Better tissue biomarkers are needed to improve diagnosis and prognosis, guide molecularly targeted therapy, and monitor activity and therapeutic response across many cancers. Proteomics methods, based on mass spectrometry, hold great promise for the discovery of novel biomarkers that might form the foundation of a new clinical test. This review will focus on potential tissue biomarkers with utility for prognosis in breast cancer.

By Dr Liping Chung

Tissue biomarkers in breast cancer
Breast cancer is the leading cause of mortality among women worldwide. It is a complex and heterogeneous disease and includes several subtypes, which have different prognoses and responses to therapy. Recent molecular characterization of some breast cancer subtypes has led to the development of personalized options for treatment targeting [1].

One of the major advantages of biomarker research for individuals with cancer is the availability of tumour tissue for analysis and the possibility that potential tissue biomarkers can be detected in histological samples. In conjunction with tumour grading and measurement of lymphovascular invasion, several tissue biomarkers are now used with prognostic significance in daily practice including estrogen receptor (ER), progesterone receptor (PR), the type 2 epidermal growth factor receptor (HER2 or erbB-2), and Ki67 [1, 2].

The identification of protein biomarkers in easily accessible biological fluids has potential for the development of minimally invasive procedures for early diagnostics, but the analysis of body fluids such as plasma, serum and urine is complicated by their wide dynamic range of protein expression, the variation in their composition and their sensitivity to sample handling. Many serum biomarkers are not very specific or sensitive [1]. Analysis of tissue homogenates using the well-established and extremely powerful conventional techniques of differential proteomics has the advantage of covering the lower range of protein expression in such samples than in biological fluids [3].

Prognosis and response prediction
Different from diagnostic markers that detect the potential for developing a malignancy or test for the presence of a malignancy, biological markers that predict prognosis once a cancer has occurred are of great importance because they may influence major therapeutic recommendations. For breast cancer, these markers have become part of contemporary clinical practice. Among established tissue marker proteins in breast cancer, ER and HER2 are not diagnostic but have the greatest predictive utility [2]. It is generally accepted that estrogen receptor-positive (ER+) and ER-negative (ER−) breast cancers represent different disease entities. ER- tumours tend to be of high grade, have more frequent p53 mutations, and have worse prognosis compared with ER+ disease. Both ER+ and ER- tumours can be either HER2 positive or negative. Low-grade tumours are typically ER positive, and almost always HER2 non-amplified. The approximately 15% of patients with breast cancer who have HER2 overexpressing and amplified tumours are typically treated with a combination of trastuzumab, a monoclonal antibody targeting HER2, and adjuvant chemotherapy [4]. HER2 amplification and overexpression are generally associated with a poor prognosis. The prognostic significance of HER2 overexpression in tumour tissue has been evaluated in several clinical trials, suggesting that HER2 positivity is correlated with worse prognosis in untreated breast cancer patients, including node-negative populations [5].

The search for breast tissue biomarkers by mass spectrometry-based proteomics
Proteomic approaches, particularly those involving mass spectrometry (MS), have been widely used in breast cancer biomarker discovery, although to date no new markers based on proteomic discovery have been adopted for use in clinical practice. Using laser capture microdissection (LCM) for tissue samples, an extensive tissue study was performed by MALDI-MS (matrix-assisted laser desorption/ionization mass spectrometry) analysis on an average of 2000 cells from 122 invasive mammary carcinomas and 167 samples of normal breast epithelium [6]. Among clusters of protein/peptide peaks that were used to discriminate cancer from normal tissue with high sensitivity and specificity were ubiquitin, S100A6 (calcyclin) and S100A8 (calgranulin A). To confirm cDNA expression profiling of breast tissues, Brozkova et al. also analysed whole tissue lysates rather than serum of 105 breast carcinomas on IMAC30 protein chips by SELDI-TOF MS (surface-enhanced laser desorption/ionization, time-of-flight mass spectrometry) [7]. They compared this analysis to cDNA expression profiling of the same tumours and found similar clustering, providing supporting evidence for the effectiveness of this technique in identifying and classifying tumours.

Most clinical tissue samples are conserved as formalin-fixed paraffin-embedded (FFPE) samples. In particular, cancer tissues contain several different cell types at various developmental stages. It was generally believed that proteins in FFPE tissues were altered and inaccessible for analysis by mass spectrometry until recent developments have shown it is possible to access the protein in imaging mass spectrometry (IMS) experiments following antigen retrieval [8]. The direct analysis of cancer tissues by IMS preserves the spatial proteomic information. Consequently, it is holds great promise for the discovery of highly specific biomarkers. A recent study demonstrated the potential of MALDI-imaging MS for HER2 status of clinical parameters in cases of breast cancer based on protein patterns. This potentially allows the selection of patients likely to respond to trastuzumab treatment. Comparing the HER2-positive (HER2+) vs HER2-negative (HER2−) breast cancer protein profiles, the authors found a specific proteomic signature of seven species, able to accurately classify the HER2 status with a sensitivity of 83%, a specificity of 92% and an overall accuracy of 89% [9].

Protein biomarkers and conventional pathologic features
In a very recent study, using protein extracts of breast tissues (n=171), we have used SELDI-TOF MS to discover two proteins that, in combination, show high discrimination between breast cancer and healthy breast tissue samples [10]. These putative breast cancer biomarkers were verified on an independent sample set, and identified as ubiquitin and a novel truncated form of the S100 protein family member, S100P. Interestingly, the combined panel of two protein markers was significantly associated with tumour histologic grade, size, and lymphovascular invasion (LVI), and also with ER-positive (ER+) and PR-positive (PR+) status and HER2 overexpression. In particular, as shown in Figure 1, significant positive associations were seen between a previously unreported short form of S100P (9.2kDa) and tumour size, high grade, LVI and lymph node involvement (LN), and also associated with hormone receptor positive status and HER2 overexpression (unpublished data). These results implicate that a protein biomarker panel may indicate a HER2-enriched breast cancer subtype with poor prognosis, and that measurement of S100P may be valuable both in the classification of breast cancer and as a possible target for treatment. Furthermore, in another very recent study, the prognostic value of S100P was also tested for FFPE tissue obtained from 85 breast cancer patients with a median follow up of 17 years. High immunocytochemical staining of breast tumour sections for S100P has been associated with poor long-term patient survival [11].

Conclusion and future prospects
In this era of using new high-throughput methods, many new protein biomarkers have been reported for both prognostic and predictive purposes. However, none of these have been widely accepted in routine clinical practice, possibly due to a lack of sufficient validation to meet the criteria of the American Society of Clinical Oncology’s tumour marker utility grading system and guideline recommendations [1]. Identification of novel markers based on gene expression and proteomic profiling has led to more definitive insights into tumour biology. The accurate evaluation of the status of clinical parameters in cases of breast cancer is of primary importance for prognostic value and therapeutic decision. Different methodologies successfully used for breast cancer prognostic information and therapy outcome prediction may suggest that the future diagnostics and consequent individualization of therapy will become much more wide-ranging.

References
1. Harris L, Fritsche H, Mennel R, Norton L, Ravdin P, Taube S, Somerfield MR, Hayes DF, Bast RC, Jr. American Society of Clinical Oncology 2007 update of recommendations for the use of tumour markers in breast cancer. J Clin Oncol 2007; 25(33): 5287–5312.
2. Chung L, Baxter RC. Breast cancer biomarkers: proteomic discovery and translation to clinically relevant assays. Expert Rev Proteomics 2012; 9(6): 599–614.
3. Danova M, Delfanti S, Manzoni M, Mariucci S. Tissue and soluble biomarkers in breast cancer and their applications: ready to use? Journal of the National Cancer Institute Monographs 2011; 2011(43): 75–78.
4. Cheang MC, Chia SK, Voduc D, Gao D, Leung S, Snider J, Watson M, Davies S, Bernard PS, Parker JS, et al. Ki67 index, HER2 status, and prognosis of patients with luminal B breast cancer. Journal of the National Cancer Institute 2009; 101(10): 736–750.
5. Andrulis IL, Bull SB, Blackstein ME, Sutherland D, Mak C, Sidlofsky S, Pritzker KP, Hartwick RW, Hanna W, Lickley L, et al. neu/erbB-2 amplification identifies a poor-prognosis group of women with node-negative breast cancer. Toronto Breast Cancer Study Group. J Clin Oncol 1998; 16(4): 1340–1349.
6. Sanders ME, Dias EC, Xu BJ, Mobley JA, Billheimer D, Roder H, Grigorieva J, Dowsett M, Arteaga CL, Caprioli RM. Differentiating proteomic biomarkers in breast cancer by laser capture microdissection and MALDI MS. J Proteome Res 2008; 7(4): 1500–1507.
7. Brozkova K, Budinska E, Bouchal P, Hernychova L, Knoflickova D, Valik D, Vyzula R, Vojtesek B, Nenutil R. Surface-enhanced laser desorption/ionization time-of-flight proteomic profiling of breast carcinomas identifies clinicopathologically relevant groups of patients similar to previously defined clusters from cDNA expression. Breast Cancer Res 2008; 10(3): R48.
8. Gustafsson JO, Oehler MK, McColl SR, Hoffmann P. Citric acid antigen retrieval (CAAR) for tryptic peptide imaging directly on archived formalin-fixed paraffin-embedded tissue. J Proteome Res 2010; 9(9): 4315–4328.
9. Rauser S, Marquardt C, Balluff B, Deininger SO, Albers C, Belau E, Hartmer R, Suckau D, Specht K, Ebert MP, et al. Classification of HER2 receptor status in breast cancer tissues by MALDI imaging mass spectrometry. J Proteome Res 2010; 9(4): 1854–1863.
10. Chung L, Shibli S, Moore K, Elder EE, Boyle FM, Marsh DJ, Baxter RC. Tissue biomarkers of breast cancer and their association with conventional pathologic features. Br J Cancer 2013; 108(2): 351–360.
11. Maciejczyk A, Lacko A, Ekiert M, Jagoda E, Wysocka T, Matkowski R, Halon A, Gyorffy B, Lage H, Surowiak P. Elevated nuclear S100P expression is associated with poor survival in early breast cancer patients. Histol Histopathol 2013; 28(4): 513–524.

The author
Liping Chung PhD
Kolling Institute of Medical Research,
University of Sydney, Royal North Shore Hospital, NSW 2065, Australia
E-mail: liping.chung@sydney.edu.au

The discovery of reliable biomarkers, which are eligible for the prediction of both disease progression and response to treatment, means a great challenge in the management of multiple sclerosis (MS), a devastating disease of the central nervous system. The results of recent proteomic findings from the cerebrospinal fluid of MS patients hold promise of finding ideal biomarkers in the near future.

by Dr J. Füvesi, Dr C. Rajda, Dr D. Zádori, Dr K. Bencsik, Prof. Dr L. Vécsei and Prof. Dr J. Bergquist

Multiple Sclerosis
Multiple sclerosis is a demyelinative disorder of the central nervous system that affects mainly young adults. It has a great impact on quality of life, social and family life, and the careers of the patients.

In the majority of cases the disease starts with a relapsing–remitting (RR) phase. After a variable period of time it turns into a secondary progressive (SP) phase characterized by the gradual accumulation of residual symptoms. In 10–15% of cases a continuous progression is observed from the very beginning, this is the primary progressive (PP) form. In very rare fulminant cases frequent relapses with incomplete remissions cause severe disability or even death in a short duration of time.

The diagnosis of multiple sclerosis is still mainly clinical, supported by MRI and cerebrospinal fluid (CSF) analysis findings. The revised McDonald Criteria [1] allow earlier diagnosis, especially in PP MS. The routine diagnostic CSF analysis in MS includes the detection of oligoclonal bands and quantitative IgG analysis. Isoelectric focusing (IEF) on agarose gels followed by immunoblotting is considered the ‘gold standard’ for detecting the presence of oligoclonal bands [2]. The sensitivity of the method is above 95% and the specificity is more than 86%. An increased IgG index and the presence of oligoclonal bands in the CSF support an MS diagnosis.

Although the diagnosis is quite straightforward in most cases, taking into account clinical findings and paraclinical tests, there are still no specific biomarkers to confirm the diagnosis nor do we have any validated prognostic markers to follow the progression of the disorder.

At the time of diagnosis, major problems include the identification of the different clinical forms of the disease and the identification of patients with a potential rapid progression before disability evolves; the differential diagnosis of clinically isolated syndrome (CIS) with optic neuritis as the presenting symptom, where neuromyelitis optica (NMO) spectrum disorder may be an alternative diagnosis. Markers of disease progression are needed to distinguish CIS patients with a high probability to develop clinically definite MS.

There is also a need for biomarkers of response to treatment and biomarkers for better understanding the underlying pathological processes of the disease. This is especially important with the growing variety of treatment options: now it is possible to change therapy in the case of an inadequate treatment response and to escalate MS treatment to more aggressive alternatives. In the near future individualized treatment choices need better classification of patient characteristics.

In order to discover new biomarkers in MS, one should analyse the whole protein content of body fluids, preferentially CSF. Because of its proximity to the central nervous system (CNS), CSF may reflect changes in the CNS that may help differentiate normal and pathological conditions.

Proteomics
Proteomics is the study of protein expression in an organism. There are excellent reviews on proteomic approaches [3–5], so we will discuss here only certain aspects of these methods relevant to multiple sclerosis biomarker research. Mass-spectrometry (MS in Italic to distinguish from multiple sclerosis in this paper) based protein identification strategies include whole-protein analysis (‘top-down’ proteomics) and analysis of enzymatically produced peptides (‘bottom-up’ proteomics) [4]. The latter is the standard for large-scale or high-throughput analysis of highly complex samples, and digestion with trypsin is the most common method. The separation of peptides and proteins is an important element of both approaches.

Mass spectrometry measures the mass-to-charge ratio (m/z) of ionized molecules, and, as multiple distinct peptides can have very similar or identical molecular masses, it can be difficult to identify the overlapping peptides [3]. The use of separation techniques, therefore, reduces the cases of coincident peptide masses simultaneously introduced into the mass spectrometer. One of the most commonly used separation techniques is high-performance liquid chromatography (HPLC) with a capillary column. Peptides of similar molecular mass but different hydrophobicity elute from the LC column and enter the mass spectrometer at different time points, no longer overlapping in the initial MS analysis. Liquid chromatography coupled to mass spectrometry reduces the complexity of the sample and allows more precise protein identification.

In order to limit the risk of systematic errors and achieve a high sample throughput, labelling by means of isobaric tags for relative and absolute quantification (iTRAQ) may be used [6]. Multiple samples may be processed in parallel with this multiplexed approach. The main advantage is that the samples are analysed under exactly the same conditions. The relative abundance of labelled peptides indicates relative changes in protein expression.

LC-MS experiments generate an enormous amount of data, making data analysis one of the most challenging parts of proteomic analysis. Protein identification and quantification is achieved by database searching. Programs, such as Mascot etc., compare observed spectra to predicted spectra for candidate peptides from a protein database. In a recent study Schutzer et al. established a database of the normal human CSF proteome [7].

Proteomics in multiple sclerosis
In recent years a number of papers appeared describing proteomic analysis of CSF or brain tissue of multiple sclerosis patients [8–12]. The first papers in the field analysed pooled samples from a relatively small group of patients [8, 9]. Hammack et al. [8] reported the analysis of a pooled sample of three relapsing–remitting MS patients and a pooled sample of three patients with non-MS inflammatory CNS disorders using two-dimensional gel electrophoresis (2-DE) and peptide mass fingerprinting. They identified four proteins in the gels containing MS CSF that were not reported previously in normal human CSF: CRTAC-1B (cartilage acidic protein), tetranectin (a plasminogen-binding protein), SPARC-like protein (a calcium binding cell signalling glycoprotein) and autotaxin t (a phosphodiesterase).

In the study of Dumont et al. [9] CSF samples from five MS patients (4 RR, one SP) were analysed by 2-DE followed by liquid chromatography tandem mass spectrometry. With this method 15 proteins have been identified that were not previously observed in non-multiple sclerosis CSF 2-DE gels. These proteins were: psoriasin, calmodulin-related protein NB-1, annexin 1, EWI-2, Niemann–Pick disease type C2 protein (NPC-2), semenogelin 1 (SEM1), semenogelin 2 (SEM2), complement factor H-related protein 1 (FHR-1), procollagen C-proteinase enhancer protein (PCPE), aldolase A, N-acetyllactosaminide β-1,3-N-acetylglucosaminyl-transferase, tetranectin, cystatin A, superoxide dismutase 3 and glutathione peroxidase.

Later, publications started to focus on the differentiation of the clinical forms of the disease. Lehmensiek et al. compared CSF samples from RR MS and clinically isolated syndrome (CIS) patients with controls using two-dimensional difference gel electrophoresis (2-D-DIGE) and matrix-assisted laser desorption/ionization – time of flight (MALDI-TOF) mass spectrometry [10]. In RR MS Ig kappa chain NIG93 protein was increased in concentration, while transferrin isoforms, alpha 1 antitrypsin isoforms, alpha 2-HS glycoprotein, Apo E and transthyretin decreased. In a study of Stoop et al. [11] significant differences were observed comparing the peak lists of spectra from CSF of MS patients and patients with other neurological diseases (OND), and also clinically isolated syndrome (CIS) vs OND. Three differentially expressed proteins were identified in the CSF of MS patients compared to CSF of patients with OND: chromogranin A, clusterin and complement C3.

The same group compared proteome profiles of CSF from RR and PP multiple sclerosis and found that they overlap to a large extent [13]. The main detected difference was that protein jagged-1 was less abundant in PP MS compared to RR MS, whereas vitamin D-binding protein was only detected in the RR MS CSF samples. Ottervald et al. found an increased CSF level of vitamin-D-binding protein in SP MS compared to the control [14]. Recently, impaired vitamin D homeostasis has been linked to multiple sclerosis [15]: high serum levels of 25-hydroxyvitamin D correlated with a reduced risk of MS [16] and vitamin D supplementation was proposed as an add-on therapy [17].

Biomarkers of disease progression are emerging as new targets of proteomics. In our recently published paper we analysed the CSF of a rare fulminant case of MS and compared it with RR MS and control samples [18]. The aim of this study was to identify proteins related to rapid progression. The presented bottom-up strategy, based on isobaric tag labelling in conjunction with enzymatic digestion followed by nanoLC coupled off-line to MALDI TOF/TOF MS resulted in the identification of 78 proteins. Seven proteins were found to be upregulated in both fulminant MS samples but not in the relapsing–remitting case compared to the control. These proteins included Ig kappa and gamma-1 chain C region, complement C4-A, fibrinogen beta chain, serum amyloid A protein, neural cell adhesion molecule 1 and beta-2-glycoprotein 1. These proteins are involved in the immune response, blood coagulation, cell proliferation and cell adhesion.

Disease progression may be examined by analysing CSF samples from CIS patients who remain CIS and CIS patients who convert to clinically definite multiple sclerosis. Comabella et al. [19, 20] analysed pooled CSF samples with
isobaric labelling and mass spectrometry. They found that chitinase 3-like 1, ceruloplasmin and vitamin D-binding protein were upregulated in CSF of patients converted to clinically definite MS. In order to validate their results, the authors determined the levels of these selected proteins by enzyme-linked immunosorbent assay (ELISA) in individual CSF samples. Only chitinase 3-like 1 was validated. In a second validation step CSF chitinase 3-like 1 levels were measured in an independent CIS cohort and its level was again significantly increased in CIS patients who later converted to MS, compared to patients who remained as CIS. High CSF levels of this protein significantly correlated with the number of gadolinium enhancing and T2 lesions on baseline brain MRI scans and disability progression during follow-up.

The search for biomarkers that are able to identify patients at high risk of rapid progression becomes increasingly important with the appearance of more aggressive treatment possibilities. In another ongoing study we currently analyse LC-Fourier transform ion cyclotron resonance (FTICR) MS [20–22] data of a larger set of CSF samples from a variety of clinical forms of MS and matched controls.

Despite the increasing number of studies investigating potential biomarkers of MS disease progression and response to therapy, there is still no protein that is repeatedly identified and validated by different groups. This may be due to the relatively small sample sizes and the heterogeneity of the methods applied. Large scale multi-centre projects using standard methods for collecting, storing and analysing the samples are necessary to validate these preliminary results and integrate candidate biomarkers into the pathomechanism of the disease.

A great step in this direction is the BIOMS project, which aims a standardized sample collection, storage and processing during the preanalytical steps to rule out the differences occurred by sample preparation [23–25] and test the different methods and hypotheses on a great sample number in multiple centres to shed light on the sources of errors using different methods. One of these initiatives was the neurofilament validation study, which is a candidate biomarker in multiple sclerosis [26]. Another validation study tested two different methods of detecting the neutralizing antibodies against interferon-beta therapy, which is a biomarker of therapy in MS [27].

In the future multi-centre studies on standardized samples and methods can bring us closer to solve the questions regarding the pathological processes and the classification of patients to the most appropriate therapy.

Acknowledgement
TÁMOP-4.2.2.A-11/1KONV/-2012-0052 and The Swedish Research Council 621-2011-4423 are gratefully acknowledged for financial support.

References
1. Polman CH, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011; 69: 292–302.
2. Freedman MS, et al. Recommended standard of cerebrospinal fluid analysis in the diagnosis of multiple sclerosis: a consensus statement. Arch Neurol 2005; 62: 865–870.
3. Karpievitch YV, et al. Liquid Chromatography Mass Spectrometry-Based Proteomics: Biological and Technological Aspects. Ann Appl Stat 2010; 4: 1797–1823.
4. Han X, et al. 3rd Mass spectrometry for proteomics. Curr Opin Chem Biol 2008; 12: 483–490.
5. Becker CH, Bern M. Recent developments in quantitative proteomics. Mutat Res 2011; 722: 171–182.
6. Ross PL, et al. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol Cell Proteomics 2004; 3: 1154–1169.
7. Schutzer SE, et al. Establishing the proteome of normal human cerebrospinal fluid. PLoS One 2010; 5: e10980.
8. Hammack BN, et al. Proteomic analysis of multiple sclerosis cerebrospinal fluid. Mult Scler 2004; 10: 245–260.
9. Dumont D, et al. Proteomic analysis of cerebrospinal fluid from multiple sclerosis patients. Proteomics 2004; 4: 2117–2124.
10. Lehmensiek V, et al. Cerebrospinal fluid proteome profile in multiple sclerosis. Mult Scler 2007; 13: 840–849.
11. Stoop MP, et al. Multiple sclerosis-related proteins identified in cerebrospinal fluid by advanced mass spectrometry. Proteomics 2008; 8: 1576–1585.
12. Han MH, et al. Proteomic analysis of active multiple sclerosis lesions reveals therapeutic targets. Nature 2008; 451: 1076–1081.
13. Stoop MP, et al. Proteomics comparison of cerebrospinal fluid of relapsing remitting and primary progressive multiple sclerosis. PLoS One 2010; 5: e12442.
14. Ottervald J, et al. Multiple sclerosis: Identification and clinical evaluation of novel CSF biomarkers. J Proteomics 2010; 73: 1117–1132.
15. Cantorna MT, Mahon BD. Mounting evidence for vitamin D as an environmental factor affecting autoimmune disease prevalence. Exp Biol Med 2004; 229: 1136–1142.
16. Raghuwanshi A, et al. Vitamin D and multiple sclerosis. J Cell Biochem 2008; 105: 338–343.
17. §Myhr KM. Vitamin D treatment in multiple sclerosis. J Neurol Sci 2009; 286: 104–108.
18. Füvesi J, et al. Proteomic analysis of cerebrospinal fluid in a fulminant case of multiple sclerosis. Int J Mol Sci 2012; 13: 7676–7693.
19. Comabella M, et al. Cerebrospinal fluid chitinase 3-like 1 levels are associated with conversion to multiple sclerosis. Brain 2010; 133: 1082–1093.
20. Bergquist J. FTICR mass spectrometry in proteomics. Curr Opin Mol Ther 2003; 5: 310–314.
21. Ramstrom M, et al. Protein identification in cerebrospinal fluid using packed capillary liquid chromatography Fourier transform ion cyclotron resonance mass spectrometry. Proteomics 2003; 3: 184–190.
22. Ramstrom M, et al. Cerebrospinal fluid protein patterns in neurodegenerative disease revealed by liquid chromatography-Fourier transform ion cyclotron resonance mass spectrometry. Proteomics 2004; 4: 4010–4018.
23. Teunissen CE, et al. A consensus protocol for the standardization of cerebrospinal fluid collection and biobanking. Neurology 2009; 73: 1914–1922.
24. Teunissen CE, et al. Short commentary on ‘a consensus protocol for the standardization of cerebrospinal fluid collection and biobanking’. Mult Scler 2010; 16: 129–132.
25. Tumani H, et al. Cerebrospinal fluid biomarkers in multiple sclerosis. Neurobiol Dis 2009; 35: 117–127.
26. Petzold A, et al. Neurofilament ELISA validation. J Immunol Methods 2010; 352: 23–31.
27. Bertolotto A, et al. Development and validation of a real time PCR-based bioassay for quantification of neutralizing antibodies against human interferon-beta. J Immunol Methods 2007; 321: 19–31.

The authors
Judit Füvesi1 PhD, MD; Cecilia Rajda1 PhD, MD; Dénes Zádori1 PhD, MD; Krisztina Bencsik1 PhD, MD; László Vécsei1,2 PhD, MD; and Jonas Bergquist3,4* PhD, MD

1 Department of Neurology, Faculty of Medicine, Albert Szent-Györgyi Clinical Center, University of Szeged, Szeged, Hungary
2 Neuroscience Research Group of Hungarian Academy of Sciences and University of Szeged, Szeged, Hungary
3 Analytical Chemistry, Department of Chemistry-Biomedical Center, Uppsala University, Uppsala, Sweden
4 Science for Life Laboratory (SciLife Lab), Uppsala University, Uppsala, Sweden

*Corresponding author
E-mail: jonas.bergquist@kemi.uu.se