OLR1 Promotes Pancreatic Cancer Metastasis Via Increased C-Myc

Published OnlineFirst February 4, 2020; DOI: 10.1158/1541-7786.MCR-19-0718

MOLECULAR CANCER RESEARCH | CANCER GENES AND NETWORKS

OLR1 Promotes Pancreatic Cancer Metastasis via Increased c-Myc Expression and Transcription of HMGA2 A C Gang Yang1, Guangbing Xiong1,2, Mengyu Feng1, Fangyu Zhao1, Jiangdong Qiu1, Yueze Liu1, Zhe Cao1, Huanyu Wang1, Jinshou Yang1, Lei You1, Lianfang Zheng3, Taiping Zhang1,4, and Yupei Zhao1

ABSTRACT ◥ Pancreatic cancer is one of the most lethal human malignancies, pancreatic cancer cells in vitro and in vivo. Mechanistically, partly because of its propensity for metastasis. However, the OLR1 increased HMGA2 transcription by upregulating c-Myc mechanisms of metastasis in pancreatic cancer remain unclear. expression to promote the metastasis of pancreatic cancer cells. In Oxidized low-density lipoprotein receptor 1 (OLR1), a lectin-like addition, patients with pancreatic cancer with high expression of scavenger receptor that recognizes several ligands, such as oxidized OLR1–c-Myc–HMGA2 axis showed worse prognosis compared low-density lipoprotein, was previously reported in cardiovascular with patients with low expression of OLR1–c-Myc–HMGA2 axis. and metabolic diseases. The role and mechanism of OLR1 in pancreatic cancer is unclear. In this study, we found that OLR1 Implications: Our ﬁndings suggested that the OLR1–c-Myc– expression was signiﬁcantly higher in pancreatic cancer tissues than HMGA2 axis promotes metastasis of pancreatic cancer cells and that in adjacent normal tissues and closely associated with reduced may serve as potential therapeutic targets and prognosis markers for overall survival. OLR1 promoted proliferation and metastasis of patients with pancreatic cancer.

Introduction Oxidized low-density lipoprotein receptor 1 (OLR1) is a lectin-like scavenger receptor that recognizes several ligands, such as oxidized Pancreatic cancer is the fourth and sixth leading cause of cancer- low-density lipoprotein, polyanionic chemicals, and anionic phospho- related death in the United States and China, respectively, with an lipids (5). OLR1 is mainly distributed in vascular endothelial cells and overall survival (OS) rate of less than 9% in both countries (1, 2). tissues with abundant blood vessels such as renal, pulmonary, and Furthermore, the mortality rate of pancreatic cancer is increasing, and neuronal tissues. Previous studies on OLR1 mainly focused on its role pancreatic cancer is projected to be the second most common cause of in cardiovascular and metabolic diseases, such as atherosclerosis, cancer-related death by 2030 in the United States (3). Because of the hypertension, diabetes, and hyperlipemia (6–8). Upregulation of atypical clinical symptoms of pancreatic cancer, the majority of ORL1 in cardiovascular and metabolic diseases increased DNA dam- patients already show metastases at diagnosis and thus are unable to age caused by intracellular reactive oxygen species, promoted the be treated by radical operation (4). Despite treatment with surgical release of inﬂammatory cytokines, and enhanced hypoxia and angio- resection, most patients with pancreatic cancer eventually develop genesis (5, 9). Notably, these processes are also closely related to cancer metastases because of recurrence or chemoresistance. The current development (10). Indeed, upregulation of OLR1 was recently found in understanding of the underlying mechanism of metastasis in pancre- multiple cancers and shown to be involved in tumorigenesis, cancer atic cancer remains limited. Therefore, clarifying the metastasis development, and metastasis (11–14). For example, activation of the mechanisms is important for developing new treatment strategies for TNFa/NF-kB pathway upregulates OLR1 to promote the migration of pancreatic cancer and prolonging patient survival. breast cancer cells (11, 15). In return, upregulated OLR1 also activates NF-kB signaling to induce cellular transformation in breast cancer and 1Department of General Surgery, Peking Union Medical College Hospital, Chinese trigger epithelial–mesenchymal transition (EMT) to induce metastasis Academy of Medical Sciences and Peking Union Medical College, Beijing, China. in prostate cancer (13, 16). OLR1 also facilitates metastasis of gastric 2Department of Biliary-Pancreatic Surgery, Afﬁliated Tongji Hospital, Tongji cancer through driving PI3K/Akt/GSK3b activation (17). Recently, Medical College, Huazhong University of Science and Technology, Wuhan, Hubei 3 our group found that OLR1 was upregulated by lncRNA GSTM3TV2 Province, China. Department of Nuclear Medicine, Peking Union Medical Col- through competitively sponging let-7 to promote gemcitabine resis- lege Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China. 4Clinical Immunology Center, Chinese Academy of tance in pancreatic cancer (18). Together these studies suggest that Medical Sciences and Peking Union Medical College, Beijing, China. OLR1 may be a link between metabolic diseases and cancer. However, the role of OLR1 in pancreatic cancer has not been completely Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/). revealed. The proto-oncogene c-Myc regulates DNA synthesis, cellular pro- Corrected online June 17, 2021. liferation, differentiation, survival, and immortalization through its Corresponding Authors: Yupei Zhao, Peking Union Medical College Hospital, function as a transcription factor that binds canonical DNA-binding Chinese Academy of Medical Sciences, 1# Shuai Fu Yuan Hu Tong, Beijing motifs (E box 50-CACGTG-30) in the promoters of target genes to 100730, China. Phone: 008-669-156-007; Fax: 008-669-156-007; E-mail: [email protected]; and Taiping Zhang, [email protected] regulate gene expression. c-Myc plays an important role in the initiation and development of cancer by its placement at the crossroads Mol Cancer Res 2020;18:685–97 of many cancer-related signaling pathways (19, 20). c-Myc promotes doi: 10.1158/1541-7786.MCR-19-0718 cell metastasis in many cancers and is regulated by many factors (19). 2020 American Association for Cancer Research. However, the relationship between OLR1 and c-Myc is unclear.

AACRJournals.org | 685

Yang et al.

High mobility group AT-hook 2 (HMGA2), a member of high Western blot analysis mobility group A, plays a critical role in growth during embryonic Western blot analysis was performed as described previously (28). development but is expressed at low levels in adult tissues (21). Briefly, protein lysates (20 mg) were separated using SDS-PAGE, and However, HMGA2 has been reported to be upregulated in many primary and secondary antibodies were applied to detect target cancers (21–23). HMGA2 mainly functions as a transcription cor- proteins. An enhanced chemiluminescence assay was used to visualize egulator by modifying chromosomal architecture or recruiting other protein bands. Primary antibodies targeting OLR1 and HMGA2 were transcription-associated proteins to regulate the differentiation and purchased from Proteintech Group and c-Myc and b-actin antibodies proliferation of cells (24). HMGA2 was recently shown to promote were obtained from Cell Signaling Technology. metastasis in cancer by regulating EMT and enhancing TGFb signaling (21–23, 25). qRT-PCR In this study, we investigated the role and potential function of TRizol reagent (Invitrogen) was used for the extraction of total OLR1 in pancreatic cancer. We found that OLR1 expression was RNA and cDNA synthesis was performed using the Takara First highly expressed in pancreatic cancer tissues and that OLR1 pro- Strand cDNA Synthesis Kit following the manufacturer's instruc- moted metastasis of pancreatic cancer cells through increasing tions. qRT-PCR assays were performed independently at least three transcription of HMGA2 through c-Myc. Our results identified a times. The PCR primers are listed in the Supplementary Materials novel OLR1-c-Myc-HMGA2 axis and show its potential as a and Methods. combined biomarker of pancreatic cancer patient prognosis. RNA-sequencing and data analysis A RNeasy MINI KIT (Qiagen) was used for the extraction of Materials and Methods total RNA. The directional (stranded) libraries were generated Cell lines and clinical specimens using an Illumina platform for the paired-end sequencing of The human pancreatic cancer cell lines Bxpc-3 and Mia PaCa-2 Bxpc-3 cells. Cuffdiff was used to obtain the sequence count data were purchased from the ATCC and cultured in DMEM or RPMI1640 [fragments per kilobase of transcript per million mapped reads with 10% FBS (HyClone) in a 5% CO2 cell culture incubator at 37 C. (FPKM) values] to conduct differential expression analysis. Shang- All cell lines were authenticated using high-resolution small tandem hai Biotechnology Corporation performed the library construction repeats (STR) profiling at Department of General Surgery Laboratory. and sequencing. The mycoplasma of cells was examined by MycoSensor PCR Assay Kit (Catalog No. C0301S; Beyotime). Cells were grown for 15 passages and Chromatin immunoprecipitation assay then replaced with fresh stocks. The 99 clinical pancreatic cancer Chromatin immunoprecipitation (ChIP) assays were performed specimens were obtained from patients who received surgery in the using a ChIP Assay Kit (Millipore, 17-295) according to the Peking Union Medical College Hospital from January 2004 to Novem- manufacturer's instructions. Briefly, c-Myc protein was immuno- ber 2008; the specimens were used for the construction of a TMA. This precipitated from whole cell lysates by A/G magnetic beads con- research was approved by the Ethics Committee of Peking Union jugated to anti-c-Myc antibody (Cell Signaling Technology). Medical College Hospital and written informed consent was obtained Quantitative ChIP was performed using an ABI StepOne Plus with from each patient. SYBR Green dye. Primers spanning every region comprising c-Myc binding elements 1 or 2 were designed and used for ChIP-qPCR. IHC analysis The primer sequences are listed in the Supplementary Materials Antibodies against OLR1, HMGA2 (Proteintech Group), and c- and Methods. Myc (Cell Signaling Technology) were used to perform IHC analysis in the TMA as described previously (26). Two experienced pathol- Dual luciferase reporter assay ogists independently assessed the results. Staining intensities were A Dual Luciferase Reporter Assay System (Promega) was performed graded as 0 (negative), 1 (low), 2 (medium), or 3 (high), whereas as described previously (22). Briefly, cells (1 103 cells/well) were the staining extent was scored from 0% to 100%. The final IHC plated in a 96-well plate. After 24 hours, the cells were cotransfected staining score was calculated as intensity score percentage score with 0.2 mg expression vector plasmid or 0.02 nmol siRNA, 0.2 mg 100. promoter reporter plasmid, and 0.002 mg SV40 plasmid using Lipofectamine 3000 (Invitrogen) according to the manufacturer's TCGA, GEO, and ENCODE datasets instructions. After 48 hours, cells were lysed using lysis buffer and Data from the TCGA, the GEO database (accession numbers: the lysates were centrifuged at maximum speed for 1 min in an GSE28735 and GSE71729) and ENCODE were used (27). Eppendorf microcentrifuge. The Dual Luciferase Reporter Assay System was used to detect the relative luciferase activity of the Constructs and cell transfection samples, and the transfection efficiencies were normalized accord- siRNA targeting OLR1 and a scramble control siRNA were pur- ing to Renilla activity. chased from Guangzhou RiboBio Co., Ltd. The full-length OLR1 cDNA was subcloned into the pcDNA 3.1 empty vector to generate Proliferation assays the pcDNA3.1-OLR1 expression construct. Transient transfections Cell Counting Kit-8 (CCK-8) reagent (Dojindo) was used to were performed using Lipofectamine 3000 (Invitrogen) according to perform growth assays following the manufacturer's protocol. Cells the manufacturer's protocol. The OLR1 overexpression construct and were seeded into 96-well plates (1,000 cells/well). At 0, 24, 48, 72, and OLR1 shRNA were packaged into lentivirus, and cells were infected to 96 hours later, CCK-8 (10 mL/well) was added and cells were cultured establish cells that stably express OLR1 or shRNA targeting OLR1. for an additional 2.5 hours at 37C. The optical density (OD) was Cells were then selected by puromycin (1 mg/mL) for 7 days. The measured as the relative cell number at a wavelength of 450 nm siRNA sequences are listed in Supplementary Materials and Methods. (OD450) with a microplate reader.

686 Mol Cancer Res; 18(5) May 2020 MOLECULAR CANCER RESEARCH

OLR1–c-Myc–HMGA2 Promotes Metastasis in Pancreatic Cancer

Cell migration and invasion assays GSE71729) and found that the level of OLR1 mRNA was upregulated Cell migration and Matrigel cell invasion assays were conducted in pancreatic cancer tissues compared with adjacent normal tissues as described previously (26). In brief, polycarbonate membranes with (Fig. 1D and E). 8-mm pores were coated with 0.5 mg/mL Matrigel (BD Biosciences) for We next evaluated the correlation between the level of OLR1 6 hours. After the membranes were rehydrated, 4.0 104 pancreatic and clinicopathologic features. There was no significant correlation cancer cells in DMEM without FBS were placed into the upper of OLR1 expression with gender, age, tumor location, or T stage in chamber of a Transwell unit, and DMEM with 10% FBS was applied patients with pancreatic cancer. However, high expression of OLR1 as a chemoattractant in the lower chamber. After 24 hours, the was correlated with poor differentiation (P < 0.05), increased migrated or invaded cells were fixed by 4% paraformaldehyde and lymph node metastasis (P < 0.001), and advanced TNM staging stained with hematoxylin and eosin. After the membranes were dried, (P < 0.001; Table 1), indicating that OLR1 level is related to the numbers of migrated/invaded cells were counted in 10 low-power metastasis. fields (magnification, 100). The mean values were determined in Kaplan–Meier analysis revealed that patients with pancreatic cancer triplicate assays. All experiments were performed at least twice with high OLR1 expression had a significantly worse median OS than independently. those with low OLR1 expression (10 months vs. 35 months, respectively; P ¼ 0.0095; Fig. 1F). Analysis of patients with pancreatic cancer Wound-healing assay in the Cancer Genome Atlas (TCGA) and GSE28735 also revealed that Cells were seeded into a 12-well plate and then incubated overnight high OLR1 mRNA expression correlated with poor prognosis of in complete medium. A scratch wound was created in the cell patients with pancreatic cancer (P < 0.05; Fig. 1G and H). Univariate monolayer with a sterile 200 mL pipette tip and the detached cells and multivariate analyses showed that OLR1 level was negatively were removed. The cells were then incubated in 1% FBS medium, and correlated with OS in both analyses, indicating that high OLR1 images were captured at 6, 12, 18, and 24 hours later. The relative expression is an independent prognostic risk factor for patients with distance of cell migration from the wound at 0 hour was determined as pancreatic cancer (Table 2). the migration distance, as published previously (26). OLR1 promotes proliferation, migration, and invasion of Xenograft tumor mouse models pancreatic cancer cells All animal experiments were performed according to the institu- To determine the effect of OLR1 overexpression on cell prolifer- tional ethical guidelines of Peking Union Medical College (Beijing, ation, CCK-8 assay was performed as described in Materials and China). For in vivo imaging, human pancreatic cancer cells were Methods by measuring the OD at a wavelength of 450 nm as the infected by lentivirus carrying a firefly luciferase cassette as well as relative cell number after adding CCK-8 agent into cells and culturing puromycin resistance gene. Cells expressing firefly luciferase (1 106) for an additional 2.5 hours at 37C. The growth rates of Bxpc-3 and were inoculated in situ or intravenously or intrasplenically into Mia PaCa-2 pancreatic cancer cells transfected with OLR1 overexpres- mice and monitored by an IVIS imaging system, as published sion plasmids were significantly faster compared with control cells (P < previously (26, 29). 0.0001; Fig. 2A) and the growth was inhibited when OLR1 was knocked down (P < 0.0001; Fig. 2B). Transwell migration and invasion Statistical analysis assays showed that OLR1 upregulation promoted migration and Each experiment was repeated at least three times. Data were invasion of Bxpc-3 and Mia PaCa-2 cells and OLR1 knockdown in measured by Student t test (two-tailed) and analyzed using one-way Bxpc-3 and Mia PaCa-2 cells inhibited cell migration and invasion (P < ANOVA. All statistical tests were performed using the Statistical 0.0001; Fig. 2C). Wound-healing assays also revealed that OLR1 Program for Social Sciences (SPSS16.0 for Windows) and were two- overexpression increased but OLR1 knockdown decreased the migra- sided. P value of <0.05wasregardedasstatisticallysignificant. tory activity of both Bxpc-3 and Mia PaCa-2 cells compared with GraphPad Prism 6 software (GraphPad) was used to create all control cells (Fig. 2D and E). The effect of OLR1 knockdown and graphs. upregulation on OLR1 protein levels was verified by Western blot analysis (Fig. 2F). Ethics approval and consent to participate Together these findings demonstrated that OLR1 promotes the All experiments were approved by the institutional ethical com- proliferation, migration, and invasion of pancreatic cancer cells mittee of Peking Union Medical College Hospital (Beijing, China). in vitro.

OLR1 promotes the growth and invasion of pancreatic cancer Results cells in subcutaneous pancreatic cancer mouse models OLR1 expression is upregulated in pancreatic cancer specimens To explore the effect of OLR1 inhibition of pancreatic cancer We first detected the expression of OLR1 in a tissue microarray proliferation and invasion in vivo, we used a lentivirus system expres- (TMA) containing 99 pancreatic cancer tissues and 62 normal pan- sing shRNA to inhibit OLR1 expression (Lenti-shOLR1) or control creatic tissues using IHC. OLR1 was mainly expressed in the cell lentivirus (Lenti-shCon) in Bxpc-3 cells. Lenti-shOLR1- and Lenti- cytoplasm and cell membrane (Fig. 1A) consistent with the cellular shCon-infected Bxpc-3 cells were subcutaneously injected into the immunofluorescence result (Supplementary Fig. S1A). The expression posterior flanks of nude mice. The growth of xenograft tumors in mice of OLR1 was significantly higher in the cancer tissues than in adjacent was monitored, and the mice were sacrificed 5 weeks after transplan- normal pancreatic tissues according to the IHC scores (P < tation. Compared with control mice, the Lenti-shOLR1 group showed 0.001; Fig. 1B). The paired analysis of OLR1 IHC scores in 62 of the significantly suppressed tumor growth (P < 0.05; Fig. 3A). Tumor 99 cases with both cancer and adjacent tissues showed the same result weights were also lower in the OLR1 silenced group than in the control (P < 0.001; Fig. 1C). We also investigated the expression of OLR1 in group (P < 0.0001; Fig. 3B). H&E staining showed that OLR1 down- pancreatic cancer samples from public databases (GSE28735 and regulation decreased the invasion of pancreatic cancer cells to adjacent

AACRJournals.org Mol Cancer Res; 18(5) May 2020 687

Yang et al.

Figure 1. Increased expression of OLR1 is correlated with pancreatic cancer progression. A, Representative images of IHC staining of OLR1 in paired pancreatic cancer tissues and adjacent normal pancreatic tissues. Magniﬁcation, 50; scale bars, 200 mm. Enlarged areas are shown in boxes: magniﬁca- tion, 200; scale bars, 100 mm. B and C, Analysis of OLR1 expression in the TMA specimens according to staining scores: (B) 99 pancreatic cancer and 62 normal tissues and (C) 62 paired pancreatic cancer and normal tissues. D and E, OLR1 mRNA expression levels in the GSE28735 (D) and GSE71729 datasets (E). F, Comparison of OS of patients with different OLR1 levels in TMA using the Kaplan–Meier method (n ¼ 96; P ¼ 0.0095). G and H, Kaplan– Meier analysis of OS in patients with pancreatic cancer according to OLR1 mRNA expression level in the TCGA database (n ¼ 177; P ¼ 0.0201; G)and GSE28735 (n ¼ 42; P ¼ 0.0120; H; , P < 0.05; , P < 0.01; , P < 0.001).

muscle tissues (Supplementary Fig. S1B). We performed similar OLR1 inhibition suppresses the metastasis of pancreatic cancer experiments using a lentivirus system overexpressing OLR1 (Lenti- cells in orthotopic, intrasplenic, and intravenous pancreatic OLR1). The tumor growth and tumor weights in the OLR1 upregulated cancer models mouse group were signiﬁcantly higher than those in the control group To investigate whether OLR1 inhibition suppresses pancreatic (Fig. 3C and D). The invasion of pancreatic cancer cells to adjacent cancer metastasis, we injected Bxpc-3 cells infected with Lenti-shOLR1 muscle tissues was also more apparent when OLR1 was increased or Lenti-shCon orthotopically into the pancreas of SCID mice and (Supplementary Fig. S1B). OLR1 overexpression and knockdown was monitored the growth of transplanted tumors by bioluminescence veriﬁed by IHC staining and Western blot analysis (Supplementary imaging. After 5 weeks, the group with OLR1 knockdown showed Figs. S1B–S1D). suppressed xenograft tumor growth compared with controls (Fig. 3E). Together, these data demonstrated that OLR1 promotes the pro- In mice injected with Lenti-shOLR1-expressing Bxpc-3 cells, the liferation and invasion of Bxpc-3 pancreatic cancer cells in subcuta- numbers of metastases in the liver, kidney, lung, spleen, abdominal, neous xenograft mouse models. and gut decreased dramatically compared with control mice (Fig. 3F

688 Mol Cancer Res; 18(5) May 2020 MOLECULAR CANCER RESEARCH

OLR1–c-Myc–HMGA2 Promotes Metastasis in Pancreatic Cancer

Table 1. Correlations of OLR1 levels in tissues and and G). All the metastases were validated by H&E staining (Fig. 3H) clinicopathologic parameters. and the downregulation of OLR1 expression was verified by IHC staining (Supplementary Fig. S1E). OLR1 expression To confirm the above results, we further injected Bxpc-3 cells Low group High group infected with Lenti-shOLR1 or Lenti-shCon into the spleen of mice. Variables n (n ¼ 30) (n ¼ 66) Pa value The bioluminescence imaging and dissection results revealed fewer Gender 1.000 and smaller metastases in liver when OLR1 was knocked down Male 63 20 43 compared with controls (Fig. 3I and J). And the metastases were Female 33 10 23 validated by H&E staining (Fig. 3K and L). Similarly, the biolu- Age 0.658 minescence imaging and dissection results of the intravenous <60 45 15 30 mice also showed fewer and smaller metastases in lung when OLR1 ≥ 60 50 14 36 was knocked down compared with controls (Supplementary Location 0.097 Fig. S2A–S2D). Head 60 22 38 Body/tail 32 6 26 Together, these results indicate that inhibition of OLR1 leads to the Differential degree 0.034 decrease of metastatic capacity in pancreatic cancer cells. Low 31 5 26 High/moderate 65 25 40 OLR1 promotes the metastasis of pancreatic cancer cells by T staging 0.660 regulating c-Myc and HMGA2 T1/2 57 17 40 To identify the transcriptome signature that may be involved in T3 38 13 25 OLR1-mediated pancreatic cancer metastasis, we performed whole < N staging 0.001 transcriptome RNA-sequencing (RNA-seq) analysis of Bxpc-3 cells N0 27 16 11 with OLR1 overexpression or inhibition. In pancreatic cancer cells N1/N2 60 11 49 < with OLR1 overexpression, there were 181 upregulated and 203 TNM staging 0.001 ≥ I/IIA 26 16 10 downregulated genes compared with control cells (by 2.0-fold P < IIB/III 61 11 50 change and 0.05; Fig. 4A; Supplementary Table S1). In cells with downregulation of OLR1, 179 genes were upregulated and 166 genes Note: P < 0.05 indicates statistical significance of differences. were downregulated compared with controls (Fig. 4A; Supplemen- aPearson chi-square tests. tary Table S1). To narrow down the genes specifically related to OLR1 expression, 26 genes positively correlated with OLR1 and 39 genes negatively correlated with OLR1 in both Lenti-OLR1 and

Table 2. Univariate and multivariate analyses for prognosis factors of pancreatic cancer.

Univariate Multivariate Variables n HR (95% CI) P HR (95% CI) P

Gender 0.314 Male 63 1 Female 33 0.774 (0.470–1.274) Age 0.456 <60 45 1 ≥60 50 1.194 (0.749–1.904) Location 0.463 Head 60 1 Body/tail 32 1.201 (0.736–1.960) Differential degree <0.001 <0.001 Low 31 1 1 High/moderate 65 0.36 (0.221–0.587) 0.39 (0.238–0.639) Tumor staging 0.643 T1/2 57 1 T3 38 0.892 (0.549–1.449) Lymph node staging 0.190 N0 27 1 N1/N2 60 1.443 (0.834–2.495) TNM staging 0.183 I/IIA 26 1 IIB/III 61 1.461 (0.836–2.551) OLR 1 expression 0.013 0.044 Low 30 1 1 High 66 2.001 (1.157–3.461) 1.768 (1.016–3.078)

Note: P < 0.05 indicates the statistical signiﬁcance of differences. Abbreviations: HR, hazard ratio; CI, conﬁdence interval.

AACRJournals.org Mol Cancer Res; 18(5) May 2020 689

Yang et al.

Figure 2. OLR1 promotes proliferation, migration, and invasion of pancreatic cancer cells. A, Comparison of the growth of Bxpc-3 (top) and Mia PaCa-2 cells (bottom) transfected with pcDNA3.1 control or pcDNA3.1-OLR1 plasmids using CCK8 assay. B, Comparison of the growth of Bxpa-3 (top) and Mia PaCa-2 (bottom) cells transfected with siRNA control (siNC) or siOLR1 for 48 hours using CCK8 assay. C, In vitro transwell migration and invasion assays were performed in Bxpc-3 (top) and Mia PaCa-2 cells (bottom) with OLR1 plasmid overexpression or silencing by siRNA (siOLR1) for 24 hours. Representative images are shown. Magnification, 100. Cell numbers of migrated and invasive cells are shown in the right panels. D, Wound-healing assays were performed in Bxpa-3 (left) and Mia PaCa-2 cells (right) with OLR1 overexpression. Wounds were photographed, and the wound closure percentage from a representative experiment (n ¼ 3) was measured using AxioVision software. Magnification, 100. E, Wound-healing assays were performed in Bxpa-3 (left) and Mia PaCa-2 (right) cells with OLR1 silencing. Wounds were photographed, and the percentage of wound closure (n ¼ 3) was measured using AxioVision software. Magnification, 100. F, Western blot analysis of OLR1 in cells with OLR1 knockdown and upregulation (, P < 0.05; , P < 0.01; , P < 0.001).

690 Mol Cancer Res; 18(5) May 2020 MOLECULAR CANCER RESEARCH

OLR1–c-Myc–HMGA2 Promotes Metastasis in Pancreatic Cancer

Figure 3. OLR1 promotes the growth of pancreatic cancer cells in pancreatic cancer mouse models. A, Photographs of dissected tumors from nude mice injected with Lenti- shCon- and Lenti-shOLR1-expressing Bxpc-3 cells (left); the tumor volumes are shown in the right panel. B, Tumor weights were calculated at the end of the experiment (right, n ¼ 7). C, Photographs of dissected tumors from nude mice injected with Lenti-OLR1 and Lenti-Con-expressing Bxpc-3 cells (left); the tumor volumes are shown in the middle panel (n ¼ 7). D, Tumor weights were calculated at the end of the experiment (right, n ¼ 7). E, Knockdown of OLR1 decreased the tumor growth in orthotopic xenograft models established using Bxpc-3 cells. Representative IVIS images of nude mice are shown (n ¼ 5). F, Dissected primary pancreas tissues, livers, lungs, and kidneys. G, Numbers of metastatic nodes and invasive organs are shown. H, Metastatic and invasive organs were examined by H&E staining. Magnification, 50, scale bars, 200 mm. The images in the top right or bottom right panels represent magnified views of the boxed regions. Magnification, 200; scale bars, 100 mm. I and J, Inhibition of OLR1 decreased the metastasis of Bxpc-3 cells as observed in intrasplenic pancreatic cancer xenograft models. Representative IVIS images of nude mice are shown (n ¼ 5). K, Representative dissected liver images in Lenti-shOLR1 and Lenti-shCon groups. L, Representative H&E images of the liver are shown in Lenti-shOLR1 and Lenti-shCon groups; black arrows show the metastases in livers. Magnification, 40; scale bars, 500 mm. The images in the right panels represent magnified views of the boxed regions. Magnification, 100; scale bars, 200 mm(, P < 0.05; , P < 0.01; , P < 0.001).

AACRJournals.org Mol Cancer Res; 18(5) May 2020 691

Yang et al.

Figure 4. OLR1 increases the metastasis of pancreatic cancer cells by regulating c-Myc and HMGA2. A, Analysis of gene expression profiles identified 26 genes positively correlated with OLR1 levels and 39 genes negatively correlated with OLR1 levels in Bxpc-3 cells. B, Cluster heatmap of mRNA expression profiles in Bxpc-3 cells with OLR1 overexpression or knockdown. C and D, RT-qPCR of c-Myc (C) and HMGA2 mRNA (D) in Bxpc-3 cells. GAPDH mRNA served as a loading control. E, Western blot analysis of c-Myc and HMGA2 in Bxpc-3 cells. F and G, c-Myc (F) and HMGA2 (G) expression positively correlated with OLR1 expression in the TCGA database of patients with pancreatic cancer (n ¼ 178). H–K, Rescue assays were performed in in vitro transwell and invasion migration experiments in Bxpc-3 cells. Representative images are shown. Magnification, 100. Cell numbers of migrated and invasive cells are shown in the right panels (, P < 0.05; , P < 0.01; , P < 0.001).

692 Mol Cancer Res; 18(5) May 2020 MOLECULAR CANCER RESEARCH

OLR1–c-Myc–HMGA2 Promotes Metastasis in Pancreatic Cancer

Lenti-shOLR1 cells were chosen for further analysis (Fig. 4B). activities were significantly inhibited, whereas the luciferase activity Among these genes, COL4A4, HMGA2, and MYC are related to of HMGA2 Mut1 showed no changes with coexpression of sic-Myc cancer metastasis, but only HMGA2 and MYC were positively (Supplementary Fig. S4C). correlated with OLR1 (22, 30). Together, these results show that c-Myc promotes the transcription We next explored the relationships between OLR1 with c-Myc and of HMGA2 by directly binding to binding element 1 in the promoter of HMGA2. qRT-PCR and Western blot analysis confirmed that c-Myc HMGA2. and HMGA2 mRNA and protein expressions were upregulated and inhibited when OLR1 was overexpressed or knocked down, OLR1 upregulates HMGA2 through c-Myc to promote pancreatic respectively, in Bxpc-3 and Mia PaCa-2 cells (Fig. 4C–E; Supplemen- cancer metastasis tary Figs. S3A–S3C). To further determine the clinical relevance, we We next examined whether OLR1 increased transcription of investigated the expression of OLR1 along with c-Myc or HMGA2 in HMGA2 through c-Myc is involved in the effects of OLR1 in pro- patients with pancreatic cancer with RNA-seq data available from moting pancreatic cancer metastasis. We first examined whether c- TCGA and confirmed that c-Myc and HMGA2 mRNA expressions Myc promotes invasion and metastasis of pancreatic cancer through were positively correlated with OLR1 expression (r ¼ 0.24, P < increasing HMGA2. We transfected siHMGA2 in c-Myc- 0.001, Fig. 4F; r ¼ 0.31, P < 0.001, Fig. 4G; ref. 31). overexpressing Bxpc-3 and Mia Paca-2 cells and found that the Rescue experiments in pancreatic cancer cells in vitro showed that increased migration and invasion activities from c-Myc overexpres- the increased migration and invasion capability in Bxpc-3 cells medi- sion were significantly inhibited in the absence of HMGA2 (Fig. 6A; ated by OLR1 overexpression was reduced upon transfection with Supplementary Fig. S5A). HMGA2 overexpression also increased siRNA targeting c-Myc (sic-Myc) or HMGA2 (siHMGA2; Fig. 4H migration and invasion activities of Bxpc-3 and Mia Paca-2 cells with and I). Similar results were obtained in transwell and invasion assays, c-Myc knockdown (Fig. 6B; Supplementary Fig. S5B). These results which showed that c-Myc or HMGA2 overexpression prevented OLR1 suggested that HMGA2 is required for the effect of c-Myc in promot- inhibition-mediated effects on cell migration and invasion in Bxpc-3 ing metastasis of pancreatic cancer cells. Previous studies indicated cells (Fig. 4J and K). We also observed similar results in Mia PaCa-2 that EMT is closely related to metastasis, and therefore we explored cells (Supplementary Figs. S3D–S3G). whether HMGA2 regulates EMT to promote metastasis of pancreatic Taken together, these data indicate that OLR1 promotes the metas- cancer cells. The results demonstrated that HMGA2 enhanced the tasis of pancreatic cancer cells partly by inducing c-Myc and HMGA2 expression of b-catenin and snail and inhibited E-cadherin level in expression. Bxpc-3 and Mia PaCa-2 cells (Fig. 6C), indicating that HMGA2 may promote metastasis by inducing EMT of pancreatic cancer cells. c-Myc promotes the transcription of HMGA2 We next examined whether OLR1 regulation of HMGA2 occurred Our results showed that ORL1 enhances c-Myc and HMGA2 through c-Myc. The rescue result showed that HMGA2 induction by expression at both mRNA and protein levels. c-Myc is a classical OLR1 overexpression was compromised by sic-Myc in Bxpc-3 and Mia transcription factor and oncogene (32). Therefore, we speculated that PaCa-2 cells (Fig. 6D; Supplementary Fig. S5C). OLR1 may promote the transcription of HMGA2 through c-Myc. To Together these data indicated that OLR1 increases the invasion and verify our hypothesis, we examined the expression of HMGA2 in metastasis of pancreatic cancer cells by promoting the transcription of Bxpc-3 and Mia PaCa-2 cells with c-Myc silencing or overexpression HMGA2 through c-Myc activity. and found that c-Myc increased the expression of HMGA2 at both mRNA and protein levels (Fig. 5A and B; Supplementary Fig. S4A and High OLR1, c-Myc, and HMGA2 expression is associated with S4B). TCGA data also showed a positive correlation between c-Myc poor prognosis of patients with pancreatic cancer and HMGA2 (r ¼ 0.24; P < 0.001; Fig. 5C). To determine whether c- We next examined the clinical relationship of OLR1, c-Myc, and Myc directly binds to the promoter of HMGA2, we first analyzed the HMGA2 levels by analyzing the expression of c-Myc and HMGA2 in human HMGA2 promoter region (2,000 bp þ1 bp) and identified TMAs containing the same patient tissues as the TMA used for OLR1 two potential c-Myc binding sites containing the E box 50-CACGTG-30 analysis. The expressions of c-Myc and HMGA2 were significantly sequence (Fig. 5D). ENCODE (Encyclopedia of DNA Elements) c- higher in the OLR1 highly expressed group, and HMGA2 was signif- Myc ChIP-seq data showed that c-Myc was enriched at HMGA2 icantly increased in the c-Myc highly expressed group, indicating that promoter regions in MCF10A human breast cells, HepG2 hepatocel- they are positively correlated with each other (Fig. 6E–H). IHC lular carcinoma cells and Gm12878 lymphocyte cells (Fig. 5E; ref. 27). staining of subcutaneous and orthotopic xenograft tumors also To verify whether c-Myc binds directly to these potential binding sites revealed that the expressions of c-Myc and HMGA2 were positively in the HMGA2 promoter, we performed a ChIP-qPCR assay in 293T, correlated with the expression of OLR1 (Supplementary Figs. S6A and Bxpc-3 and Mia PaCa-2 cells. In chromatin fractions pulled down by S6B). Kaplan–Meier survival analysis demonstrated that high expres- anti-c-Myc antibody, the c-Myc binding elements 1 and 2 in the sion of c-Myc and HMGA2 were each negatively correlated with the HMGA2 promoter were significantly increased compared with sam- prognosis of patients with pancreatic cancer in our cohort and TCGA ples immunoprecipitated with control antibody (Fig. 5F and G). (Supplementary Fig. S7A–S7D). When OLR1, c-Myc, and HMGA2 To further determine whether c-Myc activates the promoter of were combined, Kaplan–Meier survival analysis showed the prognosis HMGA2, we constructed three luciferase reporter vectors driven by of patients with high expression of all three proteins was worse than the HMGA2 wild-type promoter (WT) or HMGA2 mutant pro- that of patients with two highly expressed proteins, and the prognosis moters (Mut1 and Mut2; Fig. 5D). Dual luciferase assays showed of patients with high expression of two of these proteins was worse that the activities of HMGA2 WT and Mut2 promoters significantly compared with patients with high expression of only one or none of increased when c-Myc was overexpressed (Fig. 5H). However, these proteins (Fig. 6I; Supplementary Figs. S7E–S7G). Analysis of activity of the HMGA2 Mut1 promoter showed no significant TCGA data also revealed a similar result (Fig. 6J; Supplementary Figs. change with coexpression of c-Myc. In cells were cotransfected S7H–S7J). These data revealed that the OLR1–c–Myc–HMGA2 axis with HMGA2 WT and Mut2 reporters and sic-Myc, the luciferase may serve as prognostic markers for patients with pancreatic cancer.

AACRJournals.org Mol Cancer Res; 18(5) May 2020 693

Yang et al.

Figure 5. c-Myc promotes the transcription of HMGA2. A and B, The expression of HMGA2 regulated by c-Myc was validated in Bxpc-3 cells by RT-qPCR (A) and Western blot analysis (B). C, HMGA2 expression positively correlated with c-Myc expression in the TCGA database of patients with pancreatic cancer (r ¼ 0.24; P < 0.001). D, Schematic of the HMGA2 gene promoter. Two potential c-Myc binding elements and their respective locations (top) and the design of mutants for each binding element are shown (middle and bottom). E, Evaluation of ENCODE c-Myc ChIP-sequencing data. Peaks represent c-Myc enrichment at the HMGA2 promoter in immortalized MCF10A mammary epithelial cells, HepG2 hepatocellular carcinoma, and Gm12878 lymphocyte cells. F and G, ChIP with antibodies against c-Myc or isotype control IgG and qRT-PCR for c-Myc binding element 1 (F)and2(G) was performed in 293T, Bxpc-3, and Mia PaCa-2 cells. H, Dual luciferase assay was performed in 293T, Bxpc-3, and Mia PaCa-2 cells with c-Myc overexpression and control cells transfected with pGL3 empty vector, pGL3-HMGA2 WT, pGL3-HMGA2- Mut1, or pGL3-HMGA2-Mut2. The results are shown as relative luciferase activity to pGL3 empty vector in corresponding cells after normalization with Renilla luciferase activity (, P < 0.05; , P < 0.01; , P < 0.001).

694 Mol Cancer Res; 18(5) May 2020 MOLECULAR CANCER RESEARCH

OLR1–c-Myc–HMGA2 Promotes Metastasis in Pancreatic Cancer

Figure 6. OLR1 upregulates HMGA2 through c-Myc to promote pancreatic cancer cell metastasis, and combined OLR1, HMGA2, and c-Myc are associated with poor prognosis of patients with pancreatic cancer. A and B, Rescue assays were performed with in vitro transwell and invasion experiments in Bxpc-3 cells. Representative images are shown. Magnification, 100; numbers of migrated and invasive cells are shown in the right panels. C, E-cadherin, Snail, and b-catenin level were assessed by Western blot analysis in Bxpc-3 and Mia PaCa-2 cells with HMGA2 overexpression or knockdown. D, c-Myc and HMGA2 protein levels were assessed in Bxpc-3 cells by rescue assay using Western blot analysis. E, Representative images of IHC staining of OLR1, c-Myc, and HMGA2 in the same pancreatic cancer tissue. All proteins are expressed at low levels in the two cases on the left and overexpressed in the two cases on the right. Magnification, 50; scale bars, 200 mm. The images in the top right panels represent magnified views of the boxed regions. Magnification, 200; scale bars, 100 mm. F–H, IHC score analysis results revealed that OLR1, c-Myc, and HMGA2 positively correlated with each other. I and J, Kaplan–Meier prognosis analysis of patients with pancreatic cancer with combined OLR1, c-Myc, and HMGA2 in TMA (H) and TCGA database (I). Group 1, one protein (OLR1, c-Myc, or HMGA2) is highly expressed; Group 2, two proteins (OLR1, c-Myc, and/or HMGA2) are highly expressed; Group 3, all three proteins (OLR1, c-Myc, and HMGA2) are highly expressed (, P < 0.05; , P < 0.01; , P < 0.001).

AACRJournals.org Mol Cancer Res; 18(5) May 2020 695

Yang et al.

Discussion Previous studies reported that OLR1 as a metabolism- and inflam- mation-related factor activates the NF-kB and PI3K/Akt path- In this study, we explored the expression and role of OLR1 in ways (11, 17), which is related to the regulation of c-Myc (20), pancreatic cancer. We first examined the expression of OLR1 in indicating OLR1 may regulate c-Myc through these pathways. For pancreatic cancer tissues and found that OLR1 expression was sig- example, previous studies identified OLR1 as one of the most upre- nificantly higher in pancreatic cancer than in adjacent normal gulated lipid metabolism genes during cellular transformation, and tissues. Clinicopathologic analysis showed that increased OLR1 inhibition of OLR1 reduced the activation of NF-kB through inhibi- expression was closely associated with poor differentiation, tion of IkBa phosphorylation in cancer cells (16). Furthermore, increased lymph node metastasis, and TNM staging, as well as another study showed that overexpression of OLR1 induced the reduced OS. Inhibition of OLR1 expression decreased the prolif- activation of NF-kB (9). However, the transcription of c-Myc is eration and metastasis of pancreatic cancer cells both in vitro and directly regulated by NF-kB, the activation of which could promote in vivo. Mechanistically, OLR1 upregulated the expression of c-Myc the transcription of c-Myc (20, 33). OLR1 was reported to drive PI3K/ to increase the transcription of HMGA2 gene, promoting the Akt/GSK3b activation (17), which could enhance the expression of c- metastasis of pancreatic cancer. Finally, we revealed that the Myc (20). In addition, OLR1 may regulate the expression of HMGA2 expression of OLR1, c-Myc, and HMGA2 were positively correlated and c-Myc through other mechanisms, such as miRNAs and with each other in pancreatic cancer tissues and negatively corre- lncRNAs (34). For instance, HMGA2, c-Myc, and OLR1 are targets lated with the prognosis of patients with pancreatic cancer. Alto- of let-7 (18, 35), and OLR1 may attenuate the inhibitory effect of let-7 gether, these data demonstrated that OLR1 increases the prolifer- on c-Myc and HMGA2 by competitive binding with let-7, thus ation and metastasis of pancreatic cancer cells by promoting the promoting the expression of c-Myc and HMGA2. However, further transcription of HMGA2 through c-Myc. studies are required to explore this possibility. OLR1 has been closely related with cardiovascular and metabolic Although some lipid-lowering agents were reported to inhibit the diseases. However, OLR1 was also recently reported to play an level of OLR1, as a potential link between metabolic-cardiovascular important role in cancer development, and the overexpression of diseases and cancer, no effective targeted drug for OLR1 is avail- OLR1 has been reported in multiple cancers, including pancreatic able (36). The development of specific inhibitors targeting OLR1 will cancer. For example, Jie and colleagues reported that OLR1 is a poor be useful to study the feasibility of OLR1 as a therapeutic target for prognostic indicator for pancreatic cancer and induced EMT, which is pancreatic cancer in the future. consistent with our result. However, the authors did not explore the In conclusion, we identified the OLR1–c-Myc–HMGA2 axis as a function and mechanism of OLR1 in pancreatic cancer metastasis (14). pathway that promotes metastasis of pancreatic cancer and that may To elucidate the mechanism of OLR1 in the metastasis of pancreatic serve as potential therapeutic targets of metastasis and prognosis cancer cells, we performed whole transcriptome RNA-seq analysis of markers for patients with pancreatic cancer. Bxpc-3 cells with OLR1 stable overexpression or inhibition and found that c-Myc and HMGA2 were positively correlated with the level of Disclosure of Potential Conflicts of Interest OLR1. We further established that OLR1 promotes the metastasis of No potential conflicts of interest were disclosed. pancreatic cancer through c-Myc and HMGA2. Previous reports showed that c-Myc and HMGA2 promote Authors’ Contributions cancer metastasis as oncogenic factors (22, 30), but the relationship Conception and design: G. Yang, T. Zhang, Y. Zhao between c-Myc and HMGA2 is rarely reported. Because c-Myc is a Development of methodology: G. Xiong, M. Feng, H. Wang transcription factor, we speculated that OLR1 increased HMGA2 Acquisition of data (provided animals, acquired and managed patients, provided expression through c-Myc to promote metastasis of pancreatic facilities, etc.): G. Yang, J. Yang cancer. We first demonstrated that the mRNA and protein levels Analysis and interpretation of data (e.g., statistical analysis, biostatistics, of HMGA2 were regulated by c-Myc. Dual luciferase assay and computational analysis): G. Yang, F. Zhao, J. Qiu, Z. Cao fi Writing, review, and/or revision of the manuscript: G. Yang, H. Wang, Y. Zhao ChIP assay con rmed that c-Myc promoted the transcription of Administrative, technical, or material support (i.e., reporting or organizing data, HMGA2 by directly binding to its promoter through one of two constructing databases): G. Xiong, M. Feng, F. Zhao, Y. Liu, L. You, L. Zheng potential binding sites. We also showed that HMGA2 was required Study supervision: Y. Zhao for c-Myc-promoted invasion and metastasis of pancreatic cancer cells and that HMGA2 promoted EMT in pancreatic cancer cells. Acknowledgments Together these results revealed that OLR1 increases invasion and T. Zhang was supported by grants from the National Natural Science Foundation metastasis of pancreatic cancer cells by promoting the transcription of China (Nos. 81772639 and 81972258) and CAMS Innovation Fund for Medical Sciences (CIFMS; No. 2016-I2M-1-001). Y. Zhao was supported by grants from the of HMGA2 through c-Myc. fi fi National Natural Science Foundation of China (No. 81974376) and the Non-pro t The clinical signi cance of OLR1, c-Myc, and HMGA2 was eval- Central Research Institute Fund of Chinese Academy of Medical Sciences (Nos. uated in pancreatic cancer tissues and TCGA database and the results 2018PT32014 and 2018PT32002); G. Yang was supported by grants from CAMS showed that OLR1, c-Myc, and HMGA2 were positively correlated Innovation Fund for Medical Students (No. 2017-1002-1-16); PUMC Youth Fund with each other and negatively correlated with the prognosis of and the Fundamental Research Funds for the Central Universities (No. 2017320027). pancreatic cancer patients. In addition, patients with pancreatic cancer We thank Dr. Y. Zhao for assistance with the ChIP assay. with high expression of all proteins showed worse prognosis compared The costs of publication of this article were defrayed in part by the payment of page with patients with high expression of two or fewer proteins. These data charges. This article must therefore be hereby marked advertisement in accordance suggest that OLR1, c-Myc, and HMGA2 might serve as potential with 18 U.S.C. Section 1734 solely to indicate this fact. prognosis markers for patients with pancreatic cancer. This study has several limitations. For example, the mechanism by Received July 13, 2019; revised November 23, 2019; accepted January 31, 2020; which OLR1 promotes the expression of c-Myc is still unclear. published first February 4, 2020.

696 Mol Cancer Res; 18(5) May 2020 MOLECULAR CANCER RESEARCH

OLR1–c-Myc–HMGA2 Promotes Metastasis in Pancreatic Cancer

References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin 2019;69: 18. Xiong G, Liu C, Yang G, Feng M, Xu J, Zhao F, et al. Long noncoding RNA 7–34. GSTM3TV2 upregulates LAT2 and OLR1 by competitively sponging let-7 to 2. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, et al. Cancer statistics in promote gemcitabine resistance in pancreatic cancer. J Hematol Oncol 2019; China, 2015. CA Cancer J Clin 2016;66:115–32. 12:97. 3. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian LM. 19. Dang CV. MYC on the path to cancer. Cell 2012;149:22–35. Projecting cancer incidence and deaths to 2030: the unexpected burden of 20. Grumont RJ, Strasser A, Gerondakis S. B cell growth is controlled by phospha- thyroid, liver, and pancreas cancers in the United States. Cancer Res 2014;74: tidylinosotol 3-kinase-dependent induction of Rel/NF-kappaB regulated c-myc 2913–21. transcription. Mol Cell 2002;10:1283–94. 4. Stathis A, Moore MJ. Advanced pancreatic carcinoma: current treatment and 21. Morishita A, Zaidi MR, Mitoro A, Sankarasharma D, Szabolcs M, Okada Y, et al. future challenges. Nat Rev Clin Oncol 2010;7:163–72. HMGA2 is a driver of tumor metastasis. Cancer Res 2013;73:4289–99. 5. Chen M, Masaki T, Sawamura T. LOX-1, the receptor for oxidized low-density 22. Dong J, Wang R, Ren G, Li X, Wang J, Sun Y, et al. HMGA2-FOXL2 axis regulates lipoprotein identified from endothelial cells: implications in endothelial dys- metastases and epithelial-to-mesenchymal transition of chemoresistant gastric function and atherosclerosis. Pharmacol Ther 2002;95:89–100. cancer. Clin Cancer Res 2017;23:3461–73. 6. Chui PC, Guan HP, Lehrke M, Lazar MA. PPARgamma regulates adipocyte 23. Hristov AC, Cope L, Reyes MD, Singh M, Iacobuzio-Donahue C, Maitra A, et al. cholesterol metabolism via oxidized LDL receptor 1. J Clin Invest 2005;115: HMGA2 protein expression correlates with lymph node metastasis and 2244–56. increased tumor grade in pancreatic ductal adenocarcinoma. Mod Pathol 7. Pothineni NVK, Karathanasis SK, Ding Z, Arulandu A, Varughese KI, Mehta JL. 2009;22:43–9. LOX-1 in atherosclerosis and myocardial ischemia: biology, genetics, and 24. Lee YS, Dutta A. The tumor suppressor microRNA let-7 represses the HMGA2 modulation. J Am Coll Cardiol 2017;69:2759–68. oncogene. Genes Dev 2007;21:1025–30. 8. Kita T. LOX-1, a possible clue to the missing link between hypertension and 25. Chen RX, Chen X, Xia LP, Zhang JX, Pan ZZ, Ma XD, et al. N(6)- atherogenesis. Circ Res 1999;84:1113–5. methyladenosine modification of circNSUN2 facilitates cytoplasmic export 9. CominaciniL,PasiniAF,GarbinU,DavoliA,TosettiML,CampagnolaM, and stabilizes HMGA2 to promote colorectal liver metastasis. Nat Commun et al. Oxidized low density lipoprotein (ox-LDL) binding to ox-LDL receptor- 2019;10:4695. 1 in endothelial cells induces the activation of NF-kappaB through an 26. Du Y, Liu Z, You L, Hou P, Ren X, Jiao T, et al. Pancreatic cancer progression increased production of intracellular reactive oxygen species. J Biol Chem relies upon mutant p53-induced oncogenic signaling mediated by NOP14. 2000;275:12633–8. Cancer Res 2017;77:2661–73. 10. Gao S, Geng YJ. LOX-1: a male hormone-regulated scavenger receptor for 27. ENCODE Project Consortium. An integrated encyclopedia of DNA elements in atherosclerosis. Vascul Pharmacol 2013;59:138–43. the human genome. Nature 2012;489:57–74. 11. Wang B, Zhao H, Zhao L, Zhang Y, Wan Q, Shen Y, et al. Up-regulation of 28. Zhao T, Jiang W, Wang X, Wang H, Zheng C, Li Y, et al. ESE3 inhibits OLR1 expression by TBC1D3 through activation of TNFalpha/NF-kappaB pancreatic cancer metastasis by upregulating E-cadherin. Cancer Res 2017; pathway promotes the migration of human breast cancer cells. Cancer Lett 77:874–85. 2017;408:60–70. 29. Yang G, Wang H, Feng M, You L, Zheng L, Zhang T, et al. Integrated analysis of 12.JiangL,JiangS,LinY,YangH,ZhaoZ,XieZ,etal.Combinationofbody gene expression and methylation profiles of novel pancreatic cancer cell lines mass index and oxidized low density lipoprotein receptor 1 in prognosis with highly metastatic activity. Sci China Life Sci 2019;62:791–806. prediction of patients with squamous non-small cell lung cancer. Oncotarget 30. Nowak DG, Cho H, Herzka T, Watrud K, DeMarco DV, Wang VM, et al. MYC 2015;6:22072–80. drives Pten/Trp53-deficient proliferation and metastasis due to IL6 secretion and 13. Gonzalez-Chavarría I, Fernandez E, Gutierrez N, Gonzalez-Horta EE, Sandoval AKT suppression via PHLPP2. Cancer Discov 2015;5:636–51. F, Cifuentes P, et al. LOX-1 activation by oxLDL triggers an epithelial mesen- 31. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. Integrative chymal transition and promotes tumorigenic potential in prostate cancer cells. analysis of complex cancer genomics and clinical profiles using the cBioPortal. Cancer Lett 2018;414:34–43. Sci Signal 2013;6:pl1. 14. Zhang J, Zhang L, Li C, Yang C, Li L, Song S, et al. LOX-1 is a poor prognostic 32. Kress TR, Sabo A, Amati B. MYC: connecting selective transcriptional control to indicator and induces epithelial-mesenchymal transition and metastasis in global RNA production. Nat Rev Cancer 2015;15:593–607. pancreatic cancer patients. Cell Oncol 2018;41:73–84. 33. Li W, Zhang L, Guo B, Deng J, Wu S, Li F, et al. Exosomal FMR1-AS1 facilitates 15. Liang M, Zhang P, Fu J. Up-regulation of LOX-1 expression by TNF-alpha maintaining cancer stem-like cell dynamic equilibrium via TLR7/NFkappaB/c- promotes trans-endothelial migration of MDA-MB-231 breast cancer cells. Myc signaling in female esophageal carcinoma. Mol Cancer 2019;18:22. Cancer Lett 2007;258:31–7. 34. Buendia MA, Bourre L, Cairo S. Myc target miRs and liver cancer: small 16. Hirsch HA, Iliopoulos D, Joshi A, Zhang Y, Jaeger SA, Bulyk M, et al. A molecules to get Myc sick. Gastroenterology 2012;142:214–8. transcriptional signature and common gene networks link cancer with lipid 35. Manier S, Powers JT, Sacco A, Glavey SV, Huynh D, Reagan MR, et al. The metabolism and diverse human diseases. Cancer Cell 2010;17:348–61. LIN28B/let-7 axis is a novel therapeutic pathway in multiple myeloma. Leukemia 17. Li C, Zhang J, Wu H, Li L, Yang C, Song S, et al. Lectin-like oxidized low-density 2017;31:853–60. lipoprotein receptor-1 facilitates metastasis of gastric cancer through driving 36. Khaidakov M, Mitra S, Kang BY, Wang X, Kadlubar S, Novelli G, et al. Oxidized epithelial-mesenchymal transition and PI3K/Akt/GSK3beta activation. Sci Rep LDL receptor 1 (OLR1) as a possible link between obesity, dyslipidemia and 2017;7:45275. cancer. PLoS One 2011;6:e20277.

AACRJournals.org Mol Cancer Res; 18(5) May 2020 697

Correction: OLR1 Promotes Pancreatic Cancer Metastasis via Increased c-Myc Expression and Transcription of HMGA2

In the original version of this article (1), Fig. 2C contains the following errors:

* the Bxpc-3/migration/pcDNA3.1 image is an erroneous duplication of the Bxpc-3/migration/siNC&HMGA2 image in Fig. 4K; * the same image was used to represent the Mia PaCa-2/invasion/OLR1 and Mia PaCa-2/migration/siOLR1 images; and * an errant label reading “72 hours” appears amid images in the panel.

The inadvertently reused images have been replaced with the correct versions and the misplaced label has been removed in the latest online HTML and PDF versions of the article. The authors regret these errors.

Reference 1. Yang G, Xiong G, Feng M, Zhao F, Qiu J, Liu Y, et al. OLR1 promotes pancreatic cancer metastasis via increased c-Myc expression and transcription of HMGA2. Mol Cancer Res 2020;18:685–97.

Published online August 4, 2021. Mol Cancer Res 2021;19:1438 doi: 10.1158/1541-7786.MCR-21-0435 Ó2021 American Association for Cancer Research

AACRJournals.org | 1438 Published OnlineFirst February 4, 2020; DOI: 10.1158/1541-7786.MCR-19-0718

OLR1 Promotes Pancreatic Cancer Metastasis via Increased c-Myc Expression and Transcription of HMGA2

Gang Yang, Guangbing Xiong, Mengyu Feng, et al.

Mol Cancer Res 2020;18:685-697. Published OnlineFirst February 4, 2020.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-19-0718

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2020/02/04/1541-7786.MCR-19-0718.DC1

Cited articles This article cites 36 articles, 11 of which you can access for free at: http://mcr.aacrjournals.org/content/18/5/685.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/18/5/685.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/18/5/685. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.