Glial and neuronal markers in cerebrospinal fluid in different types of multiple sclerosis

    Glial and neuronal markers in cerebrospinal fluid in different types of multiple sclerosis ´ M. Alba Ma˜ne´ -Mart´ınez, Bob Olsson, Laura Bau, Elisabet Matas, Alvaro Cobo-Calvo, Ulf Andreasson, Kaj Blennow, Lucia Romero-Pinel, Sergio Mart´ınez-Y´elamos, Henrik Zetterberg PII: DOI: Reference:

S0165-5728(16)30182-5 doi: 10.1016/j.jneuroim.2016.08.004 JNI 476411

To appear in:

Journal of Neuroimmunology

Received date: Revised date: Accepted date:

12 February 2016 17 July 2016 4 August 2016

Please cite this article as: Ma˜ n´e-Mart´ınez, M. Alba, Olsson, Bob, Bau, Laura, Matas, ´ Elisabet, Cobo-Calvo, Alvaro, Andreasson, Ulf, Blennow, Kaj, Romero-Pinel, Lucia, Mart´ınez-Y´elamos, Sergio, Zetterberg, Henrik, Glial and neuronal markers in cerebrospinal fluid in different types of multiple sclerosis, Journal of Neuroimmunology (2016), doi: 10.1016/j.jneuroim.2016.08.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Title: Glial and neuronal markers in cerebrospinal fluid in different types of multiple sclerosis

T

Authors:

IP

M.Alba Mañé-Martíneza,b a

SC R

Joan XXIII University Hospital, Universitat Rovira i Virgili, Tarragona, Spain.

b

Bellvitge University Hospital, Universitat de Barcelona, L’Hospitalet de Llobregat,

Barcelona, Spain

NU

Bob Olssonc c

Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of

MA

Gothenburg, Möndal, Sweden Laura Baub b

D

Bellvitge University Hospital, Universitat de Barcelona, L’Hospitalet de Llobregat,

Elisabet Matasb b

TE

Barcelona, Spain

CE P

Bellvitge University Hospital, Universitat de Barcelona, L’Hospitalet de Llobregat,

Barcelona, Spain

Álvaro Cobo-Calvob b

AC

Bellvitge University Hospital, Universitat de Barcelona, L’Hospitalet de Llobregat,

Barcelona, Spain

Ulf Andreassonc c

Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of

Gothenburg, Möndal, Sweden Kaj Blennowc c

Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of

Gothenburg, Möndal, Sweden Lucia Romero-Pinelb b

Bellvitge University Hospital, Universitat de Barcelona, L’Hospitalet de Llobregat,

Barcelona, Spain

ACCEPTED MANUSCRIPT Sergio Martínez-Yélamosb* b

Bellvitge University Hospital, Universitat de Barcelona, L’Hospitalet de Llobregat,

Barcelona, Spain

T

Henrik Zetterbergc,d*

IP

c

Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of

SC R

Gothenburg, Möndal, Sweden d

Institute of Neurology, University College London, United Kingdom

NU

*These authors contributed equally to the manuscript. Corresponding author: M.Alba Mañé Martínez

MA

Unitat de Esclerosis Múltiple “EMxarxa”. Servei de Neurologia. Hospital Universitari de Bellvitge. Edif. Tècnic-quirúrgic. Planta 1 Mòdul E.

D

Freixa Llarga s/n. PC 08907. L’Hospitalet de Llobregat. Barcelona. Spain

TE

Tel 932607412. Fax 932607778.

CE P

@: [email protected]

Abbreviations

AC

Aβ38, Aβ40, Aβ42: 38, 40 and 42 amino acid long fragments of amyloid β; α-sAPP: αcleaved soluble amyloid-precursor protein; β-sAPP: β-cleaved soluble amyloidprecursor protein; CIS: Clinically isolated syndrome; CNS: Central nervous system; CSF: Cerebrospinal fluid; EDMUS: European database for multiple sclerosis; EDSS: Expanded disability status scale; GFAP: glial fibrillary acidic protein; IDIBELL: Bellvitge Biomedical Research Institute; MCP-1: monocyte chemoattractant protein; MS: multiple sclerosis; NFL: Neurofilament light; OPLS-DA: orthogonal projection to latent structures discriminant analysis; p-tau: tau phosphorylated at threonine 181; PPMS: primary progressive multiple sclerosis; RRMS: Relapsing remitting multiple sclerosis; SPMS: secondary progressive multiple sclerosis; S-100B: S-100B protein; t-tau: total tau protein; YKL-40: human chitinase 3-like 1 protein.

ACCEPTED MANUSCRIPT Glial and neuronal markers in cerebrospinal fluid in different types of multiple sclerosis 1. Introduction

T

Multiple sclerosis (MS) is classically described as a demyelinating disease of the

IP

central nervous system (CNS) due to the histological findings of white matter lesions in

SC R

brains. However, there is evidence that not only myelin and oligodendrocytes are implied in MS pathogenesis, but also neuronal damage and astroglial activation (Eng and Ghirnikar,1994; Bonneh-Barkay et al., 2010; Trapp et al., 1998; Filippi et al., 2003).

NU

Clinically, MS is considered a chronic autoimmune condition. Even though, the involvement of the immune system has been thoroughly described in MS, there is also

MA

evidence that diffuse neurodegenerative processes takes part in the MS pathogenesis from the early stages of the disease (Frischer et al., 2009). Probably, the

D

predominance of autoimmune activation at the onset of the disease explains the

TE

inflammatory course of the relapsing-remitting forms of MS. While the predominance of neuronal and glial degeneration would be associated with progressive forms of MS.

CE P

Ideally, knowing the underlying histopathological process in each MS patient could lead to an individual treatment for a specific target. Notwithstanding, it is not feasible to

AC

obtain brain biopsies. Thus, MS patients are classified by their clinical course as clinically isolated syndrome (CIS), relapsing-remitting MS (RRMS), secondary progressive MS (SPMS) and primary progressive MS (PPMS), and they are treated according to the international standards. Cerebrospinal fluid (CSF) is the closest body fluid to brain tissue (Blennow et al., 2010). Therefore, CSF from a diagnostic lumbar puncture could reflect the features of brain damage at MS onset and could differentiate the different patterns of MS. The purpose of the present study was to investigate glial and neuronal biomarkers in CSF samples from patients with different types of MS and to test whether a correlation among the biomarkers exists and whether the profile of CSF biomarkers varies among

ACCEPTED MANUSCRIPT the different types of MS. Hence, we analysed biomarkers related to axonal damage (neurofilament light protein: NFL) (Trapp et al., 1998), neuronal injury (total-tau: t-tau and tau phosphorylated at threonine 181: p-tau) (Binder et al., 1985), glial activation

IP

T

(human chitinase 3-like 1 protein: YKL-40 and monocyte chemoattractant protein: MCP-1) (Bonneh-Barkay et al., 2010; Van Der Voorn et al., 1999), astrocytic damage

SC R

(glial fibrillary acidic protein: GFAP and S-100B protein: S-100B) (Eng and Ghirnikar, 1994; Donato, 2001), and amyloid metabolism (α-cleaved soluble amyloid-precursor

NU

protein: α-sAPP; β-cleaved soluble amyloid-precursor protein: β-sAPP; 38, 40 and 42 amino acid long fragments of amyloid β: Aβ38, Aβ40, Aβ42) (Gehrmann et al., 1995;

MA

Ferguson et al., 1997). 2. Materials and Methods

D

2.1 Patients and clinical assessments

TE

The present observational study was approved by the ethics committee of Bellvitge

CE P

University Hospital, L’Hospitalet de Llobregat, Spain and informed consent was obtained from all patients. Samples were obtained from the Bellvitge Biomedical Research Institute (IDIBELL), the CSF biobank MS Unit collection. Samples were

AC

matched with clinical data from patients recruited and prospectively followed at the MS Unit, Bellvitge University Hospital. All clinical data were entered into the European Database for Multiple Sclerosis (EDMUS) (Confavreux et al., 1992). Patients were diagnosed according to the Poser and McDonald criteria as appropriate and were classified as having CIS, RRMS, SPMS or PPMS according to the disease course when the lumbar puncture was performed (Poser et al., 1983; McDonald et al., 2001; Polman et al., 2005). The cohort of CIS and RRMS patients collective in the present study was the same as was evaluated in the previous published study (Mañé-Martínez et al., 2015). For the inclusion into the relapsing group, the first signs of relapse had to start within one month of sampling. The neurological deficits were scored with the Expanded Disability Status Scale (EDSS) (Kurtzke, 1983).

ACCEPTED MANUSCRIPT 2.2 CSF sampling and biochemical analyses CSF samples were collected in 109 CIS patients, 192 RRMS patients, 6 SPMS patients

T

and 17 PPMS patients by lumbar puncture into polypropylene tubes, centrifuged at

IP

2200 x g for 10 min, aliquoted into 1 mL cryo tubes that were stored at -80ºC pending

SC R

analyses. Concentrations of CSF biomarkers were analysed in blind fashion at the Clinical Neurochemistry Laboratory, Institute of Neuroscience and Physiology, University of Gothenburg, using enzyme-linked immunosorbent assays as described

NU

(Mañé-Martínez et al., 2015).

MA

2.3 Statistics

Continuous variables were described by their mean and standard deviation or median and interquartile range, depending on their distribution, and categorical variables by

D

numbers and percentages. Differences between groups were analysed with KruskalWallis test followed by pairwise post hoc comparisons using the Mann-Whitney U test,

TE

p values ≤ 0.05 were considered significant. Spearman’s test was used to analyse correlations between demographics and CSF biomarker concentrations, p values ≤

CE P

0.05 were considered significant. Correlations among different biomarkers were analysed using age-adjusted Spearman’s test and p values were adjusted using Bonferroni (Holms) correction. Univariate statistical analyses were prepared using

AC

Statistical Package for the Social Sciences 20.0 (SPSS Inc, Chicago, IL). Multivariate analysis was performed to find differences between the relapsing and remitting phases, using orthogonal projection to latent structures discriminant analysis (OPLS-DA) implemented in the software SIMCA-P+ v. 12 (Umetrics, Umea, Sweden) (Andreasson et al., 2012). The OPLS-DA algorithm finds the projection direction, score vector, that gives the largest covariance between the variables and the pre-defined classes (ie relapsing and remitting phases) and that maximizes the separation between the classes. The variables that are found to have an influence on the projection (VIP: Variable importance on the projection plots) and that contribute to discriminate between the classes are summarized in the VIP plot. The higher the VIP bar, the more influential is the variable on the model. The VIP plot also gives a 95% confidence interval (CI) for the contribution of each variable, and a large inaccuracy (Hjalmarsson et al., 2014). 3. Results

ACCEPTED MANUSCRIPT Demographics and clinical characteristics of MS patients are shown in Table 1. 3.1 CSF biomarker concentrations and demographics

T

The correlation between age at lumbar puncture and CSF biomarker concentrations

IP

was evaluated in a total of 324 patients. A positive correlation was found between age

SC R

and GFAP, YKL-40, S-100B, p-tau, α-sAPP and β-sAPP , whereas a negative one was found for NFL (Spearman’s test correlation index: NFL: - 0.21, p < 0.0001; GFAP: 0.24, p < 0.0001; YKL-40: 0.16, p = 0.004; S-100B: 0.14, p = 0.01; p-tau: 0.15, p =

NU

0.007; α-sAPP: 0.16, p = 0.004; β-sAPP: 0.15, p = 0.005); no significant correlation was found for MCP-1, t-tau, Aβ38, Aβ40 and Aβ42. Similar correlations were found in

MA

the subgroup of CIS and RRMS patients (n = 301) (Spearman’s test correlation index: NFL: - 0.16, p = 0.003; GFAP: 0.21, p < 0.0001; YKL-40: 0.14, p = 0.02; S-100B: 0.15,

D

p = 0.01; p-tau: 0.14, p = 0.02; α-sAPP: 0.12, p = 0.04; β-sAPP: 0.12, p = 0.03). There

TE

were no significant correlations between biomarker concentrations and the duration of the disease at LP time. In relation to gender, significantly higher CSF concentrations of

CE P

MCP-1, YKL-40, GFAP, p-tau, α-sAPP, β-sAPP, Aβ38, Aβ40, Aβ42 were found in males vs females (Table 2). The subgroup of relapsing-remitting patients (n = 301)

AC

showed similar data except for YKL-40 (males: 105 ng/mL (80-165); females 94 ng/mL (67-146), p = 0.04) and GFAP (no significant differences in CSF concentrations between males and females), while no differences were found in the PPMS subgroup. There were no significant correlations between the EDSS at the time of the lumbar puncture and CSF biomarker concentrations except for YKL-40 (CI: 0.17, p = 0.002); αsAPP (CI: -0.15, p = 0.006) and β-sAPP (CI: -0.14, p = 0.01). Correlation results for EDSS were similar in the subgroup of relapsing-remitting forms of MS. 3.2 CSF biomarker concentrations in different types of MS Table 2 and Figure 1 show the CSF biomarker concentrations for each type of MS. NFL concentrations were significantly higher in relapsing-remitting forms, while GFAP

ACCEPTED MANUSCRIPT concentrations were significantly higher in progressive forms. YKL-40 concentrations were higher in the PPMS group compared to the CIS group (Table 2 and Figure 1).

T

3.3 CSF biomarker concentrations and relapses

IP

Patients from CIS and RRMS groups were evaluated together and were classified to be

SC R

in the relapsing (n = 138) or remitting (n = 162) phase. CSF concentrations of GFAP, MCP-1, t-tau, p-tau, α-sAPP, β-sAPP, Aβ38, Aβ40, Aβ42 were significantly lower during relapse vs remission (Table 2). There were no differences in YKL-40 and S-

NU

100B concentrations between the relapsing and remitting stages (Table 2). CSF NFL concentrations were significantly increased only when the relapse period was extended

MA

to three months prior to sampling (relapse: 1455 ng/L, 677 – 2670; remission: 830 ng/L, 520 – 1835; p = 0.009). No differences were seen for t-tau, p-tau, GFAP, S-100B, YKL-

D

40, MCP-1, α-sAPP, β-sAPP, Aβ38, Aβ40 and Aβ42 in the subanalysis of the relapse

TE

period. Hence, the difference in CSF biomarker concentrations between the relapsing and remitting phase was larger for GFAP, MCP-1 and NFL, while smaller for YKL-40

CE P

(Figure 2). Patients of the CIS group were classified depending on their first neurological relapse: long tracts, brainstem, spinal cord and optic neuritis. No

AC

association was found between biomarker concentrations and the type of clinical presentation at the time of first relapse (data not shown). 3.4 Correlations among glial markers In RRMS and CIS patients, significant positive correlations were observed between GFAP and YKL-40 (CI = 0.44, p = 0.006) (Figure 3), between GFAP and MCP-1 (CI = 0.38, p = 0.006) and GFAP and S-100B (CI = 0.31, p = 0.006). In progressive forms of MS no significant correlations among glial biomarkers were found. 3.5 Correlations among neuronal markers There was a weak but significant positive correlation between NFL ant t-tau (CI = 0.41, p = 0.006), while no correlation was found between NFL and p-tau. P-tau and t-tau had

ACCEPTED MANUSCRIPT significant positive correlations in all groups, the strongest in the PPMS group (CI = 0.90, p = 0.006).

T

3.6 Correlations between glial and neuronal markers

IP

Significant positive correlations were observed between YKL-40 and NFL, the

SC R

strongest correlation in the CIS group (CI = 0.59, p = 0.006) (Figure 3). GFAP correlated weakly but significantly positive with NFL in RRMS and CIS patients (CI 0.29, p = 0.006). No correlations were found between MCP-1 and NFL. In relapsing-

NU

remitting forms of MS, t-tau and p-tau significantly correlated with GFAP (t-tau: CI = 0.48, p = 0.006; p-tau: CI = 0.33, p = 0.006) and MCP-1 (t-tau: CI = 0.22, p = 0.004; p-

MA

tau: CI = 0.21, p = 0.03), as well as t-tau with YKL-40 (CI = 0.34, p = 0.005), while no significant correlations were found between p-tau and YKL-40. In PPMS patients, no

D

significant correlations were found between GFAP, YKL-40 or MCP-1 and NFL, t-tau

TE

and p-tau.

CE P

3.7 Correlations among neuronal, glial and amyloidal biomarkers No correlations were found between NFL and amyloidal metabolism biomarkers. While t-tau and p-tau showed significant positive correlations with α-sAPP and β-sAPP (t-tau:

AC

CI > 0.21 p = 0.02, p-tau: CI > 0.45, p = 0.004 for all correlations) and with Aβ38, Aβ40 and Aβ42 (t-tau: CI > 0.37 p = 0.005, p-tau: CI > 0.65, p = 0.004 for all correlations). No correlations were found between YKL-40 and amyloidal biomarkers. Furthermore, correlations between GFAP or MCP-1 and amyloidal markers were weak and only significant in the RRMS group (data not shown). Concentrations of α-sAPP and βsAPP correlate strongly (CI = 0.93, p = 0.003). Aβ38, Aβ40 and Aβ42 correlate strongly amongst themselves (CI > 0.92, p = 0.003 for all correlations) and with α-sAPP (CI > 0.47, p < 0.003 for all correlations) and β-sAPP (CI > 0.51, p < 0.003 for all correlations). 4. Discussion

ACCEPTED MANUSCRIPT 4.1 CSF biomarker profile in different types of multiple sclerosis In the present study we analysed 12 different biomarkers related to neurons, glia and

T

amyloid metabolism in 324 CSF diagnostic samples of different types of MS. The CSF

IP

biomarker profile was similar among different types of MS except for NFL, which

SC R

showed higher concentrations in relapsing-remitting forms of MS, and GFAP and YKL40 which showed higher concentrations in progressive forms of MS. These results are in agreement with previously published data (Teunissen et al., 2009; Axelsson et al.,

NU

2011). The CIS and RRMS groups showed the same profile of biomarkers. Our cohort of CIS patients included not only patients that presented a first neurological episode

MA

suggestive of MS but also fulfilled MR Barkhof criteria (Barkhof et al., 1997). 4.2 CSF biomarkers and the influence of clinical relapses.

D

Although clinical relapses do not represent all CNS inflammatory events, we

TE

considered clinical relapses as an objective and feasible marker of acute inflammation.

CE P

Thus, we evaluated the influence of the relapses in the CSF biomarker concentrations. NFL showed the highest concentrations in the relapsing group when the relapse period was extended to 3 months prior to lumbar puncture. This finding is in agreement with

AC

previous reports and could be explained because the inflammatory process, first provokes demyelination and is later followed by axonal degeneration that ends with the release of NFL. Thus, there is a delay in the increase of CSF NFL concentrations (Malmeström et al., 2003). This time pattern was not observed in any other biomarker. In contrast, MCP-1 showed increased concentrations in the stable stages. There is evidence that in MS lesions, MCP-1 is mainly expressed by reactive astrocytes and its overexpression mediates the recruitment of brain resident immune cell and the recruitment of infiltrating monocytes from the systemic bloodstream to the lesion sites as an initial mechanism of inflammatory response to a brain injury (Conductier et al., 2010). However, the sustained release of MCP-1 prolongs the inflammation and provokes cytotoxicity (Conductier et al., 2010). Therefore, we hypothesize that CSF

ACCEPTED MANUSCRIPT concentrations of MCP-1 increase during the relapse period as a response to tissue damage and they continue increasing beyond the period of time considered as clinical relapse. Furthermore, high concentrations of MCP-1 could be associated with

IP

T

increased risk of a next relapse. GFAP, tau proteins and amyloidal markers also showed increased concentrations during the stable stages, probably as a product of

SC R

cell death after severe inflammatory processes or continuous neurodegeneration as explained above. YKL-40 and S-100B were the biomarkers that were least influenced

NU

by relapses as previously reported (Mañé-Martínez et al., 2015). Some studies have proposed the regulation of inflammation and the inhibition of apoptosis as the main

MA

roles of YKL-40 in inflammatory diseases (Bonneh-Barkay et al., 2010; Prakash et al., 2013). Moreover, YKL-40 seems to be involved in tissue remodelling (Prakash et al., 2013). Thus, the biological functions of YKL-40 could explain its continuous release by

TE

D

activated astrocytes in MS, independently of the MS stage. The influence of relapses on CSF biomarker concentrations should be confirmed in future studies, since could

CE P

affect clinical trials and MS studies. 4.3 CSF biomarkers and demographics

AC

The influence of age in CSF biomarker concentrations in healthy controls and MS patients has been described with controversial results, probably because the influence of the disease on biomarker concentrations is stronger than the influence of the age (Bonneh-Barkay et al., 2010; Mañé-Martínez et al., 2015; Malmeström et al., 2003; Kuhle et al., 2013). In the age range of our study, we observed weak correlations between age and the different biomarkers, being still less significant in relapsingremitting forms of MS. However, neither older nor younger ages have been explored. Therefore, we observed a negative correlation between NFL concentrations and age, while GFAP showed a positive correlation with age. This could be explained because younger patients were those in the early stages of the relapsing-remitting forms of MS and they probably were in a more inflammatory phase of the disease, while older

ACCEPTED MANUSCRIPT patients were those in the progressive forms of MS. Herein, the results of correlations among CSF biomarker concentrations were age-adjusted, although the differences on Spearman’s rho regarding raw data were lower than 0.05 for all biomarkers. In relation

IP

T

to gender, glial and amyloidal markers showed higher concentrations in males, however these differences were less significant in the relapsing-remitting group, except

SC R

for MCP-1 which showed significantly higher concentrations in males. We hypothesize that these differences could be related to the higher prevalence of more severe MS

NU

forms in males associated with a prolonged inflammation and neurodegeneration (Ribbons et al., 2015). Contrary to what we expected, the influence of the disability rate

MA

on the CSF biomarker concentrations was weak. This could be because CSF biomarker concentrations reflect more cumulative underlying pathological processes than disability scores at any given time point, especially if the scores are focused on

TE

D

motor tasks such as EDSS (Kurtzke, 1983). CSF samples were obtained close to the time of MS diagnosis. Only 11 patients (7 RRMS and 4 SPMS) were under treatment at

CE P

the time of lumbar puncture. Although we can not exclude that the treatment influenced biomarker concentrations in SPMS group, we do not consider that treatment influenced

AC

the results of the whole study. 4.4 Correlations among different biomarkers Looking at the correlations among different biomarkers, we have to consider that some biomarkers are products of axonal degeneration (NFL), astrogliosis (GFAP) or glial activation (YKL-40 and MCP-1), products that are released into the environment and the CSF; while other markers like tau proteins and amyloid metabolism products aggregate and their influx to CSF is reduced. This could explain stronger correlations among glial markers than those between axonal and neuronal markers. Another factor to take into account is the size of the cohorts, which could explain, for instance, that no correlations were found among glial markers in progressive forms of MS. Nevertheless, in spite of these limitations, interesting correlations were found. Thus, the

ACCEPTED MANUSCRIPT concentrations of YKL-40 correlated significantly with the NFL concentrations, especially in the CIS group. Both markers are associated with early immune processes where YKL-40 is secreted by activated astrocytes that regulate inflammation and NFL

IP

T

are a product of axonal degradation. This finding is in agreement with previous studies, which showed that CIS patients with high concentrations of both NFL and YKL-40 were

SC R

associated with a shorter time before a second relapse (Mañé-Martínez et al., 2015; Comabella et al., 2010). Moreover, significant correlations between YKL-40 and GFAP

NU

were observed in CIS and RRMS patients, which were expected since YKL-40 is released by activated astrocytes and GFAP is a product of mature astrocyte

MA

degradation after immune responses. Furthermore, this finding is in agreement with our previous work which showed that high concentrations of both YKL-40 and GFAP were independent risk factors for disability progression in RRMS patients (Mañé-Martínez et

5. Conclusion

TE

D

al., 2015).

CE P

The present study is the first that evaluates a variety of biomarkers from the same CSF sample in patients with different MS types and analyses of demographic, clinical and

AC

biological data. We observed higher CSF concentrations of NFL and a strong correlation between YKL-40 and NFL and between YKL-40 and GFAP in relapsingremitting forms of MS.

ACKNOWLEDGEMENTS The authors thank Dr. Txomin Arbizu for setting up the MS Unit and Susana Pobla Müller, Ana González, Nuria Iranzo Papiol, Mª Isabel León Moreno and Mª Teresa Anguix Bricio for their technical support in the biobank and MS Unit. We also thank Monica Christiansson, Dzemila Secic and Jonas Söderblom for excellent technical assistance in laboratory work. We would also like to acknowledge the patients for their

ACCEPTED MANUSCRIPT participation. This research would not have been possible without them. M.Alba Mañé Martínez received research support from the Fundació Hospital Universitari de Tarragona Joan XXIII and Fundació Institut d’Investigació Biomèdica de Bellvitge

SC R

IP

T

(IDIBELL).

REFERENCES

NU

Andreasson, U., Portelius, E., Pannee, J., Zetterberg, H., Blennow, K., 2012.

MA

Multiplexing and multivariate analysis in neurodegeneration. Methods. 56, 464-470.

Axelsson, M., Malmeström, C., Nilsson, S., Haghighi, S., Rosengren, L., Lycke, J., 2011. Glial fibrillary acidic protein: a potential biomarker for progression in multiple

TE

D

sclerosis. J. Neurol. 258, 882-888.

CE P

Barkhof, F., Filippi, M., Miller, DH., Scheltens, P., Campi, A., Polman, CH., Comi, G., Adèr, HJ., Losseff, N., Valk, J., 1997. Comparison of MRI criteria at first presentation to

AC

predict conversion to clinically definite multiple sclerosis. Brain. 120, 2059-2069.

Blennow, K., Hampel, H., Weiner, M., Zetterberg, H., 2010. Cerebrospinal fluid and plasma biomarkers in Alzheimer disease. Nat. Rev. Neurol. 6, 131-144.

Binder, LI., Frankfurter, A., Rebhun, LI., 1985. The distribution of tau in the mammalian central nervous system. J. Cell. Biol. 101, 1371-1378.

Bonneh-Barkay, D., Wang, G., Starkey, A., Hamilton, RL., Wiley, CA., 2010. In vivo CHI3L1 (YKL-40) expression in astrocytes in acute and chronic neurological diseases. J. Neuroinflammation. 7, 34.

ACCEPTED MANUSCRIPT Comabella, M., Fernández, M., Martin, R., Rivera-Vallvé, S., Borrás, E., Chiva, C., Julià, E., Rovira, A., Cantó, E., Alvarez-Cermeño, JC., Villar, LM., Tintoré, M., Montalban, X., 2010. Cerebrospinal fluid chitinase 3-like 1 levels are associated with

IP

T

conversion to multiple sclerosis. Brain. 133, 1082-1093.

SC R

Conductier, G., Blondeau, N., Guyon, A., Nahon, JL., Rovère, C., 2010. The role of monocyte chemoattractant protein MCP1/CCL2 in neuroinflammatory diseases. J.

NU

Neuroimmunol. 224, 93-100.

MA

Confavreux, C., Compston, DA., Hommes, OR., McDonald, WI., Thompson, AJ., 1992. EDMUS, a European database for multiple sclerosis. J. Neurol. Neurosurg. Psichiatry.

TE

D

55, 671-676.

Donato, R., 2001. S100: a multigenic family of calcium-modulated proteins of the EF-

33, 637-638.

CE P

hand type with intracellular and extracellular functional roles. Int. J. Biochem Cell Biol.

AC

Eng, LF., Ghirnikar, RS., 1994. GFAP and astrogliosis. Brain Pathol. 4, 229-237.

Ferguson, B., Matyszak, MK., Esiri, MM., Perry, VH., 1997. Axonal damage in acute multiple sclerosis lesions. Brain. 120, 393-399.

Filippi, M., Bozzali, M., Rovaris, M., Gonen, O., Kesavadas, C., Ghezzi, A., Martinelli, V., Grossman, RI., Scotti, G., Comi, G., Falini, A., 2003. Evidence for widespread axonal damage at the earliest clinical stage of multiple sclerosis. Brain. 126, 433-437.

ACCEPTED MANUSCRIPT Frischer, JM., Bramow, S., Dal-Bianco, A., Lucchinetti, CF., Rauschka, H., Schmidbauer, M., Laursen, H., Sorensen, PS., Lassmann, H., 2009. The relation between inflammation and neurodegeneration in multiple sclerosis brains. Brain. 132,

IP

T

1175-1189.

SC R

Gehrmann, J., Banati, RB., Cuzner, ML., Kreutzberg, GW., Newcombe, J., 1995. Amyloid precursor protein (APP) expression in multiple sclerosis lesions. Glia. 15, 141-

NU

151.

MA

Hjalmarsson, C., Bjerke, M., Andersson, B., Blennow, K., Zetterberg, H., Aberg, D., Olsson, B., Eckerström, S., Bokemark, L., Wallin, A., 2014. Neuronal and glia-related biomarkers in cerebrospinal fluid of patients with acute ischemic stroke. J Cent Nerv

TE

D

Syst Dis. 6, 51-68.

CE P

Kuhle, J., Plattner, K., Bestwick, JP., Lindberg, RL., Ramagopalan, SV., Norgren, N., Nissim, A., Malaspina, A., Leppert, D., Giovannoni, G., Kappos, L., 2013. A comparative study of CSF neurofilament light and heavy chain protein in MS. Mult

AC

Scler. 19, 1597-1603.

Kurtzke, JF., 1983. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology. 33, 1444-1452.

Malmeström, C., Haghighi, S., Rosengren, L., Andersen, O., Lycke, J., 2003. Neurofilament light protein and glial fibrillary acidic protein as biological markers in MS. Neurology. 23, 1720-1725.

Mañé-Martínez, MA., Olsson, B., Bau, L., Matas, E., Cobo-Calvo, A., Andreasson, U., Blennow, K., Romero-Pinel, L., Martínez-Yélamos, S., Zetterberg, H., 2015. Glial and

ACCEPTED MANUSCRIPT neuronal markers in CSF predicts progression in multiple sclerosis. Mult Scler. 21, 550561.

IP

T

McDonald, WI., Compston, A., Edan, G., Goodkin, D., Hartung, HP., Lublin, FD., McFarland, HF., Paty, DW., Polman, CH., Reingold, SC., Sandberg-Wollheim, M.,

SC R

Sibley, W., Thompson, A., Van Den Noort, S., Weinshenker, BY., Wolinsky, JS., 2001. Recommended diagnostic criteria for multiple sclerosis: guidelines from the

NU

International Panel on the diagnosis of multiple sclerosis. Ann. Neurol. 50, 121-127.

MA

Polman, CH., Wolinsky, JS., Reingold, SC., 2005. Multiple sclerosis diagnostic criteria: three years later. Mult Scler. 11, 5-12.

TE

D

Poser, CM., Paty, DW., Scheinberg, L., McDonald, WI., Davis, FA., Ebers, GC., Johnson, KP., Sibley, WA., Silberberg, DH., Tourtellotte, WW., 1983. New diagnostic

231.

CE P

criteria for multiple sclerosis: guidelines for research protocols. Ann. Neurol. 13, 227-

AC

Prakash, M., Bodas, M., Prakash, D., Nawani, N., Khetmalas, M., Mandal, A., Eriksson, C., 2013. Diverse pathological implications of YKL-40: answers may lie in “outside-in” signalling. Cell Signal. 25, 1567-73.

Ribbons, KA., McElduff, P., Boz, C., Trojano, M.,, Izquierdo, G., Duquette, P., Girard, M., Grand’Maison, F., Hupperts, R., Grammond, P., Oreja-Guevaro, C., Petersen, T., Bergamaschi, R., Giuliani, G., Barnett, M., Van Pesch, V., Amato, MP., Iuliano, G., Fiol, M., Slee, M., Verheul, F., Cristiano, E., Fernandez-Bolanos, R., Saladino, ML., Rio, ME., Cabrera-Gomez, J., Butzkueven, H., Van Munster, E., Den Braber-Moerland, L., La Spitaleri, D., Lugaresi, A., Shaygannejad, V., Gray, O., Deri, N., Alroughani, R., Lechner-Scott,J., 2015. Male sex is independently associated with faster disability

ACCEPTED MANUSCRIPT accumulation in relapse-onset MS but not in primary progressive MS. Plos One. 10, e0122686.

IP

T

Teunissen, CE., Iacobaeus, E., Khademi, M., Brundin, L., Norgren, N., Koel-Simmelink, MJ., Schepens, M., Bouwman, F., Twaalfhoven, HA., Blom, HJ., Jakobs, C., Dijkstra,

SC R

CD., 2009. Combination of CSF N-acetylaspartate and neurofilaments in multiple

NU

sclerosis. Neurology. 72, 1322-1329.

Trapp, BD., Peterson, J., Ransohoff, RM., Rudick, R., Mörk, S., Bö, L., 1998. Axonal

MA

transection in the lesions of multiple sclerosis. N Engl. J. Med. 338, 278-85.

Van Der Voorn, P., Tekstra, J., Beelen, RH., Tensen, CP., Van Der Valk, P., De Groot,

TE

D

CJ. 1999. Expression of MCP-1 by reactive astrocytes in demyelinating multiple

FOOTNOTES

AC

Figure 1

CE P

sclerosis lesions. Am. J. Pathol. 154, 45-51.

Title: Cerebrospinal fluid biomarker concentrations in different types of multiple sclerosis.

Footnote: Box plot depicts concentrations of NFL (A), GFAP (B) and YKL-40 (C) in different types of multiple sclerosis (MS). Median, interquartile range (IQR); whiskers represent minimum and maximum 1.5 IQR. Patients were categorized into different groups according to the disease course when the lumbar puncture was performed. CIS: clinically isolated syndrome, RRMS: relapsing-remitting MS, SPMS: secondary progressive MS, PPMS: primary progressive MS. Significant differences in CSF

ACCEPTED MANUSCRIPT biomarker concentrations between groups are shown. Maximum extreme values are

T

not shown above, hence NFL (CIS: 28.380 ng/L; RRMS: 35.560 ng/L).

IP

Figure 2

SC R

Title: Importance of CSF biomarkers to discriminate active relapse from stable phase in relapsing-remitting MS.

NU

Footnote: Variable importance on the projection plots. The graph shows the relative contribution of the biomarkers to discriminate between patients in active relapse vs

MA

stable phases in the relapsing-remitting forms of MS group. The error bars represent

D

95% confidence intervals.

TE

Figure 3

CE P

Title: Correlations among biomarker concentrations. Footnote: Correlations between NFL and YKL-40 (A) and YKL-40 and GFAP (B) in

AC

CIS (clinically isolated syndrome), RRMS (relapsing-remitting multiple sclerosis), PPMS (primary progressive multiple sclerosis) and SPMS (secondary progressive multiple sclerosis) patients are shown. Correlations among different biomarkers were analysed using Spearman’s test and p values were adjusted using Bonferroni (Holms) correction.

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Figure 1

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Figure 2

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Figure 3

ACCEPTED MANUSCRIPT Table 1. Demographics and clinical characteristics of MS patients at baseline. CIS (n = 109)

b

Disease duration at LP, y, mean (SD) EDSS at LP, median (IQR)

c

d

Total (n = 324)

121 (63%)

4 (67%)

7 (41%)

207 (64%)

31.1 (9.8)

28.8 (8.8)

33.6 (13.9)

45.1 (11.3)

30.5 (10.1)

32 (9.8)

34.8 (8.9)

43.5 (10)

0.3 (0.5)

5.5 (6.2)

9.4 (7.1)

2 (0 - 2.0)

2 (1.0 – 2.5)

0

7

Patients under treatment before LP, n

T

Age at LP, y, mean (SD)

a

PPMS (n = 17)

75 (68%)

44.2 (23.5)

34.5 (10.8)

4.6 (3.1)

3.8 (5.5)

4.5 (4.0 – 6.0)

3 (2.0 – 3.5)

2 (1.0 – 2.5)

4

0

11

IP

Age at MS onset, y, mean (SD)

SPMS (n = 6)

SC R

Gender, Female, n (%)

RRMS (n = 192)

a

NU

Abbreviations: CIS = clinically isolated syndrome; EDSS = Expanded Disability Status Scale; IQR = interquartile range; LP = lumbar puncture; No = n = number of cases; PPMS = primary progressive multiple sclerosis; RRMS = relapsing-remitting multiple sclerosis; SD = standard deviation; SPMS = secondary progressive multiple sclerosis; y = years. At MS onset, CIS and RRMS patients were younger compared to PPMS (p < 0.0001 for both comparisons).

b

MA

At LP time, CIS patients were younger compared to RRMS (p = 0.001), SPMS (p = 0.01) and PPMS (p < 0.0001). RRMS patients were younger compared to SPMS (p = 0.03) and PPMS (p < 0.0001). c

Disease duration at lumbar puncture was shorter in CIS patients compared to RRMS, SPMS and PPMS patients (p < 0.0001 for all comparisons). d

AC

CE P

TE

D

EDSS at LP was lower in CIS vs RRMS (p = 0.004), SPMS (p < 0.0001) and PPMS (p < 0.0001). EDSS at LP was lower in RRMS vs SPMS (p < 0.0001) and PPMS (p = 0.001). EDSS was lower in PPMS vs SPMS (p = 0.01).

ACCEPTED MANUSCRIPT

GFAP (ng/L)c

YKL-40 (ng/mL)d

MCP-1 (pg/mL)

S100B (μg/L)

t-tau (pg/mL)

1165 (560-2305)

310 (210-450)

102 (72-152)

366 (286-453)

0.51 (0.32-0.67)

81 (75-134)

1150 (550-2080)

290 (200-430)

93 (67-146)

342 (278-415)

0.48 (0.31-0.67)

1190 (565-2545)

340 (235-480)

113 (82-169)

409 (319-509)

0.54 (0.36-0.71)

1150 (525-2425)

270 (190-400)

90 (66-147)

358 (264-444)

1330 (615-2312)

310 (220-457)

103 (74-152)

370 (292-462)

1040 (560-2300)

350 (260-470)

103 (80-147)

1440 (620-2660)

240 (150-370)

92 (65-152)

T

NFL (ng/L)b

MA N

Table 2. CSF biomarker levels at baseline. sAPPα (ng/mL)

sAPPβ (ng/mL)

Aβ38 (pg/mL)

Aβ40 (pg/mL)

Aβ42 (pg/mL)

260 (173-377)

99 (66-149)

1051 (788-1370)

7082 (5659-9284)

656 (478-877)

20 (15-28)

237 (170-355)

92 (62-148)

1010 (736-1340)

6925 (5096-9097)

638 (450-855)

93 (75-146)

23 (16-30)

289 (178-419)

112 (71-161)

1154 (869-1464)

7550 (6271-9811)

709 (544-915)

78 (75-128)

22 (16-29)

235 (172-366)

92 (63-140)

1020 (793-1383)

6952 (5883-9401)

630 (494-892)

0.49 (0.29-0.65)

83 (75-135)

21 (15-29)

264 (158-376)

100 (62-149)

1056 (746-1357)

7162 (5100-9176)

672 (441-897)

389 (323-485)

0.49 (0.31-0.67)

97 (75-145)

23 (17-30)

287 (197-455)

112 (77-172)

1113 (828-1396)

7548 (5910-9634)

710 (510-908)

322 (236-417)

0.53 (0.35-0.67)

75 (75-116)

18 (15-26)

222 (127-327)

87 (51-134)

987 (741-1281)

6721 (5137-8783)

583 (431-847)

All patients

21 (15-29)

US

N = 324

CR

IP

p-tau (pg/mL)

Femalea N = 207

TE D

Malea

RRMS

0.53 (0.37-0.69)

AC

N = 109

CE P

N = 117

CIS

75 (75-127)

N = 192 Remissione N = 162

Relapsee

ACCEPTED MANUSCRIPT

T

N = 138

455 (280-722)

92 (75-241)

387 (264-452)

0.38 (0.24-0.65)

122 (75-169)

550 (475-710)

420 (320-605)

140 (91-200)

377 (311-478)

0.52 (0.35-0.72)

82 (75-133)

N=6

323 (206-462)

122 (96-162)

1114 (858-1362)

7138 (6213-9029)

654 (569-830)

20 (16-28)

337 (254-428)

122 (76-176)

1047 (864-1415)

7663 (6245-9221)

712 (571-827)

MA N

N = 17

US

PPMS

a

28 (21-38)

CR

655 (317-1992)

IP

SPMS

Significantly higher CSF levels of MCP-1 (p < 0.0001), YKL-40 (p = 0.006), GFAP (p = 0.04), p-tau (p = 0.02), α-sAPP (p = 0.02), β-sAPP (p = 0.02), Aβ38 (p = 0.02), Aβ40 (p = 0.02), Aβ42 (p = 0.02) were

TE D

found in males vs females. b

c

d

CE P

CSF levels of NFL were significantly higher in RRMS patients compared to PPMS (p = 0.001) and in CIS compared to PPMS (p = 0.02). CSF levels of GFAP were lower in RRMS than PPMS (p = 0.03), in CIS

than PPMS (p = 0.006) and in CIS than SPMS (p = 0.03). CSF levels of YKL-40 were higher in the PPMS group vs RRMS (p = 0.06) and significantly higher in PPMS vs CIS (p = 0.02). No significant differences

AC

were found for any of the biomarkers between the RRMS and CIS groups, between RRMS and SPMS or between the SPMS and PPMS groups. MCP-1, S1000B, t-tau, p-tau, α-sAPP, β-sAPP and amyloidal cleavage proteins Aβ38, Aβ40, Aβ42 showed no significant differences among the different MS types. e

NFL levels were higher but not significantly during relapses (p = 0.06). Significant lower levels were found of GFAP, MCP-1, t-tau, p-tau, α-sAPP and β-sAPP during the relapsing vs remitting phase: p <

0.0001 for all comparisons. YKL-40 and S100B: Mann-Whitney U test not significant. Lower levels of Aβ38 (p = 0.02), Aβ40 (p = 0.03), Aβ42 (p = 0.009) during the relapsing vs remitting phase.

ACCEPTED MANUSCRIPT Abstract

T

In the present study, CSF concentrations of NFL, t-tau, p-tau, GFAP, S-100B, YKL-40,

IP

MCP-1, α-sAPP, β-sAPP, and Aβ38, Aβ40, Aβ42 were measured in 324 MS patients to

SC R

test whether a correlation among the biomarkers exists and whether the profile of CSF biomarkers varies among the different types of MS. The CSF concentrations of NFL were significantly higher in RRMS while CSF concentrations of GFAP were higher in

NU

PPMS. CSF concentrations of NFL correlated with YKL-40 in CIS patients while CSF

AC

CE P

TE

D

MA

concentrations of GFAP correlated with YKL-40 in RRMS patients.

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Graphical abstract

ACCEPTED MANUSCRIPT Highlights Glial and neuronal markers were analysed in 324 CSF samples of different types of MS

T

Neurofilament concentrations were higher in relapsing forms of MS.

IP

Glial fibrillary acidic protein concentrations were higher in progressive forms of MS.

SC R

CSF concentrations of NFL and YKL-40 correlated in CIS and RRMS patients.

AC

CE P

TE

D

MA

NU

CSF concentrations of GFAP and YKL-40 correlated in relapsing-remitting patients.