Regulation of DNA Replication after Heat Shock by Replication Protein A-Nucleolin Interactions*

Heat shock inhibits replicative DNA synthesis, but the underlying mechanism remains unknown. We investigated mechanistic aspects of this regulation in melanoma cells using a simian virus 40 (SV40)-based in vitro DNA replication assay. Heat shock (44 °C) caused a monotonic inhibition of cellular DNA replication following exposures for 5–90 min. SV40 DNA replication activity in extracts of similarly heated cells also decreased after 5–30 min of exposure, but returned to near control levels after 60–90 min of exposure. This transient inhibition of SV40 DNA replication was eliminated by recombinant replication protein A (rRPA), suggesting a regulatory process targeting this key DNA replication factor. SV40 DNA replication inhibition was associated with a transient increase in the interaction between nucleolin and RPA that peaked at 20–30 min. Because binding to nucleolin compromises the ability of RPA to support SV40 DNA replication, we suggest that the observed interaction reflects a mechanism whereby DNA replication is regulated after heat shock. The relevance of this interaction to the regulation of cellular DNA replication is indicated by the transient translocation in heated cells of nucleolin from the nucleolus into the nucleoplasm with kinetics very similar to those of SV40 DNA replication inhibition and of RPA-nucleolin interaction. Because the targeting of RPA by nucleolin in heated cells occurs in an environment that preserves the activity of several essential DNA replication factors, active processes may contribute to DNA replication inhibition to a larger degree than presently thought. RPA-nucleolin interactions may reflect an early step in the regulation of DNA replication, as nucleolin relocalized into the nucleolus 1–2 h after heat exposure but cellular DNA replication remained inhibited for up to 8 h. We propose that the nucleolus functions as a heat sensor that uses nucleolin as a signaling molecule to initiate inhibitory responses equivalent to a checkpoint.

DNA replication is at the center of the eukaryotic cell cycle, and its regulation under conditions of stress, or after exposure to DNA-damaging agents, is under the control of intricate pathways that are only now beginning to be characterized. Although the mechanisms regulating DNA replication after DNA damage are studied intensively and a considerable amount of information is now available (reviewed in Ref. 1), the mechanisms that regulate DNA replication in response to other types of stress, such as heat stress, are much less studied and understood. However, heat stress severely compromises cellular function and survival, constitutes a likely event in the life cycle of an organism, and is a form of stress organisms had to deal with during evolution.
It is well documented that exposure of cells to heat strongly inhibits DNA replication, and it is thought that this inhibition underlies the increased heat sensitivity of S-phase cells, as compared with cells in G 1 or G 2 (2)(3)(4)(5). The formation of chromosome aberrations in heated S-phase cells is often considered a consequence of adverse effects on steps of DNA replication (2), as heat is inefficient in inducing DNA damage directly (6 -9). Indeed, heat inhibits several nuclear processes associated with semiconservative replication of DNA including the incorporation of radiolabeled precursors into acid-insoluble material (10,11), the initiation of new replicons (3,4,12,13), the elongation of the DNA fiber at the replication fork (5), the synthesis and deposition of new histones into chromatin, and the reorganization of nascent DNA into mature chromatin (14,15).
Although the above studies identify heat-sensitive steps of DNA replication, they provide only limited information as to the molecular mechanism(s) involved. It is possible that the observed inhibition is a passive consequence of heat-induced inactivation of DNA replication factors, as well as of damage induced in chromatin by the precipitation of denatured protein on the nuclear matrix (14 -18). Alternatively, however, some of the inhibitory effects could be the consequence of a process, equivalent to a checkpoint, which actively stalls DNA replication to prevent grave errors and facilitate repair, much the way ionizing radiation inhibits replicon initiation by activating the ataxia telangiectasia mutated pathway (1, 19 -21).
In an effort to characterize putative regulatory components of DNA replication in cells exposed to heat, we introduced the simian virus 40 (SV40) 1 -based in vitro DNA replication assay. In this assay, the replication of plasmids carrying the SV40 origin of DNA replication (ori) is accomplished in vitro using either crude cytoplasmic extracts from human cells, or proteins purified from such extracts (22)(23)(24)(25)(26). Only one protein of viral origin is required, the SV40 large tumor antigen. The assay provides a well controlled and molecularly defined system to study regulatory aspects of DNA replication, and allows the separation of regulatory effects targeting DNA replication factors from effects of heat on the substrate DNA.
Using this assay we reported (27) that cytoplasmic extracts prepared from heated cells had a reduced DNA replication activity, as compared with extracts prepared from non-heated cells, and that this reduction in DNA replication activity was specifically caused by an inactivation of replication protein A (RPA) (28 -30). Addition of recombinant RPA (rRPA) in these reactions completely reversed the inhibition. Because RPA levels were similar in extracts of heated and non-heated cells and because RPA was relatively resistant to direct heat inactivation, we proposed that the observed response reflected the activation of a checkpoint that targeted and inactivated RPA. A regulatory role for RPA in DNA replication has also been suggested in cells exposed to ionizing radiation (31)(32)(33)(34)(35).
RPA is an essential component of eukaryotic DNA metabolism and has activity as the primary single-stranded DNAbinding protein (for reviews, see Refs. 36 and 37). It is a heterotrimer composed in humans of 70-kDa (hRPA1), 29-kDa (hRPA2), and 14-kDa (hRPA3) subunits. Genetic and biochemical studies suggest a role for RPA in the initiation and elongation stages of DNA replication (36,37). RPA is also required for nucleotide excision repair and homologous DNA recombination (38 -41). RPA2 is phosphorylated in a cell cycle-dependent manner and in response to DNA damage, but the regulatory significance of these phosphorylations has not been established (36,37). RPA also interacts with a variety of proteins involved in DNA metabolism including Rad51 (42), Rad52 (43,44), XPA (45)(46)(47), and p53 (48 -50). Finally, RPA is phosphorylated by the DNA-dependent protein kinase and by ataxia telangiectasia mutated (51)(52)(53)(54)(55). Several of these interactions and modifications are compatible with a regulatory function for RPA in DNA replication, and with a role in the cellular responses to DNA damage, although direct evidence is still lacking.
Because significant phosphorylation of RPA2 was not observed after exposure to heat, it was speculated that if RPA is a target of processes regulating DNA replication, it should be modified by mechanisms other than phosphorylation (27). Recently, evidence has been presented that interaction with nucleolin, a major nucleolar protein involved in ribosome biogenesis and diverse other biological processes (for reviews, see Refs. 56 and 57), can regulate the replication functions of RPA (58).
Here, we extend our studies on the regulation of DNA replication in cells exposed to heat shock and inquire whether the previously reported effects of heat on in vitro DNA replication and the associated inactivation of RPA (27) derive from an interaction with nucleolin. The results demonstrate RPAnucleolin interactions perfectly correlating with in vitro DNA replication activity and suggest that this interaction reflects an early step of the cellular response to heat stress.

MATERIALS AND METHODS
Cell Culture-A human malignant melanoma cell line (SK-MEL-28) was used for all experiments. Cells were obtained from ATCC (HTB-72) and were grown in Dulbecco's modified minimum essential medium supplemented with 25 mM glucose, 10% heat-inactivated calf serum, and antibiotics at 37°C in an atmosphere of 5% CO 2 . For experiments 5 ϫ 10 5 cells/ml were seeded in either 75-cm 2 (15 ml) or 150-cm 2 (30 ml) tissue culture flasks and allowed to grow for 3 days. At this time, the growth medium was changed and cells were incubated for 24 h in fresh growth medium before use in experiments. Cells were exposed to heat by submerging sealed flasks in a water bath set at the desired temperature with an accuracy of 0.05°C. Temperature was monitored with a calibrated mercury thermometer. After treatment cells were trypsinized, collected, counted in a cell counter, and used for extract preparation.
Measurement of DNA Replication in Vivo-To evaluate DNA replication either by incorporation of radioactively labeled precursors into total DNA, or by alkaline sucrose gradient centrifugation (see below), 5 ϫ 10 5 SK-MEL-28 cells from a growing culture were seeded into 75-cm 2 tissue culture flasks (10 ml) and were allowed to grow for 3 days. This protocol yields results equivalent to those obtained using the growth protocol described above, which was selected when large numbers of cells were needed for extract preparation. Cells were exposed to heat for the indicated periods of time and promptly returned to 37°C.
At various times thereafter, depending on the experimental protocol, 3 Ci/ml [ 3 H]thymidine (PerkinElmer Life Sciences) was added and cells were incubated at 37°C for 15 min. Total DNA synthesis was measured as described previously (59). Briefly, cells were trypsinized, loaded on glass microfiber filters (Whatman GF/A), washed with cold trichloroacetic acid, rinsed with deionized water, and counted for 3 H activity. DNA synthesis was calculated as the percentage of incorporation in sham-treated controls.
The size distribution of nascent DNA was measured in heated and non-heated cells in 5-20% linear alkaline sucrose gradients (59). They were prepared in a buffer containing 0.1 M NaOH, 0.9 M NaCl, 0.01 M EDTA, pH 12.5, in 20-ml polyallomer tubes for a total volume of 17 ml. Cells (2.5 ϫ 10 5 in 50 l) were lysed for 3 h by gently layering on 0.3 ml of lytic solution (0.5 M NaOH, 0.02 M EDTA, pH 12.5, and 0.1% Nonidet P-40), in a cut 1-ml syringe. Subsequently, the cell lysate was gently layered on the top of the gradient and centrifuged for 90 min in a Sorval RC60 centrifuge equipped with an AH 627 rotor, at 26,000 rpm, 20°C. After centrifugation, gradients were fractionated (1 ml) and incorporated activity measured as described above. Results are presented as percentage of total [ 3 H]thymidine activity per fraction after correction for the relative inhibition of total DNA synthesis resulting from heat exposure.
Cytoplasmic Extract (S100) Preparation-The method used to prepare cytoplasmic cell extracts for in vitro SV40 DNA replication has been described previously (27,60). Briefly, cells (1-2 ϫ 10 8 ) were washed in PBS, resuspended in hypotonic buffer (10 mM HEPES, pH 7.5, 1.5 mM MgCl 2 , 10 mM KCl, 0.2 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol), and disrupted either by three freeze/thaw cycles, or in a Dounce homogenizer (20 strokes, B pestle). The cell lysate was adjusted to 140 mM KCl by adding 10% volume of 10ϫ cytoplasmic buffer (0.3 M HEPES, pH 7.5, 1.4 M KCl, 0.03 M MgCl 2 ) and was incubated in ice for 15 min. Subsequently, the cell lysate was centrifuged first at 3,300 ϫ g for 15 min and then at 100,000 ϫ g for 1 h (S100), and the cleared supernatant was dialyzed against 100 volumes of dialysis buffer (20 mM HEPES, pH 7.5, 20% glycerol, 50 mM KCl, 0.2 mM EDTA, 0.2 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol) at 4°C. The dialysate was centrifuged at 15,000 ϫ g for 20 min and protein concentration determined by the Bradford assay (Bio-Rad). The extract was either used immediately or snap-frozen and stored at Ϫ80°C for later use. In some experiments the final KCl concentration was adjusted to 200 mM by appropriately increasing salt concentration in the 10ϫ cytoplasmic buffer.
In Vitro DNA Replication Assay-The SV40-based in vitro DNA replication assay was described previously (60). Briefly, reaction mixture (25 l) contained 40 mM HEPES, pH 7.5; 8 mM MgCl 2 ; 0.5 mM dithiothreitol; 3 mM ATP; 200 M each CTP, GTP, and UTP; 100 M each dATP, dGTP, and dTTP; 40 M [␣-32 P]dCTP (0.001 Ci/ml, 3000 cpm/pmol; PerkinElmer Life Sciences); 40 mM creatine phosphate; 1.25 g of creatine phosphokinase; 0.15 g of superhelical plasmid DNA; 100 -200 g of cytoplasmic extract; and 0.5 g of SV40 large T antigen. The reaction mixture was incubated at 37°C for 1 h. Reactions were terminated by addition of 20 mM EDTA. Subsequently, 25 g of denatured salmon sperm carrier DNA and 1 ml of 10% cold trichloroacetic acid were added to each sample. The resulting precipitate was collected onto Whatman GF-C glass fiber filters, and 32 P incorporation into DNA was determined by liquid scintillation counting. To analyze replication products, gel electrophoresis was performed. Samples were treated with 0.1% sodium dodecyl sulfate, followed by 15-min digestion with RNase A (Sigma, 20 g/ml) at 37°C. Proteinase K was then added (Sigma, 200 g/ml), and samples were incubated at 37°C for 30 min. DNA in the sample was purified in a QIAquick column (Qiagen) and electrophoresed in 1.2% agarose gel at 6.5 V/cm.
Detection of RPA-Nucleolin Complex by Immunoprecipitation-Interactions between RPA and nucleolin were investigated by immunoprecipitation carried out using the IMMUNOcatcher kit (CytoSignal, Irvine, CA) according to the manufacturer's instructions. Briefly, SK-MEL-28 cells (2 ϫ 10 6 ) were treated and collected for whole cell extract preparation by incubation (4°C for 1 h) in 200 l of extraction buffer (supplied with the kit). The supernatants obtained by centrifugation were incubated with protein A/G beads pre-absorbed with pre-immune serum for 30 min at room temperature and centrifuged. Clarified supernatants were then incubated with anti-RPA2 antibody (PAb34A, a gift from Dr. J. Hurwitz) for 1 h at room temperature (or overnight at 4°C). Thirty l of protein A/G beads was added to the supernatants and incubated for 1 h at room temperature. Beads were collected, washed three times in extraction buffer, and washed once in PBS, and the immunoprecipitates subjected to Western blotting.
Western Blot Analysis-Proteins were separated using 10% SDS-polyacrylamide gel electrophoresis and transferred to PVDF-Plus membrane (Osmonics Inc.). The membrane was blocked for 1.5 h with 5% powdered milk in PBS, washed twice with PBS, and incubated overnight at 4°C with primary antibody. Subsequently, the membrane was washed three times in PBS containing 0.1% Tween 20, incubated with a horseradish peroxidase-conjugated goat anti-mouse, or goat antirabbit, secondary antibody for 1.5 h and washed five times in PBS containing 0.1% Tween 20. Proteins were detected with ECL plus (Amersham Pharmacia Biotech) and visualized in a Storm 860 Phosphor-Imager (Molecular Dynamics). RPA1 and RPA2 were detected using monoclonal antibodies PAb34A and PAb70A (obtained from Dr. J. Hurwitz). RPA3 was detected using polyclonal antibody Ab11 (obtained from Dr. S.-H. Lee). Nucleolin was detected either with a polyclonal antibody 134 (obtained from Dr. P. Bouvet) or with a commercially available monoclonal antibody C23 (Santa Cruz Biotechnology Inc., Santa Cruz, CA). Immunofluorescence Microscopy-SK-MEL-28 cells grown on coverslips were washed once with cold PBS and cold cytoskeleton buffer (10 mM Hepes, pH 7.4, 300 mM sucrose, 100 mM NaCl, and 3 mM MgCl 2 ), as described previously (61). After extraction for 2 min on ice with 0.5% Triton X-100 in cytoskeleton buffer containing proteinase inhibitors (61), cells were fixed for 20 min at room temperature with 4% formaldehyde in PBS, rinsed with PBS, and incubated with 0.5% Nonidet P-40 in PBS for 5 min. Primary antibody (anti-nucleolin, C23) was then added and incubated for 1 h at room temperature. After rinsing three times with 0.5% Tween 20 in PBS, secondary antibody (donkey antimouse IgG conjugated with Texas Red; Molecular Probes, Eugene, OR) was applied and incubated for 1 h at room temperature. Coverslips were then rinsed three times with 0.5% Tween 20 in PBS and mounted onto glass slides. Conventional fluorescence or confocal microscopy was used to detect antigens.

RESULTS
Transient Inhibition of DNA Replication after Exposure of Cells to Heat-As a first step in our investigations, we characterized the effect of heat on DNA replication in SK-MEL-28 cells. Cultures were exposed to 44°C for various periods of time and the effect on DNA replication determined. Fig. 1A shows the results obtained as a function of time. In line with earlier observations (see Introduction), exposure to heat strongly inhibited ongoing DNA replication. The inhibition was biphasic, with a steep component that dominated in the first 20 min and reduced DNA synthesis to 20% of control, followed by a shallow component evident between 20 and 90 min that further reduced, to nearly zero, DNA synthesis. The effect of heat shock on DNA replication was transient and recovered to control levels 8 h after a 15-min exposure to 44°C (Fig. 1B).
The effect of heat shock on the replicon initiation and chain elongation stages of DNA replication was evaluated with alkaline sucrose density-gradient centrifugation. Fig. 1C shows the results obtained. Exposure to heat inhibited replicon initiation manifested by a reduction in the activity of early fractions, which represent the top of the gradient and therefore low molecular weight material. However, heat also inhibited, to a similar degree, chain elongation manifested by a reduction in material incorporated in the remaining portion of the gradient. Although only results obtained after exposure of cells to 44°C are presented here, the results in Fig. 2A indicate that an effect on DNA replication was clearly observed at temperatures above 40°C.
Inhibition of SV40 DNA Replication in Extracts of Heated Cells-Several processes may contribute to the above inhibition of DNA replication in cells exposed to heat shock. First, the coordinated function of the components of DNA replication machinery may be impaired at temperatures above 40°C, without associated irreversible changes in either the chromatin substrate, or the DNA replication factors. Second, one or more key DNA replication factors may be heat-sensitive and may be rapidly inactivated in cells exposed to heat. Third, heat-induced alterations in chromatin structure may modify the substrate DNA and may inhibit its efficient replication. Fourth, cells may sense heat damage in chromatin and other cellular structures and activate signaling pathways, equivalent to a checkpoint, that actively inhibit DNA replication.
To discriminate between these possibilities, we tested DNA replication activity in extracts of heated cells using the SV40 in vitro DNA replication assay. By measuring DNA replication with untreated plasmid DNA and tumor antigen, the assay removes the effects of heat on DNA replication substrate and allows the study of effects on key DNA replication factors. These effects can include direct heat inactivation, as well as the targeted inactivation by regulatory processes. Fig. 2A shows SV40 DNA replication activity in extracts (S100 fraction) prepared from SK-MEL-28 cells exposed for 30 min to temperatures between 38°C and 44°C, normalized to SK-MEL-28 cells were exposed to 44°C for 15 min and returned to 37°C. At the indicated times, a 15-min pulse with [ 3 H]thymidine was given and incorporation into acid-insoluble material measured. The mean and standard error from two experiments are shown. Panel C, alkaline sucrose density gradient profiles of SK-MEL-28 cells exposed to 44°C for 0 min (circles), 5 min (triangles), 10 min (inverted triangles), 20 min (squares), and 30 min (diamonds). Cells were exposed to a 15-min pulse with [ 3 H]thymidine immediately after heat shock. Plotted is the percentage of total activity in each fraction as a function of the fraction number. For samples exposed to heat, the incorporation values for each fraction have been multiplied by the observed inhibition of DNA replication. Sedimentation is from left to right. Early fractions contain small nascent DNA produced in the early stages of DNA replication. Shown for calibration purposes are also the locations of sedimentation maxima for T7 (37 kilobase pairs) and T2 (167 kilobase pairs) phage DNA. the activity of extracts from non-treated cells. Heat shock suppressed in vitro DNA replication activity in a temperature-dependent manner only above 40°C. The levels of DNA replication inhibition in vitro paralleled those in vivo, suggesting that the assay reproduces essential components of the cellular processes underlying DNA replication inhibition.
The effect of different exposures to 44°C on SV40 DNA replication is shown in Fig. 2B. A rapid reduction in activity was observed in cells exposed to 44°C for 5-30 min. Surprisingly, however, the replication activity increased and reached near control levels after exposures for 60 and 90 min. A similar response was also observed when the products of DNA replication were analyzed by agarose gel electrophoresis (Fig. 2C). There was no difference in the distribution of DNA replication products in the different samples, suggesting that mainly initiation events were reduced, or that all steps of DNA replication were equally affected. A specific effect on chain elongation would have increased the abundance of slowly migrating replication intermediates and can therefore be excluded. Results obtained after addition of rRPA to the reactions are discussed below.
The high levels of DNA replication activity in extracts of cells exposed to 44°C for 60 -90 min suggests that the transient inhibition between 5 and 30 min reflects a regulatory process targeting an essential replication factor(s) rather than the irreversible denaturation of heat-sensitive DNA replication factors; the latter would have persisted as exposure times were increased beyond 30 min.
Inactivation of Replication Factors and Inhibition of DNA Replication after Heat Shock in Vitro-We inquired whether the high levels of DNA replication activity after heat shock for 60 -90 min reflected an inherent heat resistance of the DNA replication factors. Extracts of non-heated cells were exposed to heat and tested for in vitro SV40 DNA replication activity. Fig.  3A shows that, contrary to the exquisite resistance observed when heating was administered in intact cells, heating in vitro for only 10 min at 41°C led to a Ͼ80% loss of DNA replication activity. The inhibition was not directly reversible, as subsequent incubation at 37°C for 30 min did not recover DNA replication activity (Fig. 3A). Additionally, DNA replication per se was sensitive to heat as incubation of reactions assembled with non-heated cell extract to temperatures as low as 39°C reduced DNA replication activity by over 50%. Incubation at 41°C completely halted DNA replication (Fig. 3B).
RPA as a Target of a Pathway That Regulates DNA Replication in Cells Exposed to Heat-Recently, we provided evidence that RPA is the target of a pathway that regulates DNA replication in HeLa cells after heat shock (27). We inquired whether a similar mechanism was active in extracts of heated SK-MEL-28 cells. The results in Fig. 2B indicate that 1 g of rRPA restored DNA replication to near control levels and abolished heat-induced fluctuations in SV40 DNA replication activity. The effect of RPA in rescuing DNA replication was specific as bovine serum albumin was ineffective in modifying DNA replication activity (data not shown). RPA inactivation can be ruled out as the cause of the reduced DNA replication activity at 5-30-min exposures as RPA retained significant activity in extracts of cells exposed to 44°C for 60 -90 min. These results point to a regulatory pathway involving RPA, which is heat-sensitive.
RPA-Nucleolin Interactions after Heat Shock-The above results point to RPA as the target of a pathway that regulates DNA replication in heated cells and raise the question as to the mechanism involved. We explored the possibility that RPA activity is regulated by an interaction with nucleolin as described recently (58). Whole cell, rather than S100, extracts were assayed to examine interactions in a large component of the cellular protein pool. Fig. 4A shows the results obtained with extracts of heated cells following immunoprecipitation with an anti-RPA2 antibody. A substantial increase was observed in the amount of interacting nucleolin after heat shock despite the decrease in the amount of immunoprecipitated RPA (Fig. 4A). An effect was detectable even after 5 min, but a dramatic increase was observed in cells exposed for 20 -30 min. After longer exposure times, the interaction became attenuated and returned to near control levels after 90 min. Thus, the interaction between RPA and nucleolin mirrors perfectly SV40 DNA replication activity in extracts from heated cells (Fig. 2B). There were no significant changes in nucleolin, or in the three RPA subunits, in extracts of heated cells (Fig. 4B), so the observed interaction was not resulting from altered availability of the interacting proteins. Heating extracts in vitro could not reproduce the heat-induced interaction between RPA and nucleolin (Fig. 4C), suggesting the operation of processes that are perturbed or cannot be activated in the extract.
Contrary to results obtained with HeLa cells (27), an evaluation of RPA levels in S100 extracts of SK-MEL-28 cells showed reductions after heat shock that could explain the observed inhibition of SV40 DNA replication, without the need to invoke an interaction with nucleolin (Fig. 5A). The absence of this reduction in HeLa cells and its transient occurrence in SK-MEL cells raised the possibility that it is caused by the interaction with nucleolin, i.e. a lower extractability of the RPA-nucleolin complex than of RPA alone from the nucleus during S100 preparation. To address this possibility, S100 extracts were prepared from heated and non-heated SK-MEL-28 cells using either the standard concentration of 140 mM KCl for extraction during preparation (see "Materials and Methods"), or 200 mM KCl to enhance protein extraction from the nucleus. We also increased from 10 to 30 min the incubation time at 4°C for extraction after homogenization. Increasing extraction time reduced the difference in RPA levels between heated and control cells without compromising the inhibition of SV40 DNA replication (Fig. 5, compare A with B). Increased extraction by high salt further limited the observed reduction in RPA levels in S100 extracts of heated cells. At the same time, however, high salt eliminated the inhibition of DNA replication in S100 extracts and the RPA-nucleolin interaction in whole cell extracts (Fig. 5C). These results confirm that the RPA-nucleolin interaction after heat shock is rather common and occurs in different cell lines such as HeLa (58) or SK-MEL-28 cells, while the trapping of the complex in the nucleus is cell line-dependent and more pronounced in SK-MEL-28 cells.
Heat-induced Translocation of Nucleolin into the Nucleoplasm-The near exclusive localization of nucleolin into the nucleolus and of RPA in the nucleoplasm in non-stressed cells is expected to prevent an interaction between the two proteins and raises the question as to how this interaction is facilitated after heat shock. Recent studies demonstrated a translocation of nucleolin from the nucleolus into the nucleoplasm after exposures to 44°C longer than 60 min (58). We investigated whether a translocation of nucleolin can be observed in heated SK-MEL-28 cells in a time frame correlating with the observed inhibition of in vitro DNA replication. Cells were exposed to 44°C for periods ranging between 0 and 90 min and nucleolin localization determined. Fig. 6 shows representative images of the results obtained. In control cells (0-min exposure), nucleolin was localized, as expected, practically exclusively in the nucleolus. The nucleolar staining was specific for nucleolin, as pre-immune serum did not produce any staining (data not shown). There was evidence for translocation of nucleolin into the nucleoplasm even in cells exposed to 44°C for 5 min, indicated by a diffuse staining in the nuclear region. However, nucleoli remained clearly visible in these samples. The pattern of staining changed significantly after 10 min at 44°C and remained qualitatively similar for exposure times up to 30 min. The images obtained after these exposure times showed a diffuse nuclear staining with little or no definition of the nucleolus. Notably, after exposure for 60 or 90 min at 44°C, a preferential localization of nucleolin in the nucleolus was again evident. These results exactly parallel the in vitro inhibition of DNA replication (Fig. 2B) as well as the interaction between nucleolin and RPA (Fig. 4A) in cells exposed to heat stress.
The observation that a translocation of nucleolin from the nucleolus into the nucleoplasm could be observed immediately after exposure to heat indicated that the underlying mechanism is fast. As a next step we investigated, therefore, the persistence of the response in an effort to correlate it with the kinetics of DNA replication recovery observed in intact cells (Fig. 1B). Cells were exposed to 44°C for 15 min and returned to 37°C. At various times thereafter, cells were fixed and stained for nucleolin localization. The results obtained are shown in Fig. 7. Nucleolin translocation is at maximum immediately after heat and returns to the nucleolus ϳ1 h after heating, long before cellular DNA replication has recovered completely (Fig. 1B). Panel C, whole cell extracts prepared from non-heated SK-MEL-28 cells were exposed in vitro to 44°C for the indicated periods of time, immunoprecipitated, and analyzed as outlined in panel A.

RPA-Nucleolin Interactions Inhibit SV40 DNA Replication in
Extracts of Heated Cells-The results presented above confirm and extend previous observations on the effect of heat on DNA replication activity, as measured by the SV40 in vitro DNA replication assay (27). We show that even modest exposures of SK-MEL-28 cells to heat (5-10 min) cause measurable reduction in DNA replication activity (Fig. 2B). Evidently, inhibitory effects on DNA replication can precede the development of lethal events, as short exposures to heat only have a small effect on colony-forming ability (10 -20% reduction). As in the case of HeLa cells (27), inhibition of DNA replication in SK-MEL-28 cells was the result of a specific inactivation of RPA (Fig. 2B), rather than the result of nonspecific and widespread heat-induced inactivation of one or more DNA replication factors.
Inactivation of RPA in heated cells was not caused by obvious post-translational modifications, as the electrophoretic mobility in SDS-polyacrylamide gels remained unchanged for all subunits (Fig. 4B). Furthermore, addition of rRPA effectively rescued DNA replication in extracts of heated cells, a result that is compatible with the absence of RPA-modifying activi-ties. If RPA-modifying activities with inhibitory effects were present in the extract, they would have also modified added rRPA and would have reduced the efficiency of the rescuing process.
Inhibition of SV40 DNA replication in extracts of heated cells correlates well with an RPA-nucleolin interaction (Fig. 4A). Nucleolin is involved in ribosome biogenesis and is known to be required for the first step of pre-rRNA processing in which the 5Ј external transcribed spacer undergoes endonucleolytic cleavage (57,62,63). However, nucleolin is a multifunctional protein, has been found in a replication complex (64), and has been recently characterized as an RPA-interacting protein (58). The latter report showed that nucleolin inhibits SV40 DNA replication and that this inhibition can be rescued by the addition of stoichiometric amounts of rRPA. Evidence was also presented that an RPA-nucleolin interaction inhibits the initiation steps of DNA replication, while leaving unaffected the singlestranded DNA binding activity of the protein. Although not examined directly, it is likely that an RPA-nucleolin interaction also affects RPA-requiring events in the elongation stages of DNA replication. These observations in aggregate provide a mechanistic explanation for the reported effects of heat on FIG. 5. Heat shock reduced the extractability of RPA into S100 fraction in SK-MEL-28 cells. Panel A, SK-MEL-28 cells were exposed to 44°C for the indicated periods of time and S100 cell extracts prepared. Thirty g of these extracts were examined by Western blotting for the levels of RPA subunits. Panel B, effect of 200 mM KCl extraction during S100 preparation on SV40 in vitro DNA replication in extracts of heated cells. SK-MEL-28 cells were exposed to 44°C for 30 min and S100 extracts prepared using either the standard extraction protocol (140 mM KCl), or 200 mM KCl. Replication activity was compared with that of extracts of non-heated cells. Panel C, whole cell extracts prepared in the presence of KCl at concentrations similar to those used in panel B were immunoprecipitated and analyzed by Western blotting as described in Fig. 4A.  FIG. 6. Heat shock induced a relocalization of nucleolin from the nucleolus into the nucleoplasm. SK-MEL-28 cells grown on coverslips were exposed to 44°C for the indicated periods of time and were subsequently fixed and stained with an antibody against nucleolin and a Texas Red-conjugated secondary antibody as described under "Materials and Methods." Cells were analyzed using a confocal microscope. Shown are representative images of cells exposed for different times to 44°C .  FIG. 7. Heat-induced translocation of nucleolin from the nucleolus into the nucleoplasm is transient. SK-MEL-28 cells grown on coverslips were exposed to 44°C for 15 min and returned to 37°C. At various times thereafter, cells were stained and analyzed as outlined in the legend of Fig. 6. Shown are representative images of cells allowed to recover from heat shock for the indicated times. SV40 DNA replication (27) and point to nucleolin as a potential regulator.
Effects of Heat on Nucleolin Localization in Vivo-The possible relevance of the above biochemical observations to the regulation of DNA replication in heated cells is indicated by the observed intracellular relocalization of nucleolin (Fig. 6). Although nucleolin exclusively localizes in the nucleolus of unstressed cells (56,57), exposure to heat (58) or other types of stresses (65) leads to its relocalization into the nucleoplasm. This relocalization potentially enables an interaction with RPA, which according to the results presented here, and those previously published (58), is expected to inhibit DNA replication. In line with this expectation, nucleoplasmic regions with increased accumulation of nucleolin show reduced incorporation of DNA synthesis precursors (58). A potentially relevant observation that hints to intriguing regulatory mechanisms is that RPA3, the RPA subunit that directly interacts with nucleolin (58), localizes in the nucleolus of non-heated cells without being associated with RPA1 and RPA2 (66).
Although earlier observations suggested that nucleolin translocation and RPA-nucleolin interactions require exposures to 44°C longer than 60 min (58), our results demonstrate effects at the earliest heat exposure times measured (5 min). This places the sensitivity of nucleolin translocation and of RPA-nucleolin interaction in the same exposure range as the inhibition of DNA replication (Fig. 1A) and is in line with a regulatory role for nucleolin in DNA replication under conditions of stress. Further evidence for a regulatory role of nucleolin in DNA replication is provided by the similarities between the kinetics of recovery of in vitro DNA replication (27) and the kinetics of intracellular nucleolin relocalization after exposure to heat (Fig. 7); both process are nearly completed within 1-2 h.
On the basis of these observations, a model can be developed for the regulation of DNA replication in heated cells whereby translocation of nucleolin from the nucleolus into the nucleoplasm inhibits DNA replication by a direct interaction with RPA. Although the ultimate target of this nucleolin-mediated inhibitory process, which may be equated to a checkpoint response, is RPA, the signal that initiates the pathway and the up-stream events that lead to nucleolin relocalization remain unknown. We previously speculated (27) that heat-induced chromatin damage (67) might be generating the initiating event. However, other mechanisms of activation cannot be ruled out at present.
One can only speculate on the upstream events that lead to nucleolin translocation. Release of nucleolin from the nucleolus can be mediated by a loss of its pre-rRNA binding substrate, which occurs in heated cells as heat exposure decreases rRNA synthesis (68,69). However, because nucleolin is a phosphoprotein that can be modified by several kinases including Cdc2, protein kinase C , and casein kinase II (56,57), and can also be methylated and ADP-ribosylated, it is possible that its translocation into the nucleoplasm is the result of post-translational modifications. Indeed, phosphorylation of nucleolin regulates some of its functions during cell growth and in the cell cycle, and may also modulate its activities in response to heat stress. Heat shock may alter the phosphorylation state of nucleolin by activating an unidentified kinase, or a phosphatase, and may mediate in this way its translocation into the nucleoplasm. The signal leading to this modification must be transient as nucleolin returns to the nucleolus 1-2 h after heat exposure (Fig. 7).
It is particularly interesting that the nucleolin-dependent pathway of DNA replication regulation is itself heat-sensitive and can be inactivated after prolonged (60 -90 min) exposures to heat (Fig. 2B). This was demonstrated by the reduced interaction of nucleolin with RPA (Fig. 4A) and by the reduced intracellular relocalization of nucleolin (Fig. 6) in cells exposed to heat for 60 min or longer. Thus, an excessive amount of stress can cause a collapse of the protective responses mounted by the cell, suggesting that experiments to characterize them should be carried out under carefully selected conditions.
The results in Fig. 4C indicate that RPA may not constitutively interact with nucleolin and that this interaction may not be induced by heating extracts in vitro. It is possible that for a productive interaction between nucleolin and RPA a particular set of posttranslational modifications is required and/or the presence of nuclear organizing structures such as the nuclear matrix, to facilitate these interactions. It may be relevant in this regard that nucleolin is found to bind matrix attachment region elements on chromosomal DNA (70) and that the DNA replication machinery is thought to be attached to the nuclear matrix. An auxiliary role of DNA in RPA-nucleolin interactions is indicated by the observation that RPA-nucleolin interactions can be disrupted at relatively low salt concentrations (200 mM KCl; Fig. 5C) in cellular extracts, whereas they require nearly 500 mM salt in an RPA column made with single-stranded DNA-cellulose (58). Elucidation of the specific requirements for RPA-nucleolin interactions will enhance our understanding of the underlying regulatory mechanism(s).
Other Mechanisms in the Regulation of DNA Replication in Heated Cells-Although the translocation of nucleolin into the nucleoplasm explains the initial inhibition of DNA replication, a comparison between the kinetics of nucleolin relocalization into the nucleolus (Fig. 7) and of the recovery of DNA synthesis in vivo (Fig. 1B) suggest that other processes must be involved in maintaining the initial inhibition. Nucleolin, by virtue of its fast translocation, may be a first line of defense activated immediately after exposure of cells to heat and causing an immediate reduction in DNA replication activity. This fast response may prevent or reduce DNA replication on a heatmodified nuclear matrix (67) that may cause complications undermining genomic stability, or even cell survival. However, this translocation, or the resulting inhibition of DNA replication, may serve as a signal for the activation of longer-lived regulatory processes that maintain the inhibition of DNA replication for up to 8 h (Fig. 1B). According to this model, the nucleolus may serve as a sensor of heat damage (71) and nucleolin as the signaling molecule.
The release of nucleolin into the nucleoplasm helps to explain earlier observations that the dissolution of the nucleolus in response to mitosis-promoting factor during premature chromosome condensation was deficient in heated cells (72). It has been reported that nucleolin is involved in chromatin decondensation by virtue of its ability to interact with histone H1 (73) and might also be involved in the dissolution of the nucleolus during mitosis. Its release from the nucleolus after heat may thus affect this process.
The Heat Resistance of DNA Replication Factors-A surprising but potentially very significant observation of the work presented here is that supralethal exposures to heat do not measurably alter the DNA replication activity of several factors required for in vitro SV40 DNA replication. Obviously, this statement holds only for factors required for SV40 DNA replication and excludes key components of the eukaryotic DNA replication machinery such as the components of the origin recognition complex and the mini chromosome maintenance proteins that are not required for in vitro SV40 DNA replication. The presence of active SV40 DNA replication factors at nearly normal concentrations and activities in extracts of cells exposed to 44°C for 60 or 90 min suggests that these factors are neither irreversibly inactivated nor somehow precipitated on cellular structures (67). As these factors completely lose their activities for SV40 DNA replication when heated in vitro, even at a temperature as low as 41°C (Fig. 3A), it becomes evident that the cellular environment must provide a high level of protection against heat-induced denaturation and inactivation. It will be particularly important to characterize the determinants of this heat resistance.
Finally, the possibility that the interaction between nucleolin and RPA increases the amount of protein on the nuclear matrix, where DNA replication occurs and therefore RPA is expected to accumulate, is intriguing. It suggests that, in addition to the passive process of protein precipitation on the nuclear matrix as a result of heat-induced denaturation (67), active processes involved in either the signaling of a stress response or the repair of the induced damage can contribute to this effect.
The above observations in aggregate suggest the operation of an active mechanism in the regulation of DNA replication in cells exposed to heat shock. The mechanism operates by the inactivation of a key DNA replication factor, RPA, mediated via direct interaction and association with nucleolin. Although nucleolin is normally found exclusively in the nucleolus in non-stressed cells, it translocates into the nucleoplasm after heat exposure, thus enabling the interaction with RPA. RPA targeting by nucleolin takes place in an environment preserving the activity of several essential DNA replication factors. It is likely, therefore, that the contribution of regulatory processes, such as the one relying on RPA-nucleolin interactions, to DNA replication inhibition after heat shock is more significant than presently thought. Characterization of the upstream components in this pathway will not only enhance our understanding of DNA replication and its regulation under conditions of stress but will also help to identify targets for therapeutic intervention, or parameters predicting the response of tumor cells to heat treatment. Hyperthermia is used as an adjuvant to radiation therapy (74), and this information should be helpful in the development of improved therapeutic strategies.