Heparanase 2 Attenuates Head and Neck Tumor Vascularity...

Tumor and Stem Cell Biology

Heparanase 2 Attenuates Head and Neck TumorVascularity and GrowthMiriam Gross-Cohen1, Sari Feld1, Ilana Doweck2, Gera Neufeld1, Peleg Hasson3,Gil Arvatz1, Uri Barash1, Inna Naroditsky4, Neta Ilan1, and Israel Vlodavsky1

Abstract

The endoglycosidase heparanase specifically cleaves theheparan sulfate (HS) side chains on proteoglycans, an activitythat has been implicated strongly in tumor metastasis and angio-genesis. Heparanase-2 (Hpa2) is a close homolog of heparanasethat lacks intrinsic HS-degrading activity but retains the capacityto bind HS with high affinity. In head and neck cancer patients,Hpa2 expression was markedly elevated, correlating with pro-longed time to disease recurrence and inversely correlating withtumor cell dissemination to regional lymph nodes, suggestingthat Hpa2 functions as a tumor suppressor. The molecularmechanism associated with favorable prognosis followingHpa2 induction is unclear. Here we provide evidence thatHpa2 overexpression in head and neck cancer cells markedlyreduces tumor growth. Restrained tumor growth was associated

with a prominent decrease in tumor vascularity (blood andlymph vessels), likely due to reduced Id1 expression, a tran-scription factor highly implicated in VEGF-A and VEGF-C generegulation. We also noted that tumors produced by Hpa2-overexpressing cells are abundantly decorated with stromalcells and collagen deposition, correlating with a markedincrease in lysyl oxidase expression. Notably, heparanase enzy-matic activity was unimpaired in cells overexpressing Hpa2,suggesting that reduced tumor growth is not caused by hepar-anase regulation. Moreover, growth of tumor xenografts byHpa2-overexpressing cells was unaffected by administration ofa mAb that targets the heparin-binding domain of Hpa2,implying that Hpa2 function does not rely on heparanase orheparan sulfate. Cancer Res; 76(9); 1–11. �2016 AACR.

IntroductionHeparanase is an endoglycosidase that specifically cleaves

heparan sulfate (HS) side chains, a class of glycosaminoglycansabundantly present in the extracellular matrix (ECM) and on thecell surface. Heparanase activity is strongly implicated in tumorangiogenesis and metastasis attributed to remodeling of thesubepithelial and subendothelial basement membranes (1, 2).Augmented level of heparanase was documented in an increasingnumber of human carcinomas and hematologic malignancies(3, 4). In many cases, heparanase induction correlates withincreased tumor metastasis, vascular density, and shorter survivalpostoperation, thus providing a strong clinical support for theprotumorigenic function of the enzyme (4–6). These studiesdepict compelling evidence for the clinical relevance of hepar-anase, making it an attractive target for the development ofanticancer drugs (5, 7, 8). Cloning of a single human heparanase

cDNA sequence was independently reported by several groups(9–12) implying that one active heparanase enzyme exists inmammals. Further analysis of human genomic DNA led research-ers to conclude that the heparanase gene is unique, and that theexistenceof relatedproteins is unlikely.On thebasis of amino acidsequence, McKenzie and colleagues nonetheless reported thecloning of heparanase homolog termed heparanase 2 (Hpa2;ref. 13). Unlike heparanase, Hpa2 lacks intrinsic HS-degradingactivity, the hallmark of heparanase, but retains the capacity tobind heparin/HS (14). In fact, Hpa2 exhibits even higher affinitytoward heparin/HS than heparanase, suggesting that Hpa2 mayinhibit heparanase activity by competition for the HS substrate(14). Clinically, Hpa2 expression was markedly elevated in headand neck carcinoma patients, correlating with prolonged time todisease recurrence (follow-up to failure) and inversely correlatingwith tumor cell dissemination to regional lymph nodes, suggest-ing that Hpa2 functions as a tumor suppressor (14). The molec-ular mechanism associated with favorable prognosis followingHpa2 overexpression is unclear. Here, we provide evidence thatHpa2 overexpression in head and neck cancer cells markedlyreduces tumor growth. Restrained tumor growth was associatedwith a prominent reduction in tumor vascularity (blood andlymph vessels) likely due to reduced Id1 expression, a transcrip-tion factor highly implicated in VEGF-A and VEGF-C gene regu-lation (15). Notably, heparanase enzymatic activity was notimpaired in cells overexpressing Hpa2, suggesting that reducedtumor growth is not due to heparanase regulation. Moreover,growth of tumor xenografts produced by Hpa2-overexpressingcells was not affected by a mAb that targets a heparin-bindingdomain of Hpa2, implying that Hpa2 functions in heparanase-,and HS-independent manner.

1Cancer and Vascular Biology Research Center, The Bruce RappaportFaculty of Medicine, Technion-Israel Institute of Technology, Haifa,Israel. 2Department of Otolaryngology, Head and Neck Surgery, Car-mel Medical Center, Haifa, Israel. 3Department of Anatomy and CellBiology, The Bruce Rappaport Faculty of Medicine, Technion-IsraelInstitute of Technology, Haifa, Israel. 4Department of Pathology, Ram-bam Health Care Campus, Haifa, Israel.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Author: Israel Vlodavsky, Technion-Israel Institute of Technol-ogy, 1 Efron st., Haifa 31096, Israel. Phone: 972-4829-5410; Fax: 972-4851-0445;E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-15-1975

�2016 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org OF1

Research. on March 22, 2019. © 2016 American Association for Cancercancerres.aacrjournals.org Downloaded from

Published OnlineFirst March 24, 2016; DOI: 10.1158/0008-5472.CAN-15-1975

http://cancerres.aacrjournals.org/

Materials and MethodsCells and cell culture, immunoblotting, andheparanase activityassay

Cal27 tongue carcinoma and FaDu pharyngeal carcinoma cellshave been described previously (16–18). Cells were grown inDMEM (Biological Industries) supplemented with 10% FBS andantibiotics. Cells were passed in culture no more than 2 monthsafter being thawed from authentic stocks. Cells were infected withcontrol empty vector (Vo) or Hpa2 gene construct, selected withpuromycin (2 mg/mL; Invitrogen), expended and pooled. Cellclones were isolated by limiting dilution and clones exhibitinghigher Hpa2 expression versus the pool of cells were evaluated byimmunoblotting, carried out essentially as described (14). Prep-aration of dishes coated with sulfate-labeled ECM and determi-nation of heparanase enzymatic activity (i.e., release of sulfate-labeled HS degradation fragments) were carried out essentially asdescribed previously (14).

Antibodies and reagentsAnti-Hpa2 polyclonal (Ab 58) and monoclonal (20c5, 1c7)

antibodies have been described previously (14). Immunohisto-chemical grade anti-LOX and anti-LOXL2 polyclonal antibodieshave been described elsewhere (19, 20). Antibodies directedagainst Ki67, LYVE-1, LOX, cytokeratin 13, cytokeratin 15, b3-tubulin, cleaved caspase-3, hypoxia-inducible factor-1a (HIF1a),and carbonic anhydrase IX were purchased from Abcam; Rat anti-mouse CD31 was from Dianova; anti-VEGF-A and anti-VEGF-Cantibodies were from Santa Cruz Biotechnology. Anti-actin andanti-smooth muscle actin mAbs, Masson/Trichrome staining kitand b-3-aminopropionitrile fumarate (BAPN) were purchasedfrom Sigma. Human Id1 gene construct was purchased fromAddgene.

Real-time PCRReal-time PCR analyses were performed using ABI PRISM 7000

Sequence Detection System employing SYBR Green PCR MasterMix (Applied Biosystems). The following primers were used:

Actin F: 50-CGCCCCAGGCACCAGGGC, R: 50-GCTGGGGTGTT-GAAGGT;

VEGF-A F: 50-TCTACCTCCACCATGCCAAGT, R: 50 TGTCCACC-AGGGTCTCGATT;

VEGF-C F: 50-GCCAACCTCAACTCAAGGAC, R: 50-CCCACATC-TGTAGACGGACA;

LOX F: 50-GTACGTGCAGAAGATGTCC, R: 50-CTGAGCAGCACC-CTGTGATC;

LOXL2 F: 50-TCGAGGTTGCAGAATCCGATT, R 50 TTCCGTCTCT-TCGCTGAAGGA;

Id1 F: 50-CTGCTCTACGACATGAACGG, R 50-GAAGGTCCCTGA-TGTAGTCGAT.

Tumorigenicity and IHCCells were detached with trypsin/EDTA, washed with PBS, and

brought to a concentration of 5 � 107 cells/mL. Cell suspension(5�106/0.1mL)was inoculated subcutaneously at the rightflankof 6-week-old NOD/SCID mice. Xenografts size was determinedby externallymeasuring tumors in twodimensions using a caliper.At the end of the experiment, mice were sacrificed; tumor

xenografts were removed, weighed, and fixed in formalin. Paraf-fin-embedded 5-mm sections were subjected to immunostainingwith the indicated antibody using the EnVision Kit according tothe manufacturer's (Dako) instructions, as described previously(14). Immunocytochemistry was carried out essentially asdescribed (21). Lymph and blood vessels density was quantifiedby the Image Pro software orwere counted under lightmicroscopeat high (�40) magnification.

Statistical analysisData are presented as means � SE. Statistical significance was

analyzed by two-tailed Student t test. Values of P < 0.05 wereconsidered significant. Data sets passed D'Agostino–Pearson nor-mality (GraphPad Prism 5 utility software). All experiments wererepeated at least three times with similar results.

ResultsHpa2 overexpression attenuates tumor growth

To reveal the function of Hpa2 in head and neck cancer, FaDupharyngeal carcinoma cells were infected with control (Vo) orHpa2 gene constructs and expression was confirmed by immu-noblotting (Supplementary Fig. S1A; Pool) and immunofluores-cent staining (Supplementary Fig. S1B). Tumor xenografts pro-duced by FaDu cells overexpressing Hpa2 were markedly smallerby volume and weight compared with control tumors (Fig. 1A;P ¼ 0.001). Histologic examination showed that xenograftsproduced by control cells were highly necrotic (Fig. 1B, left). Incontrast, xenografts produced by cells overexpressing Hpa2 wereby far less necrotic and were decorated with large cysts (Fig. 1B,right). Similar robust cysts formation was evident in Cal27 cellsoverexpressing Hpa2 (Fig. 1B, bottom). In head and neck cancerpatients, high levels of Hpa2 expression were associated withreduced lymph nodes metastasis and prolonged survival rates(14). We therefore examined the occurrence of lymph vessels intumor xenografts produced by control and Hpa2 overexpressingcells. We observed a significant 2- to 3-fold decrease in lymphan-giogenesis following Hpa2 overexpression (Fig. 1C and D; P ¼0.002), associatedwith a comparable decrease in the expressionofVEGF-C (Fig. 1D, bottom), a predominant factor for the prolif-eration of lymphatic endothelial cells (22, 23).

Tumor growth and vascularity are markedly reduced by Hpa2-overexpressing cell clones

To further delineate the impact of Hpa2 on tumor growth, weselected cell clones that exhibit high levels of Hpa2 expression.Three such cell clones were isolated from Hpa2-infected cells(clones #6, 60, 64; Supplementary Fig. S1A) and their tumorigeniccapacity was compared with that of three randomly selectedcontrol cell clones (clones #3, 5, 8). Tumor xenografts producedby Hpa2-overexpressing clones were strikingly, 10-fold smallercompared with xenografts produced by control cell clones,decrease that is statistically highly significant (Fig. 2A and Sup-plementary Fig. S1C). Proliferating, Ki67-positive cells weredetected in the entire non-necrotic tumor mass produced bycontrol clones (Fig. 2B, left). Endothelial cells and tumor cellsresiding in lymphatics were also stained positive for Ki67 incontrol tumors (Supplementary Fig. S2A). In contrast, Ki67 reac-tivity was restricted to the tumor periphery in tumors produced byHpa2 clones (Fig. 2B, right) and was decreased significantly evenin these areas (Fig. 2B, bottom). In addition, tumors produced by

Gross-Cohen et al.

Cancer Res; 76(9) May 1, 2016 Cancer ResearchOF2




Hpa2 clones exhibited higher levels of apoptotic cell deathevident by caspase-3 staining (Supplementary Fig. S2B), altogeth-er attenuating tumor growth. Histologically, tumors produced bycontrol cell clones exhibitedmassive areas of necrosis versus largecysts that decorated the tumors produced by Hpa2 clones (Fig.2C), in agreement with the histologic phenotype of tumorsproduced by cell pools (Fig. 1B). Careful pathologic examinationrevealed that tumors produced by Hpa2 clones exhibit higherdegree of differentiation (Fig. 2C, right), and this was confirmedby a marked increase in cytokeratin 13 and cytokeratin 15 immu-nostaining (Fig. 2D) and immunoblotting (Supplementary Fig.S2C). Furthermore, a noticeable decrease in lymph vessels (pos-itive for LYVE-1) density was quantified in tumor xenograftsproduced by Hpa2 clones compared with controls (Fig. 3A,second and fifth left panels), associating with a 3-fold decreasein VEGF-C expression levels (Fig. 3B). This agrees with decreasedlymphangiogenesis and VEGF-C expression observed in the cellpools (Fig. 1C and D). Notably, while tumor cells were readilydetected within lymph vessels of control tumors (Fig. 3A, secondleft panels), Hpa2 overexpression resulted not only in fewer

lymph vessels but also reduced tumor load in lymphatics (Fig.3A, second right panels). While this observation is in accordancewith our previous experimental (Fig. 1C and D) and clinicalresults (14), it cannot explain the robust attenuation of tumorgrowth. We therefore examined also the density of blood vesselsin tumors produced by control andHpa2-overexpressing cells.Wefound that blood vessel density was prominently reduced intumors produced by Hpa2-overexpressing cells compared withcontrol tumors (Fig. 3A, third and fifth right panels), associatingwith a comparable decrease in VEGF-A expression (Fig. 3C andSupplementary Fig. S1D). Consequently, tumors produced byHpa2 clones exhibited augmented tumor hypoxia evident byincreased HIF1a (Fig. 3A, fourth panels) and carbonic anhydraseIX (CAIX; Supplementary Fig. S2D) staining intensity.

Hpa2 enhances collagen deposition and LOX inductionHistologic examination revealed that tumors produced by Hpa2

cells are decorated extensively with host, stromal cells (Supplemen-tary Fig. S3A). Masson/Trichrome staining showed a marked in-crease in collagen staining following Hpa2 overexpression (Fig. 4A,

Figure 1.Hpa2 overexpression attenuates tumor growth. A, control (Vo) and Hpa2-overexpressing FaDu cells (5 � 106) were implanted subcutaneously in SCID mice andtumor volume was inspected (left). At termination, tumor xenografts were collected, weighed (right), and formalin-fixed. Paraffin-embedded 5-mm sections weresubjected to histologic examination. B, shown are representative images of hematoxylin and eosin staining at low (top) and high (middle) magnifications. Bottom,hematoxylin andeosin staining of tumor xenografts producedby control (Vo) andHpa2-overexpressingCal27 oral carcinomacells is shown.Notemassive necrosis incontrol tumors versus cysts structures in tumors overexpressing Hpa2. Original magnifications, top,�4; middle and bottom,�10. C and D, lymph angiogenesis andVEGF-C expression. Five-micron sections of tumor xenografts produced by control (Vo) and Hpa2-overexpressing FaDu cells were stained with anti-LYVE-1antibody, a marker for lymphatic endothelial cells (C); quantification of lymph vessel density is shown graphically in D (top). Original magnification,�40. Extracts ofcontrol and Hpa2-overexpressing cells were subjected to immunoblotting, applying anti-VEGF-C and anti-actin antibodies (D, bottom).

Hpa2 Attenuates Tumor Growth

www.aacrjournals.org Cancer Res; 76(9) May 1, 2016 OF3




top, blue). A similar increase in collagen staining was observed intumors produced by Cal27 cells overexpressing Hpa2 (Supplemen-tary Fig. S3B, second panels).We therefore examined the expressionof lysyl oxidase (LOX), an enzyme that is strongly implicated incollagen deposition and tissue fibrosis (24, 25). We found a prom-inent increase in LOX immunostaining in cancer cells that surroundthe stromal elements within tumors produced by Hpa2 clonescompared with control (Fig. 4A, second panel and SupplementaryFig. S4A, top). LOXupregulationbyHpa2was further ascertainedbyreal-time PCR (Fig. 4B). In contrast, the expression of LOX-like 2(LOXL2)wasdecreased inHpa2 clones (Fig. 4C).Notably, however,LOXL2was predominantly localized to the nuclei of cancer (Fig. 4A,bottom right panels) and stromal cells (Supplementary Fig. S3C,arrows). Similar phenotypes were observed in tumors produced byCal27 cells overexpressingHpa2 (Supplementary Fig. S3B, bottom).

Hpa2 mode of actionTo reveal the mechanism underlying the potent suppression of

tumor growth by Hpa2, we first evaluated heparanase activity incontrol and Hpa2-overexpressing cells, because heparanase exhi-bits protumorigenic properties inmany types of cancers includinghead and neck (26), and Hpa2 can inhibit heparanase enzymaticactivity (14).Notably, heparanase activity appeared unchanged incontrol versusHpa2-overexpressing cells (Fig. 5A), suggesting thatthe decrease in tumor growth by Hpa2 is not due to heparanaseinhibition. Despite its lack of HS-degrading activity, Hpa2 exhi-bits high affinity toward heparin/HS (14). We therefore sought toexamine whether binding and clustering of HS proteoglycans(HSPG) by Hpa2 on the cell surface is responsible for tumorsuppression. To this end, we screened a panel of anti-Hpa2 mAbsfor inhibition of cellular binding of Hpa2 shown previously to be

Figure 2.Marked inhibition of tumor growth byHpa2-overexpressing cell clones. Cellclones were isolated by limited dilutionand clones that showed Hpa2expression significantly higher than thecell pool were selected for subsequentexperiments (#6, #60, #64). Threeclones from control (Vo) cells wereselected randomly (#3, #5, #8). A,tumor growth. The indicated clone orpool cells were implantedsubcutaneously in SCID mice andtumor growth was inspected over time(top). At termination, tumors werecollected, photographed (insets),weighed (bottom), and fixed informalin for histologic evaluation. B,cell proliferation. Five-micron sectionsof control (Vo #3) and Hpa2-overexpressing (#6) cell clones weresubjected to immunostaining, applyinganti-Ki67 antibody. Shown arerepresentative photomicrographs atlow (�4; top) and high (�40; bottom)magnifications. Quantification of Ki67-positive cell is shown in the lower panel.C and D, cell differentiation. Five-micron sections of tumor xenograftsproduced by control (#3) and Hpa2-overexpressing (#6) cell clones weresubjected to histologic evaluation. C,shown are representativephotomicrographs of hematoxylinand eosin staining at low (�4; left)and high (�40; right) magnifications.D, sections were subjected toimmunostaining, applying anti-cytokeratin 13 (top) and cytokeratin15 (bottom) antibodies. Note,increased differentiation of tumorcells overexpressing Hpa2. Originalmagnifications, �40.

Gross-Cohen et al.





mediated by cell membrane HSPG such as syndecans (14). Wefound that mAb 1c7 efficiently inhibits cellular binding of Hpa2in a dose-dependent manner (Fig. 5B), suggesting that this mAbtargets a functional HS-binding domain of Hpa2. Importantly,tumor growth was not altered by administration of mAb 1c7 tomice inoculated with Hpa2 clones (Fig. 5C and SupplementaryFig. S5A). Interestingly, the formation of cysts within tumorsappeared to be increased in mice treated with mAb 1c7 (Fig.5C and Supplementary Fig. S5A). Immunoblotting revealedreduced levels of VEGF-A in Hpa2-overexpressing clones (Fig.5D, top and Supplementary Fig. S1D), in agreement with real-time PCR (Fig. 3C), but VEGF-A levels were not affected bytreatment with mAb 1c7 (Fig. 5D, top, þ). Taken together, itappears that inhibition of tumor growth and vascularity by Hpa2does not involve regulation of heparanase activity or the activa-

tion of cellmembraneHSPG. Furthermore, treatmentwith BAPN,which inhibits the enzymatic activity of all LOXs (27) reducedtumor fibrosis (i.e., collagen content) but had no effect on tumorsize (Fig. 5E and Supplementary Fig. S5B), implying that reducedtumor growth by cells overexpressing Hpa2 is not due to theelevation of LOX activity. We therefore considered the reducedtumor vascularity as the main cause for tumor growth inhibitionby Hpa2. Interestingly, we found that expression of Id1, a tran-scription factor implicated in the induction of VEGF-A and VEGF-C gene transcription (15, 28), is reduced in Hpa2-overexpressingcells (Fig. 6A). Overexpression of Id1 in Hpa2 clones (Fig. 6B)resulted in a 2.5- to 3-fold increase in tumor growth (Fig. 6C) anda comparable increase in tumor vascularity (Fig. 6D and Supple-mentary Fig. S5C), implying that downregulationof Id1mediates,at least in part, the tumor suppressor function of Hpa2.

Figure 3.Hpa2 overexpression is associated withdecreased tumor vascularity. A,immunostaining. Sections of tumorxenografts produced by control (#3, #5)and Hpa2-overexpressing (#6, #64) cellclones were subjected to immunostaining,applying anti-Hpa2 (top), anti-LYVE-1(second panel), anti-CD31 (third panel),and anti-HIF1a (fourth panel) antibodies.Note, reduced tumor vascularity andincreased tumor hypoxia following Hpa2overexpression. Original magnifications,top, third and fourth panels, �40; secondpanel, �20. Bottom, quantification oflymph (LYVE-1-positive) and blood (CD31-positive) vessels is shown graphically. Band C, real-time PCR. Total RNA wasextracted from the indicated cell clonesand corresponding cDNAs were subjectedto quantitative real-time PCR analyses.Expression of VEGF-C (B) and VEGF-A (C)are presented graphically in relation to thelevels in control clone #3, set arbitrarily to avalue of 1.






DiscussionCompelling preclinical and clinical evidence tie heparanase

with cancer initiation and progression (8, 29–31), making it apromising drug target (5, 7). In striking contrast, the biologicsignificance of its close homolog Hpa2 in tumorigenesis islargely obscure. We have reported previously that Hpa2 levelsare increased in head and neck carcinoma compared with

adjacent normal tissue, and that patients with high levels ofHpa2 are endowed with reduced lymph node metastasis andprolonged survival rates (14). Similarly, increased Hpa2 expres-sion was associated with prolonged survival of gastric cancerpatients (32) but themechanismunderlying this favorable effecthas not been resolved yet. Here, we provide evidence that highlevels of Hpa2 are associated with reduced lymph vessel density.

Figure 4.Hpa2 induces LOX expression and LOXL2 nuclear localization. A, Masson/Trichrome and immunostaining. Five-micron sections from tumor xenografts producedby the indicated control (#3, #5) and Hpa2-overexpressing (#60, #64) cell clones were stained with Masson/Trichrome reagents (top). Blue staining denotescollagen type I deposition. Corresponding tumor sections were subjected to immunostaining, applying anti-LOX (second panel) and anti-LOXL2 (third panel)antibodies. Note increased collagen deposition by cells overexpressing Hpa2, associating with increased LOX levels and nuclear localization of LOXL2.Original magnification, �40. B and C, real-time PCR. Total RNA was extracted from the indicated cell clones and corresponding cDNAs were subjected toquantitative real-time PCR analyses. Expression levels of LOX (B) and LOXL2 (C) are shown graphically in relation to the levels in control clone #3, set arbitrarilyto a value of 1.

Gross-Cohen et al.





This was noted in pool of FaDu cells infected with Hpa2 geneconstruct (Fig. 1C andD) andwasmost evident in three differentcell clones selected for higher levels of Hpa2 expression (Fig.3A). Importantly, these clones express Hpa2 levels comparablewith patients exhibiting high levels of Hpa2, (SupplementaryFig. S4B; ref. 14) and are thus considered relevant to the disease.Lymph angiogenesis is significant clinically for many types ofcancers but most appealing in head and neck cancer becauselymph node metastasis is the most important clinical parameterfor this malignancy (33). We found reduced lymph vesseldensity in Hpa2 tumors combining with a comparable decreasein VEGF-C expression (Fig. 3B), a most potent prolymphaticmediator secreted by tumor cells (34). While decreased VEGF-Cexpression and lymph angiogenesis by Hpa2 nicely agrees with

the clinical results (14, 32), it cannot explain the prominentdecrease in tumor growth following Hpa2 overexpression (Fig.2A and Supplementary Fig. S1C). Notably, we found that notonly VEGF-C but also VEGF-A expression is reduced by Hpa2(Figs. 3C and 5D and Supplementary Fig. S1D), associating withreduced blood vessel density (Fig. 3A) that likely attenuatestumor growth. This agrees with increased levels of hypoxia intumors produced by Hpa2 clones, evident by higher levels ofHIF1a and carbonic anhydrase IX (Fig. 3A and SupplementaryFig. S2D), well-recognized markers of hypoxic conditions.Tumors produced by Hpa2-overexpressing clones appearedsignificantly less necrotic despite being poorly vascularized. Thisis due to a marked decrease in tumor cell proliferation (Fig. 2B)and increased apoptosis (Supplementary Fig. S2B), thus

Figure 5.Hpa2 mode of action. A, heparanase activity. Control (#3) and Hpa2-overexpressing (#6) cell clones were plated on dishes coated with sulfate-labeled ECM andgrown for 3 days. Conditioned mediumwas then collected and evaluated for heparanase activity as described in Materials and Methods. B, characterization of mAb1c7. Hpa2 (1 mg/mL) was added to 293 cells together with the indicated anti-Hpa2 mAb or control mouse IgG (5 mg/mL; 30 minutes, 37�C) under serum-freeconditions. Medium was then aspirated, cells were washed twice with PBS, and lysate samples were subjected to immunoblotting, applying anti-Hpa2 (top) andanti-actin (second panel) antibodies. Decreased cellular binding of Hpa2 by mAb 1c7 is shown to be dose-dependent (third panel). C, tumor xenografts. Hpa2-overexpressing clone #6 cells were implanted (5� 106) subcutaneously in SCIDmice (n¼ 7) andmicewere administratedwithmAb 1c7 (250 mg/mouse) three timesa week. At termination, tumors were collected, weighed (left), and processed for histologic examination. Shown are representative images of whole sectionsstained with Masson/Trichrome reagents and scanned by 3DHISTECH Pannoramic MIDI System attached to HITACHI HV-F22 color camera (3DHISTECH). Note thatmAb 1c7 does not affect tumor growth but appears to enhance cyst formation. D, VEGF-A expression. The indicated cell clones were incubated without (�)or with (þ) mAb 1c7 (5 mg/mL, 18 hours, 37�C) and lysate samples were subjected to immunoblotting, applying anti-VEGF-A (top) and anti-actin (bottom)antibodies. Note decreased VEGF-A levels in cells overexpressing Hpa2, which is not affected by mAb 1c7. E, LOX inhibitor. SCID mice (n ¼ 7) were inoculatedsubcutaneously with Hpa2-overexpressing clone #6 cells (5 � 106) and mice were administrated with LOX inhibitor BAPN (0.75% in drinking water). Attermination, tumors were collected, weighed (left), and processed for histologic examination. Shown are representative images of whole sections stainedwith Masson/Trichrome reagents and scanned by 3DHISTECH Pannoramic MIDI System (right). Note that LOX inhibitor does not affect tumor growth butreduces tumor fibrosis (i.e., collagen content).






balancing tumor mass and blood vessel density closer to theratio occurring in normal tissues.

The mechanism responsible for VEGF-A and VEGF-C down-regulation is not entirely clear but likely involves Id1. Thistranscription factor is implicated in the induction of VEGF-A andVEGF-C gene expression (35, 36) and is decreased followingHpa2overexpression (Fig. 6A). Notably, transfection of Hpa2 cloneswith Id1 gene construct, resulting in 2-fold increase in its mRNAlevels (not shown), restored tumor vascularity and growth (Fig.6C and D and Supplementary Fig. S5C), thus supporting thenotion that reduced Id1 levels underlie tumor inhibition byHpa2and suggesting thatmodest changes in the expression levels of thistranscription factor are highly effective in modulating tumorprogression. This appears relevant clinically, because Id1 expres-sion has been shown to be elevated in a variety of primary humantumors includinghead andneck cancer, correlatingwith increased

tumor aggressiveness (37). It should be noted that Id1 is thoughtto function as a cellular proto-oncogene (37) and regulates thetranscription of several other genes implicated in tumor growthsuch asmetalloproteinases and integrins (35).Decreased Id1 geneexpression by Hpa2 may therefore affect tumor growth also bymeans other than reduced vascularity. Interestingly,we found thatthe expression of cytokeratin 13 and 15 was reduced markedly intumors produced by Hpa2 clones overexpressing Id1 (Supple-mentary Fig. S6), whereas LOX expression was unchanged (Sup-plementary Fig. S7, bottom). Accordingly, collagen deposition,evident by Masson/Trichrome staining, was not affected by Id1overexpression (Supplementary Fig. S7, third panel), butappeared disrupted in the large necrotic areas that accompaniedaccelerated tumor growth (Supplementary Fig. S7, top and secondpanel), suggesting that collagen deposition preceded tumornecrosis. These results may imply that cytokeratins are also

Figure 6.Overexpressionof Id1 restores tumor growth andvascularity. A, Id1 expression. Real-timePCRanalysis of Id1 expression in tumor xenografts producedby control (Vo)andHpa2-overexpressing cells. Expressionof Id1 is presented graphically in relation to the levels in control clone#3, set arbitrarily to avalue of 1. B, Id1 overexpression.The indicated Hpa2 clone cells were stably transfected with Id1 gene construct (Addgene) or control empty vector (Vo) and cell lysates were subjected toimmunoblotting, applying anti-Id1 (top) and anti-actin (bottom) antibodies. C, tumor growth. Control (Vo) and Id1 overexpressing cells were inoculatedsubcutaneously in SCID mice and tumor growth was inspected over time (top). At termination, tumors were collected, weighed (bottom), and processed forhistologic examination. D, five-micron sections of tumors produced by control (Vo) and Id1-overexpressing clone #60 cells were subjected to immunostaining,applying anti-LYVE-1 (left) and anti-CD31 (right) antibodies. Quantification of LYVE-1- and CD31-positive vessels is shown graphically in the lower panels.Original magnifications, left, �40; right, �20.

Gross-Cohen et al.





regulated, directly or indirectly, by Id1. An inverse correlationbetween Id1 levels and cytokeratin expressionwas also reported inovarian carcinoma cells (38).

Overexpression of Hpa2 affect the resulting tumors in severalaspects other than reduced tumor vascularity. Despite the reducedvascularity, tumor xenografts produced by Hpa2-overexpressingcells were by far less necrotic but rather contained large cystsstructures. These structures, observed in FaDu and Cal27 cellmodels (Figs. 1B and 2C), were negative for endothelial cellmarker (CD31, Supplementary Fig. S4A, right bottom panel),thus nullifying their being abnormal vessels. Cysts are most oftennonmalignant structures but can appear in primary tumors andmetastases (39). The molecular mechanism responsible for cystformation by Hpa2 is obscure but may be related to increasedcellular differentiation. Histologic examination revealed thattumor xenografts produced by Hpa2-overexpressing cells andclones exhibit higher degree of differentiation (Fig. 2C) and thiswas evidently confirmed by a marked increase in cytokeratinslevels (Fig. 2D and Supplementary Fig. S2C). Cytokeratin 15staining was most intense in cells lining, or in close proximitywith the cyst structure (Fig. 2D, bottom), but the cause andrelationship between increased cell differentiation and cyst for-mation is yet to be resolved. Alternatively, increased formation ofcyst structures in Hpa2 tumors may possibly reflect adaptation toreduced lymph vessels density and the formation of edema.

We have also noted that tumors produced by Hpa2 cloneshad far more nerve bundles than control tumors (b3-tubulin;Supplementary Fig. S4A, second panels). While the contribu-tion of neurogenesis to tumorigenesis is questionable (40), apossible role for Hpa2 in neurogenesis appears relevant tourofacial syndrome (UFS). This congenital disorder featuressomatic motor neuropathy, obvious in the face but occasionallymore widespread, and an autonomic neuropathy causing blad-der dysfunction leading to renal failure (41). Notably, biallelicmutation in the HPSE2 gene is held responsible for some casesof UFS in families from different ethnic groups (42–44),resulting in frameshift or nonsense, postulated functionallynull, Hpa2 variants (45). Moreover, mice carrying mutant Hpa2exhibit bladder dysfunction (44, 46), associating with increasedbladder fibrosis (46). The relevance of our findings with Hpa2-overexpressing head and neck carcinoma cells (i.e., increasedcell differentiation, neurogenesis, LOX expression) to UFS andthe lethality phenotype of Hpa2-mutant mice (46) are yet to beresolved.

Augmented collagen deposition was observed in tumor xeno-grafts produced byHpa2 clones (Fig. 4), correlatingwith amarkedincrease in LOX expression (Fig. 4). This enzyme catalyzes thecross-linking of collagen and/or elastin in the ECM, and is highlyimplicated in ECM remodeling and tissue fibrosis (24, 25). Intumorigenesis, LOX was first proposed to act as a tumor suppres-sor, inhibiting the transformation effect of HRAS. Subsequentstudies revealed nonetheless that LOX plays a far more complexrole in cancer. For example, the LOX propeptide (LOX-PP)domainhasbeen shown tobear tumor-suppressor activity,where-as the active enzyme is thought to promote metastasis (27, 47).Tumor suppressor properties have also been attributed to theLOX-like 2 enzyme, LOXL2. For example, a significant decrease inLOXL2 expression was observed in prostate and non–small lungcancers compared with matched normal tissue (27) and lowLOXL2 levels were associated with a more advanced pathologictumor–node–metastasis stage and poorly differentiated lung

adenocarcinomas (48). Not only the expression levels but alsothe cellular localization of LOXL2 appears important clinically.LOXL2 was noted to reside in the nucleus of normal esophagus,and esophageal cancer patients having nuclear LOXL2 survivelonger than patients with cytoplasmic LOXL2 (49).We found thatLOXL2 localizes to the nucleus of Hpa2-overexpressing cancercells (Fig. 4) and cells that comprise the tumormicroenvironment(Supplementary Fig. S3C). This agrees with the notion thatnuclear localization of LOXL2 attenuates tumor growth by tran-scriptional regulation of tumor (50) and stromal cells that mayaffect tumor growth (51). Interestingly, administrationof the LOXinhibitor BAPN to the drinking water of mice inoculated withHpa2 clones did not reverse the strong attenuation of tumorgrowth by Hpa2 (Fig. 5E and Supplementary Fig. S5B). This mayimply that LOX functions extracellularly in amanner that does notinvolve enzymatic activity (52), or that nuclear LOXL2 is respon-sible for tumor attenuation. It should be noted that in contrastwith the above examples, other studies find that LOX and LOXL2promote tumor growth and metastasis, making these enzymes atarget for the development of anticancer therapeutics (24, 27).Clearly, more work is required to delineate the mode by whichHpa2 regulates LOX and LOXL2 expression and cellular locali-zation, the clinical significance of LOX/LOXL2 expression andlocalization in head and neck cancer, and the implication of aHpa2–TGFb–LOX axis (20, 53) on tumor growth, metastasis, andfibrosis. It should be noted that while hypoxic conditions havebeen reported to induce LOX expression (54), increased LOXtranscription was noted in Hpa2 cells grown in vitro undernormoxic conditions (Fig. 4B), suggesting that Hpa2 regulatesLOX gene expression directly.

The mode by which Hpa2 exerts its tumor suppressor proper-ties is not entirely clear. An instrumental tool for deciphering thisquestionwas the finding thatmAb1c7 inhibits cellular binding ofHpa2 (Fig. 5B). Like heparanase, Hpa2 is subjected to rapid andefficient cellular binding (14). This process is due primarily to theinteraction of heparanase/Hpa2 with HSPG on the cell surfacebecause addition of heparin to cell cultures competes with cellmembrane HS, resulting in accumulation of heparanase/Hpa2 inthe cell culture medium (14). Unlike heparanase (18), the hep-arin-binding domains of Hpa2 have not been characterized yet.However, the ability ofmAb 1c7 but no other anti-Hpa2mAbs, toefficiently block cellular binding of Hpa2 (Fig. 5B) stronglyimplies that this mAb targets a functional domain required forthe interaction of Hpa2 with HS. The consequence of the high-affinity interaction of Hpa2 with cell membrane HS has not beensufficiently resolved but likely affects cellular behavior (i.e., celladhesion; ref. 55). Notably, tumor growth inhibition by Hpa2does not appear to involve HS because treatment with mAb Ic7did not affect tumor size (Fig. 5C and Supplementary Fig. S5A).Moreover, heparanase activity appeared practically identical incontrol andHpa2-overexpressing cells (Fig. 5A). These resultsmaysuggest that Hpa2 exerts its tumor suppressor properties in a HS-and heparanase activity-independent manner, possibly involvingas yet unrecognized Hpa2-binding protein.

Taken together, our results support the notion that Hpa2functions as a tumor suppressor in head and neck cancer. More-over, we provide mechanistic insights to this function of Hpa2and describe for the first time that Hpa2 modulates gene tran-scription and thereby affects tumor vascularity, growth and fibro-sis in apparently heparanase activity-, and HS-independent man-ners, thus expanding Hpa2 repertoire and mode of action.






Note Added in ProofsThe current paper is supported by the following paper, pub-

lished online onMarch 9, 2016: Gross-Cohen et al., Heparanase 2expression inversely correlates with bladder carcinoma grade andstage. Oncotarget, PMID: 26968815.

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

Authors' ContributionsConception and design: N. Ilan, I. VlodavskyDevelopment ofmethodology:M.Gross-Cohen, G.Neufeld, G. Arvatz, N. Ilan,I. VlodavskyAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): M. Gross-Cohen, S. Feld, I. Doweck, U. Barash,N. Ilan, I. Vlodavsky

Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): M. Gross-Cohen, I. Doweck, P. Hasson, U. Barash,I. Naroditsky, N. Ilan, I. VlodavskyWriting, review, and/or revision of the manuscript: M. Gross-Cohen,I. Doweck, P. Hasson, N. Ilan, I. VlodavskyAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): P. Hasson

Grant SupportThis study was supported by research grants awarded to I. Vlodavsky by the

Israel Science Foundation (grant no. 601/14); NCI, NIH (grant no. CA106456);the United States-Israel Binational Science Foundation (BSF); the Israel CancerResearch Fund (ICRF); and the Rappaport Family Institute Fund.

The costs of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked advertisementin accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received July 19, 2015; revised January 24, 2016; accepted February 26, 2016;published OnlineFirst March 24, 2016.

References1. Dempsey LA, Brunn GJ, Platt JL. Heparanase, a potential regulator of cell-

matrix interactions. Trends Biochem Sci 2000;25:349–51.2. Parish CR, Freeman C, Hulett MD. Heparanase: a key enzyme involved in

cell invasion. Biochim Biophys Acta 2001;1471:M99–108.3. Vlodavsky I, Beckhove P, Lerner I, Pisano C, Meirovitz A, Ilan N, et al.

Significance of heparanase in cancer and inflammation. Cancer Microen-viron 2012;5:115–32.

4. Vreys V, David G. Mammalian heparanase: what is the message? J CellMol Med 2007;11:427–52.

5. Hammond E, Khurana A, Shridhar V, Dredge K. The role of heparanase andsulfatases in the modification of heparan sulfate proteoglycans within thetumormicroenvironment and opportunities for novel cancer therapeutics.Front Oncol 2014;4:195.

6. Ilan N, Elkin M, Vlodavsky I. Regulation, function and clinical significanceof heparanase in cancer metastasis and angiogenesis. Intl J Biochem CellBiol 2006;38:2018–39.

7. Vlodavsky I, Ilan N, Naggi A, Casu B. Heparanase: structure, biologicalfunctions, and inhibition by heparin-derived mimetics of heparan sulfate.Curr Pharm Des 2007;13:2057–73.

8. Weissmann M, Arvatz G, Horowitz N, Feld S, Naroditsky I, Zhang Y, et al.Heparanase-neutralizing antibodies attenuate lymphoma tumor growthand metastasis. Proc Natl Acad Sci U S A 2016;113:704–9.

9. Hulett MD, Freeman C, Hamdorf BJ, Baker RT, Harris MJ, Parish CR.Cloning of mammalian heparanase, an important enzyme in tumorinvasion and metastasis. Nat Med 1999;5:803–9.

10. Kussie PH, Hulmes JD, Ludwig DL, Patel S, Navarro EC, Seddon AP, et al.Cloning and functional expression of a human heparanase gene. BiochemBiophys Res Commun 1999;261:183–7.

11. Toyoshima M, Nakajima M. Human heparanase. Purification, character-ization, cloning, and expression. J Biol Chem 1999;274:24153–60.

12. Vlodavsky I, FriedmannY, ElkinM,AingornH, AtzmonR, Ishai-Michaeli R,et al. Mammalian heparanase: gene cloning, expression and function intumor progression and metastasis. Nat Med 1999;5:793–802.

13. McKenzie E, Tyson K, Stamps A, Smith P, Turner P, Barry R, et al. Cloningand expression profiling of Hpa2, a novel mammalian heparanase familymember. Biochem Biophys Res Commun 2000;276:1170–7.

14. Levy-Adam F, Feld S, Cohen-Kaplan V, Shteingauz A, Gross M, Arvatz G,et al. Heparanase 2 interacts with heparan sulfate with high affinity andinhibits heparanase activity. J Biol Chem 2010;285:28010–9.

15. Fong S, Debs RJ, Desprez PY. Id genes and proteins as promising targets incancer therapy. Trends Mol Med 2004;10:387–92.

16. Cohen-Kaplan V, Doweck I, Naroditsky I, Vlodavsky I, Ilan N. Hepar-anase augments epidermal growth factor receptor phosphorylation:correlation with head and neck tumor progression. Cancer Res 2008;68:10077–85.

17. Cohen-Kaplan V, Jrbashyan J, Yanir Y, Naroditsky I, Ben-Izhak O, IlanN, et al. Heparanase induces signal transducer and activator of

transcription (STAT) protein phosphorylation: preclinical and clin-ical significance in head and neck cancer. J Biol Chem 2012;287:6668–78.

18. Levy-Adam F, Abboud-Jarrous G, Guerrini M, Beccati D, Vlodavsky I, IlanN. Identification and characterization of heparin/heparan sulfate bindingdomains of the endoglycosidase heparanase. J Biol Chem 2005;280:20457–66.

19. Akiri G, Sabo E, Dafni H, Vadasz Z, Kartvelishvily Y, Gan N, et al. Lysyloxidase-related protein-1 promotes tumor fibrosis and tumor progressionin vivo. Cancer Res 2003;63:1657–66.

20. Kutchuk L, Laitala A, Soueid-Bomgarten S, Shentzer P, Rosendahl AH, EilotS, et al. Muscle composition is regulated by a Lox-TGFb feedback loop.Development 2015;142:983–93.

21. Shteingauz A, Ilan N, Vlodavsky I. Processing of heparanase is mediated bysyndecan-1 cytoplasmic domain and involves syntenin and a-actinin. CellMol Life sci 2014;71:4457–70.

22. Adams RH, Alitalo K. Molecular regulation of angiogenesis and lymphan-giogenesis. Nat Rev Mol Cell Biol 2007;8:464–78.

23. Saharinen P, Tammela T, KarkkainenMJ, Alitalo K. Lymphatic vasculature:development, molecular regulation and role in tumor metastasis andinflammation. Trends Immunol 2004;25:387–95.

24. Barry-HamiltonV, Spangler R,MarshallD,McCauley S, RodriguezHM,OyasuM, et al. Allosteric inhibition of lysyl oxidase-like-2 impedes the developmentof a pathologic microenvironment. Nat Med 2010;16:1009–17.

25. Rodriguez C, Rodriguez-Sinovas A, Martinez-Gonzalez J. Lysyl oxidase as apotential therapeutic target. Drug News Perspect 2008;21:218–24.

26. Doweck I, Kaplan-Cohen V, Naroditsky I, Sabo E, Ilan N, Vlodavsky I.Heparanase localization and expression by head and neck cancer: corre-lation with tumor progression and patient survival. Neoplasia 2006;8:1055–61.

27. Barker HE, Cox TR, Erler JT. The rationale for targeting the LOX family incancer. Nat Rev Cancer 2012;12:540–52.

28. Lyden D, Young AZ, Zagzag D, Yan W, Gerald W, O'Reilly R, et al. Id1 andId3 are required for neurogenesis, angiogenesis and vascularization oftumour xenografts. Nature 1999;401:670–7.

29. Barash U, Zohar Y, Wildbaum G, Beider K, Nagler A, Karin N, et al.Heparanase enhances myeloma progression via CXCL10 downregulation.Leukemia 2014;28:2178–87.

30. Boyango I, Barash U, Naroditsky I, Li JP, Hammond E, Ilan N, et al.Heparanase cooperates with Ras to drive breast and skin tumorigenesis.Cancer Res 2014;74:4504–14.

31. Shteingauz A, Boyango I, Naroditsky I, Hammond E, Gruber M, Doweck I,et al. Heparanase enhances tumor growth and chemoresistance by pro-moting autophagy. Cancer Res 2015;75:3946–57.

32. Zhang X, Xu S, Tan Q, Liu L. High expression of heparanase-2 is anindependent prognostic parameter for favorable survival in gastric cancerpatients. Cancer Epidemiol 2013;37:1010–3.


Gross-Cohen et al.




33. YuM, Liu L, Liang C, Li P, Ma X, Zhang Q, et al. Intratumoral vessel densityas prognostic factors in head and neck squamous cell carcinoma: a meta-analysis of literature. Head Neck 2014;36:596–602.

34. Alitalo K, Carmeliet P. Molecular mechanisms of lymphangiogenesis inhealth and disease. Cancer Cell 2002;1:219–27.

35. Benezra R, Rafii S, Lyden D. The Id proteins and angiogenesis. Oncogene2001;20:8334–41.

36. Dong Z, Wei F, Zhou C, Sumida T, Hamakawa H, Hu Y, et al.Silencing Id-1 inhibits lymphangiogenesis through down-regulationof VEGF-C in oral squamous cell carcinoma. Oral Oncol 2011;47:27–32.

37. Sikder HA, Devlin MK, Dunlap S, Ryu B, Alani RM. Id proteins in cellgrowth and tumorigenesis. Cancer Cell 2003;3:525–30.

38. Strait KA, Dabbas B, Hammond EH, Warnick CT, Iistrup SJ, FordCD. Cell cycle blockade and differentiation of ovarian cancer cellsby the histone deacetylase inhibitor trichostatin A are associatedwith changes in p21, Rb, and Id proteins. Mol Cancer Ther 2002;1:1181–90.

39. Mokhtari S. Mechanisms of cyst formation in metastatic lymph nodes ofhead and neck squamous cell carcinoma. Diagn Pathol 2012;7:6.

40. Zhao CM, Hayakawa Y, Kodama Y, Muthupalani S, Westphalen CB,Andersen GT, et al. Denervation suppresses gastric tumorigenesis. SciTransl Med 2014;6:250ra115.

41. Woolf AS, Stuart HM, Roberts NA,McKenzie EA, Hilton EN, NewmanWG.Urofacial syndrome: a genetic and congenital disease of aberrant urinarybladder innervation. Pediatr Nephrol 2014;29:513–8.

42. Daly SB, Urquhart JE, Hilton E,McKenzie EA, Kammerer RA, LewisM, et al.Mutations in HPSE2 cause urofacial syndrome. Am J HumGenet 2010;86:963–9.

43. Pang J, Zhang S, Yang P, Hawkins-Lee B, Zhong J, Zhang Y, et al. Loss-of-function mutations in HPSE2 cause the autosomal recessive urofacialsyndrome. Am J Hum Genet 2010;86:957–62.

44. Stuart HM, Roberts NA, Hilton EN, McKenzie EA, Daly SB, Hadfield KD,et al. Urinary tract effects of HPSE2 mutations. J Am Soc Nephrol 2015;26:797–804.

45. ArvatzG, Shafat I, Levy-AdamF, IlanN, Vlodavsky I. Theheparanase systemand tumor metastasis: is heparanase the seed and soil?Cancer MetastasisRev 2011;30:253–68.

46. GuoC, Kaneko S, SunY,HuangY, Vlodavsky I, Li X, et al. Amousemodel ofurofacial syndrome with dysfunctional urination. Hum Mol Genet 2015;24:1991–9.

47. Bais MV, Ozdener GB, Sonenshein GE, Trackman PC. Effects of tumor-suppressor lysyl oxidase propeptide on prostate cancer xenograft growthand its direct interactions with DNA repair pathways. Oncogene 2015;34:1928–37.

48. Zhan P, Shen XK, Qian Q, Zhu JP, Zhang Y, Xie HY, et al. Down-regulationof lysyl oxidase-like 2 (LOXL2) is associated with disease progression inlung adenocarcinomas. Med Oncol 2012;29:648–55.

49. Li TY, Xu LY, Wu ZY, Liao LD, Shen JH, Xu XE, et al. Reduced nuclear andectopic cytoplasmic expression of lysyl oxidase-like 2 is associated withlymph node metastasis and poor prognosis in esophageal squamous cellcarcinoma. Hum Pathol 2012;43:1068–76.

50. Iturbide A, Garcia de Herreros A, Peiro S. A new role for LOX and LOXL2proteins in transcription regulation. FEBS J 2015;282:1768–73.

51. Scherz-Shouval R, Santagata S, Mendillo ML, Sholl LM, Ben-Aharon I, BeckAH, et al. The reprogramming of tumor stroma byHSF1 is a potent enablerof malignancy. Cell 2014;158:564–78.

52. Lugassy J, Zaffryar-Eilot S, Soueid S, Mordoviz A, Smith V, Kessler O, et al.The enzymatic activity of lysyl oxidas-like-2 (LOXL2) is not required forLOXL2-induced inhibition of keratinocyte differentiation. J Biol Chem2012;287:3541–9.

53. Roberts NA, Woolf AS, Stuart HM, Thuret R, McKenzie EA, NewmanWG, et al. Heparanase 2, mutated in urofacial syndrome, mediatesperipheral neural development in Xenopus. Hum Mol Genet 2014;23:4302–14.

54. Cox TR, Rumney RM, Schoof EM, Perryman L, Hoye AM, Agrawal A, et al.The hypoxic cancer secretome induces pre-metastatic bone lesions throughlysyl oxidase. Nature 2015;522:106–10.

55. Levy-Adam F, Ilan N, Vlodavsky I. Tumorigenic and adhesive properties ofheparanase. Semin Cancer Biol 2010;20:153–60.






Published OnlineFirst March 24, 2016.Cancer Res Miriam Gross-Cohen, Sari Feld, Ilana Doweck, et al. and GrowthHeparanase 2 Attenuates Head and Neck Tumor Vascularity

Updated version

10.1158/0008-5472.CAN-15-1975doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2016/03/24/0008-5472.CAN-15-1975.DC1

Access the most recent supplemental material at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://cancerres.aacrjournals.org/content/early/2016/04/20/0008-5472.CAN-15-1975To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-15-1975

http://cancerres.aacrjournals.org/content/suppl/2016/03/24/0008-5472.CAN-15-1975.DC1

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2016/04/20/0008-5472.CAN-15-1975


Heparanase 2 Attenuates Head and Neck Tumor Vascularity...

Documents

Transcript of Heparanase 2 Attenuates Head and Neck Tumor Vascularity...