The Histochemical Journal 34: 223–231, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.

The expression of Ras–GTPase activating protein SH3 domain-binding proteins, G3BPs, in human breast cancers Juliet French1,† , Ren´ee Stirling2,† , Michael Walsh3 & Hendrick Daniel Kennedy2,∗ Institute for Molecular Bioscience, University of Queensland, St. Lucia, 4072, Australia 2 School of Biomolecular and Biomedical Science, Griffith University, Nathan, 4111, Australia 3 Queensland Institute of Medical Research, Royal Brisbane Hospital, Herston, 4029, Australia

1

∗

Author for correspondence

†

Contributed equally

Received 10 June 2002 and in revised form 12 September 2002

Summary Ras–GTPase activating protein SH3 domain-binding proteins 1 and 2 (G3BP1 and G3BP2) have recently been reported to be encoded by two separate genes on human chromosomes 5 and 4 respectively and have been implicated in Ras signalling, NFkappaB signalling, the ubiquitin proteosome pathway and RNA processing. In addition, G3BP1 has recently been implicated in cancer biology. The transcripts for these genes have been shown to be universally expressed; however, this is not the case for the proteins which appear to be tissue and cell type specific. We report here the expression of G3BP1 and 2 in human breast cancers and present the first data showing that G3BP2 expression is specific in human breast cancer tissue and was over-expressed in 88% of tumours examined (n = 58). Introduction Human Ras–GAP SH3-domain binding protein (G3BP) was first identified by its co-immunoprecipitation with the SH3 domain of Ras–GAP (Parker et al. 1996), a key regulatory molecule in the signalling pathway downstream of the oncogene ras (Malumbres & Pellicer 1998). Around the same time, a second gene with high homology to G3BP was isolated from mice during a general screen for proteins containing an RNA recognition motif (RRM) (Kennedy et al. 1996). This gene was originally thought to be the mouse homologue of human G3BP, but subsequent identification of its human homologue revealed that this is not the case and that G3BP is a member of a larger gene family. The two genes were later termed G3BP1 and G3BP2, in order of their discovery. To date, the G3BP family in mammals is comprised of two genes, G3BP1 and G3BP2, which map to chromosomes 5q33.1–5q33.3 and 4q12–4q24, respectively in humans (Kennedy et al. 2001). G3BP1 and G3BP2 are highly homologous with 61% identity at the nucleotide level and 59% identity at the amino acid level. G3BP2 has two spliced isoforms which are known to produce proteins (G3BP2a and b). Human G3BP1, G3BP2a and G3BP2b are 466, 482 and 449 amino acids long, respectively (Parker et al. 1996, Kennedy et al. 2001). G3BP1, 2a and 2b all contain an N-terminal NTF2-like domain (Suyama et al. 2000), a minimal putative Src homology 3 (SH3) domain binding sequence (Saksela et al. 1995), an acid rich domain and two domains commonly found in RNA binding proteins, an RRM (Kenan et al. 1991) and an RGG domain (Siomi & Dreyfuss 1997).

Evidence to date suggests that G3BPs may have a general role to play in cell proliferation. Proliferative vitreoretinopathy is a visual impairment characterized by dedifferentiation and proliferation of a range of cells including the retinal pigment epithelial cells and it has been shown that these cells over-express G3BP1 (Kociok et al. 1999). On further investigation into the role G3BP1 may play in cell proliferation, Guitard et al. (2000) found that G3BP1 was up-regulated in a range of cancer tissues as compared to adjacent normal tissue by Western blot analysis. In addition, it was shown that G3BP1 promotes S-phase entry in serum starved fibroblasts (Guitard et al. 2001). Transient transfection with G3BP1 led to an increase in cell proliferation, as measured by incorporation of the thymidine analogue bromodeoxy uridine (BrdU). However, co-transfection with G3BP1 lacking the RNA binding domain (G3BPRNAb) significantly reduced the G3BP1 induced increase in BrdU uptake. Therefore, G3BPRNAb acts as a dominant negative mutant (Guitard et al. 2001) and it is possible that this mutant competes for a vital partner, but is then unable to fulfil its function due to the lack of an RNA binding domain. A similar mechanism has been shown for the RNA binding protein, Sam68, whereby a splice variant in which part of the RNA binding domain is missing actually inhibits serum-stimulated progression into S-phase of the cell cycle (Barlat et al. 1997). This suggests that the RNA binding domain of a range of proteins may play a role in cell proliferation. G3BPs have been implicated in several pathways that are known to be involved in cancer biology including Ras signalling (Parkeret al. 1996, Pazman et al. 2000), c-myc mRNA

224 turnover (Gallouzi et al. 1998, Tourriere et al. 2001), NFkappaB signalling (Prigent et al. 2000), the ubiquitin proteosome system (Soncini et al. 2001) and Her2 signalling (Barnes et al. 2002). Ras is a small GTP binding protein which, on activation by receptor tyrosine kinases, induces signal transduction cascades to the nucleus via mitogen-activated protein (MAP) kinases and other pathways (Malumbres & Pellicer 1998). Ras is switched between the active GTP state and the inactive GDP state by the action of guanine nucleotide exchange factors (GEFs), which increase GTP loading and GTPase activating proteins (GAPs), which stimulate the hydrolysis of GTP. Therefore, Ras–GAP is a negative regulator of Ras activity (reviewed in Tocque et al. (1997)). Nevertheless, evidence suggests that Ras–GAP also acts as an effector of Ras. Ras–GAP is a 120 kDa protein, the carboxy (C)-terminal domain of which binds Ras while the amino (N)-terminal contains an SH3 domain flanked by two SH2 domains (reviewed in Tocque et al. (1997)). The SH3 domain of Ras–GAP appears to be essential for triggering downstream signals. Antibodies directed against the SH3 domain of Ras–GAP block activation of maturation promoting factor by Ras without blocking the MAP kinase pathway (Duchesne et al. 1993). Therefore, it appears that Ras activates a signal transduction pathway via the SH3 domain of Ras–GAP and independent of the MAP kinase pathway. G3BP1 binds the SH3 domain of Ras–GAP only when Ras is active and is therefore an attractive candidate for an effector of Ras (Parker et al. 1996, Pazman et al. 2000). The putative RNA binding activity of G3BP1 was first addressed by Gallouzi et al. (1998) through investigation of the interaction of G3BP1 with c-myc mRNA. The transcription factors c-myc, c-fos and c-jun are early-response protooncogenes that are immediately up-regulated upon addition of growth factors to the cell. Over-expression of c-myc, c-fos or c-jun lead to uncontrolled cellular proliferation and in particular, inhibition of c-myc prevents cell proliferation even in the presence of growth factor stimulation (Almendral et al. 1988, Cross & Dexter 1991). In quiescent cells only, G3BP1 has been shown to be hyperphosphorylated, and exhibits a phosphorylation-dependent and specific RNase activity on the 3 untranslated region (UTR) of c-myc mRNA (Gallouzi et al. 1998). It was recently discovered that G3BP1 expression is regulated through the Her2 receptor (Barnes et al. 2002), an epidermal growth factor receptor which is over-expressed in 10–40% of human breast carcinomas (Menard et al. 2000). Heregulin, a ligand for Her2 was shown to induce mRNA and protein expression of G3BP1 in breast cancer cells. By immunoblot, increased Her2 expression was associated with increased G3BP1 in a small tumour set. Heregulin treatment also promoted G3BP1 phosphorylation and association with Ras–GAP. Translocation to the nucleus and co-localization with acetylated histone H3, a hallmark of active transcription sites, was also observed following stimulation of cells with heregulin. Each of these responses to heregulin could be blocked by pre-treatment with an antibody that blocked the Her2 site (Barnes et al. 2002).

J. French et al. Evidence to date suggests that both G3BP1 and G3BP2 have some role to play in cell proliferation and cancer formation or progression. We report here the first data that illustrates the specific over-expression and nuclear localization of G3BP2 in human breast tumours by immunohistochemistry.

Materials and methods Antibodies Anti-G3BP2 antibodies were raised as previously described (Kennedy et al. 2001). The commercial monoclonal G3BP1 antibody (BD Biosciences, Sydney, Australia) was used to monitor G3BP1 expression. Patients Fifty-nine cases of invasive breast carcinoma diagnosed at the Department of Pathology, Royal Brisbane Hospital, between 1981 and 1990 were randomly selected for inclusion in this study. Archival paraffin blocks were accessed subject to ethics approval from the Royal Brisbane Hospital and had been previously characterized as part of a larger study of MUC1 epithelial mucin expression (McGuckin et al. 1995). Histological classification and grading of the tumours was performed in accordance with the Nottingham modification of the Bloom and Richardson system (Elston & Ellis 1990). Data including nodal status and oestrogen receptor status as determined by biochemical dextran-coated charcoal method were obtained from clinical charts and pathology records. Immunohistochemistry of breast cancer sections Breast tumour sections (3–4 µm) were affixed to adhesive slides and dried at 37 ◦ C overnight. The sections were dewaxed and rehydrated through descending graded ethanols to deionized water using standard protocols. Sections to be stained for G3BP1 were subjected to antigen heat retrieval by autoclaving at 120 ◦ C for 20 min in 1 mM EDTA, pH 8.0. G3BP2 samples were subjected to antigen heat retrieval by boiling in 25 mM Tris–HCl, pH 9.0–9.2, for 5 min in a microwave, and repeating the process using fresh Tris–HCl buffer. All sections were allowed to cool to room temperature (20–30 min) and then washed in Tris–buffered saline, pH 7.4 (TBS). Endogenous peroxidase activity was blocked by incubating the sections in 1.0% H2 O2 , 0.1% NaN3 in TBS for 10 min. Sections were washed in TBS and subsequently incubated in 4% non-fat skim milk powder in TBS for 15 min. Sections were rinsed briefly in TBS and then incubated with 10% normal goat serum (NGS) for 20 min in a humidified chamber. Excess normal serum was decanted and primary antibody (or TBS as negative control) was applied overnight at room temperature. Sections were washed in TBS and then incubated with secondary antibody (DAKO, Glostrup Denmark, EnVision Kit) for

G3BPs in breast cancer 45 min. Sections were washed in TBS and colour was developed in 3,3 -diaminobenzidine (DAB) with H2 O2 as substrate. Sections were washed in gently running tap water then lightly counterstained in Mayers’ Haemotoxylin, dehydrated through ascending graded ethanols, cleared in xylene, and mounted in DPX mounting medium. Cell cycle synchronization of NIH 3T3 cells Cell cycle synchronization of NIH 3T3 cells was performed using the serum deprivation method (Tobey et al. 1988). NIH 3T3 cells were seeded onto coverslips at sub-confluent conditions in 10% FCS. Following 24 h, the cells were washed 3 times with phosphate buffered saline (PBS) and serum free medium was added back to the cultures and left for 24 h. The serum free medium was then replaced with medium containing 10% foetal calf serum (FCS). Coverslips were removed from the media during serum starvation, and at 2, 5, 9 and 12 h after serum stimulation. The coverslips were then processed for immunofluorescence to examine the expression of G3BP2.

225 SV40 transformed normal breast epithelial cell line SVCT, the breast cancer cell line MDA-MB-435 and the cervical cancer cell line HeLa were maintained in vitro in DMEM or RPMI supplemented with 10% FCS. The cells were harvested by trypsinization, washed twice with PBS and resuspended in HNTG buffer (50 mM Hepes, pH 7.5, 150 mM NaCl, 1% Triton X-100, 10% glycerol, 1 mM MgCl2 , 1 mM EGTA, 1 mM Na3 VO4 , 10 mM Na4 P2 O7, 10 mM NaF, 1 mM PMSF and 1 X mammalian protease inhibitor cocktail no P8340) (Sigma, Castle Hill, Australia). Lysates were cleared by centrifugation at 15 000 rpm for 10 min and the protein concentration was determined using the Pierce BCA Protein Assay (Rockford, USA). A total of 50 µg of protein was resolved by 8% SDS–PAGE and transferred to an Immobilon-P PVDF membrane (Millipore, Sydney, Australia) for Western analysis using the antibodies described above. Proteins were visualized by horseradish peroxidase (HRP) conjugated antirabbit or mouse antibodies using an ECL system (Amersham Biosciences, Sydney, Australia).

Results

Immunofluorescence of cultured cells

G3BP2 is over-expressed in 88% of breast tumours

NIH 3T3 cells were grown on coverslips, treated as described above, and washed 3 × 2 min with PBS and dried overnight at room temperature. The cells were fixed with 100% cold acetone for 5 min, allowed to dry, then rehydrated by washing the coverslips with PBS 3 × 5 min. The cells were then permeabilized by incubating the coverslips in 0.1% Triton-X 100 in PBS for 5 min. The detergent was then removed by washing the coverslips 3 × 5 min with PBS. Primary antibody, diluted to the appropriate concentration in PBS, was applied and left overnight at 4 ◦ C. The following morning the coverslips were washed 3 × 5 min in 1% NGS/1% bovine serum albumin (BSA) in PBS. Secondary antibody was then applied and incubated for 1 h at room temperature. The secondary antibody was anti-rabbit IgG conjugated with either a FITC or Rhodamine fluorescent tag (Molecular Probes, Eugene, USA) and was diluted in 0.1% Triton-X 100 in PBS at the dilution specified by the manufacturer. The coverslips were then washed 2 × 5 min in 0.1% Triton-X 100 in PBS followed by 2 × 5 min PBS. Finally, coverslips were mounted onto slides with 50% glycerol–50% PBS and sealed with nail polish. Images were generated using an Olympus Provis AX-70 and captured in digital format with a DAGE-MTI CCD camera using Scion Image 1.62 frame grabber software. Images were analysed using Adobe Photoshop 5 image-processing software (Adobe systems incorporated, Eastman Kodak Company, 1996).

It was recently discovered that G3BP1 is regulated through the Her2 receptor and it was shown, by Western blot, to be over-expressed in a small tumour set (Barnes et al. 2002). We examined the expression of G3BP1 in 24 breast tumour cases by immunohistochemistry. Of these, 22 sections were infiltrating ductal carcinomas (IDCs) and two were cases of infiltrating lobular carcinoma (ILCs). All sections were counterstained with Haematoxylin, which stains nuclei blue and the expression of G3BP1 was visualized using HRP seen as brown reaction product (see Figure 1, Panels A–C). Most normal cells exhibited detectable cytoplasmic expression of G3BP1 (see Figure 1, Panel C). Two normal ducts (ND) are seen in Figure 1, Panel C, and cytoplasmic expression of G3BP1 is apparent as seen by the distinct brown staining. G3BP1 staining is also evident in the IDC sections shown in Figure 1, Panel A and B, but the adjacent connective tissue (CT) does not express G3BP1 at detectable levels. The tumour cells in Panels A and B appear to express higher levels of G3BP1 in the cytoplasm as compared to that seen in the normal ducts of Panel C. In many cases tumour staining was heterogeneous and in some cases G3BP1 appeared to localize more prominently to one side of the cell. This can be seen quite clearly in some of the tumour cells in Panel A (indicated by the arrow). There was no nuclear staining present in any normal cells, although two of the 24 tumour cases contained distinct nuclear staining in less than 10% of tumour cells. Table 1 shows the results of all breast tumours examined for G3BP1 expression. Also listed is the available information on each of the breast tumours including its oestrogen receptor status, the grade of the tumour, and tumour stage. In summary, most normal breast cells expressed G3BP1 but all of the tumours examined appeared to over-express G3BP1 to

Cell extracts, SDS–PAGE and Western blotting The expression of G3BP in human cell lines was examined by Western blot. The human embryonic kidney cell lines HEK293T and HEK-ER-293 (Stratagene, USA), the

226

J. French et al.

Figure 1. Immunohistochemical staining of breast tumour sections and immunofluorescence of synchronized NIH 3T3 cells with antibodies specific for G3BPs. For immunohistochemistry, G3BP staining was visualized with horseradish peroxidase (brown colour) and sections were counterstained with Haematoxylin (blue colour). Panels A–C are breast tumour sections stained using an antibody specific for G3BP1. Panels A and B represent stained IDC while Panel C shows a small section of stained normal ducts (ND). G3BP1 expression is evident in normal tissue, but appears to be more prominent in tumour cells. Note also the heterogeneous nature of the staining indicated by the arrow in Panel A. Panels D–O show immunohistochemistry of G3BP2 in human breast tumours. Panel D shows a normal lobe, Panels E and F show normal ducts cut transverse and longitudinally, respectively (CT denotes connective tissue). Panels G and H show a normal duct adjacent to an IDC (DC denotes ductal carcinoma). Panel I shows an IDC which does not express G3BP2. Panel J shows an IDC adjacent to normal connective tissue. The arrow indicates cells within the connective tissue which stain positive for G3BP2 in the nucleus. Panel K shows a lower magnification of an IDC (left-hand side) adjacent to normal connective tissue. Panels L–O illustrate a variety of G3BP2 subcellular localizations in human breast cancer. All panels are ductal carcinomas from different patients. Panel L shows cytoplasmic localization of G3BP2. Panels M and N show nuclear localization of G3BP2 in two different cases of breast cancer; cytoplasmic expression is also observed in these sections. Panel O shows G3BP2 expression around the nuclear envelope region; cytoplasmic staining is also observed. Panels P–T show the immunofluorescence of synchronized NIH 3T3 cells. Cells were synchronized by serum starvation and subsequently induced to

G3BPs in breast cancer

227

Table 1. Expression of G3BP1 in 24 breast tumour sections. DCIS = ductal carcinoma in situ, IDC = infiltrating ductal carcinoma, LCIS = lobular carcinoma in situ, ILC = infiltrating lobular carcinoma. Grade is assigned according to a range of factors—a well-differentiated tumour is generally assigned a grade 1 while a poorly differentiated tumour is assigned a grade 3. ER = oestrogen receptor status. Node status refers to the presence (+) or absence (−) of tumour in the lymph nodes. NG = not graded. ND = not determined. The column labelled cytoplasm indicates the level of expression of G3BP1 in the cytoplasm of cells (1+ = low, 2+ = medium, 3+ = high). Unless stated otherwise, staining was present in greater than 75% of the cell population. The column labelled nucleus indicates the presence (+) or absence (−) of G3BP1 in the nucleus of cells. Tumour case number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 59 18 19 20 21 22 23

G3BP-1 Tumour tissue

Normal tissue

Cytoplasm

Nucleus

Cytoplasm

2+ 2+ 2+ 1–2+ 2–3+ 2–3+ 2–3+ 2+ 2–3+ 3+ 3+ 2+ 3+ 2+ 3+ 2–3+ 2–3+ 2–3+ 2–3+ 2+ 3+ 1–2+ 2+ 2–3+

Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative <10%+ Negative Negative Negative Negative Negative <10%+ Negative Negative Negative Negative Negative

Not present Not present 1+ Negative Not present 1–2+ Negative 1–2+ Negative Not present 1+ Negative 1–2+ Negative 1–2+ Negative 1+ Negative 1+ Negative 1–2+ Negative 1+ Negative 1+ Negative 1+ Negative 1–2+ Negative Not present 1–2+ Negative Not present 1–2+ Negative Not present 1+ Negative 1–2+ Negative

some extent. No significant relationship was found between G3BP1 over-expression and clinicopathological parameters of breast cancer such as lymph node involvement, hormone receptor status or nuclear or histological grade. A total of 58 breast tumour cases were examined by immunohistochemistry for altered G3BP2 expression. Of these, 54 tumours were IDCs and four were ILCs. As with G3BP1, all sections were counterstained with Haematoxylin and the expression of G3BP2 was visualized using HRP (see Figure 1, Panels D–O). Unlike that for G3BP1, the immunohistochemistry showed no detectable expression of G3BP2 in normal lobes of the breast (see Figure 1, Panel D) including the lobular and ductal epithelium and surrounding connective tissue. Panels E and F of Figure 1 show a higher magnification of two different ducts, Panel E shows a transverse section of a duct and Panel F shows a longitudinal section. As can be seen, there is no detectable expression of G3BP2 in normal ducts of the breast or within cells of the surrounding connective tissue. Immunohistochemistry revealed that G3BP2

Nucleus

Histological type and (grade)

Node status

ER

IDC (2) ILC and LCIS (2) IDC (3) IDC (2) IDC and DCIS (NG) IDC and DCIS (2) IDC (NG) ILC (2) IDC and DCIS (1) IDC and DCIS (3) IDC and DCIS (2) IDC and DCIS (2) IDC (3) IDC and DCIS (3) IDC (2) IDC (3) IDC (3) IDC (1) IDC (3) IDC (3) IDC and DCIS (1) IDC (3) IDC and DCIS (3) IDC (1)

ND + + ND − + + + − + + ND − + + − + + + − + − + −

+ − + ND ND − − + + − ND + − − + − ND + − − + ND − +

is over-expressed in breast tumours. Figure 1, Panel G shows a normal duct adjacent to an IDC. As can be seen by the brown staining, G3BP2 is highly expressed in the tumour but not expressed in the normal duct. This can be seen more clearly in Panel H which shows an IDC at higher magnification adjacent to a normal duct. Again, the normal duct does not express G3BP2 and the IDC highly expresses G3BP2. The over-expression of G3BP2 in breast tumours was not seen in all breast tumours examined (12%). Panel I of Figure 1 shows one case of IDC that does not express G3BP2. Another interesting observation noted when examining the expression of G3BP2 in human breast tumours was that in some cases G3BP2 is expressed in the nucleus of normal cells within the connective tissues lying between the tumours (marked by the arrow), but not in the cells within connective tissue away from the tumour (see Figure 1, Panel J). Figure 1, Panel K shows an example of a lower magnification of a tumour and adjacent connective tissue. As can be seen, G3BP2 is expressed in cells within the connective tissue peripheral to the tumour

enter the cell cycle by serum stimulation. Cells were stained for G3BP2 using immunofluorescent technique at several time intervals following serum starvation. Panel P shows the sub-cellular localization of G3BP2 in cells in G0 phase (time = 0). The time after serum stimulation and hence cell cycle commencement is 2, 5, 9 and 12 h for Panels Q, R, S and T, respectively.

228 and the expression becomes lower the further away the cells are from the tumour. These cells are most likely infiltrating lymphocytes as there seems to be a greater population of these cells around the tumours. This could suggest that G3BP2 expression is induced in response to a factor secreted by some tumours or that G3BP2 produces a chemotaxis-like effect. Table 2 shows the results of all breast tumours examined for G3BP2 expression. Also listed is the available information on each of the breast tumours including its oestrogen receptor status, tumour grade and stage. In summary, 88% of all tumours examined over-express G3BP2 and no significant relationship was found between G3BP2 over-expression and clinicopathological parameters of breast cancer such as stage, hormone receptor status or nuclear or histological grade. In the majority of human breast tumours that were screened, G3BP2 is over-expressed and in many cases shows a distinct nuclear localization (see Figure 1, Panels M–O). Panels L–O show four different cases of IDC with three different sub-cellular localizations. Panel L is an example of a breast tumour where G3BP2 is exclusively cytoplasmic. Panels M and N are two examples of breast tumours where G3BP2 is found in the nucleus and in the cytoplasm. Panel O shows G3BP2 localized around the nuclear envelope region. These tumours also show a cytoplasmic distribution for G3BP2. This is the first case in which G3BP2 has been found in the nucleus in situ. Approximately 50% of all tumours that express G3BP2 have G3BP2 in the nucleus, although it should be noted that it is possible that nuclear staining was not observed in some cells where G3BP2 is expressed at low levels, due to masking by the Haematoxylin counterstain. The nuclear staining varied between cases. Some cases had G3BP2 in the nucleus of all cells, whereas others had less than 10% of cells with nuclear staining. The percentage of cells which express G3BP2 in the nucleus does not correlate with the grade of the tumour or level of metastasis. In conclusion, a total of 58 breast tumours have been examined for altered G3BP2 expression. 88% were shown to express G3BP2, and of these G3BP2 was found in the nucleus of 50% of tumours. As with G3BP1, no correlation was found between G3BP2 expression in breast cancer and clinicopathological parameters. It was found that G3BP2 over-expression in breast cancer is more prominent than that of G3BP1. Localization of G3BP2 is cell cycle specific The immunohistochemistry of human breast cancers revealed that G3BP2 can also be localized to the nucleus of cells. As these cells are undergoing rapid cell proliferation, we sought to investigate whether the localization of G3BP2 is cell cycle dependent. NIH 3T3 cells were synchronized and the localization of G3BP2 was examined by immunofluorescence at different stages of the cell cycle (Figure 1, Panels P–T). During serum starvation (G0 quiescence) G3BP2 is predominantly localized to the cytoplasm of cells (Figure 1, Panel P). Panels Q and R show cells 2 h and 5 h after serum stimulation (entry into the cell cycle). As can be seen, G3BP2 has moved

J. French et al. Table 2. Expression of G3BP2 in 58 breast tumour sections. DCIS = ductal carcinoma in situ, IDC = infiltrating ductal carcinoma, LCIS = lobular carcinoma in situ, ILC = infiltrating lobular carcinoma. Grade is assigned according to a range of factors – a well-differentiated tumour is generally assigned a grade 1 while a poorly differentiated tumour is assigned a grade 3. ER = oestrogen receptor status. Node status refers to the presence (+) or absence (−) of tumour in the lymph nodes. NG = not graded. ND = not determined. The column labelled cytoplasm indicates the level of expression of G3BP2 in the cytoplasm of cells (1+ = low, 2+ = medium, 3+ = high). Unless stated otherwise, staining was present in greater than 75% of cell population. The column labelled nucleus indicates the presence (+) or absence (−) of G3BP2 in the nucleus of cancer cells. Tumour case number

Tumour tissue Cytoplasm

Nucleus

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

3+ 1+ 2+ 2+ 2+ 2+ 25–50% 1+ 1+ 3+ 2+ 2+ 2+ 2+ 1+ 2–3+ 1+ 3+ 2+ 3+ 1+ 2+ 2+ 2+ 1+ 1+ 1+ 2+ 3+ 2+ 1+ − 2+ 1+ − 2+ 1+ 2+ 1+ 1+ 1+ − 2+ 2+ 3+ 2+ 2+ 3+ 1+ 1+ 1+ 1+

<10% + <10% + + <25% + − 25–50% + − <25% + + <10% + 25–50% + <10% + <25% + <10% + − − <10% + 25–50% + <10% + − <25% + <25% + 50–75% + − − <10% + <10% + − + 25–50% + − 25–50% + − − − 50–75% + − − 25–50% + − − − − + − − − − − − −

Histological type and (grade)

Node status

ER

IDC (2) ILC and LCIS (2) IDC (3) IDC (2) IDC and DCIS (NG) IDC and DCIS (2) IDC (NG) ILC (2) IDC and DCIS (1) IDC and DCIS (3) IDC and DCIS (2) IDC and DCIS (2) IDC (3) IDC and DCIS (3) IDC (2) IDC (3) IDC (3) IDC (3) IDC (3) IDC and DCIS (1) IDC (3) IDC and DCIS (3) IDC (1) IDC (NG) IDC + DCIS (1) IDC + DCIS (2) IDC + DCIS (1) IDC + DCIS (NG) IDC + DCIS (2) IDC (2) IDC + DCIS (NG) IDC (NG) IDC + DCIS (NG) IDC + DCIS (1) IDC (NG) IDC (3) IDC + DCIS (2) IDC + DCIS (1) IDC (2) IDC + DCIS (2) IDC + DCIS (NG) IDC (2) IDC (2) IDC + DCIS (2) IDC (3) IDC (NG) IDC + DCIS (3) ILC + DCIS (1) ILC + LCIS (2) IDC (2) IDC + DCIS (3)

ND + + ND − + + + − + + ND − + + − + + − + − + − + − + − + − + + + − − + + + − − − + − + − − + − − − + −

+ − + ND ND − − + + − ND + − − + − ND − − + ND − + − + + − − − − + − − − − − + + − + + ND + + − + − + − − −

G3BPs in breast cancer

229 75 kDa

Table 2. (Continued) Tumour case number

Tumour tissue Cytoplasm

Nucleus

52 53 54 55 56 57 58

− 1+ 2+ − − 3+ −

− + + − − + −

Histological type and (grade)

Node status

ER

IDC + DCIS (3) IDC (3) IDC (3) IDC (3) IDC + DCIS (3) IDC + DCIS (3) ILC + LCIS (2)

− ND ND ND ND ND ND

− + + + + + +

G3BP1 50 kDa 75 kDa G3BP2 50 kDa 293T

into the nucleus and by 5 h after serum stimulation G3BP2 is predominantly found in the nucleus. Nine hours after serum stimulation (Figure 1, Panel S), G3BP2 is found at similar levels in the nucleus and the cytoplasm suggesting that some G3BP2 has moved out of the nucleus. A similar localization is seen at 12 h after serum stimulation (Figure 1, Panel T). In summary, G3BP2 localizes to the nucleus and the cytoplasm of proliferating cells in a manner that suggests that G3BP2 shuttles between the nucleus and the cytoplasm as the cell cycle progresses. G3BP is expressed in a range of human cell lines In order to monitor the expression of G3BPs in a range of human cells, the human embryonic kidney cell lines HEK293T and HEK-ER-293 (Stratagene), the SV40 transformed normal breast epithelial cell line SVCT, the breast cancer cell line MDA-MB-435 and the cervical cancer cell line HeLa were lysed, resolved by SDS–PAGE and transferred to a membrane for analysis by Western blotting. The commercial monoclonal G3BP1 antibody (BD Biosciences) was used to assess G3BP1 expression, while the polyclonal G3BP2 antibody (Kennedy et al. 2001) was used to examine the expression of G3BP2 in these cell lines. Previously, total protein lysates of normal adult mouse tissues were examined by Western blot and it was found that the G3BP proteins are expressed in a tissue-specific manner (Kennedy et al. 2001). In addition, G3BP2 is not expressed in normal human breast epithelium. Despite this, as shown in Figure 2, it is apparent that all cell lines (including those derived from normal breast epithelium) express significant levels of both G3BP1, present as a single distinct band, and G3BP2, present as two distinct bands, representing two different isoforms. Discussion Immunohistochemistry of breast cancer sections showed that G3BPs are over-expressed in tumour tissue. G3BP2 is overexpressed to a higher degree than G3BP1, and in fact appears to be absent from all adjacent normal tissue. G3BP1 expression in breast tumour tissue is distinct from that of G3BP2 in that G3BP1 is heterogeneous and in some cases appears to localize more prominently to one side of the cell. This is reminiscent of the Drosophila homologue of G3BP1, Rin, which

ER293 SVCT

435

HeLa

Figure 2. Expression of G3BP in a range of human cell lines. Total protein (50 µg/lane) from human embryonic kidney cell lines HEK293T and HEK-ER-293, the SV40 transformed normal breast epithelial cell line SVCT, the breast cancer cell line MDA-MB-435 and the cervical cancer cell line HeLa (labelled 293T, ER293, SVCT, 435 and HeLa, respectively) were resolved by SDS–PAGE, transferred to a membrane and probed with either the polyclonal G3BP2 antibody or the commercial G3BP1 antibody as indicated. All cells expressed G3BP1 and both isoforms of G3BP2.

localizes apically in photoreceptors in a similar manner to Drk and Sos, factors important in Ras signalling (Pazman et al. 2000). To date, 58 breast tumours have been examined for altered G3BP2 expression. 88% of the breast tumours examined over-express G3BP2 compared to normal breast tissue which does not express G3BP2. The expression of G3BP2 is therefore turned on during tumour progression. No significant relationship was found between G3BP2 over-expression and clinicopathological parameters of breast cancer such histological grade, lymph node or hormone receptor status. These histological results suggest that the inappropriate expression of G3BP2 in breast cancer is an early event in tumour progression. There are many possible reasons for the altered expression of G3BP2 in breast cancer. G3BP2 expression may be deregulated as a result of genetic defects in the G3BP2 gene itself, leading to erroneous feedback control of gene expression. Alternatively, G3BP2 up-regulation in breast cancer may be due to defects in upstream or downstream signalling factors which determine translational control of G3BP2. Conversely, G3BP2 expression in breast tumours may be collateral damage caused by the generalized deregulation of cell signalling that occurs during tumourigenesis. The results presented in this paper also indicate that G3BP2 is not only localized to the cytoplasm of cells but also may be found in the nucleus. This was first observed in the breast tumour samples, where approximately half of all breast tumours examined showed a distinct nuclear localization. To determine whether the nuclear localization of G3BP2 correlates with cell cycle progression, the expression of G3BP2 was examined in proliferating cells. The results presented show that the localization of G3BP2 changes during progression through the cell cycle. During cell quiescence (serum starvation) G3BP2 is predominantly localized to the cytoplasm of NIH 3T3 cells. Upon serum stimulation, when cells

230 move into the cell cycle, G3BP2 translocates from the cytoplasm into the nucleus of cells. G3BP2 was shown predominantly localized to the nucleus five hours after serum stimulation. Further into the cell cycle, G3BP2 was found in both the nucleus and the cytoplasm suggesting that G3BP2 moves back into the cytoplasm. In contrast to the distinct cytoplasmic localization of G3BP2 in serum starved NIH 3T3 cells, it was recently found that G3BP1 is localized to the nucleus of serum starved NIH3T3 rasGAP+/− cells (Tourriere et al. 2001). This suggests the possibility that G3BP1 and 2 nuclear localization are both regulated by the cell cycle, but since G3BP1 and G3BP2 localization patterns are not the same, cell cycle regulation must differ between the two proteins. In summary, the sub-cellular localization of G3BP2 changes during progression through the cell cycle. G3BP2 is predominantly found in the cytoplasm of quiescent cells and appears to shuttle in and out of the nucleus during cell cycle progression in a manner different to that reported for G3BP1. This is consistent with the breast tumour data which shows nuclear localization of G3BP2 in many cases of breast cancer. Future experiments are required to determine the localization of G3BP2 at particular stages of cell cycle in a variety of cell lines. The nuclear localization of G3BP2 in breast cancer was shown to vary significantly. Several tumours show nuclear localization of G3BP2 in only a fraction of cells, suggesting that the nuclear localization corresponds to a particular stage of the cell cycle. In contrast, G3BP2 localizes to the nucleus of almost all cells in some tumours. The significance of this is unclear; however, it is possible that localization or shuttling of G3BP2 is regulated through signalling pathways such as the Ras signalling pathway. The variation in G3BP2 localization may be a result of varying genetic defects, between cases, in upstream or downstream signalling molecules. Alternatively, in these cases, it is conceivable that mutations in G3BP itself lead to the unusual sub-cellular localization. For example, the RNA binding domain of G3BP2 has been shown to retain G3BP2 in the cytoplasm (Prigent et al. 2000). Therefore, if this domain is dysfunctional, in addition to any number of effects, massive nuclear localization of G3BP2 may occur. It is interesting that although G3BP1 and 2 exhibit tissue and cell type specific expression (Kennedy et al. 2001) and G3BP2 is not expressed in normal breast epithelium, both proteins are expressed in all cancer tissue and cell lines tested to date, including human normal breast epithelial cells. These results suggest that G3BP expression, like that of hnRNPB1, a potential biomarker for early detection of human lung carcinoma (Sueoka et al. 1999), may be proliferation related. It is yet to be determined whether G3BP is required for cellular immortalization. This data, along with the fact that G3BP1 is known to promote cell cycle (Guitard et al. 2001), suggest that G3BPs may be necessary for cell proliferation and hence may also be required factors for tumour formation or progression. In summary, the expression of G3BP1 and 2 in human breast cancers was investigated and it was found that G3BP2 was more prominently over-expressed in breast cancer tissues than G3BP1, suggesting that G3BP2 has a prolifera-

J. French et al. tive effect and may be related to breast cancer development or progression. It would be interesting to speculate that the observed up-regulation of G3BP2 in breast cancer may be a response to increased cell cycle in those cancers. Therefore, future experiments need to be performed to determine if up-regulation of G3BP2 can promote cell cycle progression or if G3BP2 is essential for breast cancer progression. Furthermore, evidence was presented suggesting that G3BP2 shuttles between the nucleus and cytoplasm during cell cycle progression. Given that G3BP2 is an RNA-binding protein, it could be suggested that G3BP2 may be acting as an RNA transporter, shuttling transcripts involved in cell cycle control between the nucleus and cytoplasm. This could be a mechanism by which G3BP2 regulates cell cycle progression. This would be consistent with G3BP2 containing an NTF2 domain which could be responsible for the observed nucleocytoplasmic transport (Quimby et al. 2000). It has already been found that G3BP1 promotes S-phase entry in serum-starved cells (Guitard et al. 2001) and is regulated through the Her2 growth factor receptor (Barnes et al. 2002). Since G3BP2 over-expression in breast cancer is more prominent than that of G3BP1, it may be a more appropriate target for future research into breast cancer, Her2 signal transduction and cell cycle progression. Acknowledgements We wish to acknowledge Kelin Ru for technical assistance, J. Alejandro L´opez for his critical review of this article, as well as John Mattick for his patience and support during the course of this project. Pegah Rouhipour deserves credit for her work in preparing some of the breast cancer sections. This work was supported by an NHMRC project grant. Derek Kennedy is an adjunct research fellow at the Institute for Molecular Bioscience which is a federally funded Australian research centre. Ren´ee Stirling is supported by a Queensland Cancer Foundation scholarship. References Almendral JM, Sommer D, Macdonald-Bravo H, Burckhardt J, Perera J, Bravo R (1988) Complexity of the early genetic response to growth factors in mouse fibroblasts. Mol Cell Biol 8: 2140–2148. Barlat I, Maurier F, Duchesne M, Guitard E, Tocque B, Schweighoffer F (1997) A role for Sam68 in cell cycle progression antagonized by a spliced variant within the KH domain. J Biol Chem 272: 3129–3132. Barnes CJ, Li F, Mandal M, Yang Z, Sahin AA, Kumar R (2002) Heregulin induces expression, ATPase activity, and nuclear localization of G3BP, a Ras signaling component, in human breast tumors. Cancer Res 62: 1251–1255. Cross M, Dexter TM (1991) Growth factors in development, transformation, and tumorigenesis. Cell 64: 271–280. Duchesne M, Schweighoffer F, Parker F, Clerc F, Frobert Y, Thang MN, Tocque B (1993) Identification of the SH3 domain of GAP as an essential sequence for Ras-GAP-mediated signaling. Science 259: 525–528. Elston CW, Ellis IO (1990) Pathology and breast screening. Histopathology 16: 109–118. Gallouzi IE, Parker F, Chebli K, Maurier F, Labourier E, Barlat I, Capony JP, Tocque B, Tazi J (1998) A novel phosphorylationdependent RNase activity of GAP-SH3 binding protein: A potential link between signal transduction and RNA stability. Mol Cell Biol 18: 3956–3965.

G3BPs in breast cancer Guitard E, Parker F, Millon R, Abecassis J, Tocque B (2001) G3BP is overexpressed in human tumors and promotes S phase entry. Cancer Lett 162: 213–221. Kenan DJ, Query CC, Keene JD (1991) RNA recognition: Towards identifying determinants of specificity. Trends Biochem Sci 16: 214–220. Kennedy D, Wood SA, Ramsdale T, Tam PP, Steiner KA, Mattick JS (1996) Identification of a mouse orthologue of the human ras-GAPSH3-domain binding protein and structural confirmation that these proteins contain an RNA recognition motif. Biomed Pept Proteins Nucleic Acids 2: 93–99. Kennedy D, French J, Guitard E, Ru K, Tocque B, Mattick J (2001) Characterization of G3BPs: Tissue specific expression, chromosomal localisation and rasGAP(120) binding studies. J Cell Biochem 84: 173–187. Kociok N, Esser P, Unfried K, Parker F, Schraermeyer U, Grisanti S, Toque B, Heimann K (1999) Upregulation of the RAS-GTPase activating protein (GAP)-binding protein (G3BP) in proliferating RPE cells. J Cell Biochem 74: 194–201. Malumbres M, Pellicer A (1998) RAS pathways to cell cycle control and cell transformation. Front Biosci 3: d887–912. McGuckin MA, Walsh MD, Hohn BG, Ward BG, Wright RG (1995) Prognostic significance of MUC1 epithelial mucin expression in breast cancer. Hum Pathol 26: 432–439. Menard S, Tagliabue E, Campiglio M, Pupa SM (2000) Role of HER2 gene overexpression in breast carcinoma. J Cell Physiol 182: 150–162. Parker F, Maurier F, Delumeau I, Duchesne M, Faucher D, Debussche L, Dugue A, Schweighoffer F, Tocque B (1996) A Ras-GTPase-activating protein SH3-domain-binding protein. Mol Cell Biol 16: 2561–2569. Pazman C, Mayes CA, Fanto M, Haynes SR, Mlodzik M (2000) Rasputin, the Drosophila homologue of the RasGAP SH3 binding protein, functions in ras- and Rho-mediated signaling. Development 127: 1715–1725.

231 Prigent M, Barlat I, Langen H, Dargemont C (2000) IkappaBalpha and IkappaBalpha/NF-kappa B complexes are retained in the cytoplasm through interaction with a novel partner, RasGAP SH3-binding protein 2. J Biol Chem 275: 36441–36449. Quimby BB, Lamitina T, L’Hernault SW, Corbett AH (2000) The mechanism of ran import into the nucleus by nuclear transport factor 2. J Biol Chem 275: 28575–28582. Saksela K, Cheng G, Baltimore D (1995) Proline-rich (PxxP) motifs in HIV-1 Nef bind to SH3 domains of a subset of Src kinases and are required for the enhanced growth of Nef+ viruses but not for down-regulation of CD4. EMBO J 14: 484–491. Siomi H, Dreyfuss G (1997) RNA-binding proteins as regulators of gene expression. Curr Opin Genet Dev 7: 345–353. Soncini C, Berdo I, Draetta G (2001) Ras-GAP SH3 domain binding protein (G3BP) is a modulator of USP10, a novel human ubiquitin specific protease. Oncogene 20: 3869–3879. Sueoka E, Goto Y, Sueoka N, Kai Y, Kozu T, Fujiki H (1999) Heterogeneous nuclear ribonucleoprotein B1 as a new marker of early detection for human lung cancers. Cancer Res 59: 1404–1407. Suyama M, Doerks T, Braun IC, Sattler M, Izaurralde E, Bork P (2000) Prediction of structural domains of TAP reveals details of its interaction with p15 and nucleoporins. EMBO Rep 1: 53–58. Tobey RA, Valdez JG, Crissman HA (1988) Synchronization of human diploid fibroblasts at multiple stages of the cell cycle. Exp Cell Res 179: 400–416. Tocque B, Delumeau I, Parker F, Maurier F, Multon MC, Schweighoffer F (1997) Ras-GTPase activating protein (GAP): A putative effector for Ras. Cell Signal 9: 153–158. Tourriere H, Gallouzi IE, Chebli K, Capony JP, Mouaikel J, van der Geer P, Tazi J (2001) RasGAP-associated endoribonuclease G3BP: Selective RNA degradation and phosphorylation-dependent localization. Mol Cell Biol 21: 7747–7760.