- Split View
-
Views
-
Cite
Cite
Raghavan Raju, Marinos C. Dalakas, Gene expression profile in the muscles of patients with inflammatory myopathies: effect of therapy with IVIg and biological validation of clinically relevant genes, Brain, Volume 128, Issue 8, August 2005, Pages 1887–1896, https://doi.org/10.1093/brain/awh518
- Share Icon Share
Abstract
To explore the biological significance of gene expression in the pathogenesis of inflammatory myopathies, we performed microarray experiments followed by real-time PCR and immunohistochemistry on muscle biopsies obtained before and after therapy from patients with dermatomyositis (DM) who improved and patients with inclusion body myositis (sIBM) who did not improve after controlled trials with three monthly intravenous immunoglobulin (IVIg) infusions. The pretreatment biopsies showed high expression of immunoglobulin, adhesion molecules, chemokines and cytokine genes in both sIBM and DM (sIBM > DM). In the repeated biopsies of DM patients who clinically improved, 2206 genes were downregulated more than 1.5-fold; in contrast, 1700 of the same genes remained unchanged in sIBM patients who did not improve. Genes markedly downregulated in DM, but not sIBM, were interleukin 22, Kallmann syndrome 1 (KAL-1), an adhesion molecule shown for the first time in muscle, ICAM-1, complement C1q, and several structural protein genes. Because mRNA for KAL-1 was selectively upregulated in vitro by transforming growth factor (TGF) β1, a fibrogenic cytokine immunolocalized in the endomysial connective tissue of pretreatment DM muscles, the downregulation of both TGF-β and KAL-1 after IVIg only in DM suggests that these molecules have a functional role in connective tissue proliferation and fibrosis. The improved muscles of DM, but not sIBM, showed upregulation of chemokines CXCL9 (Mig) and CXCL11, and several immunoglobulin-related genes, suggesting an effect on muscle remodelling and regeneration. The results suggest that IVIg modulates several immunoregulatory or structural muscle genes, but only a subset of them associated with inflammatory mediators, fibrosis and muscle remodelling are connected with the clinical response. Gene arrays, when combined with clinical assessments, may provide important information in the pathogenesis of inflammatory myopathies.
Introduction
The inflammatory myopathies comprise three distinct subsets: dermatomyositis (DM), polymyositis (PM) and sporadic inclusion body myositis (sIBM) (Hohlfeld and Engel, 1994; Dalakas, 1998; Dalakas and Hohlfeld, 2003; Engel and Hohlfeld, 2004). DM is characterized by early deposition of complement on the intramuscular capillaries, leading to capillary destruction, perivascular inflammation and intense upregulation of MHC antigens, cytokines and adhesion molecules (Dalakas, 1998; Dalakas and Hohlfeld, 2003). PM and sIBM are characterized by activated T cells that invade muscle fibres expressing MHC class I antigens, intense activation of immunoregulatory T-cell pathways, and upregulation of cytokines, chemokines and adhesion molecules. Although DM and sIBM are clinicopathologically distinct and respond differently to therapies, they share common inflammatory and other immunoregulatory markers. This conclusion has been corroborated at the transcriptional level, where several genes, such as immunoglobulin, T-cell receptor and interferon-inducible genes are upregulated more than 10-fold in both DM and sIBM (Greenberg et al., 2002; Tezak et al., 2002). The significance of the amplified genes in the pathogenesis of DM and sIBM, however, is unclear because of lack of biological validation and correlation with responsiveness to immunotherapies.
To assess the biological relevance of these genes in the immunopathogenesis of DM and sIBM, we examined muscle biopsies obtained before and after therapy with IVIg, a potent immunomodulatory agent that exerts a combined effect on suppressing complement and downregulating cytokines, chemokines, adhesion molecules and markers of T cell activation (Dalakas, 2004). We took advantage of the noted discrepancy in the clinical response to intravenous immunoglobulin (IVIg) between DM and sIBM and the availability of muscle biopsy specimens obtained before and after therapy from patients enrolled in the previously published controlled clinical trials (Dalakas et al., 1993, 2001); while DM patients improved dramatically after three infusions with IVIg, given every month (Dalakas et al., 1993), sIBM patients did not respond (Dalakas et al., 2001). These specimens also provide the opportunity to study for the first time the transcriptional profile associated with muscle remodelling and regeneration in muscles from patients whose strength has been restored after immunotherapy. Collectively, the goals of the present study were the following: (i) explore the immunoregulatory genes activated in the muscles of patients with DM and sIBM using microarray technique; (ii) examine in blinded analysis which of these genes are clinically relevant, based on their correlation with clinical response; (iii) shed light on the genes most readily modified by IVIg and provide data on the mode of action of this drug not only in DM but also in various neurological and haematological conditions; and (iv) search for alteration of genes that denote muscle regeneration based on changes observed in the repeated biopsies from patients whose muscle strength had improved compared with those whose strength remained unchanged.
Material and methods
Patients
Muscle biopsies were obtained from three DM patients (three females; ages 41–48 years, average age 46) and four sIBM patients (three males and one female; ages 63–78 years, average age 73) before and after IVIg therapy, as described (Dalakas et al., 1993; Dalakas, 2004). These patients participated in controlled trials conducted at the NIH Clinical Center under Institutional Review Board-approved clinical protocols and signed informed consent. All three DM patients had a major clinical improvement and underwent a repeated muscle biopsy at the end of three IVIg infusions, when their muscle strength had normalized (Dalakas et al., 1993). The four sIBM patients were treated with IVIg infusions the same way as the DM patients, but their muscle strength had not changed when repeated muscle biopsy specimens were obtained after 3 months of therapy with IVIg (Dalakas et al., 2001). The muscle specimens were processed for routine histology, muscle enzyme histochemistry and immunocytochemistry, as reported (Dalakas et al., 1993, 2001). Muscle biopsies from two patients with non-specific symptoms and normal histology (one male and one female; ages 25 and 60) were used as pretreatment controls. All biopsies were from the biceps muscle except for one IBM patient who had pre- and post-treatment biopsies from the quadriceps.
Microarray analysis and data filtration
Total RNA from muscle biopsies were reverse-transcribed (Trizol; Invitrogen, CA, USA) and biotinylated cRNA probes were generated by in vitro transcription (Ambion, CA, USA). Fifteen micrograms of fragmented cRNA was hybridized to a Human Genome U133A array containing oligonucleotide probe sets representing transcripts derived from approximately 16 000 well-substantiated human genes (Hwang et al., 2004). The hybridized gene chips were scanned to quantitate the gene expression and data analysis was performed using Microarray Suite and Data Mining Tool (Affymetrix, Santa Clara, CA, USA) as well as Genespring (Silicon Genetics, CA, USA).
The Affymetrix U133A gene chip contains an array of 22 283 probe sets; each probe set consists of 11 oligonucleotide probes directed against the same target transcript, along with 11 control probes with single base-pair mismatches; these function as a combined unit to assess transcript levels and determine the background hybridization for each probe, thereby establishing internal controls for the hybridization signals of each gene. The detection algorithm uses probe pair intensities to generate a detection P-value and assign a ‘present’, ‘marginal’, or ‘absent’ call for each gene. The determination is based on the ‘discrimination score’ as elaborated at http://www.affymetrix.com/support/technical/technotes/statistical_reference_guide.pdf. Absent calls were used for the initial elimination of a large number of irrelevant genes that produced insignificant signals in the data set, comprising 22 283 probe sets.
The data, after normalization, were initially selected for genes that had ‘present’ or ‘marginal’ detection values, as indicated in seven or more of the 14 samples. All samples were coded and analysed blindly. The mean fold difference in expression of the selected genes between normal controls and pre- and post-treatment biopsies in DM and sIBM was calculated from pairwise fold difference in gene expression after therapy in each patient. The data were further filtered for biological relevance by selecting those genes that demonstrated more than 1.5-fold difference (Zhu et al., 2003) from the post-treatment muscles of patients with DM who had improved; the same genes were then compared with the post-treatment biopsies of sIBM patients who had not improved. Gene Ontology (GO) for molecular functional pathways was analysed with the software DAVID (Database for Annotation, Visualization and Integrated Discovery; NIAID, NIH; website URL: http://apps1.niaid.nih.gov/david/) using Affymetrix gene ID (Dennis et al., 2003). Gene Ontology represents a dynamic controlled vocabulary that can be applied to the functions of genes and proteins in all organisms even as knowledge continues to accumulate and change. Functional classifications used are those included in the Locus Report provided by the National Center for Biotechnology Information (NCBI). The Affymetrix gene ids of all the genes that demonstrated more than 1.5-fold change in DM after therapy were uploaded to the DAVID database and the software categorized the genes according to their molecular functions or biological activity. In the results presented in the tables, we looked for a consistent trend, up- or downregulation, across the DM sample sets and included only those that demonstrated the same trend in at least two of the three DM patients. Most of the data presented relied mainly on two of the three DM patients, who generated consistent data sets.
Real-time PCR
The microarray data were selectively corroborated with real-time PCR experiments. The Taqman primer/probe set was purchased from Applied Biosystems (Santa Clara, CA, USA) and the protocol used was the one provided by the manufacturer. The real-time PCR was performed using an Opticon II thermocycler (MJ Research, MA, USA). Expression levels of two representative genes, CXCL9 (Mig) and ICAM-1, were tested by real-time PCR on the muscle biopsies of DM and sIBM patients before and after therapy. These genes were selected because they cover quantitatively the spectrum of altered gene expression, from a very dramatic reduction, as seen for the CXCL9 gene in sIBM and for ICAM-1 in DM, to a marginal change, as seen for ICAM-1 in sIBM muscles (see below). Glyceraldehyde 3-Phosphate Dehydrogenase (GADPH) was used as a control for gene amplification.
Immunohistochemistry
In order to compare the microarray data with the protein expression of certain relevant genes, we performed immunohistochemistry to immunolocalize ICAM-1 as a representative gene product in the muscle biopsies of DM and sIBM patients. We used standard immunoperoxidase technique and anti-ICAM-1 antibody (Immunotech) on fresh-frozen muscle biopsy sections from DM and sIBM patients before and after IVIg therapy, as previously described (Dalakas et al., 1993).
Functional studies
Because one of the genes, anosmin-1 (KAL-1), was not only found for the first time in human muscle but was also significantly upregulated in DM and downregulated after improvement with IVIg (see Results), functional studies were performed to assess if its expression is regulated by cytokines. For these studies, the human skeletal muscle-derived cell line CCL136, which is considered parental, if not identical, to TE 671 (http://www.atcc.org), was cultured in the presence or absence of transforming growth factor (TGF)-β1 (10 ng/ml), tumour necrosis factor α (TNF-α; 1 ng/ml), interleukin (IL)-1β (10 ng/ml), interferon γ (IFN-γ; 500 U/ml) and bone morphogenetic protein 4 (BMP-4; 10 ng/ml), for 8 h in serum-free medium, as described (Nagineni et al., 2003; Raju et al., 2003). RNA was isolated and cDNA synthesized. Real-time PCR was performed using Taqman primer/probe sets for anosmin-1, and three housekeeping genes, GAPDH, β actin and Hypoxanthine phosphoribosyl transferase (HPRT). As the expression of GAPDH was consistent, the anosmin-1 expression was assessed in relation to GAPDH and the fold difference was calculated from the negative control.
Statistical analysis
The Wilcoxon matched pairs test was used to calculate the significance of the difference in the level of expression of immunological genes in DM versus sIBM and the Mann–Whitney test was used in all other cases. GraphPad Prism 4 (San Diego, CA, USA) software was used for all statistical analyses.
Results
Upregulation of genes in DM and sIBM before therapy, compared with controls
The correlation of normalized hybridization signal intensities of the two separate experimental sets from separate tissue sections of the two normal controls demonstrated a Pearson correlation coefficient of 0.93 and 0.94 respectively, confirming remarkable reproducibility of the results between experiments. A large number of immunologically relevant genes were upregulated at baseline in the muscles of both DM and sIBM patients. Among the genes that were upregulated up to 60-fold in DM and sIBM were proteasome subunit (PSMB8/LMP 7, Affy ID, 209040_s_at), STAT 1, chemokines, DEC205 and MHC class I polypeptides, confirming previous observations (Greenberg et al., 2002). Overall, the upregulation of immunologically relevant genes was more intense in sIBM than in DM. These data as well as most of the subsequent data described below were based on the analysis of two of the three DM patients as the quality of the array from the third patient was poor. Among the chemokine and cytokine-related genes that were found to be increased at least 4-fold from baseline, the level of expression was significantly higher in sIBM compared with DM (P < 0.0001; data not shown).
Altered gene expression in the muscles of DM and sIBM patients after IVIg
Among the total of 22 283 probe sets assessed, 2955 probe sets demonstrated more than 1.5-fold change in their expression level after IVIg therapy in the muscles of patients with DM. From these, genes representing 2206 probe sets were downregulated in the repeated biopsies of DM patients who improved after therapy; in contrast, 1700 of them remained unchanged in the muscles of sIBM patients who did not improve. Similarly, 749 probe sets showed upregulation of at least 1.5-fold in DM, whereas 563 of the same probe sets remained unchanged in sIBM.
To appreciate the effect of IVIg on the various genes according to their importance in the immunopathogenesis of the disease and the restoration of muscle cytoarchitecture, we categorized the differentially altered genes according to their functional pathways. We searched for altered genes associated with defined molecular and biological functions relevant to inflammatory myopathies using DAVID. We found altered expression in a number of genes related to catalytic function (661 genes), transporter (273 genes), signal transduction (239 genes), transcription regulator activities (167 genes), enzyme regulator activity (104 genes), and various genes related to immunological functions, cytokines,growth factors and chemokines. Accordingly, the following groups of transcriptional profiles were studied and compared.
A. Genes for chemokines, cytokines and growth factors
As shown in Table 1, several cytokine and growth factor genes were downregulated in the muscles of patients with DM who improved, but not in patients with sIBM who did not improve, implying biological relevance. Among these cytokines, the gene for IL-22 was mostly altered. In contrast, genes for the chemokines CCL18 and CCL13 were almost equally downregulated in both DM and sIBM (Table 2), suggesting unclear biological functions. Certain genes for inflammatory chemokines, however, such as CXCL 9 (Mig) and CXCL 11 (interferon-inducible T-cell α-chemoattractant), were, surprisingly, upregulated in the improved muscles of patients with DM, but significantly downregulated or unaltered in the muscles of patients with sIBM who did not improve (Table 2). Similarly, other genes for cytokines, growth factors and chemokines, such as IL-1α, TGF-α, CCL19 and CXCL14, also increased in DM muscles after IVIg therapy (Tables 1 and 2).
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
221165_s_at | NA | −2.96 | NA | NA | Interleukin 22 | ||
214321_at | +2.35 | −2.62 | NA | NA | Nephroblastoma overexpressed gene | ||
205110_s_at | NA | −2.27 | NA | NA | Fibroblast growth factor 13 | ||
206254_at | NA | −1.99 | NA | NA | Epidermal growth factor (β-urogastrone) | ||
217728_at | +4.61 | −1.75 | +5.01 | −2.29 | S100 calcium binding protein A6 (calcyclin) | ||
208308_s_at | NA | −1.73 | −2.11 | NA | Glucose phosphate isomerase | ||
200896_x_at | +1.7 | −1.65 | NA | NA | Hepatoma-derived growth factor | ||
202544_at | +2.75 | −1.59 | +2.09 | NA | Glia maturation factor β | ||
209466_x_at | NA | +1.5 | NA | NA | Pleiotrophin | ||
210141_s_at | NA | +1.54 | NA | +2.63 | Inhibin α | ||
207687_at | NA | +1.69 | NA | NA | Inhibin βC | ||
209101_at | 1.84 | −1.92 | NA | NA | Connective tissue growth factor | ||
208200_at | NA | +1.94 | NA | NA | Interleukin 1α | ||
205016_at | NA | +2.09 | +2.94 | NA | Transforming growth factor α | ||
207426_s_at | NA | +2.31 | NA | +1.61 | Tumour necrosis factor (ligand) superfamily member 4 | ||
210513_s_at | −2.25 | +2.71 | NA | NA | Vascular endothelial growth factor | ||
206176_at | +2.39 | +2.73 | +3.28 | NA | Bone morphogenetic protein 6 | ||
39402_at | +1.74 | +2.83 | +2.25 | NA | Interleukin 1β |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
221165_s_at | NA | −2.96 | NA | NA | Interleukin 22 | ||
214321_at | +2.35 | −2.62 | NA | NA | Nephroblastoma overexpressed gene | ||
205110_s_at | NA | −2.27 | NA | NA | Fibroblast growth factor 13 | ||
206254_at | NA | −1.99 | NA | NA | Epidermal growth factor (β-urogastrone) | ||
217728_at | +4.61 | −1.75 | +5.01 | −2.29 | S100 calcium binding protein A6 (calcyclin) | ||
208308_s_at | NA | −1.73 | −2.11 | NA | Glucose phosphate isomerase | ||
200896_x_at | +1.7 | −1.65 | NA | NA | Hepatoma-derived growth factor | ||
202544_at | +2.75 | −1.59 | +2.09 | NA | Glia maturation factor β | ||
209466_x_at | NA | +1.5 | NA | NA | Pleiotrophin | ||
210141_s_at | NA | +1.54 | NA | +2.63 | Inhibin α | ||
207687_at | NA | +1.69 | NA | NA | Inhibin βC | ||
209101_at | 1.84 | −1.92 | NA | NA | Connective tissue growth factor | ||
208200_at | NA | +1.94 | NA | NA | Interleukin 1α | ||
205016_at | NA | +2.09 | +2.94 | NA | Transforming growth factor α | ||
207426_s_at | NA | +2.31 | NA | +1.61 | Tumour necrosis factor (ligand) superfamily member 4 | ||
210513_s_at | −2.25 | +2.71 | NA | NA | Vascular endothelial growth factor | ||
206176_at | +2.39 | +2.73 | +3.28 | NA | Bone morphogenetic protein 6 | ||
39402_at | +1.74 | +2.83 | +2.25 | NA | Interleukin 1β |
Change reflects at least 1.5-fold from the pretreatment biopsy. NA = not altered. + = upregulation; − = downregulation.
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
221165_s_at | NA | −2.96 | NA | NA | Interleukin 22 | ||
214321_at | +2.35 | −2.62 | NA | NA | Nephroblastoma overexpressed gene | ||
205110_s_at | NA | −2.27 | NA | NA | Fibroblast growth factor 13 | ||
206254_at | NA | −1.99 | NA | NA | Epidermal growth factor (β-urogastrone) | ||
217728_at | +4.61 | −1.75 | +5.01 | −2.29 | S100 calcium binding protein A6 (calcyclin) | ||
208308_s_at | NA | −1.73 | −2.11 | NA | Glucose phosphate isomerase | ||
200896_x_at | +1.7 | −1.65 | NA | NA | Hepatoma-derived growth factor | ||
202544_at | +2.75 | −1.59 | +2.09 | NA | Glia maturation factor β | ||
209466_x_at | NA | +1.5 | NA | NA | Pleiotrophin | ||
210141_s_at | NA | +1.54 | NA | +2.63 | Inhibin α | ||
207687_at | NA | +1.69 | NA | NA | Inhibin βC | ||
209101_at | 1.84 | −1.92 | NA | NA | Connective tissue growth factor | ||
208200_at | NA | +1.94 | NA | NA | Interleukin 1α | ||
205016_at | NA | +2.09 | +2.94 | NA | Transforming growth factor α | ||
207426_s_at | NA | +2.31 | NA | +1.61 | Tumour necrosis factor (ligand) superfamily member 4 | ||
210513_s_at | −2.25 | +2.71 | NA | NA | Vascular endothelial growth factor | ||
206176_at | +2.39 | +2.73 | +3.28 | NA | Bone morphogenetic protein 6 | ||
39402_at | +1.74 | +2.83 | +2.25 | NA | Interleukin 1β |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
221165_s_at | NA | −2.96 | NA | NA | Interleukin 22 | ||
214321_at | +2.35 | −2.62 | NA | NA | Nephroblastoma overexpressed gene | ||
205110_s_at | NA | −2.27 | NA | NA | Fibroblast growth factor 13 | ||
206254_at | NA | −1.99 | NA | NA | Epidermal growth factor (β-urogastrone) | ||
217728_at | +4.61 | −1.75 | +5.01 | −2.29 | S100 calcium binding protein A6 (calcyclin) | ||
208308_s_at | NA | −1.73 | −2.11 | NA | Glucose phosphate isomerase | ||
200896_x_at | +1.7 | −1.65 | NA | NA | Hepatoma-derived growth factor | ||
202544_at | +2.75 | −1.59 | +2.09 | NA | Glia maturation factor β | ||
209466_x_at | NA | +1.5 | NA | NA | Pleiotrophin | ||
210141_s_at | NA | +1.54 | NA | +2.63 | Inhibin α | ||
207687_at | NA | +1.69 | NA | NA | Inhibin βC | ||
209101_at | 1.84 | −1.92 | NA | NA | Connective tissue growth factor | ||
208200_at | NA | +1.94 | NA | NA | Interleukin 1α | ||
205016_at | NA | +2.09 | +2.94 | NA | Transforming growth factor α | ||
207426_s_at | NA | +2.31 | NA | +1.61 | Tumour necrosis factor (ligand) superfamily member 4 | ||
210513_s_at | −2.25 | +2.71 | NA | NA | Vascular endothelial growth factor | ||
206176_at | +2.39 | +2.73 | +3.28 | NA | Bone morphogenetic protein 6 | ||
39402_at | +1.74 | +2.83 | +2.25 | NA | Interleukin 1β |
Change reflects at least 1.5-fold from the pretreatment biopsy. NA = not altered. + = upregulation; − = downregulation.
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
32128_at | +16.25 | −4.36 | +40.0 | −2.88 | CCL18 (PARC) | ||
206407_s_at | +19.53 | −2.16 | +29.58 | −1.9 | CCL13 | ||
214974_x_at | −1.87 | +1.57 | NA | NA | CXCL 5 | ||
203915_at | +7.90 | +1.61 | +12.88 | −7.8 | CXCL9 | ||
203666_at | NA | +1.74 | NA | NA | CXCL12 | ||
211122_s_at | +1.52 | +1.80 | +2.25 | −1.99 | CXCL11 | ||
218002_s_at | +2.89 | +1.98 | +3.49 | NA | CXCL14 | ||
210072_at | +2.15 | +2.69 | +1.98 | NA | CCL19 | ||
204470_at | −1.74 | +3.8 | NA | +1.64 | CXCL1 |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
32128_at | +16.25 | −4.36 | +40.0 | −2.88 | CCL18 (PARC) | ||
206407_s_at | +19.53 | −2.16 | +29.58 | −1.9 | CCL13 | ||
214974_x_at | −1.87 | +1.57 | NA | NA | CXCL 5 | ||
203915_at | +7.90 | +1.61 | +12.88 | −7.8 | CXCL9 | ||
203666_at | NA | +1.74 | NA | NA | CXCL12 | ||
211122_s_at | +1.52 | +1.80 | +2.25 | −1.99 | CXCL11 | ||
218002_s_at | +2.89 | +1.98 | +3.49 | NA | CXCL14 | ||
210072_at | +2.15 | +2.69 | +1.98 | NA | CCL19 | ||
204470_at | −1.74 | +3.8 | NA | +1.64 | CXCL1 |
− = downregulation; + = upregulation. PARC = pulmonary- and activation-regulated chemokine; NA = not altered.
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
32128_at | +16.25 | −4.36 | +40.0 | −2.88 | CCL18 (PARC) | ||
206407_s_at | +19.53 | −2.16 | +29.58 | −1.9 | CCL13 | ||
214974_x_at | −1.87 | +1.57 | NA | NA | CXCL 5 | ||
203915_at | +7.90 | +1.61 | +12.88 | −7.8 | CXCL9 | ||
203666_at | NA | +1.74 | NA | NA | CXCL12 | ||
211122_s_at | +1.52 | +1.80 | +2.25 | −1.99 | CXCL11 | ||
218002_s_at | +2.89 | +1.98 | +3.49 | NA | CXCL14 | ||
210072_at | +2.15 | +2.69 | +1.98 | NA | CCL19 | ||
204470_at | −1.74 | +3.8 | NA | +1.64 | CXCL1 |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
32128_at | +16.25 | −4.36 | +40.0 | −2.88 | CCL18 (PARC) | ||
206407_s_at | +19.53 | −2.16 | +29.58 | −1.9 | CCL13 | ||
214974_x_at | −1.87 | +1.57 | NA | NA | CXCL 5 | ||
203915_at | +7.90 | +1.61 | +12.88 | −7.8 | CXCL9 | ||
203666_at | NA | +1.74 | NA | NA | CXCL12 | ||
211122_s_at | +1.52 | +1.80 | +2.25 | −1.99 | CXCL11 | ||
218002_s_at | +2.89 | +1.98 | +3.49 | NA | CXCL14 | ||
210072_at | +2.15 | +2.69 | +1.98 | NA | CCL19 | ||
204470_at | −1.74 | +3.8 | NA | +1.64 | CXCL1 |
− = downregulation; + = upregulation. PARC = pulmonary- and activation-regulated chemokine; NA = not altered.
B. Adhesion molecule genes
As shown in Table 3, 25 genes associated with cell adhesion were downregulated and 13 others were upregulated by IVIg in the DM muscles. These included genes described for the first time in immune-mediated diseases, such as the Kallmann syndrome 1 (KAL-1) sequence, which was the gene most profoundly downregulated in DM but unaffected in sIBM. The KAL-1 sequence, or anosmin-1, is an extracellular matrix glycoprotein linked to Kallmann syndrome, which has chemoattractive and cell adhesion activity (Franco et al., 1991; Soussi-Yanicostas et al., 2002). The expression of this gene was substantially decreased in all the DM patients who had improved after IVIg, but it was not altered in the muscles of patients with sIBM who did not improve. Two other relevant genes differentially modified in this category were integrin β-like and ICAM-1 (Table 3). In reference to ICAM-1, it is worth mentioning that, before therapy, this molecule was upregulated 3.2-fold in DM and 7.6-fold in sIBM, compared with controls; after IVIg treatment, however, it was downregulated 2.6-fold (almost to the basal level) in DM compared with only a 1.8-fold decrease in sIBM (Fig. 1). This effect implies that the degree of downregulation in reference to baseline may be also of biological significance, as discussed later.
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
205206_at | 2.08 | −3.74 | NA | NA | Kallmann syndrome 1 sequence | ||
205422_s_at | NA | −2.66 | −1.67 | NA | Integrin β-like 1 (with EGF-like repeat domains) | ||
215485_s_at | 3.19 | −2.55 | 7.63 | −1.84 | Intercellular adhesion molecule 1 (ICAM) | ||
222073_at | 3.36 | −2.19 | 2.4 | +1.68 | Collagen, type IV α3 (Goodpasture antigen) | ||
203477_at | 1.57 | −2.17 | NA | NA | Collagen type XV α1 | ||
202351_at | 1.93 | −2.06 | NA | NA | Integrin αV (CD51) | ||
202119_s_at | 1.54 | −2.02 | NA | NA | Copine III | ||
201645_at | 16.74 | −2.01 | 14.88 | −2.8 | Tenascin C (hexabrachion) | ||
215000_s_at | NA | −2.00 | NA | NA | Fasciculation and elongation protein ζ2 (zygin II) | ||
204726_at | NA | −2.00 | NA | NA | Cadherin 13, H-cadherin (heart) | ||
216178_x_at | −2.67 | −1.84 | −14.86 | +1.52 | Integrin β1 | ||
201028_s_at | NA | −1.76 | NA | NA | CD99 antigen | ||
210495_x_at | 2.17 | −1.70 | 1.56 | NA | Fibronectin 1 | ||
208405_s_at | NA | −1.69 | NA | NA | CD164 antigen, sialomucin | ||
213071_at | 2.68 | −1.66 | 3.86 | −1.88 | Dermatopontin |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
205206_at | 2.08 | −3.74 | NA | NA | Kallmann syndrome 1 sequence | ||
205422_s_at | NA | −2.66 | −1.67 | NA | Integrin β-like 1 (with EGF-like repeat domains) | ||
215485_s_at | 3.19 | −2.55 | 7.63 | −1.84 | Intercellular adhesion molecule 1 (ICAM) | ||
222073_at | 3.36 | −2.19 | 2.4 | +1.68 | Collagen, type IV α3 (Goodpasture antigen) | ||
203477_at | 1.57 | −2.17 | NA | NA | Collagen type XV α1 | ||
202351_at | 1.93 | −2.06 | NA | NA | Integrin αV (CD51) | ||
202119_s_at | 1.54 | −2.02 | NA | NA | Copine III | ||
201645_at | 16.74 | −2.01 | 14.88 | −2.8 | Tenascin C (hexabrachion) | ||
215000_s_at | NA | −2.00 | NA | NA | Fasciculation and elongation protein ζ2 (zygin II) | ||
204726_at | NA | −2.00 | NA | NA | Cadherin 13, H-cadherin (heart) | ||
216178_x_at | −2.67 | −1.84 | −14.86 | +1.52 | Integrin β1 | ||
201028_s_at | NA | −1.76 | NA | NA | CD99 antigen | ||
210495_x_at | 2.17 | −1.70 | 1.56 | NA | Fibronectin 1 | ||
208405_s_at | NA | −1.69 | NA | NA | CD164 antigen, sialomucin | ||
213071_at | 2.68 | −1.66 | 3.86 | −1.88 | Dermatopontin |
− = downregulation; + = upregulation; NA = not altered.
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
205206_at | 2.08 | −3.74 | NA | NA | Kallmann syndrome 1 sequence | ||
205422_s_at | NA | −2.66 | −1.67 | NA | Integrin β-like 1 (with EGF-like repeat domains) | ||
215485_s_at | 3.19 | −2.55 | 7.63 | −1.84 | Intercellular adhesion molecule 1 (ICAM) | ||
222073_at | 3.36 | −2.19 | 2.4 | +1.68 | Collagen, type IV α3 (Goodpasture antigen) | ||
203477_at | 1.57 | −2.17 | NA | NA | Collagen type XV α1 | ||
202351_at | 1.93 | −2.06 | NA | NA | Integrin αV (CD51) | ||
202119_s_at | 1.54 | −2.02 | NA | NA | Copine III | ||
201645_at | 16.74 | −2.01 | 14.88 | −2.8 | Tenascin C (hexabrachion) | ||
215000_s_at | NA | −2.00 | NA | NA | Fasciculation and elongation protein ζ2 (zygin II) | ||
204726_at | NA | −2.00 | NA | NA | Cadherin 13, H-cadherin (heart) | ||
216178_x_at | −2.67 | −1.84 | −14.86 | +1.52 | Integrin β1 | ||
201028_s_at | NA | −1.76 | NA | NA | CD99 antigen | ||
210495_x_at | 2.17 | −1.70 | 1.56 | NA | Fibronectin 1 | ||
208405_s_at | NA | −1.69 | NA | NA | CD164 antigen, sialomucin | ||
213071_at | 2.68 | −1.66 | 3.86 | −1.88 | Dermatopontin |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
205206_at | 2.08 | −3.74 | NA | NA | Kallmann syndrome 1 sequence | ||
205422_s_at | NA | −2.66 | −1.67 | NA | Integrin β-like 1 (with EGF-like repeat domains) | ||
215485_s_at | 3.19 | −2.55 | 7.63 | −1.84 | Intercellular adhesion molecule 1 (ICAM) | ||
222073_at | 3.36 | −2.19 | 2.4 | +1.68 | Collagen, type IV α3 (Goodpasture antigen) | ||
203477_at | 1.57 | −2.17 | NA | NA | Collagen type XV α1 | ||
202351_at | 1.93 | −2.06 | NA | NA | Integrin αV (CD51) | ||
202119_s_at | 1.54 | −2.02 | NA | NA | Copine III | ||
201645_at | 16.74 | −2.01 | 14.88 | −2.8 | Tenascin C (hexabrachion) | ||
215000_s_at | NA | −2.00 | NA | NA | Fasciculation and elongation protein ζ2 (zygin II) | ||
204726_at | NA | −2.00 | NA | NA | Cadherin 13, H-cadherin (heart) | ||
216178_x_at | −2.67 | −1.84 | −14.86 | +1.52 | Integrin β1 | ||
201028_s_at | NA | −1.76 | NA | NA | CD99 antigen | ||
210495_x_at | 2.17 | −1.70 | 1.56 | NA | Fibronectin 1 | ||
208405_s_at | NA | −1.69 | NA | NA | CD164 antigen, sialomucin | ||
213071_at | 2.68 | −1.66 | 3.86 | −1.88 | Dermatopontin |
− = downregulation; + = upregulation; NA = not altered.
Functional role of KAL-1 (anosmin-1). To explore the role of the KAL-1 gene, we examined its expression by real-time PCR in six other muscle biopsies from DM and 15 non-autoimmune control biopsies. KAL-1 was hyperexpressed 2-fold in DM (P < 0.03) compared with non-autoimmune muscle biopsies (data not shown). The KAL-1 gene was also detected in cell types such as in vitro-cultured human skeletal muscle cells, skin fibroblasts and dermal microvascular endothelial cells. The effect of cytokines and growth factors in upregulating KAL-1 was further tested in vitro by treating human muscle cultures with TGF-β1, TNF-α, IL-1β, IFN-γ and BMP-4 and quantifying the upregulated mRNA by real-time PCR. We found a significant increase in anosmin-1 gene expression in response to TGF-β1 and BMP-4 (Fig. 2), indicating that this is a functional molecule in muscle tissue.
C. Muscle development and remodelling genes
The muscles of DM patients who improved after IVIg, but not sIBM muscles from patients who did not improve, demonstrated substantial modulation of genes encoding proteins involved in muscle architecture and the stability of muscle membrane (Table 4). The most prominent genes downregulated by IVIg were myosin binding protein H, calpain 3, γ-sarcoglycan and GalNAc-T1. Myosin binding protein H binds to myosin and is probably involved in interaction with thick myofilaments in the A-band, while GalNAc-Ts initiates mucin-type O-linked glycosylation in the Golgi apparatus by catalysing the transfer of GalNAc to serine and threonine residues on target proteins and stabilizing the muscle membrane. Among the upregulated genes in this category were tropomyosin 4, myosin and γ1 laminin (Table 4).
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
206304_at | 23.09 | −3.34 | 23.01 | −1.80 | Myosin binding protein H | ||
214475_x_at | −2.83 | −3.26 | −11.8 | +6.31 | Calpain 3, (p94) | ||
207302_at | NA | −2.82 | NA | +2.15 | Sarcoglycan, gamma | ||
201237_at | NA | −2.37 | NA | NA | Capping protein (actin filament) muscle Z-line α2 | ||
200866_s_at | −1.57 | −2.3 | −2.76 | NA | Prosaposin | ||
209541_at | +1.53 | −2.13 | NA | NA | Insulin-like growth factor 1 | ||
206394_at | NA | −1.99 | NA | +1.80 | Myosin binding protein C, fast type | ||
201722_s_at | NA | −1.92 | NA | NA | GalNAc-T1 | ||
209340_at | NA | −1.91 | NA | NA | UDP-N-acetylglucosamine pyrophosphorylase 1 | ||
221051_s_at | NA | −1.85 | NA | NA | Muscle-specific β1 integrin binding protein | ||
201438_at | +2.13 | −1.79 | NA | NA | Collagen type VI α3 | ||
205120_s_at | +1.61 | −1.73 | NA | NA | Sarcoglycan α | ||
212535_at | NA | −1.73 | NA | NA | MADS box transcription enhancer factor 2, polypeptide A | ||
206770_s_at | NA | −1.67 | NA | −1.81 | Solute carrier family 35 (UDP-N-acetylglucosamine) | ||
206115_at | +6.03 | −1.65 | +4.08 | −1.78 | Early growth response 3 | ||
209200_at | NA | −1.65 | NA | NA | MADS box transcription enhancer factor 2, polypeptide C | ||
218660_at | NA | −1.6 | NA | NA | Dysferlin, limb girdle muscular dystrophy 2B | ||
213519_s_at | NA | −1.53 | NA | NA | Laminin α2 (merosin, congenital muscular dystrophy) | ||
200770_s_at | NA | +1.52 | NA | NA | Laminin γ1 (formerly LAMB2) | ||
34471_at | +32.84 | +1.96 | +28.75 | NA | Myosin heavy polypeptide 8, skeletal muscle, perinatal | ||
212481_s_at | −2.14 | +2.03 | NA | NA | Tropomyosin 4 |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
206304_at | 23.09 | −3.34 | 23.01 | −1.80 | Myosin binding protein H | ||
214475_x_at | −2.83 | −3.26 | −11.8 | +6.31 | Calpain 3, (p94) | ||
207302_at | NA | −2.82 | NA | +2.15 | Sarcoglycan, gamma | ||
201237_at | NA | −2.37 | NA | NA | Capping protein (actin filament) muscle Z-line α2 | ||
200866_s_at | −1.57 | −2.3 | −2.76 | NA | Prosaposin | ||
209541_at | +1.53 | −2.13 | NA | NA | Insulin-like growth factor 1 | ||
206394_at | NA | −1.99 | NA | +1.80 | Myosin binding protein C, fast type | ||
201722_s_at | NA | −1.92 | NA | NA | GalNAc-T1 | ||
209340_at | NA | −1.91 | NA | NA | UDP-N-acetylglucosamine pyrophosphorylase 1 | ||
221051_s_at | NA | −1.85 | NA | NA | Muscle-specific β1 integrin binding protein | ||
201438_at | +2.13 | −1.79 | NA | NA | Collagen type VI α3 | ||
205120_s_at | +1.61 | −1.73 | NA | NA | Sarcoglycan α | ||
212535_at | NA | −1.73 | NA | NA | MADS box transcription enhancer factor 2, polypeptide A | ||
206770_s_at | NA | −1.67 | NA | −1.81 | Solute carrier family 35 (UDP-N-acetylglucosamine) | ||
206115_at | +6.03 | −1.65 | +4.08 | −1.78 | Early growth response 3 | ||
209200_at | NA | −1.65 | NA | NA | MADS box transcription enhancer factor 2, polypeptide C | ||
218660_at | NA | −1.6 | NA | NA | Dysferlin, limb girdle muscular dystrophy 2B | ||
213519_s_at | NA | −1.53 | NA | NA | Laminin α2 (merosin, congenital muscular dystrophy) | ||
200770_s_at | NA | +1.52 | NA | NA | Laminin γ1 (formerly LAMB2) | ||
34471_at | +32.84 | +1.96 | +28.75 | NA | Myosin heavy polypeptide 8, skeletal muscle, perinatal | ||
212481_s_at | −2.14 | +2.03 | NA | NA | Tropomyosin 4 |
− = downregulation; + = upregulation. NA = not altered.
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
206304_at | 23.09 | −3.34 | 23.01 | −1.80 | Myosin binding protein H | ||
214475_x_at | −2.83 | −3.26 | −11.8 | +6.31 | Calpain 3, (p94) | ||
207302_at | NA | −2.82 | NA | +2.15 | Sarcoglycan, gamma | ||
201237_at | NA | −2.37 | NA | NA | Capping protein (actin filament) muscle Z-line α2 | ||
200866_s_at | −1.57 | −2.3 | −2.76 | NA | Prosaposin | ||
209541_at | +1.53 | −2.13 | NA | NA | Insulin-like growth factor 1 | ||
206394_at | NA | −1.99 | NA | +1.80 | Myosin binding protein C, fast type | ||
201722_s_at | NA | −1.92 | NA | NA | GalNAc-T1 | ||
209340_at | NA | −1.91 | NA | NA | UDP-N-acetylglucosamine pyrophosphorylase 1 | ||
221051_s_at | NA | −1.85 | NA | NA | Muscle-specific β1 integrin binding protein | ||
201438_at | +2.13 | −1.79 | NA | NA | Collagen type VI α3 | ||
205120_s_at | +1.61 | −1.73 | NA | NA | Sarcoglycan α | ||
212535_at | NA | −1.73 | NA | NA | MADS box transcription enhancer factor 2, polypeptide A | ||
206770_s_at | NA | −1.67 | NA | −1.81 | Solute carrier family 35 (UDP-N-acetylglucosamine) | ||
206115_at | +6.03 | −1.65 | +4.08 | −1.78 | Early growth response 3 | ||
209200_at | NA | −1.65 | NA | NA | MADS box transcription enhancer factor 2, polypeptide C | ||
218660_at | NA | −1.6 | NA | NA | Dysferlin, limb girdle muscular dystrophy 2B | ||
213519_s_at | NA | −1.53 | NA | NA | Laminin α2 (merosin, congenital muscular dystrophy) | ||
200770_s_at | NA | +1.52 | NA | NA | Laminin γ1 (formerly LAMB2) | ||
34471_at | +32.84 | +1.96 | +28.75 | NA | Myosin heavy polypeptide 8, skeletal muscle, perinatal | ||
212481_s_at | −2.14 | +2.03 | NA | NA | Tropomyosin 4 |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
206304_at | 23.09 | −3.34 | 23.01 | −1.80 | Myosin binding protein H | ||
214475_x_at | −2.83 | −3.26 | −11.8 | +6.31 | Calpain 3, (p94) | ||
207302_at | NA | −2.82 | NA | +2.15 | Sarcoglycan, gamma | ||
201237_at | NA | −2.37 | NA | NA | Capping protein (actin filament) muscle Z-line α2 | ||
200866_s_at | −1.57 | −2.3 | −2.76 | NA | Prosaposin | ||
209541_at | +1.53 | −2.13 | NA | NA | Insulin-like growth factor 1 | ||
206394_at | NA | −1.99 | NA | +1.80 | Myosin binding protein C, fast type | ||
201722_s_at | NA | −1.92 | NA | NA | GalNAc-T1 | ||
209340_at | NA | −1.91 | NA | NA | UDP-N-acetylglucosamine pyrophosphorylase 1 | ||
221051_s_at | NA | −1.85 | NA | NA | Muscle-specific β1 integrin binding protein | ||
201438_at | +2.13 | −1.79 | NA | NA | Collagen type VI α3 | ||
205120_s_at | +1.61 | −1.73 | NA | NA | Sarcoglycan α | ||
212535_at | NA | −1.73 | NA | NA | MADS box transcription enhancer factor 2, polypeptide A | ||
206770_s_at | NA | −1.67 | NA | −1.81 | Solute carrier family 35 (UDP-N-acetylglucosamine) | ||
206115_at | +6.03 | −1.65 | +4.08 | −1.78 | Early growth response 3 | ||
209200_at | NA | −1.65 | NA | NA | MADS box transcription enhancer factor 2, polypeptide C | ||
218660_at | NA | −1.6 | NA | NA | Dysferlin, limb girdle muscular dystrophy 2B | ||
213519_s_at | NA | −1.53 | NA | NA | Laminin α2 (merosin, congenital muscular dystrophy) | ||
200770_s_at | NA | +1.52 | NA | NA | Laminin γ1 (formerly LAMB2) | ||
34471_at | +32.84 | +1.96 | +28.75 | NA | Myosin heavy polypeptide 8, skeletal muscle, perinatal | ||
212481_s_at | −2.14 | +2.03 | NA | NA | Tropomyosin 4 |
− = downregulation; + = upregulation. NA = not altered.
D. Immunoglobulin and complement related genes
Before therapy, the most markedly upregulated genes in this category (>7-fold difference) in both DM and sIBM (sIBM > DM), were seven genes that code for immunoglobulin chain segments (Table 5). Of interest, these immunoglobulin genes were upregulated further in DM after IVIg, but did not significantly change in sIBM (Table 5).
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
214677_x_at | +38.69 | +9.26 | +175.06 | NA | Ig λ joining 3 | ||
221671_x_at | +25.04 | +3.71 | +64.51 | NA | Human Ig rearranged γ chain mRNA | ||
211430_s_at | +22.86 | +5.54 | +68.38 | NA | Ig heavy constant λ3 | ||
215379_x_at | +16.85 | +2.82 | +37.86 | NA | Ig λ joining 3 | ||
209138_x_at | +9.73 | +4.01 | +35.89 | NA | Ig λ chain | ||
214669_x_at | +9.28 | +1.98 | +16.44 | +1.56 | Human Ig rearranged γ chain mRNA | ||
221651_x_at | +7.79 | +3.05 | +21.15 | NA | Human Ig rearranged γ chain mRNA | ||
215121_x_at | +5.07 | +3.55 | +12.7 | NA | Human rearranged Ig λ light chain |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
214677_x_at | +38.69 | +9.26 | +175.06 | NA | Ig λ joining 3 | ||
221671_x_at | +25.04 | +3.71 | +64.51 | NA | Human Ig rearranged γ chain mRNA | ||
211430_s_at | +22.86 | +5.54 | +68.38 | NA | Ig heavy constant λ3 | ||
215379_x_at | +16.85 | +2.82 | +37.86 | NA | Ig λ joining 3 | ||
209138_x_at | +9.73 | +4.01 | +35.89 | NA | Ig λ chain | ||
214669_x_at | +9.28 | +1.98 | +16.44 | +1.56 | Human Ig rearranged γ chain mRNA | ||
221651_x_at | +7.79 | +3.05 | +21.15 | NA | Human Ig rearranged γ chain mRNA | ||
215121_x_at | +5.07 | +3.55 | +12.7 | NA | Human rearranged Ig λ light chain |
− = downregulation; + = upregulation. NA = not altered.
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
214677_x_at | +38.69 | +9.26 | +175.06 | NA | Ig λ joining 3 | ||
221671_x_at | +25.04 | +3.71 | +64.51 | NA | Human Ig rearranged γ chain mRNA | ||
211430_s_at | +22.86 | +5.54 | +68.38 | NA | Ig heavy constant λ3 | ||
215379_x_at | +16.85 | +2.82 | +37.86 | NA | Ig λ joining 3 | ||
209138_x_at | +9.73 | +4.01 | +35.89 | NA | Ig λ chain | ||
214669_x_at | +9.28 | +1.98 | +16.44 | +1.56 | Human Ig rearranged γ chain mRNA | ||
221651_x_at | +7.79 | +3.05 | +21.15 | NA | Human Ig rearranged γ chain mRNA | ||
215121_x_at | +5.07 | +3.55 | +12.7 | NA | Human rearranged Ig λ light chain |
Affy ID . | Fold change in DM . | . | Fold change in sIBM . | . | Gene . | ||
---|---|---|---|---|---|---|---|
. | From control . | After IVIg . | From control . | After IVIg . | . | ||
214677_x_at | +38.69 | +9.26 | +175.06 | NA | Ig λ joining 3 | ||
221671_x_at | +25.04 | +3.71 | +64.51 | NA | Human Ig rearranged γ chain mRNA | ||
211430_s_at | +22.86 | +5.54 | +68.38 | NA | Ig heavy constant λ3 | ||
215379_x_at | +16.85 | +2.82 | +37.86 | NA | Ig λ joining 3 | ||
209138_x_at | +9.73 | +4.01 | +35.89 | NA | Ig λ chain | ||
214669_x_at | +9.28 | +1.98 | +16.44 | +1.56 | Human Ig rearranged γ chain mRNA | ||
221651_x_at | +7.79 | +3.05 | +21.15 | NA | Human Ig rearranged γ chain mRNA | ||
215121_x_at | +5.07 | +3.55 | +12.7 | NA | Human rearranged Ig λ light chain |
− = downregulation; + = upregulation. NA = not altered.
A gene relevant to DM that demonstrated a significant reduction after IVIg treatment was the complement-related C1q gene, which was upregulated 2.0-fold in DM patients, compared with controls, and was downregulated 2.0-fold after therapy. In contrast, in sIBM the C1q gene remained normal.
Corroboration of microarray data with real-time PCR and immunocytochemistry
We chose two genes, CXCL9 (Mig) and ICAM-1, to corroborate the gene array data with quantitative PCR and with protein expression in the muscles before and after therapy. The CXCL9 was chosen because it was increased 7-fold in DM and 12-fold in sIBM at baseline, but after IVIg treatment it was downregulated by 7.8-fold in sIBM and upregulated by 1.6-fold in DM. In real-time PCR, Mig expression was observed at cycle 25 in sIBM before therapy and at cycle 31 after therapy (Fig. 3); in contrast, the expression in DM before and after therapy were seen at cycles 32 and 29 respectively (Fig. 3), confirming the substantial upregulation of Mig in the IVIg-treated muscles of DM patients.
The second gene chosen was ICAM-1 because it was previously shown that its protein expression is downregulated in DM muscles after IVIg therapy (Dalakas et al., 1993). Indeed, by real-time PCR, ICAM-1 amplification in the pretreatment samples became visible at cycle 34 in sIBM but at cycle 38 in DM, indicating a higher level in sIBM. After therapy, ICAM-1 remained unchanged in sIBM; in contrast, in DM its amplification was visible by the 40th cycle, indicating downregulation of ICAM-1 equivalent to at least two cycles or 4-fold (Fig. 3), thereby confirming the gene array data (Fig. 1). Immunohistochemistry confirmed further the downregulation of ICAM-1 expression in the repeated muscles of DM patients (Fig. 4). In contrast, in sIBM where ICAM-1 gene was only slightly downregulated after IVIg, its protein expression did not change (data not shown).
Discussion
We studied muscle biopsy tissues obtained before and after IVIg therapy from patients with DM who improved after treatment and compared them with tissues from patients with sIBM who were treated the same way as DM but did not improve, based on previously published controlled trials (Dalakas et al., 1993, 2001). We found that IVIg modifies a number of genes substantially, but only certain genes involved in various functional pathways are biologically relevant. Although the sample size is not large enough to obtain true variations of each gene, the noted changes were significant enough to draw conclusions about the biological relevance of certain genes and to obtain information on gene transcription modified by IVIg. The latter could be informative in understanding how IVIg works in so many different and immunologically diverse disorders (Kazatchkine and Kaveri, 2001; Dalakas, 2004).
Compared with non-inflammatory muscle controls, at baseline the DM and sIBM muscles showed very high expression of several genes associated with immunoregulatory pathways, adhesion molecules and the cytoskeletal network, confirming previous observations (Greenberg et al., 2002). Although sIBM does not consistently respond to immunomodulating therapies and our study patients did not improve with IVIg, very high expression of inflammatory genes was noted in their muscles, at levels even higher than those seen in the muscles of patients with DM who responded to therapy. For example, chemokines such as CCL13, CCL18 and CXCL9 were significantly increased in both DM and sIBM by 8- to 30-fold, while STAT 1 was increased 25-fold in DM and 42-fold in sIBM. Cytokines, such as IL-22, the adhesion molecules ICAM-1 and anosmin-1, MHC-class I, complement C1q and various molecules associated with structural membrane or cytoskeletal genes, were also upregulated at baseline to a varying degree in both DM and sIBM.
After IVIg therapy, there was a substantial alteration in the expression of genes involved in the inflammatory network but to a different degree among patients with DM who improved compared with sIBM patients who did not improve. Of note is the significant downregulation of the HLA class I (1.6- to 2-fold) and C1q (2-fold) molecules in DM but not in sIBM, supporting the previously noted suppression of MHC-1 protein expression and marked complement inhibition in the post-IVIg muscle biopsy specimens (Basta et al., 1996; Dalakas, 2004). Among chemokine genes, the one most significantly downregulated in DM was CCL18, also known as PARC (pulmonary- and activation-regulated chemokine) (Table 2). CCL18 is one of the most abundant chemokines produced by immature dendritic cells (DCs) residing in non-lymphoid organ. After the capture of an antigen, immature DCs undergo maturation and migrate to secondary lymphoid organs, where they prime naive T cells. The upregulation of CCL18 in sIBM and DM that we noted therefore indicates an active role of immature DCs at the endomysial inflammatory sites. CCL18, a known chemoattractant for T and B cells, is selectively downregulated during the maturation process induced by lipopolysaccharide, TNF and CD40 ligand (Adema et al., 1997; Hieshima et al., 1997; Lindhout et al., 2001). The noted suppression of the CCL18 gene after IVIg therapy, especially in patients with DM who improved, suggests the involvement of DCs in the immune-mediated process of the disease. In contrast to CCL18, genes for the chemokines CXCL9, CXCL11 and CXCL14 were significantly upregulated at baseline in DM and IBM, confirming previous observations (Greenberg et al., 2002; Raju et al., 2003). The upregulation of these genes even further in the muscles of patients with DM who improved after IVIg, but not in sIBM who did not improve, implies a physiological adaptive mechanism that is probably failing in IBM muscles.
One of the mechanisms of action of IVIg is its ability to upregulate the inhibitory FCγRIIB receptor or block the activation receptor, FCγRIII, on effector cells (Samuelsson et al., 2001; Bayry et al., 2003). In our study, IVIg induced a selective and significant (2.1-fold) elevation of FCγRIIIA, but not FCγRIIB, expression only in patients with DM who improved. The significance of FCγRIIIA upregulation remains unclear.
Among the biologically relevant adhesion molecule genes modulated by IVIg were ICAM-1 and KAL-1, a neuronal adhesion molecule, attributed to Kallmann syndrome (Franco et al., 1991; Soussi-Yanicostas et al., 2002). ICAM-1, the principal ligand for β2 integrin CD11a/CD18 (LFA-1), is a functional molecule because antibodies to ICAM-1 reduce acute and chronic inflammation in a number of animal models (Carlos and Harlan, 1994). ICAM-1 is overexpressed on the non-necrotic muscle fibres and autoinvasive T cells in sIBM and the endothelial cells of perimysial arterioles and perifascicular capillaries in DM, and facilitates the transmigration of T cells across the endothelial cell wall (De Bleecker and Engel, 1994; Dalakas, 1998, 2004; Dalakas and Hohlfeld, 2003). The downregulation of ICAM-1 gene after IVIg therapy almost to the normal level only in DM, compared with a slight reduction in sIBM, as confirmed by immunohistochemistry and real-time PCR, is consistent with the dramatic suppression of inflammation noted in the repeated biopsies in DM but not sIBM patients. Because, at baseline, the ICAM-1 gene was upregulated 7.6-fold in sIBM compared with 3.2-fold in DM, the results suggest that the degree of downregulation in reference to its baseline level may be biologically relevant.
The increased expression of KAL-1 at baseline was unexpected. KAL-1, reported in adult muscle tissues for the first time, may bind to a cognate receptor, which in turn activates cytoskeletal proteins, and interacts with cell surface receptors at the extracellular matrix (ECM), such as heparan sulphate, to stabilize the cell membrane. Although KAL-1 has not been studied in immunological disorders, its marked suppression in DM but not sIBM suggests biological relevance. The KAL-1 gene appears to have functional properties in human muscle because we observed almost a 3-fold augmentation of KAL-1 mRNA expression in human muscle cells in vitro after treatment with TGF-β and BMP-4, a TGF-β family protein, but not IFN-γ, TNF-α or IL1-β (Fig. 2). A more profound upregulation of KAL-1 was also observed in fibroblasts treated with TGF-β (data not shown). Because TGF-β, a fibrogenic cytokine, is markedly increased in the connective tissue of DM patients and significantly suppressed after IVIg treatment (Amemiya et al., 2000), the KAL-1/TGF-β interaction may be involved in the fibrogenic process associated with chronic inflammation, as typically seen in the DM muscles. The connection between cell adhesion, fibrosis and TGF-β, along with their suppression after IVIg, provides the opportunity to search for genes that inhibit the fibrogenic effect of chronic inflammation. Several other ECM proteins, such as metalloproteinases, agricanase-2 (ADAM-TS5) with the TSP-1 motif, were also downregulated (1.7-fold) in DM, but not sIBM, after IVIg therapy. Aggrecanases such as ADAM-TS4 and ADAM-TS5 are key matrix-degrading enzymes of the ECM that have been found to be increased in rheumatoid arthritis synovium (Malfait et al., 2002). Of interest, some of the cell surface glycosaminoglycans are also necessary for anosmin-1 function to stabilize the composition of the ECM. Collectively, KAL-1 and other ECM genes may be key molecules that play a regulatory role in the cellular structure and signal transduction network that activates the various ECM proteins in human muscle.
Some genes involved in the cytoskeletal network were downregulated in the muscles of DM patients who improved, but not in sIBM (Table 4). During the active inflammatory process there might have been an enhanced compensatory mechanism to restore the membrane network of the cytoskeletal proteins, but after successful treatment such enhanced molecular activities may have become subdued. In this context, the decreased gene expression of calpain 3, myosin binding protein H or GalNAc-T1 may not be surprising. For example, calpain 3, which was downregulated more than 3-fold (Table 4) in DM, but not sIBM, is considered a modulator protease, involved in various cellular processes. Calpain-mediated remodelling of cytoskeletal membrane interactions, such as those occurring during myoblast fusion and muscle repair, may be involved in regulation of muscle-specific filamin C–sarcoglycan interactions (Guyon et al., 2003). Consistent with the reduction in calpain 3 was the observed suppression in sarcoglycan γ (2.8-fold), sarcoglycan β (1.7-fold) and N-acetylgalactosamine-4-O-sulphotransferase (GALNAC- 4-ST1), an enzyme that adds sulphate to the non-terminal GalNAc residues in chondroitin or dermatan (Xia et al., 2000; Kang et al., 2001).
Previous studies (Greenberg et al., 2002) have demonstrated increased immunoglobulin gene expression in both IBM and DM, with greater expression in IBM. The augmentation of immunoglobulin gene transcription after IVIg therapy in the muscles of DM patients who improved is intriguing. Whether these immunoglobulins are due to a rebound effect or are associated with muscle repair and regeneration remains unclear.
In conclusion, IVIg administration modulates a subset of clinically relevant immunoregulatory or structural genes associated with clinical improvement and restoration of muscle cytoarchitecture. The effect is most likely due to IVIg and not to a difference in the pathogenic mechanisms between DM and IBM or to the different ages of the patients studied, because several of the same genes were equally upregulated in both groups before therapy but were altered only in patients with DM who improved. Gene expression profiling may therefore help us identify markers associated with response to therapies and point to the genes that are most significant in the immunopathogenesis of inflammatory myopathies. Notwithstanding its importance, however, our study has certain limitations. Various stages of data filtration and normalization strategies were employed to study the genes considered relevant among the total of over 20 000 probe sets. Though this process is expected to yield the most reliable information on a wide array of genes involved in different functional pathways, it is also possible that several important genes might have been overlooked. In addition, the sample size was very small. However, this is a unique group of patients who have improved (or not improved) with the same drug and repeated biopsies were obtained with an unbiased method before the study code was broken (Dalakas et al., 1993; Dalakas, 2004). Because the clinical results were robust, our observations, even in a very small number of specimens, clearly point towards genes of important biological relevance.
The authors thank Catherine Campbell, Bioinformatics Division, NINDS, for help with data analysis, C. Nagineni (NEI, NIH) for the in vitro-cultured and treated epithelial cells, and Wei Shi for some technical assistance.
References
Adema GJ, Hartgers F, Verstraten R, de Vries E, Marland G, Menon S, et al. A dendritic-cell-derived CC chemokine that preferentially attracts naive T cells.
Amemiya K, Semino-Mora C, Granger RP, Dalakas MC. Downregulation of TGF-b mRNA and protein in the muscles of patients with inflammatory myopathies after treatment with high-dose intravenous immunoglobulin.
Basta M, Illa I, Dalakas MC. Increased in vitro uptake of the complement C3b in the serum of patients with Guillian-Barre Syndrome, myasthenia gravis and dermatomyositis.
Bayry J, Misra N, Latry V, Prost F, Delignat S, Lacroix-Desmazes S, et al. Mechanisms of action of intravenous immunoglobulin in autoimmune and inflammatory diseases.
Dalakas MC. Molecular immunology and genetics of inflammatory muscle diseases.
Dalakas MC. Intravenous immunoglobulin in autoimmune neuromuscular diseases.
Dalakas, MC, Dambrosia JM, Soueidan SA, Stein DP, Otero C, Dinsmore ST, et al. A controlled trial of high-dose intravenous immunoglobulin infusions as treatment for dermatomyositis.
Dalakas MC, Koffman B, Fuji M, Spector S, Sivakumar K, Cupler E. A controlled study of intravenous immunoglobulin combined with prednisone in the treatment of IBM.
De Bleecker JL, Engel AG. Expression of cell adhesion molecules in inflammatory myopathies and Duchenne dystrophy.
Dennis G Jr, Sherman BT, Hosack DA, Yang J, Baseler MW, Lane HC, et al. DAVID: Database for Annotation, Visualization, and Integrated Discovery.
Engel AG, Hohlfeld R. The polymyositis and dermatomyositis syndromes. In: Engel AG, Franzini-Armstrong C, editors. Myology. New York: McGraw-Hill;
Franco B, Guioli S, Pragliola A, Incerti B, Bardoni B, Tonlorenzi R, et al. A gene deleted in Kallmann's syndrome shares homology with neural cell adhesion and axonal path-finding molecules.
Greenberg SA, Sandoudou D, Haslett JN, Kohane IS, Kunkel LM, Beggs AH, et al. Molecular profiles of inflammatory myopathies.
Guyon JR, Kudryashova E, Potts A, Dalkilic I, Brosius MA, Thompson TG, et al. Calpain 3 cleaves filamin C and regulates its ability to interact with gamma- and delta-sarcoglycans.
Hieshima K, Imai T, Baba M, Shoudai K, Ishizuka K, Nakagawa T, et al. A novel human CC chemokine PARC that is most homologous to macrophage-inflammatory protein-1 α/LD78 and chemotactic for T lymphocytes, but not for monocytes.
Hwang K-B, Kong SW, Greenberg SA, Park PJ. Combining gene expression data from different generations of oligonucleotide arrays.
Kang HG, Evers MR, Xia G, Baenziger JU, Schachner M. Molecular cloning and expression of an N-acetylgalactosamine-4-O-sulfotransferase that transfers sulfate to terminal and non-terminal beta 1,4-linked N-acetylgalactosamine.
Kazatchkine MD, Kaveri SV. Immunomodulation of autoimmune and inflammatory diseases with intravenous immune globulin.
Lindhout E, Vissers JLM, Hartgers FC, Huijbens RJF, Scharenborg NM, Figdor CG, et al. The dendritic cell-specific CC-chemokine DC-CK1 is expressed by germinal center dendritic cells and attracts CD38-negative mantle zone B lymphocytes.
Malfait AM, Liu RQ, Ijiri K, Komiya S, Tortorella MD. Inhibition of ADAM-TS4 and ADAM-TS5 prevents aggrecan degradation in osteoarthritic cartilage.
Nagineni, CN, Samuel W, Nagineni S, Pardhasaradhi K, Wiggert B, Detrick B, et al. Transforming growth factor-beta induces expression of vascular endothelial growth factor in human retinal pigment epithelial cells: involvement of mitogen-activated protein kinases.
Raju R, Vasconcelos O, Granger R, Dalakas MC. Expression of IFN-g inducible chemokines in inclusion body myositis.
Samuelsson A, Towers TL, Ravetch JV. Anti-inflammatory activity of IVIg mediated through the inhibitory Fc receptor.
Soussi-Yanicostas N, de Castro F, Julliard AK, Perfettini I, Chedotal A, Petit C. Anosmin-1, defective in the X-linked form of Kallmann syndrome, promotes axonal branch formation form olfactory bulb output neurons.
Tezak Z, Hoffman EP, Lutz JL, Fedczyna TO, Stephan D, Bremer EG, et al. Gene expression profiling in DQA1*0501+ children with untreated dermatomyositis: a novel model of pathogenesis.
Zhu G, Reynolds L, Crnogorac-Jurcevic T, Gillett E, Dublin EA, Marshall JF, et al. Combination of microdissection and microarray analysis to identify gene expression changes between differentially located tumour cells in breast cancer.