[CANCER RESEARCH 49, 499-502, January 15, 1989]

Meeting Report

Workshop Report from the Division of Cancer Etiology, National Cancer Institute,National Institutes of HealthProtease Inhibitors as Cancer Chemopreventive Agents1

Abstract

This workshop was organized to discuss the current state of researchon anticarcinogenic protease inhibitors with regard to their potential useas human cancer chemopreventive agents. Previous studies have indicatedthat protease inhibitors can be powerful anticarcinogenic agents foranimals and cells in culture and that human populations known to havehigh concentrations of protease inhibitors in the diet have low overallcancer mortality rates. In the workshop discussions, emphasis was placedon certain dietary protease inhibitors, such as the soybean-derived Bow-man-Birk inhibitor and chymotrypsin inhibitor 1 from potatoes and someof the highly purified protease inhibitors of microbial origin provided bythe Japan Society for the Promotion of Science, which have already beenshown to contain anticarcinogenic activity in laboratory studies. Potentialadverse side effects of dietary protease inhibitors were also considered,specifically, their possible effects on the pancreas and in causing decreased growth rates in young organisms. It was pointed out that thepancreata of a few species, notably rats and chicks, are extraordinarilysensitive to dietary protease inhibitors. Rats fed diets containing highconcentrations of soybean-derived protease inhibitors (raw soy flour) hadenlarged pancreata; increased pancreatic growth is thought to acceleratecancer development in the pancreas. The effect of raw soy flour on thegrowth of the rat pancreas has not been shown to occur in most otherspecies tested (examples include hamsters, mice, dogs, pigs, and monkeys) and is not expected to occur in humans. There is no evidence thatdietary protease inhibitors have adverse effects on the human pancreas.In fact, it has been observed that human populations with high levels ofdietary protease inhibitors have decreased rates of pancreatic cancer.Dietary concentrations of protease inhibitors which have been shown tobe anticarcinogenic have not produced decreased growth rates in animalsor any type of pancreatic pathology. In general, there was a high level ofenthusiasm at the workshop for the further development of protease

Received 8/11/88: accepted 9/26/88.'This workshop was held December 14 and 15 at the Endicott House in

Dedham, MA.Workshop Members: Paul Billings, Harvard School of Public Health, Boston,

MA; Yehudith Birk, Faculty of Agriculture, The Hebrew University of Jerusalem,Rehovot, Israel; Donald E. Bowman, Indiana University Medical Center, Indianapolis, IN; David Brandon, Western Regional Research Center, United StatesDepartment of Agriculture, Albany, CA; Janice Chang, Harvard School of PublicHealth, Boston, MA; Leonard Cohen, Naylor Dana Institute for Disease Prevention, American Health Foundation, Valhalla, NY; Pelayo Correa, Louisiana StateUniversity Medical Center, New Orleans, LA; Carol Grieve Phillips, ChemsynScience Laboratories, Lenexa, KS; Thomas Finlay, New York University MedicalCenter, New York, NY; Peter Flecker, Institut fürOrganische Chemie derUniversitätKarlsruhe, Karlsruhe, Federal Republic of Germany; Krystyna Fren-kel, New York University Medical Center, New York, NY; Seymour J. Garte,New York University Medical Center, New York, NY; Alfred L. Goldberg,Harvard Medical School, Boston, MA; Ronald H. Goldfarb, Pittsburgh CancerInstitute. Pittsburgh, PA; Dietrich Hoffmann, Naylor Dana Institute for DiseasePrevention, American Health Foundation, Valhalla, NY; Ann R. Kennedy, University of Pennsylvania Medical School, Philadelphia, PA; Mortimer Levitz, NewYork University Medical Center, New York, NY; Daniel Longnecker, DartmouthMedical School, Hanover, NH; Christopher Nelson, Kemin Industries, Inc., DesMoines, lA; Mark S. Pasternaek, Massachusetts General Hospital, Boston, MA;James P. Quigley, State University of New York, Stony Brook, NY; Elmer J.Reist, SRI International, Menlo Park, CA; Bill D. Roebuck, Dartmouth MedicalSchool, Hanover, NH; Clarence Ryan, Washington State University, Pullman,WA; Wei-Chiang Shen, University of Southern California School of Pharmacy,Los Angeles, CA; John E. Smart, Hoffmann-La Roche, Inc., Nutley, NJ; Carl E.Smith, National Cancer Institute, Bethesda. MD; Walter Troll, New York University Medical Center, New York, NY; Kejuo Umezawa, Microbial ChemistryResearch Foundation. Institute of Microbial Chemistry, Tokyo, Japan; Eric vonHofe, Harvard School of Public Health, Boston, MA; Lee W. Wattenberg,University of Minnesota, Minneapolis, MN; Hanspeter Witschi, University ofCalifornia, Davis, CA: and Jonathan Yavelow, Rider College, Lawrenceville, NJ.

inhibitors as chemopreventive agents. Recommendations for future research include: (a) research and development of sources of proteaseinhibitors; (b) analysis of human foods for protease inhibitor content; (<•)

evaluation of cancer incidence data in relation to protease inhibitorcontent and characteristics in the diet of human populations; (d) animalstudies on the efficacy of protease inhibitors in cancer prevention; and(<?)studies on the mechanism of action of anticarcinogenic proteaseinhibitors.

The first beneficial trace components recognized in humanfoods were vitamins necessary for growth and to prevent specificdiseases. Trace components were studied with renewed interestas anticarcinogens when it was noted that vegetarian populations had a lower rate of cancer occurrence.

The first identified compounds to prevent promotion andexperimental animal breast and colon cancers were proteaseinhibitors. Protease inhibitors are in an awkward position incomparison to vitamins, in that there is no demonstrated nutritional requirement for protease inhibitors. In fact, they limitgrowth in young animals by interfering with protein digestion.This workshop was organized to consider the usefulness ofprotease inhibitors as chemopreventive agents and to identifyareas of additional experimentation necessary to discover newprotease inhibitors relevant to cancer prevention, determinetheir range of preventive activity, understand their basic mechanisms of action, and expedite their availability to the researchcommunity.

The investigators met to discuss the usefulness of proteaseinhibitors from several sources as cancer chemopreventiveagents. The meeting was chaired by D. Hoffmann and L. W.Wattenberg and was organized by A. R. Kennedy and W. Troll,with D. G. Longfellow (Chief) and C. E. Smith (ProgramDirector) of the Biological and Chemical Prevention Program,Chemical and Physical Carcinogenesis Branch, Division ofCancer Etiology, National Cancer Institute.

Kennedy and Troll reviewed data related to the anticarcinogenic activity of protease inhibitors in humans, animals, andtissue culture. The epidemiological data on people who eatfoods rich in protease inhibitors (rice, maize, corn, bread,cereals) show a significant decrease in the incidence of breast,colon, prostatic, oral, and pharyngeal cancers. Protease inhibitors prevent or suppress many experimentally induced animaltumors. The application of synthetic protease inhibitors (thechymotrypsin inhibitor tosylphenylalanine chloromethyl ke-tone; the trypsin inhibitors tosyl-L-lysine chloromethyl ketone,and tosyl-L-arginine methyl ester; and leupeptin) to mouse skinsuppressed skin tumorigenesis initiated with 7,12-dimethyl-benz(fl)anthracene and promoted by TPA.2 Feeding leupeptin

(0.1%) in the diet suppressed experimentally induced colon andbreast carcinogenesis in rats and leukemia and skin carcinogen-esis in mice. Soybean diets rich in protease inhibitors suppressed skin tumorigenesis initiated with nitroquinoline oxideand promoted by TPA in mice and breast tumors in Sprague-

!The abbreviations used are: TPA, 12-O-tetradecanoylphorbol-13-acetate;

BBI, Hemman Hirk inhibitor.

499

on June 22, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MEETING REPORT

Dawley rats treated with ionizing radiation. Colon carcinogen-esis induced by 1,2-dimethylhydrazine was suppressed by i-aminocaproic acid, a plasminogen activator/inhibitor derivedfrom lysine by removal of the a-amino group, and by a soybeanextract containing a high concentration of the Bowman-Birk

inhibitor.Many vegetables contain protease inhibitors which could be

useful as cancer chemopreventive agents. Several studies havebeen performed which show that the soybean-derived BBI canprevent both carcinogenesis in vivo and transformation in vitro.The chymotrypsin inhibitory activity of BBI appears to benecessary for the anticarcinogenic effect of this inhibitor. Inaddition, the vegetable-derived chymotrypsin inhibitors chickpea inhibitor and chymotrypsin inhibitor 1 from potatoes havebeen shown to have the ability to inhibit malignant transformation. BBI has been studied by many investigators as ananticarcinogenic agent. The workshop was fortunate in havingpresent both D. E. Bowman and Y. Birk, the discoverers of theinhibitor. This was their first personal meeting even thoughthey worked together extensively on the identification andisolation of the inhibitor.

Birk reviewed the data on the widespread distribution ofprotease inhibitors in plant products and discussed their nutritional properties. The relatively high concentrations of proteaseinhibitors in many different constituents of the human diet haveraised questions about whether there might be potentially deleterious effects on human health. The information on the effectsof the dietary protease inhibitors on health has come primarilyfrom animal experiments. Two possibly harmful effects ofprotease inhibitors have been observed in animal studies: (a)decreased growth rates in young rats due to interference withprotein digestion; and (b) enlargement of the rat pancreas. Thislatter effect is thought to accelerate pancreatic cancer development caused by the pancreatic carcinogen azaserine in rats.

B. D. Roebuck described some recent results in the azaserine-induced pancreatic cancer system in the rat. A synthetic protease inhibitor, camostate, increased pancreatic growth throughstimulation of cholecystokinin through the inhibition of trypsin.Roebuck noted that while this inhibitor may accelerate an earlypromotional stage of cancer, it inhibits a late stage of the samedisease. This was demonstrated by using a cholecystokininanalogue as the promoter, causing enlargement of the pancreas,while the protease inhibitor camostate blocked a later stage incancer development.

Birk pointed out that long-term feeding experiments of rawsoy flour did induce growth of the rat pancreas, but this effecthas not been observed in similar studies in hamsters, mice,dogs, pigs, or monkeys.

P. Correa commented that the epidemiológica! studies havenot identified the causes of human pancreatic cancer, with thepossible exception of smoking. Protease inhibitors are unlikelycandidates since pancreatic cancer occurs with a lower thannormal frequency in vegetarian populations. In Seventh-DayAdventists protease inhibitor-containing lentils, peas, and soybeans have been identified as pancreatic cancer-suppressing

foods. Birk concluded that since protease inhibitors carry negligible health risks, the study of protease inhibitors as cancer-preventive agents offers opportunities for suppressing cancer inhumans. The progress made in identifying those inhibitorshaving anticarcinogenic activity and their mechanism(s) of action were the central themes of the Workshop.

D. L. Brandon reported on the successful measurement ofthe heat-labile protease inhibitor in soybeans, Kunitz trypsininhibitor, using monoclonal antibodies and enzyme-linked im-

munoassays. It was thought that such measurements of proteaseinhibitors by immunological techniques would be useful toepidemiologists as they attempt to establish the consumptionrate of protease inhibitors in diets and correlate these data withcancer incidence rates.

P. Flecker and J. Smart discussed the synthesis of the Bowman-Birk inhibitor through recombinant DNA techniques.Studies performed by Flecker on site-directed mutagenesis ofBBI have offered opportunities to identify the essential structural features of this protease inhibitor. Moreover, the possiblesynthesis of protease inhibitors by recombinant DNA techniques may make it possible to obtain sufficient quantities ofpure natural inhibitors for animal and human studies to evaluatetheir full potential as anticarcinogenic agents.

Identification of the genes which code for protease inhibitorsin potato and tomato plants was reported by C. A. Ryan. Amongthe richest sources of protease inhibitors are potato tubers. Forexample, tubers of the Russet Burbank variety contain over10% of their soluble proteins as two families of inhibitors,called Inhibitors I and II (monomer M, 8,100 and 12,300,respectively). The highest level of protease inhibitors in a natural food product is found in fruit from the wild tomato speciesLycopersicon esculentus. The unripened fruit of this species,although very small, contains over 50% of its soluble proteinsas Inhibitors I and II. In this fruit, as in potato tubers, synthesisof the inhibitors is under developmental regulation and is organspecific. In potato and tomato leaves the two inhibitors areunder environmental regulation, in that they accumulate in theleaves of the plants in response to insect attack and can reachlevels of 1-2% of the soluble protein in leaves within 24-48 hfollowing the initial attack. The genes coding for Inhibitors Iand II have been isolated from tomato and potato genomes.Complementary DNAs also have been isolated which code forthe wild tomato fruit protease inhibitors. Both traditional breeding and genetic engineering methodology can now be used toincrease the levels of Inhibitors I and II in potato tubers and/or tomato fruit. Thus, potato tubers and wild tomato fruit canbe considered potential sources for isolating large quantities ofprotease inhibitors to be studied as anticarcinogenic agents.

Perhaps the best present sources of pure protease inhibitorsfor anticarcinogenic studies are those inhibitors of microbialorigin made available through the work of H. Umezawa and K.Umezawa, as well as T. Aoyagi. The generosity of the JapanSociety for the Promotion of Science has made many of theseinhibitors available to investigators in the United States andhas helped confirm the findings that pure protease inhibitorsare capable of suppressing many animal cancers, neoplastictransformation, and oncogene expression. K. Umezawa presented this extensive list of protease inhibitors now availablefor study (Table 1).

Umezawa's report on the clinical use of bestaun in the

maintenance therapy of acute nonlymphocytic leukemia presents the first use of a protease inhibitor as a therapeutic agentin a clinical trial. His work emphasizes the usefulness of havingpure protease inhibitors available as possible therapeutic agents.The in vitro application of these pure protease inhibitors frombacteria has demonstrated that low concentrations of theseagents are capable of suppressing neoplastic transformation.

Kennedy showed that protease inhibitors such as antipainand the Bowman-Birk inhibitor are capable of suppressingmalignant transformation in C3H lOT'/z cells induced by chem

ical carcinogens and ionizing radiation. Not only were theseagents capable of suppressing transformation at low concentrations, but they also irreversibly blocked transformation when

500

on June 22, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MEETING REPORT

Table I Proteases and their inhibitors ofmicrobial origin

Peptidases Proteinase Inhibitors

Proteases -

—¿�Endo- —¿�

—¿�Serine- and cysteine-

Aspartic-

Metal-Peptidase

- Exo- -

—¿�Amino-

—¿�Dipeptidylamino-

1Carboxy-

•¿�Dipeptidylcarboxy-

Lcupeptin. antipain, chy-mostatin, elastatinal,Acetylleucylarginylal, Z-diprotin

Pepstatin, pepstanone, hy-droxypepstatin

Phosphoramidon

Amastatin. actinonin, ar-phamenines A and B.bestatin, ebelactones Aand B, formestin, probes-tin

Acetylleucylarginylal, anti-pain, leupeptin, diprotinsA and B

(.V)rr BL'II/>Imalk-acid,

histarginForoxymithine, histargin

given for only a short time. This suggests that a cellular proteaseis inactivated at a crucial time necessary for it to play a role inneoplastic transformation. Several candidate proteases weredescribed.

The direct contribution of proteases, the enzymes inhibitedby protease inhibitors, to carcinogenesis is still unclear. J.Yavelow reported that he has identified a chymotrypsin-likeprotease in the membranes of C3H lOT'/z cells. This protease

could be acting as a receptor for anticarcinogenic proteaseinhibitors. Consistent with this hypothesis, he recently reportedthe following results: (a) the chymotrypsin-like enzyme is anintegral plasma membrane protease; and (b) the enzyme hydro-

lyzes chymotrypsin substrates but differs from chymotrypsin inseveral respects; e.g., its pH optimum is 6.5 versus 7.5 forchymotrypsin.

The thrombin substrate i-butoxycarbonylvalylprolylarginine7-amido-4-methylcoumarin was utilized in studies by P. Billings

to identify a protease which is inhibited by anticarcinogenicprotease inhibitors, including BBI. He also searched for crucialcellular proteases using an affinity column; in these studies hehas found two proteases (M, 50,000-60,000), occurring inmouse and human cells, which interact directly with anticarcinogenic protease inhibitors. These enzymes may represent multiple cellular receptors for the protease inhibitors which suppress neoplastic transformation.

J. P. Quigley showed that a monoclonal antibody, specificallyreacting with avian plasminogen activator and inhibiting itsenzymatic activity, inhibits Rous sarcoma virus-induced neo

plastic transformation. The monoclonal antibody inhibition ofthe plasminogen activator is a precise demonstration that protease activity is necessary for the event leading directly tomalignant transformation. Plasminogen activator is anotherprotease which may interact with anticarcinogenic proteaseinhibitors.

K. Frenkel described a possible role of a chymotrypsin receptor in the formation of active oxygen species (Superoxide aniónradical, hydrogen peroxide, hydroxyl radical) in human neutro-

phils. Chymotrypsin inhibitor I completely blocks the effect ofthe phorbol ester tumor promoter TPA in its ability to induceH2Û2formation in cells, while the double-headed chymotryp-

sin/trypsin inhibitors BBI, turkey ovomucoid, and lima beanprotease inhibitors partially inhibit this TPA effect and trypsininhibitors, such as the soybean trypsin inhibitor, are virtuallyinactive. Since oxygen radicals contribute to tumor promotion,it has been hypothesized that the effective antipromoting activ

ity of chymotrypsin inhibitors is due to this prevention ofoxygen radical formation.

Another possible way in which protease inhibitors couldinhibit carcinogenesis is by suppressing oncogene expression.S. Garte showed that the protease inhibitors antipain, leupeptin,and f-aminocaproic acid inhibit transformation in NIH3T3cells after transfection with an activated H-ras oncogene. Hisresults suggest that transformation of NIH3T3 cells by a singlemutated oncogene may involve multiple stages and that part ofthis process is susceptible to inhibition by protease inhibitors.

J. Chang reported that c-myc RNA levels are reduced inproliferating normal, but not in transformed, mouse fibroblastsby antipain, leupeptin, BBI, and EP-475. Chang's data suggest

that the protease inhibitor suppression of c-myc expression isrelated to the ability of these agents to prevent the first stepinvolved in the malignant transformation of cells. The hypothesis that chemopreventive agents act by suppressing oncogeneexpression is a novel concept. Garte also reported that, likeprotease inhibitors, all-fra/w-retinoic acid will suppress ras-induced transformation in NIH3T3 cells. These results suggestthat the mechanism of action of many chemopreventive agentsmay involve effects on gene expression. It was suggested thatthe suppression of oncogene expression in cells growing inculture may serve as a test system for identifying and measuringthe potency of potential chemopreventive agents.

Proteases are known to play a role in many physiologicalprocesses. That proteases may be responsible for mediatingpathological conditions during differentiation and developmentwas indicated by studies demonstrating the antiteratogenic action of antipain. E. von Hofe showed that X-ray-induced ex-encephaly in mice is reduced if X-irradiated animals subsequently receive antipain. The mechanism of this antiteratogeniceffect may relate to the ability of antipain to block an abnormalproliferative response in the neuroepithelium of irradiated fetuses.

Proteases are known to play a major role in protein turnover.Dr. A. Goldberg has been working with proteases activated byATP that remove abnormal proteins selectively. Both animaland bacterial cells contain an ATP-dependent pathway thatselectively degrades abnormal proteins as well as certain normalregulatory proteins with short half-lives (e.g., oncogene products). The ATP-dependent protease in bacteria, and presumablyits mammalian counterpart, is known to hydrolyze two ATPsfor each peptide bond cleaved in proteins. In mammalian cells,the major neutral proteolytic activity is a M, 700,000 multi-catalytic protease complex which Goldberg calls a proteasome.In this complex a number of specific proteolytic activities havenow been identified. Several endoproteases exist in the mammalian cytosol whose functions are unknown, including: (a) apost-prolyl protease; (b) the "insulin-hydrolyzing metallopro-teinases"; and (c) the enzyme that hydrolyzes f-butoxycarbon-ylvalylprolylarginine 7-amido-4-methylcoumarin and is sensitive to the anticarcinogenic protease inhibitors (the proteasedescribed by Billings and mentioned above). The identificationof such proteases which interact with the anticarcinogenic protease inhibitors will be a major challenge of future studies inthis area of research.

R. H. Goldfarb has identified proteases which have a role inmetastasis and also contribute to killer cell activity; thus, theseproteases have a role in cancer cell invasion as well as in thekilling of cancer cells. The plasminogen activator thrombin andtype IV collagenase perhaps act in a cascade fashion in whichthey could contribute to the destruction of basement membranes. Protease inhibitors have been shown to prevent this

501

on June 22, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MEETING REPORT

action as well as metastasis of human tumor cells in nude mice.The immune response caused by interleukin 2 produces plas-minogen activator and other proteases that are used by naturalkiller cells to attack cancer cells. In this system, the proteolyticactivity must be viewed as anticarcinogenic, reminiscent of thedouble role of oxygen radicals in killing bacteria while at thesame time causing DNA damage to normal cells.

M. Pasternack has identified a serine protease from cytotoxicT-lymphocytes using the substrate /V-benzylcarbonyl-L-lysinethiobenzyl ester. This enzyme is part of the perforin system.The hypothesized exclusive role of serine esterases in the cyto-lytic process has been called into question recently by resultsfrom several different lines of investigation. Some proteaseinhibitors known to have anticarcinogenic activity are ineffective in blocking the perforin immune response of natural killercells. For example, tosyl-L-lysine chloromethyl ketone does notinhibit the proteases of natural killer cells.

In MCF-7 cells, a breast cancer cell line rich in estrogenreceptors, M. Levitz and T. Finlay observed that estradiolinduces two activities in these cells: (a) an estradiol esterasegiving rise to increased estradiol concentration, arising fromestradiol esters, and (b) an «i-antichymotrypsin either throughde novo synthesis or through the release of this inhibitor fromthe cell surface of a protease receptor complex. Whether therelease or synthesis of the chymotrypsin inhibitor has a role inmaintaining the transformed state of MCF-7 cells is not known.

Hoffmann raised the important point of the need for abiological marker that will indicate the consumption of proteaseinhibitors in humans, as has been successfully used in identifying smokers with nicotine metabolites. It may be possible touse the combination of trypsin and chymotrypsin in feces,identified by high performance liquid chromatography by Yav-elow and Troll, as a marker. Roebuck suggested that the induction of cholecystokinin in the pancreas by trypsin inhibitorsmay serve as a marker. More investigation to identify a biological marker is clearly necessary.

The final discussion, led by Wattenberg, focused on essentialwork required before protease inhibitors could be utilized ashuman cancer chemopreventive agents. It was noted at theoutset that the major requirements for such future researchwere the availability of preparations with a known proteaseinhibitor content and highly pure preparations of individual

protease inhibitors with defined characteristics. One immediatesource, for example, among others cited above, is foods thatare known to have high levels of protease inhibitors, such aslegumes and other vegetables. Such natural products are candidates for use in studies designed not only to investigate basicaspects of experimental anticarcinogenesis but also potentiallyfor use in studies directed at the prevention of cancer in humans.In summary, six major areas of research were envisioned asessential to further study protease inhibitors as potential humancancer chemopreventive agents: (a) research and developmentof sources of protease inhibitors and means of expediting theiravailability to the scientific community in both small and largequantities in defined states of content and purity; (A) analysisof human foods, including development of appropriate methodology where necessary, for protease inhibitor content; (c)réévaluationof epidemiológica! data regarding cancer occurrence in relation to known protease inhibitor content andcharacteristics in human foods; (tí)greatly expanded animalstudies on efficacy of protease inhibitors in cancer prevention.Studies should include, among other things, target organ definition, many more types of protease inhibitors, and dose-response characteristics in prevention; (e) investigations onpotentially adverse side effects of protease inhibitors, withstudies from the whole animal to the molecular level, includinginvestigations on possible effects of anticarcinogenic/antitrans-forming protease inhibitors on critical protease activities essential for normal physiological processes; (/) greatly expandedstudies on mechanisms of anticarcinogenesis of protease inhibitors. Among others, such studies might include investigationson the identification and characterization of target proteases ofanticarcinogenic protease inhibitors, mechanisms of antitransformation, effects on oncogene expression, possible importantrelevant interactions for prevention with host immune andendocrinological systems, and the role of suppression of oxygenradical production by protease inhibitors in anticarcinogenesis.

Walter Troll (ed.)New York University Medical CenterNew York, New York 10016

Ann R. Kennedy (ed.)University of PennsylvaniaSchool of MedicinePhiladelphia, Pennsylvania 19104

502

on June 22, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1989;49:499-502. Cancer Res     Inhibitors as Cancer Chemopreventive AgentsCancer Institute, National Institutes of Health: Protease Workshop Report from the Division of Cancer Etiology, National

  Updated version

  http://cancerres.aacrjournals.org/content/49/2/499

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

  .pubs@aacr.orgDepartment at

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/49/2/499To request permission to re-use all or part of this article, use this link

on June 22, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from