(12) Patent Application Publication (10) Pub. No.: US 2008/0311185 A1 Kollias Et Al

US 20080311 185A1 (19) United States (12) Patent Application Publication (10) Pub. No.: US 2008/0311185 A1 Kollias et al. (43) Pub. Date: Dec. 18, 2008

(54) DEREGULATED GENES AND/OR Publication Classification PROCESSES IN INFLAMMATORY (51) Int. Cl. ARTHRTIS A63L/7088 (2006.01) CI2O I/68 (2006.01) (76) Inventors: Georgios Kolias, Aghia Paraskevi A 6LX 9/27 (2006.01) (GR); Vassilios Aidinis, Nea A636/06 (2006.01) Smurni (GR); Yoshishide A638/12 (2006.01) Hayashizaki, Saitama (JP); Piero AOIK 67/027 (2006.01) Carninci, Saitama (JP) GOIN 33/53 (2006.01) A638/00 (2006.01) Correspondence Address: A6IP 9/02 (2006.01) HOFFMANN & BARON, LLP (52) U.S. Cl...... 424/450: 435/6: 514/44; 424/195.15; 6900 UERCHO TURNPIKE 514/9: 800/13: 800/9: 800/3: 514/12 SYOSSET, NY 11791 (US) (57) ABSTRACT The invention provides means and methods for determining (21) Appl. No.: 12/090,999 whethera sample is derived from an individual suffering from inflammatory arthritis or from an individual at risk of suffer ing there from, said method comprising measuring in said (22) PCT Filed: Oct. 19, 2006 sample the level of expression of at least one gene product of a gene of table A, table B or and comparing said expression (86). PCT No.: PCT/EP2006/O10228 level with a reference value. Further provided is a non-human animal model comprising an altered expression of a gene of S371 (c)(1), table A, table B or when compared to a reference non-human (2), (4) Date: Aug. 5, 2008 animal, and use of this model as or in an arthritis model. Further provided is an invitro assay for screening compounds (30) Foreign Application Priority Data for effect on a phenotype of synovial fibroblasts. The synovial fibroblast are preferably derived from a human or non-human Oct. 19, 2005 (GR) ...... 2005O100.526 animal suffering from inflammatory arthritis.

tatt re

8: Rab 14 w top ww. s tra is: &or Algy is: Airp13 six Mrp3 Saa3 ::::: Apf x:

Manco s: cles Pstrip Ciss w

s i: s Estat Tim: sis: s Gapdhan cdc.g f gap gax p Pagis s s spf2

irpfi

Asbp1 - wns

insif r tw Naif 8 at re cert s gf Ca Plann 9 t Gin is: Cacing Siae is is Patent Application Publication Dec. 18, 2008 Sheet 1 of 37 US 2008/0311185 A1

Fig. 1

LDDC:

s Patent Application Publication US 2008/0311185 A1

0.001 Gs

0.01 0.01 OA

0.01 Cd4 larco Patent Application Publication Dec. 18, 2008 Sheet 3 of 37 US 2008/0311185 A1

Wit SFS RASFS

Fig. 3 Patent Application Publication Dec. 18, 2008 Sheet 4 of 37 US 2008/0311185 A1

A m Synovial Hyperplasia/inflammation B Cartilage Destruction C E Bore Erosion O 4 as 2 3 2 2 O s O 3 2

hTNF*/gsn" htnF/-gsn'/.. h7 NF//gsn'?

Fig. 4 Patent Application Publication Dec. 18, 2008 Sheet 5 of 37 US 2008/0311185 A1

Cofilin

4 : 3 ORA s Wt

O , --, - Cl 50 25 12.5 50 25 12.5 ngful ngful ngful ngful ngful ngful WJ SF

Fig. 5

2.5

5c) 2.0 8 1.5 X C) 9 1.0

C)Y 0.5

WTWJ RA-WJ WT-SF RA-SF

Fig. 6 Patent Application Publication Dec. 18, 2008 Sheet 6 of 37 US 2008/0311185 A1

SF Adhesion Assay + FBS

wkw. s 5 sts w s s a.

SF Adhesion Assay-FBS

Fig. 7A Patent Application Publication Dec. 18, 2008 Sheet 7 of 37 US 2008/0311185 A1

Migration (FBS)

: s S w

e es s

w s e

Migration(-FBS)

Fig. 7B Patent Application Publication Dec. 18, 2008 Sheet 8 of 37 US 2008/0311185 A1

SF proliferation - 48h 35000 3OOOO 25000

20000 CP (?o 15000 10000 5000 O 8 8 8 8 L t t s o s o Treatment

Fig. 7C Patent Application Publication Dec. 18, 2008 Sheet 9 of 37 US 2008/0311185 A1

24h

24h TgSF

Fig. 7D Patent Application Publication Dec. 18, 2008 Sheet 10 of 37 US 2008/0311185 A1

CHXToxicity

O D5 Chx concentration (gfm)

0.5 g/ml. CHX

25 50 TNFConcentration ng/ml

Fig. 7E Patent Application Publication Dec. 18, 2008 Sheet 11 of 37 US 2008/0311185 A1

110 a wt 100

O 3O 60 1 OO 200 300 Etoposide (uM)

Fig. 7F

125

1 O gtg

2 5

O 0.01 0.05 0.1 0.25 0.5 1 Staurosporine (puM)

Fig. 7G Patent Application Publication Dec. 18, 2008 Sheet 12 of 37 US 2008/0311185 A1

Adhesion Assay

1 2 A C

1 B EU,S. 0.8 D -A DTg197 O s O6 B Lat-A 5 C ganti-TNFa 3. 0,4 D Wt o CC 0.2

O - - Tg197 anti-TNFa Wt Cell Type

Fig. 8A

Migration Assay

B C 0.35 S 03 Si 0.25 -A Tg 197 s 0.2 Bolatsia Lat-A s C E anti-TNFa c) 0, 15 DW't B 2s 0,1 0.05 2.

Tg 197 Lat-A anti-TNFa Wt Cell Type

Fig. 8B Patent Application Publication Dec. 18, 2008 Sheet 13 of 37 US 2008/0311185 A1 SF proliferation-FBS

; ;5 -; ; sy ; Treatment Fig. 9A

SF proliferation +FBS

75000

50000

25000

O c) D 3d Z ge H H S s Treatment Fig. 9B Patent Application Publication Dec. 18, 2008 Sheet 14 of 37 US 2008/0311185 A1

125 -A-Tg s 100 -O-Wt a 75 o (U 's 50 8 25

O O 0.002 0.01 0.05 Staurosporine (uM) Fig. 10A

125 -- Tg S --0- Lat-A E 75 . . o. anti-TNF -O-Wt 's 50 3 25

O O O.002 0.01 0.05 Staurosporine (uM) Fig. 10B Patent Application Publication Dec. 18, 2008 Sheet 15 of 37 US 2008/0311185 A1

Fig.11 (a) (b)

100

X 9. 10 C O : 1 t

0.1

Gelsolin Gelsolin

p C 0.04 p < 0.01 OO 100

X x 9. 3. E 10 E 10 5 5 () 1 (l) i 0.1 0.1

LST1 LST1

100

X X 2 () 10 5 5 " 3 3. io 0.1 CSTB CSTB Patent Application Publication Dec. 18, 2008 Sheet 16 of 37 US 2008/0311185 A1

Fig. 11, contd. (a) (b)

p < 0.01 100 100

O 10

0.1 0.1

100 100 X 6 t O 10

1. 8 1 o t 0.1 0.1

100 100

Dusp1 Dusp1 Patent Application Publication Dec. 18, 2008 Sheet 17 of 37 US 2008/0311185 A1

Fig. 11, contd. (a) (b)

p < 0.01 10 100 E 10 5 5

0.1 0.1

LSP1 LSP 1

p < 0.01 p = 0.01 10

X X Es G 1 5 5 : 0.1 O. O. f 0.01

Rab14 Rab14

p = 0.2937 p = 0.1607 1000 s g C o

SHIP2 SHIP2 Patent Application Publication Dec. 18, 2008 Sheet 18 of 37 US 2008/0311185 A1

Supplementary Fig. 1 Tester library "Driver" library Total RNA total RNA | Amplified M-FLC Amplified A-FLC cDNA library cDNA Library Phage DNA extraction in vivo plasmid excision Phage DNA extraction l Single stranded, anti-Sense Single stranded Single stranded, anti-sense cDNA insert (PCR amplified) plasmid preparation cDNA insert (PCR amplified) N-1 N-1 Low Cof hybridisation High CoT hybridisation (Normalization) (Substraction)

HAP Column 2nd strand synthesis Nortalized/Substracted Plamid cDNA libra Transformation Colony picking 3' end sequencing EST clustering Homology searches GO assignment Differential expression statistics

Subtractedlibrary Testerlibrary libraryDriver SequencinE"9 Clusters 0 RASF wt SF 7869 4097 L1 wt SF RASF 8602 4.186 2 wt WJ RAW 7753 3331 L7 RAWJ wt W. 3287 1740 LO,L1, L2, L7 27511 91.76

Subtractive cDNA libraries and large scale sequencing. a. Outline of experimental strategy for the preparation of subtracted cDNA libraries and the analysis of differential expression. b. Summary of normalized/subtracted cDNA libraries and Sequencing results, RA: Rheumatoid Arthritis, SF: Synovial Fibroblasts, WJ: Whole joints, Sequencing reads: N of clones sequenced, Clusters: N' of (redundant) clusters/genes, "Total number of non-redundant clusterslgenes Patent Application Publication Dec. 18, 2008 Sheet 19 of 37 US 2008/0311185 A1

Supplementary Fig. 2

99.00% 99.99%

: c . 9 5 N : 5 - : 25 g is N to Y ed 25 as 93 .

R R al

Comparison Specificity R-Value Cut offs LOWs. L1 99.00% 2,21715 L0 Vs. L1 99.99% 5.784.86 2 vs. L7 99.00% 2.40962 L2 Vs. L7 99.99% 5.70563 b

(a) Distribution of the R statistic for the different cutoffs and simulated pairwise comparisons, for the indicated significance levels (i.e. expected percentage of false positive calls per comparison). (b) Tabulated R-value Cut offs required to reach a (100-specifity) % false positive hits for each pair wise library Comparison made. Patent Application Publication Dec. 18, 2008 Sheet 20 of 37 US 2008/0311185 A1

Supplementary Fig. 3

Hybridization intensities of 20 PM & 20 MM oligonucleotides per gene y Affymetrix MAS5 Detection call & Signal intensity (background Substracted, average) Affymetrix MAS5 Scaling

Bioconductor Quantile normalization Y Gene Spring Empirical filtering Fold Change Model Differentially Expressed genes (at 90 and 95% significance levels)

Microarray data normalization and analysis outline. Patent Application Publication Dec. 18, 2008 Sheet 21 of 37 US 2008/0311185 A1

Supplementary Fig. 4 (a) Quantiles nomalization on separated oligonucleotide clip-subunits. (b) Reproducibility of technical dublicate samples in Affymetrix hybridisations. (c)"Sample type'specific Fold Change Models-derived from replicated chips.

Mu11K subunit A Mut1K subunit 8

2 : . . . . s. 6 if 2 3 '' is a s : o it 2.

oogolpôô?çokö?? •{•}{T}*****************??ae --—+º+TT*************F Patent Application Publication Dec. 18, 2008 Sheet 22 of 37 US 2008/0311185 A1

Supplementary Fig. 4, contd.

pSF - Specific Fold Change Model

S : 3. FC (pSF normal) Minimum expression . Fog5%) value -pSF FCM

WJ - Specific Fold Change Model

8 9. | so 5 O is 9 s FC (pSF diseased) FC (pSF normal) Minimum expression- FC (95%) value -pSF FCM Patent Application Publication Dec. 18, 2008 Sheet 23 of 37 US 2008/0311185 A1

Supplementary Fig. 5 Deregulated genes.previously reported in the literartire toExperiment be associated Assistant" with Rheumatoid text-mining Arthritis, software. as (a)these Pubmedds were identified of corresponding through the Biolab. publications; WJ: Whole Joints; SF: Synovial Fibroblasts (b) Schematic representation of text mining results. Red and green-colours indicaté-upregulated and downregulated and blue indicate WJ and SF respectively.

PUBMID

Bsg, 634773 11856830, 11733367, 11920402 1930331, 11801661,2504878, 3260544, 1713188, 1571092, 11549372, 10686511, 1378495, 7679058, 8162637, 80.12331, 1257.1842, 1691032,12746896, 11812022, 10685794, 1958435, 12384915, 1309559,7526870, 12417058, 10562301, 1616358, 8546720, 8546719,8849348, 11.014344, 981 1059, 1380987, 8568271, 89.12515,7686370, 9851271, 897.6638,9849315, 7561064,8316769, 7562767,928 1388, 10614777, 15020327, 12672180, 10219258, 8639.181, 12070677,9592864, 14749527, 15034063, 14730603, 12949960, 15142274, 146O7961, 9048866, 15146415, 1211.5225, 1237.5337, 15158620, 15140775, 15196219, 2575080,9058650, 14632917, 109.01283

15938744, 15081107, 12935777, 12869136, 12546136, 1246 1526

1590558

10O25098 10860865 10765925, 110972O7, 80 15285, 11062605, 11035124, 10990217, 11248661, 10914841, 12005365, 11040455, 11561108, 8030738, 8086445,8129774, 12237470, 12296867, 12296865, 10943872, 11824946, i2O51403, 11665967, 12571631, 11147173, 11762950, 11371663,7808967, 8137899, 12379519, 10693865, 12508384, 12447632, 12447637, 12879774, 12509619, i2O11369, 11508425, 12139676, 12682616,8546529,8849537, 8651985, 11710706, 12586482,788.0117,7612046, 12939814, 10724258, 10752493,892 1246, 10888709, 12009332, 11792884, 9297.577,9049977,887.7931, 8919 199, 9396371, 10415727,9177919,7490368, 12810425, 10365570, 9704635, 97.04636, 7639798,9645358, 12616343, 99.72954,9236673, 15040007, 99.18237, 10025098,12945803, 14611107,15320915, 1476O791, 15194590, 14740449, 9232430, 12913922, 8814070, 14523226, 10807502, 14705229, 10513800, 9513609, 46.96674, 10880256, 15130753, 11399.00, 15229943, 10397973, 10388526, 14714889, 9344881, 15144129, 10834863, 10464548, 12858344 3462903,7534965, 109055, 7688480,749768.9881748, 15146.432

6719058,6723125, 39.04845, 39.04889, 3484437,3484727, 2935508, 3954461, 3.954803, 2420896,3698338, 3084278, 35i 8030,3486637,2423091,352.1439, 3487899, 3743149,3638367,3462903,353.0991, 3532285, 3766605, 3778542,2946496,2946590,3782971, 3538345, 3788578, 3789816, 3791699, 3823794, 3552978, 3579385, 3495479,2820320,2443529, 3668974,2823368, 34997.07, 3499907, 3689.000, 3500676, 3423618, 350 1355,2963127,3257873,283O890, 3442962,2833,185, 32598O3,2837251,3133151,3389208,3261029, 30.58055, 3264781, 3065568,2650989, 2785142,278.5367,2469949, 2500276,274.1472,2472921, 2474900, 2764813, 2790893,2477604, 2797372,2805423,2810276, 2582890, 2591122, 2613159,2624928, 2319523, 258476,2339901, 2111124, 234.6005, 2349438,219,409, 2141558, 1974926, 1697.427, 2401 118, 21.70287, 2241286, 2146995,2147105, 1991211, 1996481,2077471, 1706901, 2024711, 2068540,2070569, 1872.046, 1680256, 1657009, 1657271, 1747138, 1768157, 1732316, 2980.129, 1839528, 1316744, 15.93589, 1595002, 1632661, 1395130, 1464075, 1484.414, 1295608, 307351 1285067, B2300 10,8252159,8287029, 8269781, 1285262,830205, 811 1198, 8208996, 8055195,8085060,7939 135,7955597,786.0508,7870343,7534965, 1366144,7548882, 10725071, 10759759, 109 11799, 10937386, 11052177, 11125324, 11149553, 11174137, 11244314, 11315937, 11352263, 11354304, 11354.563, 11550967, 1857.338, 11953970, 12063122, 12077276, 12139377, 12142347, 12212115, 1222O103, 1245.3313, 1247,5006, 12514522, 12682268, 12695150, 12723199. 12871189,7702398, 7747149, 7774O64,7621584, 8705012, 8553595,8572737, 8740816, 8742620, 8850166, 8902492, 9038384, 9093774, 9134821, 9150088, 9256031, 9402863, 94.71394, 9472668, 9548.303, 9550483, 9558172,9676177,97O6420,9782535,9794428,9805851,984.1702, 10423787, 104.87137, 10510406,5059642, 4114708, 4796805, 1078825, 1092716,806270, 806271,50336, 808229, 169295, 1099964, 11857.43, 1081730, 1202608, 1215899, 1082751, 1277573,819459, 950631, 961078, 786171,61163,970989,980923, 825053, 830776, 1022870, 849359, 853126, 857743, 140691, 301046, 327037, 881695, 302538,578869, 919557,921822, 922113, 597373, 412488,338220,591304,339851, 145832,629.093, 415670, 659.731,307274, 675210,278.177,99092, 82340, 718271, 364409,364609,364636, 105740, 374.144, 155958443249, 109097, 312646, 377474, 114636, 315208, 40532, 315842,517464, 93146,392746, 73.52929, 526442,394545, 119858, 7360040, 7362682,7389202, 7415585, 6774672, 8158571,6903558,81596.20, 744.1656,6160859, 7459014,7221787, 6908792, 6939879, 6788011, 6788211, 6166689, 67890 16,7258747, 7024339,6456623, 7027433,6794352,7306228,6914.866, 7035034, 7337970, 6978407,7073345,7340719, 6210468, 6978853, 7085116,7092.168, 6212656, 7099337, 7100806, 7051254,698 1226,6981473, 6981475, 6288942, 7125712, 6982590, 6983116,7174.786,6184771, 7178855, 7160109,6339939, 6339966, 6220848, 6221484, 66.01932, 6342551, 6847725,6857.174,6602812, 6603297, 6409476, 6879.093, 6879.095,6411401, 6883809, 6684168,6616816, 6623009, 6628574, 6316601, 66.06402, 6418086, 6652392, 6608359, 6608422,6201023, 6324537,6712297, 12761187, 12902784, 14748705, 14768941, 14987271,1281 1508,15067519, 15054157, 15241561, 15320901, 15310311, 6375688, 6376381, 66 11139, 638O835,638.1253,6384505,6384506, 6333061, 6333715, 6392549, 6514386, 6520829, 6395323, 3918543, 3158395,3892637, 4017290, 3849467,3875589, 4045842, 4048760, 4050141,3376836,2932466 98925O1, 15142265, 14872511 10753945, 98925O1 10765925, 12571631, 1453248,9794437, 12296887, 11207308, 122374.70, 12005365 15024704, 12746451

1092207s 1048656, 95.62601, 9022114, 1776991 Patent Application Publication Dec. 18, 2008 Sheet 24 of 37 US 2008/0311185 A1

Supplementary Fig. 5, contd.

Patent Application Publication Dec. 18, 2008 Sheet 25 of 37 US 2008/0311185 A1

Supplementary Fig. 6

Validation of expression profiling results (a, b, c) Semi-quantitative RT-PCR of the indicated genes from different cDNA amounts of arthritic?wt whole joint tissue (WJ), primary synovial fibroblasts (pSF) and immortalized synovial fibroblasts (ISF). Ethidium bromide stained PCR products were quantified with GelWorks 1D Advanced (v4.01) software and were normalized against the expression of B2M. (d) Western blot with a-gelsolin and a actin antibodies of whole cell extracts from arthritic and wt immortalized SFs upon SDS/PAGE and transfer to nitrocellulose. (e) Real time RT-PCR of the indicated genes from the indicated samples.

RA wt RA wt RA Wt 50ng 25ng 12,5ng

2, 5

15 Patent Application Publication Dec. 18, 2008 Sheet 26 of 37 US 2008/0311185 A1

Supplementary Fig. 6, contd.

B2M is is a RAWt RAWt RAWt 50ng 25ng 12,5ng

Patent Application Publication Dec. 18, 2008 Sheet 27 of 37 US 2008/0311185 A1

Supplementary Fig. 6, Contd.

iSF

as it is is e is RAWt RAwt RAwt 50ng 25ng 12,5ng

5 10 E. 9) O is a S. O,6 9 is SF is 9 Wt RA 1. O2 GSn Op is Patent Application Publication Dec. 18, 2008 Sheet 28 of 37 US 2008/0311185 A1

Supplementary Table 1

Differential expression results from subtractive libraries at 99.99% specificity. Only known genes are shown. SF: Synovial Fibroblasts, WJ: Whole joints, Up: Upregulated genes, Down: Downregulated genes. R refers to the differential expression likehood value of any gene to its corresponding library (SF: LOL1, WJ: L2L7). p30.0001

Expression Genbank - Gene description SF Up NM_011 198 Mm,3137 prostaglandin-endoperoxide synthase 2 10,341 SF Up NM008198 Mm653 histocompatibility 2, complement component factor B H2-Bf 8,1254. SF Up AK045288 Mm.1381.58 Lipoma preferred partner (Homo sapiens) homolog LPP homolog 7,3867 SF Up NM_023158 Mn.46424 Cxc chemokine ligand 16 Cxc16 5,9094 SFUp NM_011595 Mn4871 tissue inhibitor of metalloproteinase 3 Timp3 5,9094 SF Up NM008292 Mm,3195 hydroxysteroid (17-beta) dehydrogenase 4 21,924 SFoown NM 009636 Mn4665 A binding protein 1 11,693 SF own NM 012052 Mm.147693 ribosomal protein S3 10,623 SFDown NM010288 Mm.4504 gap junction membrane channel protein alpha 1 Gjal 10,394 SFDown NM 016685 Mm.4507 cartilage pligomeric matrix protein Comp 7,9199 SF Down NM_011354 Mm,227209 small EDRK-rich factor 2 Self? 7,1457 SFDown NM007475 Mm.5286 acidic ribosomal phosphoprotein PO Arbp 6,8872 SFDown NM009263 Mn260317 secreted phosphoprotein 1 Sppi 6,777 SF Down BF122010 Mm.196614 Elongation factor 1A-17 elongation factor Tu eEFA-1/EF-Tu 6,7577 SF Down NM_007653 Mm.4426 Cods3 antigen Cd53 6,6686 SFDown NM 016959 Mm,6957 ribosomat protein S3a Rps3a 6,5696 SF down NM 011653 Mr. 231463 tubulin, alpha 1 Tuba 6,0836 SFDown NM_026007 Mm.42960 Eukaryotic translation elongation factor 1 gamma Eef1g 5,8465 SF Down NM_008590 Mm.1089 mesodem specific transcript Mest 5,8465 SF Down NM 031167. Mm.882 interleukin 1 receptor antagonist 1?t 8,4823 W.J. Up AK031964 Mm.10017 RIKEN cDNA 5730406M06 gene 57304O6MO6Rik 6,0588 WJ Up NM 019467 Mm, 10747 allograft inflammatory factor 1 Af 8,4823 W.J. Up NM008598 Mnn, 1216 0-6-methylguanine-DNA methyltransferase Mgnt 8,4823 W.J. Up NM_023380 Mm.131406 SAM domain, SH3 domain and nuclear localisation signals, 1 Sams 9,6941 W.J. Up NM 152816 Mm.140013 dynamin 1-like Dm 5,8215 WJ Up NM_016852 Mm.1405 W domain binding protein 2 Wbp2 9,6941 W.J Up NM_010496 Mm, 1466 inhibitor of DNA binding 2 ldb2 10,906 WJ Up NM172119 Mr. 154427 deiodinase, iodothyronine type ill Dio3 7,2706 WJUp NM_0291.86 Mt. 159290 RIKEN cDNA 4930538D 17 gene 4930538D17Rik 7,2706 WJUp AKO14762 Mm, 59562 RIKEN cDNA 4930463G05 gene 4930463G05Rik 6,0588 WJ Up AKO45480 Mm.159895 RIKEN cDNAC030011JC8 gene CO30011J08Rik 7,2706 WJ Up NM 007974 Mm,614 coagulation factor (thrombin) receptor-like 1 F2r 7,2706 W.J. Up NM 011637 Mm, 17632 three prime repair exonuclease 1 Trex1 10,436 WJUp NM_010239 Mr. 776 ferritin heavy chain Fth 6,0588 WJUp NM 138303 MT. 178115 RIKEN cDNA 130001OK09 gene 13.00010K09Rik 6,0588 WJUp NM_018864 Mm.183042 inositol (myo-1 or 4)monophosphatase 1 impa1 16,965 WJUp NM_019643 Mm.18637 teratocarcinoma expressed, serine rich Tera 9,6941 W.J Up NM_009778 Mm. 19131 complement component 3 C3 6,397 WJUp NM 146252 Mm.19172 TBC1 domain family, member 13 Tbc1d3 8,4823 WJUp NM 145605 Mm, 196549 Kelch domain containing 4 Khdca 7,2706 WJUp NM 053196 Mm.197430 sideroflexin 2 Sfxn2 6,0588 WJUp NM_022323 Mr.202763 modulator of apoptosis 1 Moap 9,6941 WJUp BU526131 M.203870 Expressed sequence A413582 A43582 10,906 WJ Up NM009223 Mirth,203898 stannin Sn 14,541 W.J. Up NM 172465 Mt. 207367 RIKEN cDNA 9530098M12 gene 95.30098M12REk 7,2706 WJ Up NM009564 Mm.2095 zinc finger protein 64 Zip64 7,2706 WJUp NMOO9831 Mn.2103 cyclin G1 Cong1 6,0588 WJUp BY734237 Mirn 331072 weakly similar to B34087 hypothetical protein (H.sapiens) B34087 6,0588 WJUp NM_026506 Mm...21764 small nuclear ribonucleoprotein polypeptide G Snrpg 7,2706 WJUp NM_024233 Mm.21911 small fragment nuclease Smfn WJUp NM 133991 Mm,227141 Fts homolog(E. coli) Fis W. Up AK045814 M.22770 RIKEN cDNAB230312C02 gene B230312CO2Rik Patent Application Publication Dec. 18, 2008 Sheet 29 of 37 US 2008/0311185 A1

Supplementary Table 1, contd.

AKO891.94 Mm,2404.24 Expressed sequence A597.013 A597013 6,0588 WJ Up NM078479 Mm,24108 mitochondrial ribosomal protein S21 Mrps21 6,0588 WJ Up NM_023143 Mm,24276 complement component 1, r subcomponent C1r 6,0588 WJ Up NM 133801 Mm,24632 expressed sequence C76800 C76800 8,4823 WJUp NM 172609 Mm.246435 translocase of outer mitochondr. membrane 22 homolog (yeast) Tomm22 6,0588 WJUp AK050649 Mn247488 DNA segment, Chr 1, ERATO Doi 251, expressed 8,4823 WJUp NM_020329 Mm.25412 RIKEN cDNA 0610011H20 gene 6,0588 WJUp NM025667 Mm.25581 DNA segment, Chr 4, ERATO Doi 196, expressed 16,965 WJ Up AK082431 Mn258951 Splicing factor 3a, subunit 1 6,0588 WJUp NM 134116 Mm.26584 expressed sequence AA960287 7,2706 WJUp NM_017370 Mm,26730 haptoglobin Hp 8,4823 WJUp AK08904 Mm.27001 RIKEN cDNA C130032F08 gene C130O32F08Rik 9,6941 W.J Up NM_012025 Mm.27141 Rac GTPase-activating protein 1 Racgap1 8,4823 WJUp NM_013919 Mm,27510 ubiquitin specific protease 21 Usp21 10,906 WJUp NM 011026 Mm.27861 purinergic receptor P2X, ligand-gated ion channel 4 6,0588 WJUp NM_024471 Mr.28591 lipoic acid synthetase Lias 7,2706 WJUp NM 016692 Mm,29755 inner centromere protein Incenp 6,0588 WJUp NM_026521 Mm3035 RIKEN cDNA3110006P09 gene 3110006P09Rik 8,4823 WJUp NM024442 Mm30504 cytochrome P450, family 4, subfamily f, polypeptide 16 Cyp4f6 7,2706 WJUp NM 054102 Mm33764 Nd Nd1 6,0588 WJ Up NM026609 Mm34212 RIKEN cDNA 1520402O14 gene 1520402O14 Rik 7,2706 WJUp AK079163 Mm34220 RIKEN cDNA 9430.093107 gene 9430.093107Rik 6,0588 WJUp NM_026784 Mm34242 RIKEN cDNA 1110011E12 gene 1110011E12Rik 6,0588 WJUp NM_027878 Mm35450 RIKEN cDNA 1200002N14 gene 1200002N14Rik 9,6941 W.J Up NM_021281 Mm3619 cathepsin S Ctss 7,2706 WJ Up BQ031965 Mill.3925 S100 calcium binding protein A4 S100aa 6,0588 WJ Up BE987499 Mm397.52 RIKEN cDNA 29.00041A09 gene 29.00041A09Rik 7,2706 WJUp NM_029849 Mm.45078 RIKEN cDNA 4632433K11 gene 4632433K11Rik 7,2706 WJ Up NM_007659 Mm.4761 cell division cycle 2 homolog A (S. pombe) Codc2a 7,2706 WJ Up NM_007888 Mm.5114 dishevelled 2, dsh homolog (Drosophila) Dwi2 9,6941 WJ Up NM 008079 Mm.5120 galactosylceramidase Galc 7,2706 WJ Up NM 008307 Mm.5128 Hpall tiny fragments locus 9c H9C 7,2706 WJUp NM_026260 Mm60758 RIKEN cDNA 4930521E07 gene 493.0521E07Rik 8,4823 WJUp AK081497 Mm.72070 hypothetical protein C13002120 C13002120 6,908 WJ Up NM_010240 Mm.7500 ferritin light chain 1 Ft 7,2706 WJUp NM 011593 Mm.8245 tissue inhibitor of metalloproteinase 1 Timp1 14,541 WJUp NM 054088 Mm,86557 adiponutrin Adpn 7,2706 WJUp NM 010279 Mm.8B367 lial celline derived neurotrophic factor family receptor c1 Gfa1 9, 1872 WJDown NM_011403 Mm.7248 solute carrier family 4, member 1 Scala1 7,0671 W.J. Down NM 023644 Mm.154589 methylcrotonoyl-Coenzyme A cartboxylase 1 (alpha) Mcco1 7,0671 WJDown NM_008443 Mm.4535 kinesin family member 3A 6,3604 WJDown NM_007431 Mm.1265 alkaline phosphatase 2, liver Akp2 6,3604 WJDown AKO09375 Mm67737 DNA segment, Chr 11, ERATO Doi 736, expressed D11Erdf36e NM_017397 Mm,26703 OEADH (Asp-Glu-Ala-Asphis) box polypeptide 20 Dox20 NM 027445 MT.29092 RIKEN conA5730408C10 gene 5730408C10Rik Patent Application Publication Dec. 18, 2008 Sheet 30 of 37 US 2008/0311185 A1

Supplementary Table 2

Differential expressed genes common in two animal models of Rheumatoid Arthritis from oligonucleotide DNA microarray hybridizations at 95% specificity. SF: Synovial Fibroblasts, WJ: Whole joints, Up: Upregulated genes, Down: Downregulated genes. FC: Fold Change, ps0.05

FC Expression Genbank UniGene Description Common Tg.97 11,4 22,4 SF Up M96827 M.26730 haptoglobin Hp 10,1 25.5 SF Up W41745 Mim22673 Fc receptor, lge, high affinity, gamma polypeptide FCerig 10,0 16,6 SFUp u02298 M.3370 small inducible cytokine A5; rantes Cel5; Scya5; RANTES 9,1 11,7 SF Up M82831 Mm.2055 matrix metalloproteinase 12 Mmp12 78 10,0 SFUp L23.108 Mm.18628 CD36 antigen Co36 7,6 76 SF Up ET63426 Mr. 1293 tumor necrosis factor alpha Trfa 74 26,6 SF Up U19463 Mm,56971 tumor necrosis factor, alpha-induced protein 3 Tnfaip3 56 20 SF Up D13664 Mm.10681 osteoblast specific factor 2 (fasciclin-like) OSf2 5,6 3,3 SF Up z27231 Mm.4408 matrix metalloproteinase 9 Mmp9 52 19,0 SF Up L20315 Mm,3999 macrophage expressed gene 1 Mpeg1 48 23.2 SFUp AI326280 Mn.45436 lysozyme Lyzs 44 24,9 SF Up aa255186 M.3619 cathcpsin S Css 3,7 10,5 SF Up r14215 Mn22119 FC receptor, IgG, low affinity ill Fcgr3 36 19 SF Up X66473 Mm.5022 matrix metalloproteinase 13 Mmp13 33 6,3 SF Up aa178190 Mn.,281.07 ectonucleotide pyrophosphatase/phosphodiesterase 2 npp2 3.0 3,6 SF Up AAO59550 Mm28.107 ectonucleotide pyrophosphatase/phosphodiesterase 2 Enpp2 33 5,6 SF Up aa8.38539 Mm.34609 OZ onzin 33 2,6 SF Up C7742 Mm.216195 mVL30-1 retroelement mRNA sequence 12.00015M12Rik 3,1 42 SF Up x16490 M.5019 serine (or cysteine) proteinase inhibitor, clade B, member 2 Serpinb2 3.0 7.9 SF Up x80367 Mn2450 retinolbinding protein 1 Rbp1 27 17 SF Up aa874986 Mn.31098 expressed sequence Al451450 A451450 2,6 2,0 SF Up x03479 M. 14277 serum amyloid A3 Saa3 25 6,2 SF Up 38444 MT. 15793 T-cell specific GTPase Tgtip 24 5,2 SF Up x15591 Mn301.44 cytotoxictlymphocyte-associated protein 2alpha Cta2a 17 17 SF Up w/1236 Mm.2654 WD repeat domain 1 Wor1 16 17 SF Up AA038511 Mim28044 filamin-like protein Fina 14 9.3 SFU aa816578 Mn.6718 interferon-induced protein with tetratricopeptide repeats 1 fit 06 0.2 SFDown AAC97051 Mt. 1583 lymphocyte antigen 6 complex, locus C 04 0.2 SFDown AA027619 Mm.1583 lymphocyte antigen 6 complex, locus C 0.3 0.2 SFDown D86232 Mr. 1583 lymphocyte antiger 8 complex, locus C 03 0.3 SFDown ET61114 Mr. 1583 lymphocyte antigen 6 complex, locus C 05 0.3 SFDown x04120 P12894 Probable POL polyprotein pol 0.5 0,4 SFDown u58494 C61213 gag (intracisternal A-particle) gag 04 0,4 SFDown U58494 C61213 gag (intracisternal A-particle) gag 04 0.2 SFDown L02914 Mm.18625 aquaporin 1 Aqp1 04 0.4 SF Down X949.98 Mn41573 fibromodulin Fmod 03 0.3 SFDowth AA028265 Mm.41573 fibromodulin Frmod 0,1 03 SFDown M31418 Mm,218770 interferor activated gene 202A f202a 66,1 55.7 WJUp XO3505 Mm 14277 serum amyloid A (SAA)3 Saa3 C-type (calcium dependent, carbohydrate recognition domain) 43.8 73.3 WJUp C76739 Mm.24920 lectin, superfamily member 8 Cecsf8 40,6 19,4 WJ Up x03479 Mm.14277 serum amyloid A3 Saa3 35,6 16,6 WJUp m92649 Mm.2893 nitric oxide synthase 2, inducible, macrophage NOS2 27.7 36,6 WJUp U27287 Mm, 4660 chemokine (C-X-C motif)igand 5 Cxc5 23,3 10.9 WJU 38281 Mm.4562 immunoresponsive gene 1 1. Patent Application Publication Dec. 18, 2008 Sheet 31 of 37 US 2008/0311185 A1

Supplementary Table 2, Contd.

21,4 14,0 wJUp n19681 platelet-derived growth factor-inducible protein (JE) JE 17,9 142 WJ Up W77035 MT.41196 Tenascin N Trn 17.7 8,1 WJUp AA10355 M.22378 integrin alphaX tgax 17, 1 22, 1 WJUp c9779 M.2370 RIKEN conA 120.0011C15 gene 1200011C15Rik 13,4 10,4 WJUp W71124 Mm.10729 neutrophil cytosolic factor 2 NCf2 12.7 26,8 WJUp u02298 small inducible cytokine A5; rantes Cc5; Scya5; RANTES 122 11.9 WJUp M31690 Mr.3217 argininosuccinate synthetase 1 ASS1 120 84 WJUp m 1513 Mm,222830 interleukin 1 beta 1b. 11.0 99 WJUp L38281 Mm,4662 immunoresponsive gene 1 Irg1 10,8 13,9 WJUp AA038079 Mn2534 proline-serine-threonine phosphatase-interading protein 1 Pstpip1 10,6 73 wJUp 12297 Mm.16091 chemokine (C-C motif) ligand 7 Cov phospholipase A2 group will (platelet-activating factor 9,9 48 WJUp U3427 Mrn.9277 acetylhydrolase, plasma) Pla2g7 9.5 11,0 WJUp M82831 Mm,2055 matrix metalloproteinase 12 Mmp12 9,3 10.9 WJUp x66402 Mrm.4993 matrix metalloproteinase 3 Mmp3 9,2 13.7 WJUp AA44430 Mm 20225 nuclear factor ofkappa light polypeptide, p49?p100 Nfko2 9,0 43,8 WJUp jo4596 Mm21013 chemokine (C-X-C motif) ligand 1 Cxc1 8.2 19.7 WJUp U43384 Mm,200362 cytochrome b-245, beta Cybb 7.7 7,8 WJUp AA133105 Mm.195498 transmembrane 4 superfamily member 7 Tm4sff 7,7 29 WJUp aa137292 Mn.30205 chemokine (C-C motif) ligand 19 Cod 19 7s 12,8 WJUp AA039124 Mm.16783 phosphoglycerate mutase 1 Pgam1 7.2 71 WJUp 925.65 Mm3205 chemokine (C-X3-C motif) ligand 1 Cx3c.1 89 84 WJUp U925.65 Mm320s chemokine (C-X3-C motif) ligand 1 Cx3c.1 6.9 10,4 WJUp aab59528 Mm.28146 mevalonate (diphospho) decarboxylase Mwd 6,4 394 WJUp W5975 Tubulin Beta Tibb 6,3 52 WJ Up AAO15291 Mm.4.1331 RIKEN cDNA 111 0012EO6 gene 1110012E06Rik 6,1 44 WJUp m31236 Mm.26859 gap junction membrane channel protein beta 5 Gib5 6.0 53 WJUp Li2O315 Mm3999 macrophage expressed gene 1 Mpeg1 5.9 49 WJUp 21795 interleukin-2 receptor gamma chain II2rg; CD 132 5.9 9,4 WJUp AA10972 Mm,28088 mevalonate kinase Mwk 59 55 WJUp C80469 Mm,294.83 RIKEN conA 1600025H15 gene 1600025h 15Rik 5,8 6.4 WJUp f74294 Mm.882 interleukin receptor antagonist in 5,6 6,7 WJ Up AA6873.71 Mn21267.1 C1q and tumor necrosis factor related protein 3 C1qtnf3 5,5 6,0 WJUp AA089333 Mm,3619 cathepsin S Ctss 5,3 6,3 WJUp L16462 Mm.196731 B-cellleukemia/lymphoma 2 related protein A1a Bc2a1a 5.2 52 WJUp M65027 Mm.196617 glycoprotein 49A Gp49a 5.2 12, 1 WJUp W87123 similar to macrophage migration inhibitory factor 5.2 24,0 WJUp U20159 Mm.1781 lymphocyte cytosolic protein 2 5.2 64 WJUp AAO3783 Mm.45558 RIKEN conA 2010.008K16 gene 5,0 73 WJUp J47281 G-protein coupled receptor, bradykinin b1 subtype receptor 49 3,3 WJUp U59B07 stefin B; cystatin B

H2-M alpha chain, H2-Mbeta 2 chain, H2-M beta 1 chain, low W.J. Up U35323 molecular weight protein 2 2-Ma/b2b.1; Limp2 WJUp x80367 Mm.2450 retinolbinding protein 1, cellular Rbp 1 WJ Up AA051341 Mn3509 tumor necrosis factor induced protein 6 WJUp x52264 Mm.28973 intercellulaf adhesion molecule Carn1 WJUp w1156 Mm30241 interferon gamma inducible protein 30 fi30 WJUp w11156 Mm.30241 interferon gamma inducible protein 30 ifi30 WJUp D3564 Mm, 1068 osteoblast specific factor 2 WJUp W50.465 homologous to Class Il Major Histocompatibility antigen, M-a chain precursor WJUp x0.4648 Mm.10809 Fc receptor, IgG, low affinity Ib Fcgr2b WJUp x13333 Mm3460 CD14 antigen C14 WJUp M73748 Mm,2976 glycoprotein 38 Gp38 WJUp AA033858 Mm3020.1 ras homologgene family, member G Arhg WJUp AA04193 Mm326 RIKEN conA 1110038L14 gene 1110.038L14Rik Patent Application Publication Dec. 18, 2008 Sheet 32 of 37 US 2008/0311185 A1

Supplementary Table 2, contd.

3.9 4,6 WJUp Z27231 Mm.4406 matrix metalloproteinase 9 3,6 W. Up W67046 Mrm. 195508 chemokine (C-C motif) ligand 21c (serine) 2,4 W. Up U3S277 Mm.8884 nuclear factor of kappallight chain inhibitor, alpha colony stimulating factor 2 receptor, alpha, low-affinity 52 WJUp M85078 Mm.156264 (granulocyte-macrophage) sema domain, immunoglobulin domain (g), transmembrane 2,7 WJUp aa189937 Mm.33903 domain (TM) and short cytoplasmic domain, (semaphorin 4D Semaad 7,6 WJUp 73004 Mm, 1395 secretory leukocyte protease inhibitor Spi 3.2 WJUp AA106931 Mm30241 interferon gamma inducible protein 30 30 6,2 WJUp KO1925 Mm.175310 histocompatibility 2, class II antigen A, alpha H2-Aa 3.2 WJUp M5F890 MS53 histocompatibility 2, complement component factor B H2-Bf 3.2 WJUp AA16604 Mm.156919 cathepsin Z Ctsz 3.7 WJup M99054 acid phosphatase type 5 Acp5; TRAP 36 WJUp 90355 fascir Fscn1 49 WJUp AA023099 Mm.29879 deoxyuridine triphosphatase Dutp 36 WJUp 3.2370 Mm,302 SFFW proviral integration 1 Sipil 36 WJUp X66473 Mn.,502.2 matrix metalloproteinase 13 Mmp13 28 WJUp k01496 M653 histocompatibility 2, complement component factor B H2-Bf 25 WJUp 54984 Mm.1994.5 matrix metalloproteinase 14 (membrane-inserted) Mmp14 26 WJ Up cB0103 Mm.195387 DNA segment, Chr 14, ERATO Doi 310, expressed 33 WJUp wO1527 major histocompatibility class I antigen, l-A-beta. 2.9 WJUp ETS1664 FcgammarilB 26 WJUp AA013831 similar to Nudear Factor kappa BP100 Subunit 3, 1 WJUp W00832 Mm.175310 histocompatibility 2, classil antigen A, alpha H2-Aa 2.9 WJUp abO)9287 Macrosialin Code,8; gp10; Scard 44 WJUp 14215 Mm,22119 Fc receptor, IgG, low affinity II Fegr3 2,6 WUp 14951 Mm.1137 integrin beta 2 tgb2 15,3 WJUp AA060835 Mm.35016 histocompatibility 2, T region locus 23 H2-23 3.5 WJUp X94.444 Mm,3109 cathepsin K Ctsk 53 WJup 5242 Mm.16373 histocompatibility 2, class, locus DMa H2-DMa 3,3 WJUp AA0643 Mr.4.1 fatty acid binding protein 5, epidermai Fabp5 3.2 WJUp AA146437. Mm,3619 cathepsin S Ctss 48 WJUp AA161799 Mm.16783 phosphoglycerate mutase 1 Pgam1 3,3 WJUp AA146008 Mm.10729 neutrophil cytosolic factor 2 Mc2 34 WJUp 69135 Mm.144413 uncoupling protein 2, mitochondrial Ucp2 44 WJUp x0958 Mm,22564 histocompatibility 2, class antigen Ebeta H2-Eb 26 WJUp W10936 Mm.2970 lectin, galactose binding, soluble 3 gals3 28 WJ Up KO1923 Mm.175310 histocompatibility 2, class antiger A, alpha 2-Aa 27 WJUp X12905 Mm.8308 properdin factor, complement Pfc 30 WJUp k01923 Mm.175310 histocompatibility 2, class antigen A, alpha H2-Aa 2.9 WJ Up U19482 M.2271 chemokine (C-C motif) ligand 9 C9 2,4 WJUp X91144 Mim.22173 selectin, platelet (p-selectin) ligand Selpi 3,7 WJUp X84940 superoxide dismutase 3 Sod3 2,4 WJ Up W83564 similar to 5-lipoxygenase activating protein 3.7 W. Up W41745 Mm.22673 Foreceptor, IgE, high affinity, gamma polypeptide Fcertg 3.0 WJUp W40735 Mm301.69 EH-domain containing 1 Ed 2,9 WJUp AA037605 similar to vacuolar H(t)-ATPase 3,8 WJUp W13283 Mm.3109 cathepsin K Ctsk 2.9 WJUp W33721 Mm.90587 enolase 1, alpha non-neuron Ero1 18 WJUp M84324 Mm.29564 matrix metalloproteinase 2 Mmp2 2, 1 WJUp M96827 Mm.26730 haptoglobin hp 3.2 WJUp OOT55 Mm.8245. tissue inhibitor of metalloproteinase 1 Timp1 Patent Application Publication Dec. 18, 2008 Sheet 33 of 37 US 2008/0311185 A1

Supplementary Table 2, contd.

0,6 WJDown X67140 Mm35134 ATPase, Ca++ transporting, cardiac muscle, fast twitch 1 Atp2a1 0.5 0.5 WJ Down X61431 Mm.285 diazepam binding inhibitor Dbi 0,5 04 WJ Down X61452 Mm2214 septin 4 Sept.4 0.5 0.5 WJDown W34687 Mm214950 actin, alpha 1, skeletal muscle Acta WJDown aa)98356 Mm,536 actinin alpha 3 Acting solute carrier family 25 (mitochondrial carrier; adenine 0.4 0.5 WDown U27315 M.16228 nucleotide translocator), member 4 Slc25a4 04 04 WJ Down 3444 Mm.1155 tubulin, alpha 4 Tuba4 0,4 0,4 WDown U62545 Mm,284.80 FK506 binding protein 3 (25kD) Fkbp3 04 0.3 WDown M13227 Mn2899 preproenkephalin 1 Penk1 04 0,4 WJ Down W85566 similar to myosin light chain 2 0,4 0.3 WDown X67685 Mm.197678 myosin light chain, alkali, cardiac ventricles Mylc 04 0,3 WJDown x12507 Mm.16323 eukaryotic translation initiation factor 4A2 0,4 0.2 W. Down x53825 Mn.6417 CD24a antigen 0,4 0.5 WDown W36455 M4407 adipsin Adn 04 0,3 WDown X90875 M4038 fragile X mental retardation gene 1 autosomal homolog 0,4 0.6 WJDown W29418 Mm.1716 troponin C, fast skeletal Tcs 0,4 04 WJDown aaO983.56 Mm.5316 actinin alpha 3 Actn3 0,4 0.5 WJ Down W34915 M.39046 RIKEN cDNA 1110055E19 gene 1110C5519 Rik 0.4 0,4 WJ Down M13366 glycerophosphate dehydrogenase Gdc1 04 0,4 WJOown AA145371 Mm.46382 RIKEN cDNA 231 OO61 N23 gene 2310.061 N23Rik 0.4 04 WJ Down w15862 Mn., 18089 diacylglycerol O-acyltransferase 2 Dgat2 0.4 0.3 WDown m3227 M.2899 preproenkephalin 1 Perk 0,4 0,4 WJ Down x81600 beta-enolase subunit no3; Eno-3 0.3 0,5 WJDown k02241 alkali myosin light chains 0.3 0.5 WJ Down W11481 Mr.12829 calsequestrin 1 CaSq1 0.3 0.2 WJDown x67141 MT.2766 parvalbumin Pva 0.3 0.3 WJ Down AAO38O82 Mm,5316 actinin alpha 3 Acting 0.3 0,4 WJDown W294.66 MT-141702 myozenin 1 Myoz1 03 0,3 WJDown W11082 M.26550 phosphofructokinase, muscle Pikm 0.3 0.5 WJOown K00988 Mt.35531 myosin, heavy polypeptide 4, skeletal muscle Myh4 0.3 0,4 WJDown AA611262 M.26722 N-myc downstream regulated 2 Nd2 03 0.3 WJDown M14537 Mm.86.425 cholinergic receptor, nicotinic, beta polypeptide 1 (muscle) Chino1 03 0.4 WJ Down AA034827 similar to muscle creatine kinase 03 04 WJ Down ET63025 ADP-ribosyltransferase 2a 03 0.5 WJ Down aa522276 Mm.3.126 four and a halfLM domains 1 0.3 0,4 WJ Down AAO49662 Mm.2605 retinolbinding protein 4, plasma 0.3 0,3 WJDown aa(57994 Mn23742 nestin Nes 0.3 0,3 WJDown Z22661 apolipoprotein C1 Apoc1 0.3 0,3 WDown c8 1439 Mn200850 expressed sequence C81439 C81439 0.3 0.3 WJDown u08439 Mm.43824 cytochrome c oxidase, subunit Va. polypeptide 2 Cox6a2 03 0.5 WJDown W17917 M26550 phosphofructokinase, muscle Pikm 0.3 0.4 WJOown x04673 adipsin, Adn; DF 0.3 0.4 WJ Down U08439 Mm.43824 cytochrome c oxidase, subunit Va. polypeptide 2 Cox6a2 0.3 0.5 WJDown U77943 Mm, 14526 myosin light chain, phosphorylatable, fast skeletal muscle Mylpf 0.3 0.4 WDown 751.31 Mm,27881 upregulated during skeletal muscle growth 4 USmg4 0.3 0,4 WDown aac)33394 Mm.27806 muscle glycogen phosphorylase Pygm 0.3 0,4 WJ Down M27796 Mm300 carbonic anhydrase 3 Car3 03 0.3 WJ Down L06234 M.4418 calcium channel, voltage-dependent, L type, alpha 1S subunit Cacna 1s 0.3 0.2 WJDown w08218 M.7.11 troponin T1, skeletal, slow T1 0.3 0,4 WJ Down c79089 Mn22231 synuclein, gamma Sncg 03 0.5 WJDown u08.188 Mm.1721 lipase, homone sensitive Lipe 0,3 0,3 WJDown AA163065 Mm,5316 actinin alpha 3 Acting 0.3 0.3 WJDown W13030 Mm399.68 histidine rich calcium binding protein Hrc 0,3 0,3 WJ Down 23108 Mr.18628 CD36 antigen C38

WJDown W50088 Homologous to Sodium. /Potassiu?-transporting ATPase C-2 Patent Application Publication Dec. 18, 2008 Sheet 34 of 37 US 2008/0311185 A1

Supplementary Table 2, contd.

0,3 04 WJDown 2.8835 peroxisome membrane protein Pxmp2; PMP22 0,3 0,1 WJDown AA014295 Mm.30262 Kruppel-like factor 5 Kf5 0.3 0.3 WDown x83932 Mm.4519 ryanodine receptor 1, skeletal muscle Ryr1 0,3 0,1 WJDown W14545 Mm.3327 expressed sequence AW121776 AW121776 0.3 04 WJDown W29562 Mm.582 fatty acid binding protein 4, adipocyte fabp4 0,3 0.5 WJDown x59382 Mm.2766 parvalbumin Pya 0.3 0,4 WDown X95280 Mm.3283 G0/G1 switch gene 2 GOs2 03 0.5 WJDown JO4992 Mm.39469 troponini, skeletal, fast 2 inn2 0.3 0.3 WJDown aa462409 Mm.27227 RIKEN cDNA 5730469M10 gene 573O469M10Rik 0.2 0,3 WJ Down x67141 Mm.2766 parvalbumin Pva 02 03 WJ Down aao60862 Mm.286.38 SH3-binding domain glutamic acid-rich protein O2 0,2 WDown u31510 Mrn.31099 ADP-ribosyltransferase 1 At1 0.2 0.3 WJDown AA028501 Mm3841 cytochrome c oxidase, subunit willb Cox8b 0.2 0,4 WJDown u37222 Mm.3969 adipocyte complement related protein of 30 kDa Acrp30 02 0,3 WDown AA288376 Mm.1181 resistin Retin O2 0.3 WJOOwn Z22866 Mm.4103 myomesin i 0.2 0,6 WJ Down U39192 heparin-binding epidermal growth factor-like growth factor 0.2 0.3 WDown U58883 Mm.29030 sortin and SH3 domain containing 1 Sorbs1 0.2 0,4 WJDown ms3170 Mm.178 crystallin, alpha B Cryab 0.2 0,4 WJDown M73741 alpha-B2-crystallin Cryab; Crya2; Crya-2 0.2 0,4 WDown L264.79 Mm.2645 eukaryotic translation elongation factor 1 alpha2 Eefa2 0.2 0,3 WJDown aa888542 Mm,26722 N-myc Downstream regulated 2 Nd2 0.2 0.5 WJDown n29793 M.712 troponin C, cardiacskow skeletal c Serpina3c, 1A1; Kalbp; 0.2 0,1 WJ Down X61597 kallikrein-binding protein. Kkbp 0.2 0,3 WJ Down M61737 Mm.10026 fat specific gene 27 0.2 0,4 WJ Down U77039 Mn3126 four and a half LM domains 1 0.2 0,4 WJ Down 49470 Mm.14546 troponin T3, skeletal, fast 0.2 0.2 WJDown M23383 Mm.10661 solute carrier family 2, member 4 0.2 0,3 WJDown u04541 Mm.17306 tropomyosin 3, gamma 0.2 0,4 WJDown X73959 Mr. 83266 teasci? XB 0.2 0,4 WJ Down x51433 Mm.4550 ATPase, Na+K+ transporting, beta 1 polypeptide 0.2 0.2 WJDown AA162395 Mm,35531 myosin, heavy polypeptide 4, skeletal muscle 0.2 0,3 WJDown u76371 Mm.897.27 CD8beta opposite strand 02 0.2 WJDown U76618 Mn.6384 nebulin-related anchoring protein Nap 0.2 0.2 WJ Down 30844 fragment for repetitive element B2. 0.2 0,3 WJDown M23383 Mm.10661 solute carrier family 2, member 4 0.2 0,3 WJOown AA061886 Mm,57093 calcium channel, voltage-dependent, gamma subunit 1 Cacing 0.2 0,3 WJDown x03986 Mm.4583 cholinergic receptor, nicotinic, alpha polypeptide 1 (muscle) Chrina 1 0.2 0.2 WJ Down AA 137667 Mm.39046 RIKEN cDNA 1110055E19 gene 1110055E19Fk 0.2 0,4 WJ Down W29506 Mm.27822 apolipoprotein Bediting complex 2 Apobec2 0.2 0.2 WJ Down U15541 MT,3841. cytochrome c oxidase, subunit VIIIb Cox8to 0.2 0.2 WJDown W12396 Mm.18983 ESTs 0.2 0.2 WJDown AA120674 Mm.28407 acylphosphatase 2, muscle type Acyp2 0,1 0.2 WJDown D45913 Mm.4640 leucine rich repeat protein 1, neuronal Litt1 0.1 0,1 WJDown W18950 similar to meat shock protein 25 0.1 0,1 WJDown AA688923 Mm,7124 RIKEN cDNA 2400004E04 gene 2400004E04Rik 0.1 0,3 WDown x63815 Mn4857 calciumical modulin-dependent protein kinase I, beta Camk2b 0.1 0.2 WJDown AA162315 Mm.35531 myosin, heavy polypeptide 4, skeletal muscle 0.1 0.3 WJDown AA008531 similar to creatine kinase mRNA 0.1 0.2 WJDown m17.192 Mm.147 homeoboxA7 Hoxa7 0. 0.3 WJ down AA106126 Mn.34425 myosin, heavy polypeptide 2, skeletal muscle, adult Myh2 0.1 0.3 WJDown K02241 Mouse alkali myosin light chains 0. 03 WJDown W53582 similar to o-actinin2, skeletal muscle Patent Application Publication Dec. 18, 2008 Sheet 35 of 37 US 2008/0311185 A1

Supplementary Table 2, contd.

0,1 0,1 WJDown u05342 zinc finger protein 91 Zip91 0,1 0,4 WJDown W34.429 Mn., 195067 actinin alpha2 Actin2 0,1 0,0 WDown X61397 Mr.156583 carbonic anhydrase-like sequence 1 Cas1 0,1 0.4 WJDown M916O2 Mm,529 myosin light chain, phosphorylatable, cardiac ventricles Mypc 0,1 O,3 WJDown M916O2 Mm.529 myosin light chain, phosphorylatable, cardiac ventricles Mypc 0,1 0.2 WJDown AA210313 Mm.5290 popeye 3 Pop3-pending 0,1 0,4 WJDown WO8774 Mm.17235 cysteine-rich protein 3 Csrp3 0,1 0.2 WJ Down W70714 M.40323 growth differentiation factor 10 Gdf10 0,1 0.3 WDown abO.04.048 M.40956 neuroratin Nnat 0, 0.3 WJDown AAO38095 similar to mitochondrial creatine kinase Patent Application Publication Dec. 18, 2008 Sheet 36 of 37 US 2008/0311185 A1

Supplementary Table 3

Gene Ontology (GO) terms frequency of differentially expressed genes in Arthritic synovial fibroblasts (SF) and whole joints (WJ), selected from both subtractive libraries and microarrays.

Arthritic synovial fibroblasts (SF) List Chip GO term freq. freq. p value 30574;collagen catabolism 2 9 9,98E-08 42044;fluid transport 1 8 3.37E-05 6833; water transport 1 8 3.37E-05 6957;complement activation, alternative pathway 1 14 000010917 7015;actin filament organization l 15 0,000 12588 6958;complement activation, classical pathway 1 21 000025.077 6950;response to stress 3 308 0.00032571 9056;catabolism 4. 578 0,0003319 6955;immune response 3 314 000035055 7266;Rho protein signal transduction 26 000038683 30036;actin cytoskeleton organization and biogenesis 39 000087446 6952;defense response 3 411 0.000968 6508;proteolysis and peptidolysis 3 412 000097683 9057;macromolecule catabolism 3 420 0001.04953 3.0029;actin filament-based process 1 43 0,00106284 6956;complement activation 1 44 000111271 16064;humoral defense mechanism 1 45 0.0011637 7018;microtubule-based movement 1 46 0,0012 158 9607;response to biotic stimulus 3 455 000141324 45087;innate immune response l 66 0.00248681 7010;cytoskeleton organization and biogenesis 2 301 0.00437733 9605;response to external stimulus 3 624 000447236 6414;translational elongation 1 92 0,0047709 6006;glucose metabolism l 100 0.00561211 6996;organelle organization and biogenesis 2 365 0,0.0748304 19318:hexose metabolism l 122 000824837 7017;microtubule-based process 1 135 001002212 Arthritic whole joints (WJ) GO term freq.List freq.Chip p value 9617;response to bacteria l 1 O 7011;regulation of cytoskeleton l 1. O 6952, defense response 9 41 2,21E-07 6955;immune response 8 314 2.49E-07 9607;response to biotic stimulus 9 455 5,64E-07 6692;prostanoid metabolism 2 11 3,89E-06 6693;prostaglandin metabolism 2 11 3,89E-06 9056;catabolism 9 578 4,89E-06 9605; response to external stimulus 9 624 9,62E-06 Patent Application Publication Dec. 18, 2008 Sheet 37 of 37 US 2008/0311185 A1

Supplementary Table 3, contd.

6508;proteolysis and peptidolysis 412 2.24E-05 6690;eicosanoid metabolism 19 2.25E-05 9057;macromolecule catabolism 420 2.57E-05 16066;cellular defence response 5, 10E-05 6937;regulation of muscle contraction 30 9,20E-05 6936;muscle contraction 85 0000111 8220;necrosis 0,000 127 16271;tissue death 0,000 127 6950;response to stress 308 0,000264 6928;cell motility 197 0,000270 7160;cell-matrix adhesion 47 0,000355 6935;chemotaxis 48 0,000377 15669;gas transport 13 0,000652 15671;oxygen transport 13 0.000652 15674;di-, tri-valent inorganic cation transport 60 0,000729 6629;lipid metabolism 259 0,000940 45087;innate immune response 66 0,000963 6958;complement activation, classical pathway 21 0.001729 6511;ubiquitin-dependent protein catabolism 107 0003845 19941;protein-ligand dependent protein catabolism 1 10 0.004155 9100;glycoprotein metabolism 34 0004509 6118;electron transport 248 0.006004 6956;complement activation 44 0,007463 16064;humoral defence mechanism 45 0007796 7229;integrin-mediated signalling pathway 46 0008135 7155;cell adhesion 282 0009370 US 2008/031 1185 A1 Dec. 18, 2008

DEREGULATED GENES AND/OR tion, development and disease. However, the small number of PROCESSES IN INFLAMMATORY samples employed limits significantly statistical analysis and ARTHRTIS precludes the use of False Discovery Rates methodologies, as well as of multivariate techniques, which would be appropri ate for polygenic diseases such as RA. As it is, results of 0001. The invention relates to the field of molecular diag expression profiling need to be confirmed by an independent nostics and medicine. The invention in particular relates to the biological method, which largely diminishes the high deregulation of genes and/or processes in arthritis. The inven throughput nature and discovery rate of any given approach. tion utilizes this deregulation in a number of ways. 0006 To overcome these problems the invention in one 0002 The invention is exemplified by means of Rheuma aspect provides a novel, twin high throughput approach, com toid arthritis (RA), however, the invention is not limited to this posed of subtractive cDNA libraries and oligonucleotide disease but is applicable to all inflammatory arthritis diseases. DNA microarray hybridizations. Both methodologies, due to RA is a chronic destructive arthropathy that affects 1-3% of experimental setup, are governed by different constraints: the general population. RA imposes Substantial personal, random chance for the libraries, robotic clone selection and Social and economic costs. It is characterized by prolonged chip design for the microarrays. Consequently, both method inflammation of the joints, eventually leading to destruction ologies employ different, independent statistical analyses, are of the cartilage and bone. Inflammation is initially localized completely uncorrelated towards the propagation of error and in the synovial lining, a monolayer of synovial cells that lines conceptually have entirely different False Discovery Rates. diarthoidal joints. In RA, the synovial lining becomes mark As a result of this, the intersection of their statistically sig edly thickened due to synovial cell proliferation and infiltra nificant deregulated gene lists has a very low false discovery tion by inflammatory cells. This proliferative mass, the pan rate. To illustrate this and to exhibit the validity of the twin nus, invades and destroys articular cartilage and bone, leading high throughput approach, a number of representative genes to irreversible destruction of joint structure and function 1. from the commonly selected list was confirmed with a num Current therapies of RA rely mainly on symptomatic treat ber of different methods, such as automated literature search, ment with non-steroidal anti-inflammatory drugs and/or with semi-quantitative and quantitative RT-PCR in both mouse disease-modifying anti-rheumatic drugs. However, even the and human samples. Thus is illustrated that coupling of two best available treatments (like targeting TNF and TNF signal different high throughput approaches renders independent ing) do not cure the disease and do not even Sufficiently retard confirmation unnecessary, although it can of course still be progression in the majority of the patients, while they often done in addition. Consequently the number of likely deregu exhibit adverse side effects 2. lated genes is increased. In one embodiment the invention 0003. Despite intense efforts, the etiology and pathogen thus provides a method for identifying genes involved in a esis of RA remain poorly understood. Traditional paradigms disorder, comprising performing separately a subtractive for RA have implicated a variety of mechanisms that contrib cDNA library methodology and a oligonucleotide DNA ute to the initiation and perpetuation of synovial inflamma microarray hybridization methodology and optionally tion, including autoantibodies and immune complexes, T-cell another high throughput expression profiling methodology mediated antigen-specific responses, persistence of cytokine on a sample of an individual and then taking an intersection of networks and other pro-inflammatory molecules, genetic bias results of said methodologies. and sex predisposition, as well as proliferative behaviour of 0007 To validate the generated deregulated gene list even the arthritic synovium 3. Animal models of RA share many further, as well as to bridge two different animal models of clinical features with the human disease and hence constitute RA, expression profiling in the invention was combined with valuable tools in deciphering the pathogenetic mechanisms genetic linkage analysis. All differentially expressed genes that govern disease activation and perpetuation 4. Among were mapped to the chromosomes, together with the known them, the TNF-Tg mouse 5 has been instrumental in dem QTLS for an induced animal model of Arthritis, Collagen onstrating the role of TNF in the development of the disease Induced Arthritis (CIA) 14. Eight deregulated genes map to and foreshadowed the introduction and success of anti-TNF CIA QTLS (FIG. 1; WJ: Ctss, Pitpnm, Ncf1, Psmb8, Siat8e. therapies that transformed the effective management of the SF: Rab14, Aqp1, and GSn), leading to the conclusion that disease 6. In this model, chronic over expression of human these genes have a dominant influence in arthritic processes, TNF results in a chronic, erosive, symmetric polyarthritis, irrespective of the inciting stimulus (autoimmune or inflam with 100% phenotypic penetrance, timed disease onset and matory). Interestingly, several of the genes cluster together in progressive histological symptoms that resemble closely adjacent regions of the chromosomes (4, 7, 10, 11 and 13). human RA 5-7. These loci could define new QTLS or imply common regula 0004. The completion of the Human Genome Project and tory control at the chromatin level. the emergence of functional genomics and high-throughput 0008 Deregulated genes express their one or more gene technologies offer new possibilities for research into the eti products (or their enzymatic activities) either at a lower or at ology of RA. Amongst them, expression profiling creates a a higher level when compared to “normal cells’ obtained molecular fingerprint of the disease status, by the simulta from a healthy individual. This altered expression can be used neous quantification of the expression levels of thousand of to scrutinize samples and determine whethera sample is from genes. Similarly, reverse genetics (the genetic modification of an individual suffering from inflammatory arthritis (IA) or is a particular gene in a search of its function) allow for the at risk of suffering from IA. The invention thus provides a creation of animal models of disease, valuable tools for the method for determining whether a sample is derived from an dissection of pathogenetic mechanisms in vivo. individual suffering from inflammation arthritis or from an 0005 Expression profiling, the relative quantification of individual at risk of Suffering from IA, said method compris the expression levels of thousand of genes simultaneously, is ing measuring in said sample the level of expression or enzy an approach for understanding mechanisms of differentia matic activity of at least one gene product of a gene of table A, US 2008/031 1185 A1 Dec. 18, 2008

table B, or table 4 and comparing said expression level with a direction in which expression is affected in the arthritic reference value. The mentioned tables indicate whether in IA, samples is indicated in tables A, B and C(SFUP, SF down, WJ expression or enzymatic activity of said gene product is up, WJ down, where SF means that expression as indicated higher than “normal' or lower than “normal'. Thus by mea for synovial fibroblast cells and WJ means that expression is Suring the expression level or enzymatic activity of at least indicated for whole joint cells. one gene product, and comparing it with a reference value, 0010 Genes can have several different gene products. one can identify the sample as being derived from an indi Typically, though not necessarily, a gene is transcribed in vidual suffering from IA, or at risk of suffering from IA, or mRNA, which in turnistranslated into protein. A gene may be derived from an individual that is not suffering from IA or is transcribed in several mRNAs which in turn can be translated not at risk of suffering from IA. It is possible to identify a into several proteins. As used herein, the term "gene product’ sample in this way by analyzing expression of gene products encompasses both RNA and protein products. In a preferred or their enzymatic activity of one of said genes. In this case, a embodiment a gene product in a method of the invention sample identifies as being derived from an individual suffer comprises RNA. Expression levels of RNA are typically ing from IA, or at risk of suffering from IA when the level of expression or enzymatic activity is altered in the indicative determined with a probe that specifically hybridises with said direction given in table A, B, 3, 4 and/or C by at least a factor RNA (Northern blot) and/or with a set of primers that spe of 1.5. Thus a gene is said to be deregulated in the context of cifically anneal to said RNA (Real Time RT-PCR). Expres the invention when the expression level or enzymatic activity sion levels of proteins are typically determined either directly of a gene product of said gene is altered in the indicative with antibodies that specifically bind to the protein encoded direction as given in tables A, B, 3, 4 and/or C by a factor of by the gene or indirectly by the enzymatic/catalytically activ at least 1.5. When expression levels or enzymatic activity of ity of the protein of interest. Many different kinds of probes gene products of more than one gene of table A, table B, table are presently known in the art. Probes are often nucleic acids, 3 or table 4 are determined, the accuracy of the diagnostic however, alternatives having the same binding specificity in method increases allowing comparison of patterns of expres kind, not necessarily in amount are available to the person sion levels in the sample with patterns of a reference. In these skilled in the art, such alternatives include but are not limited cases, samples are identified as being derived from an indi to peptide nucleic acids (PNA). Similarly, many different vidual suffering from IA, or at risk of suffering from IA when specific binding bodies are available. Of old, antibodies are the pattern of expression levels or enzymatic activity matches used. However, currently many different parts, derivatives the pattern of a reference derived from an individual suffering and/or analogues of antibodies are in use. Non-limiting from IA, or at risk of suffering from IA. Or alternatively, examples of Such parts, derivatives and/or analogues are, identifies as normal, when the pattern matches the pattern of single chain Fv-fragments, monobodies, VHH, Fab-frag a reference. A reference as used in the invention is preferably ments and the like. A common denominator of Such specific from an arthritic sample, however, a normal sample can also binding bodies is the presence of an affinity region (a binding be used. Preferably the reference is derived from at least an peptide) that is present on a structural body that provides the arthritic sample and one from a normal sample. Statistical correct structure for presenting the binding peptide. Binding methods to analyze patterns of expression levels are available peptides are typically derived from or similar to CDR in the art. A non-limiting statistical method for determining sequences (typically CDR3 sequences) of antibodies, similarity of expression profiles, that can be used advanta whereas the structure providing body is typically derived geously in the present invention is referred to as the Rosetta from or similar to framework regions of antibodies. For Error Model (Hughes T. R. Marton M J. Jones A R et al. review of suitable probes and specific binding bodies see any Functional discovery via a compendium of expression pro of the following references: files. Cell. 2000; 102:109-126). A practical case is described 0011 Diagnosing early-onset rheumatoid arthritis: the on page 6 line 3-page 7 line 19 of WO 02/103320. role of anti-CCP antibodies, Vasishta A, Am Clin Lab. 2002 0009. It is of course possible to enlarge the set of genes August-September; 21 (7):34-6. Keys to unlocking the mys from which the expression level or enzymatic activity of gene teries of rheumatic autoimmune disease. Moser KL, Gaffney products is determined. In a preferred embodiment, expres P. Peterson E. Behrens T. Minn Med. 2004 May: 87(5):46-51: sion levels of gene products of at least 2 and more preferably A new tool for rheumatology: large-scale analysis of gene at least 5 genes of table A, table B, or table 4 are analyzed. expression Lequerre T. Coulouarn C, Derambure C. Lefebvre Analyzing said 5 genes provides a good correlation with the G. Vittecoq O. Daveau M. Salier J. P. Le Loet X. Joint Bone absence of presence of IA. Increasing the number of analyzed Spine, 2003 August: 70(4):248-56. genes of table A or table B, further increases the correlation. 0012 Expression levels of gene products can be altered as Thus in a particularly preferred embodiment gene products or result of a variety of mechanisms. These include but are not enzymatic activity of at least 10 genes of table A, B, or 4 are limited to, altered level of transcription/translation, altered analyzed. In a preferred embodiment at least one gene prod stability, altered degradation of the protein and/or RNA. The uctorenzymatic activity is of agene of table A and at least one altered expression level can also be the result of a combina other gene product or enzymatic activity is of a gene of table tion of processes in a cell. The sample to be analyzed is B or C. In a particularly preferred embodiment said one or typically derived from an individual suspected of suffering more, said at least 2, said at least 5 or said at least 10 genes are from IA or suspected of being at risk of suffering from IA. The genes of table A, optionally supplemented by gene products sample can be either synovial membrane or synovial fluid or enzymatic activity of one or more genes of table B or table isolated during an arthroscopy, and/or serum isolated from C. As mentioned, the set of genes analyzed may be expanded the blood of said individual. to other genes, preferably genes known to be deregulated in 0013 An individual is typically a human although the IA. Thus in a preferred embodiment said analysis includes the expression level or enzymatic activity of the genes of interest analysis of gene products of at least one gene of table C. The could be determined in animal models to determine drug US 2008/031 1185 A1 Dec. 18, 2008

efficacy. Therefore an individual is preferably a mammal, can either be a stage of IA that said disease has reached in an preferably a rodent or a primate. More preferably a human. individual and/or it can be a measurement of severity of IA in 0014) A particular class of deregulated genes in the an individual. present invention is involved with cytoskeleton arrangement. 0017 Deregulation of genes is a process that occurs in The invention therefore in one aspect provides a method for many different diseases. Each of those diseases has a charac determining whether a sample is derived from an individual teristic gene or multiple genes that are deregulated in said suffering from inflammatory arthritis, preferably rheumatoid diseases. IA is such a disease. IA is a generic term, which arthritis, or from an individual at risk of suffering thereof, said includes specific types of arthritis. Specific types of arthritis method comprising measuring in said sample the level of are typically associated with different characteristic genes expression of at least one gene product of a gene involved in that are deregulated. Deregulation of a gene will result in cytoskeleton arrangement and comparing said expression alteration of an expression level of a gene product of a gene level with a reference value. and thus in an altered amount of a gene productorin alteration 0015 Cytoskeleton arrangement in a cell comprises both of a composition of a gene product. In one embodiment the the structural processes involved with assembly and organi invention provides the use of an expression level of a gene sation of cytoskeleton and the signaling machinery in a cell product of a gene of table A, table B, or table 4 for typing that directs the structural processes. Assembly includes both inflammatory arthritis in an individual. In a preferred embodi gain and loss of cytoskeleton. Organisation typically includes ment expression levels of gene products of at least 2 and more the apparent three dimensional appearance of the cytoskel preferably at least 5 genes of table A, table B, or table 4 are eton in a cell. The signaling machinery involved in cytoskel analyzed. Analyzing said 5 genes gives a good picture of the eton organisation has an extracellular component and an involvement of characteristic genes of a type of IA. Increas intracellular component, thus in a preferred embodiment said ing the number of analyzed genes of table A, table B, or table gene is involved in upstream/downstream extracellular or 4, further increases the certainty about the type of IA con intracellular signaling to the cytoskeleton. A preferred cerned. Thus in a particularly preferred embodiment gene embodiment of the invention is particularly related to deregu products of at least 10 genes of table A, table B, or table 4 are lation of genes involved in assembly of actin cytoskeleton. analyzed. In another preferred embodiment a gene product of Thus in a preferred embodiment, said gene is involved in a gene of table A is analyzed. In a further preferred embodi binding and/or rearrangement of actin cytoskeleton. In a pre ment expression levels of gene products of at least 2 and more ferred embodiment said gene is involved in cytoskeleton preferably at least 5 genes of table A are analyzed. In a arrangement in a synovial fibroblast. In a particularly pre particularly preferred embodiment gene products of at least ferred embodiment said gene is any one of the following: 10 genes of table A are analyzed. A sample is typed as derived tubulin alpha 1, RAB14, Gelsolin, phosphatidylinositol from an individual or non-human animal Suffering from or at membrane-associated, matrix metalloproteinase 3, matrix risk of suffering from IA when the expression level of the metalloproteinase 13, tissue inhibitor of matrix metallopro tested genes is changed in the direction as indicated in tables teinase 13, cathepsin S. matrix gamma-carboxylglutamate, A, B or C(SFUP for genes that are over-expressed in synovial integrin alpha X, myosin light chain, troponin T1 skeletal fibroblast cells in IA and SF-down for genes that are under slow, troponin Iskeletal fast 2, VASP, Gesolin and/or table 4. expressed in said cells in IA, WJ UP for genes that are The individual is suffering from inflammatory arthritis when over-expressed in whole joint cells in IA and WJ down for there is at least a differences of at least 1.5 in the level of genes that are under-expressed in whole joint cells in IA). expression of said gene product of a gene involved in cytosk 0018 Inflammatory arthritis includes a number of specific eleton arrangement in said sample and said reference value. arthritis conditions such as but not limited to Rheumatoid Preferably, said method comprises measuring in said sample Arthritis, Systemic Lupus Erythematosus, Juvenile Rheuma the level of expression of a gene product of at least two of said toid Arthritis, Relapsing Polychondritis, Spondyloarthropa genes involved in cytoskeleton arrangement and comparing thie, Ankylosing Spondylitis, Reiter's Syndrome, Psoriatic said expression levels with reference values. In a particularly Arthritis, Enteropathic Arthritis, Crystal Deposition Disease, preferred embodiment the level of expression of a gene prod Gout Chondrocalcinosis, Calcium Hydroxyaptite Crystal uct of at least five of said genes involved in cytoskeleton Deposition Disease. In a preferred embodiment the invention arrangement is measured and compared with a reference is concerned with Rheumatoid Arthritis (RA). Thus when the value. invention refers to inflammatory arthritis in general, the pre 0016 While measuring at least one gene products expres ferred application is in rheumatoid arthritis. sion level or enzymatic activity and comparing it with a 0019 Typing IA according to the invention preferably reference value, according to a method of the invention, one comprises determining a pattern of expression levels of gene can determine a magnitude of a difference between said products (i.e. higher than normal SF UP, WJup or lower than expression level and said reference level. In one embodiment normal SF down, WJ down in tables A, B and C), as a pattern a method according to the invention is provided, wherein a of measured values is more specific than individual measured magnitude of a difference between said expression level and values. The invention thus provides a use according to the said reference value is normative for a degree in which said invention, wherein said typing comprises determining a pat inflammation arthritis is present. A magnitude of a difference tern of expression levels of gene products of at least two genes as used herein comprises at least a significant difference in a sample and wherein at least one of said genes is a gene of between the measured value and the reference value. A mag table A, table B, or table 4. In a preferred embodiment at least nitude of difference can for example be determined by mea one gene product is of a gene of table A and at least one other Suring fluorescence intensities. A difference between said gene product is of a gene of table B or C. In a particularly expression level and said reference value is normative in that preferred embodiment said one or more, said at least 2, said at it ascertains a degree in which inflammatory arthritis is least 5 or said at least 10 genes are genes of table A, optionally present. A degree in which inflammatory arthritis is present Supplemented by gene products of one or more genes of table US 2008/031 1185 A1 Dec. 18, 2008

B or table C. As mentioned, the set of genes analyzed may be were studied in order to gather valuable clues and to prioritize expanded to other genes, preferably genes known to be targets for therapeutic intervention of RA. In this context, deregulated in IA. Thus in a preferred embodiment said deregulated processes were determined as these are formal analysis includes the analysis of gene products of at least one ized through the controlled vocabulary of the Gene Ontology gene of table C. Consortium 17. GO term frequencies were calculated in the 0020. In one embodiment the invention provides a use selected gene list and their statistical significance was esti according to the invention, wherein said typing comprises mated. The resulting table (Supplementary Table 3) includes indicating one of the following types: Rheumatoid Arthritis, a number of functions that encompass accumulated knowl Systemic Lupus Erythematosus, Juvenile Rheumatoid edge about the pathogenesis of the disease. Such as collagen Arthritis, Relapsing Polychondritis, Spondyloarthropathie, catabolism, complement activation, immune and stress Ankylosing Spondylitis, Reiter's Syndrome, Psoriatic Arthri response. More importantly, 5 out of 26 predicted deregulated tis, Enteropathic Arthritis, Crystal Deposition Disease, Gout GO functions concerned (directly or indirectly) the actin Chondrocalcinosis, Calcium Hydroxyaptite Crystal Deposi cytoskeleton. In order to increase the experimental evidence tion Disease. In a particularly preferred embodiment of the F-actin stress fibers were visualized in vitro on arthritic as invention said typing comprises indicating an inflammatory well as wt SFs, both primary and immortalized. As evident in arthritide. Said types of arthritis are all included in the generic FIG.3a, there was a striking difference between arthritic and term Inflammatory Arthritis as used in the invention. normal SF's, with the arthritic SFs exhibiting pronounced 0021. In one embodiment of the invention genes and/or the stress fibers. Stress fibers are known to appear in differenti expression level of gene products encoded thereby are ated fibroblasts. These differentiated fibroblasts are called detected. The expression level of their products or their enzy myofibroblasts. These myofibroblasts are specialized con matic activity are determined with a probe or a set of probes tractile fibroblasts, with an important role in establishing or appropriate enzymatic assays. The invention thus provides tension during wound healing and pathological contracture a kit of parts comprising a probe or a set of probes or enzy 21.22. The invention hereby provides the first direct indica matic assays, wherein at least one of said probes is a probe tion about the presence of myofibroblasts in the arthritic specific for one of the genes listed in table A, table B, or table synovium. Consistent with this notion, arthritic SFs were 4. Preferably a set of probes comprises two or more probes observed to have more pronounced FAK positive islands that are specific for one of the genes listed in table A, table B. (data not shown). FAK positive islands are a prominent fea or table 4. In another embodiment the invention provides a kit ture of myofibroblasts 22. The actin cytoskeleton interacts of parts according to the invention, comprising a probe or a set bi-directionally with the ECM through receptors (mainly of probes, wherein at least one of said probes is a probe integrins) that possess extracellular binding sites for laminin, specific for one of the genes listed in table A. The invention collagen, fibronectin and other ECM components. The for further provides a kit of parts according to the invention, mation of intimate, extensive adhesive contacts between cells comprising at least two probes, wherein at least one of said and ECM results from cooperation between adhesive systems probes is a probe specific for one of the genes listed in table A, and the actin cytoskeleton and the generation of force across table B, or table 4 and at least one of said probes is specific for regions of the cell 24. In this context, myofibroblasts are one of the genes listed in table C. Also provided by the thought to exert increased tension to the Substratum, through invention is a kit of parts according to the invention, compris their increased adhesive capacity, which results to a more ing at least two probes, wherein at least one of said probes is flattened, elongated cell shape 22. Accordingly, we have a probe specific for one of the genes listed in table A and at observed that arthritic SFs adhere to various ECM compo least one of said probes is specific for one of the genes listed nents with increased affinity in vitro (FIG. 3b), resulting in a in table B or in table C. Any of the probe sets according to the more elongated shape in vivo (FIG.3c), further corroborating invention can comprise probes specific for a gene that is not the existence of myofibroblasts in the arthritic synovium. listed in table A, B or C. Such a probe is preferably specific for 0024. The invention thus surprisingly found that deregu a gene that is known to be involved in inflammatory arthritis, lation of genes involved in cytoskeleton arrangement in a cell more preferably known to be involved in Rheumatoid Arthri is a prominent feature in arthritic SF's. A new therapeutic and tis. In another preferred embodiment such a probe is specific diagnostic entrance is hereby established. In one embodiment for a gene known to be involved in cytoskeleton arrangement. the invention provides the use of a compound for altering 0022. The present invention has surprisingly found that in cytoskeleton or its organization in a cell for reducing stress IA particular genes involved with cytoskeleton arrangement fibers in a synovial cell. are deregulated. The invention thereby presents a novel 0025 Recent experiments have shown that the cytoskel means of therapy. In one embodiment the invention hereto eton plays a critical role in tumorigenesis 31. Since the role provides the use of a compound for altering cytoskeleton or of the cytoskeleton in cancer has been established, com its organization in a cell for the manufacture of a medicament pounds that are directed to the cytoskeleton have been devel for alleviating at least one symptom of inflammatory arthritis oped for the treatment of cancer. In one embodiment the in an individual. An individual with IA can present either one invention provides the use of a drug for the treatment of symptom or multiple symptoms. Alleviating at least one cancer for altering cytoskeleton or its organizationina cell for symptom will make the condition of an individual with IA the manufacture of a medicament for alleviating at least one better, thereby improving a patient’s quality of life. When symptom of inflammatory arthritis in an individual. In reference is made to alleviating a symptom of a disease, it is another embodiment the invention provides the use of a drug meant that the severity of a symptom is at least reduced. for the treatment of cancer for altering cytoskeleton or its 0023 Rheumatoid arthritis is a polygenic disease. Study organization in a cell for reducing stress fibers in a synvovial ing multiple genes therefore provides more valuable informa cell or for the treatment of inflammatory arthritis. Non-lim tion about RA than concentrating on one gene. Thus, in the iting examples of drugs that are directed to the cytoskeleton invention concerted gene functions and cellular processes are: Cytochalasin D. Phalloidin, Jasplakinolide, Chondra US 2008/031 1185 A1 Dec. 18, 2008

mides, Dolostatin 11, Latrunculin A, Swinholide A, Misaki glutamate, integrin alphaX, myosin light chain, troponin T1 nolide A, Mycalolide B, Shinxolide C. Scytophycins, skeletal slow, troponin I skeletal fast 2, VASP, ADF/Cofilin Goniodomin A, Pseudopterolide, Lysophosphatidylcholine, and/or is gene of table 4. Preferably said gene is Gelsolin, Lysophosphatidyl acid. In one embodiment the invention pro VASP and or ADF/Cofilin. These genes are further discussed vides the use of a compound according to the invention, in the examples, tables and figures. wherein said compound is one of the following: Cytochalasin 0028. An alternative means to create an animal wherein a D. Phalloidin, Jasplakinolide, Chondramides, Dolostatin 11, gene is over-expressed or under-expressed is the introduction Latrunculin A, Swinholide A, Misakinolide A, Mycalolide B, of a foreign gene into the genome of said animal. Such an Shinxolide C, Scytophycins, Goniodomin A, animal is a transgenic animal. The invention in one embodi Pseudopterolide, Lysophosphatidylcholine, Lysophosphati ment provides a non-human animal according to the inven dyl acid. tion, wherein said animal is transgenic for said gene or homo 0026. Many different animal models are used in studies of logue thereof that is over-expressed or under-expressed. A RA. An animal model typically represents a patient with RA. foreign gene can be constructed using recombinant DNA In order to make an animal represent an individual Suffering methodology. A preferred non-human animal as provided by from RA different strategic routes can be applied. Disease can the invention is rodent or a non-human primate. Preferably, for example be induced by injection of an immunogenic said non-human animal is a mouse. Characteristics of these Substance. Alternatively an animal can be bred that has an animals are well known and these animals are practical for altered gene expression. A gene in Such an animal can either laboratory circumstances. Mice and rats are often used in be knocked-out, over-expressed or under-expressed. The disease models for RA. In a preferred embodiment the inven invention provides an animal with an altered gene expression. tion provides the use of a non-human animal according to the In one embodiment the invention provides a non-human ani invention as a disease model. Preferably said disease repre mal wherein at least one of the genes listed in table A, table B. sented by said model is rheumatoid arthritis. Alternatively table 3 or table 4 is knocked-out, over-expressed or under said disease is a disease in which deregulated genes as deter expressed. In another embodiment the invention provides a mined by the invention are involved. Inflammatory and/or non-human animal wherein at least one of the genes listed in rheumatoid arthritis in an animal can either occur spontane table A is knocked-out, over-expressed or under-expressed. ously, by known or unknown internal causes, or can be The invention further provides a non-human animal wherein induced. In a preferred embodiment the invention provides at least one of the genes listed in table A, table B, table 3 or the use of a non-human animal as a disease model for rheu table 4 and wherein at least one of the genes listed in table C matoid arthritis, wherein said rheumatoid arthritis is induced. ortable 6 is knocked-out, over-expressed or under-expressed. Induction can proceed directly or indirectly. Direct induction Furthermore the invention provides a non-human animal for example occurs when an immunogenic composition that wherein at least one of the genes listed in table A and wherein causes arthritis is injected. Examples of Such immunogenic at least one of the genes listed in table B, table C, table 3, table compositions include but are not limited to collagen. Indirect 4 or table 6 is knocked-out, over-expressed or under-ex induction is for example due to a genetic defect, because of pressed. Naturally the invention also provides an animal which an animalata certain stage of its life develops arthritis. wherein, besides a gene listed in table A, B, 3, 4, 6 or C. Preferred examples of such induction are indicated in table 5. another gene is knocked-out, over-expressed or under-ex 0029. Further preferred is a non-human animal of the pressed. Said another gene preferably is a gene known to be invention wherein said at least one of the genes listed in table involved in RA. Another preferred gene in this context is a A, table B, table C, table 3, table 4 or table 6 is knocked-out, gene known to be involved in cytoskeleton rearrangement. over-expressed or under-expressed, in an arthritis model. The invention in one embodiment provides a non-human Preferably a genetic arthritis model, preferably a genetic animal wherein at least one gene known to be involved in model as indicated in table 5, and-or in a Tg 197 mouse. cytoskeleton rearrangement is knocked-out, over-expressed Provided is preferably a Tg 197 mouse wherein Gelsolin, or under-expressed. VASP and/or ADF/Cofilin is knocked-out, over-expressed or 0027. In RA one of the cell types that plays a very impor under-expressed. Preferably specifically in the myeloid lin tant role in the disease process is the synovial fibroblast. eage. Preferably, Gelsolin and/or ADF/Cofilin are knocked Therefore, in a preferred embodiment, the invention provides out or under-expressed. VASP is preferably over-expressed. a non-human animal according to the invention, wherein said 0030 The invention further provides the use of a non cytoskeleton arrangement is in a synovial fibroblast. Cytosk human animal of the invention as an inflammatory arthritis eleton organization is typically influenced by diverse signals. model. Preferably, said non-human animal is suffering from Therefore the invention in one embodiment provides a non inflammatory arthritis. This is typically achieved by provid human animal according to the invention, wherein said gene ing a non-human animal of the invention with a composition is involved in extracellular or intracellular signaling to the that induces said inflammatory arthritis. Such compositions cytoskeleton. In a preferred embodiment the invention pro are well known in the art and include but are not limited to vides a non-human animal according to the invention, collagen (are there other well known examples). wherein said gene is involved in binding and/or rearrange 0031. Inflammatory inflammation being a severe disease ment of the actin cytoskeleton. The invention provides genes for which no definite cure has been found, there is an ongoing that are involved in cytoskeleton arrangement. The invention need for effective medication for diverse types and stages of thus in a preferred embodiment provides a non-human animal RA. In the process of searching for effective medication a according to the invention, wherein said gene is any one of the reliable test for compounds comprised in this medication is following: tubulin alpha 1, RAB14, Gelsolin, phosphatidyli essential. The invention provides a method for determining nositol membrane-associated, matrix metalloproteinase 3. whether a compound is effective in alleviating a symptom of matrix metalloproteinase 13, tissue inhibitor of matrix met IA, comprising testing said compound in a model system for alloproteinase 13, cathepsin S, matrix gamma-carboxyl said symptom, said method characterized in that said com US 2008/031 1185 A1 Dec. 18, 2008 pound is a compound for altering the cytoskeleton or its herein alters the expression of said deregulated genes in said organization in a cell. As in the development of cancertherapy individual in such a way that the condition of said individual compounds have been found that are involved in cytoskeleton is improved. Alleviating a symptom of IA in a patient will arrangement, said compounds are used in a method of the usually imply up-regulating a down-regulated gene or down invention. Thus, in one embodiment the invention provides a regulating an up-regulated gene (see table A, B or CSF or WJ method for determining whether a compound is effective in down and SF or WJUP respectively), such that the expression alleviating a symptom of IA, comprising testing said com of said gene approaches or equals a reference value of a pound in a model system for said symptom, said method normal, i.e. healthy, state. In a preferred embodiment the characterized in that said compound is a compound for alter invention provides a method for alleviating at least one symp ing the cytoskeleton or its organization in a cell, wherein said tom of rheumatoid arthritis in an individual, comprising compound is effective in cancer therapy. Said compound modifying expression in said individual of a gene that is preferably is one of the following: Cytochalasin D. Phalloi involved in a cytoskeleton arrangement, wherein expression din, Jasplakinolide, Chondramides, Dolostatin 11, Latruncu of said gene is modified in a synovial fibroblast. In one lin A, Swinholide A, Misakinolide A, Mycalolide B, Shinx embodiment of the invention said gene is involved in extra olide C. Scytophycins, Goniodomin A, Pseudopterolide, cellular or intracellular signaling to the cytoskeleton. In Lysophosphatidylcholine, Lysophosphatidyl acid. another embodiment said gene is involved in binding and/or 0032. In the invention specific genes have been found to be rearrangement of the actin cytoskeleton. The invention dis deregulated in IA. In the search process for compounds with closes genes that are involved in cytoskeleton arrangement therapeutic value for treatment of IA, the expression of these and deregulated in IA. In one embodiment the invention thus genes is a useful parameter. For other diseases in which the provides a method according to the invention, wherein said deregulated genes that are found in the invention are deregu gene is any one of the following: tubulin alpha 1, RAB14, lated, said parameter can be used as well. The similar deregu Gelsolin, phosphatidylinositol membrane-associated, matrix lation for instance occurs because of a similar causal mecha metalloproteinase 3, matrix metalloproteinase 13, tissue nism that acts alike in these diseases. The invention provides inhibitor of matrix metalloproteinase 13, cathepsin S, matrix a method for determining whether a compound has an effect gamma-carboxylglutamate, integrin alpha X, myosin light on an expression of one or more of the genes listed in table A, chain, troponin T1 skeletal slow, troponin I skeletal fast 2, table B, or table 4, comprising testing said compound in an VASP, ADF/Cofilin and/or a gene of table 4. A method of the environment wherein said gene is expressed and comparing a invention is also applicable in another disorder than IA resulting expression level with a reference value. Preferably wherein genes involved in cytoskeleton arrangement, such as reference value indicating a higher than normal expression aforementioned genes, are deregulated. In a preferred level for the genes as listed in table A, B or C by SFUP and/or embodiment said gene in a method according to the invention WJup or a lower than normal expression level as indicated in codes for Gelsolin, VASP or ADF/Cofilin. Preferably Gelso table A, B or C by SF down, and/or WJ down. The invention lin. further provides a method for determining whether a com 0034 Gelsolin is an actin binding protein 33 that has pound has an effect on an expression of one or more of the been implicated, amongst others, in the transduction of Sig genes listed in table A, comprising testing said compound in nals into dynamic rearrangements of the cytoskeletal archi an environment wherein said gene is expressed and compar tecture. In the presence of calcium, gelsolin severs preexist ing a resulting expression level with a reference value. In one ing actin filaments and caps them, thereby preventing embodiment the invention provides a method for determining monomer addition to their fast-growing ends. The barbed end whether a compound has an effect on an expression of one or cap is highly stable, even in the absence of calcium, unless more of the genes listed in table A, table B, or table 4 and on displaced by interactions with regulatory phospholipids Such one or more of the genes listed in table C, comprising testing as phosphatidylinositol-4,5-bisphosphate (PIP2). In the pres said compound in an environment wherein said genes are ence of a large pool of profilin (another actin binding protein), expressed and comparing a resulting expression level with a or under depolymerizing conditions, these gelsolin-capped reference value. In another embodiment the invention pro ends allow the disassembly of populations of actin filaments vides a method for determining whether a compound has an by subunit loss from the pointed ends 34. GSn-/-fibroblasts effect on an expression of one or more of the genes listed in were found to have excessive stress fibers in vitro 18, very table A and on one or more of the genes listed in table B or in similar to the ones observed in arthritic SFs, where gelsolin table C, comprising testing said compound in an environment was found down-regulated with a variety of methods. Knock wherein said genes are expressed and comparing a resulting ing out gSn expression from the arthritic mice resulted in expression level with a reference value. Preferably an effect exacerbation of the disease (FIG. 4), therefore proving the on multiple genes that are listed in table A, B or C is deter participation of the actin cytoskeleton rearrangement in the mined. pathophysiology of the disease. Gelsolin was found to be one 0033 Genes that have been found to be deregulated in the of the most striking downregulated markers upon malignant invention are in one embodiment of the invention used in a transformation of fibroblasts by Ras 35, while overexpres disease model of IA. In such an embodiment said genes are sion of a gelsolin mutant was shown to Suppress Ras-induced deregulated on purpose. In contrast, in an individual with IA, transformation 36. Its expression was found undetectable or said genes already are deregulated. In this context, the pur reduced in a majority of human gastric, bladder, lung, colon pose is to regulate said genes in order to improve the condition and breast tumours 37-39. of said individual. In one embodiment the invention therefore 0035. The genes that are involved in cytoskeleton arrange provides a method for alleviating at least one symptom of ment are only one group of genes that, in the invention, are rheumatoid arthritis in an individual, comprising modifying found to be deregulated in IA. Other genes are, amongst expression in said individual of a gene that is involved in a others, listed in table A, table B, table 3, table 4 or C. The cytoskeleton arrangement. Modifying expression as used invention in one embodiment provides a method for alleviat US 2008/031 1185 A1 Dec. 18, 2008

ing at least one symptom of inflammatory and/or rheumatoid other than listed in table A, B or C. In a further embodiment arthritis in an individual, comprising modifying the level of said method comprises modifying the level of expression of at expression of at least one gene product of a gene of table A, least one gene product of a gene of table A or table B and at table B, table 3, table 4 or C in cells of said individual. In least one gene product of a gene of table C in cells of said another embodiment said method comprises modifying the individual and modifying the level of expression of at least level of expression of at least one gene product of a gene of one gene product of a gene known to be deregulated in IA table A in cells of said individual. In a further embodiment other than listed in table A, B or C. Further the invention said method comprises modifying the level of expression of at provides a method according to the invention, wherein said least one gene product of a gene of table A, table B, or table 4 method comprises modifying the level of expression of at and at least one gene product of a gene of table C in cells of least one gene product of a gene of table A and at least one said individual. Further the invention provides a method gene product of a gene of table B or table C in cells of said according to the invention, wherein said method comprises individual and at least one gene product of a gene known to be modifying the level of expression of at least one gene product deregulated in IA other than listed in table A, B or C. of a gene of table A and at least one gene product of a gene of 0037. In a preferred embodiment, the invention provides a table B or table C in cells of said individual. In preferred method for alleviating at least one symptom of inflammatory circumstances a method according to the invention alleviates and/or rheumatoid arthritis in an individual according to the two or more symptoms. Preferably the level of expression of invention, wherein modifying said expression level com two, three or more gene products of genes of table A, B, C, or prises lowering expression of a gene that is over-expressed in 4 are modified. The direction in which an expression level is said individual compared to a comparable healthy individual, modified is dependent on the direction in which an expression or elevating expression of a gene that is under-expressed in level of a gene product of a gene is deregulated (see table A, said individual compared to a comparable healthy individual B or C). The extent to which modification is applied depends (table A, B or C). on the degree in which said level of expression of said gene 0038 Raising expression of an under-expressed gene is product of said gene is deregulated. The expression level of a for instance effected by the addition of a transcription factor gene product of a gene is preferably modified to match or and/or by activating an enhancer-sequence. Lowering expres approach a reference value that represents a healthy status, sion of an over-expressed gene can for example be realized by thus a normal level. Alternatively, the expression level of a applying antisense therapy or by administering a repressor gene product of a gene is modified to match a physiologically protein that binds to the promoter of said aberrantly expressed acceptable level of an individual. Physiologically acceptable gene. In a preferred embodiment, the invention provides a levels will often be levels that are approximately normal method according to the invention, wherein said lowering levels. Normal levels are levels that can be found in a healthy said expression comprises utilizing an antisense-Sequence. individual of the same age and constitution. Physiologically There are many different anti-sense approaches to down acceptable levels can fall outside a normal range but do still regulate production of a gene product available to a person provide a sufficient function for an individual. skilled in the art. Antisense technology exploits oligonucle 0036. A gene product in a method of the invention encom otide analogs to bind to target RNAs via Watson-Crick passes both RNA and protein products. In one embodiment hybridization. After binding, the antisense agent either dis the invention provides a method for alleviating at least one ables or induces the degradation of the target RNA. Antisense symptom of inflammatory and/or rheumatoid arthritis in an agents can also alter splicing. The basic mechanisms of anti individual according to the invention, wherein said gene sense technology have been extensively explored in the last product comprises protein. A cell in which the level of expres ten years. Much has been learned about the medicinal chem sion of a gene product of a gene is modified in a method of the istry and the pharmacologic, pharmacokinetic, and toxico invention, is a cell that is affected by the disease process of logic properties of antisense molecules. Antisense technol inflammatory and/or rheumatoid arthritis in an individual. ogy has proven valuable in gene functionalization and target Preferably said cell is a cell that makes contact with a joint of validation. said individual. In a preferred embodiment of the invention, a 0039. Non-limiting examples of antisense approaches are method according to the invention is provided, wherein said interference RNA (RNAi), microRNA and splice interference cells are cells of a joint or precursors thereof. These cells can techniques such as exon-skipping. An antisense sequence can be part of any component of the joint. Precursor cells can be administered as a single stranded molecule or be part of a already be present in or in the Surroundings of said joint or hairpin construct-molecule. The anti-sense sequence can be they can for example be derived from the blood. In a preferred administered directly or be produced in a cell by means of an embodiment said cells comprise synovial membrane cells or expression cassette. Today RNA interference (RNAi), is a cells of synovial fluid. Preferably in a method according to the technique in which exogenous, double-stranded RNAs (dsR invention the level of expression of at least two gene products NAs) that are complimentary to known mRNAs, are intro is modified. In a preferred embodiment a method according to duced into a cell to specifically destroy that particular mRNA, the invention comprises modifying the level of expression of thereby diminishing or abolishing gene expression. The tech at least one gene product of a gene of table A, table B, or table nique has proven effective in Drosophila, Caenorhabditis 4 in cells of said individual and modifying the level of expres elegans, plants, and recently, in mammalian cell cultures. For sion of at least one gene product of a gene known to be research purposes in cultured mammalian cells, Small inter deregulated in IA other than listed in table A, B or C. Further fering RNAs (siRNAs) are delivered into the cell. SiRNAs are provided is a method according to the invention, wherein said dsRNAs of approximately 21-25 nucleotides. This is done method comprises modifying the level of expression of at with transfection reagents, solutions optimized for allowing least one gene product of a gene of table A in cells of said DNA and RNA to be absorbed by cultured cells. In a preferred individual and modifying the level of expression of at least embodiment the invention provides a method according to the one gene product of a gene known to be deregulated in IA invention, wherein lowering said expression comprises uti US 2008/031 1185 A1 Dec. 18, 2008

lizing RNAi. Preferably siRNA is used for the induction of gene products. An array can be specific for a whole genome, RNAi. In another embodiment shRNA is used. A classical but in the invention preferentially is specific for a part of a antisense sequence can be complementary to at least 10% and genome, i.e. a Subgenome. Another preference of the inven preferably at least 25 of the mRNA it is specific for. SiRNA, tion is an array specific for a transcriptome or a part of a miRNA, RNAi and exon-skipping antisense sequence have transcriptome, i.e. a Subtranscriptome. A transcriptome of an the normal overlap that is typically used in the field. They organism as used herein, are all DNA transcripts, mostly typically have between 15-40 bases overlap, preferably mRNAs, being produced at a certain point. A certain point is between 15-30. SiRNAs have typically 21-25 nucleotides. specified in its location and/or in time. Thus in a preferred The antisense can be provided as single-stranded or double embodiment of the invention, an array is a Subgenome array Stranded molecules, depending on their effectivity as is com or a Subtranscriptome array. In a preferred embodiment the monly known in the art. The antisense sequence can be com invention provides an array comprising at least two probes, prised in a nucleic acid or a functional equivalent thereof. wherein at least one probe is specific for at least one gene 0040. An anti-sense sequence can for example be used in listed in table A, table B, or table 4. The invention further a compound that is applied to modify an expression level of a provides an array according to the invention, wherein at least gene product of a gene. In a preferred embodiment the inven one probe is specific for at least one gene listed in table A. tion provides the use of a compound for the manufacture of a Also provided by the invention is an array according to the medicament for alleviating at least one symptom of inflam invention, wherein at least one probe is specific for at least one matory and/or rheumatoid arthritis in an individual, said com gene listed in table Aortable Band wherein at least one probe pound comprising an antisense-sequence of one of the genes is of at least one gene listed in table C. In another embodiment listed in table A or table B. In another embodiment the inven the invention provides an array according to the invention, tion provides a use according to the invention, wherein said wherein at least one probe is specific for at least one gene compound comprises an antisense-sequence of one of the listed in table A, and wherein at least one probe is specific for genes listed in table A. Further provided is a use according to at least one gene listed in table B or in table C. In a preferred the invention, wherein said compound comprises an anti embodiment an array of the invention is a microarray. Said sense-sequence of one of the genes listed in table A or table B probes specific for said genes are thus preferably comprised and an antisense-sequence of one of the genes listed in table in a microarray. Said microarray preferably comprises oligo C. In another use according to the invention, said compound nucleotides or a functional equivalent thereof. Probes or oli comprises an antisense-Sequence of one of the genes listed in gonucleotides are preferably between 15-80 nucleotides or table A and an antisense-sequence of one of the genes listed in functional equivalents thereoflong. More preferably between table Borintable C. Preferably antisense-sequences of two or 15-40. more genes listed in table A, B or C are comprised in a use of 0043. An array as provided by the invention is in one the invention. In addition to antisense-Sequences of afore embodiment used in a method of the invention. Provided is mentioned genes, antisense-sequences of other genes known the use of an array according to the invention in a method to be involved in IA can be comprised in said compound. In according to the invention for determining whether a sample preferred embodiment an antisense is provided specific for is derived from an individual suffering from inflammatory mRNA from a gene that is identified in table A, B or C by SF and/or rheumatoid arthritis or from an individual at risk of UP or WJ up. Suffering there from, said method comprising measuring in 0041. Before a suitable drug can be developed, a drug said sample the level of expression of at least one gene prod target has to be identified and validated. Drug targets are uct of a gene of table A or table Band comparing said expres mainly proteins, DNA or RNA. The invention provides such sion level with a reference value. Further provided is the use drug targets as it has identified genes and gene products that of an array according to the invention for determining are deregulated in IA. Thus in one embodiment the invention whether an individual is suffering from IA or is at risk of provides the use of a protein encoded by a gene of table A, suffering from IA. table B, or table 4 as a drug target for selecting candidate 0044 Protein expression levels can be measured in vari drugs for a treatment of inflammatory arthritis. Further pro ous ways such as ELISA, Western blotting or another sort of vided is a use according to the invention, wherein said protein antibody based calorimetric or fluorescent quantitation of is encoded by a gene of table A. In one embodiment a first protein. In a preferred embodiment the invention provides an protein encoded by a gene of table Aortable B is used as a first array comprising at least two binding bodies, wherein at least drug target and a second protein encoded by a gene of table C one binding body is specific for at least one gene product of a is used as a second drug target. In another embodiment said gene of table A, table B, or table 4. In on embodiment said first protein encoded by a gene of table A is used as a first drug binding body is an antibody. Said gene product is RNA or a target and a second protein encoded by a gene of table B or protein. In one embodiment the invention provides an array table C is used as a second drug target. By a use of the according to the invention, wherein said gene product is a invention said drug targets are thus validated. After validation protein. In a preferred embodiment an array according to the the drugs are tested further, first in a pre-clinical stage and invention comprises at least one binding body that is specific then in a clinical setting. for at least one gene product of a gene of table A. Further 0042. As already mentioned, the invention provides provided is an array according to the invention, wherein at probes specific for deregulated genes in IA. Probes thus pro least one binding body is specific for at least one gene product vided, are also applicable for disorders other than IA in which of a gene of table A. In one embodiment the invention pro said genes are deregulated. In order to make the addressing of vides an array according to the invention, wherein at least one a gene more specific, more then one probe is preferentially binding body is specific for at least one gene product of a gene used. More then one probe, as used in the invention, is pref of table A, table B, or table 4, and wherein at least one gene erably comprised in an array. An array is used to analyze one product is of a gene of table C. In another embodiment the or more samples at the same time for one or more genes or invention provides an array according to the invention, US 2008/031 1185 A1 Dec. 18, 2008 wherein at least one binding body is specific for at least one particularly Suited for high throughput Screening, particularly gene product of a gene of table A, and wherein at least one for screening candidate drugs and/or nucleic acids of the gene product is of a gene of table B or table C. Said array invention for potential therapeutic effect in IA. In a preferred according to the invention preferably is a microarray. Said embodiment said nucleic acid codes for a protein produced by array is in one embodiment used in a method according to the a gene of table A, B, C, or 4, wherein said gene is identified in invention. Provided is the use of an array according to the table A, B or C by SF down and/or WJ down or said gene is invention in a method according to the invention for deter gelsolin or ADF/Cofilin. An antisense sequence is preferably mining whether a sample is derived from an individual Suf specific for a gene that is identified in table A, B or C by SF or fering from inflammatory and/or rheumatoid arthritis or from WJUP, or VASP. Said (poly)peptide is preferably an antibody an individual at risk of suffering there from, said method or functional equivalent thereof such as but not limited to a comprising measuring in said sample the level of expression single chain antibody, a monobody, a FAB fragment or the of at least one gene product of a gene of table A or table Band like. A peptide preferably comprises at least 6, and preferably comparing said expression level with a reference value. Fur at least 8 amino acids in peptide linkage. A peptide is prefer ther provided is the use of an array according to the invention ably not longer than 40 amino acids. A polypeptide comprises for determining whether an individual is suffering from IA or 41 or more amino acids in peptidic linkage. A (poly)peptide is at risk of suffering from IA. may of the invention may comprise modifications such as 0045. Where in the description reference is made to table chemical modifications and/or post translational modifica Aortable B this reference also includes the reference totables tions such as in a protein. A (poly)peptide may be linear or 3 and 4. In these cases a preference is indicated for tables A, circular. B or table 3. Preferred is Table A or table B, Preferred is at 0048. The present invention further provides the use of a least table A. gene delivery vehicle comprising the coding region of a gene 0046. The invention further provides a method for deter of table A, B or C identified by SF and/or WJ down, Gelsolin mining whethera compound, a (poly)peptide or a nucleic acid or ADF/Cofilin, for delivering said gene to cells of a joint. molecule or functional equivalent thereof affects a phenotype Preferably said gene is delivered to a synovial fibroblast cell. of synovial fibroblast cells of an individual or non-human In another aspect the invention provides the use of a gene animal Suffering from inflammatory arthritis or at risk of delivery vehicle comprising the coding region of a gene of Suffering thereof, comprising culturing synovial fibroblast tables A, B or C identified by SF or WJ down, Gelsolin or cells in vitro from a sample obtained from an affected joint or Cofilin, for the preparation of a composition for delivering a joint at risk of being affected from said individual or non said gene to cells of a joint (WJ down) and/or a synovial human animal, in the presence of said compound and/or fibroblast cell (SF down). Preferably said gene delivery providing said synovial fibroblast cells with said nucleic acid vehicle is used for the treatment of inflammatory arthritis. In or said functional equivalent thereof and determining whether a particularly preferred embodiment said gene is Gelsolin, a phenotype of said cultured synovial fibroblast cells is ADF/Cofilin. The invention thus further provides a gene altered. Preferably, wherein said synovial fibroblast cells are delivery vehicle comprising a coding region for Gelsolin obtained from a joint of a mouse comprising the genetic and/or ADF/Cofilin. Further provided is the use of a gene alteration of the Tg 197 mouse, a non-human animal accord delivery vehicle comprising a nucleic acid of a functional ing to the invention and/or a genetic model as depicted in table equivalent thereof comprising an antisense-Sequence for a 5. In a preferred embodiment said compound, nucleic acid or gene of table A, B or C identified by SF and/or WJ UP or functional equivalent thereof is a compound for altering VASP or comprising an expression cassette therefore, for cytoskeleton or its organization in a cell. Preferably said delivering said nucleic acid or functional equivalent thereof to compound is effective in cancer therapy, as described herein cells of a joint, or to a synovial fibroblast cell. Preferably above. Preferably said compound is one of the following: wherein said antisense-Sequence comprises an shRNA, RNAI and/or siRNA. Preferably wherein said gene delivery 0047 Cytochalasin D. Phalloidin, Jasplakinolide, Chon vehicle comprises a liposome, or a viral vector. Preferably an dramides, Dolostatin 11, Latrunculin A, Swinholide A, Mis adenovirus, an adeno-associated virus vector or a lentivirus akinolide A, Mycalolide B, Shinxolide C. Scytophycins, vector. The gene delivery vector may be delivered/adminis Goniodomin A, Pseudopterolide, Lysophosphatidylcholine, tered systemically. In a particularly preferred embodiment Lysophosphatidyl acid. In a preferred embodiment said said gene delivery vehicle is administered locally, preferably nucleic acid comprises an antisense-Sequence, preferably an intrarticular. Said expression cassette preferably comprises a RNAi, an siRNA, oran shRNA. Said antisense sequence can promoter to drive expression of the coding region and/or be provided in a nucleic acid or a functional equivalent antisense sequence. Preferably said promoter is a pol II pro thereof. Said nucleic acid or functional equivalent thereof preferably comprises a morpholino, locked nucleic acid moter or a pol III promoter. (LNA) or a peptide nucleic acid (PNA). In a particularly BRIEF DESCRIPTION OF THE DRAWINGS preferred embodiment said antisense-Sequence is an anti sense sequence of a gene listed in table A, B, C, or table 4. 0049 FIG. 1. Chromosomal localization of identified wherein said gene is identified in either of said tables by deregulated genes in the RA animal model, together with the expression SF Up or WJ Up or said gene is VASP. In a Quantitative Trait Loci (QTL) for Collagen Induced Arthritis particularly preferred embodiment said antisense sequence is (CIA). Red/Green arrows indicate up/down-regulated genes specific for a gene of table A indicated in table A by expres respectively. Black/Blue lettering refers to deregulated genes sion SF UP or WJ up. It is preferably determined whether in whole joints/synovial fibroblasts respectively. stress fiber formation, existence is affected. In another pre 0050 FIG. 2. Confirmation of deregulated expression in ferred embodiment said phenotype comprises adhesion, human patient samples. Quantitative RT-PCR for the indi migration, proliferation, wound-healing and/or apoptosis of cated genes for 19 RA and 8 OA samples. Values were nor said synovial fibroblast cells. The present in vitro assay is malized to the expression of B2m and were expressed as US 2008/031 1185 A1 Dec. 18, 2008

expression index. Similar results were obtained upon normal under specific pathogen-free conditions, in compliance with ization to L32. p-values were obtained in accordance to non the Declaration of Helsinki principles. (Temperature 20-22 parametric Mann-Whitney statistical test. C., Humidity 55+5%, 12 hours light/dark cycle, Water ad 0051 FIG. 3. Arthritic SFs have a rearranged cytoskeleton libitum). “Arthritic” transgenic mice (Tg 197, hTNF"), car and increased ECM adhesion. a. Immunofluoresence of ried the human TNF gene where the 3' UTR was replaced by arthritic (RA) or normal (wt) primary and immortalized syn the corresponding one from b-globin 5. Arthritic' knock in ovial fibroblasts (pSFs and iSFs respectively) for F-actin. b. mice (TNF^' on 129/C57BL6) expressed the endog Adhesion assays of arthritic or normal primary and immor enous mTNF gene, where 69 bp encompassing the TNF ARE talized SFs on purified ECM components. c. Transmission (AU-rich elements) at the 3'UTR have been deleted, resulting electron microscopy (TEM, magnificationx5000) of synovial in increased message stability and translational efficiency 8. fibroblasts (SFs) from ankle joints isolated from wt and Overall, 86 mice were used for the experiments shown in this arthritic mice. N, nuclei; r-ER, rough endoplasmic reticulum; study (since some experiments were only representative Arrow heads indicate dilation of r-ER, while arrows point to ones), as listed explicitly where appropriate. swollen mitochondria with distorted cristae. 0061 Cell isolation and culture. Synovial Fibroblasts 0052 FIG. 4. Knocking out gelsolin expression results in (VCAM") were isolated from 6-8 weeks old mice essentially disease (RA) exacerbation. a. Histopathological scores. * as previously described 7.40). Fibroblasts were selected by p=0.025 n=8-11. b. Representative histopathological analysis continuous culturing for at least 21 days and a minimum of (H/E) of arthritic joints. Shown images were assembled from four passages. No macrophage markers could be detected by multiple overlapping sections. Arrows indicate the synovial FACS analysis. Cells were grown at 37°C., 5% CO in com membrane. plete DMEM medium (Gibco/BRL) supplemented with 10% 0053 FIG. 5. SQ-RT-PCR analysis of non-muscle cofilin FBS and 100 Units/ml of penicillin/streptomycin. The cre in the Tg 197 animal model. ation and culture conditions (33°C., 10 Units/ml of murine 0054 FIG. 6. Q-RT-PCR analysis of mVASP in Tg197 rIFN-Y) of conditionally immortalized synovial fibroblast cell versus wild-type whole joint and synovial fibroblasts. lines and has been described previously 7. 0055 FIG. 7. In vitro assays studying the properties of Tg 0062 RNA extraction. Total RNA was extracted from sub SF’s versus Wt SF's (A) and (B) Tg SFs exhibit increased confluent (70-80%) cultured Synovial Fibroblasts (primary adhesion and migration to fibronectin. Mean averages of trip or conditionally immortalized) with the RNAwiz reagent licates with mean background (adhesion to BSA) values sub (Ambion Inc.), followed by single passage through an RNe tracted. Representative experiment out of three (t-test; P-0. asy column (QIAGEN, Hilden, Germany) according to 05). (C) DNA synthesis/proliferation assay in the presence or manufacturer's instructions. Total RNA was extracted from absence of 10% FBS, 48 h after treatment. Values represent whole joints using the Guanidinium isothiocynate/acid phe 3HThymidine incorporation (x10). Mean average of trip nol protocol 41, followed by single passage through an licates (t test; P-0.05). (D) Assay of wound-healing treated RNeasy column. RNA integrity was assessed by electro with Mytomycin C for 2 h. Pictures were taken at Oh and 24 phoresis on denaturing 1.2% agarose/formaldehyde gels. hafter the wound (Oh not shown). (E) TNF cytotoxicity assay RNA quantity/quality was calculated based on OD readings with cyclohexamide. After 24 h survival was determined at 260/280 nm. using the crystal violet assay. (F) Apoptosis assay using 30-300 LMetoposide. (G) Apoptosis assay using 0.01-1 LM 0063 Generation of full-length cDNA libraries. 1 strand staurosporine. This figure shows that one of the properties of cDNA synthesis was performed using SUPERSCRIPT II Tg-SF's is that they exhibit increased proliferation compare (Life technologies/Invitrogen) at 56°C. in the presence of to Wt-SF's (in addition to increased Adhesion, Migration and trehalose and sorbitol. The cap structure at the 5' end of Resistance to apoptosis). mRNAs was biotinylated and full length cDNAs were 0056 FIG. 8. Adhesion/Migration assays showing the selected, after RNAse I treatment, using streptavidin-coated effect of Lat-A on Tg 197 actin-cytoskeleton properties. beads (42). 2" strand synthesis was primed by the single 0057 FIG. 9. Proliferation assays revealing the effect of strand linker ligation method (SSLLM), where a double Lat-A on apoptosis of Tg 197 synovial fibroblasts (p<0.05, stranded DNA linker (with random 6-bp, dN6 or dGN5, 3 Student t test). overhangs) is ligated to the single-stranded cDNA (43). The 0058 FIG. 10. Apoptosis assays revealing the effect of 2 strand is made by primer extension using mixtures of Lat-A on apoptosis of Tg 197 synovial fibroblasts (p<0.05, long-range thermostable polymerases, followed by restric Student t test). tion digestion (BamHI nd XhoI) and ligation to the W-FLCI 0059 FIG. 11. DEGs validation of expression on human phagemid vector 43.44. After packaging, cDNA libraries patient samples. Q-RT-PCR analysis for the indicated genes were amplified on solid phase, as previously described 45. for 12 OAS (osteoarthritis) and 21 RA samples. Values were 0064 cDNA libraries normalization and subtraction (out normalized to the expression of (a) B2M, (b)L32 and lined in Supplementary FIG. 1a). Amplified, phagemid expressed as expression index. Nonparametric Mann-Whit W-FLCI cDNA libraries were used to infect the Cre-express ney statistical tests were used to derive p-values. ing bacterial cell line (BNN-132) and the excised plasmids were isolated 44. Double stranded plasmid DNA was nicked EXAMPLES by the site specific (fl origin) endonuclease GeneII and con Verted to circular single stranded by digestion with Exonu Example 1 clease III 45. Circular, single stranded plasmid cDNA libraries (tester libraries) were then subjected to a single step Materials and Methods normalization-subtraction step, with PCR-derived single 0060 Animals. All transgenic mice were bred and main strand anti-sense DNA drivers produced from these libraries tained on a mixed CBAXC57BL/6 genetic background and 45. Normalization refers to low CoT hybridization (CoT-2) kept at the animal facilities of the B.S.R.C. Al. Fleming with drivers produced from the “self library and aims to US 2008/031 1185 A1 Dec. 18, 2008 decrease the representation of highly expressed mRNAs. For each and every gene, the common gene transcript fre Subtraction refers to high CoT hybridization (CoT=100-200) quency, f, is calculated with the help of the formula above. with drivers produced from different libraries. Hybridized This number corresponds to the expected frequency of gene double stranded cDNAs were removed with two passes transcripts under the null hypothesis of no difference in the through a Hydroxyapatite (HAP) column. Non-hybridized, abundance of gene jacross all libraries. 2) The parameters of single stranded cDNAs were converted to double stranded the Poisson distributions giving the sampling distribution of and were subsequently electroporated into DH.10b bacterial clone abundance for each gene and library are calculated cells, where only tester cDNAs are able to propagate (due to under the null hypothesis. This parameter is equal to N, f, for the presence of a replication origin and antibiotic resistance) gene j in library i, where N, is the total number of clones 45. (taking into account ALL genes) sequenced in library i and f, 0065 High-throughput sequencing and sequence analy was calculated in Step 1. This value is equal to the expected sis. Colony picking, cDNA sequencing and sequence analysis absolute number of clones of the j" gene in the i' library were performed essentially as previously described 46-48. under the null hypothesis. 3) Then, for each library compared Sequences were filtered for primer and vector sequences 48 and each gene, a random number is generated from the Pois and masked for rodent specific and mammalian wide repeats son distribution of step 2. This number is a simulated count with RepeatMasker (http://www.phrap.org). EST clustering compatible with the null hypothesis i.e. gene frequency for a was performed with stackPACK and d2 cluster (word size 6. particular gene is the same across all libraries. This step uses similarity cutoff 0.98, minimum sequence size 50, window the random generating functions of the computer to create an size 200)49.50. Homology searches with known genes were artificial dataset corresponding to performing the actual performed with BLAST 51.52 at Unigene (http://www. experiment of library creation and analysis for each of the incbi.nlm.nih.gov/UniGene), FANTOM2 (http://fantom2. libraries compared under the null hypothesis. 4) For genejin gSc.riken.go.jp) 53 and SWISSPROT (http://www.ebi.ac. the artificial dataset of Step 3 we calculated the test statistic R uk/swissprot/) databases. BLAST results were associated by Substituting for X, the random numbergenerated in Step 3, with Gene Ontology (GO) terms (http://www.geneontology. for f, the frequency calculated in Step 1 and for N, the total org) at http://source. Stanford.edu/cgi-bin/sourceSearch (for number of clones sequenced. 5) The R values calculated in Unigene), ftp://ftp.ebi.ac.uk/pub/databases/GO/goa/SPTR/ Step 4 are sorted in descending order, and for a range of values gene association.goa sptrigZ (for SWISSPROT) and http:// for specificity (i.e. the true negative rate) we find the corre fantom2.g.sc.riken.go.jp (for FANTOM2). The detailed sponding R statistic value. These are by definition true nega results (including clone numbers, clusters/genes, blast results tives since they were obtained from libraries under the null and Evalues, accession numbers and GO assignments) can be hypothesis. 6) Steps 3-5 are repeated 1000 times allowing us found in the corresponding author's web site, at http://www. to construct the histogram shown in Supplementary FIG.2a. fleming.gr/en/investigators/Aidinis/data.html. Most of the These histograms depict the distribution of the R value as a sequencing data, have been already Submitted to public data function of the required true negative rate cutoff. The mean of bases 47, in the context of the ongoing FANTOM (Func this (empirical) distribution is used as the R-value cutoff for tional annotation of the mouse) project 53.54. the analysis of the experimental dataset. 7) The experimental 0066. Differential gene expression statistical analysis of dataset R values are computed for all the comparative librar subtractive libraries. The differential gene expression/abun ies using the observed clone counts. Genes with R values dance among the different subtracted cDNA libraries was greater than the previously established cutoff are considered calculated with a single statistical test, designed especially for to be differentially regulated between the libraries. Calcula this purpose 55. Essentially, the used formula is the entropy tions were performed in the Computer Algebra System Math of a partitioning of genes among cDNA libraries and is ematica 4.2 (www.wri.com) with the MathStatica add-on described by an R-value, which is the log likelihood ratio 56), notebooks available under request. statistic, that follows (2R) an asymptotic x distribution55). 0068 High density oligonucleotide array hybridization. The formula for the statistic R, for the j" gene is given by the cRNA probes were generated from 5 ug of total RNA and express1on: hybridized to the Mu1 1 K (subunits A and B) chip set accord ing to the manufacturer's instructions (Affymetrix, Santa Clara, Calif., USA) and as previously described 7. The chip set is designed to collectively recognize 13179 distinct where murine transcripts, where the expression level of any given gene is interrogated by 40 oligonucleotides, 20 with a perfect match sequence (PM) and 20 carrying a single mismatch X kid (MM). Hybridized chips were washed and stained in a Flu idics Station (Affymetrix) following standard protocols; sub XE Ni sequently fluorescence intensities were readby an Affymetrix scanner. All MIAME-compliant microarray data can be where m is the number of cDNA libraries, N, is the total downloaded from the ArrayExpress database (accession number of clones sequenced in the i' library, y, the counts number: E-MEXP-255), as well as from the corresponding (transcript copies) of thei" gene in the i' library, and f, is the author's web site (http://www.fleming.gr/en/investigators/ frequency of gene transcripts copies of the "gene in all the Aidinis/data.html). libraries. 0069 Microarray data pre-processing and normalization 0067. The significance of the R statistic was established (outlined in Supplementary FIG.3). Low-level analysis of the utilizing a re-sampling method, which tries to establish an resulting scanned image (as background Subtraction and “optimal' cut-off value using simulated library datasets based computation of individual probe cell intensities) was auto on the observed counts. The method proceeds as follows: 1) matically performed by MicroArray Suite 5.0 (MAS5) soft US 2008/031 1185 A1 Dec. 18, 2008 ware (Affymetrix). Briefly, the quantitative level of expres Change Model (FCM) was derived from the 10 modeling sion (Signal value), as well a qualitative measure of points using a least squares linear fit in log-log plots. The expression (Detection call), of any given gene is calculated by following modeling parameters have been obtained for the proprietary Affymetrix algorithms from the combined, back two FCM (Supplementary FIG. 4c): a slope of -0.327 and an ground-adjusted hybridization intensities of 20 PM and 20 mM oligonucleotides (probe set). All signal values from all intercept of 2.356 for the SF FCM and a slope of -0.346 and chips in the experiment were scaled to reach target intensity an intercept of 2.678 for the WJFCM. This means that WJare (TGT) of 500 or 2500, following Affymetrix recommenda intrinsically more variable than SF (as expected). Using tions for the individual chip sets used. Scaled values were slopes (a90% or a95%) and intercepts (b90% or b95%) of the then Submitted to a normalization step that has the aim to resulting regression curves, we could obtain for each given force each chip subunit's Signal distribution to have an iden sample type (SF or WJ) and for any given minimum expres tical overall shape across chips of the same subunit. Gene sion level a 90% as well as a 95% significance threshold: expression values were reorganized into two distinct data matrices, one for each chipset Subunit (A,B), where rows and columns represented genes and chips respectively. The col --- ag5%:8.95% umns of the data matrix were normalized to have the same FCoso, min(x1, x2) e quantiles using BioConductor 57. Each quantile of each (0072 For each of the 4 data matrices (i.e. Tg197/SF, TNF column was set to the mean of that quantile across arrays (Supplementary FIG. 4a). The rows of the two data matrices ^'/SF, Tg 197/WJ, TNF^'/WJ) an observed average fold were then merged and the columns of the resulting data change between experimental conditions (diseased or nor matrix were split according to the corresp9)ting animal mal) was calculated for each gene in the following way: model of rheumatoid arthritis (Tg 197, TNF") and to the FC-2-11-121/min (11,12) type of sample (SF or WJ), thus obtaining a total of 4 distinct data matrices. Finally, genes that were not called Present where 1, was the average expression level of a given gene in (detection call) or that had a normalized Signal lower than one thei" experimental condition. If no replicates were available, fifth of the TGT in at least one sample were not further the single expression value of the given gene was used instead considered. Subsequent analysis was therefore performed on of the average. For each gene, this observed fold change was the remaining 8021 probe sets. then compared to the fold change that was expected at 90% 0070 Microarray data analysis. Comparisons of Affyme (or 95%) significance level using the corresponding 'sample trix chips hybridizations and real time PCR have indicated type-specific FCM, given the observed values of , of that that chip analyses are accurate, reliable and underestimate differences in gene expression 58. Nevertheless, in order to gene. assess the system's performance, the reproducibility of the FCooo...min(Ill.,--- 12) ag0%b90%e (technical) duplicate samples was examined, interms of Pear son correlation coefficient of normalized Signals, as well as FC95%-min(1, ya95% seb95% Detection concordance (expressed as the percentage of probe Finally, genes with FV>FCoo (or >FCoso) were selected sets that were consistently detected with a “Present or an as differentially expressed at a significance level of 90% (or “Absent call in both chips). As it can be seen in Supplemen 95%, respectively). tary FIG. 4b, Affymetrix data were highly reproducible and reliable. 0073 Visualization of quantitative trait loci (QTL) and 0071. As previously reported 59,60, the significance of gene-expression data was performed with Expression view fold changes (that are commonly used as a measure of differ 62 at http://ensemblpZr.uni-rostock.de/Mus musculus/ex ential expression) is highly dependent on the expression level pressionview.QTL data were derived from Serrano-Fernadez of the corresponding genes. Therefore, in order to identify etal 63. genes that are significantly differentially expressed, we 0074 Functional clustering and determination of statisti decided not to use a fixed fold change threshold, but instead cal significance. Biological annotation, in the form of Gene model intensity-dependent fold changes between replicated Ontology (GO) “Biological process' term 17, for each of chips using a variant of a previously described approach 61. the genes (probe sets) in table 2, was obtained from the Briefly, we first calculated fold changes in pair-wise compari NetAffx portal (http://www.affymetrix.com). We then calcu Sons of replicated chips representing the transcriptome of the lated the observed GO term frequencies, as means to discover same sample type (SF or WJ) measured under the same deregulated functions (Supplementary Table 3). The statisti experimental condition using the following definition: cal significance for GO term frequencies was determined essentially as has been previously described 64. Briefly, the hypergeometric distribution was used to obtain the chance where X, was the normalized expression level of a given gene probability of observing a given number of genes annotated in in thei" chip. Genes were then ranked according to the minor NetAffx with a particular GO term and thus calculating of their two expression values (i.e. min(X,X)) and the overall appropriate p-values. More specifically, the probability of expression range was partitioned into 10 intervals, each con observing at least (k) probe sets annotated with a particular taining an equal number of probe sets. In each partition the GO term within a list of selected probe sets of size (n) was median of all (min(x,x)) values as well as the 90" and the calculated 64 as: 95" percentile fold change were determined, thus obtaining 10 distinct modeling points for each sample type (SF or WJ) and for each of the two significance levels (90% or 95%). Based on the observation that measurement variability of high-density oligonucleotide microarrays depends on signal X'i.it. intensity following a power law 11, a continuous Fold US 2008/031 1185 A1 Dec. 18, 2008

where (f) was the total number of genes within a functional (0078 Semi-quantitative PCR was performed by 20-25 category and (g) was the total number of probe sets on the chip cycles of denaturation at 95°C. for 30s, annealing at 57-62 set (13,179). C. (depending on the Tm of each individual set of primers) for 0075 Adhesion assays were performed on Cytometrix adhesion strips (Chemicon) coated with human fibronectin, 30s and extension at 72°C. for 1 min, in a final volume of 20 vitronectin, laminin and collagen I, according to the manu ul. The products were separated by electrophoresis on 1.5% facturer instructions. agarose gel and stained with ethidium bromide. Product 0076. Immunofluorescence. Cells were fixed using 4% intensity was quantified with GelWorcs 1D Advanced (V.4. paraformaldehyde/PBS and stained by standard methods. For 01) and normalized to the intensity of B2m and/or L32. The visualizing F-actin cells were stained with Alexa594-phalloi primers were selected to span two exons, while the two con din (Molecular Probes). P-tyrosine was detected using the 4G10 antibody (Upstate Biotechnology), FAK was detected trol primers were chosen from the Primer Bank database using a monoclonal antibody (clone 77, Transduction Labo (http://pga.mgh.harvard.edu/primerbank/). Primer sequences ratory). Antibodies against gelsolin were raised by immuniz (5'->3', S: sense; as: anti-sense) and product sizes (in bp) were ing rabbits with recombinant mouse gelsolin, immune serum as follows:

Aqpl (s: TCACCGGCAACTTCTCAAAC as : AGCTCTGAGACCAGGAAACA 4 OO)

BSg (s: ATGAGAAGAGGCGGAAGCCA as: CCACTCCACAGGGCTGTAGT, 426),

CD14 (s: CAATCCTGAATTGGGCGAGA as : CGAGTGGGATTCAGAGTCCA, 100),

CDC42 (S: AAGTGGCCCAGATCCTGGAA as: AGCACTGCACTTTTGGGGTT 380),

GSn (s: TGCAGGAAGACCTGGCTACT as : ATGGGTTGGTCCTTACTCAG, 3 OO)

: GAAGCAATGGGTGAACACAG as: GGGGTGGAGAACGACCTTCT 331),

Marco (S: CACAGGAATTCAAGGACAGA as : ATTGTCCAGCCAGATGTTCC, 397),

Miglap (s: CAGTCCCTTCATCA ACAGGA as : CTGCAGGAGATATAAAACGA, 274)

Mo (s: TCACACGCCACCAAGCACAA as : TGGGCTCCAGGGTAACACTG, 354),

Ptgis (s: TCACAGATGACCACACTCCC as : GCAGTAGGACGACAAATTGT, 403),

B2M (s: TTCTGGTGCTTGTCTCACTGA as: CAGTATGTTCGGCTTCCCATTC, 104) and

L32 (s: TTAAGCGAAACTGGCGGAAAC as: TTGTTGCTCCCATAACCGATG 1 OO). was used at 1:500 dilution. For western blot analysis, 10ug of 0079 Quantitative real-time PCR was performed using total cell protein was separated by SDS/PAGE, transferred to the iCycler iQ Real-Time detection system and the IQTM Immobilion-Pand probed for gelsolin and re-probed for actin SYBRGreen Supermix (Bio-Rad Laboratories), according to as an internal loading control. the manufacturer's instructions, for 4 min at 94° C. and 40 0077 RT-PCR. Total RNA were extracted from synovial cycles of denaturation\ annealing (50s at 95° C.; 50 s at fibroblast and whole joints tissue using TRIZol reagent (Life 57-62 C.). Primers were chosen from exons separated by Technologies, Rockville, Md.) in accordance with the manu large introns, and the PCR quality and specificity was verified facturer's instructions. RNA yield and purity were deter by melting curve analysis and gel electrophoresis. Values mined spectro-photometrically at 260/280 nm. 1 strand were normalized to B2m and/or L32 (using the same primers cDNA synthesis was performed using the MMLV reverse as for the semi-quantitative PCR). Mouse (m) and human (h) transcriptase and oligo-dTs (Promega). primer sequences and expected lengths were as follows:

mAqp 1 (s: TCACCCGCAACTTCTCAAAC as: TCATGCGGTCTGTGAAGTCG 123), mGSn (s: TGCAGGAAGACCTGGCTACT as : TCGATGTACCGCTTAGCAGA, 13 O). hAqp 1 (s: CTCCCTGACTGGGAACTCG as: GGGCTACAGAGAGGCCGAT, 182), hBSg (s: TTCCTGGGCATCGTGGCTGA as: GCGGACGTTCTTGCCTTTGT, 159), US 2008/031 1185 A1 Dec. 18, 2008

- Continued hCo.14 (S: CGGCGGTCTCAACCTAGAG as: GCCTACCAGTAGCTGAGCAG, 142), hcdc42 (s: AATTGATCTCAGAGATGACC as: TTTAGGCGTTTCTGTGTAAG, 15O) hCGsn (s: GGTGTGGCATCAGGATTCAAG as : TTTCATACCGATTGCTGTTGGA, 199), hMarco (S: TGGGACGAGATGGAGCAAC as: CCCTTAGTTCCAGTTTCCCCTT, 193 ) . hM (s: TTGGTGCTGAACGTCTGGG as: CTGTGCCAGGGGCTTAATCTC 249), h2M (s: CTGAAAAAGATGAGTATGCC as: ACCCTACATTTTGTGCATAA, 2O2) and hL32 (s: TTAAGCGTAACTGGCGGAAAC as: GAGCGATCTCGGCACAGTAA, 210)

0080 Cycle threshold (Ct; the first cycle that amplification (Tg 197, hTNF+/-)5 and their normal (wt) littermates. Each can be detected) values were obtained from the iCycler iQ sample (out of 4: RA SF, wt SF, RAWJ, wt WJ) consisted of software for each gene of interest (GOI) and control house equimolar amounts of RNA pooled from 4 (2 male and 2 keeping genes (HKG, L32 and/or b2m), together with ampli female) mice (16 mice altogether), thus equalizing (to an fication efficiencies (m; 80-120%). For the mouse samples, extent) biological diversity. Similar samples (biological rep we calculated the relative expression of the samples (S) to wt licates: different extractions, same pooling strategy) were controls as reference samples (R) using the gene expres used for both the creation of subtractive cDNA libraries, as sion relative quantification Microsoft(R) Excel add-on well as for the hybridization of oligonucleotide microarrays. macro (Bio-Rad Laboratories) that utilizes the following for I0084 Subtractive cDNA libraries and large scale sequenc mulas: Relative expression=2^^), where ACt=GOI ing. Full-length, cDNA libraries were constructed with the Ct-HKG Ct. For the human samples, Ct values were con cap-trapper method, by cloning in the w-FLCI vector, as Verted to concentration values (ng/ml) utilizing the standard described in Materials and Methods. Following solid phase curve made by serial dilutions (in duplicates) of a reference amplification and plasmid excision, the four different cDNA sample. Values were normalized to the corresponding values libraries (RASF, wt SF, RAWJ, wt WJ) were normalized and of the reference (housekeeping) gene(s) and presented (in subtracted to each other, as outlined in Supplementary FIG. logarithmic scale for visualization purposes) as expression 1a. Normalization refers to low CoT hybridization of the index. tester library with drivers produced from the “self library 0081 Arthritic score and Histopathology Paraffin-embed and aims to decrease the representation of highly expressed ded joint tissue samples were sectioned and stained with mRNAs. Subtraction refers to high CoT hybridization with haematoxylin and eosin. Arthritic histopathology was drivers produced from different libraries and aims to remove assessed (in a blinded fashion) separately for synovial hyper mRNAs common in both populations. Due to experimental plasia, existence of inflammatory sites, cartilage destruction design, the resulting subtracted libraries (LO, L1, L2 and L7) and bone erosion using a semi-quantitative (0-5) scoring as contain cDNAs from the tester cDNA library only and there described previously 65. fore constitute libraries of upregulated genes in the corre 0082 Transmission electron microscopy. Ankle joints sponding tester library (or of down-regulated genes in the (dissected from the right hind leg of each mouse 3 Tg 197 driver library). 27511 cDNAS/clones from the subtracted and 3 wt) were split open longitudinally through the midline libraries were sequenced and clustered to 9176 clusters/ between the tibia and the talus, were prefixed with 2.5% genes, as described in Materials and Methods and Summa glutaraldehyde in PBS (pH 7.4) O/N. After postfixing with rized in Supplementary FIG. 1b. Each gene was then anno 1% osmium tetroxide in PBS for 2 h, they were dehydrated in tated through BLAST homology searches at Unigene, a graded series of ethanol and processed into Araldite. Semi FANTOM and SWISSPROT databases, as described in Mate thin sections (1.0) were cut and stained with methylene blue rials and Methods. In Summary, among the 9176 genes found, to observe the orientation under the light microscope. 7977 corresponded to known genes, while 1199 had Ultrathin sections (80-90 nm) were then cut with an ultrami sequences not reported previously. Each gene was repre crotome (Riechert-Jung Ultracut E), mounted on copper sented by a different number of clones in almost all libraries, grids, counterstained with uranyl acetate and lead citrate, and directly proportional to Subtraction efficiency and transcript evaluated through electron microscope (CM120 Biotwin, FEI abundance. The relative distribution of each gene in each Company). library is the true measure of differential expression, which can be obscured by Sampling errors arising by chance when Results the clones are selected. Therefore and in order to identify the truly differentially expressed genes, a likehood value (R) was 0083) To discover genes and cellular pathways that par assigned to each gene from pair-wise comparisons of the ticipate in the pathogenesis of RA on a large scale, we per relative libraries (SF/LOL1, WJ/L2L7). The statistical signifi formed a twin high throughput approach, comprising of two cant thresholds were then calculated, as described in Materi entirely different methodologies, governed by different con als and Methods and presented in Supplementary FIG.2. Two strains and analyzed by different statistics. Total RNA significance levels were selected (Summarized in Table 1a): a samples were extracted from whole joints (WJ) and synovial very high one (99.99%; ps0.0001) to report the results inde fibroblast (SF) ex vivo cultures, isolated from 6 weeks old pendently, and a lower one (99%; pa 0.01) for the comparison arthritic (RA; disease severity index 3, see FIG. 4) mice with the corresponding results from the DNA microarrays US 2008/0311185 A1 Dec. 18, 2008 hybridizations. Known (with a Unigene cluster ID) differen from WJs and 5 from SFs, ii) different direction and degree of tially expressed genes at 99.99% significance level are pre deregulated expression: 6 upregulated and 5 down-regulated sented in Supplementary Table 1. and iii) biological interest and potential follow up. Moreover, 0085 Oligonucleotide microarray hybridizations. Fluo two of them (SF: Gsn, Aqp 1) were also confirmed by Real rescent-labeled cRNA probes made from similar (biological time RT-PCR (Supplementary FIG. 6e, summarized in Table replicates, see above) samples of total RNA (RASF, wt SF, 2). RAWJ, wt WJ) were utilized to hybridize (in duplicates) the I0088. In an attempt to combine gene expression analysis Affymetrix MullKoligonucleotide DNA chipset (8 chipsets, with genetic linkage analysis, all differentially expressed 16 chips total). Furthermore, similar samples (equimolar genes were mapped to the chromosomes, together with the amounts of RNA pooled from 4-2 male and 2 female-6 known Quantitative Trait Loci (QTLs, chromosomal regions/ weeks old arthritic mice; severity index 3) of total RNA (from genes segregating with a quantitative trait) for an induced SF and WJ) from another animal model of Arthritis (sponta animal model of Arthritis, Collagen Induced Arthritis (CIA) neous, knock-in, TNFAARE+/-) 8 were used for additional 14. As graphically represented in FIG. 1 (and summarized chip hybridizations (4 chipsets, 8 chips total). The MIAME in Table 2), eight genes mapped to CIA QTLs (WJ: Ctss, compliant 9 microarray data (which can be found at the Pitpnm, Ncf1, Psmb8, Siat8e, SF: Rab14, Aqp1, and Gsn), ArrayExpress database 10, accession number: E-MEXP proving the validity of the selected gene list even further. The 255) were normalized and analyzed as outlined in Supple expression level of seven of the genes found deregulated in mentary FIG. 3 and as described in detail in Materials and the arthritic mice and confirmed by RT-PCR in mouse Methods and supplementary FIG. 4. Differentially expressed samples, was also examined in human patient's RNA samples genes (DEG) were selected, as described in Materials and with Real-time RT-PCR analysis. Due to the lack of normal Methods, utilizing a sample-specific Fold Change Model (wt) human synovium samples, we compared the expression (FCM)11 at different significance levels (90%/ps 0.1 and of 19 RA samples with 8 Osteoarthritic (OA) samples as 95%/ps 0.05), where selected DEG have always an observed controls, a consensus strategy for differential expression fold change higher than the expected fold change. The results analysis in Arthritis 15, 16. As shown in FIG. 2 and summa for both animal models are summarized in Table 1b. Sample rized in Table 2, the deregulated expression of six (out of specific differentially expressed genes common in both ani Seven assayed) of the genes was confirmed in humans as well, mal models (significance: 95%, p<0.05) are presented in including four with high statistical significance (Gsn, Aqp1. Supplementary Table 2. Bsg, Mb.), thus extending the validity and utility of the mouse I0086) Cross platform comparison and validation. Both model generated deregulated gene list to the human disease. differential expression analysis methods presented above I0089 Arthritic synovial fibroblasts have a rearranged produced lists of deregulated genes of high statistical signifi cytoskeleton. The twin high throughput expression profiling cance. In order to validate the results independently and in approach described above, yielded a large number of disease order to avoid performing numerous RT-PCRs, differential implicated, deregulated genes, of high statistical significance. expression results by both platforms of analysis were com Furthermore and in order to: i) prove the validity and extend pared to each other, for the same animal model (Tg 197, the utility of the expression data analysis even further, ii) to hTNF+/-) and sample type (W.J. SF), as well as for the same infer deregulated biological functions from the gene expres direction of deregulated expression (up or down). The com sion data and iii) to define functional criteria for further gene parison, at significance levels 99% for the libraries and 90% Selection, the selected genes were annotated in the form of the for the microarrays (selected based on similar output gene Gene Ontology (GO)17 term "Biological process”. GO numbers), was performed through the NetAffx database term frequencies in the selected gene lists were then calcu (www.affymetrix.com). Remarkably (since there have been lated and their statistical significance was estimated, as Very few examples of expression profiling cross-platform described in Materials and Methods. As shown in Supple overlaps 12,13), 46 (15 for SF; 31 for WJ) genes were com mentary Table 3, predicted deregulated functions in SFs monly predicted as up/down-regulated by both methods include, as expected, collagen catabolism, complement acti (Table 2) (with a combined p value of 0.001). Vation, immune and stress response. Interestingly, 5 out of 26 I0087. In order to verify the validity of the twin high Statistical significant (p<0.01) deregulated GO functions con throughput approach and to prove that the reported gene list is cerned (directly or indirectly) the actin cytoskeleton, suggest self-validated, a number of the predicted deregulated genes, ing that arthritic SFs have a rearranged actin cytoskeleton. In were confirmed with different methods. 40 genes were also order to confirm the prediction, F-actin was visualized in vitro predicted deregulated in the knock-in disease model (TNF on arthritic as well as wt SFs (both primary and immortal AARE+/-, Table 2). 20 genes were previously reported in the ized). As evident from FIG. 3a, arthritic SFs exhibit pro literature to be associated with RA (Supplementary FIG. 5: nounced stress fibers, thus validating the in silico, expression summarized in Table 2), as these were identified through the based hypothesis. Biolab Experiment AssistantTM text-mining software (BEA, 0090 Stress fibers within fibroblasts allow them to exert BIOVISTA, Greece). 11 representative genes (SF: Gsn, tension on the Extra Cellular Matrix (ECM) surrounding Aqp1, mglap, cdc42 hom. Hip: WJ: Marco, Hp, CD14, Mb, them an essential process in wound healing. It is well Bsg, Ptgis) were further confirmed by semi-quantitative (SQ) expected that differences in the actin cytoskeleton reflect RT-PCR (Supplementary FIG. 6 a,b,c, summarized in Table altered ECM attachment properties and/or vice versa. Indeed, 2). All SQ-RT-PCRs were performed in the linear-range of the arthritic SFs were shown to adhere to different proteins of the reaction (at three different concentrations normalized against ECM (fibronectin, vitronectin, laminin and collagen) with housekeeping genes) in biological replicates (same pooling increased affinity in vitro (FIG. 3b). strategy) of the samples used for both the subtractive libraries I0091 Attachment to the ECM (and associated cytoskel and the microarrays hybridizations. Representative genes etal changes), mediated mainly through engagement and were selected on the basis of i) different sample source: 6 clustering of transmembrane integrin molecules, largely US 2008/031 1185 A1 Dec. 18, 2008

define cell shape and morphology, as well as their behavior straight line. Cells were then allowed to continue to grow and and fate. Increased adhesion to the ECM is expected to lead to pictures were taken at 0, 24 and 48 h, after the scrapping. a more elongated shape. In order to confirm the increased adhesion of the arthritic fibroblasts in vivo, we examined Apoptosis Assays ankle joints from arthritic or wt littermate mice on an ultra 0096 Confluent mouse synovial fibroblasts were washed structural level with transmission electron microscopy (FIG. twice with phosphate-buffered saline (PBS) and placed in 3c). In wt mice, SF's contained prominent nuclei, abundant DMEM without serum for 24 h. The following day 1.5x10" rough endoplasmic reticulum (r-ER) and mitochondria of cells were plated into each well of a 96-well flat-bottom plate different shapes and sizes. In contrast, remarkable modifica (Costar, Corning Inc.) in 100 ul of media plus serum. After 24 tion of the SFs was noticed in the joints of arthritic mice, h, cells were treated with 12.5-100 ng/ml TNFC. plus 0.5-3 where randomly flattened cells had (most of them) elongated ug/ml cyclohexamide (CHX), 30-300 uM Etoposide and shape, characterized by dilation of the r-ER and by swollen 0.01-1 LM Staurosporine, for 24 h at 37° C. overnight. Cell mitochondria with distorted cristae. Survival was determined using the crystal violet assay. The 0092. We thus found a new pathogenic mechanism in RA: absorbance at 570 nm was determined on a microplate reader The promotion of actin polymerization and rearrangement of (SPECTRAmax PLUS, Molecular Devices, Sunnyvale, Calif., USA), as previously described (Aidinis et al., 2003). the actin cytoskeleton. Gelsolin (GSn) is a gene that maps in Results are expressed as the means-S.E. of one representative one of the CIA QTLS (FIG. 1) and whose expression was experiment performed in triplicate. Statistical analysis was found downregulated in arthritic SFs by both subtractive performed using Student's t test, with a level of significance libraries and microarray hybridizations. Its deregulated set at p-0.05. expression was confirmed with Real-Time RT-PCR in both mouse (Supplementary FIGS. 5b, c and e) and human Drug Treatment. In Vitro Assays samples (FIG.2), as well as with western blot in immortalized SFs (Supplementary FIG. 5d). The gene encodes for an actin 0097 Disruption of actin cytoskeleton was induced with 4 binding protein with filament severing properties 18 and ug/ml of Latrunculin A (Sigma) for 2 hat 37° C. as previously Gsn-f- fibroblasts have excessive actin stress fibers 18. described (Spector et al., 1989). As control of TNFC. involve very similar to the ones observed in arthritic SFs (FIG.3a). In ment in actin-cytoskeleton organisation 1 Jug/ml of anti-TNF order to prove the involvement of cytoskeleton organization (Infliximab) was added on the cells for the same period of in the pathogenesis of RA and to highlight the role of gelsolin time. in it, the arthritic mice (Tg 197, hTNF--/-) were mated with the gSn knock out mice (gSn-/-) 18. Predictably, knocking Results outgSn expression from SFs would promote RA pathogenesis Validation of Deregulated Genes (DEGS), Identified by by inhibiting the severing activity of gSn. As shown in FIG. 4. Microarrays, in the Tg 197 Mouse Model Using SQ-RT-PCR abolishing gSn expression resulted in hyperplasia of the Syn (0098 Total RNA was extracted from 5-8 Tg 197 mice ovial membrane and exacerbation of the disease. (from both whole joint tissue and ex-vivo synovial fibroblast cell cultures) and pooled. For the comparison group, total Example 2 RNA was extracted from wild-type litter mates (from both whole joint tissue and ex-vivo synovial fibroblast cell cul Materials and Methods tures) and pooled. The results are depicted in table 3. 0099 Preferred proteins that affect cytoskeleton arrange 0093. Unless mentioned herein under example 2, the ment include actin binding proteins, controlling the nucle materials and methods of example 1 where used. ation, assembly, disassembly, crosslinking of actin filaments and/or the size, localization and dynamics of the un-polymer SQ and Q-RT-PCR Analysis of Mouse Cofilin 1 and VASP ized actin. A protein of table 4 is preferred in the present invention. A particularly preferred protein includes a protein 0094) Details on cell culture, RNA extraction from both expressed by the gelsolin gene, VASP and/or Cofilin (Galler whole joint tissue and ex-vivo synovial fibroblast cell cultures et al., 2006: Hotulainer et al., 2005; Paavilainen et al., 2004). and RT-PCR conditions can be found at Aidinis et al., 2005. Others, like Autotaxin and SHIP2 (Song et al., 2005; von The primers specific for mGofilin 1 were s:TGGGT Wichert et al., 2003). These proteins are major players in GAAACTCTGGGAGAT as: CCATAGTGGAAGCCT signaling pathways leading to re-organization of the actin GATGC, 399 bp and for mVASP s: CTTGCCAAGGAT cytoskeleton. GAAGTCG as: GCTTCACCCTCTCCAAGT, 98 bp. 0100 Cofilin 1 (non-muscle cofilin) promotes actin dynamics by depolymerising and severing actin filaments, in Wound-Healing Assays a similar manner like gelsolin (Hotulainer et al., 2005). As FIG. 5 shows, cofilin 1 was down-regulated in both Tg 197 0.095 Cells cultured at confluency (3x10 per well) were whole-joint and synovial fibroblasts compare to wild-type seeded in 6-well plates (in DMEM supplemented with 10% OS. FCS and 100 Units/ml of penicillin/streptomycin at 37°C., 0101. In order to evaluate the gene expression status of 5% CO) and allowed to adhere for 24 h. The following day VASP in Tg SF’s we have performed Q-RT-PCR in both cells were treated with 10 ug/ml Mitomycin C (Sigma) in whole joint samples and ex-vivo synovial fibroblast cells culture medium at 37° C. for 2 h, to block cell proliferation. culture (same material as for the SQ-RT-PCR). Surprisingly, Assays of wound healing were performed by Scrapping the VASP is found to be over-expressed in the whole joint of Tg cell monolayer with the edge of a pipette tip, forming a mice compare to wild-type littermates, while it is under US 2008/031 1185 A1 Dec. 18, 2008 expressed in the SF's (FIG. 6). This finding although unex while they possess greater motility compared with wild-type pected could probably Suggest the presence of another cell mice. This increased adhesion leads to a more elongated type within the arthritic joint that might be linked with VASP shape of the SF's. For this reason a number of in vitro assays expression. Interestingly, there are some indications of the were developed/standardised, including Adhesion, Migra possible presence of myofibroblasts in the arthritic synovium tion, Proliferation, Wound Healing and Apoptosis (FIG. 7). (Aidinis et al., 2005). Myofibroblasts are differentiated fibro 0108. These in vitro assays demonstrate the distinct prop blasts with a crucial role in manifesting tension during erties of TgSFs, including increased adhesion, migration and wound-healing and pathological contracture (Tomasek et al. apoptosis resistance compare to Wt SF's, while proliferation 2002). is increased. Targeting the expression cytoskeleton related 0102 KO mice for genes known to be involved in the genes (such as preferably Gelsolin, VASP, Cofilin 1 and/or cytoskeleton or crossed with Tg 197 and/or with Collagen Aquaporin 1) affects cytoskeleton arrangement. Knocking VI-Cre knockout with the purpose of (i) exacerbate, (ii) ame down the expression of Such genes has an effect in the actin liorate the disease and (iii) make the disease specific for cytoskeleton machinery of the SF's and this is revealed by the synovial fibroblasts. above assays. 0103 VASP is a key player in the actin cytoskeleton as it is 0109. In vitro analysis on the effect of Latrunculin A (Lat shown in vitro to promote F-actin formation while in vivo A), an inhibitor of F-actin polymerization (Giganti and regulates and promotes actin polymerization (Hauser et al. Friederich, 2003), has indicated the potential of actin-drugs in 1999). In the present invention a cross of VASP KO with altering the cytoskeleton properties of Tg 197 synovial fibro Tg 197 is expected to lead to amelioration of the disease. blasts. In detail, treatment with Lat-A has resulted in 0104 Cofilin depolymerises and severs actin filaments, in decreased adhesion of TgSF's, while migration seems to be a similar manner like gelsolin, in the present invention a increased compare to Wt (FIG. 8) This is in keeping with crossing of Cofilin KO with Tg 197 is expected to result in previous findings of Lat-A treatment and myofibrils (Wang et exacerbation of the disease (Gurniak et al., 2005). al, 2005). At the same time Tg SF's treated with anti-TNF do 0105 LST-1 was found to be over-expressed in whole not show altered adhesion/migration properties, suggesting joints from Tg 197 (Aidinis et al., 2005), while in vitro studies that this effect is not TNF-driven. In addition, treatment with have correlated this gene with a function in cytoskeleton, by Lat-A seems to have a dramatic effect on the proliferation creation of long filopodia and thus accumulation of F-actin levels of TgSFs which decline lower even than the levels of (Raghunathan et al., 2001). Creation of a KO of this gene and WtSFs, while DMSO and TNFO. do not have an effect on crossing with Tg 197 provides further evidence of the involve TgSFs rate of proliferation (FIG. 9). Most importantly, apo ment of this gene in cytoskeleton-related rheumatoid arthritis. ptosis assays using staurosporine as inducer of cell death RhoA, Rac and Cdc42 belong to the Rho family of GTP revealed that Lat-A disruption of the cytoskeleton causes binding proteins known to induce actin polymerization decreased resistance to apoptosis at the levels of the Wt SFs (Lamarche et al., 1996). Conditional ablation of these genes (FIG. 10). This result comes in concordance with previous (Chrostek et al., 2006) and crossing with ColVI-Cre and findings that show that treatment with Cytochalasin D (an Tg-197 would ameliorate the disease specific for synovial inhibitor of rapid actin polymerization) enhances apoptosis in fibroblasts. A similar outcome is predicted from crosses with HeLa cells, while phalloidin (an inducer of actin polymeriza the serum response factor (SRF) KO, which is a transcription tion) does not affect apoptosis (Kulms et al., 2002). factor regulating the expression of several actin-cytoskeleton genes (RhoA, Vinculin, Actin) and it’s ablation was shown to Validation of Selected Deregulated Genes (DEGs), on result in low F-actin levels and a reduced adhesive and migra Human RA and OA Samples by Q-RT-PCR Analysis tory capacity (Wiebel et al., 2002). 0110. The patent states that “analyzing by quantitative 0106. A conditional KO model for Autotaxin (ATX) will be also utilized to further support the role of cytoskeleton RT-PCR (Q-RT-PCR) DEGs selected from tables A, B and C re-modeling in RA. Activated hyperplastic RA Synovial may provide a good correlation with the absence or presence Fibroblasts (SFs) exhibit reorganization of their actin cytosk of inflammatory arthritis (IA). For this purpose a number of eleton and deregulated ECM adhesion that could explain the genes were selected (i.e. C3, CstB, Dusp1, LSP1, PITPNM, tumor-like behavior of the arthritic synovium. Lysophospha RAB14, SHIP2, TGFB114, TIMP1, TIMP3, TUBA1, C1GA, tidic acid (LPA), the enzymatic product of autotaxin SHIP, LST1) based on the following criteria: a) DEGs that (ATX)—a lysophospholipase D highly expressed in a variety showed the greatest level of differential expression following of tumors, is an important lipid mediator that has been found validation in the Tg 197 mouse model, by SQ-RT-PCR analy to induce cytoskeletal reorganization, proliferation and sis; b) DEGs identified in the Tg 197 mouse microarray data migration in many cell types. Preliminary studies have shown and are key players in the regulation of actin cytoskeleton and that the expression of ATX is upregulated in SFs from Tg 197, c) candidate genes of interest that were not identified by the and ex vivo cultured RASFs secreted significantly more ATX microarray experiments (SHIP and SHIP2) but are known to protein than their wild-type counterparts. be involved in the mechanism of actin cytoskeleton. 0111. In order to carry out the above task a collaboration was established with Dr Wilson's lab at the University of Development of InVitro Assays to Test Properties of Ex-Vivo Sheffield, UK, who agreed to provide us with sufficient Primary Synovial Fibroblasts (Migration, Adhesion Prolif amounts of cDNA from biopsies of RA and OA patients eration and Apoptosis) (Ethics no: SSREC01/333; Project title “Regulation of cytok 0107. It is provided here that TgSF's exhibit distinct phe ine genes in RA. In detail, Synovial tissue was obtained notype compared to those of wildtype mice. It was shown that during joint replacement Surgery from patients with RA Tg SFs have rearranged actin cytoskeleton and exhibit pro (n=21), defined according to the 1987 revised criteria of the nounced stress fibers. We also showed that Tg SF's adhere American College of Rheumatology (formerly, the American with increased affinity to the proteins of the ECM in vitro, Rheumatism Association). Synovial tissue was also obtained US 2008/031 1185 A1 Dec. 18, 2008

from patients with osteoarthritis (OA) (n=12), who were I0121 8. Machesky LM, etal (1999) Scar, a WASp-related undergoing knee joint replacement Surgery. Patient's protein, activates nucleation of actin filaments by the informed consent was obtained according to the national Arp2/3 complex. Proc Natl AcadSci USA96:3739-44. ethical rules, as these were implemented by the correspond I0122) 9. Gautreau A. et al (2002) ERM proteins and NF2 ing local ethic commission. tumor Suppressor: the Yin and Yang of cortical actin orga 0112 Quantitative real-time PCR was performed as pre nization and cell growth signalling. Curr Opin Cell Biol viously described (Aidinis et al., 2005). Primers were chosen 14:104-19. from either exons separated by large introns or exon bound (0123 10. Hoffman L N, et al (2006) J Cell Biol Genetic aries and the PCR quality and specificity was verified by ablation of Zyxin causes Mena/VASP mislocalization, melting curve analysis and gel electrophoresis. Values were increased motility, and deficits in actin remodeling 172: normalized to both B2M and L32. Human primer sequences 771-82. and expected lengths were as follows (listed in the 5' to 3' (0.124 11. Weaver A M, et al (2001) Cortactin promotes direction and designated ass; sense and as; antisense): hC3 (S. and stabilizes Arp2/3-induced actin filament network for GGCCTTTGTTCTCATCTCGCT; as, ATAGGATGTGGC mation. Curr Biol 11:370-4. CTCCACGTT, 274 bp), hCSTB (s, CCTGTGTTTAAGGC (0.125 12. Schluter K, etal (1997) Profilins as regulators of CGTGTCAT, as, GGAGAGATTGGAACACTCGCAG, 115 actin dynamics. Biochim Biophys Acta 1359:97-109. bp), hidUSP1 (s, ACAAGGCAGACATCAGCTCCTG; as, (0.126 13. Carlier M, et al., (1993) Modulation of the inter TAAGCAAGGCAGATGGTGGCT, 130 bp), hLSP1 (s, action between G-actin and thymosin beta 4 by the ATP/ CAACAATTAAGAGCACCCCATC; as, TGGCA ACAG ADP ratio: possible implication in the regulation of actin GAAGCAACTTCT, 223 bp), hRAB14 (s, GCTTATTGTTC dynamics. Proc Natl AcadSci USA 90:5034-8 CTCGAAGCGAG; as, ATCCAAGCTTCCATCCTGAATG, I0127. 14. Rohatgi R, et al (2001) Nck and phosphatidyli 107 bp), hSHIP2 (s, GGCTCGGAGGAACCAAAACTAC: nositol 4,5-bisphosphate synergistically activate actin as, TGAGGTCCCCAAACCAGAAGA, 119 bp), polymerization through the N-WASP-Arp2/3 pathway. J hTGFB114 (s, TAGCTCCTCTGGTGCAAGTGTG.; as, Biol Chem 276:26448. TCCACTTCTTCTCTGACCGCA, 102 bp), hLST1 (s, CGT I0128 15. Foger N, etal (2006) Requirement for coronin 1 GATGAGGAACTTGAGG; as, GCGATAACATTAGGT in T lymphocyte trafficking and cellular homeostasis. Sci CAGGG, 114 bp). ence 313:839-42. 0113. Of the 8 cytoskeleton-related DEGs studied, 5 I0129. 16. Nyakern-Meazza M, et al (2002) Tropomyosin showed the same pattern of deregulated expression in human and gelsolin cooperate in controlling the microfilament samples and therefore merit further investigation (FIG. 11). system. J Biol Chem 277:28774-9. Also another important conclusion is the significant associa I0130. 17. Kordowska J, et al (2006) Phosphorylated tion of down-regulation of gene expression of Gelsolin in RA 1-caldesmon is involved in disassembly of actin stress patients, in another population. This strengthens the associa fibers and postmitotic spreading. Exp Cell Res 312:95-110. tion of Gelsolin and actin-cytoskeleton re-arrangements in I0131 18. Cooper J A and Schafer DA (2000) Control of the pathogenesis of RA patients. actin assembly and disassembly at filament ends. Curr Opin Cell Biol 12:97-103. REFERENCES TO EXAMPLE 2 I0132) 19. Fujisawa T. etal (2001) The shapes and sizes of two domains of tropomodulin, the P-end-capping protein 0114 1. Paavilainen V O, et al (2004) Regulation of cytoskeletal dynamics by actin-monomer-binding pro of actin-tropomyosin. FEBS Lett 498:67-71. teins. Trends Cell Biol 14:386-394. I0133). 20. Rodal AA, et al (1999) Aiplp interacts with cofilin to disassemble actin filaments. J Cell Biol 145: 0115 2. Hotulainer P. et al (2005) Actin-depolymerizing 1251-64. factor and cofilin-1 play over-lapping roles in promoting I0134) 21. Moseley J B, et al (2006) Twinfilin is an actin rapid F-actin depolymerization in mammalian nonmuscle filament-severing protein and promotes rapid turnover of cells. Mol Biol Cell 16:649-64. actin Structures in vivo J Cell Sci 119:1547-57. 0116 3. von Wichert G. etal (2003) Force-dependent inte I0135 22. DeMali KA, et al (2002) Recruitment of the grin-cytoskeleton linkage formation requires down-regu Arp2/3 complex to Vinculin: coupling membrane protru lation of focal complex dynamics by Shp2. Embo J Sion to matrix adhesion. J Cell Biol 159:881-91. 22:5O23-35. I0136. 23. Ohta Y. etal (2006) FilCAP, a Rho- and ROCK 0117. 4. Song J, et al (2005) Autotaxin (lysoPLD/NPP2) regulated GAP for Rac binds filamin A to control actin protects fibroblasts from apoptosis through its enzymatic remodelling. Nat Cell Biol 8:803-14. product, lysophosphatidic acid, utilizing albumin-bound I0137 24. Pawlak G and Helfman DM (2002) MEK medi substrate. Biochem Biophys Res Commun 337:967-75. ates v-Src-induced disruption of the actin cytoskeleton via 0118 5. Svitkina TM, et al (1999) Arp2/3 complex and inactivation of the Rho-ROCK-LIM kinase pathway. Biol actin depolymerising factor/cofilin in dendritic organiza Chem 277:26927-33. tion and treadmilling of actin filament array in lamellipo I0138 25. Schratt G. etal (2002) Serum response factor is dia. J Cell Biol 1.45:1009-26. crucial for actin cytoskeletal organization and focal adhe 0119) 6. Rohatgi R, et al (1999) The interaction between sion assembly in embryonic stem cells. J Cell Biol 156: N-WASP and the Arp2/3 complex links Cdc42-dependent 737-50. signals to actin assembly. Cell 97:221-31. I0139 26. Lamarche N. etal (1996) Rac and cdc42 induce 0120) 7. Gallego MD, et al (2005) WIP and WASP play actin polymerization and G1 cell cycle progression inde complementary roles in T cell homing and chemotaxis to pendently of p65PAK and the JNK/SAPK MAP kinase SDF-1a. Inter Immunol 18:221-32. cascade. Cell 87:519-29. US 2008/031 1185 A1 Dec. 18, 2008

0140) 27. Trentham D. E. et al (1977). Autoimmunity to 0158 45. Hauser W. etal (1999) Megakaryote hyperplasia type II collagen an experimental model of arthritis. J Exp and enhanced agonist-induced platelet activation in Mod 146:857-68. vasodilator-stimulated phosphoprotein knockout mice. 0141) 28. Elson CJ, etal (1992). Mediators of joint swell Proc. Natl. Acad. Sci. USA. 96:8120-5. ing and damage in rheumatoid arthritis and pristane 0159 46. Gurniak CB, etal (2005) The actin depolymeris induced arthritis. Autoimmunity 13:327-31. ing factor n-cofilin is essential for neural tube morphogen 0142. 29. Glant T. et al (2003). Proteoglycan-induced esis and neural crest cell migration. Develop Biol 278:231 arthritis: immune regulation, cellular mechanisms, and 41. genetics. Crit. Rev Immunol 23:199-250. 0160 47. Aidinis V, et al (2005) Cytoskeletal rearrange 0143. 30. van den Broek MF, et al (1988) Streptococcal ments in Synovial fibroblasts as a novel pathophysiological cell wall-induced arthritis and flare-up reaction in mice determinant of modeled rheumatoid arthritis. PloS Genet. induced by homologous or heterologous cell walls. Am 1:e48. Pathol 133:139-49. 0144 31. van Lent, etal (2003) Expression of the dendritic 0.161 48. Raghunathan A. etal (2001) Functional analysis cell-associated C-type lectin DC-SIGN by inflammatory of B144/LST1: A gene in the tumor necrosis factor cluster matrix metalloproteinase-producing macrophages in rheu that induces formation of long filopodia in eukaryotic cells. matoid arthritis synovium and interaction with intercellu Exp Cell Res 268:230-44. lar adhesion molecule 3-positive T cells. Arthritis Rheum (0162 49. Chrostek A, et al (2006) RacI is crucial for hair 48:360-9. follicle integrity but is not essential for maintenance of the 0145 32. Schalkwijk et al. (1985) Cationization of cata epidermis. Mol Cell Biol 26:6957-70. lase, peroxidase, and Superoxide dismutase. Effect of (0163 50. Wiebel F F, et al (2002) Generation of mice improved intraarticular retention on experimental arthritis carrying conditional knockout alleles for the transcription in mice. J Clin Invest 76:198-205. factor SRF. Genesis 32:124-6. 0146 33. Hultqvist M. etal (2004) Enhanced autoimmu (0164. 51. Wang J, et al (2005) Differential effects of nity, arthritisandencephalomyelitis in mice with a reduced Latrunculin-A on myofibrils in cultures of skeletal muscle oxidative burst due to a mutation in the Ncfl gene. Proc cells: Insights into mechanisms of myofibrillogenesis. Cell Natl Acad Sci USA 101:12646-51. Motility Cytoskeleton 62:35-47. 0147 34. Sakaguchi N, et al (2003) Altered thymic T-cell 0.165 52. Kulms D, et al (2002) Apoptosis induced by selection due to a mutation of the ZAP-70 gene causes disruption of the actin cytoskeleton is mediated via activa autoimmune arthritis in mice. Nature 426:454-60. tion of CD95 (Fas/APO-1). Cell Death Differentiation 0148 35. Hang L, etal (1982) A spontaneous rheumatoid 9:598-608. arthritis-like disease in MRL/1 mice. J Exp Med 155:1690 (0166 53. Xia X G, et al. (2006) Pol II-expressed shRNA 1701. knocks down Sod2gene expression and causes phenotypes 0149 36. Kouskoff V, et al (1996) Organ-specific disease of the gene knockout in mice. PloS Genet. 2:e 10. provoked by systemic autoimmunity. Cell 87:811-22. 0167 54.... Song,SOh9, A. (2003) RNA interf1nterference target1ngrgeting FFaS 0150 37. Iwakura Y, et al (1991) Induction of inflamma protects mice from fulminant hepatitis. Nat Med 9:347 tory arthropathy resembling rheumatoid arthritis in mice 0168 55. Soutschek J, et al (2004) Therapeutic silencing transgenic for HTLV-I. Science 253:1026-28. of an endogenous gene by Systemic administration of 0151. 38. NikiYetal (2001) Macrophage- and neutrophil modified siRNAS. Nature 432:173. dominant arthritis in human IL-1 alpha transgenic mice. J (0169 56. Lechman E. R. et al (2003) The contralateral Clin Invest 107:1127-35. effect conferred by intra-articular adenovirus-mediated 0152 39. Horai R, et al (2000) Development of chronic gene transfer of viral IL-10 is specific to the immunizing inflammatory arthropathy resembling rheumatoid arthritis antigen. Gene Ther 24:2029-35. in interleukin 1 receptor antagonist-deficient mice. J Exp (0170) 57. Harper S Q, et al. (2005) RNA interference Med 191:313-20. improves motor rand neuropathological abnormalities in a 0153. 40. Atsumi T, et al (2002) A point mutation in of Tyr-759 in interleukin 6 family cytokine receptor subunit Huntington's disease mouse model. PNAS 102:5820 gp130 causes autoimmune arthritis. J Exp Med 196:979 (0171 58. Makarov SS, etal (1996) Suppression of experi 90. mental arthritis by gene transfer of interleukin 1 receptor 0154 41. Keffer J, et al (1991) Transgenic mice express antagonist cDNA. Proc Natl Acad Sci USA. 93:402. ing human necrosis factor: A predictive genetic model of (0172 59. Aidinis V, et al (2003) Functional analysis of an arthritis. EMBO J. 10:4025-31. arthritogenic synovial fibroblast. Arthritis Res Ther (O155 42. Kontoyiannis D, et al (1999) Impaired On/off 5:R140-57. regulation of TNF biosynthesis in mice lacking TNF AU (0173 60. Spector I, et al (1989) Latrunculins: novel rich elements: Implications for joint and gut-associated marine macrollides that disrupt microfilament organization immunopathologies. Immunity 10:387-98. and affect cell growth: I. Comparison with Cytochalasin D. 0156 43. Taylor GA, et al (1996) A pathogenetic role of Cell Motill Cytoskeleton 13:127-44. TNF alpha in the syndrome of cachexia, arthritis and (0174 61. Giganti A and Friederich E (2003) The actin autoimmunity resulting from tristetraprolin (TTP) defi cytoskeleton as a therapeutic target: State of the art and ciency. Immunity 4:445-54. future directions. Prog Cell Cycle Res 5:511-25. (O157 44. Amano T, et al (2006) Synoviolin/Hrdl, an E3 (0175 62. Tomasek JJ, et al (2002) Myofibroblasts and ubiquitin ligase, as a novel pathogenic factor for arthropa mechano-regulation of connective tissue remodeling. Nat thy. Genes Dev 17:2436-49. Rev Mol Cell Biol 3:349-63. US 2008/031 1185 A1 Dec. 18, 2008 20

TABLE 1. Summary results of subtractive libraries and microarray hybridizations a. Subtractive libraries Total genes Unigene FANTOM novel

Sample 99% 99.99% 99% 99.99% 99% 99.99% 99% 99.99%

SF 288 23 2648 20* * 14 1 10 2 WJ 653 95 552* 778: 30 3 71 15 b. Oligonucleotide microarray hybridizations Tg197 TNFAARE Common

Sample 90% 95% 90% 95% 90% 959,

SF 277: 98 752 363 95 38* * * WJ 830: 498 1226 636 413 22 78 : a. Summary results of subtractive libraries. Total genes: N of statistically significant deregulated clusters genes; Unigene: N of total genes with a Unigene cluster ID: FAN TOM:N of total genes with a FANTOM accession N. b. Summary results of oligonucleotide microarray hybridizations. Numbers correspond to Affymetrix probesets. a and b. SF: Synovial Fibroblasts, WJ: Whole joints; Percentages indicate statistical speci ficities. *Genes used for the cross-platform comparison presented in Table 2: **Genes presented in Supplementary Table 1: ***Probesets corresponding to the genes presented in Supplementary Table 2:

TABLE 2 Differentially expressed genes in arthritic Whole Joints (WJ) or Synovial fibroblasts (SF) commonly identified by subtractive libraries and oligonucleotide DNA microarray hybridizations. m SQ- m Q- h Q Liter- RT RT, RT Rank R* FC: Genbank Name Description TNFAARE+/- ature QTL, PCR PCR PCR 1 5.2 (4) 11.4 (1) M96827 Hp haptoglobin ------SF 2.9 (2) 2.7 (4) a.a674986 AI451450 expressed sequence -- up AI451450 3 2.8 (1) 2.6 (5) x66402 Mmp3 matrix metallo- -- proteinase 3 4 4.4 (5) 3.2 (3) k02782 C3 complement -- -- component 3 5 3.3 (8) 3.6 (2) X66473 Mmp13 matrix metallo- -- -- proteinase 13 6 5.9 (3) 1.4 (7) z30970 Timp3 issue inhibitor of -- -- metalloproteinase 3 7 3.7 (7) 2.6 (6) x03479 Saa3 serum amyloid A3 -- -- 8 3.7 (6) 1.3 (8) W41762 Rab14 RAB14. (GTPase) -- -- +f- member RAS oncogene family 1 6.7 (1) 0.3 (1) W44091 Eef1a1 elongation factor -- -- SF -alpha down 2 3.3 (4) 0.3 (2) AAO96992 cdc42 cell division -- -- hom cycle 42 homolog 3 6.6 (2) 0.7 (5) M13445 Tuba.1 tubulin. alpha 1 -- +f- 4 2.6 (5) 0.5 (4) D00613 Mglap matrix gamma- -- -- carboxyglutamate (gla) protein 5 2.3 (7) 0.4 (3) L02914 Aqp1 aquaporin 1 ------similar to -glyceraldehyde 3-phosphate 6 4.7 (3) 0.8 (7) M32599 Gapdh. dehydrogenase hom 7 2.6 (6) 0.7 (6) JO4953 GSn Gelsolin ------* 1 9.7 (1) 5.5/2.7 (5) AAO89333 Ctss cathepsin S ------WJ proline-serine- up threonine phosphatase interacting US 2008/031 1185 A1 Dec. 18, 2008 21

TABLE 2-continued Differentially expressed genes in arthritic Whole Joints (WJ) or Synovial fibroblasts (SF) commonly identified by subtractive libraries and oligonucleotide DNA microarray hybridizations. m SQ- m Q- h Q Liter RT RT, RT R* FC: Genbank Name Description TNFAARE+ ature QTL PCR PCR PCR 4.8 (5) 10.8 (2) AAO38079 Pstpip1 protein 1 -- 3.6 (9) 17.7 (1) AA17O3SS Itgax integrin alpha X 4.8 (6) 4.2 (4) x13333 Co14 CD14 antigen ------7.3 (2) 2.1 (15) M96827 Hp haptoglobin -- -- 2.4 (15) 7.3 (3) U18424 Marco macrophage -- -- receptor with collagenous Structure 3.7 (8) 2.7 (11) Ncf1 neutrophil cytosolic actor 1 3.6 (10) 3.4 (10) U72643 List1 eukocyte specific transcript 1 3.6 (12) 2.0/4.9 (9) AAO15331 Cstb 44 (7) 2.1 (14) X80478 Aebp1 : protein 1 11 6.1 (4) 2.0 (18) complement component 1. r Subcomponent tumor necrosis actor receptor Superfamily. member 12 2.4 (17) 3.8 (6) S9378 b 13 2.4 (16) 4.3/2.9/3.5 (7) aa140446 transmembrane 7 : Superfamily member 1 14 7.3 (3) 1.9 (20) wOO755 Timp1 issue inhibitor of metallo proteinase 1 broteasone (prosome. macropain) subunit. beta 15 3.6 (11) 2.5/2.0 (13) I11145 PSmb8: type. 8 Lmp7 16 2.4 (18) 3.5 (8) af)06467 Pitpnm phosphatidyl inositol

17 3.6 (13) 2.0 (17) m89956 LSp1 ymphocyte specific 1 18 2.4 (19) 2.4 (12) u19482 chemokine (C-C motif) igand 9 19 2.6 (14) 2.0/1.8 (19) M86736 Gm granulin -- 2O 2.4 (20) 2.0 (16) aaS18531 Mylc2b Myosin light chain. regulatory B 21 2.4 (21) 1.8 (21) AA154294 Dusp1 dual specificity phosphatase 1 22 2.4 (22) 1.8 (22) XS8861 complement component 1. q Subcomponent. alpha 4.6 (1) 0.3/0.2 (4) W10526 Cacing 1 calcium channel. WJ voltage down dependent. gamma subunit 1 3.5 (3) 0.3 (2) wO8218 Tnint1 troponin T1. skeletal. slow 3.2 (6) 0.3 (3) JO4992 Tnni2 troponin I. skeletal. fast 2 3.5 (2) 0.6 (8) X62940 transforming growth factor beta 1 induced transcript 4 US 2008/031 1185 A1 Dec. 18, 2008 22

3.2 (5) 0.4 (5) x14194 Nid1 nidogen 1 -- 2.8 (9) 0.2 (1) X98O14 Siat86 alpha-2.8-sialyl -- -- transferase. 3.2 (4) 0.7 (9) XO4405 Mb myoglobin ------* 2.8 (7) 0.4/0.4 (7) ABOO1607 Ptgis prostaglandin I2 -- -- (prostacyclin) synthase 2.8 (8) 0.4 (6) AA152678 Bsg basigin -- -- * R refers to the differential expression likehood of any gene to its corresponding library at 99% significance level. FC refers to fold change over control, as identified by the microarray hybridizations of the Tg197 samples at 90% significance level. Multiple entries indicate multiple microarray probesets directed against the same gene. Numbers in parentheses indicate the corresponding ranks. The column Rank lists the average rank. The differential expression was validated with a number of different methods as described in the text. TNF^"; deregulated expression in the knock in mouse model of RA; literature: genes associated with Arthritis upon automated text mining of literature; QTL: genes co-localized with CIA QTLs: (S)Q-RT-PCR: genes confirmed with (Semi) Quantitative RT-PCR in mouse (m) or human (h) samples; +/- indicate agreement disagreement respectively. *denotes statistical significance

TABLE 3 m SQ RT Gene Genbank Description Expression PCR WT SF Hp SF M96827 haptoglobin SF Up Timp3 Z30970 tissue inhibitor of metalloproteinase 3 SF Up N Down Saa3 xO3479 serum amyloid A3 SF Up NA Rab14 W41762 RAB14, (GTPase) member RAS oncogene SF Up family Mmp3 x664O2 matrix metalloproteinase 3 SF Up Up Mmp13 X66473 matrix metalloproteinase 13 SF Up AI451450 aa.674986 expressed sequence AI451450 SF Up NA C3 kO2782 complement component 3 SF Up Up codc42 AAO96992 cell division cycle 42 homolog SF Down y hom Gapdh. M32599 similar to -glyceraldehyde-3-phosphate SF Down NA hom dehydrogenase Mglap DOO613 matrix gamma-carboxyglutamate (gla) SF Down protein Tuba.1 M13445 tubulin. alpha 1 SF Down GSn JO4953 gelsolin SF Down Eefla1 W44091 elongation factor 1-alpha SF Down Down Aqp1 LO2914 aquaporin 1 SF Down Psmb8: I11145 proteasome (prosome. macropain) WU Up Lmp7 subunit. beta type. 8 Lst1 U72643 eukocyte specific transcript 1 Tm7sf1 aa140446 transmembrane 7 Superfamily member 1 NA Tnfrsflb S9378 tumor necrosis factor receptor Superfamily. Up member 1b complement component 1. q Down Subcomponent, alpha Grin M86736 granulin Hp WJ M96827 haptoglobin Timpl wOO755 issue inhibitor of metalloproteinase 1 Up Pstpip1 AAO38079 proline-serine-threonine phosphatase NA interacting protein 1 Pltpnm af)06467 phosphatidylinositol membrane-associated 1 Up Ncf1 L1145S neutrophil cytosolic factor 1 Up Mylc2b aaS18531 Myosin light chain, regulatory B Down Marco u18424 macrophage receptor with collagenous Structure Lsp1 m89956 ymphocyte specific 1 Down Itgax AA17O3SS integrin alpha X NA Dusp1 AA154294 dual specificity phosphatase 1 Up Cstb AAO15331 cystatin B Up C1r aa238O81 complement component 1, r s Subcomponent US 2008/031 1185 A1 Dec. 18, 2008 23

TABLE 3-continued

m SQ RT Gene Genbank Description Expression PCR WT SF

Ccl9: u19482 chemokine (C-C motif) ligand 9 WJ Up Y UP MIP-1y Cd14 x13333 CD14 antigen WJ Up Y CtSS AAO89333 cathepsin S WJ Up NA Aebp1 X80478 AE binding protein 1 WJ Up Y Up Nid1 x14194 nidogen 1 WJ Down Y Slat8e X98O14 alpha-2.8-Sialyltransferase. WJ Down Y Tnint1 wO8218 troponin T1... skeletal. slow WJ Down Y Tnni2 JO4992 troponin I. Skeletal. fast 2 WJ Down Y Tgfbli4 X62940 transforming growth factor beta 1 induced WJ Down Y SF transcript 4 Down Ptgis AB001607 prostaglandin 12 (prostacyclin) synthase WJ Down Mb XO4405 myoglobin WJ Down Cacing 1 W10526 calcium channel, voltage-dependent, WJ Down gamma Subunit 1 Bsg AA152678 basigin WJ Down Y

TABLE 4 Key regulators of actin-cytoskeleton dynamics and organization.

Gene Description Reference Arp2/3 complex Crucial for initiation of actin assembly, nucleation Svitkina et al., 1999 of new filaments and actin polymerisation WASP, WAVE WIP Activates the Arp2/3 complex Rohatzi et al., 1999; Gallego et al., 2005 SCAR Promotes actin nucleation via the Arp2/3 complex Machesky et al., 1999 ERM (Ezrin, A predominant role in intracellular positioning of Gautreau et al., 2002 Radixin, Moesin) actin polymerisation and signalling Zyxin A predominant role in intracellular positioning of Hoffman et al., 2006 actin polymerisation and signalling WASP Promotes F-actin formation and regulates actin Galler et al., 2006 polymerisation Cortactin Stabilises actin filaments, promotes actin Weaver et al., 2001 nucleation Profilin Sequesters actin monomers, releases G-actin Schluter et al., 1997 ATP, promotes actin polymerisation Thymosin Co-operates with profilin to control the pool of un Carlier et al., 1993 polymerised actin PI(4,5)P, Activates actin polymerisation through the N Rohatzi et al., 2001 WASP-Arp2/3 pathway Coronin 1 inhibits Arp2/3 complex at the leading edge and Foger et al., 2006 blocks F-actin accumulation Tropomyosin inhibits F-actin formation through the Arp2/3 Nyakern-Meazza, 2002 COICX Caldesmon inhibits F-actin formation through the Arp2/3 Kordowska et al., 2006 COICX CapZ. (+)end capping protein, inhibits actin Cooper and Schafer, 2000 polymerization Tropomodulin (-)end capping protein, inhibits actin Fujisawa et al., 2001 polymerization ADFCofilin Binds to G-actin-ADP, inhibits actin re Hotulainer et al., 2005 polymerization AIP-1 interacts with cofilin and caps the (+)end of a Rodal et al., 1999 cofilin-severed filament Twinfilin inhibits the assembly of actin monomers Moseley et al., 2006 Vinculin A focal adhesion protein that guides the Arp2/3 DeMali et al., 2002 complex to the cell-substratum RhoA, Rac and RhoA GTPases promote actin polymerisation by Lamarche et al., 1996 Cdc42 activating nucleation by Arp2/3 complex and by inhibiting depolymerisation by ADF/cofilins ROCK A Rho-associated kinase that activates LIM Pawlak et al., 2002 kinase, which dephosphorylates cofilin, induces actin formation FGAP A Rho- and ROCK-regulated GAP for Rac binds Ohta et al., 2006 filamin A to control actin remodelling US 2008/031 1185 A1 Dec. 18, 2008 24

TABLE 4-continued Key regulators of actin-cytoskeleton dynamics and organization.

Gene Description Reference SRF A transcription factor regulating the expression of Schratt et al., 2002 Several actin-cytoskeleton genes inducing actin polymerization Autotaxin Hyrdolyses LPC to LPA, a crucial lipid mediator Song et al., 2005 functioning in cell proliferation, migration and survival SHIP2 Regulates localised changes in PIP3, instructs von Wichert et al., 2003 actin remodelling, membrane ruffling and promotes cell adhesion

TABLE 5

mouse models of Rheumatoid Arthritis

Mouse model Description Reference Induced Collagen-induced arthritis T and B cell immunity to autologous Trentham et al., 1977 (CIA) type II collagen Pristane-induced arthritis Systemic (intra-peritoneal) injection of Elson et al., 1992 pristine Proteoglycan-induced arthritis Systemic injection of BALBic mice with Glant et al., 2003 proteoglycan aggrecan emulsified in DDA adjuvant Streptococcal cell-wall induced intra-articular injection of cell walls from Van den Broek et al., 1988 arthritis Streptococci is pyogenes Immune-complex induced intravenous injection with heat Van Lent et al., 2003 arthritis inactivated polyclonal rabbit anti ysozyme serum, followed by injection with poly-L-lysine-coupled lysozyme in he joint Zymosan-induced arthritis intra-articular injection of Zymosan A Schalkwijk et al., 1985 and development of irritant-induced arthritis Genetic Ncf1 Mutation in Ncfl leading to truncated Hultqvist eta, 2004 protein and lower levels of reactive oxygen species in NADPH oxidase expressing cells Zap70 Altered thymic T-cell selection due to a Sakaguchi et al., 2003 mutation of the ZAP-70 gene causes autoimmune arthritis in mice Fas-related arthropathy A spontaneous rheumatoid arthritis-like Hang et al., 1982 disease in MRLI mice KBXN KRNT-cell receptor transgenic mouse Kouskoffet al., 1996 on the C57BL/6xNOD background HTLV-1 Induction of inflammatory arthropathy Iwakura et al., 1991 resembling rheumatoid arthritis in mice transgenic for HTLV-I hIL-1a Transgenic mice expressing hIL-1 alpha Niki et al., 2001 mRNA in various organs, had high serum levels of hIL-1 alpha, and developed a severe polyarthritic phenotype at 4 weeks of age. IL-1 ra IL-1ra deficient mice, on a BALB.cA Horai et al., 2000 background, develop chronic inflammatory polyarthropathy IL-6 A point mutation of Tyr-759 in Atsumi et al., 2002 interleukin 6 family cytokine receptor Subunit gp130 causes autoimmune arthritis Transgenic mice expressing high levels Keffer et al., 1991 of both soluble and membrane-bound hTNF-C develop chronic inflammatory polyarthritis US 2008/031 1185 A1 Dec. 18, 2008 25

TABLE 5-continued

mouse models of Rheumatoid Arthritis Mouse model Description Reference TNFAARE Deletion of a TNF AUUUA motif results Kontoyiannis et al., 1999 in TNF overexpression and chronic inflammatory arthritis TTP Tristetraprolin-deficient mice develop Taylor et al., 1996 inflammatory arthritis due to de stabilisation of TNF-C. mRNA Synoviolin Hrd 1 Over-expression of human Amano et al., 2006 Synoviolin/Hrdlleads to spontaneous arthropathy

TABLE 6 New mouse models of Rheumatoid arthritis from crosses with Tg197 and involvement of cytoskeletal organization in the pathogenesis of RA (KO. knockout: GT, Gene Trap). Mouse model Source Crossing Phenotype WASP KO Ulrich Walter, Germany Tg197 Amelioration COFILIN 1 KO Walter Witke, Italy Tg197 Exacerbation LST-1 GT SIGTR, Sanger Institute, UK Tg 197 Amelioration Rac1 conditional KO Cord Brakebusch, Germany ColVI-Cre/Tg197 Amelioration. SF specific RhoA conditional KO CordBrakebusch, Germany ColVI-Cre/Tg197 Amelioration. SF specific Cdc42 conditional KO Cord Brakebusch, Germany ColVI-Cre/Tg197 Amelioration. SF specific SRF conditional KO Alfred Nordheim, Germany ColVI-Cre/Tg197 Amelioration. SF specific ATX conditional KO Vasilis Aidinis, Greece ColVI-Cre/Tg197 Amelioration. SF specific

TABLE A m SQ- m Q- h Q Liter- RT RT, RT R* FC: Genbank Name Description TNFAARE'' ature QTL, PCR PCR PCR 2.9 (2) 2.7 (4) a.a674986 AI451450 expressed sequence AI451450 -- SF up 3.3 (4) 0.3 (2) AAO96992 cdc42 hom cell division cycle 42 -- -- SF homolog down 2.6 (5) 0.5 (4) D00613 Mglap matrix gamma-carboxy- -- -- glutamate (gla) protein 4.7 (3) 0.8 (7) M32599 Gapdhhom similar to -glyceraldehyde 3-phosphate dehydrogenase 3.6 (9) 17.7 (1) AA170355 Itgax integrin alpha X -- WJ 2.4 (15) 7.3 (3) U18424 Marco macrophage receptor with -- -- up collagenous structure 3.6 (10) 3.4 (10) U72643 Lst1 leukocyte specific -- transcript 1 4.4 (7) 2.1 (14) X80478 Aebp1 AE binding protein 1 -- 6.1 (4) 2.0 (18) aa238,081 C1r complement component 1. r -- Subcomponent 2.4 (17) 3.8 (6) mS9378 Tnfrsflb tumor necrosis factor -- receptor Superfamily. member 1b 2.4 (16) 4.3/2.9/3.5 (7) aa140446 Tm7sf1 transmembrane 7 Super- -- family member 1 3.6 (13) 2.0 (17) m89956 Lsp1 lymphocyte specific 1 -- 2.4 (20) 2.0 (16) aaS18531 Mylc2b Myosin light chain, -- regulatory B 2.4 (22) 1.8 (22) X58861 C1qa complement component 1. -- q Subcomponent alpha 4.6 (1) 0.3/0.2 (4) W10526 Cacing 1 calcium channel. Voltage- -- WJ dependent, gamma Subunit 1 down 3.5 (3) 0.3 (2) w08218 Tnint1 troponin T1 skeletal. slow -- 3.2 (6) 0.3 (3) JO4992 Tnni2 troponlin I. Skeletal. -- fast 2 3.5 (2) 0.6 (8) X62940 Tgfbli4 transforming growth factor -- beta 1 induced transcript 4 3.2 (5) 0.4 (5) x14194 Nid1 nidogen 1 -- 2.8 (7) 0.4/0.4 (7) ABOO1607 Ptgis prostaglandin I2 -- -- (prostacyclin) synthase US 2008/031 1185 A1 Dec. 18, 2008 26

TABLE B m SQ- m Q- h Q Liter- RT RT, RT R* FC: Genbank Name Description TNFAARE+/- ature QTL, PCR PCR PCR 3.7 (6) 1.3 (8) W41762 Rab14 RAB14. (GTPase) member -- -- +f- SF RAS oncogene family up 2.6 (6) 0.7 (6) JO4953 Gsn Gelsolin -- -- - * SF down 3.6 (11) 2.5/2.0 (13) I11145 PSmb8; proteasome (prosome. -- -- WT Lmp7 macropain) subunit. up beta type. 8 2.4 (18) 3.5 (8) aft)06467 Pitpnm phosphatidylinositol -- -- membrane-associated 2.8 (9) 0.2 (1) X98014 Siat8e alpha-2.8- -- -- WT sialyltransferase. down

TABLE C m SQ- m Q- h Q Liter- RT RT, RT R* FC: Genbank Name Description TNFAARE'' ature QTL, PCR PCR PCR 5.2 (4) 11.4 (1) M96827 Hp haptoglobin ------SF 2.8 (1) 2.6 (5) x66402 Mmp3 matrix metalloproteinase 3 -- up 4.4 (5) 3.2 (3) k02782 C3 complement component 3 -- -- 3.3 (8) 3.6 (2) X66473 Mmp13 matrix metalloproteinase 13 -- -- 5.9 (3) 1.4 (7) z30970 Timp3 tissue inhibitor of -- -- metalloproteinase 3 3.7 (7) 2.6 (6) x03479 Saa3 serum amyloid A3 -- -- 6.7 (1) 0.3 (1) W44091 Eefla1 elongation factor 1-alpha -- -- SF 6.6 (2) 0.7 (5) M13445 Tubal tubulin. alpha 1 +f- down 2.3 (7) 0.4 (3) L02914 Aap1 aquaporin 1 ------9.7 (1) 5.5/2.7 (5) AAO89333 CtSS cathepsin S ------WJ 4.8 (5) 10.8 (2) AAO38079 Pstpip1 proline-serine-threonine -- -- up phosphatase-interacting protein 1 4.8 (6) 4.2 (4) x13333 Cd14 CD14 antigen ------7.3 (2) 2.1 (15) M96827 Hp haptoglobin ------3.7 (8) 2.7 (11) L11455 Nc neutrophil cytosolic ------actor 1 3.6 (12) 2.0/4.9 (9) AAO15331 Cstb cystatin B -- -- 7.3 (3) 1.9 (20) v00755 Timpl issue inhibitor of -- -- metalloproteinase 1 2.4 (19) 2.4 (12) u19482 Ccl9; chemokine (C-C motif) -- -- MIP-1, ligand 9 2.6 (14) 2.0/1.8 (19) M86736 Grin granulin -- -- 3.2 (4) 0.7 (9) XO4405 Mb myoglobin -- -- * WJ 2.8 (8) 0.4 (6) AA152678 Bsg basigin -- -- +* down

REFERENCES 0182 7. Aidinis V, Plows D. Haralambous S. Armaka M, (0176 1. Choy E. H. Panayi GS (2001) Cytokine pathways Papadopoulos P. et al. (2003) Functional analysis of an and joint inflammation in rheumatoid arthritis. N Engl J arthritogenic synovial fibroblast. Arthritis Res Ther 5: Med 344: 907-916. R140-157. (0177 2. Smolen JS, Steiner G (2003) Therapeutic strate 0183 8. Kontoyiannis D. Pasparakis M. Pizarro TT, Com gies for rheumatoid arthritis. Nat Rev Drug Discov 2: 473 inelli F. Kollias G. (1999) Impaired on/off regulation of 488. TNF biosynthesis in mice lacking TNF AU-rich elements: (0178. 3. Firestein GS (2003) Evolving concepts of rheu implications for joint and gut-associated immunopatholo matoid arthritis. Nature 423: 356-361. gies. Immunity 10:387-398. 0179 4. Lindqvist A KB, Bockermann R. Johanson AC 0.184 9. Brazma A, Hingamp P. Quackenbush J. Sherlock M, Nandakumar KS, Johannensson M, et al. (2002) Mouse G, Spellman P. et al. (2001) Minimum information about a models for rheumatoid arthritis. Trends Genet. 18: S7-S13. microarray experiment (MIAME)-toward standards for 0180 5. Keffer J. Probert L. Cazlaris H, Georgopoulos S, microarray data. Nat Genet. 29: 365-371. Kaslaris E. et al. (1991) Transgenic mice expressing 0185. 10. Brazma A, Parkinson H, Sarkans U, Shojatalab human tumour necrosis factor: a predictive genetic model M. Vilo J, et al. (2003) ArrayExpress—a public repository of arthritis. Embo J 10: 4025-4031. for microarray gene expression data at the EBI. Nucleic 0181 6. Li P. Schwarz E M (2003) The TNF-alpha trans Acids Res 31: 68-71. genic mouse model of inflammatory arthritis. Springer 0186 11. Pavelka N, Pelizzola M, Vizzardelli C, Capoz Semin Immunopathol 25: 19-33. Zoli M, SplendianiA, etal. (2004) A power law global error US 2008/031 1185 A1 Dec. 18, 2008 27

model for the identification of differentially expressed (0202) 27. von Muhlen CA, Tan E M (1995) Autoantibod genes in microarray data. BMC Bioinformatics 5: 203. ies in the diagnosis of systemic rheumatic diseases. Semin 0187. 12. Irizarry R A, Warren D, Spencer F, Kim I F, Arthritis Rheum 24:323-358. Biswal S, et al. (2005) Multiple-laboratory comparison of 0203. 28. Van Boekel M A. Vossenaar E. R. van den Hoo microarray platforms. Nat Methods 2: 345-350. gen F H. van Venrooij W J (2002) Autoantibody systems in 0188 13. Marshall E (2004) Getting the noise out of gene rheumatoid arthritis: specificity, sensitivity and diagnostic arrays. Science 306: 630-631. value. Arthritis Res 4: 87-93. (0189 14. Trentham D. E., Townes AS, Kang A H (1977) 0204 29. Presland RB, Kuechle M K, Lewis SP. Fleck Autoimmunity to type II collagen an experimental model man P. Dale BA (2001) Regulated expression of human of arthritis. J Exp Med 146: 857-868. filaggrin in keratinocytes results in cytoskeletal disrup 0190. 15. Neumann E. Kullmann F. Judex M. Justen H P. tion, loss of cell-cell adhesion, and cell cycle arrest. Exp Wessinghage D. etal. (2002) Identification of differentially Cell Res 270: 199-213. expressed genes in rheumatoid arthritis by a combination (0205 30. Vossenaar ER, van Venrooij W J (2004) Citrul of complementary DNA array and RNA arbitrarily primed linated proteins: Sparks that may ignite the fire in rheuma polymerase chain reaction. Arthritis Rheum 46: 52-63. toid arthritis. Arthritis Res Ther 6: 107-111. (0191) 16. van der Pouw Kraan T C, van Gaalen F A, (0206. 31. Pawlak G, Helfman D M (2001) Cytoskeletal Kasperkovitz PV, Verbeet N L, Smeets TJ, et al. (2003) changes in cell transformation and tumorigenesis. Curr Rheumatoid arthritis is a heterogeneous disease: evidence Opin Genet Dev 11: 41-47. for differences in the activation of the STAT-1 pathway 0207. 32. Firestein G. S (1996) Invasive fibroblast-like between rheumatoid tissues. Arthritis Rheum 48: 2132 synoviocytes in rheumatoid arthritis. Passive responders or 2145. transformed aggressors? Arthritis Rheum 39: 1781-1790. 0208. 33. dos Remedios C G, Chhabra D. Kekic M, (0192 17. Ashburner M, Ball CA, Blake J. A. Botstein D, Dedova IV. Tsubakihara M. et al. (2003) Actin binding Butler H. et al. (2000) Gene ontology: tool for the unifica proteins: regulation of cytoskeletal microfilaments. tion of biology. The Gene Ontology Consortium. Nat Physiol Rev 83: 433-473. Genet. 25: 25-29. (0209 34. McGough AM, Staiger CJ, Min J. K. Simonetti (0193 18. Witke W. Sharpe A H, Hartwig J H, AZuma T. KD (2003) The gelsolin family of actin regulatory pro Stossel T. P. et al. (1995) Hemostatic, inflammatory, and teins: modular structures, versatile functions. FEBS Lett fibroblast responses are blunted in mice lacking gelsolin. 552: 75-81. Cell 81: 41-51. 0210 35. Vandekerckhove J. Bauw G, Vancompernolle K, (0194 19. Olofsson P. Holmberg J, Tordsson J, Lu S, Aker Honore B, Celis J (1990) Comparative two-dimensional strom B, et al. (2003) Positional identification of Ncfl as a gel analysis and microsequencing identifies gelsolinas one gene that regulates arthritis severity in rats. Nat Genet. 33: of the most prominent downregulated markers of trans 25-32. formed human fibroblast and epithelial cells. J Cell Biol 0.195. 20. Saadoun S. Papadopoulos MC, Hara-Chikuma 111: 95-102. M. Verkman A S (2005) Impairment of angiogenesis and 0211 36. Mullauer L, Fujita H, Ishizaki A, Kuzumaki N cell migration by targeted aquaporin-1 gene disruption. (1993) Tumor-suppressive function of mutated gelsolin in Nature 434: 786-792. ras-transformed cells. Oncogene 8: 2531-2536. (0196. 21. Gabbiani G (2003) The myofibroblast in wound 0212. 37. Asch HL, Head K, Dong Y. Natoli F. Winston J healing and fibrocontractive diseases. J Pathol 200: 500 S. et al. (1996) Widespread loss of gelsolin in breast can 503. cers of humans, mice, and rats. Cancer Res 56:4841-4845. (0197) 22. Tomasek J.J. Gabbiani G, Hinz B, Chaponnier C, 0213 38. Tanaka M, Mullauer L. Ogiso Y. Fujita H, Brown RA (2002) Myofibroblasts and mechano-regula Moriya S, et al. (1995) Gelsolin: a candidate for suppressor tion of connective tissue remodelling. Nat Rev Mol Cell of human bladder cancer. Cancer Res 55: 3228-3232. Biol 3: 349-363. 0214 39. Dosaka-Akita H. Hommura F. Fujita H, (0198 23. Mattey D L, Dawes PT. Nixon N B, Slater H Kinoshita I, Nishi M. etal. (1998) Frequent loss of gelsolin (1997) Transforming growth factor beta 1 and interleukin 4 expression in non-Small cell lung cancers of heavy Smok induced alpha Smooth muscle actin expression and myofi erS. Cancer Res 58: 322-327. broblast-like differentiation in human synovial fibroblasts 0215 40. Aidinis V. Plows, D., Haralambous, S., Armaka, in vitro: modulation by basic fibroblast growth factor. Ann M. Papadopoulos, P., Kanaki, M-Z. Koczan, D. Thiesen, Rheum Dis 56: 426-431. H. J. and Kollias, G. (2002) Differential gene expression (0199 24. Gumbiner B M (1996) Cell adhesion: the analysis of an arthritic synovial fibroblast reveals its func molecular basis of tissue architecture and morphogenesis. tional properties. Submitted. Ce1184; 345-357. 0216 41. Chomczynski P. Sacchi N (1987) Single-step (0200. 25. Kasperkovitz P V. Timmer TC, Smeets T J, method of RNA isolation by acid guanidinium thiocyan Verbeet N L, Tak P Petal. (2005) Fibroblast-like synovio ate-phenol-chloroform extraction. Anal Biochem 162: cytes derived from patients with rheumatoid arthritis show 156-159. the imprint of synovial tissue heterogeneity: evidence of a 0217 42. Carninci P. Hayashizaki Y (1999) High-effi link between an increased myofibroblast-like phenotype ciency full-length cDNA cloning. Methods Enzymol 303: and high-inflammation synovitis. Arthritis Rheum 52:430 19-44. 441. 0218 43. Shibata Y. Carninci P. Watahiki A, Shiraki T, 0201 26. Shrivastav M. Mittal B. Aggarwal A, Misra R Konno H. et al. (2001) Cloning full-length, cap-trapper (2002) Autoantibodies against cytoskeletal proteins in selected cDNAS by using the single-strand linker ligation rheumatoid arthritis. Clin Rheumatol 21: 505-510. method. Biotechniques 30: 1250-1254. US 2008/031 1185 A1 Dec. 18, 2008 28

0219) 44. Carninci P. Shibata Y. Hayatsu N, Itoh M, 0229. 54. Kawai J, Shinagawa A, Shibata K. Yoshino M, Shiraki T, etal. (2001) Balanced-size and long-size cloning Itoh M. et al. (2001) Functional annotation of a full-length of full-length, cap-trapped cDNAs into vectors of the novel mouse cDNA collection. Nature 409: 685-690. lambda-FLC family allows enhanced gene discovery rate 0230 55. Stekel DJ, Git Y. Falciani F (2000) The com and functional analysis. Genomics 77: 79-90. parison of gene expression from multiple cDNA libraries. 0220 45. Hirozane-Kishikawa T. Shiraki T. Waki K, Genome Res 10: 2055-2061. 0231 56. Rose C, Smith MD (2002) Mathematical Sta Nakamura M, Arakawa T. et al. (2003) Subtraction of cap tistics with Mathematica; Verlag S, editor: Springer Verlag. trapped full-length cDNA libraries to select rare tran 0232 57. Ihaka R. Gentleman R (1996) A Language for scripts. Biotechniques 35: 510-516, 518. Data Analysis and Graphics. Journal of Computational and 0221) 46. Shibata K, Itoh M. Aizawa K, Nagaoka S. Sasaki Graphical Statistics 5: 299-314. N, et al. (2000) RIKEN integrated sequence analysis 0233 58. Yuen T. Wurmbach E. Pfeffer RL, Ebersole BJ, (RISA) system 384-format sequencing pipeline with 384 Sealfon SC (2002) Accuracy and calibration of commer multicapillary sequencer. Genome Res 10: 1757-1771. cial oligonucleotide and custom cDNA microarrays. 0222 47. Carninci P. Waki K, Shiraki T, Konno H, Shibata Nucleic Acids Res 30: e48. K, et al. (2003) Targeting a complex transcriptome: the 0234 59. Tsien C L. Libermann T.A., Gu X, Kohane I S construction of the mouse full-length cDNA encyclopedia. (2001) On reporting fold differences. Pac Symp Biocom Genome Res 13: 1273-1289. put: 496-507. 0223 48. Konno H, Fukunishi Y., Shibata K, Itoh M. Carn 0235 60. TuY. Stolovitzky G, Klein U (2002) Quantitative inci P. etal. (2001) Computer-based methods for the mouse noise analysis for gene expression microarray experi full-length cDNA encyclopedia: real-time sequence clus ments. Proc Natl Acad Sci USA99: 14031-14036. tering for construction of a nonredundant cDNA library. 0236 61. Mutch DM, Berger A. Mansourian R, Rytz A, Genome Res 11: 281-289. Roberts MA (2002) The limit fold change model: a prac 0224. 49. Miller RT, Christoffels AG, Gopalakrishnan C, tical approach for selecting differentially expressed genes Burke J. Ptitsyn A A. et al. (1999) A comprehensive from microarray data. BMC Bioinformatics 3:17. approach to clustering of expressed human gene sequence: 0237 62. Fischer G, Ibrahim S M, Brockmann G. A. the sequence tag alignment and consensus knowledge Pahnke J. Bartocci E, et al. (2003) Expression view: visu base. Genome Res 9: 1143-1155. alization of quantitative trait loci and gene-expression data 0225 50. Burke J, Davison D, Hide W. (1999) d2 cluster: in Ensembl. Genome Biol 4: R77. Epub 2003 October a validated method for clustering EST and full-length 2010. cDNAsequences. Genome Res 9: 1135-1142. 0238 63. Serrano-Fernandez P. Ibrahim, S.M., Godde, R., 0226 51. Altschul S F, Gish W. Miller W. Myers E. W. Epplen, J. & Moller, S. (2004) Intergenomic Consensus in Lipman DJ (1990) Basic local alignment search tool. JMol Multifactorial Inheritance Loci: the Case of Multiple Scle Biol 215: 403-410. rosis. Genes and Immunity 5: 615-620. 0227 52. Altschul S F, Madden T L. Schaffer AA, Zhang 0239) 64. Tavazoie S, Hughes JD, Campbell MJ, Cho RJ, J, Zhang Z. et al. (1997) Gapped BLAST and PSI-BLAST: Church G M (1999) Systematic determination of genetic a new generation of protein database search programs. network architecture. Nat Genet. 22: 281-285. Nucleic Acids Res 25: 3389-3402. 0240 65. Mould AW, Tonks I D, Cahill MM, Pettit AR, 0228 53. Okazaki Y. Furuno M. Kasukawa T. Adachi J. Thomas R, et al. (2003) Vegfb gene knockout mice display Bono H. et al. (2002) Analysis of the mouse transcriptome reduced pathology and synovial angiogenesis in both anti based on functional annotation of 60,770 full-length gen-induced and collagen-induced models of arthritis. cDNAS. Nature 420:563-573. Arthritis Rheum 48: 2660-2669.

SEQUENCE LISTING

<16 Oc NUMBER OF SEO ID NOS: 66

<210 SEQ ID NO 1 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer sequence Agp1 (sense)

<4 OO SEQUENCE: 1 to accc.gcaa cittct caaac

<210 SEQ ID NO 2 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Aqp1 (antisense) US 2008/031 1185 A1 Dec. 18, 2008 29

- Continued

<4 OO SEQUENCE: 2 agct Ctgaga C caggaaa.ca

<210 SEQ ID NO 3 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Bsg (sense) <4 OO SEQUENCE: 3 atgaga agag gcggaa.gc.ca

<210 SEQ ID NO 4 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer - Bsg (antisense) <4 OO SEQUENCE: 4 c cactic caca gggctgtagt

<210 SEQ ID NO 5 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer CD14 (sense) <4 OO SEQUENCE: 5 Caatcc taa ttgggcgaga

<210 SEQ ID NO 6 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer CD14 (antisense) <4 OO SEQUENCE: 6 cgagtgggat t cagagtc.ca

<210 SEQ ID NO 7 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer CDC42 (sense) <4 OO SEQUENCE: 7 aagtggcc.ca gatcctggaa

<210 SEQ ID NO 8 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer CDC42 (antisense) <4 OO SEQUENCE: 8 agcactgcac ttittggggtt US 2008/031 1185 A1 Dec. 18, 2008 30

- Continued

<210 SEQ ID NO 9 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Gsn (sense) <4 OO SEQUENCE: 9 tgcaggaaga Cctggctact

<210 SEQ ID NO 10 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Gsn (antisense) <4 OO SEQUENCE: 10 atggcttggit cottacticag

<210 SEQ ID NO 11 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Hp (sense) <4 OO SEQUENCE: 11 gaagcaatgg gtgaacacag

<210 SEQ ID NO 12 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Hp (antisense) <4 OO SEQUENCE: 12 ggggtggaga acgaccttct

<210 SEQ ID NO 13 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Marco (sense) <4 OO SEQUENCE: 13 Cacaggaatt Caaggacaga

<210 SEQ ID NO 14 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Marco (antisense) <4 OO SEQUENCE: 14 attgtc.ca.gc cagatgttcc

<210 SEQ ID NO 15 <211 LENGTH: 2O US 2008/031 1185 A1 Dec. 18, 2008 31

- Continued

&212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Miglap (sense)

<4 OO SEQUENCE: 15 cagt coctitc atcaac agga

<210 SEQ ID NO 16 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Miglap (antisense) <4 OO SEQUENCE: 16 Ctgcaggaga tataaaacga

<210 SEQ ID NO 17 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Mb (sense) <4 OO SEQUENCE: 17 t cacacgc.ca c caagcacaa

<210 SEQ ID NO 18 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Mb (antisense) <4 OO SEQUENCE: 18 tgggct coag ggtaac actg

<210 SEQ ID NO 19 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Ptgis (sense) <4 OO SEQUENCE: 19 t cacagatga ccacactic cc

<210 SEQ ID NO 2 O <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Ptgis (antisense) <4 OO SEQUENCE: 2O gcagtaggac gacaaattgt

<210 SEQ ID NO 21 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer B2M (sense) US 2008/031 1185 A1 Dec. 18, 2008 32

- Continued

<4 OO SEQUENCE: 21 ttctggtgct tdt ct cactg a 21

<210 SEQ ID NO 22 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer B2M (antisense) <4 OO SEQUENCE: 22 cagtatgttc ggctt.cccat t c 22

<210 SEQ ID NO 23 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer L32 (sense) <4 OO SEQUENCE: 23 ttaa.gcgaala Ctggcggalaa C 21

<210 SEQ ID NO 24 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer L32 (antisense) <4 OO SEQUENCE: 24 ttgttgct co cataac cqat g 21

<210 SEQ ID NO 25 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer mouse Aqp1 <4 OO SEQUENCE: 25 t caccc.gcaa citt citcaaac

<210 SEQ ID NO 26 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer mouse Aqp1 (antisense) <4 OO SEQUENCE: 26 t catgcggtc. tctgaagt cq

<210 SEQ ID NO 27 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer Mouse GSn (sense) <4 OO SEQUENCE: 27 tgcaggaaga Cctggctact US 2008/031 1185 A1 Dec. 18, 2008 33

- Continued

<210 SEQ ID NO 28 <211 LENGTH: &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: &223> OTHER INFORMATION: Synthetic Primer mouse GSn (antisense)

<4 OO SEQUENCE: 28 tcgatgtacc gct tag caga

<210 SEQ ID NO 29 <211 LENGTH: 19 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: &223> OTHER INFORMATION: Synthetic Primer human Aqp1 (sense)

<4 OO SEQUENCE: 29

CtcCctgact gggaacticg 19

<210 SEQ ID NO 3 O <211 LENGTH: 19 &212> TYPE: DNA <213> ORGANISM: Artificial sequence &220s FEATURE: &223> OTHER INFORMATION: Synthetic Primer human Agg1.

<4 OO SEQUENCE: 3 O gggctacaga gaggc.cgat 19

<210 SEQ ID NO 31 <211 LENGTH: &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: &223> OTHER INFORMATION: Synthetic Primer human Bsg

<4 OO SEQUENCE: 31 titcCtgggca t cqtggctga

<210 SEQ ID NO 32 <211 LENGTH: &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: &223> OTHER INFORMATION: Synthetic Primer human Bsg (antisense)

<4 OO SEQUENCE: 32 gcggacgttc ttgcctttgt

<210 SEQ ID NO 33 <211 LENGTH: 19 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: &223> OTHER INFORMATION: Synthetic Primer human Ca14 (sense)

<4 OO SEQUENCE: 33 cggcggtc.t.c aacctagag 19

<210 SEQ ID NO 34 <211 LENGTH: US 2008/031 1185 A1 Dec. 18, 2008 34

- Continued

TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Synthetic Primer human Ca14 (antisense) SEQUENCE: 34 gcctaccagt agctgagcag

SEO ID NO 35 LENGTH: 2O TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Synthetic Primer human cac42 (sense) SEQUENCE: 35 aattgat citc agagatgacc

SEQ ID NO 36 LENGTH: 2O TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Synthetic Primer human cac42 (antisense) SEQUENCE: 36 tittaggcctt totgtgtaag

SEO ID NO 37 LENGTH: 21 TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Synthetic Primer human Gsn (sense) SEQUENCE: 37 ggtgtggCat Caggattcaa g 21

SEQ ID NO 38 LENGTH: 22 TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Synthetic Primer human Gsn (antisense) SEQUENCE: 38 titt cataccg attgctgttg ga 22

SEO ID NO 39 LENGTH 19 TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Synthetic Primer human Marco (sense) SEQUENCE: 39 tgggacgaga tiggagcaac 19

SEQ ID NO 4 O LENGTH: 22 TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Synthetic Primer human Marco (antisense) US 2008/031 1185 A1 Dec. 18, 2008 35

- Continued

<4 OO SEQUENCE: 40 c cct tagttc cagttt cocc tt 22

<210 SEQ ID NO 41 <211 LENGTH: 19 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human Mb (sense) <4 OO SEQUENCE: 41 ttggtgctga acgtctggg 19

<210 SEQ ID NO 42 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human Mb (antisense) <4 OO SEQUENCE: 42

CtgtgcCagg ggcttaat ct c 21

<210 SEQ ID NO 43 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human B2M (sense) <4 OO SEQUENCE: 43 Ctgaaaaaga tigagtatgcc

<210 SEQ ID NO 44 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human B2M (antisense) <4 OO SEQUENCE: 44 accctacatt ttgttgcataa

<210 SEQ ID NO 45 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human L32 (sense) <4 OO SEQUENCE: 45 ttaa.gc.gtaa Ctggcggalaa C 21

<210 SEQ ID NO 46 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human L32 <4 OO SEQUENCE: 46 gagggat.ct c gCacagtaa US 2008/031 1185 A1 Dec. 18, 2008 36

- Continued

<210 SEQ ID NO 47 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer mouse Cofilin 1 (sense) <4 OO SEQUENCE: 47 tgggtgaaac totgggagat 2O

<210 SEQ ID NO 48 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer mouse Cofilin 1 (antisense) <4 OO SEQUENCE: 48 c cat agtgga agcctgatgc 2O

<210 SEQ ID NO 49 <211 LENGTH: 19 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer mouse VASP (sense) <4 OO SEQUENCE: 49 Cttgccaagg atgaagttcg 19

<210 SEQ ID NO 50 <211 LENGTH: 18 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer mouse VASP <4 OO SEQUENCE: 5 O gctt caccct citccaagt 18

<210 SEQ ID NO 51 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human C3 (sense 5'-3') <4 OO SEQUENCE: 51 ggcc tttgtt ct catctogc t 21

<210 SEQ ID NO 52 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human C3 (antisense 5'-3') <4 OO SEQUENCE: 52 ataggatgtg gccticcacgt t 21

<210 SEQ ID NO 53 <211 LENGTH: 22 US 2008/031 1185 A1 Dec. 18, 2008 37

- Continued

&212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic primer human CSTB (sense 5'-3') <4 OO SEQUENCE: 53 Cctgttgttta aggc.cgtgtc. at 22

<210 SEQ ID NO 54 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human CSTB (antisense 5'-3') <4 OO SEQUENCE: 54 ggagagattg galacacticgc ag 22

<210 SEQ ID NO 55 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human DUSP1 (sense 5'-3') <4 OO SEQUENCE: 55 acaaggcaga cat cagct co to 22

<210 SEQ ID NO 56 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequenc &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human DUSP1 (antisense 5'-3') <4 OO SEQUENCE: 56 talagcaaggc agatggtggc t 21

<210 SEQ ID NO 57 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human LSP1 (sense 5'-3') <4 OO SEQUENCE: 57 caacaattaa gag caccc.ca to 22

<210 SEQ ID NO 58 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human LSP1 (antisense 5'-3') <4 OO SEQUENCE: 58 tggcaa.cagg aagcaact tc t 21

<210 SEQ ID NO 59 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human RAB14 US 2008/031 1185 A1 Dec. 18, 2008 38

- Continued

<4 OO SEQUENCE: 59 gctt attgtt cct cqaag.cg ag 22

<210 SEQ ID NO 60 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human RAB14 (antisense 5'-3') <4 OO SEQUENCE: 60 atccaa.gctt coat cotgaa td 22

<210 SEQ ID NO 61 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human SHIP2 <4 OO SEQUENCE: 61 ggct cqgagg aaccaaaact ac 22

<210 SEQ ID NO 62 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human SHIP2 (sense 5'-3') <4 OO SEQUENCE: 62 tgaggit cocc aaaccagaag a 21

<210 SEQ ID NO 63 <211 LENGTH: 22 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human TGFB1I4 (sense 5'-3') <4 OO SEQUENCE: 63 tagctic ct ct ggtgcaagtg td 22

<210 SEQ ID NO 64 <211 LENGTH: 21 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human TGFB1I4 (antisense 5'-3")

<4 OO SEQUENCE: 64 tccactitctt citctgaccgc a 21

<210 SEQ ID NO 65 <211 LENGTH: 19 &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human LST1 (sense 5'-3') <4 OO SEQUENCE: 65 US 2008/031 1185 A1 Dec. 18, 2008 39

- Continued cgtgatgagg aacttgagg 19

<210 SEQ ID NO 66 <211 LENGTH: 2O &212> TYPE: DNA <213> ORGANISM: Artificial Sequence &220s FEATURE: <223> OTHER INFORMATION: Synthetic Primer human LST1 (antisense 5'-3") <4 OO SEQUENCE: 66 gcgataac at tagg to aggg

1. A method for determining whether a compound, a (poly) in table A, B, C, or table 4, wherein said gene is identified in peptide or a nucleic acid molecule or functional equivalent either of said tables by expression SF Up or WJ Up or said thereof for altering cytoskeleton or its organization in a cell, gene is VASP. affects a phenotype of synovial fibroblast cells of an indi 9. A method according to claim 1, wherein said phenotype vidual or non-human animal Suffering from inflammatory comprises adhesion, migration, proliferation, wound-healing arthritis or at risk of Suffering thereof, comprising culturing and/or apoptosis of said synovial fibroblast cells. synovial fibroblast cells in vitro from a sample obtained from 10. A method according to claim 1, wherein said phenotype comprises stress fiber formation. an affected joint or a joint at risk of being affected from said 11. A method according to claim 1, in a high-throughput individual or non-human animal, in the presence of said com Setting. pound and/or providing said synovial fibroblast cells with 12. A method according to claim 1, wherein said nucleic said (poly)peptide or nucleic acid or said functional equiva acid codes for a protein produced by a gene of table A, B, C, lent thereof and determining whether a phenotype of said 3 or 4, wherein said gene is identified in either of said tables cultured synovial fibroblast cells is altered. by SF down or WJ down or said gene is gelsolin or ADF/ 2. A method for determining whether a compound, a (poly) Cofilin. peptide or a nucleic acid molecule or functional equivalent 13. Use of a gene delivery vehicle comprising the coding thereof affects cytoskeleton or its organization of synovial region of a gene of table A, B or C identified by SF down or fibroblast cells of an individual or non-human animal suffer WJ down, a gene of table 4, Gelsolin or ADF/Cofilin, for ing from inflammatory arthritis or at risk of suffering thereof, delivering said gene to cells of a joint. comprising culturing synovial fibroblast cells in vitro from a 14. Use of a gene delivery vehicle according to claim 13, sample obtained from an affected joint or a joint at risk of for delivering said gene to a synovial fibroblast cell, or a cell being affected from said individual or non-human animal, in of a whole joint. the presence of said compound and/or providing said synovial 15. Use of a gene delivery vehicle comprising the coding fibroblast cells with said nucleic acid or said functional region of a gene of table A, B or C identified by SF down or equivalent thereof and determining whether cytoskeleton or WJ down, Gelsolin or Cofilin, for the preparation of a com its organization of said cultured synovial fibroblast cells is position for delivering said gene to cells of a synovial fibro altered. blast cell (SFUP) and/or a joint (WJUP). 3. A method according to claim 1, wherein said synovial 16. Use according to claim 15, for the treatment of inflam fibroblast cells are obtained from a joint of a mouse compris matory arthritis. ing the genetic alteration of the Tg 197 mouse, a non-human 17. Use according to claim 67, wherein said gene is Gelso animal according to any one of claims 44-56 and/or a genetic lin, ADF/Cofilin. model as depicted in table 5. 18. Use of a gene delivery vehicle comprising a nucleic 4. A method according to claim 1, wherein said compound acid of a functional equivalent thereof comprising an anti is effective in cancer therapy. sense-sequence for a gene of table A, B or C identified by SF UP or WJUP or VASP, or comprising an expression cassette 5. A method according to claim 1, wherein said compound therefore, for delivering said nucleic acid or functional is one of the following: Cytochalasin D. Phalloidin, Jasplaki equivalent thereof to a synovial fibroblast cell (SFUP) or to nolide, Chondramides, Dolostatin 11, Latrunculin A, Swin cells of a joint (WJ LTP). holide A, Misakinolide A, Mycalolide B, Shinxolide C. Scy 19. Use according to claim 18, wherein said antisense tophycins, Goniodomin A, Pseudopterolide, sequence comprises an shRNA, RNAi and/or siRNA. Lysophosphatidylcholine, Lysophosphatidyl acid. 20. Use according to claim 13, wherein said gene delivery 6. A method according to claim 1, wherein said nucleic acid vehicle comprises a liposome, or a viral vector. comprises an antisense-Sequence, preferably an RNAi, an 21. Use according to claim 20, wherein said viral vector is siRNA, oran shRNA. an adenovirus, an adeno-associated virus vector or a lentivi 7. A method according to claim 1, wherein said nucleic acid ruS VectOr. or functional equivalent thereof comprises a morpholino, 22. Use according to claim 19, wherein said gene delivery locked nucleic acid (LNA) or a peptide nucleic acid (PNA). vehicle is administered systemically. 8. A method according to claim 6 or claim 7, wherein said 23. Use according to claim 19, wherein said gene delivery antisense-Sequence is an antisense sequence of a gene listed vehicle is administered intrarticular. US 2008/031 1185 A1 Dec. 18, 2008 40

24. Use according to claim 13, wherein said expression holide A, Misakinolide A, Mycalolide B, Shinxolide C. Scy cassette is iven by a pol II promoter or a pol III promoter. tophycins, Goniodomin A, Pseudopterolide, 25. A method for determining whether a sample is derived Lysophosphatidylcholine, Lysophosphatidyl acid. from an individual suffering from inflammatory arthritis or 44. A non-human animal wherein at least one of the genes from an individual at risk of suffering therefrom, said method listed in table A, table B, or table 4 is knocked-out, over comprising measuring in said sample the level of expression expressed or under-expressed. of at least one gene product of a gene of table A, table B, or 45. A non-human animal wherein at least one of the genes table 4 and comparing said expression level with a reference listed in table A, or table 4 is knocked-out, over-expressed or value. under-expressed. 26. A method according to claim 25, wherein said gene 46. A non-human animal wherein at least one of the genes product comprises RNA. listed in table A, table B, or table 4 is knocked out, over 27. A method according to claim 25, wherein said sample expressed or under-expressed and wherein at least one of the is derived from a joint of said individual. genes listed in table C or table 6 is knocked-out, over-ex 28. A method according to claim 25, wherein said sample pressed or under-expressed. comprises cells of synovial membrane or synovial fluid. 47. A non-human animal wherein at least one of the genes 29. A method according to claim 25, wherein said measur listed in table A, or table 4 is knocked out, over-expressed or ing is in cells of said sample. under-expressed and wherein at least one of the genes listed in 30. A method according to claim 25, wherein at least two table B, in table Cortable 6 is knocked-out, over-expressed or gene products are measured. under-expressed. 31. A method according to claim 25, wherein at least one gene product is of a gene of table A, and/or table 4. 48. A non-human animal wherein at least one gene known 32. A method according to claim 25, wherein at least one to be involved in cytoskeleton rearrangement is knocked-out, gene product is of a gene of table A, table B, and/or table 4 and over-expressed or under-expressed. at least one other gene product is of a gene of table C. 49. A non-human animal according to claim 48, wherein 33. A method according to claim 25, wherein at least one said cytoskeleton arrangement is in a synovial fibroblast. gene product is of a gene of table A, or table 4 and at least one 50. A non-human animal according to claim 48, wherein other gene product is of a gene of table B or C. said gene is involved in extracellular or intracellular signaling 34. A method for determining whether a sample is derived to the cytoskeleton. from an individual suffering from inflammatory arthritis or 51. A non-human animal according to claim 48, wherein from an individual at risk of suffering thereof, said method said gene is involved in binding and/or rearrangement of the comprising measuring in cells from said sample the level of actin cytoskeleton. expression of at least one gene product of a gene involved in 52. A non-human animal according to claim 48, wherein cytoskeleton arrangement and comparing said expression said gene is any one of the following: tubulin alpha 1, RAB14, level with a reference value. Gelsolin, phosphatidylinositol membrane-associated, matrix 35. A method according to claim 34, wherein said gene is metalloproteinase 3, matrix metalloproteinase 13, tissue involved in cytoskeleton arrangementina Synovial fibroblast. inhibitor of matrix metalloproteinase 13, cathepsin S, matrix 36. A method according to claim 34, wherein said gene is gamma-carboxylglutamate, integrin alpha X, myosin light involved in extracellular or intracellular signaling to the chain, troponin T I skeletal slow, troponin I skeletal fast 2, cytoskeleton. VASP, ADF/Cofilin and/or is a gene of table 4. 37. A method according to claim 34, wherein said gene is 53. A non-human animal according to claim 44, wherein involved in binding and/or rearrangement of actin cytoskel said animal is transgenic for said gene or homologue thereof etOn. that is over-expressed or under-expressed. 38. A method according to claim 34, wherein said gene is 54. A non-human animal according to claim 44, wherein any one of the following: tubulin alpha 1, RAB14, Gelsolin, said animal is a mouse or a rat. phosphatidylinositol membrane-associated, matrix rnetallo 55. A non-human animal according to claim 44, wherein proteinase 3, matrix metalloproteinase 13, tissue inhibitor of said at least one of the genes listed in table A, table B, table 3, matrix metalloproteinase 13, cathepsin S. matrix gamma table 4 or table 6 is knocked-out, over-expressed or under carboxylglutamate, integrin alpha X, myosin light chain, expressed, in a genetic model for arthritis. troponin T1 skeletal slow, troponin I skeletal fast 2, VASP. 56. A non-human animal according to claim 55, wherein ADF/Cofilin and/or is a gene of table 4. said genetic model 10 comprises the specific genetic modifi 39. A method according to claim 25, wherein said inflam matory arthritis is rheumatoid arthritis. cation of a Tg 197 mouse ora of genetic model as depicted in 40. Use of a compound for altering cytoskeleton or its table 5. organization in a cell for the manufacture of a medicament for 57. Use of a non-human animal according to claim 44, as a alleviating at least one symptom of inflammatory arthritis in disease model. an individual. 58. Use according to claim 57, wherein said disease is 41. Use of a compound for altering cytoskeleton or its inflammatory arthritis, and preferably rheumatoid arthritis. organization in a cell for reducing stress fibers in a synovial 59. Use according to claim 58, wherein said inflammatory cell. and/or rheumatoid arthritis is induced. 42. Use according to claim 40, wherein said compound is a 60. Use according to claim 59, wherein said induction is drug for the treatment of cancer. achieved using an inducer as depicted in table 5. 43. Use according to claim 40, wherein said compound is 61. A method for determining whether a compound is one of the following: Cytochalasin D. Phalloidin, Jasplaki effective in alleviating a symptom of inflammatory arthritis, nolide, Chondramides, Dolostatin 11, Latrunculin A, Swin comprising testing said compound in a model system for said US 2008/031 1185 A1 Dec. 18, 2008

symptom, said method characterized in that said compound is 69. Use of a compound for the manufacture of a medica a compound for altering the cytoskeleton or its organization ment for alleviating at least one symptom of inflammatory in a cell. and/or rheumatoid arthritis in an individual, said compound 62. A method according to claim 61, wherein said com comprising an antisense-Sequence of one of the genes listed pound is effective in cancer therapy. in table A, table B, or table 4. 63. A method according to claim 61, wherein said com pound is one of the following: Cytochalasin D. Phalloidin, 70. Use according to claim 69, wherein said gene is iden Jasplakinolide, Chondramides, Dolostatin 11, Latrunculin A, tified in either of said tables by expression SF Up or WJUp or Swinholide A, Misakinolide A, Mycalolide B, Shinxolide C, said gene is VASP. Scytophycins, Goniodomin A, Peeudopterolide, Lysophos 71. Use according to claim 69, wherein said compound phatidylcholine, Lysophosphatidyl acid. comprises an antisense-Sequence of one of the genes listed in 64. A method for determining whether a compound has an table A, or table 4. effect on an expression of one or more of the genes listed in 72. Use according to claim 69, wherein said compound table A, table B, or table 4, comprising testing said compound comprises an antisense-Sequence of one of the genes listed in in an environment wherein said gene is expressed and com table A or table B and an antisense-sequence of one of the paring a resulting expression level with a reference value. genes listed in table C. 65. A method for determining whether a compound has an 73. Use according to claim 69, wherein said compound effect on an expression of one or more of the genes listed in comprises an antisense-Sequence of one of the genes listed in table A, or table 4, comprising testing said compound in an table A and an antisense-Sequence of one of the genes listed in environment wherein said gene is expressed and comparing a table B or in table C. resulting expression level with a reference value. 74. Use of a protein encoded by a gene of table A, table B, 66. A method for determining whether a compound has an or table 4 as a drug target for selecting candidate drugs for a effect on an expression of one or more of the genes listed in treatment of inflammatory and/or rheumatoid arthritis. table A, table B, or table 4 and on one or more of the genes listed in table C, comprising testing said compound in an 75. Use according to claim 74, wherein said protein is environment wherein said genes are expressed and compar encoded by a gene of table A. ing a resulting expression level with a reference value. 76. Use according to claim 74, wherein a first protein 67. A method for determining whether a compound has an encoded by a gene of table A or table B is used as a first drug effect on an expression of one or more of the genes listed in target and wherein a second protein encoded by a gene of table A, or table 4 and on one or more of the genes listed in table C is used as a second drug target. table B or in table C, comprising testing said compound in an 77. Use according to claim 74, wherein a first protein environment wherein said genes are expressed and compar encoded by a gene of table A is used as a first drug target and ing a resulting expression level with a reference value. whereina second protein encoded by a gene of table Bortable 68. A method for alleviating at least one symptom of C is used as a second drug target. inflammatory and/or rheumatoid arthritis in an individual, 78. Use according to claim 74, wherein said gene is iden comprising modifying the level of expression of at least one tified in either of said tables by expression SF Up or WJUp or gene product of a gene of table A, table B, or table 4 in cells said gene is VASP. of said individual, wherein said gene product comprises pro tein.