Changes of myocardial gene expression and protein composition in patients with dilated

Transcription

1 Online Resource: Electronic Supplemental Material 5 Changes of myocardial gene expression and protein composition in patients with dilated cardiomyopathy after immunoadsorption with subsequent immunoglobulin substitution Sabine Ameling 1,7#, Gourav Bhardwaj 1,7#, Elke Hammer 1,7, Daniel Beug 2, Leif Steil 1, Yvonne Reinke 2,7, Kerstin Weitmann 3,7, Markus Grube 4, Christiane Trimpert 2, Karin Klingel 5, Reinhard Kandolf 5, Wolfgang Hoffmann 3,7, Matthias Nauck 6,7, Marcus Dörr 2,7, Klaus Empen 2, Stephan B. Felix 2,7,* and Uwe Völker 1,7,* 1 Interfaculty Institute for Genetics and Functional Genomics, University Medicine Greifswald, Greifswald, Germany 2 Department of Internal Medicine B (Cardiology), University Medicine Greifswald, Greifswald, Germany 3 Institute for Community Medicine, University Medicine Greifswald, Greifswald, Germany 4 Center of Drug Absorption and Transport (C_DAT), Department of Pharmacology, University Medicine Greifswald, Greifswald, Germany 5 Department of Molecular Pathology, University Hospital Tübingen, Germany 6 Institute of Clinical Chemistry and Laboratory Medicine, University Medicine Greifswald, Greifswald, Germany 7 DZHK (German Centre for Cardiovascular Research), partner site Greifswald, Greifswald, Germany # both authors contributed equally * Corresponding authors: Uwe Völker, PhD Interfakultäres Institut für Genetik und Funktionelle Genomforschung, Universitätsmedizin Greifswald Friedrich-Ludwig-Jahn-Straße 15a D Greifswald, Germany Phone: Fax: voelker@uni-greifswald.de Stephan B. Felix Klinik und Poliklinik für Innere Medizin B Universitätsmedizin Greifswald Ferdinand-Sauerbruch-Straße D Greifswald, Germany Phone: Fax: felix@uni-greifswald.de

2 ESM5 Supplemental Figure S1 Supplemental Figure 1 qrt-pcr measurements of HF marker genes. Mean of normalized relative quantities and standard deviation of natriuretic peptides A (NPPA) and B (NPPB), corin (CORIN), natriuretic peptide receptor 3 (NPR3), angiotensin converting enzyme 2 (ACE2) and periostin (POSTN) were displayed for responders (R=9) and non-responders (NR=5) at baseline (BL, solid bars) and follow up (FU, open bars) IA/IgG. A significant IA/IgG effect on gene expression was assumed with p-value <0.05 (rankproduct test based on FU/BL ratios). Differences in gene expression between responders and non-responders at BL were analysed using Mann-Whitney-t-test (p<0.05).

3 Supplemental Figure S2 Supplemental Figure 2 Overlap of genes affected by IA/IgG in responders and non-responders.

4 Supplemental Figure S3 Supplemental Figure 3 Linear model with adjustment for the co-factors inflammation and fibrosis score. Displayed are p-values for molecules found to be altered significantly upon therapy (impact of therapy, FU vs. BL) without adjustment (x-axis) or with adjustment for inflammation (number of CD3 + T- lymphocytes and /or CD68 + macrophages) or fibrosis grade, respectively (y-axis). The linear model with adjustment for the factor (A) inflammation revealed a decrease in the number of significant p- values in responders but not non-responders. The linear model adjusted for the factor (B) fibrosis grade revealed a decrease in the number of significant p-values in responders and marginally in nonresponders.

5 Supplemental Figure S4 Supplemental Figure 4 IA/IgG affects transforming growth factor beta signalling in responders. Based on the observed changes in gene expression and protein levels in myocardial biopsies of R and NR due to IA/IgG, as well as differences between the subgroups, Upstream analysis in Ingenuity pathway analysis software predicted a lower expression/ activity of transforming growth factor beta in responders than in non-responders.

6 Supplemental Figure S5 Supplemental Figure 5 Overlap of DCM-associated genes and genes affected upon IA/IgG. Upor down-regulated genes of A) responders and B) non-responders upon IA/IgG were depicted. The overlaps of altered genes after IA/IgG with DCM associated genes (R vs. Co (Control), NR vs. Co) in the respective subgroups (Ameling et al, 2013) are rather small.

7 Supplemental Figure S6 Supplemental Figure 6 Selection of literature based genes in responders and non-responders at baseline (BL) and 6 months after IA/IgG (FU). Mean of normalized signal intensities and standard deviation for responders (R, n=20) and non-responders (NR, n=13) before (BL, solid bars) and after (FU, open bars) IA/IgG are displayed for selected genes measured on the HG-U-133-Plus-2.0 Affymetrix array. Brain (NPPB) and atrial (NPPA) natriuretic peptides, myosin heavy chain 6 (MYH6)

8 and 7 (MYH7), sarcoplasmatic calcium ATPase (ATP2A2), phospholamban (PLN) and corresponding ratios are depicted. Significant differentially expressed genes in the comparison of each subgroup (responder, non-responder) at baseline or follow-up were assumed by a p-value <0.05 (Mann- Whitney-U-test). Significant altered genes in responders or non-responders were assumed by q<0.05 based on rankproduct test using FU/BL ratios of each subgroup (q FDR-value, ns. - not significant).

9 Supplemental Figure S7 100 LVEF % disease duration (months) Supplemental Figure 7 Disease duration in the subgroups responders and non-responders as defined by their relative LVEF change ( LVEF %).

10 Supplemental Figure S8 Supplemental Figure 8 Overlap of transcriptome and proteome analyses. A) Around 90% of 669 proteins identified showed an overlap with annotated genes (present in 80% of individuals, Responders genes, Non-responders genes). B) Protein profiling revealed data for 5.6 % (R) and 5.5 % (NR) of the genes measured. The overlap of significantly altered proteins and genes in IA/IgG treated responders (R FU/BL) and non-responders (NR FU/BL) was marginal. C) The comparison R FU/BL vs. NR FU/BL ratios revealed an overlap of 7 proteins/genes. ID mapping of identified proteins and genes was performed on gene symbols which were derived from NETAFFX 10/2013 according to HG-U133-Plus-2.0 array (affymetrix) and UniProt ( respectively. For approximately 90 percent of the 669 proteins identified (Figure 8A) gene expression levels were measured. The proteins not covered in the transcriptome measurements include antibody fragments, plasma proteins obviously not synthesized in the heart but also proteins whose transcripts were not significantly detected in the minimum of 80 % of the patients (either in R or NR). Furthermore, the protein profiling revealed only data for 5.6 % (R) and 5.5 % (NR) of the transcripts measured. For significantly altered genes and proteins a very low overlap was found for each subgroup (1 candidate for each patient subgroup, Figure 8B). Other proteins found to be altered and covered by gene expression analyses (R -11 and NR -33) were mostly not changed in their transcript level at all. It is known that mrna and protein levels do not essentially correlate, because the latter one depend e.g. on mrna stability, translation rate, protein degradation rate and other factors. In contrast to the rather moderate alterations found for the patient subgroups R and NR (FU vs BL), differences became more obvious when the IA/IgG effect between the subgroups was compared (Figure 8C). The number of molecules in the overlap increased (7) and further 2 of the 5 altered genes covered by the proteome profiling showed the same trend (HBA1/HBA2 FC 1.22 and transgelin FC 1.29 p < 0.05, (cut-off protein FC 1.3).Thus, gene expression and protein analyses data overlap only

11 to a minor extent but the results point to a common theme - a difference in the myocardial remodelling and fibrosis.