REVIEW

Osteosarcoma metastasis: prospective role of ezrin

Yuanjin Zhang & Ling Zhang & Guofu Zhang & Songbai Li & Jun Duan & Jie Cheng &

Guozhen Ding & Chibing Zhou & Jie Zhang & Pengcheng Luo & Dongbing Cai &Lianghong Kuang & Yichun Zhou & Liqun Tong & Xiangdong Yu & Lifang Zhang &

Lijun Xu & Li Yu & Xiaomei Shi & Aihong Ke

Received: 1 November 2013 /Accepted: 25 February 2014 /Published online: 9 March 2014# International Society of Oncology and BioMarkers (ISOBM) 2014

Abstract Osteosarcoma is the most commonly diagnosedprimary malignant bone tumor, with similar global incidencerate across childhood and adolescence. Patients with localizeddisease have a 5-year survival period of 80 %; however, theprognosis is poor in those with metastatic osteosarcoma. Theorigin of the primary tumor is most frequently themetaphyseal (actively growing) regions of the distal femur,proximal tibia, and proximal humerus, although the tumor candevelop in any bone, and the most likely sites for metastasisare the lungs and bone. Ezrin is a member of the ezrin-radixin-moesin (ERM) family of proteins that functions as a cross-linker between the actin cytoskeleton and the plasma mem-brane, and ezrin also plays a positive role in maintaining cellshape and polarity and facilitates membrane-trafficking path-ways, cell migration, cell signaling, growth regulation, anddifferentiation. There is strong evidence to suggest that ezrin isnecessary for osteosarcoma metastasis. The objective of thecurrent review is to summarize the know-how about metasta-tic progression in osteosarcoma, with a focus on ezrin. Despitethe promise that preliminary studies on ezrin have shown,there is a great need to further analyze the role of ezrin inosteosarcoma metastasis and to determine its usefulness as abiomarker for the disease.

Keywords Ezrin . Osteosarcoma . Pathophysiology ofosteosarcoma .Metastatic progression . Bone sarcoma

Introduction

Osteosarcoma is the most commonly diagnosed form of pri-mary malignant bone tumor [1]. Osteosarcoma is the thirdmost commonmalignancy in children and adolescents and thefifth most commonmalignancy in young adults (15–19 years),with the majority of osteosarcoma cases occurring betweenthe ages of 10–19 years and with a predilection for males overfemales [2]. Although the tumor can form in any bone, theorigin is most frequently in the metaphyseal regions of thedistal femur, proximal tibia, and proximal humerus [3–5]. It iscommonly believed that osteosarcoma is a tumor ofosteoprogenitor cells, which are multipotential, hormone-responsive stromal cells in the periosteum and marrow, whichdifferentiate into various lineages depending on the environ-mental signals [6]. This also helps explain why osteosarcomasdevelop more often in adolescents in the age range whichincludes puberty and why osteosarcomas are often found inthose regions where bone is actively growing.

Osteosarcoma pathology and metastasis

Until 1970, the only treatment option for a high-grade osteo-sarcoma was amputation, with about 80 % of these patientsdying as a result of metastatic disease [3]. With the advent anddevelopment of chemotherapy protocols, both induction(preoperative) and adjuvant (postoperative) therapies, in addi-tion to surgical resection of the primary tumor, the long-termsurvival rate has increased to 60–80 % [7–9]. In addition tothis marked increase in survivability, limb salvage techniqueshave been developed that help maintain function while notaffecting the overall survival rate. Such techniques are possi-ble in 90–95 % of cases [8, 10, 11]. Overall, those patientswho present with primary disease only have a much betterprognosis than those who present with metastatic disease.

Y. Zhang (*) : L. Zhang :G. Zhang : S. Li : J. Duan : J. Cheng :G. Ding :C. ZhouDepartment of Orthopedics, Huangshi Central Hospital, AffiliatedHospital of Hubei Polytechnic University, No. 141 Tianjin Road,Huangshi City 435000, Hubei Province, Chinae-mail: yjz_hsch@163.com

J. Zhang : P. Luo :D. Cai : L. Kuang :Y. Zhou : L. Tong :X. Yu :L. Zhang : L. Xu : L. Yu :X. Shi :A. KeHubei Key Laboratory of Kidney Disease Pathogenesis andIntervention, Huangshi City, Hubei Province, China

Tumor Biol. (2014) 35:5055–5059DOI 10.1007/s13277-014-1799-y

The osteosarcoma primary lesion almost always originatesin the growing region of the long bones and grows radially,forming a ball-like mass. When the lesion penetrates the bonycortex, it compresses the surrounding muscles into apseudocapsular layer referred to as the reactive zone. Insidethe reactive zone, there may also be microextensions of theprimarymass, and these are referred to as satellites [5]. If somecells from the primary lesion are able to break loose andsurvive and effectively initiate the growth of secondary le-sions, then the tumor is said to metastasize. The tumor maymetastasize regionally (within the same extremity) or system-ically (to other organs). Tumor nodules growing outside thereactive rim but within the same bone (intraosseous metasta-ses) or across a neighboring joint (transarticular metastases)are termed skip lesions [5, 12]. Those metastases that occursystemically almost always arise in the lungs, while distantbone metastases are possible, but usually, only after pulmo-nary metastasis had occurred [13–15].

In all osteosarcoma cases, patients undergo plain-film ra-diography, CT scanning, and physical examination in order todetect the presence of recurrence/metastasis, especially pul-monary metastases. After surgical resection of nonmetastaticprimary disease, X-rays and CT scans are repeated every 3 to6 months for 2 years, the point by which the majority ofsystemic metastases appear posttreatment termination, andthen every 6 months for 3 years and then yearly indefinitely[5]. Those patients who present with metastatic pulmonarydisease have a much poorer prognosis; however, in a smallnumber of cases where patients are responsive to chemother-apy and a pulmonary metastasectomy is performed, a cure canbe achieved [14–17].

Ezrin in osteosarcoma metastasis

What is ezrin?

One gene identified as differentially expressed between highlymetastatic and nonhighly metastatic osteosarcoma cell lines,using complementary DNAmicroarray, is ezrin [18], which isencoded by the VILLIN2 gene. Ezrin is a member of the ezrin-radixin-moesin (ERM) family of proteins that functions as across-linker between the actin cytoskeleton and the plasmamembrane [19], thus allowing the cell to interact directly withits microenvironment [20]. Ezrin also plays a positive role inmaintaining cell shape and polarity and facilitates membrane-trafficking pathways, cell migration, cell signaling, growthregulation, and differentiation [21].

The ERM family of proteins shares a highly homologousN-terminal domain, called FERM. Through this N-terminaldomain, ERM proteins can bind directly to the cytoplasmictail of specific proteins with single transmembrane domains,including CD43, CD44, CD95 (Fas), ICAM-1, ICAM-2,

ICAM-3, and syndecan-2 [19]. ERM proteins bind to actinprimarily via a 34-amino acid region at the C-terminal end[19]. Until activated by a poorly defined mechanism, thoughtto involve conformational changes triggered by interactionswith phospholipids and kinases that phosphorylate threonineand tyrosine residues, ezrin remains in a dormant, inactiveconformation in the cytoplasm [20]. The active sites in theprotein are hidden by intramolecular interactions between theN-terminal (FERM) and C-terminal domains, so ezrin canneither bind to actin nor transmembrane proteins [19]. Uponactivation, ezrin molecules are directed to the cell membranewhere they can immediately form a bridge between F-actinand a number of transmembrane proteins [20]. One of themost likely modes of ezrin activation is through interactionwith the small GTPase Rho, which controls actin cytoskeletonremodeling and related cellular activities. Interestinglyenough, it appears that ezrin, once activated by Rho, plays arole in regulating Rho [19, 22, 23].

Studies on role of ezrin in osteosarcoma metastasis

Based on the findings of microarray studies showing ezrin tobe differentially expressed in highly metastatic cell lines, ezrinhas been selected for further study by a number of groups(Table 1) trying to work out the role ezrin plays in osteosar-coma metastasis [20, 24]. Evidence suggests that ezrin playsan integral role in osteosarcoma metastasis and that ezrinmight be necessary for metastasis to even occur [20]. Theexact role that ezrin plays is still under investigation; however,there are some preliminary studies that have gained someinsight as to what this role might be (Fig. 1).

There is strong evidence to suggest that ezrin is necessaryfor osteosarcoma metastasis. Khanna et al. were able to sup-press the ezrin protein by full-length antisense ezrin (ezrin-AS), and this significantly diminished metastases in micewhile not affecting in vitro cell culture or the developmentof primary tumors in vivo [20]. They were also able to mirrorthese effects by creating a dominant-negative inhibitor ofezrin, by changing the amino acid 567 from threonine toalanine, which inhibited ezrin from becoming phosphorylatedand activated. Ezrin-T567A expression eliminated the devel-opment of metastases, while metastases developed in 69 % ofthe control mice [20]. These same affects were noticed in ezrinsuppression studies in rhabdomyosarcoma, as suppression ofezrin induced a tenfold reduction in pulmonarymetastases andaddition of ezrin to poorly metastatic cell lines led to a tenfoldincrease in pulmonary metastases [24]. These correlationssuggest that ezrin might play a similar role in multiple sarco-mas. This group also looked at the timing effects of ezrinsuppression on metastasis and found that ezrin significantlyreduced the ability of previously metastatic cells to survive atthe early stages of metastasis to the lungs [20].

5056 Tumor Biol. (2014) 35:5055–5059

Other studies have demonstrated the importance of Akt andMAPK, both signal pathways mediated by ezrin, in associa-tion with ezrin and early metastasis [25, 26], so the Khannagroup investigated further the relationship between ezrin andthese kinase pathways. They found that ezrin suppression ledto decreased phosphorylation and kinase activity of both Aktand p44/42MAPK; however, in their studies, only the MAPKpathway appeared to impact metastasis associated with ezrin,as activation of the MAPK pathway under ezrin suppressionconditions resulted in the recovery of metastatic potential,while activation of the Akt did not recover metastatic potential[20]. Other groups, however, have shown that Akt actuallydoes play a key role in osteosarcoma metastasis. One group

showed that Akt signaling plays an important role through itsassociation to matrix metalloproteinases (MMPs). Inhibitionof the PI3K-Akt signaling suppressed the secretion of MMPand led to a drastic decrease in metastases [27]. The fact thatthese reports disagree provides evidence for the need to furtherstudy bothMAPK and Akt in osteosarcomametastasis as wellas their association to ezrin in the formation of metastases.

In addition to the MAPK and Akt pathways, there are anumber of other pathways with which ezrin might be involvedthat help promote the formation of metastases. The first ofthese is the hyaluronic acid receptor, CD44. It is known thatezrin can bind to the cytoplasmic tail of CD44, and it ispossible that this interaction can promote metastasis. CD44has been shown to be prometastatic in some cancer types andhas been shown to correlate with increased ezrin activationand invasion in osteosarcoma [28, 29]. Additionally, ezrin hasbeen implicated in metastatic processes through its interac-tions with the hepatocyte growth factor receptor (HGFR),which is the product of the Met gene, and has been shownto be involved in the progression of osteosarcoma in bothhumans and dogs [30, 31]. Both ezrin and HGFR have beenshown to be involved in the MEK/ERK pathway, which alsoinvolves CD44 [32]. These associations with CD44 andHGFR represent just one of the many complex metastaticpathways for which ezrin is a key regulator.

As a part of its participation in the formation of cell-surfacecomplexes that mediate cell-cell and cell-extracellular matrixattachments, ezrin interacts with E-cadherin and integrins [33,34]. E-cadherin is important because it is a negative regulatorof CD44 activity, and thus, it can suppress CD44-mediatedinvasion and branching morphogenesis [35]. The loss of E-cadherin function has been shown to result in the increase ofmetastatic dissemination, and it has been suggested that ezrinoverexpression might functionally substitute for E-cadherin

Table 1 Studies looking at the role of ezrin in metastasis of osteosarcoma

Author Year Model Major finding

Khanna et al. [18] 2001 More aggressive (K7M2) and less aggressive(K12) murine osteosarcoma models

First to report the involvement of ezrin in metastasisof osteosarcoma; suggested involvement of MAPK pathway

Yu et al. [24] 2004 Rhabdomyosarcoma cell lines Ezrin was a critical regulator of metastasis in rhabdomyosarcoma

Fukaya et al. [27] 2005 LM8 cell line PI3K-Akt pathway is involved in osteosarcoma metastasis

Peng et al. [29] 2003 Human tissue specimens CD44s, MMP-9, and Ki-67 are involved with growth andlocal invasion in osteosarcomaMartin et al. [28] 2003

Park et al. [38] 2006 Human tissue specimens Elevated levels of ezrin only in high-grade osteosarcomas whileno expression in low-grade osteosarcomas; however, ezrinexpression increases during progression of tumors

Weng et al. [39] 2005 Human tissue specimens Ezrin has a causative role in soft tissue dissemination and can beused as an additional prognostic marker; similar studies arewarranted for osteosarcoma

Wan et al. [40] 2005 K7M2 murine model of osteosarcoma Blocking the mTOR/S6K1/4E-BP1 pathway downregulatesezrin and may be an appropriate target for strategies to reduceosteosarcoma metastasis

Fig. 1 Potential roles of ezrin as a signaling intermediate during meta-static progression. HGF hepatocyte growth factor

Tumor Biol. (2014) 35:5055–5059 5057

loss [24, 33]. Past research has indicated that integrins play acomplex role in metastasis, as some integrin complexes havebeen shown to be prometastatic, while others seem to inhibitmetastasis [36]. It is known that ezrin can disrupt integrinsignaling, and this might play a role in tumor cell survivalduring the metastatic process; however, little is understoodabout this currently [37].

Finally, it is possible that there is a link between ezrin andmerlin. Merlin is the product of the NF2 tumor suppressorgene, and loss of this gene leads to a variety of tumors inhumans, and mice bearing a mutant NF2 gene are very proneto highly metastatic osteosarcoma [22, 23]. Although there islittle evidence to suggest a direct functional relationship be-tween ezrin and merlin, the possibility remains because thetwo share overlapping subcellular localization and commoninteracting partners and have physical interaction [22]. If arelationship is established, it is likely that ezrin expression issomehow (causally or responsively) related to the absence ofmerlin. Research suggests that ezrin expression is elevated inhigh-grade osteosarcomas when compared to low-grade oste-osarcomas. One group found that 43.75 % (14/32) of high-grade osteosarcomas expressed elevated levels of ezrin, whileno (0/21) low-grade osteosarcomas expressed ezrin, and thatezrin expression increases during progression of tumors [38].Another group linked high ezrin levels in primary tumors to asignificantly shorter median disease-free interval in dogswhen compared to dogs with osteosarcomas with low levelsof ezrin [20]. This group also found a similar link in samplesfrom pediatric patients with localized appendicular osteosar-coma and also found that the risk of metastatic relapse was80 % greater for those patients with elevated levels of ezrin[20].

Furthermore, another group found that patients with ezrin-positive tumors were far more likely to develop distant me-tastases, while the presence of ezrin does not seem to affect thedevelopment of local recurrence [39]. This research suggeststhat ezrin might be a useful prognostic tool as a biomarker forosteosarcoma progression [38], and specifically, ezrin immu-noreactivity might be a viable tool [39]; however, it is not sureif high levels of ezrin directly contribute to metastasis or if it isan indication of cell response to external stimuli [20]. Basedon available research of ezrin and its role in osteosarcomametastasis, Wan et al. analyzed the efficacy of rapamycin ininhibiting the metastatic behavior of ezrin in osteosarcoma.They found that rapamycin, and its analogue, cell cycleinhibitor-779 (CCI-779), inhibits ezrin-mediated metastaticbehavior, and they were able to elucidate portions of themetastatic mechanism of ezrin [40]. Rapamycin has antican-cer activity because it targets the mTOR/S6K1/4E-BP1 path-way, which plays a fundamental role in ribosome biogenesisand cap-dependent translation, and has been linked to reducetranslational production and signaling of vascular endothelialgrowth factor [41–43]. S6K1 and 4E-BP1 are downstream

targets of mTOR, and these were shown to play critical rolesin ezrin-mediated metastatic signaling, as well as to be linked,downstream, to Akt [40]. Ezrin downregulation by siRNA andtransfection with dominant-negative (dn) mutant ezrin led todecreased phosphorylation and activity of both S6K1 and4EBP1 and subsequent proteasomal degradation of the twomolecules and also led to inhibition of metastasis [40].Treatment with rapamycin had similar effects on S6K1 and4E-BP1 phosphorylation as did downregulation of ezrin, soezrin-linked S6K1 and 4E-BP1 phosphorylation is rapamycinsensitive [40]. Because of these findings, rapamycin wasstudied in the osteosarcoma murine model, and it was foundthat rapamycin and CCI-779 treatment prolonged survival andreduced the size and number of pulmonary metastases [40]. Itis thought that ezrin regulates S6K1 and 4E-BP1 throughactivation of the mTOR/S6K1/4E-BP1 pathway indirectlythrough direct activation of the phosphatidylinositol 3-kinase/Akt pathway [40].

Concluding comments

There appears to be overwhelming evidence that implicatesezrin as a key molecule in the progression of osteosarcomadisease, as there is great potential that ezrin coordinates and/oramplifies a number of metastasis-associated cell surface sig-nals and can alter the balance of many intracellular signalingcascades that help promote metastasis [37]. Furthermore,studies with rapamycin or its analogue have been very en-couraging, and further multicenter studies are required tospecifically address the role of these inhibitors in targetinglow- and high-grade osteosarcoma. More detailed study willalso have a clinical significance in determining if ezrin can beused as a prognostic marker, especially in those with metasta-tic osteosarcoma. Because of this and the fact that osteosar-coma metastasis remains a critical issue, ezrin is definitelyworthy of further study. It is imperative to confirm publishedresults on ezrin levels and metastatic potential and to furtherthese studies by further elucidating the efficacy of using ezrinas a marker for osteosarcoma.

Acknowledgments None.

Conflicts of interest None.

References

1. Damron TA,WardWG, Stewart A. Osteosarcoma, chrondrosarcoma,and Ewing’s sarcoma. National Cancer Data Base Report. ClinOrthop Relat Res. 2007;459:40–7.

2. Gurney JG, Swensen AR, Bulterys M. Malignant bone tumors. In:Ries LA, Smith MA, Gurney JG, et al., editors. Cancer incidence andsurvival among children and adolescents: United States SEER

5058 Tumor Biol. (2014) 35:5055–5059

Program 1975–1995 (Pub #99-4649). Bethesda, MD: SEER pro-gram, National Cancer Institute; 1999. p. 99–110.

3. Marcove RC, Mike V, Hajack JV, Levin AG, Hutter RV. Osteogenicsarcoma under the age of twenty-one. A review of one hundred andforty-five operative cases. J Bone Joint Surg. 1970;52:411–23.

4. McKenna RJ, SchwinnCP, Soonh KY, HiginbothamN. Sarcomata ofosteogenic series (osteosarcoma, fibrosarcoma, chondrosarcoma,parosteal osteosarcoma, and sarcomata arising in abnormal bone):an analysis of 552 cases. J Bone Joint Surg. 1966;48-A:1–26.

5. Wittig JC, Bickels J, Priebat D, Jelinek J, Kellar-Graney K, ShmooklerB, et al. Osteosarcoma: a multidisciplinary approach to diagnosis andtreatment. Am Fam Physician. 2002;65(1123–1132):1135–6.

6. Bianco P, Riminucci M, Gronthos S, Robey P. Bone marrow stromalcells: nature, biology and potential applications. StemCells. 2001;19:180–92.

7. Bickels J, Jelinek BM, Shmookler BM, Neff RS, Malawar MM.Biopsy of musculoskeletal tumors. Current concepts. Clin OrthopRelat Res. 1999;368:212–9.

8. Huth JF, Eilber FR. Patterns of recurrence after resection of osteo-sarcoma of the extremity. Strategies for treatment of metastases. ArchSurg. 1989;124:122–6.

9. Meyers PA, Heller G, Healey JH, Huvos A, Applewhite A, Sun M,et al. Osteogenic sarcoma with clinically detectable metastasis atinitial presentation. J Clin Oncol. 1993;11:449–53.

10. Rubert CK, Malawer MM, Kellar KL. Modular endoprosthetic re-placement of the proximal humerus: indications, surgical technique,and results. Semin Arthroplast. 1999;10:142–53.

11. Malawer MM, Chou LB. Prosthetic survival and clinical results withuse of large segment replacements in the treatment of high-gradebone sarcomas. J Bone Joint Surg. 1995;77:1154–65.

12. Bacci G, Picci P, Briccoli A, Avella M, Ferrari S, Femino FP, et al.Osteosarcoma of the extremity metastatic at presentation: resultsachieved in 26 patients treated with combined therapy (primarychemotherapy followed by simultaneous resection of the primaryand metastatic lesions). Tumori. 1992;78:200–6.

13. Bohndorf K, Reiser M, Lochner B, Feaux de Lacroix W, SteinbrichW. Magnetic resonance imaging of primary tumours and tumour-likelesions of bone. Skelet Radiol. 1986;15:511–7.

14. Sundaram M, McGuire MH, Herbold DR. Magnetic resonance im-aging of osteosarcoma. Skelet Radiol. 1987;16:23–9.

15. Belli L, Scholl S, Livartowski A, Ashby M, Palangie T, Levasseur P,et al. Resection of pulmonary metastases in osteosarcoma. A retro-spective analysis of 44 patients. Cancer. 1989;63:2546–50.

16. Enneking W, Spanier S, Goodman M. A system for the surgicalstaging of musculoskeletal sarcoma. Clin Orthop. 1980;153:106–20.

17. Mialou V, Philip T, Kalifa C, Perol D, Gentet JC, Marec-Berard P,et al. Metastatic osteosarcoma at diagnosis: prognostic factors andlong-term outcome—the French pediatric experience. Cancer.2005;104:1100–9.

18. Khanna C, Khan J, Nguyen P, et al. Metastasis-associated differencesin gene expression in a murine model of osteosarcoma. Cancer Res.2001;61:3750–9.

19. Bretscher A, Edwards K, Fehon RG. ERM proteins and merlin:integrators at the cell cortex. Nat Rev Mol Cell Biol. 2002;3:586–99.

20. Khanna C, Wan X, Bose S, et al. The membrane-cytoskeleton linkerezrin is necessary for osteosarcoma metastasis. Nat Med. 2004;10:182–6.

21. Tsukita S, Yonemura S. ERM family: from cytoskeleton to signaltransduction. Curr Opin Cell Biol. 1997;9:70–5.

22. Curto M, McClatchey AI. Ezrin a metastatic determinant? CancerCell. 2004;5:113–4.

23. McClatchey AI. Merlin and ERM proteins: unappreciated roles incancer development? Nat Rev Cancer. 2003;3:877–82.

24. Yu Y, Khan J, Khanna C, et al. Expression profiling identifies thecytoskeletal organizer ezrin and the developmental homeoproteinSix-1 as key metastatic regulators. Nat Med. 2004;10:175–81.

25. Gautreau A, Poullet P, Louvard D, Arpin M. Ezrin, a plasma mem-brane microfilament linker, signals cell survival through the phos-phatidylinositol 3-kinase/Akt pathway. Proc Natl Acad Sci U S A.1999;96:7300–5.

26. WardY,WangW,Woodhouse E, Linnoila I, Liotta L, Kelly K. Signalpathways which promote invasion andmetastasis: critical and distinctcontributions of extracellular signal-regulated kinase and Ral-specificguanine exchange factor pathways. Mol Cell Biol. 2001;21:5958–69.

27. FukayaY, Ishiguro N, Senga T, et al. A role for PI3K-Akt signaling inpulmonary metastatic nodule formation of the osteosarcoma cell line,LM8. Oncol Rep. 2005;14:847–52.

28. Martin TA, Harrison G, Mansel RE, Jiang WG. The role of theCD44/ezrin complex in cancer metastasis. Crit Rev Oncol Hematol.2003;46:165–86.

29. Peng TS, Qiu JS, Wu HX, Liang HZ, Luo CQ. Expressions ofCD44s, MMP-9, and Ki-67: possible association with invasion,metastasis, and recurrence of osteosarcoma. Ai Zheng. 2002;21:745–50.

30. Crepaldi T, Gautreau A, Comoglio PM, Louvard D,ArpinM. Ezrin isan effector of hepatocyte growth factor-mediated migration and mor-phogenesis in epithelial cells. J Cell Biol. 1997;138:423–34.

31. Ma PC, Maulik G, Christensen J, Salgia R. c-Met: structure, func-tions and potential for therapeutic inhibition. Cancer Metastasis Rev.2003;22:309–25.

32. Orian-Rousseau V, Chen L, Sleeman JP, Herrlich P, Ponta H. CD44 isrequired for two consecutive steps in HGF/c-Met signaling. GenesDev. 2002;16:3074–86.

33. Pujuguet P, Del Maestro L, Gautreau A, Louvard D, Arpin M. Ezrinregulates E-cadherin-dependent adherens junction assembly throughRac1 activation. Mol Biol Cell. 2003;14:2181–91.

34. Tsukita S, Oishi K, Sato N, Sagara J, Kawai A, Tsukita S. ERMfamily members as molecular linkers between the cell surface glyco-protein CD44 and actin-based cytoskeletons. J Cell Biol. 1994;126:391–401.

35. Xu Y, Yu Q. E-cadherin negatively regulates CD44-hyaluronan in-teraction and CD44-mediated tumor invasion and branching morpho-genesis. J Biol Chem. 2003;278:8661–8.

36. Vo HP, Lee MK, Crowe DL. Α2β1 integrin signaling via the mitogenactivated protein kinase pathway modulates retinoic acid-dependenttumor cell invasion and transcriptional downregulation of matrix me-talloproteinase 9 activity. Int J Oncol. 1998;13:1127–34.

37. Hunter KW. Ezrin, a key component in tumor metastasis. Trends MolMed. 2004;10:201–4.

38. Park HR, JungWW, Bacchini P, Bertoni F, Kim YW, Park YK. Ezrinin osteosarcoma: comparison between conventional high-grade andcentral low-grade osteosarcoma. Pathol Res Pract. 2006;202:509–15.

39. Weng WH, Ahlén J, Aström K, Lui WO, Larsson C. Prognosticimpact of immunohistochemical expression of ezrin in highly malig-nant soft tissue sarcomas. Clin Cancer Res. 2005;11:6198–204.

40. Wan X, Mendoza A, Khanna C, Helman L. Rapamycin inhibitsezrin-mediated metastatic behavior in a murine model of osteosarco-ma. Cancer Res. 2005;65:2406–11.

41. Brown EJ, Beal PA, Keith CT, Chen J, Shin TB, Schreiber SL.Control of p70 s6 kinase by kinase activity of FRAP in vivo.Nature. 1995;377:441–6.

42. Brown EJ, Schreiber SL. A signaling pathway to translational con-trol. Cell. 1996;86:517–20.

43. Guba M, von Breitenbuch P, Steinbauer M, et al. Rapamycin inhibitsprimary and metastatic tumor growth by antiangiogenesis: involve-ment of vascular endothelial growth factor. NatMed. 2002;8:128–35.

Tumor Biol. (2014) 35:5055–5059 5059