LOX-1 is a poor prognostic indicator and induces epithelial-mesenchymal transition and metastasis in pancreatic cancer patients

Abstract

Purpose

Pancreatic cancer (PC) is an aggressive type of cancer that exhibits a rapid progression. Previously LOX-1, which is a type II trans-membrane glycoprotein that is expressed in endothelial cells, has been found to be involved in the development of several types of cancer. As yet, however, the expression of LOX-1 and its functional consequences in PC have not been documented. The present study was aimed at investigating the prognostic relevance of LOX-1 expression in PC patients and at resolving its role in PC metastasis.

Methods

LOX-1 expression was assessed by immunohistochemistry on a tissue microarray containing samples from 98 PC patients. Kaplan-Meier analyses were performed to compare survival curves, whereas Cox regression analyses were performed to explore the independent prognostic value of LOX-1 expression on the overall survival (OS) of PC patients. Harrel’s concordance index was applied to calculate the predictive accuracy of established models. In addition, in vitro scratch wound healing and Transwell assays were used to assess the effect of LOX-1 expression silencing and over-expression on PC cell migration and invasion, whereas Cell Counting Kit-8 (CCK8) and Flow Cytometry (FCM) assays were used to assess its effects on PC cell proliferation and apoptosis.

Results

We found that LOX-1 is highly expressed in the PC tumor tissues tested and is related to the occurrence of lymph node metastases, higher TNM stages and a poor OS. We also found that LOX-1 expression may serve as an independent prognostic factor for the OS of PC patients. Our in vitro assays revealed that LOX-1 expression may promote the migration and invasion of PC cells through epithelial-mesenchymal transition (EMT). No effect on PC cell proliferation was noted.

Conclusions

From our data we conclude that a high LOX-1 expression in PC tissues is indicative for the occurrence of lymph node metastases, high TNM stages and a poor prognosis. LOX-1 may serve as an independent prognostic biomarker. Our in vitro assays additionally revealed that LOX-1 may enhance the migration and invasion of PC cells through EMT. LOX-1 may also serve as a novel therapeutic target.

Electronic supplementary material

The online version of this article (10.1007/s13402-017-0360-6) contains supplementary material, which is available to authorized users.

Introduction

Pancreatic cancer (PC) is an aggressive disease with a 5-year survival rate of only 8%. This poor survival rate is mainly attributed to the fact that most patients are diagnosed at advanced stages due to a lack of typical symptoms at early stages of the disease. [1]. As yet, surgery remains the only curative option, but merely ~15 to 20% of all PC patients are deemed eligible for this treatment option due to the occurrence of metastases and rapid disease progression [2, 3]. Also, patients with similar clinicopathological features may have diverse clinical outcomes. Together, these findings warrant an exploration of novel effective biomarkers that may improve the prognostic accuracy and therapy of PC patients [4–6].

Lectin-like oxidized low density lipoprotein receptor-1 (LOX-1) is a type II trans-membrane glycoprotein that was first identified as a receptor of oxidized LDL (ox-LDL) in endothelial cells [7, 8]. Under physiological conditions basal LOX-1 levels are low, but they may be up-regulated by pro-inflammatory and pro-oxidative stimuli which, in turn, may be associated with an increased risk of cancer development [9–11]. It has been reported that LOX-1 activation in endothelial cells may promote nuclear factor κB (NF-κB) signaling and increase intracellular reactive oxygen species (ROS) levels, leading to endothelial dysfunction [12]. In addition, it has been reported that endothelial cell dysfunction in intra-tumoral blood or lymphatic vessels may deteriorate the condition of PC patients [13]. Therefore, LOX-1 is thought to act as a key intermediator linking endothelial dysfunction with PC development. Indeed, several studies have highlighted a positive correlation between LOX-1 expression and carcinogenesis [14, 15] and that high LOX-1 expression may serve as an important indicator of tumor progression in advanced-stage prostate cancer [16], colorectal cancer [17] and squamous non-small cell lung cancer [18]. As yet, however, no study has been reported on the prognostic value of LOX-1 in PC, as well as its putative mechanistic role in PC metastasis.

Here, we performed immunohistochemistry (IHC) to assess LOX-1 expression patterns in PC tissues and cells, and analyzed their relationships with various clinicopathological features. In addition, we evaluated the in vitro effects of LOX-1 expression silencing or over-expression on the migrative, invasive and EMT activities of PC cells. From our data we conclude that LOX-1 may promote PC metastasis and, as such, may serve as a novel prognostic biomarker and as a target for the treatment of PC.

Materials and methods

Patient samples and tissue microarray construction

A total of 176 tissue samples, including 78 pairs of pancreatic tumor/non-tumor samples and an additional 20 pairs of primary PC tumor/non-tumor samples, were collected from patients who underwent surgery from September 2004 to December 2011 at the Department of General Surgery, Zhongshan Hospital, Fudan University, Shanghai, P.R. China. The 176 samples were employed for the construction of a tissue microarray (TMA) using routine methods. Relevant clinicopathological features, such as age, gender, tumor size, tumor location, tumor grade, tumor differentiation and TNM stage, were collected from each patient. Tumor stages were classified according to the seventh edition of the American Joint Committee on Cancer TNM classification. Overall survival (OS) was defined as the interval between the date of surgery and the date of death or last visit. This research was approved by the Research Medical Ethics Committee of Fudan University and informed consent was obtained from each patient.

Immunohistochemistry analysis and evaluation

TMA samples were prepared for immunohistochemistry (IHC) according to standard protocols, blocked by UltraVision Hydrogen Peroxide Block (Thermo Scientific, Fremont, CA, USA) and UltraVision Protein Block (Thermo Scientific) and, subsequently, incubated with an anti-LOX-1 antibody (1:200, Proteintech Group, Chicago, IL, USA.). For antibody staining UltraVision Quanto Detection System horseradish peroxidase (HRP) Polymer and DAB Quanto (both from Thermo Scientific) were used, whereas hematoxylin was used for counterstaining. The results were assessed by an experienced and independent pathologist blinded to the patient statuses. Staining intensities were scored from 0 to 3. The heterogeneity of the staining was scored from 0 to 4, depending on the percentage of tumor cells that were positively stained. To obtain an IHC score that takes into account the IHC signal intensity and the frequency of positive cells, we generated a composite expression score (CES) with a full range from 0 to 12 (Table S1) [19]. The optimal cutoff value (CES = 6) was determined using Receiver Operating Characteristic (ROC) curve analysis (Supplementary Fig. S1). CES > 6 indicates a high LOX-1 expression and CES ≤ 6 indicates a low LOX-1 expression.

TCGA and GEO datasets

Publicly available data from the Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) database (accession number: GSE15471) were used. For the TCGA dataset, all level-3 data were downloaded using TCGA-Assembler software. The mRNA expression levels in the TCGA dataset were assessed using RNA sequencing V2. The RSEM (RNA-Sequencing by Expectation-Maximization) counts were further normalized by TMM (trimmed mean of M value) to estimate the relative mRNA expression levels using edgeR software (empirical analysis of digital gene expression in R). For the GEO dataset, the relative mRNA expression levels were retrieved through the Oncomine database.

Cell lines and culture conditions

Five human PC-derived cell lines (SW1990, CFPAC-1, BxPC-3, AsPC-1 and CAPAN-2) were obtained from the Cell Bank of Type Culture Collection of the Chinese Academy of Science (Shanghai, China). All cells were cultured in RPMI-1640 medium, supplemented with antibiotics and 10% fetal bovine serum, at 37 °C in a humidified atmosphere containing 5% CO₂. The culture medium and its supplements were purchased from Sigma (St. Louis, Mo, USA) and fetal bovine serum was purchased from Gibco (catalogue # 16000–044; Grand Island, NY, USA).

Plasmid construction and cell transfection

A shRNA construct targeting LOX-1 and a scramble control shRNA construct were purchased from Shanghai Genechem Co., LTD (Shanghai, China). A pcDNA3.1-LOX-1-Myc/His expression construct was generated through insertion of a full-length LOX-1 cDNA into a pcDNA 3.1/Myc-His(−)A vector (Invitrogen, Carlsbad, CA, USA). Transient transfections were carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. For the establishment of cells stably expressing LOX-1 shRNA, infection with LOX-1 shRNA packaging lentiviral particles was carried out followed by puromycin selection (1 μg/ml) for 3–4 days.

Western blotting

Protein samples, obtained by cell lysis, were separated by SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride (PVDF) membranes. Next, the membranes were incubated with primary antibodies (1:1000) followed by incubation with a secondary antibody (1:2000). Protein expression was visualized using an enhanced chemiluminescene (ECL) assay. As primary antibodies anti-Cadeherin (Cat # 3195), anti-β-catenin (Cat # 8480), anti-Vimentin (Cat # 5741) and anti-Slug (Cat # 9585) were used, whereas as secondary antibody Horseradish peroxidase (HRP)-conjugated goat anti-rabbit was used (Cell Signaling Technology, Beverly, MA, USA).

Quantitative RT-PCR

Total RNA was extracted from PC cells using TRIzol (Invitrogen, Carlsbad, CA, USA) and processed for reverse transcription and quantitative PCR using a Takara RNA PCR Kit and a SYBR Premix Ex Taq (Takara, Tokyo, Japan) according to the manufacturer’s instructions.

Transwell migration and invasion assays

Transwell assays were performed using 12-well Transwell plates (8 μm pore size) according to the manufacturer’s instructions (Millipore, Billerica, MA, USA). For the migration assays 1 × 10⁵ cells in serum-free culture medium were seeded into the upper chamber, whereas the lower chamber was filled with culture medium containing 20% FBS. For the invasion assays, the bottom of the Transwell chamber was coated with BD Matrigel Basement Membrane Matrix (BD Biosciences, San Diego, CA, USA). Cells were again seeded in the upper chamber and after incubation for 24 h and 48 h at 37 °C, respectively, non-invading cells on the upper side of the chamber were removed. The cells on the lower surface of the membrane were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet, after which the number of invading cells was counted in five randomly selected microscopic fields.

Scratch wound healing assay

Cells were transfected as indicated and seeded in 6-well plates. After reaching 90% confluence, the culture medium was removed and scratches were made in the centers of the wells using a sterile P10 pipette tip. Next, the cells were washed twice with 1xPBS after which 2 ml serum-free RPMI-1640 medium was added to each well. Images were taken at 0, 24 and 48 h post-scratch.

Cell viability assay

Cell viability was quantified using a CCK-8 assay (Dojindo, Japan) according to the manufacturer’s instructions. The cells were seeded at a density of 3000 cells per well in 96-well plates, after which CCK-8 assays were carried out at 0 h, 24 h, 48 h and 72 h after transfection through measuring absorbance at 450 nm.

Cell cycle and apoptosis assays

Cell cycle alterations and the occurrence of apoptosis were assessed using flow cytometry. Cells were collected by trypsinization and washed twice with PBS. For cell cycle analysis, the collected cells were stained with propidium iodide (PI) using a Cell Cycle Staining Kit (Lianke Bio, Hangzhou, China). Apoptosis was assessed using Annexin V-PE and 7-AAD staining in conjunction with a PE Annexin V Apoptosis Detection Kit I (BD Biosciences, CA, USA) and flow cytometry. Calculations were performed using the FlowJo software (TreeStar, Ashland, OR, USA) tool.

Statistical analyses

Statistical analyses were performed using SPSS 22.0 (SPSS Inc., Chicago, IL, USA), Prism software (GraphPad, La Jolla, CA, USA) and R software version 3.2.3. with the “rms” package (R Foundation for Statistical Computing, Vienna, Austria). Pearson’s chi-square test was performed for categorical variables, whereas continuous variables were analyzed by the Student’s t test to evaluate associations between LOX-1 expression levels and clinicopathological features. The clinical end points of the patients were calculated using Kaplan-Meier analysis and compared using a log-rank test. Subsequent univariate and multivariate Cox regression analyses were performed to determine the prognostic value of LOX-1 expression levels on overall survival (OS). Hazard ratios (HRs) obtained by Cox regression analysis are reported as relative risks with corresponding 95% confidence intervals (CIs). Nomograms were constructed using R software version 3.2.3 and calibration plots were generated to examine the performance of the constructed nomograms. Through receiver operating characteristic (ROC) curve analysis concordance indexes (C-indexes) were generated to assess the predictive accuracy of different prognostic models. Gene set enrichment analyses were performed using GSEA software 2–2.2.4. Correlations between two groups were analyzed using a nonparametric Spearman’s test. A two-sided p value of < 0.05 was considered statistically significant.

Results

LOX-1 expression is up-regulated in pancreatic cancer tissues

In order to assess the expression pattern of LOX-1, IHC staining was performed in both tumor and adjacent non-tumor tissues from 78 pancreatic cancer (PC) patients. By doing so, we found that the LOX-1 protein was mainly localized in the membranes and cytoplasms of the tumor cells (Fig. 1a). Subsequent quantitative analyses revealed that LOX-1 protein expression was enhanced in tumor tissues compared to normal tissues (6.936 ± 0.295 versus 2.910 ± 0.245, p < 0.001; Fig. 1b). In addition, we analyzed LOX-1 mRNA expression levels in the PC samples included in dataset GSE15471 (n = 39) and in 20 additional pairs of tumor/non-tumor PC samples, and found that in both cohorts the LOX-1 mRNA expression levels were higher in the tumor tissues than in the adjacent non-tumor tissues (Fig. 1c, d).

Fig. 1.

LOX-1 expression is increased in human PC tissues. a LOX-1 expression in tumor tissues and paired non-tumor tissues from PC patients. b Quantitative analyses of LOX-1 IHC staining in pancreatic cancer tissues. c, d LOX-1 mRNA expression levels in the GSE15471 cohort (c) and in 20 unrelated PC patients (d)

LOX-1 expression correlates with clinicopathological features and clinical prognosis

To assess correlations between LOX-1 expression and clinicopathological PC features, composite expression score (CES) analyses were performed for evaluating LOX-1 expression in IHC stained sections according to its intensity and frequency (see Supplementary Table 1). The optimal cutoff value of CES was determined by ROC curve analysis (Supplementary Fig. S1) and found to be 6, i.e., CES > 6 indicated a high LOX-1 expression and CES ≤ 6 indicated a low LOX-1 expression. We found that a high LOX-1 expression correlated with a high occurrence of lymph node metastases and a high TNM stage (Table 1). We also compared LOX-1 expression in TNM I stage patients versus TNM II-IV stage patients, and found that the expression was higher in the advanced-stage patients than in the early-stage patients (Fig. 2a, b; 7.309 ± 0.363 versus 6.209 ± 0.380, p = 0.04). This difference was confirmed by the TCGA database set (Fig. 2c; 8.868 ± 0.154 versus 7.098 ± 0.561, p < 0.001). Moreover, Kaplan-Meier survival analysis of data from the experimental set and the TCGA data set revealed that the OS of patients with a high LOX-1 expression was significantly shorter than that of patients with a low LOX-1 expression (Fig. 2d, e). Interestingly, in both the TNM I group (Supplementary Fig. S2a, p = 0.004) and the TNM II-IV group (Supplementary Fig. S2b, p < 0.0001) the survival of patients with a high LOX-1 expression was found to be shorter than that of patients with a low LOX-1 expression.

Table 1.

Relation between LOX-1 expression and clinical characteristics

Characteristic	Patients		LOX-1 expression		P a-value
	n	%	Low	High
All patients	98	100	41	57
Gender					0.295
Male	61	62.2	28	33
Female	37	37.8	13	24
Age					0.539
≤ 61	49	50.0	22	27
> 61	49	50.0	19	30
Tumor size					0.940
≤ 3 cm	28	27.8	12	16
> 3 cm	69	72.2	29	40
Tumor grade					0.650
G1-G2	67	68.4	40	27
G3	31	31.6	17	14
T classification					0.781
T1-T2	77	78.6	32	45
T3	20	21.4	9	11
N classification					< 0.001
N0	56	57.1	34	22
N1	42	42.9	7	35
M classification					0.813
M0	96	98.0	40	56
M1	2	2.0	1	1
TNM stage					< 0.001
I	43	43.9	27	16
II-IV	55	56.1	14	41

P < 0.05 indicates statistical significance of differences

^aPearson chi-square tests

Fig. 2.

LOX-1 expression is associated with TNM stage and overall survival (OS) of PC patients. a LOX-1 expression in early-stage and advanced-stage PC patients. b, c Quantitative analyses of LOX-1 expression in TNM I or TNM II-IV PC patients. d, e Kaplan-Meier OS analysis of PC patients according to LOX-1 expression

High LOX-1 expression is an independent indicator of a poor prognosis

Univariate and multivariate Cox regression analyses were used to evaluate the prognostic value of LOX-1 expression in PC patients. Our univariate analysis revealed that LOX-1 expression (HR, 0.227; 95% CI, 0.136–0.376; p < 0.001), lymph node metastasis (HR, 0.444; 95% CI, 0.265–0.744; p = 0.002), TNM stage (HR, 0.460; 95% CI, 0.281–0.753; p = 0.002) and tumor grade (HR, 0.420; 95% CI, 0.235–0.749; p = 0.003) may serve as risk factors for the overall survival (OS) of PC patients (Table 2). Subsequent multivariate Cox regression analysis revealed that both LOX-1 expression (HR, 4.204; 95% CI, 2.302–7.676; p < 0.001) and tumor grade (HR, 2.369; 95% CI, 1.425–3.938; p = 0.001) may serve as independent prognostic indicators for the OS of PC patients (Fig. 3a).

Table 2.

Univariate Cox regression analysis of overall survival of pancreatic cancer patients

Characteristic		HR(95%CI)	P-value
Age			0.408
> 61	≤ 61	0.813 (0.498–1.327)
Gender			0.618
Female	Male	1.221 (0.742–2.011)
LOX-1 expression			< 0.001
High	Low	0.227 (0.136–0.376)
Tumor size			0.842
> 3 cm	≤ 3 cm	1.056 (0.618–1.805)
T classification			0.785
T3	T1 + T2	1.088 (0.592–1.999)
N classification			0.002
N1	N0	0.444 (0.265–0.744)
M classification			0.369
M1	M0	0.409 (0.058–2.876)
TNM stage			0.002
II + IV	I	0.460 (0.281–0.753)
Tumor grade			0.003
G3	G1 + G2	0.420 (0.235–0.749)

HR Hazard ratio, CI Confidence interval

P < 0.05 indicates the statistical significance of differences

Fig. 3.

Nomograms and calibration plots for the prediction of overall survival (OS) of PC patients. a Multivariate Cox regression analysis of independent prognostic factors for OS of PC patients. b Nomogram model to predict the probabilities of 1-year and 3-year survival of PC patients. c Calibration of nomogram-predicted survival of PC patients for 1 or 3 years

A predictive nomogram for the stratification of PC patients

In order to establish a more quantitative analysis method for a better stratification of PC patients, we constructed a predicted nomogram, integrating the independent risk factors LOX-1 expression and tumor grade. In this nomogram model, a higher total score predicts a poorer OS at 1 and 3 years (Fig. 3b). The calibration plots illustrate that the predicted probabilities of the nomogram closely align with the actual survival estimates (Fig. 3c). The gray line represents the performance of an ideal nomogram, in which the predicted outcomes perfectly match with the actual outcomes. The red line was calculated from our dataset and represents the performance of our nomogram.

Combined TNM stage and LOX-1 expression increases prognostic accuracy

Based on the independent prognostic value of LOX-1 expression and its correlation with TNM stage, we set out to integrate LOX-1 expression with TNM staging to obtain a better prognostic accuracy. Areas under ROC curve (AUC) were used to compare the prognostic accuracies. We found that integration of LOX-1 expression with TNM stage (AUC = 0.853; 95% CI, 0.774–0.931; p < 0.001) showed a better prognostic accuracy than TNM stage (AUC = 0.685; 95% CI: 0.565–0.806; p = 0.005) or LOX-1 expression (AUC = 0.825; 95% CI: 0.733–0.916; p < 0.001) alone (Fig. 4). This result was further confirmed using C-index and AIC analyses (Table 3).

Fig. 4.

Receiver operating characteristic (ROC) curve analyses for overall survival (OS) prediction of PC patients. ROC curve analysis of the sensitivity and specificity for the prognosis of OS using a LOX-1 expression model, a TNM stage model and a LOX-1 expression + TNM stage combination model

Table 3.

Comparison of prognostic accuracy of LOX-1 expression and TNM stage

Model	C-index	AIC
TNM	0.589	572.135
LOX-1	0.663	545.907
LOX-1 + TNM	0.681	546.076

A larger C-index represents a better prognostic power

C-index: Harrell’s concordance index, AIC Akaike information criterion

LOX-1 over-expression promotes migration and invasion of PC cells

Since our clinicopathological analyses indicated that LOX-1 up-regulation may be associated with the occurrence of lymph node metastasis, we further explored the pro-metastatic role of LOX-1 in PC cells. Based on quantitative RT-PCR and Western blot analyses, SW1990 cells with a high LOX-1 expression and AsPC-1 cells with a low LOX-1 expression (Fig. 5a, b) were selected for in vitro analyses. In order to assess the effect of LOX-1 on PC metastasis, we exogenously over-expressed LOX-1 in AsPC-1 cells and silenced LOX-1 expression in SW1990 cells (Fig. 5c). By subsequently using Transwell assays, we found that that AsPC-1 cells in which LOX-1 was up-regulated exhibited enhanced migratory and invasive capacities, whereas SW1990 cells in which LOX-1 was silenced exhibited reduced migratory and invasive capacities (Fig. 5d, e). Additionally, we found by using scratch-wound healing assays that a higher LOX-1 expression enhanced the migratory activity of PC cells (Fig. S3). Using CCK-8 assays we found that LOX-1 expression had no effect on the viability of PC cells (Fig. 5f). Subsequent flow cytometry analyses revealed that LOX-1 expression had no overt effect on cell cycle progression and/or PC cell proliferation (Fig. 5g, h).

Fig. 5.

LOX-1 promotes migration and invasion of PC cells. a LOX-1 mRNA levels in different PC cells determined by RT-PCR. b LOX-1 protein levels in different PC cells determined by Western blotting. c Exogenous Myc-LOX-1 overexpression efficiency in AsPC-1 cells and shRNA-mediated knock-down efficiency in SW1990 cells. d, e Effects of LOX-1 expression on cell migration and invasion determined by Transwell assays. f Effects of LOX-1 expession on cell viability determined by CCK-8 assays. g Cell cycle analysis of LOX-1 overexpressing and LOX-1 silenced cells by flow cytometry (FCM). h Apoptosis analysis using Annexin V-FITC/PI staining and FCM

LOX-1 over-expression promotes EMT of PC cells

EMT is known to promote the migration and invasion potential of cancer cells and, as such, to play a critical role in tumor progression. In SW1990 cells in which LOX-1 was silenced, we found that the morphology was changed from spindle mesenchymal-like cells to cobblestone epithelial-like cells, while the morphology of LOX-1 over-expressing AsPC-1 cells was found to be changed in the opposite direction (Fig. 6a). Moreover, we found by correlation analysis that LOX-1 expression was negatively related to the mRNA expression levels of the mesenchymal marker Vimentin (VIM) and the transcriptional repressors Slug (SNAI2) and β-catenin (CTNNB1) (Fig. 6b). Based on these preliminary results, we set out to assess the potential effect of LOX-1 on the expression of these factors at the protein level. Using Western blot analysis, we found that in LOX-1 over-expressing AsPC-1 cells the expression level of the epithelial marker E-cadherin was up-regulated, while the expression levels of the mesenchymal marker Vimentin and the transcriptional repressors Slug and β-catenin were markedly reduced (Fig. 6c, d). These results were further confirmed in LOX-1 silenced cells. Together, these data suggest that LOX-1 over-expression may facilitate the migration and invasion of PC cells through EMT.

Fig. 6.

LOX-1 promotes EMT of PC cells. a Morphologies of LOX-1 overexpressing AsPC-1 cells and LOX-1 silenced SW1990 cells. b Correlation analysis of OLR1 mRNA levels with those of CDH1, VIM, SNAI2 and CTNNB1 from TCGA database. c Western blot analysis of E-cadherin, Vimentin, Slug and β-catenin expression in LOX-1 overexpressing AsPC-1 cells and LOX-1 silenced SW1990 cells. d Quantitative analyses of relative protein expression levels

LOX-1 expression correlates with IL-6 expression in PC

Inflammation has been identified as a significant factor in the development of cancer, including pancreatic cancer. In order to explore a putative relationship between LOX-1 expression and inflammation responses, RNA was extracted from 20 pairs of PC/adjacent normal tissue samples and subjected to quantitative RT-PCR. Subsequent correlation analyses revealed that LOX-1 expression was significantly correlated with IL-6 expression, whereas little correlation was found with IL-1α, IL-1β and TGF-β1 (Fig. S4a). This correlation could be validated in SW1990 cells in which LOX-1 was silenced (Fig. S4b). These results suggest that LOX-1 may form a bridge between inflammation and PC development.

Discussion

LOX-1, a receptor of oxidized LDL (ox-LDL), has been reported to be involved in various human disorders, including atherogenesis, diabetes, obesity and metabolic syndromes [9, 14, 20]. These disorders are known to be associated with the development and aggressiveness of PC. In particular, metabolic anomalies have been found to cause metastasis and death in almost 100% of the patients with PC [21]. It has also been reported that in vivo LOX-1 targeting may elicit a protective anti-tumor CD8⁺ T cell response [22]. Our TMA analysis revealed that LOX-1 is highly expressed in PC tissues and is related to the occurrence of lymph node metastases, suggesting that LOX-1 may serve as a diagnostic tool and as a tool for monitoring PC metastasis formation.

Pancreatic cancer ranks among the most malignant cancers for which the sole curative option is surgical resection. Unfortunately, this option is applicable to no more than 15% of the cases [23]. Due to the early occurrence of lymph node metastases, most patients are diagnosed at an advanced stage [24, 25]. Hence, there is an urgent need for the identification of novel metastasis-related targets for the design of effective curative therapies. LOX-1 up-regulation has been reported to contribute to EMT in diabetic pulmonary fibrosis by activating the TGF-β1/KLF6 pathway [26]. EMT is a process of cell remodeling that is critical to normal embryogenesis and tissue development. During tumor development, EMT results in loss of the polarized organization of epithelial cells and the acquisition of migratory and invasive capacities by these cells, enabling them to metastasize [27]. In accordance with these observations, we found that LOX-1 expression up-regulation in AsPC-1 cells led to increased migrative and invasive potentials and the acquisition of EMT potential, whereas LOX-1 silencing led to decreases in the migrative and invasive potentials of PC cells. These results suggest that LOX-1 may induce PC progression via EMT induction.

Endothelial dysfunction is a major risk factor for PC [13]. An important pathophysiological event related to endothelial dysfunction is the increased adhesion of monocytes to endothelial cells, which can be induced by C-reactive protein (CRP) [28, 29]. It has also been reported that blockage of LOX-1 may lead to the inhibition of CRP-induced monocyte adhesion [30] and pro-inflammatory cytokine release, such as IL-6, IL-1α and IL-1β [30–33], which is required for PC progression [34–36]. Although our current results indicate that LOX-1 may promote PC development and progression, the exact mechanism underlying its pro-metastatic function requires further investigation.

In conclusion, we found that LOX-1 over-expression is related to a poor prognosis of PC patients and, importantly, that the prognostic accuracy of the conventional TNM staging system can be improved through the integration of LOX-1 expression. In addition, we found that exogenous LOX-1 over-expression imposes invasion/metastasis potential and EMT onto PC-derived cells. These findings may be instrumental for the design of novel therapeutic strategies targeting LOX-1.

Author contributions

Jie Zhang designed and carried out experiments, analyzed data and wrote the manuscript; Lei Zhang collected and analyzed clinical data; Can Li, Caiting Yang and Lili Li performed experiments; Shushu Song and Hao Wu analyzed data; Jianxin Gu contributed the materials and experimental equipment; Lan Wang and Fenglin Liu conceived the research idea and took responsibility for this part of the project.

Compliance with ethical standards

Conflict of interest

The authors declare no conflict of interest.