SPOCK2 Expression Correlates with Non-Small Cell Lung Cancer Proliferation, Invasion, and Tumorigenesis

Purpose: Lung cancer is one of the leading causes of death worldwide. Sparc (osteonectin), cwcv, and Kazal-like domains proteoglycan 2 (SPOCK2) play important roles in the development and progression of various types of human cancers. However, the role of SPOCK2 in non-small-cell lung cancer (NSCLC) is unclear. Hence, we aimed to elucidate the role of SPOCK2 in NSCLC. Methods: Real-time PCR and immunohistochemistry were used to study the relationship between SPOCK2 expression and the clinicopathology of NSCLC. SPOCK2 expression was modulated using siRNA (knockdown) or pcmv6-myc-DDK-SPOCK2 (overexpression). The invasion and migration abilities of NSCLC cells were assessed using a Transwell chamber assay, and cell proliferation was studied using the MTT and colony formation assays. Furthermore, axilla and tail vein inoculation using nude mice helped elucidate the function of SPOCK2 in vivo. The expression of related proteins was analyzed using western blotting. SPOCK2 methylation was detected using real-time PCR following the treatment of cells with decitabine. Results: The mRNA and protein levels of SPOCK2 were lower in NSCLC tissues than in adjacent normal lung tissues. SPOCK2 overexpression inhibited cell proliferation, migration, and invasion in A549, H1299 cells, and nude mice. Notch and Erk signaling pathways were inhibited by SPOCK2, whereas the Hippo pathway was activated. SPOCK2 was methylated in NSCLC cell lines. SPOCK2 expression was negatively correlated with NSCLC progression. Conclusion: Taken together, our ndings indicate that SPOCK2 methylation may result in its low expression, and the levels of SPOCK2 methylation may be associated with poor prognosis in NSCLC.


Introduction
Lung cancer is a major public health issue (Mao et al. 2016) and a leading cause of cancer-related deaths globally (Nasim et  During the past decades, several small molecular inhibitors, which signi cantly improve the overall prognosis of lung cancer, have been introduced clinically (Lovly 2015; Reck 2013). Epigenetic differences among patients are considered a critical determinant of poor response to drug therapy (Chen 2015).
Among the epigenetic changes associated with human cancers, DNA methylation is most widely studied (Mehta et al. 2015). The role of altered DNA methylation in cancer was identi ed in 1983. Within 30 years of this discovery, DNA methylation inhibitors were developed to clinically treat a variety of cancers, such as myelodysplastic syndrome (MDS). Furthermore, our understanding of the epigenetic changes associated with lung cancer has improved greatly in the past two decades (Duruisseaux and Esteller 2018). Currently, DNA methylation has been widely accepted as an important biomarker in the clinical management of lung cancer as DNA methylation provides useful information that enables early diagnosis, staging, prognosis, and therapy-response prediction (Fleischhacker et al. 2013).
SPARC is a multifunctional glycoprotein and extracellular matrix protein that regulates the interaction of the extracellular matrix, binds collagen and bronectin to endothelial cells, and inhibits cell proliferation.
Piersigilly et al. (2016) reported various potential biomarkers for bronchopulmonary dysplasia, including SPOCK2, using "omics" technologies. Furthermore, Hadchouel et al. (2011) showed that the SPOCK2 mRNA levels markedly increase during the alveolar stage of lung development in new-born rat lungs, and exposure to hyperoxia results in higher SPOCK2 expression than observed in air-exposed controls. Using gene difference analysis, Helena et al. showed that when AT2 cells are isolated, cultured, and stimulated to differentiate into AT1 cells, SPOCK2 expression is increased. Lastly, Ren et al. (2011aRen et al. ( , 2014b showed that epigenetic inactivation caused by the methylation of the SPOCK2 promoter is associated with malignant transformation of ovarian endometriosis. Tumorigenesis is a process of dedifferentiation.
Hence, in this study, we hypothesized that SPOCK2 might be methylated in lung cancer and play a role in the development of NSCLC.

Patients and specimens
Two hundred and sixty-two cases of NSCLC (157 men and 105 women; age, 34-78 years old; operated between 2013 and 2014) were randomly selected from the Pathology Archive of the First A liated Hospital of China Medical University for immunohistochemistry, and 40 samples of NSCLC tumor and paracancerous lung tissue from operations in 2015 were randomly selected for real-time PCR and 16 samples of NSCLC tumor and paracancerous lung tissue from operations in 2015 were randomly selected for western blotting. None of the patients whose data were subjected to the pathological correlation analysis had received preoperative chemotherapy or radiotherapy. Clinicopathological information was obtained from patient records. This study was approved by the Medical Research Ethics Committee of China Medical University, and informed consent was obtained from all patients (AF-SOP007-1.1-01). The animal care and experimental protocols were approved by the Medical Research Ethics Committee of the China Medical University (CMU2020406).
For the demethylation studies, the cells were treated with 2 µM decitabine (prepared in DMSO; Sigma-Aldrich) for 24 h; the control cells were treated with the same volume of DMSO.

Page 5/21
Cell were prepared for western blotting as described previously (Xu et al. 2018). All the primary antibodies used are shown in Table 1. The cell proliferation and colony formation assays were performed as described previously (Xu et al. 2018).

Cell migration and invasion analysis
The cell migration and invasion assays were carried out in 24-well Transwell chambers containing inserts with a pore size of 8-µm (Costar, Washington, DC, USA). For the invasion assay, the inserts were coated with 100 µL of Matrigel (1:9 dilution; BD Bioscience). Cells were trypsinized 24 h after transfection, and 5 × 10 4 cells resuspended in 100 µL of a medium supplemented with 2% FBS were transferred to the upper Transwell chamber, whereas 600 µL of the medium supplemented with 10% FBS was added to the lower chamber. After incubation for 24 h, cells on the upper membrane surface were removed with a cotton tip, and those that passed through the membrane were xed with polyformaldehyde and stained with hematoxylin. The number of migrated/invaded cells was counted in 10 randomly selected elds under a microscope at high magni cation. All experiments were repeated at least three times independently under identical conditions.

Gene set enrichment analysis (GSEA)
GSEA was performed using GSEA v2.2.2 (www.Broadinstitute.org/gsea). The SPOCK2 expression level was strati ed into low and high categories to annotate the phenotype, and we used lung cancer-related gene sets from KEGG and GO. All other parameters were set based on their default values.

Statistical analysis
SPSS 16.0 software was used for data analysis. The correlations between SPOCK2 expression and clinicopathological features were determined using the Chi-square test, and the differences between cell groups were analyzed using paired t-tests. Results with a two-sided P value < 0.05 were considered signi cant.

Clinical signi cance of SPOCK2 downregulation
Analysis of SPOCK2 protein expression by immunocytochemistry indicated that SPOCK2 is localized in both the cytoplasm and the nuclei of A549, H1299, Calu1, H292, H460, H661, and SK-MES-1 cell lines and normal bronchial epithelial HBE cells (Fig. 1A).
Overall, 40 pairs of clinical lung cancer tissues and corresponding normal lung tissue samples were collected for RNA extraction. SPOCK2 mRNA expression levels were detected by real-time quantitative PCR. SPOCK2 mRNA expression levels in the lung cancer tissues were signi cantly lower than those in the corresponding normal lung tissues (single sample paired t-test, P = 0.0176) (Fig. 1B).
Western blotting showed that SPOCK2 protein expression was lower in 14 of the 16 lung cancer tissues than in the matched normal tissues (Fig. 1C). Immunohistochemical analysis in 262 patients showed that SPOCK2 was expressed at low levels in clinical lung cancer tissues, whereas in normal bronchi and alveoli, SPOCK2 was expressed at high levels. In the correlation analysis (cross-table analysis) with clinicopathological factors, SPOCK2 expression did not correlate with age (P = 0.242) or sex (P = 0.319), but was positively correlated with histological type (P = 0.045) and differentiation (P = 0.000), and inversely correlated with tumor size (P = 0.000), lymph node metastasis (P = 0.000), and p-TNM stage (P = 0.000). Statistical analyses are shown in Table 2. The data from SPOCK2 differential expression analyses are shown in Fig. 1D. Immunohistochemistry of 156 clinical samples showed that low SPOCK2 expression is associated with poor prognosis (Fig. 1E).

SPOCK2 methylation correlates with poor prognosis
Given the methylation of SPOCK2 in endometrial carcinoma (Ren et al. 2011a, 2014b), we analyzed whether the low expression of SPOCK2 in NSCLC is also associated with methylation. First, we analyzed the methylation level of SPOCK2 in 830 NSCLC and 74 normal lung tissue samples using the sample information hosted in the TCGA. The expression level of SPOCK2 was negatively correlated with its methylation level (Fig. 1F). The methylation level of SPOCK2 in tumor samples was signi cantly higher than that in normal lung tissues (Fig. 1G). A high level of SPOCK2 methylation was associated with poor prognosis of NSCLC (Fig. 1H).

SPOCK2 inhibits the proliferation of lung cancer cells both in vitro and in vivo
26. To study the role of SPOCK2 in NSCLC, we rst investigated its expression in various cancer cell lines ( Fig. 2A). There was high expression of SPOCK2 in H661 cells, low expression in Calu1, H292, H460, and SK-MES-1, and moderate expression in A549 and H1299 cell lines. We selected A549 and H1299 to investigate the role of SPOCK2 in the cell lines. SPOCK2 was knocked down using siRNA transfection and upregulated by transfection with the pcmv6-myc-DDK-SPOCK2 plasmid. All in vitro experiments were carried out under the ideal conditions for ensuring interference and transfection e ciency (Fig. 2B). We analyzed the effect of SPOCK2 expression on colony formation as a read out of its effect on cell proliferation. In A549 and H1299 cells, SPOCK2 downregulation promoted colony formation, whereas SPOCK2 upregulation inhibited colony formation (Fig. 2C). In the MTT assay, SPOCK2 overexpression inhibited cell proliferation (Fig. 2D). Analysis of the proteins involved in cell proliferation by western blotting indicated that the expression of cyclin D1 (Pestell et al. 1999;Sherr 1996), CDK2 (Morgan 1995), and C-myc (Grandori et al. 2000) was downregulated upon SPOCK2 overexpression, whereas that of p21 (Pestell et al. 1999) was upregulated; SPOCK2 downregulation had the opposite effect (Fig. 3E). In subcutaneous tumorigenicity experiments, animals transplanted with A549 and H1299 cells stably transfected with SPOCK2 exhibited signi cantly lower tumor volume than those in the control group ( Fig. 2F and G).

SPOCK2 inhibits the migration and invasion of lung cancer cells both in vitro and in vivo
As SPOCK2 expression is correlated with lymph node metastasis, we speculated that SPOCK2 is related to the migration and invasion of lung cancer cells. In both A549 and H1299 cells, Transwell assays showed that SPOCK2 expression was associated with the promotion of migration (Fig. 3A) and invasion (Fig. 3B). Variations in SPOCK2 protein levels also affect the expression of proteins involved in migration and invasion (Coussens et al. 2002;McCormack et al. 2013;Totsukawa et al. 2000). The expression of RhoA, Rock1, MMP2, and MMP9 was downregulated, whereas that of RhoC and E-cadherin was upregulated by SPOCK2 overexpression, whereas SPOCK2 knockdown caused the opposite effects ( Fig. 3C). In nude mice tail vein tumorigenesis experiment, animals transplanted with H1299 cells stably expressing SPOCK2 exhibited less tumorigenesis in the lung (Fig. 3D-i and E) as well as in the liver (Fig. 3D-ii). Collectively, these ndings demonstrate that SPOCK2 inhibits the migration and invasion of lung cancer cells.

SPOCK2 methylation in NSCLC
Given the methylation of SPOCK2 in TCGA data, we treated our NSCLC cell lines with the demethylation drug decitabine (HY-A0004, MCE, NJ, USA) (Gangat et al. 2016;Hagemann et al. 2011;Lai et al. 2018). In each cell line, the SPOCK2 mRNA levels in the experimental group were higher than those in the control group after decitabine treatment (Fig. 4F). Decitabine is a broad-spectrum anti-methylation drug, which cannot reverse the methylation of SPOCK2, but the results can still re ect the methylation of SPOCK2 in NSCLC cell lines.

Discussion
In this study, we revealed the relationship between SPOCK2 and the clinicopathology of NSCLC. A high expression of SPOCK2 was found to be negatively correlated with tumor size, lymph node metastasis, and pTNM stage, whereas low SPOCK2 expression indicated poor prognosis. Furthermore, we found that SPOCK2 can negatively regulate the proliferation, migration, and invasion of NSCLC cells both in vitro and in vivo.
It has been previously reported that SPOCK2 expression increases during the transformation from AT2 to AT1 cells (Johansson et al. 2014); thus, SPOCK2 plays a positive regulatory role in the process of relative stem cell differentiation into mature cells. During tumorigenesis, a process involving immature cells, we found that SPOCK2 expression was low. Furthermore, low SPOCK2 expression correlated with differentiation. Thus, we showed that SPOCK2 has the ability to inhibit cell proliferation, migration, and invasion in NSCLC.
The results of GSEA are consistent with our experiments. SPOCK2 acts as a tumor suppressor gene in NSCLC by downregulating p-ERK1/2, p-JNK, and Notch signaling pathways and promoting the Hippo signaling pathway. These signaling pathways interact to form a network, which can jointly affect the proliferation, migration, and invasion of NSCLC by acting on target proteins, i.e., CyclinD1, CDKs, p21, RhoA, E-Cadherin, and MMPs. These ndings indicate the signi cance of SPOCK2 in NSCLC.
However, the reason for the low expression of SPOCK2 in NSCLC is still unknown. We thus explored the level of SPOCK2 methylation using TCGA. The methylation level of SPOCK2 in NSCLC was signi cantly higher than that in the normal lung tissues, and hypermethylation of SPOCK2 has been shown to be associated with poor prognosis. This high methylation level, i.e., low SPOCK2 level, corresponds to our results, indicating that low SPOCK2 mRNA and protein expression was associated with poor prognosis.
Decitabine is an effective demethylation agent used in the treatment of MDS (Gangat et al. 2016). The introduction of decitabine to inhibit methylation strongly suggested the methylation of SPOCK2. However, it remains unclear how SPOCK2 methylation occurs. Whether the elimination of methylation can reverse the occurrence and development of NSCLC is also worthy of follow-up study.
In summary, our ndings indicate that SPOCK2 plays a critical role in regulating the progression of NSCLC through the inhibition of cell proliferation, migration, and invasion both in vitro and in vivo. This inhibition may be counteracted by the methylation of SPOCK2 in NSCLC. Expression of SPOCK2 in NSCLC is associated with poor prognosis (a) Immuno uorescence assays were performed to detect SPOCK2 localization in NSCLC cell lines and human normal bronchial epithelial cells (b) Relative expression of SPOCK2 mRNA in 40 paired NSCLC samples (red bars) and adjacent normal lung tissue (green bars) (c) SPOCK2 protein expression in 16 lung cancer tissues compared to that in the matched normal tissues (d) SPOCK2 protein expression analyzed by immunohistochemistry in normal bronchial epithelial cells (i) and alveolar (ii), well-differentiated adenocarcinoma (iii), and squamous cell carcinoma (iv), poorly differentiated adenocarcinoma (v), and squamous cell carcinoma (vi).

Declarations
Magni cation 200× (e) Survival of patients with NSCLC with high and low SPOCK2 expression (f) The expression level of SPOCK2 and its methylation level (g) The methylation level of SPOCK2 in tumor samples and normal lung tissues (h) Survival of patients with NSCLC with high and low SPOCK2 methylation Figure 2 Effect of SPOCK2 expression on NSCLC proliferation (a) SPOCK2 level in HBE cells and seven NSCLC cell lines by western blotting (b) Interference and transfection e ciency of SPOCK2 in A549 and H1299 cell lines (c) Colony formation assay. A549: si-SPOCK2 vs. control siRNA, P = 0.003, p-SPOCK2 vs. control, P =0.003, H1299: si-SPOCK2 vs. control siRNA, P = 0.000, p-SPOCK2 vs. control, P = 0.022 (d) MTT assay. A549: si-SPOCK2 vs. control siRNA, P = 0.002, p-SPOCK2 vs. control, P = 0.039. H1299: si-SPOCK2 vs.
control siRNA, P = 0.0022, p-SPOCK2 vs. control, P = 0.0073 (e) Changes in the expression of cell proliferation-related proteins in A549 and H1299 cells (f) Xenograft tumor volumes were negatively affected by SPOCK2 in A549 cells, the upper row was the SPOCK2(+) group, and the lower row was the control group (g) Xenograft tumor volumes were negatively affected by SPOCK2 in H1299 cells, the upper row was the control group, and the lower row was the SPOCK2(+) group Effect of SPOCK2 expression on the migration and invasion of NSCLC cells (a) Cell migration analyzed by the Transwell migration assay; cells that migrated to the lower chamber were stained with hematoxylin and counted. *P < 0.05; **P < 0.01. A549 migration: A549: si-SPOCK2 vs. control siRNA, P = 0.000, p-SPOCK2 vs. control, P =0.001, H1299: si-SPOCK2 vs. control siRNA, P = 0.002, p-SPOCK2 vs. control, P = 0.042 (b) Cell invasion analyzed by the Transwell assay. A549: si-SPOCK2 vs. control siRNA, P = 0.005, p-