Supplementary Figure 1

Transcription

1 Supplementary Figure 1 Supplementary Figure 1. ADARs expression and efficiency of ADAR1 knockdown in endothelial cells. (a) Expression of ADAR1, ADAR2 and ADAR3 in HUVECs according to RNA sequencing (RNA-seq). ADAR3 was not detected in HUVECs. The mean expression and ± s.e.m of two independent experiments are depicted. (b) Immunostaining showing ADAR1 (green) and nuclear localization (nuclei: DAPI, blue) in HUVECs. F-actin (red) was used to stain the cytoskeleton. Scale bar: 50 µm. (c) qrt-pcr validation of ADAR1 expression in HUVECs and ADAR1 knockdown with short-interfering RNA targeting ADAR1 (n = 4 biological independent replicates). (d) siadar1-transfected endothelial cells were subjected to RNAsequencing and the efficiency of ADAR1 silencing is shown in FPKM values. ADAR1 expression is 78% reduced, while siadar2 showed no effect on ADAR1 mrna expression (n = 2). Data are mean ± s.e.m. of n independent biological experiments as stated. Two experimental groups were compared using a two-tailed non-parametric Mann-Whitney U-test. *P<0.05 vs. scrambled or control.

2 Supplementary Figure 2 Supplementary Figure 2. Knockdown efficiency of two additional sirnas targeting ADAR1 and effect on angiogenesis in vitro. (a) qrt-pcr validation of ADAR1 expression in HUVECs and specific ADAR1 knockdown with two additional short-interfering RNAs against ADAR1 (n = 3-4). (b) Representative images and statistical summary of 3D in vitro angiogenesis assay with two additional ADAR1 or scrambled sirna-transfected ECs. The mean sprout length was quantified (n = 3). (c) Representative images and quantification of 2D capillary tube formation assay with two additional ADAR1 or scrambled sirna-transfected ECs. The mean angiogenic net length was quantified (n = 3). Data are mean ± s.e.m. of n independent biological experiments as stated. Two experimental groups were compared using a two-tailed nonparametric Mann-Whitney U-test. *P<0.05 vs. scrambled or control.

3 Supplementary Figure 3 Supplementary Figure 3. Efficiency of ADAR2 knockdown in endothelial cells. (a) RNAsequencing data of ADAR2 expression after transfection of endothelial cells with siadar2. siadar1 does not affect the expression of ADAR2 mrna (n = 2). (b) qrt-pcr of ADAR2 expression in HUVECs after ADAR2 knockdown with sirna targeting ADAR2. HUVECs were transfected with Scrambled, siadar1 or siadar2 (55 nm) for 48 h and ADAR2 mrna expression was quantified by qrt-pcr (n = 3). (c) ADAR2 protein expression in HUVECs after ADAR2 knockdown using sirna in HUVECs 72 h after transfection. Data are mean ± s.e.m. of n independent biological experiments as stated. Two experimental groups were compared using a two-tailed non-parametric Mann-Whitney U-test. *P<0.05 vs. scrambled or control.

4 Stellos K, et al. Supplementary Information Supplementary Figure 4 Supplementary Figure 4. CTSS 3'UTR mrna contains many HuR-binding motifs. Predicted HuR-binding motifs within the 3'UTR of CTSS mrna are depicted in yellow, validated RNA edited sites are shown in red font and predicted RNA edited sites (according to DARNED) are depicted in blue. The two Alu regions, AluJo (genomic coordinates: ) and AluSx+ (genomic coordinates: ), are indicated in black font.

5 Supplementary Figure 5 Supplementary Figure 5. Efficiency of HuR knockdown in endothelial cells and effect of ADAR1 knockdown on HuR expression. (a, b) Western blot analysis of the presence of HuR protein in endothelial cells transfected with scrambled or sihur. (b) Relative quantification of HuR protein levels obtained by densitometric analysis of the western blot after normalization with anti-tubulin (n = 3). (c) HuR silencing regulates CTSS mrna expression. (d-f) RNA-seq analysis (d), RT-qPCR (e) and western blot (f) showing that ADAR1 knockdown does not affect HuR mrna expression and protein expression in endothelial cells. Data are mean ± s.e.m. of n independent biological experiments as stated. Two experimental groups were compared using a two-tailed non-parametric Mann-Whitney U-test. *P<0.05 vs. scrambled or control.

6 Supplementary Figure 6 Supplementary Figure 6. Recombinant HuR preferentially binds to in vitro transcribed and edited CTSS 3'UTR RNA. (a) The in vitro transcribed CTSS 3'UTR RNA from the T7 promoterdriven construct was subjected to in vitro RNA editing assay in order to generate the unedited and the edited form of the RNA (see Methods for details). The in vitro RNA editing efficiency reaction is shown as mean RNA editing rate per edited nucleotide within the CTSS 3'UTR (n = 3-5). (b) Both unedited and edited recombinant CTSS RNAs were incubated with recombinant HuR protein. In one set of experiments (d) HeLa nuclear extracts were added in order to provide a more physiological environment in the reaction. Then, HuR was immunoprecipitated and the respective RNA was recovered, reversed transcribed and 3'UTR of CTSS was detected by qrt- PCR. (c, d) Detection of 3'UTR of CTSS in HuR-immunoprecipitates after incubation of the recombinant RNA with HuR or control vehicle in the absence (c) or presence (d) of HeLa nuclear extracts (n/e). In both cases the fold change (vs. control) of the HuR-binding of the edited form of the 3'UTR CTSS RNA is increased in comparison to the unedited one. The mean expression and ± s.e.m of two independent experiments are depicted.

7 Supplementary Figure 7 Supplementary Figure 7. iclip maps RNA interactions of HuR protein in living cells. HUVEC cells were transfected with ADAR1- or scrambled sirna (+/ ADAR1-knockdown). (a) After UV crosslinking HuR was immunoprecipitated and samples were subjected to RNase A treatment, using two different RNase concentrations (H high, L low RNase). The coimmunoprecipitated RNA was dephosphorylated, ligated with a 3'-RNA linker and 5'- radiolabeled. Samples were size-separated by SDS-PAGE and transferred to a nylon membrane.

8 Protein/RNA-complexes were visualized by autoradiography. The boxed regions were cut from the membrane, proteins were digested with Proteinase K and RNA was subjected to iclip library preparation and sequencing (see Materials and Methods for details). (b) Bioinformatic analysis of the iclip sequencing data showing the total number of the uniquely mapped reads in scrambled (control) and siadar1-transfected cells. (c,d) Snapshot from Integrate Genome Viewer (IGV) showing the overview of the reads from the iclip sequencing data lying within the 3'UTR of CTSS region. In total 7 regions containing known HuR binding sites were identified (region 3 and 4 are shown in larger magnification along with the respective genomic coordinates, lower part) in scrambled-transfected HUVECs. The presence of HuR binding sites is reduced in siadar1-transfected HUVECs. (e) Quantification of the reduction of the mean number of reads present in all the 7 identified HuR binding regions on 3'UTR CTSS mrna compared to scrambled transfected HUVECs.

9 Supplementary Figure 8 Supplementary Figure 8. HuR interacts with seven regions of the 3'UTR of CTSS in HUVECs. Sequences obtained from iclip sequencing data are highlighted in yellow with the known HuR binding motifs in green. A-to-I RNA editing events are marked in red. Two regions (Region 6 and 7) belong to the AluSx with 4 predicted edited sites (marked in blue font) according to DARNED database. One region (Region 4) belongs to the AluJo + and two other regions (Region 1 and 2) belong to the AluSx + which is the most highly edited region in CTSS 3'UTR. One region (Region 3) is located between the edited AluJo + (AluJo) and AluSx + regions.

10 Supplementary Figure 9 Supplementary Figure 9. Human and mouse 3'UTR of CTSS pre-mrna is not conserved. (a) Snapshot obtained from Ensemble database showing the comparative analysis of human and murine CTSS gene. (b) Murine 3'UTR of CTSS mrna is much shorter than the human one (human vs. murine 3'UTR length: 2850 vs. 220 nucleotides).

11 Supplementary Figure 10 Supplementary Figure 10. Nucleotides found to be commonly highly edited at the 3'UTR CTSS mrna among the in vitro studies, the cells and human tissues. Editing rate determined by RNA editing studies (see Methods for details) of the commonly edited nucleotides in CTSS 3'UTR of (a) HUVECs, (b) the in vitro transcribed (IVT) edited transcript (n = 3-5), (c) the luciferase construct (whole 3'UTR cloned into a luciferase expressing vector- see Methods for details, n = 6), (d) PBMCs from patients with coronary artery disease (stable CAD and AMI: total n = 45), (e) Carotid atherosclerotic plaques from the BiKE cohort (n = 20), (f) Thoracic aortic aneurysm tissues (n = 35).

12 Supplementary Figure 11 Supplementary Figure 11. HuR binding motif enrichment in 3'UTRs of edited transcripts compared to all transcripts, especially in regions in close proximity to RNA editing sites. (a) Bioinformatic analysis of the mean the number of HuR binding motifs (ATTTA, TTTTT, TTTTG) in 3'UTR-edited transcripts and total transcripts divided by the respective transcript number in endothelial cell transcriptome. The edited transcripts in HUVEC ribo-minus RNA-seq data were identified as described in the Methods. (b) Bioinformatic analysis of the absolute number of HuR binding motifs in the neighboring up- and downstream 100nt-sequences of each edited site observed in the endothelial cell transcriptome (ribo-minus RNA-seq).

13 Supplementary Figure 12 Supplementary Figure 12. Schematic illustration of the impact of RNA editing on cathepsin S mrna expression during homeostasis and in disease. ADAR1 binds to double-stranded RNAs inducing A-to-I RNA editing events in Alu elements present in 3'UTR of CTSS mrna. RNA editing alters mrna secondary structure and binding of the stabilizing RNA-binding protein HuR. Hypoxia and atherosclerotic heart disease upregulate A-to-I RNA editing. The increased RNA editing under pathophysiologic conditions is accompanied by higher CTSS expression levels. Conditions and human diseases that are linked to high CTSS levels are inflammatory diseases like autoimmune and cardiovascular diseases, atherosclerotic plaque vulnerability with subsequent initiation of thrombosis and acute ischemic events (acute myocardial infarction,

14 ischemic stroke, acute peripheral arterial thrombosis), arterial aneurysm formation with subsequent dissection and bleeding (hemorrhagic stroke, aortic aneurysm), tumor angiogenesis, growth and metastasis and finally tumor and cardiovascular mortality.

15 Supplementary Table 1 A-to-I RNA Editing of Cathepsin S mrna Blue predicted according to DARNED, but not confirmed editing site Green true according to Sanger sequencing, but not predicted editing site Red both predicted and confirmed editing site 3'UTR sequence of CTSS mrna (5'è 3') / A nucleotide is marked aggaucucuccuuuuuauaacaaaucaagaaauaugaagcacuuucucuu 50 aacuuaauuuuuccugcuguauccagaagaaauaauugugucaugauuaa 100 uguguauuuacuguacuaauuagaaaauauaguuugaggccgggcacggu 150 ggcucacgccuguaaucccaguacuugggaggccaaggcaggcauaucaa 200 cuugaggccaggaguuaaagagcagccuggcuaacauggugaaaccccau 250 cucuacuaaaaauacaaaaaauuagccgagcacgguggugcaugccugua 300 aucccagcuacuugggaggcugaggcacgagauuccuugaacccaagagg 350 uugaggcuauguugagcugagaucacaccacuguacuccagccuggauga 400 cagaguggagacucuguuucaaaaaaacagaaaagaaaauauaguuugau 450 ucuucauuuuuuuaaauuugcaaaucucaggauaaaguuugcuaaguaaa 500 uuaguaauguacuauagauauaacuguacaaaaauuguucaaccuaaaac 550 aaucuguaauugcuuauuguuuuauuguauacucuuugucuuuuuaagac 600 cccuaauagccuuuuguaacuugauggcuuaaaaauacuuaauaaaucug 650 ccauuucaaauuucuaucauugccacauaccauucuuauuccuaggcaac 700 uauuaauaaucuauccugagaauauuaauugugguauucuggugaugggg 750 uuuagcaacuuugauggaagaaaauauuaggcuauaaauguccuaaggac 800 ucagauuguaucuuuguacagaagaggauucaaaacgccacguguagugg 850 cucaugccuguaaucccaacacuuugggaggcugaaguaggaggaucguc 900 uugagcccaggaguucaagaccagccuggacaacauagugagaccuuguc 950 uccacaaaaauaaaaaagaaacuauccaggagugguggugugugccugug 1000 gucccugcuaugcagaugucuaagacaggaggaucacaagagcccaggag 1050 guugagaaugcagugagcuuguaauugcaccacugcacuccagccugggu 1100 gacagagcaagacccugucuuaaaaaaagaggauucaacacauauuuuua 1150 uauuauguuaaaguaaagaaaugcauaaaagacaagcacuuuggaagaau 1200 uauuuuaaugaucaacaauuuaauguauuaguccaaauuauuuuuacgua 1250 gucaucaacaauuugaccagggccuuuauuuggcaaauaacugagccaac 1300 cagaauaaaauaaccaauacuccacugcucauauuuuuaucuaauucaga 1350 uggaucuuccuuacaacugcucuagauuaguagaugcaucuaagcaggca 1400 gcaggaacuuuaaauuuuuuaaguucaugucuaugacaugaacaaugugu 1450

16 gggauaaugucauuaauauauccuaaauuaaccuaaacguauuucacuaa 1500 cucuggcuccuucuccauaaagcacauuuuaaggaacaagaauugcuaaa 1550 uauaaaaacauaaauaauaccauaauacauggcuaucaucaaaaguguau 1600 agaauauuauaguuuaaaaguauuuaguugauuacuuuucaguuuuguuu 1650 uguuuuuugagacggagucucacucuguugcccaggcuggagugcagugg 1700 caccaucucaguucacugcaacuucugccucccgaguucaagcgauucuc 1750 cugccucagccucccgaguagcuggaauuauaggcgugcaccaccacgcc 1800 cagcuaauuuuuguauuuuuaguaaagacaggguuuugccacauuagcca 1850 ggcuggucucaaacuccugaccucaggugauccacccaccccagccuccc 1900 aaagugcuaagauuacaggcgugagccacugagcccagccuacuuuucag 1950 uuuuuaacauaauuuuuguuuuauccacaacuuuucaaguauugaaagua 2000 gaauaaaaacauggguucuuagucuuuagcuaucuguuaaagccuaugaa 2050 ugccuucuuaaaaucauguuuuuaaaugcauaaaauauauaggauuacaa 2100 aggaaucuaauuauaucgaaauacaguuauuaaaauguuaaaagauaagu 2150 uuguuauauauuaauaugcaugcuucuuuauaaaugcauuaaauaagagu 2200 uaauagcuauccuaaauuugaaauagugauaagcauaaugaaaauagaug 2250 caaaaaacuaaugugauaugaaaauaucuggguuuuucuuuugaugauga 2300 aguauugcuaauauuaccgugguuuaugaacuauguucagaauugaagaa 2350 aauccuaacuuucaguuagagguuagugacgggguucaggacacccuaca 2400 caaaauacagcacuuugacauauugaauauuuuaagcugaaggcauuuga 2450 ggaaauugcagaagcaggaaggugacucugaccuucugccugcuguucuc 2500 cccagaagcagccauaaaaccugggaaggauuuucugaccuuccccugaa 2550 guagaucauaagacugucauguaagaggugcucuccuggcacccagagaa 2600 aaggagcauccuuaccuccaaaagcacagggacacaaagaggaaucuaaa 2650 caaacaggccucucaguuucccccaguuuauuacauuuagcuuguucaca 2700 cuuugcccuaugacauuucuacaucacuggcugcucuucaucaaaccuac 2750 uauaaaaaacauucaaguucaacuguuucuuugggccuuuauuuccuuau 2800 ggagccccucgugucguguaaaacuuauauuaaauaaaugugcaugcuuu 2850

17 Supplementary Table 2. Alignment of sequencing products after cloning of the edited AluJo and AluSx + followed by Sanger sequencing. Note: The editing rate per nucleotide position is noted in green font.

18 Supplementary Table 3. Characteristics of the Pilot PBMC RNA-seq cohort. Parameter 1 Overall Controls CAD ICM *P-value n=12 n=4 n=4 n=4 Age, years - 29 (2.8) 58.5 (7.5) 62.3 (14.5) Male gender, n n.a. RNA deep sequencing: 3 UTR RNA editing of cathepsin S mrna CTSS mrna (FPKM) 308 (136) 238 (86) 328 (189) 357 (119) CV risk factors Arterial hypertension 8 (66.7) 0 (0) 4 (100) 4 (100) Diabetes mellitus 3 (25) 0 (0) 2 (50) 1 (25) Smoking 7 (58.3) 2 (50) 2 (50) 3 (75) Hyperlipidemia 8 (66.7) 0 (0) 4 (100) 4 (100) BMI, kg/m (3.7) 23 (1.2) 26.9 (3.5) 27.1 (4.8) Clinical parameters Cholesterol, mg/dl 161 (44) 190 (35) 122 (37) 171 (35) HDL cholesterol, mg/dl 49 (12) 58 (8) 45 (11) 44 (14) LDL cholesterol, mg/dl 88 (36) 112 (32) 56 (29) 95 (28) TGL, mg/dl 122 (49) 102 (46) 104 (32) 162 (51) hscrp, mg/dl 0.2 (0.3) 0.06 (0.07) 0.07 (0.03) 0.45 (0.36) Creatinine, mg/dl 0.99 (0.12) 0.99 (0.14) 0.91 (0.07) 1.07 (0.11) HbA1c, % 5.7 (0.8) 5 (0.3) 6.1 (0.8) 5.9 (0.7) NT-proBNP, pg/ml 310 (484) 34 (17) 166 (103) 731 (694) LVEF, % (5) 27.3 (6.3) NYHA I/II/III class, n 4/1/2 0 2/0/1 2/1/ History of PCI 8 (66.7) 0 (0) 4 (100) 4 (100) WBCs, number/µl 6962± ± ± ± Medication Aspirin, n (%) 8 (66.7) 0 (0) 4 (100) 4 (100) ACE-I/ARB, n (%) 8 (66.7) 0 (0) 4 (100) 4 (100) Aldosterone antagonist 5 (41.7) 0 (0) 1 (25) 4 (100) Statins, n (%) 8 (66.7) 0 4 (100) 4 (100) All continuous variables are presented as mean (SD) and ordinal variables are presented as n (%) unless otherwise stated. *P-value for trend as derived from Kruskal-Wallis test for continuous variables or Fischer s Exact test for nominal variables. Controls, healthy young subjects; CAD, patients with stable coronary heart disease; ICM, patients with chronic ischemic heart failure; SD, standard deviation; 3'UTR, 3-untranslated region; FPKM, Fragments per kilobase of exon per million fragments mapped; BMI, body-mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein; TGL, triglycerides; hscrp, high-sensitivity C-reactive protein; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association classification of heart failure symptoms; PCI, percutaneous coronary intervention; WBCs, white blood cells; ACE-I, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker.

19 Supplementary Table 4. Descriptive baseline characteristics of the total PBMC cohort (n=97). Parameter 1 Non-CAD Stable CAD AMI P-value n=52 n=22 n=23 Age, years 52.9 (11.3) 66.3 (12.3) 60.7 (8.8) <0.001 Female gender 37 (71) 2 (9.1) 1 (4.3) <0.001 Smoking 31 (59.6) 8 (36.4) 16 (69.6) BMI, kgr/m (4.5) 25.8 (3.8) 27.4 (4.7) Arterial hypertension 13 (25) 16 (72.7) 6 (26.1) <0.001 Hyperlipidemia 25 (48.1) 12 (54.5) 10 (43.5) Type 2 diabetes mellitus 0 (0) 9 (40.9) 7 (30.4) <0.001 SBP, mmhg (14.3) (22.4) (21.4) DBP, mmhg 72.4 (9.2) 76.4 (7.1) 79 (10.5) hscrp, mg/l 2.4 (2.7) 21.8 (26.1) 26.2 (38.6) <0.001 Troponin I, ng/ml n.a. n.a. 3.2 (5.7) n.a. Coronary artery disease Number of vessels 1(%) / 2 (%) / 3 (%) n.a. 8 (36.4) / 7 (31.8) / 10 (43.5) / 6 (26.1) / Left ventricular ejection fraction (%) 7 (31.8) 7 (30.4) n.a (10.4) 47.5 (7.5) Kidney disease Creatinine, mg/dl 0.7 (0.2) 1.5 (1.3) 1.1 (0.6) <0.001 Chronic renal failure 0 (0) 6 (27.3) 1 (4.3) <0.001 Medication Antiplatelets 2 (3.8) 21 (95.5) 3 (13.6) (100) 2 <0.001 Statins 7 (13.5) 19 (86.4) 7 (31.8) (100) 2 <0.001 RAS inhibitors 7 (13.5) 12 (54.5) 7 (31.8) (65.2) 2 <0.001 Anticoagulants 0 (0.0) 1 (4.5) 0 (0) (34.8) Beta blockers 7 (13.5) 17 (77.3) 6 (27.3) (69.6) 2 <0.001 Ca antagonists 2 (3.8) 7 (31.8) 2 (9.1) (13.0) Nitrates 0 (0.0) 5 (22.7) 1 (4.5) (17.4) 2 <0.001 Antidiabetics 0 (0.0) 4 (18.2) 5 (22.7) (0.0) Insulin 0 (0.0) 5 (22.7) 2 (9.1) (8.7) Diuretics 4 (7.7) 4 (18.2) 1 (4.5) (4.5) All continuous variables are presented as mean (SD) and ordinal variables are presented as n (%) unless otherwise stated. 2 Second parenthesis indicates the percentage of patients with ACS receiving in-hospital treatment at the day of discharge. P values indicate significance for differences among the 3 groups by ANOVA for continuous variables and by chi-squared test for categorical/binary variables. Communitybased cohort without clinically overt cardiovascular disease. CAD, coronary artery disease; AMI, acute myocardial infarction; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; LDL, low-density lipoprotein; HDL, high density lipoprotein; hscrp, high-sensitivity C-reactive protein; RAS, renin angiotensin system; n.a., not applicable.

20 Supplementary Table 5. Descriptive baseline characteristics of the study population with measured extent of peripheral subclinical atherosclerosis (n=84). Parameter 1 Non-CAD Stable CAD AMI P-value n=51 n=16 n=17 Age, years 53.2 (11.0) 66.6 (11.3) 58.5 (8.1) <0.001 Female gender 37 (72.5) 2 (12.5) 0 (0) <0.001 Smoking 30 (58.8) 5 (31.3) 12 (70.6) BMI, kgr/m (4.5) 26.5 (4.1) 26.4 (2.8) Arterial hypertension 13 (25.5) 11 (68.8) 2 (11.8) Hyperlipidemia 25 (49.0) 9 (56.3) 5 (29.4) Type 2 diabetes mellitus 0 (0) 8 (50) 3 (17.6) <0.001 SBP, mmhg (14.4) (22.6) (22.2) <0.001 DBP, mmhg 72.7 (9.2) 76.1 (8.1) 77.5 (10.8) Aortic SBP, mmhg (11.9) (16.9) (19.0) <0.001 Aortic DBP, mmhg 74.3 (8.3) 75.7 (9.0) 75.8 (9.1) hscrp, mg/l 2.2 (2.6) 12.7 (18.3) 30.7 (43.1) <0.001 Troponin I, ng/ml n.a. n.a (6.22) n.a. Coronary artery disease Number of vessels 1(%) / 2 (%) / 3 (%) n.a. 6 (37.5) / 5 (31.3) / 8 (47.1) / 5 (29.4) / Left ventricular ejection fraction (%) 5 (31.3) 4 (23.5) n.a (9.4) 48.8 (7.0) Peripheral subclinical atherosclerosis Presence of carotid plaque 14 (27.5) 13 (81.3) 17 (100) <0.001 Presence of femoral plaque 16 (32) 14 (87.5) 15 (93.8) <0.001 Presence of carotid and/or femoral 23 (45.1) 14 (87.5) 17 (100) <0.001 plaque Number of carotid segments with 0.4 (0.7) 2.1 (1.5) 2.1 (1.1) <0.001 plaque Number of carotid and femoral 0.8 (1.1) 3.6 (2) 3.4 (1.4) <0.001 segments with plaque Number of femoral segments with 0.5 (0.8) 1.5 (0.7) 1.4 (0.6) <0.001 plaque PWV, m/sec 8.5 (1.4) 12.1 (3.1) 9.5 (2.2) <0.001 Combined carotid IMT, mm 1 (0.2) 1.5 (0.5) 1.5 (0.4) <0.001 Common carotid IMT, mm 0.8 (0.2) 1.1 (0.4) 1 (0.4) Carotid bulb IMT, mm 1.1 (0.3) 1.8 (0.6) 1.9 (0.8) <0.001 Internal carotid IMT, mm 0.9 (0.4) 1.6 (0.8) 1.5 (0.6) <0.001 Femoral IMT, mm 1.2 (0.9) 2.3 (1.2) 2.5 (1) <0.001 Kidney disease Creatinine, mg/dl 0.6 (0.2) 1.7 (1.5) 1 (0.3) <0.001 Chronic renal failure 0 (0) 5 (31.3) 1 (5.9) <0.001 Medication Antiplatelets 2 (3.9) 16 (100) 2 (11.8) (100) 2 <0.001 Statins 7 (13.7) 14 (87.5) 3 (17.6) (100) 2 <0.001 RAS inhibitors 7 (13.7) 9 (56.3) 4 (23.5) (70.6) Anticoagulants 0 (0.0) 0 (0.0) 0 (0.0) (47.1) 2 n.a. Beta blockers 6 (11.8) 13 (81.3) 4 (23.5) (76.5) 2 <0.001 Ca antagonists 2 (3.9) 7 (43.8) 1 (5.9) (11.8) 2 <0.001 Nitrates 0 (0.0) 1 (6.3) 1 (5.9) (11.8) Antidiabetics 0 (0.0) 4 (25.0) 3 (17.6) (0.0)

21 Insulin 0 (0.0) 4 (25.0) 0 (0.0) (0.0) 2 <0.001 Diuretics 3 (5.9) 3 (18.8) 0 (0.0) (0.0) All continuous variables are presented as mean (SD) and ordinal variables are presented as n (%) except if stated otherwise. 2 Second parenthesis indicates the percentage of patients with ACS receiving in-hospital treatment at the day of discharge. P values indicate significance for differences among the 3 groups by ANOVA for continuous variables and by chi-squared test for categorical/binary variables. Communitybased cohort without clinically overt cardiovascular disease. CAD, coronary artery disease; AMI, acute myocardial infarction; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; PWV, pulse wave velocity; IMT, intima-media thickness; LDL, low-density lipoprotein; HDL, high density lipoprotein; CRP, C-reactive protein; RAS, renin angiotensin system; n.a., not applicable.

22 Supplementary Table 6. Association of CTSS and ADAR1 mrna expression with extent of cardiovascular disease and markers of peripheral subclinical atherosclerosis in the PBMC cohort. Total cohort (n=97) Age-matched cohort (n=45) XY correlation matrix CTSS expression ADAR1 expression CTSS expression ADAR1 expression Extent of cardiovascular disease Presence of stable CAD or < AMI 1, * Presence of CAD 2, ** < Number of diseased coronary arteries* Left ventricular ejection -0.2 (0.181) (0.639) (0.089) (0.384) fraction hscrp 0.39 (<0.001) 0.35 (0.001) 0.31 (0.071) 0.25 (0.146) Number of carotid segments with plaque Number of carotid and femoral segments with plaque Number of femoral segments Markers of peripheral subclinical atherosclerosis 0.38 (<0.001) 0.29 (0.008) 0.45 (0.005) 0.36 (0.026) 0.32 (<0.003) 0.22 (0.042) 0.41 (0.011) 0.33 (0.044) 0.19 (0.083) 0.14 (0.199) 0.3 (0.070) 0.22 (0.190) with plaque Presence of carotid plaque* Presence of carotid and/or femoral plaque* Presence of femoral plaque* PWV 0.28 (0.014) 0.29 (0.011) 0.35 (0.047) 0.26 (0.152) Combined carotid IMT mm 0.22 (0.045) 0.16 (0.142) 0.36 (0.026) 0.24 (0.140) Common carotid IMT, mm 0.18 (0.097) 0.14 (0.196) 0.22 (0.189) 0.13 (0.421) Carotid bulb IMT, mm 0.27 (0.014) 0.24 (0.023) 0.42 (0.009) 0.36 (0.026) Internal carotid IMT, mm 0.05 (0.649) (0.801) 0.22 (0.181) 0.13 (0.425) Femoral IMT, mm 0.23 (0.035) 0.19 (0.080) 0.33 (0.044) 0.2 (0.234) CTSS, cathepsin S; ADAR1, Adenosine deaminase acting on RNA-1; hscrp, high-sensitivity C-reactive protein; CAD, coronary artery disease; AMI, acute myocardial infarction; PWV, pulse wave velocity; IMT, intima-media thickness. 1 Group analysis where non-cad=0, stable CAD=1, AMI=2; 2 Group analysis where non-cad=0, CAD (stable CAD or AMI)=1. Correlations were tested for significance with the Spearman rank correlation coefficient test. All values with P<0.1 are marked in bold to indicate a trend or a significant association (P 0.05) between CTSS or ADAR1 mrna expression and cardiovascular disease markers. *P values are shown for nominal/ordinal parameters calculated by Kruskal-Wallis test, ** P values are shown for binary parameters calculated by Mann-Whitney U test.

23 Supplementary Table 7. Association of the extent of RNA editing rate of specific nucleotides in CTSS 3'UTR with CTSS and ADAR1 mrna expression in the PBMC cohort. Total cohort (n=97) Age-matched cohort (n=45) All (n=97) XY correlation matrix, r (pvalue) CTSS expression ADAR1 expression CTSS expression ADAR1 expression % mean editing rate 1 ADAR1 expression 0.79 (<0.001) n.a (<0.001) n.a. n.a. nt (0.636) 0.03 (0.747) (0.277) (0.815) 9.5 (2.2) nt (0.180) 0.16 (0.125) 0.09 (0.561) 0.2 (0.193) 57.4 (4.4) nt (0.323) 0.07 (0.489) 0.02 (0.912) 0.28 (0.058) 9.8 (5.2) nt (0.214) 0.11 (0.305) (0.941) 0.34 (0.023) 10.5 (4.1) nt (0.172) 0.22 (0.034) 0.18 (0.241) 0.36 (0.014) 43.2 (5.3) nt (0.347) 0.26 (0.010) 0.18 (0.233) 0.43 (0.004) 12.8 (4.3) nt (0.012) 0.27 (0.007) 0.37 (0.012) 0.45 (0.002) 58.0 (4.2) nt (0.779) 0.15 (0.144) 0.16 (0.306) 0.34 (0.026) 10.9 (4.1) nt (0.551) 0.18 (0.084) 0.23 (0.126) 0.38 (0.012) 12.6 (3.4) nt (0.087) 0.19 (0.062) 0.34 (0.026) 0.35 (0.021) 35.1 (5.5) nt (0.089) 0.25 (0.016) 0.34 (0.027) 0.39 (0.009) 30.7 (4.0) nt (0.081) 0.27 (0.009) 0.19 (0.213) 0.36 (0.018) 15.5 (2.8) nt (0.958) 0.11 (0.295) 0.19 (0.239) 0.33 (0.033) 20.7 (5.6) nt (0.234) 0.21 (0.041) 0.28 (0.070) 0.44 (0.004) 29.4 (8.7) nt (0.108) 0.09 (0.409) 0.05 (0.773) 0.11 (0.502) 70.9 (3.5) nt (0.149) 0.1 (0.362) 0.25 (0.118) 0.23 (0.141) 64.8 (5.6) nt (0.034) 0.23 (0.029) 0.04 (0.807) 0.13 (0.402) 21.0 (4.3) nt (0.407) 0.08 (0.456) 0.16 (0.306) 0.2 (0.200) 31.2 (4.0) nt (0.013) 0.22 (0.041) 0.31 (0.051) 0.36 (0.021) 33.0 (4.7) nt (0.038) 0.14 (0.204) 0.22 (0.167) 0.2 (0.212) 49.4 (4.1) nt (0.005) 0.22 (0.037) 0.12 (0.475) 0.15 (0.364) 30.9 (4.8) nt (0.671) 0.07 (0.513) 0.04 (0.792) 0.01 (0.934) 33.0 (3.2) nt (0.848) 0.03 (0.796) (0.778) (0.991) 28.2 (5.0) nt (0.076) 0.12 (0.273) 0.25 (0.137) 0.15 (0.388) 65.8 (5.6) nt (0.617) 0.06 (0.606) (0.581) 0.11 (0.512) 21.7 (5.3) CTSS, cathepsin S; ADAR1, Adenosine deaminase acting on RNA-1; nt, nucleotide position in 3 UTR of CTSS mrna where an A-to-I RNA editing was observed. 1 Mean % (SD) A-to-I RNA editing rate per nucleotide in the 3'UTR of CTSS mrna in the total PBMC cohort. Correlations were tested for significance with the Spearman rank correlation coefficient test. All values with P<0.1 are marked in bold to indicate a trend or a significant association (P 0.05) between CTSS or ADAR1 mrna expression and cardiovascular disease markers.

24 Supplementary Table 8. Association of RNA editing rate of specific nucleotides in CTSS 3'UTR with extent of cardiovascular disease and markers of peripheral subclinical atherosclerosis in the PBMC cohort. XY correlation matrix nt 1770 nt 1806 nt 1807 nt 1826 nt 1843 nt 1850 nt 1802 nt 1846 nt 1909 Extent of cardiovascular disease Presence of stable CAD or AMI* Presence of CAD** Number of diseased coronary arteries,* Left ventricular ejection fraction (0.102) (0.276) (0.154) (0.230) -0.3 (0.055) (0.871) (0.322) (0.240) (0.948) hscrp 0.18 (0.115) 0.01 (0.930) 0.2 (0.075) 0.19 (0.097) 0.12 (0.309) 0.15 (0.189) 0.09 (0.440) 0.12 (0.294) 0.17 (0.167) Markers of peripheral subclinical atherosclerosis Number of carotid segments with plaque Number of carotid and femoral segments with plaque Number of femoral segments with plaque Presence of carotid plaque* Presence of carotid and/or femoral plaque* Presence of femoral plaque* 0.26 (0.016) 0.3 (0.006) 0.25 (0.023) 0.26 (0.019) 0.19 (0.094) 0.07 (0.566) 0.39 (<0.001) 0.31 (0.005) 0.15 (0.200) 0.23 (0.042) 0.28 (0.013) 0.26 (0.023) 0.23 (0.050) 0.22 (0.053) 0.15 (0.185) 0.13 (0.258) 0.12 (0.294) 0.13 (0.281) 0.32 (0.004) 0.25 (0.024) 0.12 (0.305) 0.2 (0.078) 0.22 (0.053) 0.21 (0.075) 0.2 (0.090) 0.18 (0.126) 0.13 (0.292) PWV 0.06 (0.636) (0.807) 0.13 (0.265) 0.07 (0.538) 0.04 (0.774) 0.18 (0.150) (0.983) 0.14 (0.257) 0.11 (0.373) Combined carotid IMT 0.23 (0.039) 0.23 (0.042) 0.35 (0.001) 0.23 (0.044) 0.18 (0.114) 0.08 (0.473) 0.27 (0.014) 0.15 (0.187) 0.13 (0.293) Common carotid IMT 0.09 (0.396) 0.05 (0.646) 0.1 (0.387) 0.05 (0.696) 0.05 (0.655) 0.16 (0.173) 0.1 (0.380) 0.04 (0.759) 0.07 (0.538) Carotid bulb IMT 0.16 (0.139) 0.18 (0.107) 0.28 (0.012) 0.13 (0.264) 0.11 (0.337) (0.962) 0.23 (0.043) 0.08 (0.476) 0.09 (0.450) Internal carotid IMT 0.18 (0.100) 0.19 (0.086) 0.29 (0.008) 0.26 (0.021) 0.18 (0.120) 0.1 (0.395) 0.21 (0.057) 0.2 (0.077) 0.13 (0.290) Femoral IMT 0.19 (0.099) 0.01 (0.901) 0.11 (0.320) 0.24 (0.035) 0.1 (0.390) 0.12 (0.300) 0.05 (0.637) 0.13 (0.277) 0.05 (0.669) CTSS, cathepsin S; ADAR1, Adenosine deaminase acting on RNA-1; hscrp, high-sensitivity C-reactive protein; CAD, coronary artery disease; AMI, acute myocardial infarction; PWV, pulse wave velocity; IMT, intima-media thickness. *P values are shown for nominal/ordinal parameters calculated by Kruskal-Wallis test, **P values are shown for binary parameters calculated by Mann-Whitney U test. Correlations were tested for significance with the Spearman rank correlation coefficient test. All values with P<0.1 are marked in bold to indicate a trend or a significant association between extent of RNA editing of specific nucleotides and cardiovascular disease markers.

25 Supplementary Table 9. Descriptive baseline characteristics of the thoracic aortic aneurysm cohort. Parameter 1 Non-dilated aorta Dilated aorta P-value n 8 35 Age, years 70 (15) 64 (12) Female gender 3 (37.5) 13 (37.1) 0.99 Smoking, current smokers 0 (0) 3 (8.6) BMI, kg/m (5.4) 26.8 (3.0) Triglycerides, mmo/l 0.81 (0.28) 1.20 (0.57) <0.05 Cholesterol, mmo/l 5.34 (0.84) 5.06 (0.87) SBP, mmhg (22.1) (17.3) DBP, mmhg 79.9 (16.2) 77.4 (15.5) hscrp, mg/l 2.87 (1.79) 5.28 (13.24) Aortic aneurysm and valve disease Max ascending aortic diameter, mm 32.5 (5.0) (6.94) < Aortic valve stenosis 5 (62.5) 1 (2.9) < Aortic valve regurgitation 3 (37.5) 23 (65.7) Medication Beta blockers 3 (37.5) 16 (45.7) Ca antagonists 0 (0) 4 (11.4) ACE inhibitors 4 (50) 11 (31.4) Statins 1 (12.5) 11 (31.4) All patients had a tricuspid aortic valve. BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; hscrp, high-sensitive C-reactive protein; ACE, Angiotensin-converting enzyme. 1 All continuous variables are presented as mean (SD) and binary variables are presented as n (%). P values indicate significance for differences between the two groups using a Mann-Whitney test for continuous variables and a Chi-square test for nominal variables.

26 Supplementary Table 10. Association of the extent of RNA editing rate of specific nucleotides in CTSS 3'UTR with CTSS and ADAR1 mrna expression in the thoracic aortic aneurysm cohort. XY correlation matrix, r (p-value) CTSS expression Total cohort (n=43) ADAR1 expression Aortic aneurysms (n=35) CTSS expression ADAR1 expression CTSS (0.006) (0.009) nt (0.042) (0.371) 0.41 (0.015) (0.354) nt (0.002) 0.32 (0.053) 0.47 (0.005) 0.29 (0.100) nt (0.012) (0.944) 0.42 (0.013) (0.860) nt (0.025) (0.372) (0.010) (0.336) nt (0.085) 0.25 (0.140) 0.28 (0.111) 0.37 (0.034) nt (0.070) (0.477) 0.35 (0.048) (0.659) nt (0.015) 0.05 (0.794) 0.46 (0.009) 0.14 (0.454) nt (0.007) (0.794) 0.61 (0.001) 0.18 (0.385) nt (0.018) -0.3 (0.128) 0.49 (0.017) (0.322) nt (0.003) (0.188) 0.62 (0.002) (0.553) nt (0.066) -0.4 (0.061) 0.49 (0.034) (0.170) nt (0.011) (0.754) 0.43 (0.039) (0.490) nt (0.033) (0.115) 0.5 (0.017) (0.388) CTSS, cathepsin S; ADAR1, Adenosine deaminase acting on RNA-1; nt, nucleotide position in 3'UTR of CTSS mrna where an A-to-I RNA editing was observed. Correlations were tested for significance with the Spearman rank correlation coefficient test. All values with P<0.1 are marked in bold to indicate a trend or a significant association between RNA editing rate and CTSS or ADAR1 mrna expression.

27 Supplementary Table 11. Baseline characteristics of the BiKE patients cohort used in RNA editing experiments. Parameters 1 n 20 Age, years mean (SD) 71.7 (6.845) Gender (male/female), n 15/5 BMI mean (SD) (5.083) Smoking yes/no/former, n 3/6/11 Symptoms Asymptomatic 2 (10%) Minor stroke 8 (40%) Transitory ischemic attack 5 (25%) Amaurosis fugax 5 (25%) Comorbidities Previous infarction (estimated by computer tomography) 6 (30%) Inflammatory diseases (rheumatism, psoriasis, multiple sclerosis, inflammatory 3 (15%) bowel disease) Arterial hypertension 17 (85%) Diabetes mellitus (all type II) 3 (15%) Biochemic lab measurements Creatinine, mg/dl 80.4 (22.44) Cholesterol, mmol/l (0.839) LDL, mmol/l (0.6418) HDL, mmol/l (0.411) hscrp, mg/l (8.281) HbA1c, mmol/mol (17.97) Hb, g/dl (18.85) Medication Hypolipidemic therapy (ezetimib, HMG-CoA reductase inhibitors) 20 (100%) Antidiabetics 3 (15%) Antihypertensives (ACE inhibitors, beta-blockers, diuretics, angiotensine II 17 (85%) blockers) BMI, body mass index; hscrp, high-sensitive C-reactive protein; ACE, Angiotensin-converting enzyme. 1 All continuous variables are presented as mean (SD) and binary variables are presented as n (%).

28 Supplementary Table 12. Association of the extent of RNA editing rate of specific nucleotides in CTSS 3'UTR with CTSS and ADAR1 mrna expression in the BiKE cohort. XY correlation matrix, r (p-value) CTSS expression Total cohort (n=28) ADAR1 expression Atherosclerotic plaques (n=20) CTSS expression ADAR1 expression CTSS (<0.0001) (<0.0001) nt (<0.0001) 0.83 (<0.0001) 0.5 (0.027) 0.57 (0.008) nt (0.001) 0.61 (0.001) 0.23 (0.339) 0.29 (0.217) nt (0.001) 0.64 (<0.0001) 0.29 (0.217) 0.37 (0.110) nt (<0.0001) 0.8 (<0.0001) 0.36 (0.121) 0.55 (0.013) nt (<0.0001) 0.68 (<0.0001) 0.64 (0.003) 0.65 (0.002) nt (0.017) 0.52 (0.009) 0.34 (0.145) 0.39 (0.086) nt (0.001) 0.65 (<0.0001) 0.44 (0.056) 0.49 (0.028) nt (0.091) 0.47 (0.052) 0.41 (0.091) 0.47 (0.052) nt (0.002) 0.64 (<0.0001) 0.52 (0.022) 0.65 (0.002) nt (0.112) 0.35 (0.084) 0.49 (0.027) 0.54 (0.015) nt (0.576) 0.33 (0.226) (0.894) 0.30 (0.370) nt (0.006) 0.71 (0.002) 0.25 (0.489) 0.44 (0.200) nt (<0.0001) 0.85 (<0.0001) 0.14 (0.736) 0.31 (0.456) CTSS, cathepsin S; ADAR1, Adenosine deaminase acting on RNA-1; nt, nucleotide position in 3'UTR of CTSS mrna where an A-to-I RNA editing was observed. Correlations were tested for significance with the Spearman rank correlation coefficient test. All values with P<0.1 are marked in bold to indicate a trend or a significant association between RNA editing rate and CTSS or ADAR1 mrna expression.

29 Supplementary Table 13. Selected genes that are edited in 3'UTR, have at least one HuR binding motif in the 3'UTR and are downregulated after ADAR1 knockdown in endothelial cells under basal conditions (ribo-minus RNA-seq; n=3 per condition). Gene ID Name Fold expression siadar1/scrambled under basal conditions (mean±sem) ENSG EBF ± 0.12 ENSG CSMD ± 0.09 ENSG RNF113B 0.37 ± 0.19 ENSG AL ± 0.09 ENSG GBP ± 0.06 ENSG IGSF ± 0.11 ENSG DRAXIN 0.48 ± 0.08 ENSG FXYD ± 0.29 ENSG PLAC ± 0.05 ENSG CTSS 0.52 ± 0.14 ENSG IQGAP ± 0.14 ENSG HIVEP ± 0.20 ENSG MOBP 0.55 ± 0.09 ENSG PM20D ± 0.28 ENSG DCLK ± 0.10 ENSG APOBEC3F 0.62 ± 0.02 ENSG SPRY ± 0.09 ENSG ZNF ± 0.26 ENSG SFXN ± 0.03 ENSG UNC5C 0.68 ± 0.08 ENSG KCTD ± 0.13 ENSG VGLL ± 0.18 ENSG KLHDC ± 0.05 ENSG KRBA ± 0.08 ENSG FBXW ± 0.04 ENSG STON ± 0.10 ENSG FAM118A 0.72 ± 0.06 ENSG SGCD 0.72 ± 0.72 ENSG ZNF ± 0.04 ENSG ZNF ± 0.03 ENSG TNRC6C 0.73 ± 0.03 ENSG IFNAR ± 0.11 ENSG PPARA 0.74 ± 0.09 ENSG KLHL ± 0.03 ENSG ZNF ± 0.08 ENSG MOCOS 0.74 ± 0.10 ENSG APOL ± 0.12 ENSG SNAPC ± 0.03 ENSG KIAA ± 0.04 ENSG ZFHX ± 0.14

30 ENSG UTY 0.76 ± 0.07 ENSG ZNF ± 0.17 ENSG ZNF ± 0.15 ENSG DHTKD ± 0.02 ENSG ZC3H ± 0.27 ENSG PLEKHG ± 0.06 ENSG C9orf ± 0.06 ENSG NOX ± 0.03 ENSG GOLGA ± 0.04 ENSG MBD ± 0.08 ENSG IBA ± 0.09 ENSG ZNF ± 0.08 ENSG PLEKHA ± 0.14 ENSG AP3S ± 0.09 ENSG DMD 0.79 ± 0.14 ENSG LOH12CR ± 0.09 ENSG C8orf ± 0.18 ENSG KIAA1324L 0.79 ± 0.08 ENSG TMC ± 0.11 ENSG FOXK ± 0.06

31 Supplementary Table 14. Selected genes that are edited in 3'UTR, have at least one HuR binding motif in the 3'UTR and are downregulated after ADAR1 knockdown in endothelial cells under hypoxic conditions (ribo-minus RNA-seq; n=3 per condition). Gene ID Name Fold expression siadar1/scrambled under hypoxic conditions (mean±sem) ENSG GPC ± 0.21 ENSG CTSS 0.39 ± 0.10 ENSG AL ± 0.06 ENSG ITGA ± 0.13 ENSG PDE4C 0.46 ± 0.24 ENSG UNC5C 0.47 ± 0.06 ENSG NHS 0.47 ± 0.10 ENSG ZNF ± 0.09 ENSG GBP ± 0.17 ENSG SPECC ± 0.02 ENSG UPK1B 0.55 ± 0.55 ENSG NRG ± 0.15 ENSG VSIG ± 0.19 ENSG PNPLA ± 0.29 ENSG CTF ± 0.17 ENSG KCNE ± 0.08 ENSG FBXO ± 0.38 ENSG ARMC ± 0.05 ENSG LCTL 0.64 ± 0.22 ENSG APOL ± 0.11 ENSG TFDP ± 0.13 ENSG MAP7D ± 0.14 ENSG AL ± 0.16 ENSG EIF1AY 0.67 ± 0.14 ENSG FAM118A 0.68 ± 0.07 ENSG CASP ± 0.22 ENSG SNAPC ± 0.08 ENSG LRRK ± 0.27 ENSG AS3MT 0.70 ± 0.08 ENSG ZNF ± 0.15 ENSG TTLL ± 0.30 ENSG ZNF ± 0.04 ENSG TTC ± 0.21 ENSG LDLRAD ± 0.16 ENSG MYLK ± 0.13 ENSG PARD6G 0.71 ± 0.06 ENSG CREBRF 0.71 ± 0.06 ENSG C14orf ± 0.04 ENSG C4orf ± 0.01 ENSG KLHL ± 0.06

32 ENSG ANKRA ± 0.08 ENSG PANX ± 0.19 ENSG ACSBG ± 0.05 ENSG GPR ± 0.17 ENSG KRBA ± 0.08 ENSG GOLGA ± 0.11 ENSG UBOX ± 0.17 ENSG EFCAB ± 0.03 ENSG ZNF ± 0.05 ENSG ZNF ± 0.05 ENSG EFNB ± 0.12 ENSG CPPED ± 0.04 ENSG GIPC ± 0.09 ENSG C11orf ± 0.06 ENSG ZNF ± 0.09 ENSG CLN ± 0.05 ENSG AJUBA 0.78 ± 0.12 ENSG THOC ± 0.10 ENSG BVES 0.78 ± 0.03 ENSG SDK ± 0.06 ENSG TSPAN ± 0.11 ENSG POLH 0.79 ± 0.04 ENSG TEP ± 0.12 ENSG YIPF ± 0.06 ENSG TMEM ± 0.09 ENSG MBD ± 0.13