Cleavage of Cystatin C Is Not Associated with Multiple Sclerosis Piero Del Boccio, PhD,1–3 Damiana Pieragostino, BSc,1–3 Alessandra Lugaresi, MD, PhD,1,4 Maria Di Ioia, MD,1,4 Barbara Pavone, BSc,1–3 Daniela Travaglini, MD,1,4 Simona D’Aguanno, PhD,5,6 Sergio Bernardini, MD, PhD,5,6 Paolo Sacchetta, PhD,1,2 Giorgio Federici, MD, PhD,3,5,6 Carmine Di Ilio, PhD,1,2 Domenico Gambi, MD,1,4 and Andrea Urbani, PhD1–3 Recently, Irani and colleagues proposed a C-terminal cleaved isoform cystatin C (12.5kDa) in cerebrospinal fluid as a marker of multiple sclerosis. In this study, we demonstrate that the 12.5kDa product of cystatin C is formed by degradation of the first eight N-terminal residues. Moreover, such a degradation is not specific in the cerebrospinal fluid of multiple sclerosis, but rather is given by an inappropriate sample storage at ⫺20°C. We conclude that the use of the 12.5kDa product of cystatin C in cerebrospinal fluid might lead to a fallacious diagnosis of multiple sclerosis. Preanalytical validation procedure is mandatory for proteomics investigations. Ann Neurol 2007;62:201–204 Multiple sclerosis (MS) is an often disabling demyelinating disease of the central nervous system. The cause of MS has not been fully elucidated, although genetic, environmental, and immunological factors play a pathogenetic role.1– 4 A definite diagnosis is still difficult to ascertain, especially at disease onset, for the lack of reliable molecular markers. A clinical proteomics investigation on biofluids is currently largely pursued to tentatively identify protein profiles that are diagnostic or prognostic indicators of disease course.5–7 Linear matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) has already been used recently, in the version of surface-enhanced laser desorption ionization TOF-MS, by several groups to detect potential novel biomarkers of several neurological disorders8 –11 in cerebrospinal fluid (CSF) samples. However, pattern-profiling proteomics methodology may not be quite reproducible, because a number of clinical and preanalytical chemistry factors constitute major sources of variability and bias.12,13 Recently, Irani and colleagues14 reported a selective C-terminal cleavage of cystatin C with a mass signal in the spectrum at 12.5kDa, in the CSF of a subgroup of patients with MS and clinically isolated syndromes. In this work, surface-enhanced laser desorption ionization TOF-MS was used to investigate the differential protein profiling from various neurological diseases. This cystatin C isoform was 100% specific for a subgroup of MS and clinically isolated syndromes patients, and the authors suggest an adaptive host response that may also be able to differentiate these patients’ diagnoses from other inflammatory diseases. In contrast, Carrette and coworkers15 showed an N-terminal truncated cystatin C at m/zcalc ⫽ 12.5kDa in human CSF as a result of long-term storage at ⫺20°C in healthy subjects and patients with dementia of various origin. These two reports raise many controversies about the actual role of truncated cystatin C and also about the site of truncation. We further investigated this finding using linear MALDI-TOF-MS technology to study the CSF from MS patients compared with healthy control subjects and patients with other neurological diseases. Subjects and Methods Patients and Cerebrospinal Fluid Collection From the 1Centro Studi sull’Invecchiamento, Fondazione “G. D’Annunzio,” Chieti; 2Dipartimento di Scienze Biomediche, Università “G. D’Annunzio” di Chieti e Pescara; 3Centro Europeo Ricerca sul Cervello, IRCCS-Fondazione S. Lucia, Roma; 4Dipartimento di Oncologia e Neuroscienze, Università “G. D’Annunzio” di Chieti e Pescara; 5Dipartimento di Medicina di Laboratorio, Policlinico di Tor Vergata; and 6Dipartimento di Medicina Interna, Università di Roma Tor Vergata, Roma, Italy. Received Jun 12, 2006, and in revised form Jul 25. Accepted for publication Aug 4. This article includes supplementary materials available via the internet at http://www.interscience.wiley.com/jpages/0364-5134/suppmat Published online Sep 27, 2006, in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.20968 Address correspondence to Dr Urbani, Centro Studi sull’invecchiamento (Ce.S.I.), Centre of Excellence on Aging, Università degli Studi di Chieti e Pescara, Via Colle dell’ Ara, Chieti Scalo, 66100, Chieti, Italy. E-mail: email@example.com Thirty-seven patients with definite MS16,17 were included in this study (complete clinical data are available online in the supplementary material). Expanded Disability Status score was obtained at the time of CSF acquisition. Clinical diagnosis was confirmed by cerebral and spinal magnetic resonance imaging studies and the presence of oligoclonal bands in CSF. Patients had not been treated with steroids in the month before study entry. Thirty CSF samples were used as control group: 5 healthy control subjects (ie, subjects undergoing spinal anesthesia for orthopaedic noninflammatory conditions; 1 female and 4 male subjects); 11 patients with dementia, including 3 patients with suspected Alzheimer’s disease; 6 patients with Guillain–Barré syndrome; 5 stroke patients; 1 patient with amyotrophic lateral sclerosis; 1 case of encephalomyelitis; and 1 case of Leber’s disease. The diagnosis in each of these patients was defined according to individual disease diagnostic criteria. The study was conducted in accordance with the © 2006 American Neurological Association Published by Wiley-Liss, Inc., through Wiley Subscription Services 201 principles of the Helsinki Declaration (World Medical Association, 1997). After lumbar puncture, each sample was centrifuged at 10,000g for 10 minutes, transferred in polypropylene tubes, and stored at ⫺80°C or ⫺20°C depending on the specific experiment. Linear Matrix-Assisted Laser Desorption Ionization Time-of-Flight Mass Spectrometric Analysis Sample (10l CSF with 10l TFA 0.1%) preparation was performed by ZipTip (Millipore tip size P10; Millipore, Bedford, MA) C4 tips, as reported previously.11 All analyses were performed with a Reflex IV MALDI-TOF mass spectrometer (Bruker Daltonics, Bremen, Germany). Cerebrospinal Fluid Protein Purification and Identification CSF samples stored at ⫺20°C for 3 months were fractionated by ultrafiltration using Microcon (Millipore) to enrich proteins with a molecular mass in the 10- to 30kDa range. This liquid portion was separated and fractionated by high-performance liquid chromatography mass spectrometry. Forty liquid fractions were collected from 24 to 50 minutes of the chromatographic run. Fractions containing the 12,536 Thompson (Th) protein signal were digested with trypsin and used for protein identification by mass spectrometric analysis, as reported previously.11 The routine FindPept Tool software (www.expasy.org/cgi-bin/findpept.pl) was used to calculate the accurate mass differences in the protein sequence. Results Cystatin C Signals in Cerebrospinal Fluid MatrixAssisted Laser Desorption Ionization Mass Spectra In Figure 1, the mass spectra is highlighted in the region of the cystatin C signal obtained from the CSF samples of a MS patient and healthy control subjects, analyzed after 3 months of storage at ⫺80°C (see Figs 1A, B) and after 3 months at ⫺20°C (see Figs 1C, D). In Figures 1C and D, an unidentified peak at m/zcalc ⫽ 12,536 Th appears that is absent in the mass spectra acquired after storage at ⫺80°C. This signal was associated with a truncated isoform of cystatin C, as Carrette and coworkers15 and Irani and colleagues14 reported. In these two works, conflicting data are presented about the site of truncation and the functional role of this modified isoform. Our experiments univocally assign the 12,536 Th present in ⫺20°C stored samples as cystatin C (National Center for Biotechnology Information accession number gi兩14278690). Further experiments demonstrate that the site of truncation is in the N-terminal region of the protein and result from the loss of the 1-8 peptide (SSPGKPPR) (complete mass spectrometry characterization available online in the supplementary materials). 202 Annals of Neurology Vol 62 No 2 August 2007 Fig 1. Mass spectra region of the cystatin C signals obtained from the same cerebrospinal fluid (CSF) samples of a multiple sclerosis (MS) patient and a healthy control subject analyzed after 3 months of storage at ⫺80°C (A: MS patient; B: healthy control subject) and after 3 months at ⫺20°C (C: MS patient; D: healthy control subject). (C, D) The peak at m/zcalc ⫽ 12,536 Th is highlighted, associated with a truncated isoform of cystatin C, as reported previously. N-Truncated Cystatin C as a Marker of Ex Vivo Degradation Processes The relative intensity ratio of the m/zcalc ⫽ 12,536 Th versus m/zcalc ⫽ 13,360 Th forms of cystatin C in the CSF of MS patients has been investigated on a cohort of 21 subjects. These data have been compared with a control group. The t test was applied to discriminate the two groups investigated showing a nonincreased relative ratio of this truncated form ( p ⫽ 0.28) in MS patients (Fig 2). However, the N-truncated cystatin C is always absent in spectra obtained from CSF samples stored at ⫺80°C, and the relative intensity ratio was measured considering the background level at the specific m/z value. Subsequently, we analyzed eight selected CSF samples from MS patients after a 3-month storage at ⫺80°C versus aliquots of the same samples kept 3 months at ⫺20°C. The relative intensity ratio of the 12,536 signal versus 13,360 signal reported in Figure 3A clearly shows the selective ex vivo generation of this isoform when CSF samples are stored at ⫺20°C ( p ⬍ 0.005). The in vitro N-truncation of cystatin C occurred in truncated form of cystatin C at m/zcalc ⫽ 12,536 Th, which was always absent in the fresh samples or in samples kept at ⫺80°C, in all CSF samples investigated when left for more than 10 days at ⫺20°C. This truncation occurred in all the CSF samples analyzed regardless of the underlying neurological status, indicating a storage-related phenomenon rather than physiological or pathological processing of the protein. Our data are in agreement with Carrette and coworkers’ findings15 in dementia, and demonstrate that such a degradation occurs at ⫺20°C also in MS patients, causing an erroneous interpretation of this truncated form of cystatin C as a biomarker for MS. Subsequent investigations allowed the identification of the site of Fig 2. Relative intensity ratio distribution of cystatin C truncated form/cystatin C full length (12,536/13,360) in the investigated groups (multiple sclerosis [MS] vs control group [CG]). The CG includes 5 healthy control subjects, 10 patients with dementia, 5 patients with Guillain–Barré syndrome, 2 stroke patients, 1 patient with amyotrophic lateral sclerosis, and 1 case of encephalomyelitis. Samples stored at ⫺80°C were analyzed immediately after thawing to avoid oxidation or other artifact modifications. all CSF samples stored at ⫺20°C independently from the diagnosis. In fact, the truncated form of the protein appeared (see Fig 3B) in samples kept ⫺20°C already after 10 days, either belonging from MS patients (n ⫽ 16) or control group (n ⫽ 6). Discussion Irani and colleagues14 describe a C-terminal cleaved cystatin C in CSF as a marker of MS by measuring the relative intensity ratio of the truncated form 12.5kDa versus the intact molecule at 13.6kDa by surfaceenhanced laser desorption ionization TOF-MS. In a previous study, Carrette and coworkers15 reported an N-truncated isoform of cystatin C with mass of 12.5kDa in CSF samples from patients with various dementia diseases when the CSF specimens were stored 3 months at ⫺20°C. However, addressing the same cleaved form at 12.5kDa cystatin C, these data are in contrast either for the site of the truncation or for the clinical functional insight. Here, we have pursued a thorough characterization of this phenomenon, exploring both the distribution and preanalytical requirement of the truncated cystatin C in CSF of MS patients and patients with other neurological diseases. The linear MALDI-TOF-MS data region of the cystatin C (see Fig 1) was not characterized by any statistical difference in the protein profiling in all patients investigated regardless of the diagnosis, not reproducing the data showed in the previous work of Irani and colleagues.14 Our results show an N-terminal eight-amino acid– Fig 3. (A) Box and whisker plot of the relative intensity ratio of the 12,536 versus 13,360 signal on 8 selected cerebrospinal fluid (CSF) samples from multiple sclerosis (MS) patients after 3 months of storage at ⫺80°C (A) versus aliquots of the same samples kept 3 months at ⫺20°C (B). (B) Comparison of the relative intensity ratio of CSF samples kept more than 3 months at ⫺20°C versus samples stored more than 3 months at ⫺80°C. Control group (CG) included patient affected by dementia (n ⫽ 1), patients with cerebral vascular diseases (n ⫽ 3), patients with Guillain–Barré syndrome (n ⫽1), and Leber’s hereditary optic neuropathy case (n ⫽ 1). Del Boccio et al: Cystatin C Degradation in CSF 203 truncation in the N-terminal region, resulting in the loss of the 1-8 peptide (SSPGKPPR). In conclusion, direct MALDI-TOF-MS approach for proteomics investigations may lead to the development of novel laboratory tests for clinical molecular markers, but preanalytical processing standardization is a mandatory step for clinical proteomics studies, to avoid erroneous identification of biomarkers. This study was supported by the Italian National Grant Program (“Progetto Ricerca Finalizzata; IRCCS-S. Lucia #186,” G.F., C.D.I.) and by Italian Minister of Research (PRIN 2004055300_003, G.F., A.U.). References 1. Mc Farlin DE, Mc Farlan HF. Multiple sclerosis. N Engl J Med 1982;307:1183–1188. 2. Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG. Multiple sclerosis. N Engl J Med 2000;343:938 –952. 3. White RF. Emotional and cognitive correlates of multiple sclerosis. J Neuropsychiatry Clin Neurosci 1990;4:422– 428. 4. Brassington JC, Marsh NV. Neuropsychological aspects of multiple sclerosis. Neuropsychol Rev 1998;2:43–77. 5. Del Boccio P, Urbani A. Homo sapiens proteomics: clinical perspectives. Ann Ist Super Sanita 2005;41:479 – 482. 6. Veenstra TD, Conrads TP, Hood BL, et al. Biomarkers: mining the biofluid proteome. Mol Cell Proteomics 2005;4: 409 – 418. 7. Maddalo G, Petrucci F, Iezzi M, et al. Analytical assessment of MALDI-TOF imaging mass spectrometry on thin histological samples. An insight in proteome investigation. Clin Chim Acta 2005;357:210 –218. 204 Annals of Neurology Vol 62 No 2 August 2007 8. Dekker LJ, Boogerd W, Stockhammer G, et al. MALDI-TOF mass spectrometry analysis of cerebrospinal fluid tryptic peptide profiles to diagnose leptomeningeal metastases in patients with breast cancer. Mol Cell Proteomics 2005;4:1341–1349. 9. Carete O, Demalte I, Scherl A, et al. A panel of cerebrospinal fluid potential biomarkers for the diagnosis of Alzheimer’s disease. Proteomics 2003;3:1486 –1494. 10. Sanchez JC, Guillaume E, Lescuyer P, et al. Cystatin C as a potential cerebrospinal fluid marker for the diagnosis of Creutzfeldt-Jakob disease. Proteomics 2004;4:2229 –2233. 11. Biroccio A, Del Boccio P, Panella M, et al. Differential posttranslational modifications of transthyretin in Alzheimer’s disease: a study of the cerebral spinal fluid. Proteomics 2006;6: 2305–2313. 12. Villanueva J, Philip J, Chaparro CA, et al. Correcting common errors in identifying cancer-specific serum peptide signatures. J Proteome Res 2005;4:1060 –1072. 13. Carlson SM, Najmi A, Whitin JC, Cohen HJ. Improving feature detection and analysis of surface-enhanced laser desorption/ ionization-time of flight mass spectra. Proteomics 2005;5: 2778 –2788. 14. Irani DN, Anderson C, Gundry R, et al. Cleavage of cystatin C in the cerebrospinal fluid of patients with multiple sclerosis. Ann Neurol 2006;59:237–247. 15. Carrette O, Burkhard PR, Hughes S, et al. Truncated cystatin C in cerebrospinal fluid: technical [corrected] artefact or biological process? Proteomics 2005;5:3060 –3065. 16. McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–127. 17. Frohman EM, Goodin DS, Calabresi PA, et al. Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. The utility of MRI in suspected MS: Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 2003;61:602– 611.