The clinicopathological significance of the expression of Granzyme B in oral squamous cell carcinoma

Oral Oncology 46 (2010) 185–189

Contents lists available at ScienceDirect

Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology

The clinicopathological significance of the expression of Granzyme B in oral squamous cell carcinoma Nádia Lago Costa a, Rita de Cássia Gonçalves Alencar b, Marize Campos Valadares c, Tarcília Aparecida Silva d, Elismauro Francisco Mendonça a, Aline Carvalho Batista a,* a

Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás, Goiânia, Brazil Division of Anatomopathology and Cytopathology, Araújo Jorge Hospital, Association of Cancer Combat of Goiás, Goiânia, Brazil Laboratory of Cellular Pharmacology and Toxicology, Faculty of Pharmacy, Federal University of Goiás, Goiânia, Brazil d Department of Oral Surgery and Pathology, Dental School, Federal University of Minas Gerais, Belo Horizonte, Brazil b c

a r t i c l e

i n f o

Article history: Received 18 October 2009 Received in revised form 28 November 2009 Accepted 30 November 2009 Available online 8 January 2010 Keywords: Granzyme B Oral cancer Oral squamous cell carcinoma Apoptosis

s u m m a r y Granzyme B (GB) is a serine protease synthesized by activated cytotoxic T-lymphocytes and natural killer cells that induces neoplastic cells apoptosis. The expression of GB in the tumor microenvironment has been considered a favorable prognostic factor in several types of human cancers. Thus, the aim of this study was to evaluate the density of GB+ cells in samples of oral cavity squamous cell carcinoma (OCSCC), as well as their relationship with clinical and microscopic parameters. GB expression was analyzed in 55 cases of OCSCC and metastatic and non-metastatic lymph nodes by means of immunohistochemistry. The high density of GB+ cells demonstrated an association with the high percentage of Bax+ and annexin V+ neoplastic cells. In addition, the number of peritumoral GB+ cells was significantly higher in the OCSCC group without lymph node metastasis, when compared with the metastatic OCSCC group. Moreover, patients with OCSCC with a high density of peritumoral GB+ cells showed a longer survival time when compared with patients with a lower density of these cells. In lymph node tissues, the density of GB+ cells was significantly higher in non-metastatic lymph nodes than in metastatic lymph nodes. Our findings suggest that the increased of expression of GB in the tumor microenvironment of OCSCC and in lymph nodes may have beneficial effect against neoplastic cells, contributing to apoptosis of these cells and increased survival of patients. Ó 2009 Elsevier Ltd. All rights reserved.

Introduction Granzyme B (GB) is a member of a serine protease family, secreted mainly by Cytotoxic T-lymphocytes (CTLs) and natural killer (NK) cells, which kill abnormally proliferating cells due to its ability to induce apoptosis.1–3 GB induces apoptosis directly by transforming procaspases, relatively inactive endocellular proteins of the apoptotic system, to caspases.2,3 In addition, GB can independently destroy cytoplasmic and nuclear proteins necessary for cell viability and activate proapoptotic proteins, which are involved in the mitochondrial apoptosis process.1–6 The increased number of the GB positive (GB+) cells in the tumor microenvironment has been considered a favorable prognostic factor in several types of human cancers, such as cutaneous squamous cell carcinoma (SCC), colorectal carcinomas, cervical intraepithelial neoplasia and melanoma.7–10 However, the clinicopathological sig* Corresponding author. Address: Disciplina de Patologia Geral e Bucal, Faculdade de Odontologia, Universidade Federal de Goiás, Praça Universitária S/N, Setor Universitário CEP: 74605-220, Brazil. Tel./fax: +55 62 3209 6327. E-mail addresses: [email protected] (N.L. Costa), alencar.ritadecassia@ gmail.com (Rita de Cássia Gonçalves Alencar), [email protected] (M.C. Valadares), [email protected] (T.A. Silva), [email protected] (E.F. Mendonça), [email protected] (A.C. Batista). 1368-8375/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2009.11.016

nificance of this protease in patients with oral cavity squamous cell carcinoma (OCSCC) has not yet been totally defined.11 Recently, we demonstrated that patients with oral SCC that had a high density of peritumoral CTLs showed a low frequency of lymph node metastasis, a lower neoplastic proliferative index and longer survival time, when compared with patients with low counts of these cells.12 Considering these findings, we hypothesized that the expression of GB, which is a primary molecular mediator of apoptosis, specifically expressed by CTL, contributes to more favorable local cytotoxic immune responses against oral neoplastic cells. In line with this, the present study was undertaken to evaluate the density of GB+ cells in the tumor microenvironment of OCSCC as well as their relationship with clinicopathological factors such as survival data, lymph node metastasis, tumor localization, tumor size, tumoral proliferation, WHO histological grading, intensity of inflammation and apoptosis. Materials and methods Samples The samples of this retrospective study consisted of surgicallyexcised specimens from 55 patients with primary OCSCC; and

186

N.L. Costa et al. / Oral Oncology 46 (2010) 185–189

these samples were obtained from files of the Anatomopathology Division of Araujo Jorge Hospital, Association of the Combat of Cancer of Goias State, Brazil. Samples of lymph nodes (n = 10) were also taken from the patients with oral tumor without cervical lymph node metastases (negative). In addition, in patients with cervical lymph node metastases, samples were obtained from both lymph nodes with (positive) (n = 10) and without metastases (negative) (n = 10). All patients in this study were submitted to surgical treatment consisting of cervical lymph node removal and none received radiotherapy, chemotherapy or any other treatment prior to surgery. Clinical data (gender, age, ethnic group, tobacco and alcohol consumption, tumor location, and tumor size) and follow-up information (survival data and death) were obtained from medical records. This study was approved by the Institutional Ethics Committee for human subjects. Light microscopy All specimens were fixed in 10% buffered formalin (pH 7.4) and paraffin embedded. The microscopic features were evaluated from the analysis of one 5 lm section of each sample, stained with hematoxylin and eosin. All the OCSCC sections were graduated according to the World Health Organization (WHO) classification of tumors.13 In addition, the intensity of associated inflammatory infiltrate (mild, moderate or intense) was evaluated.14 Immunohistochemistry and immunofluorescence techniques Paraffin-embedded tissues were sectioned (3 lm) and collected in series on glass slides coated with 2% 3-aminopropyltriethsilane (Sigma–Aldrich, St. Louis, MO, USA). The sections were deparaffinized by immersion in xylene, and this was followed by immersion in alcohol and then incubation with 3% hydrogen peroxide for 40 min. For antigen retrieval, the sections were immersed in citrate buffer (pH 6.0) for 20 min. Afterwards, the sections were incubated with 3% normal goat serum at room temperature, for 20 min. The slides were incubated with one the following primary antibodies: monoclonal mouse anti-human Granzyme B (clone 11F1; Novocastra, Newcastle, UK) at 1:100; monoclonal mouse anti-human cyclin B1 (clone 7A9, Novocastra) at 1:40; monoclonal mouse anti-human bcl-2 (clone 124, DAKO, Glostrup, Denmark) at 1:500 and polyclonal rabbit anti-human Bax (clone A3533, DAKO) at 1:500, at 4 °C overnight in humidified. After washing in TBS, the sections were treated with Novolink Max Polymer Detection System (Novocastra). Finally, the sections were stained with Mayer’s hematoxylin and were covered. Samples of tonsil were used as positive controls for all markers. Negative controls were obtained by the omission of primary antibodies, which were substituted by 1% PBS–BSA and by non-immune mouse (X501-1, DAKO) and rabbit (X0902, DAKO) serum. Apoptotic cells were estimated by binding of FITC-conjugated annexin V (TACS™ Annexin V Kits, 4830-250-K, R&D Systems, Gaithersburg, Maryland), which were evaluated under fluorescence microscopy. Cell counting and statistical analysis The densities (per mm2) of GB+ cells in the stroma near the invasion front of OCSCC (peritumoral) and within the cancer nests (intratumoral) were determined. All counts were performed in 10 alternate microscopic high-power fields (400) using an integration graticule (4740680000000-Netzmik-rometer 12.5, Carl Zeiss, Göttingen, Germany) and, at this magnification, each field of integration graticule had an area of 0.0961 mm2. Descriptive analyses were expressed as mean ± standard deviation (SD) of n observations/mm2. Comparative analyses were performed using the non-

parametric Mann–Whitney’s test. The assessment of tumoral proliferation (cyclin B1+ neoplastic cells) and of regulatory apoptotic proteins (bcl-2+ and Bax+ neoplastic cells) was calculated as the proportion of positive cells of the total neoplastic cell population at the invasion front of OCSCC. Additionally, annexin V was considered low when the majority of the fields (>7 fields) demonstrated up to 5 points of fluorescence and high when the majority of fields (>7 fields) demonstrated more than 5 points of fluorescence. The correlation and/or association between GB+ cells and microscopic (WHO histological grading, inflammatory infiltrate and proportions of cyclin B1+, bcl-2+, Bax+ and Anexin V+ neoplastic cells) and clinical characteristics (T stage) was calculated using Spearman correlation coefficients and/or Mann–Whitney’s test. The Kaplan–Meier methods were used for survival analysis. The survival time was calculated from surgical resection until the last followup appointment or the death of the patient. The samples of OCSCC were dichotomized by the median value of GB+ cells and differences in survival between groups were evaluated by the log-rank test. The level of statistical significance was accepted at P < 0.05.

Results The analysis of 55 patients with OCSCC revealed a predominance of males (70%) and an age range that varied from 35– 90 years (mean = 59.96). With regard to the last follow-up, the mean survival time was 41 months (95% CI = 14.3–67.9) for patients with non-metastatic OCSCC and 35.2 months (95% CI = 20.4–50.1) for patients with metastatic OCSCC. Details concerning clinical and microscopic findings are also summarized in Table 1. Cox regression analysis showed that, considering all clinical and microscopic parameters (Table 1), only cervical lymph node metastasis (presence and absence) was associated with a significant effect on overall survival of OCSCC patients, where absence of metastasis positively influenced the patient’s survival (Cox regression, b = 3.3, P = 0.009). Thus, in this study, we considered the following groups separately: non-metastatic OCSCC (n = 20) and metastatic OCSCC (n = 35). The immunohistochemical analysis revealed that GB+ cells were found in all cases displaying lymphocyte appearance and exhibited a granular cytoplasmic staining pattern. These cells were diffusely distributed throughout the tumoral stroma, mostly concentrated adjacent to the tumor front (peritumoral) and infiltrating on tumor parenchyma (intratumoral) (Fig. 1A and B). Our results revealed that the OCSCC group with a high number of GB+ cells on the tumoral front, in close contact with tumoral cells, showed a low percentage of bcl-2+ neoplastic cells (antiapoptotic protein) (P = 0.004) and a high percentage of Bax+ neoplastic cells (proapoptotic protein) (P = 0.031), when compared with the group with a low number of peritumoral GB+ cells (Fig. 2). In accordance with these data, the OCSCC group with the high number of peritumoral GB+ cells showed a higher binding of annexin V on neoplastic cells (Fig. 3). Our results also demonstrated that the peritumoral GB+ cell density was significantly higher in the non-metastatic OCSCC group (50.24 ± 9.99), when compared with metastatic OCSCC (28.10 ± 3.18) (P = 0.031). In the intratumoral region, the GB+ cell density was also higher in the non-metastatic OCSCC group (18.76 ± 3.96), when compared with metastatic OCSCC (10.78 ± 1.53); however, this difference was not statistically significant (P = 0.10). In addition, peri- and intratumoral GB+ cell densities were dichotomized by median values (high and low groups) and a log-rank test showed that the mean survival time for patients with high levels of peritumoral GB+ cells was significantly higher (51.5 months, 95% CI = 29.5–73.5), when compared with patients

N.L. Costa et al. / Oral Oncology 46 (2010) 185–189

187

Table 1 Main clinical and microscopic findings (%) of patients with OCSCC (n = 55). Clinical and microscopic features

OCSSC (%)

Age <60 years >60 years

54 46

Gender Male Female

70 30

Ethnic group Caucasian Non caucasian

51 49

Location Oral tongue Floor of the mouth Others

52 33 15

Tobacco Yes No

84 16

Alcohol Yes No

60 40

T stage T1–2 T3–4

25 75

Clinical outcome Dead Alive (overall survival)

41 59

Survival time Alive: P48 months Alive: <48 months

30 70

Metastasis Yes No

68 32

Cyclin B1 proportion (mean = 10.45%; IC 95% = 8.72–14.18) <10.45% >10.45%

52 48

bcl-2 proportion (mean = 0.60%; IC 95% = 0.31–0.88) <0.60% >0.60%

38 62

Bax proportion (mean = 6.65%; IC 95% = 1.24–14.55) <3.16% >3.16%

50 50

Intensity of inflammation Mild Moderate Intense

25 40 35

Histological grading (WHO grade) Grades I–II (well and moderately differentiated) Grades III–IV (poorly differentiated and undifferentiated)

38 62

with lower densities (19.82 months, 95% CI = 9.1–30.5) of these cells (Kaplan–Meier, log rank; P = 0.024) (Fig. 4). The log-rank test showed no difference in survival between high and low GB+ cells in the intratumoral region (P = 0.78). In lymph node tissues, the density of GB+ cells was significantly higher in non-metastatic lymph nodes of the patients without metastases (277.8 ± 53.4) than metastatic lymph nodes (72.5 ± 13.7) (P = 0.014). Non-metastatic lymph nodes (negative) (190.5 ± 23.1) exhibited higher GB+ cells numbers than metastatic lymph nodes (positive) in the same patients (P = 0.001). The density of GB+ cells, both in peri and in intratumoral tissues, was higher in the T1–T2 group (39.44 ± 7.21 and 14.61 ± 3.60, respectively), when compared with the T3–T4 group (31.03 ± 3.76 and 11.90 ± 1.88, respectively), however, this difference was not statistically significant (P > 0.05). No association between the GB+ cell densities and other clinical and microscopic

Figure 1 (A) Granzyme B (GB) positive cells (brown stain) in the stroma near the invasion front of oral cavity squamous cell carcinoma. (B) Shows a close-up view of the detached area in (A), showing GB positive cells in close contact with tumoral cells displaying lymphocyte appearance and granular cytoplasmic staining. Immunohistochemical staining, original magnification 400 (A) and 1000 (B). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Figure 2 Proportions of bcl-2+ and Bax+ neoplastic cells in low and high Granzyme B (GB) densities groups. Results are expressed as mean ± SD. The Mann–Whitney test was used to compare groups at a 0.05 significance level. Represents significant difference in low GB group when compared with high GB group.

188

N.L. Costa et al. / Oral Oncology 46 (2010) 185–189

Figure 3 Annexin V+ cells (arrows) in intratumoral regions (detached areas) in oral cavity squamous cell carcinoma. Immufluorescence staining, original magnification 100.

Figure 4 Kaplan–Meier survival curves, according to the density status of GB+ cells in peritumoral stroma in oral cavity squamous cell carcinoma. GB+ cells in peritumoral stroma were dichotomized by median values (log rank; P = 0.024).

It has been established in the literature that the high apoptotic activity of oral SCC neoplastic cells is associated with a lower lymph node involvement and higher survival time of patients.16,17 The presence of regional and distant metastasis in head and neck SCC patients has been also related to a decreased cytotoxicity of peripheral blood mononuclear cells due to the downregulated expression of GB, perforin and FasL.15 Despite Cruz et al. observed a low proportion of peri- and intratumoral GB+ cells in oral SCC, however no clinicopathologic associations were investigated by these authors.11 In the present study, we found a lower number of GB+ cells in the peritumoral microenvironment of the OCSCC group with lymph node metastasis, when compared with the non-metastatic OCSCC group. Corroborating this, for the first time, we showed that a higher peritumoral expression of GB positively influenced the survival time of OCSCC patients. However, it is important to consider that the association of GB with survival may be via the association with nodal status. In addition, our results showed a higher density of GB+ cells in non-metastatic lymph nodes than in metastatic lymph nodes. Corroborating our findings, Pretscher et al. observed, in patients with oro- and hypopharyngeal carcinoma, a higher number of CD8+ and GB+ cells in non-metastatic lymph nodes, when compared with metastatic lymph node tissues.21 Thus, it is possible that a more effective antitumoral response in the lymph node and in the microenvironment of the primary tumor may contribute to a favorable outcome in patients with head and neck cancer. In agreement with our data, the high expression of GB in the tumor environment has been associated with a favorable prognosis in cervical intraepithelial neoplasia,7 cutaneous SCC,8 colorectal carcinomas10 and melanoma.9 On the other hand, high expression of GB correlates to poor clinical outcome of patients with cervical and nasopharyngeal carcinomas.19,20 In conclusion, our findings suggest that the increased expression of GB in the tumor microenvironment of OCSCC and in lymph nodes may have a beneficial effect against neoplastic cells, contributing to apoptosis of these cells, lower lymph node involvement and increased survival time of patients.

Conflict of Interest Statement None declared.

features of OCSCC, such as tumor localization (oral tongue, floor of the mouth and others), tumoral proliferation (cyclin B1 proportion), WHO histological grading (grades I–II and III–IV) and intensity of inflammation (mild, moderated and intense) was found.

Discussion GB may induce apoptosis through mitochondrial-independent mechanisms via caspases2,3 and also to lead to the proteolysis of the protein, Bid, generating a gtBid fragment, that translocates to mitochondria, where it recruits the proapoptotic protein, Bax, to the external mitochondrial membrane culminating in cell apoptosis.2,3,5,6 Nevertheless, the apoptotic activity of GB is blocked by overexpression of protein bcl-2.6,18 In line with these data, our findings revealed an association between the higher expression of GB on the tumoral front of the OCSCC and higher expression of proteins Bax and annexin V binding by neoplastic cells. Interestingly, in the OCSCC group with high GB expression and high apoptotic activity, a low expression of anti-apoptotic protein bcl-2 was found. From these results, we can speculate that GB may contribute to destruction of neoplastic cells via activation of the mitochondrial apoptosis pathway, which may be favored by the low expression of bcl-2 by neoplastic cells.

Acknowledgment This work was supported by a grants from the National Council for Scientific and Technological Development (CNPq) (471878/ 2006-5 to A.C.B.) and from the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) (Pró-Equipamento/2008). References 1. Graubert TA, Ley TJ. How do lymphocytes kill tumor? Clin Cancer Res 1996;2(5):785–9. 2. Elmore S. Apoptosis: a review of programmed cell death. Toxicol Pathol 2007;35(4):495–516. 3. Sintsov AV, Kovalenko EI, Khanin MA. Apoptosis induced by Granzyme B. Russ J Bioorganic Chem 2008;34(6):647–54. 4. Trapani JA, Smyth MJ. Functional significance of the perforin/Granzyme cell death pathway. Nat Rev Immunol 2002;2(10):735–47. 5. Heibein JA, Goping IS, Barry M, Pinkoski MJ, Shore GC, Green DR, et al. Granzyme B-mediated cytochrome c release is regulated by the Bcl-2 family members Bid and Bax. J Exp Med 2000;192(10):1391–401. 6. Pinkoski MJ, Waterhouse NJ, Heibein JA, Wolf BB, Kuwana T, Goldstein JC, et al. Granzyme B-mediated apoptosis proceeds predominantly through a Bcl-2inhibitable mitochondrial pathway. J Biol Chem 2001;276(15):12060–7. 7. Kondo MC, Ribalta JCL, da Silva IDG, Alves MTS, de Azevedo-Fochi GR, Martins NV, et al. Granzyme B as a prognostic marker of cervical intraepithelial neoplasia. Eur J Oncol 2005;26(1):87–9.

N.L. Costa et al. / Oral Oncology 46 (2010) 185–189 8. Batinac T, Zamolo G, Hadzisejdic I, Zauhar G. A comparative study of Granzyme B expression in keratoacanthoma and squamous cell carcinoma. J Dermatol Sci 2006;44(2):109–12. 9. van Houdt IS, Sluijter BJR, Moesbergen LM, Vos WM, de Gruijl TD, Molenkamp BG, et al. Favorable outcome in clinically stage II melanoma patients is associated with the presence of activated tumor infiltrating T-lymphocytes and preserved MHC class I antigen expression. Int J Cancer 2008;123(3):609–15. 10. Mulder WMC, Bloemena E, Stukart MJ, Kummer JA, Wagstaff J, Scheper RJ. T cell receptor-f and Granzyme B expression in mononuclear cell infiltrates in normal colon mucosa and colon carcinoma. Gut 1997;40(1):113–9. 11. Cruz I, Meijer CJ, Walboomers JM, Snijders PJ, Van Der Waal I. Lack of MHC class I surface expression on neoplastic cells and poor activation of the secretory pathway of cytotoxic cells in oral squamous cell carcinoma. Br J Cancer 1999;81(5):881–9. 12. Zancope E, Costa NL, Junqueira-Kipnis AP, Valadares MC, Silva TA, Leles CR, et al. Differential infiltration of CD8+ and NK cells in lip and oral cavity squamous cell carcinoma. J Oral Pathol Med, in press. 13. Pindborg JJ, Reichard PA, Smith CJ, Van Der Wall I. Histological typing of cancer and precancer of the oral mucosa. World Health Organization classification of tumours. Berlin: Springer-Verlag; 1997.. p. 21–31. 14. Costa NL, Oton-Leite AF, Cheim-Júnior AP, Alencar RC, Bittar GO, Silva TA, et al. Density and migration of mast cells in lip squamous cell carcinoma and actinic cheilitis. Histol Histopathol 2009;24(4):457–65. 15. Bose A, Chakraborty T, Chakraborty K, Pal S, Baral R. Dysregulation in immune functions is reflected in tumor cell cytotoxicity by peripheral blood

16.

17.

18.

19.

20.

21.

189

mononuclear cells form head and neck squamous cell carcinoma patients. Cancer Immunol 2008;8:10. Zhang M, Zhang P, Zhang C, Sun J, Wang L, Li J, et al. Prognostic significance of Bcl-2 and Bax expression in the patients with oral squamous cell carcinoma. J Oral Pathol Med 2009;38(3):307–13. Camisasca DR, Honorato J, Bernard V, Silva LE, Fonseca EC, Faria PA, et al. Expression of Bcl-2 family proteins and associated clinicopathologic factors predict survival outcome in patients with oral squamous cell carcinoma. Oral Oncol 2009;45(3):225–33. Jans DA, Sutton VR, Jans P, Froelich CJ, Trapani JA. BCL-2 blocks perforininduced nuclear translocation of granzymes concomitant with protection against the nuclear events of apoptosis. J Biol Chem 1999;274(7):3953–61. Oudejans JJ, Harijadi H, Kummer JA, Tan IB, Bloemena E, Middeldorp JM, et al. High numbers of Granzyme B/CD8-positive tumour-infiltrating lymphocytes in nasopharyngeal carcinoma biopsies predict rapid fatal outcome in patients treated with curative intent. J Pathol 2002;198(4):468–75. Guzman VB, Silva IDCG, Brenna SMF, Carvalho CRN, Ribalta JCL, GerbaseDeLima M. High levels of Granzyme B expression in invasive cervical carcinoma correlates to poor response to treatment. Cancer Invest 2008;26(5):499–503. Pretscher D, Distel LV, Grabenbauer GG, Wittlinger M, Buettner M, et al. Distribution of immune cells in head and neck cancer: CD8+ T-cells and CD20+ B-cells in metastatic lymph nodes are associated with favourable outcome in patients with oro- and hypopharyngeal carcinoma. BMC Cancer 2009;9:292.