Predictive Markers and Cancer Prevention
Increased expression of amyloid precursor protein in oral squamous cell carcinoma
Version of Record online: 20 MAY 2004
Copyright © 2004 Wiley-Liss, Inc.
International Journal of Cancer
Volume 111, Issue 5, pages 727–732, 20 September 2004
How to Cite
Ko, S.-Y., Lin, S.-C., Chang, K.-W., Wong, Y.-K., Liu, C.-J., Chi, C.-W. and Liu, T.-Y. (2004), Increased expression of amyloid precursor protein in oral squamous cell carcinoma. Int. J. Cancer, 111: 727–732. doi: 10.1002/ijc.20328
- Issue online: 8 JUL 2004
- Version of Record online: 20 MAY 2004
- Manuscript Accepted: 15 MAR 2004
- Manuscript Received: 6 NOV 2003
- National Research Program for Genome Medicine. Grant Number: 91GMP004-1
- National Science Council, Taiwan. Grant Number: NSC 91-3112-B-075-001
- APP expression;
- oral carcinogenesis;
In our previous study, we identified amyloid precursor protein (APP) in an oral squamous cell carcinoma (OSCC)-enriching subtractive hybridization library. Our present study attempts to define the significance of APP expression in the genesis of OSCC. RT-PCR analysis showed increase in APP mRNA expression for more than 2-fold in 76% of OSCC (n = 55) relative to corresponding non-cancerous matched tissues (NCMT). The majority of esophageal SCCs also had increase in APP mRNA expression. OSCC patients exhibiting increase in APP mRNA expression had significantly lower survival rate compared to patients exhibiting the opposite status. Western blotting analysis identified APP751 and APP770 as the major APP isoforms in oral keratinocytes. A high correlation between mRNA and protein expressions of APP was noted in OSCC/NCMT pairs. Immunohistochemistry further showed a remarkable increase of APP in OSCC tissue relative to NCMT. Treatment with an antisense oligonucleotide against APP reduced cellular and secreted APP as well as growth in an OSCC cell line. Our study provides novel clues that APP expression is involved in the proliferation and carcinogenesis of OSCC. Correlated with such pathogenesis was the survival of its victims. The degree of APP expression could serve as an invaluable marker for oral carcinogenesis. © 2004 Wiley-Liss, Inc.
Amyloid precursor protein (APP) is a Type I integral membrane protein1 that is encoded by a gene containing 19 exons2 and is located in chromosome 21q21.3 Exons 7 and 8 of APP could be spliced to produce 100–135 kDa proteins including isoforms APP695, APP751 and APP770.4, 5APP695 expression is high in brain tissues whereas APP751 and APP770 polypeptides are widely expressed in non-neuronal cells.6 Mutant APP is related to the pathogenesis of Alzheimer's disease, but the role of APP in normal cells and carcinogenic process remains unclear.
APP has been linked to proliferation in a variety of cells. The secreted form APP (sAPP) is related to the growth of keratinocyte and thyroid cells.7, 8, 9, 10 Hoffmann et al.10 have demonstrated the growth-promoting effects of sAPP on skin keratinocytes. APP expression has also been linked to the malignant progression or growth of neuronal and colorectal carcinoma cells.11, 12, 13 sAPP also acts as a regulator of dendrite motility and melanin release in epidermal melanocytes and melanoma cells.14
Oral squamous cell carcinoma (OSCC) is the third most common malignancy in developing nations. It is also prevalent in Taiwan due to the geographically-linked habit of areca chewing.15, 16, 17 The prognosis for OSCC remains dismal, as more than a half of victims succumb to the disease progression and its complications after treatment. Identification of markers for OSCC should facilitate the monitoring or prevention of OSCC. Our previous subtraction hybridization assay had identified APP in an OSCC-enriching library.17 We verified the increase in expression of APP in the vast majority of OSCCs and examined the underlying mechanisms. Moreover, OSCC patients with increased APP mRNA expression had poor prognoses.
MATERIAL AND METHODS
Fifty-five pairs of OSCC and corresponding non-cancerous match tissue (NCMT) samples were obtained for analysis of mRNA expression as approved by an ethics review board. The ages of the patients ranged from 32–78 years with a mean of 49 years. The most common primary site was buccal mucosa (53%, 29 cases). In histopathological grading, 36% (20 cases) OSCCs were well-differentiated. Fifty-one percent (28 cases) of patients presented with lymph node metastasis. Forty-seven percent (26 cases) of patients had Stage IV tumor. The mean follow-up period was 24.4 ±2.1 months. cDNAs from 10 pairs of esophageal squamous cell carcinoma (ESCC) and their NCMT tissues were kindly provided by Dr. F-H Wong. A tissue array containing 32 primary OSCC and 19 corresponding NCMT tissues was used. The array had been generated from 0.3 × 0.3 cm2 rectangles of representative sample fraction.
RNA extraction and cDNA synthesis
Total RNA was extracted using a Tri-reagent® RNA isolation kit (Molecular Research Center, Cincinnati, OH). The RNA was treated with DNase I (Stratagene, La Jolla, CA) to remove any contaminating DNA. Five micrograms of total RNA were reversely transcribed to cDNA using oligo-dT(18) primer and Stratascript reverse transcriptase (Stratagene).
cDNA was amplified by PCR in 25 μl of reaction mixture containing 1 U of Prozyme DNA polymerase (Protech, Taipei), 0.4 mM dNTP, and 0.4 mM primers. Primers used are listed in Table I. Primer Set 1 differentially amplifies APP and generates amplicons of 657 bp, 599 bp and 432 bp for APP770, APP751 and APP695, respectively.10 The APP primer Set 2 amplifies a common region or both APP770 and APP751 isoforms and achieves a single band of 289 bps.18 Amplification of GAPDH was carried out as the control.19 The cDNA were pre-denatured for 5 min at 95°C, denatured at 95°C for 1 min, annealed at the predetermined temperature (Table I) for 1 min, and then extended at 72°C for 1 min for 26–30 cycles. The cycle number of PCR had been optimized by pilot studies to assure that the amplification was within the logarithmic phase. PCR products were resolved on a 2% agarose gel and visualized by an imaging system (Viber Lourmat, Marne La Valle, France). The densities of the signals were measured by a densitometer (Amersham, Piscataway, NJ). The relative APP mRNA expression was determined semi-quantitatively by normalizing with GAPDH expression. OSCC exhibiting an APP mRNA expression ratio of ≥2-fold compared to the corresponding NCMT was defined as having increased mRNA expression. At least 2 independent RT-PCR analyses were carried out on each sample to confirm the expression status.
|Name||Sequence (5′–3′)||Amplicon (bp)||Annealing temp (°C)|
|APP 1||forward||AATGTGGATTCTGCTGATGCGGAG||657, 599 or 432||57|
SYBR LightCycler real-time PCR assays were used to elucidate changes in APP mRNA expression. Primer pair 3 in Table I was used for the analysis. The standard procedures followed the protocols we established previously to achieve ΔΔCT (ΔCT-OSCC − ΔCT-NCMT) that represents the relative n-values in OSCC as compared to those in NCMT;18 ±2−n represented the difference(s) in APP mRNA expression in OSCC relative to that in NCMT (+, increase in mRNA expression). The quantitative data shown were means from triplicate experiments.
The OSCC cell lines, OECM-1 and SCC25, were grown in RPMI-1640 and DMEM/F12 (1:1) media (Life Technologies, Gaithersburg, MD) routinely supplemented with 10% FCS except in cases of specific experimental demands. Normal human oral keratinocyte (NHOK) was grown in low calcium keratinocyte serum free medium (KSFM) (Life Technologies). OC3 cells20 were cultured in a medium composed of DMEM containing 10% FCS and KSFM at a 1:2 ratio. Antibiotic regimens of 100 U/ml of penicillin and 100 μg/ml streptomycin together with 0.25 μg/ml amphotericin B were routinely added to the culture media.
Cell lysates and conditioned media were harvested. Conditioned media were concentrated 100× by vivaspin tube (Sartorius, Goettingen, Germany). Proteins (30 μg) were resolved by 10% SDS-page gel, and then transferred onto nitrocellulose membranes (PALL Corp., Ann Arbor, MI). Membranes were blocked with non-fat milk and then incubated with primary antibodies overnight at 4°C. Primary antibodies used were mouse monoclonal clone 22C11 for APP (Chemicon, Temecula, CA) at 1:2,000 dilutions and mouse monoclonal antibody for GAPDH (Chemicon) at 1:10,000 dilutions. The secondary antibody was horseradish peroxidase-conjugated anti-mouse antibody 1:1,000 (Amersham). The signals were detected by Western lightning chemiluminescence reagent plus kit (Perkin-Elmer, Wellesley, MA). The densities of the signals were measured by a densitometer (Amersham). Quantification of the APP signal was achieved by normalization with that of GAPDH or the number of cultured cells.
Immunohistochemistry was carried out on OSCC and NCMT in tissue arrays. After regular processing, sections were incubated with the anti-APP antibody, which was the same as that used for Western blotting, at a dilution of 1:200 at 25°C for 2 hr in a humid chamber. After rinsing with PBS, standard IHC staining was carried out using a LSAB2 streptavidin-biotin complex system (Dako, Carpinteria, CA) with AEC as the chromogen. The slides were counterstained with hematoxylin and mounted with Clearmount (Zymed, San Francisco, CA). An identical tissue array but without the addition of the primary antibody was processed the same was to serve as a negative control. The extent of immunoreactivity was independently scored based on the percentage of positive cells found in each arrayed tissue: <10%, absent, (−); 10–50%, weak, (+) and >50%, strong, (++).
Morpholino antisense treatment
The sequences of morpholino antisense-oligonucleotide (ODN) for APP and random-ODN (standard control) were 5′-AAACCGGGCAGCATCGCGACCCTGC-3′ and 5′-CCTCTTACCTCAGTTACAATTTATA-3′, respectively (Gene Tools, Philomath, OR). OECM-1 cells were cultured in T-75 flask till 80% confluency. Antisense-ODNs (10 μM) were used to transfect cells. The scrape delivery method, as outlined in the manufacturer's protocol (Gene Tools), was used to treat the cells. After treatment, OECM-1 cells were grown in serum-free medium for various time periods. A minor portion of cells was subjected to the trypan blue dye exclusion assay and BrdU incorporation assay. The majority of cells were for Western blotting analysis.
Trypan blue dye exclusion assay
Treated cells were seeded on 6-well plates at a concentration of 105/well and cultured for 24, 48 and 72 hr. Cell viability was determined by the ability of cells to exclude 0.5% trypan blue (Biological Industries, Te Aviv, Israel). An equal volume of trypan blue dye solution (0.1% w/v), PBS, and cell slurry were combined and allowed to sit for 5 min at room temperature. The sample was loaded onto a hemocytometer and the cells were scored as living or dead based on uptake of dye. The results shown are the means and SE of triplicate experiments.
BrdU incorporation assay
Treated cells were seeded on 8-well chamber slides at a density of 104/well and incubated for 48 hr. BrdU (Sigma, St. Louis, MO) was added to the culture medium to a final concentration of 10 μM, and labeling was carried out for 2 hr. Cells were then fixed in 95% ethanol, permeabilized with 0.1% Triton X-100, and a fluorescein-conjugated anti-BrdU mouse monoclonal antibody (Boehringer-Becton Dickinson, San Jose, CA) was used to detect the cells in the S phase. All slides were finally washed with PBS containing Hoechst 33258 (final concentration: 1 μg/ml; Sigma), rinsed with water and mounted. The fluorescent signal was visualized using an Axioscope microscope (Zeiss, Heinrich, Bunger, Germany). In each experiment, at least 300 cells were counted and compared. The results shown are the means and SE from triplicate experiments.
The t-test, Fisher's exact test and Kaplan-Meier survival analysis were carried out. Differences between the values were considered significant when p < 0.05.
APP mRNA expression
RT-PCR analysis was carried out in pairs on oral and esophageal tissue pairs to assess APP mRNA expression. PCR was carried out with APP primer Set 1 on selected cases to confirm the presence of both APP751 and APP770 in the samples. APP primer Set 2, which can amplify both isoforms to generate a single band, was used in subsequent analyses for better semi-quantitation. Representative results of oral tissue pairs exhibiting clear distinction in semi-quantification are shown in Figure 1a. Selected tissues exhibiting very weak APP mRNA expression in OSCC or NCMT (W-1∼W-10) that might cause difficulties in semi-quantitation were subjected to real-time RT-PCR analysis. Upper panel of Figure 1b shows the results of the RT-PCR and the lower panel represents the real-time quantitative data of the corresponding samples. In general, real-time RT-PCR measurements correlated well with changes in mRNA expression. Among the 55 oral tissue pairs analyzed by RT-PCR, 39 (70.9%) OSCC had increase in APP mRNA expression as they exhibited ≥2-fold higher APP mRNA expression than the corresponding NCMT. Figure 2 depicts the normalized mRNA expression ratios in each group of tissue analyzed by RT-PCR. A statistically significant difference was found between NCMT and OSCC (0.22 ± 0.04 vs. 0.54 ± 0.11, p < 0.01, paired t-test) (Fig. 2a). Increase in APP mRNA expression in ESCC compared to NCMT was also observed in 6 of 10 esophageal tissue pairs (Fig. 1c). A statistically significant difference was also found between NCMT and ESCC (0.37 ± 0.04 vs. 0.53 ± 0.04, p = 0.01, paired t-test) (Fig. 2b).
APP mRNA expression and clinicopathological parameters
A significant association of worse prognosis with OSCC subjects having increase in APP mRNA expression (increase cases in Fig. 3) was found (p = 0.01, Kaplan-Meier survival analysis, Fig. 3). The frequency of increased mRNA expression did not differ significantly with respect to age, differentiation status, lymph node metastasis and clinical stage (detailed analyses not shown).
APP protein expression
RT-PCR analysis and Western blotting were carried out on NHOK and multiple OSCC cell lines including OC3, OECM-1 and SCC25, to determine the expressions of APP. The OSCC cell lines had ∼1.8–∼3.5-fold increases in APP mRNA expression relative to that in the NHOK (control). The antibody against APP detected 2 isoforms (upper band for 770 and lower band for 751 in Fig. 4b) in cell lysates. The expression of APP751 was generally higher than that in APP770. Quantitative analysis indicated that the OSCC cell lines had higher APP protein expression than the NHOK cells by ∼1.5–∼3.3-fold. OC3 cells had the highest APP mRNA and protein expressions among the cell lines (Fig. 4a,b). To examine the correlation of mRNA and protein expression of APP in tissues, 3 oral tissue pairs were analyzed. Figure 4c shows that OSCC 16 and 19 had consistent increased expressions of APP at both the mRNA and protein levels. OSCC 20 had marked decrease in APP mRNA expression relative to the corresponding NCMT although the protein expression of APP was barely detectable in tumor tissue. The results also suggested a general agreement in protein and mRNA expressions with respect to APP. Interestingly, OSCC tissues might have remarkable increases in APP770 (Fig. 4d).
Immunohistochemistry was carried out with antibodies validated by Western blotting on tissue arrays containing 19 NCMT and 32 OSCC tissues. Immunoreactivity scoring was based on proportions of cells exhibiting cytoplasmic APP immunoreactivity. In 16 NCMT, faint APP immunoreactivity was detected in the lower segment of the epithelium and was accorded the grading of APP (+) or APP (++) (Fig. 5a). In contrast, 3 NCMTs were APP (−). Interestingly, all OSCC had extensive APP immunoreactivity scoring of APP (++) (Fig. 5b). The difference in the APP immunoreactivity between NCMT and OSCC tissues was significant (p < 0.05, Fisher's exact test). APP immunoreactivity was also observed in fibroblasts, endothelial cells and some inflammatory cells in stromal tissue.
APP and OECM-1 cell growth
OECM-1 cells having endogenous APP expression were treated with ODNs to access the importance of APP in OSCC proliferation. Western blotting showed that cellular APP and sAPP in OECM-1 cells were significantly downregulated ∼66% and ∼43% after antisense-ODN treatment for 48 hr (p < 0.001, t-test, Fig. 6). Trypan blue dye exclusion assay showed a significant reduction in OECM-1 viability of ∼19% was observed at 72 hr after treatment with APP antisense-ODN, (p < 0.01, t-test, Fig. 7a). There was no increase in cellular death observed in the supernatants after treatment. BrdU incorporation assay showed treatment with antisense-ODN for 72 hr led to an inhibition of proliferation in OECM-1 cells for ∼56% (p = 0.04, t-test, Fig. 7b). Antisense-ODN treatment of OC3 cells resulted only in an inhibition of cell viability of only ∼6%, with no remarkable changes in cellular APP and sAPP due most likely to redundancy in transfection (detailed analysis not shown).
sAPP has been linked to the growth of keratinocyte and thyroid cells.7, 8, 9, 10 The involvements of APP in non-neuronal cancer cells have been investigated only to a limited extent. No study has ever addressed the in vivo expression status of APP in cancer tissues. Our current study has provided, for the first time, molecular evidences that mRNA expression of APP is elevated prominently in OSCC and ESCC (Figs. 1,2).
Neuronal cells express APP695. Our analyses indicating the presence of mRNA for APP770 and APP751 but barely detectable APP695 mRNA in oral tissue, were compatible with findings in other epithelial cells.7, 10 That the upregulation of APP gene expression might through the increase in transcription or mRNA stability. Western blotting confirmed that APP751 and APP770 were expressed in oral keratinocytes and oral tissues with a notable increase of expressions in most OSCC cell lines and tissues (Fig. 4). Such differences in protein expression agreed with the results of mRNA analyses. A subsequent immunohistochemistry study found a tremendous cytoplasmic immunoreactivity of APP in OSCC (Fig. 5).
The function of APP in tumors other than neuronal is rather unclear.12 Hoffmann et al.10 have shown the growth-promoting effects of sAPP on skin keratinocytes. Schmitz et al.7 have demonstrated that sAPP can induce the migration of keratinocytes. Our present study indicated that OSCC patients with increase in APP mRNA expression had significantly worse prognosis than the contrasting group (Fig. 3). This is the first time that APP expression has been shown to be a potential prognostic marker in epithelial tumors. Our study, by demonstrating the significant reduction in the viability and proliferation of OECM-1 after the blockage of APP expression, provided clues for the role of APP in OSCC growth (Fig. 7). The cause–effect relationship was further supported by the decease in viability and proliferation occurring at 72 hr, whereas the reduction of APP can occur earlier at 48 hr (Fig. 6). A hypothetic APP receptor exists on cell membrane that is responsible for the autocrine activation of MAP kinase and the resulting phenotypic changes has been proposed.7 It was believed that the reduction of APP could result in the decrease of sAPP generation. We have proven the concordant decrease of cellular APP and sAPP in our knock-down system. This knock-down strategy might have therapeutic application for OSCC treatment.
Oncogenic stress could stimulate the PKC or Ras pathway. Both PKC and Ras pathways can upregulate APP expression.21, 22 Furthermore, PKC could activate α-secretase that cleaves APP to sAPP.21, 23 Although the complexity in the regulation of APP activity in epithelial cells has yet to be clarified, we have provided evidences substantiating that Alzheimer amyloid precursor protein was increased in OSCC. This could further the exploration of common regulatory pathways for oral carcinogenesis or new modalities for cancer intervention.
We gratefully acknowledge Professor Y.-C. Shum for English editing.
- 23Muscarine enhances soluble amyloid precursor protein secretion in human neuroblastoma SH-SY5Y by a pathway dependent on protein kinase C(alpha), src-tyrosine kinase and extracellular signal-regulated kinase but not phospholipase C. Brain Res Mol Brain Res 2002; 102: 62–72., , , , .