Cysteine Protease Activities and Tumor Development in ...or equal to the standard deviation of the...
7
(CANCER RESEARCH 49. 3809-3814. July 15. I989| Cysteine Protease Activities and Tumor Development in Human Colorectal Carcinoma1 Kieran Sheahan, Sania Shuja, and Mary Jo M umane2 Mallory Institute of Pathology ¡K.S., M. J. M.J, and Departments of Pathology ¡K.S., S. S., M, J. M.] and Biochemistry [M. J. M.J, Boston University School of Medicine, Boston, Massachusetts 021 IS ABSTRACT Many studies of malignant cells or tissues in culture have implicated cysteine proteases in the progression of malignancy. We have extended these observations by measuring quantitative and qualitative changes in the expression of cathepsin B-like and L-like cysteine proteases during the growth and development of human coloréela!carcinomas. Data derived from matched pairs of normal colorectal mucosa and carcinoma tissue from 27 patients demonstrated that both cathepsin B-like and cathepsin L-like specific activities were significantly elevated (/' < 0.005) in the carcinoma tissue, while levels of endogenous cysteine protease inhibitor remained constant. Correlation of cathepsin enzyme activities with different stages of colorectal cancer demonstrated significantly higher cysteine protease activities in individuals with Dukes' A tumors (tumors confined to the bowel wall) than in patients with more advanced tumors (Dukes' B, C, or D tumors) (P < 0.01-0.05). The relative proportion of activities contained in tumor epithelial and si roma Ielements remains to be elucidated. These results suggest an important role for cysteine proteases in the early progression of human colorectal carci noma. INTRODUCTION Proteases have long been known to play a role in malignancy. Their ability to digest extracellular matrix proteins provides a rationale for the contribution of proteases to tumor invasion and metastasis. In addition, many proteases are known to act as mitogens by stimulating DNA synthesis or cell division of nonprolili-niting cultured fibroblast cells (1, 2). Intracellular proteases have been implicated in the activation of hormones and growth factors (2, 3), as well as in protein turnover, and thus may contribute to the regulation of cell growth and metab olism. We have studied the expression of papain-like cysteine pro teases in human colorectal adenocarcinoma. These enzymes, known as cysteine cathepsins in mammalian tissues, are capable of degrading extracellular matrix components, including colla gen, fibronectin, proteoglycans, and elastin (4-8). Elevated acid protease activity in the extracellular milieu of malignant cells (9) has largely been traced to a cathepsin B-like cysteine pro tease (10, 11). Both intracellular and extracellular cathepsin B activity is frequently increased in cultured malignant cells. It may also be present in malignant tissues in an altered pheno- typic form as demonstrated by an abnormal isozyme pattern, increased stability to alkaline pH or high temperature (10), or enhancement in the plasma membrane and nuclei compared to normal cells (12-14). Oncogene studies have also implicated the related cysteine protease, cathepsin L, in malignancy by demonstrating that this enzyme is the major induced protein of raj-transformed Received 11/29/88; revised 3/27/89; accepted 4/17/89. 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 with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported in part by American Cancer Society Grants PDT- 349 and IN-97. 2To whom requests for reprints should be addressed, at Department of Pathology, L804, Boston University School of Medicine, 80 E. Concord St., Boston, MA 02118. NIH3T3 cells (15, 16), and the "major excreted protein" of cells transformed by Kirsten, Harvey, or Maloney sarcoma viruses and by SV40 virus (14, 17, 18). In terms of ability to degrade stromal proteins, cathepsin L is more active against collagen (19, 20) and 100-fold more active against elastin (8) than is cathepsin B. Despite this evidence from animal and cell culture models for the role of cysteine proteases in malignancy, only a few studies have attempted to correlate cathepsin activities in hu man tumor tissue with clinicopathological parameters. In one study of colorectal carcinoma, Durdey et al. (21) reported that of four proteases examined (cathepsin B, cathepsin H, and 2 collagenases), only cathepsin B activity was significantly cor related with local tumor invasion. We have investigated this relationship further by measuring cathepsin B-like and cathep sin L-like specific activities in matched tissue samples (normal mucosa and adenocarcinoma) from 27 patients. We have also measured levels of cysteine protease inhibitor in these tissues in order to determine whether cathepsin enzyme activities might reflect increases or decreases in protease inhibitor levels with malignancy. Utilizing an alkaline pH treatment we have also explored whether the cathepsin B-like activity in the cancer tissues is biochemically distinguishable from the enzyme activ ity in the normal mucosa. Finally, we have compared changes in the expression of cysteine proteases with stages of tumor development and with tumor differentiation, including muci- nous/nonmucinous histológica! subtypes. MATERIALS AND METHODS Tissue Specimens. Fresh tissue from colorectal resections was ob tained within 4 h of surgery at the Mallory Institute of Pathology, Boston, MA. To minimize possible differences due to genetic variation, we compared normal mucosa and tumor from the same patient. In sampling tumors, care was taken to avoid grossly necrotic areas and fat. Samples of normal colorectal tissue were obtained at least 10 cm from the tumor. The mucosa was then separated from the muscle layer, serosa, and surrounding fat. All samples were stored at —¿ 80"C until extraction. Normal mucosa was also obtained from non-tumor-bearing colorectal resections in a similar manner. The clinical details of the patients investigated and the corresponding pathological data are sum marized in Table 1. Dukes1 Classification. Individual tumors were staged according to the Dukes' classification (22) as modified by Turnbull et al. (23). Dukes' A tumors are confined to the bowel wall; Dukes' B tumors have spread beyond the wall without involving lymph nodes; Dukes' C tumors are associated with regional lymph node metastasis; and Dukes' D tumors are associated with distant metastasis. Tissue Extraction. To minimize test variation, care was taken to extract and assay each pair of normal mucosa/tumor samples at the same time. Tissue samples (60-80 mg) were homogenized in 500 //I of distilled, deionized water, frozen and thawed three times, and centri- fuged for 50 min at 4°Cat 12,000 rpm (17,210 x g) in a Sorvall 5B centrifuge (24). Supernatants were removed and utilized for the follow ing assays. Enzyme Assays. To distinguish cathepsin B-like enzyme activity from cathepsin L-like activity, the cathepsin B specific substrate, Z-Ala-Arg- 3809 Research. on October 7, 2020. © 1989 American Association for Cancer cancerres.aacrjournals.org Downloaded from
Transcript of Cysteine Protease Activities and Tumor Development in ...or equal to the standard deviation of the...
(CANCER RESEARCH 49. 3809-3814. July 15. I989|
Cysteine Protease Activities and Tumor Development in Human ColorectalCarcinoma1
Kieran Sheahan, Sania Shuja, and Mary Jo M umane2
Mallory Institute of Pathology ¡K.S., M. J. M.J, and Departments of Pathology ¡K.S., S. S., M, J. M.] and Biochemistry [M. J. M.J, Boston University School ofMedicine, Boston, Massachusetts 021 IS
ABSTRACT
Many studies of malignant cells or tissues in culture have implicatedcysteine proteases in the progression of malignancy. We have extendedthese observations by measuring quantitative and qualitative changes inthe expression of cathepsin B-like and L-like cysteine proteases duringthe growth and development of human coloréela!carcinomas. Dataderived from matched pairs of normal colorectal mucosa and carcinomatissue from 27 patients demonstrated that both cathepsin B-like andcathepsin L-like specific activities were significantly elevated (/' < 0.005)
in the carcinoma tissue, while levels of endogenous cysteine proteaseinhibitor remained constant. Correlation of cathepsin enzyme activitieswith different stages of colorectal cancer demonstrated significantlyhigher cysteine protease activities in individuals with Dukes' A tumors
(tumors confined to the bowel wall) than in patients with more advancedtumors (Dukes' B, C, or D tumors) (P < 0.01-0.05). The relative
proportion of activities contained in tumor epithelial and si roma Ielementsremains to be elucidated. These results suggest an important role forcysteine proteases in the early progression of human colorectal carcinoma.
INTRODUCTION
Proteases have long been known to play a role in malignancy.Their ability to digest extracellular matrix proteins provides arationale for the contribution of proteases to tumor invasionand metastasis. In addition, many proteases are known to actas mitogens by stimulating DNA synthesis or cell division ofnonprolili-niting cultured fibroblast cells (1, 2). Intracellular
proteases have been implicated in the activation of hormonesand growth factors (2, 3), as well as in protein turnover, andthus may contribute to the regulation of cell growth and metabolism.
We have studied the expression of papain-like cysteine proteases in human colorectal adenocarcinoma. These enzymes,known as cysteine cathepsins in mammalian tissues, are capableof degrading extracellular matrix components, including collagen, fibronectin, proteoglycans, and elastin (4-8). Elevated acidprotease activity in the extracellular milieu of malignant cells(9) has largely been traced to a cathepsin B-like cysteine protease (10, 11). Both intracellular and extracellular cathepsin Bactivity is frequently increased in cultured malignant cells. Itmay also be present in malignant tissues in an altered pheno-typic form as demonstrated by an abnormal isozyme pattern,increased stability to alkaline pH or high temperature (10), orenhancement in the plasma membrane and nuclei compared tonormal cells (12-14).
Oncogene studies have also implicated the related cysteineprotease, cathepsin L, in malignancy by demonstrating that thisenzyme is the major induced protein of raj-transformed
Received 11/29/88; revised 3/27/89; accepted 4/17/89.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 inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported in part by American Cancer Society Grants PDT-349 and IN-97.
2To whom requests for reprints should be addressed, at Department ofPathology, L804, Boston University School of Medicine, 80 E. Concord St.,Boston, MA 02118.
NIH3T3 cells (15, 16), and the "major excreted protein" of
cells transformed by Kirsten, Harvey, or Maloney sarcomaviruses and by SV40 virus (14, 17, 18). In terms of ability todegrade stromal proteins, cathepsin L is more active againstcollagen (19, 20) and 100-fold more active against elastin (8)than is cathepsin B.
Despite this evidence from animal and cell culture modelsfor the role of cysteine proteases in malignancy, only a fewstudies have attempted to correlate cathepsin activities in human tumor tissue with clinicopathological parameters. In onestudy of colorectal carcinoma, Durdey et al. (21) reported thatof four proteases examined (cathepsin B, cathepsin H, and 2collagenases), only cathepsin B activity was significantly correlated with local tumor invasion. We have investigated thisrelationship further by measuring cathepsin B-like and cathepsin L-like specific activities in matched tissue samples (normalmucosa and adenocarcinoma) from 27 patients. We have alsomeasured levels of cysteine protease inhibitor in these tissuesin order to determine whether cathepsin enzyme activities mightreflect increases or decreases in protease inhibitor levels withmalignancy. Utilizing an alkaline pH treatment we have alsoexplored whether the cathepsin B-like activity in the cancertissues is biochemically distinguishable from the enzyme activity in the normal mucosa. Finally, we have compared changesin the expression of cysteine proteases with stages of tumordevelopment and with tumor differentiation, including muci-nous/nonmucinous histológica! subtypes.
MATERIALS AND METHODS
Tissue Specimens. Fresh tissue from colorectal resections was obtained within 4 h of surgery at the Mallory Institute of Pathology,Boston, MA. To minimize possible differences due to genetic variation,we compared normal mucosa and tumor from the same patient. Insampling tumors, care was taken to avoid grossly necrotic areas andfat. Samples of normal colorectal tissue were obtained at least 10 cmfrom the tumor. The mucosa was then separated from the muscle layer,serosa, and surrounding fat. All samples were stored at —¿�80"C until
extraction. Normal mucosa was also obtained from non-tumor-bearingcolorectal resections in a similar manner. The clinical details of thepatients investigated and the corresponding pathological data are summarized in Table 1.
Dukes1 Classification. Individual tumors were staged according to theDukes' classification (22) as modified by Turnbull et al. (23). Dukes' Atumors are confined to the bowel wall; Dukes' B tumors have spreadbeyond the wall without involving lymph nodes; Dukes' C tumors areassociated with regional lymph node metastasis; and Dukes' D tumors
are associated with distant metastasis.Tissue Extraction. To minimize test variation, care was taken to
extract and assay each pair of normal mucosa/tumor samples at thesame time. Tissue samples (60-80 mg) were homogenized in 500 //I ofdistilled, deionized water, frozen and thawed three times, and centri-fuged for 50 min at 4°Cat 12,000 rpm (17,210 x g) in a Sorvall 5B
centrifuge (24). Supernatants were removed and utilized for the following assays.
Enzyme Assays. To distinguish cathepsin B-like enzyme activity fromcathepsin L-like activity, the cathepsin B specific substrate, Z-Ala-Arg-
3809
Research. on October 7, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from
CYSTEINE PROTEASES IN HUMAN COLORECTAL CARCINOMA
Arg-MNA' (Enzyme Systems Products, CA), was used (25). A specific
substrate for cathepsin L is not yet available. However, the substrateZ-Phe-Arg-MNA (Enzyme Systems Products, Dublin, CA) is cleavedmuch more readily by cathepsin L than by cathepsin B and so iscommonly used to detect cathepsin L-like activity, with the knowledgethat a lesser portion of this activity may represent cathepsin B activityin an extract containing both enzymes (26). To generate a more specificenzyme assay for cathepsin L we have utilized pH-sensitive differencesfor cathepsin B-like and L-like activity against these substrates.
Cathepsin B-like activity was determined by a modification of themethods of MacGregor et al. (27) and Barrett and Kirschke (26). Cellextracts were incubated in 0.1 M MES buffer, pH 5.8, containing 1 miviZ-Ala-Arg-Arg-MNA as substrate, 1 mM DTT, and 1 m\i EDTA at37"C. The assay was started by the addition of 20 n\ of tissue extract
(2-6 Mg/V protein) to the substrate/buffer mixture. At 10 min thereaction was terminated by the addition of 50 n\ l N HCI in 2% TritonX-100. Fast Blue B (O-dianisidine tetrazotized) (Sigma Chemical Co.,St. Louis, MO) was added and color developed for 10 min beforereading at A520in a Gilford spectrophotometer. For each tissue samplethree extractions were done and each extract was assayed in duplicatewith a mean value determined for inclusion in Table 1. For each of theenzymes assayed the intra) issue standard deviation was always less thanor equal to the standard deviation of the interi issue values.
Cathepsin L-like activity was measured by using the synthetic substrate Z-Phe-Arg-MNA. Tissue extracts (20 ^1) were incubated in 0.1M MES buffer, pH 3.5, containing 1 mM Z-Phe-Arg-MNA, 1 mM DTT,1 mM EDTA at 30"C for 10 min. Subsequent steps were similar to
those used in the cathepsin B assay./3-Glucuronidase activity was assayed according to the methods of
Kolodny and Mumford (28) and Beaudet et al. (29). Tissue extractsidentical to those prepared for cathepsin B and L assays were added(20 ^I/tube) to 0.1 mM sodium acetate buffer, pH 4.4. The enzymereaction was started by addition of 70 n\ substrate (final concentration,2.5 HIM4-methylumbelliferyl /3-D-glucuronide) and the reaction mixtures were incubated in aluminum foil-wrapped tubes for 6 min at 37°C.
The reaction was stopped by adding 3 ml of high pH buffer (0.1 MNH4OH-EDTA, pH 10.5) and fluorescence was measured in a Perkin-
Elmer fluorimeter.Specific activities of cathepsin B-like and L-like enzymes and of ß-
glucuronidase are expressed as nmol of substrate hydrolyzed per minper mg protein. Protein content of the tissue extracts was determinedby the method of Lowry et al. (30), using bovine serum albumin as astandard. The mean level of total soluble protein in our normal colo-rectal mucosa extracts was 3.11 ±1.01 (SD) compared to 3.22 ±1.04mg/ml total soluble protein in matched colorectal carcinoma tissues<n = 28).
pH Curves. To generate the pH curve, stock buffer solutions of thedesired pH were prepared by adjusting the pH of an MES solution (0.1M final concentration) with increasing amounts of NaOH. To test theability of the buffers to maintain pH, we measured the actual pH ofreaction mixtures after all the ingredients (including the tissue extract)had been added. The final pH corresponded to the expected pH of thebuffer.
pH Sensitivity. Sensitivity of normal and tumor extracts to alkalinepH (8.0) was examined by preincubating 40 »ilof extract in a Tris buffer(0.1 M final concentration), pH 8.0, at 4°Cfor l h (12). Subsequent
enzyme assays were performed as described above.Endogenous Cysteine Protease Inhibitor. To determine the levels of
CPI present in normal colorectal mucosa and carcinoma tissues, 200fi\ of tissue extract were treated to destroy endogenous proteases andto release cysteine protease inhibitor. This treatment included adjustingextracts to pH 2.5 with 0.5 N HCI, heating at 80°Cfor 1 min, adjusting
to pH 6.0 with 0.5 N NaOH, and centrifuging at 8000 x g for 10 min(31). Of the resulting extract, 100 ><\of the supernatant were assayedfor CPI levels by incubation with 20 ^1 of commercial papain (0.0175
1The abbreviations used are: Z-Ala-Arg-Arg-MNA. benzoyl-alanine-arginine-arginine-4-methoxy-2-naphlhylamide; Z-Phe-Arg-MNA, benzoyl-phenylalanine-arginine-4-methoxy-2-naphthylamide; CPI, cysleine protease inhibitor; MES,morpholinoethanesulfonicacid; DTT, dithiothreitol; E-64, L-rrans-epoxysuccinyl-leucylamido(4-guanido)butane: p21, M, 21,000 protein.
ng/n\) (Boehringer-Mannheim) at room temperature for 5 min (31).Levels of CPI in the extract were determined by comparing the extentof inhibition of papain activity [hydrolysis of a-A'-benzoyl-DL-arginine-2-naphthylamide (32)] in the presence or absence of treated extracts.One unit of inhibitor is defined as the amount causing a 1 unit decreasein papain activity.
E-64 Inhibition Assay. E-64 (Sigma) is a synthetic cysteine protease-specific inhibitor (33). Six normal colorectal mucosa plus six matchedcolorectal carcinoma extracts were tested for sensitivity of cathepsin fluke and L-like activities to E-64. A 10 mM stock solution was preparedin dimethyl sulfoxide and stored at 4 ( until needed. To determine theextent of protease inhibition by E-64, 20-25 /jl of tissue extract werepreincubated in 400 ¿ilof MES-EDTA-DTT buffer (see enzyme assaysabove) at 37"C for 5 min in the presence of 10 ><ME-64 (34). After 5
min of E-64 treatment, the extract was tested for either cathepsin fluke or cathepsin L-like activity, using the substrates and protocolsdescribed above.
Statistical Methods. Summary data are expressed as the mean ±SD,unless otherwise noted. For comparing data on matched pairs of normaland carcinoma tissue, the Wilcoxon signed rank test was used tocalculate the test statistic (35). To compare data with two means, aone-tailed Student's t test for the difference between means was used
to establish statistical significance (36). Probability values of 0.05 orless were considered significant.
RESULTS
Quantitative Enzyme and Patient Data
Table 1 shows the levels of cathepsin B-like and cathepsin L-like enzyme activities measured in colorectal carcinoma and inthe distant normal mucosa of the same patient for 28 adeno-carcinomas from 27 individuals. Table 1 also contains theclinical data of the patient study group, including age, sex,Dukes' stage, and tumor differentiation. Only cases in which
samples of both normal and tumor tissue were available fortesting were included in the analysis.
Cathepsin B-like Activity
As shown by the data in Table 1, mean cathepsin B-likeactivity in the carcinoma tissue (79.5 ±33.8) was significantlyhigher (P < 0.005) than cathepsin B-like activity in normalmucosa (57.9 ±24.3), using a two-sample t test to calculate thedifference between means (36). In 14 of 28 cases analyzed, thecathepsin B-like specific activity in tumor tissue was notablyhigher than that of the matched normal mucosa, as reflected bya cancer/normal ratio greater than 1.44 (see Table 1). Theaverage ratio for the individual pairs was 1.48.
Although considerable variation is observed in cathepsin B-like enzyme levels in both normal and carcinoma tissues, only2 of 28 normal tissues express levels of cathepsin B-like activitygreater than 86 nm/min/mg protein, while 13 of 28 tumorextracts have activity in the range of 90-140 nm/min/mgprotein. Among those 15 tumor tissues with relatively lowenzyme activity levels (less than 86 nm/min/mg protein), 6have a cancer/normal ratio greater than 1.44. Thus, in 19 of 28colorectal carcinomas an abnormal expression of cathepsin B-like activity could be detected by quantitative assay as eitherhigh total activity or a high cancer/normal ratio.
Cathepsin L-like Activity
As we observed for cathepsin B-like activities, cathepsin L-like activity in colorectal carcinoma tissue (13.1 ±3.0) wassignificantly higher than cathepsin L-like activity in matchednormal mucosa (10.4 ±2.2), using the two-sample / test to testthe difference in means (P < 0.005). In 8 of 17 pairs analyzed,
3810
Research. on October 7, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from
CYSTEINE PROTEASES IN HUMAN COLORECTAL CARCINOMA
Table 1 Calhepsin B-like and L-like activity in extracts of paired samples of normal coloréela!mucosa and carcinoma: distribution according to Dukes ' stage
Cathepsin B-like (nM/min/mgprotein)Dukes'
stageCaseA
3571133B
1812161723252932343538C
246913151836D
1920(¡r(u)Age758074673678498465897564556371455973586942747576715371SexFFMMMFFMFFFFMFFMFMMMMFMMMMMDifferen
tiation"MMMue,
MWPMMMMMMMMMMMMMMMMue,
PMMMue,
MMue,PMue,
PMue,MMNormal117.074.948.866.634.768.4
±28.0*80.6111.974.649.743.253.636.220.043.353.437.255.154.9
±24.384.346.963.018.328.276.531.561.051.2
±22.348.080.880.869.9
±15.5Cancer121.680.8117.5139.272.2106.3
±25.5116.7120.795.971.631.8111.534.638.343.9115.489.334.375.3
±35.398.7100.855.848.447.998.423.7124.774.8
±33.073.046.971.463.8
±12.0Cancer/
normal1.041.082.412.092.081.74
±0.61.451.081.291.440.742.080.961.921.012.162.400.621.43
±0.61.172.150.892.641.701.290.752.041.58
±0.61.520.580.880.99
±0.4Cathepsin
L-like (nM/min/mgprotein)Normal11.0012.306.357.758.569.2
±2.28.0811.2110.769.8912.5810.5910.0
±1.49.8710.568.4314.7516.257.4111.2
±3.2Cancer16.7017.7711.2814.7616.8015.5
±2.312.3011.9713.4514.1510.7110.8912.2
±1.212.6112.469.0710.2419.058.1111.9
±3.6Cancer/
normal1.521.441.781.901.901.72
±0.21.521.071.251.430.851.031.19
+0.21.281.211.080.691.171.091.08
±0.2
AU stages 57.9 ±24.3 79.5 ±33.8 1.48 ±0.6 10.4 ±2.2 13.1 ±3.0 1.31 ±0.3°Ail adenocarcinomas. W, well; M, moderate; P, poor differentiation; Mue, mucinous.* Mean ±SD.' (i) and (ii) represent 2 different adenocarcinomas Irum the same patient.
the activity in the tumor tissue was notably higher than that inthe normal tissue, as reflected by a cancer/normal ratio greaterthan 1.28. The average cancer/normal ratio for cathepsin L-like activity in the individual pairs was 1.31 ±0.3.
As with cathepsin B-like activity, variation is observed inboth normal and carcinoma tissues, but a significant shifttoward higher cathepsin L-like activity is seen in carcinomatissues such that 9 of 17 tumor extracts express activity greaterthan 12.3 nm/min/mg protein, while only 3 of 17 normalextracts have activity above this level.
Cathepsin Activities in Colorectal Mucosa of Noncancer Patients
In addition to paired comparisons of normal and carcinomatissue from specimens of patients with colorectal tumors, cathepsin B-like and L-like spécifieactivities were also determinedin colorectal mucosa from patients undergoing surgical resections for diverticular disease or abdominal trauma. Neither thecathepsin B-like activity (64.3 ±20.9; n = 9) nor cathepsin L-like activity (10.0 ±2.7; n = 5) levels in these nonneoplasticcontrol tissues were significantly different from activity levelsin the normal mucosa of patients with colorectal carcinoma.
Relationships between Tumor Morphology and Cathepsin Activity Levels
Enzyme data were analyzed to determine whether cathepsinactivity levels correlated in any way with independently assessed(a) Dukes' stage, (/>) mucinous/nonmucinous histológica! sub
type, and (c) tumor grade.
Dukes' Stage. In patients with different stages of colorectalcancer, neither the cathepsin B-like nor cathepsin L-like specificactivities varied significantly in normal tissue across the Dukes'
A, B, and C groups. However, in carcinoma tissue, cathepsinB-like specific activity in Dukes' A patients was significantlyhigher than in Dukes' B patients (P < 0.05), Dukes' C patients(P < 0.05), or Dukes' D patients (P < 0.01). For cathepsin L-like activity the pattern was similar, with Dukes' A levelssignificantly higher than Dukes' B (P < 0.01) and Dukes' C (P
< 0.05) levels of enzyme activity.Mucinous versus Nonmucinous Carcinomas. Cathepsin B-like
specific activity levels showed slightly higher cancer/normalratios in the mucinous (1.66 ±0.8) compared to nonmucinouscancers (1.43 ±0.5), but the difference was not significant.Similarly, both the 6 mucinous and 22 nonmucinous tumorsshowed almost identical cancer/normal ratios for cathepsin L-like activity (1.31 and 1.32, respectively).
Tumor Grade. Cathepsin B-like activity in 4 poorly differentiated carcinomas showed a significantly higher (P < 0.05)cancer/normal ratio (2.07 ±0.46) than in 23 moderately differentiated carcinomas (1.35 ±0.58).
pH Stability of Normal and Tumor Cathepsin B-like Activities
One means of determining whether cathepsin B-like activitiesare not only increased in cancerous tissue but are also presentin an altered form is to test the enzyme activity for changes inbiochemical characteristics. One of the strongest markers ofnormal cathepsin B activity is its extreme lability at or above
3811
Research. on October 7, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from
CYSTEINE PROTEASES IN HUMAN COLORECTAL CARCINOMA
pH 8.0 (32). However, it is also known that cathepsin B-like
activities within tumor tissues or secreted by tumor tissues areoften more stable to treatment at pH 8.0 than is the normalenzyme (10, 14). Thus, we treated 17 normal extracts and 17matched carcinoma extracts for stability to pH 8.O. The resultsare represented in Table 2 as residual cathepsin B activity inthe cancer versus normal tissue after pretreatment at pH 8.O.Even for the carcinoma extracts, treatment at pH 8.0 is remarkably destructive of cathepsin B-like activity, indicating theneed for a less alkaline pH treatment (e.g., pH 7.5) to obtainmore subtle comparisons in the future. Nonetheless, in 8 of the17 pairs of extracts tested, the cathepsin B-like activity in thecarcinoma extracts was 50 to 200% more stable to alkalinetreatment than activity in the matched normal extracts. Themean of the individual ratios of cancer to normal residualactivity is 1.42 ±0.8, which indicates a significant increase inresidual activity at pH 8.0 in the carcinoma tissues (P< 0.025).What is perhaps most interesting to note, with respect to thosepairs in which the cathepsin B-like activity is more notablystable, is that 7 of 8 were among those cases in which theabsolute level of cathepsin B-like activity in the carcinoma tissuewas not extremely high, falling below a value of 86 nm/min/mg, which roughly separates normal from abnormal cathepsinB-like activity in colorectal mucosa. In addition, 4 of the 9 withmore alkaline-stable enzyme activity had quite low ratios ofcancer/normal cathepsin B-like activity. Thus, colorectal tumorextracts which, by quantitative cathepsin B assay might lookrelatively normal, may be recognized as abnormal by measuringthe pH stability of their cathepsin B-like activity.
Determining the Specificity of Cathepsin B-like versusCathepsinL-like Activities
In order to measure cathepsin B-like enzyme activity, thesubstrate Z-Ala-Arg-Arg-MNA was used as it is highly specificfor cleavage by cathepsin B but not cathepsin L (25, 26). Thesubstrate, Z-Phe-Arg-MNA, which is cleaved more readily byL than by B, is commonly used to detect cathepsin L plus aportion of cathepsin B activity at pH 5.8 (26). To generate amore specific assay for cathepsin L, we have utilized differencesin pH activity and stability for these enzymes. Cathepsin L is
Table 2 Percentage of residual cathepsin B-like activity after pretreatment ofpaired normal mucosa and cancer extracts al pH 8.0a
Case12361617181920
(i)420
<¡¡)23323334353638Normal3.74.65.28.63.59.29.24.05.05.03.910.79.118.09.22.96.6Cancer4.64.13.77.210.08.923.211.08.07.76.010.421.13.65.42.312.6Cancer/
normal1.240.890.710.842.860.972.522.751.601.541.540.932.320.200.590.791.91
Mean ±SD 8.81 ±5.6 1.42 ±0.8'
" For details of assay, see "Materials and Methods." Mean specific activities
(nM/min/mg protein) for control and treated normal mucosa were 57.9 ±29.7and 3.34 ±1.28, respectively. Mean specific activities (nM/min/mg protein) forcontrol and treated cancers were 81.3 ±46.9 and 5.64 ±3.2, respectively.
* (i) and (ii) represent 2 different adenocarcinomas from the same patient.' Paired Student's f test (one-tailed) is significant at the P = 0.025 level.
more active and stable in a lower pH range than cathepsin Band has a lower pH optimum for most substrates (pH 5.0-5.5)(37) compared to cathepsin B (pH 5.8-6.5) (25, 26). In addition,cathepsin B is rapidly inactivated below pH 4.0 (25) whilecathepsin L retains 30% of its activity after a 1-h incubation atpH 3.5 (20). Thus, the hydrolysis of a given substrate bycathepsin L compared to cathepsin B increases from pH 6.5 topH3.5(19).
As seen in Fig. 1, the ratio of colorectal mucosa! extractactivity against Z-Phe-Arg-MNA (cathepsin L-like) versus Z-Ala-Arg-Arg-MNA (cathepsin B-like) also increases dramatically as one moves from a pH of 6.0 to 3.5. These differencesat low pH values provide a useful means of measuring cathepsinL-like specific activity in extracts containing a mixture ofactivities. This is demonstrated by the independent resultswhich we have obtained for cathepsin B-like activity (hydrolysisof Z-Ala-Arg-Arg-MNA at pH 5.8) versus cathepsin L-likeactivity (hydrolysis of Z-Phe-Arg-MNA at pH 3.5) as seen inthe matched pairs data of Table 1. Although, as a group,carcinoma tissues express higher average levels of both cathepsin B-like and L-like activities, for a given individual theseenzymes do not appear to be coordinately regulated. Thus,patients with high cathepsin B-like activity will not necessarilyhave high cathepsin L-like activity. Similarly, it is possible tomeasure a dramatic increase in cathepsin B-like specific activityin cancer IW.VM.Vnormal colorectal tissue without finding acorresponding change in cathepsin L-like specific activity inthat patient's tissues.
Inhibition of Cathepsin Activities by E-64
Twelve tissue extracts (six pairs of normal and carcinomatissue) were tested in duplicate for sensitivity to the cysteineprotease inhibitor, E-64 (34). No residual activity could bemeasured against the substrates Z-Ala-Arg-Arg-MNA and Z-Phe-Arg-MNA in the presence of E-64, irrespective of inhibitorconcentration (2 to 10 /IM), indicating that we are measuringcysteine protease activity in our assay systems and not othertypes of proteases.
/3-Glucuronidase Activity
To determine whether the increased cathepsin activities incolorectal carcinoma tissue might be explained by a nonspecificincrease in lysosomal enzyme activities, we also tested 19matched pairs of normal and carcinoma tissue for activity levelsof /3-glucuronidase, a nonproteolytic lysosomal enzyme. Fornormal mucosa samples, /3-glucuronidase specific activity was9.80 ±3.55 nm/min/mg protein, while that of the carcinoma
100
Fig. 1. Comparison of pH effects on the hydrolysis of synthetic substrates (Z-Ala-Arg-Arg-MNA. •¿�;and Z-Phe-Arg-MNA, A. Points, maximal activity ofnormal colorectal mucosa extracts at various pH values.
3812
Research. on October 7, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from
CYSTEINE PROTEASES IN HUMAN COLORECTAL CARCINOMA
tissue was 7.49 ±5.04 nm/min/mg protein, which is not asignificant difference. Thus, in contrast to the increased protease activities, levels of /3-glucuronidase in the carcinoma tissues were not different from matched normal tissues. Thesevalues for /3-glucuronidase are similar to those measured inhuman fibroblasts (7.82 nm/min/mg protein) and leukocytes(9.37 nm/min/mg protein) (29).
Levels of Cysteine Protease Inhibitor in Matched Pairs of Tumorand Normal Colorectal Mucosa
Levels of endogenous CPI were determined in 27 matchedpairs of normal mucosa and carcinoma tissue. The averagecontent (in units of inhibitor/mg protein) of CPI in colorectalmucosa (3.85 ±1.6) is the same as that in colorectal carcinoma(3.82 ±1.5).
DISCUSSION
Studies using both cell culture and animal models on the roleof cysteine proteases in malignancy have demonstrated thatincreased or altered expression of cathepsins B and L providesa marker for the transformed phenotype (4, 9-18). However,only a few studies have been done to explore whether quantitative and qualitative changes in cysteine protease activity demonstrate a reproducible pattern in human tumor tissue or acorrelation with tumor staging. To obtain these types of data,we have analyzed cysteine protease activities in matched normalcolorectal mucosa and carcinoma tissue from 27 patients. Colorectal cancer resections provide suitable material for examining the relationship of protease activity levels to tumor invasion. Advantages include the ability to obtain appropriate normal control mucosa with a relatively constant epithelial cellcomponent, and the presence of a well-tested staging system(Dukes') which closely parallels the biology of tumor invasion.
Only four previous studies of human cancer have analyzedcysteine cathepsin activities in a series of matched malignantand adjacent normal tissues. Our data are consistent with thoseof Vasishta et al. (38) and Watanabe et al. (39) in gastric tumors,with those of Durdey et al. (21) in colorectal carcinomas andwith those of Abecassis et al. in breast tumors (40), in describingsignificantly elevated cysteine protease activity in tumor compared to normal tissue. However, we have examined in greaterdetail the relationships between altered protease expression andclinicopathological parameters. We have also extended biochemical analyses to include measurements of protease stabilityand endogenous protease inhibitors.
Our finding of a statistically significant increase in cathepsinB-like and L-like activities in the earliest stage (Dukes' A) ofcolorectal cancer compared to later stages (Dukes' B, C, and
D) has not been reported previously. Durdey et al. (21) demonstrated no correlation of cysteine protease activity withDukes' stage, although importantly, the presence of only oneDukes' A tumor in that study would have precluded finding
such an association. Durdey et al. did observe surprisingly lowlevels of protease activity in Dukes' D tumors, an observation
which our data confirm.Elevated cysteine protease activities in early stage colorectal
carcinomas suggest that tumors still in the process of invadingthe bowel wall (Dukes' A) may require more protease activity
than tumors which have already achieved a greater degree ofinvasion and metastasis (Dukes' B, C, and D). This pattern of
high protease activity in early tumors which decreases in laterstages may provide an example of "phenotypic drift" of clonally
derived tumors, with continued somatic mutation generating
an admixture of distinct phenotypes (41, 42), and selection forthose cells which provide advantageous characteristics at eachstage.
High levels of protease activity in early tumors might also bedue to a transient modification of cathepsin gene expression byother factors important in the early stages of colorectal carcinoma. For example, elevated levels of the ras oncogene protein,p21, are common to earlier stages of colon tumors (Dukes' B
and C tumors) as compared to later stage primary tumors(Dukes' D) or metastatic tumor deposits (43,44). DNA analysis
has also indicated that ras gene mutations are relatively earlyevents in colorectal tumorigenesis (45). The similarity in patternfor cathepsin B-like and L-like protease activities during colo
rectal tumor progression to that described for raÃp21 may notbe coincidental. A precursor form of cathepsin L has alreadybeen shown to be the "major induced protein" (16) and the"major excreted protein" (15, 17, 18) of ras transformed NIH
3T3 cells. However no one has yet studied the relative expression of raj p21 and cysteine protease activities in tumor tissues.Colorectal carcinoma may provide a useful tissue type forstudying a possible relationship between coordinately increasedexpression of cathepsin L and ras p21 during early humantumor cell growth.
In studies of gastric adenocarcinomas, Watanabe et al. (39)demonstrated significantly higher cathepsin B and L activitiesin poorly differentiated compared to well or moderately differentiated tumors. We have also found that the 4 poorly differentiated tumors in our series contain significantly higher cathepsin B-like activity than the moderately differentiated tumors. In addition we examined whether cysteine cathepsinactivity correlated with mucinous or nonmucinous histológica!subtype, given that Vickery and Symonds (46) have reportedthat stage-matched mucinous colorectal carcinomas have apoorer 5-year survival rate than nonmucinous tumors. Ouranalysis demonstrates no significant difference in cathepsin B-like or L-like activities between mucinous and nonmucinoustumors. While the number of mucinous tumors we have assayedis small (six), our data support recent reports that mucinouscarcinomas show a similar biological behavior, with no differentprognosis, compared to their nonmucinous counterparts (47,48).
Although we found no difference in the levels of cysteineprotease inhibitor in normal versus tumor tissue, our assays donot rule out the possibility that the cysteine cathepsin activitiesin carcinoma tissue might somehow be less sensitive to inhibition by the tumor tissue CPI due to a qualitative change ineither the tumor protease or protease inhibitors of the carcinoma tissue. This type of change has been suggested by Lah etal. (49), who found the total cysteine protease inhibitory contentto be the same in liver and sarcoma tissue, but the protease/inhibitor affinity to be reduced for the sarcoma inhibitor.
Abnormalities in the expression of cathepsin B derived frommalignant tissues may also be due to altered forms of theenzyme (10,12). Our results demonstrated a significant increasein stability to pH 8.0 of cathepsin B-like activity in carcinomatissues compared to matched normal tissues for 8 of 17 patientstested. This type of alkaline-stable cathepsin B activity in asciticfluid of women with ovarian cancer or in culture medium ofhuman breast tumor expiants, has been shown to represent ahigh-molecular-weight procathepsin B which has not been fullyprocessed to the mature lysosomal form (50). Steiner et al. (3)suggest that such precursor forms of cathepsin B may be moreresponsible for the activation of hormones and growth factorsthan the mature lysosomal enzyme. Sloane et al. (14) have also
3813
Research. on October 7, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from
CYSTEINE PROTEASES IN HUMAN COLORECTAL CARCINOMA
found an atypically pH-stable cathepsin B in malignant celllines, including colorectal carcinoma cell lines, which representscathepsin B-like activity shifted from the lysosome to theplasma membrane (13), where it is presumed to function in thedigestion of the extracellular matrix.
In summary, our observations on an increase in both theamount and stability of cysteine protease activity in colorectalcarcinoma tissues indicates that cathepsins B and L should bestudied further as markers of tumor progression in this type ofmalignancy. Work remains to be done to determine how earlyin the adenoma/carcinoma sequence these protease activitiesare altered, to explore the mechanism by which abnormalcysteine protease activities are generated, and to elucidate thepredominant cell type responsible for increased cysteine protease activities in colorectal carcinoma.
ACKNOWLEDGMENTS
We thank Dr. C. O'Keane and residents of the Mallory Institute of
Pathology for assistance in obtaining surgical specimens. Dr. J. Younusfor /J-glucuronidase assays, J. Song for assistance with statistical calculations, P. Ward for typing the manuscript, and Drs. A. E. Reif, W.Gonnerman, and M. J. O'Brien for critical reading of the manuscript.
REFERENCES
1. Cunningham. D. D. Proteases as growth factors. In: R. Bascrga (ed.). TissueGrowth Factors, pp. 229-248. New York: Springer-Verlag. 1981.
2. Scher, \V. Biology of disease. The role of extracellular proteases in cellproliferation and differentiation. Lab. Invest.. 57: 607-633, 1987.
3. Steiner. D. F., Docherty. K.. and Carroll, R. Golgi/granule processing ofpcptide hormone and ncuropeptide precursors: a minireview. J. Cell.Biochem.. 24: 121-1.10. 1984.
4. Recklies. A. D..and Poole. A. R. Proteolytic mechanisms of tissue destructionin tumor growth and metastasis. In: G. Weiss (ed.). Liver Metastasis, pp.77-95. Boston; Hall, 1982.
5. Morrison, R. I. G.. Barrett. A. J., Dingle. J. T.. and Prior, D. Cathcpsins Bland D action on human cartilage proteoglycans. Biochim. Biophys. Acta,¿02:411-419. 1973.
6. Burlcigh. M. C., Barrett. A. J., and Lazarus, G. S. Cathepsin Bl. A lysosomalenzyme that degrades native collagen. Biochem. J., 137: 387-398, 1974.
7. Etherington. D. S., and Evans, P. J. The action of cathepsin B and collagen-olytic cathepsin in the degradation of collagen. Acta Biol. Mcd. Germ.. 36:1555-1563, 1977.
8. Mason. R. W., Johnson. D. A., Barrett, A. J.. and C'hapman. H. A. Elastin-olytic activity of human cathepsin L. Biochem. J.. 233: 925-927. 1986.
9. Carrel. A., and Ebeling. A. H. The fundamental properties of the fibroblastand the macrophage. III. The malignant fibroblast of sarcoma 10 of theCrocker Foundation. J. Exp. Med., 48: 105-123. 1928.
10. Mort. J. S., Recklies. A. D.. and Poole, A. R. Characterization of a thiolproteinase secreted by malignant breast tumors. Biochim. Biophys. Acta,614: 134-143. 1980.
11. Szego, C. M., Seeler. B. J.. and Smith. R. E. Lysosomal cathepsin B,: partialcharacterization in rat preputial gland and recompartmentation in responseto estradiol-17-beta. Eur. J. Biochem.. 69: 463-474. 1976.
12. Pietras. R. J., and Roberts, J. A. Cathepsin B-like enzymes. Subcellulardistribution and properties in neoplastic and control cells from humanectocervix. J. Biol. Chem.. 256: 8536-8544, 1981.
13. Sloane. B. F.. Rozhin. J.. Johnson. K.. Taylor. H., Crissman. J. D.. andHonn. K. V. Cathepsin B: association with plasma membrane in metastatictumors. Proc. Nati. Acad. Sci. USA, 83: 2483-2487. 1986.
14. Sloane. B. F.. Rozhin. J., Halfield. J. S.. Crissman, J. D., and Honn, K. V.Plasma membrane-associated cysteine protcinascs in human and animaltumors. Exp. Cell Biol., 55: 209-224, 1987.
15. Gal. S.. and Gottesman. M. M. The major excreted protein (MEP) oftransformed mouse cells and cathepsin L have similar protease specificity.Biochem. Biophys. Res. Commun., 139: 156-162, 1986.
16. Joseph. L., Lapid. S.. and Sukhatme. V. The major ras induced protein inNIH3T3 cells is cathepsin L. Nucleic Acids Res., IS: 3186, 1987.
17. IK-iih.mil D. T.. Hamilton, R. T., Parfelt. C. K. J.. Edwards. D. R.. St.Pierre, R.. Waterhouse. P., and Nilsen-Hamilton, M. Close relationship ofthe major excreted protein of transformed murine fibroblasts to thiol-de-pendent cathepsins. Cancer Res., 46: 4590-4593, 1986.
18. Portnoy. D. A., Erickson. A. H., Kochan. J.. Ravelch. J. V.. and Unkeless,J. C. Cloning and characterization of a mouse cysteine proteinase. J. Biol.Chem.. 261: 14697-14703. 1986.
19. Kirschke. H.. Kembhavi. A. A.. Bohley. P.. and Barrett. A. J. Action of rat
liver cathepsin L on collagen and other substrates. Biochem. J., 201: 367-372. 1982.
20. Mason, R. W., Green, G. D.. and Barrett. A. J. Human liver cathepsin L.Biochem. J., 226: 233-241. 1985.
21. Durdey, P.. Cooper. J. C.. Switala, A.. King, R. F. G. J.. and Williams. N.S. The role of peptidases in cancer of the rectum and sigmoid colon. Br. J.Surg.. 72:378-381. 1985.
22. Dukes. C. E. The classification of cancer of the rectum. J. Pathol. Bacterio!.,35:323-332, 1932.
23. Turnbull, R. B., Jr.. Kyle. K.. Watson, F. R., and Spratt. J. Cancer of thecolon: the influence of the no-touch isolation technic on survival rates. Ann.Surg.. 166: 420-427, 1967.
24. Nichols, E. A., and Ruddle. F. H. A review of enzyme polymorphism, linkageand electrophoretic conditions for mouse and somatic cell hybrids in starchgels. J. Histochem. Cytochem.. 21: 1066-1081. 1973.
25. Takahashi. T., Yonezawa. S.. Dehdarani. A. H.. and Tang, J. Comparativestudies of two cathepsin B isozymes from porcine spleen. J. Biol. Chem.,261: 9368-9374. 1986.
26. Barrett. A. J.. and Kirschke. H. Cathepsin B. cathepsin H. cathepsin L.Methods Enzymol.. 80: 535-561. 1981.
27. MacGregor. R. R., Hamilton. J. W., Shofstall, R. E.. and Cohn. D. V.Isolation and characterization of porcine parathyroid cathepsin B. J. Biol.Chem.. 254:4423-4427. 1979.
28. Kolodny. E. H., and Mumford, R. A. Human leukocyte acid hydrolascs:characterization of eleven lysosomal enzymes and study of reaction conditions for their automated analysis. Clin. Chim. Acta. 70: 247-257. 1976.
29. Bcaudet. A. L., DiFerrante, N. M., Ferry. G. D.. Nichols, B. L.. Jr., andMullins, C. E. Variation in the phenotypic expression of beta-glucuronidasedeficiency. J. Pediatr.. 86: 388-394, 1975.
30. Lowry. O. H., Rosebrough, N. J., Farr, A. L.. and Randall, R. J. Proteinmeasurement with the Polin phenol reagent. J. Biol. Chem.. 193: 265-275,1951.
31. Sano, M.. Wada. Y., li. K.. Kominami. E.. Katunuma. N.. and Tsukagoshi.H. Immunolocalization of cathepsins B, H and L in skeletal muscle of X-linked muscular dystrophy (mdx) mouse. Acta Neuropathol.. 75: 217-225,1988.
32. Barrett, A. J. A new assay for cathepsin Bl and other thiol proteinases. Anal.Biochem., 47: 280-293, 1972.
33. Barrett, A. J., Kembhavi. A. A.. Brown. M. A., Kirschke. H., Knight, C. G.,Tamai, M., and Hanada. K. i.-/ra/w-Epoxysuccinyl-leucylamido (4-guani-dino) butane (E-64) and its analogues as inhibitors of cysteine proteinascsincluding cathepsins B. H and L. Biochem. J.. 207; 189-198. 1982.
34. Kirschke. H.. Wood. L.. Roisen. F. J.. and Bird. J. W. C. Activity of lysosomalcysteine proteinase during differentiation of rat skeletal muscle. Biochem. J..2/4:871-877. 1983.
35. Campbell, R. C. Statistics for Biologists. New York: Cambridge UniversityPress. 1967.
36. Freund. J. E. Mathematical statistics. Englewood Cliffs, NJ: Prentice-Hall,1971.
37. Kirschke. H., Langner. J., Wiederanders. B., Ansorge, S.. and Bohley, P.Cathepsin L. A new proteinase from rat-liver lysosomes. Eur. J. Biochem.,74:293-301. 1977.
38. Vasishta, A., Baker, P. R., Hopwood. D., Holley. P. M.. and Cuschieri, A.Proteinasc-like pcptidase activities in malignant and non-malignant gastrictissue. Br. J. Surg.. 72: 386-388, 1985.
39. Watanabe, M.. Higashi. T., Hashimoto, M.. Tomoda. I.. Tominaga, S.,Hashimoto, N.. Morimoto. S.. Yamauchi. Y., Nakatsukasa, H.. Kobayashi.M.. Watanabe. A., and Nagashima, H. Elevation of tissue calhepsin B andL activities in gastric cancer. Hepato-gastroenterology, 34: 120-122, 1987.
40. Abecassis. J.. Collard. R., Eber. M., Pusel. J. Fricker, J. P., and Methlin. G.Proteinases and sialytransfcrase in human breast tumors. Int. J. Cancer, 33:821-824. 1984.
41. Nowell, P. C. The clonal evolution of tumor cell populations. Science (Wash.DC). 194: 23-28, 1976.
42. Poste, G., and Fidler, I. J. The pathogenesis of cancer metastasis. Nature(Lond.). 283: 139-146. 1980.
43. Gallick, G. E.. Kurzrock, R.. Kloetzer. W. S.. Arlinghaus, R. B.. and Gutter-man. J. U. Expression of p2l ras in fresh primary and metastatic humancolorectal tumors. Proc. Nati. Acad. Sci. USA, 82: 1795-1799, 1985.
44. Horan Hand. P.. Vilasi. V., Thor, A.. Chuchi. N.. and Schlom. J. Quantita-tion of Harvey ras p21 enhanced expression in human breast and coloncarcinomas. J. Nail. Cancer Inst., 79: 59-65, 1987.
45. Vogelstein, B.. Fearon. E. R.. Hamilton. S. R., Kern, S. E., Preisinger. A.C., Leppert, M., Nakamaura, Y., White. R.. Smits, A. M. M., and Bos. J. L.Genetic alterations during colorectal tumor development. N. Engl. J. Med.,319: 525-532. 1988.
46. Symonds. D. A., and Vickery. A. L.. Jr. Mucinous carcinoma of the colonand rectum. Cancer (Phila.), 37: 1891-1900, 1976.
47. Sasaki. O., Atkin. W. S., and Jass. J. R. Mucinous carcinoma of the rectum.Histopathology(Oxf), //: 259-272. 1987.
48. Halvorscn. T. B.. and Seim. E. Influence of mucinous components on survivalin colorectal adcnocarcinomas: a multivariate analysis. J. C'lin. Pathol.. 41:1068-1072. 1988.
49. Lah, T. T., Rozhin. J., Honn, K. V.. Crissman. J. D.. and Sloane. B. F.Human tumor cysteine proteinase inhibitors: reduced affinity against cathepsin B. J. Cell Biol., /0/:483a, 1985.
50. Mort. J. S.. and Recklies. A. D. Interrelationship of active and latent secretedhuman cathepsin B precursors. Biochem. J., 233: 57-63, 1986.
3814
Research. on October 7, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from
1989;49:3809-3814. Cancer Res Kieran Sheahan, Sania Shuja and Mary Jo Murnane Colorectal CarcinomaCysteine Protease Activities and Tumor Development in Human
Updated version
http://cancerres.aacrjournals.org/content/49/14/3809
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/49/14/3809To request permission to re-use all or part of this article, use this link
Research. on October 7, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from
