increased incidence of matrix metalloproteinases in urine ... · zymogram of a urine specimen...
TRANSCRIPT
(CANCER RESEARCH 58, 1395-1399. April 1. 1998)
Advances in Brief
Increased Incidence of Matrix Metalloproteinases in Urine of Cancer Patients1
Marsha A. Moses,2 Dmitri Wiederschain, Kevin R. Loughlin, David Zurakowski, Carolyn C. Lamb, and
Michael R. FreemanLaboratory for Surgical Research ¡M.A. M.. D. W.¡,The Urologie Laboratory, Department of Urology [M. R. F.j, and Department of Research Computing ID. Z.]. Children'sHospital, Boston, Massachusetts 02115; Division of Urológica! Surgen; Brigham and Women's Hospital, Boston, Massachusetts 02115 ¡K.R.Lj: Department of Radiation
Oncology, Beth Israel Hospital. Boston, Massachusetts 02115 1C. C. L.]: and Joint Center for Radiation Therapy and Departments of Radiology ¡C.C. L. D. Z.¡and Surgery¡M.A. M., K. R. L. M. R. F.], Hart-ard Medical School, Boston. Massachusetts 02115
Abstract
Matrix metalloproteinases (MMPs) have been implicated in mechanisms of metastasis in experimental cancer models and in human malignancies. In this study, we used substrate gel electrophoresis (zymography)to determine the frequency of detection of MMPs in urine of patients witha variety of cancers. Three molecular weight classes of urinary MMPs, .*/,.72,000, \/, 92,000, and high molecular weight (,V/,.£150,000) species, weredetected reproducibly and correlated with disease status. The Vi, 72,000and .)/,. 92,000 species were identified as MMP-2 and MMP-9, respec
tively, by Western blot analysis. The presence of biologically activeMMP-2 (P < 0.001) or MMP-9 (P = 0.002) was an independent predictorof organ-confined cancer, and the high molecular weight species(/' < 0.001) was an independent predictor of metastatic cancer. This is the
first study to demonstrate that analysis of urinary MMPs may be useful indetermining disease status in a variety of human cancers, both within andoutside of the urinary tract.
Introduction
Tumor cells metastasize to distant sites by disassembling the complex extracellular matrices that delimit tissue spaces and surroundblood vessels. A specialized class of matrix-degrading enzymes, theMMPs,3 has been implicated in metastasis in experimental models and
in human tumors. Elevated production of MMP enzymes by cancercells and tumor stroma correlates with the malignant or metastaticphenotype. Overproduction of MMPs by a tumor communicating withthe vascular and lymphatic systems might result in increased levels ofMMP activity in other body fluids, such as blood or urine. Thispossibility is consistent with previous demonstrations that levels ofother regulatory molecules overproduced by tumors and measured inbody fluids of cancer patients, such as the angiogenic peptide basicfibroblast growth factor, have been shown to be independent predictors of disease status (1, 2).
Elevated levels of MMPs have been found in serum and plasma ofanimals bearing experimental tumors and in human patients. Serumand plasma levels of the Mr 92,000 collagenase (gelatinase B,MMP-9) were increased in rats bearing metastatic mammary carcino
mas (3). In humans, serum levels of the MT72,000 collagenase (gelatinase A, MMP-2) have been correlated with lung cancer metastasis
and have been seen to decline in response to combination chemotherapy (4). Serum levels of interstitial collagenase (MMP-1) have been
reported to be increased in patients with metastatic prostate cancer (5),
Received 1/19/98; accepted 2/18/98.The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported by grants from CaPCURE (to M. A. M. and M. R. F.) and
the Cancer Research Foundation of America.2 To whom requests for reprints should be addressed, at Enders Research Laboratories,
Children's Hospital, 300 Longwood Avenue, Boston, MA 02115. Fax: (617)355-7043;
E-mail: [email protected] The abbreviations used are: MMP, matrix metalloproteinase; NED, no evidence of
disease; CI, confidence interval; hMW, high molecular weight; TCC, transitional cellcarcinoma.
and gelatinase B (MMP-2) levels have been shown to be increased in
serum of patients with colon and breast cancer (6). Recently, Gohji etal. (I, 8) reported that the ratio of serum MMP-2 to tissue inhibitor ofmetalloproteinase 2 and serum levels of MMP-2 and MMP-3 (strome-lysin-1) correlated positively with recurrence of urothelial cancer
following complete tumor resection. These data are consistent withthe conclusion that MMPs are important mediators of cancer progression and that MMP levels are frequently elevated in predictable waysin body fluids of cancer patients.
There has been relatively little study of the possibility that MMPanalysis of human urine might provide important clinical information.In light of the potential of clinical urinary analysis for cancer diagnosis and prognosis, we addressed three questions in this study, (a)Can MMPs be detected in human urine in their intact, biologicallyactive forms? (b) Does the detection of urinary MMPs correlate withdisease status in cancer patients? (c) If such a correlation can bedemonstrated, does it apply to cancers outside of the urinary tract?
Materials and Methods
Patient Urine Collection. Patients were chosen for the study, their urinewas collected, and their clinical status was verified and monitored by aurologist (K. R. L.) and a radiation oncologist (C. C. L.). Urine was collectedaccording to the institutional bioethical guidelines pertaining to discardedclinical material. Urine samples were collected in one of two ways. Patients
seen in an ambulatory care setting provided a voided specimen; those whowere undergoing cystoscopic or surgical procedures were catheterized prior tothe planned procedure. All patients in the NED group were treated by surgicalresection. Prior to analysis, urine samples were tested for the presence of bloodusing Ames Multistix 7 reagent strips (Miles, Elkhart. IN), and specimenscontaining blood were not analyzed.
Sample Preparation and Substrate Gel Electrophoresis. Samples werefrozen immediately after collection and stored frozen (-20°C) until assay. The
samples were thawed, and creatinine concentrations were determined using acommercial kit (Sigma Chemical Co., St. Louis, MO) according to manufacturer's instructions. Aliquots (1 ml) of each sample were dialyzed against
double-distilled water in 18-mm dialysis tubing [Spectra/Por membrane, molecular weight cut-off (MWCO): M, 3500; Spectrum, Houston. TX] with two
exchanges of 4 liters each (per 20 urine samples). Following dialysis, urinesamples were centrifugad at 4000 rpm for 5 min at 4°C,and the supernatants
were collected. Urine samples (30 /ul) were mixed with buffer consisting of 4%SDS, 0.15 M Tris (pH 6.8), 20% (v/v) glycerol, and 0.5% (w/v) bromphenolblue. Samples were applied, without boiling, into wells of a 4% acrylamideLaemmli stacking gel/10% SDS-acrylamide separating gel containing 0.1%
(w/v) gelatin (Life Technologies, Inc., Detroit, MI) on a mini-gel apparatus as
described previously (9, 10). Gels were run at 15 mA/gel during stacking and
at 20 mA/gel during the resolving phase at room temperature. After electrophoresis, the gels were soaked in 2.5% Triton X-100 with gentle shaking for
30 min at ambient temperature with one change of detergent solution. The gelswere then rinsed and incubated overnight at 37°Cin substrate buffer [50 HIM
Tris-HCI buffer (pH 8), 5 mM CaCl2, and 0.02% NaN,]. After incubation, gelswere stained for 15-30 min in 0.5% Coomassie Blue R-250 in acetic acid,
isopropyl alcohol, and water (1:3:6): destained in acetic acid, ethanol, andwater (1:3:6); photographed; and dried for permanent record. Both proenzyme
1395
Research. on December 2, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from
MMPS IN URINE
and activated proteinases appeared as zones of substrate clearing. DifferentMMPs were distinguished from each other on the basis of their molecularweights. The identity of these MMPs was confirmed by Western blot analysisusing anti-MMP antibodies (Oncogene Science, Cambridge, MA). Antibodies
purchased from Biogenesis Ltd. (Sandown, NH) and Chemicon International,Inc. (Temecula, CA) were also tested. In all cases. anti-MMP-9 antibodiescross-reacted with MMP-2. As positive controls, purified human MMP-2 and
MMP-9 (Biogenesis Ltd.) were subjected to gel electrophoresis as describedabove. To verify that the proteolytic activities detected were metal-dependent
proteinases. samples were subjected to incubation in substrate buffer in thepresence of 1,10-phenanthroline. a MMP inhibitor.
Data Collection and Analysis. Zymograms were processed and evaluatedindependently by two investigators who had no knowledge of the clinicalstatus of the individuals from whom the urine specimens were obtained.
Statistical Analysis. MMPs detected in the urine of patients with organ-
confined disease and with metastatic cancer were compared to the normal/NED control groups. Sensitivity and specificity were calculated using standardformulas and were expressed as percentages. Ninety-five % CIs were derivedusing Pratt's method (11). The likelihood ratio for a positive test result was
determined as the fraction of true-positives divided by the fraction of false-positives [sensitivity/I 1 —¿�specificity)] to provide an indicator of the discrim
inating power for each MMP (12). Stepwise logistic regression was used toestablish the independent predictors of cancer and to estimate the probabilityfor combinations of the MMP markers in the final multivariate model (13).One-way ANOVA was performed to assess differences in creatinine levels
among the groups, with a Bonferroni correction for multiple comparisons.Fisher's exact test was used for comparison of proportions. All statistical tests
were conducted using a two-sided a level of 0.05. Data analysis was performed
using SAS for Windows (Version 6.11 ; SAS Institute Inc., Cary, NC).
Results
One hundred-seventeen urine specimens obtained at random sampling times were analyzed. Sixty-eight specimens were obtained fromindividuals with prostate, renal, bladder, breast, and a variety of othertypes of cancer (Table 1). Specimens taken from patients with organ-confined cancers were obtained prior to surgical or other therapeuticintervention. Gleason sum scores of the prostate adenocarcinomasranged from 5 to 9, with 26 of 28 (93%) of the prostate adenocarcinomas scored as Gleason grades 5-7. Prostate cancer patients had anaverage serum prostate-specific antigen value of 12.4 ±12.5 ng/ml(mean ±SD) taken near the time of sampling. The bladder cancergroup included grades I-III. The renal cancer group included grades1I-III. Twenty-one patients had metastatic cancer at the time ofsampling. All of the breast cancer patients had metastatic disease.Thirty urine specimens were obtained from healthy male and femalevolunteers (normals), and 19 samples were obtained from formercancer patients (male and female) showing NED at the time ofsampling.
Thirty /u.1of each urine sample were electrophoresed through gelatin-containing SDS-polyacrylamide gels and evaluated for the presence of zones of proteolytic activity after incubation of the gels inrenaturing conditions. Representative results are shown in Fig. IA(Lane I). There was an association between the patients' disease
250 —¿�
98 —¿�64 —¿�;
B
100—75—
so-
so—se-
so
ie"
1 2 3
250'
98-64'
50-
36"
30-
16'
Fig. l. A, substrate gel electrophoresis of MMPs in urine from a patient with metastaticcancer (Lune /). Gelatinase activity was detected at A/rs of approximately 72.(KK).92,000,and 150,000. Lanes 2 and 3, results of Western blot analysis of the same specimen usingantibodies against human MMP-2 (Lane 2) and MMP-9 (Lane 3). B. gelatinase activity ofMMP-2 (Lane I] and MMP-9 (Lane 2), demonstrated by gelatin zymography. Lane 3.
zymogram of a urine specimen containing the hMW, Mf 92,000, and Mr 72,000 species,which was incubated in substrate buffer containing I.IO-phenanthroline, a MMP inhibitor.
status (i.e., presence of cancer) and the presence of three gelatinaseactivities, a hMW form (greater than or equal to A/r 150,000), a A/r92.000 form, and a M, 72.000 species (Table 2). The frequencies ofdetection of one or more of these three urinary MMP species in eachexperimental group were: normal plus NED, 8 of 49 (16%); NED, 2of 19 (11%); normal, 6 of 30 (20%); cancer, 48 of 68 (71%); andmetastatic cancer, 19 of 21 (90%). The frequency of detection of atleast one MMP species was significantly higher for both the cancergroup and the metastatic cancer group, compared to each controlgroup (P < 0.001). There were no significant differences between the
Table 1 Patient¡xi/tulatiimNo.
of patientsMean age (yr)*
Female/maleDetectableprimary tumors
Detectable métastasesProstate28
63 ±6.40/28282Renal10
61 ±114/6103Cancer
typeBladder1064
±162/8100Breast9
49 ±119/099Other11 49 ±216/5117NED"19 68 ±6.96/13NA'
NANormal30
32 ±6.08/22NANA
" Undetectable disease.'' Values are expressed as mean ±SD.' NA, not applicable.
1396
Research. on December 2, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from
MMPS IN URINE
Table 2 Gelatinose profile
Specimens positive for any MMP(%)AllspecimensDetectable
primarytumorsDetectablemétastasesSpecimens
positive for a specific MMP(%)hMWM,
92.000M,72,000Creatinine
(mg/dl) 'Prostate757550576439131 ±70Renal404067303030134±78Cancer
typeBladder8080NA70703091
±52Breast1001001001001001085±42Other64648664643669±38Metastatic90909086812997±60NEDIINA"NA"511090±47Normal20NANA18241188±76" NA. not applicable.
Values are expressed in mean ±SD.
two control groups (P = 0.46). The difference in detection of at leastone MMP was higher for metastatic cancers than it was for organ-
confined cancers, although this did not reach statistical significance(P = 0.08). The detection frequencies of one or more of these three
MMPs were significantly higher for prostate (21 of 28, 75%), bladder(8 of 10, 80%), and breast (9 of 9, 100%) cancers as compared to thecontrol groups (all P < 0.001).
In all specimens examined, four molecular weight classes of gela-tinase-degrading activity were frequently observed: the hMW class
was seen in 48 (41%) of the samples; the Mr 92,000 form was seen in52 (44%); the Mr 72,000 form was seen in 23 (20%); and a Air 20,000form was seen in 34 (29%). Gelatinases with Mrs of 100,000-116,000
and 45,000 were detected in 4 (3.4%) and 5 (4.3%) of the specimens,respectively. A gelatinase of approximately 125,000 was detected inthe urine of five of the specimens obtained from metastatic breastcancer patients (66%) but not in any of the other urine specimens.Zymographic profiles were identical in repeat assays of the same urinesamples. Enzymes remained biologically active after two to threefreeze/thaw cycles. However, significant degradation of the intactspecies was evident after storage for over 24 h at 4°C.
MMP-2 (gelatinase A) and MMP-9 (gelatinase B) are the predom
inant MMP species detectable by gelatin zymography (14). The M,72,000 and Mr 92,000 activities were recognized by anti-MMP antibodies in Western blots, confirming their identities as MMP-2 andMMP-9 (Fig. 1A, Lanes 2 and 3). The identities of the hMW and the
Mr 20,000 gelatinases are unknown. There was no correlation betweenpresence of the A/r 20,000 species and the clinical status of thesubjects. Representation of the Mr 100,000-116,000 species and the
Mr 45,000 species was too small for statistical analysis. Incubation ofthe substrate gels in 10 HIM 1,10-phenanthroline, a metalloproteinase
inhibitor, resulted in loss of proteolytic activity, verifying that theurine proteinases are members of the metalloproteinase family (Fig.IB, Lane 3; Ref. 9).
The hMW, M, 92,000, and M, 72,000 gelatinases were each significant multivariate (independent) predictors of disease status. Step-wise logistic regression indicated that the M, 92,000 (P = 0.002) and
the Mr 72,000 (P < 0.001) gelatinases were each independent predictors of cancer (Table 3). Each of these two enzymes was also anindependent predictor of the subgroup of organ-confined cancers,
Table 3 Multivariate predictors of disease stains
MarkerMetastatic
cancershMwOrgan-confined
cancersM,92,000M,
72,000AllcancersM,92,000M,
72,000ß
coefficient3.41.572.581.892.5SE0.710.521.000.481.07Oddsratio30.54.813.16.612.395%CI7.7-120.11.7-13.33.6-93.22.6-17.01.7-87.3P<0.001<0.01<0.010.002<0.001
indicating that the Mr 92,000 and Mr 72,000 species provided the mostuseful information for discriminating between patients with localizeddisease from those who are normal or show no evidence of cancer.The hMW gelatinase, but not the Mr 92,000 or M, 72,000 species, wasa multivariate predictor of metastatic cancer (P < 0.001). Remarkably, the estimated odds for metastatic cancer when the hMW specieswas present in urine was approximately 30 times greater than whenthe marker was absent (95% CI = 7.7-120.1). This model estimates
that the probability of metastatic cancer is 74% when the hMWspecies is detected and 8% when it is not detected. For organ-confined
cancers, the probability of cancer is 96% when both the MT92,000 andMr 72,000 gelatinases are detected and 29% when neither is detected.In fact, the probability of organ-confined cancer was calculated to be
over 50% when at least one of these gelatinases was detected.The sensitivity, specificity, and likelihood ratios for the hMW, Mr
92,000, and Mr 72,000 enzymes for detection of metastatic cancer,organ-confined cancer, and all cancers (metastatic plus organ-con
fined) are shown in Table 4. Specificity values were highest for the Mr72,000 gelatinase, which correctly classified all but one of the 49normal/NED patients as being cancer-free. Given that the Mr 72,000
and the Mr 92,000 species alone were multivariate predictors ofdisease status for organ-confined cancers and for outcomes of all
cancers, we analyzed the statistical performance characteristics ofthese two markers in combination (Table 4). This analysis resulted inan increase in sensitivity in comparison to the evaluation of eachmarker alone. Calculation of likelihood ratios indicated that each ofthe hMW, M, 92,000, and Mr 72,000 gelatinases had significantdiscriminatory power when samples were grouped into metastaticcancer, organ-confined cancer, or all cancer categories.
There was no association between MMP detection in the zymo-
grams and urinary creatinine concentration (Table 2), indicating thatthe presence of gelatinase activity in urine is unlikely to be a result ofvariations in solute concentration or functional status of the kidney.
Discussion
In this study, we report that three MMPs are independent predictorsof disease status for patients with a variety of cancers. This is the firstreport demonstrating that: (a) MMPs in urine correlate with thepresence of malignant disease; (b) MMP-2 and MMP-9 are present inurine in their intact and functional forms; (e) MMP-2/MMP-9 and ahMW MMP species can serve as independent predictors of organ-
confined cancer and metastatic cancer, respectively; and (d) thesecorrelations are not confined to tumors within the genitourinary tract.
In our analysis of 117 urine specimens, we detected three predominant molecular weight classes of gelatinases (Mr 150,000. Mr 92,000.and Mr 72,000) in urine by substrate gel electrophoresis. The Mr72,000 and M, 92,000 species were identified as MMP-2 (gelatinaseA) and MMP-9 (gelatinase B) by Western blot analysis. The identities
of the hMW class, the M, 125,000 MMP species detected only in the
1397
Research. on December 2, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from
MMPS IN URINE
Table 4 Statistical performance characteristics for urine MMPs"
MarkerMetaslatic
cancershMWM,
92,000M,72,000Organ-confined
cancershMWM,
92,000M,72,000M,
92,000/72,000AllcancershMWM,
92,000M,72.000M,
92,000/72.000Sensitivity
(95%CI)81.0(58.1-94.6)76.2(52.8-91.8)28.6(11.2-52.2)53.2(38.1-67.9)59.6
(44.3-73.6)34.0(20.9-49.3)66.0(50.7-79.1)61.8(49.2-73.3)64.7
(52.2-75.9)32.4(21.5^14.8)70.6(58.3-81.5)Specificity
(95%CI)87.8
(75.2-95.4)83.7(70.3-92.7)98.0(89.2-99.9)87.8
(75.2-95.4)83.7(70.3-92.7)98.0(89.2-99.9)83.7(70.3-92.7)87.8
(75.2-95.4)83.7(70.3-92.7)98.0(89.2-99.9)83.7
(70.3-92.7)Positive
LR*2.82.06.04.23.516.03.96.05.522.06.0
" These values are based on the ability of each MMP to discriminate between the
normal and NED samples (n = 49) versus metastatic cancers (n = 21), organ-confinedcancers (n = 47). and all cancers (n = 68). The results of the analysis of the combination
of the M, 92,000 and M, 72,000 species are given because these markers were shown tobe multivariate predictors for all cancers and for organ-confined cancers by logistic
regression analysis, as shown in Table 3.h Likelihood ratio for a positive test result.
urine from breast cancer patients, and other minor species detectedinfrequently are unknown.
To evaluate potential differences in prevalence of MMPs betweenthe clinical categories shown in Table 1 and between individuals withcancer and the control groups (normal plus NED), specimens werescored as positive or negative in the three major molecular weightclasses in a double-blind manner. There was a statistically significant
difference in the prevalence of one or more of these MMPs in thecancer group (organ-confined plus metastatic cancers) in comparison
to the control groups (Table 2). One or more of these MMPs weredetected in 71% of the specimens from the cancer groups and in16.3% of the control groups. When metastatic cancer patients wereanalyzed as a separate group, MMPs were detected in 90% of thecases. These data suggest that the incidence of urinary gelatinaseactivity detectable by zymography increases when cancer is presentand increases further when metastatic cancer is present. Our data alsosuggest that this association is true for a spectrum of human tumors.
To our knowledge, the only study published to date that addressesthe possibility of using information on urinary MMPs to assessclinical status of cancer patients was conducted by Margulies et al.(15). On the basis of an analysis of 73 specimens, these investigatorsshowed that type IV collagenase was present in human urine; however, they concluded that urinary MMPs were present largely asproteolytic fragments of intact enzymes. There was no predictivevalue in detection of MMP-2 or MMP-9 within the cancer cohort, withthe exception of a Mr 43,000 MMP-2 fragment, which statistical
analysis suggested could be used as a potential marker for TCC of thebladder but not for renal or prostate carcinoma. This study alsosuggested that MMPs may only be associated with TCC and not withcancers outside the genitourinary tract. The simplest explanation forthese results is that MMPs are released from the bladder wall at thesites of malignant growth and invasion and are subsequently degraded, either by autocatalysis or by proteolysis from intraluminalenzymes.
The most frequently detected enzymes found in our study weregelatinases migrating at approximately Mr 150,000, A/r 92,000, and Mr72,000. indicating that intact MMPs are the predominant forms inhuman urine. We did not detect the frequent laddering pattern ofproteolytic species identified by Margulies et al. (15), nor was there ademonstrable correlation between clinical status and the presence ofproteolytic activities migrating at low molecular weights. It is possiblethat the Mr 45,000 proteolytic activity that we detected in some of the
specimens corresponds to the A/r 43,000 MMP-2 fragment detected
previously in urine (15) and correlates with clinical status of TCCpatients. In our study, we did not detect this fragment in any of 11TCC specimens we examined (10 bladders and 1 kidney). It isinteresting to note that the pattern of the low molecular weight speciesof gelatinases in urine appears random, whereas the pattern seen withthe higher molecular weight species, i.e., the hMW, Mr 125,000 (inbreast cancer), Mr 92,000, and Mr 72,000 gelatinases appears to benonrandom. There is ample evidence in the literature to suggest thatthe lower molecular weight forms are most likely degradative products of the higher forms that arise by autocatalysis or some otherproteolytic mechanism (16), which may occur during handling, storage, or preparation of the specimens after collection.
Our results are consistent with previous reports indicating thatMMPs may be elevated in the serum of cancer patients. However, thefinding that intact MMPs are also present in urine and that evaluationof these enzymes provides useful information on the disease status ofpatients is a unique observation with potentially important clinicalramifications. It is also significant that this information was derivedusing small volumes (30 /xl) of urine, obtained directly from patientswith no sample concentration, few processing steps, and no expensiveequipment. The sensitivity and specificity values obtained using theurinary hMW, Mr 92,000, and Mr 72,000 enzymes as clinical indicators are comparable to, or better than, those observed for a number ofserum tumor markers currently in use to monitor disease recurrence(17-19). These data support the feasibility of using MMP enzymeassays as a first-look clinical assay or as a means to provide correl
ative and supportive information in assessing the clinical course ofdisease in cancer patients.
Interestingly, the MT72,000 enzyme demonstrated the highest specificity (98%) and highest positive likelihood ratios for organ-confined
cancers (16.0) and all cancers (22.0). The likelihood ratio, which is theratio of true-positives to false-positives, is generally considered to be
the most important statistical parameter in the consideration of theusefulness of a clinical test (20). This suggests the possibility thatanalysis of urinary MMP-2 might provide important information in a
clinical setting in which malignancy is suspected but is, as yet,undetected by other diagnostic methods.
MMP enzyme analysis might also be used in combination withother standard methods to increase the specificity and predictive valueof these measurements. Our results also suggest that analysis ofurinary MMPs by zymography in subsequent studies may provideadditional, novel information. For example, zymography detects aseries of enzyme species in a single evaluation, allowing for theinteresting possibility that certain cancers may display characteristiczymographic patterns. Consistent with this possibility is our observation of a A/r 125,000 MMP species in the breast cancer urine specimens that was not detected in any of the other pathological or controlspecimens.
In conclusion, this study suggests that the evaluation of urinaryMMPs by zymography reflects disease status in patients with organ-
confined and metastatic tumors, both within and outside the urinarytract.
Acknowledgments
We thank Drs. Judah Folkman and Robert Langer for helpful discussionsand Dr. Folkman for critical review of this manuscript.
References
1. Nguyen, M., Watanabe, H., Hudson, A. E., Richie, J. P., and Folkman, J. Elevatedlevels of the angiogenic peptide basic fibroblast growth factor ¡nurine of bladdercancer patients. J. Nati. Cancer Inst. (Bethesda), 85: 241-242, 1993.
2. Nguyen, M., Watanabe, H.. Budson. A. E., Richie, J. P., Hayes, D. F., and Folkman.J. Elevated levels of an angiogenic peptide, basic fibroblast growth factor, in the urine
1398
Research. on December 2, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from
MMPS IN URINE
of patients with a wide spectrum of cancer. J. Nati. Cancer Inst. (Bethesda), 86:356-361, 1994.
3. Nakajima, M, Welch, D. R., Wynn, D. M, Tsuruo, T., and Nicolson, G. L. Serumand plasma M, 92,000 progelatinase levels correlate with spontaneous metastasis ofrat 13762NF mammary adenocarcinoma. Cancer Res., 53: 5802-5807, 1993.
4. Garbisa. S., Scaglioni, G., Masiero, L.. Di Francesco, C.. Caenazzo, C., Onisto, M..Micela. M., Stetler-Stevenson, W. G., and Liotta, L. A. Correlation of serum metal-
loproteina.se levels with lung cancer metastasis and response to therapy. Cancer Res.,52:4548-4549, 1992.
5. Baker, T., Tickle, S., Wasan, H., Docherty, A., Isenberg, D., and Waxman, J. Serummetalloproteinases and their inhibitors: markers for malignant potential. Br. J. Cancer,70:506-512, 1994.
6. Zucker, Lysik, R. M.. Zarrabi, M. H., and Moll, U. M, 92,000 type IV collagenase isincreased in plasma of patients with colon cancer and breast cancer. Cancer Res., 53:140-146, 1993.
7. Gohji, K., Fujimoto, N., Fujii, A., Komiyama, T., Okawa, J., and Nakajima, M.Prognostic significance of circulating matrix metalloproteinase-2 to tissue inhibitor ofmetalloproteinases-2 ratio in recurrence of urothelial cancer after complete resection.Cancer Res., 56: 3196-3198. 1996.
8. Gohji, K., Fujimoto, N., Komiyama, T., Fujii, A., Okawa, J., Kamidono, S., andNakajima, M. Elevation of serum levels of matrix metalloproteinase-2 and -3 as new
predictors of recurrence in patients with urothelial carcinoma. Cancer (Phila.), 78:2379-2387, 1996.
9. Herrón, G. S., Banda, M. J.. Clark, E. J.. Gavrilovic, J., and Werb, Z. Secretion ofmetalloproteinases by stimulated capillary endothelial cells. Expression of collagenase and stromelysin activities is regulated by endogenous inhibitors. J. Biol. Chem.,261: 2814-2828, 1986.
10. Braunhut, S. J., and Moses, M. A. Retinoids modulate endothelial cell production ofmatrix-degrading proteases and TIMPs. J. Biol. Chem., 269: 13472-13479, 1994.
11. Blyth, C. R. Approximate binomial confidence limits. J. Am. Stat. Assoc., 81:843-855, 1986.
12. Weinstein. M. C., and Fineberg. H. V. (eds.). Clinical Decision Analysis, pp. 84-108.
Philadelphia: W. B. Saunders, Co., 1980.13. Breslow. N. E., and Day, N. E. Statistical Methods in Cancer Research. I. The
Analysis of Case-Control Studies, pp. 192-210. Lyon, France: IARC, 1980.14. Woessner, J. F. The family of matrix metalloproteinases. Ann. N. Y. Acad. Sci., 7.J2:
11-21, 1994.15. Margulies, I. M., Hoyhtya, M., Evans, C., Stracke, M. L., Liotta, L. A., and Stetler-
Stevenson, W. G. Urinary type IV collagenase: elevated levels are associated withbladder transitional cell carcinoma. Cancer Epidemiol. Biomark. Prev., /: 467-474,1992.
16. Howard. E. W., Bullen, E. C., and Banda, M. J. Regulation of the autoactivation ofhuman 72-kDa progelatinase by tissue inhibitor of metalloproteinase-2. J. Biol.Chem., 266: 13064-13069, 1991.
17. Prestigiacomo, A. F.. Lilja, H., Pettersson. K., Wolfert, R. L., and Slamey, T. A... Acomparison of the free fraction of serum prostate specific antigen in men with benignand cancerous prostates. J. Urol., /56: 350-354, 1996.
18. Moertel. C. G., Fleming, T. R., Macdonald, J. S., Haller, D. G., Laurie, J. A., andTangen, C. An evaluation of the carcinoembryonic antigen (CEA) test for monitoringpatients with resected colon cancer. J. Am. Med. Assoc.. 270: 943-947. 1993.
19. Berek, J. S., and Bast, R. C.. Jr. Ovarian cancer screening. The use of serialcomplementary tumor markers to improve sensitivity and specificity for early detection. Cancer (Phila.), 76: 2092-2096, 1995.
20. Greenhalgh, T. Papers that repori diagnostic or screening tests. Br. Med. J., 315:540-543.
1399
Research. on December 2, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from
1998;58:1395-1399. Cancer Res Marsha A. Moses, Dmitri Wiederschain, Kevin R. Loughlin, et al. Cancer PatientsIncreased Incidence of Matrix Metalloproteinases in Urine of
Updated version
http://cancerres.aacrjournals.org/content/58/7/1395
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://cancerres.aacrjournals.org/content/58/7/1395To request permission to re-use all or part of this article, use this link
Research. on December 2, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from