. 3B). Analysis of whole cell Msh2 expression throughout the cell cycle was also monitored by flow cytometry. The mean values for Msh2-fluorescence intensity (channel number) for wt sub-populations identified by relative DRAQ5 staining were: G1 cells 132 ± 45, 134 ± 48, 138 ± 45; S phase cells, 172 ± 43, 168 ± 47, 178 ± 45 and G2/M cells, 217 ± 39, 227 ± 40, 216 ± 42 for 0 h, 1 h and 4 h exposures to colcemid (0.5 µg/ml) respectively. For the Msh2-/- cells the auto-fluorescence signal was 51 ± 55 for untreated cells, 56 ± 53 for cells treated with colcemid for 1 h, and 52 ± 58 for cells treated for 4h. The results show that the colcemid treatment did not modulate the over-

A

all levels of Msh2 protein in either the interphase G1+S phase population or the premitotic/mitotic G2+M population cells. Cell Cycle Checkpoint Activation by Colcemid in MEFs. We initially determined gross cell cycle effects, wild-type and Msh2-/- cells were exposed continuously to 0 or 0.5 µg/mL colcemid alone or pretreated for 1 h with 10 µg/mL of cisplatin. Table 1 shows the cell cycle percentages of the MMR deficient and proficient cells after the different treatments. The G2/M percentage increased from 24.2% to 50.8% for Wt, and from 19.6 to 65.8% for Msh2-/-, primarily reflecting differences in the cycling fraction of the early passage cultures. G1 empties from 53.6% to 14.9% for Wt, and 65.4 to 14% for Msh2-/- cultures. The results for cisplatin treatment combined with colcemid do not differ from those for colcemid alone for either cell type, suggesting that cisplatin does not affect cell cycle traverse in MEFs before G2. Low percentages of polyploid cells with >4n DNA content were detected in both cell cultures (3.2% for the wt and 1.6% for Msh2-/-). As expected for the aneuploidogen colcemid, treated cells increased the percentage of polyploids (10% in the wt; 6.5% in Msh2-/-) with similar absolute increases from control values suggesting that Msh2-/- status does not affect induced polyploid generation and therefore by implication the capacity to exit mitosis under spindle stress. Effects of Colcemid on Mitotic Commitment and Duration and Outcome in wt and Msh2-/- MEFs. Using time-lapse microscopy analysis the accumulated mitotic outcomes over 24 h ( (Fig. 4A and B) showed that the main type of event as a result of colcemid treated wt MEFs was poyploid formation (53%), with a similar frequency as that for daughter cell generation in the untreated counterpart (56%). In Msh2-/- cultures colcemid also abrogated the percentage of successful daughter events and with a slightly greater ratio of polyploid events. The increase level of unresolved events (up to 8%) after the colcemid treatment was in part a reflection of the increased duration of mitosis and its nonresolution within the 24 h time course. The cells classified as interphase represent a fraction of cells that did not experience a

B Figure 5. Molecular modelling of DRAQ5DNA interaction reveals preferential intercalation at AT base pairs (A) Intercalation model of DRAQ5 centrally located in a short palindromic DNA duplex of sequence TACGTA (DRAQ5 atom codes: grey, H; green, C; blue, N; red, O). (B) Characterization of DRAQ5 labelling patterns in single nuclei. Distribution of DRAQ5 (20 µM) and an AT-specific reference probe Hoechst 33342 (5 µg/mL) in the nuclei of living Wt MEF cells, together with a corresponding transmission image. Bar 10 µm.


Cell Cycle

667

MSH2 AND COLCEMID

A

B

Figure 6. Localization of Msh2 protein in single nuclei after treatment with colcemid. MEF cultures were prepared for Msh2 immuno-fluorescence detection and DNA structures stained with DRAQ5. Merge of the two fluorescence channels shows sites of co-localization. Ratio images show relative distribution of Msh2 signal per unit of DNA signal. Contrast wedge indicates the quantitative ratio parameters as represented by the corresponding color in the image. (A) Untreated conditions for a typical interphase Wt cell (B) 1 h post-colcemid treatment (0.5 µg/mL) (C) 1 h post-cisplatin treatment (10 µM). (D) Histogram representation of relative changes in the Msh2 signal per unit area of banded or segmented DNA structures (categorized according to fluorescence intensity). Significance is depicted by *p < 0.05, **p < 0.01 (t-test) (see “Materials and Methods” for details).

668

Cell Cycle


MSH2 AND COLCEMID

mitotic commitment but remained intact during the course of the time course and could be considered as arrested cells. The cisplatin treatment increased the level of these interphase cells in both wt and Msh2-/- cells as expected due to the late acting G2 arrest with successful mitotic events being evident in both cultures. The combination of cisplatin with colcemid slightly increased arrest in interphase. Both cisplatin alone and colcemid alone induced low levels of cell death which were additive upon combination of the drugs. We also sought to determine whether the mitotic spindle checkpoint was functionally activated/breached differently in Msh2-/- cells. Prolonged colcemid exposure is known to cause cells to withdraw from trapping at the spindle checkpoint and enter a G1-like state.18 Here we determined whether the duration of the compromised mitosis (Fig. 4C and D) varied in knockout cells or as a function of the period of colcemid exposure. Colcemid increased the duration of mitosis in both cell types by approximately 10-fold. The presence of DNA damage through cisplatin pretreatment did not modify the duration of colcemid-induced trapping (data not shown). Analysis of the mitotic duration versus time of exposure to colcemid reveals no significant correlation. The mean duration of the mitosis did not change significantly between the first 5 h and the 5–24 h window in any of the treatments, showing that late-delivering cells do not undergo longer periods of mitotic trapping. Control and colcemid treatments resulted in equivalent cumulative event kinetics over 24 h in both Wt cells (Fig. 4E) and Msh2-/- cells (Fig. .4F). Nuclear Mapping of DNA with DRAQ5 Identifies AT-rich Regions by Preferential Intercalation. The level of spectral discrimination achieved for the DRAQ5 and Alexa-488 signals results in robust differentiation of signal from each channel enabling quantification of co-localization. The question arises as to the nature of the chromatin regions, identified by DRAQ5, at which Msh2 relocalization occurs. We sought to establish evidence of base-pair selectivity for DRAQ5 to clarify co-localization status. Molecular modelling suggests that DRAQ5 is capable of binding to DNA through intercalation, the side chains on opposing sides of the aromatic ring structure each having the potential to stabilise the association with DNA. Molecular modelling was carried out for DRAQ5 intercalated into short DNA molecules, including a palindromic DNA duplex of sequence TACGTA, with the intercalator binding at the central GC site (Fig. 5A). The model shows DRAQ5 straddling the intercalation site, with one side chain in each of the major and minor grooves. The intercalation is stabilized by electrostatic interaction between the protonated tertiary amino group of the side chain and the phosphate backbone of the DNA. The binding energy was taken as the sum of the energies required for the conformational changes of the DNA and DRAQ5 and the interaction energy. Based on this, the interaction between DRAQ5 with pGC was considerably less favorable than that with pAT. The calculated enthalpic energy difference was in excess of 550 kJ mol-1. This order reflects the higher energy required to unwind pGC. These calculated energies do not take into account enthalpic or kinetic effects, which will be significant in terms of residence times and thus experimentally determined binding constant values. UV/Vis Scatchard analysis confirms the overall result of the modelling, placing the binding constant of pAT/DRAQ5 (K = 3.0 x 107 ± 2.4 x 104) as greater than that of pGC/DRAQ5 (1.4 x 107 ± 1.6 x 106). To relate this potential for preferential intercalation at AT base pairs to our in situ experimental conditions, colabelling experiments were performed using the bisbenzimidazole dye Hoechst 33342 as a reporter for AT-rich heterochromatic regions of chromatin.19 Sequential imaging of dual labelled samples (Fig. 5B) generated staining patterns derived from each probe in the same nucleus. The results showed that the fluorescence signals were co-localized immediately (within 5 min) indicating extensive overlap of binding sites. Intra-nuclear DNA Relocalization of Msh2 Protein After Drug Perturbations. Analysis of the Msh2 protein distribution before and after drug treatment enabled us to dissect the drug effect further by addressing whether the Msh2 protein was preferentially lost from particular DNA structures within the nucleus upon foci redistribution to the cytoplasm. The DNA structures were objectively defined on threshold values so that the entire nucleus was segmented into five distinct regions. This enabled us to extract an index value of Msh2 signal per unit area of DNA for each segmented


Figure 7. Consequences of colcemid drug treatment on mitotic commitment and checkpoint breaching after incurring DNA damage. Wt cultures were treated with colcemid (0.5 µg/ml) (▲) or with both colcemid (0.5 µg/ml) and cisplatin (10 µM) together (o); Control (♦).

region. The trend of preferential localization in control conditions (Fig. 6A) showed that the Msh2 signal formed distinct foci further organized into regions with approximately 1.5 fold more Msh2 signal in bright DNA structures. Compared to less well-labelled nucleus structures upon drug treatment the overall amount of Msh2 signal per DNA signal decreased dramatically (Fig. 6B and D). More specifically the Msh2 signal appeared to be lost at the sites of bright DRAQ5 labelled regions (t-student p < 0.01 band 5 vs. band 2). This subtle redistribution occurred independently of cytoplasmic relocation. Therefore Msh2 in bright DNA foci appeared to be more labile than other localizations. A similar response was shown for cisplatin treated cells (t-student p < 0.05 band 5 vs. band 2). (Fig. 6C and D). To test the functional impact of Msh2 relocation we have determined whether the colcemid effect abrogates the ability to mount the early MMR-dependent G2 delay observed upon exposure to a DNA damaging agent cisplatin. Figure 7E shows the early loss of mitotic delivery in cisplatin treated wt cells compared to the untreated counterpart and an inability of colcemid to block that response.

DISCUSSION Our working hypothesis was that the expression and subcellular distribution of Msh2 protein may show cell cycle-linkage, modulation by DNA damage and sensitivity to disruption by changes in microtubule dynamics. Flow cytometry indicated that the protein was not differentially expressed in the cell cycle, while immunofluorescence data were consistent with the report of Zink et al.20 who found that Msh2 complexes could be discerned as discrete foci for co-localization studies. Image analysis of the disposition of Msh2 foci in metaphase cells reveals that none of the Msh2 protein associates specifically with condensed chromosomes and that the remaining pool is subject to dispersal. Previous immunofluorescence analyses using HeLa MR cells reported by Christmann and Kaina21 showed both a cytoplasmic and nuclear location of Msh2 which became predominantly nuclearlocated upon treatment with the methylating agent MNNG. In contrast we report here that MEFs are normally highly geared for nuclear retention of the protein in the absence of stress signals. The findings suggest that it is unlikely that cytoplasm-to-nuclear transport is rate limiting for MMR processes in MEFs.

Cell Cycle

669

MSH2 AND COLCEMID

The disruption of microtubule dynamics by colcemid resulted in a reduced nuclear signal for Msh2 and increased appearance of Msh2 foci in the cytoplasm of interphase cells with no changes in overall expression per cell and no loss of nuclear membrane. The results show that shifts in microtubule dynamics alter Msh2 protein disposition within MEFs and that a similar process of nuclear-to-cytoplasm relocalization occurs upon exposure to the DNA damaging agent cisplatin. However, using a sensitive co-localization approach we have further identified a rapid and significant process of intra-nuclear relocalization of Msh2 upon stress induction. No differences could be discerned between wild-type and Msh2-/- cells for spindle checkpoint function indicating that dysfunction of MMR is unlikely to modify initial responses to mitosis arresting agents. We conclude that MEFs have a high nuclear retention of Msh2 in interphase and that the disruption of microtubule dynamics or induction of DNA damage can result in relocation of protein from AT rich chromatin sites. Importantly relocation does not appear to affect the mounting of the early Msh2-dependent G2 delay in wild-type cells in response to the DNA damaging agent cisplatin. Immunofluorescence analysis shows that colcemid induced a rapid disruption of microtubule arrays in MEFs and that Msh2 protein is not associated with microtubule structures. Extensive image analysis, addressing preferential Msh2 localization to defined nuclear structures, also showed that Msh2 displayed a punctuated distribution that also had a higher element of aggregation associated with AT base pair rich heterochromatic nuclear regions. Martin et al.22 has reported that Msh2 ATPase activity is essential for somatic hyper-mutation at A-T base pairs, which is consistent with the colocalization of Msh2 proteins at A-T rich foci. This level of discernment of co-localization was achieved through the spectral separation of the immunofluorescence probes used. We show here modelling evidence of the intercalation characteristics and the AT preference of the DRAQ5 probe. Critically since DRAQ5 does not show significant fluorescence enhancement upon binding to DNA12 we have been able to correct for DNA density at a given pixel location addressed for Msh2 content. We suggest that the results obtained using the Hoechst 33342 reference agent for AT base pair recognition19,23 also provides evidence of the DRAQ5 AT-base pair preference in situ. Hoechst 33342 dye is a minor groove binding bisbenzimidazole displaying a high degree of fluorescence enhancement upon binding thereby revealing its preferred binding sites. The extensive emission spectrum of Hoechst 33342 and those of similar UV-excited probes reach across the visible region and would severely limit the dynamic range of the derived colocalization coefficient. The cell permeant properties of the far-red DNA dye12 also opens up the possibility of Msh2 co-localization studies being pursued in live cells using GFP-tagged proteins in permissive cell systems. Comparing the duration of mitotic trapping in wt and Msh2-/cells revealed similar patterns suggesting similar functioning of the spindle checkpoint and the propensity of cells to break through the checkpoint and exit abnormally from mitosis. The increased duration of mitosis in colcemid appears to relate to the acute disruption of spindle function at mitosis since there was no correlation between duration and the period of premitotic exposure to colcemid. There is no a priori reason why colcemid exposure should induce MMR substrates or changed demands for MMR. Previous studies have shown that human and rat fibroblasts display prominent activation of the integrin-ERK pathways after application of colcemid11 and trigger p53-driven cell cycle arrest dependent upon focal adhesion formation. Induction of integrin-ERK pathways in mouse cells is 670

highly transient resulting in an inability to mount a p53-driven supra-threshold response in the presence of colcemid.24,25 Thus microtubule disruption may activate cellular stress responses at subthreshold levels while still modulating protein location. Colcemid does not affect the mounting of the cisplatin-induced early delay response in G2 suggesting that relocation does not compromise the MMRdependent triggering of checkpoint activation and by implication MMR protein remaining located at chromatin sites does not appear to be limiting for mounting a checkpoint response in MEFs. It would be of interest to track the mobility of Msh2 and other MMR proteins in human tumor cells undergoing cell cycle perturbations or unhindered passage through abrogated checkpoints. The present study reveals a physiological pattern for Msh2 expression in the cell cycle and links for the first time microtubuleintegrity with the maintenance of normal intra-nuclear Msh2 protein dynamics. It appears likely that only low levels of DNA-associated Msh2 protein are required for checkpoint function upon damage induction in MEFs and we describe here analytical approaches applicable to live cell studies. The loss of a damage-responsive premitotic MMR checkpoint but the retention of spindle checkpoint function in Msh2-defective cells has implications for exploiting the complex responses of MMR-defective tumor cells to combinations of anticancer agents. References 1. Ren B, Cam H, Takahashi Y, Volkert T, Terragni J, Young RA, et al. E2F integrates cell cycle progression with DNA repair, replication, and G2/M checkpoints. Genes Dev 2002; 16:245-56. 2. Polager S, Kalma Y, Berkovich E, Ginsberg D. E2Fs up-regulate expression of genes involved in DNA replication, DNA repair and mitosis. Oncogene 2002; 21:437-46. 3. Molinari M. Cell cycle checkpoints and their inactivation in human cancer. Cell Prolif 2000; 33:261-74. 4. Yu H. Regulation of APC–Cdc20 by the spindle checkpoint. Curr Opin Cell Biol 2002; 14:706-714. 5. Jullien D, Vagnarelli P, Earnshaw WC, Adachi Y. Kinetochore localization of the DNA damage response component 53BP1 during mitosis. J Cell Sci 2002; 115:71-79. 6. Karran P, Bignami M. Mismatch repair and cancer. In: Smith PJ and Jones CJ, ed. Recombination and Repair. New York: Oxford University Press, 1999:66-98. 7. Kolodner RD. Mismatch repair: Mechanisms and relationship to cancer susceptibility. Trends Biochem Sci 1995; 20:397-401. 8. Buermeyer AB, Deschenes SM, Baker SM, Liskay RM. Mammalian DNA mismatch repair. Ann Rev Genet 1999; 33:533-564. 9. Christmann M, Tomicic MT, Kaina B. Phosphorylation of mismatch repair proteins MSH2 and MSH6 affecting MutSalpha mismatch-binding activity. Nucleic Acids Res 2002; 30:1959-66. 10. Giannakakou P, Nakano M, Nicolaou KC, O'Brate A, Yu J, Blagosklonny MV, et al. Enhanced microtubule-dependent trafficking and p53 nuclear accumulation by suppression of microtubule dynamics. Proc Natl Acad Sci USA 2002; 99:10855-60. 11. Sablina AA, Chumakov PM, Levine AJ, Kopnin BP. p53 activation in response to microtubule disruption is mediated by integrin-Erk signaling. Oncogene 2001; 20:899-909. 12. Smith PJ, Blunt N, Wiltshire M, Hoy T, Teesdale-Spittle P, Craven MR, et al. Characteristics of a novel deep red/infrared fluorescent cell-permeant DNA probe, DRAQ5, in intact human cells analyzed by flow cytometry, confocal and multiphoton microscopy. Cytometry 2000; 40:280-91. 13. Manders EM, Verbeek FJ, Aten JA. Measurement of co-localization of objects in dual-color confocal microscopy. J Microscopy 1993; 169:375-82 14. Wiltshire M, Patterson LH, Smith PJ. A novel deep red/low infrared fluorescent flow cytometric probe, DRAQ5NO, for the discrimination of intact nucleated cells in apoptotic cell populations. Cytometry 2000; 39:217-23. 15. Epstein RJ, Watson JV, Smith PJ. Subpopulation analysis of drug-induced cell-cycle delay in human tumor cells using 90 degrees light scatter. Cytometry 1988; 9:349-58. 16. Watson JV, Chambers SH, Smith PJ. A pragmatic approach to the analysis of DNA histograms with a definable G1 peak. Cytometry 1987; 8:1-8. 17. Marquez N, Chappell SC, Sansom OJ, Clarke AR, Court J, Errington RJ, et al. Single cell tracking reveals that Msh2 is a key component of an early-acting DNA damage-activated G2 checkpoint. Oncogene 2003; 22:7642-7648. 18. Li Q, Dang CV. c-Myc overexpression uncouples DNA replication from mitosis. Mol Cell Biol 1999; 19:5339-5351. 19. Satz AL, White CM, Beerman TA, Bruice TC. Double-stranded DNA binding characteristics and subcellular distribution of a minor groove binding diphenyl ether bisbenzimidazole. Biochemistry 2001; 40:6465-6474.

Cell Cycle


MSH2 AND COLCEMID

20. Zink D, Mayr C, Janz C, Wiesmuller L. Association of p53 and MSH2 with recombinative repair complexes during S phase. Oncogene 2002; 21:4788-800. 21. Christmann M, Kaina B. Nuclear translocation of mismatch repair proteins MSH2 and MSH6 as a response of cells to alkylating agents. J Biol Chem 2000; 275:36256-36262. 22. Martin A, Li Z, Lin DP, Bardwell PD, Iglesias-Ussel MD, Edelmann W, et al. Msh2 ATPase activity is essential for somatic hypermutation at a-T basepairs and for efficient class switch recombination. J Exp Med 2003; 198:1171-1178. 23. Crissman HA, Steinkamp JA. Cell cycle-related changes in chromatin structure detected by flow cytometry using multiple DNA fluorochromes. Eur J Histochem 1993; 37:129-38. 24. Howe DM, Juliano RL. Distinct mechanisms mediate the initial and sustained phases of integrin-mediated activation of the Raf/MEK/mitogen-activated protein kinase cascade. J Biol Chem 1998; 273:27268-27274 25. Schlaepfer DD, Jones KC, Hunter T. Multiple Grb2-mediated integrin-stimulated signaling pathways to ERK2/mitogen-activated protein kinase: Summation of both c-Src- and focal adhesion kinase-initiated tyrosine phosphorylation events. Mol Cell Biol 1998; 18:2571-2581.


Cell Cycle

671