Quantitative and qualitative changes in gene expression patterns characterize the activity of plaques in multiple sclerosis

Molecular Brain Research 119 (2003) 170 – 183 www.elsevier.com/locate/molbrainres

Research report

Quantitative and qualitative changes in gene expression patterns characterize the activity of plaques in multiple sclerosis Lotti Tajouri a, Albert S. Mellick a, Kevin J. Ashton a, Anthony E.G. Tannenberg b, Rashed M. Nagra c, Wallace W. Tourtellotte c, Lyn R. Griffiths a,* a

School of Health Science, Griffith University, Gold Coast Campus, Parklands Drive, Southport QLD 4215, Australia b Queensland Medical Laboratories, P.O. Box 5410, West End QLD 4101, Australia c Human Brain and Spinal Fluid Resource Center, Neurology Research (127A), VA West Los Angeles Healthcare Center, 11301 Wilshire Boulevard, Los Angeles, CA 90073, USA Accepted 16 September 2003

Abstract Multiple sclerosis (MS) is a complex autoimmune disorder of the CNS with both genetic and environmental contributing factors. Clinical symptoms are broadly characterized by initial onset, and progressive debilitating neurological impairment. In this study, RNA from MS chronic active and MS acute lesions was extracted, and compared with patient matched normal white matter by fluorescent cDNA microarray hybridization analysis. This resulted in the identification of 139 genes that were differentially regulated in MS plaque tissue compared to normal tissue. Of these, 69 genes showed a common pattern of expression in the chronic active and acute plaque tissues investigated ( Pvalue < 0.0001, q = 0.73, by Spearman’s q analysis); while 70 transcripts were uniquely differentially expressed ( z 1.5-fold) in either acute or chronic active tissues. These results included known markers of MS such as the myelin basic protein (MBP) and glutathione S-transferase (GST) M1, nerve growth factors, such as nerve injury-induced protein 1 (NINJ1), X-ray and excision DNA repair factors (XRCC9 and ERCC5) and X-linked genes such as the ribosomal protein, RPS4X. Primers were then designed for seven array-selected genes, including transferrin (TF), superoxide dismutase 1 (SOD1), glutathione peroxidase 1 (GPX1), GSTP1, crystallin, alpha-B (CRYAB), phosphomannomutase 1 (PMM1) and tubulin b-5 (TBB5), and real time quantitative (Q)-PCR analysis was performed. The results of comparative Q-PCR analysis correlated significantly with those obtained by array analysis (r = 0.75, Pvalue < 0.01, by Pearson’s bivariate correlation). Both chronic active and acute plaques shared the majority of factors identified suggesting that quantitative, rather than gross qualitative differences in gene expression pattern may define the progression from acute to chronic active plaques in MS. D 2003 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Gene structure and function: general Keywords: Multiple sclerosis; Plaque; Chronic active; Acute; cDNA microarray; Quantitative PCR analysis

1. Introduction Multiple sclerosis (MS) is an autoimmune disease of the CNS, characterized by zones of demyelination and inflammatory plaques [43]. Symptoms include limb weakness, sensory loss, visual alterations and bladder dysfunction. MS is one of a number of inflammatory-demyelinating diseases (IDDs), which are characterized by the appearance * Corresponding author. Genomics Research Centre, School of Health Science, Griffiths, PMB 50 GCMC, Gold Coast 9726, Australia. Tel.: +6175552-8808; fax: +61-75552-8908. E-mail address: [email protected] (L.R. Griffiths). 0169-328X/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.molbrainres.2003.09.008

of lesions that are disseminated in time and space (either restricted or diffuse) [81]. Prototypic MS is characterized by an initial disease onset that is often followed by a secondary progressive (SP) course. In general terms, acute lesions show an infiltration of lymphocytes, as well as oligodendroglial hyperplasia [22]. Chronic active plaques, in contrast, possess a rich oligodendrocyte population, and are characterized by a patchwork of inflammatory and ‘silent’ zones, or regions showing some remyelination. Chronic ‘silent’ plaques show extensive remyelination, and unlike acute and chronic active lesions, lack an obvious inflammatory component [59,60]. Recent advances in array hybridization analysis have allowed the investigation of many thousands of candidate

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

171

markers of disease simultaneously. This approach has already been used successfully in the identification of the Jagged-Notch-Hes (putative oligodendrocyte differentiation pathway) as a contributor to MS development [32,56]. However, arrays by themselves have limitations that restrict analysis to particular transcripts based on relative abundance and representation on the array. In contrast, Q-PCR is a robust and sensitive method, which is frequently employed to investigate gene expression differences independent of the absolute level of gene expression [62]. In this study, we have employed array hybridization and Q-PCR analysis to examine quantitative, as well as gross qualitative changes in gene expression that may differentiate inflammatory forms of MS plaques. RNA from plaque tissue was isolated from patients who died with MS and compared with patient-matched normal white matter. The data collected in this way was then used to examine patterns of gene expression that differentiate genes and gene clusters, grouped by designated biological function. The results of microarray analysis revealed a significant correlation in gene expression pattern ( Pvalue < 0.0001, q = 0.73, by Spearman’s q correlation) for the 69 genes identified as differentially expressed in both acute and chronic active lesions as compared to normal tissue. However, for these genes the mean ‘fold’ level of gene expression difference was found to be significantly higher in acute plaques, suggesting that symptomatic variation in inflammatory lesions in MS may be the result of overall quantitative, rather than specific gross qualitative differences in gene activity. In this study, SYBR Green I, real time Q-PCR was used to validate the differential expression of array selected clones; and to examine the expression of transcripts that may be present below the level of sensitivity for array analysis.

diagnosed with secondary progressive MS. Out of these samples only one human brain section (HSB) 3131 showed signs of recent hypoxia. All control tissues were obtained from patients that had no past history of neurological disorder and experimental procedures were conducted following ethical guidelines of the National Health Medical Research Council (NHMRC), Australia. Samples were obtained with ethical consent from the NHMRC Brain Bank, University of Queensland, Australia and the Human Brain and Spinal Resource Center, Los Angeles, USA. All tissue was immediately placed in liquid nitrogen upon dissection, and stored at
80 jC until used. The demyelinating activity of specimens was made available to authors at the completion of array analysis, and all five samples were used in Q-PCR and array hybridization analysis.

2. Material and methods

2.3. RNA validation and probe preparation

2.1. Patients and tissues

The quality of RNA extracted post mortem from brain tissue can be problematic unless certain precautions are taken [29,67]. In our study, one-step RT-PCR was performed using published methods to investigate the integrity of extracted post mortem RNA [69]. The following cycling conditions were used: (1) 50 jC for 30 min ( 1); (2) 95 jC for 30 s, 57 jC for 1 min, 72 jC for 30 s ( 45); (3): 72 jC for 4 min ( 1). All products were resolved by agarose (2%) gel electrophoresis. The 6000 Nano Lab Chipk kit (Agilent Technologies, CA) in conjunction with the 2100 Bioanalyzerk (Agilent Technologies) was also used to investigate the integrity of RNA used in probe preparation [26]. Probe was prepared for hybridization analysis from 30 to 40 Ag of total RNA (plaque and matched control brain tissue) using either [Cy3]-dCTP or [Cy5]-dCTP fluorescein dyes (Amersham Pharmacia Biotech), following standard methods and using mixed an-

MS plaques, or aberrant tissue, were obtained post mortem (16.6 F 7.7 h) (Table 1). Tissue was classified based on published criteria [9,65]. Tissue was collected and 7 mm sections were imaged and snap frozen (
150 jC), prior to storage at
80 jC. Computerized serial full coronal digital images from each case were used as a guide for dissection of frozen slices; facilitated by high-speed dental equipment. The bottom half of each section was kept for histological classification: necessary for determining demyelinating activity. The mean age at death F S.D. for MS subjects (S) 1– 5 was 53.6 F 4.5 years (4 female and 1 male). Control tissue (Normal/Non-MS) was obtained and matched to MS plaques tissue by sex, age F 8 years, ethnicity and anatomical brain localization. The distribution was as follows: two acute lesions and three chronic active lesions. All patients were

2.2. RNA extraction and cDNA synthesis The protocol applied for the purification of total RNA from plaque and non-plaque tissue (0.3 – 1.1 g) was adapted from existing protocols. To maximize the quality of RNA used as probe in array hybridization each sample was first extracted in phenol-containing Trizolk (Invitrogen/Life Technologies, CA). The aqueous phase was removed and subjected to a further round of purification, following the protocol outlined in the RNeasyk Mini Kit Manual (Qiagen, CA). About 1– 5 Ag of total RNA not used in probe preparation was also converted to cDNA using 5 U/Al Superscript II RT (Invitrogen), RT buffer (50 mM Tris – HCl pH 8.3, 75 mM KCl, 3 mM MgCl2), 100 AM dNTP mix (equimolar dG/C/A/TTP) (Amersham Pharmacia Biotech), 10 AM dithiothreitol (DTT) and 25 ng/Al anchored oligo-dT primer mix [equimolar: 5V-T19(A/C/G)-3V]. Each cDNA pool (matched control and plaque) was stored at
20 jC, until used for Q-PCR analysis.

172

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

Table 1 Histopathology Quantity Locationb (g)

Sex Demyelination Causec activity

1 HSB3163 20

0.9

Subcortical White Matter

F

Chronic Active

N/Ae

57

Secondary 14 progressive

2 HSB3161 20

0.4

Periventricular F White Matter

Chronic Active

Respiratory 51 insufficiency due to pneumonia

Secondary 19 progressive

3 HSB3131 24

0.3

Subcortical White Matter

F

Chronic Active

N/Ae

53

Secondary 24 progressive

4 HSB2589

4

0.5

Medial subcortical White Matter

F

Acute

Secondary 30 progressive

5 HSB2946 15

1.1

Medial Subcortical White Matter

M

Acute

48 Respiratory insufficiency due to pulmonary embolic and pneumonia Respiratory 59 insufficiency due to pneumonia

Tissue

Timea (h)

Aged Course of (year) MS

Duration Lesion pathology of illness (years)

Secondary 16 progressive

70 – 80% axonal loss, mild associated gliosis and macrophage activity. Also, perivascular lymphocytic infiltration associated with the plaque and the adjacent white matter. 80 – 90% axonal loss. Slight increase in cellularity at the plaque border. Mild associated gliosis and macrophage activity. Perivascular lymphocytic cuffing are present. Complete demyelination in myelin stain sections. Near complete loss of myelin. Macrophage associated intracytoplasmic myelin debris. Reactive macrophages along focal regions of the plaque edges. Marked demyelination, CD68 MHC Class 2 activated macrophages, deposited evenly throughout the plaque. Marked demyelination and axonal loss and sparse, focal chronic perivascular inflammation.

a

Time after death. Location from which control and matched plaque tissue was obtained. c Cause of death. d Age of death. e N/A, not available. b

chored primers (5V-T12VN-3V) (Amersham Pharmacia) and Superscript II RT enzyme. 2.4. Array hybridization The protocol used for microarray hybridization follows previously published methods [66]. Custom-made glass arrays, containing 5000 random genes/cDNAs each (Queensland Institute of Medical Research, Australia), were used. All clones were double spotted to assess hybridization specificity, and were obtained with gene reference identifiers for independent verification of sequence, using the National Center for Biotechnology Database (http://www.ncbi.nlm. nih.gov/entrez). A complete gene list is available at http:// www.genomicsresearchcentre.org/ms.html. Following hybridization (12 h, 45 jC), each slide was scanned (428 Array scannerk, Affymetrix), and signal intensity ratios for Cy3/ Cy5 used to determine a fold value for the difference in expression between aberrant and control tissue. Over 100

control (house keeping and 3  SSC) spots on each array were used to normalize the fold expression, and to determine the degree of non-specific hybridization. Each experiment was also repeated with the alternate dye label, and the mean fold determined following statistical validation. Signal intensity ratios for duplicate spots that differed (>2-fold), were excluded from analysis. In order to examine small differences that might be associated with particular gene groups or pathways, all transcripts were provided with a functional designation based on relationship to disease from information available through the Online Mendelian Index in Man (OMIM) (http://www.ncbi.nlm.nih.gov/entrez). 2.5. Q-PCR analysis Primer design, and optimization of MS array-selected genes, followed protocols previously developed within this laboratory [62]. Where possible primers for each selected gene were designed to be intron spanning, so that genomic

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

DNA (gDNA) could be used as a positive control template (Table 2). All Q-PCRs were performed in 96 well Bio-Rad iCycler iQ plates (Bio-Rad, Sydney, Australia), and contained 15 nM Fluorescein Dye (Bio-Rad), 1  PCR buffer II [50 mM KCl and 10 mM Tris – HCl (pH8.3)], MgCl2 (see Table 2), 200 AM dNTPs (eqimolar dG/C/T/ATP), 1  Bovine Serum Albumin (BSA) (NEB, Beverly, MA); 0.5  SYBR Green I (Sigma); Taq DNA Polymerase (0.05 U/Al) (Amersham Pharmacia); as well as forward and reverse primers (0.5 AM each) (Geneworks, Australia). The following cycling conditions were used: (1) 95 jC for 5 min (  1), (2) 95 jC for 30 s, 55 jC for 1 min, 72 jC for 30 s (  45), (3) 72 jC for 4 min (  1). Detection of PCR product in real time was performed using the BioRad iCycler iQ system. All products were resolved at end point using melt curve, and 10% polyacrylamide gel electrophoresis (PAGE). Each determination of fold was conducted from three repeat determinations of cycle threshold (CT) at linearity, which were corrected using 18S rRNA as a loading and house keeping gene control (DCT) [74]. Comparative differences in gene expression levels were then obtained from the difference between DCT for control (DCTc) and test (plaque) tissue (DCTt), or 2
DDCT [40]. 2.6. Statistical analysis SPSS Version 10.1 (PC) was used to collate and analyze all data collected. Standard independent t-test analysis was used to determine whether determinations of fold based array analysis data were statistically significant ( Pvalue < 0.05). Spearman’s q analysis was used to determine whether the observed correlation in gene expression patterning between pathologies was significant ( Pvalue < 0.05). To investigate whether observed differences in gene

173

expression patterns for genes assigned to different gene groups were significant, Mann-Whitney U tests were used ( Pvalue < 0.05). Following Q-PCR analysis, the corrected differences in CT (for control and test cDNA: DCTc and DCTt) were examined by independent t-test analysis to determine whether the observed difference between them (DDCT) was statistically significant ( Pvalue < 0.05). Pearson’s bivariate correlation analysis was also used to determine whether the observed similarities between array and Q-PCR values were significant ( Pvalue < 0.05). In this study, the coefficient of variation (CV) of the values is used to indicate patient variance.

3. Results MS plaque (Fig. 1) and unaffected patient matched brain tissue was isolated. Total RNA, extracted from both, was then compared using fluorescent microarray hybridization analysis (Fig. 2). The mean fluorescence ratio was then used to determine the degree of expression compared to normal matched tissue and expressed as the mean fold for each plaque type (chronic active, or acute). In this way, 139 genes were identified as differentially regulated in the five inflammatory MS plaques examined in this study (mean z 1.5fold). Forty-six of the total number of differentially regulated transcripts (or 33.1%) represent products that are preferentially regulated in the brain, or that have previously been linked to neurological disease, such as the nerve injury-induced protein (NINJ) 1, which promotes axonal growth [4], or the brain specific creatine kinase, CKB [5]. This suggests that despite the labile nature of RNA and the variable preservation of the body post mortem [47], an adequate representation of neuronal markers, and those

Table 2 Primer sequences cDNAa (bp)

gDNAb (bp)

Gene

Gene reference

TF

Ref | NM_001063.2 |

74

567

GPX1

Ref | NM_000581.1 |

70

349

GSTP1

Ref | NM_000852.2 |

60

237

CRYAB

Ref | NM_001885.1 |

65

1139

SOD1

Ref | NM_000454.1 |

99

838

PMM1

Ref | NM_002676.1 |

74

441

TBB5

Ref | NM_006087.2 |

102

219

18S rRNA

Gb | U13369.1 |

110

110

a

Forward (5V– 3V)

Reverse (5V– 3V)

MgCl2c

ATG TGG CCT TTG TCA AGC ACT (EXd 6) GCT TCC CGT GCA ACC AGT TT (EX 1) CAG GGA GGC AAG ACC TTC AT (EX 6) CAC CCA GCT GGT TTG ACA CT (EX 1) GGT CCT CAC TTT AAT CCT CTA T (EX 3) CAG CTT CGA CAC CAT CCA CTT (EX 7) GGA TCA ACG TGT ACT ACA ACG A (EX 2) CTT AGA GGG ACA AGT GGC

GCT CAT CAT ACT GGT CCC TGT CA (EX 7) CTT GAG GGA ATT CAG AAT CTC T (EX 2) GCA GGT TGT AGT CAG CGA A (EX 7) TGA CAG AGA ACC TGT CCT TCT (EX 2) CAT CTT TGT CAG CAG TCA CAT T (EX 4) TCG GCA AAG ATC TCA AAG TCG T (EX 8) CAG AAC GGA CAG AGT CCA TGG T (EX 3) ACG CTG AGC CAG TCA GT

3.50

cDNA refers to the length of amplicon. gDNA refers to the length of the corresponding genomic DNA. c MgCl2 concentration (mM) optimized for Sybr Green I Q-PCR analysis. d EX, exon. b

3.50 3.50 3.50 3.50 3.75 3.75 3.50

174

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

Fig. 1. (A) Computerized full colour image of brain section HSB3163, showing regions of MS affected tissue. (1) Plaque, (2) normal apparent white matter, (3) non-apparent grey matter, (4) plaque, (5) normal apparent white matter, (6) normal apparent grey matter. (B) Microscopic analysis of demyelinating activity Haematoxylin and eosin staining of MS-affected white matter, from typical chronic active plaque (HSB3163). Shown infiltration of stromal cells (e.g. leukocytes) from vasculature into the area of the plaque (arrow) (magnification 20, scale bar 100 AM).

potentially significant to disease pathology, were identified following the microarray protocol outlined above. 3.1. Genes identified by array analysis as differentially regulated in both chronic active and acute plaques Sixty-nine of the 139 genes identified as differentially regulated in MS plaques were commonly regulated in either chronic active or acute plaques (Table 3), and showed a significant correlation ( Pvalue < 0.0001) in the pattern of expression, by Spearman’s q correlation. These 69 were then divided into groups based on function and these functional clusters were investigated to determine specific differences in gene expression between acute and chronic active lesions. Seventy (54.0%) of the 139 genes identified as differentially regulated in MS were uniquely regulated in either chronic active or acute plaques. Of the 69 genes differentially regulated in both inflammatory MS plaque types (see above): 34 (49.3%) showed a higher degree of expression, compared with normal white matter, or were upregulated in both chronic active and acute plaques; 33 (47.8%) showed a lower level of expression, compared with normal white matter; or were down-regulated in both chronic active and acute plaques. Enolase (ENO) 2 [58] and dolichylphosphate mannosyltransferase 1, catalytic (DPM1), which

Fig. 2. Probe preparation and fluorescent hybridization analysis. (A) Representative agarose (2%) gel electrophoresis of total RNA, isolated from MS plaque tissue and normal tissue post-mortem (prior to DNaseI treatment). Lane 1, HSB2589 (20 Ag), Lane 2, Normal tissue (20 Ag). 28S rRNA (a) and 18S rRNA (h) bands can be identified following denaturing extracted from normal and MS plaque tissue. (B) One step RT-PCR analysis of total RNA from MS plaque RNA, following DNaseI treatment. Shown agarose (2%) gel electrophoresis of PCR product. Lane 1, DnaseI treated genomic DNA control; Lane 2, HSB3163 RNA post DNaseI treatment; and Lane 3, normal brain RNA post RNA treatment. Primers used, human hactin, forward: 5V-ACC CAC ACT GTG CCC ATC TA-3V, reverse: 5V-CGG AAC CGC TCA TTG CC-3V. (C) Representative Agilentk trace (left) and chromatographic separation (right) of DNaseI treated total RNA from MS brain tissue (HSB3163), and prior to probe labelling: a, 18S rRNA (40 s); h, 28 rRNA (45 s). (D) Scatter plot: log of relative difference in fluorescence units (DRFU), following 5000-gene HSB2589/array analysis. Those cDNAs with a ratio of 1 (line) are not regulated significantly, and are not included in analysis of fold. (E) Result of fluorescent array analysis (HSB2589), following hybridisation with fluorescently labelled probe. Cy3 (Green), Cy5 (red). Each clone spotted twice on the array aids in differentiating specific and non-specific hybridisation events. Results of array analysis: m: MBP, 63.1-fold up regulated in acute tissue; and y: ITGAE, 145.9-fold down regulated in acute tissue.

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183 Table 3 Genes identified by array analysis as differentially expressed in MS plaque tissue Gene

OMIM CAa

INTEGRIN, ALPHA-E (ITGAE) CALPONIN 2 (CNN2) VILLIN 2 (VIL2) CAP PROTEIN, ACTIN, ALPHA-1 SUBUNIT (CAPPA1) DYNACTIN 2 (DCTN2)c MICROTUBULE-ASSOCIATED PROTEIN, RP/EB FAMILY, MEMBER 3 (MAPRE3)c TUBULIN, BETA-5 (TBB5)c FLIGHTLESS I, DROSOPHILA, HOMOLOG OF (FLII)c TRANSFERRIN (TF) PROTEIN-TYROSINE PHOSPHATASE, RECEPTOR-TYPE, J (PTPRJ) AlkB, E. COLI, HOMOLOG OF X-RAY REPAIR, COMPLEMENTING DEFECTIVE, IN CH, 9 (XRCC9) EXCISION-REPAIR, COMPLEMENTING DEFECTIVE, IN CH, 5 (ERCC5)c,d COLLAGEN, TYPE III, ALPHA-1 (COL3A1) TEB4 Similar to S. cerevisiae (SSM4) Homo Sapiens cDNA FLJ32847 fis, clone TESTI2003376 SUPEROXIDE DISMUTASE 1 (SOD1)c,d GLUTATHIONE PEROXIDASE (GPX1)c ESTs, Highly similar to A33507 hypothetical protein DUC-1 KIAA1046 protein GAMMA-AMINOBUTYRIC ACID RECEPTOR, GAMMA-2 (GABRG2)c,d INTERFERON REGULATORY FACTOR 1 (IRF1) CHEMOKINE, CXC MOTIF, LIGAND 10 (CXCL10)c INTERLEUKIN 10 RECEPTOR, BETA (IL10RB) ANKYRIN 2 (ANK2) MAJOR HISTOCOMPATIBILITY COMPLEX, CLASS I, A (HLA-A) ISOLEUCYL-tRNA SYNTHETASE (IARS) GUANINE NUCLEOTIDEBINDING PROTEIN, a-INHIBITING ACTIVITY 2 (GNAI2) T-LYMPHOCYTE MATURATIONASSOCIATED PROTEIN (MAL) MYELIN BASIC PROTEIN (MBP)c,d

604 682 602 373 123 900 601 580

1000.0# 81.1# 1.7z 333.3#

Acute

FDb

145.9# 333.3# 3.5z 28.6#

ADH ADH ADH CYT

607 376 125.0# 605 788 9.4#

28.5# 55.6#

CYT CYT

602 662 2.1z 600 362 107.9#

16.3z CYT 6.3# DEV

190 000 2.9z 600 925 12.8#

32.7z DEV 15.4# DEV

605 345 55.6# 602 956 20.2#

13 3530 23.2#

120 180 14.7#

57.2# 500#

55.1#

DRP DRP

DRP

36.1#

ECM

NA

9.1#

2.4#

EST

NA

6.2#

3.5#

EST

147 450 2.0z

14.0z FRM

138 320 1.4z

20z

NA

FRM

3.2z

5.2z HYP

NA 6.1# 137 164 5.0z

5# HYP 11.2z ICH

147 575 17.2#

46.3#

147 310 2.5z

16.1z IMM

123 889 2.4z

5.5z IMM

106 410 6.3# 142 800 NSD

16.1# IMM 40.1z IMM

600 709 7.9# 139 360 1.7z

181.0#

IMM

IMM

8.5z IMM

188 860 NSD

20.8z IMM

159 430 3.4z

63.1z IMM

175

Table 3 (continued ) Gene

OMIM CAa

ARYLACETAMIDE DEACETYLASE (AADAC) PALMITOYL-PROTEIN THIOESTERASE 1 (PPT1)c,d ENOLASE 2 (ENO2)d COMPLEMENT COMPONENT 1, q SUBCOMPONENT, BETA (C1QB) SOLUTE CARRIER FAMILY 10, MEMBER 1 (SLC10A1) PHOSPHOMANNOMUTASE 1 (PMM1)d GLYCOGEN PHOSPHORYLASE, BRAIN TYPE (PYGB)d DOLICHYL-PHOSPHATE MANNOSYLTRANSFERASE 1, CATALYTIC (DPM1)c BETAINE-HOMOCYSTEINE METHYLTRANSFERASE (BHMT) INOSITOL 1,4,5-TRISPHOSPHATE 3-KINASE B (ITPKB)d SOLUTE CARRIER FAMILY 25 (MITOCHONDRIAL CARRIER), A6 (SLC25A6) BETA-GALACTOSIDASE PROTECTIVE PROTEIN, INCLUDED (PPGB)c ALDOLASE A, FRUCTOSEBISPHOSPHATE (ALDOA)d CYTOCHROME C1 (CYC1) ATP SYNTHASE, H + TRANSPORTING, MITO F1 COMPLEX, BETA (ATP5B) RIBONUCLEASE A FAMILY, 1 (RNASE1) CHROMOSOME 21 OPEN READING FRAME 33 (C21ORF33)d UBIQUITIN-ACTIVATING ENZYME 1 (UBE1) UBIQUITIN A-52-RESIDUE RIBOSOMAL PROTEIN FUSION PRODUCT (UBA52) SMALL NUCLEAR RIBONUCLEOPROTEIN POLYPEPTIDE N (SNRPN) CYSTATIN C (CSTC) PHOSPHATIDYLINOSITOL 3-KINASE, REGULATORY, 4 (PIK3R)4 INSULIN-LIKE GROWTH FACTOR II RECEPTOR (IGF2R) PROTEIN C RECEPTOR (PROCR) CALPAIN, SMALL SUBUNIT 1 (CAPNS1) ATP SYNTHASE, H + TRANSPORTING, MITO F1 COMPLEX, O SUBUNIT (ATP5O)

600 338 22.5#

9.3#

MET

600 722 71.4#

48.9#

MET

131 360 1.7# 120 570 8.0z

16.7z MET 18.1z MET

182 396 2.7z

27.7z MET

601 786 1.7z

15.1z MET

138 550 1.6z

18.2z MET

603 503 9.3z

47.6#

602 888 2.3z

17.6z MET

147 522 2.6z

19.8z MET

300 151 2.1z

14.1z MET

256 540 1.9z

19.0z MET

103 850 2.0z

8.0z MET

123 980 9.1# 102 910 2.6z

9# MIT 15.7z MIT

180 440 2.5z

20.7z NUC

601 659 9.4#

19.3#

314 370 2.2z

48.7z PRA

191 321 2.1z

17.8z PRA

182 279 2.0z

52.4z PRA

604 312 1.9z 602 610 125.0#

23.7z PRA 16.1# PRA

147 280 31.9#

66.7#

600 646 23.3# 114 170 2.7z

25.6# PRA 15.7z PRA

600 828 2.6z

4.9z PRA

Acute

FDb

MET

ORF

PRA

(continued on next page)

176

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

Table 3 (continued )

Table 3 (continued ) a

Acute

FD

b

Gene

OMIM CA

RIBOSOMAL PROTEIN S4, X-LINKED (RPS4X)c INSULIN-LIKE GROWTH FACTOR-BINDING PROTEIN 6 (IGFBP6) CULLIN 5 (CUL5) BRCA1-ASSOCIATED RING DOMAIN 1 (BARD1) CYCLIN-DEPENDENT KINASE INHIBITOR 3 (CDKN3) CYCLIN B1 (CCNB1) ABELSON MURINE LEUKEMIA VIRAL ONCOGENE HOMOLOG 1 (ABL1) THYROID HORMONERESPONSIVE GENE ZAKI4 (ZAKI4)d INSULIN RECEPTOR (INSR) CYTOCHROME B-5 (CYB5)c,d CYTOCHROME P450, 51 (CYP51) HUMAN IMMUNODEFICIENCY VIRUS TYPE 1 ENHANCER-BINDING 1 (HIVEP1) FORKHEAD, DROSOPHILA, HOMOLOG-LIKE 5 (FKHL5) MICROTUBULE-ASSOCIATED PROTEIN 4 (MAP4)d DYNEIN, CYTOPLASMIC, LIGHT INTERMEDIATE POLYPEPTIDE 2 (DNCLI2) SARCOGLYCAN, ALPHA (SGCA) GLIAL FIBRILLARY ACIDIC PROTEIN (GFAP)c,d PROFILIN 2 (PFN2) DERMATOPONTIN (DPT) INTEGRIN, BETA-5 (ITGB5) DERMATAN SULFATE PROTEOGLYCAN 3 (DSPG3) MITOGEN-ACTIVATED PROTEIN KINASE 9 (MAPK9) PROTEIN PHOSPHATASE 2, REGULATORY SUBUNIT B (B56), BETA (PPP2R5B)d GDNF FAMILY RECEPTOR ALPHA-2 (GFRA2)d PERIPHERAL MYELIN PROTEIN 22 (PMP22)c,d MOESIN (MSN) TYROSINE 3-MONOOXYGENASE/ TRYPTOPHAN 5-MONOOXYGENASE ACTIVATION PROTEIN, ETA ISOFORM (YWHAH)c,d INHIBITOR OF DNA BINDING 2 (ID2) EXOSTOSIN-LIKE 1 (EXTL1) GRANULIN (GRN)d CYCLIN-DEPENDENT KINASE INHIBITOR 2A (CDKN2A) RNA-BINDING MOTIF PROTEIN, X CHROMOSOME (RBMX)

312 760 2.4z

25.0z PRA

146 735 56.6#

18.4#

PRO

601 741 50.8# 601 593 12.4#

22.7# 27.9#

PRO PRO

123 832 8.1#

24.4#

PRO

123 836 7.9# 189 980 2.3z

604 876 32.3#

147 670 250 790 601 637 194 540

11.4# 8.0# 7.9# 16.9#

6.5# PRO 5.3z PRO

11.2#

5.8# 111.1# 98.2# 52.2#

RHA

RHA RHA RHA TRE

601 089 2.3z

10.4z TRE

157 132 1.7z

CYT

NA

8.0z

CYT

600 119 40.0# 137 780 1.6#

CYT CYT

176 590 125 597 147 561 601 657

4.1z 42.0# 5.1z 5.8#

CYT ADH ADH ADH

602 896 56.8#

PRO

601 644 11.3#

PRO

601 956 4.5#

PRO

601 097 2.9z

PRO

309 845 2.9z 113 508 5.3z

PRO PRO

600 386 1.6#

PRO

601 738 9.0z 138 945 2.9z 600 160 1.9z

PRO PRO PRO

300 199 5.3#

DEV

Gene

OMIM CAa

ENDOMETRIAL BLEEDING-ASSOCIATED FACTOR (EBAF) HEAT – SHOCK 70-KD PROTEIN 1A (HSPA1A)c PEROXIREDOXIN 1 (PRDX1) BETA-2-MICROGLOBULIN (B2M)c BRADYKININ RECEPTOR B2 (BDKRB2) COMPLEMENT COMPONENT 4B (C4B) ALDO-KETO REDUCTASE FAMILY 1, MEMBER C3 (AKR1C3) CREATINE KINASE, BRAIN TYPE (CKB)c,d PHOSPHATIDYLINOSITOL 4-KINASE, CATALYTIC, BETA (PIK4CB) GLUTATHIONE S-TRANSFERASE, PI (GSTP1) ATPASE, CA(2+)-TRANSPORTING, SLOW-TWITCH (ATP2A2)c AMMA-AMINOBUTYRIC ACID RECEPTOR, EPSILON (GABRE)d RIBOSOMAL PROTEIN S13 (RPS13) INOSITOL POLYPHOSPHATE PHOSPHATASE-LIKE 1 (INPPL1) PARATHYROID HORMONE RECEPTOR 1 (PTHR1) PARATHYMOSIN (PTMS)d URACIL DNA GLYCOSYLASE (UNG) ATR-X GENE (ATRX) Ribonuclease 6 precursor (RNASE6PL) E74-like factor 4 (ets domain transcription factor) (ELF4) GTPASE-ACTIVATING PROTEIN, RHO, 1 (ARHGAP1) YES-ASSOCIATED PROTEIN 1, 65-KD (YAP1) MYOSIN VIIA (MYO7A)c TUBULIN, ALPHA, BRAIN-SPECIFICd ZONA PELLUCIDA GLYCOPROTEIN 3 (ZP3) NERVE INJURY-INDUCED PROTEIN 1 (NINJ1)c,d INTERFERON-INDUCED TRANSMEMBRANE PROTEIN 3 (IFITM3) PREGNANCY-SPECIFIC BETA-1-GLYCOPROTEIN 3 (PSG3) TRANSMEMBRANE PROTEIN 1 (TMEM1)d CRYSTALLIN, ALPHA-B (CRYAB)d

601 877 2.5#

DEV

140 550 19.6z

IMM

176 763 2.4z 109700 1.8z

IMM IMM

113 503 3.1z

IMM

120820 1.5z

IMM

603 966 4.3#

MET

123 280 3.5z

MET

602 758 15.9#

MET

134 660 2.8z

FRM

108 740 2.4z

ICH

300 093 24.3#

ICH

180 476 1.6z

PRA

600 829 61.2#

RHA

168 468 2.7z

RHA

168 440 1.5z 191 525 71.4#

RHA DRP

300 032 13.9# NA 47.6#

DRP NUC

NA

TRE

Acute

17.5z

FDb

602 732

9.6z CYT

606 608

1.7#

276 903 602 529

CYT

250# CYT 27.6z CYT

182 889

4.6#

ADH

602 062

9.7#

ADH

605 579

58.1#

IMM

176 392

41.9z IMM

602 103

16.7#

123590

19.3z IMM

IMM

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183 Table 3 (continued ) Gene

OMIM CAa

EPHRIN RECEPTOR EPHB6 (EPHB6)d PYRUVATE KINASE, LIVER AND RED BLOOD CELL (PKLR) RIBOSOMAL PROTEIN L21 (RPL21) RAS-ASSOCIATED PROTEIN RAB2 (RAB2) RELATED TO THE N TERMINUS OF TRE (RNTRE) CALCIUM-DEPENDENT ACTIVATOR PROTEIN FOR SECRETION (CADPS)d PROTEIN INHIBITOR OF ACTIVATED STAT1 (PIAS1) ALKALINE PHOSPHATASE, LIVER (ALPL)c FARNESYL DIPHOSPHATE SYNTHASE (FDPS) SOLUTE CARRIER FAMILY 6, 8 (SLC6A8)d CYTOCHROME P450, SUBFAMILY IIB, POLYPEPTIDE 6 (CYP2B6) AMYLOID BETA A4 PRECURSOR PROTEIN-BINDING, FAMILY A, 1 (APBA1)d PICTAIRE PROTEIN KINASE 3 (PCTK3) REQUIEM, APOPTOSIS RESPONSE ZINC FINGER GENE (REQ) PRKR INHIBITOR, REPRESSOR OF (PRKRIR) ADRENOMEDULLIN (ADM) ADENYLATE CYCLASEACTIVATING POLYPEPTIDE 1 (ADCYAP1)c,d METALLOTHIONEIN 1L (MT1L) ESTs, moderately similar to T00390 kiaa0614 protein ZINC FINGER PROTEIN 38 (ZNF38) 8-@OXOGUANINE DNA GLYCOSYLASE (OGG1)

602 757

63.9z DEV

266 200

12.0z DEV

603 636

6.9z PRA

179509

5.3z PRA

605 405

1000.0 # PRA

604 667

166.7 # PRA

603 566

26.3 # PRA

171 760

12.2z MET

134 629

8.1z MET

300 036

31.5z MET

605 059

11.8 # MET

602 414

7.1 # MET

169 190

20.4z PRO

601 671

17.5 # PRO

607 374

10.0 # PRO

103 275 102 980

7.4z RHA 7.2 # RHA

156 358 NA

30.4z HMB 8.1z EST

601 261

47.6 # TRE

601 982

19.2 # DRP

Acute

FDb

z, up regulated compared with normal tissue; #, down regulated when compared with normal tissue. a Chronic active. b Functional Designation: ADH, adhesion; AXT, axonal transport; CYT, cytoskeletal; DRP, DEV, developmental; DNA repair and protection; ECM, extracellular matrix component; EST, expressed sequence tag; FRM, free radical metabolism; HMB, heavy metal binding; HYP, hypothetical; ICH, ion channel; IMM, immunological; LYM, lysosomal Marker; MET, metabolic; MIT, mitochondrial; NUC, nuclease; ORF, open reading frame; PRA, protein regulatory activity; PRO, cell cycle regulation; RHA, regulation of hormone activity; RPA, apoptosis regulation; TRE, transcription factor or enhancer; and m, data not available.

is associated with severe mental psychomotor retardation [35], were the only transcripts to be down regulated in one plaque type, and up-regulated in the other. This strong correlation in expression pattern was confirmed by Spear-

177

man’s q correlation analysis (q = 0.73, Pvalue < 0.0001, n = 69). However, in general it should be noted that the mean level of up-regulation was higher in acute (mean 19.9fold, CV 70%, median 17.6-fold) compared with chronic active (mean 2.4-fold, CV 56%, median 2.3-fold) tissue. In contrast, no significant difference was identified in the mean level of genes down regulated in both acute (mean 60.4-fold, CV 99.8%, median 25.6-fold) and chronic active (median 67.1-fold, CV 173.5%, median 16.9-fold) tissues (Fig. 3). Seven (9.8%) of the genes differentially regulated in both chronic and acute tissues serve cytoskeletal, motility and adhesion functions. Of these, five (71%) were down regulated in chronic active (mean 309.7-fold, CV 130.0%, median 404.2-fold) and acute (mean 118.4-fold, CV 109.0%, median 129.5-fold) tissue: including the axonal transport regulator, dynactin (DCTN) 2 [37], and the neuronal growth regulator, microtubule associated protein, RP/ EB Family, Member 3 (MAPRE3) [53]. Both villin 2 (VIL2) (Ezrin), which is a marker of T-cell activation [63], and the neuronal embryonic factor, tubulin b5 (TBB5) [17], were both up-regulated in MS tissue. Notably, no significant difference ( Pvalue < 0.05) was observed in the expression patterning of these cytoskeletal factors between chronic active or acute tissue (Table 4). Twenty-three (32.9%) genes differentially regulated in plaque tissue have roles in development, proliferation, as well as broad metabolic functions. Nine (39.1%) of these genes were down regulated in both chronic active (mean 35.8-fold, CV 97.0%, median 17.7fold) and acute (mean 20.4-fold, CV 62.0%, median 20.6fold) tissue; including the small glycoprotein, Palmitoylprotein Thiolesterase 1 (PPT1) [27], and factors implicated in regulation of cell cycle progression, such as cyclin-dependentkinase inhibitor 3 (CDKN3) [28]. Thirteen (56.5%) were up regulated in both chronic active (mean 2.6-fold, CV 65.4%, median 2.1-fold) and acute (mean 17.7-fold, CV 40.6%, median 18.1-fold) tissue. These included glutathione peroxidase (GPX) 1, which has diverse roles in free radical metabolism [16], and factors associated with glyco/lipoprotein synthesis such as the phosphorylase, glycogen phosphorylase, brain-type (PYGB) [24]. Notably, a significant difference was observed in the expression pattern of developmental, proliferation and metabolic factors, ( Pvalue < 0.001), between chronic or acute plaques (Table 4). Nine (12.9%) of the genes differentially regulated in MS chronic active and acute lesions have immunological functions. Of these, six (66.7%) were up regulated in chronic active (mean 1.7-fold, CV 84.0%, median 2.1-fold) and acute (mean 25.7-fold, CV 85.8%, median 18.5-fold) tissue, while three were down regulated in chronic active (mean 10.5-fold, CV 56%, median 7.9-fold) and acute (mean 81.1fold, CV 108%, median 46.3-fold) tissue. The list includes: the nuclear factor, interferon regulatory factor 1 (IRF1), required for B-cell differentiation [85]; as well as several interferon inducible genes which were up-regulated in both tissues including, chemokine, CXC motif, ligand 10 (CX CL10), which has functions in adhesion and motility [3], and

178

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

Fig. 3. (A) Representative real time Q-PCR trace (iQ iCycler System) of PMM1 cDNA (HSB3163) showing product amplification. Shown: CT at linearity (arrow); Trace 1, 18S rRNA control tissue (CT, 15.9); Trace 2, 18S rRNA MS tissue (CT, 15.6); Trace 3, PMM1 control (CT, 19.5); and Trace 4, PMM1 plaque (CT, 21.2). DCTPMM1 control = 4, DCTPMM1 plaque = 5.6, DDCTPMM1 = 1.6, Putative fold determination (21.2) = 2.3-fold up. (B) Representative trace following melt curve analysis of PMM1 product at end point: 18S rRNA (85 jC), and PMM1 (94 jC). (C) 10% PAGE, PMM1 PCR product at end point: Lanes 1 and 3 cDNA control; and Lanes 2 and 4, cDNA MS plaque; Lanes 5 and 6, no template control (H2O); Lane 7, genomic DNA, positive template control. a, PMM1 genomic PCR product (441 bp); h, PMM1 PCR product; and m, 18S rRNA cDNA product (74 bp).

the inflammatory cytokine receptor, interleukin 10 receptor, b (IL10RB) [72]. Interestingly, while the major histocompatibility complex, class 1, A (HLA-A) [70] was up-regulated in acute tissue (40.1-fold), no significant difference was identified in the chronic active tissue. Several T- and B-cell markers were also up regulated, including the T-lymphocyte maturation-associated protein (MAL) [46]. While there was no significant difference in the pattern of down regulation, a

significant difference ( Pvalue < 0.01) was observed in the expression pattern of immunological factors up regulated, in acute and chronic active tissue. Fourteen (20%) genes regulated in both acute and chronic active tissue have functions in the regulation of protein function and hormone activity. Seven (50%) are up regulated in chronic active (mean 2.3-fold, CV 133.8%, median 2.2-fold) and acute (mean 26.9-fold, CV 65.0%, median

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183 Table 4 Summary gene expression patterns in acute and chronic active tissue Gene categories

n

Acute

Chronic active

Mean CV (-fold) (%) CYT/ADH

z 2 9.9 # 5 118.4 PRO/MET/DEV z 13 17.7 # 9 20.4 IMM z 6 25.7 # 3 81.1 PRA/HR z 7 26.9 # 7 47.8

Median Mean CV (-fold) (-fold) (%)

91.5 9.9 1.9 109.0 129.5 309.7 40.6 18.1* 2.6 62.0 20.6* 35.8 85.8 18.5* 1.7 108.0 46.3 10.5 65.0 23.7* 2.3 91.6 25.6 34.3

Median (-fold)

14.9 1.9 130.0 404.2 65.4 2.1* 97.0 17.7* 84.0 2.1* 56.0 7.9 133.8 2.2* 120.8 23.2

n, number of genes; z, up-regulated compared to normal tissue; #, downregulated when compared with normal tissue. * Significantly different ( P < 0.05 by Mann Whitney, U) between acute and chronic tissues.

23.7-fold) tissue. These include: members of the ubiquitin pathway, such as the X-linked ubiquitin activating enzyme (UBE) 1 [11], the cysteine proteinase, calpain small subunit (CAPNS1) [54], and cysteine protein inhibitor, cystatin C (CSTC). Seven (50.0%) of genes were down regulated in both chronic active (mean 34.3-fold, CV 120.8%, median 23.2-fold) and acute (mean 47.8-fold, CV 91.6%, median 25.6-fold) plaque cDNAs. Interestingly, the regulators of hormone activity were all down regulated: including, the insulin receptor (INSR) [68], and the neuronal thyroidhormone responsive gene (ZAKI4) [50]. While there was no significant difference in the pattern of gene down regulation, a significant difference ( Pvalue < 0.01) was identified in the pattern of gene up regulation. 3.2. Genes uniquely differentially regulated in chronic active MS plaques A list of genes 39 (32.1%) regulated specifically in chronic active plaque tissue (and not also expressed in acute plaques) is listed in Table 3. Eight (20.5%) have cytoskeletal and adhesion functions. Of these, four (50%) were up regulated in MS chronic active tissue, including: the neuronal microtubule associated protein (MAP) 4 [14]. Four (40%) genes were down regulated, including the astroglial intermediate filament protein, glial fibrillary acidic protein (GFAP) [10]. Fifteen (38.5%) genes identified in chronic active tissue have roles in development, proliferation and metabolism. Seven (46.7%) are up regulated: including the peripheral myelin protein (PMP) 22, which is linked to hereditary neuropathy [1]; activation protein, ETA isoloform (YWHAH), implicated in mediating neuronal responses to narcotics [51]; and the brain-type, creatine kinase (CKB) [12]. Eight (53.3%) were down regulated, including MAPK9, which is required for T-cell differentiation [18]. Five (12.8%) of the genes preferentially up regulated in chronic active tissue have immunological roles, including: the heat-shock 70 kDa protein 1A (HSPA1A) [15], b-2 microglobulin (B2M), required for HLA expression [61];

179

and the inflammatory mediator, bradykinin receptor B2 (BDKRB2) [30]. Five (10%) genes have functions in the regulation of protein and hormone activity, four (80.0%) of which were up regulated such as the parathyroid hormone receptor (PTHR1) [33]. 3.3. Genes uniquely differentially expressed in acute tissue A list of genes 31 (22.3%) regulated specifically in acute tissue is listed in Table 3. Six (19.4%) have cytoskeletal and adhesion functions. Both the GTPase-activating protein, Rho, 1 (ARHGAP1) [38], and the brain specific, Tubulin a [76], were up regulated in MS acute tissue. The rest of the genes in this category were down regulated. Eleven (35.5%) have roles in development, proliferation and metabolism. Of these seven (63.6%) are up regulated, including the serine/ threonine kinase, pictare protein kinase 3 (PCTK3) [57]. Four (36.6%) are down regulated including the candidate Alzheimer marker, amyloid b A4 precursor protein-binding, family A, 1 (APBA1) [7], and the interferon inducible, PRKR Inhibitor (PRKRIR) [23]. Eight (25.8%) genes have functions related to the regulation of protein and hormone activity. Those up regulated include, the GTP binding, RAS-associated protein (RAB2) [75], and the hypoxia inducible factor, adrenomedullin (ADM) [77]. The myeloid factor, related to the N terminus of TRE (RNTRE) [39], and the protein inhibitor of the activated signal transducer and activator of transcription (STAT) 1 (PIAS1) [82], were both down regulated. 3.4. Q-PCR validation of array analysis Each of the genes chosen for Q-PCR examination demonstrated a mean degree of regulation of at least 1.5-fold in MS plaques by array analysis (Table 5). The genes chosen for Q-PCR investigation included, transferrin (TF) [6], glutathione peroxidase 1 (GPX1) [16], glutathione S-transferase (GSTP1) [25], superoxide dismutase (SOD1) [79], which have important functions in regulating free radical metabolism, as well as the MS associated autoantigen, crystallin aB (CRYAB) [73], the brain specific glycosylation factor, phosphomannomutase (PMM1) [31], and TBB5 [17]. Following Q-PCR analysis the mean quantitative degree of fold regulation for all genes tested was found to be significantly ( Pvalue < 0.0001) higher in acute (mean 13.1-fold up, CV 39.2%, median 13.5-fold up), compared with chronic active (mean 1.2-fold up, CV 97.2%, median 1.4-fold up) plaque tissue. This correlated well with the results of array analysis for the same group of seven genes investigated in acute (mean 19.6-fold up, CV 35.2%, median 16.3-fold up) and chronic active (mean 2.2-fold up, CV 31.4%, median 2fold up) tissues. During analysis, the average variation between repeat determinations of CT (for the same matched control and test cDNA) on the same plate was less than 1%, although patient variation (CV) was much larger (Table 5). The average

180

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

Table 5 Comparison of the results of array hybridisation and Q-PCR analysis TF Q-PCR CAa (n = 3) Acute (n = 2)

2.3zb (47.8%)c 11.2z (2.6%)

GPX1 Array 2.9z 32.8z

Q-PCR 2.2z (13.6%) 18.9z (11.6%)

GSTP1 Array 1.4z 20.0z

SOD

CRYAB

Q-PCR

Array

Q-PCR

2.3z (4.3%) 5.5z (18.2%)

3.2z

NSDd (137.5%) 16.4z (4.3%)

NAe

Array 2.0z 14.0z

PMM1

Q-PCR

Array

Q-PCR

1.4z (78.6%) 8.1z (4.9%)

NA

NSD (125.6%) 18.4z (107.3%)

19.3

TBB5 Array 1.7z 15.1z

Q-PCR NSD (130.2%) 13.5z (112.8%)

Array 2.1z 16.3z

z, up regulated compared with normal tissue; #, down regulated when compared with normal tissue. Pearsons Bivariate correlation analysis showed a significant association between the results of Q-PCR and array hybridisation analysis across the 7 genes and two different cDNA groups tested (n = 14, r = 0.75, Pvalue < 0.01). a CA, chronic active. b Mean. c Coefficient of variation. d NSD, no significant difference ( Pvalue < 0.05, by t-test analysis). e Data not available.

difference between array and Q-PCR analysis, in chronic active cDNA, across the seven array-selected genes was 1.1fold (CV 103.8%). This represents less than a cycle difference, and is well within the error of real time detection of PCR product. By contrast, acute tissue showed a much larger difference between array and Q-PCR determination of fold (difference in the mean fold, 5.2-fold, CV 185.1%). In this case the result is strongly influenced by two key outliers: TF and CRYAB (Table 5). For these two transcripts, the relative error can be attributed to the relative abundance (rarity) of specific target, and (to a lesser degree) the quality of cDNA. Both factors would negatively influence hybridization kinetics for array analysis, and comparative Q-PCR, determination of fold [62]. Despite these outliers Pearson correlation analysis showed a significant correlation between the results of array analysis and Q-PCR (r = 0.75, Pvalue < 0.01, n = 14).

4. Discussion Multiple sclerosis lesions, with their regions of cell death, and ‘silent’ zones provide a special challenge for studies in microarray analysis [9,65,67]. Previous authors have sought to investigate global gene expression changes in MS plaques using a variety of control tissues, and array strategies. These have included both cDNA and oligo based arrays, which have been subjected to a variety of probes (fluorescent/ radiolabel) [13,41,52,83,84]. Critical to all these studies is the use of control tissues, which have varied from pooled controls to control tissues from single individuals. The result of these divergent approaches has been the identification of a range of very diverse MS associated genes. Initial array experiments used a restricted set of controls, including normal white matter from the same individual [83]. In later studies, experimental allergic encephalomyelitis (EAE) mouse models were used to validate arrayselected genes, however, it seems that such approaches favored an immunological model of MS plaque pathology [41,48,84]. In Lock et al. four MS samples were used to

compare acute with chronic lesions, using two controls and comparative cDNA arrays. Genes identified as potentially important in MS by their differential expression included the granulocyte macrophage colony stimulating factor (GMCSF), as well as the FCg receptor. In contrast, Chabas et al. [13] compared two non-normalized MS libraries (Chronic active and chronic silent pooled and acute) with one normal brain library. This led to the identification of osteopontin, as well as CRYAB as markers of MS pathology. In our study, we have used five plaques and analyzed gene expression differences for each plaque tissue with age and sex matched controls, separately. The results were then pooled resulting in the results shown in Table 5. This was done to identify gene expression differences that characterize different plaque activities (chronic active or acute), independent of variation that might be due to the use of inappropriate controls. This approach also resulted in the identification of CRYAB as significant marker of different MS plaque pathology [52,73]. Notably, osteopontin, which has been identified as an MS target gene [13] was not present on our array, and recent observations suggest that it may not be a significant marker of MS pathology [8]. CRYAB in contrast is closely associated with MS pathology, is highly immunogenic, expressed by oligodendrocytes and astrocytes and is not usually found in unaffected myelin [78]. Following Q-PCR analysis we also confirmed its expression in both chronic active and acute plaque tissues. The inability of microarray analysis to detect its expression in both tissues suggests that CRYAB may be expressed at a lower relative level of abundance in chronic active tissue. Other autoimmune related genes we have identified include components of the myelin sheath, such as the encephalitogenic, MBP [80] and HLA-A [19]. Interestingly, both MBP and HLA-A showed a common pattern of gene expression, and this may be due to the up regulation of myelin regulator proteins in the margin of plaque tissues [8]. The differential regulation of these genes in chronic active and acute tissue may reflect differences in the size of the margin (oligodendrocyte population) [8]. The inflammatory response in MS is mediated by T- and B-lymphocytes; so it

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

is not surprizing that many markers of T-cell maturity were also identified, and these included the marker of T-cell activation and autoantigen, MAL [2]. Factors, which enhance antigen presentation such as VIL2, were also up regulated in MS plaque tissue [63]. In this study, several interferon inducible factors were also differentially regulated. These included the chemotactic ligand CXCL10, found in the cerebrospinal fluid of MS patients; it acts as a ligand for the T-cell receptor, CXCR3 [71]. In contrast, other interferon inducible factors were down regulated in this study, including the interferoninduced transmembrane protein 3 (IFITM3) as well as the interferon regulator factor (IRF1) [49]. Recent studies suggest that the down regulation of interferon inducible factors, and resistance to interferon in MS, may be due to impaired signal transducer and activator of transcription (STAT) 1 signaling [21,36], and this is supported in our study by the observed decrease in expression of the protein inhibitor of activated STAT-1 (PIAS1). Interferon signaling, as a critical aspect of MS pathology is also important in regulating genes associated with the genes associated with oxidative stress and tissue damage, via the inducible nitric oxide synthase (iNOS) [20]. Several genes are expressed in response to free radical metabolism and as protection from oxidative damage, and this is an important aspect of MS pathology [42]. Candidate genes include TF, SOD1, GPX1 and GSTP1. In our study, each of these genes displayed a common pattern of regulation associated with MS plaque pathology by array analysis, and we were able to investigate their expression more sensitively using Q-PCR (Table 5). TF, which provides protection from iron, may be expressed as a response to iron accumulation in MS plaques [44]. SOD1 has also been linked to Amyotrophic lateral sclerosis (ALS) [55], and its effects may be mediated by GPX1; which is also up regulated in MS lesions [64]. While mutations in GSTP1 have been associated with MS susceptibility [34], and its link with Parkinson’s disease [45] suggests that it has a role in maintaining neuronal homeostasis. Finally, the use of QPCR analysis allowed us to investigate TBB5 as a candidate marker of MS pathology. Our result, which indicated it, was up regulated in MS patients, appears to contradict the published result of Lock et al. [41], which showed down regulation of this transcript. Q-PCR analysis provided impendent validation of our result, and suggests that all array data must be interpreted in the context of the particular method, sample size and validation approach used. In summary, many factors can contribute to MS plaque variance such as inflammation stage, plaque localization in the brain, degeneration– regeneration ratio, and the stage of the plaque. In this study, we have minimized plaque variance by selecting patients that all were assessed as suffering from the same clinical course: secondary progressive (SP). Finally, a comparison of data obtained from different microarray studies can provide a real insight in the understanding of genes and gene pathways generally

181

important in MS. This data can also be compared with published structural genetic alterations already associated with MS susceptibility, and pathology.

Acknowledgements Specimens were obtained from the Human Brain and Spinal Fluid Resource Center, Los Angeles, CA, USA, Founding Director (WWT) is an author. This research was supported by the Rebecca L. Cooper Foundation and by the Griffith University Research Fund. L.T. is supported by a Lindsay Yeo Multiple Sclerosis Research Scholarship, whilst A.M. and K.A. were supported by Australian Postgraduate Research Awards. We would like to take the opportunity to thank Sonya Webb of the University of Queensland and Stephanie Williams of QIMR for aiding in access to normal brain tissue and customized array slides, respectively.

References [1] N.K. Aarskog, C.A. Vedeler, Real-time quantitative polymerase chain reaction: a new method that detects both the peripheral myelin protein 22 duplication in Charcot-Marie-Tooth type 1A disease and the peripheral myelin protein 22 deletion in hereditary neuropathy with liability to pressure palsies, Hum. Genet. 107 (2000) 494 – 498. [2] M.A. Alonso, S.M. Weissman, cDNA cloning and sequence of MAL, a hydrophobic protein associated with human T-cell differentiation, Proc. Natl. Acad. Sci. U. S. A. 84 (1987) 1997 – 2001. [3] A.L. Angiolillo, C. Sgadari, D.D. Taub, F. Liao, J.M. Farber, S. Maheshwari, H.K. Kleinman, G.H. Reaman, G. Tosato, Human interferon-inducible protein 10 is a potent inhibitor of angiogenesis in vivo, J. Exp. Med. 182 (1995) 155 – 162. [4] T. Araki, J. Milbrandt, Ninjurin, a novel adhesion molecule, is induced by nerve injury and promotes axonal growth, Neuron 17 (1996) 353 – 361. [5] J.C. Benger, I. Teshima, M.A. Walter, M.G. Brubacher, G.H. Daouk, D.W. Cox, Localization and genetic linkage of the human immunoglobulin heavy chain genes and the creatine kinase brain (CKB) gene: identification of a hot spot for recombination, Genomics 9 (1991) 614 – 622. [6] E. Beutler, T. Gelbart, P. Lee, R. Trevino, M.A. Fernandez, V.F. Fairbanks, Molecular characterization of a case of atransferrinemia, Blood 96 (2000) 4071 – 4074. [7] G. Blanco, N.G. Irving, S.D.M. Brown, C.C.J. Miller, D.M. McLoughlin, Mapping of the human and murine X11-like genes (APBA2 and Apba2), the murine Fe65 gene (Apbb1), and the human Fe65-like gene (APBB2): genes encoding phosphotyrosine-binding domain proteins that interact with the Alzheimer’s disease amyloid precursor protein, Mamm. Genome 9 (1998) 473 – 475. [8] T. Blom, A. Franzen, D. Heinegard, R. Holmdahl, Comment on ‘‘the influence of the proinflammatory cytokine, osteopontin, on autoimmune demyelinating disease’’, Science 299 (2003) 1845. [9] L. Bo, S. Mork, P.A. Kong, H. Nyland, C.A. Pardo, B.D. Trap, Detection of MHC class II-antigens on macrophages and microglia, but not astrocyes and endothelia in active multiple sclerosis lesions, J. Neuroimmunol. 51 (1994) 135 – 146. [10] M. Brenner, A.B. Johnson, O. Boespflug-Tanguy, D. Rodriguez, J.E. Goldman, A. Messing, Mutations in GFAP, encoding glial fibrillary acidic protein, are associated with Alexander disease, Nat. Genet. 27 (2001) 117 – 120. [11] C.J. Brown, V.E. Powers, H.F. Willard, Localization of the A1S9T

182

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25] [26]

[27]

[28]

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183 gene to the proximal short arm of the X chromosome, Cytogenet. Cell Genet. 51 (1989) 970. M.G. Brubacher, J.C. Benger, G.D. Billingsley, M.H. Hofker, Y. Nakamura, R. White, D.W. Cox, A genetic linkage map of the distal region of human chromosome 14, (Abstract), Am. J. Hum. Genet. 45 (1989) A133. D. Chabas, S.E. Baranzini, D. Mitchell, C.C. Bernard, S.R. Rittling, D.T. Denhardt, R.A. Sobel, C. Lock, M. Karpuj, R. Pedotti, R. Heller, J.R. Oksenberg, L. Steinman, The influence of the proinflammatory cytokine, osteopontin, on autoimmune demyelinating disease, Science 294 (2001) 1731 – 1735. S.J. Chapin, J.C. Bulinski, Non-neuronal 21010(3) M(r) microtubule-associated protein (MAP4) contains a domain homologous to the microtubule-binding domains of neuronal MAP2 and tau, J. Cell Sci., Suppl. 98 (1991) 27 – 36. C.J. Cummings, Y. Sun, P. Opal, B. Antalffy, R. Mestril, H.T. Orr, W.H. Dillmann, H.Y. Zoghbi, Over-expression of inducible HSP70 chaperone suppresses neuropathology and improves motor function in SCA1 mice, Hum. Mol. Genet. 10 (2001) 1511 – 1518. J.B. de Haan, C. Bladier, P. Griffiths, M. Kelner, R.D. O’Shea, N.S. Cheung, R.T. Bronson, M.J. Silvestro, S. Wild, S.S. Zheng, P. Beart, P.J. Hertzog, I. Kola, Mice with a homozygous null mutation for the most abundant glutathione peroxidase, Gpx1, show increased susceptibility to the oxidative stress-inducing agents paraquat and hydrogen peroxide, J. Biol. Chem. 273 (1998) 22528 – 22536. K. Dennis, M. Uittenbogaard, A. Chiaramello, S. Moody, Cloning and characterization of the 5V-flanking region of the rat neuron-specific Class III beta-tubulin gene, Gene 294 (2002) 269. C. Dong, D.D. Yang, C. Tournier, A.J. Whitmarsh, J. Xu, R.J. Davis, R.A. Flavell, JNK is required for effector T-cell function but not for Tcell activation, Nature 405 (2000) 91 – 94. A. Dressel, J.L. Chin, A. Sette, R. Gausling, P. Hollsberg, D.A. Hafler, Autoantigen recognition by human CD8 T cell clones: enhanced agonist response induced by altered peptide ligands, J. Immunol. 159 (1997) 4943 – 4951. C. Espejo, M. Penkowa, I. Saez-Torres, J. Hidalgo, A. Garcia, X. Montalban, E.M. Martinez-Caceres, Interferon-gamma regulates oxidative stress during experimental autoimmune encephalomyelitis, Exp. Neurol. 177 (2002) 21 – 31. X. Feng, A.L. Petraglia, M. Chen, P.V. Byskosh, M.D. Boos, A.T. Reder, Low expression of interferon-stimulated genes in active multiple sclerosis is linked to subnormal phosphorylation of STAT1, J. Neuroimmunol. 129 (2002) 205 – 215. D.A. Francis, A.J. Thompson, P. Brookes, N. Davey, R.I. Lechler, W.I. McDonald, J.R. Batchelor, Multiple sclerosis and HLA: is the susceptibility gene really HLA-DR or -DQ? Hum. Immunol. 32 (1991) 119 – 124. M. Gale Jr., C.M. Blakely, D.A. Hopkins, M.W. Melville, M. Wambach, P.R. Romano, M.G. Katze, Regulation of interferon-induced protein kinase PKR: modulation of P58 (IPK) inhibitory function by a novel protein, P52(rIPK), Mol. Cell. Biol. 18 (1998) 859 – 871. T. Glaser, K.E. Matthews, J.W. Hudson, P. Seth, D.E. Housman, M.M. Crerar, Localization of the muscle, liver and brain glycogen phosphorylase genes on linkage maps of mouse chromosomes 19, 12 and 2, respectively, Genomics 5 (1989) 510 – 521. L.I. Golbe, Parkinson’s disease: nature meets nurture, Lancet 352 (1998) 1328 – 1329. E. Gottwald, O. Muller, A. Polten, Semiquantitative reverse transcription-polymerase chain reaction with the Agilent 2100 Bioanalyzer, Electrophoresis 22 (2001) 4016 – 4022. P. Gupta, A.A. Soyombo, A. Atashband, K.E. Wisniewski, J.M. Shelton, J.A. Richardson, R.E Hammer, S.L Hofmann, Disruption of PPT1 or PPT2 causes neuronal ceroid lipofuscinosis in knockout mice, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 13566 – 13571. J. Gyuris, E. Golemis, H. Chertkov, R. Brent, Cdi1, a human G1 and S phase protein phosphatase that associates with Cdk2, Cell 75 (1993) 791 – 803.

[29] P.J. Harrison, P.W.J. Burnett, P. Falkau, B. Bogerts, S.L. Eastwood, Gene expression and neuronal activity schizophrenia: a study of polyadenylated mRNA in the hippocampal formation and cerebral cortex, Schizophr. Res. 26 (1997) 93 – 102. [30] J.F. Hess, J.A. Borkowski, G.S. Young, C.D. Strader, R.W. Ransom, Cloning and pharmacological characterization of a human bradykinin (BK-2) receptor, Biochem. Biophys. Res. Commun. 184 (1992) 260 – 268. [31] L. Heykants, E. Schollen, S. Grunewald, G. Matthijs, Identification and localization of two mouse phosphomannomutase genes, Pmm1 and Pmm2, Gene 270 (2001) 53 – 59. [32] G.R. John, S.L. Shankar, B. Shafit-Zagardo, A. Massimi, S.C. Lee, C.S. Raine, C.F. Brosnan, Multiple sclerosis: re-expression of a developmental pathway that restricts oligodendrocyte maturation, Nat. Med. 8 (2002) 1115 – 1121. [33] H. Juppner, Molecular cloning and characterization of a parathyroid hormone/parathyroid hormone-related peptide receptor: a member of an ancient family of G protein-coupled receptors, Curr. Opin. Nephrol. Hypertens. 3 (1994) 371 – 378. [34] O.H. Kantarci, M. de Andrade, B.G. Weinshenker, Identifying disease modifying genes in multiple sclerosis, J. Neuroimmunol. 123 (2002) 144 – 159. [35] S. Kim, V. Westphal, G. Srikrishna, D.P. Mehta, S. Peterson, J. Filiano, P.S. Karnes, M.C. Patterson, H.H. Freeze, Dolichol phosphate mannose synthase (DPM1) mutations define congenital disorder of glycosylation Ie (CDG-Ie), J. Clin. Invest. 105 (2000) 191 – 198. [36] F. Koike, J. Satoh, S. Miyake, T. Yamamoto, M. Kawai, S. Kikuchi, K. Nomura, K. Yokoyama, K. Ota, T. Kanda, T. Fukazawa, T. Yamamura, Microarray analysis identifies interferon beta-regulated genes in multiple sclerosis, J. Neuroimmunol. 139 (2003) 109 – 118. [37] B.H. LaMonte, K.E. Wallace, B.A. Holloway, S.S. Shelly, J. Ascano, M. Tokito, T. Van Winkle, D.S. Howland, E.L. Holzbaur, Disruption of dynein/dynactin inhibits axonal transport in motor neurons causing late-onset progressive degeneration, Neuron 34 (2002) 715 – 727. [38] C.A. Lancaster, P.M. Taylor-Harris, A.J. Self, S. Brill, H.E. van Erp, A. Hall, Characterization of rhoGAP: a GTPase-activating protein for rho-related small GTPases, J. Biol. Chem. 269 (1994) 1137 – 1142. [39] L. Lanzetti, V. Rybin, M.G. Malabarba, S. Christoforidis, G. Scita, M. Zerial, P.P. Di Fiore, The Eps8 protein coordinates EGF receptor signalling through Rac and trafficking through Rab5, Nature 408 (2000) 374 – 377. [40] K.J. Livak, T.D. Schmittgen, Analysis of relative gene expression data using real-time quantitative PCR and the 2(DDC(T)) method, Methods 25 (2001) 402 – 408. [41] C. Lock, G. Hermans, R. Pedotti, A. Brendolan, E. Schadt, H. Garren, A. Langer-Gould, S. Strober, B. Cannella, J. Allard, P. Klonowski, A. Austin, N. Lad, N. Kaminski, S.J. Galli, J.R. Oksenberg, C.S. Raine, R. Heller, L. Steinman, Gene-microarray analysis of multiple sclerosis lesions yields new targets validated in autoimmune encephalomyelitis, Nat. Med. 8 (2002) 500 – 508. [42] F. Lu, M. Selak, J. O’Connor, S. Croul, C. Lorenzana, C. Butunoi, B. Kalman, Oxidative damage to mitochondrial DNA and activity of mitochondrial enzymes in chronic active lesions of multiple sclerosis, J. Neurol. Sci. 177 (2000) 95 – 103. [43] C. Lucchinetti, W. Bruck, J. Parisi, B. Scheithauer, M. Rodriguez, H. Lassmann, Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of demyelination, Ann. Neurol. 47 (2000) 707 – 717. [44] K. Mehindate, D.J. Sahlas, D. Frankel, Y. Mawal, A. Liberman, J. Corcos, S. Dion, H.M. Schipper, Proinflammatory cytokines promote glial heme oxygenase-1 expression and mitochondrial iron deposition: implications for multiple sclerosis, J. Neurochem. 77 (2001) 1386 – 1395. [45] A. Menegon, P.G. Board, A.C. Blackburn, G.D. Mellick, D.G. Le Couteur, Parkinson’s disease, pesticides, and glutathione transferase polymorphisms, Lancet 352 (1998) 1344 – 1346. [46] J. Millan, R. Puertollano, L. Fan, C. Rancano, M.A. Alonso, The MAL proteolipid is a component of the detergent-insoluble mem-

L. Tajouri et al. / Molecular Brain Research 119 (2003) 170–183

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56] [57]

[58]

[59]

[60] [61] [62]

[63]

[64]

[65]

[66]

[67]

brane subdomains of human T-lymphocytes, Biochem. J. 321 (1997) 247 – 252. C.L. Miller, R.H. Yolken, Methods to optimize the generation of cDNA from postmortem human brain tissue, Brain Res. Protoc. 10 (2003) 156 – 167. E. Mix, J. Pahnke, S.M. Ibrahim, Gene-expression profiling of experimental autoimmune encephalomyelitis, Neurochem. Res. 27 (2002) 1157 – 1163. M. Miyamoto, T. Fujita, Y. Kimura, M. Maruyama, H. Harada, Y. Sudo, T. Miyata, T. Taniguchi, Regulated expression of a gene encoding a nuclear factor, IRF-1, that specifically binds to IFN-beta gene regulatory elements, Cell 54 (1988) 903 – 913. T. Miyazaki, Y. Kanou, Y. Murata, S. Ohmori, T. Niwa, K. Maeda, H. Yamamura, H. Seo, Molecular cloning of a novel thyroid hormoneresponsive gene, ZAKI-4, in human skin fibroblasts, J. Biol. Chem. 271 (1996) 14567 – 14571. T. Muratake, S. Hayashi, T. Ichikawa, T. Kumanishi, Y. Ichimura, R. Kuwano, T. Isobe, Y. Wang, S. Minoshima, N. Shimizu, Y. Takahashi, Structural organization and chromosomal assignment of the human 14-3-3-eta chain gene (YWHAH), Genomics 36 (1996) 63 – 69. M.P. Mycko, R. Papoian, U. Boschert, C.S. Raine, K.W. Selmaj, cDNA microarray analysis in multiple sclerosis lesions: detection of genes associated with disease activity, Brain 126 (2003) 1048 – 1057. H. Nakagawa, K. Koyama, Y. Murata, M. Morito, T. Akiyama, Y. Nakamura, EB3, a novel member of the EB1 family preferentially expressed in the central nervous system, binds to a CNS-specific APC homologue, Oncogene 19 (2000) 210 – 216. S. Ohno, Y. Emori, K. Suzuki, Nucleotide sequence of a cDNA coding for the small subunit of human calcium-dependent protease, Nucleic Acids Res. 14 (1986) 5559. A. Okado-Matsumoto, I. Fridovich, Amyotrophic lateral sclerosis: a proposed mechanism, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 9010 – 9014. J.R. Oksenberg, S.E. Baranzini, L.F. Barcellos, S.L. Hauser, Multiple sclerosis: genomic rewards, J. Neuroimmunol. 113 (2001) 171 – 184. T. Okuda, J.L. Cleveland, J.R. Downing, PCTAIRE-1 and PCTAIRE3, two members of a novel cdc2/CDC28-related protein kinase gene family, Oncogene 7 (1992) 2249 – 2258. D. Oliva, L. Cali, S. Feo, A. Giallongo, Complete structure of the human gene encoding neuron-specific enolase, Genomics 10 (1991) 157 – 165. J.W. Prineas, R.O. Barnard, E.E. Kwon, L.R. Sharer, E.-S. Cho, Multiple sclerosis: remyelination of nascent lesions, Ann. Neurol. 33 (1993) 137 – 151. C.S. Raine, E. Wu, Multiple sclerosis: remyelination in acute lesions, J. Neuropathol. Exp. Neurol. 52 (1993) 199 – 205. P.J. Robinson, L. Graf, K. Sege, Two allelic forms of mouse beta-2microglobulin, Proc. Natl. Acad. Sci. U. S. A. 78 (1981) 1167 – 1170. R.B. Rose’meyer, A.S. Mellick, B.G. Garnham, G.J. Harrison, H.M. Massa, L.R. Griffiths, The measurement of adenosine and estrogen receptor expression in rat brains following ovariectomy using quantitative PCR analysis, Brain Res. Protoc. 11 (2003) 9 – 18. A. Roumier, J.C. Olivo-Marin, M. Arpin, F. Michel, M. Martin, P. Mangeat, O. Acuto, A. Dautry-Varsat, A. Alcover, The membranemicrofilament linker ezrin is involved in the formation of the immunological synapse and in T cell activation, Immunity 15 (2001) 715 – 728. T. Sakai, A. Inoue, C.S. Koh, S. Ikeda, A study of free radical defense and oxidative stress in the sera of patients with neuroimmunological disorders, Arerugi 49 (2000) 12 – 18. V. Sanders, A.J. Conrad, W.W. Tourtellotte, On classification of postmortem multiple sclerosis plaques for neuroscientists, J. Neuroimmunol. 46 (1993) 207 – 216. M. Schena, D. Shalon, R.W. Davis, P.O. Brown, Quantitative monitoring of gene expression patterns with a complementary DNA microarray, Science 270 (1995) 467 – 470. M. Schramm, P. Flakai, R. Tepest, T. Schneider-Axmann, R. Przkora, A. Waha, T. Pietsch, W. Bonte, Stability of RNA tran-

[68]

[69]

[70]

[71]

[72]

[73] [74]

[75] [76]

[77]

[78]

[79]

[80]

[81] [82]

[83]

[84]

[85]

183

scripts in post-mortem psychiatric brains, J. Neural Transm. 106 (3/4) (1999) 329 – 335. S. Seino, M. Seino, S. Nishi, G.I. Bell, Structure of the human insulin receptor gene and characterization of its promoter, Proc. Natl. Acad. Sci. U. S. A. 86 (1989) 114 – 118. S. Selvey, E.W. Thompson, K. Matthaei, R.A. Lea, M.G. Irving, L.R. Griffiths, Beta-actin—an unsuitable internal control for RT-PCR, Mol. Cell. Probes 15 (2001) 307 – 311. H. Shukla, G.A. Gillespie, R. Srivastava, F. Collins, M.J. Chorney, A class I jumping clone places the HLA-G gene approximately 100 kilobases from HLA-H within the HLA-A subregion of the human MHC, Genomics 10 (1991) 905 – 914. E. Sindern, T. Patzold, L.M. Ossege, A. Gisevius, J.P. Malin, Expression of chemokine receptor CXCR3 on cerebrospinal fluid T-cells is related to active MRI lesion appearance in patients with relapsing – remitting multiple sclerosis, J. Neuroimmunol. 131 (2002) 86 – 90. S.D. Spencer, F. Di Marco, J. Hooley, S. Pitts-Meek, M. Bauer, A.M. Ryan, B. Sordat, V.C. Gibbs, M. Aguet, The orphan receptor CRF2-4 is an essential subunit of the interleukin 10 receptor, J. Exp. Med. 187 (1998) 571 – 578. L. Steinman, Presenting an odd autoantigen, Nature 375 (1995) 739 – 740. O. Thellin, W. Zorzi, B. Lakaye, B. De Borman, B. Coumans, G. Hennen, T. Grisar, A. Igout, E. Heinen, Housekeeping genes as internal standards: use and limits, J. Biotechnol. 75 (1999) 291 – 295. E.J. Tisdale, W.E. Balch, Rab2 is essential for the maturation of preGolgi intermediates, J. Biol. Chem. 271 (1996) 29372 – 29379. S. Todd, S.L. Naylor, Dinucleotide repeat polymorphism in the human tubulin alpha 1 (testis specific) gene (TUBA1), Nucleic Acids Res. 19 (1991) 3755. T. Udono, K. Takahashi, M. Nakayama, A. Yoshinoya, K. Totsune, O. Murakami, Y.K. Durlu, M. Tamai, S. Shibahara, Induction of adrenomedullin by hypoxia in cultured retinal pigment epithelial cells, Invest. Ophthalmol. Vis. Sci. 42 (2001) 1080 – 1086. J.M. van Noort, A.C. van Sechel, J.J. Bajramovic, M. El Quagmiri, C.H. Polman, H. Lassmann, R. Ravid, The small heat-shock protein alpha-B-crystallin as candidate autoantigen in multiple sclerosis, Nature 375 (1995) 798 – 801. P. Wang, H. Chen, H. Qin, S. Sankarapandi, M.W. Becher, P.C. Wong, J.L. Zweier, Overexpression of human copper, zinc-superoxide dismutase (SOD1) prevents postischemic injury, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 4556 – 4560. K.G. Warren, I. Catz, L. Steinman, Fine specificity of the antibody response to myelin basic protein in the central nervous system in multiple sclerosis: the minimal B-cell epitope and a model of its features, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 11061 – 11065. B.G. Weinshenker, The natural history of multiple sclerosis, Neurol. Clin. 13 (1995) 119 – 146. R. Weiskirchen, M. Moser, S. Weiskirchen, M. Erdel, S. Dahmen, R. Buettner, A.M. Gressner, LIM-domain protein cysteine-and glycinerich protein 2 (CRP2) is a novel marker of hepatic stellate cells and binding partner of the protein inhibitor of activated STAT1, Biochem. J. 359 (2001) 485 – 496. L.W. Whitney, K.G. Becker, N.J. Tresser, C.I. Caballero-Ramos, V.V. Munson Prabhu, J.M. Trent, H.F. McFarland, W.E. Biddison, Analysis of gene expression in multiple sclerosis lesions using cDNA microarrays, Ann. Neurol. 46 (1999) 425 – 428. L.W. Whitney, S.K. Ludwin, H.F. McFarland, W.E. Biddison, Microarray analysis of gene expression in multiple sclerosis and EAE identifies 5-lipoxygenase as a component of inflammatory lesions, J. Neuroimmunol. 121 (2001) 40 – 48. G. Yamada, M. Ogawa, K. Akagi, H. Miyamoto, N. Nakano, S. Itoh, J. Miyazaki, S. Nishikawa, K. Yamamura, T. Taniguchi, Specific depletion of the B-cell population induced by aberrant expression of human interferon regulatory factor 1 gene in transgenic mice, Proc. Natl. Acad. Sci. U. S. A. 88 (1991) 532 – 536.