LDLR Promotes Growth and Invasion in Renal Cell Carcinoma and Activates The EGFR Pathway

Background: Previous studies identied an association of low-density lipoprotein (LDL) levels and LDL receptor (LDLR) with renal cell carcinoma (RCC) development. This study investigated the expression and roles of LDLR in RCC. Methods: LDLR expression was examined in clear cell RCC (ccRCC) and adjacent normal kidney tissues, and its clinicopathological signicance was analyzed. The role of LDLR in RCC cell proliferation, cell cycle and invasion were assessed in RCC cells with LDLR stable knockdown. Results: LDLR expression was higher in ccRCC tissues than in normal kidney tissues and increased with RCC progression. LDLR knockdown in RCC cells inhibited cell growth, migration and invasion and induced G1/S cell cycle arrest. We identied an interaction between LDLR and EGFR, and EGFR signaling protein expression was reduced after LDLR knockdown. Conclusions: Our ndings reveal that LDLR plays an important role in RCC carcinogenesis, suggesting that LDL and LDLR might be potential targets for therapeutic intervention in RCC.

Increasing studies have investigated the role of dyslipidemia, a predominant component of MS, in the initiation and development of cancers. CcRCC is characterized by sterol storage in cancer cells, which prompts an abnormality in cell lipid metabolism. Horiguchi et al. found that uvastatin, a type of statin that is an effective drug for dyslipidemia, inhibited RCC cell growth, invasion, angiogenesis and metastasis in vitro [8]. Our previous study examined the prevalence of dyslipidemia in RCC patients in a Chinese population, and we observed that abnormal low-density lipoprotein (LDL) elevation was common in RCC cases compared with controls [9]. We further explored the association between LDL receptor (LDLR) polymorphism and ccRCC risk and found that functional variants in the LDLR gene are associated with ccRCC susceptibility [10]. Therefore, we speculated that LDLR may be involved in ccRCC carcinogenesis. In this study, we examined the expression of LDLR in ccRCC tissues and explored its association with clinicopathological characteristics. Furthermore, the impact of LDLR on RCC cell growth and invasion was also investigated.

Patient samples
A total of 574 consecutive ccRCC patients who underwent surgical treatment between January 2010 and December 2013 at the Department of Urology, Fudan University Shanghai Cancer Center (FUSCC) were enrolled in this study. Tumor tissues and adjacent normal kidney tissues were obtained after surgery and stored in the FUSCC Tissue Bank. All patients were diagnosed with ccRCC based on histopathological evaluation and did not receive radiotherapy, chemotherapy or targeted therapy before surgery. Sciences, Logan, USA). All media were supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 µg/ml streptomycin (Hyclone; GE Healthcare Life Sciences, USA), and cells were maintained at 37°C in a humidi ed atmosphere containing 5% CO 2 .

Immunohistochemistry (IHC)
LDLR protein expression was detected by IHC on 5 µm thick tissue sections prepared from formalin-xed, para n-embedded tissue from a constructed 10 × 12 tissue microarray that was made by the FUSCC Tissue Bank, as described previously [11]. After depara nization, dehydration, antigen retrieval and endogenous peroxidase activity blocking, tissue sections were incubated with antibody against LDLR (Santa Cruz Biotechnology, Santa Cruz, USA) and SABC (goat IgG) detection kit (BOSTER, Wuhan, China).
The IHC staining results were independently scored by two pathologists who were blinded to patient information.
RNA extraction and quantitative real-time PCR Total RNA from tissues and cultured cells was extracted using TRIzol reagent. Single-stranded cDNA was synthesized with the RevertAid First Strand cDNA synthesis Kit (Life Technology, Carlsbad, USA). Realtime PCR was performed using the SYBR Green PCR Master Mix (Applied Biosystems, Foster City, USA). The fold change of LDLR gene expression was determined using β-actin mRNA level as an internal control. Primer sequences were as follows: LDLR-F: TCTGCGAGGGACCCAACAAG, LDLR-R: TCGTTGGTCCCGCACTCTTT; and β-actin F: ACCGAGCGCGGCTACAG, β-actin R: CTTAATGTCACGCACGATTTCC.

Western blotting
Total protein was isolated from tissues and cells using RIPA lysis buffer, and protein concentration was quanti ed by the BCA assay kit (Thermo Fisher Scienti c, Waltham, USA) according to the manufacturer's protocol. Protein samples (50 µg) were separated by 10% SDS polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The membranes were rst blocked and then incubated with primary antibodies overnight at 4°C. After washes with phosphate buffered saline (PBS), membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, USA), and proteins were visualized using an ECL detection system (Thermo Fisher Scienti c, USA) Vector construction and lentivirus production and infection Short hairpin RNA (shRNA) for LDLR was introduced into the PLKO.3G vector and transfected into HEK293T cells along with psPAX2 and PMD2-G using Lipofectamine 2000 reagent (Life Technology, Carlsbad, USA) according to the manufacturer's protocol. Forty-eight hours later, lentivirus was harvested and used to infect RCC cells. Puromycin (2 µg/ml) was added into the medium to select stable infected cell clones. The e ciency of LDLR knockdown was con rmed by qRT-PCR and western blotting. The sequences of LDLR and control shRNAs are listed in supplementary Table 1.
Cell proliferation and colony formation assays CCK-8 assay (Dojindo, Shanghai, China), EdU assay (Ribobio, Guangzhou, China) and colony formation assay were performed to measure cell proliferation. For CCK-8 assays, cells were seeded in 96-well plates (4 ×10 3 cells per well) and cultured for 24 h. Next, 10 µL of CCK-8 solution was added into each well and cells were incubated for another 2 h. The absorbance values were measured at 450 nm using a microplate reader. For EdU assays, cells were incubated in EdU solution (1:5000) for 2 h, harvested and washed using PBS mixed with TritonX-100 (200:1). Cells were then stained with the Cell-Light EdU Apollo 643 In Vitro Flow Cytometry Kit according to the manufacturer's protocol and analyzed by ow cytometry (Beckman Coulter, USA). For colony formation assays, a total of 600 cells were seeded in a 6-well plate in triplicate per experimental group and cultured at 37°C in a 5% CO 2 incubator for 14 days. The medium was replaced with 4% paraformaldehyde (1 ml/well) and cells were incubated for 60 min at room temperature. After removing the supernatant, the clones were stained using 0.5% crystal violet for 30 min and counted under a light microscope.

In vivo tumorigenicity
All animal studies were approved by the Animal Studies Ethics Committee of FUSCC. BALB-C nude mice were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. ACHN-LDLR-shRNA or ACNH-Scr cells (1×10 7 /ml cells) were implanted subcutaneously in both sides of the back region of nude mice. Tumor sizes were measured at least three times weekly. At week 5 after injection, mice were euthanized with CO 2 and the tumors were removed. Tumor weight was examined and tumor volume was calculated after sacri cing.

Cell cycle and apoptosis assays
Cell cycle and apoptosis assays were both performed using ow cytometry (Beckman Coulter, USA). Brie y, cells were cultured for 72 h at 37°C, washed with PBS three times and then xed with 75% ethanol overnight at 4°C. Propidium iodide (50 µg/mL) containing RNase was added to the cells for DNA staining.
Stained cells were subjected to ow cytometry for cell cycle analysis. For cell apoptosis assays, cells were resuspended in 100 µl buffer containing Annexin V; next, 5 µl FITC-Annexin V and 5 µl propidium iodide (BD Biosciences, Franklin Lakes, USA) were added to cells for staining. After incubating at room temperature in the dark for 15 min, 400 µl of binding buffer was added to each cell suspension, and cells were subjected to analysis by ow cytometry.

Cell migration and invasion assays
Wound healing assay and Transwell chamber assay were performed to test cell migration and invasion in vitro, respectively. For wound healing assays, cells were seeded in a monolayer in 6-well plates. A scratch was introduced in the cell monolayer in the middle of each well using a 200 µl pipette tip. Images of cells were captured under an inverted microscope at 0 and 24 h time points. For migration assays, a total of 4×10 4 cells were seeded into the upper chamber of a Transwell chamber (BD Biosciences, Franklin Lakes, USA) coated with 60 µl Matrigel (BD Biosciences, Franklin Lakes, USA) diluted with serum-free medium (1:50), and 600 µl of medium supplemented with 10% FBS was added to the lower chamber. After incubation for 24 h, cells were xed with 4% polyoxymethylene and stained with crystal violet. Cells on the upper surface of the membrane were wiped off with cotton swabs, and those that invaded through the pores were photographed and counted using an inverted microscope in ve random elds.
Immuno uorescence Cells were seeded on coverslips overnight and then xed with 4% paraformaldehyde for 30 min. After incubating in 1% BSA and 0.25% Triton X-100 in PBS, cells were incubated with EGFR antibody or LDLR antibody (Santa Cruz Biotechnology, Santa Cruz, USA) at room temperature, followed by incubation with Alexa Fluor 594 IgG donkey anti-rabbit antibody (Invitrogen, Carlsbad, USA) for 1 h at room temperature.
Nuclei detection was performed using DAPI co-staining. Fluorescence images were acquired with a laser confocal microscope.

Co-immunoprecipitation
Cells were washed with PBS and lysed using RIPA buffer containing protease inhibitors (Roche Diagnostics, Basel, Switzerland). Protein samples were incubated with speci c antibodies overnight at 4°C and then 50 µl beads (Santa Cruz Biotechnology, Santa Cruz, USA) were added to the mixture. Samples were then incubated at 4°C for 4 h. The beads were washed in ice-cold PBS, re-suspended in loading buffer, and incubated at 90°C for 10 min. After separation on a 10% Bis-Tris gel, the samples were analyzed by western blot analysis.

Statistical analyses
All data are presented as means ± standard deviation. Student's t-test was used to compare the statistical differences between variables. All statistical analyses were performed using SPSS software version 20.0 (IBM SPSS, Armonk, USA), and two-sided P < 0.05 was considered to indicate a statistically signi cant difference.

LDLR expression levels in ccRCC tissues
We examined the expression of LDLR in ccRCC tissues and kidney tissues using IHC. Tumor tissues were acquired from 286 ccRCC patients, and adjacent normal kidney tissues were available for 188 patients. The clinicopathological characteristics of the 286 patients are listed in Supplementary Table 2. The mean (± SD) patient age was 55.8 ± 12.2 years. There were 157 patients with Fuhrman I-II (low-grade) and 129 patients with Fuhrman III-IV (high-grade) disease. The expression levels of LDLR were classi ed into strongly-positive, weakly-positive and negative groups according to IHC staining (Fig. 1A-F). Stronglypositive, weakly-positive and negative staining were observed in 181, 92 and 13 ccRCC tissues and in 90, 89 and 9 normal kidney tissues, respectively. LDLR expression was higher in ccRCC tissues than in normal kidney tissues (p < 0.05) (Fig. 1G). In addition, strongly-positive, weakly-positive and negative staining were observed in 90, 59 and 8 low-grade ccRCC tissues and in 91, 33 and 5 in high-grade tissues, respectively. LDLR expression was higher in high-grade tissues than in low grade tissues, however with marginal statistical signi cance (p = 0.069) (Fig. 1H).
We also detected the mRNA expression of LDLR in a different set of 288 ccRCC tissues and 100 adjacent normal kidney tissues. Supplementary Table 3 lists the clinicopathological features of these 288 patients. Of the 288 patients, 135 presented with Fuhrman I-II (low-grade) and 153 with Fuhrman III-IV (highgrade) disease, while 246 presented with stage I-II and 42 presented with stage III-IV disease. LDLR mRNA expression was signi cantly higher in high-grade and stage III-IV diseases compared with low-grade and stage I-II diseases, respectively (p < 0.05) (Fig. 1J, K). Interestingly, we observed lower expression of LDLR in ccRCC tissues than in normal kidney tissues (p < 0.05) (Fig. 1I).

Validation of LDLR knockdown in RCC cells
To explore the biological function of LDLR in RCC, we rst measured the protein and mRNA expression levels of LDLR in ve human RCC cell lines (ACHN, 786-O, A498, 769-P and Caki-1). We observed relatively high expression of LDLR in ACHN and 786-O cells at both mRNA and protein levels ( Fig. 2A, B). We selected ACHN cells for subsequent knockdown experiments.
Using lentivirus-mediated shRNA expression, we down-regulated the expression of LDLR in ACHN cells.
The knockdown e ciencies of the shRNAs are shown in Fig. 2C. Compared with control shRNA, shRNA1 reduced the levels of LDLR. We chose this construct to establish stable knockdown cells (named ACHN-LDLR-shRNA) for further experiments.

Effects of LDLR knockdown on RCC cell proliferation
We used CCK8, EdU and colony forming assays to investigate the effects of LDLR knockdown on RCC cell proliferation in vitro. As demonstrated in Fig. 3A and 3C, cell growth was signi cantly inhibited in ACHN-LDLR-shRNA cells compared with relevant controls (p < 0.05), as shown by both CCK8 and EdU assays. In addition, compared with controls, ACHN-LDLR-shRNA cells developed fewer colonies at day 14 in colony formation assays (p < 0.05) (Fig. 3B). Taken together, these results revealed that LDLR knockdown exerted a suppressive role on RCC cell proliferation in vitro.
To examine whether LDLR knockdown also affects RCC cell growth in vivo, subcutaneous tumor models were established in ve nude mice using ACHN-LDLR-shRNA or control cells. The mice were sacri ced at day 38 after injection. We found that both tumor volume and mass were notably lower in ACHN-LDLR-shRNA tumors than in controls (p < 0.05) (Fig. 3D-F). Together, these results demonstrated that downregulated LDLR expression inhibited cell proliferation in RCC cells both in vitro and in vivo.

Effects of LDLR knockdown on RCC cell migration and invasion
Would healing assay was used to evaluate the effect of LDLR knockdown on cell migration and Transwell assay was performed to detect the invasiveness of RCC cells with LDLR knockdown. As shown in Fig. 4A-D, cell migration of ACHN-LDLR-shRNA cells was markedly reduced 24 h after wound creation. Transwell assays demonstrated that LDLR knockdown signi cantly inhibited the invasive capacity of ACHN cells compared with the control cells (Fig. 4E, F). These observations indicate that LDLR knockdown restrains the migration and invasion abilities of RCC cells.

Effects of LDLR knockdown on RCC cell cycle and apoptosis
As LDLR knockdown exerts an inhibitory effect on RCC cell proliferation, we further explored its role in cell cycle and cell apoptosis. Compared with controls, ACHN-LDLR-shRNA cells showed a notable increase in the G1 phase population, while the proportion of cells in S phase decreased (Fig. 4G, H), indicating that LDLR knockdown induced a G1/S cell cycle arrest. This arrest was accompanied by elevated expression of p21 Cip1 and p27 Kip1 as well as decreased expression of cyclin D1 and CDK4 (Fig. 5C). However, we did not observe obvious differences in cell apoptosis between ACHN-LDLR-shRNA cells and controls (data not shown).

Effects of LDLR knockdown on EGFR pathway
To further investigate the mechanisms by which LDLR affects RCC carcinogenesis, we performed mass spectrometry, co-immunoprecipitation and immunocyto uorescence analyses to identify the proteins that interact with LDLR. As shown in Fig. 5A and 5B, an interaction between LDLR and EGFR was observed. EGF and EGFR, which belong to a growth factor signaling pathway, are important participants in cancer initiation and development. We thus examined several important proteins in the EGFR signaling pathway. As indicated in Fig. 5D, the expression levels of EGFR, mTOR, AKT, pAKT and Ras were all reduced after LDLR knockdown compared with controls.

Discussion
The prevalence of obesity and obesity-related chronic diseases have dramatically increased worldwide over the past decades. Besides cardiovascular disease (CVD) and diabetes, common cancers investigated in the context of obesity include breast, colorectal, prostate, and endometrial cancers, as well as RCC [12]. Haggstrom et al. explored metabolic factors associated with RCC risk and found that obesity, hypertension and dyslipidemia contribute to increased RCC risk [13]. Based on studies suggesting that abnormal lipid metabolism may play a role in the biological process driving RCC development, we examined the expression and role of LDLR, an important molecule involved in cholesterol homeostasis, in ccRCC and found that the levels of LDLR were elevated with the progression of RCC. Down-regulation of LDLR in RCC cells remarkably inhibited cell proliferation, migration and invasion and induced G1/S cell cycle arrest. Additionally, we observed an interaction between LDLR and EGFR, suggesting that LDLR may promote growth and invasion in RCC by activating the EGFR pathway.
LDLR is a ubiquitously expressed cell membrane glycoprotein that binds and internalizes circulating cholesterol-containing lipoprotein particles. LDLR is an essential mediator for lipid metabolism, and its dysfunction has been proven to contribute to familial hypercholesterolemia and early onset coronary heart disease [14,15]. Previous genome-wide association studies have reported that LDLR genetic susceptibility is associated with serum lipid levels [16-18]. However, relatively few studies to date have investigated the role of LDLR in cancer development. Rudling et al. observed relatively lower mRNA expression of LDLR in RCC tissues compared with normal kidney tissues [19]. Consistent with Rudling's results, we also observed lower LDLR expression at the mRNA level in RCC tissues compared with normal kidney tissues. Interestingly, our IHC results showed the opposite phenomenon: LDLR expression was higher in ccRCC tissues than in normal kidney tissues. We speculate that this discrepancy might be attributed to different sampling processes. IHC staining was judged in the area of proximal convoluted tubules from where ccRCC is derived. However, the tissue ultrastructure was not separated when mRNA was tested. Hence, the LDLR mRNA levels re ected levels expressed in both distal and proximal convoluted tubules. Notably, in tumor tissues, LDLR expression was higher in high-grade disease than in low-grade disease at both mRNA and protein levels. In addition, LDLR mRNA expression was signi cantly higher in stage III-IV diseases compared with stage I-II diseases. Therefore, we speculate that LDLR exerts a role in promoting RCC development.
In breast cancer patients, higher level of intratumor cholesteryl ester displayed higher expression of LDLR at both mRNA and protein levels, and linked to cell proliferation and aggressive tumor potential [20]. Gallagher et al. used mouse models for hyperlipidemia and publicly available human datasets to determine the importance of LDLR in breast cancer. The authors found that silencing LDLR in breast cancer cells led to decreased growth of Her2Neu-overexpressing tumor cells both in vitro and in vivo, and high LDLR expression in human breast cancers was associated with decreased recurrence-free survival [21]. In our study, we observed similar roles of LDLR in ccRCC, suggesting that LDLR might be a contributing factor in obesity-related cancers. However, Gallagher et al. also found that LDLR knockdown resulted in increased caspase 3 cleavage. In mice lacking LDLR, increased cell apoptosis was also observed in the liver [22]. We did not observe an association between LDLR knockdown and RCC cell apoptosis. Tumor heterogeneity may account for this discrepancy. In addition, we did not perform followup because most of our subjects were with early-stage diseases. Further investigations are needed to explore the complete mechanisms and clinical effects of LDLR in RCC.
Increasing evidence has revealed the important role of growth factor-related signaling pathways in cancer initiation and progression. Among these pathways, EGF-EGFR is a key signaling pathway and its abnormal activity has been observed in many cancers, including breast, lung, esophageal, colorectal, head and neck cancers [23]. The EGF-EGFR pathway participates in tumor cell proliferation, apoptosis, migration, invasion and angiogenesis as well as the pathogenesis of many renal disorders [24,25]. Stumm et al. reported that EGFR is overexpressed in RCC and is associated with tumor initiation and progression [26]. Therefore, although some EGFR inhibitors such as ge tinib and erlotinib have shown therapeutic outcomes that are less than expected, targeting the EGFR signaling pathway is still an attractive and promising intervention in cancer therapy. In our study, we identi ed EGFR as a binding partner of LDLR and con rmed that LDLR interacted with EGFR. We further showed that LDLR knockdown led to decreased expressions of EGFR, mTOR, AKT, pAKT and Ras. These results suggest that LDLR might promote RCC growth and invasion by activating the EGFR signaling pathway. To date, several new drugs targeting EGFR and/or related signaling molecules have been tested for application in RCC [27,28]. However, the deeper mechanism underlying the cross-talk between LDLR and EGFR, as well as the therapeutic effect of new compounds targeting the LDLR signaling pathway, still remain to be explored.

Conclusions
In conclusion, our results revealed that LDLR promotes RCC cell growth and invasion through activating the EGFR signaling pathway. This study provides another clinical implication. Dyslipidemia is an important risk factor in the development of CVD, and recent studies have identi ed some cardiovascular biomarkers that are associated with RCC progression [29]. Thus, we speculate that CVD may share etiology with RCC, and we might bene t more from better control of abnormal lipid pro les.

Declarations Acknowledgments
We thank Liwen Bianji, Edanz Editing China (www.liwenbianji.cn/ac), for editing the English text of a draft of this manuscript.

Authors' contributions
Guiming Zhang designed the study, collected, analyzed and interpreted the data. Wei Chen, Yu Yao and Lei Luo collected part of the patients' clinical data. Lei Luo and Lijiang Sun analyzed part of the data. Wei Chen performed parts of the experimental studies. Guiming Zhang and Lijiang Sun wrote the manuscript. Guiming Zhang supervised the project and revised the manuscript. All authors vouch for the respective data and analysis, approved the nal version, and agreed to publish the manuscript.

Funding
This study was partly funded by the National Natural Science Foundation of China (Grant No. NSFC 81502195).

Availability of data and materials
The datasets used or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study is approved by Ethics Committee of Fudan University Shanghai Cancer Center.

Consent for publication
All authors agreed to publish the manuscript.    Interaction between LDLR and EGFR and effects of LDLR knockdown on EGFR pathway. An interaction between LDLR and EGFR was observed using co-immunoprecipitation (A) and immunocyto uorescence analyses (B); LDLR knockdown triggered alterations in cell cycle related molecules (C), and the EGFR signaling pathway (D).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.