altered endosome biogenesis in prostate cancer has ... · oncogenes and tumor suppressors altered...

Oncogenes and Tumor Suppressors

Altered Endosome Biogenesis in Prostate Cancer HasBiomarker Potential

Ian R.D. Johnson1, Emma J. Parkinson-Lawrence1, Tetyana Shandala1, Roberto Weigert2, Lisa M. Butler3,4,and Doug A. Brooks1

AbstractProstate cancer is the second most common form of cancer in males, affecting one in eight men by the time they

reach the age of 70 years. Current diagnostic tests for prostate cancer have significant problems with both falsenegatives and false positives, necessitating the search for new molecular markers. A recent investigation ofendosomal and lysosomal proteins revealed that the critical process of endosomal biogenesis might be altered inprostate cancer. Here, a panel of endosomal markers was evaluated in prostate cancer and nonmalignant cells and asignificant increase in gene and protein expression was found for early, but not late endosomal proteins. There wasalso a differential distribution of early endosomes, and altered endosomal traffic and signaling of the transferrinreceptors (TFRC and TFR2) in prostate cancer cells. These findings support the concept that endosome biogenesisand function are altered in prostate cancer. Microarray analysis of a clinical cohort confirmed the altered endosomalgene expression observed in cultured prostate cancer cells. Furthermore, in prostate cancer patient tissue specimens,the early endosomalmarker and adaptor protein APPL1 showed consistently altered basementmembrane histologyin the vicinity of tumors and concentrated staining within tumor masses. These novel observations on altered earlyendosome biogenesis provide a new avenue for prostate cancer biomarker investigation and suggest newmethods forthe early diagnosis and accurate prognosis of prostate cancer.

Implications: This discovery of altered endosome biogenesis in prostate cancer may lead to novel biomarkers formore precise cancer detection and patient prognosis. Mol Cancer Res; 12(12); 1851–62. �2014 AACR.

IntroductionProstate cancer is the most common form of cancer in

males from developed countries, and the incidence of thisdisease is predicted to double globally by 2030 (WorldCancer Research Fund prostate cancer statistics; http://globocan.iarc.fr). Prostate cancer affects approximately 1 in8men globally by the time they reach the age of 70 years (1).The prostate-specific antigen test is currently used forprostate cancer screening; however, this assay suffers froma high percentage of false-positive results (see for example

ref. 2), and recently, there have been recommendations toabandon this procedure, particularly in older men (3). Inaddition, the digital rectal examination, which manuallychecks the prostate for abnormalities, is limited by theinability to assess the entire gland and to some degree thesize of the tumor. Understanding the cell biology of prostatecancer is important to develop new biomarkers for the earlydiagnosis and accurate prognosis of prostate cancer.There is mounting evidence for a central role of endo-

some–lysosome compartments in cancer cell biology (seerefs. 4–6). Endosomes and lysosomes are directly involved inthe critical processes of energy metabolism (7), cell division(8) and intracellular signaling (see for example ref. 9) andwould therefore have a direct role in cancer pathogenesis.The endosome–lysosome system has a specific capacity torespond to environmental change, acting as an indicator ofcellular function and will consequently be altered in cancer(10). Moreover, the endosome–lysosome system has a crit-ical role in controlling the secretion of proteins into extra-cellular fluids (see for example ref. 11), making it an idealsystem to identify new biomarkers that are released fromcancer cells. Cumulative evidence from patient data and celllines suggested that the process of lysosomal biogenesismight be altered in prostate cancer (see for example refs. 12,13). However, we recently demonstrated that a panel oflysosomal proteins was unable to effectively discriminate

1Mechanisms in Cell Biology and Disease Research Group, School ofPharmacy and Medical Sciences, Sansom Institute for Health Research,University of South Australia, Adelaide, South Australia, Australia. 2NIDCR,NIH, Bethesda, Maryland. 3Dame Roma Mitchell Cancer Research Labo-ratories, School of Medicine, University of Adelaide, Adelaide, SouthAustralia, Australia. 4Adelaide Prostate Cancer Research Centre, Schoolof Medicine, University of Adelaide, Adelaide, South Australia, Australia.

Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).

Corresponding Author: Doug A. Brooks, Mechanisms in Cell Biology andDisease Research Group Leader, School of Pharmacy and MedicalSciences, Sansom Institute for Health Research, University of SouthAustralia, GPO Box 2471, Adelaide, SA 5001, Australia. Phone: 61-8-83021229; Fax: 61-883021087; E-mail: [email protected]

doi: 10.1158/1541-7786.MCR-14-0074

�2014 American Association for Cancer Research.

MolecularCancer

Research

www.aacrjournals.org 1851

on April 9, 2020. © 2014 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst July 30, 2014; DOI: 10.1158/1541-7786.MCR-14-0074

http://mcr.aacrjournals.org/

between a set of nonmalignant and prostate cancer cells (14).In contrast, the endosomal-related proteins cathepsin B andacid ceramidase displayed increased gene and protein expres-sion in prostate cancer cells and demonstrated some dis-criminatory capacity when compared to nonmalignant cells.Acid ceramidase was previously shown to be upregulated inprostate cancer tissues, and the overexpression of this enzymehas been implicated in advanced and chemoresistant prostatecancer (15). Importantly, we also showed that LIMP-2, acritical regulator of endosome biogenesis (16), had increasedgene and protein expression in prostate cancer cells (14),leading us to postulate that endosome biogenesis is altered inprostate cancer.Endosome biogenesis involves the synthesis and organi-

zation of structural elements of the endosome system to forman integrated set of functional organelles that eventuallyinteract with lysosomes (see for example ref. 17). There aretwo main endosomal pathways: first, from the biosyntheticcompartments (endoplasmic reticulum andGolgi apparatus)via specific vesicular traffic toward distal elements of theendosomal network, including early endosomes, late endo-somes, and multivesicular bodies; and from the cell surfacethrough early endosomes to either recycling endosomes ortoward late endosomes. In each case, the formationand movement of these dynamic vesicular compartmentsis controlled by specific targeting signals and traffickingmachinery (see for example refs. 17, 18). This vesicularmachinery can be used to define individual compartments;including, for example, the small GTPase Rab5 onearly endosomes and the small GTPase Rab7 on late endo-somes (18, 19). Here, we have investigated the gene expres-sion, amount of protein, and intracellular distribution of apanel of endosomal proteins in prostate cancer and nonma-lignant cell lines, to determine whether endosome biogenesisis altered in prostate cancer cells and to identify potential newbiomarkers.

Materials and MethodsAntibody reagentsA LIMP-2 sheep polyclonal antibody was generated

against the peptide sequence CKKLDDFVETG-DIRTMVFP (Mimotopes Pty. Ltd.). Rabbit polyclonalantibodies (Abcam PLC) were against Appl1 (0.4 mg/mL),Appl2 (0.4 mg/mL), Rab4 (1 mg/mL), TGN46 (10 mg/mL),TfR1 (1 mg/mL), and TfR2 (1 mg/mL). Akt (1:1,000) andphospho-Akt (Thr308; 1:1,000) from Cell Signaling Tech-nology Inc., and horseradish peroxidase (HRP)–conjugatedanti-GAPDH (1:20,000; Sigma-Aldrich Pty. Ltd.). Goatpolyclonal antibodies (Santa Cruz Biotechnology) wereagainst Rab5 (1 mg/mL), Rab7 (1 mg/mL), and EEA1(1 mg/mL). A LAMP-1 (1 mg/mL) mouse monoclonal BB6was provided by Prof. Sven Carlsson (Umea University,Umea, Sweden). HRP-conjugated secondary antibodies forWestern blot analysis included anti-goat/sheep (1:2,000;Merck Millipore Pty. Ltd.), anti-rabbit (1:2,000), and anti-mouse (1:2,000; Sigma-Aldrich). The secondary and otherantibody-conjugated fluorophores that were used included

Alexa Fluor 488 (1:250), Alexa Fluor 633 (1:250), Trans-ferrin-633 (1:1,000), Phalloidin-488 (1:100), and Lyso-Tracker (5 mmol/L); all from Life Technologies Pty. Ltd.

Cell lines and culture conditionsThe nonmalignant cell lines PNT1a and PNT2 and

prostate cancer cell lines 22RV1 and LNCaP (clone FCG)were obtained from the European Collection of Cell Cul-tures via CellBank Australia (Children's Medical ResearchInstitute, Westmead, NSW, Australia). These cell lines wereabsent from the list of cross-contaminated or misidentifiedcell lines, version 6.8 (March 9, 2012; ref. 20).Cell lines were cultured in 75-cm2 tissue culture flasks and

maintained in RPMI-1640 culture medium (Gibco, LifeTechnologies), supplemented with 10% fetal calf serum (InVitro Technologies Pty. Ltd.) and 2 mmol/L L-glutamine(Sigma-Aldrich Pty. Ltd.). Cells were incubated at 37�Cwith 5% CO2 in a Sanyo MCO-17AI humidified incubator(Sanyo Electric Biomedical Co., Ltd.). Cells were cultured toapproximately 90% confluence before passage, by washingwith sterile PBS (Sigma-Aldrich), trypsin treatment (Tryp-sin–EDTA solution containing 0.12% trypsin, 0.02%EDTA; SAFC; Sigma-Aldrich) to dissociate the cells fromthe culture surface and then resuspension in supplementedculture medium.

Preparation of cell extracts and conditioned culturemedia for protein detectionThe culture medium was aspirated from cultures at 80%

to 90% confluence, the cells washed once with PBS, andthen incubated with 800 mL of a 20 mmol/L Tris (pH 7.0)containing 500 mmol/L sodium chloride and 2% (w/v)SDS. Cells were harvested and an extract prepared byheating at 95�C and sonication for 1 minute. The lysatewas then passaged six times through a 25-guage needle.Total protein in the cell extracts was quantified using abicinchoninic acid assay according to the manufacturer'sinstructions (Micro BCA Kit; Pierce). Samples werequantified using a Wallac Victor optical plate-reader andWorkout software v2.0 (PerkinElmer Pty. Ltd.), using afive-point parameter standard curve. Cell extracts werestored at �20�C.Protein was recovered from conditioned culture media,

collected at the time of cell harvesting, using trichloroaceticacid precipitation. Briefly, cell debris was removed from theculture media by centrifugation (1,000� g for 10 minutes),sodium deoxycholate (Sigma-Aldrich) added to a final con-centration of 0.02% (v/v) and incubated on ice for 30minutes. Trichloroacetic acid (Sigma-Aldrich) was thenadded to a final concentration of 15% (v/v) and incubatedon ice for 2 hours. Protein was collected by centrifugation at4�C (5,500� g for 30 minutes), washed twice with ice-coldacetone and resuspended in SDS-sample buffer/PBS solu-tion, and stored at �20�C.

Gene expressionRelative amounts of mRNA from nonmalignant and

prostate cancer cell lines were defined by quantitative PCR

Johnson et al.

Mol Cancer Res; 12(12) December 2014 Molecular Cancer Research1852




(qPCR). Briefly, cells were lysed with TRI reagent (AppliedBiosystems, Life Technologies) and RNA extraction per-formed using RNeasy (Qiagen Pty. Ltd.) according to themanufacturer's instructions. Two micrograms of total RNAwas reverse-transcribed using the High Capacity RNA-to-cDNAKit (Life Technologies) following the manufacturer'sinstructions. qPCR was performed with 2 mL of a 1:25dilution of cDNA in 10 mL of reaction mixture; containing5 mL Power SYBR Green PCR Master Mix (Life Technol-ogies) and 0.5 mL of both 10 nmol/L forward and reverseprimer. qPCR was performed using a 7500 Fast Real-TimePCR System (Life Technologies). Each target was assessed intriplicate on a single plate and quantified relative toGAPDHendogenous control for each plate, with triplicate biologicreplicates run independently. Oligonucleotides (Gene-Works Pty. Ltd.) were as follows:GAPDH TGCACCACCAACTGCTTAGC (Fwd) GG-

CATGGACTGTGGTCATGAG (Rev; ref. 21);LAMP-1 ACGTTACAGCGTCCAGCTCAT (Fwd) T-

CTTTGGAGCTCGCATTGG (Rev; ref. 21);LIMP2 AAAGCAGCCAAGAGGTTCC (Fwd) GTCT-

CCCGTTTCAACAAAGTC (Rev);APPL1 ACTTGGGTACATGCAAGCTCA (Fwd) TC-

CCTGCGAACATTCTGAACG (Rev);APPL2 AGC TGATCGCGCCTGGAACG (Fwd) GG-

GTTGGTACGCCTGCTCCCT (Rev);EEA1 CCCAACTTGCTACTGAAATTGC (Fwd) TG-

TCAGACGTGTCACTTTTTGT (Rev);RAB5 AGACCCAACGGGCCAAATAC (Fwd) GCC-

CCAATGGTACTCTCTTGAA (Rev);RAB4 GGGGCTCTCCTCGTCTATGAT (Fwd) AG-

CGCATTGTAGGTTTCTCGG (Rev);RAB7 GTGTTGCTGAAGGTTATCATCCT (Fwd)

GCTCCTATTGTGGCTTTGTACTG (Rev).

Western blottingTen micrograms of total protein for whole-cell lysates or

the secreted protein from approximately 3 � 106 cells washeat-denatured (5 minutes at 100�C in NuPAGE LDSSample Buffer and reducing agent), then electrophoresedat 120 V for 1.5 hours using pre-cast gels in an XCellSureLock Mini-Cell system (Life Technologies). Theprotein was then transferred to polyvinylidene difluoridemembranes (Polyscreen; PerkinElmer). The transfermembranes were blocked for 1 hour at room temperatureusing a 5% (w/v) skim milk solution in 0.1% (v/v) TBS-Tween (blocking solution) and incubated with primaryantibody overnight at 4�C. The membranes were washedin 0.1% (v/v) TBS-Tween and then incubated with theappropriate HRP-conjugated secondary antibody dilutedat 1:2,000 in blocking solution. The membranes weredeveloped using a Novex ECL chemiluminescent sub-strate reagent kit (Life Technologies), and proteins werevisualized using an ImageQuant LAS 4000 imager, soft-ware version 1.2.0.101 (GE Healthcare Pty. Ltd.). Trip-licate samples were analyzed and images quantified relativeto a reference GAPDH loading control using Alpha-ViewSA software v3.0 (ProteinSimple Pty. Ltd.).

ImmunofluorescenceCells were cultured on 22-mm glass coverslips, fixed with

4% (v/v) formaldehyde in PBS for 20 minutes at roomtemperature, and then permeabilized with 0.1% Triton-X(v/v) in PBS for 10 minutes. Nonspecific antibody reactivitywas blocked by incubation with 5% (w/v) bovine serumalbumin (BSA) in PBS for 2 hours at room temperature.Cells were incubated with primary antibody in 5%BSA for 2hours at room temperature, followed by fluorophore-con-jugated secondary antibody in the dark for 1 hour at roomtemperature. Unbound antibody was removed by three PBSwashes and coverslips mountedwith ProLongGold AntifadeReagent containing DAPI nuclear stain (Life Technologies).Confocal microscopy was performed using a Zeiss LSM 710META NLO laser scanning microscope and associated CarlZeiss Zen 2009 software. Laser lines of 370, 488, 543, and633 nm were used for DAPI, Alexa Fluor 488, Cy3, andAlexa Fluor 633 fluorescence, respectively. Images wereexported as grayscale 16-bit TIFF files and processed usingAdobe Photoshop CS5 (Adobe Systems Inc.).

ImmunohistochemistryMatched human nonmalignant and tumor prostate tissue

sections (3 mm) were mounted on Superfrost Ultra Plusslides (Menzel-Gl€aser GmbH) and heated overnight at50�C. Sections were then dewaxed in xylene, rehydratedin ethanol, and incubated in 0.3% H2O2 in PBS for 15minutes at room temperature. Heat–Induced EpitopeRetrieval was carried out using 10 mmol/L citrate buffer(pH 6.5) in a Decloaking Chamber (Biocare Medical LLC)for 5 minutes at 125�C. Slides were blocked first using anInvitrogen Avidin/Biotin Kit (as per the manufacturer'sinstructions) and then in 5% blocking serum (Sigma-Aldrich) for 30 minutes at room temperature in a humidchamber. Sections were then incubated with primary anti-body overnight at 4�C in a humid chamber, followed byincubation with the appropriate biotinylated secondaryantibody (1:400; Dako Australia Pty. Ltd.) for 1 hour atroom temperature in a humid chamber, then streptavidin–HRP (1:500; Dako Australia Pty. Ltd.) for 1 hour at roomtemperature in a humid chamber and finally with DAB/H2O2. The tissue sections were then counterstained withLillie–Mayer hematoxylin, rinsed in water, rehydrated, andmounted on slides with DPX mounting media (MerckMillipore Pty. Ltd.). Images were obtained by scanningslides using a NanoZoomer (Hamamatsu Photonics K.K.).

Data analysisQuantities of target-gene and endogenous-control

(GAPDH) were calculated from a standard curve from serialdilutions of control material (LNCaP cDNA). Kruskal–Wallis nonparametric analysis of variance statistical analyseswere performed using Stata/SE v11.2 (StataCorp LP) todetermine the significance between nonmalignant control(PNT1a and PNT2) and prostate cancer (22RV1 andLNCaP) cell lines (95% confidence limit; P � 0.05) forintracellular or secreted protein amount, and gene expres-sion. The Taylor microarray cohort (GSE21034) consisted

Altered Endosome Biogenesis in Prostate Cancer

www.aacrjournals.org Mol Cancer Res; 12(12) December 2014 1853




of patients treated by radical prostatectomy at the MemorialSloan-Kettering Cancer Center (MSKCC; New York, NY;ref. 22), profiling 150 prostate cancer and 29 nonmalignanttissue samples that was performed using Affymetrix HumanExon 1.0 ST arrays. Statistical analysis of microarray geneexpression was performed using a two-tailed unpaired t testwith Welch correction using GraphPad Prism 5.03 (Graph-Pad Software Inc.).

ResultsIncreased endosome-related gene and protein expressionin prostate cancer cellsThe expression of endosome- and lysosome-related genes

was quantified by qPCR in control and prostate cancer cellsand normalized to the expression of GAPDH mRNA. Theamounts of LIMP2, APPL1, APPL2, RAB5A, EEA1, andRAB4 mRNA were significantly increased in prostate cancerwhen compared with nonmalignant control cell lines (P �0.05; Fig. 1). In each case, there was an approximately 2- to 3-fold increase in mRNA expression. There was no significantdifference in the amount of either RAB7 or LAMP-1mRNAdetected in prostate cancer cells compared with nonmalignantcontrols. Western blot analysis (Fig. 2A) demonstrated sig-nificant increases in the amount of LIMP-2, Appl1, Appl2,EEA1, and Rab4 protein in extracts from prostate cancer cellswhen compared with nonmalignant control cells (P �0.05; Fig. 2B). Moreover, for both LIMP-2 and Rab4, theincrease was approximately 2- to 4-fold for prostate cancerwhen comparedwith nonmalignant cells (Fig. 2B). There wasno significant difference in the amount of Rab5, Rab7, andLAMP-1 protein detected in nonmalignant compared withprostate cancer cells (Fig. 2B).

Altered distribution of endosomes and lysosomes inprostate cancer cellsRepresentative confocal images for the distribution of

endosomes and lysosome proteins (Fig. 3) show evidenceof increased staining and altered distribution in prostatecancer compared with the nonmalignant controls. In non-malignant control cells, LIMP-2 was concentrated in theperinuclear region, with some punctuate vesicular staining inthe remainder of the cytoplasm. In contrast, prostate cancercells displayed relatively smaller LIMP-2 compartments,which had an even distribution throughout the cytoplasm.Appl1-positive endosomes were detected throughout the cellcytoplasm of nonmalignant control cells, whereas in prostatecancer cells, these compartments were more concentrated atthe cell periphery, particularly near the plasma membrane incellular extensions/pseudopodia. In nonmalignant controlcells, both Rab5 and its effector EEA1 were concentrated inthe perinuclear region, whereas in prostate cancer cells, theseendosomal compartments were found throughout the cyto-plasm, with some compartments located toward the cellperiphery in cellular extensions. Rab7-positive endosomeswere located mainly in the perinuclear region of bothnonmalignant control and prostate cancer cells. In nonma-lignant control cells, LAMP-1 compartments were concen-

trated in the perinuclear region, whereas in prostate cancercells the LAMP-1 compartments were distributed away fromthe perinuclear region and concentrated in cellular exten-sions. Consistent with the LAMP-1 staining, LysoTrackerpositive acidic compartments were concentrated mainly inthe perinuclear region of nonmalignant control cells, where-as in prostate cancer cells, these compartments were detectedin both the perinuclear region and in cytoplasmic extensions(Fig. 3).

Altered distribution of endocytosed transferrin inprostate cancer cellsPrevious studies have reported increased uptake of trans-

ferrin in prostate cancer cells, prompting the investigation ofreceptor expression and transferrin endocytosis in relation tothe observed increase in endosome protein expression and

Figure 1. Quantification of endosomal and lysosomal gene expression incontrol and prostate cancer cell lines. Levels of mRNA transcripts innonmalignant control cell lines (white bars) and prostate cancer cell lines(black bars) were evaluated by qPCR in triplicate experiments. Data wereexpressed relative to GAPDH endogenous control and analyzed by theKruskal–Wallis rank-sum method. Statistical significance (P � 0.05) isrepresented by an asterisk.

Johnson et al.





altered endosome distribution. In nonmalignant controlcells, endocytosed transferrin was observed in punctateintracellular structures after 5minutes and in the perinuclearregion at 15 and 30 minutes (Fig. 4). The prostate cancercells endocytosed relatively more transferrin than the non-

malignant control cell lines within the first 5 minutesof incubation and at the 30-minute incubation point. Innonmalignant cells at 30 (Fig. 4) and 20 minutes (Fig. 5),this transferrin was tightly concentrated in close proximity tothe nucleus. After 15minutes of incubation, the internalized

Figure 2. Detection andquantification of intracellularlysosomal proteins innonmalignant control and prostatecancer cell lines. A, representativeimages from Western blot analysisof 10 mg whole-cell lysate fromnonmalignant control cell linesPNT1a and PNT2, and cancer celllines 22RV1 and LNCaP, examinedin triplicate. B, protein amount wasquantified by densitometry relativeto a GAPDH endogenous control.Datawereanalyzedby theKruskal–Wallis rank-sum method withstatistical significance representedby an asterisk. (�, P � 0.05;��, P � 0.01).






transferrin was not as concentrated in the perinuclear regionof prostate cancer cells, with more transferrin-labeled com-partments in the cell periphery and distributed throughoutthe cytoplasm when compared with the nonmalignant cells(Fig. 4). There was also a marked reduction in actin stainingfor the prostate cancer compared with the nonmalignantcontrol cell lines (Fig. 4). In the nonmalignant control cells,transferrin was clustered mainly in LIMP-2- and Rab7-positive endosomes localized in the perinuclear region (Fig.5). Although the prostate cancer cells had someLIMP-2- andtransferring-positive staining in the perinuclear region andsome colocalization of transferrin with the Golgi markerTGN46 (yellow colocalization), the majority of transferrinwas localized in different endosomal compartments (i.e.,Appl1, Rab5, and EEA1) distributed throughout the cyto-plasm and in cellular extensions (Fig. 5). The Rab4 recyclingendosomes and LAMP-1–positive lysosomes had similarpatterns of transferrin staining for the prostate cancer andnonmalignant control cell lines (Fig. 5). Further analysis ofthe transferrin receptors revealed variable gene and proteinexpression for TfR1 (TFRC) and TfR2 (TFR2; Fig. 6A andB). There was a significant increase inTFRC gene expression(P � 0.05) in prostate cancer cells when compared withnonmalignant controls (Fig. 6A), but only a qualitativeincrease in TfR1 protein in the prostate cancer cell line22RV1 and not for LNCaP (Fig. 6B). Although there wassignificantly more TfR2 protein detected in prostate cancercells when compared with the nonmalignant cells (P �0.05), there was only an increase in TFR2 gene expression

in theLNCaP cancer cell line (Fig. 6A).Colocalization ofTfR1and transferrin was observed in all cell lines, and was in aperinuclear location in nonmalignant cell lines PNT1a andPNT2 compared with a broader cytoplasmic distribution inthe cancer cell lines 22RV1 and LNCaP. Conversely, thereappeared to be no colocalization of transferrin with TfR2 innonmalignant cells and limited colocalization of transferrinwithTfR2 compartments in the prostate cancer cells (Fig. 6C).

Altered Akt signaling in prostate cancer cellsThe total amount of Akt protein detected in nonmalignant

control cells was similar to that detected in prostate cancercells (Fig. 6D andE). There were, however, differences in theamount of phosphorylated Akt in the prostate cancer lines,with 22RV1 showing a marked reduction in the amount ofphosphorylated Akt, whereas LNCaP had an increasedamount of phosphorylated Akt (Fig. 6D), a phenomenonpreviously observed by Shukla and colleagues (23) andrelated tomutations of PTEN in LNCaP.More importantly,following the addition of transferrin, there was a significantincrease in the amount of phosphorylated Akt in the non-malignant control cell lines, but no change in the amount ofphosphorylated Akt in either of the cancer cells (Fig. 6E).

LAMP-1and APPL1 mRNA expression in a prostatecancer microarray cohort and distribution of LAMP-1and Appl1 in prostate tissueTo support the hypothesis of altered endosome biogen-

esis in prostate cancer, the percentage change of mRNA

Figure 3. Confocal micrographs of endosomal markers in prostate cancer cell lines compared with nonmalignant control cell lines. Fixed cells were probed forendosome markers (green) and counterstained with DAPI nuclear stain (blue) and visualized by laser-scanning confocal microscopy. Cell outlines werevisualized by transmitted light illumination and cell periphery depicted by white dotted line. Antibody labeling was performed in triplicate experiments and aminimum of 5 cells were visualized in each cell line for each assay. Similar fluorescence staining was observed in all cells for each cell line/antibody label. Anadditional confocal micrograph from an independent labeling assay is presented in Supplementary Fig. S1.

Johnson et al.





expression for LAMP-1 and APPL1 was analyzed from theTaylor microarray cohort (Fig. 7A). LAMP-1 gene expres-sion was significantly decreased (P � 0.01) in prostatecancer tissue compared with nonmalignant prostate tissue.APPL1 gene expression was significantly increased (P �0.05) in prostate cancer tissue compared with nonmalig-nant tissue. Immunohistochemistry was used to investi-gate the distribution of LAMP-1 and Appl1 in prostatecancer patient tissue samples (Fig. 7B). The lysosomalmarker LAMP-1 showed tumor-specific staining in somepatient samples, but consistent with previous studies,there were variable results with some patient sampleshaving little or no LAMP-1 staining (data not shown forthe latter). In nonmalignant tissue, Appl1 clearly delin-eated basement membranes, whereas in the malignanttissue there was no evidence of basement membranestaining (Fig. 7B). In addition, Appl1 specifically delin-eated the cancer margins and showed increased stainingwithin the tumor mass (Fig. 7B).

DiscussionProstate cancer is one of the most frequently diagnosed

cancers in men and a leading cause of cancer-related deathsworldwide, particularly, in the United States and Australa-sian populations (24, 25). The prostate-specific antigen isstill commonly used to detect prostate cancer, but hassignificant problems in terms of miss-diagnosis and prog-nostic prediction (see for example ref. 26). Some promisingadjunct tests have recently been developed, including pros-tate cancer antigen 3 (PCA3; ref. 27), the analysis ofcholesterol sulfate (28), and a novel sequence of the geneprotein kinase C-zeta (PRKCZ), which is translated to theprotein PRKC-z-PrC (29). However, these biomarkers donot provide early and accurate detection of prostate cancer,which is needed to enable the selection of the most appro-priate therapeutic intervention and to avoid potential over-treatment (2). On the basis of our observations of alteredLIMP-2 expression (14), we investigated altered endosomalbiogenesis in prostate cancer to help provide more sensitiveand specific markers for early detection and diseaseprediction.There have been extensive protein and proteomic stud-

ies undertaken to identify potential new prostate cancerbiomarkers (see for example ref. 26); however, the idealmarker with appropriate sensitivity and specificity is yet tobe established. Interestingly, many of the early biomarkersinvestigated, and some of the recent proteins identified inproteomic studies, are either lysosomal hydrolases (e.g.,lysosomal cathepsins, acid ceramidase, and acid phospha-tase), lysosomal membrane proteins (e.g., LAMP-1-3 alsocalled CD107a, b, and CD63) or proteins that are deliv-ered from the cell surface into the endosome–lysosomesystem (e.g., sialomucin/CD164, CD1, CD47, andCD75). Additional evidence supports the concept thatendosomal–lysosomal biogenesis is altered in prostatecancer, including the altered distribution of lysosomesthat has been reported in prostate cancer cells (30).Despite these indications on lysosomal biogenesis, a setof optimal prostate cancer biomarkers have yet to bedefined. In a recent study of endosome and lysosomemarkers in prostate cancer cell lines, we also found thatlysosomal markers were unable to discriminate prostatecancer from nonmalignant cell lines, but there was evi-dence suggesting that endosome biogenesis may be alteredin prostate cancer cells (14).Here, we observed altered distribution of specific endo-

some subsets and lysosomes into the cellular periphery ofprostate cancer cells, which could have important implica-tions for cancer cell biomarker release and intracellularsignaling. Acidic extracellular pHhas been shown to enhanceorganelle redistribution, stimulating the traffic of endo-some–lysosome–related organelles to the periphery of cancercells (10, 31). This altered endosome–lysosome traffic hasbeen linked with the release of cathepsin B and tumorinvasiveness (32), presumably due to the hydrolysis ofextracellular matrix after the exocytosis of this enzyme(33). However, cathepsin B has been reported to be more

Figure 4. Time-course of transferrin uptake in prostate cell lines.Representative confocal micrographs from triplicate stainingexperiments showing increased uptake and altered distribution oftransferrin in prostate cancer cell lines compared with nonmalignantcontrol cell lines. Cell cultures were incubatedwith transferrin Alexa Fluor633 conjugate (red) for a period of 5, 15, and 30 minutes before cellfixation and F-actin labeled with phalloidin Alexa Fluor 488 (green).Arrows in the control cells depict transferrin that was concentrated inclose proximity to the nucleus at 30 minutes. Additional confocalmicrographs of transferrin uptake are presented in SupplementaryFig. S2.






enriched in endosomes (34) rather than lysosomes, whereasthe reverse is true for another proposed prostate cancerbiomarker cathepsin D (35). The movement of lysosom-al-related vesicles to the periphery of prostate cancer cells hasbeen shown to be dependent on GTPases (e.g., RhoA),microtubules, the molecular motor protein KIF5b, and toinvolve PI3K, Akt/Erk1/2 phosphorylation, and MAPKsignaling (32). Moreover, a component of the MAPKsignaling pathway, the endosome-localized MAPK/Erkkinase (MEK1) p14–MP1 scaffolding complex, has beenshown to specifically interact with and regulate the distri-

bution of endosomes via ERK signaling (36). Increases inNaþ/Hþ exchange activity (acidification), RhoA GTPaseactivity, and PI3K activation have been shown to result inexocytosis from prostate cancer cells (31). The increasedendosomal-associated gene and protein expression observedhere, together with the previously observed cathepsin Brelease, suggested that endosome-related proteins may pro-vide an important new focus for prostate cancer diseasebiomarker studies.Increased expression of the endosomal protein LIMP-2

has been shown in oral squamous cell carcinoma and was

Figure 5. Transferrin and endosome/lysosomemarker cofluorescence. Confocal micrographs and enlargements showing transferrin (red; endocytosed for 20minutes) and endosome/lysosome marker (green) in nonmalignant control cell lines PNT1a and PNT2, and prostate cancer cell lines 22RV1 and LNCaP.Colocalization of markers is depicted by yellow fluorescence.

Johnson et al.





associated with tumor metastasis (37). LIMP-2 has beenreported to have a role in endosome biogenesis and its over-expression evoked the enlargement of both early and lateendosomes (16). We observed increased gene and proteinexpression of the endosomal protein LIMP-2 in prostatecancer cell lines (14), prompting us to investigate otherendosomal proteins in prostate cancer cells. The early endo-some-associated proteins Appl1, Appl2, EEA1, and recyclingendosome protein Rab4 were significantly upregulated (geneand protein) in prostate cancer cells, supporting the hypoth-esis of altered endosome biogenesis in prostate cancer. APPL1expression was significantly increased in the Taylor prostatecancer tissue microarray, supporting the expression profiles

observed in cell lines. Furthermore, the each EEA1 and Appl1endosome subpopulations displayed altered intracellular dis-tribution consistent with altered endosome traffic and poten-tially function. Interestingly, while Rab7 expression wasunaltered, Rab7-positive compartments displayed differentialdistribution in prostate cancer compared with nonmalignantcells. Changes in subcellular localization may affect signalingin a similar manner to that which transpires through down-regulated gene/protein expression that affects prostate cancerprogression through enhanced signaling (38). Thus, theanalysis of compartment distribution may distinguish cancercell phenotypes independently of altered gene and proteinexpression.

Figure 6. Analysis of transferrinreceptor expression andlocalization with transferrin.A, quantification of transferrinreceptor 1 (TFRC) and transferrinreceptor 2 (TFR2) gene expressionin nonmalignant and prostatecancer cell lines. B, Western blotanalysis and protein quantificationof transferrin receptor 1 (TfR1) andtransferrin receptor 2 (TfR2).Quantification of gene and proteinexpression was relative to GAPDHgene and protein, respectively.C, confocal micrographs andenlargements showing transferrin(red) and transferrin receptor(green) in nonmalignant cell linesPNT1a and PNT2, and prostatecancer cell lines 22RV1 andLNCaP. Colocalization oftransferrin receptor and transferrinis represented by yellowfluorescence. D, Western blotanalysis and quantification (E) ofAKT phosphorylation relative tototal AKT, prior and subsequent totreatment of nonmalignant (PNT1aand PNT2) and prostate cancercells (22RV1 and LNCaP) withtransferrin for 20 minutes.�, P � 0.05.






The significant changes that we observed in endosome-associated gene and protein expression, together with thealtered distribution of endosome populations prompted usto investigate transferrin receptor expression together withtransferrin endocytosis, sorting, and Akt signaling as mea-sures of endosome function. Significant increases in theamount of transferrin receptor have previously been reportedin prostate cancer cells (39), and this has been linked to c-Myc activation, which alters proliferation and tumorigenesis(40). Akt signaling is also essential for regulating cell growthand survival; and this controls the cell surface expression oftransferrin and growth factor receptors (41). The transferrin

receptor has previously been observed to colocalize withRab5 and themotor proteinmyosin VI; the latter of which isinvolved in retrograde transport to the plasma membrane(42). This was consistent with our observations of endosomepopulations costaining with labeled transferrin in the cellularperiphery of prostate cancer cells. There also appeared to be aderegulation of Akt signaling in prostate cancer cells, withcontrol cells being responsive to transferrin endocytosis, butprostate cancer cells being unresponsive, despite havingvariable high or low amounts of Phospho-Akt/Akt. Thisaltered signaling may be related to the intracellular locationof the transferrin receptor that can be disturbed throughchanges in localization or depletion of PtdIns3P (43) oraffected through variable internalization resulting fromaltered Appl1 or Rab5 expression (as is the case withepidermal growth factor receptor; ref. (44), affecting recep-tor trafficking and signal modulation.Appl1 has been shown to be directly involved in insulin

signaling and the translocation of the glucose transporterGLUT-4, which is mediated by direct binding of Appl1 toPI3K and Akt (45), inducing endosome relocalization. Inprostate cancer cells, Appl1-potentiated Akt activity hasalso been shown to suppress androgen receptor transacti-vation (46). The increased gene and protein expression ofAppl1 that we observed in prostate cancer cells might beexpected to cause increased glucose uptake, due to itseffect on GLUT-4 and this could have implications forenergy metabolism in these cancer cells. Indeed, Appl1also regulates other aspects of both lipid and glucosemetabolism, activating AMP-activated kinase, p38MAPK, and PPARa (see for example ref. 45). Appl2 hasbeen shown to function as a negative regulator of adipo-nectin signaling, by competitive binding with Appl1 forinteraction with the adiponectin receptor, again regulatingenergy metabolism. The increased expression of bothAppl1 and Appl2 could therefore impact heavily onprostate cancer cell metabolism with direct significancefor increased energy utilization and prostate cancer cellsurvival. The altered Appl1 expression and effect on Aktsignaling in prostate cancer cells would be expected to alsohave significant consequence for other aspects of prostatecancer biology, due to the importance of the Appl1/PI3K/Akt signaling pathway in leading cell adhesion and cellmigration (47). Notably, Appl1 also acts as a mediator ofother signaling pathways, by interaction with the cytosolicface of integral or membrane-associated proteins either atthe cell surface or in the endosome pathway; where it isdirectly involved in endosome traffic.Rab GTPases are integrally involved in the control of

endosome traffic, cycling between the cytoplasmic GDP-bound state and the active membrane–associated GTP-bound state. Rab5 and Rab7 respectively define early- andlate-endosome compartments and during endosome matu-ration Rab5 recruits the HOPS complex as a mechanism toactivate and be replaced by Rab7. Although mVps39 isknown to be a guanine nucleotide exchange factor (GEF)that promotes the GTP-bound state on endosomal Rabs,TBC-2/TBC1D2 is a Rab GTPase-activating protein (GAP)

Figure 7. Analysis of gene expression and protein distribution inprostate cancer compared with nonmalignant tissue. A, box-and-whisker graphs showing percentage change of LAMP1 and APPL1mRNA expression in normal (n ¼ 29) and prostate cancer tissue (n ¼150) from metanalysis of the cohort by Taylor and colleagues (22).Box-and-whisker graphs were plotted with Tukey outliers (blackpoints). Statistical significance is represented by an asterisk (�, P �0.05; ��, P � 0.01). B, LAMP-1 (a and b) and Appl1 (d and e) expressionin matched human normal (a and d) and malignant (b; Gleason grade3þ3, e; Gleason grade 3þ4) prostate tissue. Both normal andmalignant mixed-tissue is stained for LAMP-1 (c; Gleason grade 3þ3)and Appl1 (f; Gleason grade 3þ4). The arrows in d and f show Appl1staining the basement membrane in nonmalignant prostate tissue.Scale bar, 100 mm in a, b, and d and 200 mm in c, e, and f.

Johnson et al.





that promotes the GDP-bound state; and in combination isused to regulate the membrane localization of Rab proteins.TBC-2/TBC1D2 is therefore thought to act as a regulator ofendosome to lysosome traffic and is required to maintain thecorrect size and distribution of endosomes (48). The altereddistribution of endosome populations that we observed inprostate cancer cells suggests that TBC-2/TBC1D2 (GAP)and or mVps39 (GEF) might be functionally impaired;particularly, as the early endosomes were routed mainlytoward the cell periphery, whereas late endosomes remainedin a perinuclear location. Interestingly,microarray analysis hasdetected increased expression of TBC-2/TBC1D2 andreduced Vps39 mRNA in relation to altered endosomal–lysosomal traffic (49). This altered GEF and GAP functionhas been shown to be critical for endosomal traffic of integrinsand there have been direct links established between alteredRab GTPase activity and cancer progression (50).The expression of endosome markers has not previously

been investigated thoroughly in prostate cancer, althoughsome lysosomal-related cell surface CD (cell differentiation)markers and LAMP-1/LAMP-2 have been used in tissuebiopsy analysis. TheGleason grading system is used to definehistologic differentiation in conjunction with marker anal-ysis to predict the course of disease in patients with prostatecancer. The lysosomal membrane proteins LAMP-1-3 andCD markers CD164, CD1, CD47, and CD75 are oftenevident in primary and metastatic cancer biopsies (51), buttheir consistency and predictive capacity for disease progres-sion is limited. We observed increased amounts of Appl1protein in malignant tissue from biopsies of patients withprostate cancer, confirming the increased gene and proteinexpression of Appl1 in prostate cancer cell lines. Appl1appeared more concentrated in the basement membranesin nonmalignant tissue, whereas in the malignant tissue,there was no basement membrane staining, indicating diag-nostic/prognostic potential for this biomarker. Furtherimmunohistochemical and patient tissue analysis of Appl1and other endosomal proteins is required to establish thevalidity and predictive value of these proteins as prognosticbiomarkers in prostate cancer.In summary, we have demonstrated increased expression

of early endosome markers and altered localization of endo-some and lysosome compartments in prostate cancer cells,which was associated with altered endocytosis and recyclingof the transferrin receptor and aberrant Akt signaling. Thealterations to the endocytic machinery that we have observed

here, may increase the amount of endocytosis in prostatecancer cells, which could increase nutrient uptake/availabil-ity, provide additional membrane for cell division (9), andalter intracellular signaling (10); which are all hallmarks ofcancer cell biology. There appeared to be a specific discon-nect between the cellular location of early endosomes (andlysosomes) in the cell periphery and late endosomes in theperinuclear region, which could affect degradative andsignaling processes in prostate cancer cells. We concludedthat endosome biogenesis and function is altered in prostatecancer cells, opening a potentially new avenue to investigatebiomarkers that aid in the diagnosis and prognosis of prostatecancer. Endosomes are directly involved in the processes ofcellular secretion and exosome release, making these newlyidentified endosomal proteins potentially available for detec-tion in patient samples, such as blood and urine.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: I.R.D. Johnson, E.J. Parkinson-Lawrence, T. Shandala,R. Weigert, L.M. Butler, D.A. BrooksDevelopment of methodology: I.R.D. Johnson, E.J. Parkinson-Lawrence,T. ShandalaAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): I.R.D. Johnson, E.J. Parkinson-LawrenceAnalysis and interpretation of data (e.g., statistical analysis, biostatistics, compu-tational analysis): I.R.D. Johnson, E.J. Parkinson-Lawrence, T. Shandala,L.M. Butler, D.A. BrooksWriting, review, and/or revision of themanuscript: I.R.D. Johnson, E.J. Parkinson-Lawrence, L.M. Butler, D.A. BrooksAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): E.J. Parkinson-Lawrence, D.A. BrooksStudy supervision: E.J. Parkinson-Lawrence, D.A. Brooks

AcknowledgmentsThe authors thank Dr. Shalini Jindal and Ms. Marie Pickering (Dame Roma

Mitchell Cancer Research Laboratories, University of Adelaide, South Australia) forexpert assistance with the immunohistochemistry and pathology of human prostatetissue samples.

Grant SupportThis project was funded by a University of South Australia Presidents Scholarship

and a University of South Australia Postgraduate Award, together with additionalsupport from University of South Australia Research SA Seeding Funds.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be herebymarked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

Received February 10, 2014; revised July 9, 2014; accepted July 10, 2014;published OnlineFirst July 30, 2014.

References1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer

J Clin 2013;63:11–30.2. Schroder FH, Hugosson J, Roobol MJ, Tammela TL, Ciatto S, Nelen V,

et al. Screening and prostate-cancer mortality in a randomized Euro-pean study. N Engl J Med 2009;360:1320–8.

3. Moyer VA. Screening for prostate cancer: U.S. preventive servicestask force recommendation statement. Ann Intern Med 2012;157:120–34.

4. Parachoniak CA, Park M. Dynamics of receptor trafficking in tumor-igenicity. Trends Cell Biol 2012;22:231–40.

5. Hurley JH, Odorizzi G. Get on the exosome bus with ALIX. Trends CellBiol 2012;14:654–5.

6. Hu CT, Wu JR, Wu WS. The role of endosomal signaling triggered bymetastatic growth factors in tumor progression. Cell Signal 2013;25:1539–45.

7. Settembre C, Fraldi A, Medina DL, Ballabio A. Signals from thelysosome: a control centre for cellular clearance and energy metab-olism. Nat Rev Mol Cell Biol 2013;14:283–96.

8. Boucrot E, Kirchhausen T. Endosomal recycling controls plasma mem-brane area duringmitosis. Proc Natl Acad Sci U S A 2007;104:7939–44.






9. Palfy M, Remenyi A, Korcsmaros T. Endosomal crosstalk: meetingpoints for signaling pathways. Trends Cell Biol 2012;22:447–56.

10. Glunde K, Guggino SE, Solaiyappan M, Pathak AP, Ichikawa Y,Bhujwalla ZM. Extracellular acidification alters lysosomal traffickingin human breast cancer cells. Neoplasia 2003;5:533–45.

11. Stow JL, Murray RZ. Intracellular trafficking and secretion of inflam-matory cytokines. Cytokine Growth Factor Rev 2013;24:227–39.

12. Henneberry MO, Engel G, Grayhack JT. Acid phosphatase. Urol ClinNorth Am 1979;6:629–41.

13. Quintero IB, Araujo CL, Pulkka AE,Wirkkala RS, Herrala AM, EskelinenEL, et al. Prostatic acid phosphatase is not a prostate specific target.Cancer Res 2007;67:6549–54.

14. Johnson IR, Parkinson-Lawrence EJ, Butler LM, Brooks DA. Prostatecell lines as models for biomarker discovery: Performance of currentmarkers and the search for new biomarkers. Prostate 2014;74:547–60.

15. Camacho L, Meca-Cortes O, Abad JL, Garcia S, Rubio N, Diaz A, et al.Acid ceramidase as a therapeutic target in metastatic prostate cancer.J Lipid Res 2013;54:1207–20.

16. Kuronita T, Eskelinen EL, Fujita H, Saftig P, HimenoM, TanakaY. A rolefor the lysosomal membrane protein LGP85 in the biogenesis andmaintenance of endosomal and lysosomal morphology. J Cell Sci2002;115:4117–31.

17. Huotari J, Helenius A. Endosome maturation. EMBO J 2011;30:3481–500.

18. Zerial M, McBride H. Rab proteins as membrane organizers. Nat RevMol Cell Biol 2001;2:107–17.

19. Jordens I, Marsman M, Kuijl C, Neefjes J. Rab proteins, connectingtransport and vesicle fusion. Traffic 2005;6:1070–7.

20. Capes-Davis A, Theodosopoulos G, Atkin I, Drexler HG, Kohara A,MacLeod RA, et al. Check your cultures! A list of cross-contaminatedor misidentified cell lines. Int J Cancer 2010;127:1–8.

21. Sardiello M, Palmieri M, di Ronza A, Medina DL, ValenzaM, GennarinoVA, et al. A gene network regulating lysosomal biogenesis and func-tion. Science 2009;325:473–7.

22. Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS,et al. Integrative genomic profiling of human prostate cancer. CancerCell 2010;18:11–22.

23. Shukla S, Maclennan GT, Hartman DJ, Fu P, Resnick MI, Gupta S.Activation of PI3K-Akt signaling pathway promotes prostate cancercell invasion. Int J Cancer 2007;121:1424–32.

24. ParkinDM,BrayF, Ferlay J, Pisani P.Global cancer statistics, 2002.CACancer J Clin 2005;55:74–108.

25. Australian Institute of Health and Welfare (AIHW). Cancer in Australia2010: an overview. Cancer series no. 60. Cat. no. CAN 56. Canberra:AIHW; 2010 Jan 1.

26. Roobol MJ, Haese A, Bjartell A. Tumour markers in prostate cancer III:biomarkers in urine. Acta Oncol 2011;50(Suppl 1):85–9.

27. Ploussard G, de laTaille A. Urine biomarkers in prostate cancer. NatRev Urol 2010;7:101–9.

28. Eberlin LS, Dill AL, Costa AB, Ifa DR, Cheng L, Masterson T, et al.Cholesterol sulfate imaging in human prostate cancer tissue bydesorption electrospray ionization mass spectrometry. Anal Chem2010;82:3430–4.

29. Yao S, Ireland SJ, Bee A, Beesley C, Forootan SS, Dodson A, et al.Splice variant PRKC-zeta(-PrC) is a novel biomarker of humanprostatecancer. Br J Cancer 2012;107:388–99.

30. Sloane BF, Moin K, Sameni M, Tait LR, Rozhin J, Ziegler G. Membraneassociation of cathepsin B can be induced by transfection of humanbreast epithelial cells with c-Ha-ras oncogene. J Cell Sci 1994;107:373–84.

31. Steffan JJ, Snider JL, Skalli O, Welbourne T, Cardelli JA. Naþ/Hþ

Exchangers and RhoA regulate acidic extracellular pH-induced lyso-some trafficking in prostate cancer cells. Traffic 2009;10:737–53.

32. Steffan JJ, Cardelli JA. Thiazolidinediones induce Rab7–RILP–MAPK–dependent juxtanuclear lysosome aggregation and reduce tumor cellinvasion. Traffic 2010;11:274–86.

33. Roshy S, Sloane BF, Moin K. Pericellular cathepsin B and malignantprogression. Cancer Metastasis Rev 2003;22:271–86.

34. Authier F, Kouach M, Briand G. Endosomal proteolysis of insulin-likegrowth factor-I at its C-terminal D-domain by cathepsin B. FEBS Lett2005;579:4309–16.

35. Zaidi N, Maurer A, Nieke S, Kalbacher H. Cathepsin D: a cellularroadmap. Biochem Biophys Res Commun 2008;376:5–9.

36. Deacon SW, Nascimento A, Serpinskaya AS, Gelfand VI. Regulation ofbidirectional melanosome transport by organelle bound MAP kinase.Curr Biol 2005;15:459–63.

37. Pasini FS, Maistro S, Snitcovsky I, Barbeta LP, Rotea Mangone FR,Lehn CN, et al. Four-gene expression model predictive of lymph nodemetastases in oral squamous cell carcinoma. Acta Oncol 2012;51:77–85.

38. Steffan JJ, Dykes SS, Coleman DT, Adams LK, Rogers D, Carroll JL,et al. Supporting a role for the GTPase Rab7 in prostate cancerprogression. PLoS ONE 2014;9:e87882.

39. Keer HN, Kozlowski JM, Tsai YC, Lee C, McEwan RN, Grayhack JT.Elevated transferrin receptor content in human prostate cancer celllines assessed in vitro and in vivo. J Urol 1990;143:381–5.

40. O'Donnell KA, Yu D, Zeller KI, Kim JW, Racke F, Thomas-TikhonenkoA, et al. Activation of transferrin receptor 1 by c-Myc enhances cellularproliferation and tumorigenesis. Mol Cell Biol 2006;26:2373–86.

41. Edinger AL, Thompson CB. Akt maintains cell size and survival byincreasing mTOR-dependent nutrient uptake. Mol Biol Cell 2002;13:2276–88.

42. Puri C, ChibalinaMV, Arden SD, Kruppa AJ, Kendrick-Jones J, Buss F.Overexpression of myosin VI in prostate cancer cells enhances PSAand VEGF secretion, but has no effect on endocytosis. Oncogene2010;29:188–200.

43. Fili N, Calleja V, Woscholski R, Parker PJ, Larijani B. Compartmentalsignal modulation: endosomal phosphatidylinositol 3-phosphate con-trols endosome morphology and selective cargo sorting. Proc NatlAcad Sci U S A 2006;103:15473–8.

44. Lee JR, Hahn HS, Kim YH, Nguyen HH, Yang JM, Kang JS, et al.Adaptor protein containingPHdomain, PTBdomain and leucine zipper(APPL1) regulates the protein level of EGFR by modulating its traffick-ing. Biochem Biophys Res Commun 2011;415:206–11.

45. Wang C, Xin X, Xiang R, Ramos FJ, Liu M, Lee HJ, et al. Yin-Yangregulation of adiponectin signaling by APPL isoforms in muscle cells.J Biol Chem 2009;284:31608–15.

46. Yang L, LinHK, Altuwaijri S, Xie S,WangL,ChangC. APPL suppressesandrogen receptor transactivation via potentiating Akt activity. J BiolChem 2003;278:16820–7.

47. Broussard JA, LinWH,Majumdar D, Anderson B, Eason B, BrownCM,et al. The endosomal adaptor protein APPL1 impairs the turnover ofleading edge adhesions to regulate cell migration. Mol Biol Cell2012;23:1486–99.

48. Chotard L, Mishra AK, Sylvain MA, Tuck S, Lambright DG, RocheleauCE. TBC-2 regulates RAB-5/RAB-7–mediated endosomal traffickingin Caenorhabditis elegans. Mol Biol Cell 2010;21:2285–96.

49. Peralta ER, Martin BC, Edinger AL. Differential effects of TBC1D15and mammalian Vps39 on Rab7 activation state, lysosomal mor-phology, and growth factor dependence. J Biol Chem 2010;285:16814–21.

50. Subramani D, Alahari SK. Integrin-mediated function of Rab GTPasesin cancer progression. Mol Cancer 2010;9:312.

51. Liu AY, Roudier MP, True LD. Heterogeneity in primary and metastaticprostate cancer as defined by cell surface CD profile. Am J Pathol2004;165:1543–56.


Johnson et al.




2014;12:1851-1862. Published OnlineFirst July 30, 2014.Mol Cancer Res Ian R.D. Johnson, Emma J. Parkinson-Lawrence, Tetyana Shandala, et al. PotentialAltered Endosome Biogenesis in Prostate Cancer Has Biomarker

Updated version

10.1158/1541-7786.MCR-14-0074doi:

Access the most recent version of this article at:

Material

Supplementary

http://mcr.aacrjournals.org/content/suppl/2014/07/31/1541-7786.MCR-14-0074.DC1

Access the most recent supplemental material at:

Cited articles

http://mcr.aacrjournals.org/content/12/12/1851.full#ref-list-1

This article cites 50 articles, 15 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/12/12/1851.full#related-urls

This article has been cited by 1 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mcr.aacrjournals.org/content/12/12/1851To request permission to re-use all or part of this article, use this link



http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-14-0074

http://mcr.aacrjournals.org/content/suppl/2014/07/31/1541-7786.MCR-14-0074.DC1

http://mcr.aacrjournals.org/content/12/12/1851.full#ref-list-1

http://mcr.aacrjournals.org/content/12/12/1851.full#related-urls

http://mcr.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://mcr.aacrjournals.org/content/12/12/1851


altered endosome biogenesis in prostate cancer has ... · oncogenes and tumor suppressors altered...

Documents