Differential expression of matrix ... - Semantic Scholar

2 downloads 0 Views 277KB Size Report
disease has not been determined. We investigated the temporal and spatial mRNA expression pattern of gelatinase B in experimental meningococcal meningitis ...
Brain (1999), 122, 1579–1587

Differential expression of matrix metalloproteinases in bacterial meningitis Bernd C. Kieseier,1 Robert Paul,2 Uwe Koedel,2 Thomas Seifert,1 John M. Clements,3 Andrew J. H. Gearing,3 Hans-Walter Pfister2 and Hans-Peter Hartung1 1Department

of Neurology, Karl-Franzens-Universita¨t, Graz, Austria, 2Department of Neurology, LudwigMaximilians-Universita¨t, Mu¨nchen, Germany and 3British Biotech Pharmaceuticals Limited, Oxford, UK

Correspondence to: Dr Bernd C. Kieseier, Department of Neurology, Karl-Franzens-Universita¨t, Auenbruggerplatz 22, 8036 Graz, Austria E-mail: [email protected]

Summary Matrix metalloproteinases (MMPs) are implicated in the pathogenesis of various inflammatory diseases of the central nervous system. Evidence is accumulating that gelatinase B (MMP-9) might be involved in the pathogenesis of meningitis, but the spectrum of different MMPs involved in the inflammatory reaction of this disease has not been determined. We investigated the temporal and spatial mRNA expression pattern of gelatinase B in experimental meningococcal meningitis in rats. In contrast to controls, increased mRNA levels with peak values 6 h after injection with menigococci were found in brain specimens of the animals. Elevated MMP9 mRNA expression was accompanied by enhanced proteolytic activity, as demonstrated by gelatin zymography, and positive immunoreactivity. The mRNA expression pattern of six other MMPs was investigated.

Collagenase-3 and stromelysin-1 mRNAs were also found to be upregulated. In contrast, mRNA levels for gelatinase A, matrilysin, stromelysin-2 and stromelysin-3 remained unchanged. As evidenced by significantly increased intracranial pressure and by leakage of intravenously injected Evans blue through the blood vessel walls into the brain parenchyma, the animals injected with meningococci revealed signs of blood–brain barrier disruption. Augmented proteolytic activity of MMP-9 could also be demonstrated in CSF samples obtained from patients with bacterial meningitis, underlining the clinical relevance of our experimental findings. Our data indicate that gelatinase B, collagenase-3 and stromelysin1 are selectively upregulated in bacterial meningitis and thus may contribute to the pathogenesis of this infectious disease of the central nervous system.

Keywords: gelatinase B; collagenase-3; stromelysin-1; bacterial meningitis; experimental meningitis Abbreviations: MMP 5 matrix metalloproteinase; PBS 5 phosphate-buffered saline; PCR 5 polymerase chain reaction; TNF-α 5 tumour necrosis factor-α; WBC 5 white blood cell

Introduction The matrix metalloproteinases (MMPs) belong to a large family of zinc-dependent proteinases which includes the collagenases, gelatinases, stromelysins, matrilysin, membrane-type metalloproteinases and metalloelastase (Woessner, 1994; Birkedal-Hansen, 1995; Yong et al., 1998). Various MMPs have been identified to date which have structural domains in common but differ in cellular sources, substrate specificity and inducibility (Goetzl et al., 1996). However, knowledge about their involvement in physiology as well as pathological conditions remains incomplete. An emerging body of evidence points to MMPs as important factors in the pathogenesis of meningitis. Increased levels of gelatinase B (MMP-9) were detectable in CSF samples from patients suffering from viral (Kolb et al., 1998) or bacterial (Gijbels et al., 1992) meningitis. Furthermore, application of © Oxford University Press 1999

the synthetic broad-spectrum MMP inhibitor batimastat (BB94) in a rat model of meningococcal meningitis significantly reduced blood–brain barrier disruption and the diseasecharacteristic increase in intracranial pressure (Paul et al., 1998). Breakdown of the blood–brain barrier is a typical histopathological feature in meningitis, giving rise to the development of vasogenic oedema (Quagliarello and Scheld, 1992). The notion that MMPs play an important role in blood–brain barrier disruption is supported by studies in experimental autoimmune encephalomyelitis. In this animal model of multiple sclerosis, non-specific inhibition of MMPs was shown to protect the blood–brain barrier (Gijbels et al., 1994; Hewson et al., 1995). Among all known MMPs, gelatinase B might be particularly involved in blood–brain barrier disturbance, as raised CSF levels of MMP-9 in MS

1580

B. C. Kieseier et al.

patients are associated with a leaky blood–brain barrier on magnetic resonance imaging (Rosenberg et al., 1996). However, while there is some evidence suggestive of increased gelatinase activity in meningitis, the mRNA expression pattern of gelatinase B and other MMPs in meningitis has not been explored so far. The present study focuses on this important issue in a rat model of meningococcal meningitis, since such information might help in the design of novel MMP-targeted therapies for a disorder that, despite improved antimicrobial therapy, is still associated with high mortality (Quagliarello and Scheld, 1997).

Material and methods Animals Male Wistar rats (Charles Rivers, Sulzfeld, Germany) weighing 250–320 g were used. The group challenged with meningococci consisted of eight animals, while four served as controls. In addition, two animals were used to assess the integrity of the blood–brain barrier. All experiments were conducted according to local state regulations for animal experimentation and were approved by the responsible authorities.

Induction of meningococcal meningitis Bacterial meningitis was induced as described previously (Pfister et al., 1990). Briefly, rats were anaesthetized with thiopental (Trapanal, Byk Gulden, Germany; 100 mg/kg, intraperitoneally), tracheotomized and artificially ventilated. The left femoral artery was cannulated for continuous monitoring of the mean arterial blood pressure and for hourly blood gas analysis. The left femoral vein was cannulated for fluid substitution. Through a burr hole at the occipital bone, a catheter was inserted into the cisterna magna for continuous intracranial pressure monitoring, injection of meningococci and the determination of white blood cell (WBC) counts in the CSF. During the experiment all animals were kept at a constant temperature of 37.5 6 0.5°C using a rectal thermometer-controlled heating pad. Three, six and nine hours after induction of the disease, given numbers of animals were killed, and each brain was flash-frozen in liquid nitrogen and stored at –70°C.

Intracisternal inoculum Heat-killed (60°C for 4 h) Neisseria meningitidis (serogroup B, strain B1940) were used for intracisternal application. Before use, the bacteria were subcultured on blood-agar plates and resuspended in phosphate-buffered saline (PBS) (Pfister et al., 1990). Rats were challenged intracisternally with 100 µl of 109 colony-forming units/ml of heat-killed meningococci. Control animals were treated with the same volume of PBS.

Competitive PCR For quantitation of rat MMP mRNA levels, a polymerase chain reaction (PCR) using a multi-competitor DNA standard was performed (Wells et al., 1996). The synthetic standard DNA contained tandem arrays of 59 and 39 priming sites for cDNAs of different matrix metalloproteinases and β-actin. Primer pairs for collagenase-3, matrilysin, gelatinases A and B, stromelysin-1, -2 and -3 and β-actin were used according to the sequences published elsewhere (Wells et al., 1996). Poly(A1) RNA was extracted from frozen brain specimens and used as a template for cDNA synthesis using AMV (avian myeloblastosis virus) reverse transcriptase (Promega, Madison, Wis., USA). Threefold serial dilutions of competitive standard DNA were combined with a fixed amount of sample cDNA, and PCR was performed in 50-µl reactions containing 1.5 mM MgCl2, 50 mM KCl, 10 mM Tris–HCl (pH 9.0) and 0.1% Triton X-100 in the presence of 200 µM dNTP (deoxynucleotide triphosphate; Pharmacia, Freiburg, Germany), 50 pmol of sense and antisense MMP primers, 1 U Taq DNA polymerase (Perkin Elmer, Branchburg, NJ, USA) and 1 µCi [α-32P]dCTP (Amersham, Braunschweig, Germany). Amplification was carried out using 35 cycles (95°C, 30 s/57°C, 30 s/72°C, 120 s) in a Hybaid Omnigene thermal cycler (MWG Biotech, Ebersberg, Germany). Ten microlitres of the reaction products was electrophoresed on a 6% polyacrylamide gel. Gels were exposed to appropriate screens, and the uptake of [α-32P]dCTP within each individual PCR product was determined using a phosphor imaging system (Storm 360; Molecular Dynamics, Krefeld, Germany). Levels of MMP mRNA were determined by plotting the ratio of sample cDNA to standard DNA against the standard dilution using a double logarithmic scale.

Gelatin zymography in the animal model Gelatinase activity was determined by SDS–PAGE (sodium dodecyl sulphate–polyacrylamide gel electrophoresis) zymography (Brown et al., 1990). In brief, 10 3 10 µm frozen sections of brain were incubated with 25 µl of Tris/ glycine SDS sample buffer (Novex, San Diego, Calif., USA), degraded by ultrasound and further homogenized by vortexing for 1 h at 4°C. The samples were applied to a 10% (w/v) polyacrylamide resolving gel containing 0.1% SDS and 0.1% gelatin type A from porcine skin (Sigma, St Louis, Mo., USA). Stacking gels were 3% (w/v) polyacrylamide. After electrophoresis, gels were washed in renaturing buffer (Novex) containing Triton X-100 to remove any SDS and incubated in developing buffer (Novex) for 18 h at 37°C. Gels were stained for 6 h in 30% methanol/10% acetic acid containing 0.5% (w/v) Coomassie brilliant blue G-250 and destained in the same buffer without dye. Gelatinase activity was detected as unstained bands on a blue background representing areas of gelatin digestion. Supernatants from astrocyte cell cultures stimulated with tumour necrosis factor-α (TNF-α) served as positive controls.

Matrix metalloproteinases in meningitis

1581

Table 1 Clinical characteristics of patients with bacterial meningitis and non-inflammatory neurological disorders (controls) Patient group

CSF WBC count (cells/µl)

CSF total protein (g/l)

Bacterial meningitis: 1st LP Bacterial meningitis: 2nd LP Control

2465 (324–8010) 315 (132–827) 2 (0–4)

1.87 (1.23–3.37) 1.09 (0.6–2.14) 0.38 (0.22–0.49)

Values are given as medians and ranges. WBC 5 white blood cells; LP 5 lumbar puncture.

Immunohistochemistry

Frozen sections (10 µm) of rat brain were air-dried and fixed for 10 min in 4% formalin and 2 min in each of 50, 100 and 50% acetone at room temperature. A monoclonal affinitypurified antibody against gelatinase B (R&D Systems, Abingdon, UK) was used at 1.2 µg/ml overnight at 4°C. Thereafter a biotinylated secondary antibody against mouse IgG and an avidin-biotinylated peroxidase complex (Dako, Hamburg, Germany) were used with 3,39-diaminobenzidine as peroxidase substrate. Endogenous peroxidase activity was suppressed by incubating the sections with 3% H2O2 in methanol for 10 min before adding the secondary antibody. Sections were counterstained with haematoxylin, dehydrated, and mounted in Eukitt (Kindler, Freiburg, Germany).

Assessment of blood–brain barrier integrity To verify disruption of the blood–brain barrier in diseased animals, an additional number of control and meningococcusinjected animals were investigated. Meningococcal meningitis was induced as described above, but 1 h prior to termination of the experiment, 1 ml of 1% (w/v) Evans blue (Sigma Chemicals, Deisenhofen, Germany) was injected intravenously. Six hours after challenge with meningococci, the animals were perfused with 100 ml of ice-cold PBS and the brain was removed and stored at –70°C for further histological evaluation. Frozen sections (10 µm) of the brains were fixed for 20 min with 100% ethanol. Evans blue was observed under green fluorescence microscopy (excitation filter 545 nm, barrier filter 590 nm) and appeared red.

Patients and CSF sampling Fourteen CSF samples were obtained from 10 patients that suffered from bacterial meningitis (Table 1). Follow-up investigations were performed on four patients. Samples from five patients with non-inflammatory neurological diseases served as controls. After lumbar puncture, CSF samples were immediately centrifuged at 3000 g for 10 min, and supernatants were aliquoted and stored at –70°C. All patients gave informed consent and the study was approved by the local ethics committee.

Gelatin substrate zymography in human CSF samples Human gelatinase activity was measured as described above using 10 µl of undiluted CSF.

Statistical analysis One-way analysis of variance and Student–Newman–Keuls multiple comparisons were used as the principal statistical tests. A P value , 0.05 was considered significant. Data are given as mean 6 standard error.

Results Induction of meningococcal meningitis Intracisternal meningococcal injection in the animals did not result in changes in p(O2), p(CO2), pH and mean arterial blood pressure compared with the control group (all results were within normal ranges; data not shown). However, a significant increase in WBC counts in CSF (4229 6 2959 cells/µl 3 h, 8250 6 2804 cells/µl 6 h and 7728 6 5634 cells/µl 9 h after intracisternal injection, compared with 30 6 11 cells/µl in controls; P , 0.05) and in intracranial pressure (15.8 6 3.1 mmHg 6 h after intracisternal injection of meningococci compared with 3.2 6 0.9 mmHg in the control group; P , 0.05) occurred in animals injected with meningococci.

Gelatinase B Quantitation of mRNA A constitutive low level of expression of gelatinase B mRNA was found in the PBS-injected animals that served as controls. In contrast, elevated levels were already detected in animals 3 h after injection with meningococci, with peak mRNA levels 6 h after intracisternal challenge. Thereafter mRNA expression declined, but remained upregulated in comparison with controls (Fig. 1).

Proteolytic activity in the animal model To demonstrate that increased expression of gelatinase B mRNA was associated with enhanced enzymatic activity, zymography of brain tissue taken from animals injected with meningococci was performed. A pronounced rise in gelatinase B activity was observed in meningococcus-injected animals compared with control animals, in line with the mRNA data obtained by competitive PCR. Thus, higher expression of mRNA was accompanied by augmented gelatin digestion. In contrast, no increase in the activity of gelatinase A, which is detected in the same experiment, could be found using this assay (Fig. 2).

1582

B. C. Kieseier et al.

Fig. 2 Detection of proteinase activity in brain specimens from Wistar rats by SDS substrate gel analysis. Zone of clearing represents proteinase activity. Tissues from animals obtained 3, 6 and 9 h after intracisternal injection of heat-killed meningococci reveal increased areas of gelatin digestion at the level of 92 kDa in comparison with brain specimens obtained from animals injected with PBS (co). Supernatant from stimulated astrocyte cell culture served as positive control (1). Molecular mass is indicated by arrows at 72 kDa for gelatinase A and 92 kDa for gelatinase B.

repeated puncture was carried out, a pronounced decrease in proteolytic activity was detected at the second time point compared with the first. In contrast, the activity of gelatinase A appeared to be unchanged between meningitis and control cases, the size of the unstained bands remaining constant (Fig. 3).

Immunohistochemistry In control animals no immunoreactivity for gelatinase B could be detected. In contrast, positive signals were noted in diseased animals, primarily around the blood vessels and localized to cells with polymorphic or round nuclei, most likely representing invading neutrophils and mononuclear cells (Fig. 4). Control sections after omission of the primary antibody showed only background staining.

Assessment of blood–brain barrier integrity

Fig. 1 MMP mRNA expression in brain specimens of Wistar rats injected with heat-killed meningococci (HKM) and PBS (control). Increased mRNA expression of (A) gelatinase B can be detected 3 h after intracisternal injection of meningococci, with maximum levels after 6 h. The mRNAs for (B) stromelysin-1 and (C) collagenase-3 were upregulated at all time points studied. Results are given as mean 6 SD. Values for 3 and 9 h represent data from two animals each, whereas data for 6 h after intracisternal injection and controls (6 h after injection of PBS) represent data from four animals.

Proteolytic activity in human CSF samples In CSF samples taken from patients with bacterial meningitis, augmented activity of gelatinase B was recorded in comparison with non-inflammatory controls. In cases where

Weak Evans blue fluorescence was demonstrable in the subarachnoid space of control rats. No extravasation of the dye could be observed in the underlying brain parenchyma. In contrast, enhanced staining within the leptomeninges and a diffuse extravasation of Evans blue dye in the underlying brain parenchyma, indicating disruption of the blood–brain barrier, was detected in meningococcus-injected rats 6 h after challenge (Fig. 5).

Other matrix metalloproteinases To investigate whether MMPs other than MMP-9 are regulated in bacterial meningitis, we studied the mRNA expression of collagenase-3, stromelysin-1, -2 and -3, gelatinase A and matrilysin. The last four MMPs were constitutively expressed at low levels, and did not exhibit any difference in mRNA levels between the experimental and the control group. The constitutive low mRNA expression pattern of gelatinase A underlines the finding that no change in proteinase activity measured by gelatin zymography could be detected for this MMP within the brain specimens.

Matrix metalloproteinases in meningitis

1583

Fig. 3 Detection of proteolytic activity in CSF samples from patients with bacterial meningitis (lanes 1–10) and controls (lanes 11–15). Follow-up investigations were performed on the first four patients (a 5 first lumbar puncture; b 5 second lumbar puncture). Molecular mass is indicated by arrows at 72 kDa for gelatinase A and 92 kDa for gelatinase B.

In contrast, mRNAs for collagenase-3 and stromelysin-1 were upregulated 3, 6 and 9 h after injection of meningococci (Fig. 1).

Discussion In the present study the expression pattern of gelatinase B and six other MMPs in experimental meningitis was investigated. In meningococcus-injected animals, upregulation of MMP-9 mRNA was found, associated with increased enzymatic activity and protein immunoreactivity. On immunohistochemistry, positive signals for MMP-9 could be localized along the meninges and to cellular structures suggestive of infiltrating polymorphonuclear and mononuclear cells. Given that monocytes (Welgus et al., 1990), T cells (Leppert et al., 1995), neutrophils (Masure et al., 1991) and macrophages (Nielsen et al., 1996) are known to be capable of producing MMP-9, we assume that these cells, and especially neutrophils, which are important effector cells in mediating the pathogenesis of bacterial meningitis, have to be considered as the primary sources of MMP-9 in the present study. Degranulation of neutrophil gelatinase B appears to be a fast process, since increased gelatinase B mRNA as well as proteolytic activity could already be detected 3 h after intracisternal challenge with meningococci. Since heat-killed meningococci were used in the present experiment, the decrease in MMP expression noticeable 9 h after intracisternal injection reflects a decrease in immunoreactivity when no bacteria are present to perpetuate the infection any further. It should be noted also that astrocytes (Wells et al., 1996), microglia (Gottschall and Yu, 1995) and endothelial cells (Herron et al., 1986) have been reported to synthesize gelatinase B after stimulation with various cytokines, such as interleukin (IL)-1 and TNF-α in vitro; however, there is no histological evidence that these cells were sources of this metalloproteinase in the current study. Gelatin zymography revealed that the upregulation of the message accompanied enhanced proteolytic activity of MMP9, underscoring the pathogenic relevance of this finding. In the CSF of animals with bacterial meningitis, increased proteolytic activity of MMP-9 and MMP-2 has been observed (Paul et al., 1998).

Upregulation of MMP-9 mRNA, augmented proteolytic activity and positive immunoreactivity were accompanied by both a significant increase in CSF WBC counts and raised intracranial pressure. These findings point to a potential role of MMP-9 in the pathogenesis of meningitis. Breakdown of the blood–brain barrier is a pathological hallmark in bacterial meningitis that (i) leads to vasogenic brain oedema with a consecutive increase in the intracranial pressure, and (ii) facilitates leucocyte recruitment across the endothelium into the CSF (Gray and Nordmann, 1997). Studies in experimental meningitis support the concept that endogenous inflammatory mediators play a crucial role in promoting impairment of the blood–brain barrier. Waage and coworkers demonstrated release of TNF-α, IL-1 and IL-6 into the CSF after intracisternal inoculation of meningococcal lipopolysaccharide into rabbits (Waage et al., 1989). Intracisternal challenge of rats by IL-1 precipitated blood–brain barrier disturbance in a dose-dependent manner, an effect which was synergistically enhanced by the additional administration of TNF-α (Quagliarello et al., 1991). The mediators thought to be involved in cytokine modulation of capillary permeability include MMPs, and it has been suggested that these proteinases play a central role in the opening of the blood– brain barrier. Intracerebral injection or induction of gelatinases A and B resulted in the breakdown of the extracellular matrix and the opening of the blood–brain barrier in rats (Rosenberg et al., 1992, 1994). Our data, showing upregulated gelatinase B mRNA and augmented proteolytic activity in association with increased intracranial pressure and Evans blue extravasation in experimental meningitis, support the concept that MMP-9 is a key mediator of blood–brain barrier breakdown. Knowledge about the mechanisms involved in the passage of leucocytes across the endothelium, leading to increased WBC counts in the CSF, is still incomplete. It is commonly accepted that the process of homing and transmigration of circulating leucocytes through the blood–brain barrier encompasses the complex interaction of a variety of adhesion molecules and different cytokines (Hartung, 1995; Archelos and Hartung, 1997). Some of these adhesion molecules and cytokines are also known to regulate MMP expression (Romanic and Madri, 1994). Various studies point to the involvement of gelatinase B in the migration of leucocytes

1584

B. C. Kieseier et al.

Fig. 4 Immunohistochemistry for gelatinase B in frozen sections of brain specimens from Wistar rats. Positive signals occur primarily along the meninges 3 h after intracisternal injection of meningococci (A), with an increase in density of immunoreactivity 6 h after injection (B), followed by decreasing positive signal intensity 9 h after intracisternal injection of meningococci (C). Control animals do not reveal any positive immunoreactivity (D). Higher magnification of the meninges in animals injected with meningococci clearly links positive immunoreactivity to infiltrating polymorphonuclear and mononuclear cells, marked with arrows; arrowheads point to cells without immunoreactivity (E). Magnification: A and C, 3200; B and D, 3400; E, 3600.

through artificial basement membrane systems in vitro (Huber and Weiss, 1989; Leppert et al., 1995). Since migration of inflammatory cells from the blood to the CNS is believed to be of paramount importance in the genesis of an inflammatory reaction along the meninges, such an action would assign MMPs a strategic role in this process. The upregulation of gelatinase B associated with a significantly increased WBC count in the CSF of diseased animals, as demonstrated in the present study, supports this notion. Finally, MMP-9 might be involved in the cleavage or regulation of proteins including type IV collagen of the

basement membrane, myelin proteins, and cytokines (Opdenakker and Van Damme, 1994; Ries and Petrides, 1995). Release of the proinflammatory cytokine TNF-α appears to depend on an MMP-like mechanism of action (Gearing et al., 1994; Mohler et al., 1994), emphasizing the complexity of interaction between cytokines and MMPs in neuroinflammation. However, the recently described TNF-αconverting enzyme (TACE) seems more likely to be a novel metalloproteinase, distinct from the classical MMPs (Black et al., 1997; Moss et al., 1997). Thus, it is difficult at present to determine the contributory effect of known MMPs to the

Matrix metalloproteinases in meningitis

Fig. 5 Assessment of blood–brain barrier integrity. Only weak Evans blue fluorescence is observed in the arachnoid space of control animals (A), whereas in meningococcus-injected rats there is enhanced staining within the leptomeninges and diffuse extravasation of Evans blue dye in the underlying brain parenchyma, indicating disruption of the blood–brain barrier (B). Magnification: A and B, 3200.

release of surface cytokines and receptors, which play an important role in the modulation of the inflammatory response. Further studies are needed to shed light on the involvement of MMPs in this process. To demonstrate that our finding of upregulated gelatinase B is not restricted to the experimental model but also has clinical relevance, we performed gelatinase zymography in CSF samples from patients with bacterial meningitis. The markedly increased proteolytic activity of MMP-9 detectable in the acutely ill patients, with an obvious decrease following treatment, and our experimental data drawn from the rat model, reveal, at least in part, similarities between this model and the human disease. This finding is consistent with previous observations in patients suffering from viral (Kolb et al., 1998) or bacterial (Gijbels et al., 1992) meningitis. To investigate whether MMPs other than gelatinase B might also be upregulated in experimental meningitis, mRNA expression patterns of six other metalloproteinases were investigated. The mRNAs of collagenase-3 and stromelysin1 were found to be upregulated in meningococcus-injected animals, to a lesser extent than the mRNA of gelatinase B, revealing constitutive upregulation at all time points investigated. In contrast, mRNAs for matrilysin and stromelysin-2 and -3 were expressed at low levels but were not regulated. Similarly, gelatinase A, thought to be primarily involved in blood–brain barrier breakdown (Rosenberg et al., 1992), was not upregulated in the present study. Thus, MMPs in the CNS do not respond during experimental meningitis in an all-or-none fashion but rather are differentially regulated. Interestingly, the response of MMPs differs among different diseases of the inflamed CNS, since the MMP pattern found

1585

in the present study deviates from the pattern of MMP mRNA expression described in experimental autoimmune encephalomyelitis (Kieseier et al., 1998), an inflammatory demyelinating disease associated with blood–brain barrier breakdown. The role of selectively upregulated collagenase-3 and stromelysin-1 in the inflammatory reaction remains conjectural. Both MMPs could be implicated in the disruption of the blood–brain barrier, since collagenase-3 is known to degrade various collagens and gelatins such as stromelysin1, which has been reported to degrade fibronectin, laminin, elastin, gelatin and different collagens (Cawston, 1998). In addition, increased mRNA expression of collagenase-3 and stromelysin-1, as seen in our study, may be involved in controlling the inflammatory reaction, as both metalloproteinases are capable of activating progelatinase B, the latent proform of MMP-9, by sequential cleavage of the propeptide (Okada et al., 1992; Knauper et al., 1997). Furthermore, it has been demonstrated that the proinflammatory cytokines IL-1 and IL-6, which have been implicated as playing a disease-promoting role in bacterial meningitis, can induce the expression of stromelysin-1, collagenase-3 and gelatinase B (Kusano et al., 1998). Along with the study by Paul and colleagues (Paul et al., 1998), demonstrating that MMP inhibition effectively reduces blood–brain barrier damage, the present study provides strong circumstantial evidence that MMPs play a critical role in experimental meningitis. Disruption of the blood–brain barrier is a crucial step in the pathogenesis of this disease, and experimental studies provide evidence that various mediators, such as nitric oxide (Koedel et al., 1995; Paul et al., 1997), reactive oxygen species (Pfister et al., 1992; Koedel and Pfister, 1997) and prostaglandin E2 (Pfister et al., 1990), are involved in this process. The present findings emphasize the role of MMPs in the disruption of the blood–brain barrier. Further research is needed to elucidate the precise mechanisms of action and the interplay between MMPs and other factors involved in damage to the blood–brain barrier. It is hoped that increased knowledge of MMP-mediated actions in the pathogenesis of meningitis will aid in the design of inhibitors targeted to specific MMPs that could be administered at critical points in the evolution of this inflammatory disease, and that such compounds may enlarge our therapeutic arsenal to improve the clinical outcome, frequently still dubious, in bacterial meningitis.

Acknowledgements We thank Professor M. Frosch, University of Wu¨rzburg, for providing Neisseria meningitidis and H. Pischel for technical assistance. This study was supported by the Gemeinnu¨tzige Hertie Stiftung (to H.P.H.) and by the Forschungsgemeinschaft (Pf 246/5-2) (to H.W.P.). References Archelos JJ, Hartung HP. The role of adhesion molecules in multiple sclerosis: biology, pathogenesis and therapeutic implications. [Review]. Mol Med Today 1997; 3: 310–21.

1586

B. C. Kieseier et al.

Birkedal-Hansen H. Proteolytic remodeling of extracellular matrix. [Review]. Curr Opin Cell Biol 1995; 7: 728–35.

acetyl-L-cysteine in pneumococcal meningitis in the rat. Neurosci Lett 1997; 225: 33–6.

Black RA, Rauch CT, Kozlosky CJ, Peschon JJ, Slack JL, Wolfson MF, et al. A metalloproteinase disintegrin that releases tumournecrosis factor-α from cells. Nature 1997; 385: 729–33.

Koedel U, Bernatowicz A, Paul R, Frei K, Fontana A, Pfister H-W. Experimental pneumococcal meningitis: cerebrovascular alterations, brain edema, and meningeal inflammation are linked to the production of nitric oxide. Ann Neurol 1995; 37; 313–23.

Brown PD, Levy AT, Margulies IM, Liotta LA, Stetler-Stevenson WG. Independent expression and cellular processing of Mr 72,000 type IV collagenase and interstitial collagenase in human tumorigenic cell lines. Cancer Res 1990; 50: 6184–91. Cawston T. Matrix metalloproteinases and TIMPs: properties and implications for the rheumatic diseases. [Review]. Mol Med Today 1998; 4: 130–7. Gearing AJ, Beckett P, Christodoulou M, Churchill M, Clements J, Davidson AH, et al. Processing of tumour necrosis factor-α precursor by metalloproteinases. Nature 1994; 370: 555–7. Gijbels K, Masure S, Carton H, Opdenakker G. Gelatinase in the cerebrospinal fluid of patients with multiple sclerosis and other inflammatory neurological disorders. J. Neuroimmunol 1992; 41: 29–34. Gijbels K, Galardy RE, Steinman L. Reversal of experimental autoimmune encephalomyelitis with a hydroxamate inhibitor of matrix metalloproteases. J Clin Invest 1994; 94: 2177–82. Goetzl EJ, Banda MJ, Leppert D. Matrix metalloproteinases in immunity. [Review]. J Immunol, 1996; 156: 1–4. Gottschall PE, Yu X. Cytokines regulate gelatinase A and B (matrix metalloproteinase 2 and 9) activity in cultured rat astrocytes. J. Neurochem 1995; 64: 1513–20. Gray F, Nordmann P. Bacterial infections. In: Graham DI, Lantos PL, editors. Greenfield’s neuropathology, Vol. 1. London: Arnold; 1997. p. 113–52. Hartung HP. Pathogenesis of inflammatory demyelination: implications for therapy [editorial]. [Review]. Curr Opin Neurol 1995; 8: 191–9. Herron GS, Werb Z, Dwyer K, Banda MJ. Secretion of metalloproteinases by stimulated capillary endothelial cells. I. Production of procollagenase and prostromelysin exceeds expression of proteolytic activity. J Biol Chem 1986; 261: 2810–3. Hewson AK, Smith T, Leonard JP, Cuzner ML. Suppression of experimental allergic encephalomyelitis in the Lewis rat by the matrix metalloproteinase inhibitor Ro31–9790. Inflamm Res 1995; 44: 345–9. Huber AR, Weiss SJ. Disruption of the subendothelial basement membrane during neutrophil diapedesis in an in vitro construct of a blood vessel wall. J Clin Invest 1989; 83: 1122–36. Kieseier BC, Kiefer R, Clements JM, Miller K, Wells GM, Schweitzer T, et al. Matrix metalloproteinase-9 and -7 are regulated in experimental autoimmune encephalomyelitis. Brain 1998; 121: 159–66. Knauper V, Smith B, Lopez-Otin C, Murphy G. Activation of progelatinase B (proMMP-9) by active collagenase-3 (MMP-13). Eur J Biochem 1997 248: 369–73. Koedel U, Pfister H-W. Protective effect of the antioxidant N-

Kolb SA, Lahrtz F, Paul R, Leppert D, Nadal D, Pfister HW, et al. Matrix metalloproteinases and tissue inhibitors of metalloproteinases in viral meningitis: upregulation of MMP-9 and TIMP-1 in cerebrospinal fluid. J Neuroimmunol 1998; 84: 143–50. Kusano K, Miyaura C, Inada M, Tamura T, Ito A, Nagase H, et al. Regulation of matrix metalloproteinases (MMP-2, -3, -9, and -13) by interleukin-1 and interleukin-6 in mouse calvaria: association of MMP induction with bone resorption. Endocrinology 1998; 139: 1338–45. Leppert D, Waubant E, Galardy R, Bunnett NW, Hauser SL. T cell gelatinases mediate basement membrane transmigration in vitro. J Immunol 1995; 154: 4379–89. Masure S, Proost P, van Damme J, Opdenakker G. Purification and identification of 91-kDa neutrophil gelatinase. Release by the activating peptide interleukin-8. Eur J Biochem 1991; 198: 391–8. Mohler KM, Sleath PR, Fitzner JN, Cerretti DP, Alderson M, Kerwar SS, et al. Protection against a lethal dose of endotoxin by an inhibitor of tumor necrosis factor processing. Nature 1994; 370: 218–20. Moss ML, Jin S-L, Milla ME, Bickett DM, Burkhart W, Carter HL, Chen W-J, et al. Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-α [published erratum appears in Nature 1997; 385: 738]. Nature 1997; 385: 733–6. Nielsen BS, Timshel S, Kjeldsen L, Sehested M, Pyke C, Borregaard N, et al. 92 kDa type IV collagenase (MMP-9) is expressed in neutrophils and macrophages but not in malignant epithelial cells in human colon cancer. Int J Cancer 1996; 65: 57–62. Okada Y, Gonoji Y, Naka K, Tomita K, Nakanishi I, Iwata K, et al. Matrix metalloproteinase 9 (92-kDa gelatinase/type IV collagenase) from HT 1080 human fibrosarcoma cells. Purification and activation of the precursor and enzymic properties. J Biol Chem 1992; 267: 21712–9. Opdenakker G, Van Damme J. Cytokine-regulated proteases in autoimmune diseases. [Review]. Immunol Today 1994; 15: 103–7. Paul R, Koedel U, Pfister H-W. 7-Nitroindazole inhibits pial arteriolar vasodilation in a rat model of pneumococcal meningitis. J Cereb Blood Flow Metab 1997; 17: 985–91. Paul R, Lorenzl S, Koedel U, Sporer B, Vogel U, Frosch M, et al. Matrix metalloproteinases contribute to the blood–brain barrier disruption during bacterial meningitis. Ann Neurol 1998; 44: 592– 600. Pfister HW, Koedel U, Haberl RL, Dirnagl U, Feiden W, Ruckdeschel G, et al. Microvascular changes during the early phase of experimental bacterial meningitis. J Cereb Blood Flow Metab 1990; 10: 914–22. Pfister H-W, Koedel U, Lorenzl S, Tomasz A. Antioxidants attenuate microvascular changes in the early phase of experimental pneumococcal meningitis in rats. Stroke 1992; 23: 1798–804.

Matrix metalloproteinases in meningitis

1587

Quagliarello V, Scheld WM. Bacterial meningitis: pathogenesis, pathophysiology, and progress. [Review]. N Engl J Med 1992; 327: 864–72.

Rosenberg GA, Dencoff JE, Correa N, Reiners M, Ford CC. Effect of steroids on CSF matrix metalloproteinases in multiple sclerosis: relation to blood–brain barrier injury. Neurology 1996; 46: 1626–32.

Quagliarello VJ, Scheld WM. Treatment of bacterial meningitis [see comments]. [Review]. N Engl J Med 1997; 336: 708–16. Comment in: N Engl J Med 1997; 337: 793–4.

Waage A, Halstensen A, Shalaby R, Brandtzaeg P, Kierulf P, Espevik T. Local production of tumor necrosis factor α, interleukin 1, and interleukin 6 in meningococcal meningitis: relation to the inflammatory response. J Exp Med 1989; 170: 1859–67.

Quagliarello VJ, Wispelwey B, Long WJ Jr, Scheld WM. Recombinant human interleukin-1 induces meningitis and blood– brain barrier injury in the rat. J Clin Invest 1991; 87: 1360–6. Ries C, Petrides PE. Cytokine regulation of matrix metalloproteinase activity and its regulatory dysfunction in disease. Biol Chem HoppeSeyler 1995; 376: 345–55. Romanic AM, Madri JA. The induction of 72-kD gelatinase in T cells upon adhesion to endothelial cells is VCAM-1 dependent. J Cell Biol 1994; 125: 1165–78. Rosenberg GA, Kornfeld M, Estrada E, Kelley RO, Liotta LA, Stetler-Stevenson WG. TIMP-2 reduces proteolytic opening of blood–brain barrier by type IV collagenase. Brain Res 1992; 576: 203–7. Rosenberg GA, Dencoff JE, McGuire PG, Liotta LA, StetlerStevenson WA. Injury-induced 92-kilodalton gelatinase and urokinase expression in rat brain. Lab Invest 1994; 71: 417–22.

Welgus HG, Campbell EJ, Cury JD, Eisen AZ, Senior RM, Wilhelm SM, et al. Neutral metalloproteinases produced by human mononuclear phagocytes. J Clin Invest 1990; 86: 1496–502. Wells GM, Catlin G, Cossins JA, Mangan M, Ward GA, Miller KM, et al. Quantitation of matrix metalloproteinases in cultured rat astrocytes using the polymerase chain reaction with a multicompetitor cDNA standard. Glia; 18: 332–40. Woessner JF Jr. The family of matrix metalloproteinases. [Review]. Ann N Y Acad Sci 1994; 732: 11–21. Yong VW, Krekoski CA, Forsyth PA, Bell R, Edwards DR. Matrix metalloproteinases and diseases of the CNS. Trends Neurosci 1998; 21: 75–80.

Received September 18, 1998. Revised March 9, 1999. Accepted March 12, 1999