| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
*
Department of Infectious Diseases, Imperial College School of Medicine, Hammersmith Hospital, London, United Kingdom;
Wellcome Trust Clinical Research Unit, Ho Chi Minh City, Viet Nam;
Center for Tropical Medicine, Nuffield Department of Clinical Medicine, John Radcliffe Hospital, Oxford, United Kingdom; and
Center for Tropical Diseases, Ho Chi Minh City, Viet Nam.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Increased MMP-9 secretion from infiltrating leukocytes may result in
cerebral injury in infectious and inflammatory conditions. MMP-9 is not
normally present in cerebrospinal fluid (CSF), but has been detected in
patients with Lyme disease (11), viral meningitis
(12), human T cell lymphotrophic virus-1-associated
myelopathy (13), and multiple sclerosis (14, 15). MMP-9 was detected in 40% of patients with HIV infection
and more commonly in cases with neurological deficit (16).
CSF MMP-9 concentrations were elevated in patients with bacterial
meningitis and fell during recovery (17). In rodents MMP-9
may cause cerebral injury by provoking leukocyte recruitment,
disrupting the integrity of the blood-brain barrier (10, 17, 18, 19) and by cleaving myelin proteins (20).
TNF-
and IL-1
, implicated in pathological opening of the BBB in
meningitis (21), induce MMP-9 secretion by macrophages
(22). Conversely, other MMPs may release proinflammatory
cytokines, such as membrane-bound TNF- (23, 24).
Furthermore, blocking MMP activity reduced intracranial pressure,
blood-brain barrier disruption, and cerebral inflammation in rodent
models of bacterial meningitis (17, 24). Direct evidence
implicating MMP-9 activity in cerebral injury has come from studies in
MMP-9 knockout mice, which were resistant to cerebral damage from
experimentally induced autoimmune encephalomyelitis (25).
In contrast, others found delayed resolution to contact
hypersensitivity in these animals (26).
To protect the host during immune responses, MMP activity is down-regulated by specific tissue inhibitors of matrix metalloproteinases (TIMPs). TIMP-1 binds to the active and latent forms of MMP-9 and is the major TIMP secreted by mononuclear phagocytes (27). It is constitutively secreted into many tissue fluids, including CSF (14, 17), and the balance between the local TIMP-1 and MMP-9 concentrations critically determines net proteolytic activity.
In this study we present novel data showing that infection with M. tuberculosis leads to a matrix-degrading phenotype in vitro and in vivo. First, the effect of infection with live, virulent M. tuberculosis on human monocytic cell gene expression and secretion of MMP-9 and TIMP-1 was investigated in vitro. These experiments involved human monocytic THP-1 cells, which resemble primary monocytes in a number of important respects. In particular, THP-1 cells have phenotypic characteristics similar to those of primary cells, actively phagocytose tubercle bacilli (28), and secrete a similar profile of MMPs (29, 30). We next examined MMP-9 and TIMP-1 secretion in vivo in CSF from patients with tuberculous meningitis. The data show significantly elevated MMP-9 activity in tuberculous meningitis compared with that in bacterial and viral meningitis. As in in vitro studies, there was good evidence for the development of a phenotype in which increased MMP-9 activity was relatively unopposed by TIMP-1 in tuberculous meningitis. Elevated MMP-9 activity was related to focal neurological damage and death.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RPMI 1640 was obtained from Life Technologies (Paisley, U.K.),
and Dubos’ enriched media was purchased from Difco (Detroit, MI). HRP
was obtained from Dako (Glostrup, Denmark), and purified pro-MMP-9 was
obtained from Calbiochem (Nottingham, U.K.). For zymogram gels, AccuGel
29:1 (30% acrylamide and acrylamide/bis-acrylamide (29/1)), ProtoGel
stacking and running buffers were obtained from National Diagnostics
(Atlanta, GA). Triton X-100 was obtained from BDH (Poole, U.K.).
Coomassie blue tablets were obtained from Pharmacia Biotech (Uppsala,
Sweden). Redi-Prime II random primer labeling system,
[-32P]dCTP,
[
-32P]dCTP, nitrocellulose membranes
(Hybond-N and Hybond-C) and Hyperfilm ECL were purchased from Amersham
(Little Chalfont, U.K.). Kodak Biomax MS-1 film (Eastman Kodak,
Rochester, New York) was used for autoradiography. TIMP-1 standard and
mAbs were gifts from Prof. Timothy Cawston (University of
Newcastle-Upon-Tyne, Newcastle-Upon-Tyne, U.K.). Sheep anti-human
pro-MMP-9 Ab and peroxidase-conjugated donkey anti-sheep IgG were
purchased from The Binding Site (Birmingham, U.K.). All other reagents
were obtained from Sigma (Poole, U.K.).
Culture of M. tuberculosis and THP-1 cells
Stocks of live virulent M. tuberculosis, strain H37-Rv (from Dr. V. Snewin, Imperial College, London, U.K.), were maintained at 37°C in Dubos’ medium enriched with albumin Cohn fraction V plus dextrose and sodium chloride (endotoxin level, <3 pg/ml). To generate single-cell suspensions, aliquots were briefly sonicated and passed eight times through a 22-gauge needle. This was confirmed by modified Kinyoun staining. The multiplicity of infection (MOI) used in experiments was quantitated by colony counting in triplicate on Middlebrook 7H10 plates.
THP-1 cells (European Collection of Animal Cell Cultures, no. 88081201; Salisbury, U.K.) were maintained in RPMI 1640, supplemented with 10% FCS (endotoxin level, <20 pg/ml), L-glutamine (2 mM), and ampicillin (100 µg/ml; ampicillin is not active against M. tuberculosis at this concentration). Cultures were incubated in a humidified 5% CO2 atmosphere at 37°C. For experiments, cells were suspended in serum-free medium at 2 x 106 cells/ml in six-well tissue culture plates. Cells were stimulated with M. tuberculosis (MOI = 1), 1 µg/ml LPS (Escherichia coli serotype 0127:B8, positive control), or inert 3-µm latex beads (negative particulate control of comparable size to tubercle bacilli).
Gelatin zymography
MMP-9 was detected by zymography using standard methodology (31). Briefly, cell culture supernatants or CSF samples were mixed with 5x loading buffer (50 mM Tris-HCl (pH 7.6), 10% glycerol, 1% SDS, and 0.01% bromophenol blue) and resolved on a 11% SDS gel impregnated with 0.12 mg/ml gelatin, overlaid with a 4% stacking gel. Gels were run for approximately 3 h at 180 V, washed for 1 h in 2.5% Triton X-100, and incubated overnight in collagenase buffer (50 mM Tris-HCl (pH 7.6), 0.2 M NaCl, and 5 mM CaCl2) at 37°C. After staining with 0.2% Coomassie blue, proteolytic activity was revealed as white bands on a dark background. Gels were digitized using a UVP Transilluminator (Cambridge, U.K.), and densitometric analysis was performed using Image 1.61 analysis program (National Institutes of Health, Bethesda, MD). A linear range for quantitation of MMP-9 (6–170 ng/ml) was determined from standard curves with known quantities of rMMP-9 (32–8000 pg). Samples containing MMP-9 activity above the upper limit of this range were diluted as necessary.
TIMP-1 determination
The TIMP-1 concentration was measured in cell culture samples and CSF specimens by ELISA (32). Ninety-six-well plates were coated overnight with RRU-T5 mAb in PBS (5 µg/ml) at 4°C. After blocking with 10 mg/ml BSA in PBS for 1 h, standards and samples were loaded. After an overnight incubation at 4°C, biotinylated polyclonal B-anti CL1 Ab (25 µg/ml in 0.5 mg/ml BSA in 0.1% Tween 20 in PBS) was added for 2 h at room temperature. Binding was detected by addition of streptavidin-HRP (1/1000) using o-phenylenediamine as a substrate, and absorbency at 492 nm was measured. The lower limit of sensitivity of this assay is 5 ng/ml.
Northern analysis
RNA was extracted from 5 x 106 cells
using Tri-Reagent according to the manufacturer’s instructions. This
protocol is a modified guanidium thiocyanate-phenol-chloroform method
(33). Aliquots of RNA (12–15 µg) were run on denaturing
1% agarose-formaldehyde gels, transferred by capillary blotting to
Hybond-N, and fixed by exposure to UV light (UV Stratalinker 1800, La
Jolla, CA). Plasmids containing cDNA probes for MMP-9 (34)
and TIMP-1 were gifts from Prof. Howard Welgus (Washington University,
St. Louis, MO) and Dr. Ian Clark (University of East Anglia, Norwich,
U.K.), respectively, and were labeled with
[-32P]dCTP using a Redi-Prime II kit (random
primer method). Blots were first prehybridized, and then hybridized
overnight, washed, and finally autoradiographed for 24–48 h at
-80°C. Densitometry was performed using NIH Image 1.61. Blots were
stripped by heating for 1 h at 65°C in 0.005 M Tris-HCl (pH
8.0), 0.0002 M EDTA, and 0.1x Denhardt’s solution and then reprobed
with a -32P end-labeled, 42-mer -actin
probe (35). Measurement of expression of this housekeeping
gene and assessment of 18/28S ribosomal RNA on agarose gels were used
to confirm uniform loading of total RNA.
Western blot analysis
Cell culture supernatants and CSF samples were run on 10% SDS gels and transferred to Hybond-C. Blots were blocked (0.1% Tween 20 in PBS and 5% nonfat milk) for 1 h at room temperature and then incubated overnight with sheep anti-human pro-MMP-9 Ab (1/1000) at 4°C. After washing, blots were incubated with peroxidase-conjugated donkey anti-sheep IgG (1/1000) for 1 h at room temperature. Protein bands were visualized on Hyperfilm ECL by chemiluminescence (Amersham, Aylesbury, U.K.).
Clinical study
All patients were adults admitted to the Clinical Research Ward at the Center for Tropical Diseases (Ho Chi Minh City, Viet Nam). Control subjects were patients undergoing clinical investigation for what subsequently transpired to be noninfectious CNS disorders. The scientific and ethical committee of the Center for Tropical Diseases, Cho Quan Hospital, approved the use of clinical samples for this project.
Patients had standard hematological and biochemical investigations, CSF examination, and computed tomography brain scanning when indicated. CSF culture (for bacteria and fungi) and Gram, India ink, and Ziehl-Neelsen’s stains were routinely performed. Patients were allocated to clinical groups as follows.
Bacterial meningitis. Diagnosis was made by bacterial culture and PCR of CSF. Bacterial meningitis was also diagnosed in cases where no causative pathogen was detected (in part due to prior antibiotic treatment) if there was a typical CSF picture (polymorphonuclear predominant leukocytosis, elevated protein, and decreased glucose concentrations) and an excellent clinical response to conventional antibiotics.
Tuberculous meningitis. Diagnosis was made by positive cultures/PCR for M. tuberculosis in CSF and/or consistent CSF investigations (50–1000 leukocytes/mm3, mononuclear cell predominance, and raised protein concentration). Supportive evidence in some cases included a positive chest x-ray or recent history of TB.
Viral meningitis. Diagnosis was made by negative cultures of blood and CSF for bacteria/M. tuberculosis, consistent CSF biochemistry (raised protein and normal glucose concentration), and absence of CSF polymorphonuclear leukocytosis. Clinical criteria were complete patient recovery without antimicrobial treatment and no evidence of active TB.
Statistics
Data from cell culture experiments were log-transformed to obtain equal variances and were analyzed by one-way ANOVA, with post-hoc adjustment using the Bonferroni method to account for multiple testing. For the in vivo study, Fisher’s exact tests were used to analyze categorical data. Normally distributed, continuous outcomes were analyzed using a one-way ANOVA. Mann-Whitney U tests (with Bonferroni adjustment) were used to analyze nonnormally distributed data, including CSF MMP-9 and TIMP-1 concentrations. In vivo data are presented as box and whisker plots; the median value is shown within each box, the lower border represents the 25th percentile, and the upper border represents the 75th percentile. Whiskers show the highest and lowest values, excluding outliers (*), which are defined as greater than 1.5 box-lengths from the upper border of the box.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Infection of THP-1 cells by M. tuberculosis (MOI =
1) caused MMP-9 secretion at 24 h similar to that following LPS
exposure (Fig. 1
A). In
contrast, there was only minor induction of MMP-9 production after
phagocytosis of latex beads, which suggests that the process of
phagocytosis per se is a weak stimulus to MMP-9 production. MMP-9
secretion from unstimulated cells was barely detectable at 24 h.
Western blot analysis using specific anti-MMP-9 Abs confirmed that
zymolytic activity in cell culture supernatants was due to MMP-9 (data
not shown). A dose-dependent effect on MMP-9 secretion was observed
after stimulation with increasing MOI of M. tuberculosis
(Fig. 1B). MMP-9 secretion was stimulated with a MOI of 0.01
(1 bacillus to 100 cells), demonstrating the potency of live M.
tuberculosis.
|
, C and D). MMP-9 concentrations
secreted from cells infected with M. tuberculosis were
approximately 10-fold greater than the low level secretion from
unstimulated cells (p = 0.001 at 24 h).
MMP-9 secretion following phagocytosis of latex beads was not
significantly different from secretion from unstimulated control cells.
MMP-9 secretion continued to increase over the 48 h of study. Effect of infection with M. tuberculosis on TIMP-1 secretion by THP-1 cells
Since the balance between the local concentrations of MMP-9 and
TIMP-1 is important in determining net proteolytic activity, TIMP-1
secretion from THP-1 cells was investigated (Fig. 2). At 24 h there was no significant
difference between TIMP-1 secretion from cells stimulated with M.
tuberculosis, LPS, or latex beads and constitutive secretion from
unstimulated controls. However, there was a modest increase in TIMP-1
secretion from M. tuberculosis-infected monocytic cells at
48 h, which was approximately 5-fold greater than that from
unstimulated cells (p = 0.035). TIMP-1
concentrations at 48 h were also increased in M.
tuberculosis-infected cells compared with those stimulated by LPS
or phagocytosing latex beads, but the differences were not significant.
LPS exposure or phagocytosis of latex beads did not significantly
increase TIMP-1 secretion compared with that in controls at
48 h.
|
To investigate the mechanisms controlling MMP-9 and TIMP-1 gene
expression, mRNA accumulation was investigated. MMP-9 mRNA was not
constitutively detectable by Northern blot analysis despite the
observed very low level MMP-9 secretion in control cultures.
Twenty-four hours after infection with M. tuberculosis,
MMP-9 mRNA was detected, and this increased further at 48 h
(relative to -actin; Fig. 3A). In contrast, TIMP-1 mRNA
was constitutively expressed, and following infection with M.
tuberculosis, TIMP-1 mRNA levels marginally increased compared
with control values (Fig. 3B). However, compared with
changes in MMP-9 gene expression, the increase in TIMP-1 mRNA was
modest.
|
To investigate whether the laboratory findings were relevant to clinical disease, we next studied CSF from adult Vietnamese patients with tuberculous (23 cases), bacterial (13 cases), and viral meningitis (20 cases). There were no significant differences among groups in age and sex distribution, fever, cardio-respiratory function, conscious level, confusion, and rigors. A pathogen was positive identified in 46% of cases of bacterial meningitis (mainly Neisseria meningitidis, with one Klebsiella species isolated) and in 22% of M. tuberculosis-infected cases. The remaining cases were defined by the clinical criteria described above. The three recorded deaths occurred in the tuberculous meningitis group. The incidence of HIV infection in Vietnam is very low (<0.01–0.1% in pregnant women), and no tuberculous meningitis patient tested was sero-positive.
Despite finding very high total leukocyte counts in some bacterial
meningitis CSF specimens, overall there was no significant difference
in total CSF leukocyte counts between groups (Table I). As expected, there were significantly
more neutrophils in bacterial meningitis CSF compared with tuberculous
meningitis. The CSF total protein concentration was significantly
greater in tuberculous than viral meningitis, but other CSF parameters
(opening pressure, mononuclear cell count, glucose and lactate
concentration) were similar in all groups. There were no significant
differences among groups in peripheral blood investigations, including
total leukocyte, neutrophil and lymphocyte counts, sodium, creatinine,
albumin, and glucose concentrations. The venous blood lactate
concentration was significantly higher in tuberculous and bacterial
meningitis compared with viral meningitis (p 
0.03).
|
MMP-9 activity in CSF from patients with tuberculous, bacterial,
and viral meningitis was measured, and representative zymograms of
patient samples are shown in Fig. 4A. MMP-2 (72-kDa gelatinase
A) is detectable using this methodology and is constitutively expressed
in CSF as reported previously, but did not differ among groups
(11, 14, 15, 16, 17). Since infiltrating leukocytes are the major
source of MMP-9 in the CSF (36, 37), MMP-9 concentrations
will reflect the numbers of leukocytes entering the CNS. To take this
into account, MMP-9 concentrations were corrected for total CSF
leukocyte number (Fig. 4B). The MMP-9/CSF leukocyte ratio
was significantly higher in tuberculous meningitis (median (range),
3.19 (0.19–31.00) ng/ml/cell) than in bacterial (0.23 (0.01–18.37)
ng/ml/cell) or viral (0.20 (0.04–31.00) ng/ml/cell) meningitis
(p < 0.01). The absolute concentration of
MMP-9 was also highest in the tuberculous meningitis group (1,032
(31–17,499) ng/ml). The absolute MMP-9 concentration was significantly
higher in tuberculous than in viral meningitis (31 31–3(31–3,806) ng/ml;
p < 0.001), but the difference between tuberculous and
bacterial meningitis (196 31–26(31–26,201) ng/ml) did not reach statistical
significance in this study. As in cellular studies, the specificity of
MMP-9 activity detected by zymography was confirmed by Western blot
analysis.
|
Consistent with in vitro data, CSF TIMP-1 is constitutively
secreted into CSF from normal individuals independently of leukocyte
influx. Similarly, TIMP-1 was constitutively secreted from human
monocytes in vitro. In contrast to the increased MMP-9 activity, TIMP-1
concentrations were not raised in CSF of tuberculous meningitis
patients compared with those in bacterial meningitis and normal
controls (Fig. 5). TIMP-1 levels were
significantly greater in tuberculous than in viral meningitis
(p < 0.05).
|
We next investigated whether the elevated MMP-9 concentrations
seen in tuberculous meningitis might be clinically significant.
Patients with signs of CNS tissue injury including unconsciousness
(defined as a Glasgow coma score of <11), confusion, or focal
neurological deficit had significantly higher levels of MMP-9/CSF
leukocyte compared with other patients (Table II). TIMP-1 concentrations did not relate
to CNS injury, which is consistent with a shift in the protease
inhibitor balance to increased proteolytic activity. Furthermore,
concentrations of MMP-9/CSF leukocyte but not TIMP-1 were significantly
higher in the three patients who died compared with survivors.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In vitro, there was induction of MMP-9 gene expression 24 h following infection with M. tuberculosis and a 10-fold rise in secretion compared with controls. More importantly, increases in TIMP-1 gene expression and secretion following infection with M. tuberculosis were modest in comparison. Since net proteolysis is partly determined by local MMP-9 and TIMP-1 concentrations, the protease:anti-protease balance may be tipped in favor of tissue degradation. A similar disproportionate increase in MMP-9 relative to TIMP-1 has been reported previously following LPS stimulation of macrophages (7, 27). At 24 h in this study the secretion of MMP-9 after infection with M. tuberculosis was comparable to that produced by LPS stimulation. However, the concentration of LPS used (1 µg/ml) is equivalent to the amount derived from E. coli at an MOI of 500 (38). In comparison, MMP-9 secretion followed stimulation by M. tuberculosis with an MOI of 0.01 and was dependent on infectious load. We have further evidence for the relative specificity of this response in that respiratory syncytial virus infection of monocytes does not result in detectable MMP-9 secretion (data not shown).
Little is known about the mechanisms that activate MMP synthesis in
monocytes infected with M. tuberculosis. M.
tuberculosis-derived lipoarabinomannan induces MMP-9 secretion by
THP-1 cells, and our unpublished data showing that heat-killed M.
tuberculosis had a similar effect as live organisms suggests that
infection with viable intact bacilli is not required to stimulate MMP-9
production (29). Moreover, the minimal response to latex
beads suggests that the process of phagocytosis per se does not induce
MMP-9 secretion. In LPS-stimulated monocytic phagocytes, MMP-9 and
TIMP-1 secretion are regulated at the level of gene transcription and
by alterations in mRNA stability (39, 40). MMP-9 mRNA was
also been found to be up-regulated in THP-1 cells stimulated with
heat-killed, avirulent M. tuberculosis (29).
Our data are consistent with a central role for transcriptional
regulation of MMP-9. TNF- stimulates MMP-9 secretion by macrophages
(22), and the delayed kinetics of MMP-9 gene expression,
which becomes detectable at 24 h, suggest that autocrine and
paracrine TNF- might play a significant role in this response; this
is an area of ongoing study. However, the fact that MMP-9 secretion was
detected before gene expression may reflect the secretion of stored,
preformed enzyme or a lack of sensitivity of Northern assays.
In the in vivo study MMP-9/CSF leukocyte levels in patients with tuberculous meningitis were significantly elevated. In addition, the absolute concentration of MMP-9 was greatest in CSF from patients with tuberculous meningitis. Such data are consistent with our cellular observations and extend the findings of a recent limited study that found raised MMP-9 concentrations in a mixed patient population with CNS infections (including fungal and tuberculous meningitis) compared with noninflammatory CNS disease (41). In our study 46% of bacterial and 22% of tuberculous cases had a definitive microbiological diagnosis. The remaining cases were diagnosed using the standard clinical criteria outlined above. However, since tuberculous and bacterial meningitis are invariably fatal if untreated, the excellent clinical response patients made to specific antimicrobial therapy suggests that these were diagnosed accurately. The significant neutrophil predominance in bacterial meningitis CSF samples further supports this. In contrast, spontaneous recovery without treatment was observed in all viral meningitis cases and is strong evidence that these diagnoses were also correct.
Since infiltrating leukocytes, rather than resident cells, are the major sources of MMP-9 in meningitis (36, 37), the MMP-9 concentration will depend upon both the number and the subsets of leukocytes recruited to the CNS. A significantly greater proportion of neutrophils were found in bacterial meningitis CSF samples than in either tuberculous or viral meningitis, suggesting that infiltrating mononuclear cells in tuberculous meningitis secrete more MMP-9 than neutrophils in bacterial meningitis. Neutrophils are shorter lived than monocytes and release proteases from cytoplasmic granules formed during cellular development, rather than synthesizing MMP-9 de novo (42). In contrast to changes in MMP-9, MMP-2 secretion was constitutive and unaffected by infection, which is consistent with previous reports (10, 11, 14, 15, 16, 17). In addition and similar to findings after M. tuberculosis infection in vitro, there was no significant increase in CSF TIMP-1 concentrations in patients with tuberculous meningitis compared with those with bacterial meningitis and controls. A possible consequence of this is that unrestricted MMP-9 activity in tuberculous meningitis may result in local tissue destruction; it would therefore be expected that this is associated with signs of neurological damage.
The striking observation in the in vivo study was that MMP-9/CSF
leukocyte, but not TIMP-1, concentrations are raised in tuberculous
meningitis patients with neurologic complications (unconsciousness,
confusion, neurological deficit). Furthermore, MMP-9/CSF leukocyte
concentrations were significantly elevated in the three patients who
died, all of whom had tuberculous meningitis, compared with those in
the survivors. In contrast, there was no difference in either MMP-9/CSF
leukocyte or TIMP-1 concentrations in patients with or without systemic
features of infection, except for systolic hypotension, which is partly
regulated in the CNS. TNF- is present in high concentrations both in
the CNS and peripherally and may be important in the development of
hypotension. TNF- may drive MMP-9 secretion in vivo, and a close
association was found between CSF concentrations of MMP-9 and TNF-
in a rodent model of bacterial meningitis (24). In
addition, MMPs may regulate TNF- activity, since MMP inhibition
reduced both TNF- levels and cerebral injury.
Taken together, these data suggest that unrestricted MMP-9 activity within the CNS causes tissue destruction in tuberculous meningitis. Tuberculous and, to a lesser extent, bacterial meningitis destroy cerebral tissue and are fatal if untreated. In contrast, viral meningitis is characteristically a nondestructive condition from which most patients recover without neurological deficit. Although monocyte-predominant infiltration occurs in viral meningitis, MMP-9 secretion may be rapidly switched off once cells have entered the CNS. Differential down-regulation of MMP-9 may partly explain the varying extent of tissue destruction produced by M. tuberculosis vs infection with other pathogens. For example, IL-10 completely blocks LPS, but not M. tuberculosis-induced MMP-9, secretion (data not shown). Furthermore, since tuberculous meningitis is a meningo-encephalitis, the wide distribution of leukocytes throughout the brain parenchyma might be potentially more harmful than in bacterial meningitis, where the inflammatory infiltrate is predominantly confined to the meninges (43). Although MMP-9 is the quantitatively most important MMP secreted from monocytes, the principal host defense to M. tuberculosis, our study does not exclude some role for other monocyte/macrophage-derived MMPs previously implicated in pathophysiology (24, 36). We were constrained from investigating these due to restricted amounts of CSF available for analysis.
In summary, we have shown that infection with M. tuberculosis stimulates gene expression and secretion of MMP-9 from human monocytic cells in a manner dependent on infectious load. TIMP-1 secretion was stimulated to a much lesser degree, which suggests that these M. tuberculosis-infected cells shift to a matrix-degrading phenotype in vitro. A similar phenotype was observed in the in vitro study, which revealed that MMP-9 activity, but not TIMP-1 concentrations, were significantly elevated in tuberculous meningitis. MMP-9 activity was also specifically increased in patients with neurologic complications and in fatal cases, but was not related to systemic signs of infection. Leukocyte-derived MMP-9 secretion may therefore be an important contributor to localized cerebral injury in tuberculous meningitis. Investigation of therapeutic strategies in non-rodent animal models of tuberculous meningitis now needs to be tested, since inhibition of MMP activity may be effective as an adjunctive treatment for tuberculous meningitis in the future.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Jon S. Friedland, Department of Infectious Diseases, Imperial College of Science, Technology and Medicine, Hammersmith Hospital, Du Cane Road, London, United Kingdom W12 ONN.
3 Abbreviations used in this paper: TB, tuberculosis; MMPs, matrix metalloproteinases; TIMPs, tissue inhibitors of matrix metalloproteinases; MOI, multiplicity of infection; CSF, cerebrospinal fluid.
Received for publication June 7, 2000. Accepted for publication January 4, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-1 selectively induce expression of 92-kDa gelatinase by human macrophages. J. Immunol. 157:4159.[Abstract]
from cells. Nature 385:729.[Medline]
, LPS, and IL-1. Science 243:1467.This article has been cited by other articles:
|
|
 |
|
  P. T. Elkington, J. A. Green, J. E. Emerson, L. D. Lopez-Pascua, J. J. Boyle, C. M. O'Kane, and J. S. Friedland Synergistic Up-Regulation of Epithelial Cell Matrix Metalloproteinase-9 Secretion in Tuberculosis Am. J. Respir. Cell Mol. Biol., October 1, 2007; 37(4): 431 - 437. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Harris, J. A. Green, P. T. Elkington, and J. S. Friedland Monocytes infected with Mycobacterium tuberculosis regulate MAP kinase-dependent astrocyte MMP-9 secretion J. Leukoc. Biol., February 1, 2007; 81(2): 548 - 556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Harris, M. Fernandez-Vilaseca, P. T. G. Elkington, D. E. Horncastle, M. B. Graeber, and J. S. Friedland IFN{gamma} synergizes with IL-1{beta} to up-regulate MMP-9 secretion in a cellular model of central nervous system tuberculosis FASEB J, February 1, 2007; 21(2): 356 - 365. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Harris, R. K. Nuttall, P. T. Elkington, J. A. Green, D. E. Horncastle, M. B. Graeber, D. R. Edwards, and J. S. Friedland Monocyte-Astrocyte Networks Regulate Matrix Metalloproteinase Gene Expression and Secretion in Central Nervous System Tuberculosis In Vitro and In Vivo J. Immunol., January 15, 2007; 178(2): 1199 - 1207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Taylor, J. M. Hattle, S. A. Dreitz, J. M. Troudt, L. S. Izzo, R. J. Basaraba, I. M. Orme, L. M. Matrisian, and A. A. Izzo Role for Matrix Metalloproteinase 9 in Granuloma Formation during Pulmonary Mycobacterium tuberculosis Infection Infect. Immun., November 1, 2006; 74(11): 6135 - 6144. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P T G Elkington and J S Friedland Matrix metalloproteinases in destructive pulmonary pathology Thorax, March 1, 2006; 61(3): 259 - 266. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. P. Simmons, G. E. Thwaites, N. T. H. Quyen, E. Torok, D. M. Hoang, T. T. H. Chau, P. P. Mai, N. T. N. Lan, N. H. Dung, H. T. Quy, et al. Pretreatment Intracerebral and Peripheral Blood Immune Responses in Vietnamese Adults with Tuberculous Meningitis: Diagnostic Value and Relationship to Disease Severity and Outcome J. Immunol., February 1, 2006; 176(3): 2007 - 2014. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. T. G. Elkington, R. K. Nuttall, J. J. Boyle, C. M. O'Kane, D. E. Horncastle, D. R. Edwards, and J. S. Friedland Mycobacterium tuberculosis, but Not Vaccine BCG, Specifically Upregulates Matrix Metalloproteinase-1 Am. J. Respir. Crit. Care Med., December 15, 2005; 172(12): 1596 - 1604. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Elass, L. Aubry, M. Masson, A. Denys, Y. Guerardel, E. Maes, D. Legrand, J. Mazurier, and L. Kremer Mycobacterial Lipomannan Induces Matrix Metalloproteinase-9 Expression in Human Macrophagic Cells through a Toll-Like Receptor 1 (TLR1)/TLR2- and CD14-Dependent Mechanism Infect. Immun., October 1, 2005; 73(10): 7064 - 7068. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Liu, T. Desta, H. He, and D. T. Graves Diabetes Alters the Response to Bacteria by Enhancing Fibroblast Apoptosis Endocrinology, June 1, 2004; 145(6): 2997 - 3003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Wright and J. S. Friedland Regulation of monocyte chemokine and MMP-9 secretion by proinflammatory cytokines in tuberculous osteomyelitis J. Leukoc. Biol., June 1, 2004; 75(6): 1086 - 1092. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. M. Price, R. H. Gilman, J. Uddin, S. Recavarren, and J. S. Friedland Unopposed Matrix Metalloproteinase-9 Expression in Human Tuberculous Granuloma and the Role of TNF-{alpha}-Dependent Monocyte Networks J. Immunol., November 15, 2003; 171(10): 5579 - 5586. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Selman, J. Cisneros-Lira, M. Gaxiola, R. Ramirez, E. M. Kudlacz, P. G. Mitchell, and A. Pardo Matrix Metalloproteinases Inhibition Attenuates Tobacco Smoke-Induced Emphysema in Guinea Pigs Chest, May 1, 2003; 123(5): 1633 - 1641. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||
