Original Paper Received: April 10, 2014 Accepted after revision: July 12, 2014 Published online: September 23, 2014

Pathobiology 2014;81:206–214 DOI: 10.1159/000366022

HuD Promotes Progression of Oral Squamous Cell Carcinoma Tomonori Sasahira a Miyako Kurihara a, b Kazuhiko Yamamoto b Nobuhiro Ueda b Chie Nakashima a, b Sayako Matsushima a Ujjal K. Bhawal a, c Tadaaki Kirita b Hiroki Kuniyasu a   

 

 

 

 

 

 

 

 

Departments of a Molecular Pathology and b Oral and Maxillofacial Surgery, Nara Medical University, Kashihara, and c Department of Biochemistry and Molecular Biology, Nihon University School of Dentistry at Matsudo, Matsudo, Japan  

 

 

Abstract Head and neck cancer, including oral squamous cell carcinoma (OSCC), ranks as the sixth most common malignancy worldwide. Overall 5-year survival rates of OSCC have not significantly improved during the past 3 decades and the 5-year survival rate is less than 50%. Several invasion grading systems have been employed in OSCC, however, their utility is still controversial. HuD belongs to the Hu protein family and acts as an RNA-binding protein involved in mRNA stability and translational regulation. Although HuD has a pivotal role for neuronal differentiation, the functional role of HuD in OSCCs is still unclear. In this study, we examined HuD expression in 82 OSCC cases. Expression of HuD was observed in 36.6% of OSCCs and significantly associated with histological differentiation, nodal metastasis and mode of invasion. HuD expression in high-metastatic HSC3 cells was high-

© 2014 S. Karger AG, Basel 1015–2008/14/0814–0206$39.50/0 E-Mail [email protected] www.karger.com/pat

er than in low-metastatic HSC4 cells, and inhibition of invasion ability and activation of caspase-3 were shown by HuD siRNA-treated HSC3 cells. Furthermore, we clarified that HuD regulates expression of vascular endothelial growth factor (VEGF)-A, VEGF-D, matrix metallopeptidase (MMP)-2 and MMP-9. These results suggest that HuD is a useful diagnostic and therapeutic target in OSCCs. © 2014 S. Karger AG, Basel

Introduction

Head and neck cancer, including oral squamous cell carcinoma (OSCC), ranks as the sixth most common malignancy worldwide [1] with an estimated 275,000 cases and 128,000 deaths annually [2]. It is the first leading cause of cancer death in southern Asia [3]. In the USA, approximately 34,000 patients are diagnosed with OSCC annually, representing about 3% of all newly diagnosed cancers [4], and mortality from OSCC is 3.7 per 100,000 deaths in Japan [5]. OSCC has a high potential for local invasion and lymph node metastasis. Hiroki Kuniyasu, MD, PhD Department of Molecular Pathology Nara Medical University 840 Shijo-cho, Kashihara, Nara 634-8521 (Japan) E-Mail cooninh @ zb4.so-net.ne.jp

Downloaded by: Freie Universität Berlin 149.126.78.65 - 7/9/2015 2:51:34 PM

Key Words Oral squamous cell carcinoma · Head and neck cancer · Cancer invasion · Invasive pattern · Yamamoto-Kohama classification · Angiogenesis

Materials and Methods Tumor Specimens Eighty-two formalin-fixed, paraffin-embedded primary OSCC specimens (64 from male and 18 from female patients, age range 45–84 years, mean 65.2, median 64.5) were randomly selected from Nara Medical University Hospital, Kashihara, Japan. All patients did not receive any preoperative therapy. Tumor staging and the histology of OSCCs were classified according to the TNM and WHO classifications, respectively. Medical records and prognostic follow-up data were obtained from the patient database managed by the hospital. All experiments with human samples were performed according to the ethical standards ex-

HuD in Oral Cancer

pressed in the Declaration of Helsinki and approved by the Ethical Committee of the Nara Medical University. Written informed consent was obtained from individual patients for use of their tissue samples. Classification of Tumor Invasion Pattern The mode of tumor invasion at the tumor-host border was also classified using the Yamamoto-Kohama (YK) classification, which includes five categories: YK-1, a well-defined borderline; YK-2, a less marked borderline; YK-3, invasion with cell nests and no distinct borderline; YK-4C, invasion with small cell nests, and YK4D, diffuse invasion with desmoplastic reaction [15, 16]. The YK classification is frequently used to predict progression, metastasis and prognosis in Japan [17, 18]. Each invasion pattern is shown in figure 1. Immunohistochemistry Consecutive 3-μm sections were cut from each block, and immunohistochemistry was performed as we described previously [5]. An immunoperoxidase technique was employed following antigen retrieval with microwave treatment (95 ° C) in citrate buffer (pH 6.0) for 45 min. After endogenous peroxidase block by 3% H2O2-methanol for 15 min, specimens were rinsed with phosphate-buffered saline (PBS) three times. Anti-HuD antibody (Santa Cruz Biotechnology Inc., Santa Cruz, Calif., USA) diluted by 0.5 μg/ml was used as the primary antibody. After being incubated for 2 h at room temperature, specimens were rinsed with PBS three times and treated for 1 h at room temperature with the secondary antibody peroxidase-conjugated anti-mouse (Medical & Biological Laboratories Co. Ltd., Nagoya, Japan) diluted to 0.5%. The specimens were then rinsed with PBS three times and color-developed with diaminobenzidine (DAB) solution (DAKO, Carpinteria, Calif., USA). After washing, the specimens were counterstained with Meyer’s hematoxylin (Sigma Chemical Co., St. Louis, Mo., USA). Immunostaining of all samples was performed under the same conditions of antibody reaction and DAB exposure. Immunoreactivity of HuD was classified according to Allred’s score (AS) [19] and we divided the immunoreactivity into 4 grades by AS: grade 0, AS 0; grade 1, AS approximately 2–4; grade 2, AS 5 or 6; grade 3, AS 7 or 8. Cases with grade 2 and 3 were regarded as immunologically positive [1, 6].  

 

Cell Culture Human OSCC cell lines, HSC3 and HSC4 cells were obtained from Health Science Research Resources Bank and maintained in  Dulbecco’s modified Eagle’s medium (DMEM; Wako Pure Chemical industries Ltd., Osaka, Japan) supplemented with 10% fetal bovine serum (Sigma Chemical Co., St. Louis, Mo., USA) under conditions of 5% CO2 in air at 37 ° C. HSC3 cells have high metastatic potential, while HSC4 cells have low metastatic ability [1, 6].  

 

Quantitative Reverse Transcription Polymerase Chain Reaction Total RNA was extracted using RNeasy Mini Kit (Qiagen Inc., Valencia, Calif., USA) and total RNA (1 μg) was synthesized with the ReverTra Ace qRT Kit (Toyobo, Osaka, Japan). Quantitative reverse transcription polymerase chain reaction (qRTPCR) was performed on StepOne Plus Real-Time PCR Systems

Pathobiology 2014;81:206–214 DOI: 10.1159/000366022

207

Downloaded by: Freie Universität Berlin 149.126.78.65 - 7/9/2015 2:51:34 PM

Over 80% of early-stage OSCCs can be cured by treatment, whereas more than 70% of advanced-stage OSCCs cannot. Overall, 5-year survival rates of OSCC have not significantly improved during the past 3 decades and the 5-year survival rate is less than 50% [5, 6]. Therefore, early detection is important and an understanding of the detailed molecular mechanism of OSCC is urgently required. We previously reported that the high-mobility group box-1 receptor for advanced glycation end products and melanoma inhibitory activity signal plays a pivotal role in tumor progression and nodal metastasis, and is implicated in a worse prognosis in OSCC [1, 3, 5, 6]. However, it remains unclear whether or not it is a specific diagnostic and therapeutic marker of OSCC, and further research is needed. Members of the Hu protein family, also known as the ELAVL (embryonic lethal, abnormal vision, drosophila, homolog-like) protein family, include HuR (gene symbol ELAVL1), HuB (gene symbol ELAVL2), HuC (gene symbol ELAVL3) and HuD (gene symbol ELAVL4) proteins [7]. Hu proteins act as RNA-binding proteins involved in mRNA stability and translational regulation. They contain three RNA recognition motifs, of which two domains are Au-rich elements present in the 3′-untranslated region and the third domain resides in the poly(A) tail [8, 9]. The target mRNAs of HuD include GAP-43, acetylcholine transferase, c-myc, N-myc, Notch3 and vascular endothelial growth factor (VEGF)-A, etc. [9–12]. HuD proteins normally express in neurons and associate with neuronal differentiation [7, 9]. They also express in small cell carcinoma and carcinoid tumor of the lung [13, 14]. However, the expression and functional role of HuD in OSCC remain unknown. In the present study, we examined the relationship between HuD expression levels and clinicopathological parameters using OSCC samples, and also the functional role of HuD based on in vitro analysis.

b

c

d

Color version available online

a

e

Fig. 1. Invasion pattern of OSCC. Invasion patterns of YK-1 (a), YK-2 (b), YK-3 (c), YK-4C (d) and YK-4D (e) by YK classification. Original magnification. ×100.

208

Pathobiology 2014;81:206–214 DOI: 10.1159/000366022

Small Interfering RNA Stealth Select RNAi (siRNA) for ELAVL4 (HuD) was purchased from Invitrogen (Carlsbad, Calif., USA). AllStars Negative Control siRNA (Qiagen Inc.) was used for the control. According to the provider’s instructions, 20 nM of siRNA was transfected with Lipofectamine 2000 (Invitrogen). Cell Growth Assay The cells were seeded at a density of 2,000 cells per well in 96-well tissue culture plates and incubated for 48 h at 37 ° C. Cell growth was assessed by MTT assay using the incorporation of  

Sasahira  et al.  

 

Downloaded by: Freie Universität Berlin 149.126.78.65 - 7/9/2015 2:51:34 PM

(Applied Biosystems, Foster City, Calif., USA) using TaqMan Fast Universal PCR Master Mix (Applied Biosystems) and analyzed using the relative standard curve quantification method. The PCR conditions were according to the manufacturer’s instructions and GAPDH mRNA level were amplified for the internal control. TaqMan gene expression assays of ELAVL4 (HuD), c-myc, N-myc, Notch3, VEGF-A, VEGF-C, VEGF-D, matrix metallopeptidase (MMP)-2, MMP-9 and GAPDH were purchased from Applied Biosystems. PCR products were electrophoresed on 2% agarose gel and all PCRs were performed in triplicate.

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (Sigma Chemical Co.). The experiments were performed in triplicate. In vitro Invasion Assay A modified Boyden chamber assay was performed using BD BioCoat Cell Culture Inserts glued to type IV collagen (BectonDickinson Labware, Bedford, Mass., USA), as described previously. Cells were suspended in 500 μl of DMEM and placed in the insert. After 48 h of incubation at 37 ° C, the filters were stained with hematoxylin. The stained cells were counted in whole inserts at 100× magnification. Each experiment was repeated at least three times.  

 

Caspase-3 Assay Activation of caspase-3 was detected using the CaspACE assay system, Colorimetric (Promega, Madison, Wisc., USA), according to the manufacturer’s protocol. The experiments were performed in triplicate. Immunocytochemistry The cells grown in a monolayer on the glass slides and were fixed for 24 h with 10% buffered formalin at 4 ° C and immunostained with anti-Ki-67 antibody (DAKO). Peroxidase-conjugated secondary antibody and DAB were employed for detection. For the counterstaining, Meyer’s hematoxylin was used. We observed 20 microscopic fields at 200-fold magnification and counted the tumor cells per case. The results are expressed as the percentage of tumor cells with positive nuclei, and these values are defined according to the Ki-67 labeling index.  

 

Statistical Analysis Statistical analysis was carried out with JMP8 (SAS Institute, Cary, N.C., USA). Statistical differentiation was calculated with Pearson’s r2 test and Student’s t test. Disease-free survival was calculated using the Kaplan-Meier method and differences between groups were tested by means of a log-rank test. Univariate analysis for disease-free survival was calculated by log-rank test. For multivariate analysis, a Cox proportional hazards model was used [described as the risk ratio with 95% confidence intervals (CI), together with the p value]. p values 65 years 29 Site Tongue 34 Other 18 Histological differentiation Good 43 Moderate/poor 9 T classification T1 6 T2 14 T3 19 T4 13 Clinical stage I 6 II 14 III 21 IV 11 Nodal metastasis Negative 37 Positive 15 Mode of invasion1 1 4 2 9 3 32 4C 4 4D 3

HuD-negative

0.8

positive 24 6

0.7458

18 12

0.1689

14 16

0.0975

18 12

0.0234

0.6

HuD-positive

0.4 0.2 p = 0.0033

2 6 9 13

0.3777

2 6 13 9

0.6853

11 19

0.0023

2 3 8 9 8

0.0015

0

0

400 800 1,200 Disease-free survival (days)

1,600

Fig. 3. Disease-free survival curve of HuD-negative or HuD-posi-

The relationship between the expression of HuD and parameters were calculated by Pearson’s χ2 test. T classification and clinical stage were classified according to the TNM classification. 1 According to the YK classification.

Discussion

HuD belongs to the Hu protein family and plays a pivotal role in neuronal differentiation [7, 9]. Although pulmonary neuroendocrine tumor is expressed by HuD, such as small cell carcinoma and carcinoid tumor [13, 14], the role of HuD in other tumors has not been well documented. In this study, we revealed that HuD promotes nodal metastasis and diffuse invasion of the OSCC. HuD is also related with histological grade and poor OSCC prognosis. Although it was recently reported that HuD in Oral Cancer

1.0

p value

tive OSCC cases. Disease-free survival was calculated by the Kaplan-Meier method.

the other Hu family protein, HuR, localizes not only the nucleus, but also the cytoplasm by extranuclear transportation [21], we did not observe cytoplasmic localization of HuD in OSCC, and we also confirmed no expression  of  HuD by immunoblotting using extranuclear protein extracted from OSCC cells (data not shown). Thus, we surmised that HuD is not able to transport to the extra nucleus in OSCC, while further studies of extranuclear transportation of HuD in OSCCs needs to be validated. HSC3 cells possess high metastatic ability, whereas HSC4 cells have low metastatic potential [22]. Compared to HSC4 cells, HSC3 cells are characterized by adhesion to type-IV collagen, colony formation in a type-I collagen matrix [22], upregulated MMP-2/9 [20] and higher expression of VEGF-A/C/D [1, 6]. In this study, HSC3 cells showed a higher expression of HuD than HSC4 cells. Although growth ability was not affected by HuD siRNA treatment, HuD regulated invasion of the HSC3 cells. De Giorgio et al. [23] reported that the number of apoptosis cells is increased by antiHuD antibody treatment in neuroblastoma cells. We also ascertained the increase of caspase-3 activity by HuD siRNA treatment in HSC3 cells, and this result supports their data on the whole. HuD is an RNA-binding protein and it is known that the target genes of HuD are c-myc, N-myc, Notch3 and VEGF-A, etc. [10–12], and we investigated whether expression of these factors Pathobiology 2014;81:206–214 DOI: 10.1159/000366022

211

Downloaded by: Freie Universität Berlin 149.126.78.65 - 7/9/2015 2:51:34 PM

HuD expression, n

Survival rate

Parameters

Color version available online

Table 1. Relationship between HuD expression (n  = 82 OSCC specimens) and clinicopathological parameters

d

1.5

*

1.0 0.5 0

HSC3

35 30 25 20 15

*

10 5

b

HSC4

HuD siRNA

Negative siRNA

e

120

50

100

40

80

Ki-67 LI

Relative cell growth (%)

2.0

Relative caspase-3 activity (%)

RQ (HuD/GAPDH) Invading cells per well

a

2.5

60 40

100 90 80 70 60 50 40 30 20 10 0

20 10

20 0

30

HuD siRNA

Negative siRNA

c

0

HuD siRNA

Negative siRNA

*

HuD siRNA

Negative siRNA

Fig. 4. Effect of knockdown of HuD in OSCC cells. a HuD mRNA expression in HSC3 and HSC4 human OSCC cells. Effect of cell growth ability (b), Ki-67 labeling index (c), invasion ability (d) and activation of caspase-3 (e) by HuD siRNA-treated HSC3 cells. RQ = Rela-

tive quantification; LI = labeling index. * p < 0.05.

Table 2. Univariate and multivariate analysis of disease-free survival

Age (65 years) Gender (male vs. female) Site (tongue vs. other) Histology (good vs. moderate/poor) T factor (T1–2 vs. T3–4) Stage (I–II vs. III–IV) Nodal metastasis (negative vs. positive) YK classification (1–3 vs. 4C, 4D) HuD (negative vs. positive)

are regulated by HuD. Moreover, we examined the activation of VEGF-C/D or MMP-2/9 because significant correlation was found between HuD expression and nodal metastasis, mode of invasion and invasive ability, respectively, and HSC3 cells showed higher expression of VEGF-C/D and MMP-2/9 [1, 6, 20]. This investigation clarified that HuD regulates not only activation of VEGF-A, but also VEGF-D, MMP-2 and MMP-9 in HSC3 cells. Furthermore, overexpression of HuD was 212

Pathobiology 2014;81:206–214 DOI: 10.1159/000366022

Univariate analysis

Multivariate analysis

p value

p value

0.822 0.9394 0.5485 0.0017 0.0789 0.0895

HuD promotes progression of oral squamous cell carcinoma.

Head and neck cancer, including oral squamous cell carcinoma (OSCC), ranks as the sixth most common malignancy worldwide. Overall 5-year survival rate...
4MB Sizes 4 Downloads 8 Views