MMP Inhibition and Downregulation by Bisphosphonates

453

MMP Inhibition and Downregulationby Bisphosphonates

OLLI TERONEN,

a,e,h

PIA HEIKKILÄ,

d

YRJÖ T. KONTTINEN,

c,e

MINNA LAITINEN,

g

TUULA SALO,

f

ROELAND HANEMAAIJER,

i

ANNELI TERONEN,

d

PÄIVI MAISI,

e

AND TIMO SORSA

b

Department of Oral and Maxillofacial Surgery,

a

Periodontology,

b

and Anatomy,

c

Institute of Dentistry,

d

and Helsinki University Central Hospital,

e

University of Helsinki, Helsinki, Finland

f

Department of Oral Pathology, Institute of Dentistry, University of Oulu, Oulu, Finland

g

Institute of Medical Technology, Surgery, University of Tampere, Tampere, Finland

e

Faculty of Veterinary Medicine, University of Helsinki, Helsinki, Finland

i

Gaubius Laboratory, TNO-PG, PO Box 2215, 2301 CE Leiden, the Netherlands

ABSTRACT: Bisphosphonates are a group of drugs capable of inhibitingbone resorption, and are thus used for the treatment of bone diseases, such asPaget’s disease, osteoporosis, and for bone metastases of malignant tumors.Their primary cellular target is considered to be the osteoclast. The molecularmechanisms responsible for the downregulation of bone resorption by bispho-sphonates have remain unclear. We have discovered that various matrix met-alloproteinases (MMPs) are inhibited

in vitro

by several bisphosphonates. Thisnovel finding may, in part, explain the efficacy of bisphosphonates in their cur-rent indications in humans. In enzyme activity tests using purified and recom-binant enzymes, we have observed the inhibition of MMP-1, -2, -3, -7, -8, -9,-12, -13, and -14 by clondronate, alendronate, pamidronate, zolendronate,nedrinate, and clodrinate. The IC

50

s range from 50 to 150

�

M. We have alsoshown that clodronate can downregulate the expression of MT1-MMP proteinand mRNA in several cell lines. Additionally, several bisphosphonates de-crease the degree of invasion of malignant melanoma (C8161) and fibrosarco-ma (HT1080) cells through artificial basement membrane (Matrigel) in cellcultures at IC

50

s of 50–150

�

M and below. Having low toxicity and proven tobe well tolerated after several years in human use, bisphosphonates have thepotential to become one of the main MMP-inhibitors for MMP-related humansoft and hard tissue–destructive diseases in the near future.

It has been clear for some time that inhibition of MMPs by pharmacologic agents incertain pathologic conditions would be beneficial to downregulate the tissue-destructive course in certain inflammatory and malignant diseases. Therefore the de-velopment of new MMP inhibitors to drugs it is hoped will be useful in the future isof great interest. Although the therapeutic utility of a collagenase/MMP inhibitorhad already been recognized briefly after the initial discovery of vertebrate-type col-

h

Address for correspondence: Dr. Olli Teronen, Department of Oral and MaxillofacialSurgery, Institute of Dentistry, P.O. Box 41, FIN-00014 University of Helsinki, Finland. Phone,(358)-9-191 27 257; fax, (358)-9-191 27 265; e-mail, [email protected]

454 ANNALS NEW YORK ACADEMY OF SCIENCES

lagenase or MMP-1,

1

it is somewhat surprising that no safe pharmacologic productwithout adverse side-effects is currently available for therapeutic use as an MMP in-hibitor. The discovery that tetracyclines can control tissue-eroding MMPs was madeby Golub

et al.

more than 15 years ago.

2

In clinical use, low nonmicrobial doses oftetracyclines, especially doxycycline, have proved to be effective in the treatment ofperiodontal disease and arthritides.

3–9

Tetracycline derivatives lacking antimicrobi-al properties also called chemically modified tetracyclines (CMTs), have been pro-duced and tested for the treatment of periodontitis and arthritis in animal models. Itremains to be seen whether CMTs and tetracyclines (doxycycline) develop undesir-able side effects, such as gastrointestinal disturbances and potential antibiotic resis-tance, during long-term treatment.

Some active side-targeted or peptidomimetic MMP inhibitors have been designedand tested.

10

Although very effective

in vitro,

at even nanomolar concentrations,these new “designer” inhibitors require higher doses than expected

in vivo,

and stillhave a long way to go to become clinically ready and safe pharmaceutical productsowing to their early stage of development and the multiple steps a new drug has toundergo before approval for clinical use.

10

Several studies during recent years have described beneficial effects of bisphos-phonates in malignant human diseases such as breast cancer and multiple myelo-ma.

11–14

Other effects of bisphosphonates include inhibition of breast cancer celladhesion

in vitro,

15

inhibition of tooth-root resorption in rats,

16

reduction of the hu-man metastastic cancer burden in nude mice,

17

and a more effective decrease in os-teolytic bone metastasis with a combination of TIMP-2 and bisphosphonateibandronate than with either alone.

18

Finally, treatment with bisphosphonates haltedperiodontal tissue destruction and the progression of bone destruction during thecourse of periodontitis in monkeys, beagle dogs, and humans.

19–21

These phenome-na strongly mimic the outcome following use of any matrix metalloproteinase inhib-itor.

The bisphosphonates are a class of drugs developed in the past three decades foruse against various diseases involving disturbances in bone and calcium metabolism.Bisphosphonates are synthetic compounds that exhibit strong affinity for the hy-droxyapatite crystal of bone. The P–C–P bond of bisphosphonates is stable in heatand most chemical reagents, and completely resistant to enzymatic hydrolysis, butcan be hydrolyzed in solution by ultraviolet light. The major effect of pharmacolog-ically active bisphosphonates has so far been their recognized ability to inhibit boneresorption. Bisphosphonates are generally well tolerated and have low toxicity.

22,23

At first, the antiosteolytic action of bisphosphonates was supposed to be due to thediminished dissolution of the hydroxyapatite crystals to which bisphosphonates areattached. Today, the mechanisms of this inhibition have not been completely clari-fied, but several lines of evidence indicate that the inhibition of bone resorption bybisphosphonates is a result of one or more of the following mechanisms: (a) directinhibition of osteoclast function; (b) physicochemical incorporation into the skeletalmatrix, thereby interfering with the actual process of bone resorption; or (c) directinhibition of osteoblast-mediated cytokine production. Rodan and Fleisch

24

con-clude that although the detailed mechanism of action of bisphosphonates has notbeen elucidated, it is clear that at the tissue level all bisphosphonates inhibit boneresorption, bone turnover, and therefore, bone loss. At the cellular level bisphospho-

455TERONEN

et al.

: MMP INHIBITION BY BISPHOSPHONATES

nates are considered to act directly and/or indirectly to inactivate osteoclastic boneresorption.

24

Vacuolar ATPase, squalene synthetase, and protein tyrosine phos-phatase are inhibited with certain bisphosphonates, but conclusions as to the impactof these findings on bone resorption or other possible biological effects

in vivo

re-mains to be established.

24–27

In this work we describe the inhibition of catalytic activities of several geneticallydistinct, but structurally related MMP family members by various bisphosphonatessuch as clodronate, alendronate, pamidronate, and zolendronate. This inhibition atleast partly explains the effects of bisphosphonates when used for treatment of dis-eases involving significant bone and soft tissue destruction. This therefore suggeststhat certain MMPs may indeed play an important role in bone resorption. Further-more, our findings may lead to new medical indications and use of these safe, famil-iar drugs in numerous tissue-destructive inflammatory and malignant diseases.

MATERIALS

Chemicals

Pamidronate and zolendronate were donated by manufacturers for this purpose aswas alendronate (Merck, Sharp & Dohme, West Point, PA, USA). Clodronate waspurchased from Leiras, Turku, Finland.

Matrix Metalloproteinases (MMPs)

For enzymatic assays of MMP activities in human biological fluids and extracts,saliva, gingival crevicular fluid (GCF), peri-implant sulcus fluid, and jaw cyst as wellas gingival tissue extracts were used. The various types of MMPs were studied byWestern blotting with specific poly- and monoclonal antibodies.

28–31

Purified MMPsfrom human peripheral blood neutrophils as well as human gingival fibroblasts andkeratinocytes (MMP-1, -2, -8, and -9) were used as described in earlier papers.

28–33

Recombinant MMPs (MMP-1, -3, -7, -12, -13, -14, and -20), kindly provided by Drs.G. Murphy, V. Knäuper, and C. López-Otín were also used in the experiments.

METHODS

Cell Cultures

Human osteosarcoma MG-63, U2-OS, human melanoma C8161, and humanfibrosarcoma HT1080 cell lines were purchased from American Type CultureCollection (ATCC, Rockville, MD). Cells were cultured in a humid atmosphere of5% CO

2

and 95% air at 37°C in Dulbecco’s modified eagle’s medium with 1000 mgglucose/

L

and sodium bicarbonate (DMEM, Sigma Chemical Co, St Louis, MO)supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gibco BRL,Paisley, Scotland), 2.2 mM to which were added

L

-glutamine, 1% non-essential ami-no acids (cat. no. 11140-035, Gibco BRL), penicillin (500 IU/ml)(Gibco BRL), andstreptomycin (50 mg/ml)(Gibco BRL). After reaching confluence, the cells were


harvested by trypsinization with 0.25% trypsin-EDTA (Gibco-BRL) and reseededfor further use.

Invasion Assay

Tumor cell invasion was studied using 6.4-mm-diameter Boyden chambers pre-coated with Matrigel (Becton Dickinson, Belford, MA); the layer of Matrigel matrixserves as reconstituted basement membrane

in vitro

. This uniform layer occludes thepores of the membrane, blocking noninvasive cells from migrating through the mem-brane. By contrast, invasive cells are able to detach themselves from and migratethrough the Matrigel matrix-treated membrane.

Quantification of Matrigel Invading Cells

Cells were preincubated at 37

°

C in a 5% CO

2

atmosphere for 2 hr in the presenceof different concentrations of bisphosphonate (0–500

µ

M) in the media. Each wellwas plated with 100,000 cells in a 500-

µ

l volume, in fresh serum-containing media.After cells were cultured for 16–20 hr, they were fixed in methanol, washed, andstained in toluidene blue. Cells were removed from the upper surface of the mem-brane with a cotton swab and the cells migrated in the membrane were quantitatedby calculating under microscope.

Cell Culture and Gelatin Zymography

For MMP assays cells were trypsinized and plated at the desired density (1

×

10

5

cells per well) in the 24-well tissue culture plate (Greiner GmbH, Frickenhausen,Germany). Confluent cell layers were washed with fresh medium and cultured for 24hours with fresh medium containing 1

×

10

–7

M phorbol 12–myristate 13–acetate(PMA) (Sigma). The cell layers were then washed twice with serum-free mediumand re-fed with serum-free medium containing 0, 0.1, 1.0, or 2.5 mM clodronate(Leiras, Turku, Finland). Cultures were incubated for 24 hours, and the conditionedmedia harvested and analyzed for MMPs by zymography and immunoassays.

Gelatin Zymography

Zymograms were used for the detection of distinct molecular forms of MMP-2 incell culture media in varying concentrations of bisphosphonates reflecting the pro-teolytic activation of proMMP-2 from higher molecular form to lower ones. Sampleswere incubated in Laemmli’s sample buffer for 30 minutes at 22

°

C, after which 20

µ

lof the sample was loaded onto each lane of a polyacrylamide (10% sodium dodecylsulfate) gel containing gelatin. A separate lane was run with molecular weight stan-dards (high and low molecular weight prestained standards from BioRad Lab.,Richmond, CA, USA). Electrophoresis was carried out in precooled pool buffer(50 mM Tris; 0.38 M glysine; 0.1% SDS; pH 8.3) at a continuous voltage of 110 V.Gels were washed three times for 10 minutes in 50 mM Tris, buffer, pH 7.5, and sup-plemented with 2.5% Tween and 0.02% NaN

3

; then three times for 10 minutes in thesame buffer supplemented with 1

µ

M ZnCl

2

and 5 mM CaCl

2

; and overnight at 37

°

Cin 50 mM Tris buffer with 5 mM CaCl

2

, 1

µ

M ZnCl

2

, and 0.02% NaN

3

, pH 7.5. Thereaction was stopped with Coomassie Brilliant Blue R 250 staining followed by

457TERONEN

et al.


destaining in 5% acetic acid and 10% methanol in water. The gelatinases from thesample solubilize their substrate in the gel and therefore were visualized as whitebands against the blue-stained surrounding gel, where the substrate is intact andheavily stained.

RNA Extraction and Nothern Blot Analysis

Total RNA was prepared from 5

×

10

7

cells treated with 10

–7

PMA using Quick-Prep

®

total RNA extraction kit (Pharmacia Biotech, Uppsala, Sweden). Northernanalysis of MT-1-MMP expression was carried out as described.

53

Measurement of Collagenase Activity

Samples (biological fluids, tissue extracts, purified human MMPs, recombinanthuman MMPs, or cell culture media) with or without 1 mM APMA-treatment (an op-timal organomercurial proMMP activator) were incubated with pure type I collagenin the presence of varying concentrations of bisphosphonates. Subsequently, thesamples were separated by gel electrophoresis, and the type I collagen degradationwas quantitated by densitometry.

28

Measurement of

�

-caseinolytic Activity

Samples were incubated with 52

µ

M

β

-casein for 1 hr at 22

°

C.

34

The enzyme re-action was then halted, and samples were electrophoresed and analyzed using a den-sitometer (BioRad Model GS-700 Imiging Densitometer). The disappearance of21-kDa

β

-casein was regarded as MMP activity and quantitated.

34

Colorimetric Assay for Matrix Metalloproteinases Using Modified Pro-urokinase as Substrate

The recombinant catalytic domain of human MT1-MMP and other pure humanMMPs

35,36

was incubated with 0–2000

µ

M clodronate and other bisphosphonates(1 hour preincubation) and assayed for MT1-MMP and other MMPs’ activity ac-cording to Verheijen

et al.,

36

with the results expressed as activity units, in which oneunit is 1000

× ∆

A

405

/hr.

2

Measurement of Gelatinase Activity by Soluble Radioactive Type I Gelatin Substrate

Gelatinolytic activities of the samples and purified MMPs were assayed with

125

I-labeled gelatin as a substrate.

37

Specimens were incubated with

125

I-labeledgelatin for 1 hour at 37

°

C, after which the undegraded gelatin was precipitated with20% TCA. The supernatants and precipitates were counted separately by a gammascintillation counter.

38


TA

BL

E 1

.T

he e

ffec

ts o

f va

riou

s bi

spho

spho

nate

s on

MM

P a

ctiv

itie

s, a

ctiv

atio

n, p

rodu

ctio

n, a

nd m

alig

nant

cel

l inv

asio

n

Dow

nreg

ulat

ion

or i

nhib

itio

n of

:C

lond

rona

te

Pam

idro

nate

Ale

ndro

nate

Zol

endr

onat

eN

edri

nate

Clo

drin

ate

Til

undr

onat

e

MM

P-1

act

ivit

y+

++

++

++

++

MM

P-2

act

ivit

y+

++

+

MM

P-3

act

ivit

y+

++

++

++

++

MM

P-8

act

ivit

y+

++

++

++

++

MM

P-9

act

ivit

y+

++

++

++

+

MM

P-1

2 ac

tivi

ty+

++

++

+

MM

P-1

3 ac

tivi

ty+

++

++

+

MM

P-1

4 ac

tivi

ty+

++

MM

P-2

0 ac

tivi

ty+

++

U2-

OS

MM

P-2

act

ivat

ion

++

+

MG

63 M

MP

-2 a

ctiv

atio

n+

+

MG

63 M

T1-

MM

P e

xpre

ssio

n+

++

MG

63 M

T1-

MM

P p

rodu

ctio

n+

++

HT

1080

cel

l in

vasi

on+

++

++

++

++

++

+

C81

61 c

ell

inva

sion

++

++

++

++

++

++

459TERONEN

et al.


RESULTS

T

ABLE

1 summarizes the effects of distinct bisphosphonates on MMP activities,MMP expression, and MMP production. We have also observed dose-dependent de-crease in invasion of malignant cell lines through experimental basement membrane(Matrigel) in the presence of bisphosphonates.

Inhibition of MMP Activities, Expression, and Production by Bisphosphonates

Clodronate inhibited neutrophil MMP-8–mediated degradation both of

β

-caseinand of type I collagen substrates. Inhibition of type I collagen as well as

β

-casein deg-radation were both dose-dependent (T

ABLE

1).

29

Furthermore, not only was the puri-fied neutrophil MMP-8 inhibited, but the corresponding enzyme in complexbiological/inflammatory fluids such as gingival crevicular fluid (GCF) and peri-implant sulcular fluid (PISF) was inhibited as well.

29

Clodronate at 200 and 400

µ

Mconcentrations efficiently inhibited recombinant MMP-1.

30

Although significant andspecific degradation of type I collagen by buffer-treated MMP-1 from jaw cyst and hu-man fibroblast sources occurred, no formation of the specific

α

A (3/4) cleavage prod-ucts of type I collagen occurred after pre-treatment of MMP-1 with clodronate attherapeutically attainable levels.

30

Enzymatic activities of MMP-9 and MMP-2 arealso inhibited by clodronate (T

ABLE

1).Pamidronate and alendronate, each in the concentration range of 250 to 1000 µM,

dose-dependently inhibited the degradation of 21-kDa β-casein by human recombi-nant MMP-13 (TABLE 1).39 We further found that alendronate and pamidronate canalso inhibit human recombinant stromelysin (MMP-3),39 which has been closely re-lated to various malignancies. Additionally, pamidronate and alendronate inhibit hu-man MMP-1, -2, -8, -9, -12, and MMP-20 at corresponding concentrations(TABLE 1).39 Zolendronate inhibited MMP-3, -12, -13, and -20 dose-dependently.

Clodronate, in the concentration range of 200 to 2000 µM, inhibited the activity ofthe catalytic domain of human recombinant MT1-MMP (TABLE 1).39 The IC50 wasextrapolated to be 80–120 µM.

Zymography showed that, PMA-induced MG-63 and U2-OS cell culture mediacell lines contained MMP-2 in both proform (72 kDa) and proteolytically activatedform (68 kDa). When the concentration of clodronate was increased, the active (68kDa) in relation to the proform (72 kDa) decreased significantly (TABLE 1; FIG. 1Aand B).

Immunologic staining of the MG-63 monolayer in glass slides showed clear de-crease in the amount of pericellular MT1-MMP immunoreactivity. Further, Northernblotting showed reduced MT1-MMP mRNA in clodronate-treated PMA-inducedHT1080 fibrosarcoma cells compared with non–clodronate-treated ones (data notshown).

When cultured on Matrigel, human melanoma C8161 and human fibrosarcomaHT1080 cell lines efficiently invade the Matrigel membrane. Invasion assays showedthat, in the presence of human normal serum, bisphosphonates (clodronate, pamidr-onate, alendronate, nedrinate, and clodrinate) constantly and dose-dependent reducedthe invasion of these malignant cell lines; the IC50s for bisphosphonate-induced re-ductions of in vitro invasion of human malignant cell lines studied were 30–80 µM(FIG. 2).


FIGURE 1B. Clodronate decreases MMP-2 conversion in U2-OS osteosarcoma cellculture media. Seven parallel experiments of the percentage of the 68-kDa form of totalMMP-2 activities in zymography were calculated by scanning. Block graph shows the meanand standard deviation of these calculations.

FIGURE 1A. Zymogram of PMA-induced U2-OS osteosarcoma cell culture media.72-kDa gelatinolytic band represents proMMP-2 and 68-kDa its porteolytically convertedfrom; both are confirmed to be MMP-2 by Western blotting. Clodronate (concentrations in-dicated below lanes) dose-dependently decreases the processing of proMMP to the 68-kDaform.

461TERONEN et al.: MMP INHIBITION BY BISPHOSPHONATES

DISCUSSION

Nakaya and co-workers have independently confirmed our original findings,29,30

reporting on the inhibition of MMP-1 and MMP-3 by tilundronate.40,41 Overall, themechanism of the MMP inhibition may involve the ability of bisphosphonates to actas cation-chelators. For the MMP inhibition by tetracyclines and CMTs, the samemechanism has been suggested.2,3,32,42–44 Noteworthy is the fact that this inhibitionoccurs at a substantially low concentration (about 100 µM), obtained in targeted ar-eas by in vivo medication, and bisphosphonates are known to seek in the bone andespecially under osteoclasts.23,48 Nevertheless, the specific inhibition of distinct ma-trix metalloproteinases seems to differ with various bisphosphonates, a finding verysimilar to that observed for TCs and CMTs.3,45

From the the viewpoint of pharmacologic intervention, of interest is the predom-inant role and apparent imbalances between MMPs and tisssue inhibitors of matrixmetalloproteinases (TIMPs) in malignant and inflammatory tissue-destructive dis-eases. Finding safe MMP inhibitors for long-term clinical use has become a majorgoal in the design of optimal treatment procedures and medications to halt ongoingpathologic tissue-loss in such diseases. In addition to the classic involvement ofMMPs in cancer and metastatic tissue degradation, the role of MMPs have beenshown in recent studies to play a part in molecular steps involving the growth of theprimary tumor, in angiogenesis, in the initiation of growth at an ectopic site, and in

FIGURE 2. Block graph presents mean and standard deviation of five parallel experi-ments in invasion assay. Melanoma (C8161) and fibrosarcoma (HT1080) cells were cul-tured on Matrigel membrane. Invasion of the cells in the membrane were calculated in thepresence of different concentrations of alendronate. The amount of cells invaded were cal-culated under the microscope and compared to the identical control without alendronate as100% invasion.


the sustained growth of metastatic foci which are becoming clinically detectable tu-mors.46 Membrane-type MMPs have been under growing interest in the recent yearsand their role in malignant cell invasion has been well established owing to theirability to target proteolytic action at specific direction through activation, as haveother MMPs, at cell surface upon their secretion. The suprisingly good results in thetreatment of multiple myeloma and breast cancer with bisphosphonates may thus inpart be explained by the inhibition of MMPs, and especially inhibition of proMMP-2activation and MT1-MMP activity/expression at a cellular level, as shown in our ex-periments. Furthermore, the invasive potential of the several different types of hu-man malignant cell lines studied could be downregulated by bisphosphonates.

The pharmacologic MMP inhibition can be beneficial in arthritides, in variouskinds of ulceration, several skin diseases, in oral aphtous lesions, in periodontal dis-eases, dental peri-implantitis, in jaw cysts, and in prevention of the loosening of totalhip replacement (and other) prostheses/implants. The impact of such a clinically safeinhibitor in the long-term treatment of these soft- and hard-tissue-destructive diseas-es would be immense. The increasing interest in MMP inhibitors has led to the de-velopment of various new active-site-targeted peptide inhibitors (anti-MMPpeptidomimietics) in addition to the previously known tetracyclines and their non-antimicrobial CMT-derivatives.47 Despite the promising prospects of these effectivepeptidomimetic inhibitors, they are burdened with long testing periods prior to pos-sible approval for their use clinically.10 It is noteworthy that these anticollagenolyticpeptidomimetics are very efficient in vitro MMP inhibitors at pico- and nanomolarlevels, but their beneficial anti-tissue-destructive effects in vivo have been obtainedat micromolar serum levels, thus, making them comparable to bisphosphonates andtetracyclines.2,3,10,48

Bisphosphonates have been shown to be beneficial in the treatment of bone de-struction in malignancies with increased metastatic potential as well as in thetreatment of inflammatory tissue-destructive diseases in humans.11–13,49,50 The mo-lecular mechanism(s) of the ability of bisphosphonates to prevent and inhibit tissuedestruction and cancer dissemination have, however, remained unclear.24 Most re-cently, it has been shown that alendronate reverses the ability of TIMP-2 to inhibitMMP-2 from degradation by plasmin in vitro.51 The inhibition of MMPs by bispho-sphonates should be considered potentially one of the main mechanisms explaining,at least partially, their antitumor effects, such as reduction of the actual tumor burdenand the spread not only to bony tissues but also to soft tissues.49 In fact, recent stud-ies by Ramamurthy et al.52 show that in the endotoxin (LPS)-induced rat periodon-titis model, when compared to single use of doxycycline or bisphosphonate,combination treatment with these broad-spectrum MMP-inhibitors (doxycycline/CMTs with bisphosphonate) lead in synergistically enhanced reduction of periodon-tal soft- and bone-tissue destruction, associated with inhibition and downregulationof gingival tissue MMPs and serine proteinases.52

SUMMARY

Bisphosphonates have the ability to inhibit MMPs at several levels starting fromexpression of MMPs to cellular events known to be related to MMPs’ actions. New


findings show the novel potential of bisphosphonates to inhibit major tissue-destruc-tive enzymes and may help to clarify the beneficial molecular mechanisms of theseagents in inhibiting and preventing hard- and soft-tissue destruction. These observa-tions can be useful in future development of drugs as clinically relevant MMP inhib-itors/downregulators alone and/or in different combinations in the treatment ofdiseases involving pathologic excessive extracellular matrix degradation.

Bisphosphonates are thus among the prime candidates for pharmacologic use asMMP inhibitors locally or systemically for conditions such as periodontal disease,dental peri-implantitis, rheumatoid arthritis and other arthitides, loosening of totalhip replacement prostheses, several skin diseases, oral aphtous ulcers, and even forcancer. Bisphosphonates which are already well tolerated in humans, may thus soonbe used in clinical medicine, sooner than will the use of inhibitors currently at thedevelopmental stage. Having low toxicity and proven to be well tolerated after sev-eral years in human use, bisphosphonates have the potential to become one of themain MMP inhibitors for MMP related human diseases in the near future.

ACKNOWLEDGMENTS

This study was supported by grants from the Finnish Dental Society, theAcademy of Finland, and by the EVO Clinical Research Grant (TKILO19).

REFERENCES

1. GROSS, J. & C.M. LAPIÉRE. 1962. Collagenolytic activity in amphibian tissue cultureassay. Proc. Natl. Acad. Sci. USA 54: 1197–1204.

2. GOLUB, L.M. et al. 1983. Minocycline reduces gingival collagenolytic activity duringdiabetes. Preliminary observations and a proposed new mechanism of action. J. Peri-odont. Res. 18: 516–526.

3. GOLUB, L.M. et al. 1992. Host modulation with tetracyclines and their chemicallymodified analogues. Curr. Opin. Dent. 2: 80–90.

4. GOLUB, L.M. et al. 1997. A matrix metalloproteinase inhibitor reduces bone-type col-lagen degradation fragments and specific collagenases in gingival crevicular fluidduring adult periodontitis. Inflamm. Res. 46: 310–319.

5. INGMAN, T. et al. 1993. Tetracycline inhibition and the cellular source of collagenase ingingival crevicular fluid in different periodontal diseases. J. Periodontol. 64: 82–88.

6. LAUHIO, A. et al. 1994. Reduction of matrix metalloproteinase 8-neutrophil collage-nase levels during long-term doxycycline treatment of reactive arthritis. Antimicrob.Agents Chemother. 38: 400–402.

7. LAUHIO, A. et al. 1994. In vivo inhibition of human neutrophil collagenase (MMP-8)activity during long-term combination therapy of doxycycline and NSAID in acutereactive arthritis. Clin. Exp. Immunol. 98: 21–28.

8. LAUHIO, A. et al. 1995. Tetracyclines in the treatment of rheumatoid arthritis. Lancet.340: 645–646.

9. NORDSTRÖM, D. et al. 1998. Anti-collagenolytic mechanism of action of doxycyclinetreatment in rheumatoid arthritis. Rheumatol. Int. 5: 175–180.

10. BROWN, P.D. & R. GIAVAZZI. 1995. Matrix metalloproteinase inhibition: a review ofanti-tumour activity. Ann. Oncol. 6: 967–974.

11. DELMAS, P.D. 1996. Bisphosphonates in the treatment of bone diseases. N. Engl. J.Med. 335: 1836–1837.


12. HORTOBAGYI, G.N. et al. 1996. Efficacy of pamidronate in reducing skeletal complica-tions in patients with breast cancer and lytic bone metastases. Protocol 19 ArediaBreast Cancer Study Group. N. Engl. J. Med. 335: 1785–1791.

13. BANKHEAD, C. 1997. Bisphosphonates spearhead new approach to treating bonemetastases. J. Natl. Cancer Inst. 89: 115–116.

14. DIEL, I.J. et al. 1998. Reduction in new metastases in breast cancer with adjuvant clo-dronate treatment. N. Engl. J. Med. 339: 357–363.

15. VAN DER PLUIJM, G. et al. 1996. Bisphosphonates inhibit the adhesion of breast cancercells to bone matrices in vitro. J. Clin. Invest. 98: 698–705.

16. IGARASHI, K. et al. 1996. Inhibitory effect of the topical administration of a bisphos-phonate (risedronate) on root resorption incident to orthodontic tooth movement inrats. J. Dent. Res. 75: 1644–1649.

17. SASAKI, A. et al. 1995. Bisphosphonate risendronate reduces metastatic human breastcancer burden in bone in nude mice. Cancer Res. 55: 3551–3557.

18. YONEDA, T. et al. 1997. Inhibition of osteolytic bone metastasis of breast cancer bycombined treatment with the bisphosphonate ibandronate and tissue inhibitor of thematrix metalloproteinase-2. J. Clin. Invest. 99: 2509–2517.

19. JEFFCOAT, M.K. & M.S. REDDY. 1996. Alveolar bone loss and osteoporosis: Evidencefor a common mode of therapy using the bisphosphonate alendronate. In: BiologicalMechanisms of Tooth Movement and Craniofacial Adaptation. Z. Davidowitch &L.A. Norton, Eds.: 365–373. Harvard Society for Advancement of Orthodontics,Boston, MA.

20. REDDY, M.S. et al. 1995. Alendronate treatment of naturally occuring periodontitis inbeagle dogs. J. Periodontol. 66: 211–217.

21. WEINREB, M. et al. 1994. Histomorphometrical analysis of the effects of the bisphos-phonate alendronate on bone loss caused by experimental periodontitis in monkeys.J. Periodont. Res. 29: 35–40.

22. FLEISCH, H. et al. 1991. Bisphosphonates. Pharmacology and use in the treatment oftumour-induced hypercalcaemic and metastatic bone diseases. Drugs. 42: 919–944.

23. FLEISCH, H. et al. 1997. Bisphosphonates in bone disease. From laboratory to thepatient. The Parthenon Publishing Group. New York.

24. RODAN, G.A. & H.A. FLEISCH. 1996. Bisphosphonates: mechanisms of action. J. Clin.Invest. 97: 2692–2696.

25. FELIX, R. et al. 1976. The effect of several diphosphonates on acid phosphohydrolasesand other lysosomal enzymes. Biochim. Biophys. Acta 429: 429–438.

26. SKOREY, K. et al. 1997. How does alendronate inhibit protein-tyrosine phosphatases?J. Biol. Chem. 272: 22472–22480.

27. SMITH, A. et al. 1996. Protein-tyrosine phosphatase activity regulates osteoclast for-mation and function: inhibition by alendronate. Proc. Natl. Acad. Sci. USA 93:3068–3073.

28. TERONEN, O. et al. 1995. Characterization of interstitial collagenases in jaw cyst walls.Eur. J. Oral Sci. 103: 141–147.

29. TERONEN, O. et al. 1997. Inhibition of matrix metalloproteinase-1 by dichloromethyl-ene bisphosphonate (clodronate). Calcif. Tissue Int. 61: 59–61.

30. TERONEN, O. et al. 1997. Human neutrophil collagenase MMP-8 in peri-implant sulcusfluid and its inhibition by clodronate. J. Dent. Res. 76: 1529–1537.

31. SORSA, T. et al. 1994. Effects of tetracyclines on neutrophil, gingival, and salivary col-lagenases. A functional and western-blot assessment with special reference to theircellular sources in periodontal diseases. Ann. N.Y. Acad. Sci. 732: 112–131.

32. SORSA, T. et al. 1997. Activation of type IV procollagenenase by human tumor-associ-ated trypsin-2. J. Biol. Chem. 272: 21067–21074.


33. TERONEN, O. et al. 1995. Identification and characterization of gelatinases/type IV col-lagenases in jaw cysts. J. Oral Pathol. Med. 24: 78–84.

34. DECARLO, A. 1994. Matrix metalloproteinase activation and induction in keratinocytesby a purified thiol-proteinase from Porphyromonas gingivalis. PhD thesis, Univer-sity of Alabama at Birmingham, Alabama, USA.

35. LICHTE, A. et al. 1996. The recombinant catalytic domain of membrane-type matrixmetalloproteinase-1 (MT1-MMP) induces activation of progelatinase A and proge-latinase A complexed with TIMP-2. FEBS Lett. 397: 277–282.

36. VERHEIJEN, J.H. et al. 1997. Modified proenzymes as artificial substrates for pro-teolytic enzymes: colorimetric assay of bacterial collagenase and matrix metallopro-teinase activity using modified pro-urokinase. Biochem J. 323: 603–609.

37. RISTELI, L. & J. RISTELI. 1987. Analysis of extracellular matrix proteins in biologicalfluids. Meth. Enzymol. 145: 391–411.

38. MÄKELÄ, M. et al. 1994. Matrix metalloproteinases (MMP-2 and MMP-9) of oral cav-ity: cellular origin and relationship to periodontal status. J. Dent. Res. 73: 1397–1406.

39. TERONEN, O. 1998. Jaw cyst matrix metalloproteinases (MMPs) and inhibition ofMMPs by bisphosphonates. PhD thesis, University of Helsinki, Helsinki, Finland.

40. NAKAYA, H. et al. 1997. Effects of bisphosphonate on MMP-1 activity and mRNAexpression. J. Dent. Res. 76: A2725.

41. NAKAYA, H. et al. 1997. Effects of bisphosphonates on MMP-3 activity and mRNAexpression [abstract]. Annual meeting of the American Academy of Periodontology.

42. GOLUB, L.M. et al. 1985. Further evidence that tetracyclines inhibit collagenase activ-ity in human crevicular fluid and from other mammalian sources. J. Periodont. Res.20: 12–23.

43. RYAN, M.E. et al. 1996. Matrix metalloproteinases and their inhibition in periodontaltreatment. Curr. Opin. Periodont. 3: 85–96.

44. JONAT, C. et al. 1996. Transcriptional downregulation of stromelysin by tetracycline.J. Cell. Biochem. 60: 341–347.

45. SUOMALAINEN, K. et al. 1992. Tetracycline inhibition identifies the cellular origin ofinterstitial collagenases in human periodontal diseases in vivo. Oral Microbiol.Immunol. 7: 121–123.

46. CHAMBERS, A.F. & L.M. MATRISIAN. 1997. Changing views of the role of matrix met-alloproteinases in metastasis. J. Natl. Cancer Inst. 89: 1260–1270.

47. SEFTOR, R.E. et al. 1998. Chemically modified tetracyclines inhibit human melanomacell invasion and metastasis. Clin. Exp. Metastasis. 16: 217–225.

48. SATO, M. et al. 1991. Bisphosphonate action. Alendronate localization in rat bone andeffects on osteoclast ultrastructure. J. Clin. Invest. 88: 2095–2105.

49. SIRIS, E.S. 1997. Breast cancer and osteolytic metastases: Can bisphosphonates help?Nat. Med. 2: 151–152.

50. DELMAS, P.D. 1996. Bisphosphonates in the treatment of bone diseases. N. Engl. J.Med. 335: 1836–1837.

51. FARINA, A.R. et al. 1998. Tissue inhibitor of metalloproteinase-2 protection of matrixmetalloproteinase-2 from degradation by plasmin is reversed by divalent cation che-lator EDTA and the bisphosphonate alendronate. Cancer Res. 58: 2957–2960.

52. RAMAMURTHY, N.S. et al. 1999. A suboptimal CMT-8 clodronate combination therapyinhibits alveolar bone resorption in LPS induced periodontitis in rats. Ann. N.Y.Acad. Sci. This volume.

53. HANEMAAIJER, R. et al. 1997. Matrix metalloproteinase-8 is expressed by rheumatoidsynovial fibroblasts and endothelial cells. J. Biol. Chem. 272: 31504–31509.

MMP Inhibition and Downregulation by Bisphosphonates

Documents