Professional Documents
Culture Documents
13381
Aim: The aim of this study was to analyse allelic loss PTEN expression was higher in MSD (P = 0.002) and
of the phosphatase and tensin homologue deleted on OSCC (P = 0.0259) compared with the C group; addi-
chromosome 10 (PTEN) gene and its protein tionally, a higher expression was observed in MSD
immuno-expression in dysplastic oral lesions and oral (P = 0.0035) and OSCC (P = 0.049) than MD.
squamous cell carcinomas (OSCCs). Regarding FISH analysis, a higher hemizygous (single
Methods and results: Samples were collected from copy) loss was observed in OSCC than in C
153 patients [20 ranulas used as a control (C); 30 (P = 0.0467) and in OSCC than in MD (P = 0.0175),
leucoplakias with mild dysplasia (MD); 30 leu- as well as a higher homozygous deletion in OSCC
coplakias with moderate to severe dysplasia (MSD); compared with C (P = 0.0159) and OSCC than MD
73 oral squamous cell carcinoma (OSCC)]. PTEN pro- (P = 0.0145).
tein expression was investigated using immunohisto- Conclusion: The results of this work suggest that
chemistry, and PTEN allelic loss was analysed by PTEN allelic loss is an important mechanism in the
fluorescence in-situ hybridisation (FISH). Differences late stage of the development of oral potentially
among groups were evaluated using the v2 test. malignant lesions into oral cancer.
Keywords: gene deletion, immunohistochemistry, leucoplakia, oral cancer, PTEN phosphatase
malignant lesions, such as oral leucoplakia, oral sub- Materials and methods
mucous fibrosis and oral erythroplakia.
TISSUE SAMPLES
The presence and degree of epithelial dysplasia are
used to assess the risk of leucoplakia progression to Samples were collected from the Service of Oral
malignancy; however, histological grading has lim- Pathology of the Jo~ ao de Barros Barreto University
ited value in predicting the individual risk of each Hospital (Belem, Brazil). The local ethics committee
case. Furthermore, the criteria reproducibility of approved this study (process no. 1 426 757).
microscopic grading of dysplastic lesions is very low, The specimens were obtained from 153 patients [20
and poses significant difficulties among patholo- samples of morphologically normal epithelium obtained
gists.3–5 Therefore, the understanding of the molecu- from ranulas with inflammatory infiltrated located dis-
lar mechanisms that govern the evolution of oral tant from the overlying epithelium to be used as a control
potentially malignant lesions into OSCC is highly rel- group (group C); 30 oral leucoplakias with mild dysplasia
evant for clinical practice, leading potentially to (MD); 30 leucoplakias with moderate to severe dysplasia
the identification of genes and proteins suitable for (MSD); and 73 OSCC]. Histological sections (5 lm) of all
therapeutic targeting and predictors of prognostic samples were stained routinely with haematoxylin and
determination.5–7 eosin (H&E) and analysed under light microscopy. Two
Akt/protein kinase B is a major downstream target independent oral pathologists without prior knowledge of
of growth factor receptor tyrosine kinase that signals the clinical data assessed the stained sections to classify
via phosphatidylinositol-3 kinase (PI3K), and is the histological grades of oral dysplasia according to a
activated frequently in different human malignan- previously proposed binary system criteria of classifica-
cies.6,8–10 Phospho-Akt (p-Akt) controls a variety of tion.3 In cases of disagreement, pathologists discussed the
critical cellular pathways during the carcinogenic findings and achieved a final agreement.
process, including those leading to apoptosis inhibi-
tion and increased cell proliferation, as well as
IMMUNOHISTOCHEMISTRY
enhanced tumour cell invasion, angiogenesis and cell
metabolism, through glucose metabolism.6,11 Phos- Immunohistochemical reactions were performed in
phatase and tensin homologue deleted on chromo- 3 lm sections that were dewaxed with xylene and rehy-
some 10 (PTEN) is a tumour suppressor gene, and its drated in ethanol series. Antigen retrieval was performed
major substrate is phosphatidylinositol (3,4,5)-tripho- by immersing the slides in an ethylenediamine tetraace-
sphate (PIP-3), which dephosphorylates PIP3 to tic acid (EDTA) solution (pH 8.0) for 15 min in a micro-
regenerate PIP2 and which antagonises PI3K func- wave oven. Peroxidase activity was blocked with a 6%
tion. Consequently, PTEN down-regulates Akt activity hydrogen peroxide and methanol (1:1) solution in two
by dephosphorylating and regulating PI3K signalling baths of 15 min each at room temperature. After wash-
negatively.12 Conversely, loss of PTEN expression ing in Tris buffer (pH 7.4), the slides were incubated
results in increased Akt activity and continued cell with primary antibody (polyclonal anti-human PTEN;
survival and proliferation. Invitrogen, Carlsbad, CA, USA) in a dilution of 1:100
Recent studies have proposed that activation of Akt overnight for 18 h at 4°C. The slides were subsequently
has been shown to be associated with advanced exposed to the avidin–biotin complex [LSAB-
stages and poor prognosis in OSCC;7,13–15 further- Kit + horseradish peroxidase (HRP); Dako Cytomation,
more, studies using OSCC cell lines have shown that Glostrp, Denmark] and to the 3,30 -diaminobenzidine
induced PTEN overexpression down-regulated Akt chromogen (DAB+; Dako Cytomation). The sections
phosphorylation significantly and induced apopto- were counterstained with Mayer’s haematoxylin, dehy-
sis.16 Regarding oral potentially malignant lesions, drated in ethanol, cleared in xylene and mounted. Pros-
several studies have shown that p-Akt may play an tate cancer tissue was used as a positive control; the
important role in the conversion of these lesions to its negative control was obtained by omitting the primary
malignant counterpart.6,7,15,17 specific antibody during the reaction. The results were
The expression of PTEN in potentially malignant considered positive when brown-coloured nuclear and
oral lesions remains to be investigated, and its impor- cytoplasmic stainings were observed, characterising the
tance to the oral carcinogenesis process deserves to presence of DAB in the immunohistochemical reaction.
be understood more clearly. Therefore, the aim of this The scoring system used has been described previ-
study was to determine the expression and the allelic ously in the literature.18 Briefly, the analysis was
loss of PTEN in oral leucoplakia with different degrees based on both the intensity and distribution of stain-
of epithelial dysplasia and in OSCC. ing. The distribution of stained cells was organised as
© 2017 John Wiley & Sons Ltd, Histopathology, 72, 330–338.
332 L A N Miyahara et al.
follows: 0 (0% of cells stained), 1 (1–50% of cells Table 1. Clinicopathological profile of OSCC samples
stained) and 2 (51–100% of cells stained). The inten-
sity of staining was rated as follows: 0 (no staining), Characteristics Data (%)
1 (mild staining), 2 (moderate staining) and 3 (strong Age (years)
staining). The final record for each case was defined
Average 61.13
by the sum of the values obtained as follows: FR0,
FR2, FR3, FR4 and FR5. Finally, FR0 and FR2 were Range 33–87
considered negative staining, while FR3, FR4 and
Sex
FR5 were considered positive staining.
Male/female 44/29 (60.27%/39.73%)
Table 2. Clinicopathological profile of leucoplakia samples was sometimes observed. In normal oral mucosa and
dysplastic tissues, the immunoreaction was found pre-
Characteristics Data (%) dominantly in the parabasal and basal layers,
Age (years) whereas OSCC showed a diffuse distribution in neo-
plastic islands (Figure 1).
Average 55.63
In the C group, 35% (seven cases) were positive for
Range 14–84 PTEN immuno-expression, whereas 65% (13 cases)
were negative. In the oral leucoplakia group, MD
Sex
showed that 43.33% (13 cases) were positive and
Male/female 28/32 (46.67%/53.33%) 56.67% (17 cases) were negative, whereas MSD
Tobacco consumption showed positivity in 80% of the cases (24 cases) and
20% (six cases) demonstrated negativity. Regarding
Smoker/non-smoker 25/33 (41.67%/55%) OSCC samples, 63.01% (46 cases) were positive and
Not specified 2 (3.33%) 36.99% (27 cases) were negative.
PTEN expression was higher in MSD (P = 0.002)
Anatomical location and OSCC (P = 0.0259) compared with the C group.
Tongue 32 (53.33%) Additionally, a higher expression was observed in
MSD (P = 0.0035) and OSCC (P = 0.049) than MD
Floor of the mouth 10 (16.67%)
(Table 3).
Others (jugal mucosa, retromolar 18 (30%)
region, palate, attached gingiva,
FLUORESCENCE IN-SITU HYBRIDISATION
alveolar ridge
ANALYSIS
This study comprised 73 patients (44 male and 29 All samples of normal epithelium did not demonstrate
female) with OSCC. An average age of 61.13 years allelic loss. In MD, only 3.33% (one of 30) showed
was observed, with a range of 33–87 years. Fifty-one homozygous deletion; in MSD, 3.33% (one of 30)
patients (69.86%) were smokers and 22 (30.14%) showed a PTEN hemizygous (single copy) deletion,
were non-smokers. Tumours were localised mainly in whereas 10% (three of 30) showed homozygous dele-
the floor of the mouth (36.99%), tongue (31.50%), tion. In OSCC samples, single-copy loss was observed
and alveolar ridge (19.18%). Tumours classified as in 13.70% (10 of 73) of the cases, and bi-allelic loss
T4 represented the majority of the cases (32.87%), occurred in 20.55% (15 of 73) (Figure 2). Statisti-
followed by T3 (28.77%), T2 (23.29%) and T1 cally, a higher hemizygous loss was observed in OSCC
(15.07%). A large proportion of the patients had than C (P = 0.0467) and OSCC than MD
regional or locoregional involvement (52.05%), and (P = 0.0175), as well as a higher homozygous dele-
35 of 73 (47.95%) were free of lymph node involve- tion in OSCC than C (P = 0.0159) and OSCC than
ment. Distant metastases were not observed in these MD (P = 0.0145) (Tables 4 and 5).
samples. Following the AJCC criteria, patients were
categorised as stage IV tumours in 46.57% of the
cases, stage III in 27.40%, stage II in 15.07% and
Discussion
stage I in 10.96%. Despite advances in therapeutic approaches, the mor-
Regarding oral leucoplakia, 60 cases were retrieved bidity and mortality associated with OSCC have not
(30 with MD and 30 with MSD). Of these samples, improved significantly in the last 30 years, with the
28 (46.67%) were male and 32 (53.33%) female. An 5-year overall survival rate varying between 30%
average age of 55.63 years was observed, with a and 50%.13 OSCC may be preceded by a potentially
range of 14–84 years. The proportion of smokers/ malignant lesion, and the presence of epithelial dys-
non-smokers was 25/33 (ratio 1:1.3). The majority plasia seems to be the most important prognostic
of the cases affected the tongue (53.33%), followed indicator of its potential for malignant transforma-
by the floor of the mouth (16.67%). tion;20 however, the incidence of malignant change
varies in different studies, possibly as a result of differ-
ences in ethnic and environmental factors.6 More-
IMMUNOHISTOCHEMICAL STAINING
over, the low reproducibility of current histological
PTEN immunoreaction showed, predominantly, a criteria for grading epithelial dysplasia is largely
cytoplasmatic staining. However, nuclear staining recognised and remains to be improved. These facts
© 2017 John Wiley & Sons Ltd, Histopathology, 72, 330–338.
334 L A N Miyahara et al.
A B
C D
Figure 1. Immunoexpression of phosphatase and tensin homologue deleted on chromosome 10 (PTEN) in normal, dysplastic and malignant
epithelium (streptavidin–biotin). A, Normal oral epithelium; B, epithelium with mild dysplasia; C, epithelium with moderate to severe dys-
plasia; D, oral squamous cell carcinoma.
Table 3. v2 test; P-value combination among groups at pathways that govern the progression of oral carcino-
immunohistochemistry genesis.6,21,22
PTEN is a tumour suppressor that opposes PI3K
C MD MSD OSCC function by dephosphorylating PIP3 to regenerate
C – NS 0.002 0.0259 PIP2. It is believed that down-regulation of PTEN
leads to an accumulation of PIP3, leading to
MD NS – 0.0035 0.049
increased activation of PI3K and pAkt. pAkt partici-
MSD NS NS – NS pates in the regulation of the cell cycle, proliferation,
apoptosis, cell adhesion, migration, invasion and
OSCC NS NS NS –
metastasis during cancer progression by regulating
C, control group; MD, mild dysplasia; MSD, moderate/severe dys- various downstream substrates.
plasia; OSCC, oral squamous cell carcinoma; NS, not significant. Therefore, in this study, we investigated the allelic
loss of PTEN and its immuno-expression in dysplastic
oral lesions and in OSCC. We employed the binary
highlight the necessity of identifying molecular mark- system to graduate dysplasia because this system has
ers that could predict oral leucoplakia aggressiveness been proved to be a reliable tool, reducing variability
and clinical course more clearly. For this reason, it is and enhancing the reproducibility of epithelial dyspla-
crucial to understand the deregulated molecular sia.3,23 Our results revealed that PTEN expression
© 2017 John Wiley & Sons Ltd, Histopathology, 72, 330–338.
PTEN in oral lesions 335
A B
C D
Figure 2. Vysis LSI phosphatase and tensin homologue deleted on chromosome 10 (PTEN) SpectrumOrange/CEP 10 SpectrumGreen Probes
hybridise to band 10q23 (SpectrumOrange LSI PTEN) and to the centromere, band region 10p11.1–q11.1 (SpectrumGreen CEP 10) of
human chromosome 10. Fluorescence in-situ hybridization (FISH) image of PTEN in normal, dysplastic and malignant epithelium. A, Normal
oral epithelium; B, epithelium with mild dysplasia; C, epithelium with moderate to severe dysplasia; D, oral squamous cell carcinoma.
Table 4. v2 test; P-value combination among groups at Table 5. v2 test; P-value combination among groups at
hemizygous deletion homozygous deletion
C – NS NS 0.0467 C – NS NS 0.0159
MD NS – NS 0.0175 MD NS – NS 0.0145
MSD NS NS – NS MSD NS NS – NS
OSCC NS NS NS – OSCC NS NS NS –
C, control group; MD, mild dysplasia; MSD, moderate/severe dys- C, control group; MD, mild dysplasia; MSD, moderate/severe dys-
plasia; OSCC, oral squamous cell carcinoma; NS, not significant. plasia; OSCC, oral squamous cell carcinoma; NS, not significant.
was higher in MSD (P = 0.002) and in OSCC MD. In addition, a higher hemizygous loss was
(P = 0.0259) when both were compared with the C observed in OSCC than C (P = 0.0467) and OSCC
group; moreover, a higher expression was observed than MD (P = 0.0175), as well as a higher homozy-
in MSD (P = 0.0035) and OSCC (P = 0.049) than gous deletion in OSCC when compared with C
© 2017 John Wiley & Sons Ltd, Histopathology, 72, 330–338.
336 L A N Miyahara et al.
(P = 0.0159) and in OSCC when compared with MD Here we also investigated the loss of PTEN copy
(P = 0.0145). To the best of our knowledge, this is number in oral dysplastic lesions. Our results
the first study analysing the expression and PTEN showed that the loss of PTEN copy number in oral
allelic loss in oral leucoplakias and in OSCC together. leucoplakias can represent an important mechanism
The molecular mechanism underlying malignant for the development of oral potentially malignant
transformation of leucoplakias has not yet been elu- lesions into progression to oral cancer when the loss
cidated. It is now accepted widely that there are of PTEN copy number occurred in 34.25% of our
divergent molecules and pathways that affect the sample (10 hemizygous deletion and 15 homozygous
process of transformation from leucoplakia to oral deletion). As PTEN is a tumour suppressor gene, for
squamous cell carcinoma sequentially.24,25 Recently, its complete inactivation there is a need for both
the tumour microenvironment was shown to carry copies to be deleted. In the cases of OSCC in which
out an important role in oral leucoplakia and oral there was hemizygous deletion, evaluation of the
cancer progression.26,27 In addition, several factors mutation or hypermethylation of the PTEN promoter
including age, sex, smoking and oral subsite, as region in the remaining alleles may elucidate
well as dysplasia grade, have been suggested as whether the loss of PTEN expression in OSCC sam-
predictors of the risk of malignant transformation of ples would be occurring through genetic or epige-
oral leucoplakia.26–28 These lines of evidence sug- netic events. Some authors have already addressed
gest that carcinogenic transformation of a leu- this issue, demonstrating that mutation in PTEN is
coplakia is multifactorial and is patient-specific.29 a rare event in HNSCC,36–41 whereas others have
Therefore, it is possible that inter-individual and shown that hypermethylation of the PTEN promoter
interpopulation differences in risk could be region is frequent in this tumour,36,42 which is an
explained partially by different distributions of important mechanism of epigenetic silencing that
genetic variants that may cause variation in the may contribute to cancer.
ability to metabolise carcinogens and/or effective Some papers highlight that homozygous deletion is
repair of the damage caused by them.30 The above unlikely to play a key role in the PTEN inactivation
observations contribute to understanding of 20% of process in OSCC.43,44 However, in this study bi-allelic
negative expression of PTEN in MSD group and loss occurred in 20.55% of the OSCC group. This
37% of negative expression in OSSC group. We homozygous loss of PTEN alleles was statistically
must also keep in mind that some oral cancers can higher in OSCC samples when compared with lesions
develop from lesions that lack dysplastic that presented in MD and the C group. An important
changes.31,32 On this basis we can explain, in some increase in loss of both alleles was observed in OSSC
measure, the proportion of cases in the C group when compared with MSD, although there was no
with expression-positive PTEN. statistically significant difference.
Recently, Chaves et al. studied p-Akt function in In conclusion, taking into consideration the fact
oral epithelial dysplasia and in OSCC, observing that that longitudinal studies are more indicated to attri-
p-Akt immunostaining was significantly higher in the bute predictive value for malignant transformation,
malignancy group than in dysplasias and controls.15 this cross-sectional study permits only speculation, to
Other studies investigated p-Akt in the process of some extent, that PTEN bi-allelic loss can be an
malignant transformation of oral epithelium, suggest- important mechanism in the late stage of develop-
ing that activation of the PI3K-AKT signal pathway ment of oral potentially malignant lesions into oral
is involved in the malignant phenotype acquisition cancer. Moreover, we observed an increase in the
process from its early stages.6,7,33 In addition, Akt expression of PTEN in the dysplastic lesions, suggest-
expression was shown to be a prognostic determinant ing an attempt of PTEN to control the increased
for patients affected by OSCC.13,34 In this regard, in expression of Akt. After acquisition of the malignant
our work the increase in PTEN expression observed phenotype, a decrease in PTEN expression was
in the evolution of dysplastic lesions can be under- observed, which may be explained by the increased
stood as an attempt to control the increased expres- deletion in this group.
sion of Akt described in several studies of oral
dysplastic lesions.6,7,15,33 In addition, a decrease in
PTEN expression was observed in OSCC. Reduced Conflicts of interest
PTEN protein expression in OSCC has also been
The authors state that they have no potential con-
observed in previous studies,21,35,36 suggesting a
flicts of interest.
PTEN functional inactivation in this malignancy.
© 2017 John Wiley & Sons Ltd, Histopathology, 72, 330–338.
PTEN in oral lesions 337
squamous cell carcinoma regulates phenotype switch of macro- 37. Cohen Y, Goldenberg-Cohen N, Shalmon B et al. Mutational
phages. Oncotarget 2016; 7; 81305–81321. analysis of PTEN/PIK3CA/AKT pathway in oral squamous cell
28. Sakata J, Yoshida R, Matsuoka Y et al. Predictive value of the carcinoma. Oral Oncol. 2011; 47; 946–950.
combination of SMAD4 expression and lymphocyte infiltration 38. Snaddon J, Parkinson EK, Craft JA, Bartholomew C, Fulton R.
in malignant transformation of oral leukoplakia. Cancer Med. Detection of functional PTEN lipid phosphatase protein and
2017; 6; 730–738. enzyme activity in squamous cell carcinomas of the head and
29. Narayan TV, Shilpashree S. Meta-analysis on clinicopathologic neck, despite loss of heterozygosity at this locus. Br. J. Cancer
risk factors of leukoplakias undergoing malignant transforma- 2001; 84; 1630–1634.
tion. J. Oral Maxillofac. Pathol. 2016; 20; 354–361. 39. Mavros A, Hahn M, Wieland I et al. Infrequent genetic alter-
30. Shridhar K, Aggarwal A, Walia GK et al. Single nucleotide ations of the tumor suppressor gene PTEN/MMAC1 in squa-
polymorphisms as markers of genetic susceptibility for oral mous cell carcinoma of the oral cavity. J. Oral Pathol. Med.
potentially malignant disorders risk: review of evidence to date. 2002; 31; 270–276.
Oral Oncol. 2016; 61; 146–151. 40. Henderson YC, Wang E, Clayman GL. Genotypic analysis of
31. Silverman S Jr, Gorsky M, Lozada F. Oral leukoplakia and tumor suppressor genes PTEN/MMAC1 and p53 in head and
malignant transformation. A follow-up study of 257 patients. neck squamous cell carcinomas. Laryngoscope 1998; 108;
Cancer 1984; 53; 563–58. 1553–1556.
32. Holmstrup P, Vedtofte P, Reibel J, Stoltze K. Long-term treat- 41. Shao X, Tandon R, Samara G et al. Mutational analysis of the
ment outcome of oral premalignant lesions. Oral Oncol. 2006; PTEN gene in head and neck squamous cell carcinoma. Int. J.
42; 461–474. Cancer 1998; 77; 684–688.
33. Watanabe S, Sato K, Okazaki Y, Tonogi M, Tanaka Y, Yamane 42. Sushma PS, Jamil K, Kumar PU, Satyanarayana U, Ramakr-
GY. Activation of PI3K-AKT pathway in oral epithelial dys- ishna M, Triveni B. PTEN and p16 genes as epigenetic
plasia and early cancer of tongue. Bull. Tokyo Dent. Coll. 2009; biomarkers in oral squamous cell carcinoma (OSCC): a study
50; 125–133. on south Indian population. Tumour Biol. 2016; 37; 7625–
34. Pickhard A, Gr€ ober S, Haug AK et al. Survivin and pAkt as 7632.
potential prognostic markers in squamous cell carcinoma of 43. Chen Q, Samaranayake LP, Zhou H, Xiao L. Homozygous dele-
the head and neck. Oral Surg. Oral Med. Oral Pathol. Oral tion of the PTEN tumor-suppressor gene is not a feature in oral
Radiol. 2014; 117; 733–742. squamous cell carcinoma. Oral Oncol. 2000; 36; 95–99.
35. Jasphin SS, Desai D, Pandit S, Gonsalves NM, Nayak PB, Iype 44. Gasparotto D, Vukosavljevic T, Piccinin S et al. Loss of
A. Immunohistochemical expression of phosphatase and tensin heterozygosity at 10q in tumors of the upper respiratory tract
homolog in histologic gradings of oral squamous cell carci- is associated with poor prognosis. Int. J. Cancer 1999; 84;
noma. Contemp. Clin. Dent. 2016; 7; 524–528. 432–436.
36. Kurasawa Y, Shiiba M, Nakamura M et al. PTEN expression
and methylation status in oral squamous cell carcinoma.
Oncol. Rep. 2008; 19; 1429–1434.