Head and Neck Pathol DOI 10.1007/s12105-014-0565-1
ORIGINAL PAPER
Expression of p8 in Human Oral Squamous Cell Carcinoma Christopher Bingham • Douglas Dickinson • James Cray • Komal Koli • Kalu U. E. Ogbureke
Received: 13 May 2014 / Accepted: 14 August 2014 Ó Springer Science+Business Media New York 2014
Abstract The present study investigated the expression of p8, a transcription factor upregulated in some human cancers, in oral squamous cell carcinomas (OSCCs). Immunohistochemical analysis of p8 expression was carried out on 20 archived surgical specimens of human OSCCs, and expression correlated with clinical outcome parameters in a retrospective study. Expression of p8 in a number of OSCC cell lines also was investigated by western blot and RT-PCR analyses. p8 was expressed in 80 % (16/20) of the samples with levels of expression exhibiting a significant difference (v2 = 8.352, df = 3, p = 0.039) by patient age. Furthermore, greater levels of p8 immunoreactivity was significantly associated with advanced tumor grade (p = 0.008). p8 also was upregulated in OSCC cell lines. p8 is expressed in a significant proportion of OSCCs,
Electronic supplementary material The online version of this article (doi:10.1007/s12105-014-0565-1) contains supplementary material, which is available to authorized users. C. Bingham Department of Periodontics, College of Dental Medicine, Georgia Regents University, Augusta, GA, USA D. Dickinson College of Graduate Studies, Georgia Regents University, Augusta, GA, USA J. Cray Department of Oral Biology, College of Dental Medicine, Georgia Reagents University, Augusta, GA, USA K. Koli K. U. E. Ogbureke (&) Department of Diagnostic and Biomedical Sciences, University of Texas School of Dentistry at Houston, 7500 Cambridge Street, SOD 5460, Houston, TX 77054, USA e-mail: [email protected]
and in human OSCC cell lines, suggesting a potential value of p8 as a diagnostic and/prognostic tool for oral cancers. Keywords p8 expression Oral squamous cell carcinoma Immunohistochemistry OSC2 cells
Introduction p8, a gene activated in rat pancreas during the acute phase of pancreatitis, is a transcription factor of the basic helixloop-helix (bHLH) family identified initially as playing a significant role in rat pancreatic development and regeneration. p8 also promotes cellular growth [1]. The human and mouse homologues of rat p8 subsequently were cloned and characterized, and the complete structure and mapping of human p8 to chromosome 16 at position p11.2 was determined [1, 2]. Three exons separated by two introns constitute the human p8 gene [1, 2]. The rat and mouse p8 polypeptides are each 80 amino acids long with a predicted molecular weight of 8kD, while the length of human p8 is 82 amino acids [1, 2]. The rat p8 shows 75 and 91 % homology with human and mouse p8 respectively [1, 2]. Published findings indicating that p8 is upregulated in a number of human cancers including breast, pancreatic, and lung cancer suggest that p8 might perform growth-promoting functions [3–5]. Reports of other studies by Vasseur et al. indicated that p8 might play a crucial role in tumor metastasis [6, 7]. For example, the identification of p8 in human breast cancer cell lines correlates with growthpromotion in early development and the ability to establish primary carcinomas [3–5]. However, its role in the later stages of tumorigenesis or metastasis of human breast cancer cells remains doubtful and unclear [3–5]. Interestingly, the p11.2 region on chromosome 16 (the location of
123
Head and Neck Pathol
human p8 gene) is frequently amplified in breast cancer [6, 7]. Other studies have reported high expression of p8 in pituitary tumors, and that RNAi silencing of p8 impaired pituitary tumorigenesis in mice [8]. Similarly, in vitro studies show that p8 mRNA is expressed in a human hepatoma HepG2 cell line in vitro [2]. Recently published evidence has suggested that p8 regulates autophagy and apoptosis and may be relevant to cardiovascular pathologies and cancers associated with dysregulated autophagy [9]. It also was determined to regulate DNA repair, thereby adding another layer to a chromatin regulatory network [10, 11]. Thus, these reports collectively indicate that p8 may play vital roles in metastasis, chemoresistance, survival, or apoptosis of, at least some carcinomas. However, the expression and potential functional implications of p8 in human OSCC have not been reported. Here, we report findings of our investigation of the expression of p8 in human OSCC.
Materials and Methods Case Selection for Immunohistochemistry The specific samples used in this study have been described in a previous publication [12]. A retrospective cohort analysis of the immunohistochemical expression of p8 in 20 cases of de-identified surgical biopsied human OSCC maintained in the archives of the Department of Pathology, Georgia Health Sciences University was performed. Cases selected were those from patients seen between January 2004 and December 2007. Thereafter, expression was correlated with clinical prognostic and outcome parameters of OSCC. The Inclusion criteria included the availability of a pathology record indicating a histologic diagnosis of OSCC on hematoxylin/eosin sections independently verified by two oral pathologists, and excision/resection of primary tumor following histopathologic diagnosis. The Exclusion criteria included a stipulation on a Pathology Request Form of the presence of neoplasm of other anatomic sites/systems at the time of diagnosis of primary OSCC, previous diagnosis of and treatment for OSCC, and prior history, diagnosis, or treatment for cancers of other organs/regions. The cases satisfying the inclusion criteria were randomly selected from archived cases in the Department of Pathology diagnostic database using the search words: ‘‘oral squamous cell carcinoma.’’ Two pathologists were blinded to the clinicopathologic details of each case including whether or not patients presented with recurrent disease, or post-surgical resection, at primary site. Cases were randomly selected until 20 cases satisfying the inclusion criteria were identified.
123
Histopathologic Analysis of Hematoxylin and Eosin Sections The presence of invasive OSCC on the primary resection/ excision specimen was determined by a combination of established architectural and cytologic parameters including: nuclear pleomorphism; increased and/or abnormal mitosis; hyperchromatism; basal cell hyperplasia; abnormal maturation sequence; and stromal invasion. Immunohistochemistry Five lm sections of human OSCC were obtained from archived paraffin blocks of tissues from cases selected and immunostained with p8 antibody using the Nemesis 7200 automated system supplied with Super-Picture-Perfect Broad-Spectrum HRP-Polymer and Single-Solution-AEC reagents from BioCare (San Francisco, CA) as described previously [12–14]. The p8 polyclonal antibody was a kind donation from Dr. Juan L. Iovanna (Marseille Cedex, France). In brief, after manually dewaxing sections in three 5-min xylene washes, and rehydrating through graded ethanol (100, 95, and 75 %) and water, endogenous peroxidase activity was prevented by treating the sections with 3 % hydrogen peroxide (in methanol) for 30 min. Sections thereafter were washed 3 times in phosphate-buffered saline (19 PBS) for at least 5 min each and covered with PBS ? 0.05 % Tween-20 (PBS-T) before loading the slides on to the preprogrammed Nemesis automated immunohistochemistry machine. Incubation of sections for 1 h with p8 antibody diluted in 10 % normal goat serum in PBS was performed before washing (4 9 1 min) with PBS-T followed by incubation with SuperPicTure Polymer HRP-conjugated broad-spectrum secondary antibody (# 87-8963, Zymed Lab. Inc., San Francisco, CA, USA) for 10 min. The chromogenic substrate, aminoethylcarbazole (AEC Single Solution, # 00-1122, Zymed Lab. Inc., San Francisco, CA, USA), was then applied for 2 min, rinsed in tap water, and manually counterstained with Mayer’s hematoxylin for 10 s. All steps were performed at room temperature. Finally, an overlay of Clearmount glaze was applied over the sections and left to dry overnight at room temperature before coverslips were applied with the aid of Histomount (Zymed Lab. Inc., San Francisco, CA, USA). A negative control consisted of the substitution of p8 with non-immune rabbit serum or mouse IgG control (# 08-6599, Zymed Lab. Inc., San Francisco, CA, USA), while a positive control consisted of human pancreatic tissue sections (known to be immunoreactive for p8) stained with p8 antisera. Representative photographic images were captured using the Axioplan2 Universal microscope equipped with an Axiovision digital camera and Axiovision program (Carl Zeiss Gmbh, Jena, Germany).
Head and Neck Pathol
Scoring of Immunohistochemistry Results Proportions of immunostaining for p8, based on the combined intensity of cytoplasmic staining (CS) and nuclear staining (NS), was assigned semi-quantitative scores by the two independent observers earlier calibrated for scoring criteria. Scoring was as: negative (0, immunoreactivity not detectable/faint staining in less than 10 % of tumor cells); 1 (10–50 % of immunoreactive tumor cells); 2 (50–75 % of immunoreactive tumor cells); and 3 (widely and intensely expressed in tumor cells). Statistical Analysis Non-parametric correlation, Kendall Tau, Fisher’s exact test, Kruskal–Wallis and ordinal regression were utilized to determine trends and significant differences for the response variable, p8 immunostain (p8IHC) score, by the independent or predicting variables, Staining Pattern, Subsite, Age, Gender, Ethnicity, TNM stage, Lymphatic and Vascular Invasion (LVI), Perineural (PN) spread, Radiotherapy (RT) and Recurrence, respectively. Odds ratios, where informative, were also included. Differences were considered significant if p \ 0.05. All analyses were performed using SPSS 19.0 (IBM Corporation) and SAS 9.1 (SAS Institute). All graphs were created in SPSS 19.0. Human Cell Lines and Culture Conditions SCC25 and OSC2 (OSCC), and DOK (dysplastic oral keratinocytes), and A818-6 (Human Pancreatic Adenocarcinoma) human cell lines have been published [15–18] and were initially obtained from American Type Culture Collection. The human oral keratinocyte (HOK) whole lysate was purchased from ScienCellTM Research Laboratories (cat. #2616; San Diego, CA). Endogenous expression levels of p8 protein and mRNA were evaluated by Western blot and semiquantitative-reverse transcriptase (RT)-PCR analyses using whole-cell lysates and total RNA extracts, respectively, from the above cell lines. The human pancreatic adenocarcinoma whole cell lysate and extract was used as positive control. The cell lines were routinely cultured as monolayers in DMEM/F12 medium containing 10 % FBS (Invitrogen, Carlsbad, CA) supplemented with 1 % Penicillin/Streptomycin and 500 ng/ml Hydrocortisone (Sigma Aldrich, St. Louis, MO) and maintained in the presence of 5 % CO2 humidified air at 37 °C. Semiquantitative RT-PCR Following the manufacturer’s instructions, total RNA was extracted from a third of the cell pellets from each group (experimental and control) using the RNEasy Plus Mini kit
(Qiagen, CA). 1 lg of total RNA was used to produce cDNA by reverse-transcription using the High Capacity Reverse Transcription Kit (Applied BioSystems, CA) in a conventional Thermo cycler. PCR reaction was performed using the Promega GoTaq Green MasterMix reagent (cat. #M7122; Promega, Madison WI), and the p8 primer pair (sc-40792-PR; exact primer sequences are proprietary and therefore were not available from company; Santa Cruz Biotechnology), and 100 ng of cDNA in a 25 ll PCR reaction. The primer design amplified a 407 base amplicon. GAPDH was used as a normalizing control. At the end of the reaction, PCR products were electrophoresed on a 2 % agarose gel in Tris–acetate-EDTA (TAE) buffer, stained with ethidium bromide, and photographed. Densitometric analysis of p8 mRNA levels was performed and results represented by a histogram. Western Blot Analysis Total cell lysates prepared from 2/3rd of the cell pellets from each group (experimental and controls) were resolved on SDS–PAGE gel using the Mini-Protean Tetra Cell unit at 100 V (Bio-Rad, CA) by standard protocol. Briefly, 25 lg of protein was loaded in each lane and resolved on 7 % polyacrylamide gels before transferring to polyvinylidene (PVDF) membranes (Millipore, MA) at 100 mA constant current for 75 min in the Bio-Rad Transfer unit. The membrane was then incubated in blocking solution consisting of 1:1 diluted Sea Block Blocking solution (LICOR, Biosciences, Lincoln, Nebraska) in 19 PBS, for 1 h at RT. Membrane was thereafter incubated in anti-p8 polyclonal antibody and anti-b-actin monoclonal antibody (Sigma, St. Louis, MO) diluted in blocking buffer (1:1,000 and 1:5,000, respectively) overnight at 4 °C before washing with 19 PBS-T for 5 9 5 min. The membrane was then incubated with IRDye 600 goat anti-mouse secondary antibody (L-ICOR Biosciences) at 1:5,000 dilution for 1 h at RT, washed with 19 PBS-T 4 9 5 min, and with 19 PBS for 5 min. Signal was detected using the Infrared LICOR Imaging system (LI-COR Biosciences).
Results p8 Expression in Primary OSCCs The cellular expression of p8 was determined by immunohistochemistry in successive paraffin slides from each of the 20 OSCC cases selected for this study. Representative immunoreactivity results are shown in Fig. 1 with 1A representing a semiquantitative score of ‘‘1’’, 1B a score of ‘‘2’’, and 1C a score of ‘‘3’’. Figure 1d show a lack of expression of p8 in normal oral mucosa obtained from
123
Head and Neck Pathol
Fig. 1 Representative p8 immunoreactivity of human OSCC sections. a A semiquantitative score of ‘‘1’’ indicating a \10 % population of tumor cells positive for p8, while (b) illustrates a score of ‘‘2’’ where more than 50 % but \75 % of tumors cells are immunoreactive to p8. c Score of ‘‘3’’ indicating more than 75 %
diffuse immunoreactivity of tumor cells. d Lack of expression of p8 in normal oral mucosa, whereas e is a representative pre-immune negative control on OSCC. Pink–Brown stain (arrow) represents positive immunoreactive tumor cells for p8. (Images are at mag. 9 10 and corresponding insets for a–c, e are at mag. 9 40)
surgical wastes of non-cancer patients, whereas Fig. 1e is a representative negative control (non-immune serum) in OSCC section. In some tumor cells, p8 immunoreactivity was confined to the cytoplasm and the perinuclear perimeters, while in others staining of the nucleus was evident. The connective tissue stroma in all cases showed negative immunoreactivity for p8. Table 1 summarizes the clinical data of patients and the semiquantitative scores for p8 immunoreactivity of the study samples. As shown, p8 is expressed in 16 (80 %) of the 20 cases of OSCCs studied.
v2 = 8.352, df = 3, p = 0.039). A post hoc Nemenyi test showed marginal differences upon multiple comparisons (p8 expression levels 1–2, and 1–3 approached p = 0.05). This is most likely due to the unequal samples sizes, but may be indicative of a type 1 error. Figure 3 represent the cumulative probabilities level of p8 expression by each of the predictors. Of note is the ordinal regression models showing a significant correlation (p = 0.008) between level of p8 immunoreactivity and the histologic differentiation status of OSCC (well-, moderate-, poorly-differentiated) as seen on hematoxylin and eosin (HE-DIFF) stained sections. This suggests a relationship between p8 expression and OSCC differentiation status. Thus, poorlyand moderately-differentiated OSCCs upregulate p8 more than their well differentiated counterparts. No other patient characteristics or clinical variables were found to be significantly associated with p8 expression. As previously reported for this same cohort of the 20 cases [12], 9 (45 %) had clinical documentation of recurrence of OSCC (Table 1); a rate not significantly different from published estimated population recurrence-rate of 50 % (p = 0.664) [12]. Time between initial surgical
p8 Expression and Patients Clinical Variables Table 2 shows the results of a nonparametric bivariate (Kendall Tau) correlation of all variables with p8 immunoreactivity, while Table 3 summarizes the results for the Fisher’s Exact test for each variable compared to p8 immunoreactivity. Table 4 represents the results of the ordinal regression modeling for p8 immunoreactivity by each variable, 1 factor ordinal regression. With respect to mean age, the level of p8 expression as shown in Fig. 2 showed a significant difference (Kruskal–Wallis Test
123
Head and Neck Pathol Table 1 Semiquantitative scores for p8 immunoreactivity on oral squamous cell carcinoma (N = 20) Serial no.
Histologic characteristics of primary tumor (PT) HEDIFF
Clinical characteristic/parameters of patients
P8 IHC scorec
Staining pattern
Oral subsite location
Age/gender/ ethnicitya
TNM stage
LV invasion
PN spread
RTb
Recurrence N
1
WD
3
CS ?NS
TG
58//M/AA
0
1
1
0
2
PD
1
CS
FOM
63/M/C
0
0
0
0
N
3
PD
2
CS ?NS
TG
44/M/C
2
0
0
0
Y (17mo)
4
PD
0
NS
TG
54/F/AA
0
0
1
0
N
5
PD
0
NS
BOT
51/M/C
0
0
0
0
Y (24mo)
6
WD
2
NS
TG
47/F/AA
2
1
2
0
N
7
MD
1
NS
TG
50/M/C
0
1
0
0
N
8
PD
1
NS
FOM
53/M/C
2
0
0
0
N
9
PD
2
NS
SP
45/F/AA
3
1
0
0
N
10
WD
2
CS
TG
48/M/C
2
1
0
0
N
11
MD
2
CS
BM
52//F/other
2
0
1
0
Y (11.5mo)
12
MD
3
CS
FOM
85/F/C
2
0
1
1
N
13
PD
1
NS
BM
68/F/C
2
1
1
0
Y (13mo)
14
MD
3
NS
TG
68/F/C
0
0
0
2
Y (12mo)
15
PD
3
NS
TG
51/F/AA
1
1
1
0
Y (9mo)
16
WD
2
CS ?NS
RMT
53/F/AA
0
1
0
3
Y (7mo)
17
PD
1
NS
BM
58/F/C
3
0
1
0
N
18
WD
0
NS
RMT
49/M/C
2
1
1
3
N
19
PD
0
NS
FOM
68/F/C
2
1
1
0
Y (7mo)
20
MD
2
NS
FOM
54/M/C
0
1
1
0
Y (6mo)
F female, M male, C caucasian, AA African American, N absence of recurrence at least 48 months post-primary treatment, LV presence (1) or absence (0) of lymphatic and/vascular invasion, PN spread presence (1) or absence (0) of perineural spread, RT number of times adjunct radiotherapy was given post-surgical treatment Oral Subsites: BM buccal mucosa, FOM floor of mouth, RMT retromolar trigone, BOT base of tongue, TG lateral tongue Degree of tumor differentiation on hematoxylin and eosin stain (HE-DIFF): WD well-differentiated, MD moderately-differentiated, PD poorlydifferentiated a b c
Mean age = 56 years Average recurrence free-time (ARFT) = 10.5 months Immunohistochemical scoring was semiquantitative using prior published criteria; p8 Staining pattern: CS cytoplasmic, NS nuclear
resection of primary tumor and initial recurrence (Recurrence-Free Time; RFT) ranged from 6 to 17 months with an average RFT (ARFT) of 10.5 months (Table 1). The correlation values for Recurrence approached 0.40, considered as a moderate correlation. p8 Upregulation in OSCC and Oral Dysplastic Epithelial Cell Lines Further investigation of p8 expression in two human OSCC cell lines (OSC2 and SCC-25), the human dysplastic oral keratinocyte cell line (DOK), and the human primary oral keratinocyte (HOK) cells by Western blot (WB) and semiquantitative reverse transcriptase (RT)-PCR analyses showed significant upregulation in OSC2, SCC25 and
DOK compared with HOK cells at both the translational (Fig. 4a) and transcriptional (Fig. 4b) levels. Positive control consisted of whole-cell lysates and total RNA extracts from pancreatic cancer cell line, A818-6. With respect to the oral cancer cell lines, these in vitro findings are consistent with immunohistochemistry results (Fig. 1) indicating up-regulation of p8 in OSCC specimens from patients.
Discussion The aims of this study were to determine the expression of p8 in human OSCC. This is with a view to laying the foundation for extended inquiry into the functional and
123
Head and Neck Pathol
mechanistic role of p8 in the biology of oral cancer. A search of the English language literature suggests that the present report represents the first demonstrating the Table 2 Correlation of variables with p8 immunoreactivity Variable
Kendal Tau
p value
HE-DIFF
0.27
0.20
Stain
0.28
0.21
Subsite
0.33
0.10
Age
0.05
0.80
Gender
0.10
0.64
Race
0.08
0.70
TNM
0.01
0.97
LV
0.11
0.60
PN
0.07
0.72
RT
0.14
0.51
Recurrence
0.36
0.11
upregulation of p8 in a significant proportion of OSCC resection specimens as well as in the OSCC cell line, OSC2 and SCC-25, and in DOK cells but absent in normal oral mucosal tissue (as evidenced by immunohistochemistry), although basal levels are seen in HOK cells by western blot and RT-PCR analyses. Our observation that p8 is at all expressed by HOK cells (albeit at basal level) was unexpected because of an assumption that the HOK cells are comparable to the normal oral mucosa, which lacked p8 expression. Since p8 is also known to be a stress-related protein whose expression may be enhanced under stress conditions [19], we speculated that the low level expression of p8 in HOK cells may be in response to cell cultureinduced stress. Except for histologic differentiation status, p8 expression and/or levels, as determined by immunohistochemistry, pointed to no apparent correlation with clinical outcome in OSCCs patients.
A nonparametric bivariate correlation by way of Kendall Tau was performed for all variables with p8 immunoreactivity. Note that there are no significant correlations detected. However, the correlation value for recurrence is of note. HE-DIFF histological differentiation; TNM tumor size (T), presence/absence of regional node metastasis (N), and presence/absence of distant metastasis (M), LV lymphatic and vascular invasion, PN perineural invasion, RT radiotherapy
Table 3 Fisher’s exact test for variable compared to p8 immunoreactivity
Distribution of values for p8 immunoreactivity was investigated by each independent variable by way of Fisher’s Exact test. Note that there are no significant differences in distribution of p8 immunoreactivity detected
Variable
p value
HE-DIFF
0.21
Stain Subsite
0.88 0.64
Gender
0.92
Race
0.91
TNM
0.92
LV
0.88
PN
0.26
RT
0.36
Recurrence
0.16
Fig. 2 Plot of mean age (± SD) by p8-IHC category. Shows a significant difference (v2 = 8.352, df = 3, p = 0.039)
Table 4 Ordinal regression modeling for p8 immunoreactivity Cox snell R2
Variable
Model
Goodness of fit
AIC
Parallel lines
HE-DIFF
v2 = 12.807, p = 0.008
v2 = 2.121 p = 0.908
57.019
v2 = 12.807, p = 0.046
0.463 0.368
2
Stain
v = 7.332, p = 0.291
v = 2.564 p = 0.861
37.596
v2 = 7.524, p = 0.736
Subsite
v2 = 4.688, p = 0.455
v2 = 9.632, p = 0.473
55.372
v2 = 13.070, p = 0.220
0.219
2
2
Gender
v = 0.239, p = 0.625
v = 0.554 p = 0.758
57.606
v2 = 0.554, p = 0.758
0.012
Race
v2 = 0.169, p = 0.919
v2 = 3.024 p = 0.554
57.682
v2 = 3.659, p = 0.454
0.009
2
2
TNM
v = 2.342, p = 0.719
v = 4.228, p = 0.646
57.884
v2 = 5.188, p = 0.520
0.068
LV
v2 = 0.268, p = 0.875
v2 = 3.543, p = 0.471
57.583
v2 = 4.349, p = 0.361
0.014
2
2
2
2
PN
v = 3.072, p = 0.381
v = 8.590 p = 0.198
57.528
v = 7.223, p = 0.301
0.149
RT
v2 = 7.560, p = 0.056
v2 = 1.132 p = 0.970
56.016
v2 = 1.949, p = 0.924
0.343
Recurrence
v2 = 3.313, p = 0.069
v2 = 3.364, p = 0.186
17.252
v2 = 4.295, p = 0.117
0.168
123
Head and Neck Pathol
Fig. 3 Cumulative probabilities for each independent RT (a), Recurrence (b), HE-DIFF (c), Staining characteristics (d), oral subsite (e), Gender (f), Ethnicity (g), TNM (h), LV (i), and PN (j) variable plotted by p8 IHC (semiquantitative p8 immunostaining) category. Note the difference for Recurrence by p8 IHC category, specifically category 3. HE-DIFF = Degree of histologic differentiation as seen on hematoxylin and eosin stained sections: 1 = Well-
Differentiated (WD); 2 = Moderately-Differentiated (MD); 3 = Poorly-Differentiated (PD). N = Absence of recurrence at least 48 months post-primary treatment; LV = presence (1) or absence (0) of lymphatic and/vascular invasion; PN spread = presence (1) or absence (0) of perineural spread; RT = number of times adjunct radiotherapy was given post-surgical treatment
Previous reports on the expression of p8 in cancers, other than oral cancer, and correlation of expression with outcome and prognostic parameters of these cancers have provided conflicting and paradoxical results. Thus, several antitumor functions attributable to p8 have been reported in spite of its notable tumor-promoting activity. Except for marginal correlation with recurrence shown by the ordinal regression and cumulative probabilities analysis, results of our present study indicated no apparent correlation of p8 expression with outcome parameters. The production of multiple and even opposing effects by p8, despite its relatively small size (8-12kd) and the absence of known interaction domains, make it a puzzling biological molecule. It promotes tumor growth and aggressiveness and protects tumor cells from apoptosis; yet
at the same time it can act as a tumor suppressor [1, 2, 19]. In breast cancers, for example, p8 expression resulted in poor disease outcome [20], whereas upregulation in prostate cancers was associated with favorable disease outcome [20]. In human prostate cancer cells, p8 expression is lower compared with levels in normal epithelial cells [21, 22]. Furthermore, expression levels inversely correlated with the invasiveness and growth potential of prostate cancer cells in vitro, and overexpression of p8 reduced the growth rate of prostate tumors in vivo [21, 22]. Thus, p8 was proposed to behave as a potential tumor suppressor for human prostate cancers; an effect partially mediated by the interaction of p8 with the peroxisome proliferator-activated receptor-c coactivator PGC-1 [21, 22]. Our observation in the present study that poorly-differentiated (PD) OSCCs,
123
Head and Neck Pathol
Fig. 4 a Western blot (WB) and densitometric analyses show over twofold increase in the levels of p8 expression in OSC2, SCC25, and DOK cells compared with levels in HOK cells. The pancreatic carcinoma cell human A818-6 known to express abundant p8 was used as positive control. Normalization was with b-actin. b Semiquantitative RT-PCR analysis shows p8-mRNA expression in OSC2, SCC25, DOK, and HOK cells were proportionately comparable with
WB protein levels in (a). Normalization of RT-PCR was with GAPDH. Cells used in study: OSC2 is a cell line from a regional lymph node metastases of human tongue primary SCC; SCC25 is a human tongue primary SCC; DOK is a human oral epithelial dysplastic cell line from the tongue; and A818-6 is a human pancreatic adenocarcinoma cell line
which also tend to be more aggressive, downregulate p8 seems to fit these paradox models observed with prostate, breast, and pancreatic cancers discussed above. In summary, the data presented in this study show that p8 is expressed in a significant proportion of OSCCs as well as in OSCC carcinoma cell lines, although expression levels do not appear to impact clinical outcome and prognostic parameters. Any ascribed role for p8 in oral cancer is significant because, as a secreted protein, it may represent an ideal target for therapeutic intervention. While a detailed study of the mechanism of p8-oral cancer tumorigenesis was beyond the scope of this study, the data provides a framework for future studies elaborating on a p8 mechanistic network involved in oral cancer biology. This, in turn, is with a view to identifying target points for the design of potent biomimetics for intervention in oral cancer.
2. Vasseur S, Mallo GV, Garcia-Montero A, et al. Structural and functional characterization of the mouse p8 gene: promotion of transcription by the CAAT-enhancer binding protein alpha (C/ EBPalpha) and C/EBPbeta trans-acting factors involves a C/EBP cis-acting element and other regions of the promoter. Biochem J. 1999;343:377–83. 3. Ree AHM, Tvermyr O, Engebraaten M, et al. Expression of a novel factor in human breast cancer cells with metastatic potential. Cancer Res. 1999;59:4675–80. 4. Ree AH, Pacheco MM, Tvermyr M, Fodstad O, Brentani MM. Expression of a novel factor, com1, in early tumor progression of breast cancer. Clin Cancer Res. 2000;6:1778–83. 5. Pommier RM, Gout J, Vincent DF, et al. The human NUPR1/P8 gene is transcriptionally activated by transforming growth factor b via the SMAD signalling pathway. Biochem J. 2012;445:285–93. 6. Courjal F, Theillet C. Comparative genomic hybridization analysis of breast tumors with predetermined profiles of DNA amplification. Cancer Res. 1997;57:4368–77. 7. Gruel N, Lucchesi C, Raynal V, Rodrigues MJ, et al. Lobular invasive carcinoma of the breast is a molecular entity distinct from luminal invasive ductal carcinoma. Eur J Cancer. 2010;46:2399–407. 8. Mohammad HP, Seachrist DD, Quirk CC, Nilson JH. Reexpression of p8 contributes to tumorigenic properties of pituitary cells and appears in a subset of prolactinomas in transgenic mice that hypersecrete luteinizing hormone Mol. Endocrinol. 2004;18:2583–93. 9. Kong DK, Georgescu SP, Cano C, et al. Deficiency of the transcriptional regulator p8 results in increased autophagy and apoptosis, and causes impaired heart function. Mol Biol Cell. 2010;21:1335–49.
References 1. Mallo GV, Fiedler F, Calvo EL, Ortiz EM, et al. Cloning and expression of the rat p8 cDNA, a new gene activated in pancreas during the acute phase of pancreatitis, pancreatic development, and regeneration, and which promotes cellular growth. J Biol Chem. 1997;272:32360–9.
123
Head and Neck Pathol 10. Gironella M, Malicet C, Cano C, et al. p8/nupr1 regulates DNArepair activity after double-strand gamma irradiation-induced DNA damage. J Cell Physiol. 2000;221:594–602. 11. Sambasivan R, Cheedipudi S, Pasupuleti N, Saleh A, Pavlath GK, Dhawan J. The small chromatin-binding protein p8 coordinates the association of anti-proliferative and pro-myogenic sproteins at the myogenin promoter. J Cell Sci. 2009;122: 3481–91. 12. Ogbureke KUE, Weinberger PM, Looney SW, Fisher LW. DSPP, OPN, and MMP-9 expression at histologically-negative resection margins predict Oral squamous cell carcinoma recurrence. Oncotarget. 2012;3:286–98. 13. Ogbureke KUE, Nikitakis NG, Warburton G, Ord R, Sauk JJ, Fisher LW. Upregulation of SIBLING proteins and correlation with cognate MMP expression in oral cancer. Oral Oncol. 2007;43:920–32. 14. Ogbureke KUE, Fisher LW. SIBLING expression patterns in ductal epithelia reflect the degree of metabolic activity. J Histochem Cytochem. 2007;55:403–9. 15. Hsu S, Yamamoto T, Borke J, et al. Green tea polyphenolinduced epidermal keratinocyte differentiation is associated with coordinated expression of p57/KIP2 and caspase 14. J Pharmacol Exp Ther. 2005;312:884–90.
16. Joshi R, Tawfik A, Edeh N, McCloud V, Looney S, Ogbureke KUE. Dentin sialophosphoprotein (DSPP) gene-silencing inhibits key tumorigenic activities in human oral cancer cell line, OSC2. PLoS ONE. 2010;5:e13974. 17. Gasparoni A, Fonzi L, Schneider GB, et al. Comparison of differentiation markers between normal and two squamous cell carcinoma cell lines in culture. Arch Oral Biol. 2004;49:653–64. 18. Chang SE, Foster S, Betts D, Marnock WE. DOK, a cell line established from human dysplastic oral mucosa, shows a partially transformed non-malignant phenotype. Int J Cancer. 1992;52:896–902. 19. Goruppi S, Bonventre JV, Kyriakis JM. Signaling pathways and late-onset gene induction associated with renal mesangial cell hypertrophy. EMBO J. 2002;21:5427–36. 20. Chowdhury UR, Samant RS, Fodstad O, Shevde LA. Emerging role of nuclear protein 1 (NUPR1) in cancer biology. Cancer Metastasis Rev. 2009;28:225–32. 21. Jiang WG, Davies G, Kynaston H, Mason MD, Fodstad O. Does the PGC-1/PPARgamma pathway play a role in Com-1/p8 mediated cell growth inhibition in prostate cancer? Int J Mol Med. 2006;18:1169–75. 22. Jiang WG, Davies G, Martin TA, et al. Com-1/p8 acts as a putative tumour suppressor in prostate cancer. Int J Mol Med. 2006;18:981–6.
123
ORIGINAL PAPER
Expression of p8 in Human Oral Squamous Cell Carcinoma Christopher Bingham • Douglas Dickinson • James Cray • Komal Koli • Kalu U. E. Ogbureke
Received: 13 May 2014 / Accepted: 14 August 2014 Ó Springer Science+Business Media New York 2014
Abstract The present study investigated the expression of p8, a transcription factor upregulated in some human cancers, in oral squamous cell carcinomas (OSCCs). Immunohistochemical analysis of p8 expression was carried out on 20 archived surgical specimens of human OSCCs, and expression correlated with clinical outcome parameters in a retrospective study. Expression of p8 in a number of OSCC cell lines also was investigated by western blot and RT-PCR analyses. p8 was expressed in 80 % (16/20) of the samples with levels of expression exhibiting a significant difference (v2 = 8.352, df = 3, p = 0.039) by patient age. Furthermore, greater levels of p8 immunoreactivity was significantly associated with advanced tumor grade (p = 0.008). p8 also was upregulated in OSCC cell lines. p8 is expressed in a significant proportion of OSCCs,
Electronic supplementary material The online version of this article (doi:10.1007/s12105-014-0565-1) contains supplementary material, which is available to authorized users. C. Bingham Department of Periodontics, College of Dental Medicine, Georgia Regents University, Augusta, GA, USA D. Dickinson College of Graduate Studies, Georgia Regents University, Augusta, GA, USA J. Cray Department of Oral Biology, College of Dental Medicine, Georgia Reagents University, Augusta, GA, USA K. Koli K. U. E. Ogbureke (&) Department of Diagnostic and Biomedical Sciences, University of Texas School of Dentistry at Houston, 7500 Cambridge Street, SOD 5460, Houston, TX 77054, USA e-mail: [email protected]
and in human OSCC cell lines, suggesting a potential value of p8 as a diagnostic and/prognostic tool for oral cancers. Keywords p8 expression Oral squamous cell carcinoma Immunohistochemistry OSC2 cells
Introduction p8, a gene activated in rat pancreas during the acute phase of pancreatitis, is a transcription factor of the basic helixloop-helix (bHLH) family identified initially as playing a significant role in rat pancreatic development and regeneration. p8 also promotes cellular growth [1]. The human and mouse homologues of rat p8 subsequently were cloned and characterized, and the complete structure and mapping of human p8 to chromosome 16 at position p11.2 was determined [1, 2]. Three exons separated by two introns constitute the human p8 gene [1, 2]. The rat and mouse p8 polypeptides are each 80 amino acids long with a predicted molecular weight of 8kD, while the length of human p8 is 82 amino acids [1, 2]. The rat p8 shows 75 and 91 % homology with human and mouse p8 respectively [1, 2]. Published findings indicating that p8 is upregulated in a number of human cancers including breast, pancreatic, and lung cancer suggest that p8 might perform growth-promoting functions [3–5]. Reports of other studies by Vasseur et al. indicated that p8 might play a crucial role in tumor metastasis [6, 7]. For example, the identification of p8 in human breast cancer cell lines correlates with growthpromotion in early development and the ability to establish primary carcinomas [3–5]. However, its role in the later stages of tumorigenesis or metastasis of human breast cancer cells remains doubtful and unclear [3–5]. Interestingly, the p11.2 region on chromosome 16 (the location of
123
Head and Neck Pathol
human p8 gene) is frequently amplified in breast cancer [6, 7]. Other studies have reported high expression of p8 in pituitary tumors, and that RNAi silencing of p8 impaired pituitary tumorigenesis in mice [8]. Similarly, in vitro studies show that p8 mRNA is expressed in a human hepatoma HepG2 cell line in vitro [2]. Recently published evidence has suggested that p8 regulates autophagy and apoptosis and may be relevant to cardiovascular pathologies and cancers associated with dysregulated autophagy [9]. It also was determined to regulate DNA repair, thereby adding another layer to a chromatin regulatory network [10, 11]. Thus, these reports collectively indicate that p8 may play vital roles in metastasis, chemoresistance, survival, or apoptosis of, at least some carcinomas. However, the expression and potential functional implications of p8 in human OSCC have not been reported. Here, we report findings of our investigation of the expression of p8 in human OSCC.
Materials and Methods Case Selection for Immunohistochemistry The specific samples used in this study have been described in a previous publication [12]. A retrospective cohort analysis of the immunohistochemical expression of p8 in 20 cases of de-identified surgical biopsied human OSCC maintained in the archives of the Department of Pathology, Georgia Health Sciences University was performed. Cases selected were those from patients seen between January 2004 and December 2007. Thereafter, expression was correlated with clinical prognostic and outcome parameters of OSCC. The Inclusion criteria included the availability of a pathology record indicating a histologic diagnosis of OSCC on hematoxylin/eosin sections independently verified by two oral pathologists, and excision/resection of primary tumor following histopathologic diagnosis. The Exclusion criteria included a stipulation on a Pathology Request Form of the presence of neoplasm of other anatomic sites/systems at the time of diagnosis of primary OSCC, previous diagnosis of and treatment for OSCC, and prior history, diagnosis, or treatment for cancers of other organs/regions. The cases satisfying the inclusion criteria were randomly selected from archived cases in the Department of Pathology diagnostic database using the search words: ‘‘oral squamous cell carcinoma.’’ Two pathologists were blinded to the clinicopathologic details of each case including whether or not patients presented with recurrent disease, or post-surgical resection, at primary site. Cases were randomly selected until 20 cases satisfying the inclusion criteria were identified.
123
Histopathologic Analysis of Hematoxylin and Eosin Sections The presence of invasive OSCC on the primary resection/ excision specimen was determined by a combination of established architectural and cytologic parameters including: nuclear pleomorphism; increased and/or abnormal mitosis; hyperchromatism; basal cell hyperplasia; abnormal maturation sequence; and stromal invasion. Immunohistochemistry Five lm sections of human OSCC were obtained from archived paraffin blocks of tissues from cases selected and immunostained with p8 antibody using the Nemesis 7200 automated system supplied with Super-Picture-Perfect Broad-Spectrum HRP-Polymer and Single-Solution-AEC reagents from BioCare (San Francisco, CA) as described previously [12–14]. The p8 polyclonal antibody was a kind donation from Dr. Juan L. Iovanna (Marseille Cedex, France). In brief, after manually dewaxing sections in three 5-min xylene washes, and rehydrating through graded ethanol (100, 95, and 75 %) and water, endogenous peroxidase activity was prevented by treating the sections with 3 % hydrogen peroxide (in methanol) for 30 min. Sections thereafter were washed 3 times in phosphate-buffered saline (19 PBS) for at least 5 min each and covered with PBS ? 0.05 % Tween-20 (PBS-T) before loading the slides on to the preprogrammed Nemesis automated immunohistochemistry machine. Incubation of sections for 1 h with p8 antibody diluted in 10 % normal goat serum in PBS was performed before washing (4 9 1 min) with PBS-T followed by incubation with SuperPicTure Polymer HRP-conjugated broad-spectrum secondary antibody (# 87-8963, Zymed Lab. Inc., San Francisco, CA, USA) for 10 min. The chromogenic substrate, aminoethylcarbazole (AEC Single Solution, # 00-1122, Zymed Lab. Inc., San Francisco, CA, USA), was then applied for 2 min, rinsed in tap water, and manually counterstained with Mayer’s hematoxylin for 10 s. All steps were performed at room temperature. Finally, an overlay of Clearmount glaze was applied over the sections and left to dry overnight at room temperature before coverslips were applied with the aid of Histomount (Zymed Lab. Inc., San Francisco, CA, USA). A negative control consisted of the substitution of p8 with non-immune rabbit serum or mouse IgG control (# 08-6599, Zymed Lab. Inc., San Francisco, CA, USA), while a positive control consisted of human pancreatic tissue sections (known to be immunoreactive for p8) stained with p8 antisera. Representative photographic images were captured using the Axioplan2 Universal microscope equipped with an Axiovision digital camera and Axiovision program (Carl Zeiss Gmbh, Jena, Germany).
Head and Neck Pathol
Scoring of Immunohistochemistry Results Proportions of immunostaining for p8, based on the combined intensity of cytoplasmic staining (CS) and nuclear staining (NS), was assigned semi-quantitative scores by the two independent observers earlier calibrated for scoring criteria. Scoring was as: negative (0, immunoreactivity not detectable/faint staining in less than 10 % of tumor cells); 1 (10–50 % of immunoreactive tumor cells); 2 (50–75 % of immunoreactive tumor cells); and 3 (widely and intensely expressed in tumor cells). Statistical Analysis Non-parametric correlation, Kendall Tau, Fisher’s exact test, Kruskal–Wallis and ordinal regression were utilized to determine trends and significant differences for the response variable, p8 immunostain (p8IHC) score, by the independent or predicting variables, Staining Pattern, Subsite, Age, Gender, Ethnicity, TNM stage, Lymphatic and Vascular Invasion (LVI), Perineural (PN) spread, Radiotherapy (RT) and Recurrence, respectively. Odds ratios, where informative, were also included. Differences were considered significant if p \ 0.05. All analyses were performed using SPSS 19.0 (IBM Corporation) and SAS 9.1 (SAS Institute). All graphs were created in SPSS 19.0. Human Cell Lines and Culture Conditions SCC25 and OSC2 (OSCC), and DOK (dysplastic oral keratinocytes), and A818-6 (Human Pancreatic Adenocarcinoma) human cell lines have been published [15–18] and were initially obtained from American Type Culture Collection. The human oral keratinocyte (HOK) whole lysate was purchased from ScienCellTM Research Laboratories (cat. #2616; San Diego, CA). Endogenous expression levels of p8 protein and mRNA were evaluated by Western blot and semiquantitative-reverse transcriptase (RT)-PCR analyses using whole-cell lysates and total RNA extracts, respectively, from the above cell lines. The human pancreatic adenocarcinoma whole cell lysate and extract was used as positive control. The cell lines were routinely cultured as monolayers in DMEM/F12 medium containing 10 % FBS (Invitrogen, Carlsbad, CA) supplemented with 1 % Penicillin/Streptomycin and 500 ng/ml Hydrocortisone (Sigma Aldrich, St. Louis, MO) and maintained in the presence of 5 % CO2 humidified air at 37 °C. Semiquantitative RT-PCR Following the manufacturer’s instructions, total RNA was extracted from a third of the cell pellets from each group (experimental and control) using the RNEasy Plus Mini kit
(Qiagen, CA). 1 lg of total RNA was used to produce cDNA by reverse-transcription using the High Capacity Reverse Transcription Kit (Applied BioSystems, CA) in a conventional Thermo cycler. PCR reaction was performed using the Promega GoTaq Green MasterMix reagent (cat. #M7122; Promega, Madison WI), and the p8 primer pair (sc-40792-PR; exact primer sequences are proprietary and therefore were not available from company; Santa Cruz Biotechnology), and 100 ng of cDNA in a 25 ll PCR reaction. The primer design amplified a 407 base amplicon. GAPDH was used as a normalizing control. At the end of the reaction, PCR products were electrophoresed on a 2 % agarose gel in Tris–acetate-EDTA (TAE) buffer, stained with ethidium bromide, and photographed. Densitometric analysis of p8 mRNA levels was performed and results represented by a histogram. Western Blot Analysis Total cell lysates prepared from 2/3rd of the cell pellets from each group (experimental and controls) were resolved on SDS–PAGE gel using the Mini-Protean Tetra Cell unit at 100 V (Bio-Rad, CA) by standard protocol. Briefly, 25 lg of protein was loaded in each lane and resolved on 7 % polyacrylamide gels before transferring to polyvinylidene (PVDF) membranes (Millipore, MA) at 100 mA constant current for 75 min in the Bio-Rad Transfer unit. The membrane was then incubated in blocking solution consisting of 1:1 diluted Sea Block Blocking solution (LICOR, Biosciences, Lincoln, Nebraska) in 19 PBS, for 1 h at RT. Membrane was thereafter incubated in anti-p8 polyclonal antibody and anti-b-actin monoclonal antibody (Sigma, St. Louis, MO) diluted in blocking buffer (1:1,000 and 1:5,000, respectively) overnight at 4 °C before washing with 19 PBS-T for 5 9 5 min. The membrane was then incubated with IRDye 600 goat anti-mouse secondary antibody (L-ICOR Biosciences) at 1:5,000 dilution for 1 h at RT, washed with 19 PBS-T 4 9 5 min, and with 19 PBS for 5 min. Signal was detected using the Infrared LICOR Imaging system (LI-COR Biosciences).
Results p8 Expression in Primary OSCCs The cellular expression of p8 was determined by immunohistochemistry in successive paraffin slides from each of the 20 OSCC cases selected for this study. Representative immunoreactivity results are shown in Fig. 1 with 1A representing a semiquantitative score of ‘‘1’’, 1B a score of ‘‘2’’, and 1C a score of ‘‘3’’. Figure 1d show a lack of expression of p8 in normal oral mucosa obtained from
123
Head and Neck Pathol
Fig. 1 Representative p8 immunoreactivity of human OSCC sections. a A semiquantitative score of ‘‘1’’ indicating a \10 % population of tumor cells positive for p8, while (b) illustrates a score of ‘‘2’’ where more than 50 % but \75 % of tumors cells are immunoreactive to p8. c Score of ‘‘3’’ indicating more than 75 %
diffuse immunoreactivity of tumor cells. d Lack of expression of p8 in normal oral mucosa, whereas e is a representative pre-immune negative control on OSCC. Pink–Brown stain (arrow) represents positive immunoreactive tumor cells for p8. (Images are at mag. 9 10 and corresponding insets for a–c, e are at mag. 9 40)
surgical wastes of non-cancer patients, whereas Fig. 1e is a representative negative control (non-immune serum) in OSCC section. In some tumor cells, p8 immunoreactivity was confined to the cytoplasm and the perinuclear perimeters, while in others staining of the nucleus was evident. The connective tissue stroma in all cases showed negative immunoreactivity for p8. Table 1 summarizes the clinical data of patients and the semiquantitative scores for p8 immunoreactivity of the study samples. As shown, p8 is expressed in 16 (80 %) of the 20 cases of OSCCs studied.
v2 = 8.352, df = 3, p = 0.039). A post hoc Nemenyi test showed marginal differences upon multiple comparisons (p8 expression levels 1–2, and 1–3 approached p = 0.05). This is most likely due to the unequal samples sizes, but may be indicative of a type 1 error. Figure 3 represent the cumulative probabilities level of p8 expression by each of the predictors. Of note is the ordinal regression models showing a significant correlation (p = 0.008) between level of p8 immunoreactivity and the histologic differentiation status of OSCC (well-, moderate-, poorly-differentiated) as seen on hematoxylin and eosin (HE-DIFF) stained sections. This suggests a relationship between p8 expression and OSCC differentiation status. Thus, poorlyand moderately-differentiated OSCCs upregulate p8 more than their well differentiated counterparts. No other patient characteristics or clinical variables were found to be significantly associated with p8 expression. As previously reported for this same cohort of the 20 cases [12], 9 (45 %) had clinical documentation of recurrence of OSCC (Table 1); a rate not significantly different from published estimated population recurrence-rate of 50 % (p = 0.664) [12]. Time between initial surgical
p8 Expression and Patients Clinical Variables Table 2 shows the results of a nonparametric bivariate (Kendall Tau) correlation of all variables with p8 immunoreactivity, while Table 3 summarizes the results for the Fisher’s Exact test for each variable compared to p8 immunoreactivity. Table 4 represents the results of the ordinal regression modeling for p8 immunoreactivity by each variable, 1 factor ordinal regression. With respect to mean age, the level of p8 expression as shown in Fig. 2 showed a significant difference (Kruskal–Wallis Test
123
Head and Neck Pathol Table 1 Semiquantitative scores for p8 immunoreactivity on oral squamous cell carcinoma (N = 20) Serial no.
Histologic characteristics of primary tumor (PT) HEDIFF
Clinical characteristic/parameters of patients
P8 IHC scorec
Staining pattern
Oral subsite location
Age/gender/ ethnicitya
TNM stage
LV invasion
PN spread
RTb
Recurrence N
1
WD
3
CS ?NS
TG
58//M/AA
0
1
1
0
2
PD
1
CS
FOM
63/M/C
0
0
0
0
N
3
PD
2
CS ?NS
TG
44/M/C
2
0
0
0
Y (17mo)
4
PD
0
NS
TG
54/F/AA
0
0
1
0
N
5
PD
0
NS
BOT
51/M/C
0
0
0
0
Y (24mo)
6
WD
2
NS
TG
47/F/AA
2
1
2
0
N
7
MD
1
NS
TG
50/M/C
0
1
0
0
N
8
PD
1
NS
FOM
53/M/C
2
0
0
0
N
9
PD
2
NS
SP
45/F/AA
3
1
0
0
N
10
WD
2
CS
TG
48/M/C
2
1
0
0
N
11
MD
2
CS
BM
52//F/other
2
0
1
0
Y (11.5mo)
12
MD
3
CS
FOM
85/F/C
2
0
1
1
N
13
PD
1
NS
BM
68/F/C
2
1
1
0
Y (13mo)
14
MD
3
NS
TG
68/F/C
0
0
0
2
Y (12mo)
15
PD
3
NS
TG
51/F/AA
1
1
1
0
Y (9mo)
16
WD
2
CS ?NS
RMT
53/F/AA
0
1
0
3
Y (7mo)
17
PD
1
NS
BM
58/F/C
3
0
1
0
N
18
WD
0
NS
RMT
49/M/C
2
1
1
3
N
19
PD
0
NS
FOM
68/F/C
2
1
1
0
Y (7mo)
20
MD
2
NS
FOM
54/M/C
0
1
1
0
Y (6mo)
F female, M male, C caucasian, AA African American, N absence of recurrence at least 48 months post-primary treatment, LV presence (1) or absence (0) of lymphatic and/vascular invasion, PN spread presence (1) or absence (0) of perineural spread, RT number of times adjunct radiotherapy was given post-surgical treatment Oral Subsites: BM buccal mucosa, FOM floor of mouth, RMT retromolar trigone, BOT base of tongue, TG lateral tongue Degree of tumor differentiation on hematoxylin and eosin stain (HE-DIFF): WD well-differentiated, MD moderately-differentiated, PD poorlydifferentiated a b c
Mean age = 56 years Average recurrence free-time (ARFT) = 10.5 months Immunohistochemical scoring was semiquantitative using prior published criteria; p8 Staining pattern: CS cytoplasmic, NS nuclear
resection of primary tumor and initial recurrence (Recurrence-Free Time; RFT) ranged from 6 to 17 months with an average RFT (ARFT) of 10.5 months (Table 1). The correlation values for Recurrence approached 0.40, considered as a moderate correlation. p8 Upregulation in OSCC and Oral Dysplastic Epithelial Cell Lines Further investigation of p8 expression in two human OSCC cell lines (OSC2 and SCC-25), the human dysplastic oral keratinocyte cell line (DOK), and the human primary oral keratinocyte (HOK) cells by Western blot (WB) and semiquantitative reverse transcriptase (RT)-PCR analyses showed significant upregulation in OSC2, SCC25 and
DOK compared with HOK cells at both the translational (Fig. 4a) and transcriptional (Fig. 4b) levels. Positive control consisted of whole-cell lysates and total RNA extracts from pancreatic cancer cell line, A818-6. With respect to the oral cancer cell lines, these in vitro findings are consistent with immunohistochemistry results (Fig. 1) indicating up-regulation of p8 in OSCC specimens from patients.
Discussion The aims of this study were to determine the expression of p8 in human OSCC. This is with a view to laying the foundation for extended inquiry into the functional and
123
Head and Neck Pathol
mechanistic role of p8 in the biology of oral cancer. A search of the English language literature suggests that the present report represents the first demonstrating the Table 2 Correlation of variables with p8 immunoreactivity Variable
Kendal Tau
p value
HE-DIFF
0.27
0.20
Stain
0.28
0.21
Subsite
0.33
0.10
Age
0.05
0.80
Gender
0.10
0.64
Race
0.08
0.70
TNM
0.01
0.97
LV
0.11
0.60
PN
0.07
0.72
RT
0.14
0.51
Recurrence
0.36
0.11
upregulation of p8 in a significant proportion of OSCC resection specimens as well as in the OSCC cell line, OSC2 and SCC-25, and in DOK cells but absent in normal oral mucosal tissue (as evidenced by immunohistochemistry), although basal levels are seen in HOK cells by western blot and RT-PCR analyses. Our observation that p8 is at all expressed by HOK cells (albeit at basal level) was unexpected because of an assumption that the HOK cells are comparable to the normal oral mucosa, which lacked p8 expression. Since p8 is also known to be a stress-related protein whose expression may be enhanced under stress conditions [19], we speculated that the low level expression of p8 in HOK cells may be in response to cell cultureinduced stress. Except for histologic differentiation status, p8 expression and/or levels, as determined by immunohistochemistry, pointed to no apparent correlation with clinical outcome in OSCCs patients.
A nonparametric bivariate correlation by way of Kendall Tau was performed for all variables with p8 immunoreactivity. Note that there are no significant correlations detected. However, the correlation value for recurrence is of note. HE-DIFF histological differentiation; TNM tumor size (T), presence/absence of regional node metastasis (N), and presence/absence of distant metastasis (M), LV lymphatic and vascular invasion, PN perineural invasion, RT radiotherapy
Table 3 Fisher’s exact test for variable compared to p8 immunoreactivity
Distribution of values for p8 immunoreactivity was investigated by each independent variable by way of Fisher’s Exact test. Note that there are no significant differences in distribution of p8 immunoreactivity detected
Variable
p value
HE-DIFF
0.21
Stain Subsite
0.88 0.64
Gender
0.92
Race
0.91
TNM
0.92
LV
0.88
PN
0.26
RT
0.36
Recurrence
0.16
Fig. 2 Plot of mean age (± SD) by p8-IHC category. Shows a significant difference (v2 = 8.352, df = 3, p = 0.039)
Table 4 Ordinal regression modeling for p8 immunoreactivity Cox snell R2
Variable
Model
Goodness of fit
AIC
Parallel lines
HE-DIFF
v2 = 12.807, p = 0.008
v2 = 2.121 p = 0.908
57.019
v2 = 12.807, p = 0.046
0.463 0.368
2
Stain
v = 7.332, p = 0.291
v = 2.564 p = 0.861
37.596
v2 = 7.524, p = 0.736
Subsite
v2 = 4.688, p = 0.455
v2 = 9.632, p = 0.473
55.372
v2 = 13.070, p = 0.220
0.219
2
2
Gender
v = 0.239, p = 0.625
v = 0.554 p = 0.758
57.606
v2 = 0.554, p = 0.758
0.012
Race
v2 = 0.169, p = 0.919
v2 = 3.024 p = 0.554
57.682
v2 = 3.659, p = 0.454
0.009
2
2
TNM
v = 2.342, p = 0.719
v = 4.228, p = 0.646
57.884
v2 = 5.188, p = 0.520
0.068
LV
v2 = 0.268, p = 0.875
v2 = 3.543, p = 0.471
57.583
v2 = 4.349, p = 0.361
0.014
2
2
2
2
PN
v = 3.072, p = 0.381
v = 8.590 p = 0.198
57.528
v = 7.223, p = 0.301
0.149
RT
v2 = 7.560, p = 0.056
v2 = 1.132 p = 0.970
56.016
v2 = 1.949, p = 0.924
0.343
Recurrence
v2 = 3.313, p = 0.069
v2 = 3.364, p = 0.186
17.252
v2 = 4.295, p = 0.117
0.168
123
Head and Neck Pathol
Fig. 3 Cumulative probabilities for each independent RT (a), Recurrence (b), HE-DIFF (c), Staining characteristics (d), oral subsite (e), Gender (f), Ethnicity (g), TNM (h), LV (i), and PN (j) variable plotted by p8 IHC (semiquantitative p8 immunostaining) category. Note the difference for Recurrence by p8 IHC category, specifically category 3. HE-DIFF = Degree of histologic differentiation as seen on hematoxylin and eosin stained sections: 1 = Well-
Differentiated (WD); 2 = Moderately-Differentiated (MD); 3 = Poorly-Differentiated (PD). N = Absence of recurrence at least 48 months post-primary treatment; LV = presence (1) or absence (0) of lymphatic and/vascular invasion; PN spread = presence (1) or absence (0) of perineural spread; RT = number of times adjunct radiotherapy was given post-surgical treatment
Previous reports on the expression of p8 in cancers, other than oral cancer, and correlation of expression with outcome and prognostic parameters of these cancers have provided conflicting and paradoxical results. Thus, several antitumor functions attributable to p8 have been reported in spite of its notable tumor-promoting activity. Except for marginal correlation with recurrence shown by the ordinal regression and cumulative probabilities analysis, results of our present study indicated no apparent correlation of p8 expression with outcome parameters. The production of multiple and even opposing effects by p8, despite its relatively small size (8-12kd) and the absence of known interaction domains, make it a puzzling biological molecule. It promotes tumor growth and aggressiveness and protects tumor cells from apoptosis; yet
at the same time it can act as a tumor suppressor [1, 2, 19]. In breast cancers, for example, p8 expression resulted in poor disease outcome [20], whereas upregulation in prostate cancers was associated with favorable disease outcome [20]. In human prostate cancer cells, p8 expression is lower compared with levels in normal epithelial cells [21, 22]. Furthermore, expression levels inversely correlated with the invasiveness and growth potential of prostate cancer cells in vitro, and overexpression of p8 reduced the growth rate of prostate tumors in vivo [21, 22]. Thus, p8 was proposed to behave as a potential tumor suppressor for human prostate cancers; an effect partially mediated by the interaction of p8 with the peroxisome proliferator-activated receptor-c coactivator PGC-1 [21, 22]. Our observation in the present study that poorly-differentiated (PD) OSCCs,
123
Head and Neck Pathol
Fig. 4 a Western blot (WB) and densitometric analyses show over twofold increase in the levels of p8 expression in OSC2, SCC25, and DOK cells compared with levels in HOK cells. The pancreatic carcinoma cell human A818-6 known to express abundant p8 was used as positive control. Normalization was with b-actin. b Semiquantitative RT-PCR analysis shows p8-mRNA expression in OSC2, SCC25, DOK, and HOK cells were proportionately comparable with
WB protein levels in (a). Normalization of RT-PCR was with GAPDH. Cells used in study: OSC2 is a cell line from a regional lymph node metastases of human tongue primary SCC; SCC25 is a human tongue primary SCC; DOK is a human oral epithelial dysplastic cell line from the tongue; and A818-6 is a human pancreatic adenocarcinoma cell line
which also tend to be more aggressive, downregulate p8 seems to fit these paradox models observed with prostate, breast, and pancreatic cancers discussed above. In summary, the data presented in this study show that p8 is expressed in a significant proportion of OSCCs as well as in OSCC carcinoma cell lines, although expression levels do not appear to impact clinical outcome and prognostic parameters. Any ascribed role for p8 in oral cancer is significant because, as a secreted protein, it may represent an ideal target for therapeutic intervention. While a detailed study of the mechanism of p8-oral cancer tumorigenesis was beyond the scope of this study, the data provides a framework for future studies elaborating on a p8 mechanistic network involved in oral cancer biology. This, in turn, is with a view to identifying target points for the design of potent biomimetics for intervention in oral cancer.
2. Vasseur S, Mallo GV, Garcia-Montero A, et al. Structural and functional characterization of the mouse p8 gene: promotion of transcription by the CAAT-enhancer binding protein alpha (C/ EBPalpha) and C/EBPbeta trans-acting factors involves a C/EBP cis-acting element and other regions of the promoter. Biochem J. 1999;343:377–83. 3. Ree AHM, Tvermyr O, Engebraaten M, et al. Expression of a novel factor in human breast cancer cells with metastatic potential. Cancer Res. 1999;59:4675–80. 4. Ree AH, Pacheco MM, Tvermyr M, Fodstad O, Brentani MM. Expression of a novel factor, com1, in early tumor progression of breast cancer. Clin Cancer Res. 2000;6:1778–83. 5. Pommier RM, Gout J, Vincent DF, et al. The human NUPR1/P8 gene is transcriptionally activated by transforming growth factor b via the SMAD signalling pathway. Biochem J. 2012;445:285–93. 6. Courjal F, Theillet C. Comparative genomic hybridization analysis of breast tumors with predetermined profiles of DNA amplification. Cancer Res. 1997;57:4368–77. 7. Gruel N, Lucchesi C, Raynal V, Rodrigues MJ, et al. Lobular invasive carcinoma of the breast is a molecular entity distinct from luminal invasive ductal carcinoma. Eur J Cancer. 2010;46:2399–407. 8. Mohammad HP, Seachrist DD, Quirk CC, Nilson JH. Reexpression of p8 contributes to tumorigenic properties of pituitary cells and appears in a subset of prolactinomas in transgenic mice that hypersecrete luteinizing hormone Mol. Endocrinol. 2004;18:2583–93. 9. Kong DK, Georgescu SP, Cano C, et al. Deficiency of the transcriptional regulator p8 results in increased autophagy and apoptosis, and causes impaired heart function. Mol Biol Cell. 2010;21:1335–49.
References 1. Mallo GV, Fiedler F, Calvo EL, Ortiz EM, et al. Cloning and expression of the rat p8 cDNA, a new gene activated in pancreas during the acute phase of pancreatitis, pancreatic development, and regeneration, and which promotes cellular growth. J Biol Chem. 1997;272:32360–9.
123
Head and Neck Pathol 10. Gironella M, Malicet C, Cano C, et al. p8/nupr1 regulates DNArepair activity after double-strand gamma irradiation-induced DNA damage. J Cell Physiol. 2000;221:594–602. 11. Sambasivan R, Cheedipudi S, Pasupuleti N, Saleh A, Pavlath GK, Dhawan J. The small chromatin-binding protein p8 coordinates the association of anti-proliferative and pro-myogenic sproteins at the myogenin promoter. J Cell Sci. 2009;122: 3481–91. 12. Ogbureke KUE, Weinberger PM, Looney SW, Fisher LW. DSPP, OPN, and MMP-9 expression at histologically-negative resection margins predict Oral squamous cell carcinoma recurrence. Oncotarget. 2012;3:286–98. 13. Ogbureke KUE, Nikitakis NG, Warburton G, Ord R, Sauk JJ, Fisher LW. Upregulation of SIBLING proteins and correlation with cognate MMP expression in oral cancer. Oral Oncol. 2007;43:920–32. 14. Ogbureke KUE, Fisher LW. SIBLING expression patterns in ductal epithelia reflect the degree of metabolic activity. J Histochem Cytochem. 2007;55:403–9. 15. Hsu S, Yamamoto T, Borke J, et al. Green tea polyphenolinduced epidermal keratinocyte differentiation is associated with coordinated expression of p57/KIP2 and caspase 14. J Pharmacol Exp Ther. 2005;312:884–90.
16. Joshi R, Tawfik A, Edeh N, McCloud V, Looney S, Ogbureke KUE. Dentin sialophosphoprotein (DSPP) gene-silencing inhibits key tumorigenic activities in human oral cancer cell line, OSC2. PLoS ONE. 2010;5:e13974. 17. Gasparoni A, Fonzi L, Schneider GB, et al. Comparison of differentiation markers between normal and two squamous cell carcinoma cell lines in culture. Arch Oral Biol. 2004;49:653–64. 18. Chang SE, Foster S, Betts D, Marnock WE. DOK, a cell line established from human dysplastic oral mucosa, shows a partially transformed non-malignant phenotype. Int J Cancer. 1992;52:896–902. 19. Goruppi S, Bonventre JV, Kyriakis JM. Signaling pathways and late-onset gene induction associated with renal mesangial cell hypertrophy. EMBO J. 2002;21:5427–36. 20. Chowdhury UR, Samant RS, Fodstad O, Shevde LA. Emerging role of nuclear protein 1 (NUPR1) in cancer biology. Cancer Metastasis Rev. 2009;28:225–32. 21. Jiang WG, Davies G, Kynaston H, Mason MD, Fodstad O. Does the PGC-1/PPARgamma pathway play a role in Com-1/p8 mediated cell growth inhibition in prostate cancer? Int J Mol Med. 2006;18:1169–75. 22. Jiang WG, Davies G, Martin TA, et al. Com-1/p8 acts as a putative tumour suppressor in prostate cancer. Int J Mol Med. 2006;18:981–6.
123
