Abstract

Objective. To obtain a global view of the immunological alterations occurring in early systemic sclerosis (SSc) by transcriptional profiling of peripheral blood cells (PBCs).

Methods. Oligonucleotide microarrays were used to compare PBC gene expression profiles in 18 SSc cases (<2 yr duration) and 18 controls matched for race, gender and ethnicity. SSc cases had no prior or current exposure to cytotoxic drugs. PAXgene tubes were used to stabilize RNA during phlebotomy. Changes in gene expression were independently validated by real-time polymerase chain reaction.

Results. SSc PBCs demonstrated differential expression of 18 interferon-inducible genes. Six of these genes were identical to the interferon signature genes in lupus peripheral blood mononuclear cells. Notably, SSc PBCs also had increased expression of allograft inflammatory factor (AIF1) and several selectins and integrins involved in cellular adhesion to the endothelium. Global analysis of 284 known biological pathways revealed that 13 were differentially regulated in SSc PBCs, including two pathways (IL2RB and GATA3) that lead to TH2 polarization.

Conclusions. Transcriptional profiling reliably discriminates between PBCs from SSc and normal donors despite the fact that they represent a heterogeneous cell population. Multiple biological pathways were differentially regulated in SSc PBCs, but a common thread across these pathways was alterations in protein tyrosine kinase 2β and mitogen-activated protein kinase signalling. Although the SSc PBC gene expression profile demonstrated some parallels with the lupus interferon gene signature, there was also increased expression of transcripts encoding proteins that target PBCs to the endothelium, which might be relevant to the vasculopathy of SSc.

Endothelial cell dysfunction and immune dysregulation can be viewed as sentinel events in the pathophysiology of systemic sclerosis (SSc) because they often precede the fibrosis that is the most widely recognized clinical feature of this disease [1]. Raynaud's phenomenon and nailfold capillary abnormalities are some of the earliest clinical signs of vascular pathology in SSc, while intimal proliferation and vascular obliteration occur in later stages of the disease [2]. In gene profiling studies of dermal fibroblasts, important effector cells involved in the pathogenesis of SSc, we found that one of the most discriminating features of SSc fibroblasts was the increased expression of transcripts for collagen XVIIIα1 (COL18A1), whose C-terminal fragment contains endostatin, a potent inhibitor of angiogenesis and endothelial proliferation. This was accompanied by decreased expression of vascular endothelial growth factor B (VEGFB) compared with normal fibroblasts, thus raising the possibility that the SSc fibroblast might contribute to the microvascular abnormalities seen in SSc [3].

Dysregulation of humoral immunity in SSc is best represented by the variety of autoantibodies to nuclear and nucleolar components spontaneously produced by SSc patients, some of which are highly disease-specific. Strong correlations exist between certain SSc-specific autoantibodies and clinical subtypes of disease [4]. In addition, recent reports suggest that anti-topoisomerase antibodies directly bind fibroblasts, and that their serum levels correlate with SSc disease activity [5, 6]. These studies, together with observations in systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA), indicate that disease-specific autoantibodies precede the first clinical signs of disease by several years, providing strong circumstantial evidence suggesting that autoantibodies might be pathogenic [7, 8].

Abnormalities in cellular immunity are apparent from the increased numbers of T cells bearing markers of activation in the peripheral blood of SSc patients [9]. In the early stages of SSc, there is a prominent dermal mononuclear infiltrate, often found in a perivascular distribution [2]. It has been shown that the T cells infiltrating the skin and the lungs of SSc patients have a restricted T-cell receptor repertoire [10–12]. In studies of bronchoalveolar lavage fluids, the TH2 cytokine pattern, mainly IL-4, has been associated with worsening forced vital capacity [13, 14].

In this report, we outline the results of transcriptional profiling of peripheral blood cells (PBCs) from patients with early SSc. Our hypothesis is that since immune dysregulation is an early event in SSc, analysis of global gene expression patterns of these effector cells might provide novel perspectives into its pathogenesis. Because substantial changes in gene expression have been well documented in blood cells due to ex vivo handling, we used a commercially available method for bedside RNA stabilization to give a more accurate reflection of the PBC gene expression in vivo [15–17].

Methods

Study sample

All study subjects provided written informed consent and the study was approved by the University of Texas-Houston Committee for the Protection of Human Subjects. The clinical characteristics of the patients are shown in Table 1. All SSc patients were in the early, oedematous phase of skin involvement and none of the patients was on any cytotoxic therapy at the time of phlebotomy, although seven patients were on low-dose prednisone (5–10 mg per day). Blood was drawn by phlebotomy directly into PAXgene™ tubes (Qiagen, Valencia, CA, USA) and all samples were handled in a standardized fashion according the manufacturer's instructions. Total RNA was extracted from 5 ml whole blood using the PAXgene Blood RNA Kit (Qiagen). RNA quality was assessed using A260/A280 ratios and the presence of intact 28S and 18S bands on agarose gel electrophoresis.

Table 1.

Clinical and serological characteristics of SSc and healthy PBC donors

Clinical characteristics SSc patients Controls 
Age (mean ± s.d.; median) 43.6 ± 13.7 yr; 43.5 yr 43.2 ± 13.6 yr; 42 yr 
Sex distribution 7 men; 11 women 7 men; 11 women 
Ethnic distribution (Caucasian:African-American:Asian) 14:3:1 14:3:1 
Disease duration from 1st non-Raynaud's symptom (mean ± s.d.; median) 2.1 ± 1.4 yr; 2 yr – 
Diffuse clinical subtype 16 (89%) – 
Interstitial lung disease 10 (56%) – 
Modified Rodnan skin score (mean ± s.d.; median) 24.8 ± 12.4; 27 – 
Serological characteristics  – 
    Anti-nuclear antibody (ANA)-positive 18 (100%) – 
    Anti-topoisomerase I 9 (50%) – 
    Anti-RNA polymerase I/III 1 (5.6%) – 
    Anti-RNP 1 (5.6%) – 
    ANA of undefined specificity 7 (39%) – 
Clinical characteristics SSc patients Controls 
Age (mean ± s.d.; median) 43.6 ± 13.7 yr; 43.5 yr 43.2 ± 13.6 yr; 42 yr 
Sex distribution 7 men; 11 women 7 men; 11 women 
Ethnic distribution (Caucasian:African-American:Asian) 14:3:1 14:3:1 
Disease duration from 1st non-Raynaud's symptom (mean ± s.d.; median) 2.1 ± 1.4 yr; 2 yr – 
Diffuse clinical subtype 16 (89%) – 
Interstitial lung disease 10 (56%) – 
Modified Rodnan skin score (mean ± s.d.; median) 24.8 ± 12.4; 27 – 
Serological characteristics  – 
    Anti-nuclear antibody (ANA)-positive 18 (100%) – 
    Anti-topoisomerase I 9 (50%) – 
    Anti-RNA polymerase I/III 1 (5.6%) – 
    Anti-RNP 1 (5.6%) – 
    ANA of undefined specificity 7 (39%) – 

Microarrays

A reference experimental design was used for these studies (Human Universal Reference RNA, Stratagene, La Jolla, CA, USA). Since class comparison and class prediction were the primary aims, we favoured studying larger numbers of biological samples rather than fewer samples with multiple technical replicates [18]. The microarray used is a custom-printed array representing 16 659 human genes from the Qiagen Array-Ready Oligo Set version 1.1. The 3DNA Array 900 Detection Kit (Genisphere, Montvale, NJ, USA) was used to synthesize labelled probes from 1 μg of peripheral blood mononuclear cell (PBMC) total RNA (Cy5) or reference RNA (Cy3), and hybridization and washes were carried out according to the vendor's protocol. The features were extracted from the TIFF files using fixed-circle segmentation (QuantArray v. 3.1; Perkin-Elmer) and preprocessed as previously described [3].

Features with a poor signal-to-noise ratio (<3) or that were obvious artefacts were flagged for exclusion. To filter out non-expressed genes, we used the t-test to compare the signal for each gene with the signals for 30 negative control features on the array across all experiments for each channel. Any gene whose signal was not significantly different from the negative control (P>0.05) was excluded. In addition, any gene in which the expression data had been filtered out in 20% of the experiments was also excluded from further analysis. The signal intensities from both channels were transformed into log2 ratios and local weighted least squares regression or lowess normalization was used to adjust for differences in labelling intensities of the two fluorescent dyes [19]. The program BRB ArrayTools v. 3.2, designed by Richard Simon and Amy Peng Lam (Biometric Research Branch, National Cancer Institute, USA), was used for data analysis [3]. This program implements a variety of analytical and data mining methods, including classification, class prediction, quantitative trait analysis [20], significance analysis of microarrays (SAM) [21], and comparative gene ontology and pathway analysis. To estimate the actual number of false discoveries at a given significance level, we used the spacings locally weighted regression smoother histogram (SPLOSH) to calculate the conditional false discovery rate (cFDR) at each significance level [22].

Quantitative real-time polymerase chan reaction

The quantitative real-time polymerase chain reaction (real-time PCR) was performed to confirm microarray results. Real-time PCR was performed using validated TaqMan® Gene Expression Assays on an Applied Biosystems 7900HT Fast Real-Time PCR System (Applied Biosystems, CA, USA). cDNA was synthesized using the High-Capacity cDNA Archive Kit from Applied Biosystems. The transferrin receptor gene (TFRC/CD71) was used as an endogenous control to normalize transcript levels of total RNA of each sample. The data were analysed with SDS 2.2 software using the comparative CT method (2-ddCT method). Fold change was calculated as 2−ΔΔCT. The Mann–Whitney U-test was used to assess significant differences in transcript levels.

Results

Global gene expression profiles from SSc and normal donors are significantly different

A total of 8541 genes (51%) passed the filtering and preprocessing criteria (see Methods) when all 36 samples were considered together. This is comparable to the percentage present calls in methods that use globin reduction protocols prior to linear RNA amplification (Affymetrix Technical Note; http://www.affymetrix.com/support/technical/technotes/blood2_technote.pdf).

We initially screened for differentially regulated genes using SAM and found 504 (5.9%) genes whose expression level was significantly different between SSc and normal PBCs. To further refine the gene list, we used the random variance t-test to discover differentially regulated genes, and SPLOSH to estimate the cFDR at each significance level. By restricting the cFDR rate to ≤10%, we found that 382 genes (4.5%) out of 8541 expressed genes were potentially differentially regulated in SSc PBCs compared with normal PBCs. This included 244 genes whose transcripts were significantly increased and 138 genes that were significantly decreased in SSc PBCs. Although the fold changes were modest (of the order of 2- to 3-fold), they were highly statistically significant, ranging from P = 0.0059 to P = 5 × 10−7. The global probability of finding 382 genes differentially expressed by chance alone if there was no real difference between SSc and normal PBCs is less than 1 in 1000 (P = 0.0007), suggesting that these genes are likely to be biologically relevant to SSc.

We then performed unsupervised hierarchical cluster analysis of all the samples with these 382 genes (Fig. 1). There was a distinct separation of the 32 SSc and normal PBMC samples into two major groups, although four of the samples (two controls and two SSc patients) were misclassified based on this gene set. The cluster demonstrated very good reproducibility, having a robustness index after 1000 permutations of R = 0.961 [23]. Figure 1 shows that two large clusters of co-regulated genes (clusters 37 and 71) separated the SSc and normal PBCs. Cluster 37 contains 73 genes overexpressed in SSc PBCs, and they include: SELL, SELPLG, ITGB2 (CD18), S100A12, S100A9 and CFLAR as well as several interferon (IFN)-inducible genes (IFI30, STAT1, TAP1, ISG20 and IFNGR2). Smaller clusters that are also overexpressed in the SSc samples include cluster 15, which contains another group of three IFN-related genes, and cluster 27, which contains S100A8, CD52 and AIF1 (Fig. 1). Cluster 71 contains 40 genes significantly underexpressed by SSc PBCs, including ADAM9, EGFR, ZNF259, FGF4, CTGF and several transcription factors (NFYA, NR3C1, POU3F1 and RB1). Quantitative differences in gene expression between SSc and normal PBCs for selected genes are shown in Table 2 and a comprehensive list is publicly available at http://www.uth.tmc.edu/scleroderma.

Fig. 1.

Hierarchical clustering of genes and samples. The gene panel used comprised the 382 genes differentially expressed at a conditional false discovery rate of ≤0.10 (P<0.006). Centred correlation metric and average linkage clustering was used. The genes constituting several clusters of interest (clusters 15, 27, 37, 64 and 71) are shown to the right of the heatmap. The complete list of genes that constitute each gene cluster is available at http://www.uth.tmc.edu/scleroderma.

Fig. 1.

Hierarchical clustering of genes and samples. The gene panel used comprised the 382 genes differentially expressed at a conditional false discovery rate of ≤0.10 (P<0.006). Centred correlation metric and average linkage clustering was used. The genes constituting several clusters of interest (clusters 15, 27, 37, 64 and 71) are shown to the right of the heatmap. The complete list of genes that constitute each gene cluster is available at http://www.uth.tmc.edu/scleroderma.

Table 2.

Selected genes differentially expressed by SSc PBCs ranked by fold difference within each groupa

Qiagen ID Description Genbank Symbol P-valueb Microarray SSc:control ratio 
IFN signaturec      
    H003786_01 Interferon α-inducible protein (clone IFI-6–16) AB019565 G1P3 1.05 × 10−4 2.25 
    H002062_01 Ribonuclease, RNase A family, 2 (liver, eosinophil-derived neurotoxin) X55987 RNASE2 1.10 × 10−4 2.02 
    H001850_01 Interferon α-inducible protein (clone IFI-15K) M13755 G1P2 (ISG15) 6.32 × 10−4 1.94 
    H000063_01 Phospholipid scramblase 1 AB006746 PLSCR1 2.08 × 10−3 1.41 
    H000397_01 Myxovirus (influenza) resistance 1, homologue of murine M33882 MX1 1.45 × 10−3 1.36 
    H002785_01 Serine (or cysteine) proteinase inhibitor, clade G (C1 inhibitor), member 1 X54486 SERPING1 2.70 × 10−2 1.32 
IFN regulated genes      
    H000825_01 S100 calcium-binding protein A8 (calgranulin A) X06234 S100A8 1.02 × 10−4 2.75 
    H003284_01 Myeloid cell nuclear differentiation antigen M81750 MNDA 5.26 × 10−3 1.75 
    H003225_01 Interferon regulatory factor 7 AF076494 IRF7 7.04 × 10−5 1.73 
    H001857_01 Transporter, ATP-binding cassette, major histocompatibility complex 1 X57522 TAP1 1.63 × 10−4 1.64 
    H000399_01 Interferon γ-inducible protein 30 J03909 IFI30 6.78 × 10−4 1.62 
    H000670_01 Interferon stimulated gene (20 kDa) U88964 ISG20 3.72 × 10−4 1.60 
    H002950_01 Signal transducer and activator of transcription 1, 91 kDa M97935 STAT1 1.03 × 10−4 1.49 
    H002834_01 Interferon γ receptor 2 (interferon γ transducer 1) U05875 IFNGR2 8.31 × 10−4 1.40 
    H009254_01 Solute carrier family 11 (proton-coupled divalent metal ion transporters), member 1 (NRAMP1) L38593 SLC11A1 1.28 × 10−3 1.27 
    H000893_01 Major histocompatibility complex, class I, E X56841 HLA-E 2.12 × 10−3 1.36 
    H004048_01 Interferon induced with helicase C domain 1 AL080107 IFIH1 2.35 × 10−3 0.70 
    H000600_01 Suppressor of cytokine signalling 3 AB004904 SOCS3 1.17 × 10−3 0.68 
Vasculotrophic      
    H002931_01 Selectin L (lymphocyte adhesion molecule 1) M25280 SELL 1.28 × 10−3 2.02 
    H001931_01 Glycoprotein Ib (platelet), β polypeptide U59632 GP1BB 3.25 × 10−4 1.94 
    H002233_01 Selectin P ligand (CD162) NM_003006 SELPLG 5.00 × 10−3 1.69 
    H002227_01 Integrin gamma 2b (platelet glycoprotein IIb of IIb/IIIa complex, antigen CD41B) NM_000419 ITGA2B 9.56 × 10−4 1.67 
    H003139_01 Integrin beta 2 (antigen CD18 (p95), lymphocyte function-associated antigen 1; macrophage antigen 1 M15395 ITGB2 5.58 × 10−3 1.58 
    H000389_01 Allograft inflammatory factor 1 Y14768 AIF1 5.17 × 10−4 1.55 
    H003834_01 Integrin alpha 5 X53002 ITGB5 2.64 × 10−3 1.49 
    H002394_01 Integrin alpha 6 X53586 ITGA6 5.05 × 10−3 1.31 
GTP-protein/ cytoskeleton      
    H002450_01 Gardner–Rasheed feline sarcoma viral (v-fgr) oncogene homologue AL031729 FGR 2.02 × 10−5 1.77 
    H014630_01 Rho GTPase activating protein 9 AF161339 ARHGAP9 1.32 × 10−3 1.70 
    H002358_01 A disintegrin and metalloproteinase domain 8 (CD156) D26579 ADAM8 4.79 × 10−3 1.64 
    H003989_01 Inositol 1,3,4-triphosphate 5/6 kinasec U51336 ITPK1 5.00 × 10−7 1.62 
    H008261_01 RAS guanyl releasing protein 2 (calcium and DAG-regulated) Y12336 RASGRP2 2.87 × 10−4 1.59 
    H004486_01 Endothelial differentiation, G-protein-coupled receptor 6 AJ000479 EDG6 8.33 × 10−4 1.58 
    H003200_01 Mitogen-activated protein kinase kinase 3 D87116 MAP2K3 8.13 × 10−4 1.56 
    H001757_01 IQ motif containing GTPase activating protein 1 L33075 IQGAP1 6.37 × 10−5 1.54 
    H003706_01 Ras homologue gene family, member G (rho G) X61587 ARHG 2.47 × 10−3 1.51 
    H012023_01 Tubulin α 1 (testis-specific) X06956 TUBA1 3.17 × 10−3 1.50 
    H004281_01 Wiskott–Aldrich syndrome (eczema-thrombocytopenia) U12707 WAS 5.22 × 10−3 1.49 
    H002895_01 Mitogen-activated protein kinase 14 L35263 MAPK14 4.11 × 10−3 1.43 
    H007053_01 HRAS-like suppressor 2 AK000563 HRASLS2 4.41 × 10−3 1.40 
    H000214_01 Ras homologue gene family, member A L09159 ARHA 1.09 × 10−3 1.35 
    H002763_01 Mitogen-activated protein kinase kinase kinase 8 D14497 MAP3K8 2.63 × 10−3 1.33 
    H005429_01 Guanine nucleotide binding protein (G protein) alpha inhibiting activity polypeptide 2 X04828 GNAI2 5.48 × 10−3 1.33 
    H001360_01 Rho-associated, coiled-coil containing protein kinase 2 NM_004850 ROCK2 2.74 × 10−3 0.78 
    H013369_01 Ral GEF with PH domain and SH3 binding motif 1 AB002349 RalGPS1A 5.37 × 10−3 0.76 
    H004391_01 G protein-coupled receptor 52 AL022171 GPR52 5.60 × 10−3 0.70 
    H004031_01 RAB3D, member RAS oncogene family NM_004283 RAB3D 2.01 × 10−3 0.70 
    H000683_01 Fibroblast growth factor receptor 1 (fms-related tyrosine kinase 2, Pfeiffer syndrome) M63889 FGFR1 1.80 × 10−3 0.69 
    H002756_01 Fibroblast growth factor receptor 4 L03840 FGFR4 3.60 × 10−3 0.69 
    H002040_01 Epidermal growth factor receptor (avian erythroblastic leukaemia viral (v-erb-b) oncogene homologue) X00588 EGFR 3.54 × 10−3 0.67 
    H009451_01 Rho GTPase activating protein 11A D87717 ARHGAP11A 1.52 × 10−3 0.65 
    H003304_01 G protein-coupled receptor 22 U66581 GPR22 1.66 × 10−5 0.61 
    H000632_01 Guanine nucleotide binding protein (G protein) α inhibiting activity polypeptide 3 M27543 GNAI3 8.70 × 10−4 0.60 
    H003000_01 RAB5A, member RAS oncogene family M28215 RAB5A 6.45 × 10−4 0.59 
    H005279_01 Gamma tubulin ring complex protein (76p gene) AK001639 76P 5.50 × 10−5 0.56 
Immunity/inflammation      
    H003335_01 S100 calcium-binding protein A9 (calgranulin B) M26311 S100A9 1.00 × 10−6 2.62 
    H001290_01 S100 calcium-binding protein A12 (calgranulin C) D83664 S100A12 6.74 × 10−5 2.62 
    H000403_01 Fc fragment of IgE, high affinity I, receptor for; γ polypeptide M33195 FCER1G 2.36 × 10−4 2.28 
    H000603_01 TYRO protein tyrosine kinase binding protein (DAP12) AF019562 TYROBP 1.87 × 10−3 1.95 
    H003124_01 Fc fragment of IgG, receptor, transporter, α U12255 FCGRT 3.00 × 10−6 1.92 
    H006273_01 B cell RAG associated protein AB011170 BRAG 5.93 × 10−4 1.88 
    H002831_01 Fc fragment of IgG, low affinity IIIb, receptor for (CD16) J04162 FCGR3B 4.92 × 10−3 1.80 
    H000438_01 CASP8 and FADD-like apoptosis regulator Y14039 CFLAR 6.69 × 10−4 1.75 
    H000866_01 CDW52 antigen (CAMPATH-1 antigen) X62466 CDW52 3.95 × 10−3 1.73 
    H004405_01 Chemokine-like factor superfamily 6 AK000403 CKLFSF6 8.00 × 10−4 1.66 
    H002964_01 B-cell CLL/lymphoma 6 (zinc finger protein 51) U00115 BCL6 5.03 × 10−3 1.66 
    H016180_01 T-cell, immune regulator 1, ATPase, H+ transporting, lysosomal V0 protein a isoform 3 XM_006100 TCIRG1 1.11 × 10−3 1.63 
    H002342_01 IL-1 receptor antagonist U65590 IL1RN 6.60 × 10−4 1.56 
    H002638_01 IL-10 receptor β Z17227 IL10RB 2.57 × 10−3 1.50 
    H007856_01 Chemokine-like factor, alternatively spliced NM_016951 CKLF 1.25 × 10−3 1.45 
    H001343_01 Transcription factor 7 (T-cell specific, HMG-box) X59871 TCF7 4.35 × 10−3 0.73 
    H002944_01 IL-8 M17017 IL8 3.68 × 10−3 0.69 
    H005604_01 B-cell receptor-associated protein 29 AK000878 BCAP29 1.26 × 10−3 0.68 
    H005608_01 IL-23 p19 subunit AB030000 IL23A 1.35 × 10−4 0.67 
    H002800_01 Connective tissue growth factor X78947 CTGF 8.47 × 10−4 0.56 
    H000375_01 ELK4, ETS-domain protein (SRF accessory protein 1) M85164 ELK4 1.50 × 10−3 0.55 
    H001032_01 Cytotoxic T-lymphocyte-associated protein 4 L15006 CTLA4 1.46 × 10−4 0.55 
Qiagen ID Description Genbank Symbol P-valueb Microarray SSc:control ratio 
IFN signaturec      
    H003786_01 Interferon α-inducible protein (clone IFI-6–16) AB019565 G1P3 1.05 × 10−4 2.25 
    H002062_01 Ribonuclease, RNase A family, 2 (liver, eosinophil-derived neurotoxin) X55987 RNASE2 1.10 × 10−4 2.02 
    H001850_01 Interferon α-inducible protein (clone IFI-15K) M13755 G1P2 (ISG15) 6.32 × 10−4 1.94 
    H000063_01 Phospholipid scramblase 1 AB006746 PLSCR1 2.08 × 10−3 1.41 
    H000397_01 Myxovirus (influenza) resistance 1, homologue of murine M33882 MX1 1.45 × 10−3 1.36 
    H002785_01 Serine (or cysteine) proteinase inhibitor, clade G (C1 inhibitor), member 1 X54486 SERPING1 2.70 × 10−2 1.32 
IFN regulated genes      
    H000825_01 S100 calcium-binding protein A8 (calgranulin A) X06234 S100A8 1.02 × 10−4 2.75 
    H003284_01 Myeloid cell nuclear differentiation antigen M81750 MNDA 5.26 × 10−3 1.75 
    H003225_01 Interferon regulatory factor 7 AF076494 IRF7 7.04 × 10−5 1.73 
    H001857_01 Transporter, ATP-binding cassette, major histocompatibility complex 1 X57522 TAP1 1.63 × 10−4 1.64 
    H000399_01 Interferon γ-inducible protein 30 J03909 IFI30 6.78 × 10−4 1.62 
    H000670_01 Interferon stimulated gene (20 kDa) U88964 ISG20 3.72 × 10−4 1.60 
    H002950_01 Signal transducer and activator of transcription 1, 91 kDa M97935 STAT1 1.03 × 10−4 1.49 
    H002834_01 Interferon γ receptor 2 (interferon γ transducer 1) U05875 IFNGR2 8.31 × 10−4 1.40 
    H009254_01 Solute carrier family 11 (proton-coupled divalent metal ion transporters), member 1 (NRAMP1) L38593 SLC11A1 1.28 × 10−3 1.27 
    H000893_01 Major histocompatibility complex, class I, E X56841 HLA-E 2.12 × 10−3 1.36 
    H004048_01 Interferon induced with helicase C domain 1 AL080107 IFIH1 2.35 × 10−3 0.70 
    H000600_01 Suppressor of cytokine signalling 3 AB004904 SOCS3 1.17 × 10−3 0.68 
Vasculotrophic      
    H002931_01 Selectin L (lymphocyte adhesion molecule 1) M25280 SELL 1.28 × 10−3 2.02 
    H001931_01 Glycoprotein Ib (platelet), β polypeptide U59632 GP1BB 3.25 × 10−4 1.94 
    H002233_01 Selectin P ligand (CD162) NM_003006 SELPLG 5.00 × 10−3 1.69 
    H002227_01 Integrin gamma 2b (platelet glycoprotein IIb of IIb/IIIa complex, antigen CD41B) NM_000419 ITGA2B 9.56 × 10−4 1.67 
    H003139_01 Integrin beta 2 (antigen CD18 (p95), lymphocyte function-associated antigen 1; macrophage antigen 1 M15395 ITGB2 5.58 × 10−3 1.58 
    H000389_01 Allograft inflammatory factor 1 Y14768 AIF1 5.17 × 10−4 1.55 
    H003834_01 Integrin alpha 5 X53002 ITGB5 2.64 × 10−3 1.49 
    H002394_01 Integrin alpha 6 X53586 ITGA6 5.05 × 10−3 1.31 
GTP-protein/ cytoskeleton      
    H002450_01 Gardner–Rasheed feline sarcoma viral (v-fgr) oncogene homologue AL031729 FGR 2.02 × 10−5 1.77 
    H014630_01 Rho GTPase activating protein 9 AF161339 ARHGAP9 1.32 × 10−3 1.70 
    H002358_01 A disintegrin and metalloproteinase domain 8 (CD156) D26579 ADAM8 4.79 × 10−3 1.64 
    H003989_01 Inositol 1,3,4-triphosphate 5/6 kinasec U51336 ITPK1 5.00 × 10−7 1.62 
    H008261_01 RAS guanyl releasing protein 2 (calcium and DAG-regulated) Y12336 RASGRP2 2.87 × 10−4 1.59 
    H004486_01 Endothelial differentiation, G-protein-coupled receptor 6 AJ000479 EDG6 8.33 × 10−4 1.58 
    H003200_01 Mitogen-activated protein kinase kinase 3 D87116 MAP2K3 8.13 × 10−4 1.56 
    H001757_01 IQ motif containing GTPase activating protein 1 L33075 IQGAP1 6.37 × 10−5 1.54 
    H003706_01 Ras homologue gene family, member G (rho G) X61587 ARHG 2.47 × 10−3 1.51 
    H012023_01 Tubulin α 1 (testis-specific) X06956 TUBA1 3.17 × 10−3 1.50 
    H004281_01 Wiskott–Aldrich syndrome (eczema-thrombocytopenia) U12707 WAS 5.22 × 10−3 1.49 
    H002895_01 Mitogen-activated protein kinase 14 L35263 MAPK14 4.11 × 10−3 1.43 
    H007053_01 HRAS-like suppressor 2 AK000563 HRASLS2 4.41 × 10−3 1.40 
    H000214_01 Ras homologue gene family, member A L09159 ARHA 1.09 × 10−3 1.35 
    H002763_01 Mitogen-activated protein kinase kinase kinase 8 D14497 MAP3K8 2.63 × 10−3 1.33 
    H005429_01 Guanine nucleotide binding protein (G protein) alpha inhibiting activity polypeptide 2 X04828 GNAI2 5.48 × 10−3 1.33 
    H001360_01 Rho-associated, coiled-coil containing protein kinase 2 NM_004850 ROCK2 2.74 × 10−3 0.78 
    H013369_01 Ral GEF with PH domain and SH3 binding motif 1 AB002349 RalGPS1A 5.37 × 10−3 0.76 
    H004391_01 G protein-coupled receptor 52 AL022171 GPR52 5.60 × 10−3 0.70 
    H004031_01 RAB3D, member RAS oncogene family NM_004283 RAB3D 2.01 × 10−3 0.70 
    H000683_01 Fibroblast growth factor receptor 1 (fms-related tyrosine kinase 2, Pfeiffer syndrome) M63889 FGFR1 1.80 × 10−3 0.69 
    H002756_01 Fibroblast growth factor receptor 4 L03840 FGFR4 3.60 × 10−3 0.69 
    H002040_01 Epidermal growth factor receptor (avian erythroblastic leukaemia viral (v-erb-b) oncogene homologue) X00588 EGFR 3.54 × 10−3 0.67 
    H009451_01 Rho GTPase activating protein 11A D87717 ARHGAP11A 1.52 × 10−3 0.65 
    H003304_01 G protein-coupled receptor 22 U66581 GPR22 1.66 × 10−5 0.61 
    H000632_01 Guanine nucleotide binding protein (G protein) α inhibiting activity polypeptide 3 M27543 GNAI3 8.70 × 10−4 0.60 
    H003000_01 RAB5A, member RAS oncogene family M28215 RAB5A 6.45 × 10−4 0.59 
    H005279_01 Gamma tubulin ring complex protein (76p gene) AK001639 76P 5.50 × 10−5 0.56 
Immunity/inflammation      
    H003335_01 S100 calcium-binding protein A9 (calgranulin B) M26311 S100A9 1.00 × 10−6 2.62 
    H001290_01 S100 calcium-binding protein A12 (calgranulin C) D83664 S100A12 6.74 × 10−5 2.62 
    H000403_01 Fc fragment of IgE, high affinity I, receptor for; γ polypeptide M33195 FCER1G 2.36 × 10−4 2.28 
    H000603_01 TYRO protein tyrosine kinase binding protein (DAP12) AF019562 TYROBP 1.87 × 10−3 1.95 
    H003124_01 Fc fragment of IgG, receptor, transporter, α U12255 FCGRT 3.00 × 10−6 1.92 
    H006273_01 B cell RAG associated protein AB011170 BRAG 5.93 × 10−4 1.88 
    H002831_01 Fc fragment of IgG, low affinity IIIb, receptor for (CD16) J04162 FCGR3B 4.92 × 10−3 1.80 
    H000438_01 CASP8 and FADD-like apoptosis regulator Y14039 CFLAR 6.69 × 10−4 1.75 
    H000866_01 CDW52 antigen (CAMPATH-1 antigen) X62466 CDW52 3.95 × 10−3 1.73 
    H004405_01 Chemokine-like factor superfamily 6 AK000403 CKLFSF6 8.00 × 10−4 1.66 
    H002964_01 B-cell CLL/lymphoma 6 (zinc finger protein 51) U00115 BCL6 5.03 × 10−3 1.66 
    H016180_01 T-cell, immune regulator 1, ATPase, H+ transporting, lysosomal V0 protein a isoform 3 XM_006100 TCIRG1 1.11 × 10−3 1.63 
    H002342_01 IL-1 receptor antagonist U65590 IL1RN 6.60 × 10−4 1.56 
    H002638_01 IL-10 receptor β Z17227 IL10RB 2.57 × 10−3 1.50 
    H007856_01 Chemokine-like factor, alternatively spliced NM_016951 CKLF 1.25 × 10−3 1.45 
    H001343_01 Transcription factor 7 (T-cell specific, HMG-box) X59871 TCF7 4.35 × 10−3 0.73 
    H002944_01 IL-8 M17017 IL8 3.68 × 10−3 0.69 
    H005604_01 B-cell receptor-associated protein 29 AK000878 BCAP29 1.26 × 10−3 0.68 
    H005608_01 IL-23 p19 subunit AB030000 IL23A 1.35 × 10−4 0.67 
    H002800_01 Connective tissue growth factor X78947 CTGF 8.47 × 10−4 0.56 
    H000375_01 ELK4, ETS-domain protein (SRF accessory protein 1) M85164 ELK4 1.50 × 10−3 0.55 
    H001032_01 Cytotoxic T-lymphocyte-associated protein 4 L15006 CTLA4 1.46 × 10−4 0.55 

aComplete list available at http://www.uth.tmc.edu/scleroderma; bsignificance by the random variance t-test; cIFN-regulated genes that constitute the IFN signature according to Baechler et al. [24].

Multiple interferon-inducible genes and biological pathways are differentially regulated in SSc PBCs

Closer examination of the data shows that SSc PBCs have significantly increased expression of at least seven type I IFN-inducible genes, including G1P3, G1P2, MNDA, IRF7, TAP1, ISG20 and MX1 (Table 2). In addition, there was also differential expression genes typically activated by IFN-γ, a type II IFN (IFI30, IFNGR2, HLA-E, SLC11A1 and S100A8) (Table 2). In contrast, SOCS3, an important negative regulator of IFN-γ signalling, was significantly decreased in SSc PBCs, while STAT1, which is involved in both type I and type II IFN signalling, was significantly increased in SSc PBCs.

Next, using BRB Array Tools, we examined known biological pathways in order to discover those that might be differentially regulated between SSc and control PBCs. The results are summarized in Table 3. Thirteen out of 284 biological pathways in publicly available (KEGG and BioCarta) databases were found to be differentially regulated in SSc PBCs. It should be pointed out that these pathways are not mutually exclusive and some of the individual genes were represented in more than one biological pathway. The complete list of genes constituting these pathways is available at http://www.uth.tmc.edu/scleroderma. The most significantly differentially regulated pathway was the Pyk2 pathway (P = 1 × 10−5), with eight of 15 genes showing differential expression. These included increased expression of GRB2 (1.5-fold, P = 0.0007), PTK2B (1.40-fold, P = 0.007), RAF1 (1.6-fold, P = 0.006), MAPK14 and MAP2K3, and decreased expression of RBL2 (1.4-fold, P = 0.01), CRKL (1.5-fold, P = 0.008) and MAP2K4 (1.3-fold, P = 0.007).

Table 3.

Pathways which discriminate SSc and normal PBCs sorted by significancea

Biocarta pathway IDb Description Number of genes LS permutationcP-value KS permutationdP-value 
h_pyk2 Pathway Links between Pyk2 and map kinases 16 0.00001 0.0017 
h_igf1 Pathway IGF-1 signalling pathway 0.00065 0.0007 
h_egf Pathway EGF signalling pathway 11 0.00073 0.013 
h_At1r Pathway Angiotensin II mediated activation of JNK pathway via Pyk2-dependent signalling 16 0.00076 0.0043 
h_insulin Pathway Insulin signalling pathway 0.0011 0.012 
h_spry Pathway Sprouty regulation of tyrosine kinase signals 0.0017 0.008 
h_GATA3 Pathway GATA3 participates in activating Th2 cytokine gene expression 0.0023 0.021 
h_gh Pathway Growth hormone signalling pathway 11 0.0024 0.068 
h_cardiacegf Pathway Role of EGF receptor transactivation by GPCRs in cardiac hypertrophy 0.0034 0.039 
h_creb Pathway Transcription factor CREB and its extracellular signals 11 0.0035 0.014 
h_il2rb Pathway IL-2 receptor beta chain in T cell activation 23 0.0043 0.091 
h_igf1r Pathway Multiple anti-apoptotic pathways from IGF-1R signalling lead to BAD phosphorylation 11 0.0045 0.027 
h_pdgf Pathway PDGF signalling pathway 11 0.0048 0.023 
Biocarta pathway IDb Description Number of genes LS permutationcP-value KS permutationdP-value 
h_pyk2 Pathway Links between Pyk2 and map kinases 16 0.00001 0.0017 
h_igf1 Pathway IGF-1 signalling pathway 0.00065 0.0007 
h_egf Pathway EGF signalling pathway 11 0.00073 0.013 
h_At1r Pathway Angiotensin II mediated activation of JNK pathway via Pyk2-dependent signalling 16 0.00076 0.0043 
h_insulin Pathway Insulin signalling pathway 0.0011 0.012 
h_spry Pathway Sprouty regulation of tyrosine kinase signals 0.0017 0.008 
h_GATA3 Pathway GATA3 participates in activating Th2 cytokine gene expression 0.0023 0.021 
h_gh Pathway Growth hormone signalling pathway 11 0.0024 0.068 
h_cardiacegf Pathway Role of EGF receptor transactivation by GPCRs in cardiac hypertrophy 0.0034 0.039 
h_creb Pathway Transcription factor CREB and its extracellular signals 11 0.0035 0.014 
h_il2rb Pathway IL-2 receptor beta chain in T cell activation 23 0.0043 0.091 
h_igf1r Pathway Multiple anti-apoptotic pathways from IGF-1R signalling lead to BAD phosphorylation 11 0.0045 0.027 
h_pdgf Pathway PDGF signalling pathway 11 0.0048 0.023 

aComplete list of genes constituting each pathway available at http://www.uth.tmc.edu/scleroderma; bhttp://cgap.nci.nih.gov/Pathways/BioCarta/; cFisher's least significant statistic after 100 000 permutations; dKolmogorov–Smirnov statistic after 100 000 permutations.

The IL-2RB pathway (P = 0.0043) contains the largest number of genes among the differentially regulated pathways, with 23 genes in all (Table 3). These included increased expression of CFLAR (1.8-fold, P = 0.0007), GRB2, RAF1 and RPS6KB1 (1.4-fold, P = 0.02) and decreased expression of SOCS3, IRS1 (1.6-fold, P = 0.002), and CRKL. In another pathway for TH2 activation, GATA3 (P = 0.002), three of five genes were significantly differentially regulated, including JUNB (1.4-fold, P = 0.002), MAPK14 (1.4-fold, P = 0.004) and MAP2K3 (1.6-fold, P = 0.0008).

Independent validation of array data with real-time PCR

We then selected 11 genes differentially regulated on the microarray between SSc and normal PBCs for independent validation by real-time PCR (Fig. 2). Four of these genes, ITGA2B, G1PBB, SELL and AIF1, code for proteins that are important for endothelial cell adhesion, while IRF7, G1P3 and S100A8 are IFN-inducible genes. Overall, the real-time PCR data showed good concordance with the microarray results (Fig. 2).

Fig. 2.

Comparative expression levels of genes in SSc vs controls by real-time PCR. Real-time PCR of 11 genes selected from microarray data was performed on cDNA of SSc patients and controls. The box plots above show the normalized transcript levels and the median is indicated by the horizontal line. The error bars indicate the 90th and 10th percentiles, and the boxed areas indicate the 25th and 75th percentiles. The dots represent the highest and lowest expression levels for that gene. The Mann–Whitney U-test was used to assess the differences for statistical significance.

Fig. 2.

Comparative expression levels of genes in SSc vs controls by real-time PCR. Real-time PCR of 11 genes selected from microarray data was performed on cDNA of SSc patients and controls. The box plots above show the normalized transcript levels and the median is indicated by the horizontal line. The error bars indicate the 90th and 10th percentiles, and the boxed areas indicate the 25th and 75th percentiles. The dots represent the highest and lowest expression levels for that gene. The Mann–Whitney U-test was used to assess the differences for statistical significance.

Lack of correlation between gene expression patterns and clinical parameters

We performed a subanalysis of the SSc patients using with SAM and the random variance t-test to detect any genes that may be correlated with clinical features. We were unable to detect any genes that correlated with modified Rodnan skin score, anti-topoisomerase antibody status, interstitial lung disease, or low dose prednisone use.

Discussion

To our knowledge, this is the first report of transcriptional profiling of PBCs in SSc patients. Previous studies have been carried out in PBMCs in SLE, RA, ANCA-associated glomerulonephritis and multiple sclerosis [24–28]. However, there is a significant trade-off with the use of purified cells in that substantial changes in gene expression have been documented in PBCs with ex vivo handling [15, 16]. Indeed, Baechler et al. [15] has reported that out of 4566 genes expressed in PBMCs, 2034 undergo significant changes in expression under environmental stress, and these had to be excluded from analysis [24]. By using a commercially available method to immediately stabilize blood RNA upon phlebotomy, we could better preserve the immunological alterations that are present in vivo, with minimal interference from artefacts that arise from blood handling, storage and separation [17]. The disadvantage of this approach is that once the whole blood is stabilized it is not possible to subsequently separate the different cell populations. However, we felt that this was a reasonable compromise, given that the present approach makes sample collection practical in a busy out-patient clinical setting, and data retention is maximized by immediately stabilizing the blood RNA at bedside.

Our results show that transcriptional profiling can reliably discriminate between PBCs from SSc and normal donors despite the fact that PBCs are a heterogeneous cell population of lymphocytes, neutrophils, monocytes, macrophages, NK cells, dendritic cells and mast cells. These differences clearly highlight the signatures of immunological dysregulation of SSc and provide a framework for subsequent investigations into disease mechanisms or biomarkers. These data, taken together with our previous observations of increased expression of COL18A1, and decreased expression of VEGFB in cultured, non-lesional dermal fibroblasts from patients with early diffuse SSc, suggest intriguing mechanisms in which the immune system and the fibroblasts can each uniquely contribute to the vasculopathy of SSc [3].

Relevant to the vasculopathy of SSc is that PBCs from SSc patients have increased expression of a set of genes that appear to target these cells to the endothelium. These include GP1BB, ITGA2B, AIF1, SELL, SELPLG, ITGB2, ITGA6 and ITGB5 [29, 30]. SELL and ITGA6 encode key leucocyte adhesion molecules involved in initial tethering to the endothelial cell. SELPLG, a major ligand for selectin-P, has been shown to be involved in leucocyte rolling on endothelial cells and the homing of T lymphocytes to the skin. On the other hand, ITGA2B and GP1BB are receptors for von Willebrand factor (vWF) which have been shown to be abnormally expressed in SSc [31–33]. AIF1 is a gene previously reported to be expressed by macrophages in response to various cytokines (e.g. IFN-γ, TGF-β), and in response to vascular injury it enhances vascular smooth muscle cell migration and intimal hyperplasia [34, 35]. Differential regulation of multiple transcripts involved in the actin cytoskeleton was also noted. Indeed, a common thread among all the differentially regulated pathways in Table 3 is protein tyrosine kinase 2β and/or mitogen-activated protein kinase (MAP kinase) signalling. These genes play a critical role in the actin cytoskeletal remodelling required for leucocyte migration, inflammation and other biological functions [29]. Regarding inflammation, it was also interesting to find a striking increase in transcript levels for S100A8, S100A9 and S100A12 in SSc PBCs (Fig. 2). The S100/calgranulins are a group of pro-inflammatory calcium-binding proteins made and secreted by phagocytes [36]. S100A12, in particular, is involved in a novel inflammatory signal amplification pathway through interaction with the receptor for advanced glycation end products (RAGE) [37].

The first array studies of PBMCs in SLE patients reported increased expression of multiple type I and type II IFN-inducible genes [24, 25]. It is interesting to note that, of the 14 ‘IFN signature’ genes described by Baechler et al. in SLE PBMCs [24], six were also found to be differentially regulated in SSc PBCs, including G1P3, G1P2, MX1, RNASE2, PLSCR1 and SERPING1. Recently it has been reported that mononuclear cells from SLE patients had coordinated overexpression of a number of type I IFN genes and that these gene signatures correlated with disease activity [38]. In our study, the observed induction of IRF7 is of interest since this is an important regulator of the type I IFN response [39]. However, at this point it would be premature to conclude that there is a dominant type I IFN gene signature in SSc, since IRF7 can be activated by IFN-independent pathways [40]. Nonetheless, the parallels with the IFN signature in SLE remain striking.

The fact that SSc patients demonstrate activation of T cells is well described in the literature [9]. From this standpoint, it was interesting to find significantly increased expression of CFLAR in SSc PBCs (2-fold, P = 0.0007), since constitutive overexpression of CFLAR has been shown to drive the polarization of T-cell responses towards TH2 [41]. Conversely, we found that transcripts for CTLA4 were significantly reduced (2-fold, P = 0.0001) in SSc PBCs. CTLA4 is a costimulatory molecule expressed by CD4+ and CD8+ T cells that terminates the T-cell response by inhibiting activation signals delivered by CD28 [42]. Along the same lines, we observed significant differential regulation of multiple genes in the IL2RB pathway and the GATA3 pathway (Table 3). Both intermediate- and high-affinity isoforms of the IL-2 receptor contain a β subunit; these isoforms are expressed in activated T cells, and are involved receptor-mediated endocytosis and transduction of mitogenic signals from IL-2 [43]. GATA3 is an important transcription factor in the regulation of human TH2 cell differentiation in vivo [44]. Thus, differential regulation of these pathways is consistent with the observed T-cell activation in SSc.

Although limited to the transcript level, the large amount of data obtained from gene expression studies provides a framework for future investigations into the complex cellular and genetic interactions that lead to SSc. From the point of view of understanding disease pathogenesis, it will be important to dissect out which cell types are responsible for the observed gene signatures in future studies. Such cell purification studies will need to systematically take into account and control for the changes in gene expression induced by the process of storage and purification of living cells [15, 16].

Another important question to be addressed in future studies is whether these IFN and vasculotrophic gene signatures are unique to SSc. Although comparison with published data on SLE reveal some parallels, a formal study using the same array platform with various autoimmune connective tissue diseases (e.g. SLE and RA) would be needed. From the standpoint of disease biomarkers, it would be important to determine whether these gene expression signatures could identify different clinical subsets of SSc or correlate with disease outcome or prognosis. We are unable to detect any differences in the SSc patients in or subanalysis. This is probably due to the loss of statistical power as a result of sample partitioning and because we are now assessing for subtle differences in subjects within the SSc group, which generally are more phenotypically similar to each other than are healthy controls. With a sufficiently large cohort of SSc patients, it is likely that such subtle differences would be detectable. All of these future studies will require careful case ascertainment and matching with healthy controls as well as disease controls to minimize confounding variables, such as age, gender and pharmacological interventions. Correlation with disease variables will require serial PBC samples and assessments of SSc disease activity in a sufficiently large longitudinal cohort. Whole-blood RNA stabilization makes undertaking these studies feasible, especially if multiple academic centres are involved. Such studies will require considerable effort and resources, but hold promise for the identification of candidate biomarkers for SSc disease activity or severity in the future.

graphic

This work was supported by grants NIH-NCRR 3M01RR02558-12S1, NIH-NIAMS IP50AR4488, NIH-NIAMS N01-AR-02251 and NIH 5T35 DK007676, and grants from the Scleroderma Foundation.

The authors have declared no conflicts of interest.

References

1
Prescott RJ, Freemont AJ, Jones CJ, Hoyland J, Fielding P. Sequential dermal microvascular and perivascular changes in the development of scleroderma.
J Pathol
 
1992
;
166
:
255
–63.
2
Korn JH. Pathogenesis of systemic sclerosis. In: Koopman WJ, Moreland LW, eds.
Arthritis and allied conditions
 . Philadelphia: Lippincott Williams & Wilkins,
2005
:
1621
–32.
3
Tan FK, Hildebrand BA, Lester MS et al. Classification analysis of the transcriptosome of nonlesional cultured dermal fibroblasts from systemic sclerosis patients with early disease.
Arthritis Rheum
 
2005
;
52
:
865
–76.
4
Silver RM, Medsger TA Jr, Bolster MB. Systemic sclerosis and scleroderma variants: clinical aspects. In: Koopman WJ, Moreland LW, eds.
Arthritis and allied conditions
 . Philadelphia: Lippincot Williams & Wilkins,
2005
:
1633
–80.
5
Henault J, Tremblay M, Clement I, Raymond Y, Senecal JL. Direct binding of anti-DNA topoisomerase I autoantibodies to the cell surface of fibroblasts in patients with systemic sclerosis.
Arthritis Rheum
 
2004
;
50
:
3265
–74.
6
Hu PQ, Fertig N, Medsger TA Jr, Wright TM. Correlation of serum anti-DNA topoisomerase I antibody levels with disease severity and activity in systemic sclerosis.
Arthritis Rheum
 
2003
;
48
:
1363
–73.
7
Arbuckle MR, McClain MT, Rubertone MV et al. Development of autoantibodies before the clinical onset of systemic lupus erythematosus.
N Engl J Med
 
2003
;
349
:
1526
–33.
8
Nielen MM, van SD, Reesink HW et al. Specific autoantibodies precede the symptoms of rheumatoid arthritis: a study of serial measurements in blood donors.
Arthritis Rheum
 
2004
;
50
:
380
–6.
9
Sakkas LI, Platsoucas CD. Is systemic sclerosis an antigen-driven T cell disease?
Arthritis Rheum
 
2004
;
50
:
1721
–33.
10
Sakkas LI, Xu B, Artlett CM, Lu S, Jimenez SA, Platsoucas CD. Oligoclonal T cell expansion in the skin of patients with systemic sclerosis.
J Immunol
 
2002
;
168
:
3649
–59.
11
Yurovsky VV, Sutton PA, Schulze DH et al. Expansion of selected V delta 1+ gamma delta T cells in systemic sclerosis patients.
J Immunol
 
1994
;
153
:
881
–91.
12
Yurovsky VV, Wigley FM, Wise RA, White B. Skewing of the CD8+ T-cell repertoire in the lungs of patients with systemic sclerosis.
Hum Immunol
 
1996
;
48
:
84
–97.
13
Atamas SP, Yurovsky VV, Wise R et al. Production of type 2 cytokines by CD8+ lung cells is associated with greater decline in pulmonary function in patients with systemic sclerosis [see comments].
Arthritis Rheum
 
1999
;
42
:
1168
–78.
14
Luzina IG, Atamas SP, Wise R et al. Occurrence of an activated, profibrotic pattern of gene expression in lung CD8+ T cells from scleroderma patients.
Arthritis Rheum
 
2003
;
48
:
2262
–74.
15
Baechler EC, Batliwalla FM, Karypis G et al. Expression levels for many genes in human peripheral blood cells are highly sensitive to ex vivo incubation.
Genes Immun
 
2004
;
5
:
347
–53.
16
Tanner MA, Berk LS, Felten DL, Blidy AD, Bit SL, Ruff DW. Substantial changes in gene expression level due to the storage temperature and storage duration of human whole blood.
Clin Lab Haematol
 
2002
;
24
:
337
–41.
17
Rainen L, Oelmueller U, Jurgensen S et al. Stabilization of mRNA expression in whole blood samples.
Clin Chem
 
2002
;
48
:
1883
–90.
18
Simon RM, Dobbin K. Experimental design of DNA microarray experiments.
BioTechniques
 
2003
;
34
:
S16
–S21.
19
Yang YH, Dudoit S, Luu P et al. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation.
Nucleic Acids Res
 
2002
;
30
:
e15
.
20
Wright GW, Simon RM. A random variance model for detection of differential gene expression in small microarray experiments.
Bioinformatics
 
2003
;
19
:
2448
–55.
21
Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response.
Proc Natl Acad Sci USA
 
2001
;
98
:
5116
–21.
22
Pounds S, Cheng C. Improving false discovery rate estimation.
Bioinformatics
 
2004
;
20
:
1737
–45.
23
McShane LM, Radmacher MD, Freidlin B, Yu R, Li MC, Simon R. Methods for assessing reproducibility of clustering patterns observed in analyses of microarray data.
Bioinformatics
 
2002
;
18
:
1462
–9.
24
Baechler EC, Batliwalla FM, Karypis G et al. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus.
Proc Natl Acad Sci USA
 
2003
;
100
:
2610
–5.
25
Bennett L, Palucka AK, Arce E et al. Interferon and granulopoiesis signatures in systemic lupus erythematosus blood.
J Exp Med
 
2003
;
197
:
711
–23.
26
Olsen N, Sokka T, Seehorn CL et al. A gene expression signature for recent onset rheumatoid arthritis in peripheral blood mononuclear cells.
Ann Rheum Dis
 
2004
;
63
:
1387
–92.
27
Yang JJ, Pendergraft WF, Alcorta DA et al. Circumvention of normal constraints on granule protein gene expression in peripheral blood neutrophils and monocytes of patients with antineutrophil cytoplasmic autoantibody-associated glomerulonephritis.
J Am Soc Nephrol
 
2004
;
15
:
2103
–14.
28
Bomprezzi R, Ringner M, Kim S et al. Gene expression profile in multiple sclerosis patients and healthy controls: identifying pathways relevant to disease.
Hum Mol Genet
 
2003
;
12
:
2191
–9.
29
Worthylake RA, Burridge K. Leukocyte transendothelial migration: orchestrating the underlying molecular machinery.
Curr Opin Cell Biol
 
2001
;
13
:
569
–77.
30
Kitayama J, Ikeda S, Kumagai K, Saito H, Nagawa H. Alpha 6 beta 1 integrin (VLA-6) mediates leukocyte tether and arrest on laminin under physiological shear flow.
Cell Immunol
 
2000
;
199
:
97
–103.
31
Hirata T, Furukawa Y, Yang BG et al. Human P-selectin glycoprotein ligand-1 (PSGL-1) interacts with the skin-associated chemokine CCL27 via sulfated tyrosines at the PSGL-1 amino terminus.
J Biol Chem
 
2004
;
279
:
51775
–82.
32
Kahaleh MB, Osborn I, Leroy EC. Increased factor VIII/von Willebrand factor antigen and von Willebrand factor activity in scleroderma and in Raynaud's phenomenon.
Ann Intern Med
 
1981
;
94
:
482
–4.
33
Konttinen YT, Mackiewicz Z, Ruuttila P et al. Vascular damage and lack of angiogenesis in systemic sclerosis skin.
Clin Rheumatol
 
2003
;
22
:
196
–202.
34
Autieri MV. cDNA cloning of human allograft inflammatory factor-1: tissue distribution, cytokine induction, and mRNA expression in injured rat carotid arteries.
Biochem Biophys Res Commun
 
1996
;
228
:
29
–37.
35
Autieri MV, Kelemen SE, Wendt KW. AIF-1 is an actin-polymerizing and Rac1-activating protein that promotes vascular smooth muscle cell migration.
Circ Res
 
2003
;
92
:
1107
–14.
36
Roth J, Vogl T, Sorg C, Sunderkotter C. Phagocyte-specific S100 proteins: a novel group of proinflammatory molecules.
Trends Immunol
 
2003
;
24
:
155
–8.
37
Schmidt AM, Yan SD, Yan SF, Stern DM. The multiligand receptor RAGE as a progression factor amplifying immune and inflammatory responses.
J Clin Invest
 
2001
;
108
:
949
–55.
38
Kirou KA, Lee C, George S, Louca K, Peterson MG, Crow MK. Activation of the interferon-alpha pathway identifies a subgroup of systemic lupus erythematosus patients with distinct serologic features and active disease.
Arthritis Rheum
 
2005
;
52
:
1491
–503.
39
Honda K, Yanai H, Negishi H et al. IRF-7 is the master regulator of type-I interferon-dependent immune responses.
Nature
 
2005
;
434
:
772
–7.
40
Ning S, Huye LE, Pagano JS. Regulation of the transcriptional activity of the IRF7 promoter by a pathway independent of interferon signalling.
J Biol Chem
 
2005
;
280
:
12262
–70.
41
Tseveleki V, Bauer J, Taoufik E et al. Cellular FLIP (long isoform) overexpression in T cells drives Th2 effector responses and promotes immunoregulation in experimental autoimmune encephalomyelitis.
J Immunol
 
2004
;
173
:
6619
–26.
42
Noel PJ, Boise LH, Thompson CB. Regulation of T cell activation by CD28 and CTLA4.
Adv Exp Med Biol
 
1996
;
406
:
209
–17.
43
Tsudo M, Kitamura F, Miyasaka M. Characterization of the interleukin 2 receptor beta chain using three distinct monoclonal antibodies.
Proc Natl Acad Sci USA
 
1989
;
86
:
1982
–6.
44
Skapenko A, Leipe J, Niesner U et al. GATA-3 in human T cell helper type 2 development.
J Exp Med
 
2004
;
199
:
423
–8.

Author notes

Division of Rheumatology, Department of Internal Medicine, UT Houston Medical School, Houston, TX and 1Center for Genome Information, University of Cincinnati, Cincinnati, OH, USA.

Comments

0 Comments