- Email: [email protected]
Matrix metalloproteinase-9 deficiency impairs host defense mechanisms against Streptococcus pneumoniae in a mouse model of bacterial meningitis
Neuroscience Letters 338 (2003) 201–204 www.elsevier.com/locate/neulet
Matrix metalloproteinase-9 deficiency impairs host defense mechanisms against Streptococcus pneumoniae in a mouse model of bacterial meningitis Tobias Bo¨ttcher, Annette Spreer, Ivo Azeh, Roland Nau*, Joachim Gerber Department of Neurology, Georg-August-University, Robert-Koch-Strasse 40, D-37075 Go¨ttingen, Germany Received 26 September 2002; received in revised form 22 November 2002; accepted 22 November 2002
Abstract Matrix metalloproteinase-9 (MMP-9) appears to contribute to blood – brain barrier damage and neuronal injury in bacterial meningitis. To further explore the function of MMP-9 in meningeal inflammation, we injected 104 colony forming units (CFU) of a Streptoccocus pneumoniae type 3 strain into the right forebrain of MMP-9 deficient mice (MMP-92/2 , n ¼ 16) and wild-type controls (129 £ B6, n ¼ 15). The clinical course of the disease, leukocyte recruitment into the subarachnoid space and bacterial titers in the brain did not differ. Yet, clearance of the bacteria from blood (log CFU/ml 4.7 [3.8/5.4] vs. 3.6 [3.0/4.0]; P ¼ 0:005) and spleen homogenates (log CFU/ml 5.3 [4.8/5.5] vs. 4.0 [2.8/4.7]; P ¼ 0:01) was reduced in MMP-9 deficient mice. A reduced systemic bacterial clearance of MMP-92/2 mice was confirmed in experimental S. pneumoniae peritonitis/sepsis. This implies a compromised systemic, but not intracerebral host response against S. pneumoniae in MMP-9 deficiency. q 2002 Published by Elsevier Science Ireland Ltd. Keywords: Matrix metalloproteinases; Meningitis; Sepsis; Streptococcus pneumoniae
Matrix metalloproteinases (MMP) comprise a large family of zinc-dependent endopeptidases with substrate affinity for a variety of components of the extracellular matrix (ECM) [17]. Upon stimulation, they are produced by neutrophils, neurons, glial, vascular smooth muscle and endothelial cells [9]. Released as inactive proforms and regulated by endogenous tissue inhibitors of metalloproteinases they play a role in health and disease. Since components of the ECM contribute to the integrity of the blood – brain barrier (BBB) [14], the influence of MMPs on BBB damage and neuronal injury in various diseases of the central nervous system (CNS), especially inflammatory disorders, has been studied. MMPs interact with well known mediators of inflammation, especially tumor necrosis factor-a (TNF-a), interleukin-b1, interleukin-6 and interleukin-10 [12]. In addition to a marked inflammatory reaction of the subarachnoid and ventricular space, acute bacterial meningitis is characterized by the breakdown of the BBB. Raised levels of MMP-9 in the cerebrospinal fluid (CSF) have been * Corresponding author. Tel.: þ 49-551-398455; fax: þ 49-551-398405. E-mail address: [email protected] (R. Nau).
observed in a rabbit model of pneumococcal meningitis [2] as well as in humans suffering from bacterial meningitis [10, 18]. In rat models of pneumococcal and meningococcal meningitis elevated MMP-9 mRNA expression in brain tissue and CSF cells, respectively, was demonstrated [6 – 8]. There is emerging evidence that particularly MMP-9 is involved in BBB disturbance during meningitis and other inflammatory neurological disorders [2,13]. In a rat model of meningococcal meningitis, inhibition of MMPs led to reduced BBB disruption [10]. Decreased concentrations of TNF-a in the CSF, reduced mortality and decreased cortical neuronal injury were observed after MMP inhibition in an infant rat model of Streptococcus pneumoniae meningitis [7,8]. In the present study we investigated the influence of the lack of MMP-9 on the course of experimental S. pneumoniae meningitis in mice. Sixteen MMP-92/2 mice on a 129 £ B6 background (gift of S. Nourshargh, Imperial College School of Medicine, Hammersmith Hospital, London, England; originally generated by R. Senior, Washington School of Medicine, St. Louis, MI) and 15 wild-type controls
0304-3940/02/$ - see front matter q 2002 Published by Elsevier Science Ireland Ltd. doi: 1 0 . 1 0 1 6 / S 0 3 0 4 - 3 9 4 0 ( 0 2 ) 0 1 4 0 6 - 4
202
T. Bo¨ttcher et al. / Neuroscience Letters 338 (2003) 201–204
Fig. 1. Gelatin substrate zymogram of EDTA-blood illustrating the lack of MMP-9 activity in MMP-9 deficient mice (3–5) compared to wild-type controls (1 þ 2). Note the preserved MMP-2 activity in all samples. Blood from a healthy human volunteer served as positive control (þ ). Marker lane (M); molecular mass is indicated by arrows at 92 kDa for MMP-9 and 72 kDa for MMP-2.
(129 £ B6) were anaesthetized with ketamine (100 mg/kg) and xylazine (20 mg/kg) and infected by injecting 4.3 ^ 0.4 log colony-forming units (CFU) of a S. pneumoniae type 3 strain (gift of M.G. Ta¨uber, Dept. of Infectious Diseases, University of Berne, Switzerland), dissolved in 12.5 ml sterile saline, into the right forebrain. MMP-9 deficiency was verified by zymography as follows. 8% sodium dodecyl sulfate (SDS)-polyacrylamide gel (60 £ 80 £ 1 mm) was copolymerized with 0.05% gelatin. EDTA-blood was prediluted 1:32 in H2O. 15 ml of this sample was mixed with 5 ml of 4 £ sample buffer (600 mM Tris – HCl pH 6,8 containing SDS 10%, glycerol 40%, and bromophenol blue) and then run on SDS-polyacrylamide gel electrophoresis at 4 8C. Washing of the gel was carried out for two 30 min periods with buffer (50 mM Tris – HCl, pH 7.5, containing 10 mM CaCl2 and 2.5% Triton X-100) followed by incubation for 12 h at 37 8C in the above buffer but containing only 1% Triton X-100. The gel was stained for 2 h with 0.2% Coomassie blue R-250 in 30% methanol þ 7.5% acetic acid, followed by destaining for two 30 min periods in 30% methanol þ 7.5% acetic acid. A blood preparation from a healthy human volunteer served as positive control (Fig. 1). According to ethical regulations mice unable to walk were killed. The animal experiments were approved by the Animal Care Committee of the University Hospital Go¨ttingen and the District Government, Braunschweig, Germany. The health status of the mice was assessed by weighing, by a clinical score and by a tight rope test score. Clinical and tight rope test score have been described in detail [5,15]. In brief, animals were assessed for their spontaneous, voluntary movement, ranging from zero (no apparent behavioral abnormality) up to four (dead). The tight rope test assesses motor activity. Animals were placed in the middle of a 60 cm long rope, tightly spanned between two platforms. The time to reach a platform at one end of the rope was measured and given a score, reaching from zero (mouse reaches end of the rope within 6 s) up to 20 (mouse falls from the rope immediately). Body weight, clinical score and tight rope test were performed prior to and 10 h, 24 h and 30 h after infection, respectively. Thirty hours after infection, animals were anaesthetized
and killed by decapitation. Four animals in each group were killed or died prematurely (between 24 h and 30 h after infection). Blood was sampled and the whole brain and spleen removed. The cerebellum and one half of the spleen were homogenized in saline. Serial tenfold dilution of blood, cerebellar and spleen homogenates were plated on blood-agar plates to determine bacterial titers. The cerebrum was fixed in 4% formaldehyde for histological examination. Haematoxylin-eosin (HE) stains of formaldehyde-fixed coronal brain sections were performed to evaluate meningeal inflammation. The degree of meningeal inflammation was estimated by a score, counting the number of granulocytes in one high-power field ( £ 40 objective) in different brain regions as described earlier [5]. The regional scores were added to reach a minimum of zero (no granulocytes) and a maximum of 21 (more than 50 granulocytes in every region investigated). To validate impaired systemic host defense in MMP-9 deficient animals we induced experimental S. pneumoniae peritonitis and sepsis by intraperitoneal challenging 6 MMP-92/2 mice and six wild-type controls with 100 CFU of the S. pneumoniae type 3 strain, dissolved in 100 ml sterile saline. Mice were killed 36 h after infection and bacterial titers in blood and spleen homogenates were determined as described above. Data are expressed as medians [25th/75th percentile]. Statistical comparisons were performed by Mann – Whitney U-test and P , 0:05 was considered significant. Thirty hours after intracerebral infection there were no differences between MMP-92/2 animals and 129 £ B6 mice in the clinical score (1.8 [1.5/2.5] vs. 1.5 [1.0/2.0]), the tight rope test score (17.0 [7.5/19.2] vs. 17.0 [3.0/20.0]) and in the change of body weight (2 2.8 g [2 3.5/ 2 1.9] vs. 2 2.5 g [2 3.2/ 2 1.5]), respectively (Table 1). The number of animals which were killed or died because of the severity of the disease before the end of the experiment did not differ either (n ¼ 4 in each group). The score of meningeal/ ventricular inflammation did not differ significantly between the two experimental groups (8.5 [7.0/11.1] vs. 9.0 [7.0/ 11.0]) (Fig. 2A). In cerebellar homogenates the bacterial titers (expressed as log CFU/ml) did not differ between MMP-92/2 mice and wild-type controls (6.5 [5.9/6.7] vs. 6.0 [5.0/6.5]) (Fig. 2B). In contrast, in the blood (4.7 [3.8/ Table 1 Severity of the disease in MMP-9 deficient mice and wild type controls (129 £ B6) during Streptococcus pneumoniae type 3 meningitis (30 h after infection)a
Body weight loss (g) Clinical score Tight rope test score a
129 £ B6 (n ¼ 15)
MMP-92/2 (n ¼ 16)
22.5 (23.2/ 2 1.5) 1.5 (1.0/2.0) 17.0 (3.0/20.0)
22.8 (23.5/ 2 1.9) 1.8 (1.5/2.5) 17.0 (7.5/19.2)
Data are expressed as medians (25th/75th percentile). Groups were compared by Mann– Whitney U-test. No significant differences were detected.
T. Bo¨ttcher et al. / Neuroscience Letters 338 (2003) 201–204
203
Fig. 2. Inflammation score in the subarachnoid space and bacterial titers in cerebellum, spleen and blood of MMP-9 deficient mice (MMP-92/2 ; n ¼ 16) and wild-type controls (129 £ B6; n ¼ 15) in Streptococcus pneumoniae type 3 meningitis. (A) Inflammation score: sum of the numbers of neutrophilic granulocytes in the subarachnoid space (no granulocyte – score 0; 1–10, score 1; 11–50, score 2; .50, score 3) in seven defined brain regions (three meningeal, the two temporobasal and the interhemispherical region and the third ventricle). No significant differences between both groups were detected (P ¼ 0:78). (B) Bacterial titers in cerebellar homogenates 30 h after infection: The density of viable bacteria did not differ significantly between MMP-9 deficient mice and controls (P ¼ 0:33). (C) Bacterial titers in blood 30 h after infection: bacterial titers were significantly higher in MMP-92/2 mice than in wild-type controls (P ¼ 0:005). (D) Bacterial titers in spleen homogenates 30 h after infection: MMP-92/2 mice had significantly higher titers of viable bacteria than 129 £ B6 mice (P ¼ 0:01).(Mann–Whitney U-test; **P , 0:01, *P , 0:05)
5.4] vs. 3.6 [3.0/4.0]; P ¼ 0:005) (Fig. 2C) and in spleen homogenates (5.3 [4.8/5.5] vs. 4.0 [2.8/4.7]; P ¼ 0:01) (Fig. 2D) bacterial titers were significantly higher in MMP-9 deficient than in control animals. In experimental peritonitis the bacterial titers in spleen homogenates differed significantly between MMP-92/2 mice and wild-type controls (7.3 [7.0/7.77] vs. 6.0 [5.5/6.2]; P ¼ 0:04), while in the blood a strong trend towards an impaired bacterial clearance in MMP-9 deficient animals was detected (6.6 [6.1/7.6] vs. 4.0 [2.0/6.0]; P ¼ 0:06). With respect to previously published data we hypothesized that animals deficient for MMP-9 would develop less severe meningeal inflammation and clinical symptoms. Surprisingly, we found no differences in the severity of the disease between the experimental groups during pneumococcal meningitis as assessed by monitoring body weight by repeated weighing, motor performance by tight rope test and voluntary movement activity by a clinical score. All tests demonstrated a similar degree of disease-related clinical deterioration, suggesting equal severity of meningitis in both groups. Histological examination of the brains revealed leukocyte recruitment into the subarachnoid space in MMP-
92/2 mice which was not distinguishable from wild-type animals. Obviously, neutrophil recruitment to the site of inflammation was not disturbed in MMP-92/2 animals. Furthermore, there were no differences in the densities of viable bacteria in cerebellar homogenates between wildtype and MMP-92/2 mice. On the contrary, the clearance of S. pneumoniae from the bloodstream was severely affected in MMP-9 deficient mice resulting in bacterial titers in the blood and spleen homogenates which were more than 1 log higher than in wild-type controls. This impaired host defense in MMP-9 deficiency was independently demonstrated in S. pneumoniae-induced peritonitis and sepsis, also resulting in higher bacterial titers in blood and spleen homogenates in MMP-92/2 animals. The present data confirm that the CNS has characteristics of a physiologically immunocompromised site. Even an immunocompetent organism does not manage to inhibit bacterial growth in the subarachnoid space. Rabbits, rendered artificially leukopenic by nitrogen mustard showed comparable bacterial growth rates in the CSF as control animals in experimental S. pneumoniae meningitis [4]. TNF-a deficiency did not influence bacterial titers in
204
T. Bo¨ttcher et al. / Neuroscience Letters 338 (2003) 201–204
cerebellar homogenates and leukocyte recruitment into the subarachnoid space in a mouse model of pneumococcal meningitis [16]. The finding of unaffected CNS inflammation in MMP-92/2 mice contrasts with the increased bacterial titers in blood and spleen in these animals. Obviously, MMP-9 deficiency leads to an impaired immune response with reduced clearing of bacteria in the systemic circulation. Promoting the migration of granulocytes and other immunocompetent cells across physiological barriers such as endothelium and basement membranes is one of the main function of MMPs [1,11]. Yet, the lack of MMP-9 did not influence leukocyte migration in our study, as evidenced by unaffected granulocyte recruitment into the subarachnoid space. Similar results have been demonstrated in vitro where neutrophils from MMP-9 deficient mice did not show any defects in transendothelial migration [1]. Disturbances in the inflammatory cytokine network might explain the immunocompromising effect of MMP-9 deficiency since MMPs interact with these cytokines in a tightly regulated manner [12]. Higher bacterial titers in blood and spleen despite comparable CNS inflammation were also observed in meningitic TNF-a deficient mice [16]. Similarly, pneumococcal meningitis in TLR2-deficient mice led to an increased inflammatory response with higher clinical severity scores and lower survival compared to control animals, while leukocyte count and bacterial titers in the CSF remained unchanged [3]. Although the action of MMPs in the pathogenesis of bacterial meningitis appears deleterious for the host, MMP9 must be considered an important defense mechanism for the clearance of bacteria from the bloodstream in S. pneumoniae infection thereby preventing septicaemic spreading of pneumococcal disease.
Acknowledgements The study was supported by a grant to T.B. from the Deutsche Akademie der Naturforscher Leopoldina, sponsored by the Federal Ministry for Education and Research, and by a grant to T.B. and R.N. from the Else Kro¨nerFresenius Foundation.
References [1] J. Allport, Y-C. Lim, J. Shipley, R. Senior, S. Shapiro, N. Matsuyoshi, D. Vestweber, F. Luscinskas, Neutrophils from MMP-9 or neutrophil elastase-deficient mice show no defect in transendothelial migration under flow in vitro, J. Leukocyte Biol. 71 (2002) 821–828. [2] I. Azeh, M. Maeder, A. Smirnov, W. Beuche, R. Nau, F. Weber, Experimental pneumococcal meningitis in rabbits: the increase of matrix metalloproteinase-9 in cerebrospinal fluid correlates with leukocyte invasion, Neurosci. Lett. 256 (1998) 127 –130.
[3] H. Echchannaoui, K. Frei, C. Schnell, S. Leib, W. Zimmerli, R. Landmann, TLR2 deficient mice are highly susceptible to Streptococcus pneumoniae meningitis because of reduced bacterial clearing and enhanced inflammation, J. Infect. Dis. 186 (2002) 798–806. [4] J. Ernst, J. Decazes, M. Sande, Experimental pneumococcal meningitis: role of leukocytes in pathogenesis, Infect. Immun. 41 (1983) 275–279. [5] J. Gerber, G. Raivich, A. Wellmer, C. Noeske, T. Kunst, A. Werner, W. Bruck, R. Nau, A mouse model of Streptococcus pneumoniae meningitis mimicking several features of human disease, Acta Neuropathol. (Berlin) 101 (2001) 499–508. [6] B. Kieseier, R. Paul, U. Koedel, T. Seifert, J. Clements, A. Gearing, H. Pfister, H. Hartung, Differential expression of matrix metalloproteinases in bacterial meningitis, Brain 122 (1999) 1579–1587. [7] S. Leib, D. Leppert, J. Clements, M. Ta¨uber, Matrix metalloproteinases contribute to brain damage in Experimental Pneumococcal Meningitis, Infect. Immun. 68 (2000) 615–620. [8] S. Leib, J. Clements, R. Lindberg, C. Heimgartner, J. Loeffler, L. Pfister, M. Ta¨uber, D. Leppert, Inhibition of matrix metalloproteinases and tumour necrosis factor alpha converting enzyme as adjuvant therapy in pneumococcal meningitis, Brain 124 (2001) 1734–1742. [9] A. Pagenstecher, A. Stalder, C. Kincaid, S. Shapiro, I. Campbell, Differential expression of matrix metalloproteinases and tissue inhibitiors of matrix metalloproteinases genes in the mouse central nervous system in normal and inflammatory state, Am. J. Pathol. 152 (1998) 729 –741. [10] R. Paul, S. Lorenzl, U. Koedel, B. Sporer, U. Vogel, M. Frosch, H. Pfister, Matrix metalloproteinases contribute to the blood–brain barrier disruption during bacterial meningitis, Ann. Neurol. 44 (1998) 592 –600. [11] G. Ratzinger, P. Stoitzner, S. Ebner, M. Lutz, G. Layton, C. Rainer, R. Senior, J. Shipley, P. Fritsch, G. Schuler, N. Romani, Matrix Metalloproteinases 9 and 2 are necessary for the migration of Langerhans cells and dermal dendritic cells from human and murine skin, J. Immunol. 168 (2002) 4361–4371. [12] C. Ries, P. Petrides, Cytokine regulation of matrix metalloproteinase activity and its regulatory dysfunction in disease, Biol. Chem. HoppeSeyler 376 (1995) 345–355. [13] G. Rosenberg, J. Dencoff, N. Correa, M. Reiners, C. Ford, Effect of steroids on CSF matrix metalloproteinases in multiple sclerosis: relation to blood – brain barrier injury, Neurology 46 (1996) 1626–1632. [14] I. Sapsford, J. Buontempo, R. Weller, Basement membrane surfaces and perivascular compartments in normal human brain and glial tumors: a scanning electron microscope study, Neuropathol. Appl. Neurobiol. 9 (1983) 181– 194. [15] A. Wellmer, C. Noeske, J. Gerber, U. Munzel, R. Nau, Spatial memory and learning deficits after experimental pneumococcal meningitis in mice, Neurosci. Lett. 296 (2000) 137 –140. [16] A. Wellmer, J. Gerber, J. Ragheb, G. Zysk, T. Kunst, A. Smirnov, W. Bruck, R. Nau, Effect of deficiency of tumor necrosis factor alpha or both of its receptors on Streptococcus pneumoniae central nervous system infection and peritonitis, Infect. Immun. 69 (2001) 6881–6886. [17] J. Woessner, The family of matrix metalloproteinases, Ann. N.Y. Acad. Sci. 732 (1994) 11–21. [18] M. Yushenko, F. Weber, M. Ma¨der, U. Scho¨ll, M. Maliszweska, H. Tumani, K. Felgenhauer, W. Beuche, Matrix metalloproteinase-9 (MMP-9) in human cerebrospinal fluid (CSF): elevated levels are primarily related to CSF cell count, J. Neuroimmunol. 110 (2000) 244 –251.
Matrix metalloproteinase-9 deficiency impairs host defense mechanisms against Streptococcus pneumoniae in a mouse model of bacterial meningitis Tobias Bo¨ttcher, Annette Spreer, Ivo Azeh, Roland Nau*, Joachim Gerber Department of Neurology, Georg-August-University, Robert-Koch-Strasse 40, D-37075 Go¨ttingen, Germany Received 26 September 2002; received in revised form 22 November 2002; accepted 22 November 2002
Abstract Matrix metalloproteinase-9 (MMP-9) appears to contribute to blood – brain barrier damage and neuronal injury in bacterial meningitis. To further explore the function of MMP-9 in meningeal inflammation, we injected 104 colony forming units (CFU) of a Streptoccocus pneumoniae type 3 strain into the right forebrain of MMP-9 deficient mice (MMP-92/2 , n ¼ 16) and wild-type controls (129 £ B6, n ¼ 15). The clinical course of the disease, leukocyte recruitment into the subarachnoid space and bacterial titers in the brain did not differ. Yet, clearance of the bacteria from blood (log CFU/ml 4.7 [3.8/5.4] vs. 3.6 [3.0/4.0]; P ¼ 0:005) and spleen homogenates (log CFU/ml 5.3 [4.8/5.5] vs. 4.0 [2.8/4.7]; P ¼ 0:01) was reduced in MMP-9 deficient mice. A reduced systemic bacterial clearance of MMP-92/2 mice was confirmed in experimental S. pneumoniae peritonitis/sepsis. This implies a compromised systemic, but not intracerebral host response against S. pneumoniae in MMP-9 deficiency. q 2002 Published by Elsevier Science Ireland Ltd. Keywords: Matrix metalloproteinases; Meningitis; Sepsis; Streptococcus pneumoniae
Matrix metalloproteinases (MMP) comprise a large family of zinc-dependent endopeptidases with substrate affinity for a variety of components of the extracellular matrix (ECM) [17]. Upon stimulation, they are produced by neutrophils, neurons, glial, vascular smooth muscle and endothelial cells [9]. Released as inactive proforms and regulated by endogenous tissue inhibitors of metalloproteinases they play a role in health and disease. Since components of the ECM contribute to the integrity of the blood – brain barrier (BBB) [14], the influence of MMPs on BBB damage and neuronal injury in various diseases of the central nervous system (CNS), especially inflammatory disorders, has been studied. MMPs interact with well known mediators of inflammation, especially tumor necrosis factor-a (TNF-a), interleukin-b1, interleukin-6 and interleukin-10 [12]. In addition to a marked inflammatory reaction of the subarachnoid and ventricular space, acute bacterial meningitis is characterized by the breakdown of the BBB. Raised levels of MMP-9 in the cerebrospinal fluid (CSF) have been * Corresponding author. Tel.: þ 49-551-398455; fax: þ 49-551-398405. E-mail address: [email protected] (R. Nau).
observed in a rabbit model of pneumococcal meningitis [2] as well as in humans suffering from bacterial meningitis [10, 18]. In rat models of pneumococcal and meningococcal meningitis elevated MMP-9 mRNA expression in brain tissue and CSF cells, respectively, was demonstrated [6 – 8]. There is emerging evidence that particularly MMP-9 is involved in BBB disturbance during meningitis and other inflammatory neurological disorders [2,13]. In a rat model of meningococcal meningitis, inhibition of MMPs led to reduced BBB disruption [10]. Decreased concentrations of TNF-a in the CSF, reduced mortality and decreased cortical neuronal injury were observed after MMP inhibition in an infant rat model of Streptococcus pneumoniae meningitis [7,8]. In the present study we investigated the influence of the lack of MMP-9 on the course of experimental S. pneumoniae meningitis in mice. Sixteen MMP-92/2 mice on a 129 £ B6 background (gift of S. Nourshargh, Imperial College School of Medicine, Hammersmith Hospital, London, England; originally generated by R. Senior, Washington School of Medicine, St. Louis, MI) and 15 wild-type controls
0304-3940/02/$ - see front matter q 2002 Published by Elsevier Science Ireland Ltd. doi: 1 0 . 1 0 1 6 / S 0 3 0 4 - 3 9 4 0 ( 0 2 ) 0 1 4 0 6 - 4
202
T. Bo¨ttcher et al. / Neuroscience Letters 338 (2003) 201–204
Fig. 1. Gelatin substrate zymogram of EDTA-blood illustrating the lack of MMP-9 activity in MMP-9 deficient mice (3–5) compared to wild-type controls (1 þ 2). Note the preserved MMP-2 activity in all samples. Blood from a healthy human volunteer served as positive control (þ ). Marker lane (M); molecular mass is indicated by arrows at 92 kDa for MMP-9 and 72 kDa for MMP-2.
(129 £ B6) were anaesthetized with ketamine (100 mg/kg) and xylazine (20 mg/kg) and infected by injecting 4.3 ^ 0.4 log colony-forming units (CFU) of a S. pneumoniae type 3 strain (gift of M.G. Ta¨uber, Dept. of Infectious Diseases, University of Berne, Switzerland), dissolved in 12.5 ml sterile saline, into the right forebrain. MMP-9 deficiency was verified by zymography as follows. 8% sodium dodecyl sulfate (SDS)-polyacrylamide gel (60 £ 80 £ 1 mm) was copolymerized with 0.05% gelatin. EDTA-blood was prediluted 1:32 in H2O. 15 ml of this sample was mixed with 5 ml of 4 £ sample buffer (600 mM Tris – HCl pH 6,8 containing SDS 10%, glycerol 40%, and bromophenol blue) and then run on SDS-polyacrylamide gel electrophoresis at 4 8C. Washing of the gel was carried out for two 30 min periods with buffer (50 mM Tris – HCl, pH 7.5, containing 10 mM CaCl2 and 2.5% Triton X-100) followed by incubation for 12 h at 37 8C in the above buffer but containing only 1% Triton X-100. The gel was stained for 2 h with 0.2% Coomassie blue R-250 in 30% methanol þ 7.5% acetic acid, followed by destaining for two 30 min periods in 30% methanol þ 7.5% acetic acid. A blood preparation from a healthy human volunteer served as positive control (Fig. 1). According to ethical regulations mice unable to walk were killed. The animal experiments were approved by the Animal Care Committee of the University Hospital Go¨ttingen and the District Government, Braunschweig, Germany. The health status of the mice was assessed by weighing, by a clinical score and by a tight rope test score. Clinical and tight rope test score have been described in detail [5,15]. In brief, animals were assessed for their spontaneous, voluntary movement, ranging from zero (no apparent behavioral abnormality) up to four (dead). The tight rope test assesses motor activity. Animals were placed in the middle of a 60 cm long rope, tightly spanned between two platforms. The time to reach a platform at one end of the rope was measured and given a score, reaching from zero (mouse reaches end of the rope within 6 s) up to 20 (mouse falls from the rope immediately). Body weight, clinical score and tight rope test were performed prior to and 10 h, 24 h and 30 h after infection, respectively. Thirty hours after infection, animals were anaesthetized
and killed by decapitation. Four animals in each group were killed or died prematurely (between 24 h and 30 h after infection). Blood was sampled and the whole brain and spleen removed. The cerebellum and one half of the spleen were homogenized in saline. Serial tenfold dilution of blood, cerebellar and spleen homogenates were plated on blood-agar plates to determine bacterial titers. The cerebrum was fixed in 4% formaldehyde for histological examination. Haematoxylin-eosin (HE) stains of formaldehyde-fixed coronal brain sections were performed to evaluate meningeal inflammation. The degree of meningeal inflammation was estimated by a score, counting the number of granulocytes in one high-power field ( £ 40 objective) in different brain regions as described earlier [5]. The regional scores were added to reach a minimum of zero (no granulocytes) and a maximum of 21 (more than 50 granulocytes in every region investigated). To validate impaired systemic host defense in MMP-9 deficient animals we induced experimental S. pneumoniae peritonitis and sepsis by intraperitoneal challenging 6 MMP-92/2 mice and six wild-type controls with 100 CFU of the S. pneumoniae type 3 strain, dissolved in 100 ml sterile saline. Mice were killed 36 h after infection and bacterial titers in blood and spleen homogenates were determined as described above. Data are expressed as medians [25th/75th percentile]. Statistical comparisons were performed by Mann – Whitney U-test and P , 0:05 was considered significant. Thirty hours after intracerebral infection there were no differences between MMP-92/2 animals and 129 £ B6 mice in the clinical score (1.8 [1.5/2.5] vs. 1.5 [1.0/2.0]), the tight rope test score (17.0 [7.5/19.2] vs. 17.0 [3.0/20.0]) and in the change of body weight (2 2.8 g [2 3.5/ 2 1.9] vs. 2 2.5 g [2 3.2/ 2 1.5]), respectively (Table 1). The number of animals which were killed or died because of the severity of the disease before the end of the experiment did not differ either (n ¼ 4 in each group). The score of meningeal/ ventricular inflammation did not differ significantly between the two experimental groups (8.5 [7.0/11.1] vs. 9.0 [7.0/ 11.0]) (Fig. 2A). In cerebellar homogenates the bacterial titers (expressed as log CFU/ml) did not differ between MMP-92/2 mice and wild-type controls (6.5 [5.9/6.7] vs. 6.0 [5.0/6.5]) (Fig. 2B). In contrast, in the blood (4.7 [3.8/ Table 1 Severity of the disease in MMP-9 deficient mice and wild type controls (129 £ B6) during Streptococcus pneumoniae type 3 meningitis (30 h after infection)a
Body weight loss (g) Clinical score Tight rope test score a
129 £ B6 (n ¼ 15)
MMP-92/2 (n ¼ 16)
22.5 (23.2/ 2 1.5) 1.5 (1.0/2.0) 17.0 (3.0/20.0)
22.8 (23.5/ 2 1.9) 1.8 (1.5/2.5) 17.0 (7.5/19.2)
Data are expressed as medians (25th/75th percentile). Groups were compared by Mann– Whitney U-test. No significant differences were detected.
T. Bo¨ttcher et al. / Neuroscience Letters 338 (2003) 201–204
203
Fig. 2. Inflammation score in the subarachnoid space and bacterial titers in cerebellum, spleen and blood of MMP-9 deficient mice (MMP-92/2 ; n ¼ 16) and wild-type controls (129 £ B6; n ¼ 15) in Streptococcus pneumoniae type 3 meningitis. (A) Inflammation score: sum of the numbers of neutrophilic granulocytes in the subarachnoid space (no granulocyte – score 0; 1–10, score 1; 11–50, score 2; .50, score 3) in seven defined brain regions (three meningeal, the two temporobasal and the interhemispherical region and the third ventricle). No significant differences between both groups were detected (P ¼ 0:78). (B) Bacterial titers in cerebellar homogenates 30 h after infection: The density of viable bacteria did not differ significantly between MMP-9 deficient mice and controls (P ¼ 0:33). (C) Bacterial titers in blood 30 h after infection: bacterial titers were significantly higher in MMP-92/2 mice than in wild-type controls (P ¼ 0:005). (D) Bacterial titers in spleen homogenates 30 h after infection: MMP-92/2 mice had significantly higher titers of viable bacteria than 129 £ B6 mice (P ¼ 0:01).(Mann–Whitney U-test; **P , 0:01, *P , 0:05)
5.4] vs. 3.6 [3.0/4.0]; P ¼ 0:005) (Fig. 2C) and in spleen homogenates (5.3 [4.8/5.5] vs. 4.0 [2.8/4.7]; P ¼ 0:01) (Fig. 2D) bacterial titers were significantly higher in MMP-9 deficient than in control animals. In experimental peritonitis the bacterial titers in spleen homogenates differed significantly between MMP-92/2 mice and wild-type controls (7.3 [7.0/7.77] vs. 6.0 [5.5/6.2]; P ¼ 0:04), while in the blood a strong trend towards an impaired bacterial clearance in MMP-9 deficient animals was detected (6.6 [6.1/7.6] vs. 4.0 [2.0/6.0]; P ¼ 0:06). With respect to previously published data we hypothesized that animals deficient for MMP-9 would develop less severe meningeal inflammation and clinical symptoms. Surprisingly, we found no differences in the severity of the disease between the experimental groups during pneumococcal meningitis as assessed by monitoring body weight by repeated weighing, motor performance by tight rope test and voluntary movement activity by a clinical score. All tests demonstrated a similar degree of disease-related clinical deterioration, suggesting equal severity of meningitis in both groups. Histological examination of the brains revealed leukocyte recruitment into the subarachnoid space in MMP-
92/2 mice which was not distinguishable from wild-type animals. Obviously, neutrophil recruitment to the site of inflammation was not disturbed in MMP-92/2 animals. Furthermore, there were no differences in the densities of viable bacteria in cerebellar homogenates between wildtype and MMP-92/2 mice. On the contrary, the clearance of S. pneumoniae from the bloodstream was severely affected in MMP-9 deficient mice resulting in bacterial titers in the blood and spleen homogenates which were more than 1 log higher than in wild-type controls. This impaired host defense in MMP-9 deficiency was independently demonstrated in S. pneumoniae-induced peritonitis and sepsis, also resulting in higher bacterial titers in blood and spleen homogenates in MMP-92/2 animals. The present data confirm that the CNS has characteristics of a physiologically immunocompromised site. Even an immunocompetent organism does not manage to inhibit bacterial growth in the subarachnoid space. Rabbits, rendered artificially leukopenic by nitrogen mustard showed comparable bacterial growth rates in the CSF as control animals in experimental S. pneumoniae meningitis [4]. TNF-a deficiency did not influence bacterial titers in
204
T. Bo¨ttcher et al. / Neuroscience Letters 338 (2003) 201–204
cerebellar homogenates and leukocyte recruitment into the subarachnoid space in a mouse model of pneumococcal meningitis [16]. The finding of unaffected CNS inflammation in MMP-92/2 mice contrasts with the increased bacterial titers in blood and spleen in these animals. Obviously, MMP-9 deficiency leads to an impaired immune response with reduced clearing of bacteria in the systemic circulation. Promoting the migration of granulocytes and other immunocompetent cells across physiological barriers such as endothelium and basement membranes is one of the main function of MMPs [1,11]. Yet, the lack of MMP-9 did not influence leukocyte migration in our study, as evidenced by unaffected granulocyte recruitment into the subarachnoid space. Similar results have been demonstrated in vitro where neutrophils from MMP-9 deficient mice did not show any defects in transendothelial migration [1]. Disturbances in the inflammatory cytokine network might explain the immunocompromising effect of MMP-9 deficiency since MMPs interact with these cytokines in a tightly regulated manner [12]. Higher bacterial titers in blood and spleen despite comparable CNS inflammation were also observed in meningitic TNF-a deficient mice [16]. Similarly, pneumococcal meningitis in TLR2-deficient mice led to an increased inflammatory response with higher clinical severity scores and lower survival compared to control animals, while leukocyte count and bacterial titers in the CSF remained unchanged [3]. Although the action of MMPs in the pathogenesis of bacterial meningitis appears deleterious for the host, MMP9 must be considered an important defense mechanism for the clearance of bacteria from the bloodstream in S. pneumoniae infection thereby preventing septicaemic spreading of pneumococcal disease.
Acknowledgements The study was supported by a grant to T.B. from the Deutsche Akademie der Naturforscher Leopoldina, sponsored by the Federal Ministry for Education and Research, and by a grant to T.B. and R.N. from the Else Kro¨nerFresenius Foundation.
References [1] J. Allport, Y-C. Lim, J. Shipley, R. Senior, S. Shapiro, N. Matsuyoshi, D. Vestweber, F. Luscinskas, Neutrophils from MMP-9 or neutrophil elastase-deficient mice show no defect in transendothelial migration under flow in vitro, J. Leukocyte Biol. 71 (2002) 821–828. [2] I. Azeh, M. Maeder, A. Smirnov, W. Beuche, R. Nau, F. Weber, Experimental pneumococcal meningitis in rabbits: the increase of matrix metalloproteinase-9 in cerebrospinal fluid correlates with leukocyte invasion, Neurosci. Lett. 256 (1998) 127 –130.
[3] H. Echchannaoui, K. Frei, C. Schnell, S. Leib, W. Zimmerli, R. Landmann, TLR2 deficient mice are highly susceptible to Streptococcus pneumoniae meningitis because of reduced bacterial clearing and enhanced inflammation, J. Infect. Dis. 186 (2002) 798–806. [4] J. Ernst, J. Decazes, M. Sande, Experimental pneumococcal meningitis: role of leukocytes in pathogenesis, Infect. Immun. 41 (1983) 275–279. [5] J. Gerber, G. Raivich, A. Wellmer, C. Noeske, T. Kunst, A. Werner, W. Bruck, R. Nau, A mouse model of Streptococcus pneumoniae meningitis mimicking several features of human disease, Acta Neuropathol. (Berlin) 101 (2001) 499–508. [6] B. Kieseier, R. Paul, U. Koedel, T. Seifert, J. Clements, A. Gearing, H. Pfister, H. Hartung, Differential expression of matrix metalloproteinases in bacterial meningitis, Brain 122 (1999) 1579–1587. [7] S. Leib, D. Leppert, J. Clements, M. Ta¨uber, Matrix metalloproteinases contribute to brain damage in Experimental Pneumococcal Meningitis, Infect. Immun. 68 (2000) 615–620. [8] S. Leib, J. Clements, R. Lindberg, C. Heimgartner, J. Loeffler, L. Pfister, M. Ta¨uber, D. Leppert, Inhibition of matrix metalloproteinases and tumour necrosis factor alpha converting enzyme as adjuvant therapy in pneumococcal meningitis, Brain 124 (2001) 1734–1742. [9] A. Pagenstecher, A. Stalder, C. Kincaid, S. Shapiro, I. Campbell, Differential expression of matrix metalloproteinases and tissue inhibitiors of matrix metalloproteinases genes in the mouse central nervous system in normal and inflammatory state, Am. J. Pathol. 152 (1998) 729 –741. [10] R. Paul, S. Lorenzl, U. Koedel, B. Sporer, U. Vogel, M. Frosch, H. Pfister, Matrix metalloproteinases contribute to the blood–brain barrier disruption during bacterial meningitis, Ann. Neurol. 44 (1998) 592 –600. [11] G. Ratzinger, P. Stoitzner, S. Ebner, M. Lutz, G. Layton, C. Rainer, R. Senior, J. Shipley, P. Fritsch, G. Schuler, N. Romani, Matrix Metalloproteinases 9 and 2 are necessary for the migration of Langerhans cells and dermal dendritic cells from human and murine skin, J. Immunol. 168 (2002) 4361–4371. [12] C. Ries, P. Petrides, Cytokine regulation of matrix metalloproteinase activity and its regulatory dysfunction in disease, Biol. Chem. HoppeSeyler 376 (1995) 345–355. [13] G. Rosenberg, J. Dencoff, N. Correa, M. Reiners, C. Ford, Effect of steroids on CSF matrix metalloproteinases in multiple sclerosis: relation to blood – brain barrier injury, Neurology 46 (1996) 1626–1632. [14] I. Sapsford, J. Buontempo, R. Weller, Basement membrane surfaces and perivascular compartments in normal human brain and glial tumors: a scanning electron microscope study, Neuropathol. Appl. Neurobiol. 9 (1983) 181– 194. [15] A. Wellmer, C. Noeske, J. Gerber, U. Munzel, R. Nau, Spatial memory and learning deficits after experimental pneumococcal meningitis in mice, Neurosci. Lett. 296 (2000) 137 –140. [16] A. Wellmer, J. Gerber, J. Ragheb, G. Zysk, T. Kunst, A. Smirnov, W. Bruck, R. Nau, Effect of deficiency of tumor necrosis factor alpha or both of its receptors on Streptococcus pneumoniae central nervous system infection and peritonitis, Infect. Immun. 69 (2001) 6881–6886. [17] J. Woessner, The family of matrix metalloproteinases, Ann. N.Y. Acad. Sci. 732 (1994) 11–21. [18] M. Yushenko, F. Weber, M. Ma¨der, U. Scho¨ll, M. Maliszweska, H. Tumani, K. Felgenhauer, W. Beuche, Matrix metalloproteinase-9 (MMP-9) in human cerebrospinal fluid (CSF): elevated levels are primarily related to CSF cell count, J. Neuroimmunol. 110 (2000) 244 –251.