TargetingMatrilysinandItsImpactonTumorGrowth In vivo ...possibility that can be explored further by...

TargetingMatrilysin and Its Impact onTumor Growth In vivo:The Potential Implications in Breast CancerTherapyWenG. Jiang,1Gaynor Davies,1TraceyA. Martin,1Christian Parr,1Gareth Watkins,1

Malcolm D.Mason,1KefahMokbel,2 and Robert E. Mansel1

Abstract Introduction:Matrilysin (MMP-7) is a metalloproteinase that is involved in the degradation ofextracellular matrix, invasion, and tumor progression. The current study examined if targetingmatrilysin using retroviral ribozyme transgenes may have an impact on breast cancer cells andmay have clinical implications.Experimental Design: Retroviral hammerhead ribozyme transgenes were designed to specifi-cally target human matrilysin mRNA. The breast cancer cell MDA-MB-231was transfected witheither a retroviral matrilysin transgene or a control retroviral transgene. Stably transfected cellswere tested for their invasiveness and migratory properties in vitro. The cells were also used increatinga tumormodel inathymicnudemice inwhich the growthof tumors and levels ofmatrilysinwere assessed. In addition, levels of both protein and mRNA of matrilysin were investigated in acohort of humanbreast tumors.Results: Expression of matrilysin in MDA-MB-231was successfully eliminated by the retroviralhammerhead ribozyme transgene for matrilysin as revealed by reverse transcription-PCR. Matri-lysin transgene ^ transduced cancer cells (MDA-MB-231DMatrilysin) exhibited a significantlylower degree of invasion (number of invading cells 16.0 F 2.5) compared with wild type(MDA-MB-231WT; 26.2 F 6.2, P < 0.05) or control transgene-transduced cancer cells (MDA-MB-231pRevTRE; 25.3 F 4.2, P < 0.01). However, the rate of growth of the cells in vitro was notsignificantly affected. In the in vivo tumor model, MDA-MB-231DMatrilysin tumors, which hadvery low levels of immunoreactive matrilysin, grew at a significantly lower rate (0.24 F 0.03cm3, 4 weeks after inoculation) compared with the wild-type MDA-MB-231WT (1.46 F 0.04cm3) and MDA-MB-231pRevTRE (1.12 F 1.0 cm3) tumors. In human breast tumors, breast cancercells stained matrilysin at a significantly higher density, compared with normal mammary epithe-lium. The highest level of matrilysin was seen in high-grade tumors and that from patients withmoderate and poor prognosis. Finally, high levels of matrilysin were significantly linked with apoor long-term survival (P = 0.0143).Conclusion:Matrilysin, which is aberrantly expressed inhumanbreast tumors, canbe effectivelyeliminated from breast cancer cells by way of hammerhead ribozyme transgene. Elimination ofmatrilysin is associated with low invasiveness and slow tumor growth.Taken together, the studysuggests that targetingmatrilysinmay have important therapeutic implications.

Matrilysin (MMP-7 , putative metalloproteinase I, PUMP1)gene was identified through studies of collagenase-relatedconnective tissue–degrading metalloproteinases produced byhuman tumors (1). MMP-7, by acting as a proteolytic enzyme,

has been indicated in the invasiveness and progression ofcancer cells. Matrilysin cleaves extracellular matrix and base-ment membrane proteins, such as fibronectin, collagen type IV,laminin, and, particularly, elastin, entactin, osteopontin, andcartilage proteoglycan aggregates. Furthermore, matrilysinseems to mediate the proteolytic processing of other molecules(e.g., tumor necrosis factor a precursor, urokinase plasminogenactivator; refs. 2, 3).

The promoter region of the matrilysin gene contains typicalMMP promoter elements, such as AP-1 and PEA3, which mediateresponsiveness to growth factors, oncogenes, and phorbol esters.The transcription relation of matrilysin also requires the LEF-1/h-catenin pathways (4, 5). Matrilysin protein localizes tosecretory and ductal epithelium in the endometrium and invarious exocrine glands. In the mouse, high constitutive levels ofmatrilysin mRNA are found in epithelial cells in the uterus, smallintestine, and extratesticular ducts, suggesting that matrilysinmay have a specific role in normal gland and organ function, a

Cancer Therapy: Preclinical

Authors’ Affiliations: 1Metastasis and Angiogenesis Research Group,WalesCollege of Medicine, Cardiff University, Cardiff, United Kingdom and 2Departmentof Surgery, St George’s Medical School, London, United KingdomReceived 2/7/05; revised 5/24/05; accepted 6/2/05.Grant support: Breast Cancer Campaign and Cancer Research Wales.The costs of publication of this article were defrayed in part by the payment of pagecharges.This article must therefore be hereby marked advertisement in accordancewith18 U.S.C. Section1734 solely to indicate this fact.Requests for reprints:Wen G. Jiang, Metastasis and Angiogenesis ResearchGroup, University Department of Surgery,Wales College of Medicine, CardiffUniversity, Heath Park, Cardiff CF14 4XN, United Kingdom. Phone: 44-29-2074-2895; Fax: 44-29-2076-1623; E-mail: [email protected].

F2005 American Association for Cancer Research.doi:10.1158/1078-0432.CCR-05-0275

www.aacrjournals.orgClin Cancer Res 2005;11(16) August15, 2005 6012

Cancer Research. on February 19, 2021. © 2005 American Association forclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

possibility that can be explored further by the genetic manipu-lation of matrilysin levels in vivo. Antisense oligonucleotide tomatrilysin has been shown to reduce the metastasis of coloncancer to liver in a mouse model (6).

Matrilysin is also able to cleave integrins on the surface of cancercells, including breast cancer cells (7). Matrilysin forms a complexwith CD44 at the cell surface of cancer cells, possibly to coordinatethe matrix degradation process by cancer cells (8). Matrilysincan also cleave E-cadherin to induce the invasive potential ofcancer cells (9, 10). Matrilysin is involved in the development of

mammary gland (11, 12). In breast cancer, matrilysin wasclaimed to have no association with other classic clinicalparameters, such as age, menopausal status, stage, size, nodalstatus, vascular infiltrate, necrosis, steroid receptors, metastasis,and survival (n = 81, follow-up 40 months; ref. 13). In breasttissues, matrilysin was expressed by morphologically normalepithelial ducts within tumors and in tissue from reductionmammaplasties and by epithelial-derived tumor cells (14).

Despite the progress in this area, the potential role of MMP-7in human breast cancer, and particularly in clinical breastcancer, has not been thoroughly investigated. For example, thelimited information available in the literature was based on asmall cohort and a short period of follow-up. There have beenno conclusive reports to show the link between matrilysin andclinical outcome. In addition, no reports are yet available onattempting to target matrilysin using ribozyme transgenes. Inthe current study, we have shown that by knocking-down theexpression of matrilysin from breast cancer cells using ribozymetransgenes, the invasiveness of the cancer was greatly reduced.Furthermore, the matrilysin knockdown cells displayed asignificant reduction in the growth in a nude mouse model.The study went on to show that patients with breast cancerhad a significantly poor long-term (10 years) survival whenmatrilysin was expressed at a higher level.

Materials andMethods

Human mammary cancer cells, MDA-MB-231 was from theEuropean Collection of Animal Cell Culture (Salisbury, England) andwere routinely maintained in DMEM F12 with 10% FCS. Recombinanthuman hepatocyte growth factor/scatter factor was a gift from Dr. T.Nakamura (Osaka University Medical School, Osaka, Japan). Matrigel(reconstituted basement membrane) was purchased from CollaborativeResearch Products (Bedford, MA). Doxycycline was from ClontechLaboratories (Palo Alto, CA). A rabbit and a mouse anti-humanmatrilysin, anti –cytokeratin-19 (CK19), anti-Ki67, and peroxidaseconjugated anti-IgG were from Santa Cruz Biotechnologies (SantaCruz, CA) and Sigma (Poole, Dorset, England, United Kingdom),respectively. A chemiluminescence detection kit for Western blottingand protein A/G conjugate were from Santa Cruz Biotechnologies. Atranswell plate equipped with a porous insert (pore size 8 Am) was fromBecton Dickinson Labware (Oxford, England, United Kingdom). DNArestriction enzymes and T4 DNA ligase were obtained from NewEngland Biological Laboratories (Hertfordshire, England, UnitedKingdom). DNA gel extraction and plasmid extraction kits were fromQiagen (Crawley, England, United Kingdom).

Primary breast cancer tissues (n = 120) and normal tissues that wereaway from tumor tissues and free from cancer cells (n = 32) werecollected immediately after surgery and stored in the deep freezer untiluse. Patients were routinely followed clinically after surgery. The medianfollow-up period was 120 months. The presence of tumor cells in thecollected tissues was verified by examination of frozen sections usingH&E staining. Details of the samples were previously described (15).

Construction of retroviral hammerhead ribozyme transgenes targeting

human matrilysin and generation of active viral hammerhead ribozymes.The procedure has been previously reported (16, 17). Briefly, thesecondary structure of human matrilysin was generated using Zuker’sRNA mFold software (18). A suitable GUC site within the matrilysinmRNA was identified. Touch-down PCR was used to generatePCR-based ribozyme (17) using the following primers: 5V-CTGACTG-CAGTAGGTCTGATGAGTCCGTGAGGACGAAACCACTTTGCTGAT-G A G T C C G T G A G G A C G A A A A G T C C A T T T G G G C T A - 3V a n d5V-CCATACTAGCCACTATTTTTCGTCCTCACGGACTCATCAGCAAA-TAGCCCAAAATGGACTTTTCGTCCTCACGGACTCATCA-3V. Pst I andSpeI restriction sites were introduced during the PCR reaction.

Fig. 1. A, retroviral ribozyme transgenes successfully eliminated expression ofmatrilysin transcript from breast cancer cells. MDA-MB-231wild-type cellsexpressedmatrilysin transcript (top). A pREV-TRV-Matrilysin transgene successfullyknocked down the expression of the transcript from the cells.B, matrilysin andinvasiveness of breast cancer cells.

TargetingMatrilysin in Breast Cancer

www.aacrjournals.org Clin Cancer Res 2005;11(16) August15, 20056013



The above-generated antisense-hammerhead ribozymes, flanked by

PstI and SpeI sites, were first T-A cloned into a pCR2.1 TOPO cloningvector (Invitrogen, Paisley, Scotland, United Kingdom), followed by

amplification in the OneShot E. coli (Invitrogen). Bacterial clones with

correct oriented insert were verified using PCR. Plasmid was subse-quently purified from the bacterial preparation and dually digested

using PstI and SpeI. The digests were separated on a 2% agarose gel,followed by purification of the respective ribozyme inserts from the gel.

Gel purified ribozymes (with PstI and SpeI overhangs) were thenligated into either pLXSN (RetroX, K1060) or pRevTRE292 vector(RevTet-On, K1267, Clontech Laboratories) using T4 DNA ligase. Thesevectors were modified in our laboratory by inserting a U1 promoter infront of the multicloning site from our previous work (a generous giftfrom Dr. J. Laterra, Department of Neurology, Kennedy Krieger Instituteand Johns Hopkins School of Medicine, Baltimore, MD; ref. 19).Ligated products were used to transform JM109 E. coli . Bacterial cloneswith correctly ligated antimatrilysin ribozyme, termed here as pRevTRE-MAT, and control plasmid (termed pRevTRE-cont) was subsequentlyidentified. pRevTRE carried a tetracycline-responsive gene and wereused throughout the in vitro and in vivo studies. pLXSN ribozymes were

not used in the subsequent studies. The direction and sequence wereverified using a plasmid-specific primer LXSNF, 5V-CCCTTGAA-CCTCCTCGTTCGACC-3V, and U1-specific primers 5V-GGATCCGC-CAACCGAAAGT-3V (UBAMHF) and 5V-GTACGATTAACAACTAAGA-3V(UBAMHR).

Retroviral packaging and transduction of cells. Plasmid, extracted

and purified using a plasmid extraction kit (Qiafilter, Qiagen), wasintroduced to a retroviral packaging cell line, PT67, using electro-

poration as previously described (19), followed by selection with G418

containing-medium for over 3 weeks. Viral titers from stably transfectedPT67 cells were tested using NIH3T3 cells, and were found to be on

average 8 � 105 cfu/mL. Active viral stocks (a combination of pRevTRE-MAT or control stock) and the pRevTet-On viral stock (separately

generated from PT67) were used to transduce MDA-MB-231 mammary

cancer cells in the presence of polybrene (8 Ag/mL final concentration).Each transduction lasted 24 hours and three consecutive transductions

were carried out. Transduced cells were subject to dual selection withG418 (for pRevTRE-MAT) and hygromycin (for pRevTet-On; Calbio-

chem, Nottingham, England, United Kingdom), each at 100 Ag/mL for

over 3 weeks to obtain stably transduced strains that carried both the

Fig. 2. Reduction of tumor growth by theretroviral ribozyme transgene.A, growth oftumor cell variants over a 72-hour period.Shown are absorbances from stained cells.No differencewas seen between threevariant cells.B, tumor weight in grams at theend of the experiment. C, the volume oftumors over the experimental period.*, P < 0.05 versus tumor fromMDA-MB-231WT.D, immunohistochemical staining ofmatrilysin in nude mouse tumors. a, tumorfromMDA-MB-231WT; b, tumor fromMDA-MB-231pRevTRE; c, tumor fromMDA-MB-231DMatrilysin.





inducing vector, pRevTet-On and expression vector pRevTRE-MAT (or

control). These stably transduced and subsequently verified cell variantswere designated the following names and are used throughout the

text: MDA-MB-231WT, MDA-MB-231 wild type; MDA-MB-231pRevTRE,

MDA-MB-231 transduced with pRevTRE empty vector; MDA-MB-231DMatrilysin, MDA-MB-231 transduced with pRevTre-Matrilysin ribo-

zyme transgene.RNA preparation and reverse transcription-PCR. RNA from cells and

tissues was extracted using an RNA extraction kit (AbGene, Ltd., Surrey,

England, United Kingdom) and quantified using a spectrophotometer(Wolf Laboratories, York, England, United Kingdom). cDNA was

synthesized using a first-strand synthesis with an oligo(dT) primer

(AbGene). PCR primers were as follows: for matrilysin, MATF1,5V-ATGCGACTCACCGT-3V, and MATR1, 5V-TGAATTACTTCTCTTTC-3V.The PCR was done using sets of primers with the following conditions:

5 minutes at 95jC, and then 20 seconds at 94jC to 25 seconds at 56jC,50 seconds at 72jC for 36 cycles, and finally 72jC for 7 minutes.

b-actin was amplified simultaneously using the following primers5V-GCTGATTTGATGGAGTTGGA-3V and 5V-TCAGCTACTTGTTCTT-

GAGTGAA-3V. PCR products were then separated on a 0.8% agarose

gel, visualized under UV light, photographed using a Unisave camera(Wolf Laboratories) and documented with Photoshop software.

Quantitative analysis of matrilysin. The level of matrilysin tran-

scripts from the above-prepared cDNA was determined using a real-time quantitative PCR, based on the Amplifluor technology, modified

from previous reports (15, 20). Briefly, pairs of PCR primers weredesigned using the Beacon Designer (Palo Alto, CA) software (version

2), but to one of the primer, an additional sequence, known as the

Z sequence (5V-ACTGAACCTGACCGTACA-3V), which is complementaryto the universal Z probe (Intergen, Inc., Oxford, England, United

Kingdom). A Taqman detection kit for h-actin was purchased from

Perkin-Elmer (Beaconsfield, Bucks, England, United Kingdom). Thereaction was carried out using the following: Hot-start Q-master mix

(AbGene), 10 pmol of specific forward primer (5V-CTACAGTGGGAA-

CAGGCTCAG-3V, mmp7f3), 1 pmol reverse primer that has the Zsequence (underlined; 5V-ACTGAACCTGACCGTACAATCTCCTTGAG-

TTTGGCTTC-3V, mmp7ZR), 10 pmol of FAM-tagged probe (Intergen),

and cDNA from f50 ng RNA. The reaction was carried out usingIcyclerIQ (Bio-Rad, Hemel Hempstead, Hertfordshire, England, United

Kingdom), which is equipped with an optic unit that allows real-timedetection of 96 reactions using the following conditions: 94jC for

12 minutes, 50 cycles of 94jC for 15 seconds, 55jC for 40 seconds,

and 72jC for 20 seconds. CK19 was used for comparison ofcellularity during the analysis and primers for CK19 were 5V-CAGGTCCGAGGTTACTGAC-3V and 5V-ACTGAACCTGACCGTACA-CACTTTCTGCCAGTGTGTCTTC-3V, respectively (21). The levels of

the transcripts were generated from a standard that was simulta-

neously amplified with the samples. The levels of matrilysin transcriptwas also normalized by CK19 and is shown as matrilysin/CK19 ratio

here.In vitro invasion analysis and cell growth assay. This was done as

previously reported and modified in our laboratory (22). Briefly,

transwell inserts (upper chamber) with 8 Am pore size were coated with50 Ag/insert of Matrigel and air-dried before being rehydrated. Twenty

thousand cells were added to each well with or without hepatocyte

growth factor/scatter factor. After 72 hours, cells that had migratedthrough the matrix and adhered to other side of the insert were fixed

and stained with 0.5% (w/v) crystal violet. Cells that had invaded andstained with crystal violet were extracted with 10% (v/v) acetic acid and

absorbance was obtained using a multiplate reader.

Fig. 3. Immunohistochemical staining ofmatrilysin in humanmammary tissues. Left,normal tissue; middle, tumor tissue; right,negative control.Magnifications, from top tobottom,�40,�100,�200, and�400,respectively.





For cell growth assay, MDA-MB-231WT, MDA-MB-231pRevTRE, orMDA-MB-231DMatrilysin cells were plated into a 96-well plate at 2,500cells/well. Cells were fixed in 10% formaldehyde at the day of plating;days 1, 2, 3, 4, 5, and 6 after plating; and then stained with 0.5% (w/v)crystal violet. Following washing, stained crystal was extracted with

10% (v/v) acetic acid and absorbance determined using a multiplate

reader (using wavelength 560 nm). The growth of cells are shown here

as absorbance (mean F SD).In vivo development of mammary tumor. Athymic nude mice (Nude

CD-1) of 4 to 6 weeks old were purchased from Charles River (Margate,Kent, England, United Kingdom) and maintained in filter-toped units.

One hundred microliters of cell suspension (in 0.5 mg/mL Matrigel)were injected s.c. at the left scapula area. Each tumor group was dividedinto those receiving twice weekly injection of doxycycline or injection

of buffers. Mice were weighed and tumor sizes measured twice weeklyfor 4 weeks. Mice with weight loss over 25% and tumor size larger than

1 cm in any dimension were terminated according to the UnitedKingdom Home Office and the United Kingdom Coordinating

Committee on Cancer Research guidelines. The volume of the tumorwas determined using the following formula: tumor volume = 0.523 �width2 � length (19). At the conclusion of the experiment, animalswere terminally anesthetized and primary tumors were dissected,weighed, and frozen at �80jC. Part of the primary tumors was fixed

for histologic examination.Immunohistochemical staining of matrilysin. Frozen sections of

tissues (32 paired normal and tumor tissues as well as dissected tumor

tissues) were cut at a thickness of 6 Am using a cryostat (15). The sections

were mounted on superfrost plus microscope slides, air dried, and then

fixed in a mixture of 50% acetone and 50% methanol. The sections were

then placed in Optimax wash buffer for 5 to 10 minutes to rehydrate.

Sections were incubated for 20 minutes in 10% horse serum of blocking

solution and probed with the primary antibody. Following extensive

washings, sections were incubated for 30 minutes in the secondary

biotinylated antibody (Multilink Swine anti-goat/mouse/rabbit immu-

noglobulin; DAKO, Inc., Kidlington, Oxford, England, United King-

dom). Following washings, avidin-biotin complex (Vector Laboratories,

Peterborough, England, United Kingdom) was then applied to the

sections followed by extensive washings. Diaminobenzidine chromogen

(Vector Laboratories) was then added to the sections that were incubated

in the dark for 5 minutes. Sections were then counterstained in Gill’s

hematoxylin and dehydrated in ascending grades of methanol before

clearing in xylene and mounting under a coverslip. Cytoplasmic staining

of the respective proteins was quantified using Optimas 6.0 software as

we previously described (23, 24) and is shown here as relative staining

intensity.Statistical analysis was carried out using Mann-Whitney U test and

significant difference was taken at P < 0.05.

Results

Elimination of matrilysin expression in breast cancer cellsaffects the in vitro invasiveness and migration of breast cancercells. MDA-MB-231 cells expressed matrilysin as shown byreverse transcription-PCR (Fig. 1A). Two sublines were gener-ated by infecting the wild-type MDA-MB-231 (MDA-MB-231WT)with pRevTRE-MAT and pRevTRE-cont active viral stocksand termed MDA-MB-231DMatrilysin and MDA-MB-231pRevTRE,respectively. MDA-MB-231DMatrilysin cells almost completely lost

Fig. 4. Levels of expression of matrilysin transcript inmammary tissues. A, levels in normal (n = 32) and tumor tissues (n = 120; *, P < 0.05); B, levels in different tumorgrades (#, P = 0.05, n = 23, 41, and 56 for grade1, 2, and 3, respectively);C, levels in node negative (n = 65) and node positive (n = 55) tumors (*, P < 0.05 versus grade1).D, levels of expression of matrilysin transcript in patients with predicted outcome. E, levels of expression of matrilysin transcript and clinical outcome (bottom). *, Patients whodied of breast cancer ^ related conditions (n = 20) had higher levels of matrilysin transcript than patients who remained disease-free (n = 81). Insets, matrilysin/CK19 ratio.





the expression of matrilysin, compared with the wild-typeMDA-MB-231WT and MDA-MB-231pRevTRE (Fig. 1A). Thissuggests that the retroviral transgene was highly active intargeting matrilysin mRNA (Fig. 1A). It is interesting to note thatin the pRevTet-On ribozyme transgene system created, theknockout was highly effective, with or without doxycycline,indicating that the U1 promoter inserted in the pRevTRE292worked well, independent of the tetracycline response elements.

All three cell variants were subject to in vitro testing. Figure 1Breveals that MDA-MB-231DMatrilysin cells had significantlyreduced invasiveness into Matrigel in response to hepatocytegrowth factor (number of invading cells 16.02.5). In contrast,MDA-MB-231WT (26.2 F 6.2, P < 0.05 versus MDA-MB-231DMatrilysin) and MDA-MB-231pRevTRE (25.3 F 4.2, P < 0.01versus MDA-MB-231DMatrilysin) cells responded equally tohepatocyte growth factor. Addition of doxycycline did notsignificantly increase the effectiveness of the ribozyme trans-gene. Neither ribozymes nor control plasmid significantly

altered the growth rate of MDA MB-231 cells. Over a periodof 72 hours, MDA-MB-231WT, MDA-MB-231pRevTRE, andMDA-MB-231DMatrilysin cells showed a similar rate of growth(absorbances were 1.56 F 0.21, 1.61 F 0.47, and 1.57 F 0.36,respectively; Fig. 2A). Addition of doxycycline did not affect thegrowth rate of the cells.

Reduced tumor growth of breast tumors after elimination ofmatrilysin. In the nude mouse model, MDA-MB-231WT, MDA-MB-231pRevTRE, and MDA-MB-231DMatrilysin cells were s.c.injected. Figure 2 shows that whereas MDA-MB-231WT andMDA-MB-231pRevTRE tumors grew at similar rate, growth ofMDA-MB-231DMatrilysin tumors were significantly slower(Fig. 2C). In addition, the weight of tumors at the conclu-sion of the experiment indicated the same trend, where MDA-MB-231DMatrilysin tumors were the smallest (Fig. 2B). The sizeand rate of growth of tumors from MDA-MB-231DMatrilysin

cells were similar, with or without the injection of doxycycline(Fig. 2B and C). Primary tumors were further analyzed his-tologically. MDA-MB-231DMatrilysin tumors displayed substan-tially weaker staining of matrilysin (Fig. 2D) compared withMDA-MB-231WT and MDA-MB-231pRevTRE tumors (Fig. 2D-Bversus C, respectively).

Aberrant expression of matrilysin in human breast cancer.Matrilysin was seen to be present primarily in the mammaryepithelial cells and very little in stromal cells (Fig. 3, left). Inbreast cancer tissues, the staining became markedly stronger,again in cancer cells (Fig. 3, middle). In both cases, stromal cellsshowed no or very little staining.

We went on to analyze the transcript levels of matrilysin,using quantitative real-time PCR. There were higher levels ofmatrilysin transcript in breast tumors compared with normaltissues (P = <0.05; Fig. 4A). Interestingly, the levels of matrilysintranscript were high in high-grade tumors (P < 0.05, grade 3versus grade 1; n = 23, 41, and 56 for grades 1, 2, and 3respectively; Fig. 4C) and in node-positive tumors (n = 55; P =0.05 versus node-negative tumors, n = 65; Fig. 4B). A similartrend was seen when matrilysin transcript was normalized byCK19 (Fig. 4, inset).

Here, we have used two methods to assess the clinicalrelevance: the Nottingham prognosis index (with good,moderate, and poor predicted prognosis, for the respectiveindex <3.4, 3.4-5.4, and >5.4) and with clinical outcome:patients who remained disease-free, with metastatic disease,with local recurrence and with mortality (Fig. 4D). Tumorsfrom patients with predicted moderate and poor prognosis hadhigher levels (Fig. 4D). Tumors from patients who died ofbreast cancer (n = 20) had significantly higher levels ofmatrilysin transcript (Fig. 4E), compared with those whoremained disease-free (n = 81). The same was seen whenmatrilysin transcript was normalized by CK19 (Fig. 4, insets).

Expression of matrilysin is linked with clinical outcome. AKaplan-Meier survival analysis indicated that high levels ofMatrilysin were significantly associated with reduced survival(P = 0.0143; Fig. 5A), in that patients with higher levels ofmatrilysin had a median survival of 107.5 months (95%confidence interval, 85.1-129.9 months), compared withpatients with low levels of matrilysin, 140.5 months (95%confidence interval, 131.5-150.0). High levels of matrilysin wasalso linked to shorter disease-free survival although this is notstatistically significant (P = 0.0892; Fig. 5B) in that patientswith higher levels of matrilysin had a median survival of

Fig. 5. Matrilysin and long-term survival. A, overall survival; B, disease-freesurvival.





107.1 months (95% confidence interval, 84.6-129.5 months),compared with patients with low levels of matrilysin 132.9months (95% confidence interval, 122.5-143.5 months).

Discussion

The current study has shown that matrilysin is potentiallyinvolved in the invasion and migration of breast cancer, andthat genetical elimination of matrilysin resulted in lessaggressive cells and slower-growing tumors. Supported by theclinical information that matrilysin was significantly associateda poor survival, this study clearly shows the pivotal role ofmatrilysin in breast cancer and potential value for matrilysin asa therapeutic target.

The role of matrilysin in breast cancer and its progression haslong been debated and recognized. As well as acting as a matrixdegradation enzyme, matrilysin has been shown to enhancethe selection of cells that are resistant to apoptosis, potentiallyvia the FasL pathway (25), targeting critical cell adhesionmechanisms in cancer cells including the E-cadherin complex(9, 10, 26). Matrilysin has also been shown to cleave the IGFBP-3, and increase the bioavailability of IGF-I, thus promotingthe survival of cancer cells (27). Factors leading to the up-regulation and increased expression of matrilysin tumorsinclude the h-catenin and PEA3 pathways (28, 29). Syndican-1is also a reported target of matrilysin (30). In breast cancer, theclinical relevance of matrilysin has not been fully established.The current study has first provided evidence that matrilysin isnot only overexpressed in breast tumor tissues, but also ahigher level of matrilysin transcript is associated with a highergrade of tumor, with lymph node involvement, and withclinical mortality. Perhaps the most convincing evidence thatmatrilysin is clinically relevant is its significant link with long-term survival, both overall and disease-free. A similar observa-tion has been made to other clinical tumors such as colorectalcancer (31). This study has, therefore, established that firstly,matrilysin is linked to the aggressive nature of breast cancer

and, secondly, pointed the potential value of targeting the MMPin cancer therapy.

In the current study, we have used ribozyme transgene as ameans to manipulate the expression of matrilysin in cancer cellsand to create a cell model to test if targeting matrilysin is of valuein cancer intervention. As shown in Fig. 1, the ribozyme hassuccessfully knocked out the expression of matrilysin mRNAfrom MDA-MB-231 cell, the wild type of which highly expressesmatrilysin. Using the cells thus generated (MDA-MB-231DMatrilysin), we have shown that loss of matrilysin in thebreast cancer cell MDA-MB-231DMatrilysin is accompanied by thesignificant reduction of invasiveness of breast cancer cells in vitro.This is largely expected, given the role of matrilysin in matrixdegradation. The same cells were further tested in a in vivo modeland have shown a convincing slower growth rate, compared withcontrol and wild-type tumor cells. It is also noteworthy thatinsertion of a U1 promoter in the pRevTRE vector has enabledthe Tet-On system independent of doxycycline.

The success of the ribozyme transgene to matrilysin has beenfurther verified from the in vivo model in that the MDA-MB-231DMatrilysin tumors have also displayed low levels ofmatrilysin staining in the tumor. As well as assisting invasionby degradation of matrix proteins, matrilysin has also beenshown to potentially activate angiogenic factors or factors thatmay facilitate angiogenesis (32, 33), affecting the growth ofendothelial cells (34), degradation of matrix thus facilitatingthe angiogenic process (35). These factors may collectivelycontribute to the reduced growth of the breast tumors in thein vivo model.

In conclusion, the current study has shown a potentialaberrant level of matrilysin in human breast cancer and that thehigh levels of are linked to a poor clinical outcome. In addition,the current study has also shown that manipulating theexpression of matrilysin in breast cancer cells can reduce theinvasiveness in vitro and tumor growth in vivo. This indicatesthat targeting matrilysin may be an effective way to intervenethe aggressiveness of breast cancer.

References1. Muller D, Quantin B, Gesnel M-C, et al. The collage-nase gene family in humans consists of at least fourmembers. BiochemJ1988;253:187^92.2. Gaire M, Magbanua Z, McDonnell S, McNeil L,Lovett DH, Matrisian LM. Structure and expression ofthe human gene for the matrix metalloproteinasematrilysin. JBiol Chem1994;269:2032^40.3.Wilson CL, Matrisian LM. Matrilysin: An epithelialmatrix metalloproteinase with potentially novel func-tions. IntJBiochem Cell Biol1996;28:123^36.4. Crawford HC, Fingleton BM, Rudolph-Owen LA,et al. The metalloproteinase matrilysin is a target of h-catenin transactivation in intestinal tumors. Oncogene1999;18:2883^91.5. Crawford HC, Fingleton B, Gustavson MD, et al.ThePEA3 subfamily of Ets transcription factors synergizeswithh-catenin-LEF-1toactivatematrilysintranscriptionin intestinal tumors.MolCellBiol 2001;21:1370^83.6. Hasegawa S, Koshikawa N, Momiyama N, et al.Matrilysin-specific antisense oligonucleotide inhibitsliver metastasis of human colon cancer cells in a nudemousemodel. IntJCancer1988;76:812^6.7. Abdel-Ghany M, Cheng HC, Elble RC, Pauli BU. Thebreast cancer h(4) integrin and endothelial humanCLCA2 mediate lung metastasis. J Biol Chem 2001;276:25438^46.8. YuWH,Woessner JF Jr, McNeish JD, Stamenkovic I.CD44 anchors the assemblyofmatrilysin/MMP-7with

heparin-binding epidermal growth factor precursorand ErbB4 and regulates female reproductive organremodeling. Genes Dev 2002;16:307^23.9. Davies G, JiangWG, MasonMD. Matrilysin mediatesextracellular cleavage of E-cadherin fromprostate can-cer cells: a keymechanism in HGF/SF induced cell-celldissociation and in vitro invasion. Clin Cancer Res2001;7:3289^97.10. NoeV, Fingleton B, Jacobs K, et al. Release of aninvasion promoter E-cadherin fragment by matrilysinand stromelysin-1. JCell Sci 2001;114:111^8.11. Rudolph-Owen LA, Cannon P, Matrisian LM. Over-expression of the matrix metalloproteinase matrilysinresults in premature mammary gland differentiationand male infertility. Mol Biol Cell 1998;9:421^35.12. Rudolph-Owen LA, Matrisian LM. Matrix metallo-proteinases in remodelingof thenormal andneoplasticmammary gland. J Mammary Gland Biol Neoplasia1998;3:177^89.13. PachecoMM,MouraoM, Mantovani EB, NishimotoIN, Brentani MM. Expression of gelatinases A and B,stromelysin-3 and matrilysin genes in breast carcino-mas: clinico-pathological correlations. Clin ExpMetastasis1998;16:577^85.14. Heppner KJ, Matrisian LM, Jensen RA, RodgersWH. Expression of most matrix metalloproteinasefamily members in breast cancer represents a tumor-inducedhost response.AmJPathol1996;149:273^82.

15. JiangWG,Watkins G, LaneJ, et al. Prognostic valueof Rho family and rho-GDIs in breast cancer. ClinCan-cer Res 2003;9:6432^40.16. JiangWG, Grimshaw D, Leterra J, Abounader R,Lane J, Mansel RE. Suppression of invasiveness andcellular motility of breast cancer cells by a hammer-head ribozyme encoding cMETantisense. Clin CancerRes 2001;7:2555^62.17. JiangWG, Davies G, Fodstad O. Com-1/p8 in oes-trogen regulated growth of breast cancer cells, theER-h connection. Biochem Biophys Res Commun2005;330:253^62.18. Zuker M. Mfold web server for nucleic acid foldingand hybridization prediction. Nucleic Acids Res 2003;31:3406^15.19. JiangWG,GrimshawD, LaneJ, et al. Retroviralham-merhead transgenes to cMETand HGF/SF inhibitedgrowth of breast tumor, induced by fibroblasts. ClinCancer Res 2003;9:4274^81.20.Nazarenko IA, Bhatnagar SK, Hohman RJ. A closedtube format for amplification and detection of DNAbased on energy transfer. Nucleic Acids Res 1997;25:2516^21.21. JiangWG, Watkins G, Fodstad O, Douglas-Jones A,Mokbel K, Mansel RE. Differential expression of theCCN family members Cyr61 from CTGF and Nov inhuman breast cancer. Endocr Relat Cancer 2004;11:781^91.





22. Jiang WG, Hiscox S, Hallett MB, Horrobin DF,Mansel RE, Puntis MCA. Regulation of the expressionof E-cadherin on human cancer cells by g linolenicacid. Cancer Res1995;55:5043^8.23. Davies G, JiangWG, Mason MD. Cell-cell adhesionand signalling intermediates inhumanprostate cancer.JUrology 2000;163:985^92.24. King JAC, Ofori-Acquah AF, StevensT, et al. Prog-nostic value of ALCAMinhumanbreast cancer.BreastCancer Res 2004;6:R478^87.25. Vargo-Gogola T, Fingleton B, Crawford HC,Matrisian LM. Matrilysin (matrix metalloproteinase-7)selects for apoptosis-resistant mammary cells in vivo.Cancer Res 2002;62:5559^63.26. Crawford HC, Fingleton B, Gustavson MD, et al.The PEA3 subfamily of Ets transcription factorssynergizes with h-catenin-LEF-1to activate matrilysintranscription in intestinal tumors. Mol Cell Biol 2001;21:1370^83.

27. Miyamoto S, Yano K, Sugimoto S, et al. Matrixmetalloproteinase-7 facilitates insulin-like growthfactor bioavailability through its proteinase activity oninsulin-like growth factor binding protein 3. CancerRes 2004;64:665^71.28. Rivat C, Le Floch N, Sabbah M, et al. Synergisticcooperation between theAP-1and LEF-1transcriptionfactors in activation of the matrilysin promoter by thesrc oncogene: implications in cellular invasion. FASEBJ 2003;17:1721^3.29. McGuire JK, Li Q, Parks WC. Matrilysin (matrixmetalloproteinase-7) mediates E-cadherin ectodo-main shedding in injured lung epithelium. AmJPathol2003;162:1831^43.30. Li Q, Park PW,Wilson CL, Parks WC. Matrilysinshedding of syndecan-1regulates chemokine mobili-zation and transepithelial efflux of neutrophils in acutelung injury. Cell 2002;111:635^46.31.MasakiT, Matsuoka H, SugiyamaM, et al. Matrilysin

(MMP-7) as a significant determinant of malignantpotential of early invasive colorectal carcinomas. Br JCancer 2001;84:1317^21.32.HashimotoG, Inoki I, FujiiY, AokiT, Ikeda E,OkadaY.Matrix metalloproteinases cleave connective tissuegrowth factor and reactivate angiogenic activity ofvascular endothelial growth factor 165. J Biol Chem2002;277:36288^95.33. Patterson BC, Sang QA. Angiostatin-convertingenzyme activities of human matrilysin (MMP-7) andgelatinase B/type IV collagenase (MMP-9). J BiolChem1997;272:28823^5.34. Huo N, Ichikawa Y, Kamiyama M, et al. MMP-7(matrilysin) accelerated growth of human umbilicalvein endothelial cells. Cancer Lett 2002;177:95^100.35. YuWH,Woessner JF Jr. Heparan sulfate proteogly-cans as extracellular docking molecules for matrilysin(matrix metalloproteinase 7). J Biol Chem 2000;275:4183^91.





2005;11:6012-6019. Clin Cancer Res Wen G. Jiang, Gaynor Davies, Tracey A. Martin, et al. The Potential Implications in Breast Cancer Therapy

:In vivoTargeting Matrilysin and Its Impact on Tumor Growth

Updated version

http://clincancerres.aacrjournals.org/content/11/16/6012

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/11/16/6012.full#ref-list-1

This article cites 33 articles, 18 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/11/16/6012.full#related-urls

This article has been cited by 19 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] 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://clincancerres.aacrjournals.org/content/11/16/6012To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/content/11/16/6012.full#ref-list-1

http://clincancerres.aacrjournals.org/content/11/16/6012.full#related-urls

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

mailto:[email protected]



TargetingMatrilysinandItsImpactonTumorGrowth In vivo ...possibility that can be explored further by...

Documents

Transcript of TargetingMatrilysinandItsImpactonTumorGrowth In vivo ...possibility that can be explored further by...