LOXL2 Status Correlates with Tumor Stage and Regulates Integrin Levels to Promote Tumor Progression in Ccrcc

Published OnlineFirst August 4, 2014; DOI: 10.1158/1541-7786.MCR-14-0233

Molecular Cancer Oncogenes and Tumor Suppressors Research

LOXL2 Status Correlates with Tumor Stage and Regulates Integrin Levels to Promote Tumor Progression in ccRCC

Hiroaki Hase1, Kentaro Jingushi1, Yuko Ueda1, Kaori Kitae1, Hiroshi Egawa1, Ikumi Ohshio1, Ryoji Kawakami1, Yuri Kashiwagi1, Yohei Tsukada1, Takumi Kobayashi1, Wataru Nakata2, Kazutoshi Fujita2, Motohide Uemura2, Norio Nonomura2, and Kazutake Tsujikawa1

Abstract Clear cell renal cell carcinoma (ccRCC) is the most common histologically defined subtype of renal cell carcinoma (RCC). To define the molecular mechanism in the progression of ccRCC, we focused on LOX-like protein 2 (LOXL2), which is critical for the first step in collagen and elastin cross-linking. Using exon array analysis and quantitative validation, LOXL2 was shown to be significantly upregulated in clinical specimens of human ccRCC tumor tissues, compared with adjacent noncancerous renal tissues, and this elevated expression correlated with the pathologic stages of ccRCC. RNAi-mediated knockdown of LOXL2 resulted in marked suppression of stress-fiber and focal adhesion formation in ccRCC cells. Moreover, LOXL2 siRNA knockdown significantly inhibited cell growth, migration, and invasion. Mechanistically, LOXL2 regulated the degradation of both integrins a5 (ITGAV5) and b1 (ITGB1) via protease- and proteasome-dependent systems. In clinical ccRCC specimens, the expression levels of LOXL2 and integrin a5 correlated with the pathologic tumor grades. In conclusion, LOXL2 is a potent regulator of integrin a5 and integrin b1 protein levels and functions in a tumor-promoting capacity in ccRCC. Implications: This is the first report demonstrating that LOXL2 is highly expressed and involved in ccRCC progression by regulating the levels of integrins a5 and b1. Mol Cancer Res; 12(12); 1807–17. 2014 AACR.

Introduction controls protein levels of hypoxia-inducible factor (HIF), a Renal cell carcinoma (RCC) is the leading cause of death transcription factor and master regulator of the cellular among urological malignancies, and clear cell renal cell response to hypoxia. When VHL is absent, HIF accumu- carcinoma (ccRCC) is the most common histologic subtype lates, even under normoxic conditions, and inappropriately of RCC. Early-stage ccRCC is usually curable by clinical transactivates target gene expression. Because HIF target genes are potently tumorigenic, HIF target gene expression surgery, but a large number of early-stage ccRCC cases are ﬁ asymptomatic, with approximately one-third of all patients is considered the primary orchestrator of VHL-de cient presenting with locally metastatic cancer at the time of ccRCC tumor progression. Because of this, many diagnosis (1). Therefore, a better understanding of the approaches have been attempted that target downstream molecular mechanisms of ccRCC progression is crucial for molecules of HIF for ccRCC therapeutics. Molecular tar- the discovery of novel prognostic markers and targeted geted drugs for ccRCC therapy are developed on the basis of therapies. such ideas, but these agents only delay disease progression A genetic hallmark of ccRCC is the inactivation of the and are not curative (2). Therefore, it is necessary to clarify von Hippel–Lindau (VHL) tumor suppressor gene. VHL critical molecular mechanisms in ccRCC tumorigenesis and functions as a part of an E3 ubiquitin ligase complex that progression. LOX-like protein 2 (LOXL2) is a member of the lysyl oxidase gene family, comprising ﬁve members, LOX and 1Laboratory of Molecular and Cellular Physiology, Graduate School of four LOX-like proteins (LOXL1–4), and functions both Pharmaceutical Sciences, Osaka University, Yamadaoka, Suita, Osaka, intracellularly and extracellularly. Secreted LOXL2 catalyzes Japan. 2Department of Urology, Graduate School of Medicine, Osaka University, Yamadaoka, Osaka, Japan. the covalent crosslinking of the collagen and elastin component side chains via its lysyl oxidase activity, thereby Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/). stabilizing the proteins in the extracellular matrix (ECM; ref. 3). On the other hand, intracellular LOXL2 down- Corresponding Author: Kentaro Jingushi, Laboratory of Molecular and Cellular Physiology, Graduate School of Pharmaceutical Sciences, Osaka regulates E-cadherin expression by cooperating with the University, 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan. Phone: 81-6- transcriptional factor Snail (4, 5) or regulating the methyl- 6879-8192; Fax: 81-6-6879-8190; E-mail: [email protected] ation status of the E-cadherin gene (6), thereby promoting doi: 10.1158/1541-7786.MCR-14-0233 epithelial–mesenchymal transition (EMT). Moreover, 2014 American Association for Cancer Research. LOXL2 was recently shown to deaminate lysine 4 in histone

www.aacrjournals.org 1807

Hase et al.

H3 to transcriptionally control the E-cadherin gene (CDH1; Table 1. ccRCC clinical samples used for exon ref. 7). Through these mechanisms, LOXL2 is involved in array analysis cancer progression. However, to our knowledge, no reports have investigated the expression and function of LOXL2 in ccRCC. Patient Patient Patient Patient Integrins are ubiquitously expressed transmembrane A B C D receptors of at least 24 heterodimers composed of 18 Age, y 67 86 60 66 a-subunits and eight b-subunits (8). Integrins function in Gender Male Female Female Male both cell–cell interactions, through binding to cadherins, Pathologic stage T1 T3 T3 T4 and cell–ECM interactions, involving fibronectin, collagen, Pathologic grade G2 G2 G2 G3 and laminin (9, 10). Hence, deregulated integrin expression TNM classification I III IV IV or function correlates with cancer progression by enhancing cancer cell invasion, metastasis, and survival. For example, loss of integrin b1 expression blocks tumor progression in using 31 paired ccRCC samples and 90 tumor specimens. different tumor models (11, 12), and increased expression of Clinical and pathologic data for the clinical samples are integrin a5/b1 is associated with poor prognosis in early- presented in Tables 1 and 2 for exon array analysis and qPCR stage non–small cell lung cancer (13). However, the regu- analysis, respectively. lation mechanisms of integrin expression remain unclear in cancer cells. Exon array analysis In the present study, we found that LOXL2 was signif- Total RNA (100 ng) was subjected to cDNA synthesis icantly upregulated in ccRCC tissues compared with using the Ambion WT expression kit (Life Technologies). normal renal tissues. LOXL2 knockdown suppressed pro- The obtained 5.5 mg of cDNA were biotinylated and then liferation, migration, and invasion by inhibiting focal fragmented with the GeneChip WT terminal labeling kit adhesion complex and stress fiber formation in ccRCC (Affymetrix). After creating the hybridization cocktail, bio- cells. We also clarified that intracellular LOXL2 is involved tinylated cDNAs were hybridized with the GeneChip in the stabilization of integrin a5 and integrin b1through human Exon 1.0 ST array (Affymetrix) for 17 hours. After suppression of the protease and proteasome systems. hybridization, the GeneChip was washed and stained using Moreover, we show that the protein levels of LOXL2 and the GeneChip hybridization wash and stain kit (Affymetrix) integrin a5weresignificantly increased in ccRCC speci- and the GeneChip Fluidics Station (Affymetrix). The mens.Tothebestofourknowledge,thisisthefirst report showing that the highly expressed LOXL2 is involved in Table 2. ccRCC clinical samples used for qPCR ccRCC progression by regulating the expression of integrins a5andb1. analysis

Matched-pair Tumor only Materials and Methods sample sample Chemicals and antibodies Age, y The polyclonal anti-LOXL2 antibody was purchased Mean 65 65.5 from Abcam. The polyclonal anti-FAK, anti-integrin a5, Range 34–86 34–82 and anti-integrin b1 antibodies were purchased from Santa Gender Cruz Biotechnology. The monoclonal anti-phospho-FAK Male 20 58 576 (Tyr ) and anti-b-actin antibodies were purchased from Female 11 31 Sigma. Leupeptin was purchased from Tocris Bioscience, Pathologic stage and MG132 was purchased from Sigma. Normal 23 0 T1 19 33 Clinical specimens T2 3 23 ccRCC clinical specimens were obtained from patients T3 8 33 during primary curative resection at Osaka University Med- T4 1 1 ical Hospital in Japan. Tumor-associated normal renal Pathologic grade tissues were also obtained from a subset of these patients G1 12 26 where possible. Prior written and informed consent was G2 15 57 obtained from each patient, and the study was approved by G3 4 7 the Ethics Review Board of Osaka University Medical TNM classiﬁcation Hospital. Following excision, tissue samples were immedi- I1927 ately immersed in RNAlater (Qiagen) and stored at 20 C II 3 11 until RNA extraction. mRNA was puriﬁed using the RNeasy III 6 15 mini kit (Qiagen). An exon array was conducted using four IV 3 37 paired samples. LOXL2 expression was validated by qPCR

1808 Mol Cancer Res; 12(12) December 2014 Molecular Cancer Research

LOXL2 Regulates Integrin a5/b1 in ccRCC Cells

GeneChip was then scanned by the GeneChip Scanner 3000 Fluorescence microscopy (Affymetrix). Gene Spring 12.1 (Agilent Technologies) was Cells, plated on coverslips, were incubated for the indi- used for the data analysis. cated times and washed with PBS. The cells were fixed and permealized in 4% formaldehyde for 15 minutes at room Cell culture temperature and then washed twice with PBS. After blocking The four human ccRCC cell lines, 786-O, Caki-1, Caki- with 5% skim milk for 1 hour, the coverslips were incubated 2, and ACHN, were cultured in RPMI1640 medium with primary antibody at 4C. The coverslips were washed (Wako) supplemented with 10% FBS, 100 U/mL penicillin twice with PBS and incubated with fluorochrome-conju- G, and 0.1 mg/mL streptomycin. The HEK 293T cell line gated secondary antibody for 1 hour at room temperature. was cultured in DMEM medium (Wako) supplemented The cells were examined under a fluorescence microscope with 10% FBS, 100 U/mL penicillin G, and 0.1 mg/mL (Olympus). streptomycin. Normal human renal epithelial cells (Kurabo) were cultured with RenaLife BM medium supplemented Construction of LOXL2 expression vector with RenaLife LifeFactor (Kurabo). The LOXL2 sequence was PCR amplified from human genomic DNA followed by cloning into the pDEST26 qPCR for LOXL2 and integrin a5 mRNA expression vector (Invitrogen). The primer sequences for gene amp- 0 Total RNA was isolated by using TRIzol reagent (Invi- lification were as follows: forward, 5 -CACCATGGA- 0 0 trogen). PrimeScript RT reagent Kit (Takara) was used to GAGGCCTCTGTGCTCCCACC-3 , and reverse, 5 - prepare cDNA from 500 ng total RNA. The light cycler TCG AGG ATC CGG CCC AAT CCC TGC TTC CAC 0 (Roche) was used for qPCR analysis. Thermal cycling T-3 . conditions included an initial step at 95Cfor30seconds and 50 cycles of 94C for 15 seconds, 60Cfor10 Wound-healing assay seconds, and 72C for 15 seconds. Primer sequences for Cell migration was examined using a wound-healing gene amplification were as follows: LOXL2 forward, 50- assay. 786-O cells transfected with siRNAs were seeded in 5 TTCCCTGGGGAGAGGGACATACA-30, and reverse, a 24-well plate (1.0 10 cells/well) and incubated for 72 50-TTCTTCATGGGGTCCAGTGACA-30,integrina5 hours. A wound was created in a monolayer of approximately forward, 50-GGCAGCTATGGCGTCCCACTGTGG-30, 90% confluent 786-O cells, using a sterile 1 mL pipette tip. and reverse, 50-GGCATCAGAGGTGGCTGGAGGCT- Cell migration pictures were recorded at 0 and 12 hours, T-30, and integrin b1 forward, 50- GACAAATTACCC- using an Olympus IX71 fluorescence Microscope. CAGCCGGTCCAAC-30, and reverse, 50- GGCCAATAA- GAACAATTCCAGCAACCA-30. Cell proliferation and cell invasion assay Cell proliferation and cell invasion assay were each siRNA and DNA transfection carried out using E-plate and CIM-plate, respectively, siRNA duplexes used to downregulate LOXL2 and Integ- and analyzed by xCELLigence system (Roche) according rin a5 mRNA expression (LOXL2 siRNA #1: GCCACA- to the manufacturer. TAGGTGGTTCCTTCATT, LOXL2 siRNA #2: AGT- AAAGAAGCCTGCGTGGTCTT, Integrin a5 siRNA #1: Cell membrane biotinylation CAGAUAACUUCACCCGAAUTT, and Integrin a5 The cell membranes of 786-O cells that were transfected siRNA #2: GCAGAUUGCAGAAUCUUAUTT) were with LOXL2 siRNA or control siRNA for 24 hours were purchased from Sigma-Aldrich. The negative control biotinylated with membrane-impermeable EZ-LINK Sulfo- fi siRNA duplex (AGTAAAGAAGCCTGCGTGGTCTT) NHS-SS-Biotin (Thermo Scienti c) on ice for 30 minutes. was purchased from B-Brige. For all siRNA transfection Cells were washed with PBS, lysed by using lysis buffer (50 4 mmol/L Tris-HCl, pH7.5, 150 mmol/L NaCl, 1% Tri- studies, 786-O cells were seeded at a density of 2 10 cells/cm2 and 10 nmol/L siRNA was transfected using tonX-100), and then incubated for indicated time at 37 C. Lipofectamine RNAiMAX reagent (Life Technologies). For Biotinylated proteins were precipitated with streptavidin DNA transfection, 2 104 HEK 293T cells were seeded in beads and then subjected to Western blot analysis. a 48-well plate 24 hours before transfection. DNA transfection was performed using HilyMax transfection reagent Statistical analysis (Wako). Results are expressed as the mean SD or SE of the mean. Differences between values were statistically analyzed using a t posthoc Western blotting Student test or one-way ANOVA with Bonferroni P < Samples were separated by 7.5% to 15% SD-PAGE tests (GraphPad Prism 6.0, GraphPad Software). A value fi and then transferred to a polyvinylidene difluoride mem- 0.05 was considered statistically signi cant. brane using a semidry transfer system (Bio-Rad). Immu- noreactive proteins were visualized with ECL Prime Results Western blotting detection reagent (GE Healthcare). High expression of LOXL2 mRNA in ccRCC specimens Densitometric analysis was performed using the NIH To obtain the mRNA expression signature of ccRCC, ImageJ software. we performed an exon array analysis using four ccRCC

www.aacrjournals.org Mol Cancer Res; 12(12) December 2014 1809

Hase et al.

AB ×10–3 ×10–2 20 3

15 2 10 Figure 1. LOXL2 mRNA expression was upregulated in ccRCC 1 samples. LOXL2 mRNA expression LOXL2/GAPDH LOXL2/GAPDH 5 was examined in 31 matched-pair ccRCC clinical samples (A) and 121 0 0 ccRCC clinical samples (B–D) by Normal Tumor T1 T2 T3,T4 qPCR. Data are the relative expression normalized to GAPDH CD mRNA. The relative expression of LOXL2 mRNA was compared by T –2 –2 ×10 ×10 stages (B), pathologic stages (C), 3 3 and pathologic grades (D). Values are mean SD; , P < 0.05; , P < 0.001. 2 2

1 1 LOXL2/GAPDH LOXL2/GAPDH

0 0 I,II,III IV G1 G2 G3

specimens. Because of its novelty and importance in the were observed, and therefore, the knockdown cells were cancer biology of ccRCC, LOXL2, which showed a marked stained with anti-b-actin antibody. LOXL2 siRNAs knock- increase in the ccRCC tissues compared with the adjacent down markedly reduced the cell extensibility and stress fiber normal renal tissues (data not shown), was selected for formation, compared with the control siRNA (Fig. 2C and further study. We first conducted qRT-PCR analysis of D). Stress fibers are connected to focal adhesion complexes, LOXL2 mRNA in 31 matched-pair ccRCC samples. As consisting of integrins and focal adhesion kinase (FAK; shown in Fig. 1A, ccRCC tissues significantly expressed ref. 14). Therefore, we investigated the effect of LOXL2 more LOXL2 mRNA compared with the matched adjacent knockdown on focal adhesion formation by staining with normal renal tissues, with an approximately 4-fold increase. anti-FAK antibody and observed an unusual diffuse FAK Next, the relationship between LOXL2 mRNA levels and localization (Fig. 2D). Moreover, Western blot analysis the histopathological findings in 121 ccRCC tissue samples showed that the level of Tyr576-phosphorylated FAK, indi- was analyzed. LOXL2 mRNA expression significantly cor- cating activated state of FAK (15), was also decreased upon related with tumor (T) stage (Fig. 1B) and pathologic stage LOXL2 knockdown (Fig. 2E). These results suggest that (Fig. 1C), but not with the pathologic grades (Fig. 1D). LOXL2 regulates the formation of stress fibers and focal These results suggest a relationship between LOXL2 mRNA adhesion complexes in 786-O cells. expression and ccRCC progression. LOXL2 upregulates 786-O cell proliferation, migration, LOXL2 siRNA knockdown suppresses stress fiber and and invasion focal adhesion complex formation in 786-O cells. Because LOXL2 knockdown downregulated FAK activa- To investigate LOXL2 function in ccRCC cells, we tion, which is a key upstream regulator of vast signal evaluated LOXL2 mRNA expression levels in four ccRCC transduction (16), we examined the effect of LOXL2 knock- cell lines. Among the ccRCC cell lines, the highest LOXL2 down on additional 786-O cell phenotypes. As shown mRNA expression was observed in 786-O cells (Fig. 2A), in Fig. 3A–C, the examined LOXL2 siRNAs markedly and therefore, we analyzed LOXL2 function in 786-O cells inhibited 786-O cell proliferation, migration, and invasion. using an RNAi method. To assess the knockdown efficiency These results indicated that LOXL2 plays an important role of LOXL2 siRNAs on its endogenous expression, qPCR and in the proliferation, migration, and invasion of 786-O cells. Western blot analysis were conducted. As shown in Fig. 2B, LOXL2 expression was sufficiently reduced at both the LOXL2 regulates integrin a5 expression in 786-O cells mRNA and protein levels. During the experiment, mor- Because LOXL2 knockdown reduced focal adhesion for- phologic changes of the LOXL2-knockdowned 786-O cells mation, we then focused on the integrin family, a family of

1810 Mol Cancer Res; 12(12) December 2014 Molecular Cancer Research

LOXL2 Regulates Integrin a5/b1 in ccRCC Cells

A ×103 BC×103 25 6 siRNA CNT 20 5 4 15 3 10 2

LOXL2/GAPDH 5 LOXL2/GAPDH 1 siRNA LOXL2 #1 siRNA LOXL2 #2 (kDa) 0 0 100 LOXL2 75

786-O ACHN Caki1 Caki2 b-Actin 48 35 siRNA CNT #1 #2 DELOXL2 siRNA b-Actin FAK Merge (kDa) 130 P-FAK (Tyr576) CNT 100 FAK b-Actin

LOXL2 siRNA CNT #1 #2 #1 LOXL2

LOXL2 #2

Figure 2. LOXL2 knockdown reduced stress fiber and focal adhesion formation in 786-O cells. A, LOXL2 mRNA expression in four ccRCC cell lines was examined by qPCR. Values are mean SD of three independent experiments. B, 786-O cells were transfected with control (CNT) siRNA or two LOXL2 siRNAs for 48 hours. LOXL2 mRNA and protein expression levels were examined by qPCR and by Western blotting with anti-LOXL2 and anti-b-actin antibodies, respectively. For qPCR analysis, values are mean SD of three independent experiments. For Western blot analysis, representative pictures of three independent experiments are shown. C and D, 786-O cells were transfected with control (CNT) siRNA or two LOXL2 siRNAs for 48 hours, and then immunofluorescence staining of the cells was performed with the anti-b-actin (C) and with the anti-b-actin and anti-FAK antibodies (D). Nuclear staining was performed using 40,6-diamidino-2-phenylindole (DAPI). Representative pictures of three independent experiments are shown. White bars indicate 50 mmfor C and 10 mm for D. E, protein lysates from 786-O cells transfected with control (CNT) siRNA or LOXL2 siRNA for 48 hours were subjected to Western blot analysis with anti-phospho-FAK (Tyr576), anti-FAK, and anti-b-actin antibodies. The blots shown are representative of three independent experiments. focal adhesion core proteins that function as crucial pro- into 786-O cells and examined the phenotypes. As shown moters of various cancers (8). Our exon array data showed in Fig. 4D and E, integrin a5 knockdown reduced both that integrin a5 expression was most upregulated among the stress fibers and focal adhesion formation, which were similar integrin family molecules in all four ccRCC tissues com- phenotypes to those of LOXL2 knockdown, as shown in Fig. pared with adjacent normal renal tissues (Supplementary 2D. Although integrin a5 knockdown had no effect on Fig. S1). We confirmed integrin a5 expression levels by proliferation of 786-O cells, cell migration was significantly qPCR. Interestingly, correlation (r ¼ 0.60) was shown reduced (Fig. 4F and G). In addition to the LOXL2 gene, between LOXL2 and integrin a5 mRNA levels in ccRCC our exon array analysis also showed high LOX gene expres- specimens (Fig. 4A). Next, we investigated the effect of sion in the four ccRCC clinical samples (Supplementary LOXL2 knockdown on integrin a5 expression. LOXL2 Fig. S3). To investigate whether LOX also regulates integrin siRNAs knockdown decreased integrin a5 mRNA expres- a5 expression and cell migration, we transfected LOX sion (Fig. 4B) and markedly reduced integrin a5 protein siRNA into Caki1 cells and examined the resulting pheno- levels in 786-O cells (Fig. 4C). These results suggest that the types. LOX knockdown had no significant effect on either phenotypes of LOXL2 knockdown 786-O cells may be integrin a5 expression or cell migration ability, but it did manifested through its effects on integrin a5 expression. suppress focal adhesion formation (Supplementary Fig. S4). To test this hypothesis, we transfected integrin a5 siRNA These results indicate that LOXL2, but not LOX, maintains

www.aacrjournals.org Mol Cancer Res; 12(12) December 2014 1811

Hase et al.

AB8 45 C3 7 40 2.5 6 35 30 2 5 25 4 1.5 20 3 1 Cell index 15 Cell index 2 10

Wound healing (%) Wound 0.5 1 5 0 0 0 siRNA CNT #1 #2 siRNA CNT #1 #2 siRNA CNT #1 #2 LOXL2 LOXL2 LOXL2

Figure 3. LOXL2 knockdown suppressed proliferation, migration, and invasion of 786-O cells. 786-O cells were transfected with control (CNT) siRNA or LOXL2 siRNA for 24 hours. A, the transfected cells were inoculated in an E-plate, and the cell proliferation was monitored in real-time for 72 hours using the xCELLigence system. Values expressed as cell index are the mean SD of duplicated experiments. B, the transfected cells were replated at 90% conﬂuence, and cell migration was measured 12 hours after a wound was formed by scraping. Values are means SD of three independent experiments. C, the transfected cell suspension was added to the upper chamber of a Matrigel-coated CIM-plate. The lower chamber was ﬁlled with medium and then cultured for 18 hours in the xCELLigence system. Values expressed as cell index are mean SD of duplicated experiments.

the cytoskeleton and promotes migration of ccRCC cells by protein expression in the LOXL2 knockdown 786-O cells. regulating integrin a5 expression. As shown in Figs. 6A and B, both leupeptin and MG132 treatment resulted in an increase in integrin a5andb1 LOXL2 induces 786-O cell adhesion to fibronectin via proteins, compared with LOXL2 siRNA knockdown integrin a5b1 expression alone, in a dose-dependent manner. These results suggest Integrin a5 associates with integrin b1 to form a hetero- that the decreased LOXL2 expression led to degradation of dimeric receptor (17). Therefore, we examined integrin b1 integrin a5andb1 proteins via the protease and protea- mRNA and protein expression levels. However, LOXL2 some systems in 786-O cells. Although the precise mechan- knockdown had no effect on integrin b1 mRNA expression isms remain largely unknown, internalized integrins a5b1 (Fig. 5A). In Western blot analysis, two bands of integrin b1, are either returned to the cell membrane or routed to a premature (lower band) and a mature (upper band) form, lysosomes for degradation, and binding to fibronectin were detected. Interestingly, mature integrin b1 expression regulates the decision to recycle or degrade internalized was markedly suppressed upon LOXL2 silencing, but the integrin a5b1 by ubiquitination (19). To clarify whether premature form was unaffected (Fig. 5B). On the other LOXL2 affects integrin a5b1 recycling, we then conducted hand, LOXL2 overexpression increased the protein expres- a cell membrane biotinylation assay. As shown in Fig. 6C, sion of integrin a5 as well as both forms of integrin b1in LOXL2 knockdown did not affect integrin a5protein HEK293T cells (Fig. 5C). From these results, it was pre- levels in the cell membrane fraction. Moreover, although sumed that LOXL2 expression was associated with the LOXL2 knockdown reduced integrin a5andb1protein function of integrin a5/b1 in ccRCC cells. Because the expression in the whole-cell lysate of 786-O cells, protein integrin a5/b1 heterodimer specifically binds to fibronectin levels in the cell membrane fraction were comparable (18), we performed a fibronectin adhesion assay using 786- between LOXL2 knockdown and control cells (Fig. O cells to examine the effects of LOXL2 and integrin a5 6D). These results suggest that LOXL2 regulates the siRNAs. Both LOXL2 and integrin a5 siRNA knockdowns expression of integrin a5andb1 proteins not via recycling significantly reduced the adhesive ability of 786-O cells to a system, but via protease- and proteasome-dependent sys- fibronectin-coated plate (Fig. 5D). These results supported tems in 786-O cells. our hypothesis that LOXL2 modulates a function of integrin a5/b1 in ccRCC cells. The relationship between LOXL2 expression and integrin a5b1 expression in ccRCC clinical samples LOXL2 knockdown downregulates integrin a5 and b1 To examine the expression of LOXL2, integrin a5, and proteins in a protease- and proteasome-dependent integrin b1 in ccRCC, clinical ccRCC tumor samples manner in 786-O cells were lysed and blotted with anti-LOXL2, anti-integrin Because LOXL2 knockdown reduced the protein expres- a5, and anti-integrin b1antibodies.Compared with sion of integrin a5 and b1 subunits, we then examined the normal human renal epithelial cell lysates, integrin b1 underlying regulatory mechanism of integrin protein expres- expression protein was comparable, but both LOXL2 and sion. First, we utilized the protease inhibitor leupeptin and integrin a5 protein expression was elevated in ccRCC the proteasome inhibitor MG132 on integrin a5 and b1 samples (Fig. 7).

1812 Mol Cancer Res; 12(12) December 2014 Molecular Cancer Research

LOXL2 Regulates Integrin a5/b1 in ccRCC Cells

A BC 30 r = 0.60 (kDa) ×10–3 25 130 8 Integrin a5 20 100 b-Actin 15 6 siRNA CNT #1 #2 10 LOXL2 5 4 F siRNA CNT 0 siRNA Integrin a5 #1

ITG a 5 mRNA expression –5 0 5 10 15 20 2 –5 ITG a 5 / GAPDH siRNA Integrin a5 #2

fold change (tumor vs. normal) (tumor vs. change fold LOXL2 mRNA expression 1.5 fold change (tumor vs. normal) 0 ) D siRNA CNT #1 #2

630 nm 1.0 LOXL2 Integrin a5 - A b-Actin 0.5 450 nm Absorbance

siRNA CNT #1 #2 (A Integrin a5 0.0 E 02468 b Day siRNA -Actin FAK Merge G 60

CNT 40

Integrin a5 #1 20 Wound healing (%) Wound 0 Integrin a5 #2 siRNA CNT #1 #2 Integrin a5

Figure 4. LOXL2 knockdown downregulated integrin a5 expression in 786-O cells. A, the expression of LOXL2 and integrin a5 mRNAs was examined in 31 matched-pair samples of ccRCC by qPCR, and the relationship was evaluated. 786-O cells were transfected with control (CNT) siRNA or two LOXL2 siRNAs for 48 hours. B, integrin a5 mRNA expression was examined by qPCR and normalized to GAPDH mRNA expression. C, integrin a5 protein expression was analyzed by Western blot analysis with an anti-integrin a5 antibody and normalized to b-actin protein. The blots shown are representative of three independent experiments. D, protein lysates from 786-O cells transfected with control (CNT) siRNA or two integrin a5 siRNAs for 48 hours were subjected to Western blot analysis with integrin a5 and anti-b-actin antibodies. The blots are representative of three independent experiments. E, 786-O cells were transfected with control (CNT) siRNA or two integrin a5 siRNAs for 48 hours, and immunoﬂuorescence staining was performed with anti-b-actin and anti-FAK antibodies. Nuclear staining was performed using 40,6-diamidino-2-phenylindole (DAPI). Representative pictures of three independent experiments are shown. The white bar indicates 10 mm. F, 786-O cells were transfected with control (CNT) siRNA or two integrin a5 siRNAs for 24 hours, and then, cell proliferation was examined by WST-1 assay. Values are means SD of three independent experiments. G, the transfected cells were reseeded at 90% conﬂuence, and cell migration was measured 12 hours after a wound was formed by scraping. Values are mean SD of three independent experiments. , P < 0.05 versus control siRNA.

Discussion (3, 5). Although we conﬁrmed the reduced snail expression To develop a novel therapeutic drug for ccRCC, it is upon LOXL2 knockdown, its effects on E-cadherin and crucial to clarify the underlying molecular mechanisms N-cadherin expression were not observed in 786-O cells involved in the development and progression of ccRCC. (Supplementary Fig. S2). On the other hand, we found In the present study, we showed that LOXL2 was that LOXL2 knockdown suppressed focal adhesion complex highly expressed in ccRCC clinical samples and induced formation and reduced FAK phosphorylation (Fig. 2C–E). proliferation, migration, and invasion partly through FAK critically regulates cell migration and invasion in several the suppression of integrin a5b1degradationthrough cancers by regulating focal adhesion complex formation and a protease and proteasome-dependent system in ccRCC matrix metalloproteinase expression (20). These results cells. suggest that the highly expressed LOXL2 promotes migra- LOXL2 promotes EMT by reducing E-cadherin expres- tion and invasion of ccRCC cells not via EMT progression sion via the upregulation of snail expression, which leads to but via regulation of stress ﬁber and focal adhesion complex cell shape extension and the acceleration of invasion activity formation.

www.aacrjournals.org Mol Cancer Res; 12(12) December 2014 1813

Hase et al.

(kDa) 130 ABIntegrin b1 C LOXL2 ×10–3 100 a 20 b-Actin Integrin 5

siRNA CNT #1 #2 Integrin b1 15 LOXL2 b-Actin Mature integrin b1 10 Vector Mock LOXL2 b 1.5 Premature integrin 1

ITG b 1/GAPDH 5 D 1.0 ×105 0 50 siRNA CNT #1 #2 0.5 40 LOXL2

(normalized to b -Actin) 30 0.0 Relative integrin b 1 expression siRNA CNT #1 #2 20 LOXL2 10 Fluorescence intensity 0 siRNA CNT #1 #2 #1 #2 LOXL2 Integrin a5

Figure 5. LOXL2 regulated adhesion of 786-O cells to fibronectin via integrin a5/b1 expression. 786-O cells were transfected with control (CNT) siRNA or two LOXL2 siRNAs for 48 hours. Integrin b1 mRNA and protein expression levels were examined by qPCR and by Western blotting with anti-integrin b1 antibody, respectively. A, for qPCR analysis, values were normalized to GAPDH mRNA expression and are the mean SD of three independent experiments. B, for Western blot analysis, representative pictures of three independent experiments are shown. Values, normalized to b-actin protein, are mean SE of three independent experiments. C, HEK 293T cells were transfected with mock or LOXL2 expression vectors for 72 hours, and protein samples were collected and subjected to Western blot analysis with anti-LOXL2, anti-integrin a5, and anti-integrin b1 antibodies. Representative pictures of three independent experiments are shown. D, 786-O cells were transfected with control (CNT) siRNA, LOXL2 siRNAs, or integrin a5 siRNAs for 48 hours and recultured on fibronectin-coated wells for 30 minutes. After PBS washing, adherent cells were stained with calcein-AM, and the fluorescence intensity was measured. Values are mean SD of five independent experiments. , P < 0.05 compared with control siRNA.

LOXL2 is a member of the LOX family, which com- cells might happen irrespective of the lysyl oxidase domain prises the 5 members LOX and LOXL1–4. The LOX and may be a function of the SRCR domain. family contains a highly conserved lysyl oxidase domain, The upregulated integrin a5 mRNA expression in ccRCC which catalyzes the covalent cross-linking of component specimens correlated with LOLX2 mRNA expression (Fig. side chains of collagen and elastin (3). LOX knockdown in 4A). However, LOLX2 knockdown signiﬁcantly downre- Caki1 cells, however, did not signiﬁcantly affect either the gulated integrin a5 expression at the protein level but not the integrin a5 expression or migration ability, but it did mRNA level (Fig. 4B and C). TGFb signaling could explain suppress focal adhesion formation (Supplementary Fig. these contradictory data. The TGFb signaling pathway is S4). LOXL2 contains not only the lysyl oxidase domain, known to be involved in cell proliferation, differentiation, but also 4 scavenger receptor cysteine-rich (SRCR) apoptosis, and EMT in various cancer cell types (25). In domains (5), which are generally found in secreted or ccRCC cells, although the precise regulatory mechanism cell surface-bound proteins, and is involved in cell adhe- remains largely unknown, TGFb expression correlates with sion and cell signaling by mediating protein–protein pathologic stages and grades, and TGFb promotes the interactions, as reported in CD6 and ALCAM (21–23). establishment of bone metastasis in RCC (26, 27). More- Lugassy and colleagues have demonstrated that the SRCR over, the TGFb signaling pathway has been reported to domain, but not the enzyme activity via the lysyl oxidase upregulate both integrin a5 (28) and LOXL2 expression domain, is required for the LOXL2-induced inhibition of (29). Our preliminary data also showed that TGFb treat- keratinocyte differentiation (24). Therefore, LOXL2- ment increased the mRNA expression levels of both LOXL2 mediated regulation of integrin a5 expression in ccRCC and integrin a5 in 786-O cells (Supplementary Fig. S5).

1814 Mol Cancer Res; 12(12) December 2014 Molecular Cancer Research

LOXL2 Regulates Integrin a5/b1 in ccRCC Cells

siRNA CNT LOXL2 #1 siRNA CNT LOXL2 #1 Leupeptin (mmol/L) 0 50 100 200 0 50 100 200 MG132 (mmol/L) 0 20 50 100 0 20 50 100 Integrin a5 Integrin a5

Integrin b1 Integrin b1

b-Actin b-Actin

CD 2.0 siRNA CNT After siRNA LOXL2 #1 biotinylation (h) 024 1.5 LOXL2 LOXL2 siRNA CNT LOXL2 #1 siRNA CNT #1 CNT #1 1.0 After a biotinylation (h) 0 3 6 0 3 6 Integrin 5 0.5

Integrin a5 b Relative membrane Integrin 1 integrin a 5 expression (IP: streptavidin) 0.0

b (normalized to total integrin a 5) Integrin a5 -Actin Time (h) 0 24 (whole-cell lysate) IP: streptavidin 2.0 siRNA CNT b-Actin siRNA LOXL2 #1 a (whole-cell lysate) Integrin 5 1.5 Integrin b1 Fibronectin-coated dish 1.0 b-Actin 0.5

Whole cell lysate Relative membrane integrin b 1 expression 0.0 Time(normalized to total integrin b 1) (h) 0 24

Figure 6. LOXL2 suppressed integrin a5 and integrin b1 protein degradation in 786-O cells. 786-O cells were transfected with control (CNT) siRNA or LOXL2 siRNA and treated with indicated concentrations of leupeptin for 1 hour (A) or MG132 for 12 hour (B). The cell lysates were subjected to Western blot analysis with anti-integrin a5 and anti-integrin b1 antibodies. Representative pictures of three independent experiments are shown. C, 786-O cells were transfected with control (CNT) siRNA or LOXL2 siRNA and cultured on a ﬁbronectin-coated plate for 24 hours. The cell membrane proteins were biotinylated for 0, 3, and 6 hours, harvested by precipitation with streptavidin beads, and then subjected to Western blot analysis with anti-integrin a5 and anti-b-actin antibodies. Representative pictures of three independent experiments are shown. D, the cell membranes of 786-O cells transfected with control (CNT) siRNA or LOXL2 siRNA for 24 hours were biotinylated. After cell lysis, biotinylated proteins were harvested by precipitation with streptavidin beads and then subjected to Western blot analysis with anti-integrin a5, anti-integrin b1, and anti-b-actin antibodies. Values, normalized to integrin a5 and anti-integrin b1 protein levels in whole-cell lysate, are mean SE of three independent experiments.

These results suggested that LOXL2 and integrin a5 are of integrin a5 and b1 proteins, without affecting the half-life induced by TGFb at the transcriptional level, and LOXL2 is of their membrane-bound forms in 786-O cells (Fig. 6A, B, involved in integrin a5 protein stabilization in ccRCC cells. and D). The same results were also found in the fibronectin- Integrin turnover has been implicated in trafficking, binding experiment (Fig. 6C). Although the precise mechan- including internalization, recycling, and degradation. Integ- isms are largely unknown, Sayeed and colleagues showed rins internalized by endocytosis are either returned to the cell that insulin-like growth factor receptor stabilizes integrin membrane or routed for degradation via lysosomes and/or a5b1 by protecting it from proteasomal degradation in a the proteasome. Sorting nexin-17 (SNX17) controls recy- fibronectin-independent manner (31). Therefore, LOXL2 cling or degradation of integrin b1 by ubiquitination, and could intracellularly regulate integrin a5 and b1 protein SNX17 knockdown downregulates only the mature form of degradation regardless of the fibronectin-dependent degra- integrin b1 (30). Our data show that LOXL2 knockdown dation system in ccRCC. Further studies are required to downregulates only the mature form of integrin b1. There- resolve this issue. fore, LOXL2 might regulate integrin b1 protein expression Although the function of integrin a5 on cancer progres- via SNX17 in ccRCC cells. sion is controversial (32–35), integrin a5 has been shown to In response to fibronectin binding, both integrins a5 and contribute to the acquisition of resistance to apoptosis, b1 expressed on cell membrane undergo ubiquitination and angiogenesis, invasion, and metastasis, and it has thus subsequent proteasome-dependent degradation via endoso- attracted attention as a target for ccRCC therapy (36, mal sorting complex required for transport (ESCRT) 37). Interestingly, Mierke and colleagues have reported that machinery (19). Our data showed that LOXL2 knockdown in a 3D collagen invasion assay, integrin a5 accelerates induced protease- and proteasome-dependent degradation invasive activity by regulating the cell adhesiveness and

www.aacrjournals.org Mol Cancer Res; 12(12) December 2014 1815

Hase et al.

sion in ccRCC cells. The high expression of LOXL2 promotes migration and invasion via this novel mechanism, causing the progression of ccRCC.

fl G1 G2 G3 Disclosure of Potential Con icts of Interest N. Nonomura reports receiving commercial research grant from Takeda, Astellas, Normal human renal epithelial cell T1 T1 T1 T1 T1 T1 T2 T1 T1 T1 T3 and Astra Zeneca and has received speakers' bureau honoraria from Novartis. No LOXL2 potential conflicts of interest were disclosed by the other authors. Integrin α5 Integrin β1 Authors' Contributions Conception and design: H. Hase, K. Tsujikawa β-Actin Development of methodology: H. Hase, K. Tsujikawa Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Hase, W. Nakata, K. Fujita, M. Uemura, N. Nonomura, K. Tsujikawa a Figure 7. High expression of LOXL2 and integrin 5 was detected in Analysis and interpretation of data (e.g., statistical analysis, biostatistics, compu- ccRCC clinical samples. The expression levels of LOXL2, integrin a5, tational analysis): H. Hase, K. Jingushi, Y. Ueda, M. Uemura, K. Tsujikawa integrin b1, and b-actin in lysates of ccRCC tumor clinical samples were Writing, review, and/or revision of the manuscript: H. Hase, K. Jingushi, examined by Western blot analysis with their respective antibodies. W. Nakata, K. Fujita, K. Tsujikawa Eleven ccRCC samples and one normal renal epithelial cells were Administrative, technical, or material support (i.e., reporting or organizing data, examined. G1, G2, and G3 are pathologic grades. T1, T2, and T3 are the constructing databases): H. Hase, K. Kitae, H. Egawa, I. Ohshio, R. Kawakami, classifications of tumor by the size and extent. Y. Kashiwagi, Y. Tsukada, T. Kobayashi, M. Uemura, N. Nonomura, K. Tsujikawa Study supervision: H. Hase, K. Jingushi, W. Nakata, K. Fujita, K. Tsujikawa contractile forces in 786-O cells, instead of regulating the a Grant Support secretion of proteolytic enzyme (38). Moreover, integrin 5 This work was supported by a Grant-in-Aid for Scientific Research (25670025 and binds to c-MET fibronectin independently and promotes 25893113) and by Platform for Drug Discovery, Informatics, and Structural Life Science from the Ministry of Education, Culture, Sports, Science, and Technology of the invasion of ovarian cancer cells (39). Therefore, LOXL2 Japan and by Project MEET, Osaka University Graduate School of Medicine. might regulate cytoskeletal maintenance and invasion of The costs of publication of this article were defrayed in part by the payment of page ccRCC cells via integrin a5. charges. This article must therefore be hereby marked advertisement in accordance with In conclusion, our present study shows that intracellular 18 U.S.C. Section 1734 solely to indicate this fact. but not secreted LOLX2 transduces enhanced focal adhesion Received April 29, 2014; revised July 21, 2014; accepted July 21, 2014; signaling through the stabilization of integrin a5b1 expres- published OnlineFirst August 4, 2014.

References 1. Keefe SM, Nathanson KL, Rathmell WK. The molecular biology of renal 13. Anne-Marie CD, Vivian van den B, Bettine AV, Robert-Jan van S, Arjan cell carcinoma. Semin Oncol 2013;40:421–8. W G, Victor LT. Integrin expression profiling identifies integrin a5 and 2. Cowey CL, Rathmell WK Using molecular biology to develop drugs for b1 as prognostic factors in early stage non-small cell lung cancer. Mol renal cell carcinoma. Expert Opin Drug Discov 2008;3:311–27. Cancer 2010;9:152. 3. Holly E, Thomas R. The rationale for targeting the LOX family in cancer. 14. Sari T, Gergana G, Pekka L. Actin stress fibers-assembly, dynamics Nature Rev Cancer 2012;12:540–52. and biological roles. J Cell Sci 2012;125:1855–64. 4. Hector P, David O, Katalin C, Keith SK F, Sonia V, Maria Angela N , et al. 15. Calalb MB, Polte TR, Hanks SK. Tyrosine phosphorylation of focal A molecular role for lysyl oxidase-like 2 enzyme in Snail regulation and adhesion kinase at sites in the catalytic domain regulates kinase tumor progression. The EMBO Journal 2005;24:3446–58. activity: a role for Src family kinases. Mol Cell Biol 1995;15:954–63. 5. Schietke R, Warnecke C, Wacker I, Schodel€ J, Mole DR, Campean V , 16. David D. Schlaepfer, Christof RH, David JS. Signaling through focal et al. The lysyl oxidases LOX and LOXL2 are necessary and sufficient to adhesion kinase. Prog Biophys Mol Biol 1999;71:435–78. repress E-cadherin in hypoxia: insights into cellular transformation 17. Iain DC, Martin JH. Integrin Structure, Activation, and Interactions. processes mediated by HIF-1. J Biol Chem 2010;26:6658–69. Cold Spring Harb Perspect Biol 2011;10:1101. 6. Cano A, Santamaría PG, Moreno-Bueno G. LOXL2 in epithelial cell 18. Danen EH, Sonneveld P, Brakebusch C, Fassler R, Sonnenberg A. The plasticity and tumor progression. Future Oncol 2012;8:1095–108. fibronectin-binding integrins a5b1 and avb3 differentially modulate 7. Herranz N, Dave N, Millanes-Romero A, Morey L, Díaz VM, Lorenz- RhoA-GTP loading, organization of cell matrix adhesions, and fibro- Fonfría V , et al. Lysyl oxidase-like 2 deaminates lysine 4 in histone H3. nectin fibrillogenesis. J Cell Biol 2002;159:1071–86. Mol Cell 2012;46:369–76. 19. Lobert VH, Brech A, Pedersen NM, Wesche J, Oppelt A, Malerød L, 8. Jay SD, David A. Integrins in cancer: biological implications and et al. Ubiquitination of a5b1 integrin controls fibroblast migration therapeutic opportunities. Nature Rev Cancer 2010;10:9–22. through lysosomal degradation of fibronectin-integrin complexes. Dev 9. Gregory W, Maureen B, Douglas D. Integrins and cadherins join forces Cell 2010;19:148–59. to form adhesive networks. J Cell Sci 2011;124:1183–93. 20. Schlaepfer DD, Mitra SK. Multiple connections link FAK to cell motility 10. Marta C, Alan S, Margaret F, Valerie B. E-cadherin-integrin crosstalk in and invasion. Curr Opin Genet Dev 2004;14:92–101. cancer invasion and metastasis. J Cell Sci 2013;126:393–401. 21. Hohenester E, Sasaki T, Timpl R. Crystal structure of a scavenger 11. Angelika K, Vanessa B, Francois L, Christoph W, Karin S, Helena A , receptor cysteine-rich domain sheds light on an ancient superfamily. et al. Increased tumor cell dissemination and cellular senescence in the Nat Struct Biol 1999;6:228–32. absence of2288;b1-integrin function. EMBO J 2007;26:2832–42. 22. Nishioka T, Eustace A, West C. Lysyl oxidase: from basic science to 12. Donald EW, Natasza AK, Dongmei Z, John AH, Sandra B, Ulrich M. future cancer treatment. Cell Struct Funct 2012;37:75–80. Targeted disruption of b1-integrin in a transgenic mouse model of 23. Bowen MA, Bajorath J, Siadak AW, Modrell B, Malacko AR, Marquardt human breast cancer reveals an essential role in mammary tumor H. The amino-terminal immunoglobulin-like domain of activated leu- induction. Cancer Cell 2004;6:159–70. kocyte cell adhesion molecule binds specifically to the membrane-

1816 Mol Cancer Res; 12(12) December 2014 Molecular Cancer Research

LOXL2 Regulates Integrin a5/b1 in ccRCC Cells

proximal scavenger receptor cysteine-rich domain of CD6 with a 1:1 31. Sayeed A, Fedele C, Trerotola M, Ganguly KK, Languino LR. IGF-IR stoichiometry. J Biol Chem 1996;271:17390–6. promotes prostate cancer growth by stabilizing a5b1 integrin protein 24. Lugassy J, Zaffryar-Eilot S, Soueid S, Mordoviz A, Smith V, Kessler O. levels. PLoS ONE 2013;8:e76513. The enzymatic activity of lysyl oxidas-like-2 (LOXL2) is not required for 32. Roman J, Ritzenthaler JD, Roser-Page S, Sun X, Han S. a5b1-integrin LOXL2-induced inhibition of keratinocyte differentiation. J Biol Chem expression is essential for tumor progression in experimental lung 2012;287:3541–9. cancer. Am J Respir Cell Mol Biol 2010;43:684–91. 25. Talbot LJ, Bhattacharya SD, Kuo PC. Epithelial-mesenchymal transi- 33. Qian F, Zhang ZC, Wu XF, Li YP, Xu Q. Interaction between integrin a5 tion, the tumor microenvironment, and metastatic behavior of epithelial and fibronectin is required for metastasis of B16F10 melanoma cells. malignancies. Int J Biochem Mol Biol 2012;3:117–36. Biochem Biophys Res Commun 2005;333:1269–75. 26. Mitropoulos D, Kiroudi A, Christelli E, Serafetinidis E, Zervas A, Ana- 34. Kuwada SK, Kuang J, Li X. Integrin a5/b1 expression mediates HER-2 stasiou I. Expression of transforming growth factor b in renal cell down-regulation in colon cancer cells. J Biol Chem 2005;280:19027–35. carcinoma and matched non-involved renal tissue. Urol Res 2004; 35. Varner JA, Emerson DA, Juliano RL. Integrin a5b1 expression nega- 5:317–22. tively regulates cell growth: reversal by attachment to fibronectin. Mol 27. Kominsky SL, Doucet M, Brady K, Weber KL. TGF-b promotes the Biol Cell 1995;6:725–40. establishment of renal cell carcinoma bone metastasis. J Bone Miner 36. Zhang Z, Vuori K, Reed JC, Ruoslahti E. The a5/b1 integrin supports Res 2007;22:37–44. survival of cells on fibronectin and up-regulates Bcl-2 expression. Proc 28. Gailit J, Welch MP, Clark RA. TGF-b1 stimulates expression of kera- Natl Acad Sci 1995;92:6161–5. tinocyte integrins during re-epithelialization of cutaneous wounds. 37. Cowey CL, Rathmell WK. Using molecular biology to develop drugs for J Invest Dermatol 1994;103:221–7. renal cell carcinoma. Expert Opin Drug Discov 2008;3:311–27. 29. Sethi A, Mao W, Wordinger RJ, Clark AF. Transforming growth factor-b 38. Mierke CT, Frey B, Fellner M, Herrmann M, Fabry B. Integrin a5/b1 induces extracellular matrix protein cross-linking lysyl oxidase (LOX) facilitates cancer cell invasion through enhanced contractile forces. genes in human trabecular meshwork cells. Invest Ophthalmol Vis Sci J Cell Sci 2011;124:369–83. 2011;52:5240–50. 39. Mitra AK, Sawada K, Tiwari P, Mui K, Gwin K, Lengye E. Ligand- 30. Bottcher€ RT, Stremmel C, Meves A, Meyer H, Widmaier M, Tseng HY, independent activation of c-Met by fibronectin and a5b1-integrin et al. Sorting nexin 17 prevents lysosomal degradation of b1 integrins regulates ovarian cancer invasion and metastasis. Oncogene 2011;30: by binding to the b1-integrin tail. Nature Cell Biology 2012;14:584–92. 1566–76.

www.aacrjournals.org Mol Cancer Res; 12(12) December 2014 1817

LOXL2 Status Correlates with Tumor Stage and Regulates Integrin Levels to Promote Tumor Progression in ccRCC

Hiroaki Hase, Kentaro Jingushi, Yuko Ueda, et al.

Mol Cancer Res 2014;12:1807-1817. Published OnlineFirst August 4, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-14-0233

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2021/03/13/1541-7786.MCR-14-0233.DC1

Cited articles This article cites 39 articles, 14 of which you can access for free at: http://mcr.aacrjournals.org/content/12/12/1807.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/12/12/1807.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/12/12/1807. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.