Glypican-3 expression is silenced in human breast cancer

Abstract

Glypican-3 (GPC3) is a membrane-bound heparan sulfate proteoglycan that is mutated in the Simpson-Golabi-Behmel syndrome. This is an X-linked condition characterized by overgrowth, and various visceral and skeletal dysmorphisms. The phenotype of the Simpson-Golabi-Behmel syndrome patients and GPC3-deficient mice, as well as gene transfection experiments indicate that GPC3 can act as an inhibitor of cell proliferation and survival. It has been previously shown that GPC3 expression is downregulated in mesotheliomas and ovarian cancer. Here we report that GPC3 expression is also silenced in human breast cancer, and that this silencing is due, at least in part, to hypermethylation of the GPC3 promoter. Ectopic expression of GPC3 inhibited growth in eight out of 10 breast cancer cell lines. Collectively, these data suggest that GPC3 can act as a negative regulator of breast cancer growth.

Main

Glypicans are a family of heparan sulfate proteoglycans that are attached to the cell surface by a glycosyl-phosphatidylinositol anchor (Bernfield et al., 1999; Filmus et al., 2000; Filmus, 2001). Glypican-3 (GPC3), a member of the glypican family, is mutated in patients with the Simpson-Golabi-Behmel syndrome (SGBS) (Pilia et al., 1996). This is an X-linked condition characterized by pre- and postnatal overgrowth, and a broad spectrum of clinical manifestations that vary from a mild phenotype in carrier females to infantile lethal forms in some males (Behmel et al., 1984; Garganta et al., 1992; Golabi et al., 1984; Neri et al., 1998). The list of abnormalities observed in SGBS patients can include a distinct facial appearance, and malformations involving craniofacial development (macroglossia, cleft lip/palate, dental malocclusion), chest/skeleton (polydactyly, syndactyly, supernumerary nipples, supernumerary ribs), genitalia (hypospadias, cryptorchidism), and internal organs (cardiac defects, diaphragmatic hernias, renal cystic dysplasia) (Garganta et al., 1992; Hughes-Benzie et al., 1996; Neri et al., 1998). An increased risk for the development of pediatric tumors has also been reported (Hughes-Benzie et al., 1992).

The phenotype of SGBS patients and of GPC3-deficient mice (Cano-Gauci et al., 1999) suggests that during development GPC3 can act as an inhibitor of cell proliferation, and as an inducer of apoptosis in specific tissues. Experimental evidence has been provided supporting such a role for GPC3 in various cell lines (Duenas Gonzales et al., 1998).

Currently, the mechanism by which GPC3 regulates cell proliferation and survival is not clear. It has been proposed that GPC3 can act as a negative regulator of insulin-like growth factor II (Pilia et al., 1996), but this remains to be proven. There is also genetic evidence that, at least in certain tissues, GPC3 can regulate bone morphogenetic factors (Grisaru et al., 2001; Paine-Saunders et al., 2000), but it seems unlikely that this regulation is the direct cause of the overgrowth observed in GPC3 patients.

It has been recently reported that GPC3 expression is downregulated in mesotheliomas and ovarian cancer (Lin et al., 1999; Murthy et al., 2000). This finding is consistent with the fact that GPC3 can act as a negative regulator of cell proliferation and survival, since cancer cells tend to have an increased proliferation rate, and a reduced sensitivity to apoptosis.

The study on ovarian cancer demonstrated frequent silencing of GPC3 expression in cell lines (Lin et al., 1999). In all the cases in which GPC3 expression was inhibited, the GPC3 promoter was hypermethylated, and expression was restored by treatment with a demethylating agent. The study on mesotheliomas reported similar results. In this case the silencing of GPC3 expression was also demonstrated in primary tumors (Murthy et al., 2000). No mutations in the coding region of the GPC3 gene were found either in the ovarian tumors or in the mesotheliomas. In this regard it is important to note that since GPC3 is an X-linked gene, the silencing of a single allele is enough for complete inhibition of expression (Huber et al., 1999).

In the adult, GPC3 is expressed only in a few tissues including mesothelia, and the ovarian and mammary epithelia (Filmus, unpublished results). We decided, therefore, to investigate whether, like in ovarian cancer and mesothelioma, GPC3 expression is also downregulated in mammary cancer.

To assess GPC3 expression in normal mammary epithelium and in mammary cancer, we performed in situ hybridization in tissue sections that contained both normal and malignant tissue. In all of the 12 patients that were studied GPC3 expression was lower in the tumor cells than in the adjacent normal epithelium. Moreover, in seven of these patients GPC3 expression was undetectable in the tumor cells. Figure 1 shows the results obtained with three typical patients, including a case with cancer in situ. The number of samples studied was insufficient to establish a correlation between the degree of downregulation of GPC3 expression and tumor stage.

Figure 1
figure1

Analysis of GPC3 expression in normal human mammary gland and breast cancer by in situ hybridization. (a,b) Normal mammary gland on the right, and in situ ductal carcinoma on the left. (c,d,e,f,g) Invasive ductal carcinoma. A normal duct containing carcinoma in situ can also be observed (arrow) in e, f and g. (b,d,f) Tissue sections were hybridized with GPC3 anti-sense or (g) sense RNA probes. (a,c,e) Parallel sections were stained with hematoxylin-eosin. GPC3 sense and anti-sense probes were prepared from Bluescript KS-plasmids (Stratagene) containing the 594 bp PstI and HindIII fragment of the human GPC3 cDNA in the sense and anti-sense orientations. The RNA probes were synthesized by in vitro transcription in the presence of digoxigenin-labeled UTP (Roche) with the use of T3 and T7 RNA polymerase (Promega) following the protocol recommended by the company. In situ hybridization was performed as described previously (Xiang et al., 1994). GPC3 expression was detected in normal mammary epithelial cells in (b, right side), and (f, arrow), whereas tumor cells are negative (d,f) or weakly positive in non-invasive cancer (b, left side). Scattered positive cells within the tissue sections are infiltrating lymphocytes

Next, we investigated whether GPC3 expression is downregulated in breast cancer as a result of promoter hypermethylation, as it was described for ovarian cancer and mesotheliomas (Lin et al., 1999; Murthy et al., 2000). To select the appropriate cell lines to perform this study, we assessed GPC3 expression in 10 breast cancer cell lines by Northern blot analysis. Since GPC3 expression was undetectable in all of them (data not shown), we decided to perform RT–PCR, a more sensitive method. This analysis showed that GPC3 expression can be detected in five of the 10 cell lines tested (Figure 2a).

Figure 2
figure2

Expression and DNA methylation analysis of GPC3. (a) RT–PCR analysis of GPC3 expression. One μg of total RNA isolated from the hepatocarcinoma cell line HepG2 (GPC3 positive control), and the breast carcinoma cell lines was reverse transcribed using SuperScriptTM RNase H (GIBCO BRL), and amplified by using primers specific for GPC3 (upstream: 5′-CCAACATGCTGCTCAAGAAAGATGGAAG-3′, downstream: 5′CAAACTCAAAAGCTTGTGGAGTCAGGCT-3′), and for β-actin (upstream: 5′-TGACGGGGTCACCCACACTGTGCCCATCTA-3′, downstream: 5′-CTAGAAGCATTTGCGGTGGACGATGGAGGG-3′). PCR was performed for 35 cycles (denaturation at 94°C, 1 min; annealing at 52°C, 1 min; elongation at 72°C, 1 min; with a final extension cycle at 72°C, 5 min). β-actin RT–PCR product provides a control for the amount of intact RNA used in the reactions (bottom). The expected amplification products for GPC3 (226 bp) and β-actin (661 bp) are indicated. (b) Schematic drawing of the 5′ region of the GPC3 gene encompassing part of the promoter, the first exon, and part of the first intron. The location of the 247 bp fragment used as a probe for Southern blot analysis, and the size of the expected DNA fragments are indicated. (c) DNA methylation analysis. Twenty μg of genomic DNA were digested with EcoRI (E), or with EcoRI and EagI (EE). Digested DNA was analysed by Southern blot using the GPC3 probe shown in (b). The expected 8.0-kb and 4.2-kb fragments corresponding to the methylated and non-methylated states of the GPC3 promoter, respectively, are indicated

To assess the methylation status of the GPC3 promoter in the breast cancer cell lines, their genomic DNA was digested with EcoRI alone, or with EcoRI and EagI. Both enzymes cut within the CpG island located in the GPC3 promoter region, and EagI is methylation-sensitive (Figure 2b). It has been previously reported that the methylation of the EagI site of the GPC3 promoter correlates well with expression (Lin et al., 1999). The sensitivity of the genomic DNA to digestion with EcoRI, and a combination of EcoRI and EagI was assessed by Southern blot analysis, using as probe a DNA fragment that can detect both possible products of the digestion (Figure 2b). As shown in Figure 2c, three cell lines (MDA-MB-231, MDA-MB-435, and Hs578T) displayed complete methylation of the EagI site. As expected, GPC3 was not expressed in these cell lines (Figure 2a). On the other hand, the five cell lines that expressed GPC3 displayed significant amounts of a 4.2 Kb band, which is the expected size when the genomic DNA is not methylated at the EagI site (Lin et al., 1999). In the case of MDA-MB-468 and BT-549, DNA digestion also generated a 4.2 Kb band despite the fact that these cell lines do not show any detectable GPC3 expression. To investigate the possibility that in these two cell lines other sites within the GPC3 promoter have been subjected to methylation, we performed Southern blot analysis using BssHII, another methylation-sensitive enzyme that demonstrated an aberrant methylation pattern in the GPC3 promoter of mesothelioma cell lines (Murthy et al., 2000). This analysis showed that both MDA-MB-468 and BT-549 also display aberrant methylation of the GPC3 promoter (data not shown).

Since GPC3 is located on the X chromosome, in females one of the two alleles is inactivated by methylation in normal somatic tissues (Huber et al., 1999). Thus, one would expect to see in all breast cancer cell lines the presence of the 8 Kb band corresponding to the methylated GPC3 allele. It is interesting to note, however, that in two cell lines (MDA-MB-157 and T47D) this band was not detected. It is possible, therefore, that in these tumors the inactive allele has been lost during progression of the disease.

To determine whether the methylation of the GPC3 promoter can play a role in the silencing of GPC3 expression, we treated the GPC3-negative cell lines that showed complete methylation of the EagI site (MDA-MB-231 and MDA-MB-435) with the demethylating agent 5-aza-2′-deoxycytidine. As shown in Figure 3b, such treatment was able to restore GPC3 expression in both cell lines. This was accompanied by demethylation of the EagI site, as determined by Southern blot analysis (Figure 3a). Treatment with 5-aza-2′-deoxycitidine also restored GPC3 expression in the MDA-MB-468 and BT-549 cell lines, which displayed an abnormal pattern of methylation after digestion with BssHII (data not shown). We conclude therefore that, like in ovarian cancer and mesotheliomas, promoter hypermethylation is at least one of the mechanisms by which GPC3 expression is silenced in breast cancer.

Figure 3
figure3

Reexpression of GPC3 after demethylation. Tumor cell lines MDA-MB-231 and MDA-MB-435 were treated with 5 μM 5-aza-2′-deoxycitidine (Sigma) for 4 days. The medium was replenished every day. (a) The effect of the treatment on the methylation status of the EagI site was assessed by Southern blot as described in Figure 2. (b) GPC3 expression before and after treatment was analysed by RT–PCR as described in Figure 2

By using a colony-forming efficiency assay we have previously shown that GPC3 can inhibit growth in the breast cancer cell line MCF-7 (Duenas Gonzales et al., 1998). To investigate whether this growth-inhibitory activity is a general feature of breast cancer we performed the colony-forming efficiency assay in all the cell lines included in this study. When compared with vector control, an expression vector containing GPC3 significantly inhibited colony formation in eight out of the 10 cell lines (Table 1). Thus, although the growth-inhibitory effect of GPC3 is not restricted to MCF-7 cells, it is not a general feature of all breast cancer cells. It is well established that tumor cell lines carry numerous mutations, and that some of these mutations can affect signaling pathways of specific growth factors, inducing growth-factor independence. We speculate that the lack of growth-inhibitory activity of GPC3 in MDA-MB-231 and MDA-MB-435 indicates that the particular growth factor signaling system with which GPC3 interacts is not functioning properly in these specific cell lines.

Table 1 Effect of GPC3 in colony-forming efficiency

In conclusion, we have shown here that GPC3 expression is silenced in breast cancer, and that in most cases this silencing is due, at least in part, to promoter hypermethylation. Since ectopic expression of GPC3 can inhibit cell growth of a significant proportion of breast cancer cell lines, collectively our results suggest that GPC3 can be considered a negative regulator for breast cancer growth.

References

  1. Behmel A, Plochl E, Rosenkranz W . 1984 Hum. Genet. 67: 409–413

  2. Bernfield M, Gotte M, Park PW, Reizes O, Fitzgerald ML, Lincecum J, Zako M . 1999 Ann. Rev. Biochem. 68: 729–777

  3. Cano-Gauci DF, Song H, Yang H, McKerlie C, Choo B, Shi W, Pullano R, Piscione TD, Grisaru S, Soon S, Sedlackova L, Tanswell AK, Mak TW, Yeger H, Lockwood GA, Rosenblum N, Filmus J . 1999 J. Cell Biol. 146: 255–264

  4. Duenas Gonzales A, Kaya M, Shi W, Song H, Testa JR, Penn LZ, Filmus J . 1998 J. Cell. Biol. 141: 1407–1414

  5. Filmus J . 2001 Glycobiol. 11: 19R–23R

  6. Filmus J, Song HH . 2000 Proteoglycans. Iozzo, RV (ed) Marcel Dekker: New York pp. 161–176

  7. Garganta CL, Bodurtha JN . 1992 Am. J. Med. Genet. 44: 129–135

    CAS  Article  Google Scholar 

  8. Golabi M, Rosen L . 1984 Am. J. Med. Genet. 17: 345–358

    CAS  Article  Google Scholar 

  9. Grisaru S, Cano-Gauci D, Tee J, Filmus J, Rosenblum ND . 2001 Dev. Biol. 231: 31–46

  10. Huber R, Hansen RS, Strazzullo M, Pengue G, Mazzarella R, D'urso M, Schlessinger D, Pilia G, Gartler SM, D'Esposito M . 1999 Proc. Natl. Acad. Sci. USA 96: 616–621

  11. Hughes-Benzie RM, Hunter AGW, Allanson JE, Mackenzie AE . 1992 Am. J. Med. Genet. 43: 428–435

    CAS  Article  Google Scholar 

  12. Hughes-Benzie RM, Pilia G, Xuan JY, Hunter AGW, Chen E, Golabi M, Hurst JA, Kobori J, Marymee K, Pagon RA, Punnett HH, Schelley S, Tolmie JL, Wohlferd MM, Grossman T, Schlessinger D, Mackenzie AE . 1996 Am. J. Med. Genet. 66: 227–234

    CAS  Article  Google Scholar 

  13. Lin H, Huber R, Schlessinger D, Morin PJ . 1999 Cancer Res. 59: 807–810

  14. Murthy SS, Shen T, De Rienzo A, Lee WC, Ferriola PC, Jhanwar SC, Mossman BT, Filmus J, Testa JR . 2000 Oncogene 19: 410–416

    CAS  Article  Google Scholar 

  15. Neri G, Gurrieri F, Zanni G, Lin A . 1998 Am. J. Med. Genet. 79: 279–283

    CAS  Article  Google Scholar 

  16. Paine-Saunders S, Viviano BL, Zupicich J, Skarnes WC, Saunders S . 2000 Dev. Biol. 225: 179–187

  17. Pilia G, Hughes-Benzie RM, MacKenzie A, Baybayan P, Chen EY, Huber R, Neri G, Cao A, Forabosco A, Schlessinger D . 1996 Nature Genet. 12: 241–247

  18. Xiang YY, Tanaka M, Suzuki M, Igarashi H, Kiyokawa E, Naito Y, Ohtawara Y, Shen Q, Sugimura H, Kino I . 1994 Cancer Res. 54: 3034–3041

Download references

Acknowledgements

We thank Penny Papadakos for assistance in the preparation of this manuscript. This work has been supported by the Canadian Breast Cancer Research Initiative, the Canadian Institutes for Health Research and the Sunnybrook Trust for Medical Research.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Jorge Filmus.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Xiang, Y., Ladeda, V. & Filmus, J. Glypican-3 expression is silenced in human breast cancer. Oncogene 20, 7408–7412 (2001). https://doi.org/10.1038/sj.onc.1204925

Download citation

Keywords

  • glypicans
  • breast cancer
  • DNA methylation

Further reading

Search