Synaptonemal complex protein 3 is associated with lymphangiogenesis in non-small cell lung cancer patients with lymph node metastasis

H146, H460, H1299, H1666, H2228, H358, and H3122 cell lines were obtained from American type culture collection (ATCC, Manassas, VA). All cells were maintained in RPMI-1640 supplemented with 10% FBS in the presence of 5% CO₂ at 37 °C in a humidified incubator. H358/no insert, H358/SCP3, H1666/no insert and H1666/SCP3 cells were generated by retroviral transduction with pMSCV/empty (Clontech, Mountain View, CA) or pMSCV/SCP3 as previously described [11].

Synthetic small interfering RNAs (siRNAs) specific for green fluorescent protein (GFP) or human SCP3 (hSCP3) were purchased from the Invitrogen (Invitrogen, Carlsbad, CA); GFP, 5′-GCAUCAAGGUGAACUUCAA-3′ (sense), 5′-UUGAAGUUCACCUUGAUGC-3′ (anti-sense); hSCP3, 5′-GGAGAAGAAUCAUGAUAAU-3′ (sense), 5′-AUUAUCAUGAUUCUUC-UCC-3′ (anti-sense). For in vitro delivery, tumor cells on a six-well vessel were transfected with 300 pmol of synthesized siRNAs using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions.

To determine the expressional levels of proteins, 50 μg of protein were separated on SDS-PAGE and transferred to nitrocellulose membrane. The membranes were blocked with 5% nonfat dry milk in TBST (50 mM Tris, pH 7.5, 150 mM NaCl, 0.05% Tween-20) for 1 h, washed, and subsequently incubated overnight at 4 °C in TBST with 5% BSA containing the following antibodies: mouse monoclonal anti-SCP3 (clone# 25/SCP3; 0.5 µg/ml; BD Bioscience, San Jose, CA), goat polyclonal anti-VEGF-C (cat.# AF752; 0.5 µg/ml; R&D Systems, Minneapolis, MN), goat polyclonal anti-VEGF-D (cat.# AF286; 0.5 µg/ml; R&D Systems) and mouse monoclonal anti-β-actin (clone# AC-74; 0.5 µg/ml; Sigma-Aldrich, St. Louis, MO). Specific molecules were detected with HRP-labeled anti-mouse or anti-rabbit secondary antibodies (Pierce, Rockford, IL) and enhanced with ECL kit (Elpis Biotech, Korea). Densitometry was performed using an image analyzer Fujifilm LAS-400 (Fuji, Tokyo, Japan) and Image J densitometry software (Version 1.6, National Institutes of Health, Bethesda, MD) was used for analysis.

A total of 89 NSCLC cases with LN metastasis were prospectively selected from patients who were enrolled at Toyama University Hospital and National Hospital Organization Higashi-Ohmi General Medical Center. All patients underwent complete resections between 1988 and 2010, but did not receive neoadjuvant chemo or radio-therapy. Survival time and outcome data were available for all patients. Tumors were staged according to the International Union against cancer’s tumor-node-metastasis (TNM) classification. The histology was classified and graded according to 2004 World Health Organization guidelines [17]. The follow-up period ranged from 10 to 153 months.

Tissue microarray (TMA) was constructed from archival formalin-fixed, paraffin-embedded tissue blocks. Three 1.0 mm diameter tissue cores were arrayed on a recipient paraffin block using a tissue arrayer (Pathology Devices, Westminster, MD). Briefly, a representative tumor area was carefully selected for each tumor from hematoxylin and eosin (H&E) stained section of a donor block. TMA blocks were cut into serial 5-µm-thick sections, and then sections were deparaffinized in xylene and were rehydrated through a graded alcohol series to distilled water. Immunohistochemistry for SCP3 was performed as previously described [13]. For VEGF-A, VEGF-B, VEGF-C and VEGF-D, endogenous peroxidase was blocked using a 3% solution of aqueous hydrogen peroxide. Immunohistochemical staining was performed with the following primary antibodies: mouse monoclonal anti-VEGF-A (clone# 16F1; diluted 1:100; Thermo scientific, Waltham, MA), mouse monoclonal anti-VEGF-B (clone# MM0008-7B43; diluted 1:100; Abcam, Cambridge, MA), goat polyclonal anti-VEGF-C (Cat.# AF752; diluted 1:100; R&D Systems), and goat polyclonal anti VEGF-D (cat.# AF286; diluted 1:100; R&D systems). Antigen retrieval for VEGF-C and VEGF-D was performed using a pressure chamber (Pascal; Dako, Carpinteria, CA) with pH 6 target retrieval solution (Dako). For VEGF-A and VEGF-B antibodies, the TMA slide was pretreated with pH 9 target retrieval solution (Dako). After antigen retrieval, slides were incubated with primary antibodies for 30 min (VEGF-A) or overnight (VEGF-B, VEGF-C and VEGF-D) at 4 °C. Signals were detected with LSAB detection system (Dako) for VEGF-C and VEGF-D, and with Envision⁺ detect system (Dako) for the other antibodies. The stain was visualized using DAB (Dako) and then was lightly counterstained with hematoxylin, dehydrated in ethanol, and cleared in xylene. Negative controls were processed by omitting the primary antibodies, and TMA included testis positive control tissues. The cut-off values of histoscore were identified considering the distribution and prognostic significance of the values (Additional file 1: Figure S1).

The immunohistochemical staining was assessed using computer-assisted image analysis software, as described previously [18]. In brief, immunohistochemically stained slides were scanned by the NanoZoomer 2.0 HT (Hamamatsu Photonics K.K., Japan) at ×20 objective magnification (0.5 μm resolution). Captured images were analyzed using Visiopharm Digital Image Analysis (DIA) software (for Windows 7, version 4.5.1.324; Visiopharm, Hørsholm, Denmark). First, transformation of an image from one form to another (image processing) is carried out to enhance relative image structures for subsequent image segmentation. After training the system by digitally “painting” examples of the nucleus in the image, tumor nuclei were defined during segmentation. The cytoplasm was further defined by outlining the defined nucleus. The pixels that contribute to positive staining were identified based on a DAB color deconvolution. The algorithm evaluated each pixel on the value of 0–255. We used the mean of the pixel intensity values as the quantity of protein expression. Cut-off values of histoscore were identified considering the distribution and prognostic significance of the values (Additional file 1: Figure S1).

Given that SCP3 had been implicated in cervical cancer and metastasis to lymph nodes, we decided to examine its role in lung cancer. It has been known that VEGF-C and VEGF-D are associated with tumor lymphangiogenesis and metastasis [19, 20]. Thus, we measured the expression of SCP3, VEGF-C, and VEGF-D using human lung cancer cell lines. We observed a correlation between SCP3 and VEGF-C or VEGF-D in various human lung cancer cell lines (Fig. 1). Notably, linear regression analysis demonstrated that the level of SCP3 expression positively correlated with VEGF-C (R ² = 0.555) or VEGF-D (R ² = 0.867) in various human lung cancer cell lines (Fig. 1b). We further examined whether the relationship between SCP3 and VEGF-C or SCP3 and VEGF-D was conserved in human lung cancer cells. Consistently, we observed decreased levels of VEGF-C and VEGF-D when endogenous SCP3 high tumor cells (H1299) were treated with siRNA targeting to SCP3 (siSCP3) but not with non-specific control siRNA targeting GFP (siGFP). Conversely, increased VEGF-C and VEGF-D levels among H358 and H1666 cells expressing low levels of SCP3 were observed after over-expressing hSCP3 by a retroviral system (Fig. 1c, d).

We performed immunohistochemistry on 89 formalin-fixed paraffin-embedded tissues from NSCLC patients with LN metastasis for investigated the expression of SCP3, VEGF-A, VEGF-B, VEGF-C, and VEGF-D. Forty-six (51.7%) patients were N1 and 43 (48.3%) patients were N2–3 (Additional file 2: Table S1). Expression of SCP3 and all examined VEGFs was primarily detected in cytoplasm (Fig. 2), and the rate of high expression ranged from 24.7 to 30.3% (Table 1). No correlations between clinicopathological factors and SCP3 or VEGFs expression was observed, except for the association between VEGF-D and tumor type (P = 0.02, Table 1).

To explore the association between SCP3 and lymphangiogenetic factors, we further examined correlation between SCP3 and VEGF family member expression using Pearson Chi square test (Table 2). SCP3 expression showed significant positive correlation with VEGF-C (odds ratio [OR] = 5.600, P < 0.001) and VEGF-D (OR = 7.700, P < 0.001), whereas SCP3 expression was negatively correlated with VEGF-A (OR = 0.205, P = 0.019) and VEGF-B (OR = 0.244, P = 0.019). Furthermore, SCP3 expression inversely correlated with increased VEGF-A and VEGF-B expression, and both VEGF-A and VEGF-B expressions levels also inversely correlated with LN metastasis status (Fig. 3).

Analysis of patient survival was performed by Kaplan–Meier plots and log-rank analysis. Median survival time for patients with high SCP3 expression was significantly worse (median, 16 months) than that for patients with low expression (median, 66 months; p = 0.008; Fig. 4). No significant survival differences were observed for any tested VEGF family member. After adjustments for age, gender, cancer type, T factor and other antibodies, SCP3 remained significant (P = 0.008) and showed higher hazard ratio (HR) than T factor (HR; 1.86 with SCP3 and 1.80 with T factor) by a Cox proportional hazards regression model (Table 3).