thorax effect mycobacterium matrix · thorax 1996;51:306-311 effect ofmycobacterium tuberculosis...

Thorax 1996;51:306-311

Effect of Mycobacterium tuberculosis and itscomponents on macrophages and the release ofmatrix metalloproteinases

Jui C Chang,* Annette Wysocki, Kam-Meng Tchou-Wong, Naomi Moskowitz,Yihong Zhang, William N Rom

Background - Pulmonary tuberculosis isassociated with caseating necrosis, par-enchymal lung destruction, and cavityformation. It was hypothesised that tuber-culous lung destruction is mediated, atleast in part, by the participation ofmatrixmetalloproteinases released by mono-nuclear phagocytes.Methods - Cells of the myelomonocyticleukaemia cell line THP-1 were incubatedwith lipoarabinomannan (LAM), themajor antigenic cell wall component, andwith Mycobacterium tuberculosis andanalysed by Northern blot analysis. Twopatients with active cavitary tuberculosisalso underwent bronchoalveolar lavageand the cells were analysed by Northernblotting.Results - Incubation of THP-1 cells withLAM resulted in the stimulated releaseof matrix metalloproteinase-9 (MMP-9),a 92 kDa gelatinase, by 24 hours in a dose-dependent fashion. In addition, Northernanalysis revealed that LAM upregulatedthe gene for MMP-9 by 24 hours, but notthe gene for the 72 kDa gelatinase MMP-2. Heat killed M tuberculosis H37Ra alsoupregulated the MMP-9 gene. Broncho-alveolar lavage of the two patients withactive cavitary tuberculosis showed strik-ing upregulation of the MMP-9 gene com-pared with a normal control usingNorthern analysis. LAM also upregulatedthe type I interstitial collagenase (MMP-1) gene by 24 hours in both THP-1 cellsand peripheral blood monocytes.Conclusions - These data suggest that Mtuberculosis and its major cell antigeniccomponent, LAM, stimulate the release ofMMP-9 and upregulate the expression ofgenes for MMP-1 and MMP-9. It is pos-sible that M tuberculosis and its com-ponents contribute directly to cavityformation by their ability to stimulatemacrophages to release matrix metallo-proteinases that digest collagens I-IV, andindirecdy by stimulating the release ofthecytokines interleukin 1, and tumour nec-rosis factor a that induce fibroblasts toamplify the release of matrix metallo-proteinases.(Thorax 1996;51:306-311)

Keywords: collagenases, matrix metalloproteinases,tuberculosis, cavity.

Central caseating necrosis is the hallmark oftuberculous granulomas. Tissue caseation leads

to destruction of lung tissue with cavity forma-tion. The granulomatous lesion is tyically a cellmediated response involving macrophages andlymphocytes as the immunoresponsive cells.Fibroblasts, one of the major connective tissuecells, are also involved in the course of theacute and chronic phase of granulomatous in-flammation as evidenced by the deposition ofcollagen fibres in the granulomatous lesion.'Advanced pulmonary tuberculosis is associatedwith a locally destructive process of cavitarylesions which plays an important part in trans-mission of disease. The pathogenesis of thedestructive lung disease caused by Myco-bacterium tuberculosis is still poorly understood.Dannenberg' suggested that the tissue damagesuch as liquefaction, caseation, and cavityformation was the result of a hypersensitivityreaction - that is, excess bacillary antigen caus-ing host tissue or cells to produce cytotoxiccytokines, oxygen radicals, hydrolytic enzymesand eventual cell death.3 Since the degradationof connective tissue matrix involves an initialextracellular cleavage of insoluble proteins me-diated by neutral proteinases such as elastasesand collagenases,4 it would be important toassess neutral matrix metalloproteinase activityin host cells during tuberculosis infection.5 Totest the hypothesis that M tuberculosis or itscomponents modulate the synthesis of matrixmetalloproteinases, especially collagenases, wehave studied the effects ofM tuberculosis andlipoarabinomannan (LAM), the major anti-genic component of the mycobacterial cellwall,67 on the stimulation of matrix metallo-proteinases in human peripheral blood mono-cytes and a myelomonocytic leukaemia cell line(THP-1).7 Bronchoalveolar lavage (BAL) wasalso performed in two patients with active ca-vitary tuberculosis to obtain BAL cells to assessexpression of the matrix metalloproteinase-9gene. Since we8 and others9-" have shown thatLAM can stimulate the release of tumor nec-rosis factor oa (TNF-cx) and interleukin 1 ,B (IL-1 I) from mononuclear phagocytes, and sincethese two cytokines are known to stimulatecollagenase production in synovial cells,4 15 weevaluated gene expression of the type I inter-stitial collagenase following stimulation ofhuman lung fibroblasts.

MethodsCELL CULTURE AND STIMULATIONLipoarabinomannan (LAM) from laboratoryattenuated Mycobacterium spp was kindly pro-vided by Dr P Brennan, Fort Collins, Colorado,

Division of Pulmonaryand Critical CareMedicine,Departments ofMedicine,EnvironmentalMedicine and Surgery,and Bellevue HospitalChest Service, NewYork UniversityMedical Center, NewYori, NY 10016, USAJ C ChangA WysockiK-M Tchou-WongN MoskowitzY ZhangW N Rom

*Current address:Department of VeteransAffairs Medical Center,Mount Sinai School ofMedicine, Bronx, NewYork, USA.

Correspondence to:Dr W N Rom.

Received 24 January 1994Returned to authors6 May 1994Revised version received25 September 1995Accepted for publication23 October 1995

306

on April 27, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.3.306 on 1 M

arch 1996. Dow

nloaded from

http://thorax.bmj.com/

Effect ofM tuberculosis on macrophage matrix metalloproteinases

USA. Evaluation of tuberculosis reagents forthe presence of Gram negative bacterial endo-toxin was performed with the amoebocytelysate assay (E-Toxate Kit, Sigma). Alipopolysaccharide (LPS) standard curve wasgenerated with an assay limit of 1 pg/ml LPS.Six different batches of LAM from Myco-bacterium spp were tested and all contained<5pg/ml LPS or <5pg/pg test reagent. TheLAM had been passed through a Detoxigelcolumn to remove any potential LPS. Lipo-polysaccharide from Escherichia coli was pur-chased from Sigma;M tuberculosis H37Ra (heatkilled) was obtained from American Type Cul-ture Collection (ATCC), Rockville, Maryland;human lung fibroblasts (CCL 190) and themyelomonocytic leukaemia cell line (THP-1)were obtained from ATCC; human collagenasecDNA (pC1lase 1) was obtained from ATCCand was cloned from TPA treated human skinfibroblasts from a healthy 29 year old man; the92 kDa and 72 kDa cDNAs for matrix metallo-proteinases were a gift ofDr Gregory Goldberg,Washington University, St Louis, Missouri; andthe pHe7 cDNA probe was used as a controlhousekeeping gene.THP-1 cells were grown in RPMI 1640

medium at 37°C, 95% air/5% carbon dioxidesupplemented with 10% fetal calf serum and100 U/ml penicillin plus 100 pg/ml strepto-mycin. Peripheral blood monocytes were isol-ated from human buffy coat obtained fromtuberculin negative individuals. Mononuclearcells were isolated from blood components bylymphocyte separation medium (LSM, TeknitaCorp, Durham, North Carolina), centrifugedand washed three times with RPMI 1640.Mononuclear cells were then suspended inRPMI supplemented with 10% fetal calf serumand antibiotics. The cells were plated in a250 ml culture flask at 2 x 106 cells/ml. Afterallowing the cells to adhere for 1-5 hours at370C the cells were washed three times withice cold phosphate buffered saline (PBS). Theadherent cells, which consisted of more than95% monocytes, were exposed to test reagentsfor up to 24 hours in serum-free media.Human lung fibroblasts were cultured at

37°C in humidified 90% air/10% carbon di-oxide in DMEM supplemented with 10% fetalcalf serum and penicillin plus streptomycin. Atconfluence, fibroblasts were treated with IL-1oc (5 U/ml) or TNF-cx (10 ng/ml) in serum-free medium.

BRONCHOALVEOLAR LAVAGEBronchoalveolar lavage was performed witha flexible fibreoptic bronchoscope with localxylocaine anesthaesia. Normal saline (3 x 50 mlaliquots) was instilled into the radiographicallyinvolved site in tuberculosis patients, and5 x 20 ml aliquots were instilled in three differ-ent sites and pooled for the normal volunteer.The recovered fluid was filtered through sterilegauze. A total cell count was performed in ahaematocytometer and cell viability was de-termined by trypan blue exclusion. Broncho-alveolar lavage cells were immediately placed in5.5M guanidinium isothiocyanate buffer and

frozen at - 70°C. Two patients with activepulmonary tuberculosis confirmed by sputumculture ofM tuberculosis with radiographic cav-ities and one normal volunteer were lavaged.All three individuals were HIV negative andboth patients with active tuberculosis had re-ceived isoniazid, rifampicin, pyrazinamide, andethambutol for two weeks before lavage. Theclinical research protocol had been approvedby the human subjects review committees ofNew York University Medical Center andBellevue Hospital.

NORTHERN ANALYSISPreparation of total RNA from monocytes,THP-1 cells, or bronchoalveolar lavage cellswas carried out according to the guanidiniumisothiocyanate method. Briefly, cells werewashed three times and lysed by 5-5M guan-idinium solution containing 5 mM sodiumcitrate, 0 5% Sarkosyl, and 1% ,B-mercapto-ethanol. The RNA was then pelleted througha caesium chloride gradient and dissolved inRNA suspension buffer (0 5% SDS, 10 mMEDTA, 10mM Tris pH 8-0, 0-2 M NaCl).

Electrophoresis of the sample was performedin a 1% agarose/7% formaldehyde gel and thegel was transferred to a NYTRAN membrane(Schleicher and Schuell, Keene, New Hamp-shire, USA) by capillary blotting. NYTRANfilters were hybridised with respective cDNAslabelled with 32p using random primer labelling(Boehringer Mannheim). Hybridisations werecarried out for 20 hours at 42°C and washingwas performed three times for 30 minutes in2 x SSC/01% SDS at 22°C and then twice for30 minutes in 0 1 x SSC/0 1% SDS at 650C.Filters were exposed to Kodak X-Omat ARfilm with intensifying screens at -700C for2-8 days.

SUBSTRATE GEL ELECTROPHORESISProteinase profiles were determined by sub-strate gel electrophoresis using 8% acrylamidegels containing gelatin at a final concentrationof 4-75 mg/ml.'6 Conditioned media fromTHP-1 cells that had been stimulated withvarying concentrations of LAM or LPS(500 pg/ml, 1 ng/ml, 500 ng/ml, 1 pg/ml) weresubjected to SDS-PAGE under non-reducingconditions at 22°C. Each lane was loaded with30 ptg protein. After electrophoresis the gelswere washed twice for 30 minutes with 2-5%Triton-X-100 to remove SDS. Gels were thenbriefly rinsed with deionised water and thenincubated overnight at 370C in buffer con-taining 50mM Tris HCI, 150mM NaCl, and5 mM CaCl2 (pH 7 4). Following incubation,gels were stained with 0- 1% Coomassie brilliantblue and destained. Proteolytic activity ap-peared as clear zones against a dark blue back-ground.

ResultsDEMONSTRATION OF MMP-9 AND MMP-2Substrate gel electrophoresis demonstrated agelatinase with an apparent molecular mass of

307



j.com/


arch 1996. Dow

nloaded from


Chang, Wysocki, Tchou- Wong, Moskowitz, Zhang, Rom

Figure 1 Substrate gel electrophoresis of conditioned media collected at 24 hours fromTHP-1 cells. Lane 1, medium alone (negative control); lane 2, LAM (500pglml); lane3, LAM (500 nglml); lane 4, LAM (1 ,uglml); lane 5, LPS (500pglml); lane 6, LPS(1 nglml); lane 7, LPS (500 nglml); lane 8, LPS (1 pglml). Molecular weight markers(kDa) are indicated along the margins.

92 kDa that was stimulated by LAM and LPSin supernatants from THP-1 cells. The 92 kDagelatinase (MMP-9) was barely detectable bytwo hours (data not shown) but was prominentat 24 hours (fig 1). A dose-response analysisrevealed that the gelatinase was discernible atLAM doses of 500 pg/ml and maximally stim-ulated at 500 ng/ml. Other matrix metallo-proteinases detected in the THP- 1 super-natants were a 230 kDa gelatinase which maybe a higher molecular weight aggregate ofMMP-9. The 72 kDa gelatinase (MMP-2) wasconstitutively expressed.

NORTHERN ANALYSIS OF MMP-9 AND MMP-2Incubation of THP-1 cells for 24 hours withLAM (2,g/ml) or LPS (1 ,ug/ml) revealed up-regulation of MMP-9 by LAM or LPS by 24hours in comparison with the unstimulatedcontrol (fig 2); there was no detectable up-regulation at two hours. Equal amounts of totalRNA were loaded in each lane as shown bythe pHe 7 housekeeping gene. These data areconsistent with the increased protein releaseand time course after LAM or LPS stimulationofTHP- 1 cells by substrate gel electrophoresis.MMP-2 was constitutively expressed at thegene level (fig 2), similar to the observationmade with the zymogram.

Incubation ofTHP-1 cells withM tuberculosisH37Ra (heat killed) revealed upregulation ofMMP-9 similar to that with LPS (fig 3).Northern analysis of bronchoalveolar lavage

cells from two patients with active cavitarypulmonary tuberculosis and one normal volun-teer demonstrated striking upregulation of thegene for MMP-9 in both tuberculosis patientscompared with the normal volunteer (fig 4).A small amount of constitutive expression ofMMP-9 could be detected in the normal volun-teer and equal amounts of total RNA wereloaded in each lane as demonstrated by the -actin housekeeping gene.

Figure 2 Northern analysis of 92 kDa and 72 kDa

matrix metalloproteinases (MMP-9 and MMP-2). THP'-1 cells were incubated with L-AM (2 ,ig/ml) or LPS (1 pg/

ml) for two hours (lanes 1-3) or 24 hours (lanes 4-6).

Lanes 1 and 4, unstimulated cells; lanes 2 and 5, LAM;

lanes 3 and 6, LP'S.

Figure 3 Northern analysis ofMMP-9 in THP-1 cells.THP-1 cells were incubated with M tuberculosis (MTB)H37Ra (1 ,uglml) or LPS (1 pglml) for 24 hours andtotal RNA extracted. Lane 1, unstimulated cells; lane 2,M tuberculosis H37Ra; lane 3, LPS.

NORTHERN ANALYSIS OF MMP-1Incubation of THP-1 cells with LAM inducedan increase in type I interstitial collagenase(MMP-1) mRNA compared with unstimulatedcells as early as two hours (fig 5A) and wassignificantly upregulated at 24 hours (fig 5B)and persisted up to 48 hours of incubation.

308



j.com/


arch 1996. Dow

nloaded from



1 2 3

Normal TB TB

92 kDaMMP-9 2*05 kb --

4 [3-actinControl LAM

Figure 4 Northern analysis of bronchoalveolar lavage(BAL) cells. Total RNA from BAL cells was run on anagarose-formaldehyde gel and filters were probed withMMP-9 cDNA. Lane 1, normal volunteer; lanes 2 and3, tuberculosis patients.

Equal amounts of RNA were loaded in eachlane as demonstrated by ethidium bromidestaining of the gel. The ability of LAM toinduce MMP-1 mRNA expression was similarto that with LPS. Incubation of peripheralblood monocytes with LAM induced an in-crease in MMP-1 mRNA at 24 hours (fig 6).Ethidium bromide stained gels demonstratedequal amounts ofRNA in each lane. Incubationof human lung fibroblasts with TNF-oa, IL-

B

Control LPS LAM LPS

Figure 5 Northern analysis of type I interstitial collagenase gene (MMP-1) for (Ahours and (B) 24 hours with LPS (1 iglml) and LAM (2 pglml). THP-1 cells wegrown to a density of 5 x 10' cellslml in 10% fetal cal serum and treated with stimzvarious time points. Total RNA was then isolated and an equal amount ofRNA fteach sample, as determined by ethidium bromide gel staining, was electrophoresed tha 1% denaturing agarose gel containing 7% formaldehyde. A 2-05 kb mRNA transfor MMP-1 was noted at two hours. Upregulation of the MMP-1 gene by LPS or Awas seen at two hours and persisted to 24 hours.

LAM

1) twoFreui for,omroughscriptLAM

Figure 6 Analysis of steady state MMP-1 mRNA levelsin peripheral blood monocytes. Human peripheral bloodmonocytes were stimulated for 24 hours with LAM (2 ,ig/ml) and equal amounts of total RNA (35,g) wereextracted as determined by ethidium bromide gel staining.

1 oa, or tetradecanoyl-phorbol- 1 3-acetate (TPA)induced an increase in MMP-1 mRNA (datanot shown).

Figure 7 illustrates a proposed scheme ofhow Mycobacterium spp could stimulate mono-nuclear phagocytes to upregulate the matrixmetalloproteinase genes in macrophages andrelease human proteinases. Macrophage releaseof TNF-ot and IL-1p can lead to fibroblastsynthesis of MMP-1 which amplifies the de-gradation of collagen types I-IV. MMP- 1inhibits ot,-antitrypsin leading to enhancedelastolytic activity with digestion of elastin andfibronectin resulting in tissue destruction, case-ation necrosis, and cavity formation.

DiscussionWe have shown that M tuberculosis H37Ra andLAM can induce the formation of the 92 kDagelatinase MMP-9 in THP-1 cells and the typeI interstitial collagenase MMP-1 mRNA inTHP-1 cells and human peripheral bloodmonocytes. In two patients with active cavitarytuberculosis we found striking upregulation ofthe MMP-9 gene in bronchoalveolar lavagecells. MMP-9 can be upregulated by peptidegrowth factors and cytokines in mononuclearphagocytes.6 17 The induction of MMP-1mRNA by LAM in mononuclear phagocytecells is also considered significant for the fol-lowing reasons: (a) unstimulated cells do notexpress the type I interstitial collagenase gene18;(b) several studies have shown that immaturemononuclear phagocytes (U-937 cells and peri-pheral blood monocytes) under basal con-ditions contain predominantly neutrophilelastase and maturation of these cells into res-ident macrophages by phorbol ester is as-sociated with an increased synthesis of matrixmetalloproteinases which include MMP- 1,MMP-2, and MMP-9, and the stromelysins

A

2.05 kb -*

309



j.com/


arch 1996. Dow

nloaded from


Chang, Wysocki, Tchou-Wong, Moskowitz, Zhang, Rom

LAM (cell wall component of Mycobacterium tuberculosis)

Mononuclear phagocyte

/\ IL-1[, TNF-a

92 kDa type IV collagenase Type collagenase NO Fibroblasts(MMP-9) (MMP-1)

Inhibited by tissue inhibitor of metalloproteinases

Degrades collagen types IVV Degrades collagen types 1, II, III

Inhibits a, antiproteinase

I Enhanceed elastolytic activity

72 kDa collagenase IV Degrades elastin, fibronectin, collagens IV, V(MMP-2)

Figure 7 Scheme ofproposed mechanisms of parenchynial destruction in tuberculousgranuloma. LAM= lipoarabinomannan.

MMP-3, MMP-10 and MMP- 11912i; (c)monocytes secrete much smaller quantities ofprocollagenase upon stimulation than domacrophages.22 All of the matrix metallo-proteinases cleave gelatin and fibronectin atthe same rate but MMP- 1 is unique in itsability to cleave interstitial collagens.'7 Thus,M tuberculosis and LAM are potent agents forthe induction of MMP-1 mRNAs and MMP-9.LAM, the major antigenic cell wall com-

ponent in Mycobacterium spp,7-9 can induceMMP-1 mRNA indirectly in human lungfibroblasts through the release of TNF-o andIL-1: from monocyte macrophages. The re-

leased collagenases may participate in the de-struction of lung extracellular matrix by thefollowing mechanism. MMP-9 can degradeseveral native collagens including types IV andV. LAM stimulates mononuclear phagocytes torelease IL-i D3 and TNF-ct which stimulate lungfibroblasts to release MMP-1 as well as MMP-2 and, interestingly, fibroblasts are capable ofrelasing greater quantities of these twoMMPs. 16'7 MMP-1 degrades collagens I, II,III, and X and MMP-9 and MMP-2 cleavetype IV basement membrane collagen.23 Thisinitial collagenolytic cleavage of insoluble pro-tein is essential for the subsequent endocytosisand completion of digestion within lyosomes,followed by tissue remodelling.4"

In a normal physiological state metallo-proteinases are tightly regulated. All of themetalloenzymes are inhibited by specific tissueinhibitors of metalloproteinases (TIMP) whichare ubiquitous natural inhibitors and whichform complexes with these metalloenzymes. 24

In pulmonary tuberculosis, however, the regu-latory mechanism of collagenases is likely to bealtered in favour of collagenase activity. TNF-ci and IL-i 1B stimulate MMP gene expressioni7but do not affect the level of TIMP.25 This isin contrast to the effect ofTPA which stimulatesboth collagenase and TIMP in skin fibro-blasts.26 In addition to collagenolytic activity,MMP-1 has been shown to express serpinase

activity- that is, to hydrolyse and inactivate thetwo major serine proteinase inhibitors found inplasma, x,-proteinase inhibitor (ocIPI) and cc,-antichymotrypsin.27 Since ociPI is the majorinhibitor of elastase, in the microenvironmentwhere human collagenases are present elastaseactivity would be relatively uninhibited,resulting in further destruction of lungparenchyma.

All of the MMPs are expressed and secretedby human mononuclear phagocytes in responseto stimuli. 117202228 There are 11 members ofthe MMP family known as endopeptidases in-cluding interstitial collagenases, stromelysins,and gelatinases.'7 Both MMP-1 and MMP-9contain AP-1 binding sites in their 5' DNApromoters that respond to phorbol esters andthe cytokines TNF-ot and IL-i1 through ac-tivation of the jun gene.'729 MMP-9 is readilysecreted by monocytes.'5 It has been referredto as type IV collagenase and is capable ofdegrading not only basement membranecollagens (type IV collagen) but also elastin.28Thus, in the presence of uninhibited matrixmetalloproteinases, the destructionof lung mat-rix is the sum of collagenolytic and elastolyticactivities. We propose that M tuberculosis and itsmajor cell wall antigen LAM play an importantpart in the pathogenesis of lung damage as-sociated with pulmonary tuberculosis throughthe stimulation and release of collagenases bymononuclear phagocytes and, indirectly, byfibroblasts.

The authors acknowledge the Suntory Institute of BiomedicalResearch, Osaka, Japan for recombinant TNF-a, Dr P Brennan,Colorado State University, Fort Collins, Colorado, USA forLAM, Natalie Little for editorial assistance, and grants GCRCMO I RR00096, NIH Al 32233, HL-51494, NR 03212, HeiserFoundation, Stony Wold Herbert Foundation, and Aaron Dia-mond Foundation for support.

1 Chang JC, Jagirdar J, Lesser M. Long-term evolution ofBCG- and CFA-induced granulomas in rat lungs. Ani]Pathol 1986;125:16-27.

2 Dannenberg AM Jr. Cellular hypersensitivity and cellularimmunity in the pathogenesis of tuberculosis: specificity,systemic and local nature, and associated macrophageenzymes. Bacteriol Rev 1968;32:85-102.

3 Dannenberg AM Jr. Pathogenesis of pulmonary tuber-culosis. Am Rev Respir Dis 1982:125:25-30.

4 Werb Z, Banda MJ, Jones PA. Degradation of connectivetissue matrices by macrophages. 7 Exp Med 1980;152:1340-57.

5 Reynolds JJ. The molecular and cellular interactions in-volved in connective tissue destruction. Br _7 Dervnatol1985;112:715-23.

6 Hunter SW, Brennan PJ. Evidence for the presence of aphosphatidylinositol anchor on the lipoarabinomannanand lipomannan of Mvcobactenriun tuberculosis. _7 Biol Chen1990;265:9272-9.

7 Chatterjee D, Roberts AD, Lowell K, Brennan PJ, OrmeIM. Structural basis of capacity of lipoarabinomannan toinduce secretion of tumor necrosis factor. Infect Inimun1992;60:1249-53.

8 Zhang Y, Doerfler M, Lee TC, Guillemin B, Rom WN.Mechanisms of stimulation of interleukin-lp and tumornecrosis factor-a by Mycobacterium tuberculosis and its com-ponents. _7 Clin Invest 1993;91:2076-83.

9 Barnes PF, Chatterjee D, Abrams JS, Lu S, Wang E,Yamamura M, et al. Cytokine production induced byMycobacterium tuberculosis lipoarabinomannan. Relation-ship to chemical structure. 7 Immunol 1992;149:541-7.

10 Wallis RS, Amir-Tahmasseb M, Ellner JJ. Induction ofinterleukin 1 and tumor necrosis factor by mycobacterialproteins: the monocyte Western blot. Proc Mad Acad SciUSA 1990;87:3348-52.

11 Takashima T, Ueta C, Tsuyuguchi I, Kishimoto S. Pro-duction of tumor necrosis factor alpha by monocytes frompatients with pulmonary tuberculosis. Infect In"niun 1990;58:3286-92.

12 Moreno C, Taveme J, Mehlert A, Bate CAW, Brealey RJ,Meager A, et al. Lipoarabinomannan from M tuberculosisinduces the production of tumor necrosis factor fromhuman and murine macrophages. Clin Exp Immunol 1989;76:240-5.

310



j.com/


arch 1996. Dow

nloaded from



13 Chensue SW, Davey MP, Resnick DG, Kunkel SL. Releaseof interleukin-l by peripheral blood mononuclear cells inpatients with tuberculosis and active inflammation. InfectImmun 1986;52:341-3.

14 Dayer J-M, Beutler B, Cerami A. Cachectin/tumor necrosisfactor stimulates collagenase and prostaglandin E2 pro-duction by human synovial cells and dermal fibroblasts.J Exp Med 1985;162:2163-8.

15 Dayer J-M, de Rochemonteix B, Demczuk S, Dinarello CA.Human recombinant interleukin 1 stimulates collagenaseand prostaglandin E2 production by human synovial cells.J Clin Invest 1986;77:645-8.

16 Welgus HG, Campbell EJ, Cury JD, Eisen AZ, Senior RM,Wilhelm SM, et al. Neutral metalloproteinases producedby human mononuclear phagocytes: enzyme profile, re-gulation, and expression during cellular development. JClin Invest 1990;86:1496-502.

17 Birkedal-Hansen H, Moore WGI, Bodden MK, WindsorLJ, Birkedal-Hansen B, DeCarlo A, et al. Matrix metallo-proteinases: a review. Crit Rev Oral Biol Med 1993;4:197-250.

18 Bar-Shavit Z, Teitelbaum SL, Stricklin GP, Eisen AZ, KahnAJ, Welgus HG. Differentiation of a human leukemia cellline and expression of collagenase inhibitor. Proc NatlAcadSci USA 1985;82:5380-4.

19 Senior RM, Campbell EJ, Landis JA, Cox FR, Kuhn C,Koren HS. Elastase of U-937 monocytelike cells: com-parisons with elastases derived from human monocytesand neutrophils and murine macrophagelike cells. J ClinInvest 1982;69:384-93.

20 Welgus HG, Connolly NL, Senior RM. 12-o-Tetra-decanoyl-phorbol-13-acetate-differentiated U9327 cellsexpress a macrophage-like profile of neutral proteinases:

high levels of secreted collagenase and collagenase in-hibitor accompany low levels of intracellular elastase andcathepsin G. J Clin Invest 1986;77:1675-81.

21 Hersh CL, Yeh RK, Callaway JE, Garcia JA Jr, Gilmore-Hebers M. Induction of collagenase production in U937cells by phorbol ester and partial purification of the in-duced enzyme. Biochemistry 1986;25:2750-7.

22 Campbell EJ, Cury JD, Lazarus CJ, Welgus HG. Monocyteprocollagenase and tissue inhibitor of metalloproteinases.J Biol Chem 1987;262:15862-8.

23 Welgus HG, Fliszar CJ, Seltzer JL, Schmid TM, Jeffrey JJ.Differential susceptibility of type X collagen to cleavageby two mammalian interstitial collagenases and 72-kDatype IV collagenase. J Biol Chem 1990;265:13521-7.

24 Matrisian LM. Metalloproteinases and their inhibitors inmatrix remodeling. Trends Genet 1990;6:121-5.

25 MacNaul KL, Chartrain N, Lark M, Tocci MJ, HutchinsonNI. Discoordinate expression of stromelysin, collagenase,and tissue inhibitor of metalloproteinases-l in rheumatoidhuman synovial fibroblasts. J Biol Chem 1990;265:17238-45.

26 Clark SD, Wilhelm SM, Stricklin GP, Welgus HG. Co-regulation of collagenase and collagenase inhibitor pro-duction by phorbol myristate acetate in human skinfibroblasts. Arch Biochem Biophys 1985;241:36-44.

27 Desrochers PE, Jeffrey JJ, Weiss SJ. Interstitial collagenase(matrix metalloproteinase-l) expresses serpinase activity.J Clin Invest 1991;87:2258-65.

28 Senior RM, Griffin GL, Fliszar CJ, Shapiro SD, GoldbergGI, Welgus HG. Human 92- and 72-kilodalton type IVcollagenase are elastases. J Biol Chem 1991;266:7870-5,

29 Brenner DA, O'Hara M, Angel P, Chojkier M, Karin M.Prolonged activation of jun and collagenase genes bytumour necrosis factor-oa. Nature 1989;337:661-3.

311



j.com/


arch 1996. Dow

nloaded from


thorax effect mycobacterium matrix · thorax 1996;51:306-311 effect ofmycobacterium tuberculosis...

Documents