Author + information
- Received April 25, 2006
- Revision received February 9, 2007
- Accepted February 12, 2007
- Published online June 26, 2007.
- J. Peter van Tintelen, MD⁎,⁎ (, )
- Rene A. Tio, MD, PhD†,
- Wilhelmina S. Kerstjens-Frederikse, MD⁎,
- Jop H. van Berlo, MD‡,
- Ludolf G. Boven, BSc⁎,
- Albert J.H. Suurmeijer, MD, PhD§,
- Stefan J. White, PhD∥,
- Johan T. den Dunnen, PhD∥,
- Gerard J. te Meerman, PhD⁎,
- Yvonne J. Vos, MSc⁎,
- Annemarie H. van der Hout, PhD⁎,
- Jan Osinga⁎,
- Maarten P. van den Berg, MD, PhD†,
- Dirk J. van Veldhuisen, MD, PhD†,
- Charles H.C.M. Buys, PhD⁎,
- Robert M.W. Hofstra, PhD⁎ and
- Yigal M. Pinto, MD, PhD‡
- ↵⁎Reprint requests and correspondence:
Dr. J. Peter van Tintelen, Department of Genetics, University Medical Center Groningen, P.O. Box 30001, 9700 RB Groningen, the Netherlands.
Objectives The goal of this study was to identify the underlying gene defect in a family with inherited myocardial fibrosis.
Background A large family with an autosomal dominantly inherited form of myocardial fibrosis with a highly malignant clinical outcome has been investigated. Because myocardial fibrosis preceded the clinical and echocardiographic signs, we consider the disease to be a hereditary form of cardiac fibrosis.
Methods Twenty-five family members were clinically evaluated, and 5 unaffected and 8 affected family members were included in a genome-wide linkage study.
Results The highest logarithm of the odds (LOD) score (LOD = 2.6) was found in the region of the lamin AC (LMNA) gene. The LMNA mutation analysis, both by denaturing gradient gel electrophoresis and sequencing, failed to show a mutation. Subsequent Southern blotting, complementary deoxyribonucleic acid sequencing, and multiplex ligation-dependent probe amplification analysis, however, revealed a deletion of the start codon-containing exon and an adjacent noncoding exon. In vitro studies demonstrated that the deletion results in the formation of nuclear aggregates of lamin, suggesting that the mutant allele is being transcribed.
Conclusions This novel LMNA deletion causes a distinct, highly malignant cardiomyopathy with early-onset primary cardiac fibrosis likely due to an effect of the shortened mutant protein, which secondarily leads to arrhythmias and end-stage cardiac failure.
Dilated cardiomyopathy (DCM) is characterized by dilatation and impaired contractile function of the left ventricle. The mortality rate is high, with the 5-year survival rate around 50% (1). The incidence is about 4 of 10,000/year, making it one of the most important reasons for cardiac transplantation (2). Up to 50% of cases lack an underlying diagnosis, which justifies its classification as “idiopathic” dilated cardiomyopathy (IDCM) (3).
The familial character of IDCM has become increasingly clear in that several recent studies have shown that up to 35% of all IDCM patients have at least 1 affected (first-degree) relative (4–7). In at least 60% of these familial IDCM cases, an autosomal-dominant inheritance pattern is suggested (6–8).
Among the most frequently reported genetic causes of DCM, identified in up to 30% of cases of DCM in association with cardiac conduction disease or supraventricular arrhythmias, are mutations in the gene encoding lamin A and C (LMNA) proteins (9,10).
Dilated cardiomyopathy is clinically highly variable, encompassing the spectrum from frank dilatation with relatively little fibrosis to less dilative forms with more severe derangement of myocardial histology. Myocardial fibrosis accompanies virtually all forms of cardiomyopathy, including inherited forms, and is therefore commonly regarded to be a secondary phenomenon. The idea that fibrosis is secondary to heart failure has been fueled by the fact that genetic defects that give rise to extensive primary myocardial fibrosis have not as yet been described.
In presenting here a large family with a hereditary form of early onset cardiac fibrosis, we will be addressing the question of whether primary myocardial fibrosis can be considered as a separate disease entity or should be regarded as a subphenotype of DCM.
Investigated family and cardiological investigations
The investigated family has been known at our hospital for nearly 50 years (11). As part of our clinical genetic-counseling procedure, we asked patients to inform relatives about the possibility of genetic counseling and cardiological evaluation. Patients who were investigated gave their written informed consent. To establish the clinical phenotype, electrocardiographic recording, exercise testing, echocardiography, signal-averaged electrocardiography, and 24-h electrocardiography were performed. Owing to the extensive and early fibrosis that has been described this family (11), all family members suspected of being affected (as indicated by atrioventricular blockade or supraventricular arrhythmias on 24-h Holter monitoring) underwent a right-ventricular endomyocardial biopsy (EMB) after obtaining a fully normal coronary angiogram that included a left ventricular angiogram.
Paraffin sections of EMB and postmortem full-thickness sections of right and left ventricular myocardium were stained with hematoxylin and eosin. A Masson trichrome stain was used to determine the amount of collagen. Patients with manifest (interstitial) fibrosis were considered as being affected.
Deoxyribonucleic acid isolation
Blood samples were collected from all marked-affected and unaffected subjects of the family involved (Fig. 1).Genomic deoxyribonucleic acid (DNA) was extracted from peripheral blood leukocytes with the salting-out procedure (12).
Genome-wide mapping was performed with approximately 300 fluorescent microsatellite markers (ISOGEN Bioscience, Maarsen, the Netherlands), spread over the entire genome with an average resolution of 10 cM. DeCODE, Marshfield, and Genethon human-linkage maps were used as guidance for intermarker distances.
Standard polymerase chain-reaction (PCR) amplification was performed in 20 μl containing 150 ng of genomic DNA and 10 pmol of fluorescently labeled primers. Reaction products from 4 to 8 markers were pooled and separated on an automatic ABI 377 DNA sequencer or on a MegaBACE 1000. Marker analysis was performed with the GENESCAN v3.1 and GENOTYPER v2.5 software (Applied Biosystems, Foster City, California), respectively. Linkage analysis was performed with Gronlod (13) with a 5% recombination frequency, a penetrance of 90%, and a phenocopy rate of 10%.
Denaturing gradient-gel electrophoresis and sequence analysis of the LMNA gene
Point-mutation analysis for the LMNA gene was performed with both denaturing gradient-gel electrophoresis (DGGE) and sequencing. Thirteen amplicons were analyzed with both methods (J. P. van Tintelen, unpublished data, 2005; primers available upon request).
For sequencing, PCR products were purified with the High Pure PCR Product Purification Kit (Boehringer Mannheim, Mannheim, Germany) and then sequenced with the Thermo Sequenase Kit (Amersham Life Science, Buckinghamshire, United Kingdom) or the Sequenase kit (USB, Cleveland, Ohio).
Genomic DNA (8 μg) was digested either with BamHI and XbaI or with Bgl2 and EcoRV. The DNA fragments were separated overnight by gel electrophoresis in a 0.7% agarose gel and were transferred under denaturing conditions (0.4 mol/l sodium hydroxide and 0.6 mol/l sodium chloride) to a Hybond N+membrane. The filters were rinsed, air-dried, and baked at 80°C for 2 h. Prehybridization was carried out in a 0.5 mol/l phosphate buffer, 6% sodium dodecyl sulfate (SDS), 1.0 mmol/l ethylenediaminetetraacetic acid (EDTA) solution for 30 min at 65°C and hybridization in fresh solution overnight with [32P]dCTP-labeled LMNA complementary DNA (cDNA) probe (a 2027 base pair [bp] insert of an LMNA cDNA [IMAGp958B1610Q2]). Filters were washed twice in 2 × standard saline citrate (SSC), 0.1% SDS at 65°C for 15 min, and in 1 × SSC, 0.1% SDS for 15 min, and in 0.3 × SSC, 0.1% SDS for 10 min at 65°C. After washing the blot was exposed to X-ray Kodak XAR film with intensifying screens at −80°C for several days.
Ribonucleic acid (RNA; 2.5 μg), isolated from deletion carrier III-12, was used to make cDNA with the Ready-To-Go You-Prime-First-Strand Beads Kit (Amersham Biosciences) according to the manufacturer’s protocol. This cDNA was used to perform PCR with the exon −2 forward primer (CGGAGATCTCAGAGGCACCGAC) and an exon +3 reverse primer (GCAGCATCTCATCCTGAAGT). The PCR products were purified and sequenced as described previously.
Multiplex ligation-dependent probe amplification
For the detection of large deletions or duplications of whole exons, we also used the multiplex ligation-dependent probe amplification (MLPA) test (MRC-Holland, Amsterdam, the Netherlands) (14,15). The probe mix contains 8 coding exon probes for the LMNA gene (exons 1, 2, 3, 4, 6, 7, 8, and 10). In addition, we designed a set containing 3 adjacent non-coding exon probes (numbered −1, −2 and −3, respectively, where exon −1 is adjacent to the LMNA gene) at the 5’ region of the LMNA gene and 4 control probes specific for DNA sequences outside the LMNA gene.
Cell culture, immunoblotting, and indirect immunofluorescence
Human dermal fibroblasts from a patient and an unaffected relative were cultured at 37°C under 5% carbon dioxide in Dulbecco’s Modified Eagle’s Medium, supplemented with 10% fetal bovine serum.
Cells were counted and plated in 6-well plates, allowed to attach for 24 h, and lysed with the sample buffer. Protein samples were sheared through a 23G needle and boiled at 95°C for 10 min. Proteins were separated in 8% acrylamid gels and transferred onto polyvinylidene fluoride (PDVF) membranes. Primary antibodies against lamins A and C (Jol2, gift from Dr. C. J. Hutchison, Durham, United Kingdom) and GAPDH (RDI TRK5G4-6C5) were incubated overnight at 4°C in 3.5% Protifar plus (Nutricia, Zoetermeer, the Netherlands) in TRIS-buffered saline, 0.1% Tween. The secondary antibody, rabbit anti-mouse IgG HRP-linked, was visualized with enhanced chemoluminescence.
For immunofluorescence studies, cells were plated on coverslips, allowed to attach for 24 h, and fixed in 3.7% formaldehyde. Cells were permeabilized with 0.1% Triton and stained with primary antibody against lamin A (133a2) and counterstained with FITC-conjugated rabbit anti-mouse IgG (F0232, DAKO, Heverlee, Belgium). Immunofluorescence-labeled cells were counterstained with the DNA dye 4’-6-diamidino-2-phenylindole (DAPI). Nuclear aggregates of 100 separate nuclei were counted.
Clinical and histopathological evaluation
We evaluated 29 individuals from 3 generations (Fig. 1); generation II has been described previously (11). Twenty-five family members visited us for screening starting in 1995 (11 men, 14 women). We studied the clinical records and postmortem examinations of 4 additional persons (3 women) who had died (III-3, -7, -8, and -11) (Fig. 1). No signs of neuromuscular disease or lipodystrophy were noted. Creatine phosphokinase (CPK) values were within normal limits. Two patients (III-17 and -18) died during follow-up.
Of the 29 family members we studied, 18 were suspected to be affected either on clinical grounds or because of premature death (Fig. 1). Of these 18 subjects, 15 were histologically investigated: 13 underwent EMB (4 of these with additional postmortem or post-transplantation examinations). In 2 patients only postmortem investigations were available. Myocardial fibrosis was found in 14 of these 15 histologically investigated subjects. In 4 of these (III-12, III-15, IV-8, IV-9) (Table 1),distinct pathological fibrosis was identified as compared with the more extensive fibrosis in the remaining 10 patients. In patient IV-6, EMB showed fibrolipomatosis that could not be distinguished from arrhythmogenic right ventricular cardiomyopathy. Histological postmortem examinations of full-thickness myocardium in 5 patients (III-3, -7, -11, -17, and -18) all demonstrated extensive areas of interstitial and replacement fibrosis throughout the left and right ventricular myocardium and extensive fibrosis of the cardiac conduction system. The definite diagnosis was decided on the basis of a positive biopsy or postmortem histological proof of fibrosis.
More importantly, fibrosis (III-15, IV-2, -5, -8, and -9) or fibrolipomatosis (IV-6) was detected in 6 subjects who did not exhibit any abnormalities on echocardiography; in 4 of them (III-15, IV-2, IV-5, and IV-6) palpitations were the only complaint. In asymptomatic subjects, a first-degree atrioventricular block or supraventricular tachycardias raised the suspicion that they might be affected, which resulted in the treating physician ordering an EMB, which revealed fibrosis.
Thus, of all 14 of the affected family members with pathological interstitial myocardial fibrosis, 5 had cardiac fibrosis and 1 had fibrolipomatosis without any important loss of left ventricular function. The clinical data and survival curve for these patients are shown in Table 1and Figure 2.
Genome-wide linkage screening yielded the highest logarithm of the odds (LOD) score (LOD score = 2.6) for the region containing the LMNA gene. The haplotypes around this gene and individuals studied are shown in Table 2.Besides this region, no other regions showed a LOD score over 1.5. Although linkage was not completely conclusive, the gene was screened for point mutations with DGGE and direct-sequence analysis. These analyses did not reveal any sequence variation. Subsequent Southern blot analysis with 2 different double digests showed a deletion of the exon containing the start codon of the LMNA gene (Fig. 3).The PCR on cDNA revealed 2 PCR products of approximately 900 and 250 bp in length (Fig. 4).Subsequent sequence analysis showed that the longest cDNA fragment contained all exons, whereas the shorter band was lacking exons −1 and +1 of the LMNA gene (Fig. 5).This resulted in a deletion of 674 bp, which explains the 2 different-sized bands on the agarose gel. Subsequent MLPA analysis confirmed the deletion of exon 1 (Table 3),whereas the other coding exons had been retained. Because of technical difficulties in the design of the primers, a deletion of exon −1 could not be confirmed with this technique.
Deoxyribonucleic acid samples were available for 16 of the 18 clinically affected family members. The LMNA deletion was identified in all but 1 (IV-6) of these 16 persons.
Screening IDCM patients for LMNA deletion
Investigations in 150 additional patients with familial and non-familial DCM (with or without cardiac conduction disease and/or supraventricular arrhythmias) and in a control group of 150 unaffected persons have not revealed any further intragenic LMNA deletions.
Cell culture studies
To assess the effect of this novel deletion on the expression of lamins, we performed Western blot and indirect immunofluorescence assays. Western blot analysis of lamins A and C clearly showed decreased expression of both normal-sized lamins A and C in cells from a mutation carrier, when compared with the expression of the control cells (Fig. 6).Indirect immunofluorescence clearly showed an increased number of nuclear aggregates in cells from a lamin mutation carrier but hardly any aggregates in control cells (Fig. 7)(0.54 [SEM 0.10] in control cells vs. 6.05 [SEM 0.32] in cells from an LMNA mutation carrier [p < 0.0001]).
In a family with early onset myocardial fibrosis, we identified a deletion of the 5’ end of the LMNA gene including the start codon containing exon, located in the region of the highest LOD score. The LOD score at this locus did not exceed 2.6, because individual IV-6, referred for palpitations and initially classified as clinically affected, neither showed linkage to this region nor had the deletion. Her EMB showed fibrolipomatosis that could not be distinguished from arrhythmogenic right ventricular cardiomyopathy. We therefore consider the clinical picture of IV-6 a phenocopy unrelated to the inherited disorder in other family members.
Southern blot analysis indicated a deletion that was subsequently confirmed by PCR on cDNA and MLPA. Western blot analysis did show decreased expression of both lamins A and C in fibroblasts from a mutation carrier as compared with a non-carrier, although no alternative product from the mutated allele was seen (Fig. 6). An explanation for this might be that the truncated protein is less stable.
It is striking that fibrosis was massive and so easily encountered at biopsy. In 6 patients cardiac fibrosis was present before cardiac function was clearly affected as judged by echocardiography. Although it cannot be excluded that fibrosis is secondary to loss of a small number of myocytes as reactive fibrosis, this severe fibrosis occurs early in the disease and seems less likely to be secondary to overt failure. Severe myocardial fibrosis was also noted in the majority of the first families described by Fatkin et al. (9) as having DCM due to LMNA gene mutations, and also fibrosis of skeletal muscles has been noted before (16,17). The early occurrence of severe cardiac fibrosis in the absence of cardiac-function loss in the presented family practically excludes that cardiac fibrosis is caused by overt, severe heart failure but is initiated very early in the course of the disease. Our finding of nuclear aggregates of lamin in cultured fibroblasts from a mutation carrier substantiates the effect of this deletion. This raises the possibility that other LMNA gene mutations might also cause primary fibrosis, thus forming the underlying pathology in this cardiac phenotype.
Pathophysiologic effect of the LMNA deletion containing the start codon
The LMNA gene encodes lamins A and C by means of alternative splicing. These proteins are major components of the nuclear lamina, a fibrous network underlying the inner surface of the nuclear envelope (18). Mutations in the lamin A/C gene have been reported in a variety of disorders, such as autosomal dominant and recessive Emery-Dreifuss muscular dystrophy (19,20); limb-girdle muscular dystrophy type 1B (21); autosomal recessive Charcot-Marie-Tooth disease type 2B1 (22); Dunnigan-type familial partial lipodystrophy (23–25); mandibuloacral dysplasia (26); the Hutchinson-Gilford progeria syndrome (27,28); atypical Werner’s syndrome (29); and lipoatrophy with diabetes, hepatic steatosis, hypertrophic cardiomyopathy, and leukomelanodermic papules (30). Finally, LMNA gene mutations have been identified in patients having IDCM with conduction disease (11) or “pure” DCM (31,32). The exact pathophysiological mechanism of LMNA mutations that lead to such an impressive heterogeneous spectrum of disorders has not yet been elucidated. However, myocardial fibrosis as a secondary phenomenon has been described before in different laminopathies (10,15,33).
The identified deletion of the 5’ end of the LMNA gene containing the start codon is predicted to affect mainly the head domain of lamin (that is, it might lead to a shortened protein lacking the head domain). Assuming that the deletion does not interfere with the splicing of the other exons, new potential translation initiation sites should lie at cDNA position 559 and 598 (amino acids 187 and 200, respectively). When these proteins are produced, they lack the N-terminal region of lamin A and C, resulting in a smaller protein, which lacks the head domain. Previously, targeted mutation studies in Chinese hamster ovary cells have shown that functional disruption of the head domain has dominant negative effects, because these aberrant proteins can form aggregates of lamin in the nucleus (34). In line with this in vitro observation, we also found an increased number of lamin nuclear aggregates in cultured fibroblasts of a carrier of this deletion, as opposed to the virtual absence of these aggregates in a non-deletion carrier family member. This suggests that the mechanisms identified in vitro (34) do cause cardiac pathology in vivo possibly via a dominant negative effect. This raises the possibility that other LMNA gene mutations might also cause primary fibrosis, thus forming the underlying pathology of this cardiac phenotype.
Do lamin A/C mutations cause a specific cardiac phenotype?
The large pedigree made it possible to clinically evaluate patients at the earliest sign of being affected (that is, when demonstrating subtle cardiac (supraventricular) arrhythmias or cardiac conduction disease). The LMNA mutations can give rise to cardiac conduction disease (10). The histopathological substrate for the occurrence of conduction disease, however, remains unclear. We were, however, able to demonstrate that pathological myocardial fibrosis occurs at the first signs of conduction delay in a relatively large number of the family members affected. At postmortem examinations, extensive fibrosis of the conduction system was especially notable, because in 6 of the patients cardiac fibrosis was present before cardiac function was clearly affected as judged by echocardiography.
Another distinct finding was the small extent of dilatation for a mutant gene often described as a cause of DCM. Despite severe cardiac complications in laminopathies, this observation has been made before (35–39). Therefore these combined data suggest that LMNA mutations might cause a distinct cardiomyopathy, characterized by fibrosis and rather limited cardiac dilatation, which suggests a distinct disease entity within the spectrum of idiopathic “dilated” cardiomyopathy. This implies that distinct forms of cardiomyopathy can be encountered and that LMNA mutations might give rise to a type of cardiomyopathy in which primary fibrosis drives the specific pathophysiology on a highly malignant course. Myocardial fibrosis, therefore, might not occur as a separate disease entity but rather as a subphenotype of DCM.
The authors wish to thank J. M. C. Went for her support in collecting data and Marian Kraak for her assistance in MLPA.
This work was supported by grants to Drs. Pinto and van Berlo (2000.130, 2002T016) from the Netherlands Heart Foundation and is a part of research lines 27 and 50 of the Interuniversity Cardiology Institute of the Netherlands. The first two authors contributed equally to this article.
- Abbreviations and Acronyms
- dilated cardiomyopathy
- denaturing gradient gel electrophoresis
- deoxyribonucleic acid
- endomyocardial biopsy
- idiopathic dilated cardiomyopathy
- lamin AC
- multiplex ligation-dependent probe amplification
- polymerase chain reaction
- ribonucleic acid
- Received April 25, 2006.
- Revision received February 9, 2007.
- Accepted February 12, 2007.
- American College of Cardiology Foundation
- Michels V.V.,
- Driscoll D.J.,
- Miller F.A.,
- et al.
- Grunig E.,
- Tasman J.A.,
- Kucherer H.,
- Franz W.,
- Kubler W.,
- Katus H.A.
- Crispell K.A.,
- Wray A.,
- Ni H.,
- Nauman D.J.,
- Hershberger R.E.
- Gavazzi A.,
- Repetto A.,
- Scelsi L.,
- et al.
- Mestroni L.,
- Krajinovic M.,
- Severini G.M.,
- et al.
- Arbustini E.,
- Pilotto A.,
- Repetto A.,
- et al.
- Schouten J.P.,
- McElgunn C.J.,
- Waaijer R.,
- Zwijnenburg D.,
- Diepvens F.,
- Pals G.
- Brodsky G.L.,
- Muntoni F.,
- Miocic S.,
- Sinagra G.,
- Sewry C.,
- Mestroni L.
- Lin F.,
- Worman H.J.
- Muchir A.,
- Bonne G.,
- van der Kooi A.J.,
- et al.
- De Sandre-Giovannoli A.,
- Chaouch M.,
- Kozlov S.,
- et al.
- Cao H.,
- Hegele R.A.
- Speckman R.A.,
- Garg A.,
- Du F.,
- et al.
- De Sandre-Giovannoli A.,
- Bernard R.,
- Cau P.,
- et al.
- Caux F.,
- Dubosclard E.,
- Lascols O.,
- et al.
- Izumi M.,
- Vaughan O.A.,
- Hutchison C.J.,
- Gilbert D.M.
- Sanna T.,
- Dello Russo A.,
- Toniolo D.,
- et al.
- Karkkainen S.,
- Helio T.,
- Miettinen R.,
- et al.