Supplementary Appendix

1. Next generation sequencing

Briefly, sequencing of exonic regions and intronic boundaries of selected was performed on Illumina Hiseq1500 platform for 149 panels, and MiSeq or NextSeq platform (Illumina, USA) for 103 gene panel, with customized library probes (SureSelect Target Enrichment Kit; Agilent Technologies ) or OS-Seq™ technology with automated target DNA capture (1). Bioinformatic data analyses included software as GATK NovoAlign (Novocraft Technologies Sdn Bhd), BCFtools and SAMtools (Sanger Institute), and Annovar for variant calling and genotyping. Filtering was performed relying on data from multiple population databases (1000 genomes database , ESP database, Exome Aggregation Consortium and ClinVar) and with commonly used “in silico” prediction tools, as previously described(2, 3). Median sequencing depth in target region was 358x and 99.8% of the nucleotides reached at least 15x coverage. All selected variants was confirmed with bi-directional Sanger sequencing.

2. Gene Panel lists tested in the study 149 genes Health in Code’(http://www.healthincode.com/servicios/genetica-medica) Pan Panel AARS2, ABCC9, ACAD9, ACADVL, ACTA1, ACTA2, ACTC1, ACTN2, AGK, AGL, AGPAT2, ALMS1, ANK2, ANKRD1, ATPAF2, BAG3, BRAF, BSCL2, CALR3, CASQ2, CAV3, CBL, COQ2, COX15, COX6B1, CRELD1, CRYAB, CSRP3, CTF1, CTNNA3, DES, DLD, DMD, DNAJC19, DOLK, DSC2, DSG2, DSP, DTNA, ELN, EMD, EYA4, FAH, FHL1, FHL2, FHOD3, FKRP, FKTN, FLNA, FLNC, FOXD4, GAA, GATA4, GATA6, GATAD1, GFM1, GJA1, GJA5, GLA, GLB1, GNPTAB, GUSB, HCN4, HFE, HRAS, JAG1, JPH2, JUP, KCNH2, KCNJ2, KCNJ8, KCNQ1, KLF10, KRAS, LAMA2, LAMA4, LAMP2, LDB3, LIAS, LMNA, MAP2K1, MAP2K2, MIB1, MLYCD, MRPL3, MRPS22, MTO1, MURC, MYBPC3, MYH11, MYH6, MYH7, MYL2, MYL3, MYLK2, MYOT, MYOZ2, MYPN, NEBL, NEXN, NKX2-5, NOTCH1, NRAS, OBSL1, PDHA1, PDLIM3, PHKA1, PITX2, PKP2, PLN, PMM2, PRDM16, PRKAG2, PSEN1, PSEN2, PTPN11, RAF1, RBM20, RYR2, SCN5A, SGCA, SGCB, SGCD, SHOC2, SLC22A5, SLC25A4, SMAD3, SOS1, SPRED1, SURF1, TAZ, TBX1, TBX20, TBX5, TCAP, TGFB3, TMEM43, TMEM70, TMPO, TNNC1, TNNI3, TNNT2, TPM1, TRIM63, TSFM, TTN, TTR, TXNRD2, VCL.

103 genes Blueprint Genetics’ (http://cardiology.blueprintgenetics.com/tests/) Pan Cardiomyopathy Panel (v1.1 May 06, 2014)

ABCC9, ACADVL, ACTC1, ACTN2, AGL, ANKRD1, ATP5E, BAG3, BRAF, CALR3, CASQ2, CAV3, CBL, COA5, CRYAB, CSRP3, CTF1, CTNNA3, DES, DMD, DMPK, DNAJC19, DNM1L, DOLK, DSC2, DSG2, DSP, DTNA, EMD, EYA4, FHL1, FHL2, FKTN, FOXRED1, FXN, GAA, GATAD1, GLA, GLB1, GUSB, HFE, HRAS, ILK, JPH2, JUP, KRAS, LAMA4, LAMP2, LDB3, LMNA, MAP2K1, MAP2K2, MRPL3, MIB1, MYBPC3, MYH6, MYH7, MYL2, MYL3, MYLK2, MYOM1, MYOZ2, MYPN, NEBL, NEXN, NRAS, PDLIM3, PKP2, PLN, PRKAG2, PSEN1, PSEN2, PTPN11, RAF1, RBM20, RYR2, SCN5A, SCO2, SDHA, SGCD, SHOC2, SLC25A3, SOS1, SPRED1, SYNE1, SYNE2, TAZ, TCAP, TGFB3, TMEM43, TMEM70, TMPO, TNNC1, TNNI3, TNNT2, TPM1, TRIM63, TSFM, TTN, TTR, TXNRD2, VCL, XK

3. Genetic Evaluation Variant interpretation warrants several considerations: currently, ClinVar and ExAc Databases are becoming the gold standard of variant classification; they rely on multiple submitters, both clinical and research genetic Laboratories and databases, the vast majority of them located in US or Northern Europe (http://www.ncbi.nlm.nih.gov/clinvar/docs/submitter_list/). We must aknowledge that actually not all research/clinical/private genetic Laboratories are active submitters: given this fact, they still may not be representative of all population databases and their variant’s definitions are subjected to changes over times.

Pathogenic / likely pathogenic carriers: If identified rare variants were present in ClinVar database and classified as pathogenic or likely pathogenic, patients carrying these variants were considered pathogenic/likely pathogenic mutation carriers. Notably, familial Rare Variants absent from ExAc (http://exac.broadinstitute.org) and ClinVar, or present in ExAc with MAF < or = to 0,0004 (4) and absent from ClinVar or present in ClinVar but classified as Variant of Unknown Significance (VUS) or with conflicting interpretation , which turned to be positively co-segregating with the disease inside the family, were classified as likely pathogenic as well (e.g. FLNC truncating variants (2), PKP2 large deletion variant, OBSL1 and MYH6 variants). (4,5,6)

Possibly Pathogenic rare Variants (PPrV) Carriers: Rare Variants with MAF < or = to 0,0004 in ExAc and classified as of unknown significance/conflicting interpretation in ClinVar, or absent from ClinVar , and rare variants absent from ClinVar and ExAc, were considered PPrV. Patients with sporadic DCM, or familial DCM in which no segregation study was available, carrying these PPrV as above defined (n°40 “PPrV carriers patients”), without sufficient data to rule-out pathogenicity, were considered as still eligible in Clusters. Each PPrV was carefully evaluated with the available clinical and bioinformatics evidence (5 out of 41 PPrV resulting in premature truncation -e.g. frameshift, nonsense or consensus splice site-, 28 of the remaining 36 missense variants predicted deleterious by the majority of in silico tools applied - Polyphen-2 (v2.2.2r398) HumVar; SIFT, Mutation Taster- and 8 with conflicting interpretations). A double exception to frequency cutoff criteria was made for LDB3 PPrV rs121908338 and DSC2 PPrV rs145560678 (MAF in ExAc= 0,004 and 0,001 respectively). (see Fig. 2 and Detailed PPrV information).

4. Detailed PPrV informations of “PPrV Patients” in Clusters 3 – 5 (Gene, position, Nucleotide change, Change, eventual RS, MAF in ExAc, clinical-bionformatic comments)

4a : Cluster 3 patients

OBSL1 2: 220435575 c.380C>T p.A127V - -

This variant has not been found in ExAc. OBSL1 have been described in 3-M Syndrome, and proposed as a candidate gene for (DCM) due to its known function (cytoskeletal adaptor). However, no clinical association has been established yet.

MYPN 10 :69961746 c.3654G>C p.Arg1218Ser MAF ExAc: 0.000008259

Arg1218Ser affects a highly conserved residue, up to chicken considering 11 species, which is in the region of interaction with ACTN (amino acids 945-1320), between the Ig-like domains 3 and 4 of Myopalladin. The results of the analysis with different bioinformatics predictors suggest that it is a deleterious substitution: - Polyphen-2 (v2.2.2r398) HumVar: Probably damaging, score 0.999 (score range: 0-1). - SIFT: deleterious, score 0 (score<0.05 deleterious), low confidence. - Mutation Taster: presumably deleterious, p-value=1 (range: 0-1).

MYPN 10: 69970162 c.3913A>G p.M1305V MAF ExAc: 0.00003295

This variant has been submitted to CLinVar, classified as VUS. It affects Highly conserved residue, up to chicken considering 11 species, which is in the region of interaction with ACTN (amino acids 945-1320), between the Ig-like domains 3 and 4 of Myopalladin. The results of the analysis with different bioinformatics predictors suggest that it is a tolerated variant, still this information is not enough to confidently exclude a pathogenic role

OBSL1 2: 220426685 c.2999G>A p.Cys1000Tyr OBSL1 2: 220435300 c.655C>G p.Gln219Glu

These variants are absent in ExAc and ClinVar. They are found in double Heterozygosis. Both variants segregates with the disease inside the family. We considered the combination of these variants as likely pathogenic.

DMD X: 32632565 c.1337A>G p.H446R rs72468699 MAF ExAc: 0.0003147

In silico analyses predict the substitution deleterious. Male gender patient. It has been reported in ExAc with mutation-compatible frequency. It is classified as a variant of unknown significance in ClinVar.

NEXN 1: 78407863 c.1630delC p.Q544KfsX23

This truncating variant is absent in ExAc, and has been submitted to ClinVar (variation id: 180457). NEBL 10: 21101775 c.2441A>G p.Gln814Arg

This variant is predicted deleterious by in silico programs PolyPhen, SIFT and MutationTaster. It has not been reported in the Exome Aggregation Consortium (ExAC) data set.

DES 2:220284998 c.665G>A p.Arg222His rs367961979 MAF ExAc: 0.0003708

This variant is located at the Coil 1B region (Amino acids 152-252) of the . It affects a highly conserved amino acid, up to Frog (considering 10 species). In silico prediction are contradictory: it is considered deleterious by SIFT and MT, and benign by PolyPhen-2 (v2.2.2r398).

NEXN 1: 78383276 c.53T>C p.Val18Ala MAF ExAc: 0.000008330

This variant is absent from ClinVar. The amino acid Val18 is located in the N-terminal domain of nexilin, specifically at the first binding domain of this protein (residues 1-164). Val18Ala affects a moderately conserved amino acid (considering 10 species) In-silico predictors are inconclusive: -Polyphen-2 (v2.2.2r398): benign, score 0.060 (Range score: 0-1). -SIFT: Deleterious, score 0.01 (score <0.05 deleterious), low confidence. -Mutation taster: presumably disease causing, p-value: 0.668 (range: 0-1).

LDB3 10: 88446830 c.349G>A p.Asp117Asn rs121908338 MAF ExAc: 0.004554

This variant has conflicting interpretation in ClinVar. Orginally reported as pathogenic, it has subsequently been reported with a rare variant’s compatible frequency. It has been reported in several families with DCM and LVNC, but segregation was not always confirmed. Functional studies are conflicting. In our case, it segregates inside the family with the disease. We can not exlude a possible role in the development of DCM.

4b: Cluster 4 patients

DSC2 18: 28666574 c.907G>A p.Val303Met rs145560678 MAF ExAc: 0.001194

Reported with Conflicting Interpretation in CLinVAr. In silico predictors suggest it may be tolerated. It has been reported in literature associated with Dilated Cardiomiopathy and cardiac transplantations, with possible familial segregation. The available information to date is not sufficient to confirm or discard its pathogenicity.

PKP2 12: 33031360 c.454C>T p.Pro152Ser

This variant is absent in ExAc. This amino acid, Pro152, is located in the N-terminus of 2 (amino acids 1-340), described as the domain that interacts with other demosomal . It affects a highly conserved amino acid, up to frog (considering 8 species). The results of the evaluation in silico with different bioinformatics predictors are inconclusive: PolyPhen-2 (v2.2.2r398): Probably damaging, score 0.999 (score range:0-1), SIFT: Tolerated, score of 0.18 (score <0.05 deleterious), Mutation Taster: Presumably disease causing, p-value 0.994 (range: 0 - 1).

DSP 6: 7584025 c.4733C>A p.T1578N

The variant has not been reported in the Exome Aggregation Consortium (ExAC) .In silico predictions for this substitution are contradictory: PolyPhen and SIFT interpret it as deleterious whereas Mutation Taster benign.

DSP 6: 7583680 c.6185A>G p.H2062R

The variant has not been reported in the Exome Aggregation Consortium (ExAC). In silico analyses for the variant are contradictory: it is predicted possibly damaging by Polyphen but tolerated by SIFT and Mutation Taster.

JUP 17: 39925336 c.592G>A p.Asp198Asn JUP 17: 39919398 c.1334C>T p.Ala445Val

Variant Asp198Asn has been detected in a single carrier in ExAc Population, while variant Ala445Val is absent from ExAc. Both variants are predicted damaging by PolyPhen2 but tolerated by SIFT. Missense mutations in compound heterozygosity in this gene were described associated with cardiac disease, especially Naxos disease or arrhythmogenic cardiomyopathy. When left ventricular involvement is present, it is difficult to make a differential diagnosis with dilated cardiomyopathy.

CTNNA3 10: 67680251 c.2525G>A p.R842Q MAF ExAc: 0.00003295

In silico predictions for this substitution are contradictory: Mutation Taster interprets it as disease causing whereas SIFT benign

4c: Cluster 5 patients

MYBPC3 11:47349779 c.649A>G p.Ser217Gly rs138753870 MAF Exac: 0.002268

This variant has conflicting interpretation in ClinVar, previously associated with HCM phenotype. It is predicted as possibly damaging /disease causing by PolyPhen2 and Mutation Taster.

MYBPC3 11: 47356625 c.2873C>T p.T958I rs376504548 MAF ExAc: 0,0000651

This variant has conflicting interpretation in ClinVar, previously reported as likely pathogenic with HCM phenotype. Predicted tolerated.

MYH6 14: 23859551 c.3447C>G p.Ser1149Arg rs564367705 MAF ExAc: 0.0001318

This variant is predicted deleterious by PolyPhen2, SIFT and Mutation Taster.

MYH6 14: 23874969 c.212T>C p.V71A MAF ExAc: 0,00001647

This variant is predicted deleterious by PolyPhen2, SIFT and Mutation Taster.

MYH6 14: 23857395 c.4328C>A p.A1443D MAF ExAc: 0,0000659

This variant is predicted deleterious by PolyPhen2, SIFT and Mutation Taster

MYH7 14: 23883309 c.5562G>A p.T1854= rs368706722 MAF ExAc: 0,000066 MYH6 14: 23857507 c.4216G>A p.V1406M rs201566738 MAF ExAc: 0,0000494

In silico analyses for MYH6 rs201566738 variant are contradictory: it is predicted deleterious by PolyPhen and Mutation Taster but tolerated by SIFT. MYH7 rs368706722 variant is predicted disease causing by mutation taster, due to splice site changes.

MYH6 14: 23855783 c.4700T>G p.Phe1567Cys

This variant is reported with a single heterozygous carrier in ExAc. It is predicted deleterious by Polyphen2, Mutation Taster, SIFT.

Adjunctive phenotypical descriptions of PPrV carriers are available upon request

Supplementary References 1) Myllykangas S, Buenrostro JD, Natsoulis G, Bell JM, Ji HP. Efficient targeted resequencing of human germline and cancer genomes by oligonucleotide-selective sequencing. Nat Biotechnol 2011;29:1024–1027. 2) Ortiz-Genga M, Cuenca S, Dal Ferro M et al. Truncating FLNC mutations are associated with high-risk dilated and arrhythmogenic . J Am Coll Cardiol. 2016 Dec 6;68(22):2440-2451 3) Akinrinade O, Ollila L, Vattulainen S et al. Genetics and genotype-phenotype correlations in Finnish patients with dilated cardiomyopathy. Eur Heart J 2015;36:2327-37. 4) Begay RL, Graw S, Sinagra G, Merlo M, et al. Role of Missense Variants in Dilated Cardiomyopathy. J Am Heart Assoc. 2015 Nov 13;4(11). pii: e002645. 5) Hershberger RE, Parks SB, Kushner JD et al. Coding sequence mutations identified in MYH7, TNNT2, SCN5A, CSRP3, LBD3, and TCAP from 313 patients with familial or idiopathic dilated cardiomyopathy. Clin Transl Sci. 2008;1:21–26. 6) Norton N, Robertson PD, Rieder MJ et al. Evaluating Pathogenicity of Rare Variants From Dilated Cardiomyopathy in the Exome Era. Circ Cardiovasc Genet. 2012;5:167-174