Central nervous system (CNS) involvement has been reported to occur in 14–75% of all systemic lupus erythematosus (SLE) patients 123. However, frequency rates vary considerably, depending on the diagnostic criteria applied. CNS lupus can occur at any time during the course of SLE, and its symptoms are diverse. The features of this condition can include seizures, stroke, depression, psychoses and disordered mentation. Neuropsychiatric involvement in SLE (NPSLE) has been shown to predict a high frequency of flares, and it is considered a major cause of long-standing functional impairment as well as a cause of mortality 4. Over the past decade CNS lupus has been treated with cytotoxic drugs, which improve disease outcome 56.

Because of the multiple pathogenic mechanisms that cause manifestations of CNS lupus, no single confirmatory diagnostic test is available. Several clinical, laboratory and radiographic test findings are reported to be abnormal in some but not all patients with CNS lupus. Magnetic resonance imaging (MRI) of brain has been shown to be valuable in detecting even minor lesions caused by CNS lupus, and these correlate with CNS manifestations in SLE 7. Pleocytosis and elevated protein levels are found in some but not all patients with CNS lupus 89. Elevated concentrations of IgG in cerebrospinal fluid (CSF), IgG:albumin ratio and IgG index, and the presence of oligoclonal bands have all been described with varying frequencies in patients with NPSLE 101112. Few studies have demonstrated elevated IL-6 levels in CSF from patients with CNS lupus 1314151617. Some other reports have described increased levels of IL-1 13, IL-8 15 and interferon-γ 18 in CSF from patients with CNS lupus. We recently reported neuronal damage, astrogliosis and formation of toxic metabolic products such as Aβ42 in patients with NPSLE 19.

One of the prototypical destructive events in the human brain, initiated by the release of inflammatory cytokines and ending with tissue destruction, is production of matrix metalloproteinases (MMPs). The MMPs are a family of endopeptidases produced by a variety of inflammatory cells 20. All of the cell types that exist in the CNS are potential sources of MMPs. In vitro, neurones, astrocytes, microglia 2122 and oligodendrocytes 23 express various MMP family members, and production of MMPs by neuronal cells can be upregulated by several inflammatory cytokines. MMPs have a multitude of regulatory functions, including control of influx of inflammatory mononuclear cells into the CNS, participation in myelin destruction and disruption of the integrity of the blood–brain barrier 24. With respect to MMP-9, it was recently shown that CSF levels of this enzyme increase during bacterial meningitis and that it is associated with brain damage 2526. The aim of the present study was to measure levels of free active MMP-2 and MMP-9 in CSF of SLE patients with CNS lupus, and to relate these data to clinical and laboratory measures of brain parenchymal degradation. Our results suggest that MMP-9 but not MMP-2 actively participates in brain destruction in CNS lupus.

A total of 123 patients met criteria for SLE. Forty-three patients were found to have NPSLE (in accordance with the criteria presented in the Methods section above), and nine patients either were found to have phospholipid antibody syndrome or met the CNS lupus criteria but were excluded because of non-SLE origin of neurological events. The remaining 71 SLE patients were considered to be cerebrally healthy.

Mild pleocytosis was seen in patients with CNS lupus (9 × 10⁶ ± 6 × 10⁶ cells/l) as compared with cerebrally healthy SLE patients (2 × 10⁶ ± 0.5 × 10⁶ cells/l; not significant). As previously validated 31, we found an increased number of oligoclonal bands in CSF from the CNS lupus group (1.7 ± 0.4) as compared with SLE patients without CNS involvement (0.5 ± 0.1; P < 0.05). The mean level of CSF:serum albumin ratio was not increased in patients with NPSLE (6.1 ± 0.7 mg/dl) as compared with cerebrally healthy SLE patients (5.2 ± 0.3 mg/dl; not significant). There were no significant differences in levels of serum antibodies specific for native DNA or in complement levels (C3 and C4) between SLE patients with and those without CNS involvement.

Intrathecal MMP-9 levels were significantly increased in all SLE patients as compared with cerebrally healthy control individuals (153 ± 27 versus 0.28 ± 0.16 pg/ml; P = 0.016). In CNS lupus patients MMP-9 levels were significantly increased, both compared with cerebrally healthy SLE patients (240 ± 60 versus 100 ± 20 pg/ml; P < 0.05; Fig. 1) and compared with cerebrally healthy control individuals (240 ± 60 versus 0.28 ± 0.16 pg/ml; P = 0.0012). On stratification of all SLE patients into two groups with respect to the presence or absence of brain MRI pathology, we found increased CSF MMP-9 levels in SLE patients with MRI pathology as compared with those without (160 ± 50 versus 140 ± 30 pg/ml; not significant).

There were no statistically significant differences in CSF MMP-2 levels between NPSLE and patients without CNS lupus (499 ± 70 versus 555 ± 70 pg/ml; not significant) or compared with cerebrally healthy control individuals (499 ± 70 versus 384 ± 33 pg/ml; not significant; Fig. 2).

CSF levels of IL-6 (47 ± 25 pg/ml versus 15 ± 3 pg/ml; P < 0.008) and IL-8 (91 ± 23 pg/ml versus 45 ± 6 pg/ml; P < 0.05) were both significantly increased in CSF from patients with CNS lupus as compared with cerebrally healthy SLE patients, supporting our previous findings 1519. Importantly, intrathecal levels of IL-6 and IL-8 significantly correlated with those of MMP-9 (r = 0.30 [P < 0.002] and r = 0.47 [P < 0.0001], respectively; Table 3 and Fig. 3). A neuronal degeneration marker (protein tau) and an astrocytic degeneration marker (glial fibrillary acidic protein) were both significantly increased in CSF from NPSLE patients as compared with CSF from SLE patients who were clinically free from CNS involvement (311 ± 78 pg/ml versus 178 ± 16 pg/ml [P < 0.05] and 1288 ± 708 pg/ml versus 396 ± 30 pg/ml [P < 0.009]), in concordance with previous findings 2832. Importantly, a significant correlation was noted between intrathecal MMP-9 and levels of tau and glial fibrillary acidic protein (P < 0.05 in both cases; Table 3).

We previously reported that patients with CNS lupus express increased intrathecal levels of proinflammatory cytokines IL-6, IL-8 and interferon-γ. Furthermore, we recently demonstrated that inflammation in CNS during SLE leads to neurodegeneration, which manifests as increased levels of neuronal and astrocytic degradation products 2832. However, the mechanisms responsible for the brain damage occurring during CNS lupus have never been clarified. It was previously demonstrated 3334 that proinflammatory cytokines are able to trigger production and release of tissue-derived MMPs. It is also established that most of the resident cells in human brain have the capacity to produce MMP-2 and MMP-9 – enzymes that are known to participate in brain damage (e.g. in case of infectious meningitis) 353637.

In the present study we found significantly higher levels of MMP-9 in CSF from NPSLE patients than in CSF from SLE patients without CNS lupus and healthy control individuals. This finding may be of clinical significance because of correlation between MMP-9 and levels of neuronal/astrocytic degradation products in CSF, reflecting the potential of MMP-9 to damage brain parenchyma. Findings of a relationship between MMP-9 and brain damage in SLE are new and should be scrutinized critically. We believe that effort should be invested in analyzing free (i.e. non-tissue inhibitor of metalloproteinase bound) and metabolically active enzyme to ascertain the validity of our conclusions. The method used in the present study fulfils both of these requirements.

MMP-9 plays an essential role in the breakdown of extracellular matrix molecules at the blood–brain barrier. This, together with locally elevated levels of the chemokine IL-8, greatly facilitates the transmigration of activated inflammatory cells across the endothelium. Indeed, a significant correlation was observed between the intrathecal levels of MMP-9 and IL-8.

A broad range of cytokines that are expressed in viral meningitis are able to regulate the production of MMPs and tissue inhibitors of metalloproteinase 3839. IL-1α and IL-1β, as well as IL-6, induce or upregulate the transcription of MMP genes 40. This, along with increased IL-6 levels in CSF of patients with CNS lupus, may provide an explanation for the local synthesis of MMP-9.

CSF levels of enzymatically active MMP-2 were not increased in NPSLE patients as compared with SLE patients without CNS involvement or compared with cerebrally healthy control individuals. This finding is consistent with a previous study of viral meningitis, in which increased expression of MMP-9 but constitutive expression of MMP-2 was found 37. The reason for this discrepancy, seen in the present study as well as in the previous one, might be differential regulation of gene promotors for MMP-9 and MMP-2. Indeed, whereas the promotor regions of MMP-9, which is regulated by cytokines, harbor a nuclear factor-κB responsive element, those of MMP-2 lack a TATA box and contain two SP-1 sites, both characteristic of house keeping genes 4041.

Our present and previous findings indicate that the following chain of events takes place during NPSLE: intrathecal release of inflammatory cytokines, leading to synthesis and release of MMP-9, potentially culminating in insult to brain parenchyma, resulting in release of neuronal and astrocytic degradation products and culminating in MRI verifiable lesions and clinical states of brain deficiency. Interestingly, a recent study 42 indicated that IgG dose-dependently downregulates secretion of MMP-9 by macrophages, suggesting a possible early therapeutic manipulation in NPSLE.

None declared.

CNS = central nervous system; CSF = cerebrospinal fluid; IL = interleukin; MMP = matrix metalloproteinase; MRI = magnetic resonance imaging; NPSLE = neuropsychiatric involvement in systemic lupus erythematosus; SLE = systemic lupus erythematosus.