HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 24, 22035-22044, June 14, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/24/22035    most recent M200914200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by von Kobbe, C. Articles by Bohr, V. A. Search for Related Content PubMed PubMed Citation Articles by von Kobbe, C. Articles by Bohr, V. A. Social Bookmarking                      What's this?

Colocalization, Physical, and Functional Interaction between Werner and Bloom Syndrome Proteins^*

Cayetano von Kobbe, Parimal Karmakar, Lale Dawut, Patricia Opresko, Xianmin Zeng, Robert M. Brosh Jr., Ian D. Hickson^§, and Vilhelm A. Bohr^¶

From the  Laboratory of Molecular Gerontology, NIA, National Institutes of Health, Baltimore, Maryland 21224 and the ^§ Imperial Cancer Research Fund Laboratories, Institute of Molecular Medicine, University of Oxford, Oxford, OX3 9DS, United Kingdom

Received for publication, January 18, 2002, and in revised form, March 7, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The RecQ helicase family comprises a conserved group of proteins implicated in several aspects of DNA metabolism. Three of the family members are defective in heritable diseases characterized by abnormal growth, premature aging, and predisposition to malignancies. These include the WRN and BLM gene products that are defective in Werner and Bloom syndromes, disorders which share many phenotypic and cellular characteristics including spontaneous genomic instability. Here, we report a physical and functional interaction between BLM and WRN. These proteins were coimmunoprecipitated from a nuclear matrix-solubilized fraction, and the purified recombinant proteins were shown to interact directly. Moreover, BLM and WRN colocalized to nuclear foci in three human cell lines. Two regions of WRN that mediate interaction with BLM were identified, and one of these was localized to the exonuclease domain of WRN. Functionally, BLM inhibited the exonuclease activity of WRN. This is the first demonstration of a physical and functional interaction between RecQ helicases. Our observation that RecQ family members interact provides new insights into the complex phenotypic manifestations resulting from the loss of these proteins.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Werner syndrome (WS)¹ is a hallmark premature aging syndrome associated with increased malignancies and genomic instability (1). The protein defective in WS, Werner syndrome protein (WRN), is an ATP-dependent 3'-5'-helicase and also has a 3'-5'-exonuclease activity (2). Recently, a number of protein partners for WRN have been identified. These include proliferating cell nuclear antigen (3), replication protein A (RPA) (4), DNA topoisomerase I (3), the Ku heterodimer (5, 6), DNA polymerase  (7), and p53 (8). Some of these interactions are not only physical but are also functional. Each of these binding proteins is involved in some form of DNA metabolism, such as DNA recombination, replication, and repair, and in the resolution of alternative DNA structures (1, 2). This suggests that WRN plays a role in a number of key DNA metabolic pathways. Bloom syndrome (BS) is a highly cancer-prone disease associated with increased genomic instability and elevated sister chromatid exchanges. The protein defective in this disorder, Bloom syndrome protein (BLM), is a 3'-5'-helicase (9). BLM binds to RPA, p53, topoisomerase III, and RAD51 and is a component of the BRCA1-associated genome surveillance complex (10). This links BLM with proteins implicated in several aspects of DNA metabolism (DNA recombination, replication, and repair).

BLM and WRN both belong to the RecQ family of helicases, which are conserved from Escherichia coli to human (11). WRN is unique among the human RecQ helicases in having an exonuclease domain in the N-terminal region of the protein. A number of the RecQ helicases have been purified, and their biochemical properties have been characterized (11). There is considerable interest in the substrate specificities of each of these enzymes and in possible differences between them. However, so far there has been no indication of any interaction between different RecQ helicases and no focus on the potential modulation of their catalytic activities when present together in a complex.

Based on their shared helicase sequences, WRN and BLM are expected to exhibit similar biochemical properties (12). Our recent studies demonstrate that BLM and WRN are structure-specific helicases that demonstrate a marked similarity in DNA substrate preference (13). Also, WS and BS patients share a number of phenotypic characteristics (14), indicating a possible "cross-talk" in the intracellular pathways in which BLM and WRN function. In this context, as mentioned above, both helicases have common interacting partners (p53, RPA).

In the present work, we demonstrate that BLM and WRN colocalize in three different human cell lines and interact physically and functionally. These results suggest that these and possibly other RecQ helicases exist in complexes and that their functions should be considered in this context.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Recombinant Proteins-- Recombinant, hexahistidine-tagged BLM and WRN proteins were purified as described previously (9, 15).

Solubilization of Nuclear Matrix Proteins-- The lysis of HeLa cells (~2 × 10⁹) was performed as described previously (16). The solubilization of the nuclear insoluble pellet was as described previously (17, 18) but with some modifications. The insoluble pellet (containing chromatin, nuclear envelope, and nuclear matrix proteins) resulting from the second sucrose cushion extraction was resuspended in 4 ml (1 volume) of Tween 20 buffer (1% Tween 20, 50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 15 mM MgCl₂, 60 mM -glycerophosphate, 100 µM Na₃VO₃, 2 mM dithiothreitol, 450 mM NaCl, 2 µg/ml aprotinin, 2 µg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride). The suspension was sonicated twice (5 s each), dounced four times in a glass Dounce homogenizer (tight pestle, B type), and incubated on ice for 20 min. The suspension was then centrifuged at 100,000 × g (Ti-60 rotor, Beckman) for 30 min at 4 °C. The supernatant was diluted with 8 ml (2 volumes) of Tween 20 buffer without NaCl (150 mM NaCl final concentration). Prior to binding assays, the supernatant was centrifuged at 13,000 × g for 3 min at 4 °C. The resulting supernatant was defined as the nuclear matrix (NM) solubilized fraction.

Western Blot of the Different HeLa Subcellular Fractions-- Twelve µg of HeLa S3 nuclear extracts, nuclear matrix-solubilized fraction, and cytoplasmic extracts (S100) were loaded onto an SDS-polyacrylamide gel and transferred to polyvinylidene difluoride membranes (Bio-Rad) at 250 mA for 2 h at room temperature. Before blocking, the membranes were stained with Amido Black to verify the amount of proteins loaded into each well. Before incubation with the primary antibodies, the membranes were incubated in blocking buffer (5% non-fat milk, 0.5% Tween 20 in 1× PBS) for 1.5 h at room temperature. The membranes were then incubated with rabbit anti-BLM (1:1000, Novus), rabbit anti-WRN (1:1000, Novus), anti--Tubulin (1:200, Santa Cruz Biotechnology, S100 marker), or goat anti-lamin B (1:1000, Santa Cruz Biotechnology, NM marker) polyclonal antibodies for 2 h at room temperature. After washing two times (10 min each) with blocking buffer, the membranes were incubated with the corresponding secondary antibodies horseradish peroxidase-labeled (1:10,000, Vector Laboratories) for 1 h at room temperature. After three washes (15 min each), the horseradish peroxidase signal was detected using ECL Plus (Amersham Biosciences) following the manufacturer's instructions.

Coimmunoprecipitation Assay-- The nuclear matrix-solubilized fraction (500 µl) was precleared with protein A-Sepharose beads (30 µl) (Amersham Biosciences) for 1 h at 4 °C and then incubated with either rabbit anti-BLM antibodies (19) or rabbit anti-GST antibodies (negative control) (Amersham Biosciences) (1:100 each) for 16 h at 4 °C. After the incubation, 30 µl of protein A-Sepharose were added to each sample, and the sample was incubated for 2 h at 4 °C. Then the beads were pelleted (2000 × g, 2 min at 4 °C) and washed four times with 500 µl of Tween 20 buffer (see above). Bound proteins were eluted by boiling in protein sample buffer and were analyzed by SDS-PAGE and Western blot as described above. In this experiment, we used as primary antibodies either mouse anti-WRN (1:250, Becton Dickinson) or mouse anti-BLM (1:50 (20)).

Transfection and Immunofluorescence Assays-- HeLa, U-2 OS, and SV40-transformed WS (AG11395) cells were grown in Dulbecco's modified Eagle's medium (Invitrogen) and 10% fetal bovine serum. For expression as GFP fusion protein in mammalian cells, the PCR fragment containing the full-length sequence of WRN (aa 1-1432) was subcloned into the XhoI-XmaI sites of vector pEGFPC3 (CLONTECH). Cells growing on coverslips were transfected with a pEGFPC3-WRN (1-1432) expression vector using the Calphos mammalian transfection kit (CLONTECH). After 15 h, the cells were fixed and permeabilized with 4% paraformaldehyde (Sigma), 0.2% Triton X-100 (Sigma) in 1× PBS for 15 min at room temperature. Then the coverslips were washed three times with PBS followed by incubation with blocking buffer (0.1% Tween 20, 2% BSA in 1× PBS) for 1 h at room temperature. Finally, the coverslips were incubated with rabbit anti-BLM antibodies (1:200 dilution in blocking buffer) for 4 h at room temperature. After washing two times (10 min each, constant shaking) with blocking buffer, the coverslips were incubated with donkey anti-rabbit Texas Red (The Jackson Laboratory) (1:200 in blocking buffer) for 1 h at room temperature. After washing three times (15 min each, constant shaking), the coverslips were mounted on Vectashield (Vector Laboratories) and viewed under a laser scan confocal microscope (Zeiss 410) in separate channels (green, 488 nm; red, 568 nm). The images were then overlaid and analyzed with Metamorph imaging system 4.1 (Universal Imaging Corp.). Representative photographs from three independent experiments are shown in the figures.

Characterization of WRN-BLM Interaction by ELISA-- ELISA was done as described previously (19) but with some modifications. The blocking and binding steps were performed with the same buffer: 3% BSA, 0.1% Tween 20 in PBS. In the coating step, WRN was diluted in carbonate buffer at 2 ng/ml. Then BLM was added at 1 ng/ml in the binding step. DNase I (Calbiochem) was included (5 µg/ml) in the incubation time in the corresponding wells. The absorbance (490 nm) values were corrected for background signal (absorbance signal of anti-BLM antibodies with WRN-coated wells, optical density, 0.1). For heat-denatured BLM, the diluted protein was incubated at 100 °C for 5 min prior to adding to the corresponding wells.

Cloning and Expression of WRN Fragments as GST Fusion Proteins-- The WRN fragments were generated by PCR using the gene encoding human WRN protein kindly provided by Dr. Junko Oshima (University of Washington Medical School, Seattle, WA) as template. The PCR products were subcloned into the NcoI-BamHI and NcoI-XmaI (for fragments encoding aa 1-51 and aa 1-120) sites of vector pGEXCS. The WRN expression plasmids were sequenced to verify their integrity. The expression of GST-WRN recombinant fragments was as described previously (21).

GST Pull-down Assays-- The pull-down assays with the various GST-WRN fragments and [³⁵S]BLM were performed as described previously (21) with the exception that the buffer used for binding and washing steps was PBS (1×), 0.1% Tween 20, and 3% BSA. For the pull-down experiment using HeLa nuclear extract, 40 µl of glutathione-Sepharose beads were saturated with the different GST-WRN fragments. The recombinant fragments were then incubated with 500 µg of HeLa nuclear extract for 2 h at 4 °C. After extensive washes, the bound proteins were analyzed by Western blot with anti-BLM antibodies.

Cloning of Human BLM-- Human BLM cDNA was obtained by PCR using a cDNA library as the template and using Expand high fidelity enzyme (Roche Molecular Biochemicals). The PCR fragments were cloned into the XhoI-KpnI sites of vector pRSETA (Invitrogen) to generate the plasmid pRSETA-BLM for expression as radiolabeled, in vitro translated BLM ([³⁵S]BLM).

Helicase Assay-- The 100-bp M13mp18 partial duplex DNA substrate was constructed with a 100-mer oligonucleotide (Loftsrand Laboratories Ltd.) complementary to positions 6038-6137 of M13mp18 (New England Biolabs). The 100-mer was ³²P-labeled at its 5' end and annealed to the M13mp18 single-stranded DNA circle and subsequently purified as described previously (4). The forked DNA duplex was created by annealing the D_control and T_stem oligonucleotides (Midland Certified Reagents) as described previously (22). Helicase assay reactions were performed and analyzed as described previously (4) with the following exceptions. Reactions (20 µl) contained ~0.3 nM 100-bp M13mp18 partial duplex DNA substrate or 0.5 nM forked DNA substrate and the indicated amounts of WRN and BLM in 30 mM HEPES (pH 7.6), 5% glycerol, 40 mM KCl, 0.1 mg/ml BSA, 8 mM MgCl₂, 2 mM ATP, and 0.5 mg/ml yeast tRNA. WRN and BLM were preincubated at 24 °C for 4 min. The reactions were initiated by the addition of the DNA substrate and incubated for 30 min at 24 °C. Helicase reactions with the M13mp18 partial duplex substrates were terminated as described previously (4). Helicase reactions with the forked duplex substrates were terminated with the same buffer and 3.3 nM unlabeled oligonucleotide. Reactions were electrophoresed on 12% polyacrylamide 1× Tris borate/EDTA gels and analyzed as described (23).

Exonuclease Assay-- Single-stranded DNA oligomers (72- and 53-mer) were obtained from Invitrogen. The 53-mer (7 pmol) was 5'-labeled with [-³²P]ATP (60 µCi, 3000 Ci/mmol) and polynucleotide kinase (10 units) using standard conditions. Construction and purification of double-stranded DNA substrates containing a 5' overhang and the conditions for the exonuclease reactions were as described previously (5).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Colocalization of WRN and BLM in HeLa and ALT Cell Lines-- Colocalization of WRN and BLM has been reported in human primary fibroblasts (24), but no data have appeared using other human cell lines. We used an N-terminal GFP-tagged WRN construct to specifically label the WRN protein. After transfection with the GFP-WRN expression vector, HeLa cells were fixed, permeabilized, and probed with specific anti-BLM antibodies. As shown in Fig. 1A, GFP-WRN and endogenous BLM colocalized in nuclear foci. In the transfected cells that show GFP-WRN nuclear foci staining (~20%), we found almost 100% colocalization with BLM. A low percentage of transfected cells showed no colocalization (Fig. 1A, panel c, white arrows) of WRN and BLM proteins.

View larger version (127K): [in this window] [in a new window]   Fig. 1.   Colocalization of GFP-WRN and BLM in human cell lines. Exponentially growing HeLa cells (A), SV40-transformed WS cells (AG11395 cell line)(B), and U-2 OS cells (C) were transfected with the pEGFPC3-WRN (aa 1-1432) vector and processed as described under "Materials and Methods." Panels d, h, and l in A and panels d and h in B and C represent the transmitted image. These pictures correspond to two representative GFP-WRN transfected cells from each cell line. As shown in A-C, the calibration bar (right, bottom panel) corresponds to 10 µm. White triangles represent examples of BLM/GFP-WRN foci that are not colocalizing.

We also examined two additional human cell lines (U-2 OS and AG11395) for colocalization of WRN and BLM. As seen in Fig. 1, B and C, both cell lines showed strong colocalization of GFP-WRN and endogenous BLM in nuclear foci. However, some BLM nuclear foci and GFP-WRN nuclear foci did not colocalize (Fig. 1, panels c and g, white arrows), suggesting that the WRN-BLM association is dynamic. In the course of these experiments, we detected both a strong nuclear punctuate staining and either weaker or no nucleolar staining of GFP-WRN and BLM in U-2 OS and AG11395 cells (compare the number of nuclear foci corresponding to both GFP-WRN and the red BLM staining in Fig. 1A with Fig. 1, B and C). These cell lines are telomerase-negative cells (ALT cells), whereas HeLa cells are telomerase-positive (25, 26). The colocalization of GFP-WRN and BLM in the ALT cell lines was not affected by whether endogenous WRN was present (U-2 OS cells) or absent (AG11395, SV40-transformed WS cells). Thus, these results indicated that GFP-WRN and BLM colocalized in nuclear foci. Furthermore, this colocalization was almost 100% when both proteins were in these foci. The colocalization between GFP-WRN and BLM was more evident in ALT cells where the two proteins particularly localized to the nuclear foci.

WRN and BLM Interact Directly-- To explore whether a physical (direct) interaction exists between WRN and BLM, we performed ELISA with purified recombinant proteins. The results demonstrated that BLM specifically binds with WRN (Fig. 2). This interaction was not mediated by DNA since the binding was largely unaffected by the presence of DNase I (Fig. 2, bar 3). The WRN-BLM binding was highly dependent on the BLM conformation because when BLM was heat-denatured, the interaction was greatly reduced (bar 4). This reduced absorbance signal was not due to the lack of the recognition of heat-inactivated BLM by anti-BLM antibodies (data not shown). These experiments clearly demonstrated that WRN and BLM interact directly.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Direct interaction between WRN and BLM. Either BSA (bar 1) or purified recombinant WRN (bars 2-4) were coated onto ELISA plates. After a blocking step with BSA, the wells were incubated with recombinant BLM alone (bars 1 and 2), DNase I (5 µg/ml)-treated recombinant BLM (bar 3), or heat-denatured (H.D.) recombinant BLM (bar 4). Bound BLM was detected using anti-BLM antibodies and secondary horseradish peroxidase (HRP)-labeled antibodies. All the values are corrected for the background signal, which corresponds to the A490 signal (0.1) of anti-BLM antibodies incubated with coated WRN. All values represent the mean of three independent experiments performed in duplicate.

WRN and BLM Colocalize in a Nuclear Matrix-solubilized Fraction-- Since WRN and BLM helicases colocalized to the same subcellular fraction in vivo, we next sought to verify this colocalization biochemically. Solubilized cellular fractions representing nuclear, cytoplasmic (S100), and NM extracts from exponentially growing HeLa cells were analyzed by Western blot for the presence of WRN and BLM. The localization patterns of WRN and BLM were very similar with both proteins primarily found in the nuclear matrix-solubilized fraction (Fig. 3A, lane 2). A small amount of each protein was found in the nucleoplasm (Fig. 3A, lane 1), whereas there was little (WRN) or no (BLM) signal in the cytoplasm (S100) (Fig. 3A, lane 3). The subcellular localization pattern of BLM agrees with previous studies (27, 28).

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Intracellular distribution of BLM and WRN and co-immunoprecipitation (IP) of WRN with anti-BLM antibodies from nuclear matrix extracts. A, equal amounts of protein of HeLa S3 nuclear extracts (NE, lane 1), NM-solubilized fraction (lane 2), and cytoplasmic extracts (S100, lane 3) loaded onto an SDS-polyacrylamide gel and analyzed by Western blot with the corresponding antibodies as described under "Materials and Methods." B, one mg (~500 µl) of NM-solubilized fraction incubated with either rabbit anti-BLM antibodies (lane 2) or rabbit control antibodies (anti-GST, lane 3) for 16 h at 4 °C. The bound proteins were eluted with protein sample buffer and analyzed by SDS-PAGE and Western blot with anti-WRN monoclonal antibody and anti-BLM antibody as indicated. The asterisk indicates cross-reactivity of the antibody.

Coimmunoprecipitation of WRN Using Anti-BLM Antibodies-- The next step was to perform coimmunoprecipitation experiments using the nuclear matrix-solubilized fraction. Previous studies have failed to detect a possible WRN-BLM interaction (28), maybe because the nuclear matrix solubilization method used involved SDS, urea, and/or ammonium sulfate, which can destabilize protein-protein interactions. Here, we solubilized the nuclear matrix components in a buffer suitable for immunoprecipitation analysis (see "Materials and Methods"). Under these conditions, WRN coimmunoprecipitated with anti-BLM antibodies and not with control antibodies (Fig. 3B, lanes 2 and 3). The fraction of WRN molecules that immunoprecipitated with anti-BLM antibodies was only about 5%. This may be due to differences in WRN and BLM expression patterns during the cell cycle. WRN is expressed constitutively, whereas BLM is expressed during the S/G₂ phase (29). Alternatively, the lack of binding by the antibody could be due to blocking of the binding site by WRN. We were not able to detect the reverse immunoprecipitation (anti-WRN, blot anti-BLM), possibly for similar reasons. The complex containing WRN and BLM proteins may only contain a limited fraction of the cellular amounts of these proteins, but it is also possible that the WRN-BLM association is destabilized during the cellular fractionation steps (see "Materials and Methods"). However, as will be described later (see Fig. 4D), we were able to pull down BLM from HeLa nuclear extract using a GST-tagged C-terminal fragment of WRN. In conclusion, WRN and BLM localized to the same subcellular compartment, and fractions of the two proteins are associated with each other.

BLM Binds to the N-terminal and RecQ conserved (RQC)-containing Domains of WRN-- To map the BLM binding domain(s) of WRN, we cloned and expressed a battery of WRN fragments as GST fusion proteins (Fig. 4, A and B). Some of the bands in the gel (Fig. 4A, lanes 5 and 6) are of smaller molecular weight than the full-length GST-WRN domains (helicase and C-terminal domains, respectively). They were identified by anti-GST Western blot to be degradation products (data not shown). The recombinant proteins were used in a series of binding assays with radiolabeled, in vitro translated BLM ([³⁵S]BLM). As shown in Fig. 4C, [³⁵S]BLM bound to three different domains of WRN. Although the first 51 aa of WRN failed to bind [³⁵S]BLM (lane 2), a domain that overlaps with the exonuclease domain (first 120 aa) clearly showed binding to [³⁵S]BLM (Fig. 4C, lane 3). The [³⁵S]BLM binding efficiency of this N-terminal fragment was about 15% of the input. The acidic region of WRN was unable to bind [³⁵S]BLM (Fig. 4C, lane 4). All the proteins containing a portion of the conserved RQC terminal motif (Fig. 4, B and E, red box) bound to [³⁵S]BLM (Fig. 4C, lanes 5-8). The [³⁵S]BLM binding efficiency of the RQC-containing C-terminal fragment of WRN (lane 6) was about 30% of the input. In contrast, a C-terminal region of WRN, lacking the RQC motif, failed to bind [³⁵S]BLM (lane 9). Importantly, a small WRN region of 144 aa (aa 949-1092, lane 8) was able to bind [³⁵S]BLM, although less efficiently (about 8% of input) than the full-length C-terminal fragment (Fig. 4C, compare lanes 6 and 8). Confirming this physical interaction, the small fragment of WRN (aa 949-1092) was able to pull down BLM from HeLa cell nuclear extract (Fig. 4D). This also ruled out the possibility that the binding of this fragment to [³⁵S]BLM was restricted to recombinant proteins.

View larger version (36K): [in this window] [in a new window]   Fig. 4.   Mapping the BLM binding domains of WRN. A, single-step purified recombinant GST-WRN fragments used for the binding experiments (lanes 1-9). B, scheme of WRN sequence (numbers correspond to amino acid sequence) corresponding to each recombinant domain as GST fusion fragments. C, for the binding experiment, the GST-WRN fragments were bound to glutathione beads. After washing, the beads were incubated with in vitro translated BLM ([35S]BLM) for 2 h at room temperature. The bound [35S]BLM was eluted with protein sample buffer and analyzed in SDS-PAGE followed by fluorography. D, coprecipitation of BLM with GST-WRN (949-1092) from HeLa nuclear extract. Lane 2, GST. Lane 3, GST-WRN (949-1092). Lane 4, GST-WRN (1072-1236). Fragments were bound to glutathione beads and then incubated with HeLa nuclear extract. The presence of BLM was analyzed by Western blot. The control for equal gel loading is indicated by Amido Black staining of the membrane. E, scheme of functional domains of WRN and binding sites (solid lines) of proteins that modulate WRN exonuclease activity. Dashed lines show other possible binding sites. HRDC, helicase, RNAseD, C-terminal conserved region. NLS, nuclear localization signal.

We have recently demonstrated (30) that this same WRN region (aa 949-1042) functionally interacts with flap endonuclease 1 (FEN-1). This result, together with the binding to BLM described here, indicates that this domain plays a central role in the regulation of WRN-interacting proteins. Based on the above results, we conclude that [³⁵S]BLM binds to the N-terminal domain (aa 51-120) and to all fragments containing the RQC domain sequences of WRN. We cannot rule out that there may be additional binding sites in the WRN helicase domain (Fig. 4E). These results were confirmed with ELISA analysis, which showed a direct binding of recombinant BLM to both the exonuclease and the RQC-containing regions of WRN (data not shown).

The WRN binding domain(s) of BLM may reside in the central region of the BLM protein. A ³⁵S-labeled C-terminal truncated BLM mutant (aa 1-435) failed to bind to any of the WRN fragments, whereas the first 901 aa of BLM (aa 1-901) were able to bind both the WRN N-terminal and the RQC-containing domains (data not shown).

Lack of Synergism between WRN and BLM in the Helicase Reaction-- To investigate a functional interaction between BLM and WRN, we tested for a potential cooperation/synergism of the two helicases using several double-stranded DNA substrates. We have shown previously that WRN and BLM helicases unwind short (28-bp) M13 partial duplex substrates (4, 19). However, WRN does not unwind longer DNA duplex substrates such as a 69-bp partial duplex (4), whereas BLM can unwind a 71-bp M13 partial duplex but not a 257-bp substrate (19). The limited unwinding reactions of both WRN and BLM can be enhanced by a physical and functional interaction with RPA (4, 19). The physical interaction between WRN and BLM suggested that their reduced activity on long DNA duplexes might be enhanced if the enzymes function together as a complex during DNA unwinding. This would be consistent with the prevailing evidence that a large number of DNA helicases are likely to function as higher order oligomers to unwind double-stranded DNA (31). We tested this hypothesis using a 100-bp M13 partial duplex as the helicase substrate. WRN alone failed to unwind the substrate (Fig. 5, lanes 1 and 4). BLM unwound 37% of the 100-bp M13 partial duplex substrate at a monomer concentration of 36 nM (lanes 3 and 6). When BLM (36 nM) and WRN (18 nM, lane 2 or 36 nM, lane 5) were present together, a similar amount of unwinding (35%) was observed as compared with the reactions containing only BLM. These results suggest that neither BLM nor WRN helicase activity on the 100-bp duplex substrate was increased in the presence of the other helicase. Similar results were obtained using a short 28-bp M13 partial duplex, 19- and 34-bp forked oligonucleotide duplexes, and a 44-bp oligonucleotide duplex substrate containing a central nick (data not shown). In addition, we also tested the ability of the two helicases to individually or jointly unwind a Holliday junction recombination intermediate. This is a preferred helicase substrate for WRN and BLM (13). We found no synergistic activity of WRN and BLM on this substrate despite the ability of each enzyme to unwind the structure (data not shown). The structures tested so far are shown in Fig. 5B. The collective results indicated a lack of synergistic unwinding by WRN and BLM on a number of B-form DNA duplex substrates of varying lengths or structures.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   The presence of WRN and BLM together does not increase their helicase activities. A, WRN and BLM were preincubated at 24 °C for 4 min, and then the reactions were initiated by the addition of the DNA substrate (0.3 nM 100-bp M13mp18 partial duplex) and incubated for 30 min at 24 °C. Reactions (20 µl) contained ~0.3 nM partial duplex DNA substrate molecules (final concentration) and 36 nM BLM (monomer). Reactions in the presence of WRN contained either 18 or 36 nM WRN monomer (final concentration). Lane 1, 18 nM WRN; lane 2, 18 nM WRN + 36 nM BLM; lane 3, 36 nM BLM; lane 4, 36 nM WRN; lane 5, 36 nM WRN + 36 nM BLM; lane 6, 36 nM BLM; lane 7, no enzyme; lane 8, heat-denatured DNA substrate. B, helicase substrates used in this study. Each is a substrate for both WRN and BLM, but there is no synergy in the unwinding when both helicases are together. *, 5-[-32P]ATP-labeled end.

The helicase reactions using the 100-bp M13 partial duplex were done at 24 °C. However, similar results were achieved when reactions were performed at 37 °C on a forked duplex (Fig. 5B) with a mutant form of WRN that lacks exonuclease activity (E84A) (32) (data not shown).

The Exonuclease Activity of WRN Is Inhibited by BLM-- In contrast to the results from the helicase assay, the exonuclease activity of WRN was substantially inhibited when BLM was present in the reaction (Fig. 6A, lanes 4-7). This inhibition was evident at a 1:2.4 WRN:BLM molar ratio (lane 6) and was dependent on native BLM because heat-inactivated BLM was unable to inhibit the WRN exonuclease reaction (lane 8). To confirm that this inhibition was due to direct binding, we tested a WRN truncated protein that only contained the WRN exonuclease domain (aa 1-368) and retained exonuclease activity (N-WRN) (33). As shown in Fig. 6B, increasing amounts of BLM were also able to inhibit the N-WRN exonuclease activity. In this case, an N-WRN:BLM molar ratio of 1:1.6 (lane 7) yielded the same percentage of inhibition as with the full-length WRN (Fig. 6A, lane 6). Thus, BLM inhibited the exonuclease activity of both the full-length and the N-terminal truncated WRN proteins at similar molar ratios. Since BLM could not unwind the 5'-tailed duplex substrate used in this experiment,² the inhibition of WRN exonuclease was not due to simple unwinding of the duplex substrate. The inhibition may have resulted from BLM binding to the DNA, thus impairing WRN functions. However, gel shift analyses suggested that BLM did not bind to the WRN exonuclease substrate (data not shown). Thus, these results demonstrate that the binding of BLM to the N-terminal region of WRN (aa 1-120) is responsible for its exonuclease inhibition.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   BLM specifically inhibits the exonuclease activity of both full-length WRN and an N-WRN. A, WRN protein (40 ng) was incubated with the exonuclease substrate (3'-recessed DNA substrate, represented on the top of the figure) in the presence of increasing amounts of BLM under exonuclease reaction conditions for 1 h at 37oC. Once the reactions were stopped, the samples were loaded onto denaturating PAGE in the following order: lane 1, no enzyme; lane 2, BLM (125 nM); lane 3, WRN (25 nM); lane 4, WRN (25 nM) + BLM (15.6 nM); lane 5, WRN (25 nM) + BLM (31.2 nM); lane 6, WRN (25 nM) + BLM (62.4 nM); lane 7, WRN (25 nM) + BLM (125 nM); lane 8, WRN (25 nM) + BLM (125 nM, heat-denatured (H.D.)). B, N-WRN fragment (aa 1-368) was incubated with BLM in the same conditions as described in panel A. The samples were loaded in the following order: lane 1, no enzyme; lane 2, BLM (62.5 nM); lane 3, N-WRN (36.6 nM); lane 4, N-WRN (36.6 nM) + BLM (7.8 nM); lane 5, N-WRN (36.6 nM) + BLM (15.6 nM); lane 6, N-WRN (36.6 nM) + BLM (31.2 nM); lane 7, N-WRN (36.6 nM) + BLM (62.5 nM); lane 8, N-WRN (36.6 nM) + BLM (62.5 nM, heat-denatured (H.D.)). C, exonuclease III (ExoIII), DNA polymerase I (Klenow), or WRN were incubated with the exonuclease substrate in the presence of BLM. The reaction was carried out at 37 °C either for 1 h (for WRN and Klenow) or 10 min (for exonuclease III). The samples were loaded in the following order: lane 1, no enzyme; lane 2; BLM (125 nM); lane 3, WRN (25 nM); lane 4, WRN (25 nM) + BLM (125 nM); lane 5, ExoIII (0.1 units); lane 6, ExoIII (0.1 unit) + BLM (125 nM); lane 7, Klenow (2 units); lane 8, Klenow (2 units) + BLM (125 nM). For the Ku stimulation experiment: lane 9, WRN (25 nM) + Ku 70/80 (25 nM); lane 10, WRN (25 nM) + Ku 70/80 (25 nM) + BLM (125 nM).

We next explored whether the functional interaction between WRN and BLM was specific for WRN exonuclease. As shown in Fig. 6C, BLM did not affect the digestion activity of exonuclease III (lanes 5 and 6) or Klenow (lanes 7 and 8), whereas it inhibited the WRN exonuclease (Figs. 4A and 6C, lanes 3 and 4). Thus, the functional interaction between BLM and WRN appeared to be specific for the WRN exonuclease. As described previously (5, 6), the Ku heterodimer binds to both the N-terminal and C-terminal ends of WRN (Fig. 4E) and dramatically stimulates the exonuclease activity of WRN (2, 6). We therefore investigated whether Ku and BLM could functionally compete for the same site in the WRN N-terminal region. As shown in Fig. 6C (lanes 9 and 10), the Ku stimulation dominated the BLM inhibition when both proteins were present, even when BLM was present at higher molar ratios (1:5 Ku:BLM molar ratio). There are several possible explanations for this observation. Ku and BLM may bind to the same WRN region, but Ku may have stronger affinity for WRN than BLM. Alternatively, Ku and BLM may bind to different WRN regions, and the exonuclease stimulation might be dominant. It is also possible that Ku and BLM bind to each other and coordinately affect WRN. A genetic interaction between Drosophila BLM and Ku70 was recently reported (34), but this interaction has not been noted in human cells. Our finding that Ku stimulates the WRN exonuclease activity in the presence of BLM indicates that the inhibition of WRN exonuclease by BLM is specific and not due to the presence of a potential contaminant in the BLM preparation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In the present work, we have presented clear evidence that the RecQ helicase family members WRN and BLM are components of the nuclear matrix (nuclear foci) and interact physically and functionally. Using different techniques, we showed that: 1) WRN and BLM colocalized in three different human cell lines. 2) Both proteins interacted directly (ELISA). Using in vitro binding experiments, we showed that WRN has two binding sites for BLM. 3) BLM and WRN primarily localized in the same nuclear matrix-enriched subcellular fraction. 4) A fraction of cellular WRN coimmunoprecipitated from this fraction using specific anti-BLM antibodies. Also, a fraction of cellular BLM coprecipitated from nuclear extracts using a GST-tagged C-terminal domain of WRN. 5) Functionally, BLM inhibited the exonuclease activity of WRN.

The colocalization and coimmunoprecipitation of WRN/BLM do not necessarily imply that the proteins interact physically in vivo. However, the in vitro results indicate that these proteins can interact physically and functionally and may do so in vivo. A fraction of the cellular BLM and WRN may act together as a complex in some aspect of DNA metabolism.

Both WRN and BLM have been reported to colocalize with promyelocytic leukemia (PML) bodies (nuclear matrix-associated structures) (24, 25), which is in accordance with our observation here (GFP-WRN and BLM colocalize in nuclear foci in HeLa cells). We also find that GFP-WRN and PML colocalize in HeLa cells (data not shown). Several observations suggest that an N-terminal tag (either GFP or c-Myc, both similar in size) does not affect WRN functions: 1) Expression of a GFP-WRN construct was able to rescue the 4-NQO sensitivity of SV40-transformed WS cells (35). 2) Purified WRN containing a c-Myc tag retained helicase activity (36). 3) The GFP-WRN fusion protein localized to the nucleolus and/or in nuclear foci, as described for endogenous WRN (25, 37).³ Thus, the specific GFP labeling of the WRN protein provided an excellent and physiological tool to follow the intracellular localization of this protein.

The number of GFP-WRN/BLM nuclear foci was higher in AG11395 and U-2 OS cells than in HeLa cells. These cell lines are both telomerase-negative (ALT) (26), implying that telomere maintenance occurs independently of telomerase activity. ALT cells contain a unique type of PML body that includes the proteins PML, RPA, RAD51, RAD52, TRF1, and TRF2 (26). The higher number of GFP-WRN and BLM nuclear foci in the AG11395 and U-2 OS cells may represent these ALT-associated PML (AA-PML) bodies. Consistent with this, we have observed a higher number of AA-PML bodies in U-2 OS and AG11395 than of PML bodies in HeLa cells (data not shown). One or more of the proteins in the AA-PML body (RPA, p53, and RAD51) are known to bind to WRN and/or BLM and could function as a recruiting factor for both WRN and BLM proteins. There is evidence that cellular conditions (cell cycle phase, DNA damage) dictate which proteins are recruited to PML- and AA-PML-containing foci in a dynamic process. For example, it has been shown that the frequency of cells containing AA-PML bodies is high in the G₂/M phase of the cell cycle (38). In addition, we observed that some GFP-WRN and BLM nuclear foci did not colocalize. This suggests that the formation of the WRN-BLM protein complex is dynamic and that the colocalization that we detected is not an artifact of the expression of GFP-WRN.

What aspects of DNA metabolism may involve a WRN-BLM complex? One possibility is that WRN and BLM can function together in the processing of telomeric DNA ends. Accelerated rates of telomere shortening have been observed in some BS and WS cell lines (24). As mentioned previously, colocalization with telomere-binding proteins was detected for BLM in normal cells and in SV40-transformed human fibroblasts (24) and for WRN in telomerase-negative immortalized human fibroblasts (ALT cells) (25). Recently, the RecQ yeast homologue, Sgs1, was shown to participate in a recombination-mediated pathway for lengthening telomeres that occurs in the absence of telomerase (25). A WRN-BLM complex may function at telomeres by resolving the secondary structure to allow access to replication, repair, and/or telomere lengthening machinery. Consistent with this notion, both WRN and BLM unwind G4 tetraplex DNA structures (13), which can form in telomeric sequences. In addition, human telomeric ends have been observed to form large t-loop structures that are stabilized by D-loops (39), which may be unwound by a BLM-WRN complex. Furthermore, inhibition of WRN exonuclease by BLM may be important for preventing digestion at the telomeric end (Table I).

View this table: [in this window] [in a new window]   Table I WRN exonuclease activity modulationpossible implications in different cellular processes

WRN and BLM may also function together in other aspects of DNA metabolism. Both WS and BS cell lines exhibit hyper-recombination phenotypes (11), suggesting that WRN and BLM act in preventing illegitimate recombination. The expression of WRN or BLM genes in a Sgs1-deficient yeast strain suppresses the hyper-recombination phenotype (40). In addition, both proteins can act on recombination intermediates in vitro. Specifically, both WRN and BLM promote branch migration of Holliday junctions (41) and unwind Holliday junction structures (13, 42). In addition, both WS and BS cells exhibit replication abnormalities, including an extended S phase (11). It has been proposed that these enzymes may function during replication fork arrest to resolve collapsed or regressed replication forks (43). Furthermore, both BLM and WRN helicases interact physically and functionally with RPA (4, 19), and WRN was observed to colocalize with RPA in vivo during S phase arrest (41). Thus, it seems likely that a BLM-WRN-RPA complex may act to resolve recombination intermediates, such as those formed as a result of replication fork arrest.

As shown here, the WRN-BLM physical interaction is mediated by at least two domains of the WRN protein. The BLM binding to the N-terminal region of WRN is likely to be responsible for the WRN exonuclease inhibition. WRN differs from the other human RecQ family members by having this exonuclease domain, which resides in the N-terminal region of the protein. Thus, the specific interaction and inhibition of the exonuclease activity of WRN by BLM place this domain in a critical pathway for the regulation of WRN functions that may relate to the maintenance of genomic stability. WRN exonuclease activity is not only modulated by BLM but is stimulated by the Ku complex (5) and is inhibited by p53 (44) (Fig. 4E). Although Ku (6) and BLM interact directly with the N terminus of WRN, p53 interacts only with the C-terminal end of WRN (8). The action of the WRN exonuclease activity appears to depend on the specific protein complex in which WRN is participating (Table I). The C-terminal end of WRN has no known catalytic activity but appears to recruit several other proteins (Ku, p53, FEN-1, and BLM). Some of these may compete for the same binding site. For example, we find that BLM binds to the same small 144-aa region of WRN (aa 949-1092) to which FEN-1 also binds (30). This particular region of WRN may play an important role in modulating other proteins (FEN-1) or in recruiting WRN exonuclease-modulating proteins (BLM).

Since BLM and WRN participate in the same complex, it is possible that they could accomplish some task together that neither could accomplish alone. This could, for example, be the unwinding of some DNA structure or intermediate, a special DNA repair intermediate, an RNA/DNA duplex structure, or bypassing a specific lesion in the DNA. Thus far, our in vitro studies of the unwinding of different substrates by the two helicases together have not shown any synergy. It is possible that the key function of the WRN-BLM complex is to recruit other cellular proteins necessary for the resolution of different DNA/DNA and/or RNA/DNA structures. A recent study reported that both WRN and BLM may be involved in DNA repair in a complementary fashion. Based on measurements of hypersensitivity to DNA-damaging agents, synergism was observed when cells were deficient in both WRN and BLM proteins (45).

When one of the RecQ helicase family members is missing (WS or BS), the efficiency of forming the relevant protein complexes would be hampered. Consequently, pathways involved (telomere maintenance, recombination, DNA damage response) might be deficient, and cellular processes related to aging and cancer formation could be accelerated.

	ACKNOWLEDGEMENTS

We thank Dr. V. Colombo for help with cloning steps and J. Sommers and J. Piotrowski for technical assistance. We also thank Dr. Magdalena Juhaszova for confocal technical assistance and Drs. C. Chen and N. Souza-Pinto for comments.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Laboratory of Molecular Gerontology, NIA, National Institutes of Health, 5600 Nathan Shock Dr., Baltimore, MD 21224. Tel.: 410-558-8162; Fax: 410-558-8157; E-mail: vbohr@nih.gov.

Published, JBC Papers in Press, March 27, 2002, DOI 10.1074/jbc.M200914200

² R. M. Brosh, unpublished information.

³ C. von Kobbe and V. A. Bohr, submitted.

	ABBREVIATIONS

The abbreviations used are: WS, Werner syndrome; NM, nuclear matrix; WRN, Werner syndrome protein; N-WRN, N-terminal WRN fragment; BS, Bloom syndrome; BLM, Bloom syndrome protein; ALT, alternative lengthening telomere maintenance; PML, promyelocytic leukemia protein; AA-PML, ALT-associated PML; GFP, green fluorescent protein; ELISA, enzyme-linked immunosorbent assay; GST, glutathione S-transferase; PBS, phosphate-buffered saline; aa, amino acids; BSA, bovine serum albumin; RPA, replication protein A; RQC, RecQ conserved domain; FEN-1, flap endonuclease 1; 4-NQO, 4-nitroquinoline 1-oxide.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				N. Selak, C. Z. Bachrati, I. Shevelev, T. Dietschy, B. van Loon, A. Jacob, U. Hubscher, J. D. Hoheisel, I. D. Hickson, and I. Stagljar The Bloom's syndrome helicase (BLM) interacts physically and functionally with p12, the smallest subunit of human DNA polymerase {delta} Nucleic Acids Res., September 1, 2008; 36(16): 5166 - 5179. [Abstract] [Full Text] [PDF]

J. Grillari, H.