High tumoral maspin expression is associated with improved ... - Nature

ã

Oncogene (2000) 19, 2398 ± 2403 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc

High tumoral maspin expression is associated with improved survival of patients with oral squamous cell carcinoma Weiya Xia1,6, Yiu-Keung Lau1,6, Mickey C-T Hu1, Lei Li4, Dennis A Johnston2, Shi-jie Sheng5, AK El-Naggar3 and Mien-Chie Hung*,1 1

Department of Molecular and Cellular Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas, TX 77030, USA; 2Department of Biomathematics, The University of Texas MD Anderson Cancer Center, Houston, Texas TX 77030, USA; 3Department of Pathology, The University of Texas MD Anderson Cancer Center, Houston, Texas TX 77030, USA; 4 Department of Oral Pathology, Ninth People's Hospital, Shanghai Second Medical University, People's Republic of China; 5 Department of Pathology, Wayne State University, 9207 Scott Hall, Detroit, Michigan, MI 48201, USA

Maspin, a member of the serpin family of protease inhibitors, is known to have tumor-suppressor functions. However, the association between its expression level and survival has not been demonstrated in human cancer. Using the immunohistochemical technique to examine the expression levels of maspin in 44 cases of oral squamous cell carcinoma (SCC), we found that 66% of the cases expressed low to intermediate levels of maspin and 34% of the cases expressed high levels of maspin. We further examined maspin protein expression in a series of six SCC cell lines from the head and neck, and found that all but one expressed low or no maspin protein. We also compared the clinicopathological features of the oral SCC cases with the maspin expression level, and found that high maspin expression was associated with the absence of lymph node metastasis. More importantly, we showed that higher maspin expression was signi®cantly associated with better rates of overall survival, suggesting that high maspin expression may be a favorable prognostic marker for oral SCC. Oncogene (2000) 19, 2398 ± 2403. Keywords: cancer; maspin; protease inhibitor; tumor suppressor Introduction Maspin is a 42 kDa cytoplasmic protein that belongs to the serpin family of protease inhibitors. It is expressed in normal human mammary epithelial cells (Zou et al., 1994) and associated with secretory vesicles and cellular surface (Pemberton et al., 1997). It has also been shown that maspin has tumor suppression activity (Sager et al., 1996; Sheng et al., 1994). Transfecting human mammary carcinoma cells with the maspin gene reduced tumor induction and metastasis in nude mice as well as in vitro invasion of basement membrane (Zou et al., 1994). In primary breast cancer cells, maspin is down-regulated and its expression is inhibited in metastasis. It has been suggested that this decrease in maspin expression may

*Correspondence: M-C Hung, Box 108, Department of Molecular and Cellular Oncology, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, USA 6 The ®rst two authors contributed equally to this work Received 8 January 2000; revised 2 February 2000; accepted 22 February 2000

be due to the absence of transactivation through the Ets and Ap1 elements in the promoter of maspin (Zhang et al., 1997a). Another report showed that expression of maspin can also be repressed by a negative hormonal response element in the prostate cells (Zhang et al., 1997b). Therefore the reduction in maspin expression in cancer cells may be triggered by transcriptional down regulation. However, it is not clearly de®ned whether maspin expression could provide any prognostic value for cancer patients. In addition, in certain prostate cancer cell lines, maspin is found to have an isoleucine to valine mutation at amino acid 319 (Umekita et al., 1997). However, further studies are required to examine whether mutation of maspin gene may also play a role in the dysregulation of its function. The mechanism underlying the function of maspin as a tumor suppressor is not fully understood. Maspin seems to act on the cell membrane to inhibit cell motility and invasion (Sheng et al., 1996) and it has been found in many human organs at the epithelia (Pemberton et al., 1997). A single-chain tissue plasminogen activator (tPA) speci®cally interacts with the maspin reactive site loop peptide and forms a stable complex with the recombinant maspin, suggesting that maspin may have an inhibitory eect on tPA (Sheng et al., 1998). Although loss or downregulation of maspin has been implicated in tumor progression, a correlation between maspin expression and survival in human cancers has not been reported thus far. The molecular pathogenesis of oral squamous cell carcinoma (SCC) is still not clear. Several reports have demonstrated a high frequency of p53 gene mutations in head and neck SCC (Ahomadegbe et al., 1995; Somers et al., 1992; Watling et al., 1992). The presence of other oncogenes, such as EGFR (Yamamoto et al., 1986), cmyc (Field et al., 1989), and bcl-1 (Berenson et al., 1989) has also been reported in head and neck SCC. We have also recently shown that HER-2/neu overexpression is associated with poor survival in patients with oral SCC (Xia et al., 1997). Further understanding of the molecular alteration in oral SCC will enable us to provide more accurate and useful prognostic markers and more eective treatments. The purpose of this study was to examine the expression of maspin in oral tumor specimens and its relationship with the maspin expression and survival in patients with oral SCC. We found that decreases in maspin expression were frequent in oral SCC, and high expression of maspin was signi®cantly associated with better survival.

Expression of maspin in oral SCC W Xia et al

2399

Results Expression of maspin in oral SCC To examine the expression pattern of maspin in the head and neck SCC, we analysed the protein levels of maspin by Western blot analysis in a series of six head and neck SCC cell lines, including three oral SCC (Tu 138, 686/LN-1, and Tu 167). All of the cell lines tested, including all three oral SCC cells expressed low to no maspin as compared with the normal human oral keratinocytes (Figure 1A,B). We then examined the mRNA levels of maspin in these cell lines by Northern blot analyses. We found that most of the cell lines express two forms of maspin mRNA (Figure 1C). Two dierent maspin mRNA transcripts were also previously reported in human myoepithelial cells (Sternlicht et al., 1997). Interestingly, we noticed that the lower band of maspin mRNA is correlated better with the maspin protein level detected by Western blotting analysis (Figure 1A,B). Although further characterization is required, the results seem to suggest that the lower band of maspin mRNA may be the functional mRNA and the upper band could be a less translatable or precursor form of maspin mRNA. The expression of maspin in the 44 oral SCC patients' specimens was detected by using immunohistochemical technique. Representative examples are shown in Figure 2. A normal oral epithelial tissue taken from gingiva was used as a positive control (Figure 2D). Of 44 patient specimens examined, 29 (66%) expressed low or intermediate levels of maspin. The tumor cells in both the low and intermediate subgroups showed much weaker staining than those of the normal tissues (control). The remaining 15 (34%) expressed high levels of maspin and the tumor cells stained as strongly as the normal oral epithelial cells (control). To determine whether the expression levels of maspin detected by immunohistochemical stainings are parallel with the mRNA levels in the oral specimens, we examined fresh oral tumor and normal tissues obtained from MD Anderson Cancer Center. As shown in Figure 3A,B, mRNA levels of maspin in the normal tissues were signi®cantly higher than those of the tumor tissues. These results were correlated well with the expression levels as detected by immunohistochemistry (Figure 3C). Taken together, the results of Northern blot analyses were consistent with those of immunohistochemical stainings. Higher maspin expression is associated with the absence of lymph node metastasis We further examined the association of maspin with other clinicopathological features, i.e. patient age, sex, histological grade, TNM stage, and postsurgical treatment, in the 44 oral SCC patients. A signi®cant positive association with the presence of lymph node involvement (P=0.009) was found in the patients with low to intermediate maspin expression, but not with the other factors (Table 1). Furthermore, all of the patients who expressed high levels of maspin did not show any nodal involvement, suggesting that higher maspin expression is associated with the absence of lymph node metastasis.

Figure 1 (A) Western blot analysis of maspin expression in six head and neck SCC cell lines. A normal oral keratinocyte cell line was used as a positive control for maspin protein detection. The same membrane was stripped and reprobed with anti-actin antibody to show equal protein loading. (B) The immunoreactivity level of maspin in each cell line was quantitated as mean optical density (MOD) (Xia et al., 1997). (C) Northern blot analysis of maspin mRNA expression in these cell lines. The gel was stained with ethidium bromide before blotting to show the 28S and 18S ribosomal RNA as loading controls (bottom panel)

Higher maspin expression is associated with better survival Since maspin functions as a tumor-suppressor, we examined whether higher maspin expression is associated with better overall survival in patients with oral SCC in our series of 44 cases. As shown in Figure 4, the patients with high maspin expression had better survival rates than those who had low and intermediate levels of maspin expression (P=0.017). These results suggest that high maspin expression in oral SCC may be a marker of favorable prognosis. Oncogene


2400

Figure 2 Representative immunohistochemical stainings for expression of maspin protein in oral SCC. (A) high expression; (B) intermediate expression; (C) low expression; (D) normal oral tissue from the epidermis-dermis junction (dermis at lower left)

Discussion Functional studies have demonstrated that maspin is a tumor suppressor gene that can inhibit in vitro tumor invasion and motility (Sager et al., 1996; Sheng et al., 1994), tumor growth, and metastasis in nude mice (Zou et al., 1994). However, no clinical data have been reported regarding possible association between the expression level of maspin and the survival rate. We provide here the ®rst evidence that higher maspin expression is signi®cantly associated with better survival in oral SCC patients (P=0.017). These results support the concept that maspin is a tumor suppressor in oral SCC. Further analyses of maspin expression and the clinicopathological features in our series of 44 oral SCC cases showed that high maspin expression was associated with the absence of nodal metastasis (P=0.009). However, we did not ®nd any association between maspin protein level and distant metastasis. Tumor metastasis is a multistep process that includes local invasion, angiogenesis, disruption of adhesion to the neighboring cells and the extracellular matrix, adhesion to and transgression of endothelial cells to access into and out of the vascular circulation, attachment and proliferation at distant sites, adhesion to the extracellular matrix, and ®nally neovascularization at the distant site. Maspin has been shown to inhibit cell invasiveness and motility (Sheng et al., 1996). Our data seem to support the notion that maspin is involved in inhibiting the initial step in the metastasis of oral SCC. Oncogene

We also found that decreases in maspin expression are frequent in the established tumor cell lines as well as in human tissue specimens from patients with oral SCC. Lower levels of maspin expression in the tumor samples were associated with poor survival. The signi®cant inverse correlation between maspin expression level and lymph node involvement is consistent with the idea that loss of maspin expression may contribute to the progression of cancer. In summary, we have provided the ®rst evidence that high maspin expression is correlated with a favorable prognosis in oral SCC and may be useful as a clinical marker in the management of this disease.

Materials and methods Patients and specimens Forty-four specimens of primary oral SCC were obtained from the Department of Oral Pathology, Ninth People's Hospital, Shanghai Second Medical University, People's Republic of China. The patients from whom these samples were excised (29 men and 15 women) underwent surgical treatment in that hospital between 1989 and 1990. Their ages ranged from 32 ± 74 years (median 50). The primary tumors were graded according to the WHO classi®cation as described previously (Xia et al., 1997). Of the 44 cases, 34 were de®ned as grade I, and 10 as grade II. The tumor tissues included tongue (25 cases), cheek (two cases), ¯oor of the mouth (six cases), gum (®ve cases), lip mucosa (two cases), palate (two cases), mucosa of mandibular molar area (one case), and mandible (one case). The specimens were routinely ®xed in


C

2401

Figure 3 Expression of mRNA level of maspin is consistent with immunohistochemical staining in normal and tumor samples. (A) The mRNA level of maspin was detected by Northern blot analyses. The levels of GDPH were examined as loading controls (lower panel). (B) The immunoreactivity level of maspin in each tissue was quantitated as mean optical density (MOD). (C) The representative pictures of the immunostainings of the normal (upper panel) and tumor (bottom panel) samples are shown

formalin, and embedded in paran. After surgical treatment, these 44 patients were followed up in 3 ± 6 months either at the University Hospital clinic or local clinics if the patients were not from the city. If the patients were found to have recurrent disease in the local clinics, they would be referred back to the University Hospital. Furthermore, a questionnaire was sent to the patients to obtain information including general well-being, whether they have gone to the follow-up visit, what kind of treatment they received in their local clinics after the operation, and whether the patients were still alive. These patients were followed up for a period of more than 6 years. Two patients received chemotherapy and two patients had radiation therapy after the surgery. There are four missing data and the remaining 36 patients did not receive any post-surgical treatment. The median survival rate for these 44 patients are 61 months (ranged from 6 ± 73 months). A total of 11 patients died during the study period. Ten of the patients has died of primary tumors and one died of lung metastasis of the primary tumor. Immunohistochemical staining The immunoperoxidase staining method used was a modi®cation of the avidin-biotin complex technique as described previously (Xia et al., 1997). Brie¯y, tissue sections were

deparanized and dehydrated in a graded series of alcohol. They were then digested in 0.05% trypsin for 15 min, blocked in 0.3% hydrogen peroxide in methanol for 15 min, and treated with 1% (v/v) normal goat serum for 30 min. The slides were incubated for 3 h at room temperature with the puri®ed AbS4A polyclonal antibody against maspin (10 mg/ ml) (Sheng et al., 1996). After extensive washing with PBS, the slides were incubated for another 30 min at room temperature with the biotinylated goat anti-rabbit IgG antibody (diluted 1 : 200 in PBS). The slides were subsequently incubated for 60 min at room temperature with an avidin-biotin-peroxidase complex (diluted 1 : 100 in PBS). The peroxidase-catalyzed product was visualized with 0.125% aminoethylcarbazole chromogen stock solution (Sigma Chemical Co.). Between steps, the slides were rinsed for 2 min in PBS (pH 7.6) three times. After light counterstaining with the Mayer's modi®ed hematoxylin (Sigma Chemical Co.), the slides were mounted. Negative controls, in which PBS was used instead of the primary antibody, were run with each batch of slides stained. A section of normal oral epithelia previously identi®ed as strong staining was used as a positive control. Those tumor cells that were immunostained with red granules were considered as positive, and those cells without any immunostaining were considered as negative. Cytoplasmic, but not membrane, staining was observed. The Oncogene


2402

Table 1

Clinicopathological features of oral squamous cell carcinoma and maspin immunostaining results Maspin expression level Low and intermediatea High P valueb

Patients' sex Men Women

20 9

9 6

0.74

Age 550 years old 550 years old

11 18

10 5

0.11

Histological grade I II

21 8

13 2

0.45

Tumor size T1 ± 2 T3

27 2

14 1

1.0

Nodal stage No Yes

19 10

15 0

0.009*

Distant metastasis No Yes

21 8

13 2

0.45

Post-surgical therapyc No Yes

22 4

14 0

0.28

a

Patients with low and intermediate level (lower than normal level) were grouped together to perform a valid Fisher's exact test. bP value was obtained by Fisher's exact test. cOf the 44 patients, two received chemotherapy, and two had radiotherapy after surgery, four had missing data, and the remaining 36 did not receive any post surgical chemo- or radio-therapy

and Tu 167), have been described previously (Xia et al., 1997). The human oral keratinocytes were used as a positive control. The head and neck SCC cell lines were grown in Dulbecco's modi®ed Eagle's medium/Ham's F12 (DMEM/ F12) (Life Technologies, Inc., Grand Island, NY, USA), supplemented with 10% fetal bovine serum (FBS) and antibiotics. The human oral keratinocytes were maintained in DMEM/F12 supplemented with 10% FBS, 400 ng/ml hydrocortisone, cholera toxin 0.1 nM, epidermal growth factor 10 ng/ml, adenine 0.089 mM, insulin 5 ng/ml, and penicillin/streptomycin 100 IU/ml. Cells were maintained in a humidi®ed incubator at 378C under 5% CO2 in air. Western blot analysis Western blot analyses were performed to detect the maspin expression in the head and neck SCC cell lines. Brie¯y, proteins were extracted by RIPA lysis buer (20 mM Na2PO4 pH 7.4; 150 mM NaCl; 1% Triton X-100; 1% aprotinin; 1 mM phenylmethylsulfonyl ¯uoride; 100 mM NaF; and 2 mM Na3VO4). The protein concentration was determined by using the Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA, USA). Equal amounts (75 mg) of cell lysates were subjected to SDS-polyacrylamide gel electrophoresis, and transferred to a nitrocellulose membrane. The membrane was blocked for 1 h with 5% non-fat dry milk in PBS(PBS/Tween 20)(0.1%,v/v), and then incubated with the puri®ed polyclonal antibody against maspin (1 mg/ml) or an anti-actin antibody for loading control at 48C overnight. The membranes were then incubated with the HRP-conjugated goat-anti-rabbit IgG secondary antibody (1 : 5000 dilution) or goat anti-mouse IgC secondary antibody (1 : 5000 dilution) (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for about 30 min at room temperature, followed by the detection with the enhanced chemiluminescence (ECL) system (Amershan Corp., Arlington Heights, IL, USA). Northern blot analyses Total RNAs were isolated from the cells or tissues using the Trizol reagent (Life Technologies, Gaithersburg, MD, USA) or the Micro RNA Isolation Kit (Stratagene) according to the manufacturer's recommended protocols. Equal amounts (25 mg) of total RNAs were sized-fractionated on a 1% agarose gel containing formaldehyde. The RNAs were then transferred to a nylon membrane and hybridized with the 32Plabeled probe of 1 kb fragment of maspin cDNA using the standard Northern blot protocols. Statistical analyses

Figure 4 Survival curves for 44 patients expressing low and intermediate maspin levels vs high maspin level (log rank test, P=0.017)

immunoreactivity of maspin was ranked into three groups according to the percentage and staining intensity of the positive tumor cells: high (550% cells stained), or intermediate (20 ± 49% cells stained), or low (up to 20% cells stained). In general, the percentage of the stained cells was correlated with the intensity of staining. Cell culture The six head and neck squamous cell lines examined, including three oral squamous cell lines (Tu 138, 686/LN-1, Oncogene

Fisher's exact test was used to analyse the association of maspin expression with clinicopathological factors such as age, sex, histological grade, TNM stage, and subsequent treatment after surgery. The patients with low to intermediate maspin expression (i.e. both lower than the normal level) grouped together in order to perform a valid Fisher's exact test. Survival curves (high vs low and intermediate) were calculated by Kaplan ± Meier product limit estimate. The logrank method was used to analyse dierences in the survival time. The statistical analyses were performed by using SPSS software. A P value of less than 0.05 was set as the criterion for statistical signi®cance.

Acknowledgments We thank Dr Douglas Boyd for generously providing the human keratinocyte cell line. This study was supported in part by MD Anderson Faculty Achievement Award (to MC Hung), and MD Anderson Nellie Connally Breast Cancer Research Fund (to M-C Hung).


2403

References Ahomadegbe JC, Barrois M, Fogel S, Le Bihan ML, DoucRasy S, Duvillard P, Armand JP and Riou G. (1995). Oncogene, 10, 1217 ± 1227. Berenson JR, Yang J and Mickel RA. (1989). Oncogene, 4, 1111 ± 1116. Field JK, Spandidos DA, Stell PM, Vaughan ED, Evan GI and Moore JP. (1989). Oncogene, 4, 1463 ± 1468. Pemberton PA, Tipton AR, Pavlo N, Smith J, Erickson JR, Mouchabeck ZM and Kiefer MC. (1997). J. Histochem. Cytochem., 45, 1697 ± 1706. Sager R, Sheng S, Pemberton P and Hendrix MJ. (1996). Curr. Topics Microbiol. Immunol., 213, 51 ± 64. Sheng S, Carey J, Seftor EA, Dias L, Hendrix MJ and Sager R. (1996). Proc. Natl. Acad. Sci. USA, 93, 11669 ± 11674. Sheng S, Pemberton PA and Sager R. (1994). J. Biol. Chem., 269, 30988 ± 30993. Sheng S, Truong B, Fredrickson D, Wu R, Pardee AB and Sager R. (1998). Proc. Natl. Acad. Sci. USA, 95, 499 ± 504. Somers KD, Merrick MA, Lopez ME, Incognito LS, Schechter GL and Casey G. (1992). Cancer Res., 52, 5997 ± 6000.

Sternlicht MD, Kedeshian P, Shao ZM, Safarians S and Barsky SH. (1997). Clin. Cancer Res., 3, 1949 ± 1958. Umekita Y, Hiipakka RA and Liao S. (1997). Cancer Lett., 113, 87 ± 93. Watling DL, Gown AM and Coltrera MD. (1992). Head & Neck, 14, 437 ± 444. Xia W, Lau Y-K, Zhang H-Z, Liu A-R, Kiyokawa N, Clayman GL, Katz RL and Hung M-C. (1997). Clin. Cancer Res., 3, 3 ± 9. Yamamoto T, Kamata N, Kawano H, Shimizu S, Kuroki T, Toyoshima K, Rikimaru K, Nomura N, Ishizaki R, Pastan I, Gamou S and Shimizu N. (1986). Cancer Res., 46, 414 ± 416. Zhang M, Maass N, Magit D and Sager R. (1997a). Cell Growth. Dier., 8, 179 ± 186. Zhang M, Magit D and Sager R. (1997b). Proc. Natl. Acad. Sci. USA, 94, 5673 ± 5678. Zou Z, Anisowicz A, Hendrix MJ, Thor A, Neveu M, Sheng S, Ra®di K, Seftor E and Sager R. (1994). Science, 263, 526 ± 529.

Oncogene