collagenase inhibitor

Proc. Natl. Acad. Sci. USAVol. 82, pp. 5380-5384, August 1985Cell Biology

Differentiation of a human leukemia cell line and expression ofcollagenase inhibitor

(HL60 cells/vitamin D3/retinoic acid/phorbol ester/monocytes)

Zvi BAR-SHAVIT*t, STEVEN L. TEITELBAUMt, GEORGE P. STRICKLINt, ARTHUR Z. EISEN§,ARNOLD J. KAHN*, AND HOWARD G. WELGUS¶*Division of Cell Biology, Washington University School of Medicine, St. Louis, MO 63110; and §Division of Dermatology, Department of MedicinetDepartment of Pathology and Laboratory Medicine, and ¶Division of Dermatology, Department of Medicine, The Jewish Hospital at WashingtonUniversity Medical Center, St. Louis, MO 63110; and tDivision of Dermatology, Department of Medicine, VA Medical Center/Universityof Tennessee Center for the Health Sciences, Memphis, TN 38163

Communicated by Paul E. Lacy, April 22, 1985

ABSTRACT A human collagenase inhibitor (CI) of Mr28,500 has been extensively characterized in skin fibroblastsand identified in a variety of connective tissues. Because humanalveolar macrophages synthesize and secrete both a collagenaseand CI that are immunologically and functionally identical totheir counterparts in fibroblasts, we studied the production ofsuch proteins by an immature human cell line (HL60) that canbe induced to differentiate along monocytic or granulocyticpathways. The cells failed to synthesize collagenase under anyculture condition tested. However, upon exposure to 1,25-dihydroxyvitamin D3 or phorbol esters (PMA), both of whichpromote monocytic differentiation of HL60, these cells synthe-sized and released CI in a dose-dependent manner. Further-more, the extent of CI expression was paralleled by theacquisition by such cells of the monocytic marker 63D3,indicating that inhibitor production and differentiation areclosely correlated. This CI was immunologically and function-ally identical to that produced by human macrophages andhuman skin fibroblasts. The quantity of CI synthesized byPMA-stimulated cells was 3- to 5-fold greater than produced byhuman alveolar macrophages, 1 ,ug per 106 cells per day. Incontrast, undifferentiated HL60 cells produced little or nodetectable CI (Z10-20 ng per 106 cells per day). Interestingly,when HL60 cells were stimulated to undergo granulocyticdifferentiation by dimethyl sulfoxide or retinoic acid, they alsoproduced the "monocytic" CI.

Collagenases catalyze the initial and rate-limiting step ofcollagen degradation and thus occupy a pivotal position inboth normal connective tissue turnover and in the tissuedestruction that accompanies certain pathologic processes(1-5). Such enzymes have been isolated from a number ofmammalian connective tissue cells (6-11) and inflammatorycells (12-15), including macrophages.

It currently appears that collagenase-mediated connectivetissue turnover is in reality a complex process that involvesmodulation of collagenolysis by specific CIs (16-19). It hasrecently shown that human alveolar macrophages may re-model collagen by this mechanism as the cells produce botha collagenase and collagenase inhibitor (CI) that are immu-nologically and functionally identical to the collagenase andCI elaborated by human fibroblasts (15). Specifically, the CIis a glycoprotein of Mr 28,500 that blocks enzymatic activitywith a high affinity and a 1:1 stoichiometry.

Since human alveolar macrophages elaborate proteins thatcan directly affect collagen degradation, the present studieswere undertaken to examine the regulation of such moleculesin cellular differentiation. To accomplish this, the human

promyelocytic leukemia cell line, HL60, was employed. Thiscell line responds to chemical inducers by differentiatingalong a monocytic lineage [by 4p-phorbol 12-myristate 13-acetate (PMA; refs. 20 and 21) or 1,25-dihydroxyvitamin D3(1,25(OH)2D3; refs. 22-26)] or along a granulocytic lineage[by dimethyl sulfoxide (Me2SO; refs. 20, 21, 27, and 28) orretinoic acid (20, 21)]. We report here that untreated HL60cells failed to synthesize or release either collagenase or CI;when induced to differentiate along monocytic or granulo-cytic pathways, however, they produced CI. Furthermore,monocytic differentiation was accompanied by elaboration ofquantities of CI 3- to 5-fold greater than were produced byhuman alveolar macrophages. In contrast to alveolar macro-phages, HL60s did not secrete collagenase under any testedcondition.

MATERIALS AND METHODS

Cells. HL60 cells, a human promyelocytic leukemia cellline (20) were obtained from E. Huberman (Argonne NationalLaboratories). The cells were grown in Iscove's modifiedDulbecco's medium containing transferrin at 5 mg/ml, bo-vine serum albumin at 5 mg/ml, and 2 mM glutamine andwere maintained in humidified 95% air/5% CO2 at 37°C.

Cell Treatment. The HL60 cells were plated in flat-bot-tomed wells (16-mm diameter) at 5 x 105 cells per well per mlof culture medium. When specified, the cells were treatedwith PMA (Sigma), retinoic acid (Sigma), Me2SO (Sigma), orvitamin D3 metabolites (courtesy of Milan Uskokovic, Hoff-mann-La Roche). At the indicated times, the conditionedmedium was removed and adjusted to 0.05 M Tris HCl/0.01M CaC12, pH 7.5, or dialyzed overnight against 5 mMTris HCl/1 mM CaCl2, pH 7.5, then lyophilized and suspend-ed in water to achieve a 10-fold concentration. For thebiochemical characterization studies, cells were plated inlarge tissue culture dishes (150 mm) and then exposed to thevarious inducing agents.Flow Cytometry. HL60 cells were treated with different

vitamin D3 metabolites (see Fig. 3). Induction of monocyticdifferentiation was assessed by monitoring for the appear-ance of 63D3 (a specific human monocyte surface marker)with an EPICS V flow cytometer (Coulter) as described (22).Analysis of antibody binding (percent positive cells) wasperformed on a TERAK LSI/1123 microcomputer using theprogram IMMUNO written by C. Bruce Bagwell (Coulter).

Abbreviations: CI, collagenase inhibitor; PMA, 4,8-phorbol 12-myristate 13-acetate; 1,25(OH)2D3, 1,25-dihydroxyvitamin D3;Me2SO, dimethyl sulfoxide.This work was presented in part at the meeting of the AmericanSociety of Cell Biology (33).

5380

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. NatL Acad Scd USA 82 (1985) 5381

Chromatography. Conventional chromatography was car-ried out at 40C, with column effluents monitored at 280 nm.High pressure liquid chromatography was carried out atambient temperatures (160-20'C) using a Waters Associatessystem. Conditioned medium (200 ml) from PMA-treatedHL60 cells was dialyzed against 0.05M Tris HCl, pH 7.5, andapplied to a 2.5 x 10 cm column of phosphocellulose (P-li,Whatman). The column was washed with 0.05 M Tris-HCI,pH 7.5, at a flow rate of 25 ml/hr for 24 hr and the boundprotein was eluted with 0.05 M TrisHCl, pH 7.5/0.5 M(NH4)2SO4. The fall-through fractions contained >90% ofthetotal protein applied to the column while the eluted materialcontained >90% ofthe CI as measured by an ELISA (18, 19).An aliquot of the eluted protein was then subjected to HPLCusing a Bio-Rad TSK-250 (7.5 x 300 mm) gel filtrationcolumn. The column was equilibrated with 0.05 M Tris-HCl,pH 7.5/0.10 M Na2SO4 and eluted at 0.8 ml/min. Half-minutefractions were collected and the absorbance at 229 nm wasrecorded.

Immunoprecipitation. HL60 cells were exposed to cys-teine-free medium containing 80 uCi (1 Ci = 37 GBq) of[35S]cysteine (New England Nuclear) per ml for 24 hr. Fifty-to 150-,ul aliquots ofmedium containing labeled cysteine in atotal volume of 300 1.I (buffer = 0.04 M sodium phosphate,pH 8.0/0.15M NaCl/0.1% Triton X-100/1mM phenylmethyl-sulfonyl fluoride) were incubated with 5 pul of fibroblast CIantibody overnight at 40C. Immunoglobulin was precipitatedwith IgGsorb (The Enzyme Center). The precipitates werewashed extensively, boiled in sample buffer (65 mMTris HCl, pH 6.8/8 M urea/3% sodium dodecyl sulfate) andsubjected to polyacrylamide gel electrophoresis. Radioau-tography was performed using Kodak XR-5 X-Omat R filmdeveloped at -70'C (29).

Collagenase and CI Assays. Collagenase was assayed im-munologically by an ELISA (30) and catalytically usingreconstituted [14C]glycine-labeled guinea pig type I collagenfibrils [25,000 cpm/mg (31)]. The solubilized products werequantitated in a scintillation counter. Enzyme activity wasalso screened using monomeric human type I and III collagenas substrate (5).To determine CI activity, concentrated conditioned medi-

um was treated with EDTA (20 mM) to destroy any intrinsicenzyme, exhaustively dialyzed, and then assayed for capac-ity to inhibit pure fibroblast collagenase (9). The molarstoichiometry of inhibition was quantified by calculatingmoles of collagenase inhibited by HL60-conditioned mediumper mol ofimmunoreactive protein [determined by an ELISA(19) for fibroblast CI].

Intracellular CI was measured by an ELISA (19) from cellextracts prepared by scraping the cell layer into 1 ml of buffer(0.05 Tris HCl, pH 7.5) followed by sonication and centrifu-gation (2500 X g for 20 min).

RESULTSConditioned medium collected from untreated HL60 cells didnot possess collagenase or CI activity and similarly lackedmaterial cross-reactive with antibodies directed against anal-ogous molecules from fibroblasts. However, when the cellswere induced to differentiate along a monocytic pathway byPMA or 1,25(OH)2D3 or, alternatively, along a granulocyticpathway by retinoic acid or Me2SO, significant quantities ofimmunologically detectable CI were released into the culturemedium (Fig. 1). In all cases, by 2 days after induction morethan 90%o of the CI was secreted into the culture medium andcellular extracts contained less than 10% of the total immu-nologically detectable protein (data not shown).The kinetics of CI production following exposure ofHL60

cells to the various differentiating agents is shown in Fig. 2A(for monocytic differentiation) and Fig. 2B (for granulocytic

4

3

2

1

00

00.

1.5

1.0

0.5

PMA Retinoic acid

1 10 100 0.3 3 30 300ng/ml AM

1 ,25(OH)2D3 Me2O0

0.1 1 10 100 0.125 1.25 5nM % (vol/vol)

FIG. 1. Dose-response of CI release from HL60 cells followingexposure to differentiating agents. Conditioned media were collected4 days after the addition ofPMA or 1,25(OH)2D3 and 5 days after theaddition of retinoic acid or Me2SO. CI was measured by an ELISA(19).

differentiation). Both the time course of (Fig. 2) anddose-response to (Fig. 1) the inducing substance suggest thatCI synthesis correlates with differentiation/maturation ofthe

3

(a

= 20

0

0

0.T-

"1cL

U3

2

1~

0 1 2 3 4 5 6 7Time, days

FIG. 2. Time course of CI production in differentiating HL60cells. (A) Monocytic differentiation. *, PMA at 20 ng/ml; A, 10 nM1,25(OH)2D3. (Inset) Early kinetics of the PMA effect. o, Total CI;*, intracellular CI. (B) Granulocytic differentiation. *, 30 /AMretinoic acid; c, 2.5% (vol/vol) Me2SO.

A

1.5

0 1.0

0.5

0 B8 16 24Hours

l_/ a I IB e

/~~~~I A-

4

6

I

Cell Biology: Bar-Shavit et aL

4

5382 Cell Biology: Bar-Shavit et al.

progenitor cells. Intracellular CI appeared in PMA-treatedcells 4-6 hr after induction (Fig. 2A Inset). Similarly, thoseagents that promoted monocytic differentiation [i.e., PMAand 1,25(OH)2D3] induced maximum secretion of CI withinthe first 24-48 hr of treatment (Fig. 2A). In contrast,"granulocyte-maturing" compounds (i.e., Me2SO andretinoic acid) exerted their maximum effects after the secondday (Fig. 2B). These data correspond to the acquisition byHL60 cells of monocytic or granulocytic markers whensubjected to the various differentiating agents (20-28).The appearance of the huiran monocytic surface marker,

63D3, on the HL60 cells could be utilized to monitor theextent ofmonocytic differentiation induced by agents such as1,25(OH)2D3 (22). In Fig. 3, the capacity of different vitaminD3 metabolites to induce the expression of 63D3 on cellsurfaces and the secretion of CI into culture medium arecompared. Both 1,25(OH)2D3 and 1,25,26-trihydroxyvitaminD3 promoted monocytic differentiation (appearance of 63D3)and CI expression in a parallel fashion when HL60 cells wereexposed to concentrations of these agents ranging from 0.1nM to 0.1 ,uM (data not shown). The vitamin D metabolite25-hydroxyvitamin D3 was capable of inducing 63D3 expres-sion and CI release only when present in high concentration.Other metabolites such as 24,25- and 25,26-dihydroxyvitaminD3 totally failed to promote monocyte differentiation or CIsecretion (Fig. 3). Similar conclusions were obtained whenother markers of monocytic differentiation (adherence toplastic, morphology, and nonspecific esterase) were assessed(data not shown). Thus, in all instances, the extent of CIproduction appeared to vary precisely with the degree ofcellular differentiation.

Since HL60 cells induced to differentiate along either amonocytic or a granulocytic lineage elaborated a proteindetectable by antibody to human skin fibroblast CI (Figs.1-3), studies were undertaken to assess the extent of struc-tjiral similarity between the HL60 and fibroblast-derivedantigens. HL60 cells exposed to PMA, 1,25(0H)2D3, Me2SO,or retinoic acid were grown in the presence of [35S]cysteine(see Materials and Methods). Exposure of the radiolabeledmedium-proteins to specific antibody directed against fibro-blast CI (18, 19) resulted in the precipitation of a major bandof electrophoretic mobility (Mr -28,000) identical to that ofthe fibroblast protein. This band was observed only withCI-specific antibody and in cells exposed to each of thedifferentiating agents but not in untreated cells (Fig. 4). Tofurther assess the nature of this immunoreactive material, 0.2

AB CDEF

* X.' ..... ':.

CTI

FIG. 4. Immunoprecipitation of HL60 cell-derived CI. HL60cells either untreated or exposed to various differentiating agentswere pulsed with [35S]cysteine for 24 hr and the medium wasprocessed as detailed in Materials and Methods. Lanes A-E containinhibitor-specific antibody: A, PMA-treated cells; 14, 1,25(OH)2D3-treated cells; C, Me2SO-treated cells; D, retinoic acid-treated cells;E, untreated cells. Lane F: PMA-treated cells plus nonimmune IgG.The doublet CI band present in lanes B and D results from M,differences related to degree of inhibitor reduction as reported (18).

liters of conditioned medium from PMA-exposed cells wasprocessed through two of the chromatographic steps used topurify human skin fibroblast CI (18). Ninety percent of theimmunoreactive material was bound to phosphocellulose atpH 7.5 as compared to only 10% of the total applied protein.The protein eluted from the phosphocellulose column wasthen subjected to gel filtration using a TSK-250 HPLCcolumn. Both the immunoreactive material and functionalanti-collagenolytic activity migrated in a position identical tohuman skin fibroblast CI (Fig. 5). Furthermore, protein fromthis inhibitor peak gave a reaction of complete identity withpure fibroblast CI when tested by double immunodiffusion(data not shown).The functional collagenase inhibitory activity present in

conditioned medium from HL60 cells exposed to the variousinducing agents was next examined. As shown in Fig. 6, suchinhibitory activity invariably exhibited an approximately 1:1molar stoichiometry with respect to exogenously addedcollagenase.The ability of the HL60 cells to synthesize and secrete

collagenase following their exposure to various inducingagents was also examined. Conditioned medium thus ob-

O % 63D3+P0g Cl/106 cells

T

1,25(OH)2D3 1,25,26(0)3 25(0H)D3 25(0H)D3 24,25(O)2 25,26(0*H)203

10 nM I 1 A4m '

1.25

1.00 (aOD00

0.75-2

0.50 I

0.25

FIG. 3. Correlation ofCI production and appearance of the 63D3 marker on cell surfaces in response to various vitamin D3 metabolites. Cellswere incubated for 4 days with the indicated metabolites. Untreated cells did not produce detectable CI and only 2% expressed 63D3 on theircell surface.

100

75FCO)

0

._4

CO)0COco

0-

501-

25

Proc. Nad Acad Sd USA 82 (1985)

Proc. Natl. Acad Sci USA 82 (1985) 5383

0.5~~~~~~~~~~~~~~~~~~~~~~.

10~~~~~~~~~~~~~~~~~~02030 4

0.4-10.09~103 ~~~~~~~0oo

0.3-.~ oa'~~ ~~~~ s 0)

6.0 o6O400.2 4.0 40

2.0 3 20~EE

0.0 ~5~ 0.0 0

10 20 30 40Fraction

FIG. 5. Chromatographic behavior of HL60 cell-derived CI. Conditioned medium from PMA-treated HL60 cells was applied to aphosphocellulose column and eluted as described in Materials andMethods. An aliquot ofthe eluted material was then subjected to high pressureliquid chromatography using a Bio-Rad TSK-250 gel filtration column. The arrow denotes the elution position of pure human fibroblast CI.A229; *, immunoreactive inhibitor; o, % inhibition of collagenase activity.

tained failed to exhibit either enzymatic acmonomeric or fibnllar type I and III collagensmaterial cross-reactive with anti-human clELISA (30). Treatment of the differentiated cpolysaccharide or latex particles did not altervations.

DISCUSSIONThese studies show that differentiation of HLeither a monocytic or granulocytic pathway isby production and secretion of a CI. This pronologically and functionally identical to the CInormal human skin fibroblasts and by humanrophages.The expression of a CI in HL60 cells ap

directly related to the extent of differentiatexposure of the cells to various inducing agentsby dose-response (Fig. 1) and time-course (FFurthermore, the ability of vitamin D3 metabmote CI expression was directly correlated to Ito induce the appearance of the human monc63D3 on cell surfaces and thus to their potenctiating factors (Fig. 3). In the undifferentiate(cells secreted little or no CI (_10-20 ng per 106c

100 _

80 -

0

:E 60 -

~~60~~

20

0.5 1.0

Inhibitor/collagenase (mol/mol)

FIG. 6. Molar stoichiometry of CI. The capacity oconditioned medium from HL60 cell cultures to inhibfibroblast collagenase was quantified as detailed inMethods. An approximate 1:1 molar stoichiometryfrom medium obtained after exposure of the cells totiating agent. *, Me2SO; 0, 1,25(OH)2D3; A, PMA; A,

Ativity against Both monocytic and granulocytic differentiation resulted in a; or significant marked increase in production of the protein.ollagenase in Human alveolar macrophages produce both a collagenase-ells with lipo- and CI that are immunologically and functionally identical tor these obser- the analogous proteins elaborated by human skin fibroblasts

(15). Production of the macrophage collagenase is markedlystimulated by the addition of lipopolysaccharide to culturemedium while CI secretion is significantly enhanced byexposing the cells to either PMA or lipopolysaccharide (15).

,60 cells along Collagenase activity has also been studied in human poly-accompanied morphonuclear leukocytes (14) and such cells have been)tein is immu- shown to elaborate an enzyme that differs catalytically fromelaborated by its macrophage/fibroblast counterpart, particularly with re-alveolar mac- gard to substrate specificity for different genetic collagen

types (32). We were unable to detect any collagenase activitypeared to be or immunoreactive material in the HL60 cells, either untreat-ion following ed or exposed to agents inducing monocytic or granulocyticas evidenced differentiation, either in the presence or absence of lipopoly-

ig. 2) curves. saccharide or latex particles. Production of CI, however, wasolites to pro- approximately 3- to 5-fold greater in HL60 cells induced totheir capacity differentiate along a monocytic pathway (=1 ,ug per 106 cellscytic marker per day) as compared to human alveolar macrophages treatedy as differen- with PMA or lipopolysaccharide (0.3 ,ug per 106 cells per 24d state HL60 hr). This set of circumstances indicates that CI gene expres-cells per day). sion may be more enhanced in HL60 cells exposed to agents

promoting monocyte differentiation than in stimulated mac-rophages. In marked contrast, the HL60 cells do not acquirethe capacity to produce collagenase. These observations mayreflect fundamental differences between differentiating HL60cells and authentic macrophages or, alternatively, the rela-tive uniqueness of the alveolar macrophage as a collagenase-producing mononuclear phagocyte.HL60 is developmentally bipotential, giving rise to

nonproliferative cells with some granulocyte characteristicsin response to one set of inductive agents [Me2SO or retinoicacid (20, 21)] and to cells with monocytic features in responseto others [PMA or 1,25(OH)2D3] (20-26). While normalhuman macrophages have been shown to secrete an identicalCI to that produced by fibroblasts (15) or HL60 cells, human

L|...... I | . polymorphonuclear leukocytes do not contain immunologi-1.5 1.9 2.0 cally detectable quantities of this protein (18). Thus, it is

interesting that HL60 cells exposed to the granulocytic-maturing agents, Me2SO or retinoic acid, are induced tosecrete large amounts of CI. One possible explanation for this

fit pure human behavior is that although Me2SO and retinoic acid-treatedMaterials and HL60 cells are considered granulocytic, the differentiation iswas observed not complete and the cells despite being committed along aeach differen- granulocytic pathway may still express some monocyticretinoic acid. characteristics. Since monocytes and granulocytes are de-

Cell Biology: Bar-Shavit et aL

5384 Cell Biology: Bar-Shavit et al.

rived from a common precursor, it would be of interest todetermine whether progenitor granulocytic cells in the bonemarrow temporarily express some monocyte characteristicsas part of the normal granulocyte differentiation process.

We thank Nadean Brown and Catherine Fliszar for their excellenttechnical assistance and Jane Wodicker for secretarial assistance.This work was supported by Grants AM34401 (to Z.B.), DE05413and AM32788 (to S.L.T.), AM35805 (to H.G.W.), DE04629 (toA.J.K.), and AM12129 (to A.Z.E.) from the National Institutes ofHealth. Z.B. is a recipient of an Arthritis Foundation InvestigatorAward and a National Institutes of Health Biomedical ResearchSupport Grant award to Washington University School of DentalMedicine. H.G.W. is the recipient of National Institutes of HealthResearch Career Development Award AM01525. G.P.S. is a ClinicalInvestigator of the Veterans Administration. This work was alsosupported by a grant from The Council for Tobacco Research-U.S.A., Inc.

1. Gross, J. & Nagai, Y. (1965) Proc. Natl. Acad. Sci. USA 54,1197-1204.

2. Harris, E. D., Jr., Evanson, J. M., DiBona, D. R. & Krane,S. M. (1970) Arthritis Rheum. 13, 83-94.

3. Birkedal-Hansen, H. (1980) in Collagenase in Normal andPathological Connective Tissues, eds. Woolley, D. E. &Evanson, J. M. (Wiley, New York), pp. 128-140.

4. Welgus, H. G., Jeffrey, J. J., Stricklin, G. P., Roswit, W. T.& Eisen, A. Z. (1980) J. Biol. Chem. 255, 6806-6813.

5. Welgus, H. G., Jeffrey, J. J. & Eisen, A. Z. (1981) J. Biol.Chem. 256, 9511-9515.

6. Harris, E. D., Jr., & Krane, S. M. (1974) N. Engl. J. Med. 291,557-563.



9. Stricklin, G. P., Bauer, E. A., Jeffrey, J. J. & Eisen, A. Z.(1977) Biochemistry 16, 1607-1615.

10. Dayer, J.-M., Breard, J., Chess, L. & Krane, S. M. (1979) J.Clin. Invest. 64, 1386-1392.

11. Cawston, T. E. & Tyler, J. A. (1979) Biochem. J. 183,647-656.

12. Wahl, L. M., Wahl, S. M., Mergenhagen, S. E. & Martin,G. R. (1974) Proc. Natl. Acad. Sci. USA 71, 3598-3601.

13. Werb, Z. & Gordon, S. (1975) J. Exp. Med. 142, 346-360.14. Hasty, K. A., Hibbs, M. S., Kang, A. H. & Mainardi, C. L.

(1984) J. Exp. Med. 159, 1455-1463.15. Welgus, H. G., Campbell, E. J., Bar-Shavit, Z., Senior, R. M.

& Teitelbaum, S. L. (1985) J. Clin. Invest. 76, 219-224.16. Vater, C. A., Mainardi, C. L. & Harris, E. D., Jr. (1979) J.

Biol. Chem. 254, 3045-3053.17. Cawston, T. E., Murphy, G., Mercer, E., Galloway, W. A.,

Hazleman, B. L. & Reynolds, J. J. (1983) Biochem. J. 211,313-318.

18. Stricklin, G. P. & Welgus, H. G. (1983) J. Biol. Chem. 258,12252-12258.

19. Welgus, H. G. & Stricklin, G. P. (1983) J. Biol. Chem. 258,12259-12264.

20. Lotem, J. & Sachs, L. (1979) Proc. Natl. Acad. Sci. USA 76,5158-5162.

21. Rovera, G., Santoli, D. & Damsky, C. (1979) Proc. Natl.Acad. Sci. USA 78, 2779-2783.

22. Bar-Shavit, Z., Teitelbaum, S. L., Reitsma, P., Hall, A., Pegg,L. E., Trial, J. & Kahn, A. J. (1983) Proc. Natl. Acad. Sci.USA 80, 5907-5911.

23. McCarthy, D. M., San Miquel, J. F., Freake, H. C., Greene,P. M., Zola, H., Catovsky, D. & Goldman, J. M. (1983)Leukemia Res. 7, 51-55.

24. Mangelsdorf, D. J., Koeffler, P. H., Donaldson, C. A., Pike,W. J. & Haussler, M. R. (1984) J. Cell Biol. 98, 391-398.

25. Murao, S., Gemmell, M. A., Callahamn, M. F., Anderson,N. L. & Huberman, E. (1983) Cancer Res. 43, 4989-4996.

26. Reitsma, P. H., Rothberg, P. G., Astrin, S. M., Trial, J.,Bar-Shavit, Z., Hall, A., Teitelbaum, S. L. & Kahn, A. J.(1983) Nature (London) 306, 492-494.

27. Breitman, T. R., Selonick, S. E. & Collins, S. J. (1980) Proc.Natl. Acad. Sci. USA 77, 2936-2940.

28. Collins, S. J., Ruscetti, F. W., Gallagher, R. E. & Gallo, R. C.(1978) Proc. Natl. Acad. Sci. USA 75, 2458-2462.

29. Lasky, R. A. & Mills, A. D. (1975) Eur. J. Biochem. 56,335-341.

30. Cooper, T. W., Bauer, E. A., & Eisen, A. Z. (1982) CollagenRel. Res. 3, 205-216.

31. Nagai, Y., Lapiere, C. M. & Gross, J. (1966) Biochemistry 5,3123-3130.

32. Horwitz, A. L., Hance, A. J. & Crystal, R. G. (1977) Proc.Natl. Acad. Sci. USA 74, 897-901.

33. Bar-Shavit, Z., Teitelbaum, S. L., Eisen, A. Z., Kahn, A. J. &Welgus, H. G. (1984) J. Cell Biol. 99, 89a (abstr.).

Proc. NatL Acad Sci. USA 82 (1985)

collagenase inhibitor

Documents