Identification of Cardiac Fibrosis in Young Adults With a Homozygous Frameshift Variant in SERPINE1

Key Points

Question  What is the genetic cause of cardiac fibrosis observed in young adults from 3 related Amish families?

Findings  In this study, cardiovascular phenotyping and genotyping of 17 participants (mean age, 24 years) was performed, including cardiovascular imaging, whole-exome sequencing, and examination of stem cell–derived cardiomyocytes. An autosomal recessive phenotype of cardiac fibrosis in individuals homozygous for a frameshift variant in SERPINE1, the gene that codes for plasminogen activator inhibitor-1 (PAI-1), was identified.

Meaning  In this study, genetic deficiency of PAI-1 was associated with early-onset cardiac fibrosis, suggesting an optimal range of PAI-1 is needed for cardiac homeostasis in humans.

Importance  Cardiac fibrosis is exceedingly rare in young adults. Identification of genetic variants that cause early-onset cardiomyopathy may inform novel biological pathways. Experimental models and a single case report have linked genetic deficiency of plasminogen activator inhibitor-1 (PAI-1), a downstream target of cardiac transforming growth factor β, with cardiac fibrosis.

Objective  To perform detailed cardiovascular phenotyping and genotyping in young adults from an Amish family with a frameshift variant (c.699_700dupTA) in SERPINE1, the gene that codes for PAI-1.

Design, Setting, and Participants  This observational study included participants from 3 related nuclear families from an Amish community in the primary analysis and participants from the extended family in the secondary analysis. Participants were recruited from May 2015 to December 2016, and analysis took place from June 2015 to June 2020.

Main Outcomes and Measures  (1) Multimodality cardiovascular imaging (transthoracic echocardiography and cardiac magnetic resonance imaging), (2) whole-exome sequencing, and (3) induced pluripotent stem cell–derived cardiomyocytes.

Results  Among 17 participants included in the primary analysis, the mean (interquartile range) age was 23.7 (20.9-29.9) years and 9 individuals (52.9%) were confirmed to be homozygous for the SERPINE1 c.699_700dupTA variant. Late gadolinium enhancement was present in 6 of 9 homozygous participants (67%) with absolute PAI-1 deficiency vs 0 of 8 in the control group (P = .001). Late gadolinium enhancement patterns tended to be dense and linear, usually subepicardial but also midmyocardial and transmural with noncoronary distributions. Targeted whole-exome sequencing analysis identified that homozygosity for c.699_700dupTA SERPINE1 was the only shared pathogenic variant or variant of uncertain significance after examination of cardiomyopathy genes among those with late gadolinium enhancement. Induced pluripotent stem cell–derived cardiomyocytes from participants homozygous for the SERPINE1 c.699_700dupTA variant exhibited susceptibility to cardiomyocyte injury in response to angiotensin II (increased transforming growth factor β1 secretion and release of lactate dehydrogenase) compared with control induced pluripotent stem cell–derived cardiomyocytes. In a secondary analysis based on echocardiography in 155 individuals across 3 generations in the extended family, no difference in global longitudinal strain was observed in carriers for the SERPINE1 c.699_700dupTA variant compared with wild-type participants, supporting an autosomal recessive inheritance pattern.

Conclusions and Relevance  In this study, a highly penetrant, autosomal recessive, cardiac fibrosis phenotype among young adults with homozygous frameshift variant for SERPINE1 was identified, suggesting an optimal range of PAI-1 levels are needed for cardiac homeostasis.

Quiz Ref IDPlasminogen activator inhibitor-1 (PAI-1) plays a key role in regulating fibrinolysis through inhibition of urokinase-type plasminogen activator and tissue-type plasminogen activator.¹ Beyond this well-established role, PAI-1 regulates cell migration, senescence, and extracellular matrix deposition, each of which contribute to fibrosis.²Quiz Ref ID In experimental models, increased PAI-1 levels correlate with fibrosis in vitro and in vivo.¹ Counterintuitively, the absence of PAI-1^2,3 has been reported to cause spontaneous cardiac fibrosis in murine models, suggesting the need of an optimal range of PAI-1 for cardiac homeostasis.

Loss-of-function variants in SERPINE1, the gene that codes for PAI-1, have the potential to inform biological relevance of PAI-1 in cardiac fibrosis among humans. We recently reported the incidental finding of cardiac fibrosis in 2 members of an Amish community who are homozygous for a frameshift variant (c.699_700dupTA) in the SERPINE1 gene resulting in a premature stop codon and nonfunctional truncated protein.^4,5 To follow up on these preliminary findings and better define the link between PAI-1 and cardiac fibrosis, we undertook a systematic study of detailed cardiovascular phenotyping (multimodality cardiovascular imaging [transthoracic echocardiography and cardiac magnetic resonance imaging (CMR)]), genetic sequencing (whole-exome sequencing [WES]), and induced pluripotent stem cell–derived cardiomyocytes (iPSC-CM).

We enrolled 17 participants from 3 related Amish families in the primary study between May 2015 and December 2016. Participants underwent detailed cardiac phenotyping, WES, and urine collection to generate iPSC-CM. In a secondary analysis, we performed speckle-tracking analysis on echocardiograms in 155 participants of the extended kindred. The institutional review boards at the Indiana Hemophilia and Thrombosis Center Inc and Northwestern University approved the study protocols. All participants provided written informed consent.

All study participants in the primary and secondary studies underwent comprehensive 2-dimensional transthoracic echocardiogram with Doppler and tissue Doppler imaging using a commercially available ultrasound system with harmonic imaging (Vivid 7 or Vivid I; GE Healthcare) and a 3.5-MHz transducer. Deidentified images were analyzed offline using GE EchoPAC software version 7.0.0 (GE Healthcare). All measurements were made by a single experienced sonographer and verified by an experienced investigator (S.J.S.) (both blinded to clinical data and genotype).⁶ Cardiac magnetic resonance imaging was performed on a clinical 1.5-T scanner (MAGNETOM Aera; Siemens Healthcare) with analysis in Medis QMass MR 7.6 (Medis Medical Imaging Systems) by a single experienced researcher (B.C.B.) and verified by an experienced investigator (D.C.L.) (both blinded to clinical data and genotype).

Exome sequencing was performed at the NUSeq Core Facility at the Northwestern University Feinberg School of Medicine. Sequencing and bioinformatics analyses are detailed in the eMethods in the Supplement.

We compared baseline demographics between participants homozygous for the c.699_700dupTA SERPINE1 variant and controls using models adjusted for age, sex, and family structure. For the primary analysis, our exposure defined a priori was homozygosity for the c.699_700dupTA SERPINE1 variant and the end point of interest was late gadolinium enhancement (LGE) on CMR based on the experimental literature. We also investigated the association of homozygosity for the c.699_700dupTA SERPINE1 variant with other cardiovascular parameters. Analyses were conducted using Sequential Oligogenic Linkage Analysis Routines, version 8.1.1 (Southwest Foundation for Biomedical Research) given the high degree of relatedness between participants in this closed founder population. Two-sided P values less than .05 for Sequential Oligogenic Linkage Analysis Routines analyses were considered to indicate statistical significance. We also calculated the estimated logarithm of the odds score according to the formula recommended by the Clinical Genome Resource Consortium (eMethods in the Supplement). Analysis began June 2015 and ended June 2020.

As a proof-of-concept study, we used previously characterized iPSC lines to pursue preliminary mechanistic in vitro studies.^7-9 Cardiomyocyte differentiation of iPSCs was achieved by following published protocols^10,11 with details outlined in the eMethods in the Supplement from 1 participant homozygous for the c.699_700dupTA SERPINE1 variant and 1 wild-type control participant. Experimental results examining cardiomyocyte response to angiotensin II and metabolic stress are presented as mean (SEM). Given that the multiple cardiomyocyte differentiations (n = 5) were derived from a single participant homozygous for the c.699_700dupTA SERPINE1 variant and a single wild-type participant, no formal statistical analyses were performed to conclude significant differences.

Of 17 participants enrolled in the primary study, the mean (interquartile range) age was 23.7 (20.9-29.9) years (Table). Participants homozygous for the SERPINE1 (c.699_700dupTA) had no evidence of detectable PAI-1 antigen or activity in plasma. No participants reported symptoms or a clinical diagnosis of heart failure. During follow-up, 1 SERPINE1^−/− participant experienced sudden death, but autopsy was declined by the family. In 3 years of follow-up, none of the other SERPINE1^−/− participants have reported signs or symptoms of heart failure or sustained arrhythmias with annual in-person contact.

Cardiovascular phenotyping parameters are shown in the eTable in the Supplement. Quiz Ref IDOf 9 SERPINE1^−/− participants, 6 had LGE (Figure 1) compared with 0 of 8 in the control group, including no heterozygous participants (median [interquartile range] % of fibrosis for the 2 groups: 7.9% [0%-13.4%] vs 0%; Wilcoxon 2-sample P = .001). Late gadolinium enhancement patterns tended to be dense and linear, usually subepicardial but also midmyocardial and transmural with noncoronary distributions, supporting a nonischemic etiology.

Targeted WES analysis performed focused only on known cardiomyopathy genes did not identify any pathogenic or likely pathogenic variants in cardiomyopathy genes in any study participant. While we did identify several distinct genetic variants of uncertain significance in cardiomyopathy genes, none of these variants visually segregated with the LGE phenotype among the participants (Figure 2). Homozygosity for c.699_700dupTA SERPINE1 was the only shared variant of uncertain significance with biologic plausibility associated with cardiac fibrosis among all participants with LGE. The estimated logarithm of the odds score generated for homozygosity for the c.699_700dupTA SERPINE1 variant was 4.01 (based on the Clinical Genome Resource Consortium formula).

Quiz Ref IDPreliminary studies in iPSC-CM (eFigure 1 in the Supplement) from a participant homozygous for the c.699_700dupTA SERPINE1 variant exhibited increased transforming growth factor β1 secretion in response to angiotensin II compared with control iPSC-CMs (eFigure 2 in the Supplement). Gene expression levels for connective tissue growth factor and Smad7 were increased among SERPINE1-null iPSC-CM compared with controls (eFigure 3 in the Supplement). This profibrotic transcriptional pattern was also present in response to stress and exposure to angiotensin II. Quiz Ref IDSERPINE1-null iPSC-CMs were associated with an increased stress response to H2O2 and angiotensin II as indicated by increased release of lactate dehydrogenase when compared with control iPSC-CMs, suggesting increased susceptibility to cardiomyocyte injury with PAI-1 deficiency.

A secondary analysis to examine global longitudinal strain from echocardiograms was undertaken in 155 previously recruited participants from the same Amish community. No differences in mean (SD) global longitudinal strain was observed between carriers for the c.699_700dupTA SERPINE1 variant (n = 36; 20.6% [2.5%]) and wild-type participants (n = 119; 20.6% [3.0%]).

In this analysis, we identified a highly penetrant, autosomal recessive pattern of inheritance of a cardiac fibrosis phenotype in young adults with a frameshift variant in SERPINE1 (c.699_700dupTA) based on rigorous assessment with multimodality imaging, targeted WES analysis, and preliminary iPSC-CM experimental studies. Further supporting the clinical validity of this gene-disease association is the fact that this rare variant has not been identified in any frequency in the general population (Genome Aggregation Database that includes data from more than 140 000 exomes and genomes). These findings build on extensive experimental evidence (in vivo and in vitro) from our group and others supporting the findings that complete PAI-1 deficiency predicates and is sufficient to promote spontaneous cardiac-specific fibrosis in mammals.²

The use of CMR in this study enabled us to characterize and quantify the pattern and extent of LGE among asymptomatic young adults with absolute PAI-1 deficiency who did not harbor any other known pathogenic variants in cardiomyopathy genes. While LGE is found in several distinct forms of genetic cardiomyopathies (eg, desmoplakin and lamin A/C cardiomyopathy),^12-14 the presence of LGE in young adults is uncommon, especially in the absence of significant left ventricular dilation or hypertrophy. The closest pattern of LGE to what we observed has been previously described in Duchenne muscular dystrophy.¹⁵ Myocarditis can also result in epicardial LGE; however, the enhancement pattern is typically more localized (eg, inferolateral wall) and less dense. The presence of dense LGE reflects in vivo replacement fibrosis of the myocardium and was only identified among individuals with complete PAI-1 deficiency, supporting the role of PAI-1 as an important regulator of cardiac fibrosis.

Limitations of this study include the small sample size of our analysis, but we successfully enrolled all 9 known individuals who are homozygous for this SERPINE1 variant to inform the cosegregation and penetrance analyses. Second, preliminary experiments from iPSC-CM are presented as hypothesis generating to provide mechanistic insights but only included a single participant who was homozygous for the c.699_700dupTA SERPINE1 variant and a single wild-type control. However, the proof-of-concept nature of these in vitro studies provides additional support of the biologic plausibility of the imaging and genetic analyses. Third, we did not have adequate power to examine effect modification by sex or other rare variants in noncardiomyopathy genes.

In conclusion, we identified a highly penetrant, autosomal recessive, cardiac fibrosis phenotype among young adults with homozygous frameshift variant for SERPINE1, suggesting an optimal range of PAI-1 levels are needed for cardiac homeostasis. Further studies are needed to inform the mechanistic pathways supporting the role of PAI-1 as a broader molecular driver of human myocardial fibrosis.

Corresponding Author: Douglas E. Vaughan, MD, Department of Medicine, Northwestern University Feinberg School of Medicine, 251 E Huron St, Galter Ste 3-150, Chicago, IL 60611 (d-vaughan@northwestern.edu).

Accepted for Publication: November 13, 2020.

Published Online: January 13, 2021. doi:10.1001/jamacardio.2020.6909

Correction: This article was corrected on July 12, 2021, to fix an error in the byline.

Author Contributions: Drs Vaughan and Khan had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Khan, Shah, Flevaris, McNally, Lee, Heiman, Gupta, Shapiro, Vaughan.

Acquisition, analysis, or interpretation of data: Khan, Shah, Strande, Baldridge, Flevaris, Puckelwartz, Rasmussen-Torvik, Lee, Carr, Benefield, Afzal, Heiman, Gupta, Shapiro, Vaughan.

Drafting of the manuscript: Khan, Strande, Gupta, Shapiro, Vaughan.

Critical revision of the manuscript for important intellectual content: Shah, Strande, Baldridge, Flevaris, Puckelwartz, McNally, Rasmussen-Torvik, Lee, Carr, Benefield, Afzal, Heiman, Vaughan.

Statistical analysis: Khan, Baldridge, Puckelwartz, Rasmussen-Torvik.

Obtained funding: Flevaris, Vaughan.

Administrative, technical, or material support: Strande, McNally, Lee, Benefield, Heiman, Shapiro, Vaughan.

Supervision: Shah, Strande, Lee, Carr, Heiman, Gupta, Vaughan.

Conflict of Interest Disclosures: Dr Khan reports grants from National Institutes of Health/National Center for Advancing Translational Sciences and American Heart Association during the conduct of the study. Dr Shah reports grants from Actelion, AstraZeneca, Pfizer, Novartis, and Corvia Medical and personal fees from Amgen, Aria CV, AstraZeneca, Axon, Bayer, Bristol Myers Squibb, Boehringer Ingelheim, Cardiora, Cyclerion Therapeutics, Cytokinetics, Eisai, GlaxoSmithKline, Imara, Ionis Pharmaceuticals, Ironwood, Keyto, Merck, MyoKardia, Regeneron Pharmaceuticals, Sanofi, Shifamed, United Therapeutics, Actellion, Pfizer, and Novartis outside the submitted work. Dr Strande reports grants from National Institutes of Health outside the submitted work. Dr McNally reports grants from National Institutes of Health and American Heart Association to Northwestern University during the conduct of the study; grants from National Institutes of Health, US Department of Defense, and Solid Biosciences to Northwestern University outside the submitted work; personal fees from AstraZeneca, Amgen, Avidity, Cytokinetics, Exonics/Vertex, Pfizer, Invitae, and Janssen outside the submitted work; and is founder of Ikaika Therapeutics, unrelated to this content. Dr Lee reports grants from Abbott Laboratories outside the submitted work. Dr Carr reports grants from Siemens, Bayer, and Guerbet and personal fees from Bracco (advisory board), Siemens (advisory board), and Bayer (speaker) outside the submitted work. Dr Shapiro reports other support from Bioverativ (advisory board and research), Catalyst Biosciences (advisory board), Genentech/Roche (advisory board, speakers bureau, and research), Agios Pharmaceuticals (research), BioMarin Pharmaceutical (research), Daiichi Sankyo (research), Glover Blood Therapeutics (research), Kedrion (research), Liminal Sciences (Prometic BioTherapeutics) (consultant and research), OPKO Health (research), Octapharma Plasma (research), Pfizer (advisory board and research), Novo Nordisk (advisory board, consultant, speakers bureau, and research), Sangamo Therapeutics (research), Takeda (advisory board and research), and Sigilon (advisory board) to their institution outside the submitted work. No other disclosures were reported.

Funding/Support: This work was supported by grants from the National Institutes of Health (grants R01 HL051387 and R01 HL142761) to Dr Vaughan. Research reported in this publication was supported, in part, by the National Institutes of Health’s National Center for Advancing Translational Sciences (grant KL2TR001424) and the American Heart Association (grant 19TPA34890060) to Dr Khan.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: Dr Shah is Associated Editor of JAMA Cardiology but was not involved in any of the decisions regarding review of the manuscript or its acceptance. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Additional Contributions: We thank our Amish liaisons and the Indiana Hemophilia and Thrombosis Center’s Research clinic staff. In addition, we extend our deepest gratitude to the Berne Amish community for their time, participation, and support, without which these studies would not have been possible. A pedigree is not included in the Supplement to protect anonymity. This work received analytic supported from the Northwestern University NUSeq Core Facility.