MMPs - RERO DOC

Brain (2001), 124, 1743–1753

The expression profile of matrix metalloproteinases (MMPs) and their inhibitors (TIMPs) in lesions and normal appearing white matter of multiple sclerosis Raija L. P. Lindberg,1 Corline J. A. De Groot,3 Lisette Montagne,3 Peter Freitag,2 Paul van der Valk,3 Ludwig Kappos2 and David Leppert1,2 Departments of 1Research and 2Neurology, University Hospitals Basel, Switzerland and the 3Department of Pathology, Division of Neuropathology, University Hospital Vrije Universiteit, MS Centre for Research and Care, The Netherlands

Correspondence to: Dr R. L. P. Lindberg, Departments of Research and Neurology, University Hospitals, Pharmazentrum, Klingelbergstrasse 50, CH-4056 Basel, Switzerland E-mail: [email protected]

Summary In multiple sclerosis, matrix metalloproteinases (MMPs) are effectors of crucial pathogenetic steps, such as blood– brain barrier breakdown, invasion of brain parenchyma by immune cells and demyelination. However, only limited data are available on the types of MMPs induced in the course of multiple sclerosis, and on the role of their endogenous antagonists, the tissue inhibitors of metalloproteinases (TIMPs). We quantified the transcriptional expression of six MMPs and the four TIMPs in lesions and in normal appearing white matter (NAWM) from post-mortem multiple sclerosis brain tissue by real-time polymerase chain reaction, and compared levels with those in brain tissue from six control patients without neurological disease. The mRNA expression of MMP-7 and -9, but not of other metalloproteinases [MMP-2 and -3, and tumour necrosis factor (TNF)-α-converting-

enzyme] was equally upregulated throughout all stages of lesion formation with active inflammation, and in most of matched NAWM tissue. The transcription of cytokines TNF-α/β and IL (interleukin)-2, known modulators of MMPs, was upregulated only in distinct stages of lesion formation, while their receptors were not induced at all, which suggests that additional signalling molecules participate in the sustained upregulation of MMP-7 and -9 in multiple sclerosis. None of the TIMPs showed a significant induction over baseline expression of controls. We hypothesize that an imbalance between MMP and TIMP expression may cause a persistent proteolytic overactivity in multiple sclerosis, that may be a factor for continuous tissue destruction, and hence for secondary disease progression.

Keywords: brain tissue; matrix metalloproteinases; multiple sclerosis; real-time polymerase chain reaction; tissue inhibitors of metalloproteinases Abbreviations: GAPDH ⫽ glyceraldehyde phosphate dehydrogenase; IL ⫽ interleukin; MMP ⫽ matrix metalloproteinase; NAWM ⫽ normal appearing white matter; PP ⫽ primary progressive; RR ⫽ relapsing–remitting; RT–PCR ⫽ reverse transcriptase–polymerase chain reaction; SP ⫽ secondary progressive; TACE ⫽ TNF-α-converting enzyme; TIMP ⫽ tissue inhibitor of metalloproteinase; TNF ⫽ tumour necrosis factor

Introduction Multiple sclerosis is thought to be caused by an autoimmune response directed against one or several myelin components of the CNS. The histopathological hallmark of multiple sclerosis is the plaque, a well-demarcated white matter lesion characterized by demyelination and axonal loss. Focal blood– brain barrier leakage, followed by extravasation of immune cells into the brain parenchyma were believed to be the earliest © Oxford University Press 2001

steps in the pathogenesis of multiple sclerosis (Kermode et al., 1990; Kwon and Prineas, 1994; Hartung, 1997). However, lesion formation may not evolve stereotypically, but via distinct types of inflammatory and degenerative events (Lucchinetti et al., 2000). Moreover, longitudinal studies on early plaque formation using newer MRI techniques have demonstrated morphological changes in the so-called normal

1744

R. L. P. Lindberg et al.

appearing white matter (NAWM) that occur independently of later plaque formation (Catalaa et al., 2000; Werring et al., 2000). Matrix metalloproteinases (MMPs) are a family of at least 23 endopeptidases that act as effectors of extracellular matrix remodelling in physiological and pathological conditions (Woessner, 1998). MMPs can be subdivided into gelatinases (MMP-2 and -9), collagenases (MMP-1, -8, -13 and -18), stromelysins (MMP-3, -10 and -11) and other MMPs, according to their substrate affinity profile. Their activity is closely regulated by tissue inhibitors of metalloproteinases (TIMPs), a group of four endogenous antagonists that bind to the catalytic site of MMPs (Yong et al., 1998). There is accumulating evidence that MMPs play a key role in the pathogenesis of many neuroinflammatory diseases (Kieseier et al., 1999). In multiple sclerosis, elevated protein levels of MMP-9 are detectable in CSF (Gijbels et al., 1992; Leppert et al., 1998) and serum (Lee et al., 1999; Waubant et al., 1999). Furthermore, the appearance of new gadoliniumenhancing lesions is more likely to occur after an increase of serum MMP-9 (Waubant et al., 1999). In vitro experiments and results from animal models of multiple sclerosis indicate that metalloproteinases are effectors of blood–brain barrier disruption (Rosenberg et al., 1995), extravasation of immune cells into the brain parenchyma (Gijbels et al., 1994; Leppert et al., 1996; Xia et al., 1996), enhancement of tumour necrosis factor (TNF)-α release (Gearing et al., 1994; Moss et al., 1997) and degradation of myelin proteins (Proost et al., 1993; Chandler et al., 1995). Furthermore, enzyme inhibitors of MMPs have been shown to ameliorate the clinical course and reduce inflammatory cell infiltration in experimental autoimmune encephalomyelitis (Gijbels et al., 1994; Hewson et al., 1995; Clements et al., 1997). Histological studies have demonstrated elevated protein expression of several MMPs in macrophages, microglial cells and astrocytes within multiple sclerosis lesions (Cuzner et al., 1996; Maeda and Sobel, 1996; Cossins et al., 1997). However, the array of MMPs and TIMPs expressed in the process of plaque formation has not been elucidated and their expression levels in NAWM are not known. In this report, we quantified the transcriptional expression of six MMPs (MMP-2, -3, -7, -8, -9 and -12), four TIMPs (TIMP-1, -2, -3 and -4) and the metalloproteinase TNF-αconverting-enzyme (TACE) in different stages of multiple sclerosis lesion formation, matched NAWM, and CSF cells from multiple sclerosis patients by real-time polymerase chain reaction (PCR). We compared their expression levels with that of cytokines TNF-α, TNF-β and interleukin (IL)-2, known inducers of MMP expression (Leppert et al., 1995, 1996; Kubota et al., 1996; Johnatty et al., 1997), and their respective receptors.

Methods Human brain tissue samples Human brain tissue was obtained at autopsy (with short postmortem intervals; see detailed autopsy data in Table 1) from

seven patients with secondary progressive (SP) multiple sclerosis (diagnosed according to the Poser criteria; Poser et al., 1983), and six controls (i.e. patients who died without neurological disease). The clinical diagnosis of multiple sclerosis was confirmed neuropathologically. The postmortem samples used in this study were obtained through the rapid autopsy system of the Netherlands Brain Bank (coordinator Dr R. Ravid), which has been approved by the Ethics Committee of the University Hospital Vrije Universiteit in Amsterdam. All patients and their next-of-kin had given written consent for autopsy, and use of their brain for scientific research. Brain tissue samples from non-neurological control cases were taken from the subcortical white matter or corpus callosum. Tissue samples were snap-frozen in liquid nitrogen and stored at –196°C. From each multiple sclerosis patient, one brain sample was obtained (which contained a lesion that was representative for all the lesions found in a particular patient), except from multiple sclerosis cases 98-009 and 99-025 where two independent samples were used for analysis. In addition, macroscopically, NAWM was collected from each multiple sclerosis patient.

Neuropathological evaluation Tissue samples derived from multiple sclerosis lesions were stained with anti-myelin basic protein (IgG; Boehringer Mannheim, Germany) and with the neutral lipid marker Oil Red O to delineate areas of myelin breakdown and demyelination, with mouse anti-human KP1 (CD68, IgG1; Dako, Copenhagen, Denmark) and mouse anti-human leucocyte common antigen (CD45, IgG1; Dako) to detect leucocyte infiltration and microglial activation, with MHC class II antigens (HLA-DR; Biotest AG, Dreieich, Germany) to detect activated lymphocytes and glial cells and with anticow glial fibrillary acidic protein (Dako) to determine the extent of astrogliosis. Detection of MMP-9 was performed with the antibody 4H3 (IgG1; R&D Systems, Abingdon, UK). For immunohistochemistry, serial 5 µm thick brain tissue sections were stained by the ABC method (avidin– biotin–horseradish peroxidase complex; Vector Laboratories, Burlingame, Calif., USA) using 3,3⬘-diaminobenzidine tetrachloride as a chromogen (Boven et al., 2000). Staging of multiple sclerosis lesions was performed according to the classification described (Table 2) (Van Waesberghe et al., 1999; van der Valk and De Groot, 2000). In normal control white matter (corpus callosum or subcortical white matter), no activity of inflammatory cells was detected.

CSF samples Twenty-two patients (12 women, 10 men) with clinically definite or laboratory-supported definite multiple sclerosis (Poser et al., 1983) were included. In 17 patients, lumbar puncture was performed in clinically stable phases: four patients had primary progressive (PP) multiple sclerosis, six patients had relapsing–remitting (RR) multiple sclerosis and

MMP and TIMP expression in MS tissue

1745

Table 1 Details of control and multiple sclerosis autopsy tissue Lesional activity

Patient case no.

Age (years)/gender

Disease duration (years)

Post-mortem delay (h:min)

Colour code (Figs 1, 3, 4)

Cause of death

Reactive

98-009 96-026 96-121 97-160 97-123 96-104

70/F 69/F 53/F 40/F 46/M 72/M

27 19 19 11 23 22

6:25 9:15 7:15 7:00 3:35 4:45

Red Green Blue Orange Black Olive

98-009

70/F

27

6:25

Red

Urosepsis Respiratory failure Pneumonia Aspiration pneumonia Pneumonia Bladder/intestine carcinoma Urosepsis

99-025 98-125 98-127 94-125 95-007

64/F 58/F 56/M 51/M 54/M

25 – – – –

7:45 6:15 5:25 6:00 9:15

Magenta Black Black Black Black

94-113 94-119

82/F 51/F

– –

6:30 7:40

Black Black

Active Chronic active

Chronic inactive Controls

Pneumonia Septicaemia, pneumonia Myocardial infarction Liposarcoma, ileus Rupture of right common carotid artery Respiratory failure Sepsis

Table 2 Classification of multiple sclerosis lesions Type of lesion

Description

NAWM

Low expression of CD68 (KP1), CD45 (LCA) and HLA-DR (MHC class II antigens) on resident microglial cells. No inflammation, no loss of myelin.

Reactive

Strong expression of CD68, CD45 and HLA-DR on resident microglial cells and on clusters of microglial cells present in proximity to microvessels. Peri- and intravascular accumulation of CD45- and HLA-DR-positive lymphocytes in microvessels. A variable degree of oedema is detectable, but there is no apparent loss of myelin.

Active

Demyelinating lesion with perivascular and parenchymal macrophages that have a strong CD68, CD45 and HLA-DR expression. These phagocytes contain MBP-positive myelin degradation products or ORO/PAS-positive (neutral) lipids. A variable number of strongly CD45-expressing lymphocytes are localized pre-dominantly within the perivascular spaces of blood vessels. GFAP-positive reactive astrocytes with long processes are distributed throughout the demyelinating regions.

Chronic active

The hypocellular, demyelinated centre of the lesion contains a small number of CD68-positive macrophages with some residual ORO/PAS-positive (neutral) lipids. Few CD45-positive lymphocytes are present in the perivascular cuffs. The hypercellular rim contains perivascular and parenchymal CD68-positive macrophages with either MBPpositive myelin degradation products or ORO/PAS-positive (neutral) lipids. GFAP-positive reactive astrocytes are localized predominantly at the edge of the lesion centre and within the hypercellular rim.

Chronic inactive

The hypocellular demyelinated lesion contains largely (isomorphic) gliosis with widened extracellular spaces. A very small number of CD68-positive macrophages and CD45-positive lymphocytes can be detected in blood vessel walls.

MBP ⫽ myelin basic protein; ORO/PAS ⫽ Oil red O/periodic acid–Schiff; GFAP glial fibrillary acidic protein.

seven had SP multiple sclerosis. Five patients were lumbar punctured during their first attack and were later diagnosed as having RR disease. None of the patients had received corticosteroids or other immunosuppressive agents within 6 weeks prior to the lumbar puncture. Five CSF samples from patients with hydrocephalus (two), lumbago (one) and psychiatric diseases (two) served as controls.

Gene expression analysis Frozen brain tissue sections of multiple sclerosis lesions and control cases (10 µm thick) were collected in 1.5 ml Eppendorf tubes, suspended in 600 µl of lysis buffer

(QIAGEN AG, Basel, Switzerland) and stored at –70°C until RNA isolation. After routine analyses, CSF samples were centrifuged and cells were suspended in lysis buffer and stored at –70°C until RNA isolation. RNA was isolated according to the manufacturer’s instructions (RNeasy; QIAGEN). Total RNA (5–10 ng/µl final concentration) was first incubated with 0.5 µg of oligo(dT) at 70°C for 2 min and then reverse-transcribed at 37°C for 1 h in reaction mixture containing a final concentration of 1⫻ first strand buffer (Promega Corporation, Madison, Wis., USA), 500 µM of each deoxynucleotide triphosphate, 1 U/µl of Moloney murine leukaemia virus reverse transcriptase (Promega) and 1 U/µl

1746


of ribonuclease inhibitor (RNAsin, Promega). cDNA was used as a template for the real-time PCR analysis based on the 5⬘-nuclease assay (Gibson et al., 1996) with the ABI PRISM® 7700 Sequence Detection System (Applied Biosystems, PE Europe B.V., Rotkreuz, Switzerland). PCR primers and TaqMan probes were designed by using PrimerExpress® software (Applied Biosystems). Expression of MMP-2, -3, -7, -8, -9, -12, TIMP-1, -2, -3, -4, TNF-α, TNFβ, TNF receptor 1, TNF receptor 2, TACE, IL-2, IL-2 receptor α, IL-2 receptor β and IL-2 receptor γ was analysed. Glyceraldehyde phosphate dehydrogenase (GAPDH) was amplified from all samples on each plate as a housekeeping gene to normalize expression between different samples and to monitor assay reproducibility. Relative quantification of all targets was calculated by the comparative cycle threshold method outlined in user bulletin No. 2 provided by Applied Biosystems. A non-template control was included for each target analysed. The investigators (R.L.P.L. and D.L.) involved in measurements of gene expression and subsequent data analysis were blinded to the origin (multiple sclerosis versus control) of CSF cells and brain tissue, and the types of multiple sclerosis.

Statistical analysis Data were analysed using non-parametric statistical tests. The Spearman rank correlation of levels of gene expression in plaques and NAWM was calculated. Expression levels of genes in different types of tissue, or different stages of disease were compared using the Mann–Whitney U-test. P-values ⬍0.05 were considered significant.

Results Brain tissue Transcriptional expression of MMPs, TIMPs and TACE In control brain samples, MMP-9 was constitutively expressed over a wide range, while MMP-7 was not detectable. In multiple sclerosis plaques, MMP-9 and MMP-7 were upregulated compared with controls (PMMP-9 ⫽ 0.0032, PMMP-7 ⫽ 0.0029) (Fig. 1A and B), and the individual levels of the two MMPs were closely correlated (ρ ⫽ 0.857, P ⫽ 0.023); one sample of a chronic inactive lesion was not included (see below). However, the two chronic inactive lesions, derived from the same patient (case no. 99-025), showed significantly lower expression levels for MMP-9 than lesions with active inflammation (reactive, active, chronic active lesions) (P ⫽ 0.039), and were similar to those in controls (P ⫽ 0.180). Unexpectedly, the analysis for MMP-7 in chronic inactive lesions yielded divergent results, as transcripts for this gene were absent in one and present in high amounts in the other (Fig. 1B). An immunohistochemical re-evaluation revealed a rim of intense MHC class II (HLA-DR)immunoreactive microglial cells (without uptake of myelin)

Fig. 1 mRNA expression (normalized with GAPDH) of MMP-9 (A), -7 (B), -2 (C) and -3 (D) was measured by RT–PCR in different lesion types (closed circles) and adjacent NAWM (open circles) from multiple sclerosis patients (for colour code, see Table 1). Control patients without neurological diseases are marked with crosses. C ⫽ control patient; re ⫽ reactive, ac ⫽ active, ca ⫽chronic active, ci ⫽ chronic inactive lesions of the multiple sclerosis patients. ND ⫽ not detected.

surrounding the edge of the chronic inactive lesion in the sample with high MMP-7 expression (Fig. 2A). Therefore, this sample could not be defined entirely as chronic inactive and was not included in statistical evaluations. The expression of MMP-9 in NAWM from multiple sclerosis was closely correlated with that within matched lesions (ρ ⫽ 0.867, P ⫽ 0.015) and was also increased in the majority of samples compared with controls (P ⫽ 0.022). Immunohistochemical staining of NAWM from multiple sclerosis-affected brain tissue revealed a strong immunoreactivity for MMP-9 in blood vessel walls (Fig. 2B). We have investigated the immunohistochemical expression of MMP-9 in one control brain sample (no. 94-113), two brain samples containing NAWM (case nos 98-009 and 96-026) and in two samples containing an active demyelinating lesion (nos 96-121 and 97-160). In both the NAWM and the active lesions, the blood vessels were immunoreactive for MMP-9. No significant MMP-9 immunoreactivity was found in normal white matter of control cases (not shown). In contrast, MMP-7 was detectable only in 57% (four out of seven) of NAWM samples, but there was also a strong correlation between levels in lesions and matched NAWM (ρ ⫽ 0.804, P ⫽ 0.035). As a whole, expression levels of MMP-2 and -3 (Fig. 1C and D) in plaques, and in NAWM from multiple sclerosis were similar to those in controls (statistical data not shown). However, reactive plaques scored significantly higher than controls (PMMP-2 ⫽ 0.044, PMMP-3 ⫽ 0.046) and other lesion types (PMMP-2 ⫽ 0.040, PMMP-3 ⫽ 0.041) for these two MMPs.


1747

Fig. 2 (A) Immunohistochemical staining of an inactive multiple sclerosis lesion (case no. 99-025) with a monoclonal antibody directed against MHC class II antigen (HLA-DR, IgG1) showing numerous strongly immunoreactive resident microglial cells at the edge of the lesion (right side of panel). Within the lesion, only a few HLA-DR-immunoreactive leucocytes can be observed (left side of panel). (B) Immunohistochemical staining of the NAWM with a monoclonal antibody directed against MMP-9 (clone 4H3, IgG1) showing a strong immunoreactivity in a blood vessel wall. Magnification: A, ⫻90; B, ⫻180.

The results of the quantification of gene expression of the other targets are depicted in Table 3. Thirty-eight per cent (six out of 16) of multiple sclerosis specimens (lesions and NAWM), but no controls, contained low amounts of MMP-8 mRNA, which were at the detection limit of the RT–PCR

for most samples. Conversely, MMP-12 was constitutively expressed in controls, and its expression was decreased in multiple sclerosis (Plesions ⫽ 0.025, PNAWM ⫽ 0.038). All four types of TIMPs were constitutively expressed in controls. Although their range of expression was wider in plaques and

1748


Table 3 Expression of MMP-8 and -12, TIMPs and TACE in multiple sclerosis brain tissue Gene

Control

Multiple sclerosis NAWM

Lesion

MMP-8 P-value Median (range)

0/6 – – –

4/7 0.036 0.000002 (0–0.000003)

2/9 0.232 0 (0–0.001)

MMP-12 P-value Median (range)

6/6 – 0.0015 (0.0003–0.0034)

7/7 0.035 0.00031 (0.00001–0.0026)

7/9 0.025 0.00020 (0–0.0025)

TIMP-1 P-value Median (range)

6/6 – 0.0015 (0.0004–0.0019)

7/7 9/9 0.100 0.262 0.0028 0.0032 (0.0006–0.0340) (0.0007–0.0252)


6/6 – 0.343 (0.264–0.781)

7/7 0.668 0.511 (0.157–1.467)

9/9 0.289 0.538 (0.279–1.505)


6/6 – 0.047 (0.029–0.057)

7/7 0.775 0.039 (0.018–0.122)

9/9 0.087 0.103 (0.014–0.366)


6/6 – 0.004 (0.002–0.007)

7/7 0.884 0.004 (0.002–0.008)

9/9 0.807 0.004 (0.002–0.013)

TACE P-value Median (range)

6/6 – 0.007 (0.003–0.020)

7/7 0.772 0.008 (0.003–0.013)

9/9 0.905 0.004 (0.002–0.017)

NAWM from multiple sclerosis, none of the TIMPs were significantly overexpressed compared with controls (Table 3). Moreover, the comparison of TIMP levels with those of MMP-7 and -9 in individual tissue specimens failed to demonstrate any significant correlation. Lastly, TIMP levels were not different between lesions and NAWM from multiple sclerosis, and did not relate to a specific type of plaque (statistical data not shown). Similarly, the expression of TACE in multiple sclerosis was not different from controls and did not show a correlation with other gene targets or types of multiple sclerosis tissue (statistical data not shown).

Transcriptional expression of TNF-α/β, IL-2 and their receptors The expression levels of TNF-α (P ⫽ 0.044) and TNF-β (P ⫽ 0.046) (Fig. 3A and B), and of their corresponding receptors TNFR1 and TNFR2 (PTNFR1, TNFR2 ⫽ 0.046) (Fig. 3C and D) were increased only in reactive multiple sclerosis lesions, but were similar in other lesion types, and in all NAWM specimens from multiple sclerosis compared with controls (statistical data not shown).

In active lesions, where T-cell invasion is most prominent, and adjacent NAWM from the same patient, the cytokine IL-2 was transcriptionally upregulated. This was also the case in some other NAWM samples adjacent to other lesion types with active inflammation (Fig. 4A). The presence of lymphocytes in all these tissue samples was confirmed by positive staining for CD45 (not shown). Similarly to the pattern seen for TNF receptors (Fig. 3C and D), all three subunits of the IL-2 receptor were upregulated in a reactive and a chronic active plaque from the same patient (Fig. 4B– D), whereas the majority of tissue from multiple sclerosis patients revealed expression levels within or below the range of controls.

MMP and cytokine gene expression in CSF cells In order to delineate the contribution of blood-derived immune cells to the upregulation of MMP-7 and -9, and to confirm the lack of induction of other metalloproteinases in multiple sclerosis brain tissue, we measured the transcriptional expression of MMP-2, -3, -7 and -9 and TACE in CSF cells from lumbar punctures performed during relapses and clinically stable phases of RR/SP multiple sclerosis, and from PP multiple sclerosis. Figure 5A shows that MMP-9 transcripts were not detectable in control CSF (nil out of 5), whereas all samples from RR/SP multiple sclerosis (17 out of 17) were positive with similar levels during relapses versus clinically stable phases of the disease (medianstable ⫽ 0.003, medianrelapse ⫽ 0.002; P ⫽ 0.999). Conversely, only one out of four CSF samples from PP multiple sclerosis expressed small amounts of MMP-9 mRNA. Transcripts for MMP-7 were not detectable in either multiple sclerosis or control samples (not shown). Genes for MMP-2 (Fig. 5B), MMP-3 and TACE (not shown) were constitutively expressed in controls and were not upregulated in multiple sclerosis (statistical data not shown).

Discussion The definition of the expression profile of MMPs and TIMPs in the course of lesion formation, and their spatial distribution in brain tissue are prerequisites for the understanding of their role in disease pathogenesis, which may have therapeutic implications. Earlier histological studies (Cuzner et al., 1996; Anthony et al., 1997; Cossins et al., 1997) demonstrated an upregulation of MMP-2, -3, -7 and -9 expression in multiple sclerosis lesions. However, these results were not quantitative and were restricted mostly to acute lesions, and the involvement of NAWM has not been evaluated. Moreover, the patterns of MMP expression in animal models of multiple sclerosis, as measured by immunohistochemistry and semiquantitative PCR, were only partly congruent with those in human disease, but revealed the involvement of additional members of the MMP family (Anthony et al., 1998; Kieseier


1749

Fig. 3 mRNA expression (normalized with GAPDH) of TNF-α (A) and TNF-β (B), and of TNF receptor 1 (C) and TNF receptor 2 (D) in multiple sclerosis brain tissue. The same colour codes for individual patients as in Fig. 1 are applied. Control patients without neurological diseases are marked with crosses. C ⫽ control patient; re ⫽ reactive, ac ⫽ active, ca ⫽chronic active, ci ⫽ chronic inactive lesions of the multiple sclerosis patients.

Fig. 4 mRNA expression (normalized with GAPDH) of IL-2 (A) and IL-2 receptor subunits α (B), β (C) and γ (D) in multiple sclerosis brain tissue. The same colour codes for individual patients as in Fig. 1 are applied. C ⫽ control patient; re ⫽ reactive, ac ⫽ active, ca ⫽ chronic active, ci ⫽ chronic inactive lesions of the multiple sclerosis patients. ND ⫽ not detected.

et al., 1998; Pagenstecher et al., 1998). Finally, no data on the expression profile of TIMPs in multiple sclerosis tissue were available.

Here, we show that MMPs are regulated specifically in multiple sclerosis brain tissue: MMP-7 and -9 are strongly induced, while other MMPs showed either minimal induction,

1750


Fig. 5 mRNA expression (normalized with GAPDH) of MMP-9 (A) and MMP-2 (B) in CSF samples from PP multiple sclerosis (MS) and RR/SP multiple sclerosis patients. Closed circles ⫽ relapsing– remitting (RR); open circles ⫽ secondary progressive (SP); ND ⫽ not detected.

restricted to ‘early’ plaques (MMP-2 and -3), or were downregulated (MMP-12) compared with controls. Moreover, the transcriptional overexpression of MMP-7 and -9 is not confined to lesions, but extends into the NAWM. Immunohistochemical studies confirmed that this led to upregulated expression of the corresponding proteins in microglial cells (Cossins et al., 1997) and endothelial cells (present results). Modern MRI techniques have demonstrated diffuse structural changes indicative of myelin damage and neuronal loss in NAWM of multiple sclerosis that do not represent precursor stages of later lesion formation, but occur independently of the latter (Catalaa et al., 2000; Werring et al., 2000). Our present results support the concept of multiple sclerosis as a primarily generalized, instead of a focal, disease process at the molecular level. TIMPs are believed to regulate the enzymatic activity of MMPs in order to protect them from overactive proteolysis, and hence uncontrolled tissue destruction. It has been hypothesized that MMP-9 upregulation predominates in early lesion development, whereas in later phases, where repair processes take over, MMP-2 and TIMP-2 participate in reparative extracellular matrix remodelling (Bever and Rosenberg, 1999). However, current results do not confirm this scenario, but rather support a sustained overexpression of MMP-7 and -9 and a lack of functional counter-regulation of TIMPs throughout the course of disease. This is different from other CNS inflammatory diseases such as bacterial meningitis, in which upregulation of MMP-9 is closely related to increased TIMP-1 levels in individual patients (Leppert et al., 2000). Earlier studies in CSF and serum of multiple sclerosis patients demonstrated an upregulation of MMP-9 protein expression, whereas levels of TIMP-1 and -2 were not increased (Leppert et al., 1998; Lee et al., 1999). In serum of patients with active multiple sclerosis, an identical pattern was found for the transcriptional regulation of MMP-9 and TIMP-1 (Lichtinghagen et al., 1999). Functionally, the transient increase in the MMP-9/TIMP-1 protein ratio in longitudinal serum measurements was predictive for the

occurrence of new gadolinium-enhancing lesions in MRI, defining for the first time a prospective disease marker for the course of multiple sclerosis (Waubant et al., 1999). We propose that dyscoordinate regulation of MMPs and TIMPs is a specific feature in multiple sclerosis that occurs uniformly in the brain, and other tissue compartments. The present findings allow us only to surmise the cellular source of MMPs in multiple sclerosis brain tissue. The upregulation of MMP-7 and -9 in reactive lesions suggests that resident brain cells are the major source, as bloodderived immune cells are scarce at this stage of lesion formation (Cuzner et al., 1996; Maeda and Sobel, 1996; van der Valk and De Groot, 2000). This concept is supported by the absence of transcripts, and of protein (Leppert et al., 1998) of MMP-7 in CSF of multiple sclerosis, and points to resident cells (e.g. macrophages, microglial cells or astrocytes) as its exclusive source in brain parenchyma (Cossins et al., 1997). Conversely, the increased transcriptional (present study) and protein expression of MMP-9 in CSF and brain tissue of multiple sclerosis (Cossins et al., 1997), and the relatively low expression level in chronic inactive plaques, suggests that leucocytes, besides endothelial cells, are quantitatively the most relevant producers. Low amounts of MMP-8 were detected in the majority of lesions and NAWM, but not in controls. Initially, MMP-8 was believed to originate exclusively from neutrophils and was therefore named ‘neutrophil collagenase’. Based on this concept, the occurrence of MMP-8 transcripts is difficult to explain, as accumulation of neutrophils in the brain parenchyma is not a typical feature of multiple sclerosis. However, endothelial cells, fibroblasts and macrophages recently have also been demonstrated to produce MMP-8, and may represent sources in multiple sclerosis (Hanemaaijer et al., 1997; Prikk et al., 2001). The pathogenesis of multiple sclerosis is associated with an upregulation of cytokines such as TNF-α/β and IL-2 (Hofman et al., 1989; Selmaj et al., 1991; Wucherpfennig et al., 1992; Cannella and Raine, 1995; Raine et al., 1998).

MMP and TIMP expression in MS tissue TNF-α and TNF receptors have been localized immunohistochemically in acute and chronically active multiple sclerosis lesions, but a moderate elevation of expression was also present in adjacent NAWM (Hofman et al., 1989; Selmaj et al., 1991; Raine et al., 1998; Bitsch et al., 2000). Interestingly, the metalloproteinase TACE (ADAM-17) that transforms cell membrane-bound TNF-α, and possibly TNF receptors, into soluble forms was not induced in brain tissue, nor in CSF from multiple sclerosis. As expected, IL-2 transcripts were detected only in acute lesions and adjacent NAWM where parenchymal invasion by T cells is a typical feature, but were absent in controls and other multiple sclerosis tissue. In contrast, transcripts of all three subunits of the IL-2 receptor were constitutively expressed in high amounts in controls and were not significantly different in multiple sclerosis. In agreement with current results, others have found that TNF-α and IL-2 did not display a consistent and reproducible pattern of expression when analysed in eight specimens of different multiple sclerosis lesion types (Baranzini et al., 2000). The restricted, or lack of upregulation of cytokines and cytokine receptors contrasts with the sustained induction of MMP-7 and -9 and led us to conclude that other signalling molecules are more relevant regulators of MMP expression in multiple sclerosis. In summary, the present results add to the concept of a specific role for MMP-7 and -9 in both lesion formation and pre-lesional changes in NAWM of multiple sclerosis. Based on the lack of induction of TIMPs, we hypothesize that the failure to induce compensatory inhibition against excess proteolytic activity is a primary feature of multiple sclerosis pathogenesis. This continuous inflammatory stress could lead to neuronal loss, and may eventually promote clinical secondary chronic progression. However, brain tissue samples from other neurological (inflammatory and noninflammatory) diseases need to be analysed in order to assess better the specificity of the present findings for multiple sclerosis. From a therapeutic point of view, current findings add to the molecular explanation for the beneficial effect of interferon-β in multiple sclerosis. Besides suppressing production of MMPs in vitro (Leppert et al., 1996; Stu¨ ve et al., 1996) and in serum of multiple sclerosis patients (Trojano et al., 1999; Ozenci et al., 2000), interferon-β also increases gene transcription of TIMP-1, and may, therefore, attenuate MMP overactivity in multiple sclerosis by a dual approach. Furthermore, the lack of sufficient endogenous inhibitors containing MMP activity broadens the scientific basis for the use of hydroxamic acid-type MMP inhibitors in multiple sclerosis (Yong, 1999). The results of this and prior studies (Leppert et al., 1998; Lee et al., 1999; Waubant et al., 1999) suggest that such a drug would need to inhibit both MMP-7 and -9 and that long-term administration may be desirable. However, the roles of the more recently detected MMPs (MMP-14 to MMP-28) and unknown MMP signalling molecules in multiple sclerosis remain to be determined. Only a comprehensive analysis of gene regulation based on

1751

microarray techniques is expected to identify the complete network of multiple sclerosis disease genes.

Acknowledgements We wish to thank Caroline Grygar for excellent technical assistance, Dr R. Ravid of The Netherlands Brain Bank for supplying the human CNS tissue, and Dr W. Kamphorst for the neuropathological evaluation. This work was supported by grants (4038-52841 and 31.51084.97) from the Swiss National Science Foundation, the Swiss Multiple Sclerosis Society, the Schering Foundation, the The´ odore Ott Foundation and Swiss Life Insurance.

References Anthony DC, Ferguson B, Matyzak MK, Miller KM, Esiri MM, Perry VH. Differential matrix metalloproteinase expression in cases of multiple sclerosis and stroke. Neuropathol Appl Neurobiol 1997; 23: 406–15. Anthony DC, Miller KM, Fearn S, Townsend MJ, Opdenakker G, Wells GM, et al. Matrix metalloproteinase expression in an experimentally-induced DTH model of multiple sclerosis in the rat CNS. J Neuroimmunol 1998; 87: 62–72. Baranzini SE, Elfstrom C, Chang S-Y, Butunoi C, Murray R, Higuchi R, et al. Transcriptional analysis of multiple sclerosis brain lesions reveals a complex pattern of cytokine expression. J Immunol 2000; 165: 6576–82. Bever CT Jr, Rosenberg GA. Matrix metalloproteinases in multiple sclerosis. Targets of therapy or markers of injury? [Review]. Neurology 1999; 53: 1380–1. Bitsch A, Kuhlmann T, Da Costa C, Bunkowski S, Polak T, Bru¨ ck W. Tumor necrosis factor alpha mRNA expression in early multiple sclerosis lesions. Correlation with demyelinating activity and oligodendrocyte pathology. Glia 2000; 29: 366–75. Boven LA, Montagne L, Nottet HSLM, De Groot CJA. Macrophage inflammatory protein-1α (MIP-1α), MIP-1β, and RANTES mRNA semiquantification and protein expression in active demyelinating multiple sclerosis (MS) lesions. Clin Exp Immunol 2000; 122: 257–63. Cannella B, Raine CS. The adhesion molecule and cytokine profile of multiple sclerosis lesions. Ann Neurol 1995; 37: 424–35. Catalaa I, Grossman RI, Kolson DL, Udupa JK, Nyul LG, Wei L, et al. Multiple sclerosis: magnetization transfer histogram analysis of segmented normal-appearing white matter. Radiology 2000; 216: 351–5. Chandler S, Coates R, Gearing A, Lury J, Wells G, Bone E. Matrix metalloproteinases degrade myelin basic protein. Neurosci Lett 1995; 201: 223–6. Clements JM, Cossins JA, Wells GM, Corkill DJ, Helfrich K, Wood LM, et al. Matrix metalloproteinase expression during experimental autoimmune encephalomyelitis and effects of a combined matrix metalloproteinase and tumour necrosis factor-α inhibitor. J Neuroimmunol 1997; 74: 85–94.

1752


Cossins JA, Clements JM, Ford J, Miller KM, Pigott R, Vos W, et al. Enhanced expression of MMP-7 and MMP-9 in demyelinating multiple sclerosis lesions. Acta Neuropathol (Berl) 1997; 94: 590–8.

Kwon EE, Prineas JW. Blood–brain barrier abnormalities in longstanding multiple sclerosis lesions. An immunohistochemical study. J Neuropathol Exp Neurol 1994; 53: 625–36.

Cuzner ML, Gveric D, Strand C, Loughlin AJ, Paemen L, Opdenakker G, et al. The expression of tissue-type plasminogen activator, matrix metalloproteases and endogenous inhibitors in the central nervous system in multiple sclerosis: comparison of stages in lesion evolution. J Neuropathol Exp Neurol 1996; 55: 1194–204.

Lee MA, Palace J, Stabler G, Ford J, Gearing A, Miller K. Serum gelatinase B, TIMP-1 and TIMP-2 levels in multiple sclerosis. A longitudinal clinical and MRI study. Brain 1999; 122: 191–7.

Gearing AJ, Beckett P, Christodoulou M, Churchill M, Clements J, Davidson AH, et al. Processing of tumour necrosis factor-alpha precursor by metalloproteinases. Nature 1994; 370: 555–7. Gibson UE, Heid CA, Williams PM. A novel method for real time quantitative RT–PCR. Genome Res 1996; 6: 995–1001. 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. Hanemaaijer R, Sorsa T, Konttinen YT, Ding Y, Sutinen M, Visser H, et al. Matrix metalloproteinase-8 is expressed in rheumatoid synovial fibroblasts and endothelial cells. Regulation by tumor necrosis factor-alpha and doxycycline. J Biol Chem 1997; 272: 31504–9.

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. Leppert D, Waubant E, Bu¨ rk MR, Oksenberg JR, Hauser SL. IFN beta-1b inhibits gelatinase secretion and in vitro migration of human T cells: a possible mechanism for treatment efficacy in multiple sclerosis. Ann Neurol 1996; 40: 846–52. Leppert D, Ford J, Stabler G, Grygar C, Lienert C Huber S, et al. Matrix metalloproteinase-9 (gelatinase B) is selectively elevated in CSF during relapses and stable phases of multiple sclerosis. Brain 1998; 121: 2327–34. Leppert D, Leib SL, Grygar C, Miller KM, Schaad UB, Holla¨ nder GA. Matrix metalloproteinase (MMP)-8 and MMP-9 in cerebrospinal fluid during bacterial meningitis: association with blood– brain barrier damage and neurological sequelae. Clin Infect Dis 2000; 31: 80–4.

Hartung HP. Pathogenesis of multiple sclerosis. In: Abramsky O, Ovadia H, editors. Frontiers in multiple sclerosis. London: Martin Dunitz; 1997. p. 45–59.

Lichtinghagen R, Seifert T, Kracke A, Marckmann S, Wurster U, Heidenreich F. Expression of matrix metalloproteinase-9 and its inhibitors in mononuclear blood cells of patients with multiple sclerosis. J Neuroimmunol 1999; 99: 19–26.

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.

Lucchinetti C, Bru¨ ck W, Parisi J, Scheithauer B, Rodriguez M, Lassmann H. Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of demyelination. Ann Neurol 2000; 47: 707–17.

Hofman FM, Hinton DR, Johnson K, Merrill JE. Tumor necrosis factor identified in multiple sclerosis brain. J Exp Med 1989; 170: 607–12.

Maeda A, Sobel RA. Matrix metalloproteinases in the normal human central nervous system, microglial nodules, and multiple sclerosis lesions. J Neuropathol Exp Neurol 1996; 55: 300–9.

Johnatty RN, Taub DD, Reeder SP, Turcovski-Corrales SM, Cottam DW, Stephenson TJ, et al. Cytokine and chemokine regulation of proMMP-9 and TIMP-1 production by human peripheral blood lymphocytes. J Immunol 1997; 158: 2327–3.

Moss ML, Jin SL, Milla ME, Bickett DM, Burkhart W, Carter HL, et al. Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-α. Nature 1997; 385: 733–6.

Kermode AG, Thompson AJ, Tofts P, MacManus DG, Kendall BE, Kingsley DP, et al. Breakdown of the blood–brain barrier precedes symptoms and other MRI signs of new lesions in multiple sclerosis. Brain 1990; 113: 1477–89. 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. Kieseier BC, Seifert T, Giovannoni G, Hartung H-P. Matrix metalloproteinases in inflammatory demyelination. Targets for treatment. [Review]. Neurology 1999; 53: 20–5. Kubota S, Takeda K, Yamada T, Nomura Y, Takeda M, Iwamoto S, et al. Tumor necrosis factor alpha and lymphotoxin stimulate human myeloblastic leukemia cell (ML-1) invasion through a reconstituted basement membrane (Matrigel) with concomitant induction of 92 kDa gelatinase secretion. Cancer Lett 1996; 98: 233–40.

Ozenci V, Kouwenhoven M, Teleshova N, Pashenkov M, Fredrikson S, Link H. Multiple sclerosis: pro- and anti-inflammatory cytokines and metalloproteinases are affected differentially by treatment with IFN-beta. J Neuroimmunol 2000; 108: 236–43. Pagenstecher A, Stalder AK, Kincaid CL, Shapiro SD, Campbell IL. Differential expression of matrix metalloproteinase and tissue inhibitor of matrix metalloproteinase genes in the mouse central nervous system in normal and inflammatory states. Am J Pathol 1998; 152: 729–41. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983; 13: 227–31. Prikk K, Pirila¨ E, Sepper R, Maisi P, Salo T, Wahlgren J, et al. In vivo collagenase-2 (MMP-8) expression by human bronchial epithelial cells and monocytes/macrophages in bronchiectasis. J Pathol. 2001; 154: 232–8.

MMP and TIMP expression in MS tissue Proost P, Van Damme J, Opdenakker G. Leukocyte gelatinase B cleavage releases encephalitogens from human myelin basic protein. Biochem Biophys Res Commun 1993; 192: 1175–81. Raine CS, Bonetti B, Cannella B. Multiple sclerosis: expression of molecules of the tumor necrosis factor ligand and receptor families in relationship to the demyelinated plaque. Rev Neurol (Paris) 1998; 154: 577–85. Rosenberg GA, Estrada EY, Dencoff JE, Stetler-Stevenson WG. Tumor necrosis factor-α induced gelatinase B causes delayed opening of the blood–brain barrier: an expanded therapeutic window. Brain Res 1995; 703: 151–5. Selmaj K, Raine CS, Cannella B, Brosnan CF. Identification of lymphotoxin and tumor necrosis factor in multiple sclerosis lesions. J Clin Invest 1991; 87: 949–54. Stu¨ ve O, Dooley NP, Uhm JH, Antel JP, Francis GS, Williams G, et al. Interferon β-1b decreases the migration of T lymphocytes in vitro: effects on matrix metalloproteinase-9. Ann Neurol 1996; 40: 853–63. Trojano M, Avolio C, Liuzzi GM, Ruggieri M, Defazio G, Liguori M, et al. Changes of serum sICAM-1 and MMP-9 induced by rIFNβ1b treatment in relapsing–remitting MS. Neurology 1999; 53: 1402–8. van der Valk P, De Groot CJ. Staging of multiple sclerosis (MS) lesions: pathology of the time frame of MS. [Review]. Neuropathol Appl Neurobiol 2000; 26: 2–10. van Waesberghe JH, Kamphorst W, De Groot CJ, van Walderveen MA, Castelijns JA, Ravid R, et al. Axonal loss in multiple sclerosis

1753

lesions: magnetic resonance imaging insights into substrates of disability. Ann Neurol 1999; 46: 747–54. Waubant E, Goodkin DE, Gee L, Bacchetti P, Sloan R, Stewart T, et al. Serum MMP-9 and TIMP-1 levels are related to MRI activity in relapsing multiple sclerosis. Neurology 1999; 53: 1397–401. Werring DJ, Brassat D, Droogan AG, Clark CA, Symms MR, Barker GJ, et al. The pathogenesis of lesions and normal-appearing white matter changes in multiple sclerosis: a serial diffusion MRI study. Brain 2000; 123: 1667–76. Woessner JF Jr. The matrix metalloproteinase family. In: Parks WC, Mecham RP, editors. Matrix metalloproteinases. San Diego: Academic Press; 1998. p. 1–14. Wucherpfennig KW, Newcombe J, Li H, Keddy C, Cuzner ML, Hafler DA. T cell receptor V alpha-V beta repertoire and cytokine gene expression in active multiple sclerosis lesions. J Exp Med 1992; 175: 993–1002. Xia M, Leppert D, Hauser SL, Sreedharan SP, Nelson PJ, Krensky AM, et al. Stimulus specificity of matrix metalloproteinase dependence of human T cell migration through a model basement membrane. J Immunol 1996; 156: 160–7. Yong VW. The potential use of MMP inhibitors to treat CNS disease. Exp Opin Invest Drugs 1999; 8: 255–68. Yong VW, Krekoski CA, Forsyth PA, Bell R, Edwards DR. Matrix metalloproteinases and diseases of the CNS. [Review]. Trends Neurosci 1998; 21: 75–80. Received January 15, 2001. Revised March 27, 2001. Accepted April 26, 2001