You are here
Cerebrospinal fluid IL-1β correlates with cortical pathology load in multiple sclerosis at clinical onset
Journal of Neuroimmunology
The cerebrospinal fluid levels of interleukin-1 beta and structural magnetic resonance parameters of cortical damage, i.e., cortical lesion number and volume, and global cortical thickness, were analysed in multiple sclerosis patients at clinical onset. Cerebrospinal fluid interleukin-1 beta levels strongly correlated with cortical lesion load and cortical thickness, while correlation with white matter lesion load was modest. Interleukin-1 beta, intrathecally produced by infiltrating lymphocytes and activated microglia, may constitute a possible link between inflammation and neurodegeneration in multiple sclerosis.
- Cortical demyelinating lesions and atrophy are peculiar features of MS pathology.
- The progressive cortical damage in MS determines physical and cognitive decline.
- CSF levels of IL-1β correlate with cortical pathology in very early MS.
- IL-1β, expressed by activated microglia, may contribute to neurodegeneration in MS.
- This cytokine may play a role in linking inflammation and neurodegeneration in MS.
Abbreviations: CSF - cerebrospinal fluid, MS - multiple sclerosis, IL-1β - interleukin-1 beta, CL - cortical lesion, CTh - cortical thickness, T2WMLV - T2 white matter lesion volume, DIR - double inversion recovery, FFE - 3D fast field echo, FLAIR - 3D fluid attenuated inversion recovery, EDSS - Expanded Disability Status Scale, CIS - clinically isolated syndrome, RRMS - relapsing–remitting MS, MRI - magnetic resonance imaging.
Keywords: Multiple sclerosis, Interleukin-1β, Cortical pathology.
An extensive and progressive cortical damage occurs in multiple sclerosis (MS) and evidence is accumulating on the pivotal role played by neuronal damage in determining the physical and cognitive decline observed in MS patients (Calabrese et al, 2009 and Calabrese et al, 2011). Cortical demyelinating lesions (CLs) and atrophy are peculiar features of MS pathology and are frequently observed since the early disease stages (Dalton et al, 2004 and Calabrese et al, 2010). CLs are associated with meningeal inflammation ( Lucchinetti et al., 2011 ), and may precede the appearance of white matter (WM) lesions ( Popescu et al., 2011 ).
Although a peculiar histological feature of MS cortical pathology is the presence of activated microglia cells ( Block and Hong, 2005 ), the mechanisms by which these cells may contribute to neurodegeneration remain elusive. However, MS shares some histological characteristics with other neurodegenerative diseases (e.g., Alzheimer's disease, Parkinson's disease, etc.), such as changes in microglia number and morphology, elevated tissue expression of pro-inflammatory cytokines and oxidative stress mediators, and progressive neuronal loss. The hypothesis that microglia might be a chronic source of soluble cytotoxic factors (e.g., tumour necrosis factor and nitric oxide) that drive neuron damage, and may be implicated in the chronic nature of neurodegenerative diseases, has been suggested (Block and Hong, 2005 and Cunningham, 2012).
Among the cytokines produced by microglia, interleukin-1 beta (IL-1β) is a major pro-inflammatory cytokine that induces cyclooxygenase type 2, increases expression of adhesion molecules and synthesis of cytotoxins, and affects antigen recognition and lymphocyte function ( Dinarello, 2009 ). Moreover, IL-1β may also trigger oligodendrocyte and neuronal degeneration by inducing the upregulation of excitatory glutamatergic transmission ( Takahashi et al., 2003 ) and downregulation of inhibitory GABAergic transmission ( Galic et al., 2012 ). In primary rat and human neuronal cultures, IL-1β induces increased levels of intra- and extra-cellular glutamate, neuronal death and apoptosis ( Ye et al., 2013 ). Moreover, the subacute administration of IL-1β in rats was found to produce memory deficits, and increased expressions of APP, microglia active marker CD11b, TNFα and IL-1β, and reduce expression of astrocyte active marker glial fibrillary acidic protein, brain-derived neurotrophic factor and tyrosine kinase B ( Song et al., 2013 ).
The co-presence of IL-1β, microglia activation/proliferation and neuronal degeneration has been observed in Alzheimer's disease and in preclinical experimental models of dementia ( Cunningham, 2012 ), thus suggesting a link between IL-1β and neuron injury to death. Recently, IL-1β-containing cerebrospinal fluid (CSF) from active MS patients was found to increase spontaneous glutamate-mediated excitatory postsynaptic current frequency and glutamate-mediated neuronal swelling in vitro, thus suggesting a possible link between inflammation and excitotoxic neurodegeneration in MS ( Rossi et al., 2012 ).
We analysed the relationship between CSF levels of IL-1β and cortical inflammatory lesion load and cortical thickness in patients with clinically isolated syndromes or early MS to evaluate whether a soluble inflammatory factor produced in the brain tissue and measurable in the CSF could be associated with MRI-detectable cortical grey matter changes.
2. Materials and methods
Thirty-six consecutive individuals (21 F, 15 M; mean age ± SD: 35.5 ± 10.5 years, range: 16–57 years; mean disease duration: 6.1 ± 9.8 months, range: 0–53 months) presenting with symptoms and signs suggestive of MS, in the period from February 2009 to September 2011, were included in the study. Median Expanded Disability Status Scale (EDSS) value was 1.5 ± 0.5 (range 1.0–2.5). Twenty-two patients were at first clinical episode and had MRI evidence of dissemination in space, but not in time, of the lesions and thus were classified as clinically isolated syndrome (CIS/possible MS) (McDonald/Polman criteria). Fourteen patients, who previously had symptoms suggestive of brain or spinal cord inflammation (mean interval of time: 6.7 ± 6.3 months, range: 1–18), were classified as early relapsing–remitting MS (RRMS). As expected, the mean EDSS value was a little higher in RRMS (2.0, range 0–3) compared to CIS (1.5, range 0–3) (p = 0.012). All the patients underwent a complete diagnostic work-up including CSF analysis and magnetic resonance imaging (MRI) for white matter and grey matter parameters. No patient was treated with immunomodulatory or immunosuppressive therapies. Four patients received high-dose steroids (methylprednisolone, 1 g/day for 5 days) in the month before study entry (25.5 ± 8.2 days). Ten patients affected by neuromyelitis optica (NMO) were included in the study as inflammatory controls. All these patients were positive for the search of anti-aquaporin-4 antibodies, had a polyclonal IgG pattern in the CSF (collected during an active disease phase) and no evidence of cortical lesions at MRI (as described elsewhere, Geurts et al., 2011 ). Twenty subjects presenting with headache, sensory (e.g., paresthesias) or other subjective symptoms (e.g., dizziness), who underwent lumbar puncture for diagnostic purposes were included in the study as non-inflammatory controls. All investigations resulted normal in these subjects and neurological (inflammatory and/or degenerative) diseases were definitely excluded. The study was approved by the local Ethic Committee and an informed consent was obtained from all the patients.
2.2. CSF analysis
CSF was collected by non-traumatic lumbar puncture between 8.00 and 9.00 a.m. Routine examination on paired CSF and serum specimens included: cell count and differentiation, albumin CSF/serum ratio to estimate the integrity of the blood–brain barrier, calculation of intrathecal IgG synthesis by means of quantitative formulae (IgG index, IgG HypFunct) and demonstration of IgG oligoclonal bands (IgGOB) by isoelectric focusing and specific IgG immunofixation. After cell centrifugation, the CSF was stored at − 20 °C in 100 μL aliquots until cytokine determination.
2.3. IL-1β determination
CSF levels of IL-1β were measured by a highly sensitive immunoenzymatic assay ELISA (Avibon, catalogue #IL01b02) in undiluted samples. The assay was performed according to the manufacturer's instruction. Concentrations of IL-1β were expressed in picograms per millilitres (pg/mL).
2.4. MRI sequences
Brain MRI was performed using a 1.5 T Philips Achieva (Philips Medical Systems, Best, and the Netherlands) as previously described in detail ( Calabrese et al., 2010 ) and consisted in the following sequences: double inversion recovery (DIR), 3D fast field echo (FFE), 3D fluid attenuated inversion recovery (FLAIR), and post-contrast T1-weighted spin echo.
Determination of CL number and CL volume on DIR images and of volume of T2 WM lesions (T2WMLV) on FLAIR images was achieved by consensus of two experienced observers blind to patient identity and to the date of image acquisition, as previously described in detail ( Calabrese et al., 2010 ) and in agreement with the recent MAGNIMS Study Group's recommendations ( Geurts et al., 2011 ).
Global and regional cortical thickness (CTh) (mean of right and left hemispheres) was performed on the volumetric FFE data sets by means of Freesurfer image analysis suite, as described in detail elsewhere ( Calabrese et al., 2011 ). Spinal cord MRI included T1, T2, STIR and contrast enhancing T1 sequences.
2.5. Statistic analysis
Mann–Whitney tests were applied to assess the differences between groups. Univariate correlations were assessed using the non-parametric Spearman rank correlation coefficient. A stepwise linear regression analysis was performed to assess the relative contributions of all MRI and CSF variables (CSF-lymphocytes, CSF-proteins, IgG index, IgG, T2WMLV, CL number, CL volume and CTh) in predicting increases CSF levels of IL-1β. Backward and forward stepwise analyses were conducted using Wald's statistic as a criterion.
IgGOB and increased IgG index were demonstrated in the CSF of 94% (34/36) and 61% (22/36) patients, respectively. The albumin ratio was increased in 5/36 (13.9%) and the number of CSF lymphocytes in 21/36 (58.3%; range 0.3–70).
Detectable CSF levels of IL-1β were found in 19/36 (52.8%) patients, ranging from 1.66 to 12.75 pg/mL (mean ± SD, 4.37 ± 3.31). No correlation was observed between IL-1β levels and the lymphocyte number, the intrathecal IgG synthesis (value of the IgG Index) and the presence of increased albumin ratio (p > 0.5 for all comparisons).
Patients having IL-1β (IL-1β+) in the CSF had significantly higher CL volume (p < 0.001) and number (p < 0.001), global CTh (p = 0.007) and T2WMLV (p = 0.049) compared to IL-1β−patients ( Fig. 1 ). No difference was observed in terms of demographic parameters and EDSS score between the two groups ( Table 1 ). The difference, statistically not significant, in disease duration between IL-1β+and IL-1β−patients was explained by two IL-1β−patients having higher disease duration. The majority of IL-1β−cases had disease duration within the range of IL-1β+cases. Among the four patients having received high dose steroids in the month before CSF analysis, 3 (75%) had detectable levels of IL-1β, thus suggesting that a high-dose of steroids does not suppress intrathecal IL-1β production.
|IL-1β− patients||p-Value||IL-1β+ patients|
|IgGOB positivity (%)||94.1%||0.9||94.7%|
|Leucocyte (on μL) (mean ± SD; range)||13.0 ± 16.9||0.2||7.7 ± 8.5|
|IgG CSF (mg/L) (mean ± SD; range)||40,8 ± 16,9||0.9||39.7 ± 21.1|
|IgG index (mean ± SD; range)||0.8 ± 0.5||0.7||0.8 ± 0.3|
|EDSS (median; range)||1.5||1||1.5|
|Patients who received high dose steroids (%)||1 (6%)||0.3||3 (16%)|
|Disease duration (months) (mean ± SD; range)||7.8 ± 13.5||0.4||4.7 ± 4.2|
|CL number (median; range)||0||< 0.001||8|
|CL volume (mm3) (median; range)||0||< 0.001||650|
|T2WMLV (mm3) (mean ± SD; range)||962.4 ± 1816.2||< 0.05||2208.6 ± 1840.7|
|CTh (mm) (mean ± SD; range)||2.5 ± 0.1||< 0.01||2.3 ± 0.2|
|gad+ patients (%)||40%||0.5||54%|
Abbreviations: patients with or without any detectable level of interleukin-1 beta in CSF (respectively IL-1β+and IL-1β−patients); cerebrospinal fluid (CSF); IgG oligoclonal bands (IgGOB); expanded disability status scale (EDSS); cortical lesion (CL); cortical thickness (CTh); T2-white matter lesion volume (T2WMLV); and patients with almost one gadolinium enhancing lesion on post-contrast T1-weighted spin echo (gad+patients).
A significant correlation was observed between the CSF levels of IL-1β and CL volume (r = 0.810, p < 0.001), CL number (r = 0.784, p < 0.001), global CTh (r = − 0.485, p < 0.005), and T2WMLV (r = 0.525, p < 0.001). The stepwise linear regression analysis identified the CL volume as the best predictor of the increases levels of IL-1β in the CSF (B: 0.893, p < 0.001) ( Fig. 2 ).
When age and gender were included as covariates they resulted as not statistically significant.
Interestingly, no significant correlation was found between the presence and levels of IL-1β in the CSF and the presence of gadolinium-enhancing lesions (gad+). Indeed, gad+was observed in 48% patients and IL-1β levels in patients presenting on MRI gad+(2.99 ± 3.87, range 0.00–11.68) did not differ from those of gad−patients (1.91 ± 2.37, range 0.00–6.78) (p = 0.423). Moreover, the correlation between CL volume and IL-1β levels did not differ when gad+and gad−patients were separately considered (r = 0.854 and r = 0.851, respectively).
IL-1β was detectable in 9/16 (56.3%) patients having spinal cord lesions. No correlation was observed between lesion volume and IL-1β levels. Ten patients out of 20 (50%) without spinal cord lesions had detectable IL-1β in the CSF (p = 0.7).
No trace of IL-1β could be detected in the sera of patients and controls.
The peculiar features of the pathological process that extensively and progressively destroys the cortex of MS patients are a low degree of inflammation, microglial proliferation and neuronal death. An association between cortical and leptomeningeal inflammation has been also described, even in the early disease stages ( Lucchinetti et al., 2011 ), and the hypothesis that soluble products (e.g., cytokines) of this inflammatory process may diffuse into the CSF spaces seems likely.
However, all efforts aimed at identifying soluble markers of MS activity or progression in the CSF have failed. This may be due to (i) the heterogeneity of the patients studied, (ii) the failure in selecting the more appropriate (immune)pathological and/or MRI parameters, and (iii) the ongoing immunomodulatory or immunosuppressive therapies that downregulate the immune system. Therefore, we selected a highly homogeneous population of untreated patients with CIS highly suggestive of MS (i.e., with dissemination in space of the lesions and evidence of intrathecal synthesis of IgG) and early RRMS at the second clinical episode. The number of IL-1β+patients and IL-1β levels that we have observed is in line with literature data. The lack of correlation between IL-1β levels and leukocyte number seems to exclude the production of this cytokine by cells circulating in the CSF, and suggests its intra-tissue synthesis.
We observed a high correlation between CSF IL-1β levels and MRI parameters of cortical pathology, namely CL number and volume, and global CTh. These findings, together with the lack of correlation between CSF IL-1β and gad+lesions, suggest that the cellular source of this cytokine might be in the cortex rather than in active inflammatory white matter lesions. The lack of IL-1β in the CSF (collected during acute inflammatory phases) of patients with NMO, a disease characterised by the lack of cortical demyelination or other histological changes suggestive of chronic cortical pathology (i.e., microglia proliferation and activation) ( Popescu et al., 2010 ), is particularly interesting since it further stresses the hypothesis that the presence of IL-1β in MS CSF is probably more related to chronic pathology than to acute inflammation.
Furthermore, when a multiple linear regression was performed, CL volume was identified as the best predictor of the IL-1β levels, while all the other variables were excluded from the model. These findings are particularly relevant because, for the first time, a soluble marker of brain inflammation detectable in the CSF is found to correlate with MRI parameters of cortical damage in MS.
The most important source of IL-1β in the cerebral cortex is microglia, that is found proliferating and activated in almost all stages of MS and in strict contact with neurons in lesions (Block and Hong, 2005 and Cunningham, 2012). Microglia activation and proliferation, however, have been described in several CNS disorders, including inflammatory, degenerative and storage diseases. Although both protective and detrimental actions on neurons have been attributed to microglia, in primary neurodegenerative brain disorders, such as Alzheimer's disease and Parkinson's disease (Block and Hong, 2005 and Cunningham, 2012), the activation of these cells has been associated to local production of IL-1β and neuronal apoptosis. Thus, in the cortex of MS patients microglial-derived IL-1β may not only exert a pro-inflammatory action, but also contribute to trigger the neurodegenerative component of the disease.
Indeed, IL-1β has been demonstrated to promote oligodendrocyte death through glutamate excitotoxicity ( Takahashi et al., 2003 ), and cause synaptic hyperexcitability and glutamate mediated neuronal swelling ( Ye et al., 2013 ), thus contributing to both demyelination and neurodegeneration. Therefore, the association of IL-1β with both inflammatory (CLs) and degenerative (CTh) parameters of cortical damage suggests that this cytokine may play a role in linking inflammation and neurodegeneration in MS, and should be investigated as a possible molecular target for preventing neurodegeneration.
This work was supported by grants from the Istituto Superiore di Sanità (Project No. 3 “Biomarkers and Diagnosis” — ISS-17) and the University of Padova (ProgettoInterarea di Ateneo, Prot. CPDA099394).
- Block and Hong, 2005 M.L. Block, J.S. Hong. Microglia and inflammation-mediated neurodegeneration: multiple triggers with a common mechanism. Prog. Neurobiol.. 2005;76:77-98 Crossref
- Calabrese et al., 2009 M. Calabrese, F. Agosta, F. Rinaldi, I. Mattisi, P. Grossi, A. Favaretto, M. Atzori, V. Bernardi, L. Barachino, L. Rinaldi, P. Perini, P. Gallo, M. Filippi. Cortical lesions and atrophy associated with cognitive impairment in relapsing–remitting multiple sclerosis. Arch. Neurol.. 2009;66:1144-1150
- Calabrese et al., 2010 M. Calabrese, M.A. Rocca, M. Atzori, I. Mattisi, A. Favaretto, P. Perini, P. Gallo, M. Filippi. A 3-year magnetic resonance imaging study of cortical lesions in relapse-onset multiple sclerosis. Ann. Neurol.. 2010;67:376-383
- Calabrese et al., 2011 M. Calabrese, F. Rinaldi, I. Mattisi, V. Bernardi, A. Favaretto, P. Perini, P. Gallo. The predictive value of gray matter atrophy in clinically isolated syndromes. Neurology. 2011;77:257-263 Crossref
- Cunningham, 2012 C. Cunningham. Microglia and neurodegeneration: the role of systemic inflammation. Glia. 2012;10.1002/glia.22350 (Jun 6)
- Dalton et al., 2004 C.M. Dalton, D.T. Chard, G.R. Davies, K.A. Miszkiel, D.R. Altmann, K. Fernando, G.T. Plant, A.J. Thompson, D.H. Miller. Early development of multiple sclerosis is associated with progressive grey matter atrophy in patients presenting with clinically isolated syndromes. Brain. 2004;127:1101-1107 Crossref
- Dinarello, 2009 C.A. Dinarello. Immunological and inflammatory functions of the interleukin-1 family. Annu. Rev. Immunol.. 2009;27:519-550 Crossref
- Galic et al., 2012 M.A. Galic, K. Riazi, Q.J. Pittman. Cytokines and brain excitability. Front. Neuroendocrinol.. 2012;33:116-125 Crossref
- Geurts et al., 2011 J.J. Geurts, S.D. Roosendaal, M. Calabrese, O. Ciccarelli, F. Agosta, D.T. Chard, A. Gass, E. Huerga, B. Moraal, D. Pareto, M.A. Rocca, M.P. Wattjes, T.A. Yousry, B.M. Uitdehaag, F. Barkhof, MAGNIMS Study Group. Consensus recommendations for MS cortical lesion scoring using double inversion recovery MRI. Neurology. 2011;76(5):418-424 Crossref
- Lucchinetti et al., 2011 C.F. Lucchinetti, B.F. Popescu, R.F. Bunyan, N.M. Moll, S.F. Roemer, H. Lassmann, W. Brück, J.E. Parisi, B.W. Scheithauer, C. Giannini, S.D. Weigand, J. Mandrekar, R.M. Ransohoff. Inflammatory cortical demyelination in early multiple sclerosis. N. Engl. J. Med.. 2011;365:2188-2197 Crossref
- Popescu et al., 2010 B.F. Popescu, J.E. Parisi, J.A. Cabrera-Gómez, K. Newell, R.N. Mandler, S.J. Pittock, V.A. Lennon, B.G. Weinshenker, C.F. Lucchinetti. Absence of cortical demyelination in neuromyelitis optica. Neurology. 2010;75(23):2103-2109 Crossref
- Popescu et al., 2011 B.F. Popescu, R.F. Bunyan, J.E. Parisi, R.M. Ransohoff, C.F. Lucchinetti. A case of multiple sclerosis presenting with inflammatory cortical demyelination. Neurology. 2011;76(20):1705-1710 Crossref
- Rossi et al., 2012 S. Rossi, R. Furlan, V. De Chiara, C. Motta, V. Studer, F. Mori, A. Musella, A. Bergami, L. Muzio, G. Bernardi, L. Battistini, G. Martino, D. Centonze. Interleukin-1β causes synaptic hyperexcitability in multiple sclerosis. Ann. Neurol.. 2012;71:76-83 Crossref
- Song et al., 2013 C. Song, Y. Zhang, Y. Dong. Acute and subacute IL-1β administrations differentially modulate neuroimmune and neurotrophic systems: possible implications for neuroprotection and neurodegeneration. J. Neuroinflammation. 2013;10:59 Crossref
- Takahashi et al., 2003 J.L. Takahashi, F. Giuliani, C. Power, Y. Imai, V.W. Yong. Interleukin-1beta promotes oligodendrocyte death through glutamate excitotoxicity. Ann. Neurol.. 2003;53:588-595 Crossref
- Ye et al., 2013 L. Ye, Y. Huang, L. Zhao, Y. Li, L. Sun, Y. Zhou, G. Qian, J.C. Zheng. IL-1β and TNF-α induce neurotoxicity through glutamate production: a potential role for neuronal glutaminase. J. Neurochem.. 2013;125(6):897-908 Crossref
a Multiple Sclerosis Centre of Padova, Department of Neurosciences, University of Padova, via Giustiniani 2, 35128 Padova, Italy
b Neurosciences Laboratory, First Neurology Clinic, Department of Neurosciences, University of Padova, via Giustiniani 2, 35128 Padova, Italy
© 2014 Elsevier B.V., All rights reserved.