|Home | About | Journals | Submit | Contact Us | Français|
PCNA (proliferating cell nuclear antigen) is required for a wide range of cellular functions, including DNA replication and damage repair. To be functional, PCNA must associate with the replication and repair foci. In addition, PCNA also mediates targeting of certain replication and repair proteins to these foci. However, the mechanism is not yet known by which PCNA is imported into the nucleus, and then localized to the replication and repair foci.
We have found that an NLS (nuclear localization sequence) is present within the amino acid 101–120 segment of PCNA. An NLS-deleted PCNA was localized in the cytoplasm and showed 5-fold lower affinity for importin-β than wild-type, suggesting that PCNA may be imported into the nucleus by importin-β via its NLS. We previously reported that the functional unit of PCNA is a double trimer (as opposed to single homotrimer), and Lys-110 is essential for the formation of the double trimer complex [Naryzhny, Zhao and Lee (2005) J. Biol. Chem. 280, 13888–13894]. The present study shows that the substitution of Lys-110 within the NLS to an alanine residue did not affect its nuclear localization. However, the double-trimer-defective PCNA(K110A) was not localized at replication or repair foci. In contrast, the double-trimer-intact PCNA(K117A) mutant was targeted normally to replication and repair foci. Interestingly, in cells transfected with PCNA(K110A), but not PCNA(K117A), caspase-3-mediated chromosome fragmentation was activated.
The present study suggests that the regulation of PCNA is intimately connected with that of DNA replication, repair and cell death signals, and raises the possibility that defects in the formation of the PCNA double-trimer complex can cause apoptosis.
PCNA (proliferating cell nuclear antigen) is involved in a wide spectrum of cellular functions, including DNA replication, repair and epigenetic maintenance/inheritance (Kelman, 1997; Maga and Hubscher, 2003; Lee, 2006). The diverse functions of PCNA are at least, in part, regulated by its interactions with many different protein partners (Tsurimoto, 2006; Lee and Naryzhny, 2006), which may be mediated by the PCNA double-trimer toroidal complex (Naryzhny et al., 2005, 2006; Lee and Naryzhny, 2006). Although PCNA is mainly a nuclear protein, it does not contain a classical importin-α–importin-β heterodimer-mediated NLS (nuclear localization sequence), suggesting that it is imported into the nucleus by other mechanism(s) (Gorlich and Kutay, 1999; Mosammaparast and Pemberton, 2004; Naryzhny and Lee, 2004). Several recent studies showed that certain proteins are imported into the nucleus by importin-β alone (Moore et al., 1999; Xiao et al., 2000; Yamasaki et al., 2005; Kim and Lee, 2006). In line with these reports, we previously found that PCNA binds to importin-β, but not importin-α (Kim and Lee, 2006).
Nuclear PCNA is initially found at the nuclear matrix in mid-G1 phase, and then associated with the chromatin fraction when DNA replication starts (Naryzhny and Lee, 2004). As expected from its essential role in DNA replication and repair, PCNA is found in replication and repair foci (Leonhardt et al., 2000; Somanathan et al., 2001; Naryzhny et al., 2005). However, the mechanism as to how PCNA is targeted to these foci is not yet known. There are a few lines of evidence that ‘targeting sequences’ are required for directing proteins to replication foci. For example, DNA methyltransferase (amino acids 207–455) and DNA ligase 1 (amino acids 1–20) contain such targeting sequences (Leonhardt et al., 1992; Liu et al., 1998; Maga and Hubscher, 2003; Mortusewicz et al., 2005). Interestingly, the target sequence of DNA ligase 1 interacts with PCNA, suggesting it is the PCNA that mediates the targeting of DNA ligase 1 to replication foci (Cardoso et al., 1997; Montecucco et al., 1998). This raises the possibility that PCNA may be responsible for recruiting other PCNA-associated proteins to replication and repair foci. Consistent with this expectation, it has been found that the Williams syndrome transcription factor is also targeted to replication foci by PCNA (Poot et al., 2004). Thus PCNA is directly involved in targeting certain replication/repair proteins to replication/repair foci. Therefore, understanding of the PCNA targeting mechanism will provide important insights into the regulation of DNA replication and repair processes.
To gain a better understanding about the targeting of PCNA to replication and repair foci, we generated a series of deletion mutants and then examined their subcellular localization in transfected cells. We found that a PCNA NLS is present within the amino acid segment 101–120. Further analysis by point mutations within the NLS identified Lys-110 as a critical PCNA targeting sequence to replication and repair foci. The replacement of Lys-110 with an alanine residue (K110A) resulted in the accumulation of this mutant protein in the heterochromatin fraction. Most interestingly, the double-trimer defective PCNA(K110A) mutant induced apoptosis and caspase-3-mediated chromosome fragmentation in transfected cells.
To identify the NLS of PCNA protein, we examined the nuclear localization of WT (wild-type) and several mutant PCNA proteins transfected into CHO (Chinese-hamster ovary) cells. As expected, GFP (green fluorescent protein)–PCNA WT was mostly localized in the nucleus (Figures 1A and 1B). The GFP–PCNA deletion mutants, Δ1–20, Δ1–100, Δ121–172, Δ172–184, Δ184–196 and Δ197–261 also showed a similar nuclear distribution pattern. In contrast, the GFP–PCNA(Δ101–120) deletion mutant was exclusively localized in the cytoplasm, suggesting that the segment 101–120 contains an NLS (Figure 1B). Since the seven deletion mutants collectively cover the entire PCNA protein, the 101–120 segment contains the only NLS of PCNA. We found that the point mutants E104A, K110A and K117A within the NLS did not significantly alter the PCNA nuclear localization pattern (Figure 1A). Data obtained from in vitro cross-linking assays showed that, unlike WT, PCNA(Δ1–100) and PCNA(Δ101–120) mutants did not form a trimer complex (Figure 1C), suggesting that a homo-trimer complex is not essential for PCNA nuclear localization.
Sequence alignment showed that the segment 101–120 is well-conserved among PCNA proteins from human, mouse, rice, Drosophila and even yeast (Figure 1D). However, this conserved sequence is very different from the classical NLS recognized by importin-α (Mosammaparast and Pemberton, 2004), suggesting that PCNA may not be transported by importin-α. Consistent with this hypothesis, we previously found that an intact PCNA protein interacted with importin-β, but not with importin-α (Kim and Lee, 2006). Therefore, we examined whether the amino acid segment 101–120 is recognized and bound by importin-β. Data from an in vitro pull-down assay showed that PCNA WT, and K110A and K117A mutants have strong affinity for importin-β(Figure 2A). However, the PCNA(Δ101–120) deletion mutant showed at least a 5-fold lower affinity for importin-β, suggesting that PCNA is bound (and probably transported into the nucleus) by importin-β.
To further confirm the role of the segment 101–120 in nuclear import, we generated a plasmid construct containing a tri-GFP plus amino acids 101–120 of PCNA [tri-GFP–PCNA(101–120); ~100 kDa; tri-GFP is a tri-repeat GFP]. We then examined the subcellular localizations of tri-GFP (control) and tri-GFP–PCNA(101–120) in the CHO cells transfected with these constructs. As expected, tri-GFPs were more frequently found in the cytoplasm, whereas tri-GFP–PCNA(101–120) proteins were often found in the nucleus (Figure 2B). However, in many transfected cells, both the protein species were localized in the nucleus, as well as in the cytoplasm. Nevertheless, statistical analysis of 100–200 cells showed that the percentage of the pattern (‘N > C’) was much higher in the cells expressing tri-GFP–PCNA (101–120) than tri-GFP (Figure 2C, 24% compared with 7%). Considering the fact that a large portion of tri-GFP–PCNA(101–120) degradation products (asterisk in Figure 2D) would be localized in the cytoplasm (due to their GFP portion), the actual percentage of tri-GFP–PCNA(101–120) in the nucleus would be much higher than 24%. We, therefore, conclude that the amino acid segment 101–120 plays an essential role for PCNA nuclear localization.
We have found recently that the human Cdc7 (cell division cycle 7) nuclear retention sequence is localized within its NLS, and nuclear retention is mediated by its association with chromatin (Kim et al., 2007). Therefore, in an attempt to find PCNA amino acid residues that are important for chromatin binding, we examined subcellular localization of PCNA containing a number of different point mutations within the NLS. As shown in Figure 3(A), the majority of GFP–PCNA(WT) was co-localized with replication foci in early S-phase (panels a–c). The percentage of co-localization was approx. 77%, as calculated from the overlapped foci of the red [i.e. BrdU (bromo-deoxyuridine) labelled] and green spots [i.e. GFP–PCNA(WT)] in the enlarged images (Figure 3B). These data are in line with the previous observations that GFP–PCNA is localized at replication foci during S-phase (Somanathan et al., 2001; Naryzhny et al., 2005). Consistent with the data in Figure 1(B), the GFP–PCNA(Δ101–120) mutant was localized in the cytoplasm (Figure 3A, panels d–f). Although less efficient than WT, both the GFP–PCNA-(K110A) and GFP–PCNA(K117A) mutants were also localized in the nucleus (Figure 3A, panels g and j). The localization of the nuclear PCNA(K117A) mutant clearly showed a punctate pattern (Figure 3A, panel j), which largely coincided with DNA replication foci (Figure 3A, panels k and l). In contrast, nuclear PCNA(K110A) mutant proteins, which were diffusedly aggregated in large patches, were not co-localized with replication foci (Figure 3A, panels g–i).
Next, we examined if the PCNA(K110A) mutant proteins co-localize with endogenous PCNA. The data obtained from co-immunoprecipitation analysis showed that GFP–PCNA(WT) and the PCNA(K117A) mutant co-precipitated with endogenous PCNA, whereas co-precipitation of the PCNA(K110A) mutant with endogenous PCNA was decreased (Figure 3C). As expected, the PCNA(Δ101–120) mutant did not co-precipitate with endogenous PCNA (Figure 3C). These results suggest that PCNA(K110A) mutant may not interfere with endogenous PCNA. Taken together, our data suggest that Lys-110 is essential for targeting PCNA to replication foci.
Since PCNA(K110A) proteins were not associated with replicating DNA (Figure 3A, panels g–i), we examined their localization in more detail by a fractionation-based assay. Fractionation was carried out with CHO cells transfected with GFP–PCNA(WT), GFP–PCNA(K110A) and GFP–PCNA(K117A) constructs (Figure 4A). It was shown previously that limited digestion by micrococcal nuclease results in the release of oligonucleosomes (the S2 fraction in Figure 4A) which include transcriptionally active euchromatin (Remboutsika et al., 1999). This fraction contains only low levels of histone H1 (Remboutsika et al., 1999). The S3 fraction, which contains high levels of histone H1, mainly includes transcriptionally inactive heterochromatin (Remboutsika et al., 1999). Consistent with this previous finding, the S1 and S2 fractions contained only very low levels of histone H1, whereas the S3 fraction contained a very high level of histone H1 (Figure 4B, histone H1). In addition, the S3 fraction also contained a high level of HP1α (heterochromatin protein 1). Together, these data confirm that the S3 fraction is enriched with heterochromatin (Figure 4B).
As expected, endogenous PCNA proteins were mostly found in the euchromatin (S2) fraction (‘Endogenous PCNA’ in Figure 4B; compare the asterisk with the arrowhead). The same trend was also observed with GFP–PCNA(WT), although significant amounts of the exogenous WT PCNA proteins were also found in S1 and S3 fractions. In contrast with WT, GFP–PCNA(K110A) proteins were mainly detected in the heterochromatin (S3) fraction (‘K110A’ in Figure 4B; compare the asterisk with the arrowhead), suggesting that the mutation at Lys-110 is responsible for the accumulation of this mutant PCNA in heterochromatin region. Overall, our data are consistent with a model that PCNA(K110A) proteins are loaded on to the chromatin without any problems; however, it has lost specificity of targeting PCNA to replication sites (Figures 3 and and44).
Results shown in Figures 3 and and44 raised the possibility that Lys-110 might also be required for targeting PCNA to repair sites, since PCNA is essential for DNA repair synthesis. HeLa cells transfected with GFP–PCNA(WT), GFP–PCNA(Δ101–120), GFP–PCNA(K110A) and GFP–PCNA(K117A) constructs were treated with a low dose of cisplatin (10 μM), a chemotherapeutic agent that induces DNA damage by platinum–DNA adducts and thus phosphorylates histone H2AX (γ-H2AX) (Siddik, 2003; Huang et al., 2004; Solomon et al., 2004). At 12 h post-cisplatin treatment, cells were immunostained with an anti-(phospho-Ser-139 histone H2AX) antibody. As expected, cells transfected with WT PCNA did not show any repair foci in the absence of cisplatin (Figure 5B). However, HeLa cells transfected with WT PCNA showed a significant number of repair foci in the presence of cisplatin, which were largely co-localized with WT PCNA (Figures 5D–5F). As expected, PCNA(Δ101–120) was localized mainly in the cytoplasm (Figures 5G–5I). Similarly to the WT, the majority of PCNA(K117A) was also co-localized with repair foci (Figures 5M–5O). In contrast, the PCNA(K110A) mutant did not show a punctate pattern (Figure 5J), and was mostly not co-localized with repair foci (Figures 5J–5L). We, therefore, conclude that the Lys-110 residue is essential for targeting PCNA to repair foci.
We noted that cells transfected with the GFP–PCNA(K110A) construct underwent apoptosis when the transfected cells were maintained for a prolonged period. Therefore, we studied this phenomenon more systematically. As shown in Figure 6(A), cells transfected with GFP–PCNA(K110A), but not GFP–PCNA(WT), underwent apoptosis by 24 h post-transfection, even in the absence of cisplatin (compare panels i–iii with iv and vi). Examination of cell populations showed that approx. 15% of the cells transfected with the GFP–PCNA(K110A) construct underwent apoptosis by 24 h post-transfection, even in the absence of cisplatin, as observed by chromosome condensation and fragmentation by microscopy (White, 1996) (Figure 6B, K110A in the left-hand panel). In contrast, cells transfected with GFP, GFP–PCNA(WT) and GFP–PCNA(K117A) constructs did not show any significant cell death in the absence of cisplatin (Figure 6B, left-hand panel). The number of apoptotic cells by PCNA(K110A) in the absence of cisplatin increased to 30% by 52 h post-transfection (Figure 6C). This increase in cell death was not observed in the cells transfected with GFP alone or GFP–PCNA(WT) during the same period (Figure 6C).
As expected, cisplatin caused apoptosis in all of the cells transfected with the GFP, GFP–PCNA(WT), GFP–PCNA(K117A) and GFP–PCNA(K110A) constructs (Figure 6B, right-hand panel). It was noted that cisplatin treatment showed only additive effects of cell death in the cells transfected with the GFP–PCNA(K110A) construct (Figure 6B, K110A in the right-hand panel). The activation of caspase-3 was examined in the cells transfected with the PCNA(K110A) mutant. As shown in Figure 6(B) (lower panels), caspase-3 was activated (as shown by the generation of the p17 caspase-3 cleavage product) in the cells transfected with PCNA(K110A) in the presence or absence of cisplatin. Caspase-3 was not significantly activated in the cells transfected with GFP vector alone, PCNA(WT) or PCNA(K117A) in the absence of cisplatin (Figure 6B, left-hand lower panel). Thus PCNA(K110A) causes apoptosis in the transfected cells at 24 h or later time points. This was further confirmed by a caspase-3 activity assay in vitro. Caspase-3 activities were measured using extracts prepared from cells transfected with GFP–PCNA(WT) or GFP–PCNA(K110A). As shown in Figure 6(D), the relative caspase-3 activity of the cells transfected with the PCNA(K110A) construct was equivalent with the level of the purified caspase-3 positive control, whereas cells transfected with the PCNA(WT) construct did not show significant caspase-3 activities.
To begin addressing the mechanisms of how PCNA is transported into the nucleus, and then to the replication and repair foci, we mapped its NLS. The existence of the NLS was shown by PCNA cytoplasmic localization when the segment 101–120 was deleted (Figures 1, ,33 and and5;5; summarized in Table 1). Although it is generally thought that a small size of proteins (less than 50 kDa) can passively diffuse into the nucleus, certain small proteins are known to be imported by importin carriers. This group includes IGFBP-3 (insulin-like growth-factor-binding protein-3), Snail, PDX-1 (pancreatic and duodenal homeobox-1), CREB (cAMP-response-element-binding protein) and his-tone proteins (Schedlich et al., 2000; Johnson-Saliba et al., 2000; Forwood et al., 2001; Guillemain et al., 2004; Yamasaki et al., 2005). Our data presented here also support that the nuclear localization of PCNA is mediated by importin-β, despite its monomer being only 30 kDa (Figure 2A). The existence of an active regulation mechanism for PCNA transportation is also supported by the fact that its nuclear localization is largely cell cycle dependent (Naryzhny and Lee, 2004).
There may be two possible mechanisms with respect to the role of Lys-110 in PCNA targeting. One possibility is that Lys-110 may be involved in allowing the loading of PCNA directly on to replication or repair foci. An alternative mechanism may be that Lys-110 is involved in directing PCNA trimers that are already loaded on to chromatin to the replication or repair foci. We prefer the former model, since the PCNA trimer is known to be loaded at the 3′ end of the DNA primer at the junction of the already-initiated and to-be-replicated DNA (Fukuda et al., 1995).
We and others previously have shown that both PCNA(K117A) and PCNA(K110A) do not affect the formation of the PCNA homotrimer complex (Fukuda et al., 1995; Naryzhny et al., 2005). Furthermore, a PCNA(K110A) homotrimer complex could bind to either DNA polymerase δ or CAF-1 (chromatin assembly factor 1). However, a PCNA(K110A) complex could not bind DNA polymerase δ and CAF-1 simultaneously, because two PCNA(K110A) homotrimers could not form a double-trimer complex (Naryzhny et al., 2005). We report in the present study that both PCNA(K117A) and PCNA(K110A) mutants are efficiently imported into the nucleus (Figures 1, ,33 and and5;5; summarized in Table 1). The PCNA(K117A) mutant is then co-localized with replication and repair foci, but the PCNA(K110A) mutant is not (Figures 2 and and4).4). What makes it profoundly different in targeting replication and repair foci between these two point mutants, even though both the Lys-110 and Lys-117 residues are closely located within the NLS region? One clear difference is that Lys-117 is on the front and Lys-110 is on the back side of the PCNA trimer (Krishna et al., 1994; Gulbis et al., 1996). Furthermore, the PCNA(K117A) mutant can form a double-trimer structure, but the PCNA(K110A) mutant cannot (Naryzhny et al., 2005). Therefore, the disruption of the PCNA double trimer complex is clearly correlated with the loss of PCNA targeting to replication and repair foci.
The specificity of PCNA (complex) localization could be achieved by binding of PCNA to proteins that are already localized at specific DNA sites or structures [e.g. RFC (replication factor C) and the 3′ end of a DNA primer], for which Lys-110 may be essential. This model may not be correct if a PCNA single homotrimer is the functional unit, since no proteins binding to the back side (where Lys-110 is located) have so far been found. However, the model can explain our data if the PCNA functional unit is a double-trimer complex, as we reported previously on the basis of in vitro cross-linking studies (Naryzhny et al., 2005, 2006; Lee and Naryzhny, 2006). Together with these previous data, our present data in vivo are consistent with a model that the PCNA double-trimer structure is required for targeting PCNA to replication and repair foci.
One important discovery of the present study is that a large number of cells expressing PCNA-(K110A) mutant underwent caspase-3-mediated chromosome fragmentation after 24 h post-transfection (Figure 6). It is intriguing that PCNA(K110A) causes apoptosis, but PCNA(K117) does not. The only notable difference between these two mutants is the PCNA(K117A) can form a double trimer and PCNA(K110A) cannot (although it forms a single homotrimer complex). This suggests that the PCNA double-trimer complex is essential for cell survival, and disruption of this structure can activate cell death signal.
CHO cells were grown in minimal essential medium supplemented with 10% fetal clone II serum (HyClone, Logan, UT, U.S.A.). HeLa and HEK 293T cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal bovine serum (Hyclone). Transfection of cells was carried out using Lipofectamine Plus™ reagent according to the manufacturer’s instructions (Invitrogen).
The plasmid pEGFP-PCNA has been described previously (Leonhardt et al., 2000; Naryzhny et al., 2005). Various PCNA deletion and point mutants were generated by cloning mutant DNA fragments into the pEGFP-C1 vector, as described previously (Naryzhny et al., 2005). The bacterial expression plasmid encoding GST (glutathione transferase)–importin-β has been described previously (Moore et al., 1999; Kim and Lee, 2006).
To generate PCNA mutant for the nuclear localization studies, the following oligonucleotides were used to generate double-stranded DNAs: pEGFP-PCNA(101–120), 5′-TCGACTAGTATTTGAAGCACCAAACCAGGAGAAAGTTTCAGACTATGAAATGAAGTTGATGGAT-3′ and 5′-GATCATCCATCAACTTCATTTCATAGTCTGAAACTTTCTCCTGGTT-TGGTGCTTCAAATACTAG-3′ (encompassing amino acids 101–120, LVFEAPNQEKVSDYEMKLMD). To generate double-stranded DNAs, these complementary oligonucleotides were incubated in a hybridization buffer [20 mM Tris/HCl, 100 mM NaCl and 0.1 mM EDTA (pH 7.4)]. Subsequently, the double strands were ligated into the XhoI–BamHI site of pEGFP-C1. To generate the di-GFP construct, pEGFP-PCNA(101–120) was digested with AgeI and then ligated with 790 bp fragment generated by digesting pEGFP-C1 with AgeI and XmaI, as described previously (Kim et al., 2007). To generate tri-GFP, the di-GFP–PCNA(101–120) plasmid was digested with AgeI and then ligated with the 790 bp fragment generated by digesting pEGFP-C1 with AgeI and XmaI.
Antibodies against GFP (B2), PCNA (PC10), histone H1 and all secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, U.S.A.). Other reagents were as follows: anti-(phospho-Ser-139 histone H2AX) (JBW) antibody (Upstate Biotechnology), anti-caspase-3 (3G2) and anti-HP1α (#2616) antibodies (Cell Signalling), human recombinant caspase-3 proteins (#235417; Calbiochem) and cisplatin (Faulding Canada Inc., Montreal, QC, Canada).
The preparation of nuclear fractionation and chromatin-binding assays were as described previously (Liang and Stillman, 1997; Remboutsika et al., 1999; Macfarlan et al., 2005). Briefly, nuclei were lysed in lysis buffer [15 mM Tris/HCl (pH 7.4), 60 mM KCl, 15 mM MgCl2, 15 mM NaCl, 1mM CaCl2, 1 mM PMSF, 0.6% Nonidet P40 and 1×protease inhibitor cocktail (Roche)]. After centrifugation at 950 g in a bench-top centrifuge (Eppendorf) for 3 min, the pellet was incubated with 20 units of micrococcal nuclease (USB) at 37°C for 10 min. The sample was then centrifuged at 10 000 g (Eppendorf centrifuge 5415) for 10 min. The supernatant was collected, and the pellet was resuspended in ice-cold 2 mM EDTA (pH 8.0) solution. After centrifugation at 10 000 g in a bench-top centrifuge for 10 min, the supernatant containing solubilized chromatin was collected.
Recombinant GST–importin-β proteins used for the investigation shown in Figure 1(C) were expressed in Escherichia coli strain BL21 by induction with 1 mM IPTG (isopropyl β-D-thiogalactoside) (Roche) for 4–5 h at 28°C and purified by affinity chromatography with glutathione–Sepharose 4B (Amersham Biosciences), according to manufacturer’s instruction. For in vitro binding assay, cells transfected with GFP–PCNA constructs were lysed at 14 h post-transfection with lysis buffer [150 mM NaCl, 1% Nonidet P40, 50 mM Tris/HCl (pH 7.5), 50 mM NaF, 50 mM glycerophosphate, 2 mM EDTA, 10% glycerol and 1×protease inhibitor cocktail]. Cell lysates were then incubated for 4 h at 4°C with glutathione–Sepharose 4B beads pre-treated with GST or GST–importin-β. The beads were extensively washed with PBST (PBS plus 0.02% Triton X-100), and proteins bound to beads were analysed by SDS/PAGE Western blotting, as described previously (Kim and Lee, 2006).
Synchronization was performed as described previously (Guo et al., 2005). Briefly, CHO cells grown on a cover glass were maintained for 45 h in minimal essential medium without isoleucine, supplemented with 10% dialysed fetal bovine serum (Invitrogen). To obtain cells in early S-phase, cells that were synchronized in G0 were released into complete medium for 14–16 h (in the presence of BrdU). BrdU labelling was carried out according to manufacturer’s instructions (Roche).
Cells grown on a cover glass were fixed and permeabilized with 100% methanol for 15 min at −20°C. After treated with PBS containing 1% BSA for 2 h at room temperature (23°C), cells were incubated with primary antibodies for 90 min at room temperature. The cells were washed three times with PBS (5 min each wash), and were then incubated with rhodamine-conjugated secondary antibodies for 45 min at room temperature. Subsequently, cells were washed with PBS, mounted on to a slide glass, and then visualized by fluorescence microscopy (Axiovert 100, Carl Zeiss).
Cells were transfected with expression plasmids for 12 h, washed three times with PBS, resuspended in cold lysis buffer [50 mM Tris/HCl (pH 7.2), 250 mM NaCl, 0.1% Nonidet P40, 2 mM EDTA, 10% glycerol and 1×protease inhibitor cocktail] and incubated at 4°C for 30 min. The supernatant was separated by centrifugation at 15 000 g in a bench-top centrifuge for 10 min at 4°C, and was then incubated with agarose-conjugated anti-GFP antibody at 4°C overnight. The beads were washed four times with cold lysis buffer and resuspended in the SDS sample buffer, followed by SDS/PAGE and Western-blot analysis.
Apoptotic cell death was examined by fluorescence microscopy as described previously (Kim et al., 2002). Briefly, cells grown on a cover glass were transfected with plasmids using Lipofectamine Plus™ reagent. At various time points post-incubation, cells were fixed and permeabilized with 100% methanol for 10 min at −20°C. After two washes with PBS, cells were treated with 0.1 μg/ml Hoechst 33258 (Sigma), and then mounted on to a slide glass.
Caspase-3 activity was determined using the fluorogenic caspase-3 substrate (Z-DEVD)2-Rh110 [benzyloxycarbonyl- Asp-Glu-Val-Asp)2–rhodamine 110] (Calbiochem) with whole-cell extracts prepared from cells transfected with GFP–PCNA or PCNA mutants. In particular, the cleavage of DEVD-Rh110 by cell extracts was measured at 37°C using a Multimode Detector DTX880 (Beckman Coulter, CA, U.S.A.) according to the following specifications: excitation filter, 485 nm; emission filter, 535 nm; detection method, fluorescence bottom; integration time, 0.4 s.
This work was supported by grants from the Canadian Cancer Society [National Cancer Institute of Canada (NCIC) grant #016072] and the Canadian Institutes of Health Research (grant MOP79473) to H.L. B.J.K. was supported in part by an Ontario Graduate Scholarship and the University of Ottawa Excellence Award. This manuscript is a part of the PhD thesis of B.J.K.