Myocyte Specific Upregulation of ACE2 in Cardiovascular Disease: Implications for SARS-Cov-2 Mediated Myocarditis

Supplementary Materials for

Myocyte Specific Upregulation of ACE2 in Cardiovascular Disease: Implications for SARS-CoV-2 mediated myocarditis

Nathan R. Tucker,1,2* Mark Chaffin,1* Kenneth C. Bedi Jr.,3 Irinna Papangeli,4 Amer-Denis Akkad,4 Alessandro Arduini,1 Sikander Hayat,4 Gökcen Eraslan,5 Christoph Muus,5,6 Roby Bhattacharyya,5,7 Christian M. Stegmann,4 Human Cell Atlas Lung Biological Network, Kenneth B. Margulies,3 Patrick T. Ellinor1,7

1. Precision Cardiology Laboratory, The Broad Institute, Cambridge, MA, USA 02142 2. Masonic Medical Research Institute, Utica, NY, USA 13501 3. Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA 19104 4. Precision Cardiology Laboratory, Bayer US LLC, Cambridge, MA, USA 02142 5. The Broad Institute of MIT and Harvard, Cambridge, MA, USA 02142 6. John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA, USA 02138 7. Infectious Diseases Division, Department of Medicine, Massachusetts General Hospital, Boston, MA 02114 8. Cardiovascular Research Center, Massachusetts General Hospital, Boston, MA, USA 02114

* These authors contributed equally

Running Title: Myocyte upregulation of ACE2 in cardiovascular disease

Keywords: COVID-19, SARS-CoV-2, heart, single cell sequencing, myocarditis, cardiovascular disease

Corresponding Author: Patrick T. Ellinor, MD, PhD The Broad Institute of MIT and Harvard 75 Ames Street Cambridge, MA 02142 [email protected]

Index for Supplemental Materials

Item Page HCA Lung Biological Network Consortium Members 3 Supplemental Methods 8

Supplemental Tables 1. Baseline characteristics of study participants 12 2. Differential expression of ACE2 in left ventricular cell subtypes from individuals with dilated cardiomyopathy versus non-failing 13 controls 3. Differential expression of ACE2 in left ventricular cell subtypes from individuals with hypertrophic cardiomyopathy versus non- 14 failing controls 4. Sensitivity analysis comparing the expression of ACE2 in left ventricles samples from individuals with hypertrophic cardiomyopathy compared to non-failing controls after 15 exclusion of individuals on ACE inhibitors or Angiotensin II Receptor Blockers. 5. Comparison of the expression of ACE2 in cell subtypes of left ventricular tissue in the subset of patients with hypertrophic 16 cardiomyopathy that were (ACEi) or were not taking (No ACEi) an ACEi at the time of transplantation Supplemental Figure 1. Cell type resolved expression of SARS-CoV-2 associated proteases TMPRSS1 and CTSL in left ventricular tissue from non- 17 failing controls References 18 HCA Lung Biological Network Consortium Members

First name Last name Affiliation Nicholas E. Banovich Translational Genomics Research Institute, Phoenix, AZ. Pascal Barbry Université Côte d’Azur, CNRS, IPMC, Sophia-Antipolis, 06560, France Center for Immunology and Inflammatory Diseases, Massachusetts General Hospital, Boston, Katharine E. Black MA 02114 European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Alvis Brazma Wellcome Trust Genome Campus, Hinxton, Cambridge, CB10 1SD, UK Biosciences Institute, Faculty of Medical Sciences, Newcastle University, International Centre for Joseph Collin Life, Bioscience West Building, Newcastle upon Tyne NE1 3 BZ, UK Charité - Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt- Christian Conrad Universität zu Berlin, and Berlin Institute of Health, Charitéplatz 1, 10117 Berlin, Germany Department of Medicine and Institute for Stem Cell Biology and Regenerative Medicine, Tushar Desai Stanford University School of Medicine, Stanford, CA 94116 University of California San Diego and Rady Children’s Hospital San Diego, Department of Thu Elizabeth Duong Pediatrics Division of Respiratory Medicine Division of Pulmonary Sciences and Critical Care Medicine, Department of Medicine, University Oliver Eickelberg of Colorado, Anschutz Medical Campus, Aurora, CO, US Charité - Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt- Roland Eils Universität zu Berlin, and Berlin Institute of Health, Charitéplatz 1, 10117 Berlin, Germany Respiratory Bioinformatics and Molecular Biology, University of Technology Sydney, Sydney, Alen Faiz New South Wales, Australia. Institute of Transplant Immunology, Hannover Medical School, MHH, Carl-Neuberg Str. 1, 30625 Hannover, Germany, phone +40 511 532 9745; fax +40 511 532 8090; German Center for Christine Falk Infectious Diseases DZIF, TTU-IICH 07.801 Department of Immunology and Microbiology, The Scripps Research Institute, Jupiter, Florida, Michael Farzan USA (33458) Ian Glass Department of Pediatrics, Genetic Medicine, University of Washington, Seattle, Washington; Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, Cambridge CB10 1SA, UK; Biosciences Institute, Faculty of Medical Sciences, Newcastle University, Newcastle upon Tyne NE2 4HH, UK; Department of Dermatology and NIHR Newcastle Biomedical Research Centre, Muzlifa Haniffa Newcastle Hospitals NHS Foundation Trust, Newcastle upon Tyne NE2 4LP, UK. Synthetic and Systems Biology Unit, Biological Research Centre (BRC), Szeged, Hungary; Institute Peter Horvath for Molecular Medicine Finland, University of Helsinki Cardiovascular and Metabolic Sciences, Max Delbrück Center for Molecular Medicine in the Helmholtz Association (MDC), 13125 Berlin, Germany; DZHK (German Centre for Cardiovascular Research), Partner Site Berlin, 13347 Berlin, Germany; Berlin Institute of Health (BIH), 10178 Norbert Hubner Berlin, Germany; Charité-Universitätsmedizin, 10117 Berlin, Germany Professor of Genetics, Department of Genetics at Harvard Medical School and Department of Molecular Biology at Massachusetts General Hospital; Co-Director, Infectious Disease and Deborah Hung Microbiome Program and Core Faculty Member, Broad Institute of MIT & Harvard Naftali Kaminski Pulmonary, Critical Care and Sleep Medicine, Yale University School of Medicine Department of Pediatric Pulmonology and Pediatric Allergology, Beatrix Children’s Hospital, University of Groningen, University Medical Center Groningen (UMCG), Groningen Research Gerard Koppelman Institute for Asthma and COPD, Groningen, Netherlands Mark Krasnow Department of Biochemistry and Wall Center for Pulmonary Vascular Disease Division of Allergy, Pulmonary and Critical Care Medicine, Department of Medicine, Vanderbilt University Medical Center, Nashville, TN; Department of Veterans Affairs Medical Center, Nashville, TN; Department of Cell and Developmental Biology, Vanderbilt University, Nashville, Jonathan A. Kropski TN Malte Kuhnemund Cartana AB, Nobels vag 16, 17165 Stockholm, Sweden Division of Rheumatology, Department of Medicine, University of Pittsburgh Medical Center, Robert Lafyatis Pittsburgh, PA, USA. Biosciences Institute, Faculty of Medical Sciences, Newcastle University, International Centre for Majlinda Lako Life, Bioscience West Building, Newcastle upon Tyne NE1 3 BZ, UK Haeock Lee Department of Biomedicine and Health Sciences, The Catholic University of Korea, Seoul, Korea Université Côte d'Azur, CHU de Nice, FHU OncoAge, Department of Pulmonary Medicine and Allergology, Nice, France ; CNRS UMR 7275 - Institut de Pharmacologie Moléculaire et Cellulaire, Sylvie Leroy Sophia Antipolis, France Division of Molecular Neurobiology, Department of Medical Biochemistry and Biophysics, Sten Linnarson Karolinska Institute Joakim Lundeberg SciLifeLab, Department of Gene Technology, KTH Royal Institute of Technology Center for Immunology and Inflammatory Diseases, Massachusetts General Hospital, Boston, Benjamin Medoff MA 02114 Kerstin Meyer Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, Cambridge, UK Chichau Miao Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, Cambridge CB10 1SA, UK Alexander Misharin Division of Pulmonary and Critical Care Medicine, Northwestern University, Chicago, Illinois Department of Pathology and Medical Biology, University of Groningen, GRIAC Research Martijn Nawijn institue, University Medical Center Groningen, the Netherlands Marko Nikolic UCL Respiratory, Division of Medicine, University College London, London, UK National Heart and Lung Institute, Imperial College London, UK; British Heart Foundation Centre Michela Noseda for Research Excellence and Centre for Regenerative Medicine, Imperial College London, UK Division of Gastroenterology Boston Children's Hospital, Boston, MA, USA; Program in Ordovas Immunology, Harvard Medical School, Boston, MA, USA; Broad Institute of MIT and Harvard, Jose Montanes Cambridge, MA, USA; Harvard Stem Cell Institute, Cambridge, MA, USA. Canada Research Chair in Heart Failure, Division of Cardiology, 2C2 Walter Mackenzie Health Gavin Oudit Sciences Centre, Edmonton, Alberta, T6G 2B7, Canada Computational and Systems Biology Program, Sloan Kettering Institute, Memorial Sloan Dana Pe'er Kettering Cancer Center, New York, New York, USA Garvan-Weizmann Centre for Cellular Genomics, Garvan Institute of Medical Research, Sydney, NSW, Australia; UNSW Cellular Genomics Futures Institute, University of New South Wales, Joseph Powell Sydney, NSW, Australia Depts of Bioengineering and Applied Physics, Stanford University, and the Chan Zuckerberg Steve Quake Biohub. Harvard Stem Cell Institute, Cambridge, Massachusetts; Center for Regenerative Medicine, Jay Rajagopal Massachusetts General Hospital, Boston, Massachusetts Department of Cell Biology, Regeneration Next Initiative, Duke University School of Medicine, Purushothama Rao Tata Durham, NC, USA, 27710 Wellcome Trust/ CRUK Gurdon Institute and Department Physiology, Development and Emma Rawlins Neuroscience, University of Cambridge Klarman Cell Observatory, Broad Institute of MIT and Harvard, Howard Hughes Medical Aviv Regev Institute, Department of Biology, Massachusetts Institute of Technology, Cambridge MA 02142 Northwestern University Feinberg School of Medicine, Division of Pulmonary and Critical Care Paul Reyfman Medicine Mauricio Rojas Division of Pulmonary, Allergy and Critical Care Medicine, University of Pittsburgh Rosenblatt- Orit Rosen Klarman Cell Observatory, Broad Institute of Harvard and MIT, Cambridge, MA 02142, USA Department of Surgery, University of Cambridge and NIHR Cambridge Biomedical Research Kourosh Saeb-Parsy Centre, UK SciLifeLab, Department of Molecular Biosciences, Stockholm University, Stockholm Sweden and Christos Samakovlis Cardiopulmonary Institute, Justus Liebig University; Giessen Germany Comprehensive Pneumology Center (CPC) / Institute of Lung Biology and Disease (ILBD), Helmholtz Zentrum München, Member of the German Center for Lung Research (DZL), Munich, Herbert Schiller Germany Department for Genomics & Immunoregulation, LIMES-Institute, University of Bonn, 53115 Bonn, Germany and 2 PRECISE Platform for Single Cell Genomics & Epigenomics, German Center Joachim Schultze for Neurodegenerative Diseases and University of Bonn, Bonn, Germany Roland Schwarz Max Delbrück Center for Molecular Medicine, Berlin, DE Ayellet Segre Ocular Genomics Institute, Massachusetts Eye and Ear Infirmary, Boston, MA, USA 02114 Department of Pediatrics; Center for Genes, Environment, and Health; National Jewish Health; Max Seibold Denver, CO 80206 Jon Seidman Department of Genetics, Harvard Medical School, Boston, MA 02115, USA Department of Genetics, Harvard Medical School, Boston, MA 02115, USA; Cardiovascular Division, Brigham & Women’s Hospital, Boston, MA 02115, USA; Howard Hughes Medical Christine Seidman Institute Ragon Institute of MGH, MIT, and Harvard, Cambridge, MA, USA; Institute for Medical Engineering and Science (IMES), Koch Institute for Integrative Cancer Research, and Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA, USA; Broad Institute of Alex Shalek MIT and Harvard, Cambridge, MA, USA Center for Biological Physics and Department of Physics, Arizona State University, Tempe, AZ Douglas Shepherd USA Department of Internal Medicine, Gastroenterology, University of Michigan Medical School, Ann Arbor, MI 48109, USA; Department of Cell and Developmental Biology, University of Michigan Medical School, Ann Arbor, MI 48109, USA; Department of Biomedical Engineering, University of Jason Spence Michigan College of Engineering, Ann Arbor, MI 48109, USA. Department of Medicine, Boston University School of Medicine, Boston, MA, USA; Johnson & Avi Spira Johnson Innovation, Cambridge, MA, USA. Department of Pediatrics, Department of Biological Sciences, University of California SD, 9500 Xin Sun Gilman Dr. MC0766, San Diego, CA 92093-0766 Division of Neurodegeneration, Department of Neurobiology, Care Sciences and Society, Erik Sundström Karolinska Institutet and Stiftelsen Stockholms Sjukhem, Stockholm, Sweden Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, Cambridge, CB10 1SA, UK; Theory of Condensed Matter Group, Cavendish Laboratory/Department of Physics, University of Sarah Teichmann Cambridge, Cambridge CB3 0HE, UK Fabian J Theis, Institute of Computational Biology, Helmholtz Zentrum München and Fabian Theis Departments of Mathematics and Life Sciences, Technical University Munich, Germany Alexander Tsankov Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New York, NY USA Wellcome and MRC Cambridge Stem Cell Institute, Jeffrey Cheah Biomedical Centre, Biomedical Campus, Puddicombe Way, Cambridge CB2 0AW, UK; Department of Surgery, Cambridge Ludovic Vallier Biomedical Campus, Hills Rd, Cambridge, CB2 0QQ, UK Department of Pulmonary diseases and tuberculosis, University of Groningen, GRIAC Research Maarten van den Berge institue, University Medical Center Groningen, the Netherlands Jeffrey Whitsett Cincinnati Children’s Hospital Medical Center, Cincinnati, OHIO Ramnik Xavier Department of Molecular Biology, Massachusetts General Hospital and Broad Institute Divisions of Pulmonary Biology and Biomedical Informatics; Perinatal Institute, Cincinnati Yan Xu Children's Hospital Medical Center; University of Cincinnati College of Medicine Laure Emmanuelle Zaragosi Université Côte d’Azur, CNRS, IPMC, Sophia-Antipolis, 06560, France Biosciences Institute, Faculty of Medical Sciences, Newcastle University, International Centre for Life, Bioscience West Building, Newcastle upon Tyne NE1 3 BZ, UK; Microscopy Centre and Department of Applied Clinical Sciences and Biotechnology, University of L’Aquila, via Vetoio, Darin Zerti 67100 Coppito, L’Aquila, Italy UCSD Department of Bioengineering, 9500 Gilman Drive, MC0412, PFBH402, La Jolla, CA 92093- Kun Zhang 0412 Key Laboratory for Stem Cells and Tissue Engineering, Ministry of Education and Department of Histology and Embryology of Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou Hongbo Zhang 510080, China

Methods

Human tissue samples Procurement of human myocardial tissue was performed as described previously under protocols and ethical regulations approved by Institutional Review Boards at the University of Pennsylvania and the Gift-of-Life Donor Program (Pennsylvania, USA). Failing human hearts were procured at the time of orthotopic heart transplantation at the Hospital of the University of Pennsylvania after informed consent from all participants. Non-failing hearts were obtained at the time of organ donation from brain-dead organ donors after consent for research use of heart tissue from the next- of-kin. In all cases, hearts were arrested in situ using ice-cold cardioplegia solution and transported on wet ice prior to snap freezing of transmural left ventricular free wall biopsies in liquid nitrogen. Samples were stored at -80°C until further processing and analysis.

Bulk RNA-sequencing

Total RNA was extracted using the miRNeasy Kit (Qiagen) including DNAse treatment. RNA concentration and quality was determined using the NanoVue Plus™ spectrophotometer (GE Healthcare) and the Agilent 2100 RNA Nano Chip (Agilent). RNA sequencing libraries were prepared using the Illumina TruSeq stranded mRNA kit followed by the Nugen Ovation amplification kit. To avoid confounding by batch effects, libraries were randomly selected into pools of 32, and pools were sequenced on a Hiseq2500 to a depth of ~30 million 100-bp paired-end reads per biological sample. Fastq files were aligned against human reference (hg19/hGRC37) using the STAR aligner.1 Duplicate reads were removed using MarkDuplicates from Picard tools, and per gene read counts for Ensembl (v75) gene annotations were computed. Expression levels in counts per million (CPM) were normalized and transformed using VOOM in the LIMMA R package.2 Surrogate variables to account for sources for latent variation such as batch were calculated using the svaseq function from the R SVA package.3

Single nucleus RNA-sequencing Single nucleus RNA sequencing libraries were produced using the 10X genomics 3’ solution v3 according to manufacturer’s instructions. Single nuclear suspensions were generated using a combination of differential lysis, gradient centrifugation and sequential filtering as previously described.4 Sequencing of libraries was performed at the Genomics Platform of the Broad Institute of Harvard and MIT on a NextSeq500.

Data processing Single nuclei RNAs sequencing was performed in replicate on left ventricles from 12 individuals with dilated cardiomyopathy (DCM), 16 individuals with hypertrophic cardiomyopathy (HCM), and 16 non- failing (NF) hearts. Standard data processing with the 10x Genomics toolkit CellRanger 3.1.0 was applied with the addition of a read trimming step to remove homopolymer repeats and the template switch oligo from reads with the tool cutadapt.5 Low quality samples were removed based on the following CellRanger metrics: < 50% of reads in cells, < 75% of reads confidently mapping to transcriptome, < 90% valid barcodes, or abnormally low Q30. Additionally, the relationship between the rank of total UMI and total UMI was visually inspected per sample to ensure a sufficient ambient plateau for subsequent cell calling and background removal with CellBender remove-background.6 In total 8 libraries were excluded and we retained nuclei for at least 1 replicate of 11 DCM individuals, 15 HCM individuals, and 16 NF individuals (Supplemental Table 1). CellBender remove-background We applied the remove-background tool from CellBender v0.1 to the remaining samples to call non- empty droplets and correct for ambient background RNA contamination. One gene, MALAT1, had abnormally high counts and was removed from all downstream analysis. scR-Invex The tool scR-Invex (https://github.com/broadinstitute/scrinvex) was applied to assign reads to exons, introns, or a combination of both. For each cell barcode, we calculated the ratio of total exonic reads to total reads in the cell. Increased proportions of exonic reads may indicate a larger component of cytoplasmic material captured in a given droplet.

Nuclei QC Low quality nuclei were removed on a per-sample basis prior to jointly aggregating all samples together. In brief, the process was done in five steps:

1. Perform an initial clustering in scanpy7 at high resolution using the Louvain algorithm. Resolutions ranged from 1.6 to 2.5. 2. Calculate the median proportion of exonic reads, median fraction of UMI mapping to mitochondrial genes, and the median entropy per cluster. Entropy was calculated using Bayesian entropy estimation from the ndd python library (https://pypi.org/project/ndd/1.6.3/). Outlier clusters were generally identified as above the 3rd quartile plus 1.5 times the interquartile range for the proportion of exonic reads and the fraction of mitochondrial reads, and those below the 1st quartile minus 1.5 times the interquartile range for entropy. These clusters were removed. 3. For all remaining clusters, hierarchical clustering was applied to expression centroids and similar clusters were merged. 4. Low quality nuclei were identified on a per-cluster basis using an EllipticEnvelope (contamination = 0.05) outlier detection algorithm from scikit-learn8 in three dimensional space using the same quality control metrics as in step 2. Additionally, nuclei with log(number of genes) * entropy greater than the 90th percentile were removed to exclude potential doublets. Finally, any nuclei with less than 150 genes detected or a fraction of mitochondrial reads greater than 0.05 were removed. 5. A final clustering at more typical resolution (0.5) was performed and outlier clusters were again identified with a similar process as in step 2.

Map Aggregation Highly variable genes were selected using Seurat v3.1.0 on normalized expression using the FindVariableFeatures function. The scVI algorithm9 was applied to counts from these highly variable genes to generate a latent space with 50 latent variables. To account for biological heterogeneity between individuals, a categorical indicator variable for each biological replicate was included as a batch variable in scVI. A neighborhood graph was constructed based on this latent space in scanpy using the function sc.pp.neighbors with n_neighbors = 15. Initial clustering with the Leiden clustering algorithm (sc.tl.leiden in scanpy) revealed clusters that appear to be doublets of multiple cell types as they expressed high levels of marker genes from multiple clusters. We therefore ran the algorithm Scrublet10 on a per-sample basis using default parameters and removed any clusters that had an enriched predicted doublet score.

A neighborhood graph was constructed using the remaining 677,785 nuclei and Leiden clustering was performed at a resolution of 0.45. Marker genes were identified by calculating an area under the receiver operator curve (AUC) value for each gene comparing the expression in a target cluster versus all others. Inspection of genes with AUC > 0.70 for a given cluster was used to assign a cell type label to each cluster of nuclei.

Sub-clustering Analysis Sub-clustering was performed on each major cluster identified by selecting nuclei from a given cluster and performing a higher resolution Leiden clustering. This revealed that a population of mural cells from the initial clustering consisted of both pericytes and vascular smooth muscle cells. These two populations were treated as separate clusters for the remaining downstream analysis. Additionally, some sub-clusters from particular clusters appear to represent low quality nuclei as evidenced by an increased fraction of mitochondrial reads. These clusters were removed prior to differential expression testing.

ACE2 analysis Differential expression was tested between DCM and non-failing hearts and HCM and non-failing hearts within each cell type. To account for inter-nuclei correlation within sample, counts for each gene were summed across nuclei within each sample as motivated by Lun and Marioni, 2017.11 Summation was performed only if a sample had at least 20 nuclei for the given cluster to avoid introducing noisy counts. Technical replicates were aggregated together prior to downstream analysis. Lowly expressed genes were removed using the edgeR function filterByExpr(group=disease).12 We employed DESeq2 normalization to account for varying library size and then employed the limma-voom pipeline2 using a design of “1 + disease + sex”. P-values were adjusted for multiple testing across all genes using a

Benjamini-Hochberg correction. To visualize expression on a sample level, log2(CPM) values were extracted with the edgeR function cpm(prior.count = 3). As a sensitivity analysis for the HCM vs NF analysis, we excluded all individuals on ACE inhibitors or Angiotensin II Receptor Blockers (ARBs) and repeated the above analysis. We could not perform the same sensitivity analysis for DCM because all DCM individuals were on either an ACE inhibitor or ARB.

To assess the contribution of ACE inhibitors to ACE2 expression, we performed a differential expression analysis by ACE inhibitor status within our HCM samples. We excluded NF hearts and DCM hearts from this analysis because only 1 NF heart was on an ACE inhibitor and the only 2 DCM hearts not on ACE inhibitors were on ARBs, limiting our ability to glean information from these samples. Additionally, we excluded 1 HCM individual on an ARB. In total, we compared 6 HCM individuals on ACE inhibitors to 8 individuals not. We constructed a similar model as above, but instead used the following design “1 + ACEi + sex + age + LVEF” where LVEF was the LV ejection fraction for each individual, to better control for potential confounders.

Data availability Given the urgent nature of the current pandemic, this analysis was confined to the genes known to interact with COVID-19 as described above. The cell subtype expression levels for each of the genes described in this analysis will be available on the Single Cell Portal at the Broad Institute (https://singlecell.broadinstitute.org/single_cell) upon publication. The description of the expression changes observed between dilated cardiomyopathy, hypertrophic cardiomyopathy and non-failing controls will be the basis of a distinct analysis, and the full snRNAseq dataset from these samples will be released upon publication of that manuscript.

Table 1: Baseline characteristics of study participants

Etiology N Female (%) Age (years) LVMI (g/m2) LVEF (%) Non-failing controls 16 62.5 57 ± 10 98 ± 11 59 ± 7 Dilated cardiomyopathy 11 36.3 55 ± 9 177 ± 35 13 ± 3 Hypertrophic cardiomyopathy 15 33.3 49 ± 10 156 ± 49 47 ± 16

Data are mean ± standard deviation. Abbreviations: LVMI, left ventricular mass index; LVEF, left ventricular ejection fraction.

Table 2. Differential expression of ACE2 in left ventricular cell subtypes from individuals with dilated cardiomyopathy versus non-failing controls.

Gene Cell Type NDCM NNF GT0DCM GT0NF logFC P value ACE2 Cardiomyocyte I 11 16 23.5 5.2 1.59 1.70e-08 ACE2 Fibroblast I 11 16 0.3 17.3 -5.55 2.54e-08 ACE2 Endothelial I 11 16 0.1 0.6 NA NA ACE2 Endothelial II 6 10 1 1.6 NA NA ACE2 Pericyte 11 16 7.3 37.2 -1.83 2.52e-04 ACE2 Vascular Smooth Muscle 11 16 1.4 8 -1.86 3.19e-08 ACE2 Macrophage 11 16 0.1 0.5 NA NA ACE2 Lymphocyte I 11 16 0.1 0.6 NA NA ACE2 Lymphocyte II 9 13 0.2 0.2 NA NA ACE2 Lymphatic Endothelial 8 8 0.4 0.9 NA NA ACE2 Neuronal 9 15 0.2 2.1 NA NA ACE2 Adipocyte 5 10 0.6 3.5 NA NA

Abbreviations: N, number of individuals; DCM, dilated cardiomyopathy; NF, non-failing; GTO, percentage of cells with a non-zero count for ACE2; logFC, log transformed fold change comparing ACE2 expression in left ventricular tissue from dilated cardiomyopathy to non-failing samples.

Table 3. Differential expression of ACE2 in left ventricular cell subtypes from individuals with hypertrophic cardiomyopathy versus non-failing controls.

Gene Cell Type NHCM NNF GT0HCM GT0NF logFC P value ACE2 Cardiomyocyte I 15 16 18.5 5.2 2.00 2.84e-10 ACE2 Fibroblast I 15 16 0.3 17.3 -5.38 4.30e-10 ACE2 Endothelial I 15 16 0.2 0.6 NA NA ACE2 Endothelial II 8 10 0.4 1.6 NA NA ACE2 Pericyte 15 16 6.5 37.2 -2.28 1.40e-06 ACE2 Vascular Smooth Muscle 15 16 1.5 8 -2.31 2.39e-08 ACE2 Macrophage 15 16 0.05 0.5 NA NA ACE2 Lymphocyte I 15 16 0.2 0.6 NA NA ACE2 Lymphocyte II 9 13 0.1 0.2 NA NA ACE2 Lymphatic Endothelial 13 8 0.4 0.9 NA NA ACE2 Neuronal 15 15 1.1 2.1 NA NA ACE2 Adipocyte 11 10 2.7 3.5 NA NA

Abbreviations: N, number of individuals; HCM, hypertrophic cardiomyopathy; NF, non-failing; GTO, percentage of cells with a non-zero count for ACE2; logFC, log transformed fold change comparing ACE2 expression in left ventricular tissue from hypertrophic cardiomyopathy versus non-failing samples.

Table 4. Sensitivity analysis comparing the expression of ACE2 in left ventricles samples from individuals with hypertrophic cardiomyopathy compared to non-failing controls after exclusion of individuals on ACE inhibitors or ARBs.

Gene Cell Type NHCM NNF GT0HCM GT0NF logFC P value ACE2 Cardiomyocyte I 8 15 15.6 5.2 1.76 5.85e-8 ACE2 Fibroblast I 8 15 0.3 18.7 -5.40 2.12e-8 ACE2 Endothelial I 8 15 0.2 0.6 NA NA ACE2 Endothelial II 3 9 0.7 1.8 NA NA ACE2 Pericyte 8 15 5.6 38.0 -2.59 1.15e-4 ACE2 Vascular Smooth Muscle 8 15 1.4 7.7 -2.33 4.75e-7 ACE2 Macrophage 8 15 0.02 0.5 NA NA ACE2 Lymphocyte I 8 15 0.1 0.6 NA NA ACE2 Lymphocyte II 4 13 <0.001 0.2 NA NA ACE2 Lymphatic Endothelial 7 7 <0.001 0.9 NA NA ACE2 Neuronal 7 14 0.5 2.3 NA NA ACE2 Adipocyte 8 15 3.2 3.6 NA NA

Abbreviations: N, number of individuals; GTO, percentage of cells with a non-zero count for ACE2; logFC, log transformed fold change comparing ACE2 expression in left ventricular tissue from hypertrophic cardiomyopathy versus non-failing samples.

Table 5. Comparison of the expression of ACE2 expression in cell subtypes of left ventricular tissue in the subset of patients with hypertrophic cardiomyopathy that were (ACEi) or were not taking (No ACEi) an ACE inhibitor at the time of transplantation.

Gene Cell Type NACEi NNo_ACEi GT0ACEi GT0No_ACEi logFC P value ACE2 Cardiomyocyte I 6 8 24.5 15.6 0.70 0.09 ACE2 Fibroblast I 6 8 0.5 0.3 0.51 0.45 ACE2 Endothelial I 6 8 0.1 0.2 NA NA ACE2 Endothelial II 3 4 0.1 0.7 NA NA ACE2 Pericyte 6 8 8.9 5.6 0.71 0.24 ACE2 Vascular 6 8 0.13 0.90 Smooth Muscle 1.8 1.4 ACE2 Macrophage 6 8 0.1 0 NA NA ACE2 Lymphocyte I 6 8 0.2 0.1 NA NA ACE2 Lymphocyte II 4 4 0.2 0 NA NA ACE2 Lymphatic 5 7 NA NA Endothelial 0.9 0 ACE2 Neuronal 6 7 2.3 0.4 NA NA ACE2 Adipocyte 4 6 1.8 3.2 NA NA

Abbreviations: N, number of individuals; GTO, percentage of cells with a non-zero count for ACE2; logFC, log transformed fold change comparing individuals on ACE inhibitor vs those not on an ACE inhibitor

Supplemental Figure 1: Cell type resolved expression of SARS-CoV-2 associated proteases TMPRSS2 and CTSL in left ventricular tissue from non-failing controls. Size and hue of dots represents the percentage of nuclei with non-zero counts and log transformed counts in each respective cell type. Unlike in the joint analyses, there were no activated fibroblast observed within these non-failing tissues, so that column was omitted.

References

1. Dobin, A. et al. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013). 2. Law, C. W., Chen, Y., Shi, W. & Smyth, G. K. Voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 15, R29 (2014). 3. Leek, J. T. et al. sva: Surrogate Variable Analysis. (2019). 4. Tucker, N. R. et al. Transcriptional and Cellular Diversity of the Human Heart. bioRxiv 2020.01.06.896076 (2020). doi:10.1101/2020.01.06.896076 5. Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.journal 17, 10 (2011). 6. Fleming, S. J., Marioni, J. C. & Babadi, M. CellBender remove-background: a deep generative model for unsupervised removal of background noise from scRNA-seq datasets. bioRxiv 791699 (2019). doi:10.1101/791699 7. Wolf, F. A., Angerer, P. & Theis, F. J. SCANPY: large-scale single-cell gene expression data analysis. Genome Biol. 19, 15 (2018). 8. Pedregosa, F. et al. Scikit-Learn: Machine Learning in Python. J. Mach. Learn. Res. 12, 2825–2830 (2011). 9. Lopez, R., Regier, J., Cole, M. B., Jordan, M. I. & Yosef, N. Deep generative modeling for single-cell transcriptomics. Nat. Methods 15, 1053–1058 (2018). 10. Wolock, S. L., Lopez, R. & Klein, A. M. Scrublet: Computational Identification of Cell Doublets in Single-Cell Transcriptomic Data. Cell Syst. 8, 281-291.e9 (2019). 11. Lun, A. T. L. & Marioni, J. C. Overcoming confounding plate effects in differential expression analyses of single-cell RNA-seq data. Biostatistics 18, 451–464 (2017). 12. Chen, Y., Lun, A. T. L. & Smyth, G. K. From reads to genes to pathways: Differential expression analysis of RNA-Seq experiments using Rsubread and the edgeR quasi-likelihood pipeline [version 2; referees: 5 approved]. F1000Research 5, 1438 (2016).