Increased Osteopontin Levels in the Cerebrospinal Fluid of Patients With Multiple Sclerosis FREE

Proinflammatory cytokines and T cells reactive against myelin proteins play an important role in the pathogenesis of multiple sclerosis (MS) and its experimental model, experimental autoimmune encephalomyelitis (EAE).^1- 2

Osteopontin (OPN) is a phosphoprotein containing an arginine-glycine-aspartate integrin binding motif, with important roles in inflammation and immunity to infection.³ Osteopontin enhances interferon γ and interleukin 12 (IL-12), proinflammatory cytokines implicated in MS, and diminishes IL-10, a cytokine protective in MS and other inflammatory diseases. Osteopontin increases myelin-reactive T cell survival.⁴

Osteopontin is highly upregulated in the brains of persons with MS,^5- 6 and in the spinal cords of mice with EAE.⁵ In OPN-deficient mice, EAE shows a greater number of remissions, less progression, and higher IL-10 levels.⁵ Moreover, OPN induces relapses and progression of EAE.⁴

Osteopontin levels are higher in the serum of MS patients during relapse than during remission; however, no differences in OPN levels were found between patients with relapsing-remitting MS and control group participants.^7- 8 Plasma OPN levels are higher in patients with secondary progressive MS than in healthy controls.⁹

We measured OPN levels in the cerebrospinal fluid (CSF) of MS patients and compared them with the levels in patients with other inflammatory and noninflammatory neurological diseases.

Twenty-seven patients (17 women, 10 men; mean [SD] age, 42.1 [10.1] years; range, 23-62 years) with MS were investigated. Twenty-two had relapsing-remitting MS, 4 primary progressive MS, and 1 secondary progressive MS. We also investigated 11 patients (7 men, 4 women; mean [SD] age, 50.5 [15.8] years; range, 22-79 years) with other inflammatory neurological diseases (OIND), including 3 with chronic inflammatory demyelinating polyradiculoneuropathy, 3 with neurosarcoidosis, 2 with acute disseminating encephalomyelitis, and 1 each with ankylosing spondylitis with C1 transverse myelitis, multifocal motor neuropathy, and systemic lupus erythematosus. Twenty-four patients had noninflammatory neurological diseases (NIND) (18 women, 6 men; mean [SD] age, 42.1 [18.7] years; range, 17-74 years), comprising 11 patients with idiopathic intracranial hypertension, 4 with cerebrovascular disease, 2 with tension headaches, 2 with motor neuron disease, and 1 each with progressive ophthalmoplegia, complex partial seizures, nonepileptic seizures, dystonia, and normal-pressure hydrocephalus. The Nottingham Research Ethics Committee approved the study. All patients gave informed consent. In MS patients, disability was scored using Kurtzke's expanded disability status scale.¹⁰

Cerebrospinal fluid and plasma were collected simultaneously, aliquoted, and stored immediately at -80°C until use. Osteopontin was measured in the plasma and CSF by sandwich enzyme-linked immunosorbent assay (ELISA) (Assay Designs, Ann Arbor, Michigan; sensitivity, 3.5 ng/mL). We measured CSF matrix metalloproteinase 9 (MMP-9) (R+D Systems, Abingdon, England; sensitivity, 0.156 pg/mL), IL-10 (R+D Systems, Abingdon, England; sensitivity, 0.5 pg/mL), and IL-12p40 (Diaclone, Boldon, England; sensitivity, 20 pg/mL), all by ELISA, following the manufacturer's instructions. Cerebrospinal fluid was collected at the time of elective lumbar puncture. Multiple sclerosis patients were not in relapse or receiving disease-modifying treatment at the time of sample collection.

Cytokine levels were compared between groups using the Mann-Whitney U test. Correlations were explored using the Pearson coefficient.

The mean (SD) levels of OPN in plasma were 380 (236) ng/mL in the MS group, 386 (271) ng/mL in the OIND group, and 377 (121) ng/mL in the NIND group. Differences were not statistically significant. In the MS patients, the mean (SD) level of disability, using Kurtzke's expanded disability status scale, was 2.8 (1.6) (range, 1-6.5).

In the relapsing-remitting MS group, the mean (SD) interval between the last clinical relapse and sample collection was 5.8 (7) months.

Multiple sclerosis patients had a mean (SD) CSF OPN level of 415 (186) ng/mL, similar to OIND patients (563 [411] ng/mL) (difference not statistically significant; P = .5). Both were higher than those of the NIND patients (286 [150] ng/mL) (P = .02 for patients with MS; P = .02 for patients with OIND) (Figure). Patients with MS and patients with OIND together had higher OPN levels than the NIND control group (457 [271] ng/mL vs 289 [146] ng/mL; P = .008).

Levels of CSF IL-12p40 were detectable in 15 patients, of whom 8 had MS, 4 had OIND (1 with sarcoidosis, 1with multifocal motor neuropathy, 2 with polyradiculoneuropathy), and 3 had NIND (motor neuron disease, tension headache, or benign intracranial hypertension). Levels varied between 0.1 and 985 pg/mL (mean [SD], 71.8 [244] pg/mL), with the outlier of 985 pg/mL in a patient with neurosarcoidosis. Without this outlier, the mean (SD) IL-12p40 level was 10.9 (13) pg/mL. The median IL-12p40 level was 6.4 pg/mL. Levels of CSF OPN were higher in patients with detectable IL-12p40 than in those with undetectable IL-12p40 (mean [SD], 416 [184] pg/mL vs 235 [169] pg/mL; P = .02).

Cerebrospinal fluid IL-10 was detectable in 15 patients (mean [SD], 3.6 [2.9] pg/mL; range, 0.02-7.9 pg/mL) of whom 8 had MS (3 with simultaneously detectable IL-12p40; all with high IL-12p40 and low IL-10). Osteopontin levels did not correlate with either IL-10 levels or IL-10 detection. Levels of MMP-9 were below theELISA detection limit in all samples.

There was no correlation between OPN levels in CSF and plasma. Levels of CSF OPN appeared consistently higher than plasma levels, although differences were not statistically significant (P = .1). There was no correlation between the CSF OPN levels and MS type, though with 1 secondary progressive MS patient, analysis for secondary progressive MS was not possible. There was no correlation between OPN levels and Kurtzke's expanded disability status scale or the interval since the last relapse.

We found significantly elevated levels of OPN in the CSF of patients with MS and OIND compared with those with NIND. Levels of OPN were slightly higher in MS CSF than in plasma, suggesting a component of CSF OPN was contributed to by central nervous system (CNS)–derived OPN. This is consistent with the ability of CNS cells to express OPN,¹¹ and with the finding of high OPN levels in MS patients' brains.⁵

The fact that OPN levels were increased in the CSF of relapsing-remitting MS patients who were in clinical remission suggests continuous upregulation of OPN in the CNS of patients with MS, regardless of clinical disease activity. The same pattern of upregulation was found for MMP-9 in the CSF of patients with MS, suggesting the destructive proteolytic process is continuous.¹² Interestingly, OPN activates pro–MMP-9¹³ and its conversion to an active enzyme. Therefore, the similar patterns in the CSF of patients with MS may be explained by OPN activating the destructive effects of MMP-9. This hypothesis remains plausible despite the fact that, in our samples, levels of MMP-9 were below the ELISA detection limit. In a number of CSF samples from patients of Leppert et al,¹² MMP-9 was undetectable by ELISA but detectable by zymography; however, we had insufficient CSF to conduct zymographies.

Osteopontin is related to T helper 1–type responses. We therefore explored the correlation between OPN and IL-12p40 levels. Only 15 patients (8 with MS) had detectable IL-12p40 levels, probably reflecting the fact that our patients were in clinical remission, and consistent with previous findings associating CSF IL-12p40 detection with radiologically active disease.¹⁴ Similarly, only 8 MS patients had detectable levels of CSF IL-10, making correlation analysis difficult.

Osteopontin was increased in the CSF of OIND patients, indicating that OPN upregulation is not MS-specific but is inflammation-specific. The higher levels seen in CSF compared with plasma also suggest a contribution from CNS OPN. In these studies, we did not have the opportunity to compare CSF OPN levels between clinical relapse and remission. However, although previous studies showed no significant difference in plasma OPN levels between MS and healthy participants,⁷ we detected differences in CSF OPN levels between stable MS patients and participants in the control group. The majority of our patients had a mild to moderate disability; this, together with CSF OPN being increased regardless of the clinical disease activity, further suggests that OPN contributes continuously, and possibly from early stages, to MS pathology.

Osteopontin has been implicated in remyelination after experimental demyelination. Osteopontin was significantly upregulated in both the demyelination and re myelinating phases, suggesting a dual role.¹⁵ A positive OPN effect in remyelination was suggested by its ability to stimulate oligodendrocytes.¹⁵ Whether this dual OPN role is also played in MS, where demyelinating and remyelinating lesions coexist, remains to be established. Our data, including the correlation of CSF OPN levels with the detection of proinflammatory IL-12p40, but not of antiinflammatory IL-10, and recent observations in EAE,³ however, strongly suggest that in CNS inflammatory demyelination, OPN net effect is detrimental.

In conclusion, we found elevated OPN levels in the CSF of patients with MS and OIND which tend to exceed plasma levels, suggesting that the sources of OPN in these conditions include a contribution from CNS resident or infiltrating cells.

Correspondence: Cris S. Constantinescu, MD, PhD, Division of Clinical Neurology, University of Nottingham, Queen's Medical Centre, Nottingham NG7 2UH, England (cris.constantinescu@nottingham.ac.uk).

Accepted for Publication: October 30, 2007.

Author Contributions:Study concept and design: Constantinescu. Acquisition of data: Braitch, Nunan, Niepel, Edwards. Analysis and interpretation of data: Braitch, Nunan, Niepel, Constantinescu. Drafting of the manuscript: Braitch, Niepel, Edwards, Constantinescu. Critical revision of the manuscript for important intellectual content: Nunan, Constantinescu. Statistical analysis: Braitch. Obtained funding: Constantinescu. Administrative, technical, and material support: Braitch, Niepel, Edwards. Study supervision: Constantinescu.

Financial Disclosure: None reported.

Funding/Support: Supported in part by grants from the University of Nottingham, the Stanhope Trust, and the Multiple Sclerosis Society of Great Britain and Northern Ireland.