MOLECULAR GENETICS OF DILATED CARDIOMYOPATHY IN THE DOBERMANN DOG

Transcription

1 MOLECULAR GENETICS OF DILATED CARDIOMYOPATHY IN THE DOBERMANN DOG Polona Stabej

2 Cover design en druk: OPTIMA Grafische Communicatie, Rotterdam Photo cover: Dobermann dog Rutty (photo Blaž Kondža) ISBN:

3 MOLECULAR GENETICS OF DILATED CARDIOMYOPATHY IN THE DOBERMANN DOG MOLECULAIRE GENETICA VAN DILATERENDE CARDIOMYOPATHIE BIJ DE DOBERMANN (met een samenvatting in het Nederlands) MOLEKULARNA GENETIKA DILATATIVNE KARDIOMIOPATIJE PRI DOBERMANIH (s povzetkom v slovenščini) PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de Rector Magnificus, Prof. dr. W.H. Gispen, ingevolge het besluit van het College voor Promoties in het openbaar te verdedigen op woensdag 6 april 2005 des middags te 2.30 uur door Polona Stabej geboren op 10 december 1972, te Kranj, Slovenië

4 Promotor: Prof. dr. Bernard A. van Oost 1 Co-promotores: Dr. Peter A.J. Leegwater 1 Dr. Arnold A. Stokhof 1 Dr. Aleksandra Domanjko-Petrič 2 1 Faculty of Veterinary Medicine, Universiteit Utrecht, The Netherlands 2 Faculty of Veterinary Medicine, University of Ljubljana, Slovenia

5 Publication of this thesis was made possible by the generous support of: Universiteit Utrecht

6 CONTENTS CHAPTER 1 CHAPTER 2 CHAPTER 3 CHAPTER 4 CHAPTER 5 CHAPTER 6 CHAPTER 7 General Introduction: Molecular genetics of cardiomyopathies in dogs 9 Dilated cardiomyopathy in the Dobermann dog: survival, causes of death and a pedigree review in a related line 43 Characterization of the desmin (DES) gene and evaluation as a candidate gene for dilated cardiomyopathy in the Dobermann dog 63 Duplication of a polymorphic CA-repeat by retrotransposition in the canine genome: implications for the analysis of the association of the α-tropomyosin gene (TPM1) with dilated cardiomyopathy in the Dobermann dog 85 The canine sarcoglycan delta gene: BAC clone contig assembly, chromosome assignment and interrogation as a candidate gene for dilated cardiomyopathy in Dobermann dogs 103 Evaluation of the phospholamban gene in purebred large-breed dogs with dilated cardiomyopathy 121 Genetic epidemiological studies of DCM in the Dobermann dog point to a crucial role of titin in DCM susceptibility 137 SUMMARY / SAMENVATTING / POVZETEK 157 ACKNOWLEDGEMENTS 165 LIST OF PUBLICATIONS 171 CURRICULUM VITAE 175

7

8

9 Chapter 1 General Introduction: Molecular Genetics of Cardiomyopathies in Dogs Polona Stabej, Kathryn Meurs, Bernard A. van Oost Adapted from Molecular Genetics of Cardiomyopathies in Dogs ; invited chapter for The Dog and Its Genome (eds: E. A. Ostrander, U. Giger, and K. Lindblad-Toh); Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, USA

10 Chapter 1 1. CARDIOMYOPATHIES IN MAN AND DOG Canine familial cardiomyopathies Genotypes and phenotypes of familial DCM in human Pathophysiology of dilated cardiomyopathy DEVELOPMENT OF TOOLS IN CANINE GENETICS AND THE CHARACTERISTICS OF THE DOG GENOME The canine genetic maps The canine BAC library The dog genome sequence The dog genome characteristics DISCUSSION Dog as a model of human DCM Advances in human genetics shed light on patho-genetic backgrounds of DCM The complexity of DCM and the great potential of the dog genome 30 OBJECTIVES OF THIS THESIS 33 REFERENCES 34 10

11 Cardiomyopathies in man and dog 1. CARDIOMYOPATHIES IN MAN AND DOG Cardiomyopathies are defined as diseases of the myocardium associated with cardiac dysfunction. According to the WHO/ISFC criteria, cardiomyopathies in the human are classified by the dominant pathophysiology as dilated cardiomyopathy, hypertrophic cardiomyopathy, restrictive cardiomyopathy and arrhythmogenic right ventricular cardiomyopathy (Richardson et al. 1996). In dogs, dilated cardiomyopathy (DCM) and arrhythmogenic right ventricular cardiomyopathy (ARVC) are the most common inherited forms of myocardial disease CANINE FAMILIAL CARDIOMYOPATHIES FAMILIAL DILATED CARDIOMYOPATHY Dilated cardiomyopathy (DCM) is a primary myocardial disease characterized by cardiac enlargement and impaired contractile function of one or both ventricles. Although canine dilated cardiomyopathy is traditionally discussed as one disease, significant variation in the presenting signs, clinical evaluation and rate of progression have been observed among and within breeds (Freeman et al. 1996; Tidholm et al. 1996; Calvert et al. 1997; Kittleson et al. 1997; Brownlie et al. 1999; Meurs et al. 2001a; Sleeper et al. 2002). It is likely that as further evaluation of DCM in individual breeds is performed, diverse genetic etiologies will be identified. Dilated cardiomyopathy is generally a disease of large and medium sized dog breeds. Some breeds are clearly over represented and this seems to vary with geographical location. In North America the Dobermann, Irish Wolfhound, Great Dane and Cocker Spaniels are among the most commonly reported (Monnet et al. 1995; Sisson et al. 2000). Some European sources suggest an increased incidence in the Airedale Terrier, Dobermann, Newfoundland and English Cocker Spaniel (Tidholm et al. 1997). This difference may be related to the strong genetic influences of certain popular dogs within an area, a phenomenon termed the founder effect. Dilated cardiomyopathy is an adult onset disease, with the exception of the Portuguese Water Dog (Sleeper et al. 2002). The disease is predominantly a left ventricular dysfunction but biventricular involvement and heart failure may be noted. Arrhythmias, such as ventricular premature complexes, tachycardia and atrial fibrillation, are common. 11

12 Chapter 1 BREED VARIATIONS Cocker Spaniels Dilated cardiomyopathy has been reported in both American and English Cocker Spaniels. An association between the development of DCM and low plasma taurine levels has been reported in some American Cocker Spaniels with DCM (Kittleson et al. 1997). Affected dogs were between 5 and 13 years of age. Clinical findings included a dilatation of the left ventricle and a reduction in left ventricular contractility. It is not known if this is a familial problem in the American Cocker, although it should be noted that the strong breed specificity of decreased taurine levels may suggest a familial problem related to taurine metabolism or absorption. English Cocker Spaniels may also get a form of DCM but a relationship to taurine has not been well studied. Many reported dogs were from the same kennel, which suggests a heritable component (Gooding et al. 1986; Wotton et al. 1998). Profound evidence of left ventricular enlargement was frequently observed. Some of the reported dogs died suddenly, but many have had a prolonged, fairly asymptomatic course of disease, or a long survival (years) [Gooding et al. 1986; Wotton et al. 1998; Bonagura et al. 2000]. Dalmatians All dogs had adult onset disease (average of 6.8 years) and presented for signs consistent with left heart failure or collapse. Ventricular tachyarrhythmias were common. Male dogs appear to be over-represented in Dalmatian DCM, which may suggest an X-linked trait although large studies have not been performed (Freeman et al. 1996). Dobermanns The Dobermann is one of the most commonly reported breeds of dogs to be affected with DCM in North America and Europe (Monnet et al. 1995; Calvert et al. 1997; Calvert et al. 2000; Domanjko-Petrič et al. 2002). It is characterized by an adult (average of 7 years) onset of disease that results in the development of left and/or biventricular failure, often with atrial and/or ventricular tachyarrhythmias. The clinical stage of DCM in the Dobermann appears to be very malignant in comparison to the disease in other breeds. The median survival time for dogs once heart failure has developed is 9.6 weeks. Some Dobermanns develop collapse or die suddenly before left ventricular dilatation or contractile dysfunction ever develops. Dilated cardiomyopathy in this breed is believed to be familial, likely with an autosomal dominant mode, although this is not well documented at this point (Hammer et al. 1996). Some groups reported male predominance (Calvert et al. 12

13 Cardiomyopathies in man and dog 1982; Domanjko-Petrič et al. 2002) amongst DCM Dobermanns, yet others found no gender predisposition (O Grady and Horne 1995). Great Danes Affected Great Danes are presented at an average age of 5 years, most commonly for weight loss and/or coughing. Left and right heart failure are frequently observed. Dilated cardiomyopathy in the Great Dane appears to be a familial disease (Meurs et al. 2001b). Affected male dogs were over-represented suggesting an X-linked pattern of inheritance in at least some families. Irish Wolfhounds Atrial tachyarrhythmias frequently precede the development of clinical signs and heart failure in the Irish wolfhound with DCM. A heritable nature has been suggested (Brownlie et al. 1999). Newfoundlands Adult onset familial DCM without a gender predisposition has been reported in the Newfoundland (Tidholm et al. 1996; Lee et al. 2002). Clinical presentation may include dyspnea, cough, inappetance and ascites with left or biventricular heart failure. Portuguese Water Dogs A juvenile form of familial DCM has been reported in the Portuguese Water Dog (Dambach et al. 1999; Sleeper et al. 2002). Affected puppies died from congestive heart failure (CHF) at an average age of 13 weeks. The course was very rapid with the longest clinical course of 5 days. Standard emergency medical care to support the symptomatic puppies was unsuccessful. Affected puppies from either sex were from apparently unaffected parents, suggestive for an autosomal recessive inheritance. L-carnitine Boxer cardiomyopathy A deficiency in myocardial L-carnitine levels was observed in a family of boxers with DCM. These dogs had left ventricular dilatation and contractile dysfunction with arrhythmias that included both atrial and ventricular tachycardias. The dogs had normal to high levels of plasma L-carnitine, but low levels of myocardial carnitine (Keene et al. 1991). ARRHYTHMOGENIC RIGHT VENTRICULAR CARDIOMYOPATHY IN THE BOXER Since the early 1980 s, the term boxer cardiomyopathy has been used to describe adult Boxer dogs that present with ventricular arrhythmias, and sometimes, collapse (Basso et al. 2004). A small percentage of these dogs were noted to have left ventricular dilatation with left or biventricular heart failure and the disease was classified as a form of dilated cardiomyopathy. However, it would appear that the majority of boxers affected with this disease suffer primarily from 13

14 Chapter 1 conduction abnormalities as a result of right ventricular myocardial disease. In general, many affected dogs have fairly normal hearts at gross inspection but with significant histological changes. Recent studies have demonstrated that the disease has many similarities to Arrhythmogenic Right Ventricular Cardiomyopathy in humans. The similarities between the diseases include clinical presentation, etiology and histopathology among others. Therefore, more recently, the disease in the Boxer has become referred to as Arrhythmogenic Right Ventricular Cardiomyopathy (ARVC) [Basso et al. 2004]. Arrhythmogenic right ventricular cardiomyopathy is a familial disease in the boxer and appears to be inherited as an autosomal dominant trait (Kraus et al. 2002). Unfortunately, the disease also appears to be a disease of variable genetic penetrance and affected dogs may present in different ways including asymptomatic, syncope, sudden death and systolic dysfunction with CHF GENOTYPES AND PHENOTYPES OF FAMILIAL DCM IN HUMANS Dilated cardiomyopathy is a leading cause of cardiovascular morbidity and mortality in human medicine. No single clinical parameter has been found to distinguish familial from nonfamilial DCM. Based on family history, the cause is predicted to be genetic in at least 35% of cases (Grunig et al. 1998). Familial DCMs are heterogeneous disorders with autosomal dominant, autosomal recessive, X-linked or mitochondrial modes of transmission reported (Fatkin and Graham 2002). Over the past decade, identification of several DCM causing genes in human beings clarified some of the underlying pathophysiological mechanisms that will be discussed. Moreover, discoveries of the molecular basis of familial cardiomyopathies and their accompanying disorders can provide valuable information about the functioning of myocytes. AUTOSOMAL DOMINANT DCM The most frequent form of DCM in human beings is autosomal dominant and exhibits both clinical variability and genetic heterogeneity. The phenotypes of patients vary from mild, late onset DCM to severe DCM beginning in infancy. Premature death may result from severe heart failure or ventricular arrhythmias. Sudden death can occur at any age, irrespective of ventricular function (Fatkin and Graham 2002). Based on clinical data, two forms can be distinguished: pure DCM and DCM with conduction defects. Sixteen genes have been causally 14

15 Cardiomyopathies in man and dog implicated in autosomal dominant DCM to date. In this chapter, phenotypic characteristics resulting from specific DCM mutations are described. Patients with isolated DCM phenotype typically present with chamber dilatation and decreased contractility with progression to heart failure. The age of onset as well as the severity of the disease is highly variable. Mutated genes identified in relation with pure DCM are listed in Table 1. In some cases, different mutations in the same gene caused either pure DCM or DCM with abnormal conduction. Such examples are related to mutations in the titin and α-tropomyosin genes (Olson et al. 2001; Itoh-Satoh et al. 2002). An interesting clinical feature of patients carrying the mutation Ala743Val in the titin protein was cardiac arrhythmia (premature atrial or ventricular contraction and atrio-ventricular conduction block), whereas Gln4053ter in the cardiac specific N2B isoform of titin manifested solely with severe heart failure and cardiac dilatation (Itoh-Satoh et al. 2002). Also cardiac α-tropomyosin (TPM1) mutations cause either severe DCM with ventricular tachycardia or congestive failure without conduction system disorders (Olson et al. 2001). Several mutations within the lamin A/C gene (LMNA) were described, causing phenotypic heterogeneity in different families. The phenotypic commonality of the majority of the LMNA mutations causing DCM are defect impulse formations such as sinus bradycardia, atrial arrhythmias or atrioventricular conduction block (Fatkin et al. 1999; Arbustini et al. 2002; Herschberger et al. 2002; Sebillon et al. 2003; Taylor et al. 2003; Karkkainen et al. 2004). Sudden death due to fatal ventricular arrhythmias is often observed and frequently occurs at a young age, even before patients develop overt heart failure (Becane et al. 2000; Karkkainen et al. 2004). Also troponin T (TNNT2) mutations leading to DCM can have very different manifestations. Three mutations have been described to date: Lys210, R141W and A171S (Kamisago et al. 2000; Li et al. 2001; Stefanelli et al. 2004). The age of onset and the severity of the disease in patients with TNNT2 mutations were highly variable. Sinus bradycardia, sinus arrhythmia, incomplete right bundle branch block or prolonged QT interval were the electrocardiographic abnormalities of patients in the family with the R141W mutation (Li et al. 2001). The peculiarity of individuals carrying the A171S mutation was the disparity in the disease severity between males and females. Affected males tended to manifest greater left ventricular dilatation and systolic dysfunction. Ventricular arrhythmias (ventricular tachycardia) and the only two sudden cardiac deaths occurred in males (Stefanelli et al. 2004). DCM has also been associated with mutations in genes encoding ion channel proteins. Patients with mutations in the ABCC9 (encodes the regulatory SUR2A subunit of K ATP channel) had severely dilated hearts with compromised contractile 15

16 Chapter 1 function and ventricular tachycardia (Bienengraeber et al. 2004). The second ion channel gene causing DCM is a cardiac Na + channel gene SCN5A. Patients carrying SCN5A mutations were reported to have a very distinctive phenotype of DCM with abnormal cardiac conduction. The disease typically begins in adolescence with the onset of sinus node dysfunction manifested as asymptomatic sinus bradycardia, symptomatic sinus pauses, or arrest. Junctional bradycardia, supraventricular tachyarrhythmias and atrial flutter or fibrillation are also observed. Once the disease progresses, atrio-ventricular (AV) and bundle branch block, demonstrating AV node and His-Purkinje conduction system disorders become apparent. Ultimately, the disease progresses to dilatation of atria and ventricles with left ventricular systolic dysfunction and clinical heart failure (Olson and Keating 1996; McNair et al. 2004). AUTOSOMAL RECESSIVE DCM Recently, a mutation in the cardiac troponin I gene (TNNI3) was identified as the cause of DCM which was inherited in an autosomal recessive pattern. The consequence of the mutation was an amino acid substitution A2V. The proband and his sister were homozygous for the A2V mutation and presented with a progressive cardiac failure at 27 and 29 years of age (Murphy et al. 2004). X-LINKED DCM Two genes have so far been associated with X-linked forms of DCM: tafazzin (TAZ) and dystrophin (DYS). Mutations of TAZ cause severe infantile DCM, a disorder allelic to the Barth syndrome (D Adamo et al. 1997). The tafazzin gene appeared to code for an acyltransferase, which affects the cardiolipin levels in mitochondria. Cardiolipin is essential for the stability of the mitochondrial respiratory chain and mitochondrial dysfunction lies most likely at the basis of the cardiomyopathy symptoms in Barth Syndrome (Barth et al. 2004). Two more genes on the X-chromosome have been associated with DCM - not as a solely disease, but as an accompanying symptom of skeletal myopathy. These are the emerin gene responsible for Emery-Dreifuss muscular dystrophy (Bione et al. 1994) and the lysosome-associated membrane protein-2 gene involved in Danon s disease (Nishino et al. 2000). X-linked DCM caused by DYS mutations is allelic to Duchenne and Becker muscular dystrophies (DMD and BMD), with which the cardiac phenotype is usually associated. DCM patients lack the skeletal muscle symptoms and only display the cardiac phenotype (Nigro et al. 1994; Muntoni et al. 1997). A number of different DYS mutations found in DCM patients suggest multiple pathogenic 16

17 Cardiomyopathies in man and dog mechanisms resulting in different dystrophinopathic phenotypes. In the more common severe form, affected young men in their early twenties typically develop congestive heart failure that results in death or cardiac transplantation within 1 to 2 years. Female carriers may also develop mild symptoms of DCM later in life (Towbin et al. 1993). MATRILINEAL INHERITANCE As illustrated by Barth syndrome mentioned above, mitochondrial dysfunction is an important determinant for DCM. The best understood causes of mitochondrial deficiencies are mutations in the maternally transmitted mitochondrial DNA. Deletions and mutations in various mitochondrial transfer RNAs have been observed in association with DCM (Suomalainen et al. 1992; Marin-Garcia et al. 2001; Maternally inherited DCM has not yet been reported in the dog. 17

18 Chapter 1 Table 1. DCM in man: genes and phenotypes Autosomal dominant DCM DCM phenotype DCM SD Type of conduction system disorder Reference Gene symbol Gene product DCM+CD LMNA lamin A no yes yes AF/AVB/LBBB/SB Karkkainen et al VT/SVT/AF/PVC/PAC Brodsky et al AVB/SB/AF Fatkin et al AF/PVC/SSS/AVB/LBBB Hershberger et al AVB/PVC/PAC/LBBB/RBBB Arbustinin et al AF/LBB/AVB/VT/PVC Sebillon et al SGCD sarcoglycan delta yes no yes Tsubata et al DES desmin yes no no Li et al ACTC alpha cardiac actin yes no no Olson et al MYH7 beta myosin heavy chain yes no yes Kamisago et al. 2000; Daehmlow et al MLP muscle LIM protein yes no no Knoll et al PLN phospholamban yes no no Schmitt et al Cypher yes no no Arimura et al MYBPC3 myosin binding protein C yes no no Daehmlow et al no yes Kamisago et al TNNT2 cardiac troponin T yes yes yes VT Stefanelli et al yes no SB/RBBB/SA/ST Li et al TPM1 alpha tropomyosin yes yes no VT Olson et al VCL vinculin/meta vinculin yes no Olson et al TCAP T-cap protein yes no no Knoll et al yes no PAC/PVC/AVB Itoh-Satoh et al TTN titin yes no no Gerull et al. 2002; Siu et al ABCC9 K ATP channel subunit no yes no VT Bienengraeber et al SCN5A cardiac Na channel subunit no yes no AF/SSS/PVC/VT/SND/SB/AVB/RBBB/LBBB McNair et al. 2004; Olson and Keating 1996 Autosomal recessive DCM TNNI3 troponin I yes no no Murphy et al X-linked DCM TAZ tafazzin D Adamo et al yes no no DYS dystrophin Cohen and Muntoni 2004 (Review) DCM - pure DCM; DCM+CD - DCM with conduction disorders; SD - sudden death; AF - atrial fubrillation; AVB - atrio ventricular block; R(L)BBB - right (left) ventricular branch bundle block; SB - sinus bradycardia; PVC - premature ventricular contractions; PAC - premature atrial contractions; SSS - sick sinus sindrom; VT - ventricular tachycardia, ST - sinus tachycardia; SA - sinus arrhythmia; SND - sinus node dysfunction 18

19 Cardiomyopathies in man and dog 1.3. PATHOPHYSIOLOGY OF DILATED CARDIOMYOPATHY CARDIOMYOCYTE STRUCTURE PROTEINS OF THE SARCOMERE CONTRACTILE UNIT OF THE HEART The interior of the cardiac myocyte contains bundles of longitudinally arranged myofibrils that have a characteristic striated appearance formed by repeating sarcomeres. The sarcomere is the structural and functional unit of cardiac muscle comprised of thin (α-cardiac actin, α-tropomyosin, troponins C, I, T) and thick filaments (myosin and myosin binding proteins C, H, X). Each sarcomere has an I-band (comprised of thin filaments only), an A-band (comprised of overlapping thin and thick filaments) and a M-band (comprised of thick filaments only). The scaffolding for the thick and thin filaments is provided by the giant protein titin and myomesins (titin associated proteins) [Fatkin and Graham 2002]. Z-discs demarcate the sarcomeres and crosslink the myofilaments into a highly ordered 3-dimensional lattice. Essential components of Z-discs are barbed ends of actin capped by CapZ and cross linked by α-actinin, titin capped by T-cap protein, nebulette (binds actin and α-actinin), obscurin, cypher/zasp, muscle LIM protein, myopalladin, myopodin, cardiac restricted ankyrin repeated protein and the Ca 2+ binding protein S-100. Z-discs are linked with one another, to the sarcolemma and to the nuclear envelope by the desmin network of the cytoskeleton. The major connection of the Z-disc to the rest of the sarcomere occurs through titin. Titin s N-terminal domains insert into the Z-disc where the titin molecules from opposing sarcomeres overlap and titin s C-terminus attaches to the M-line of the sarcomere and makes contact with the head-neck interface of crossbridges through its interaction with the myosin-binding protein C (MyBP-C) [Pyle and Solaro 2004]. These Z-disc proteins function as physical anchor for myofilaments and the cytoskeleton and have a major role in reception, transduction and transmission of mechanical and biochemical signals. EXTRASARCOMERIC PROTEINS OF THE CYTOSKELETON: THE DYSTROPHIN GLYCOPROTEIN COMPLEX (DGC) The DGC is a multicomponent complex that provides a mechanical link between the intracellular and extracellular matrix. It has structural and signal transduction properties and consists of dystrophin, syntrophins, dystroglycans, sarcoglycans, caveolin-3, NO-synthase and sarcospan (Lapidos et al. 2004). Dystrophin links the F-actin and β-dystroglycan that connects to α-dystroglycan. The latter has an extracellular region attached to laminin 2 (a component of the 19

20 Chapter 1 extracellular matrix), which in turn binds to collagen IV. The sarcoglycans are connected to α-dystroglycan through α-sarcoglycan. This is how a continuous connection is formed between the cytoskeletal actin network and the extracellular matrix (Rybakova et al. 2000). Mutations in genes coding for DGC components have been identified to cause DCM in experimental animals and humans. DCM GENES AND MOLECULAR MECHANISMS OF INHERITED DILATED CARDIOMYOPATHIES Over the past few years, several molecular pathways involved in DCM have been elucidated by genetic studies. These findings have provided insight into how mutations of genes encoding for proteins of the cardiomyocytes can lead to DCM. While not all is crystal clear along the DCM pathways, several mechanisms have already been deciphered. In essence, we can distinguish three main pathways that will be discussed. DISTURBED INTEGRITY OF THE CYTOSKELETON Some dystrophin gene (DYS) mutations are mainly or exclusively associated with the heart, yet others affect predominantly the skeletal muscle. This suggests complex pathophysiological mechanisms (Cohen and Muntoni 2004). Presently, there are 16 DYS mutations causing DCM. The proposed classification based on the likely disease mechanism divides mutations into two groups. In group A are mutations affecting transcription or splicing and usually result in the absence of the dystrophin protein in the heart. In group B are mutations which affect specific domains of the dystrophin protein. At the molecular level, the loss or structural damage of dystrophin results in destabilization of the rest of the DGC complex and an impaired mechanical link between the sarcolemma and the extracellular matrix (Cohen and Muntoni 2004). A missense mutation (Ser151Ala) and a 3 bp deletion ( Lys238) found in the sarcoglycan delta gene of DCM patients were also predicted to affect the secondary structure of the protein and to disturb the integrity of the cytoskeleton (Tsubata et al. 2000). Actin is a sarcomeric thin filament, a major component of the cell cytoskeleton that participates in more protein-protein interactions than any other known protein (Dominguez 2004). One end forms cross-bridges with the myosin and the other end is anchored in the Z-disc. The two mutations identified in DCM patients (Arg312His and Glu361Gly) occur in the immobilized domain of the actin monomer that is involved in crosslinking to the anchor proteins in the Z- 20

21 Cardiomyopathies in man and dog band. Hence, DCM mutations were suggested to impair the cytoskeletal connection between actin and Z-disc proteins (Olson et al. 1998). Two DCM causing mutations in the β-myosin heavy chain gene (MYH7), i.e. Ser642Leu and Ser532Pro, are located in the part of the myosin heavy chain that interacts with the actin thin filament. These two mutations are likely to alter the stereo specific interaction between actin and myosin, which is essential for the power stroke of contraction (Kamisago et al. 2000; Daehmlow et al. 2002). The third MYH7 mutation (Phe764Leu) resides in the converter region that transmits the movement from the head of the myosin to the neck. The Phe764Leu mutations could therefore hinder the propelling of the thick filament (Kamisago et al. 2000). Desmin is the constituent of the major muscle specific intermediate filament that forms a three-dimensional scaffold around the Z-disc, makes longitudinal connections between the consecutive Z-discs and links the contractile apparatus to the subsarcolemmal cytoskeleton, the nuclei and other organelles (Paulin and Li 2004). The mutation Ile451Met in DES causing DCM in a human family induces a dramatic change in the secondary structure of the desmin carboxyterminal domain (Li et al. 1999a). In another family, the same Ile451Met mutation was found to cause isolated skeletal myopathy with no signs of cardiomyopathy (Dalakas et al. 2003). The pathogenetic mechanisms of these considerably different phenotypes associated with the same mutation remain obscure. Lamin A and C proteins, encoded by the lamin A/C gene (LMNA), are intermediate filaments forming a two-dimensional matrix at the nuclear lamina. Lamins of the A-type have essential roles in the maintenance of the lamina stability and regulation of the transcription factors required for the differentiation of adult stem cells (Hutchinson and Worman 2004). Mutations in the LMNA have been shown to cause a host of diseases, from cardiomyopathies and muscular dystrophies to lipid disorders and progeria (Burke and Stewart 2002). It is intriguing how the mutant LMNA can give rise to such diverse, tissue-restricted illnesses. Two working hypotheses have been proposed to explain laminopathies: 1. The gene expression hypothesis proposes that mutations of LMNA alter the interaction of the lamin A/C protein with various gene regulatory proteins and thereby causes disease in various tissues (Hutchinson and Worman 2004) 2. The structural hypothesis proposes that mutations in LMNA give rise to a weakened nuclear envelope. Damage to the nuclear envelope is thought to promote myocyte death and replacement of the highly differentiated cells by fatty and fibrotic tissue (Hutchinson and Worman 2004). 21

22 Chapter 1 SERCA2a Phospholamban Titin Myosin Sarcoplasmic reticulum Lamin nucleus Sarcoglycans Sarcolemma β α δ Desmin Dystrophin -tropomyosin Troponin-I Troponin-T MLP T-cap Cypher Actin Troponin-C Z-disc Figure 1. Structural network in a cardiomyocyte (full color figure on page 179). β Dystroglycans KATP channel Na channel 22

23 Cardiomyopathies in man and dog There is also increasing evidence of the existence of karyoskeletal networks that may form connections between proteins in the nucleus (A-type lamins, emerin) and cytoskeletal proteins via desmin (Ostlund and Worman 2003). If the lamin-dystrophyn-glycoprotein complex (DGC) connection indeed exists, then the LMNA mutations could cause DCM via disruption of the cytoskeletal support network. DISTURBED CALCIUM KINETICS AND SENSITIVITY Mutations of the cardiac troponin T gene (TNNT2) can cause dilated, hypertrophic and restrictive cardiomyopathy (Gomes and Potter 2004). Studies examining the effect of TNNT2 mutations Lys210 and R141W found diminished calcium sensitivity of force generation in both cases (Kamisago et al. 2000; Li et al. 2001; Lu et al. 2003). In contrast, nearly all TNNT2 mutations found in hypertrophic cardiomyopathy cause increased sensitivity to Ca 2+ that leads to abnormally enhanced heart contractility (Harada et al. 2004). The physiology of gender specific phenotypic differences observed in the patients with A171S mutation, with males being more severely affected than the females, remains to be elucidated. Gender specific differences in sarcoplasmic reticulum Ca 2+ loading, particularly in response to adrenergic stimulation, have been suggested as a possible cause (Stefanelli et al. 2004). IMPAIRED INTRACELLULAR SIGNALLING MECHANISMS The finding of mutations in MLP (muscle LIM protein) and TCAP (telethonin or titin-cap) in DCM and a parallel study of MLP deficient mice demonstrated the critical role of MLP as a component of the cardiac stretch sensor in a Z-disc-theletonin-titin complex. The research of Knoll et al. (2002) on MLP and TCAP mutations provides functional evidence that the chamber dilatation in DCM is related to an inability of the cardiomyocytes to sense the mechanical stretch stimulus and to generate a primary effect on muscle tension. This molecular model suggests that the elastic segments within the titin I-band domains serve as the intrinsic cardiac mechanical stress sensor that requires the binding of MLP and TCAP. Titin s elastic I-band structures serve as molecular springs generating tension in response to mechanical stretch following systole and determine ventricular wall distensibility. In a normal heart, the titin/z-disc complex is stretched, mechanical load is registered and activates downstream signals for cardiomyocyte hypertrophy and survival. The mutated MLP causes impaired anchoring of the Z-disc to the T-cap/titin complex and leads to a loss of elasticity of the titin elastic spring. The primary result of the latter is defective 23

24 Chapter 1 cardiomyocyte stretch sensing that leads to over-stretching of the myocytes and activation of the cell death pathways (Knoll et al. 2002). The giant molecule titin is after myosin and actin the third most abundant of the cardiac proteins. It belongs to the proteins of the sarcomeric cytoskeleton and is the largest natural protein currently known. One molecule spans half a sarcomere from the Z-disc to the M-line (Labeit et al. 1997). The titin protein is dynamic in structure and function. It is responsible for the structural integrity of the sarcomere by acting as a scaffold, gives raise to passive muscle stiffness, influences active force development, is essential for sarcomerogenesis and titin based protein complexes have a role in signalling as biomechanical sensors (Gregorio et al. 1999; reviewed in Granzier and Labeit 2004). As titin is such a large, multifunctional protein, it is a prominent target for mutations that give rise to muscle diseases. Seven mutations in the titin gene (TTN) have been implicated in DCM. The Ala743Val and Val154Met missense mutations were found in the binding domain of titin to α-actinin and to T-cap/telethonin, respectively. It was suggested that the two mutations might affect the Z-disc-theletonin-titin component of the cardiac stretch sensor (Itoh-Satoh et al. 2002; Knoll et al. 2002). The third missense mutation (Trp930Arg) was reported by Gerull et al. (2002) and is predicted to disrupt a highly conserved hydrophobic sequence of an immunoglobulin fold located in the Z-disc-I-band transition (Gerull et al. 2002). The nonsense mutation (Glu4053ter) was found in the cardiac specific N2-B region and the mutant allele encoded for a truncated non-functional molecule (Itoh-Satoh et al. 2002). The 2-bp insertion mutation in exon 326 of the titin caused a frame shift leading to truncation of the protein in the A-band region. The truncated titin is predicted to lack the titin kinase domain as well as binding sites for thick filaments, myosin binding protein C, myomesin and calpain. It is not clear how the truncation results in cardiac remodelling (Gerull et al. 2002). Phospholamban (PLN) is a small transmembrane phosphoprotein of 52 amino acids that plays an important role in cardiac contraction and relaxation. Cardiac contraction occurs with elevation of Ca 2+ concentration in the cytoplasm that is mediated by Ca 2+ release channels (ryanodine receptors; RyRs) and plasma-membrane Ca 2+ channels. Cardiac relaxation is triggered by Ca 2+ uptake into the sarcoplasmic reticulum (SR) through the Ca 2+ -ATP (SERCA2a) pump, plasma membrane Ca 2+ -ATPases (PMCAs) and Na + /Ca 2+ exchangers (NCX) that refill the SR and extracellular Ca 2+ stores. Phospholamban in its dephosphorylated state binds to and inhibits the SERCA2a pump activity, whereas the phosphorylated PLN reverses the Ca 2+ pump inhibition and enhances relaxation rates and contractility (MacLennan and Kranias 2003). Phospholamban is phosphorylated by protein kinase A (PKA), which is activated by elevated camp. The latter is formed by adenylate cyclase that is stimulated by a signal coming 24

25 Cardiomyopathies in man and dog from the β-adrenergic receptors in the cell membrane. The signal-transduction pathway is activated upon binding of adrenalin and other β-agonists binding to the receptors. The R9C mutation in PLN found in a family with DCM showed a markedly reduced phosphorylation of the PLN. The PLNR9C shows enhanced affinity for PKA, which becomes trapped in a mutant PLN-PKA complex and cannot dissociate and phosphorylate wild type PLN molecules. DCM patients with a PLN mutation therefore have a chronically inhibited SERCA2a pump, which leads to DCM already in their teenage years (MacLennan and Kranias 2003; Schmitt et al. 2003). Cypher/ZASP is a gene, which expresses in striated muscles a Z-disc associated cytoskeletal protein. A missense mutation located in the third LIM domain of Cypher causes an increased binding affinity of mutated Cypher for protein kinase C (PKC). Consistent with the findings of mutations in MLP and PLN, Cypher/ZASP mutations cause abnormal recruitment of molecules participating in intracellular signalling (Arimura et al. 2004). Mutations in two ion channel genes: SCN5A and ABCC9 (Bienengraeber et al. 2004; McNair et al. 2004) have been found in DCM. Additionally, the ryanodine receptor (ion channel) mutations have been found in arrhythmogenic right ventricular dysplasia, a phenotype related to DCM (Tiso et al. 2001). The phenotypic commonalities of mutations in the ion channel genes are the conduction disturbances associated with DCM. Cardiac K ATP channels are heteromultimers composed of a potassium channel pore protein (Kir6.2) and an ATPase-harboring ATP-binding protein (SUR2A subunit, encoded by ABCC9). Two heterozygous mutations were found in exon 38 of ABCC9 that encodes the C-terminal domain of SUR2A (Bienengraeber et al. 2000; Bienengraeber et al. 2004). The SUR2A subunit has a metabolic decoding capacity expressed by its ability to recognize and process intracellular energetic signals. DCM-causing mutant K ATP channel complexes formed functional channels with intact pore properties, but distorted the ATP-dependent pore regulation. Both DCM-causing ABCC9 mutations induced alterations in hydrolysis-driven SUR2A conformational probability that translated into abnormal ATP sensitivity of mutant channels. In this way, changes of the catalytic module of the channel complex confer susceptibility to DCM (Bienengraeber et al. 2004). SCN5A encodes the pore-forming α-subunit of the cardiac sodium channel, which is a transmembrane protein that produces the ion current responsible for the raising phase of the action potential (Viswanathan and Balser 2004). This channel has a key role in cardiac depolarisation and conduction of the impulse through the cardiac tissue (Napolitano et al. 2003). Mutations in SCN5A are known to cause several cardiac rhythm disorders. In the long-qt syndrome, the SCN5A mutation 25

26 Chapter 1 induces a gain of function of the protein with enhanced inward current, whereas the group of SCN5A mutations causing the isolated conduction disease and the Brugada syndrome (ventricular fibrillation in the absence of QT interval prolongation or structural heart disease) evoke a loss of function of the sodium channels (Viswanathan and Balser 2004). The heterozygous SCN5A mutation (G3823A) that segregated with a DCM and conduction disorder phenotype causes a D1275N substitution. This amino acid residue is highly conserved among voltage-gated sodium channels as well as calcium and potassium channels (McNair et al. 2004). The identification of the precise molecular mechanisms that lead from mutations of the ion channels to DCM is not yet clear. In conclusion, the three pathways described certainly interlink, as altered connections between the proteins of the cytoskeleton can result in impaired signalling mechanisms (mutations in MLP) or cause disturbed calcium kinetics (mutations in TNNT2). Another example is mutated phospholamban (PLNR9C) which causes disturbed Ca 2+ kinetics by hampering a signaling molecule protein kinase (PKA). Considering the incomplete knowledge of the physiology of the heart, the precise description of the pathways leading from a mutated gene to various forms of DCM are yet to be discovered. 26

27 Cardiomyopathies in man and dog 2. DEVELOPMENT OF TOOLS IN CANINE GENETICS AND THE CHARACTERISTICS OF THE DOG GENOME 2.1. THE CANINE GENETIC MAPS During the course of the PhD research described in this thesis, tremendous progress has been made in the development of genetic knowledge of the dog. At the start of 2001, an integrated linkage-radiation hybrid map comprising 724 markers on the dog genome was available (Mellersh et al. 2000). Approximately two thirds of these markers were microsatellites and one third were gene markers. At the end of 2001, Breen et al. (2001) further expanded this map by integrating it with the cytogenetic map (Breen et al. 1999). The result was a map with 1078 microsatellite, 320 gene-based and 102 chromosome specific markers (Breen et al. 2001). During the last two years, great progress has been made in the mapping and sequencing of the dog genome (reviewed in Sutter and Ostrander 2004). A 1-Mb resolution radiation hybrid map with 3270 markers was published in 2003 and it was followed by a 900-kb resolution map with 4249 markers (Guyon et al. 2003; Breen et al. 2004). Based on the latest map, a minimal screening set of 327 informative microsatellite markers has been selected for genetic linkage analysis of phenotypes against the entire genome (Clark et al. 2004). 2.2 THE CANINE BACTERIAL ARTIFICIAL CHROMOSOME (BAC) LIBRARY Because the majority of candidate genes for DCM in dogs were not mapped at the start of this DCM research PhD project, the most valuable resource for evaluation of the genes of interest was the canine genomic DNA bacterial artificial chromosome (BAC) library RP81 (Li et al. 1999b). The canine BAC library was constructed by partial digestion of the genomic DNA of a Dobermann and ligation into pbace3.6 vectors. A total of 165,888 BAC clones were arrayed in 432 plates of 384-wells each. DNA from each well of the library was spotted in duplicate onto nine hybridization filters in a specific pattern. In order to clone canine genes, the BAC library is screened with probes that are based on the cdna sequences derived from the human or mouse gene (described in chapters 3 to 5). Microsatellite markers can be isolated from the BAC DNA based on the repetitive character of the DNA sequence of these markers. Since the average insert of a BAC clone is 155 kb, the microsatellite markers which are located in the same BAC clone as the gene of interest reside either in the gene or close to it. 27

28 Chapter 1 If polymorphic, these markers are effective tools to study association of the gene with specific phenotypes because the chance of recombination between the marker and the gene has been all but eliminated. The BAC clones can also be used as probes for fluorescence in situ hybridization (FISH) mapping of the genes located on the clones (chapters 3 and 5). 2.3 THE DOG GENOME SEQUENCE Another essential tool in dog genetics that has just been made available is the dog genome sequence. In 2003, a sequence with 1.5x coverage of the dog genome has been generated from a male standard poodle (Kirkness et al. 2003). This was followed by the assembly of the DNA sequence of a female Boxer with 7.8x coverage in July 2004 (Lindblad-Toh 2004). 2.4 THE DOG GENOME CHARACTERISTICS The expansion of tools available in dog genetics enabled detailed studies of the properties of the dog genome. The purebred dog population consists of over 400 breeds. In order for a dog to become a registered member of a breed, the dog s parents must be registered members of the same breed. The rigorous breeding applied to propagate desired morphological and behavioral traits also resulted in breed-specific inherited disorders. Today, over 350 inherited disorders have been described in dogs and the genetically isolated dog breeds represent unique genetic material to identify genes for these disorders (Ostrander and Comstock 2004). An interesting study exploring genetic relationships between breeds revealed that, while each breed is a distinct genetic entity, most breeds can be grouped into a limited number of genetic clusters (Parker et al. 2004). Furthermore, the peculiar history of dog breeds with limited number of founders, population bottle necks and the use of popular sires, influenced the non-random association between markers known as linkage disequilibrium (LD). LD varies between dog breeds; ranging from <1Mb in the Golden Retriever to 3.2Mb in the Pekingese. In comparison to the LD in the human population, LD in dogs is up to x more extensive. With such extensive LD a smaller number of markers suffice for whole-genome association and linkage studies (Sutter et al. 2004). In conclusion, the 7.8x canine genome sequence ( the 4249 marker canine genome map (Breen et al. 2004) together with the findings regarding LD (Sutter et al. 2004) and genetic relationships between dog breeds (Parker et al. 2004), now provide the canine genetic community with all resources necessary for identification of disease genes. 28

29 Discussion 3. DISCUSSION 3.1 DOG AS A MODEL OF HUMAN DCM Parallels between the myocardial disease in man and dog have been drawn already in the 60s (Wagner 1968) and the dog has been suggested as a naturally occurring large animal model of human DCM (Smucker et al. 1990). It was already recognized as an inherited disease two decades ago (Staaden et al. 1981; Calvert et al. 1982). Researchers all over the world have been gathering family material of dog breeds with frequent DCM occurrence: Dobermanns (Meurs et al. 2001a; Domanjko-Petrič et al. 2002), Great Danes (Meurs et al. 2001b; Skelly et al. 2003), Irish Wolfhounds (Vollmar et al. 2000; Jakobs et al. 2004) Portuguese Water Dogs (Dambach et al. 1999) and Newfoundland dogs (Dukes-McEwan and Jackson 2002). Arrhythmogenic right ventricular cardiomyopathy as a model of its human counterpart has been described in the Boxer (Basso et al. 2004). Genome-wide linkage scans and candidate gene approaches have been used in order to identify the mutations causing DCM in various breeds (Meurs et al. 2001a; Spier et al. 2001; Dukes-McEwan and Jackson 2002; Jakobs et al. 2004; Stabej et al. 2004a). In this thesis, evaluation of five genes for their involvement in Dobermann DCM is described. Mutations in the desmin, δ-sarcoglycan and α- tropomyosin were excluded as a cause of DCM by typing of the microsatellite markers in the DCM and DCM-free Dobermanns. Mutations in the phospholamban gene were excluded by sequencing of the gene in Dobermanns, Newfoundlands and Great Danes with DCM. Described in the 7 th chapter of this thesis is the titin gene, which was found associated with DCM in the Dobermann. This is the first gene found to be involved in pathogenesis of canine DCM (Stabej et al. 2004b). The disease phenotype varies between breeds, while it is more or less homogeneous within a particular breed. This indicates that each breed with frequent DCM occurrence was enriched for a specific type of DCM. However, the Dobermann breed is an exception, since two different phenotypes have been described in this breed: congestive heart failure or sudden death due to ventricular tachyarrhythmia in adulthood. We suggest two possible explanations for this inter- and intra-breed phenotypic heterogeneity. 1. The existence of diverse DCM mutations. 2. The presence of the same mutations resulting in various phenotypes due to different genetic backgrounds. 29

30 Chapter 1 Since more is known about the human cardiomyopathies, the genetic research in dogs has not yet substantially contributed to the knowledge of the molecular genetic background of DCMs. Even though dog breeds are genetic isolates that offer an immense advantage in the searches for disease causing genes in comparison to the human, relatively few canine genetic disorders have been deciphered. The major reason lies in genetic tools like genetic maps and sequence that have only recently become available to the research community. 3.2 ADVANCES IN HUMAN GENETICS SHEDS LIGHT ON PATHOGENETIC BACKGROUNDS OF DCM In the past years, recognition of numerous DCM causing mutations boosted the understanding of molecular mechanisms leading to DCM. Moreover, it contributed considerably toward understanding the physiology of the heart. Until recently, DCM has been described as idiopathic; implicating the disease arises from an obscure or unknown cause (Dec and Fuster 1994). One of the earliest reports on familial autosomal dominant DCM with variable penetrance dates back to 1978 and suggests a mutant gene as a cause (Ross et al. 1978). Ten years later, Berko and Swift (1987) described a family with X-linked DCM, which was succeeded by identification of the first DCM causing mutation a deletion in the dystrophin promoter (Muntoni et al. 1993). The genes causing autosomal dominant DCM were somewhat harder to decipher. Mutations in the actin gene were the first to be revealed by candidate gene approach (Olson et al. 1998). In the years between 1999 and 2005, the availability of a large number of polymorphic markers in the human genome and high throughput sequencing/genotyping methods empowered identification of mutations in several genes either by the candidate gene approach or by genome wide linkage analysis (Table 1). 3.3 THE COMPLEXITY OF DCM AND THE GREAT POTENTIAL OF THE DOG GENOME Despite a large number of known DCM-causing mutations in the human, linkage studies in multiple families have suggested several disease loci in which the DCM causing gene remains to be identified. In addition, mutations in the known causative genes can be found only in a part of the DCM patient population. Therefore, more genes are expected to be discovered in various DCM families. Another unsolved DCM riddle is the observed significant variability in 30

31 Discussion phenotypic expression of DCM among individuals with identical causal mutations (Olson et al. 2001) or the same mutations causing either cardiomyopathy or isolated skeletal myopathy (Dalakas et al. 2003). Furthermore, the gender-specific difference in phenotype of DCM patients in some families also remains unexplained (Stefanelli et al. 2004). The variability of outcome of DCM patients highlights the importance of genetic background (modifier genes) and environmental factors in phenotype determination. Modifier genes are genes that are not involved in the genesis of the disease, but modify the severity of the phenotypic expression once the disease has developed (Le Corvoisier et al. 2003b). Not much is known about modifier genes and protective alleles in human DCM. The heterogeneity of the human population coupled with the generally modest effect of the modifier genes make identification of modifier genes a complex task requiring complementary approaches. Mouse models generated to develop DCM circumvent the problem of heterogeneity to some degree. Quantitative trait locus (QTL) mapping in an experimental mouse model of DCM, induced by cardiac-specific over-expression of calsequestrin (CSQ), and characterized by a strong strain-specific variability in phenotype, resulted in identification of several QTLs that differentially modify the cardiac phenotype. Although experimental models like the CSQ mouse recapitulate a number of human/dog DCM key features, the modifier loci identified in such a model are likely to be specific for DCM induced by CSQ over-expression (Le Corvoisier et al. 2003a). To overcome the limitation of experimental models and high heterogeneity of the human population, dogs seem to be a perfect naturally occurring animal model that can help disentangle the complex effects of the genetic background on the development of DCM. Studies of linkage disequilibrium (LD) in dogs demonstrated that the LD varies between breeds and extends up to 100x farther than LD in human populations. In addition, the haplotype diversity in dog breeds is relatively low with two and three haplotypes accounting for 80% of the chromosomes in each breed (Sutter et al. 2004). Extensive LD, coupled with a high degree of haplotype sharing, make wholegenome association scans more amenable in dog breeds than in human populations (Sutter et al. 2004a). Given the knowledge about the genetic structure of dog breeds (Parker et al. 2004) as well as the availability of important genetic resources as the 7.8x dog sequence ( and a highresolution genetic map (Breen et al. 2004), the cloning of DCM causing genes and modifier genes will expedited. Another disadvantage in the study of human DCM that dogs can overcome are small families. Several DCM genes have been identified by screening hundreds of unrelated patients for mutations in DCM candidate genes. For example, Olson et al. (2001) screened 350 unrelated patients for mutations in 31

32 Chapter 1 TPM1 and identified unique missense mutations in only two cases and Itoh-Satoh et al. (2002) had to screen 120 unrelated patients to identify three TTN mutations. The family material obtained with such an approach is usually limited and expanding the material to make it sufficient for linkage analysis often proves to be a daunting task (Daehmlow et al. 2002). Dog breeds can bypass the extensive phenotypic heterogeneity seen in human patients. Within most breeds predisposed to DCM, the phenotype is homogeneous. Therefore, either linkage analyses performed in closely related dogs or association studies performed in unrelated DCM patients of specific breeds are warranted. Several groups have been working on DCM in dogs and have gathered valuable DNA material. A good connection of the cardiologists and geneticists is essential in the genetic studies of DCM. Establishing good worldwide communication between DCM groups and making a common DNA database of DCM dogs, accompanied with detailed phenotypic descriptions would provide an invaluable resource for future DCM studies. 32

33 Objectives of this thesis OBJECTIVES OF THIS THESIS Canine dilated cardiomyopathy (DCM) is a disease of the myocardium associated with dilatation and impaired contraction of the ventricles. It primarily affects large and giant breed dogs with Dobermanns being one of the most frequently affected. The high prevalence of DCM in specific breeds suggests a genetic background, but causal mutations have not yet been identified. The main objective of the study described in this thesis was to identify the gene(s) involved in DCM in the Dobermann. For that purpose, the genes of interest were evaluated. Due to close similarities in human and canine DCM, mutated genes identified to cause DCM in the human are good candidate genes for DCM in dogs. At the outset of the DCM research project in 2001, most of the DCM candidate genes had not yet been mapped and the dog genome resources were limited. Therefore, bacterial artificial chromosome (BAC) clones carrying the genes of interest were isolated from the canine BAC library. Microsatellite markers isolated from the BAC clones were typed in the Dobermanns and evaluated for association with the DCM phenotype. While the genes encoding desmin, phospholamban, α-tropomyosin and δ-sarcoglycan were excluded as DCM causing genes, the genotyping results of the titin markers point to a crucial role of titin in DCM susceptibility. The genetic markers and sequencing oligonucleotide primers reported in this thesis are tools that will enable fast evaluation of the DCM candidate genes in breeds other than the Dobermann. Further research into the role of the titin in canine DCM will also lead to better understanding of cardiomyocyte physiology and pathophysiology of DCM. 33

34 Chapter 1 REFERENCES Arbustini, E., Pilotto, A., Repetto, A., Grasso, M., Negri, A., Diegoli, M., Campana, C., Scelsi, L., Baldini, E., Gavazzi, A., et al Autosomal dominant dilated cardiomyopathy with atrioventricular block: a lamin A/C defect-related disease. J Am Coll Cardiol. 39: Arimura, T., Hayashi, T., Terada, H., Lee, S.Y., Zhou, Q., Takahashi, M., Ueda, K., Nouchi, T., Hohda, S., Shibutani, M., et al A Cypher/ZASP mutation associated with dilated cardiomyopathy alters the binding affinity to protein kinase C. J Biol Chem. 279: Barth, P.G., Valianpour, F., Bowen, V.M., Lam, J., Duran, M., Vaz, F.M., Wanders, R.J X-linked cardioskeletal myopathy and neutropenia (Barth syndrome): an update. Am J Med Genet. 126: Basso, C., Fox, P.R., Meurs, K.M., Towbin, J.A., Spier, A.W., Calabrese, F., Maron, B.J., Thiene, G Arrhythmogenic right ventricular cardiomyopathy causing sudden cardiac death in boxer dogs: a new animal model of human disease. Circulation 109: Becane, H.M., Bonne, G., Varnous, S., Muchir, A., Ortega, V., Hammouda, E.H., Urtizberea, J.A., Lavergne, T., Fardeau, M., Eymard, B., Weber, S., et al High incidence of sudden death with conduction system and myocardial disease due to lamins A and C gene mutation. Pacing Clin Electrophysiol. 23: Berko, B.A., Swift, M X-linked dilated cardiomyopathy. N Engl J Med. 316: Bienengraeber, M., Alekseev, A.E., Abraham, M.R., Carrasco, A.J., Moreau, C., Vivaudou, M., Dzeja, P.P., Terzic, A ATPase activity of the sulfonylurea receptor: a catalytic function for the KATP channel complex. FASEB J. 14: Bienengraeber, M., Olson, T.M., Selivanov, V.A., Kathmann, E.C., O'Cochlain, F., Gao, F., Karger, A.B., Ballew, J.D., Hodgson, D.M., Zingman, L.V., et al ABCC9 mutations identified in human dilated cardiomyopathy disrupt catalytic KATP channel gating. Nat Genet. 36: Bione, S., Maestrini, E., Rivella, S., Mancini, M., Regis, S., Romeo, G., Toniolo, D Identification of a novel X-linked gene responsible for Emery-Dreifuss muscular dystrophy. Nat Genet. 8: Bonagura, J.D, Luis Fuentes, V Echocardiography. In Textbook of Veterinary Internal Medicine, 5 th edition (eds. S.J. Ettinger and E.C. Feldman), pp WB Saunders, Philadelphia. Breen, M., Hitte, C., Lorentzen, T.D., Thomas, R., Cadieu, E., Sabacan, L., Scott, A., Evanno, G., Parker, H.G., Kirkness, E.F., et al An integrated 4249 marker FISH/RH map of the canine genome. BMC Genomics 5: 65. Breen, M., Jouquand, S., Renier, C., Mellersh, C.S., Hitte, C., Holmes, N.G., Cheron, A., Suter, N., Vignaux, F., Bristow, et al Chromosome-specific single-locus FISH probes allow anchorage of an 1800-marker integrated radiation-hybrid/linkage map of the domestic dog genome to all chromosomes. Genome Res. 11:

35 Breen, M., Thomas, R., Binns, M.M., Carter, N.P., Langford, C.F Reciprocal chromosome painting reveals detailed regions of conserved synteny between the karyotypes of the domestic dog (Canis familiaris) and human. Genomics 61: Brodsky, G.L., Muntoni, F., Miocic, S., Sinagra, G., Sewry, C., Mestroni, L Lamin A/C gene mutation associated with dilated cardiomyopathy with variable skeletal muscle involvement. Circulation 101: Brownlie, S.E., Cobb, M.A Observations on the development of congestive heart failure in Irish wolfhounds with DCM. J Small Anim Pract. 40: Burke, B., Stewart, C.L Life at the edge: the nuclear envelope and human disease. Nat Rev Mol Cell Biol. 3: Calvert, C.A., Chapman, W.L. Jr, Toal, R.L Congestive cardiomyopathy in Doberman pinscher dogs. J Am Vet Med Assoc. 181: Calvert, C.A., Meurs, K.M CVT Update: Doberman pinscher occult cardiomyopathy. In Kirk's Current Veterinary Therapy XIII (ed. J.D. Bonagura), pp WB Saunders, Philadelphia. Calvert, C.A., Pickus, C.W., Jacobs, G.J., Brown, J Signalment, survival, and prognostic Factors in Dobermanns with end-stage cardiomyopathy. J Vet Intern Med. 11: Clark, L.A., Tsai, K.L., Steiner, J.M., Williams, D.A., Guerra, T., Ostrander, E.A., Galibert, F., Murphy, K.E Chromosome-specific microsatellite multiplex sets for linkage studies in the domestic dog. Genomics 84: Cohen, N., Muntoni, F Multiple pathogenetic mechanisms in X linked dilated cardiomyopathy. Heart 90: D'Adamo, P., Fassone, L., Gedeon, A., Janssen, E.A., Bione, S., Bolhuis, P.A., Barth, P.G., Wilson, M., Haan, E., et al The X-linked gene G4.5 is responsible for different infantile dilated cardiomyopathies. Am J Hum Genet. 61: Daehmlow, S., Erdmann, J., Knueppel, T., Gille, C., Froemmel, C., Hummel, M., Hetzer, R., Regitz-Zagrosek, V Novel mutations in sarcomeric protein genes in dilated cardiomyopathy. Biochem Biophys Res Commun. 298: Dalakas, M.C., Dagvadorj, A., Goudeau, B., Park, K.Y., Takeda, K., Simon-Casteras, M., Vasconcelos, O., Sambuughin, N., Shatunov, A., Nagle, J.W., et al Progressive skeletal myopathy, a phenotypic variant of desmin myopathy associated with desmin mutations. Neuromuscul Disord. 13: Dambach, D.M., Lannon, A., Sleeper, M.M., Buchanan, J Familial dilated cardiomyopathy of young Portuguese water dogs. J Vet Intern Med. 13: Dec, G.W., Fuster, V Idiopathic dilated cardiomyopathy. N Engl J Med. 331: Domanjko-Petrič, A., Stabej, P., Žemva, A Dilated cardiomyopathy in Dobermanns, survival, causes of death and pedigree review in a related line. J Vet Cardiol. 4: Dominguez, R Actin-binding proteins - a unifying hypothesis. Trends Biochem Sci. 29: Dukes-McEwan, J., Jackson, I.J The promises and problems of linkage analysis by using the current canine genome map. Mamm Genome 13: Fatkin, D., Graham, R.M Molecular mechanisms of inherited cardiomyopathies. Physiol Rev. 82:

36 Chapter 1 Fatkin, D., MacRae, C., Sasaki, T., Wolff, M.R., Porcu, M., Frenneaux, M., Atherton, J., Vidaillet, H.J., Spudich, S., De Girolami, U., et al Missense mutations in the rod domain of the lamin A/C gene as causes of dilated cardiomyopathy and conductionsystem disease. N Engl J Med. 341: Freeman, L.M., Michel, K.E., Brown, D.J., Kaplan, P.M., Stamoulis, M.E., Rosenthal, S.L., Keene, B.W., Rush, J.E Idiopathic dilated cardiomyopathy in Dalmations: Nine cases ( ). JAVMA 209: Gerull, B., Gramlich, M., Atherton, J., McNabb, M., Trombitas, K., Sasse-Klaassen, S., Seidman, J.G., Seidman, et al Mutations of TTN, encoding the giant muscle filament titin, cause familial dilated cardiomyopathy. Nat Genet. 30: Gomes, A.V., Potter, J.D Molecular and cellular aspects of troponin cardiomyopathies. Ann N Y Acad Sci. 1015: Gooding, J.P., Robinson, W.F., Mews, G.C Echocardiographic characterization of dilatation cardiomyopathy in the English cocker spaniel. Am J Vet Res. 47: Granzier, H.L., Labeit, S The giant protein titin: a major player in myocardial mechanics, signaling, and disease. Circ Res. 94: Gregorio, C.C., Granzier, H., Sorimachi, H., Labeit, S Muscle assembly: a titanic achievement? Curr Opin Cell Biol. 11: Grunig, E., Tasman, J.A., Kucherer, H., et al Frequency and phenotypes of familial dilated cardiomyopathy. J Am Coll Cardiol. 31: Guyon, R., Lorentzen, T.D., Hitte, C., Kim, L., Cadieu, E., Parker, H.G., Quignon, P., Lowe JK, Renier C, Gelfenbeyn B, et al A 1-Mb resolution radiation hybrid map of the canine genome. Proc Natl Acad Sci U S A. 100: Hammer, T.A., Venta, P.J., Eyster, G.E The genetic basis of dilated cardiomyopathy in Dobermanns. Anim Gen. 27: 101. Harada, K., Potter, J.D Familial hypertrophic cardiomyopathy mutations from different functional regions of troponin T result in different effects on the ph and Ca2+ sensitivity of cardiac muscle contraction. J Biol Chem. 279: Hershberger, R.E., Hanson, E.L., Jakobs, P.M., Keegan, H., Coates, K., Bousman, S., Litt, M A novel lamin A/C mutation in a family with dilated cardiomyopathy, prominent conduction system disease, and need for permanent pacemaker implantation. Am Heart J. 144: Hutchison, C.J., Worman, H.J A-type lamins: Guardians of the soma? Nat Cell Biol. 6: Itoh-Satoh, M., Hayashi, T., Nishi, H., Koga, Y., Arimura, T., Koyanagi, T., Takahashi, M., Hohda, S., Ueda, K., Nouchi, T., et al Titin mutations as the molecular basis for dilated cardiomyopathy. Biochem Biophys Res Commun. 291: Jakobs, P., Bestwick, M.L., Ludwigsen, S.J., Nelsen, S.M., Winther, M.J., Hershberger, R.E., Litt, M Genetic linkage analysis of cardiomyopathy in Irish Wolfhounds. In: 2 nd International Conference Advances in Canine and Feline Genomics, Utrecht, The Netherlands, October 14-17; page 77. Kamisago, M., Sharma, S.D., DePalma, S.R., Solomon, S., Sharma, P., McDonough, B., Smoot, L., Mullen, M.P., Woolf, P.K., Wigle, E.D., et al Mutations in sarcomere protein genes as a cause of dilated cardiomyopathy. N Engl J Med. 343: Karkkainen, S., Helio, T., Miettinen, R., Tuomainen, P., Peltola, P., Rummukainen, J., Ylitalo, K., Kaartinen, M., Kuusisto, J., Toivonen, L., et al A novel mutation, Ser143Pro, in the lamin A/C gene is common in finnish patients with familial dilated cardiomyopathy. Eur Heart J. 25:

37 References Keene, B.W., Panciera, D.P., Atkins, C.E., Regitz, V., Schmidt, M.J., Shug, A.L Myocardial L-carnitine deficiency in a family of dogs with dilated cardiomyopathy. J Am Vet Med Assoc. 198: Kirkness, E.F., Bafna, V., Halpern, A.L., Levy, S., Remington, K., Rusch, D.B., Delcher, A.L., Pop, M., Wang, W., Fraser, C.M., et al The dog genome: survey sequencing and comparative analysis. Science 301: Kittleson, M.D., Keene, B., Pion, P Results of the multicenter spaniel trial (MUST): Taurine- and carnitine- responsive dilated cardiomyopathy in American Cocker spaniels with decreased plasma taurine concentration. J Vet Intern Med. 11: Knoll, R., Hoshijima, M., Hoffman, H.M., Person, V., Lorenzen-Schmidt, I., Bang, M.L., Hayashi, T., Shiga, N., Yasukawa, H., Schaper, W., et al The cardiac mechanical stretch sensor machinery involves a Z disc complex that is defective in a subset of human dilated cardiomyopathy. Cell 111: Kraus, M.S., Moise, N.S., Rishniw, M., Dykes, N., Erb, H.N Morphology of ventricular arrhythmias in the Boxer as measured by 12-lead electrocardiography with pace-mapping comparison. J Vet Intern Med. 16: Labeit, S., Kolmerer, B., Linke, W.A The giant protein titin. Emerging roles in physiology and pathophysiology. Circ Res. 80: Lapidos, K.A., Kakkar, R., McNally, E.M The dystrophin glycoprotein complex: signaling strength and integrity for the sarcolemma. Circ Res. 94: Le Corvoisier, P., Park, H.Y., Carlson, K.M., Marchuk, D.A., Rockman, H.A. 2003a. Multiple quantitative trait loci modify the heart failure phenotype in murine cardiomyopathy. Hum Mol Genet. 12: Le Corvoisier, P., Park, H.Y., Rockman, H.A. 2003b. Modifier genes and heart failure. Minerva Cardioangiol. 51: Lee, B.H., Dukes-McEwan, J., French, A., French, A.T., Corcoran, B.M Evaluation of a novel doppler index of combined systolic and diastolic myocardial performance in Newfoundland dogs with familial prevalence of dilated cardiomyopathy. Vet Radiol Ultrasound 43: Li, D., Czernuszewicz, G.Z., Gonzalez, O., Tapscott, T., Karibe, A., Durand, J.B., Brugada, R., Hill, R., Gregoritch, J.M. et al Novel cardiac troponin T mutation as a cause of familial dilated cardiomyopathy. Circulation 104: Li, D., Tapscoft, T., Gonzalez, O., Burch, P.E., Quinones, M.A., Zoghbi, W.A., Hill, R., Bachinski, L.L., Mann, D.L., Roberts R. 1999a. Desmin mutation responsible for idiopathic dilated cardiomyopathy. Circulation 100: Li, R., Mignot, E., Faraco, J., Kadotani, H., Cantanese, J., Zhao, B., Lin, X., Hinton, L., Ostrander, E.A., Patterson, D.F., et al. 1999b. Construction and characterization of an eightfold redundant dog genomic bacterial artificial chromosome library. Genomics 58: Lindblad-Toh, K., Preliminary analysis of a high-quality draft sequence in the dog genome. In: 2 nd International Conference Advances in Canine and Feline Genomics, Utrecht, The Netherlands, October (2004); page 21. Lu, Q.W., Morimoto, S., Harada, K., Du, C.K., Takahashi-Yanaga, F., Miwa, Y., Sasaguri, T., Ohtsuki, I Cardiac troponin T mutation R141W found in dilated cardiomyopathy stabilizes the troponin T-tropomyosin interaction and causes a Ca2+ desensitization. J Mol Cell Cardiol. 35: MacLennan, D.H., Kranias, E.G Phospholamban: a crucial regulator of cardiac contractility. Nat Rev Mol Cell Biol. 4:

38 Chapter 1 Marin-Garcia, J., Goldenthal, M.J., Moe, G.W Mitochondrial pathology in cardiac failure. Cardiovasc Res. 49: McNair, W.P., Ku, L., Taylor, M.R., Fain, P.R., Dao, D., Wolfel, E., Mestroni, L., Familial Cardiomyopathy Registry Research Group SCN5A mutation associated with dilated cardiomyopathy, conduction disorder, and arrhythmia. Circulation 110: Mellersh, C.S., Hitte, C., Richman, M., Vignaux, F., Priat, C., Jouquand, S., Werner, P., Andre, C., DeRose, S., Patterson, D.F., et al An integrated linkage-radiation hybrid map of the canine genome. Mamm Genome. 11: Meurs, K.M., Magnon, A.L., Spier, A.W., Miller, M.W., Lehmkuhl, L.B., Towbin, J.A. 2001a. Evaluation of the cardiac actin gene in Dobermanns with dilated cardiomyopathy. Am J Vet Res. 62: Meurs, K.M., Miller, M.W., Wright, N.A. 2001b. Clinical features of dilated cardiomyopathy in Great Danes and results of a pedigree analysis: 17 cases ( ). J Am Vet Med Assoc. 218: Monnet, E., Orton, E.C., Salman, M., Salman, M., Boon, J Idiopathic dilated cardiomyopathy in dogs: Survival and prognostic indicators. J Vet Intern Med. 9: Muntoni, F., Cau, M., Ganau, A., Congiu, R., Arvedi, G., Mateddu, A., Marrosu, M.G., Cianchetti, C., Realdi, G., Cao, A., et al Brief report: deletion of the dystrophin muscle-promoter region associated with X-linked dilated cardiomyopathy. N Engl J Med. 329: Muntoni, F., Di Lenarda, A., Porcu, M., Sinagra, G., Mateddu, A., Marrosu, G., Ferlini, A., Cau, M., Milasin, J., Melis MA, et al Dystrophin gene abnormalities in two patients with idiopathic dilated cardiomyopathy. Heart. 78: Murphy, R.T., Mogensen, J., Shaw, A., Kubo, T., Hughes, S., McKenna, W.J Novel mutation in cardiac troponin I in recessive idiopathic dilated cardiomyopathy. Lancet 363: Napolitano, C., Rivolta, I., Priori, S.G Cardiac sodium channel diseases. Clin Chem Lab Med. 41: Nigro, G., Politano, L., Nigro, V., Petretta, V.R., Comi, L.I Mutation of dystrophin gene and cardiomyopathy. Neuromuscul Disord. 4: Nishino, I., Fu, J., Tanji, K., Yamada, T., Shimojo, S., Koori, T., Mora, M., Riggs, J.E., Oh, S.J., Koga, Y., et al Primary LAMP-2 deficiency causes X-linked vacuolar cardiomyopathy and myopathy (Danon disease). Nature. 406: O Grady, M.R., Horne, R. Occult dilated cardiomyopathy in Doberman Pinscher. Proc. of 13 th ACVIM Forum, Lake Buena Vista, FL 1995; Olson, T.M., Illenberger, S., Kishimoto, N.Y., Huttelmaier, S., Keating, M.T., Jockusch, B.M Metavinculin mutations alter actin interaction in dilated cardiomyopathy. Circulation 105: Olson, T.M., Keating, M.T Mapping a cardiomyopathy locus to chromosome 3p22-p25. J Clin Invest. 97: Olson, T.M., Kishimoto, N.Y., Whitby, F.G., Michels, V.V Mutations that alter the surface charge of alpha-tropomyosin are associated with dilated cardiomyopathy. J Mol Cell Cardiol. 33: Ostlund, C., Worman, H.J Nuclear envelope proteins and neuromuscular diseases. Muscle Nerve. 27: Ostrander, E.A., Comstock, K.E The domestic dog genome. Curr Biol. 14: R

39 References Parker, H.G., Kim, L.V., Sutter, N.B., Carlson, S., Lorentzen, T.D., Malek, T.B., Johnson, G.S., DeFrance, H.B., Ostrander, E.A., Kruglyak, L Genetic structure of the purebred domestic dog. Science 304: Paulin, D., Li, Z Desmin: a major intermediate filament protein essential for the structural integrity and function of muscle. Exp Cell Res. 301: 1-7. Pyle, W.G., Solaro, R.J At the crossroads of myocardial signaling: the role of Z- discs in intracellular signaling and cardiac function. Circ Res. 94: Richardson, P., McKenna, W., Bristow, M., Maisch, B., Mautner, B., O'Connell, J., Olsen, E., Thiene G, Goodwin J, Gyarfas I, et al Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the Definition and Classification of cardiomyopathies. Circulation 93: Ross, R.S., Bulkley, B.H., Hutchins, G.M., Harshey, J.S., Jones, R.A., Kraus, H., Liebman, J., Thorne, C.M., Weinberg, S.B., Weech, A.A. et al Idiopathic familial myocardiopathy in three generations: a clinical and pathologic study. Am Heart J. 96: Rybakova, I.N., Patel, J.R., Ervasti, J.M The dystrophin complex forms a mechanically strong link between the sarcolemma and costameric actin. J Cell Biol. 150: Schmitt, J.P., Kamisago, M., Asahi, M., Li, G.H., Ahmad, F., Mende, U., Kranias, E.G., MacLennan, D.H., Seidman, J.G., Seidman, C.E Dilated cardiomyopathy and heart failure caused by a mutation in phospholamban. Science 299: Sebillon, P., Bouchier, C., Bidot, L.D., Bonne, G., Ahamed, K., Charron, P., Drouin- Garraud, V., Millaire, A., Desrumeaux, G., Benaiche, A., et al Expanding the phenotype of LMNA mutations in dilated cardiomyopathy and functional consequences of these mutations. J Med Genet. 40: Sisson, D.D., Thomas, W.P., Keene, B.W Primary myocardial disease in the dog. In Textbook of Veterinary Internal Medicine 5th ed. (eds. S.J. Ettinger, E.C. Feldman), pp WB, Saunders, Philadelphia. Siu, B.L., Niimura, H., Osborne, J.A., Fatkin, D., MacRae, C., Solomon, S., Benson, D.W., Seidman, J.G., Seidman, C.E Familial dilated cardiomyopathy locus maps to chromosome 2q31. Circulation 99: Skelly, B Towards a molecular test for dilated cardiomyopathy in great danes. J Small Anim Pract. 44: Sleeper, M.M., Henthorn, P.S., Vijayasarathy, C., Dambach, D.M., Bowers, T., Tijskens, P., Armstrong, C.F., Lankford, E.B Dilated cardiomyopathy in juvenile Portuguese Water Dogs. J Vet Intern Med. 16: Smucker, M.L., Kaul, S., Woodfield, J.A., Keith, J.C., Manning, S.A., Gascho, J.A Naturally occurring cardiomyopathy in the Doberman pinscher: a possible large animal model of human cardiomyopathy? J Am Coll Cardiol. 16: Spier, A.W., Meurs, K.M., Coovert, D.D., Lehmkuhl, L.B., O'Grady, M.R., Freeman, L.M., Burghes, A.H., Towbin, J.A Use of western immunoblot for evaluation of myocardial dystrophin, alpha-sarcoglycan, and beta-dystroglycan in dogs with idiopathic dilated cardiomyopathy. Am J Vet Res. 62: Staaden, R.V Cardiomyopathy of English cocker spaniels. J Am Vet Med Assoc. 178: Stabej, P., Imholz, S., Versteeg, S.A., Zijlstra, C., Stokhof, A.A., Domanjko-Petric, A., Leegwater, P.A., van Oost, B.A. 2004a. Characterization of the canine desmin (DES) gene and evaluation as a candidate gene for dilated cardiomyopathy in the Dobermann. Gene 340:

40 Chapter 1 Stabej, P., Leegwater, P.A., Imholz, S., Loohuis, M., Mandigers, P., Stokhof, A.A., Domanjko-Petric, A., van Oost, B.A. 2004b. The giant myofilament protein Titin is associated with dilated cardiomyopathy in the Dobermann. In: 2 nd International Conference Advances in Canine and Feline Genomics, Utrecht, The Netherlands, October 14-17; page 28 Stefanelli, C.B., Rosenthal, A., Borisov, A.B., Ensing, G.J., Russell, M.W Novel troponin T mutation in familial dilated cardiomyopathy with gender-dependant severity. Mol Genet Metab. 83: Suomalainen, A., Paetau, A., Leinonen, H., Majander, A., Peltonen, L., Somer, H Inherited idiopathic dilated cardiomyopathy with multiple deletions of mitochondrial DNA. Lancet. 340: Sutter, N.B., Eberle, M.A., Parker, H.G., Pullar, B.J., Kirkness, E.F., Kruglyak, L., Ostrander, E.A Extensive and breed-specific linkage disequilibrium in Canis familiaris. Genome Res. 14: Sutter, N.B., Ostrander, E.A Dog star rising: the canine genetic system. Nat Rev Genet. 5: Taylor, M.R., Fain, P.R., Sinagra, G., Robinson, M.L., Robertson, A.D., Carniel, E., Di Lenarda, A., Bohlmeyer, T.J., Ferguson, D.A., Brodsky, G.L., et al Natural history of dilated cardiomyopathy due to lamin A/C gene mutations. J Am Coll Cardiol. 41: Tidholm, A., Jonsson, L A retrospective study of canine dilated cardiomyopathy (189 cases). J Am Anim Hosp Assoc 33: Tidholm, A., Jonsson, L Dilated Cardiomyopathy in the Newfoundland: a Study of 37 Cases. J Am Anim Hosp Assoc. 32: Tiso, N., Stephan, D.A., Nava, A., Bagattin, A., Devaney, J.M., Stanchi, F., Larderet, G., Brahmbhatt, B., Brown, K., Bauce, B., et al Identification of mutations in the cardiac ryanodine receptor gene in families affected with arrhythmogenic right ventricular cardiomyopathy type 2 (ARVD2). Hum Mol Genet. 10: Towbin, J.A., Hejtmancik, J.F., Brink, P., Gelb, B., Zhu, X.M., Chamberlain, J.S., McCabe, E.R., Swift, M X-linked dilated cardiomyopathy. Molecular genetic evidence of linkage to the Duchenne muscular dystrophy (dystrophin) gene at the Xp21 locus. Circulation 87: Tsubata, S., Bowles, K.R., Vatta, M., Zintz, C., Titus, J., Muhonen, L., Bowles, N.E., Towbin, J.A Mutations in the human delta-sarcoglycan gene in familial and sporadic dilated cardiomyopathy. J Clin Invest. 106: Viswanathan, P.C., Balser, J.R Inherited sodium channelopathies: a continuum of channel dysfunction. Trends Cardiovasc Med. 14: Vollmar, A.C The prevalence of cardiomyopathy in the Irish wolfhound: a clinical study of 500 dogs. J Am Anim Hosp Assoc. 36: Wagner, B.M Myocardial disease in man and dog, some properties. Ann N Y Acad Sci. 147: Wotton, P.R Cardiomyopathy in English cocker and springer spaniels: A review of 38 cases. Proceedings of the British Small Animal Veterinary Association, pp

41

42

43 Chapter 2 Dilated cardiomyopathy in the Dobermann dog: survival, causes of death and a pedigree review in a related line Aleksandra Domanjko-Petrič, Polona Stabej, A. Žemva Journal of Veterinary Cardiology 2002; 4:17-24

44 Chapter 2 ABSTRACT The objective of this study was to analyze clinical, electrocardiographic and echocardiographic characteristics, survival, cause of death and possible modes of inheritance of dilated cardiomyopathy (DCM) in Dobermanns (DO) and to compare the occurrence and survival with dogs of other breeds. Two cohorts of dogs were studied: 1. A consecutive series of 52 dogs of different breeds with DCM were included, 21 were Dobermanns and 31 dogs belonged to other breeds; 2. Twenty-eight asymptomatic Dobermanns, who were screened for DCM. Medical records of dogs with DCM were reviewed. Physical, electrocardiographic and echocardiographic examinations were performed on asymptomatic Dobermanns. Their pedigrees were reviewed. Dobermanns survived on average 52 days (range <1-180), while dogs of other breeds survived significantly longer, i.e. 240 days (<1-1230). Survival of Dobermanns in congestive heart failure (mean 62, range <1-180) was not different from survival of Dobermanns with sudden death (mean 33, range <1-105). High prevalence, short survival time and the clinical course of DCM in Dobermanns showed similarities to previous studies. Twenty-one percent of asymptomatic Dobermanns had increased left ventricular end-systolic diameter and 14% developed DCM within a year. A line of Dobermanns with multiple members affected with DCM was identified by the review of their pedigrees. Exact mode of inheritance could not be established. The prognosis of Dobermanns with DCM is poor. Further molecular genetic studies, which would enable detection and exclusion of disease carriers from the breeding, are necessary. 44

45 INTRODUCTION Dilated Cardiomyopathy in the Dobermann Dilated cardiomyopathy is a well-known and common cause of congestive heart failure and sudden death in dogs. Dobermanns have been reported to be the most commonly affected breed (Lombard 1984; Calvert 1986; Calvert 1992; Monnet et al. 1995; Hammer et al. 1996). Frequent occurrence of the disease in some lines of Dobermanns has been observed (Calvert 1986; Calvert 1992; Hammer et al. 1996; Calvert 1982). Consequently, it was hypothesized, that the disease might have a genetic basis. The mode of inheritance of DCM in Dobermanns has not been fully elucidated. One abstract in the literature suggested that the mode of inheritance of DCM in Dobermanns is autosomal dominant with reduced penetrance (Hammer et al. 1996). Genetic origin of the same disease has been suspected in other breeds (Staaden 1981; Gooding et al. 1986; Goodwin et al. 1995, Tidholm et al. 2000; Dambach et al. 1999). In addition, human studies have revealed a genetic component to DCM in a significant proportion of cases (Michels et al. 1985; Berko and Swift 1987; Emery 1987; Abelmann and Lorell 1998; Zeviani et al. 1991; Michels et al. 1992; Towbin et al. 1993; Goerss et al. 1995; Keeling et al. 1995). The purpose of this study was to analyze the clinical, electrocardiographic, radiographic and echocardiographic characteristics of DCM and the cause of death in Dobermanns and to compare the occurrence and survival with cases of DCM in other breeds of dogs. Along with this DCM group of dogs (1 st cohort), a small group of 28 asymptomatic Dobermanns (2 nd cohort) were studied by means of physical examination, electrocardiography and echocardiography to determine the prevalence of DCM in these dogs. In order to evaluate the frequency of familial occurrence of the disease and to assess the mode of inheritance of the disease, pedigree review was performed. MATERIALS AND METHODS ANIMALS Our study consisted of two cohorts of dogs. The first cohort consisted of 52 dogs (consecutive series) of different breeds with DCM (1 st cohort) who were admitted to the outpatient department of the Clinic for Small Animal Medicine and Surgery, Veterinary Faculty of Ljubljana, Slovenia within a period of five years. The second cohort was comprised of 28 asymptomatic Dobermanns (2 nd 45

46 Chapter 2 cohort), whose owners participated in the DCM research, and were screened for any signs of heart disease. PROCEDURES The medical records of dogs with DCM were reviewed. In the 2 nd cohort, a 3- minute ECG (standard leads) was recorded after a physical examination. Electrocardiographic parameters, (measured in lead II), were compared to the normal values (Tilley 1992). Thoracic radiographs were assessed subjectively (Suter and Lord 1984). Echocardiography (M - mode and two dimensional) was performed with a 5 MHz sector transducer using standard techniques (Thomas et al. 1993). Dogs were not sedated. They were examined in the right lateral recumbence on a plastic table with cutting outs to enable the transducer to be oriented from below. Simultaneous ECG was used to define timing of the measurements. Three measurements were averaged for each echocardiographic variable. M-mode was obtained from the left ventricular short axis view. In cases of atrial fibrillation we measured the beats, which occurred after longer diastolic filing and produced better contractility (Feigenbaum 1994). When PVC s were present, measuring the beats right before or after PVC s occurred was avoided. The diagnosis of DCM was made either by echocardiography or by clinical findings with pathomorphologic/pathohistologic findings (Van Fleet and Feranns 1986; Everett et al. 1999; Tidholm et al. 1998; Calvert et al. 1997; Kittleson 1998a). Post mortem findings such as an enlarged heart with severe dilatation of left or all 4 heart chambers, with thin walls, dilated atrioventricular ring and no significant changes of the valvular apparatus or other heart structures were considered diagnostic of DCM. A finding of a mild AV valve myxomatous degeneration in older dogs was considered insignificant concurrent finding. Common simultaneous findings were pulmonary edema and liver congestion, in some cases accompanied by pleural or abdominal effusion. In the 2 nd cohort, dogs were assigned as having DCM on the basis of abnormal echocardiographic findings or at least one PVC recorded on the ECG (O Grady and Horne 1995a). Echocardiographic diagnostic criteria for dogs of other breeds than Dobermanns were normal values referred by Bonagura et al. (1985), summarized in a table by Ware (1992). SF values 25 % were considered abnormal. The following echocardiographic variables of Dobermanns from the 1 st cohort and from the 2 nd cohort were considered abnormal: left ventricular fractional shortening (FS) <25, left ventricular end-diastolic diameter (LVEDD)>46mm in dogs weighing 42 kg and >50 mm in dogs weighing >42 kg, and left ventricular end-systolic diameter (LVESD) >38 mm (Calvert 1982; 46

47 Dilated cardiomyopathy in the Dobermann O Grady and Horne 1995a; O Grady and Horne 1995b; Calvert et al. 2000a; Sottiaux and Amberger 1997). Presented data are the results of the first examination. Within the 1 st cohort the percentage of different breeds, age at the time of diagnosis and length of survival in Dobermanns (DO 1 st cohort) in comparison to other breeds (OB 1 st cohort) were calculated. Survival times were calculated from the time of the 1 st presentation (when DCM diagnosis was made) to the death or euthanasia. Treatment analysis was not included in the study. Medical therapy consisted of variable drugs used to treat CHF and arrhythmia. Therapeutics and number of dogs receiving each therapeutic are listed in Table 1. Asymptomatic dogs with evidence of DCM were not treated. The cause of death was registered (congestive heart failure - CHF, sudden death - SD, other causes). We used published criteria for sudden cardiac death, i.e. dogs that were in heart or myocardial failure, must be stable one hour before death (Wilson et al. 1983). The pedigrees of Dobermanns of both groups were reviewed to find relatives with DCM and to try to determine the mode of inheritance of the disease. Table 1. Number of dogs of both groups receiving listed therapeutics DO OB n CHF n SD n Furosemide Digoxin ACE-I Propranolol Mexiletine 3 2 Lidocaine 2 1 Disopyramide 1 1 Propafenone 1 Dopamine HCl 1 1 Spironolactone 1 Hydrochlorothiazide 5 Carnitine 2 Taurine 1 Low Sodium diet 2 1 Peritoneocentesis 3 n - number, OB - other breeds, DO - Dobermanns, CHF - congestive heart failure, SD - sudden death, ACE-I - angiotensin converting enzyme inhibitor 47

48 Chapter 2 STATISTICAL ANALYSES The data was presented as the mean value, standard deviation and range. The two sample F test were used to detect significant differences in age of OB 1 st cohort versus DO 1 st cohort and age of DO 1 st cohort that were dead because of CHF versus DO 1 st cohort that died suddenly. Survival was analyzed in OB 1 st cohort vs. DO 1 st cohort and in DO 1 st cohort according to the mode of death (CHF vs. SD). In both analyses, Kaplan- Meier method was used for estimating survival curves, and Log Rank test was used for comparing survival between two groups of interest. The F test was used to compare echocardiographic data of Dobermanns in the 1 st cohort. The Chi - Square test was used to detect significant differences in the frequency of CHF dogs versus SD dogs within the group of Dobermanns from the 1 st cohort. P values < 0.05 were considered statistically significant. Alpine Dachsbracke 2% Kuvasz 2% Labrador Retriever 2% Giant Schnautzer 2% Irish Setter 2% Saint Bernard 2% Newfoundland 4% Bullterier 4% Borzoi 4% German Shepherd cross 2% English Cocker Spaniel 2% Dobermann 39% Rottweiler 6% Great Dane 6% Boxer 8% German Shepherd 13% Figure 1. Breed distribution of dogs with DCM in five years (n=52) 48

49 RESULTS Dilated Cardiomyopathy in the Dobermann FIRST COHORT According to the medical records, 52 dogs had a diagnosis of DCM. There were 21 Dobermanns (39 %) and 31 dogs (61 %) that belonged to other breeds (Fig. 1). Gender, age and survival of dogs from the 1 st cohort are presented in Tables 2 A & B and Fig. 2 & 3. A follow up analysis showed that all 21 Dobermanns (100%) and all dogs (100%) of other breeds were dead. Three dogs from OB group were censored in the survival analysis since they were lost for follow up. The survival time was significantly longer in dogs of other breeds (mean 240 days), versus Dobermanns (mean 52 days; Table 3 and Fig. 2). Table 2A. Gender and age in cohort 1 (all, other breeds, Dobermanns) All OB DO male (n) % 80.7% 84% 76.2% female (n) % 19.2% 16% 23.8% age (years) NS 7.2± ±1.9 mean ± STDEV (0.10) Table 2B. Gender and age in Dobermanns 1 st cohort (CHF vs. SD) CHF DO SD DO P value (CHF vs. SD) male n 11 5 % 78.6% 71.4% female n 3 2 NS (0.86) % 21.4% 28.6% age years (mean ± 7.1± ±1.3 NS (0.43) STDEV) OB - other breeds, DO - Dobermanns, CHF - congestive heart failure, SD - sudden death, n - number of dogs, STDEV - standard deviation, NS - not significant; P value stays immediately after groups that are compared; age refers to the time of the 1st examination 49

50 Chapter 2 Table 3. Survival times in days Survival time n Mean Median Range Dobermanns <1-180 Other breed dogs < CHF Dobermanns <1-180 SD Dobermanns <1-105 SD - sudden death, CHF - congestive heart failure, n - number of dogs Congestive heart failure (CHF) was the direct cause of death in seven of the 21 Dobermanns (33 %). Six dogs (29 %) were euthanised because of severe heart failure, seven Dobermanns (33 %) died suddenly and one (5 %) died due to gastric dilatation with mesenteric torsion. Autopsy was performed in ten Dobermanns with DCM, results are presented in Table 4. Table 4. Results of pathomorphologic and/or pathohistologic examination in DO 1st cohort n % 4 chamber dilatation pulmonary edema 6 75 pleural effusion only LA, LV dilatation Only right heart dilatation 2 25 pale heart muscle 2 25 mild AV valve degeneration myocardial lipidosis 4 50 myocyte degeneration 2 25 myocardial fibrosis 2 25 myocardial vacuolar degeneration dogs had pathomorphologic and pathohistologic examination done, 2 dogs had sole pathomorphologic and 2 dogs sole pathohistologic examination done; altogether 10 dogs. Severity of histologic lesions varied widely therefore gradation was not put in the table 50

51 Dilated cardiomyopathy in the Dobermann,0,8,6,4 Other breed dogs Survival,2 0,0 0 Days Other breed dogs - censored Dobermanns Dobermann - censored Figure 2. Survival curves comparing survival time in dogs of other breeds and survival time in Dobermanns. Dobermanns had significantly shorter survival time, (p(log Rank) = ) than dogs of other breeds. Mortality in Dobermanns was the highest in the first 39 days (full color figure on page 181). The group is divided into two categories: in the first one there are 13 dogs that died or were euthanised because of congestive heart failure and one dog which died due to gastric dilatation with mesenteric torsion but was also in heart failure prior to death. The second group included seven dogs that died suddenly. The results of medical history, physical examination, electrocardiography and radiography are summarized in Table 5, echocardiographic data are presented in Table 6. Four dogs that died suddenly, were not in congestive heart failure at the time of the 1 st echocardiographic examination, the reason for examination was exercise intolerance in one, syncope in two, fast heart beat and vomiting in one, examination before anesthesia in one. 51

52 Chapter 2,0,8,6,4 Survival,2 SD Dobermanns 0, CHF Dobermans Days Figure 3. Survival curves in Dobermanns in CHF vs. Dobermanns with SD. Survival in the SD group was shorter but the test was not significant, p (Log Rank) = (full color figure on page 181). SECOND COHORT This cohort consisted of 16 male and 12 female Dobermanns that were screened for signs of occult DCM. They were from one to 11 years old (mean age: 6.1 years). Results are presented in Table 5 and 6. One dog had PVC s during a 3 minute ECG recording. 16 dogs (57%) had increased LVEDD, six of these dogs (21%) had also increased LVESD and three of them (10,7%) had decreased FS. Within one year after the screening, four dogs developed DCM, three of them were in CHF and one died suddenly. Autopsy findings of the last dog were compatible with DCM. These four dogs had increased LVESD and LVEDD, three of them had also SF below 24% at the time of initial examination almost one year earlier. Among two other dogs with increased LVESD one died suddenly four years after examination and the other was euthanised for some other reasons, but no autopsy was undertaken in these two dogs. 52

53 Dilated Cardiomyopathy in the Dobermann Table 5. Clinical, electrocardiographic and radiographic findings in Dobermanns 1st cohort in congestive heart failure (CHF), sudden death (SD) and in DO 2nd cohort Variable DO 1 st cohort in CHF DO 1 st cohort /SD DO 2 nd cohort n % n % n % Medical history n=14 n=7 n=28 cough exercise intolerance difficulty breathing 5 36 not eating well vomiting 2 14 fast heart beat abdominal swelling 1 7 syncope no signs Physical findings n=14 n=6 n=28 murmur irregular pulse 6 43 premature beats tachycardia (>160 beats/min.) pulse deficit pulmonary crackles 3 21 ascites* 2 14 weak fem. pulse 3 50 no signs Electrocardiograms n= 12 n=6 n=28 sinus tachycardia atrial fibrillation 6 50 QRS complex >0.06 s (lead II) ST slurring R wave >3 mv (lead II) 6 50 P wave >0.04 s (lead II) P wave >0.4 mv (lead PVC s sinus arrhythmia 2 33 ventricular tachycardia 2 33 left axis deviation Normal Thoracic radiographs n=10 n=5 generalized LA, LV enlargement pulmonary edema ascites 1 10 pleural effusion 2 20 RV enlargement 1 20 Unremarkable findings 1 20 *ascites was obvious by balloting, later also confirmed by ultrasound, MEA - mean electrical axis, LA - left atrium, LV - left ventricle, RV - right ventricle. 53

54 Chapter 2 Table 6: Echocardiographic data of Dobermanns (DO 1 st coh.) in congestive heart failure - CHF (n=11), sudden death - SD (n=6) and of DO 2 nd cohort (n=28) DO 1st cohort. in CHF DO 1st cohort /SD * DO 2nd cohort Mean ± stdev Differe Mean ± stdev Mean ± stdev Variable n (Range) nce n (Range) n (Range) Weight 36.4± ± ± 5.4 (kg)* 11 ( ) NS 6 (33-42) 28 (28-47) LVEDD 70.5± ± ± 5.3 (mm) 11 (54-83) NS 6 ( ) 28 (40-62) LVESD 64.6± ± ± 4.5 (mm) 11 (44-79) NS 6 (35-60) 28 (28-47) FS 8.5±5.7 16± ± 5.3 (%) 11 (14-18) NS 6 (8-27) 28 (18-44) LVPWd 8.4± ± ± 1.3 (mm) 10 (7.0-11) NS 5 (5-11.5) 23 (6-11) LVPWs 10.2± ± ± 1.4 (mm) 7 ( ) NS 5 (6-14) 18 (9-14) IVSd 8.8±2.2 8± ± 1.6 (mm) 10 ( ) NS 5 (5-9.5) 25 (5-12) IVSs 11.2± ± ±2.3 (mm) 7 (6-16.0) NS 5 (8-13) 14 (8-17) LVEDD - left ventricular end-diastolic diameter, LVESD - left ventricular end-systolic diameter, FSfractional shortening, LVPWd - left ventricular posterior wall in diastole, LWPWs - left ventricular posterior wall in systole, IVSd - interventricular septum in diastole, IVSs - interventricular septum in systole, NS - not significant, P value<0.05, stdev - standard deviation; * 4 of dogs from SD group were not in CHF at the time of echocardiographic examination We obtained 38 pedigrees - 13 from DO 1 st cohort and 25 from DO 2 nd cohort. Review of the pedigrees showed that within the 1 st cohort, 9 dogs (numbers 1 to 9 in Fig. 4) were members of one line with a high incidence of heart disease and sudden death. Within the 2 nd cohort, 22 DO (78 %) have relatives in the last three generations who died suddenly or due to heart disease. In figure 4, only 15 DO 2 nd cohort (A1-A15) are presented, since the data of the ancestors from the rest of the dogs was scarce and presenting them in this pedigree would be non informative. The constructed pedigree with multiple affected members demonstrates that this particular line of Dobermanns in Slovenia was subjected to strong inbreeding, which probably enabled the diseased gene to be propagated. The exact inheritance pattern could not be determined due to incomplete data (the status of parents and siblings was frequently unknown). 54

55 Dilated Cardiomyopathy in the Dobermann A15 A A A A5 20 A7 A6 63 A A A A2 A A10 DCM cases diagnosed at Ljubljana Veterinary Faculty Possible DCM cases sudden death or incomplete diagnostic data Unaffected dogs asymptomatic (A1-A15) or died of other causes than DCM Dogs with no data Males Females Figure 4. Pedigree of a family of Dobermanns A A

56 Chapter 2 DISCUSSION In a consecutive series of 52 dogs, who were evaluated for DCM at our institution, 39 % were Dobermanns. The second most common breed affected with this disease was the German shepherd (13%), followed by the German boxer (8 %), the Great Dane (6%), the Rottweiler (6%), and others (refer to Fig. 1). Other authors reported that Dobermanns are among the three most commonly affected breeds (Calvert 1992; Kittleson 1998a; Sisson and Thomas 1995). According to the literature, the German shepherd is not so commonly affected. The high incidence in German shepherds may just be a reflection of the high popularity of this breed in Slovenia. In our study males predominated in all groups (OB, DO, DO with CHF and DO with SD). Male predominance was found also by Calvert with coworkers (Calvert 1982; Calvert et al. 1997). In contrast, O Grady and Horne (1995a) had not found a gender predisposition. The age difference of dogs of OB and DO with DCM was not significant. Dobermanns in CHF were not significantly older at the time of diagnosis than dogs that died suddenly. Calvert et al. (1997) found out that dogs dying suddenly are approximately one year younger than dogs developing CHF. An explanation for younger age in SD group of DO would be that some dogs with arrhythmias died due to arrhythmias early in the course of the disease. Congestive heart failure seems to be a more frequent outcome than sudden death in our Dobermanns, but the difference was not significant. From our pedigree review it was evident that the majority of the DO 1 st cohort and the DO 2 nd cohort belonged to one family line, which represented a limited gene pool. We found out that the incidence of DCM and SD in this family line of Dobermanns was high. The survival time of our Dobermanns with DCM 1 st cohort was significantly shorter than survival time of other breeds. It had been reported by Calvert (1992) that Dobermanns with CHF have shorter survival in comparison to other breeds. They die suddenly or as a consequence of heart failure. In our study Dobermanns in CHF did not live significantly longer (mean 62, range <1-180) than Dobermanns that died suddenly (mean 33, range <1-105). Veterinary and human studies reported the occurrence of sudden death to be between 10 and 64 % (Calvert 1992; Calvert et al. 1997; Calvert et al. 2000a; Wilson et al. 1983; Sisson and Thomas 1995; Hofmann et al. 1988). Incidence of sudden death in our Dobermanns was 33%. The data obtained by medical history, physical examination, thoracic radiographs and electrocardiogram are similar to those reported by others (Lombard 1984; Calvert 1986; Calvert 1992; Calvert et al. 1982; Calvert et al. 1997; Calvert and Brown 1986). We found greater end- 56

57 Dilated Cardiomyopathy in the Dobermann diastolic and end-systolic diameter and lower fractional shortening in Dobermanns who died due to congestive heart failure in comparison to those who died suddenly. The differences did not reach the level of statistical significance. The explanation for these differences in end-diastolic and end-systolic diameters would be that the 4 dogs that died suddenly were in an earlier stage of the disease. This finding was made also by other investigators (Calvert et al. 1997). All our postmortem results revealed lesions described as typical and most commonly present in DCM (Van Fleet and Feranns 1986; Everett et al. 1999; Calvert et al. 1997). Attenuated wavy fibers of myocardium which were described by Tidholm and others were not described pathohistologic lesions in our cases (Tidholm et al. 1998). The severity of lesions was not consistent with the diagnosis of DCM since it is known, that there are no differences in histologic lesions or lesion severity in dogs with mild or severe disease (Calvert et al. 1997; Kittleson 1998a). Only one dog in the group of 28 asymptomatic Dobermanns showed ventricular ectopy. This may be due to the fact that only a short ECG recording was obtained. This dog was already 11 years old at the time of examination and was 2 years later euthanised for some other reason. The relevance of this arrhythmia is unknown. Holter recording would probably detect significantly more arrhythmias as reported by Calvert et al (2000a). They detected PVC s in 52% of asymptomatic Dobermanns and consider this as an indicator of developing of DCM even before echocardiographic changes appear (Calvert et al. 2000a; Calvert et al. 2000b). O Grady and Horne (1995) found out that 15 DO that demonstrated at least one PVC during a short ECG recording, died. Hence, finding a PVC on an ECG recording in a dog may be suspicious of DCM and associated with a higher risk of sudden death. Increased LVESD which is a consistent echocardiograpic change in DCM, was found in 6 dogs (21%) in DO 2 nd cohort (Kittleson 1998b). Four dogs in this group developed DCM within a year, three of them were in CHF and one died suddenly. One dog among these six was euthanised for other reason and one died suddenly after four years. O Grady and Horne (1995a) performed an extensive screening on DCM in asymptomatic DO and they found out that 28.2% of DO in a group of 103 asymptomatic DO developed DCM within the period of 28.3 months. In a group of 30 asymptomatic DO, Sottiaux and Amberger (1997) found 28% of dogs with increased LVEDD and/or LVESD. Pedigree review demonstrated one line in which multiple family members were affected with DCM or sudden death, suggesting a genetic basis of the disease. Since DCM appears late in life, it was difficult to determine which dogs are unaffected. Dogs that died young and dogs in which no signs of asymptomatic DCM were found at screening could not be treated as unaffected, since the disease 57

58 Chapter 2 could become apparent later in life. This, coupled with incomplete data of several family members, prevented us from determining the mode of inheritance. In dogs, DCM has been suggested to be familial in Dobermanns, Boxers, Cocker Spaniels, Newfoundlands and Portuguese water dogs on the basis of frequent occurrence of the disease in certain lines (Calvert 1992; Goodwin et al. 1995; Tidholm et al. 2000; Dambach et al. 1999). Data about inheritance patterns in dogs are scarce; autosomal dominant inheritance was suggested in a family of Dobermanns and in English Cocker Spaniels (Hammer et al. 1996; Staaden 1981). Research regarding the dystrophin and α-cardiac actin gene had been done in Dobermanns, but no disease mutations were found so far (Meurs et al. 2001; Schatzberg et al. 1999). For human familial DCM, nine disease genes were identified to date: seven (α-cardiac actin, desmin, lamin A/C, δ-sarcoglycan, α-tropomyosin, cardiac β- myosin heavy chain and cardiac troponin T) for the most common autosomaldominant form and two (dystrophin and tafazzin) for the X-linked form (Olson et al. 1998; Li et al. 1999; Fatkin et al. 1999; Tsubata et al. 2000; Ortiz-Lopez et al. 1997; D Adamo et al. 1997; Kamisago et al. 2000; Olson et al. 2001). With the exception of tafazzin, whose function is not currently known, these genes encode for the proteins of the cell cytoskeleton, which transmit mechanical and chemical stimuli within and between cells and contribute to the cell stability (Seidman and Seidman 2001; Chen and Chien 1999; Hein et al 2000; Towbin 1998). DCM in dogs closely resembles the human form of the disease. It is expected that cytoskeletal proteins may play a significant role in the canine DCM as well and are therefore excellent candidate genes. Identification of the DCM gene in dogs would enable early specific diagnosis and detection of carriers. This could consequently reduce the incidence of DCM in predisposed breeds through selective breeding. Since DCM is common in Dobermanns, they could be an ideal breed for such a study. 58

59 REFERENCES Dilated Cardiomyopathy in the Dobermann Abelmann, W.H., Lorell, B.H The challenge of cardiomyopathy. J Am Coll Cardiol 13: Berko, A.B., Swift, M X-linked DCM. New Engl J Med 316: Bonagura, J.D., O Grady, M.R., Herring, D.S Echocardiography: principles of interpretation. Vet Clin North Am 15: Calvert, C.A., Chapman, W.L., Toal, R.L Congestive cardiomyopathy in Doberman Pinscher dogs. J Am Vet Med Assoc 181: Calvert, C.A Dilated congestive cardiomyopathy in Doberman Pinschers. The Comp Contin Edu 8: Calvert, C.A Update: Canine DCM, in Kirk, R.W., Bonagura, J.D. Eds., Current Veterinary Therapy XI WB Saunders Calvert, C., Brown, J Use of M-mode echocardiography in the diagnosis of congestive cardiomyopathy in Doberman Pinschers. J Am Vet Med Assoc 3: Calvert, C.A., Hall, G., Jacobs, G., Pickus, C Clinical and pathologic findings in Doberman Pinschers with occult cardiomyopathy that died suddenly or developed congestive heart failure: 54 cases ( ). J Am Vet Med Assoc 210: Calvert, C., Jacobs, G., Pickus, C.W., Smith, D.D. 2000a. Results of ambulatory electrocardiography in overtly healthy Doberman Pinschers with echocardiographic abnormalities. J Am Vet Med Assoc 217: Calvert, C., Jacobs, G., Smith, D.D., Rathbun, S.L., Pickus, C.W. 2000b. Association between results of ambulatory electrocardiography and development of cardiomyopathy during long-term follow-up of Doberman Pinschers. J Am Vet Med Assoc 216: Chen, J., Chien, K.R Complexity in simplicity: monogenic disorders and complex cardiomyopathies. J Clin Invest 103: Dambach, D.M., Lannon, A., Sleeper, M.M., Buchanan, J Familial DCM of Young Portuguese Water Dogs. J Vet Intern Med 13: D Adamo, P., Fassone, L., Gedeon, A., et al The X-linked gene G4.5 is responsible for different infantile dilated cardiomyopathies. Am J Hum Genet 61: Emery, A.E.H X-linked muscular dystrophy with early contractures and cardiomyopathy (Emery-Drefuss type). Clin Genetics 32: Everett, R.M., McGann, J., Wimberly, H.C., Althoff, J Dilated cardiomyopathy of Doberman Pinschers: Retrospective Histomorphologic Evaluation of Heart from 32 Cases. Vet Pathol 36: Fatkin, D., MacRae, C., Sasaki, T., et al Missense mutations in the rod domain of the lamin A/C gene as causes of DCM and conduction system disease. New Engl J Med 341: Feigenbaum, H Echocardiographic findings with altered electrical activity. In: Feigenbaum, H. Echocardiography. 5 th Ed. Lea & Febiger Goerss, J.B., Michels, V.V., Burnett, J., et al Frequency of familial DCM. Europ Heart J 16 (Suppl 0): 2-4. Gooding, L.P., Robinson, W.F., Mews, G.L Echocardiographic characterisation of dilatation cardiomyopathy in the English Cocker Spaniel. Am J Vet Res 47: Goodwin, J.K., Cattiny, G., Rouge, B Further charcterization of boxer cardiomyopathy. Proc. 13 th ACVIM FORUM Lake Buena Vista, FL. 59

60 Chapter 2 Hammer, T.A., Venta, P.J., Eyster, G.E The genetic basis of DCM in Doberman Pinschers. Animal Genetics (Abstract) 27 (Suppl 2):109. Hein, S., Kostin, S., Heling, A., Maeno, Y., Schaper, J The role of the cytoskeleton in heart failure (Review). Cardiovascular Res 45: Hofmann, T., Meinertz, T., Kasper, W., et al Mode of death in idiopathic DCM: A multivariate analysis of prognostic determinants. Am Heart J 116: Lombard, C.W Echocardiographic and clinical signs of canine DCM. J Small Anim Pract. 25: Kamisago, M., Sharma, D., DePalma, S.R., et al Mutations in sarcomere protein genes as a cause of dilated cardiomyopathy. New Engl J Med 843: Keeling, P.J., Gang, Y., Smith, G., et al Familial DCM in the United Kingdom. Br Heart J 73: Kittleson, M.D. 1998a: Primary Myocardial Disease Leading to Chronic Myocardial Failure (Dilated Cardiomyopathy and Related Diseases). In Kittleson, M.D., Kienle, R.D: Small Animal Cardiovascular Diseases. Mosby Kittleson, M.D. 1998b. Pathophysiology of heart failure. In Kittleson, M.D., Kienle, R.D: Small Animal Cardiovascular Diseases. Mosby Li, D., Tapscoft, T., Gonzalez, O., et al Desmin mutation responsible for idiopathic DCM. Circulation 100: Meurs, K.M., Magnon, A.L., Spier, A.W., et al Evaluation of the cardiac actin gene in Doberman Pinschers with dilated cardiomyopathy. Am J Vet Res 62: Michels, V.V., Driscoll, D.J., Miller, A.F Familial aggregation of idiopathic dilated cardiomyopathy. Am J Cardiol 55: Michels, V.V., Moll, P.P., Miller, F.A., et al The frequency of familial dilated cardiomyoathy in a series of patients with idiopathic DCM. New Engl J Med 326: Monnet, E., Orton, C., Salman, M., Boon, J Idiopathic DCM in dogs: Survival and prognostic indicators. J Vet Intern Med 9: O Grady, M.R., Horne, R. 1995a. Occult dilated cardiomyopathy in Doberman Pinscher. Proc. of 13 th ACVIM Forum, Lake Buena Vista, FL O Grady, M.R., Horne, R. 1995b. Echocardiographic findings in 51 normal Doberman Pinschers. ACVIM Abstracts. J Vet Intern Med 9: 202. Olson, T.M., Michels, V.V., Thibodeau, S.N., Tai, Y.S., Keating, M.T Actin mutations in DCM, a heritable form of heart failure. Science 280: Olson, T.M., Kishimoto, N.Y., Whitby, F.G., Michels, V.V Mutations that alter the surface charge of alpha-tropomyosin are associated with dilated cardiomyopathy. J Mol Cell Cardiol 33: Ortiz-Lopez, R., Li, H., Su, J., Goytia, V., Towbin, J.A Evidence for a dystrophin missense mutation as a cause of X-linked DCM. Circulation 95: Schatzberg, S., Olby, N., Steingold, S., et al A polymerase chain reaction screening strategy for the promoter of the canine dystrophin gene. Am J Vet Res 60: Seidman, J.G., Seidman, C The genetic basis for cardiomyopathy: from mutation identification to mechanistic paradigms. Cell 104: Sisson, D.D., Thomas, W.P Myocardial diseases. In Ettinger, S.J., Feldman, E.C. Eds. Textbook of Veterinary Internal Medicine, 4 th Ed. W. B. Saunders: Sottiaux, J., Amberger, C Normal Echocardiographic values in the Doberman Pinscher (French Breeding). European Society of Veterinary Cardiology Newsletter No

61 Dilated Cardiomyopathy in the Dobermann Staaden, R.V Cardiomyopathy of English Cocker Spaniels. J Am Vet Med Assoc 178: Suter, P.F., Lord, P.F Thoracic radiography: thoracic diseases of the dog and cat. Peter F Suter, Tidholm, A., Häggström, J., Jönsson, L Prevalence of attenuated wavy fibers in myocardium of dogs with dilated cardiomyopathy. J Am Vet Med Assoc 212: Tidholm, A., Häggström, J., Jönsson, L Detection of attenuated wavy fibers in the myocardium of Newfoundlands without clinical or echocardiographic evidence of heart disease. Am J Vet Res 61: Tilley, L.P Essentials of canine and feline electrocardiography. Lea & Febiger. Thomas, W.P., Gaber, C.E., Jacobs, G.J., et al Recommendations for standards in transthoracic two-dimensional echocardiography in the dog and cat. J Vet Intern Med 7: Towbin, J.A., Hejtmancik, J.F., Brink, P., et al X - linked DCM. Circulation 87: Towbin JA The role of cytoskeletal proteins in cardiomyopathies. Current Opin Cell Biology 10: Tsubata, S., Bowles, K., Vatta, M., et al Mutations in the human δ-sarcoglycan gene in familial and sporadic DCM. J Clin Invest 106: Van Fleet, J.F., Feranns, V.J Myocardial diseases in animals. Am J Pathol 124: 98. Ware, A.W Diagnostic tests for the cardiovascular system. In RW Nelson, CG Couto Eds., Essentials of Small Animal Internal Medicine, Mosby Wilson, J.R., Schwarz, J.S., Sutton, M.S.J Prognosis in severe heart failure: relations of hemodynamic measurements and ventricular ectopic activity. J Am Coll Cardiol 2: Zeviani, M., Gellera, C., Antozzi, C., et al Maternally inherited myopathy and cardiomyopathy: association with mutation in mitochondial DNA trna Leu(UUR). Lancet 338: ACKNOWLEDGMENTS Authors are grateful to Prof. dr. Janez Stare, Institute of Biomedical Informatics of the Faculty of Medicine, University of Ljubljana, and to Prof. dr. Peter Lazar, Veterinary faculty, University of Ljubljana, for helping us with statistical analysis. 61

62

63 Chapter 3 Characterization of the canine desmin (DES) gene and evaluation as a candidate gene for dilated cardiomyopathy in the Dobermann dog Polona Stabej, Sandra Imholz, Serge A. Versteeg, Carla Zijlstra, Arnold A. Stokhof, Aleksandra Domanjko-Petrič, Peter A.J. Leegwater, Bernard A. van Oost Gene 2004; 13:

64 Chapter 3 ABSTRACT Canine dilated cardiomyopathy (DCM) in dogs is a disease of the myocardium associated with dilatation and impaired contraction of the ventricles and is suspected to have a genetic cause. A missense mutation in the desmin gene (DES) causes DCM in a human family. Human DCM closely resembles the canine disease. In the present study, we evaluated whether DES mutations are responsible for DCM in Dobermann dogs. We have isolated Bacterial Artificial Chromosome clones (BACs) containing canine DES and determined the chromosomal location by fluorescence in situ hybrizidation (FISH). Using data deposited in the NCBI trace archive and GenBank, canine DES DNA sequence was assembled and seven single nucleotide polymorphisms (SNPs) were identified. A polymorphic marker was isolated from the BAC clones that contained canine DES. The microsatellite marker and four informative desmin SNPs were typed in a Dobermann family with frequent DCM occurrence. The disease phenotype was not associated with a desmin haplotype. We concluded that mutations in DES do not play a role in Dobermann DCM. Availability of the microsatellite marker, SNPs and DNA sequence reported in this study enable fast evaluation of DES as a DCM candidate gene in other dog breeds with DCM occurrence 64

65 INTRODUCTION Characterization of the canine DES Dilated cardiomyopathy (DCM) is a myocardial disease representing an important cause of congestive heart failure (CHF) and sudden death in dogs. It primarily affects large and giant breed dogs with Dobermanns being one of the most frequently affected (Domanjko-Petrič et al. 2002; Sisson and Thomas 1995). Data regarding the mode of inheritance in the Dobermanns is conflicting. Both autosomal dominant and autosomal recessive modes of transmission have been suggested (Hammer et al. 1996; Meurs 2002). The high prevalence of DCM in specific breeds suggests a genetic background, but causal mutations have not yet been identified. Within each breed, DCM has unique characteristics and between breeds, it is probably a genetically heterogeneous disease. DCM in the Dobermann is characterised by severe left ventricular dilatation, systolic dysfunction and the so-called occult (preclinical) phase during which ventricular and atrial premature contractions occur. Dobermanns either die suddenly or develop congestive heart failure (Calvert et al. 1997). Human DCM, which closely resembles the canine form of the disease, is a genetically heterogeneous disease with 13 disease genes identified (Knoll et al. 2002; Schmitt et al. 2003; for review see Fatkin and Graham 2002). Most of these genes code for proteins of the cell cytoskeleton and the mechanism by which mutations cause DCM is thought to be impairment of heart force production or transmission (Towbin and Bowles 2000). One of the identified disease genes is the gene encoding desmin (DES). A missense mutation (Ile451Met) was reported in a family with autosomal dominant DCM without conduction-system disease or skeletal myopathy (Li D. et al. 1999). Moreover, DES knockout mice exhibit disruption of myofibril organization with ventricular dilatation and impaired systolic contraction (Milner et al. 1999). These findings suggest that a reduction or absence of the DES function can result in DCM. The human DES has been assigned to human chromosome 2q35 and has nine exons that encode a protein of 469 amino acids (Vicart et al. 1996). In this study, we report on the isolation and characterisation of canine BAC clones containing the desmin gene, on the chromosomal location of these clones and on the identification of a microsatellite marker residing in the vicinity of the DES. By searching Canis familiaris sequences deposited in the NCBI GenBank trace archive and the 1.5x dog sequence (Kirkness et al. 2003), we determined a partial DNA sequence of the canine DES and identified seven DES SNPs. To evaluate whether the DES is involved in the Dobermann DCM, we analyzed the association of the DES haplotypes with the DCM phenotype in a family of Dobermanns with frequent DCM occurrence. 65

66 Chapter 3 MATERIALS AND METHODS DOGS INCLUDED IN THE STUDY AND DNA ISOLATION Blood samples were collected from 18 Dobermanns diagnosed with DCM and 10 DCM-free Dobermanns. Dogs were patients of the Department of Clinical Sciences of Companion Animals (Veterinary Faculty, Utrecht University) and of the Clinic for Small Animals and Surgery (Veterinary Faculty, Ljubljana University). The pedigrees of 10 affected and the 10 unaffected Dobermanns included in this study were collected and the familial relationships between these dogs were established. Some relationships were traced with the help of a Dobermann pedigree web page at The health status of relatives of investigated Dobermanns was obtained from the owners and breeders. Blood samples were collected and genomic DNA was isolated using a salt extraction method (Miller et al. 1988). DCM diagnosis was based on clinical and/or radiographic symptoms of congestive heart failure (CHF) and echocardiographic evidence of shortening fraction (FS) <25% in the absence of other CHF related lesions on two-dimensional echocardiography. Echocardiography was performed in conscious dogs, with the dog in right lateral recumbency, using a high definition ultrasound system (HDI 3000, Advanced Technology laboratories, Woerden, The Netherlands) equipped with a 5-3 MHz broad and phase array transducer. ECG electrodes were placed on the left and right foreleg and the left hind leg for simultaneous ECG recording. All measurements were performed using a trackball driven cursor and ultrasound software. From the right parasternal approach 2D guided M-mode tracings were made for measurements of the interventricular septum, the left ventricular dimension, and the left ventricular free wall, in diastole and systole, and of the aortic root and left atrial appendage diameter. These measurements were made from the leading edge of the first endocardial surface to the leading edge of the second endocardial surface. Diastolic measurements were made at the onset of the QRS complex of the ECG and systolic measurements were made at the maximum systolic excursion of the interventricular septum. The diameter of the aortic root was measured at the onset of the QRS complex of the ECG and the dimension of the left atrial appendage was measured at its maximal upward excursion near the end of systole. From the left ventricular dimension in diastole (LVDd) and systole (LVDs) the fractional shortening (FS) was calculated (FS (%) = [(LVDd LVDs)/LVDd] x 100) and from the diameter of the left atrial appendage and the aortic root the left atrium to aorta ratio (LA/Ao) was calculated. At the Clinic for Small Animals and Surgery (Veterinary Faculty, Ljubljana University), echocardiography was 66

67 Characterization of the canine DES ISOLATION OF THE CANINE DESMIN BAC CLONES In order to screen the canine BAC library (Li R. et al. 1999), two overlapping primers were designed in exon 1 of the human DES (GenBank accession no. U59167) using Overgo maker program ( Des_Ova: 5 - ATGAGCCAGGCCTACTCGTCCAGC-3, Des_Ovb: 5 - AGGAGGACTGGCTGGACG-3. Bold characters indicate the overlap. The 40bp probe was synthesized and labeled with an [α 32 P]-dATP and an [α 32 P]- dctp at 37 C for one hour using the overgo technique (Han et al. 2000). The hybridization of the canine BAC library and BAC DNA isolation were performed as described on the BacPac web-site ( To confirm the identity of the BAC clones, BAC DNA and canine genomic DNA were digested with EcoRI restriction enzyme, separated on a 0.7 % agarose gel and Southern blotted onto Hybond N + filters (Amersham). The blots were hybridized overnight at 65 C with an [α 32 P]-dATP labeled insert of the human IMAGE DES cdna clone (Lennon et al. 1996). The clone was obtained from the Resource Centre of the German Human Genome project (Berlin, Germany) and verified through DNA sequence analysis. The IMAGE clone insert was excised and purified on agarose gel (Quiaquick spin columns, Quiagen). The blots were washed with 2x SSC, 0.1% SDS at 65 C for 10 minutes. Genomic DNA blots were exposed to film at -70 C with intensifying screen overnight and BAC DNA for 5 minutes. CHROMOSOMAL LOCALIZATION OF THE CANINE DESMIN GENE BY FISH Metaphase chromosomes were prepared from concanavalin A stimulated peripheral blood lymphocytes from karyotypically normal dogs. Total DNA of a canine BAC clone 25J10, which contains the DES, was isolated and labeled with biotin-16-dutp or digoxigenin-11-dutp by nick-translation. For the identification of small dog chromosomes, biotinylated chromosome-specific paints Z, cc, and dd, developed by the Animal Health Trust and the Sanger Centre, were used (Breen et al. 1999a). Labeled BAC DNA was hybridized onto un-banded or GTG-banded canine metaphase chromosomes as previously described (Zijlstra et al. 1997). In these experiments a final probe concentration of 67

68 Chapter 3 5 ng/µl and a 50- or 100-fold excess of fragmentated total dog DNA were used. Posthybridization stringency washes were performed in 50% formamide/2x SSC at 42 C. Biotinylated probe hybridization was detected using avidin-fitc, and two additional layers of biotinylated goat-anti-avidin and avidin-fitc. Chromosomes were counterstained with propidium iodide. Hybridization of digoxigenin labeled probe was detected with mouse anti-dig (1:100, Sigma), followed by a layer of rabbit anti-mouse-tritc (1:100, Sigma) and a layer of goat anti-rabbit-tritc (1:100, Sigma). In these experiments, chromosomes were counterstained with DAPI. Dual-color FISH experiments were performed with the digoxigenin labeled DES probe and the biotinylated chromosome-specific paints Z, cc and dd. In these experiments, for half a slide, 75 ng digoxigenated DES probe was precipitated and dissolved in 15 µl hybridization buffer containing biotinylated chromosomespecific paint and competitor DNA (Langford et al. 1995). This probe/paint mixture was denatured for 7 minutes at 100 C, preannealed for 1½ hour at 37 C, and applied to denatured slides. Hybridization was carried out overnight at 37 C, and posthybridization washes were performed as described above. Specific sites of hybridization were simultaneously detected by using successive layers of avidin-fitc/mouse anti-dig, biotinylated goat-anti-avidin/rabbit anti-mouse- TRITC and avidin-fitc/goat anti-rabbit-tritc. Chromosomes were counterstained with DAPI. Metaphases showing sites of specific hybridization were captured using a Leica DMRA microscope equipped with the GENUS Image Analysis software of Applied Imaging. MICROSATELLITE MARKER ISOLATION IN DES BAC CLONES DNA from the BAC clones 42C1 and 216E1 was pooled and digested with Sau3AI. Adaptors were ligated to the ends of the restriction fragments and the restriction fragments were amplified using adaptor specific primers (Groot and van Oost, 1998). The amplified product was enriched for CA-repeats using 3 biotinylated [CA] 22 oligonucleotide. The enriched fragments were amplified and subcloned into the BamHI site of pzero-1 vector (Invitrogen) after which TOP10F competent cells were transformed according to the manufacturer s protocol (Invitrogen). Positive clones were identified by an α- 32 P-dATP-labelled [CA] 22 probe. The sequence of the clones was determined using BigDye Terminator cycle sequencing (Applied Biosystems) with M13-forward (5 GTTTTCCCAGTCACGAC-3 ) and M13-reverse (5 -CAGGAAACAGCTAT 68

69 Characterization of the canine DES GAC-3 ) primers. When the microsatellite marker UU42C1 was observed, flanking primers were designed using Primer 3 software (Rozen and Skaletsky, 2000). Primer sequences were: UU42C1-f: 5 - GAAGAAGCAAGCTGGTGGAC -3 and UU42C1-r: 5 - GGCTTTCCATAGCAACTCCA-3. For each PCR reaction, 25 ng of canine genomic DNA was amplified with 0.33 µm of each primer, 0.6 U of Platinum Taq (Invitrogen), 200 µm dntps, 1x Gibco buffer and 3.5 mm MgCl 2 in a final volume of 15 µl. The PCR program consisted of a denaturation step of 10 min at 94 C, followed by 35 cycles of 30 sec 94 C, 30 sec 62 C, 30 sec 72 C and a final extension at 72 C for 10 min. CANINE DES MICROSATELLITE MARKER TYPING IN THE DOBERMANNS The forward primer of the UU42C1 marker was labelled with HEX fluorescent dye (Eurogentec). PCR products were generated as described in the previous paragraph and run with GS500 size standards on an automated ABI3100 DNA Analyzer (Applied Biosystems, Foster City, CA). Genescan 3.1 software was used for genotype assessment. UU42C1 microsatellite marker was typed in 10 DCM-free Dobermanns and 18 Dobermanns diagnosed with DCM. CANINE DES DNA AND AMINO ACID SEQUENCE The canine DES sequence was derived from sequences deposited in Canis familiaris trace archive, DBGSS and DBWGS of the NCBI GenBank, by BLAST search ( using the human DES cdna sequence (GenBank accession no. NM_001927) as a probe. Canine DES sequences were imported as trace (.scf) or Editseq (.seq) files into Seqman (DNA Star alignment software) and aligned with one another and all nine human DES exons. Poor quality sequences were eliminated. Canine DES DNA sequence, intron-exon borders and the coding region were determined. The 1407 bp canine DES coding sequence was compared to the human (GenBank accession no. NM_001927) and the mouse (GenBank accession no. NM_010043) coding region. From the coding region, canine DES amino acid sequence was deduced and compared to its human (GenBank accession no. NP_001918) and mouse (GenBank accession no. NP_034173) counterparts using Blast2 sequences program 69

70 Chapter 3 Table 1. Canine DES SNPs SNP Positi Type of Prod. Name on c Change Forward_Primer_5'- 3' Reverse_Primer_5'- 3' length SNP_Ex3 a 1808 T/C TTGCTTGACCACTACCAGGA AGATGTTCTTAGCCGCGATG 402 bp SNP_In3 a 1851 G/C TTGCTTGACCACTACCAGGA AGATGTTCTTAGCCGCGATG 402 bp SNP_In4 a 2203 A/G CCAGCTTCAGGAACAACAGG TGATGTGATGAGAGCCAAGG 497 bp SNP_In4-1 a 2249 T/C CCAGCTTCAGGAACAACAGG TGATGTGATGAGAGCCAAGG 497 bp SNP_In6 a 3249 A/G CCTGCTCAATGTCAAGATGG CCTTGGGTACGAGTCTCTGC 498 bp SNP_In6-1 b 3423 N CCTGCTCAATGTCAAGATGG CCTTGGGTACGAGTCTCTGC 498 bp SNP_In7 a 4613 A/C AACTTCCGAGGTGAGTGCAT GAGGGTGCCTTGTAGCTCAG 376 bp a SNPs identified in the boxer sequences deposited in the NCBI Canis familiaris trace archive ( b SNP from the 1.5x standard poodle sequence SNP database: (Kirkness et al. 2003) c Position according to GenBank accession no. BK In order to detect SNPs in the assembled canine DES sequence, Canis familiaris trace archive (NCBI GenBank) sequences were aligned using Seqman (DNA Star alignment software). Six SNPs were located (Table 1). The SNP database of the 1.5x sequence: (Kirkness et al. 2003) was downloaded and searched for SNPs in the canine DES (GenBank accession no. BK005142) by local BLAST search. DES SNP TYPING IN THE DOBERMANNS To asses the presence of the DES SNPs in 18 DCM and 10 DCM-free Dobermanns, primers for SNP testing were designed using Primer 3 software (Rozen and Skaletsky, 2000). Primer sequences and PCR product lengths are given in Table 1. All primers were M13 tailed. For each PCR reaction, 25 ng genomic DNA was used as a template in a 15 µl reaction mixture consisting of 1x Gibco-BRL buffer (Life Technologies), 200 µm dntps, 1.5 mm MgCl 2, 0.6 U Platinum Taq polymerase (Invitrogen) and 0.33 µm of each primer. The PCR program consisted of denaturation for 4 min at 94 C, followed by 35 cycles of 30 s 94 C, 30 s 57 C, 30 s 72 C and a final extension for 10 min at 72 C. The reaction was diluted 15x and 1µl was used in a 10 µl tercycle reaction using 1µl Big Dye Terminator Ready Reaction Kit (Perkin Elmer ABI), 0.32 µm of the HPLC purified M13 forward primer (5 -GTTTTCCCAGTCACGAC-3 ) in 1 x sequence buffer (80 mm Tris, 2 mm MgCl 2, ph 9.0). The tercycle consisted of 25 cycles of 30 s at 96 C, 15 s at 55 C, 2 min at 60 C. Tercycle products were 70

71 Characterization of the canine DES purified using multiscreen 96-well filtration plates (Millipore) with Sephadex G- 50 (Amersham). Obtained sequences were aligned using Seqman (DNA Star Software) and SNPs were typed by visual examination. DEFINITION OF HAPLOTYPES AND ASSOCIATION ANALYSIS Haplotypes were constructed for the canine UU42C1 microsatellite marker and four DES SNP genotypes that were informative in the Dobermanns. Firstly, all dogs that were homozygous for all five markers were selected which defined two haplotypes. These two haplotypes were the most frequent haplotypes in all three groups. Additional haplotypes were derived. Deduced haplotypes were analyzed for the association with the DCM phenotype in the Dobermann family tree. RESULTS SELECTION AND CONFIRMATION OF CANINE DES BAC CLONES Screening of the canine genomic BAC library RP81 with the DES exon 1 probe yielded six positive BAC clones: 18L16, 25J10, 42C1, 216E1, 233D5 and 233F5. The EcoRI restriction fragment pattern of the BAC clones 25J10, 42C1 and 216E1 are similar and these are grouped. The BAC clones 18L16, 233D5 and 233F5 form a distinct group (Fig. 1). Southern blots with EcoRI restriction enzyme digests of DES BAC clones DNA and total dog genomic DNA were hybridised with a complete human desmin cdna probe. The probe detected a DNA fragment of the same length in the genomic DNA and in BAC clones 25J10, 42C1 and 216E1 (Fig. 2). These three BAC clones tested all positive for the DES sequences amplified with SNP_Ex3 (part of exon 3 and intron 3 are amplified for this SNP) and SNP_In7 (part of intron 7 is amplified for this SNP). The similar BAC clones 18L16, 233D5 and 233F5, did not hybridise to the DES cdna probe and were not studied further. We concluded that BAC clones 25J10, 42C1 and 216E1 contain the canine DES. 71

72 Chapter 3 bp M Figure 1. EcoRI restriction fragment patters of BAC clones isolated with a desmin overgo probe. M: Marker lane ng of λ DNA digested with Hind III; Lanes 1-6: 1g of BAC DNA digested with EcoRI. Lane 1 BAC 18L16, lane 2 BAC 25J10, lane 3 BAC 42C1, lane 4 BAC 216E1, lane 5 BAC 233D5, lane 6 BAC 233F5. Samples were run on 0.7 % agarose gel at 20V for 15 hours bp M G M Figure 2. Autoradiograms showing Southern blot analysis of dog genomic DNA (lane G) and DES BAC clone DNA EcoRI digests hybridised with a DES IMAGE cdna clone (lane 1 BAC 18L16, lane 2 BAC 25J10, lane 3 BAC 42C1, lane 4 BAC 216E1, lane 5 BAC 233D5, lane 6 BAC 233F5). The genomic DNA and BAC DNA digests were run on the same gel. 72

73 Characterization of the canine DES CHROMOSOMAL LOCALIZATION OF THE CANINE DESMIN GENE BY FISH BAC clone 25J10 was used as a probe for FISH analysis. After hybridization of the digoxigenin or biotin labeled DES probe to unbanded or GTG-banded chromosomes, clear fluorescent signals were found on the telomeric end of a small chromosome pair (Fig. 3a). Following the chromosome numbering suggested by Reimann et al. (1996), this pair could be pair 31, 33 or 37 on the basis of its G-banding pattern (insert Fig. 3a). Since the small chromosome pairs in the dog are very difficult to reliably identify by G-banding alone, we performed dual-color FISH with the DES probe and painting probes specific for CFA31+34 (paint Z), CFA33+36 (paint cc) or CFA37 (paint dd). The results of these experiments clearly show that the hybridization site of the desmin probe does not coincide with the hybridization signals of paint Z (Fig. 3b) or cc (Fig. 3c), but does with paint dd (Fig. 3d). Therefore, the DES can be assigned to CFA37qter. Figure 3. Chromosomal localization of the canine DES by FISH (full color figure on page 183). a. Partial metaphase spread showing hybridization signals (arrowheads) obtained after hybridization with the biotinylated DES probe on unbanded chromosomes, insert: G-banded chromosome before (left) and after (right) FISH, b, c, and d. Metaphase spreads after dual-hybridization with digoxigenated DES probe and biotinylated paint Z (b), paint cc (c), or paint dd (d). Arrowheads indicate hybridization signals obtained with the DES probe, and arrows indicate hybridization signals obtained with the paints. 73

74 Chapter 3 CANINE DESMIN POLYMORPHIC MARKER ISOLATION AND ANALYSIS A (CA) 16 microsatellite marker (UU42C1) was isolated from the pooled DNA of the canine DES BAC clones 42C1 and 216E1. A PCR specific for UU42C1 with individual BAC DNA as template showed that the microsatellite is located on BAC clones 25J10, 42C1 and 216E1 (Data not shown). A 5153bp contig (UU42C1 contig) surrounding the UU42C1 marker region was constructed using canine DNA sequences deposited in NCBI trace archives and GenBank. The following sequences from NCBI Genbank were used to make the UU42C1 contig: TI (trace archive), AACN and AACN BLAST analysis of the UU42C1 contig sequence identified several regions of high similarity to the human genomic contig NT_ on HSA2, about 9.5 kb downstream from the human DES. The human region that is similar to the dog DNA sequence around the UU42C1 marker does not contain any known coding or regulatory elements. No significant similarities to the sequences of other species were identified. The microsatellite marker UU42C1 was typed in 18 DCM and 10 DCM-free Dobermanns. The marker exhibited four allele sizes: 208, 213, 214 and 216 bp and was thus polymorphic (Table 2). Table 2. Canine DES SNPs and UU42C1 marker genotypes with derived haplotypes in the Dobermanns. SNP_In3 SNP_In4-1 SNP_In6 SNP_In7 Haplotype 1851 a 2249 a 3249 a 4613 a UU42C1 b 1 G T G A G T G A C C A C C C A C 208 a Position of the SNP in the BK sequence b The UU42C1 marker position was estimated to be 9.5 kb downstream DES 74

75 Characterization of the canine DES CANIS FAMILIARIS DESMIN GENE SEQUENCE We used data from dog genome projects to reconstruct the 9 exons and 8 introns of the DES. The 5 and 3 untranslated regions were not determined (Fig. 4). The DNA Sequence was deposited in the NCBI GenBank under accession no. BK The structure of the canine DES is very similar to the human DES, with introns 6 (1293 bp in the dog) and 7 (1683 bp in the dog) being the largest in both species (Fig. 4). The canine DES coding sequence is 1407 bp long and shows 92% and 90% similarity with its human and mouse counterparts, respectively. The deduced canine DES amino acid sequence (469 aa) shows a 96% similarity to the human and mouse DES amino acid sequence. CANINE DESMIN SNPS ANALYSIS Seven SNPs were identified in the canine DES sequence: two in intron 4 and intron 6 and one each in exon 3, intron 3 and intron 7 (Table 1). The codon in exon 3 with the T/C SNP codes for valine in both SNP variants. The 6 canine DES SNPs were assessed in 18 DCM and 10 DCM-free Dobermanns. SNPs in intron 3 (position 1851), intron 4 (position 2249), intron 6 (position 3249) and intron 7 (position 4613) were polymorphic in this selection of Dobermanns (Table 2). Figure 4. Canine DES genomic organization. The blocks represent the 9 exons; dashed lines not drawn to scale represent the 8 introns of indicated length. The 5 untranslated region in exon 1 and 3 untranslated region of exon 9 were not analysed (grey boxes). 75

76 Chapter 3 DESMIN GENE HAPLOTYPES AND THEIR ASSOCIATION WITH THE DCM IN THE DOBERMANN PEDIGREE First, haplotypes were deduced from Dobermanns that were homozygous for all five markers. Two different haplotypes were found: haplotypes 1 and 3 (Table 2). Haplotypes 2 and 4 were identified in dogs homozygous for the four DES SNPs and heterozygous for the UU42C1 marker. Haplotype 2 differs from haplotype 1 in the microsatellite and haplotype 4 differs from haplotype 3 also in the microsatellite only. The haplotypes in homozygous dogs exist definitely and haplotypes derived from heterozygous dogs exist most likely, since all the Dobermanns are closely related (Fig. 5). Eight DCM Dobermanns that are not presented in Fig. 5 had the following haplotypes: [1,3] three dogs, [1,4] two dogs, [3,3] two dogs and [2,3] one dog. The actual mode of inheritance could not be determined due to a lack of information regarding the relatives of the analysed dogs (Fig. 5). None of the four identified haplotypes showed association with the DCM phenotype under autosomal dominant or recessive mode of inheritance. 76

77 Characterization of the canine DES DCM confirmed by echocardiography Probable DCM cases DCM free dogs Sudden death Status unknown Figure 5. Pedigree showing the relationship between the analysed Dobermanns; squares - males, circles - females. The haplotypes of the desmin markers are given below the symbols (Table 2). 77

78 Chapter 3 DISCUSSION CHROMOSOMAL LOCALIZATION OF THE CANINE DESMIN GENE In the present study, we localized the canine DES to CFA37qter by FISH analysis. In the human genome, DES is located on HSA2q35. Independent comparative painting studies between the human and dog genome indicate that HSA2q35 shares homology with either CFA33 (Yang et al. 1999) or CFA37 (Breen et al 1999b). The discrepancy between these studies is most likely caused by difficulties in the unequivocal identification of the small canine chromosomes. In the present study, this problem was overcome by the use of specific chromosome paints that were assigned to individual dog chromosomes by the Committee for the Standardization of the Dog Karyotype (Breen et al. 1999a). At present, eight of the genes that have been mapped in the human chromosome region 2q33-2q37 have also been mapped in the dog. These genes are, from proximal to distal CD28, CTLA4, FN1, SCLC11A1, COL4A4, COL4A3, PDE6D and ALPI (Table 3). The latter four all map to CFA25, whereas the former map to CFA37. In the human genome, DES is located between SLC11A1 and COL4A4. The present localization of DES on CFA37, indicates that, compared to the human genome, there is a breakpoint between DES and COL4A4 in the canine genome. Table 3. Position of the desmin and its neighbouring genes on the human (HSA) and dog (CFA) chromosomes Position Gene Position Dog Gene Name Symbol HSA a HumanMb a CFA b TSP units b CD28 antigen CD28 2q Cytotoxic T-lymphocyteassociated protein 4 Fibronectin CTLA4 FN1 2q33 2q Solute carrier family 11 SLC11A1 2q Desmin DES 2q c not known Collagen, type IV, alpha 4 COL4A4 2q35-q not relevant Collagen, type IV, alpha 3 COL4A3 2q36-q not relevant Phosphodiesterase 6D, c-gmp specific, rod,delta Alkaline phosphatase, intestinal PDE6D ALPI 2q35-q36 2q not relevant not relevant a location obtained from LocusLink ( ) b location obtained from Guyon et al c present study 78

79 Characterization of the canine DES CANINE DESMIN GENE SEQUENCE Resources recently available from the canine genome project are the 7x redundant dog sequence deposited in the NCBI GenBank trace archive and recently published 1.5x redundant dog genome sequence (Kirkness et al. 2003). These resources enable rapid reconstruction and evaluation of the genes of interest. They have also radically improved the ability to perform comparative mapping and consequent identification of novel genes and regulatory elements. Combination of the NCBI GenBank Canis familiaris trace archive sequences and the published 1.5x dog sequence (Kirkness et al. 2003) provided the canine DES sequence. The DES amino acid sequence is strongly conserved between different species, therefore the observed high degree of similarity between dog and human or mouse was expected (Loh et al. 2000). Other useful information derived from the dog genome project concern the SNPs. For almost the complete DES DNA sequence, there were at least two overlapping trace files available. Discrepancies between the DNA sequence files indicated possible SNPs. We could distinguish differences due to poor quality data from true SNPs because the electropherograms are deposited with the trace files. Interestingly, we found 6 SNPs in the 6.4 kb long DES sequence from a single boxer, while this animal was selected to be sequenced because it has the lowest level of variation in its genome ( The use of this data mining is illustrated by the fact that we found 4 out of the 6 SNPs to be informative in a pedigree of Dobermanns. Only one SNP was found in the poodle SNP database (Kirkness et al. 2003). This could be due to lower coverage (1.5x) of the standard poodle sequence in comparison to the 7x redundant boxer sequence. ISOLATION AND LOCATION OF THE UU42C1 MICROSATELLITE MARKER We isolated a polymorphic microsatellite marker UU42C1 from the canine DES BAC clones. The sequence of this marker could not be found in the contig of the canine DES DNA sequence. It was an interesting finding that parts of a 5153 bp contig sequence surrounding the UU42C1 marker had high homology with segments of human DNA sequence at a 9.5 kb distance from the human DES because we did not find any similarities of these segments in the mouse genome. This is not surprising, because the similarity between coding and intergenic regions of dog and human genomes is greater than the similarity between the mouse and human or dog. The intergenic conserved sequence blocks we found have not yet been associated with any gene and may contain regulatory elements 79

80 Chapter 3 important for transcription. Between the dog, mouse and human, 45% of the conserved DNA sequence blocks are not associated with any genes and comparative sequence analysis has proven to be a very efficient way for identifying the functional elements encoded within the DNA (Chen et al. 2001; O Brien and Murphy, 2003). DOBERMANN FAMILY AND LINKAGE ANALYSIS OF THE DES MARKERS Several canine genetic diseases have been mapped with the use of microsatellite markers (Acland et al. 1999; Jonasdottir et al. 2000). The newly identified UU42C1 marker and SNPs enabled us a fast evaluation of the DES as a candidate gene for DCM in the Dobermanns. In the Dobermann family, the DES region was polymorphic with four haplotypes, but none of the haplotypes was associated with the DCM phenotype. For example, DCM dogs nos and 142 do not have haplotype 3 that is shared by other DCM dogs. Furthermore, two closely related DCM dogs: nos. 142 (haplotypes 1,1) and 2172 (haplotypes 3,3), do not share a single haplotype. Moreover, in the 8 singleton Dobermann dogs with DCM, 6 dogs were heterozygous and none of the haplotypes was present in all affected singleton dogs. It should be stressed that haplotypes 1 and 3 differ in the four SNPs of the haplotype and represent alleles that must have diverged long ago in the ancestral lineage. Therefore, we have excluded the DES as a gene responsible for DCM in the Dobermanns under the autosomal dominant and recessive models with a high level of confidence. However, the DES sequence, SNPs and the microsatellite marker reported in this study are tools that will enable fast evaluation of the desmin gene in breeds other than the Dobermann. 80

81 REFERENCES Characterization of the canine DES Acland, G.M., Ray, K., Mellersh, C.S., Langston, A.A., Rine, J., Ostrander, E.A., Aguirre, G.D A novel retinal degeneration locus identified by linkage and comparative mapping of canine early retinal degeneration. Genomics 59: Breen, M., Bullerdiek, J., Langford, C.F. 1999a. The DAPI banded karyotype of the domestic dog (Canis familiaris) generated using chromosome-specific paint probes. Chromosome Res. 7: Breen, M., Thomas, R., Binns, M.M., Carter, N.P., Langford, C.F. 1999b. Reciprocal chromosome painting reveals detailed regions of conserved synteny between the karyotypes of the domestic dog (Canis familiaris) and human. Genomics 61: Calvert, C.A., Hall, G., Jacobs, G., Pickus, C Clinical and pathologic findings in Dobermanns with occult cardiomyopathy that died suddenly or developed congestive heart failure: 54 cases ( ). J Am Vet Med Assoc 210: Chen, R., Bouck, J.B., Weinstock, G.M., Gibbs, R.A., Comparing vertebrate wholegenome shotgun reads to the human genome. Genome Res. 11: Domanjko-Petrič, A., Stabej, P., Žemva, A Dilated Cardiomyopathy in Dobermanns, Survival, Causes of Death and Pedigree Review in a Related Line. J. Vet. Cardiol. 4: Fatkin, D., Graham, R.M Molecular mechanisms of inherited cardiomyopathies. Physiol. Rev. 82: Groot, P.C. and van Oost, B.A Identification of fragments of human transcripts from a defined chromosomal region: representational difference analysis of somatic cell hybrids. Nucleic Acids Res. 26: Guyon, R., Lorentzen, T.D., Hitte, C., Kim, L., Cadieu, E., Parker, H.G., Quignon, P., Lowe, J.K., Renier, C., Gelfenbeyn, B., et al A 1-Mb resolution radiation hybrid map of the canine genome. Proc. Natl. Acad. Sci. USA 100: Hammer, T.A., Venta, P.J., Eyster, G.E The genetic basis of dilated cardiomyopathy in Dobermanns. Animal Genetics 27 (Suppl. 2), 109. Han, C.S., Sutherland, R.D., Jewett, P.B., Campbell, M.L., Meincke, L.J., Tesmer, J.G., Mundt, M.O., Fawcett, J.J., Kim, U.J., Deaven, L.L., Doggett, N.A., Construction of a BAC contig map of chromosome 16q by two-dimensional overgo hybridization. Genome Res. 10: Jonasdottir, T.J., Mellersh, C.S., Moe, L., Heggebo, R., Gamlem, H., Ostrander, E.A., Lingaas, F Genetic mapping of a naturally occurring hereditary renal cancer syndrome in dogs. Proc. Natl. Acad. Sci. USA 97: Kirkness, E.F., Bafna, V., Halpern, A.L., Levy, S., Remington, K., Rusch, D.B., Delcher, A.L., Pop, M., Wang, W., Fraser, C.M., Venter, J.C The dog genome: survey sequencing and comparative analysis. Science 301: Knoll, R., Hoshijima, M., Hoffman, H.M., Person, V., Lorenzen-Schmidt, I., Bang, M.L., Hayashi, T., Shiga N., Yasukawa, H., Schaper, W., McKenna, W., Yokoyama, M., Schork, N.J., Omens, J.H., McCulloch, A.D., Kimura, A., Gregorio, C.C., Poller, W., Schape, J., Schultheiss, H.P., Chien, K.R The cardiac mechanical stretch sensor machinery involves a Z disc complex that is defective in a subset of human dilated cardiomyopathy. Cell 111: Langford, C.F., Fischer, P.E., Binns, M.M., Holmes, N.G., Carter, N.P Chromosome-specific paints from a high-resolution flow karyotype of the dog. Chromosome Res. 3:

82 Chapter 3 Lennon, G.G., Auffray, C., Polymeropoulos, M., and Soares, M.B The I.M.A.G.E. Consortium: An Integrated Molecular Analysis of Genomes and their Expression. Genomics 33: Li, D., Tapscoft, T., Gonzalez, O., Burch, P.E., Quinones, M.A., Zoghbi, W.A., Hill, R., Bachinski, L.L., Mann, D.L., Roberts, R Desmin mutation responsible for idiopathic dilated cardiomyopathy. Circulation 100: Li, R., Mignot, E., Faraco, J., Kadotani, H., Cantanese, J., Zhao, B., Lin, X., Hinton, L., Ostrander, E.A., Patterson, D.F., de Jong, P.J Construction and characterization of an eightfold redundant dog genomic bacterial artificial chromosome library. Genomics 58: Loh, S.H., Chan, W.T., Gong, Z., Lim, T.M., Chua, K.L Characterization of a zebrafish (Danio rerio) desmin cdna: an early molecular marker of myogenesis. Differentiation 65: Miller, S.A., Dykes, D.D., Polesky, H.F A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16: Milner, D.J., Taffet, G.E., Wang, X., Pham, T., Tamura, T., Hartley, C., Gerdes, A.M., Capetanaki, Y The absence of desmin leads to cardiomyocyte hypertrophy and cardiac dilation with compromised systolic function. J. Mol. Cell. Cardiol. 31: Meurs, K.M Familial Canine Dilated Cardiomyopathy. In Proceedings. Advances in Canine and Feline Genomics, St. Louis, Missouri. O'Brien, S.J., Murphy, W.J Genomics. A dog's breakfast? Science 26: Reimann, N., Barnitzke, S., Bullerdiek, J., Schmitz, U., Rogalla, P., Niolte, I., Rønne, M., 1996 An extended nomenclature of the canine karyotype. Cytogenet. Cell Genet. 73: Rozen, S. and Skaletsky, H.J Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics Methods and Protocols: Methods in Molecular Biology. Humana Press, Totowa Schmitt, J.P., Kamisago, M., Asahi, M., Li G.H., Ahmad, F., Mende, U., Kranias, E.G., MacLennan, D.H., Seidman, J.G., Seidman, C.E Dilated cardiomyopathy and heart failure caused by a mutation in phospholamban. Science 28, Sisson, D.D. and Thomas, W.P., Myocardial diseases. In: Ettinger SJ and Feldman EC, eds. Textbook of veterinary internal medicine. 4 th ed. Philadelphia: WB Sounders Co Towbin, J.A., Bowles, N.E Genetic abnormalities responsible for dilated cardiomyopathy. Curr. Cardiol. Rep. 2: Vicart, P., Dupret, J.M., Hazan, J., Li, Z., Gyapay, G., Krishnamoorthy, R., Weissenbach, J., Fardeau, M., Paulin, D Human desmin gene: cdna sequence, regional localization and exclusion of the locus in a familial desmin-related myopathy. Hum. Genet. 98: Yang, F., O'Brien, P.C., Milne, B.S., Graphodatsky, A.S., Solanky, N., Trifonov, V., Rens, W., Sargan, D., Ferguson-Smith, M.A A complete comparative chromosome map for the dog, red fox, and human and its integration with canine genetic maps. Genomics 62: Zijlstra, C., Mellink, C.H.M., de Haan, N.A., and Bosma, A.A Localization of the 18S, 5.8S and 28S rrna genes and the 5S rrna genes in the babirusa and the whitelipped peccary. Cytogenet Cell Genet. 77:

83 ACKNOWLEDGMENTS Characterization of the canine DES We thank to veterinary students Debora Poldervaart and Laura Bloomer for helping gather the Dobermann blood samples, pedigrees and obtaining information on the status of dogs; Ruud Jalving for help with bioinformatics and the poodle SNP database; all dog owners for cooperation in our project. We are grateful to dr. N. Carter (The Sanger Centre, Cambridge, UK) for the canine painting probes that were used in this study, and to Lotte de Kreek and Esther van 't Veld for their participation in the FISH experiments. 83

84

85 Chapter 4 Duplication of a polymorphic CA-repeat by retrotransposition in the canine genome: implications for the analysis of the association of the α-tropomyosin gene (TPM1) with dilated cardiomyopathy in the Dobermann dog Polona Stabej, Peter A.J. Leegwater, Arnold A. Stokhof, Aleksandra Domanjko-Petrič, Bernard A. van Oost Submitted

86 Chapter 4 ABSTRACT Tropomyosin is one of the proteins of the sarcomere that plays structural, developmental and regulatory roles in cardiomyocytes. Diverse isoforms that are expressed in a tissue-specific manner are generated from four tropomyosin genes in mammals (TPM1-TPM4). Mutations in TPM1 cause dilated cardiomyopathy (DCM) in humans. As in humans, DCM is a common heart disease in dogs and mainly associates with giant and large breeds. In this paper we report the isolation of a canine BAC clone carrying the complete TPM1. The dog genome sequence assembly enabled anchoring of the TPM1 BAC clone to dog chromosome CFA30 as well as determination of the genomic structure of the canine TPM1. In addition, a BAC clone carrying a processed TPM1 pseudogene (TPM1-ψ) was isolated. A DNA sequence contig containing TPM1- ψ was found to reside on CFAX in the dog genome sequence assembly. The retrotransposition event, specific for the canine lineage, also encompassed a polymorphic CA-repeat in the 3 - untranslated region. To investigate whether mutations in TPM1 are responsible for DCM in the Dobermann, a PCR test specific for the CA-repeat adjacent to the TPM1 was developed and typed in a group of 26 Dobermanns diagnosed with DCM and 13 DCM non-affected Dobermanns. The marker displayed 4 alleles and was highly informative in the Dobermann breed. There was no association between any of the UU367B11 variants and the DCM phenotype, excluding with high probability a role for TPM1 in Dobermann DCM. 86

87 INTRODUCTION α- tropomyosin gene and pseudogene Alpha-tropomyosin is a central component of the regulatory machinery of Ca 2+ dependent muscle contraction. At low Ca 2+ concentrations (relaxed state), troponin C binds weakly to troponins T and I but troponin I binds strongly to actin, holding tropomyosin in a blocking position. Upon activation, tropomyosin moves to a non-blocking position and exposes the actin for myosin-head binding (Wolska and Wieczorek 2003). In vertebrates, tropomyosins α, β, γ and δ are encoded by a family of four genes (TPM1, -2, -3 and -4, respectively). Alternative promoters, alternative splice sites and multiple polyadenylation sites associated with alternative 3 -end processing result in expression of multiple isoforms (Lees-Miller et al. 1990). TPM1 is the predominantly expressed isoform in the heart (Perry, 2001). Defects in the TPM1 have been shown to cause dilated (DCM) and hypertrophic (HCM) cardiomyopathy in humans (Bashyam et al. 2003; Olson et al. 2001). Dilated cardiomyopathy in patients carrying TPM1 mutations is characterised by clinical signs of heart failure and sustained ventricular tachycardia. The two DCM causing TPM1 missense mutations were found to alter the surface electrostatic charge characteristics of the tropomyosin protein and may disturb the interactions between actin and tropomyosin filaments. However, the exact molecular and cellular mechanism is not yet clear (Olson et al. 2001). Dilated cardiomyopathy is also a common disease in dogs. The high prevalence of DCM in specific large and giant breeds suggests a genetic background, but causal mutations have not yet been identified. Dobermann DCM is associated with leftsided heart failure, ventricular arrhythmias, sudden death, age-related penetrance and very poor prognosis (Calvert et al. 1997). DCM in human and dog show a high similarity at the clinical level. As α-tropomyosin is one of the genes causing DCM in human, it is a good candidate to be investigated in canine DCM. For this purpose, we isolated a canine BAC clone carrying the canine α-tropomyosin gene. A pseudogene of the canine TPM1 (TPM1-ψ) residing on canine chromosome X was found and characterised. A highly polymorphic microsatellite marker derived from the TPM1 BAC clone was also found to be duplicated on the X- chromosome. The TPM1-linked repeat was tested for association with the DCM phenotype in a group of 26 DCM and 13 DCM-free Dobermanns. 87

88 Chapter 4 METHODS CANINE BAC LIBRARY RP81 SCREENING In order to isolate canine TPM1 clones, the BAC library RP81 (Li et al. 1999) was screened with a 40 bp human TPM1 exon 3 probe. Two partially complementary primers: OvA: 5 -TTCTCTGAACAGACGCATCCAGCT-3 and OvB: 5 -CAACTCTTCCTCAACCAGCTGGAT-3 were designed using the overgo maker program at complementary parts are indicated in bold. The probe was labelled with [α- 32 P]-dATP and [α- 32 P]- dctp at 37 C for one hour as described (Han et al. 2000). The hybridisation of the canine BAC library RP81 was performed as described on the BacPac web-site ( SOUTHERN BLOT ANALYSIS OF BAC CLONES 367B11 AND 416I15 Positive BAC clones 367B11 and 416I15 were grown and BAC DNA was isolated by alkaline lysis as described on the BacPac website ( To analyse the BAC clones for the presence of TPM1, BAC DNA and dog genomic DNA were digested with the restriction enzyme EcoRI, separated by electrophoresis on a 0.7% agarose gel and transferred to Hybond N+ filters (Amersham) by Southern blotting. The filters were hybridized overnight at 65 C with a [α 32 P]-dATP labelled human TPM1 cdna probe that was generated from IMAGE clone and that contained the human reference cdna sequence NM_ (Lennon et al. 1996). The clone was digested with restriction enzymes EcoRI and XhoI, the fragments were separated by agarose gel electrophoresis and stained with ethidium bromide. The 1250 bp clone insert was excised from the gel and purified using Quiaquick spin columns (Quiagen). After hybridisation, blots were washed with 2x SSC, 0.1% SDS at 65 C for 10 minutes. The blot with genomic DNA was autoradiographed at -70 C for 4 days with intensifying screens and BAC DNA blots were exposed to radiographs at room temperature for 30 minutes. The size of the EcoRI fragments hybridising to TPM1 cdna in the BAC clones and the genomic DNA were compared. 88

89 α- tropomyosin gene and pseudogene ALPHA TROPOMYOSIN GENE AND PSEUDOGENE SEQUENCES IN THE REFERENCE DOG ASSEMBLY DATABASE The Boxer dog assembled DNA sequence database was searched by BLASTN program ( using the human TPM1 cdna sequence (NM_000366) as a query. The identified contig sequences AAEX on CFA30 and AAEX on CFAX were aligned with one another by the Blast 2 Sequence BLASTN program at The DNA segments shared between the TPM1 sequences found on CFAX and CFA30 were evaluated. To determine the canine TPM1 exons and introns, the two contig sequences were aligned with the 15 exons described for the rat TPM1 (nomenclature according to Lees-Miller et al. 1990) using SeqMan (DNA Star alignment software). The open reading frame of the translated sequence in the pseudogene was determined by Open reading frame finder at (accessed on November 26, 2004). The DNA sequence 5 kb downstream of the 9d terminal exon was screened for interspersed repeats using RepeatMasker programme version at INVESTIGATION OF BAC CLONES 367B11 AND 416I15 FOR TPM1 SPECIFIC SEQUENCES The dog TPM1 and pseudogene in contigs AAEX and AAEX as well as the DNA sequences bordering the gene and the pseudogene were scanned for EcoRI restriction sites. The predicted EcoRI fragments were compared to those found by Southern blot analysis of the BAC clones. To further assess whether BAC clones 367B11 and 416I15 contain TPM1, the dog TPM1 EST27D12a (Guyon et al. 2003) and exon 8 specific PCRs (TPM1_ex8) were performed with DNA from the BAC clones as templates. Primers amplifying canine exon 8 were designed in intron 7 and intron 8 of the canine TPM1 using Primer3 software at and the AAEX contig sequence as a template (Table 1) [Rozen and Skaletsky 2000]. PCRs were performed on 2 µl of 1:10 diluted over-night BAC clone cultures using 0.33 µm of each primer, 0.6 U of Platinum Taq (Invitrogen), 200 µm dntps, 1x Gibco buffer and 3.5 mm MgCl 2 in a final volume of 15 µl. The PCR program consisted of a denaturation step of 10 min at 94 C, followed by 35 cycles of 30 sec 94 C, 30 sec 60 C, 30 sec 72 C and a final extension at 72 C for 10 min. 89

90 Chapter 4 RANDOM DNA SEQUENCE ANALYSIS OF THE BAC CLONE 367B11 The DNA of BAC clone 367B11 was digested with Sau3AI and the fragments were ligated into the BamHI site of pzero-1 vector (Invitrogen). TOP10F competent E.coli was transformed according to the protocol of the manufacturer (Invitrogen). Ninety-six clones were grown and DNA sequences of the inserts were determined using BigDye Terminator cycle sequencing (Applied Biosystems) with M13-forward (5 GTTTTCCCAGTCACGAC-3 ) and M13- reverse (5 -CAGGAAACAGCTATGAC-3 ) primers on an ABI 3100 genetic analyzer (Perkin Elmer ABI) according to the protocol of the manufacturer. The DNA sequences were compared to the dog genome assembly database using BLASTN software at The DNA sequences from subclones of BAC 367B11 were screened for repeats. One dinucleotide repeat (CA) n was found and named microsatellite UU367B11. Specific primers were designed to amplify the UU367B11 repeat sequence (Table 1). The exact location of the UU367B11 was established by alignment of the microsatellite marker sequence with the AAEX contig sequence and TPM1 exons using Seqman (DNA Star alignment software). ANIMALS AND DIAGNOSIS OF DCM Blood samples of 26 Dobermanns diagnosed with DCM and 13 DCM-free Dobermanns were collected. Genomic DNA was isolated by the salt extraction method (Miller et al. 1988). DCM Dobermanns were patients of the Department of Clinical Sciences of Companion Animals (Veterinary Faculty, Utrecht University) and of the Clinic for Small Animals and Surgery (Veterinary Faculty, Ljubljana University) between 1993 and The DCM-free Dobermanns were selected from a larger group of dogs of the same clinics at preventive check ups. DCM diagnosis was based on clinical and/or radiographic symptoms of congestive heart failure (CHF) and echocardiographic evidence of DCM in the absence of other CHF related lesions on two-dimensional echocardiography. Echocardiography was performed as described elsewhere (Stabej et al. 2004). Table 1. Oligonucleotide primers and their product lengths Primer name Forward primer 5'-3' Reverse primer 5'-3' Product length EST27D12a 1 ATGAGCTTCAGAACACTTGTAGGAC TCCATGATACCCACATATTCAAAGT 155 bp TPM1_Ex8 AGTCCTGGCGTTAATGTGCT AGGAGGATGCACTTGGATTG 317 bp UU367B11 ACTGTGTCCAGAGTGCAGCTA GATTGCTAGACTGGC 473 bp 1 Guyon et al

91 α- tropomyosin gene and pseudogene Dobermanns were confirmed as unaffected when they had no signs or symptoms indicative of DCM at an age of nine or older. Pedigrees of affected and unaffected dogs were gathered and relations between dogs in both groups were established. GENOTYPING OF MICROSATELLITE MARKER UU367B11 For each PCR reaction, 25 ng of canine genomic DNA was incubated with 0.33 µm of each primer, 0.6 U of Platinum Taq (Invitrogen), 200 µm dntps, 1x Gibco buffer and 1.5 mm MgCl 2 in a final volume of 15 µl. The PCR program consisted of a denaturation step of 10 min at 94 C, followed by 35 cycles of 30 sec 94 C, 30 sec 60 C, 30 sec 72 C and a final extension at 72 C for 10 min. The reverse primer of the UU367B11 marker was labelled with FAM fluorescent dye (Eurogentec). The PCR reactions were diluted 10x, 1 µl of the dilution was mixed with 10µl formamide and µl of size standard GS500-TAMRA (Applied Biosystems, CA) and analysed on the ABI 3100 Genetic Analyzer with filter set C (Applied Biosystems, CA). Genescan 3.1 software was used for genotype assessment. To assess whether a particular UU367B11 allele was associated with the DCM phenotype, the marker was typed in 26 Dobermanns diagnosed with DCM and in 13 DCM-free Dobermanns. Allele frequencies in the DCM and DCM-free groups of Dobermanns were compared using the Chi-square test. 91

92 Chapter 4 RESULTS BAC CLONES ISOLATED FROM THE CANINE BAC LIBRARY RP81 Screening of the canine BAC library RP81 with the TPM1 overgo probe identified 13 positive BAC clones: 22M21, 44D5, 49O22, 84B13, 285F5, 287I17, 300C4, 340F20, 367B11, 379C12, 379C13, 410J6 and 416I15. The clones 367B11 and 416I15 were further evaluated. These BAC clones had clearly different EcoRI restriction fragment patterns (Figure 1a). Southern blot analysis showed that clone 367B11 contained two and clone 416I15 a third fragment which are identical in size to EcoRI fragments observed in the Southern blot experiment with dog genomic DNA (Figures 1b and 1c). ALPHA TROPOMYOSIN GENE AND PSEUDOGENE IN THE DOG GENOME SEQUENCE ASSEMBLY Two DNA sequence contigs displaying high homology to human TPM1 cdna were identified in the dog genome assembly. Contig AAEX on CFA30 contained the complete canine TPM1 sequence with two alternate start codons (exons 1a and 1b) and 15 exons described in the rat TPM1 (Figure 2a) [Lees-Miller et al. 1990]. This chromosomal location of the canine TPM1 is in accordance with the human TPM1 location on HSA15q22.1, a region that shares genes with CFA30 (Guyon et al. 2003; Tiso et al. 1997). The dog EST27D12a cdna (GenBank AJ537298) that has been mapped to CFA30 (Guyon et al. 2003) was found to originate from the canine TPM1 by alignment of the AJ and AAEX DNA sequences. The microsatellite repeat UU367B11, isolated from BAC clone 367B11 by random sequencing (see Methods), resides 187 bp downstream of the terminal codon in exon 9d. The second contig with homology to human TPM1, AAEX , was located on CFAX. Alignment of this DNA sequence with TPM1 cdna from the rat showed that it was not interrupted by introns, indicating that it represented an intronless TPM1 pseudogene (TPM1-ψ). The TPM1-ψ spanned a DNA segment of 1,8 kb (from position 48.4 kb to 50.1 kb in the AAEX sequence). It starts 82 bp upstream the TPM1 start codon in exon 1a and ends 802 bp downstream of the stop codon in exon 9d with a poly-a tail followed by a canine LINE-1 element. TPM1-ψ contained TPM1 exons 1a, 2b, 3, 4, 5, 6a, 7, 8 and 9d (Figures 2a and 2b). The 1736 bp sequence displayed 97 % identity with the corresponding region of the original canine TPM1. The DNA sequence of the microsatellite repeat UU367B11 was also present in the pseudogene, although with a different number of CA repetitions ([CA] 25 in TPM1 and [CA] 19 in TPM1-92

93 α- tropomyosin gene and pseudogene ψ). The ORF in TPM1-ψ was disturbed by a T nucleotide insertion that causes a frameshift and a stop codon seventy-three bp downstream of the start codon. a. 1 2 M bp b. 1 M - bp c. 1 2 M - bp Figure 1. BAC clones 367B11 and 416I15 Southern blot analysis. a. EcoRI restriction fragment patterns of BAC clones isolated with a TPM1 probe. Lane 1 - BAC 367B11, lane 2 BAC 416I15. M - marker lane. Samples were run on 0.7 % agarose gel at 20V over night. b. and c. Auto-radiograms showing Southern blot analysis of the dog genomic DNA (Figure 1b, lane 1) and TPM1 BAC clone DNA (Figure 1c, lane 1-367B11, lane 2 -BAC 416I15) EcoRI digests hybridised with the human TPM1 cdna. M - marker. 93

94 Chapter 4 a. Canine TPM1 on CFA30 5 end end 1a 2a 2b 1b a 6b 7 8 9a 9b 9c 9d 4.6 kb 4.8 kb 6 kb 1.2 kb E E E E EEE EEE E E E b. Canine TPM1 pseudogene (TPM1- ψ) on CFAX 82 bp 852 bp 802 bp 2876 bp 5 end 3 end TPM1-Ψ LINE 1 4 kb E E Figure 2. The canine TPM1 and pseudogene a. Schematic representation of the canine TPM1 genomic organization on CFA30 as deducted from the AAEX contig sequence. The open boxes represent exons (numbering below the boxes according to Lees-Miller et al. 1990), which are separated by introns (intermittent lines). The intron lengths are indicated above the intermittent lines in basepairs (bp). The black boxes depict exons present in the canine TPM1-ψ on CFAX (Figure 2b). The grey part in exons 1a and 9d depict the regions upstream the start codon in exon 1a and downstream the stop codon in exon 9d showing high homology to TPM1-ψ sequences (Figure 2b). The EcoRI sites (E) along the canine TPM1 are outlined below the gene. The lengths of EcoRI restriction fragments containing exons present in the cdna clone used for Southern blot hybridization are noted in kilo basepairs (kb). b. Schematic representation of the canine TPM1-ψ on CFAX as deducted from the AAEX contig sequence. The black box represents the coding region with errors (852 bp long), the grey boxes depict sequences upstream the ATG start codon in exon 1a (82 bp) and downstream the TGA stop codon in exon 9d (802 bp) showing high similarity to the sequences upstream and downstream of the coding region in TPM1. The pseudogene ends with a poly-a tail that is immediately followed by a canine LINE-1 element. The two EcoRI sites (E) upstream and downstream TPM1-ψ give a fragment of 4 kb. 94

95 α- tropomyosin gene and pseudogene To verify the existence of the pseudogene in another breed, the Whole Genome Sequence database of the Poodle was searched by BLASTN using the TPM1-ψ identified in the Boxer breed as a query DNA sequence. A partial sequence of the canine TPM1-ψ was also present in the Poodle database under accession number CE (Kirkness et al. 2003). The CE sequence is intronless and contains 45 bp of canine TPM1 exon 5, complete exons 6, 7, 8 and 473 bp of exon 9d (complete translated region and 391 bp of 3 UTR). When compared to the boxer TPM1-ψ sequence, no differences were observed in the exonic sequences, but a different number of [CA] repeats was observed in the 3 UTR; [CA] 18 in the Poodle and [CA] 19 in the Boxer. ECORI RESTRICTION FRAGMENTS IN THE CANINE TPM1, TPM1-Ψ The canine TPM1 DNA sequence on chromosome CFA30 was scanned for EcoRI restriction sites. Thirteen restriction sites resulting in twelve EcoRI fragments were found (Figure 2a). Four of the fragments contained exonic sequences present in the IMAGE cdna clone used for Southern blot hybridization: a 4.6 kb fragment containing exons 1a and 2b, a 4.8 kb fragment with exon 3, a 6 kb fragment containing exons 4, 5, 6a, 7, 8, 9a and a 1.2 kb fragment with exon 9b (Figure 2a). Two restriction sites bracketing the pseudogene on chromosome CFAX produce a DNA fragment of 4 kb (Figure 2b). The TPM1 cdna hybridized with the 6 kb and 4.6 kb fragments in BAC 367B11 and with fragments of the same size in the genomic DNA (Figure 1b and 1c). A 4 kb band from the 416I15 BAC clone as well as from genomic DNA hybridized with the TPM1 probe, while a 8 kb band could be detected in the genomic DNA only (Figure 1b and 1c). The 8 kb fragment that was not present in the BAC clones represents most likely another TPM gene from the gene family or an incompletely digested fragment containing TPM1. The 4.8 kb fragment encompassing 134 bp of exon 3 and the 1.2 kb fragment containing 83 bp coding region of exon 9b, were not detected by the cdna probe. The absence of these two bands can be explained by the relatively short exonic sequence stretches present in the 4.8 kb and 1.2 kb band. The 6 kb and 4.6 kb fragments that did hybridize to the cdna probe contained 615 and 381 bp of exonic sequence, respectively. The BAC clone 367B11 tested positive as a template for the TPM1 PCR with EST27D12a, TPM1_Ex8 and UU367B11 oligonucleotide primers, whereas the BAC clone 416I15 tested negative for all three reactions. Based on these results we conclude that the 367B11 BAC carries the canine TPM1 and BAC 416I15 contains the TPM1 pseudogene (TPM1-ψ). 95

96 Chapter 4 ANIMALS STUDIED: CLINICAL EVALUATION AND FAMILIAL RELATIONS The 26 DCM Dobermanns included in this study were symptomatic with evidence of congestive heart failure. The affected dogs had severely dilated left ventricular chambers in systole and diastole with fractional shortening (FS) below 25 % as determined by echocardiography. Based on the birth date, none of the affected Dobermanns were siblings. The same applies to the group of unaffected dogs, but two DCM dogs with a pedigree had a sibling in the DCM-free group. The pedigrees of 11 affected and 9 unaffected dogs were reviewed. The other pedigrees were not available. Since none of these affected dogs shared both grandparents, there were 4 meioses or more in-between any two DCM dogs. The same applied to the unaffected dogs. EVALUATION OF THE MICROSATELLITE MARKER UU367B11 The 3 -UTR microsatellite UU367B11 was typed in 26 DCM and 13 DCMfree Dobermanns. Initially, the marker showed anomalous results as 3 alleles were found in a male dog, and even 4 alleles in a female dog (data not shown). With a new forward primer derived from intron 9 of the TPM1, Mendelian segregation was observed with alleles of four sizes: 469, 470, 474 and 479 bp. None of the four alleles associated with the DCM phenotype and comparison of allele frequencies in the DCM and DCM-free groups (Table 2) showed no significant differences (Chi-square test p-value = 0.88). Table 2. Microsatellite marker UU367B11 genotype frequencies of 26 DCM and 13 DCM-free dogs Allele DCM (n=52) DCM-free (n=26) (in bp) n % n %

97 DISCUSSION α- tropomyosin gene and pseudogene In the present study, we isolated BAC clones carrying the canine TPM1 and a pseudogene (TPM1-ψ). The genomic structure of the canine TPM1 resembles the rat and human TPM1. It spans a genomic DNA sequence contig of 27,7 kb and has 15 exons (Denz et al. 2004; Lees-Miller et al. 1990). Alpha-tropomyosin belongs to a family of tropomyosins encoded by four genes (TPM1 TPM4) that show a high degree of similarity to one another. The high homology of genes within one gene family coupled with the existence of pseudogenes that also display high identity to the original genes can complicate cloning of the gene of interest. Because of the risk of detection of false positive BAC clones, we carefully evaluated the identity of the BAC clones isolated. Southern blot analysis, amplification with oligonucleotides specific for the canine TPM1 and evaluation of random DNA sequences of BAC clone 367B11 all confirmed that the clone indeed contains canine TPM1. The 40bp long probe used for the screening of the BAC library was derived from human TPM1 exon 3 that shows 100% identity to exon 3 of the canine TPM1 and TPM1-ψ. It is therefore not surprising that the probe also hybridized to BAC clone 416I115 containing the TPM1-ψ. The TPM1-ψ clone was clearly different from clone 367B11 as shown by fingerprint and Southern blot analysis. Furthermore, it was negative for the intron sequences specific for the canine TPM1. The EcoRI fragment identified with the human c-dna TPM1 probe in BAC clone 416I15 was in agreement with the predicted EcoRI band (Figure 1c and Figure 2b). Pseudogenes often complicate BAC clone isolations of dog genes (van de Sluis, 1999). The availability of the dog genome assembly (Lindblad-Toh 2004) enabled instant sequence analysis of the canine TPM1-ψ that was found on CFAX. The canine TPM1-ψ described in this study lacks introns, displays high homology to its parental gene and possesses a poly-adenylate tail. These are all characteristics of retrocopies that recently integrated into the genome (Emerson et al We attempted to identify the TPM1-ψ in the mammals for which the complete genome sequence is available (human, rat and mouse) and did not identify a TPM1-ψ ortholog (data not presented). The pseudogene is not breed specific since the TPM1-ψ is present in the Boxer, Poodle and the Dobermann (canine BAC library RP-81; Li et al. 1999). Analyses of human and mouse pseudogenes have demonstrated that the mammalian X chromosome has generated a disproportionately high number of retroposed genes in comparison to the autosomes (Emerson et al. 2004). Also the amino acid sequence conservation points to a recent origin of TPM1-ψ. While human and dog TPM1 have 80% of the amino acids identical, TPM1-ψ is for 97% identical to its parental gene. As 97

98 Chapter 4 expected, the 3 -UTR microsatellite in the TPM pseudogene was also polymorphic as could be judged directly from comparisons of the whole genome sequence data for the Boxer and the Poodle breed, and the observation of 4 different alleles in a female dog. To the best of our knowledge, this is the first example of a retrotransposition of a polymorphic microsatellite. Several mutations identified in families with DCM and HCM prompted us to survey the α-tropomyosin gene (TPM1) in Dobermanns with DCM. The identified microsatellite marker UU367B11 resides in the TPM1 and is highly polymorphic. The marker is also present in the pseudogene, but the DNA sequence of the forward primer resides in intron 9d that is specific for the TPM1 and is not present in the pseudogene. Initially, the UU367B11 forward oligonucleotide primer was designed in exon 9d (present in TPM1-ψ) and led to inconclusive genotyping results. The existence of TPM1-ψ was discovered after a contiguous sequence of 6 kb around the UU367B11 repeat was assembled using the NCBI trace archives ( Two different contig sequences that were upstream UU367B11 [CA] n repeat identical until exon 9d, were assembled. After exon 9d, the TPM1 pseudogene and gene contig sequences continued with exon 8 and intron 8, respectively. The chromosomal location and complete structure of the pseudogene could be clarified upon the dog genome sequence assembly. The UU367B11 microsatellite marker showed no association of alleles with the DCM phenotype in the Dobermanns tested in our study. If mutations in the TPM1 would be responsible for DCM in the Dobermann, the DCM dogs would be expected to share a particular UU367B11 allele. Dobermanns in our DCM group displayed 4 alleles, none of which was shared by all of them. Also in comparison to the DCM-free group, no significant differences of UU367B11 allele frequencies were observed. Therefore, we have excluded mutations in TPM1 as a likely cause of DCM in Dobermanns. 98

99 α- tropomyosin gene and pseudogene REFERENCES Bashyam, M.D., Savithri, G.R., Kumar, M.S., Narasimhan, C., Nallari, P Molecular genetics of familial hypertrophic cardiomyopathy (FHC). J. Hum. Genet. 48: Calvert, C.A., Hall, G., Jacobs, G., Pickus, C Clinical and pathologic findings in Dobermanns with occult cardiomyopathy that died suddenly or developed congestive heart failure: 54 cases ( ). J. Am. Vet. Med. Assoc. 210: Denz, C.R., Narshi, A., Zajdel, R.W., Dube, D.K Expression of a novel cardiacspecific tropomyosin isoform in humans. Biochem. Biophys. Res. Commun. 320: Emerson, J.J., Kaessmann, H., Betran, E., Long, M Extensive gene traffic on the mammalian X chromosome. Science 303: Guyon, R., Lorentzen, T.D., Hitte, C., Kim, L., Cadieu, E., Parker, H.G., Quignon, P., Lowe, J.K., Renier, C., Gelfenbeyn, et al A 1-Mb resolution radiation hybrid map of the canine genome. Proc. Natl. Acad. Sci. 100: Han, C.S., Sutherland, R.D., Jewett, P.B., Campbell, M.L., Meincke, L.J., Tesmer, J.G., Mundt, M.O., Fawcett, J.J., Kim, U.J., Deaven, L.L. et al Construction of a BAC contig map of chromosome 16q by two-dimensional overgo hybridization. Genome Res. 10: Kirkness, E.F., Bafna, V., Halpern, A.L., Levy, S., Remington, K., Rusch, D.B., Delcher, A.L., Pop, M., Wang, W., Fraser, C.M., et al The dog genome: survey sequencing and comparative analysis. Science 301 : Lees-Miller, J.P., Goodwin, L.O., Helfman, D.M Three novel brain tropomyosin isoforms are expressed from the rat alpha-tropomyosin gene through the use of alternative promoters and alternative RNA processing. Mol. Cell. Biol. 10: Lennon, G.G., Auffray, C., Polymeropoulos, M. and Soares, M.B The I.M.A.G.E. Consortium: An Integrated Molecular Analysis of Genomes and their Expression. Genomics 33: Li, R., Mignot, E., Faraco, J., Kadotani, H., Cantanese, J., Zhao, B., Lin, X., Hinton, L., Ostrander, E.A., Patterson, D.F. et al Construction and characterization of an eightfold redundant dog genomic bacterial artificial chromosome library. Genomics 58: Lindblad-Toh, K Preliminary analysis of a high-quality draft sequence in the dog genome. In: 2 nd International Conference Advances in Canine and Feline Genomics, pp. 21. Utrecht, The Netherlands, October Miller, S.A., Dykes, D.D., Polesky, H.F A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16: Olson, T.M., Kishimoto, N.Y., Whitby, F.G., Michels, V.V Mutations that alter the surface charge of alpha-tropomyosin are associated with dilated cardiomyopathy. J. Mol. Cell. Cardiol. 33: Perry, S.V Vertebrate tropomyosin: distribution, properties and function. J. Muscle. Res. Cell. Motil. 22: Rozen, S. and Skaletsky, H.J Primer3 on the WWW for general users and for biologist programmers. In Bioinformatics Methods and Protocols: Methods in Molecular Biology (eds. S. Krawetz and S.Misener), pp Humana Press, Totowa, NJ. 99

100 Chapter 4 Stabej, P., Imholz, S., Versteeg, S.A., Zijlstra, C., Stokhof, A.A., Domanjko-Petrič, A., Leegwater, P.A., van Oost, B.A Characterization of the canine desmin (DES) gene and evaluation as a candidate gene for dilated cardiomyopathy in the Dobermann. Gene 340: Tiso, N., Rampoldi, L., Pallavicini, A., Zimbello, R., Pandolfo, D., Valle, G., Lanfranchi, G., Danieli, G.A Fine mapping of five human skeletal muscle genes: alphatropomyosin, beta-tropomyosin, troponin-i slow-twitch, troponin-i fast-twitch, and troponin-c fast. Biochem. Biophys. Res. Commun. 230: van de Sluis, B.J., Breen, M., Nanji, M., van Wolferen, M., de Jong, P., Binns, M.M., Pearson, P.L., Kuipers, J., Rothuizen, J., Cox, D.W., et al Genetic mapping of the copper toxicosis locus in Bedlington terriers to dog chromosome 10, in a region syntenic to human chromosome region 2p13-p16. Hum. Mol. Genet. 8: Wolska, B.M., Wieczorek, D.M The role of tropomyosin in the regulation of myocardial contraction and relaxation. Pflugers Arch. 446:

101

102

103 Chapter 5 The canine sarcoglycan delta gene: BAC clone contig assembly, chromosome assignment and interrogation as a candidate gene for dilated cardiomyopathy in Dobermann dogs Polona Stabej, Peter A.J. Leegwater, Sandra Imholz, Serge A. Versteeg, Carla Zijlstra, Arnold A. Stokhof, Aleksandra Domanjko- Petrič, Bernard A. van Oost Cytogenetics and Genome Research 2005; in press

104 Chapter 5 ABSTRACT Dilated cardiomyopathy (DCM) is a common disease of the myocardium recognized in human, dog and experimental animals. Genetic factors are responsible for a large proportion of cases in humans, and 17 genes with DCM causing mutations have been identified. The genetic origin of DCM in the Dobermann dogs has been suggested, but no disease genes have been identified to date. In this paper, we describe the characterization and evaluation of the canine sarcoglycan delta (SGCD), a gene implicated in DCM in human and hamster. Bacterial Artificial Chromosomes (BACs) containing the canine SGCD were isolated with probes for exon 3 and for exons 4-8 and were characterized by Southern blot analysis. BAC end sequences were obtained for four BACs. Three of the BACs overlapped and could be ordered relative to each other and the end sequences of all four BACs could be anchored on the preliminary assembly of the dog genome sequence ( One of the BACs of the partial contig was localized by Fluorescent In Situ Hybridization to canine chromosome 4q22, in agreement with the dog genome sequence. Two highly informative polymorphic microsatellite markers in intron 7 of the SGCD were identified. In 25 DCM-affected and 13 non DCM-affected dogs, 7 different haplotypes could be distinguished. However, no association between any of the SGCD variants and the disease locus was apparent. 104

105 Canine sarcoglycan- δ gene INTRODUCTION Dilated cardiomyopathy (DCM) is a myocardial disease that represents an important cause of congestive heart failure (CHF) and sudden death in dogs and humans (Fatkin and Graham, 2002; Sisson and Thomas 1995). The disease is characterized by dilatation and impaired contraction of the left or both ventricles (Richardson et al, 1996). The high prevalence of DCM in specific breeds suggests a genetic background, but causal mutations have not yet been identified. Within each breed, DCM has unique characteristics and between breeds, it is probably a genetically heterogeneous disease (for review see Dukes-McEwan et al. 2003). Human DCM, which closely resembles the canine form of the disease, is a genetically heterogeneous disease with 17 disease genes identified (for reviews see Fatkin and Graham 2002; Knöll et al. 2002; Murphy et al. 2004; Schmitt et al. 2003; Vatta et al. 2003). Most of these genes encode proteins of the cytoskeleton and the mechanism by which mutations cause DCM is thought to be impairment of heart force production or transmission (Towbin and Bowles, 2000). One of the genes involved in DCM encodes the sarcoglycan delta (SGCD) polypeptide of 35 kda. The SGCD is located on human chromosome (HSA) 5q33-q34 and has 9 exons that are alternatively spliced. The SGCD product is one of four subunits (α, β, γ and δ) in the sarcoglycan complex, which is a part of the dystrophinassociated protein complex (DAPC) located in the muscle plasma membrane (Nigro et al. 1996; Tsubata et al. 2000). The function of the DAPC is thought to be maintenance of the muscle membrane integrity by providing a physical connection between the actin cytoskeleton and the extracellular matrix (reviewed by Hemler, 1999). In human patients, three DCM causing mutations in the SGCD have been described: Ser151Ala, Arg71Thr and a 3-bp deletion in position 238 (del Lys238). The Ser151Ala and del Lys238 were predicted to alter the secondary structure of the SGCD and consequentially affect force transmission leading to DCM (Karkkainen et al. 2003; Tsubata et al. 2000). Prior to the discovery of the latter three mutations, SGCD has been demonstrated as the causative gene in the Syrian hamster cardiomyopathy (Nigro et al. 1997; Sakamoto et al. 1997). The aim of this study was to evaluate the role of the SGCD in DCM in the Dobermann, a dog breed with frequent DCM occurrence (Domanjko-Petrič et al. 2002). For this purpose we isolated a partial canine SGCD BAC contig. This contig could be aligned with the orthologous human chromosomal region and the preliminary dog genome sequence assembly on canine chromosome 4. A tetranucleotide DNA marker identified in the SGCD BAC contig and a dinucleotide DNA marker located in a sequence contig of the whole genome dog 105

106 Chapter 5 sequence database were used to query a possible involvement of the SGCD in canine DCM. MATERIALS AND METHODS DNA SAMPLES Blood samples of 25 Dobermanns diagnosed with DCM and 13 DCM-free Dobermanns were collected. The DCM Dobermanns were patients seen by the Department of Clinical Sciences of Companion Animals (Veterinary Faculty, Utrecht University) and by the Clinic for Small Animals and Surgery (Veterinary Faculty, Ljubljana University) between 1993 and The DCM-free Dobermanns were selected from a larger group of dogs seen by the same clinics at preventive check ups. The pedigrees of dogs were collected. Genomic DNA was isolated by the salt extraction method (Miller et al. 1988). DCM diagnosis was based on clinical and/or radiographic symptoms of congestive heart failure (CHF) and echocardiographic evidence of DCM in the absence of other CHF related lesions on two-dimensional echocardiography. Echocardiography was performed in conscious dogs, with the dog in right lateral recumbency according to standard procedures (Thomas et al. 1993). DCM was defined as left ventricular internal diastolic dimension >50 mm, left ventricular internal systolic dimension >38 mm and left ventricular fractional shortening (FS) < 25 % (Calvert et al. 2000). Dobermanns were confirmed as unaffected when they had no signs or symptoms indicative of DCM at an age of nine years or older. SGCD BAC CLONES ISOLATION Canine BAC library RP81 (Li et al. 1999) was screened with two probes. For the first probe, two overlapping primers (overlap is indicated in bold): OvA: 5 - TACTGGTGAACTTGGCCATGACCA-3 and OvB: 5 -ACCTTGAGAATCCA GATGGTCATG-3 were designed for exon 3 of the mouse SGCD cdna sequence (NCBI GenBank accession no. AB024923) using Overgo maker program ( The 40bp probe was synthesized with an [α 32 P]-dATP and an [α 32 P]-dCTP at 37 C for one hour using the overgo technique (Han et al. 2000). The IMAGE clone (obtained from the Resource Centre of the German Human Genome Project Berlin, Germany) which contains exons 2 8 of the human SGCD isoform 2 cdna (GenBank NM_172244) was the second probe (Lennon et al. 1996). The clone was digested with restriction enzymes EcoRI and NotI. Exon 4 of the SGCD contains an EcoRI 106

107 Canine sarcoglycan- δ gene restriction site and the insert of the clone was cut in two fragments. The fragment of 800bp containing 44bp of exon 4 and exons 5-8 was excised from the agarose gel and purified using Quiaquick spin columns (Quiagen). The 800 bp fragment was labelled with an [α 32 P] by nick translation (Sambrook and Russel, 2001). The hybridization of canine BAC library filters with either the overgo probe or the cdna probe was performed as described on the BacPac web-site ( SGCD BAC CLONES ANALYSIS To confirm the identity of isolated BAC clones, BAC DNA and canine genomic DNA were digested with EcoRI, fractionated on a 0.7 % agarose gel and Southern blotted onto Hybond N + filters (Amersham). The blots were hybridised overnight at 65 C with an [α 32 P]-dATP labelled partial insert (exons 4-8) of the human cdna. The filters were washed with 2x SSC, 0.1% SDS at 65 C for 10 minutes. Genomic DNA was exposed to film at -70 C for 3 days and BAC DNA blots for 90 minutes with intensifying screens. To further characterize the SGCD BAC clones, the ends of clones 21F2, 98H15, 379A9 and 409K19 were sequenced using 800 ng BAC DNA, 0.66 µm of sequencing primers (SP6: 5 - GTTTTTTGCGATCTGCCGTTTC 3, T7: 5 -TAATACGACTCACTATAGG G 3 ), 12 µl BigDye Terminator cycle sequencing (Applied Biosystems), 4 mm MgCl 2 in a total volume of 30 µl. The following cycle conditions were performed: 5 min at 95 C, followed by 31 cycles of 30 sec at 95 C, 10 sec at 55 C and 4 min at 60 C. The generated BAC end DNA sequences were submitted to the NCBI GenBank. The overlapping clones were confirmed by PCR using PCR primers generated from the BAC end sequences. Exact locations of BAC clones and SGCD exons in the dog genome were determined by comparing the BAC end sequences and human SGCD exonic DNA sequences to the dog genome assembly. The Ensembl Canis familiaris BLAST search was used ( CHROMOSOMAL LOCALIZATION OF THE CANINE SGCD FISH was performed on GTG-banded metaphase spreads, which were obtained from concanavalin-a stimulated peripheral blood lymphocytes, as previously described (van den Berg et al. 2004; Zijlstra et al. 1997). Briefly, wellbanded metaphases were captured prior to FISH using a Leica DMRA microscope equipped with the Genus Image Analysis software (Applied Imaging). Total DNA of a canine SGCD BAC clone 409K19 was labeled with biotin-16-dutp by nick translation and used as a probe. The final probe concentration was 5 ng/µl and a 107

108 Chapter 5 50-fold excess of fragmentated total dog DNA was added as competitor. Specific sites of hybridization were detected using avidin-fitc, and signals were amplified twice. Chromosomes were counterstained with propidiumiodide. Previously captured metaphases were reexamined after FISH, recaptured and analyzed using the GENUS software. Chromosomes showing specific hybridization signals were identified on the basis of their G-banding patterns according to the nomenclature of Switonski et al MICROSATELLITE REPEAT MARKERS ISOLATION FROM SGCD BAC SUBCLONES Two BAC clones were selected (BAC 409K19 and BAC 21F2) for microsatellite repeat isolation. The clones were digested with Sau3AI and fragments were ligated into the BamHI site of pzero-1 vector (Invitrogen) using standard procedures (Sambrook and Russel, 2001). TOP10F competent cells were transformed according to the manufacturer s protocol (Invitrogen). Ninetysix clones were picked at random and DNA sequences of the inserts were determined using BigDye Terminator cycle sequencing (Applied Biosystems) with M13-forward (5 -GTTTTCCCAGTCACGAC-3 ) and M13-reverse (5 - CAGGAAACAGCTATGAC-3 ) primers on an ABI 3100 genetic analyzer (Perkin Elmer ABI) according to the protocol of the manufacturer. The DNA sequences were screened for microsatellite repeats. MICROSATELLITE REPEAT IDENTIFICATION IN THE CANINE GENOME DATABASE Contig AAEX was identified to contain part of the SGCD by Blast search of the NCBI WGS database with the human SGCD mrna sequence (NM_000337) as the query sequence. A microsatellite repeat (CA) 24 was found in dog SGCD intron 7. With Primer 3 software (Rozen and Skaletsky 2000), specific oligonucleotides were designed to PCR amplify the UU21F2 and the SGCD microsatellite repeats. The distance between the microsatellite repeats and their positions in canine SGCD were determined by aligning the primer sequences of both microsatellites and SGCD human exonic sequences with the AAEX contig sequence using Seqman (DNA Star alignment software). 108

109 Canine sarcoglycan- δ gene MICROSATELLITE REPEAT AMPLIFICATION AND FRAGMENT SIZE ANALYSIS For each PCR reaction, 25 ng of canine genomic DNA was amplified with 0.33 µm of each primer, 0.6 U of Platinum Taq (Invitrogen), 200 µm dntps, 1x Gibco buffer and 1.5 mm MgCl 2 in a final volume of 15 µl. The PCR program consisted of a denaturation step of 10 min at 94 C, followed by 35 cycles of 30 sec 94 C, 30 sec 60 C, 30 sec 72 C and a final extension at 72 C for 10 min. The primer sequences of the UU21F2 microsatellite repeat were: UU21F2-f: 5 - CCTGGGTTTCCCTAAGATTC-3 and UU21F2-r: 5 -GGATGGAGGCATTTTC AAGC-3 and of the SGCD-CA microsatellite were SGCD-CA-f: 5 - TTCAATCAGAGAAAGAGAACCTC-3 and SGCD-CA-r: 5 -AGGGCACTA GACACAGTTGG-3. The forward primers of the UU21F2 and SGCD-CA markers were labelled with HEX fluorescent dye (Eurogentec). The PCR reactions were diluted 5x, 1 µl of the dilution was mixed with 10µl formamide and µl of size standard GS-500-TAMRA (Applied Biosystems, CA) and analysed on the ABI 3100 Genetic Analyzer with filter set C (Applied Biosystems, CA). Genescan 3.1 software was used for genotype assessment. DEFINITION OF HAPLOTYPES AND ASSOCIATION ANALYSIS Haplotypes were constructed for the UU21F2-GAAA and SGCD-CA microsatellite markers. The haplotypes were inferred from the genotypes in such a way as to keep the number of different haplotypes as low as possible. Haplotype frequencies in the DCM and DCM-free groups were compared using the Chisquare test. RESULTS ISOLATION AND CHARACTERIZATION OF SGCD BAC CLONES The screening of the canine genomic BAC library RPCI-81 with the SGCD overgo probe of exon 3 yielded six positive BAC clones: 7N13, 9K14, 406A6, 407M8, 409K19 and 420M13. Hybridization of the library with the human cdna probe derived of exons 4-8 yielded five additional positive BAC clones: 21F2, 98H15, 352D24, 354K17 and 379A9. The EcoRI restriction fragment patterns of the six BAC clones isolated with the exon 3 probe are similar, but not identical (Figure 1a, lanes 1-6), whereas BAC clones isolated with the partial SGCD cdna probe (Figure 1b, lanes 1-3) show a different restriction pattern when compared to BAC clones obtained with the exon 3 probe. BAC clones 98H15 and 379A9 (Figure 1b, lanes 2 and 3) show 109

110 Chapter 5 high similarity, whereas BAC clone 21F2 (Figure 1b, lane 1) has only some bands in common with BAC 379A9 (Figure 1b, lane 3). Southern blots with EcoRI restriction enzyme digests of DNA of the BAC clones and total dog genomic DNA (Figure 1a and 1b) were hybridised with a partial human SGCD cdna fragment that contained exons 4-8. The probe detected three DNA fragments in genomic DNA (Figure 1c, lane 1). DNA fragments of the same size lighted up in BAC 21F2, 98H15 and 379A9 (Figure 1c, lanes 2-4). In BAC clones that were isolated with the exon 3 overgo fragment, the probe did not hybridize to any fragments (not shown). We concluded that BAC clones isolated with exon 3 probe do not contain exons 4-8 of the SGCD, but might contain SGCD exons 1-4. For the BAC clones isolated with the exon 4-8 SGCD cdna probe, we confirmed that they overlap and carry the SGCD sequences (Figure 1c). BAC end sequences were submitted to the GSS division of the NCBI GenBank under accession numbers CL CL BLAST searches performed with BAC end sequences revealed 100% matches to the genomic regions on CFA4. After the SGCD exons were anchored to CFA4, the SGCD BAC contig was constructed and its coverage of the SGCD established (Table 1, Figure 2). CHROMOSOMAL LOCALIZATION OF THE CANINE SGCD For the chromosomal localization of the canine SGCD, BAC clone 409K19 DNA was used as probe for FISH. In total, 25 GTG-banded metaphase spreads were analyzed after hybridization. All metaphases showed specific fluorescent signals on both chromosomes of pair 4, band q22 (Figure 3). This band corresponds to CFA 4q31 according to the nomenclature of Breen et al. (1999a). IDENTIFICATION OF MICROSATELLITE REPEATS IN THE SGCD Two BAC clones were subjected to random sequencing to search for microsatellite repeats. One tetra nucleotide repeat (GAAA) 33, named UU21F2, was present in BAC clone 21F2, whereas no microsatellites were observed in 96 subclones of BAC clone 409K19. A (CA) 24 microsatellite repeat (SGCD-CA) was found in the canine genome database contig sequence AAEX Both microsatellite repeats are located in the 84,4 kb large canine SGCD intron 7. They lie 39,8 kb (UU21F2) and 73,9 kb (SGCD-CA) downstream of exon

111 Canine sarcoglycan- δ gene Table 1. Positioning of the canine SGCD BAC clones based on the BAC end sequences BAC clone BAC 409K19 BAC 21F2 BAC 379A9 BAC 98H15 BAC-end sequence 1 CL CL CL CL CL CL CL CL CFA ,888 kb 56,725 kb 56,611 kb 56,463 56,526 56,333 56,449 56,34 76 kb upstream 68 kb downstream 46,1 kb upstream 59,9 kb downstream 2,42 kb upstream 103,7 kb downstream 10,6 kb upstream 96,7 kb downstream Position 3 exon 1 exon 3 exon 5 exon 7 exon 7 exon 9 exon 8 exon 9 Overlap 4 ND no overlap no overlap 379A9 21F2 no overlap 379A9 379A9 1 NCBI GenBank accession number of the BAC end sequence. 2 Position on Canis familiaris chromosome (CFA) 4 as determined by Ensembl Canis familiaris BLAST ( Accessed on 19 November 2004) 3 Position on CFA4 in relation to the canine SGCD exons. Positions of the canine SGCD exons were determined by Ensembl Canis familiaris BLAST ( accessed on 19 November 2004) 4 Overlap as determined by amplifying the BAC DNA with the PCR primers designed on the SP6 and T7 BAC end sequences ND not determined 5 end 2 14,8 149,4 79,7 5,6 35,6 84,4 1,4 3 end BAC 409K19 BAC 21F2 115 kb BAC - 379A9 BAC - 98H15 Figure 2. Schematic representation of the human SGCD structure on CFA4. The vertical bars represent nine exons. Numbering of exons (below the black vertical bars) and exon/intron boundaries are according to Tsubata et al. (2000). The length of the introns is indicated in kilobasepairs (kb). Positions and overlap of dog SGCD BAC clones is depicted. 111

112 Chapter 5 a. b. c M M bp M bp Figure 1. Restriction fragment analysis of the SGCD and SGCD positive BAC clones. a. EcoRI restriction fragment patterns of BAC clones isolated with a SGCD exon 3 probe (lanes 1-6). Lane 1 BAC 7N13, lane 2 BAC 9K14, lane 3 BAC 406A6, lane 4 BAC 407M8, lane 5 BAC 409K19, lane 6 BAC 420M13. M marker lane. Samples were run in a 0.7 % agarose gel at 20V over night. b. EcoRI restriction fragment patterns of BAC clones isolated with a SGCD IMAGE cdna clone (lanes 1-3). Lane 1 BAC 21F2, lane 2 BAC 98H15, lane 3-379A9. M marker lane. Samples were run in a 0.7 % agarose gel at 20V over night. c. Auto-radiograms showing Southern blot analysis of the dog genomic DNA (lane 1) and SGCD BAC clone DNA (lane 2 21F2, lane 3 BAC 379A9, lane 4 98H15) EcoRI digests hybridised with the partial SGCD IMAGE cdna clone M marker lane. The genomic DNA and BAC DNA digests were run on the same gel. PATIENT SELECTION AND ASSOCIATION ANALYSIS OF SGCD HAPLOTYPES WITH DCM The 25 DCM-affected Dobermanns included in this analysis were symptomatic with evidence of congestive heart failure. All affected dogs had severely dilated left ventricular chambers in systole and diastole with fractional shortening (FS) below 25 % as determined by echocardiography. Thirteen non DCM-affected Dobermann dogs were selected on the basis of no signs or symptoms indicative of DCM at an age of nine years or older. Pedigrees of 11 affected and 9 unaffected dogs could be reviewed. None of the affected dogs shared both grandparents and the same applied to the unaffected dogs. For the other dogs no pedigrees were available. Two DCM dogs had a sibling in the DCM-free group. 112

113 Canine sarcoglycan- gene * Figure 3. Chromosomal localization of BAC clone 409K19 (full color figure on page 183). a. Representative GTG-banded metaphase spread in which chromosomes of pair 4 are indicated by arrowheads. Insert shows ideogram of CFA4 according to Swintonski et al. (1996), * indicates band 4q22 a. b. Same metaphase spread as under a. after FISH. Arrows indicate specific hybridization signals on both chromosomes of pair 4. b. The UU21F2 marker displays five alleles of 182, 186, 190, 194 and 198 bp and the SGCD-CA marker displayed alleles of 374, 378 and 380 bp in the two Dobermann subpopulations tested. Of the 15 possible haplotypes, 6 could be deduced directly because the dogs were homozygous for both or one of the markers. An additional haplotype (number 3) was inferred from the dogs heterozygous for both markers (Table 3). The haplotype frequencies in the DCM and DCM-free groups were not significantly different (Chi-square test p-value = 0.88). 113

114 Chapter 5 Table 2. Position of SGCD and its neighboring genes on human (HSA) and dog (CFA) chromosomes Position Position Position HSA5 CFA4 Gene Name Marker symbol (Mb) 1 CFA (Mb) 2 Reference Calnexin CANX N.A. Breen et al Prophet of PIT1 PROP N.A. Lantiga-van Leeuwen et al LOC EST11E ,9 Guyon et al Adrenergic alpha 1B receptor ADRA1B ,9 Guyon et al Enthoprotin EST8G9 (ENTH) ,7 Breen et al Sarcoglycan delta SGCD 156 4q31 56,4-56,8 present study Phosphodiesterase 6A PDE6A 149 4q31 68,8 Breen et al Fibroblast growth factor 1 FGF N.A. Breen et al location obtained from Entrez Gene - build 34.3 ( ). 2 Position on CFA4 as determined by Ensembl Canis familiaris BLAST ( accessed on 19 November 2004). N.A. - not applicable Table 3. Dobermann SGCD haplotypes and possible association with DCM Haplotype frequencies (%) 3 Haplotype 1 UU21F2 2 SGCD-CA 2 DCM (n=50) DCM-free (n=26) , , , , , Haplotypes were numbered in descending frequency in the non affected Dobermann group 2 alleles are numbered based on the length of the PCR fragment 3 25 DCM-affected dogs and 13 DCM-free dogs were genotyped 114

115 Canine sarcoglycan- δ gene DISCUSSION The human SGCD is spread over 433 kb and has several large introns. Intron 3 is with 163,9 kb the largest. We therefore chose to use probes from different parts of the gene to isolate BAC clones with the canine SGCD. Isolation of BAC clones from the canine BAC library supplied us with the necessary material for canine SGCD chromosomal assignment and evaluation as a DCM candidate gene in Dobermanns. We screened the canine BAC library (RP81) separately with two probes. With the 40bp long SGCD exon 3 overgo probe BAC clones containing SGCD exons 1-3 were isolated (spanning 76 kb upstream exon 1 to 68 kb downstream of exon 3). With the SGCD exon 4-9 cdna probe BAC clones were isolated that together span the region from 46,1 kb upstream of SGCD exon 5 to 96,7 kb downstream exon 9. The region that was not represented in any of the BAC clones is about 115 kb large and spans 81,4 kb of SGCD intron 4, exon 4 and 33,6 kb of intron 4 (Figure 2). It is possible that we failed to isolate BAC clones containing the 115 kb region around exon 4 because the SGCD exon 4-9 probe contained only 44 bp of exon 4 and was thus too short to hybridise to BAC clones containing the region around exon 4. The use of short overgo probes has proved to be efficient at BAC library screenings (Han et al. 2000). However, a gene like SGCD, consisting of large introns, is likely to contain multiple EcoRI (enzyme used for canine RP81 BAC library construction) restriction sites (Li et al. 1999). Such a gene is sub-cloned into several BAC clones that can be most efficiently identified by screening the BAC library using a complete cdna probe of the gene of interest. Since it has been recognized that the dog genome contains a high number of pseudogenes (van de Sluis et al. 2001), the true identity of BAC clones should always be carefully evaluated by Southern blot analysis of EcoRI fragments of dog genomic and BAC DNA. Whereas Southern blot analysis confirmed the identity of BAC clones 21F2, 98H15 and 379A9, it failed to confirm the identity of BAC clones isolated with the exon 3 probe. This can be explained by the fact that the probe used for Southern blot analysis contained SGCD exons 4-9 which were not present in the BAC clones isolated with the exon 3 probe. In order to evaluate the exact position of the BAC clones, the BAC ends were compared to the preliminary assembly of the dog genome ( Each of the BAC end sequences had its 100% match on CFA4 and the overlap of clones was in agreement with the data obtained by amplifying of the BAC DNA with the PCR primers designed on the BAC end sequences. This is a clear demonstration of how the availability of the dog genome sequence assembly can expedite assessment of canine genes. 115

116 Chapter 5 In this study, we mapped the SGCD to canine chromosome 4. Human SGCD is located on chromosome HSA5q33-q34. Comparative painting studies between dog and human chromosomes have demonstrated that the distal half of HSA5q shares homology with CFA 2, 4 and 11 (Breen et al. 1999b; Yang et al. 1999). At present, eight of the genes that are located on human chromosome region 5q31- q35 have also been mapped in the dog. These genes are listed in Table 3, together with their positions in the human and canine genome. Two of these genes have been regionally mapped thus far, PDE6A (Breen et al. 2001) and SGCD (present study). The localization of SGCD to the same chromosome band (CFA 4q31) as PDE6A further extends the conserved block of genes comprising LOC134492, ADRA1B, ENTH, and PDE6A (Table 3). Genetic diseases cause distress to dogs and their breeders and represent a challenge to researchers (Ostrander et al. 2000; Ostrander and Comstock, 2004; Patterson 2000). Understanding of genetic backgrounds of inherited diseases enables selective breeding and improved health of dog breeds as well as provides a target for genetic diagnosis and insight into pathophysiology. In this study, we isolated two polymorphic markers in the canine SGCD. Six out of seven haplotypes inferred from the genotyping data of the two markers were present in the DCM group and there was no single haplotype that was shared by all affected dogs (data not shown). In addition, there were no significant differences in the haplotype frequencies of the DCM and DCM-free groups. This indicates that mutations in SGCD are not a major cause of Dobermann DCM. Nevertheless, it remains a good candidate for DCM in breeds other than the Dobermann. 116

117 Canine sarcoglycan- δ gene REFERENCES Breen, M., Bullerdiek, J., Langford, C.F. 1999a. The DAPI banded karyotype of the domestic dog (Canis familiaris) generated using chromosome-specific paint probes. Chromosome Res 7: Breen, M., Thomas, R., Binns, M.M., Carter, N.P., Langford, C.F. 1999b. Reciprocal chromosome painting reveals detailed regions of conserved synteny between the karyotypes of the domestic dog (Canis familiaris) and human. Genomics 61: Breen, M., Jouquand, S., Renier, C., Mellersh, C.S., Hitte, C., Holmes, N.G., Cheron, A., Suter, N., Vignaux, F., Bristow, A.E. et al Chromosome-specific single-locus FISH probes allow anchorage of an 1800-marker integrated radiation-hybrid/linkage map of the domestic dog genome to all chromosomes. Genome Res. 11: Calvert, C.A., Jacobs, G.J., Smith D.D., Rathbun, S.L., Pickus, C.W Association between results of ambulatory electrocardiography and development of cardiomyopathy during long-term follow-up of Dobermanns. J Am Vet Med Assoc. 216: Domanjko-Petrič, A., Stabej, P., Žemva, A Dilated cardiomyopathy in Dobermanns: Survival, Causes of Death and a Pedigree Review in a Related Line. Journal of Veterinary Cardiology 4: Dukes-McEwan, J., Borgarelli, M., Tidholm, A., Vollmar, A.C., Haggstrom, J Proposed Guidelines for the Diagnosis of Canine Idiopathic Dilated Cardiomyopathy. Journal of Veterinary Cardiology 5: Fatkin, D., Graham, R.M Molecular mechanisms of inherited cardiomyopathies. Physiol Rev. 82(4): Guyon, R., Lorentzen, T.D., Hitte, C., Kim, L., Cadieu, E., Parker, H.G., Quignon, P., Lowe, J.K., Renier, C., Gelfenbeyn, B., et al A 1-Mb resolution radiation hybrid map of the canine genome. Proc Natl Acad Sci U S A 100: Han, C.S., Sutherland, R.D., Jewett, P.B., Campbell, M.L., Meincke, L.J., Tesmer, J.G., Mundt, M.O., Fawcett, J.J., Kim, U.J., Deaven, L.L., Doggett, N.A Construction of a BAC contig map of chromosome 16q by two-dimensional overgo hybridization. Genome Res. 10: Hemler, M.E. Dystroglycan versatility Cell 97: Karkkainen, S., Miettinen, R., Tuomainen, P., Karkkainen, P., Helio, T., Reissell, E., Kaartinen, M., Toivonen, L., Nieminen, M.S., Kuusisto, J., et al A novel mutation, Arg71Thr, in the delta-sarcoglycan gene is associated with dilated cardiomyopathy. J Mol Med. 81: Knöll, R., Hoshijima, M., Hoffman, H.M., Person, V., Lorenzen-Schmidt, I., Bang, M.L., Hayashi, T., Shiga, N., Yasukawa, H., Schaper, W., et al The cardiac mechanical stretch sensor machinery involves a Z disc complex that is defective in a subset of human dilated cardiomyopathy. Cell 111: Lantinga-van Leeuwen, I.S., Kooistra, H.S., Mol, J.A., Renier, C., Breen, M., van Oost, B.A Cloning, characterization, and physical mapping of the canine Prop-1 gene (PROP1): exclusion as a candidate for combined pituitary hormone deficiency in German shepherd dogs. Cytogenet Cell Genet 88: Lennon, G.G., Auffray, C., Polymeropoulos, M., and Soares, M.B The I.M.A.G.E. Consortium: An Integrated Molecular Analysis of Genomes and their Expression. Genomics 33:

118 Chapter 5 Li, R., Mignot, E., Faraco, J., Kadotani, H., Cantanese, J., Zhao, B., Lin, X., Hinton, L., Ostrander, E.A., Patterson, D.F., de Jong, P.J Construction and characterization of an eightfold redundant dog genomic bacterial artificial chromosome library. Genomics 58: Makinen, A., Zijlstra, C., de Haan, N.A., Mellink, C.H. and Bosma, A.A Localization of 18S + 28S and 5S ribosomal RNA genes in the dog by fluorescence in situ hybridization. Cytogenet Cell Genet. 78: Miller, S.A., Dykes, D.D., Polesky, H.F A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16: Murphy, R.T., Mogensen, J., Shaw, A., Kubo, T., Hughes, S., McKenna, W.J Novel mutation in cardiac troponin I in recessive idiopathic dilated cardiomyopathy. Lancet 363: Nigro, V., Okazaki, Y., Belsito, A., Piluso, G., Matsuda, Y., Politano, L., Nigro, G., Ventura, C., Abbondanza, C., Molinari, A.M., Acampora, D., Nishimura, M., Hayashizaki, Y., Puca, G.A Identification of the Syrian hamster cardiomyopathy gene. Hum Mol Genet. 6: Nigro, V., Piluso, G., Belsito, A., Politano, L., Puca, A.A., Papparella, S., Rossi, E., Viglietto, G., Esposito, M.G., Abbondanza, C., et al Identification of a novel sarcoglycan gene at 5q33 encoding a sarcolemmal 35 kda glycoprotein. Hum Mol Genet. 5: Ostrander, E.A., Comstock, K.E The domestic dog genome. Curr Biol. 14: R Ostrander, E.A., Galibert, F., Patterson, D.F Canine genetics comes of age. Trends Genet. 16: Patterson, D.F Companion animal medicine in the age of medical genetics. J Vet Intern Med. 14: 1-9. Richardson, P., McKenna, W., Bristow, M., Maisch, B., Mautner, B., O'Connell, J., Olsen, E., Thiene, G., Goodwin, J., Gyarfas, I., Martin, I., Nordet, P Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the Definition and Classification of cardiomyopathies. Circulation 93: Rozen, S. and Skaletsky, H.J Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics Methods and Protocols: Methods in Molecular Biology. Humana Press, Totowa, NJ Sakamoto, A., Ono, K., Abe, M., Jasmin, G., Eki, T., Murakami, Y., Masaki, T., Toyooka, T., Hanaoka, F Both hypertrophic and dilated cardiomyopathies are caused by mutation of the same gene, delta-sarcoglycan, in hamster: an animal model of disrupted dystrophin-associated glycoprotein complex. Proc Natl Acad Sci U S A 94: Sambrook, J., Russel, D.W Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 3rd ed. Schmitt, J.P., Kamisago, M., Asahi, M. et al Dilated cardiomyopathy and heart failure caused by a mutation in phospholamban. Science 299: Sisson, D.D. and Thomas, W.P Myocardial diseases. In: Ettinger SJ and Feldman EC, eds. Textbook of veterinary internal medicine. 4 th ed. Philadelphia: WB Sounders Co Switonski, M., Reimann, N., Bosma, A.A., Long, S., Bartnitzke, S., Pienkowska, A., Moreno-Milan, M.M. and Fischer, P Report on the progress of standardization of the G-banded canine (Canis familiaris) karyotype. Committee for the Standardized Karyotype of the Dog (Canis familiaris). Chromosome Res. 4:

119 Canine sarcoglycan- δ gene Thomas, W.P., Gaber, C.E., Jacobs, G.J., Kaplan, P.M., Lombard, C.W., Moise, N.S., Moses, B.L Recommendations for standards in transthoracic two-dimensional echocardiography in the dog and cat. Echocardiography Committee of the Specialty of Cardiology, American College of Veterinary Internal Medicine. J Vet Intern Med. 7: Towbin, J.A., Bowles, N.E Genetic abnormalities responsible for dilated cardiomyopathy. Curr Cardiol Rep. 2: Tsubata, S., Bowles, K.R., Vatta, M., Zintz, C., Titus, J., Muhonen, L., Bowles, N.E., Towbin, J.A Mutations in the human delta-sarcoglycan gene in familial and sporadic dilated cardiomyopathy. J Clin Invest. 106: van den Berg, L., Imholz, S., Versteeg, S.A., Leegwater, P.A., Zijlstra, C., Bosma, A.A., van Oost, B.A Isolation and characterization of the canine serotonin receptor 1B gene (htr1b). Gene 326: van de Sluis, B., Nanji, M.S., Breen, M., Pearson, P.L., Oost, B.A., Cox, D.W., Wijmenga, C Characterization and chromosomal localization of five canine ATOX1 pseudogenes. Cytogenet Cell Genet. 93: Vatta, M., Mohapatra, B., Jimenez, S., Sanchez, X., Faulkner, G., Perles, Z., Sinagra, G., Lin, J.H., Vu, T.M., Zhou, Q., et al Mutations in Cypher/ZASP in patients with dilated cardiomyopathy and left ventricular non-compaction. J Am Coll Cardiol. 42: Yang, F., O'Brien, P.C., Milne, B.S., Graphodatsky, A.S., Solanky, N., Trifonov, V., Rens, W., Sargan, D., Ferguson-Smith, M.A A complete comparative chromosome map for the dog, red fox, and human and its integration with canine genetic maps. Genomics 62: Zijlstra, C., Mellink, C.H., de Haan, N.A., Bosma, A.A Localization of the 18S, 5.8S and 28S rrna genes and the 5S rrna genes in the babirusa and the white-lipped peccary. Cytogenet Cell Genet. 77:

120

121 Chapter 6 Evaluation of the phospholamban gene in purebred large-breed dogs with dilated cardiomyopathy Polona Stabej, Peter A. Leegwater, Arnold A. Stokhof, Aleksandra Domanjko-Petrič, Bernard A. van Oost American Journal of Veterinary Research; in press

122 Chapter 6 ABSTRACT The objective of this study was to evaluate the role of the phospholamban gene in purebred large-breed dogs with dilated cardiomyopathy (DCM). Six dogs with DCM including 2 Dobermanns, 2 Newfoundlands, and 2 Great Danes. All dogs had clinical signs of congestive heart failure and a diagnosis of DCM was made on the basis of echocardiographic findings. Blood samples were collected from each dog and genomic DNA was isolated by a salt extraction method. Specific oligonucleotides were designed to amplify the promoter, exon 1, the 5 -part of exon 2 including the complete coding region, and part of intron 1 of the canine phospholamban gene via polymerase chain reaction procedures. These regions were screened for mutations in DNA obtained from the 6 dogs with DCM. No mutations were identified in the promoter, 5 untranslated region, part of intron 1, part of the 3 untranslated region, and the complete coding region of the phospholamban gene in dogs with DCM. Results indicate that mutations in the phospholamban gene are not a frequent cause of DCM in Dobermanns, Newfoundlands, and Great Danes. 122

123 INTRODUCTION Evaluation of the phospholamban gene Dilated cardiomyopathy (DCM) is a myocardial disease that is an important cause of congestive heart failure and sudden death in dogs. The disease primarily affects large and giant breeds (Sisson and Thomas 1995). Familial disease in dogs has been identified in Dobermanns, Boxers, English Cocker Spaniels, Irish Wolfhounds, Portuguese Water Dogs, Great Danes, and Newfoundlands (Staaden 1981; Goodwin and Cattiny 1995; Brownlie and Cobb 1999; Meurs et al. 2001; Dambach et al. 1999; Domanjko-Petrič et al. 2002; Dukes-McEwan and Jackson 2002). Causal mutations have not yet been identified. The high prevalence of DCM in specific breeds suggests a genetic background. Within each breed, DCM has unique characteristics and between breeds, it is probably a genetically heterogeneous disease. In Great Danes and Newfoundlands, notable clinical characteristics of DCM include ventricular dilatation, congestive heart failure (left-sided or biventricular), and atrial fibrillation, whereas in Dobermanns, DCM is associated with left-sided heart failure, ventricular arrhythmias, and sudden death (Tidholm and Johnson 1996; Calvert et al. 1997; Meurs et al. 2001). Although clinical signs of DCM usually become evident in adulthood, puppies can be affected within first weeks or months after birth, as has been described in Portuguese Water Dogs (Dambach et al. 1999) and more recently in a litter of Dobermanns (Vollmar et al. 2003). However, DCM in Dobermanns typically has a late onset (Calvert et al. 1997; Domanjko-Petrič et al. 2002). In several breeds, a preclinical phase of DCM has been recognized (Calvert et al. 1997; Meurs et al. 2001; Dukes-McEwan et al. 2003). In dogs, the diagnosis of DCM is not problematic once clinical signs are apparent, but diagnosis of the preclinical phase of DCM may be challenging. In a recent publication (Dukes-McEwan et al. 2003), application of a scoring system for the identification of dogs in preclinical stages of DCM highlighted the problem of false-positive and false-negative diagnosis of preclinical DCM in dogs without clinical signs. Difficulty in detection of preclinical DCM in dogs, coupled with late onset of the disease, frustrates early diagnosis and removal of disease carriers from the breeding population. Dilated cardiomyopathy in humans closely resembles the canine form of the disease and has a genetic background in as many as 35% of all cases (Grunig et al. 1998). To date, 15 genes have been implicated as the cause of DCM in humans (Fatkin and Graham 2002; Knoll et al. 2002; Schmitt et al. 2003). Most of these genes code for proteins of the cytoskeleton and destabilize the cardiomyocyte membrane or cytoskeleton via mechanical instability or force transduction, resulting in poor systolic cardiac function (Towbin and Bowles 2002). In contrast to genes encoding cytoskeletal proteins, humans with an Arg-to-Cys missense 123

124 Chapter 6 mutation at residue 9 (R9C) in the phospholamban protein develop dominantly inherited DCM through a mechanism involving direct disturbance of the myocellular Ca 2+ metabolism (Schmitt et al. 2003). Phospholamban is a small transmembrane phosphoprotein of 52 amino acids that plays an important role in cardiac contraction and relaxation. Cardiac contraction occurs with Ca 2+ release from the sarcoplasmic reticulum into the cytosol and relaxation begins with Ca 2+ uptake into the sarcoplasmic reticulum through the Ca 2+ -ATP (SERCA2a) pump. Whereas the wild-type phospholamban protein directly inhibits the SERCA2a pump, the phospholamban R9C protein fails to do so. The consequence is delayed decay of calcium transport in cardiac myocytes that initiates DCM in humans (Schmitt et al. 2003). The canine phospholamban gene (PLN) consists of 2 exons that are transcribed into phospholamban mrna. The exon 1 RNA sequence of 83 nucleotides and the initial RNA sequence of 101 nucleotides of exon 2 are not translated into the phospholamban protein and represent the 5 untranslated region (UTR). The 159-bp coding sequence starts at nucleotide position 102 and ends at position 260 of exon 2 (Figure 1; McTiernan et al. 1999a). The part of exon 2 after position 260 that is not translated into the protein is the 3 UTR. Gene transcription is regulated by the promoter region that has been well defined for PLN (McTiernan et al. 1999b). The purpose of the study reported here was to evaluate the role of the PLN in purebred large-breed dogs with DCM. To achieve this goal, we sequenced the promoter, the 5 UTR, the border regions of intron 1 at which splicing occurs, part of the 3 UTR, and the complete coding region of the PLN in 6 large-breed purebred dogs with DCM. MATERIALS AND METHODS DOGS Six dogs were included in the study (2 Dobermanns, 2 Newfoundlands, and 2 Great Danes). All dogs were client-owned dogs with DCM that were evaluated at the Department of Clinical Sciences of Companion Animals of Utrecht University and at the Clinic for Small Animals and Surgery of Ljubljana University. An informed consent for DNA to be collected and analyzed was obtained from the owners. Blood samples were collected from the Dobermanns (DO1 and DO2), Newfoundlands (NF1 and NF2), and Great Danes (GD1 and GD2) and genomic DNA was isolated by the salt extraction method (Miller et al. 1988). Since DCM is probably an inherited disorder in Dobermann Pinschers, Newfoundlands and Great Danes, it is highly likely that the cause of DCM is also genetic in the dogs included in this study (Meurs et al. 2001; Domanjko-Petrič et 124

125 Evaluation of the phospholamban gene al. 2002; Dukes-McEwan and Jackson 2002). Diagnosis of DCM was made on the basis of clinical or radiographic signs of congestive heart failure and echocardiographic evidence of DCM (marked left ventricular chamber dilatation and decreased fractional shortening) in the absence of other congestive heart failure-related lesions. The echocardiographic measurements were compared to breed-specific reference values (Calvert et al. 1982; Koch et al. 1996; Dukes- McEwan 1999). ECHOCARDIOGRAPHIC AND ECG EVALUATIONS After positioning each conscious dog in right lateral recumbency, echocardiography was performed at the Department of Clinical Sciences of Companion Animals of Utrecht University by use of a high definition ultrasound system (HDI 3000, Advanced Technology Laboratories, Woerden, The Netherlands) equipped with a 5- to 3-MHz broad and phase array transducer. For simultaneous ECG recording, ECG electrodes were placed on the left and right forelimb and the left hind limb. All measurements were performed by use of a trackball-driven cursor and ultrasound software. From the right parasternal approach, 2- dimensional M-mode tracings were obtained for measurements of the aortic root and left atrium diameters and during diastole and systole for measurements of the interventricular septum, the left ventricular dimension, and the left ventricular free wall. These measurements were made from the leading edge of the first endocardial surface to the leading edge of the second endocardial surface. Diastolic measurements were made at the onset of the QRS complex of the ECG tracing and systolic measurements were made at the maximum systolic excursion of the interventricular septum. The diameter of the aortic root was measured at the onset of the QRS complex of the ECG tracing and the diameter of the left atrium was measured at its maximal upward excursion near the end of systole. From values of the left ventricular internal dimension in diastole (LVIDd) and the left ventricular internal dimension in systole (LVIDs), the fractional shortening (FS) was calculated by use of the following equation: FS (%) = ([LVIDd LVIDs])/LVIDd) X 100 From the diameter of the left atrium and the aortic root the left atrium-toaorta ratio (LA:Ao) was calculated. At the Clinic for Small Animals and Surgery, Veterinary Faculty, Ljubljana University, echocardiography was performed as described (Domanjko-Petrič et al. 2002). 125

126 Chapter 6 MUTATION SCREENING By use of a canine cdna sequence (GenBank accession no. Y00399) as the query sequence, non-transcribed canine PLN sequences and sequences upstream from the PLN promoter region were derived from the Canis familiaris trace archive of the National Center for Biotechnology Information (NCBI) GenBank via a BLAST search ( Obtained sequences were aligned with the canine cdna using Seqman program from Lasergene software (DNASTAR Inc, Madison, WI). To screen the canine PLN for mutations, we amplified the canine promoter, the 5 UTR, part of intron 1 downstream from exon 1 and upstream from exon 2, the complete translated region, and part of the 3 UTR. Primers (PLN_1, PLN_2, and PLN_3) were designed by use of Primer3 computer software at (Table 1; Rozen and Skaletsky Product lengths were determined. For each polymerase chain reaction (PCR) procedure, 25 ng of genomic DNA was used as a template in a 15 µl reaction mixture consisting of 1x PCR reaction buffer, 200 µm dntps, 1.5 mm MgCl 2, 0.6 U of platinum Taq polymerase, and 0.33 µm of each primer. DNA was initially denatured at 94 C for 4 minutes and then subjected to 35 cycles consisting of 30 s at 94 C, 30 s at the annealing (Table 1) temperature and 30 s at 72 C. The final extension was 10 minutes at 72 C. The reaction was diluted (15x dilution) and 1 µl was used in a 10 µl tercycle reaction using 1 µl cycle sequencing mix with 0.32 µm of the primer and 2 mm MgCl 2. For sequencing of the PLN_1 and PLN_3 PCR products, PLN_1 and PLN_3 reverse primers were used; PLN_2 primers were M13 tailed and M13 forward primer was used for sequencing. The tercycle program consisted of 25 cycles. Each cycle had 3 steps: 30 seconds at 96 C, 15 seconds at 55 C, and 2 minutes at 60 C. Tercycle products were purified by use of multiscreen 96-well filtration plates with gel filtration media. The high-quality sequences obtained were compared with Canis familiaris trace archive sequences and the annotated canine cdna Y00399 sequence by use of computer software (DNASTAR Inc, Madison, WI). Canine sequences were compared with their human counterparts by use of a Blast 2 sequences program. Table 1. Polymerase chain reaction primers used to amplify regions of the canine phospholamban gene Ta ( C ) Product length Primer Forward primer 5'- 3' Reverse primer 5'- 3' PLN_1 AGCAACAGCAGCGACAATAC GTCAAAGATGATGCGACCCT bp PLN_2 TGAGAGAAGGAGGCAAAAGA* CTCTTCATGGGATGGCAGAT* bp PLN_3 GAGTGGTTGAGCTCACATTTG GAAGTGAACTGGTTAGCAGAG bp *Primers were M13 tailed: M13-forward, 5'-GTTTTCCCAGTCACGAC-3' and M13-reverse, 5'- CAGGAAACAGCTATGAC-3' 126

127 RESULTS Evaluation of the phospholamban gene CLINICAL EVALUATION All 6 dogs with DCM had evidence of overt congestive heart failure. Five dogs (the 2 Dobermanns, the 2 Newfoundlands and 1 Great Dane [GD2]) had clinical signs of left heart failure with pulmonary edema; 1 dog (GD1) had clinical signs of right heart failure with ascites and pleural effusion. The ECG recordings were indicative of atrial fibrillation in the Newfoundlands and Great Danes; in the 2 Dobermanns, the ECG recordings revealed widened QRS complexes and P waves, indicative of left ventricular and atrial enlargement. Echocardiographic 2-dimensional M-mode examination revealed marked left ventricular chamber dilatation with reduced shortening fraction and left atrial enlargement with increased left atrium-to-aorta ratio in all dogs (Table 2). Table 2. Echocardiographic findings in 2 Newfoundlands (NF1 and NF2), 2 Dobermanns (DO1 and DO2), and 2 Great Danes (GD1 and GD2) with dilated cardiomyopathy, compared with reference values for each variable in that breed Newfoundland Dobermann Great Dane Reference values a Reference values b, c Reference values d NF1 NF2 Mean ± SD DO1 DO2 Mean ± SD GD1 GD2 Median (range) Weight (kg) NA NA 63,6 68 NA LVIDd (mm) 76* 83* ± * 66* 49.7 ± c 96* 92* 53 (44-59) LVIDs (mm) 69* 70* ± * 56* 33.8 ± c 85* 98* 39.5 (34-45) LVWd (mm) 6* ± * 6* 10.8 ± c (10-16) LVWs (mm) 8* ± * 6,6* 16.2 ± c (11-19) IVSd (mm) 9* ± * 10, ± c (12-16) IVSs (mm) 10* 11* ± * 8.6* 16.5 ± c 22* (14-19) LA (mm) 44* 48* 24,13 +/- 4,06 40* NR 24,8 +/- 4,3 c 57* 66* 33 (28-46) AO (mm) 26* 30 29,18 +/- 2,71 23* NR 28,3 +/- 2,9 c 36* 26* 29.5 (28-34) LA/AO 1,7* 1,6* 0,83 +/- 0,15 1,7* >1,6* 0.86 c 1,6* 2,5* 1.1 ( ) FS (%) 9* 16* ± * 14,46* 25 b 11* 15* 25 (18-36) * Value deviates from the breed specific reference value. LVIDd = Left ventricle internal dimension during diastole. LVIDs = Left ventricle internal dimension during systole. LVWd = Thickness of the wall of the left ventricle during diastole. LVWs = Thickness of the wall of the left ventricle during systole. IVSd = Thickness of the interventricular septum during diastole. IVSs = Thickness of the interventricular septum during systole. FS = Fractional shortening. NA = Not applicable. NR = not on record. a Dukes-Mcewan et al. 1999; b Calvert et al. 1997; c Calvert et al. 1982; d Koch et al

128 Chapter 6 ANALYSIS OF THE CANINE PLN PROMOTER, THE 5 UTR (EXON 1), AND PART OF THE INTRON 1 REGION The genomic structure of canine phospholamban is known (Figure 1; Uyeda et al. 1987; McTiernan et al. 1999a and 1999b). Two high-quality sequences covering regions upstream and downstream from PLN exon 1 were selected from the Canis familiaris trace archive (GenBank Accession No. TI and TI ). The canine PLN promoter region sequence was determined on the basis of previously assessed sequence conservation among 5 orthologous mammalian PLN promoter regions (Uyeda et al. 1987; McTiernan et al. 1999b). In the dog, the PLN regulatory elements were identified in the sequence 200 bp upstream of the canine PLN exon 1 (McTiernan et al. 1999b). The exon 1-intron 1 splice site was based on the canine cdna sequence Y00399 and the presence of the typical splice donor site (GT) in intron 1. A DNA fragment containing 200 bp of the canine PLN promoter and 83 bp of exon 1 had 83 % identity with the human counterpart (GenBank accession no. AF177763). Sequencing of the canine PLN promoter, exon 1, and 155 bp of intron 1 with the PLN_1 reverse primer (Table 1; Figure 2) resulted in a high-quality sequence in the 6 dogs with DCM. The DNA sequence was deposited in the NCBI GenBank (accession No. AY576871). We compared the DNA sequences obtained from dogs with DCM with corresponding Canis familiaris trace file sequences and exon 1 with the annotated canine cdna sequence (Y00399) and found no variations. ANALYSIS OF THE CANINE PLN INTRON 1, 5 UTR (EXON 2), COMPLETE TRANSLATED REGION, AND THE 3 UTR From the NCBI GenBank Canis familiaris trace archive, the DNA sequences TI and TI spanning the region upstream from exon 2 and part of exon 2 were selected and aligned with the canine cdna Y00399 sequence. The beginning of exon 2 was based on the Y00399 DNA sequence and the presence of the universal splice acceptor site (AG) in intron 1. The translated region has been determined by Uyeda et al. (1987) Comparison of the 5 UTR region and the translated region of exon 2 to the same regions in human (NM_002667) revealed 84% and 92 % similarity, respectively. 128

129 Evaluation of the phospholamban gene Promoter Exon 1 Intron Exon bp 83 bp 101 bp 159 bp 2271 bp 5 UTR 5 UTR TR 3 UTR Figure 1. Structure of the canine phospholamban gene (PLN) based on the annotated mrna sequence (GenBank accession No. Y00399). The gene consists of 2 exons, preceded by a 200-bp promoter region. The 159-bp translated region (TR) is contained in exon 2. Several phospholamban mrnas are present in canine cardiac muscle; the 2614 bp transcript presented here is the major one (Uyeda et al. 1987). 1 AGCAACAGCA GCGACAATAC AATGAAAGTT ATTTGACTAA GCCAATAATA TTCTCACTCA 61 TTAAAATTCC ATATATCTAT TTTAGTTCTT CAACATTATC ACCATCAGTA CACAATTATT 121 TCTAAGCCTG AAGATTCATG AATCTTTAAA GGGAGCTTTC TACCACCCCA AACTTTTTTT 181 TTTTTTCATT TATCTCCCAT ACACTTTTAA AAATTACAAG CAAAAAAATG TGGCACAAAG 241 TGTTAATGAC CTTTCCATAC TCAAAGTAAG ATAAGTGACA TTCTAAAAGG TTTGGTTGTG 301 ATAAGACTAT GGCTAACCAA TCATAACTTC AGAATTCTTG TATGACATCA TAAGACCTCC 361 CTAGAACACT TTTCCTCCTA CACCTACTTC AATTGTTGCC ATAAGTCTGG TAACAGAGTC 421 AGAAAACTTT CTAACTAAAC ACCGATAAGA CTTCATACAA CTCACAATAC TTTATATTGT 481 AATCATCACA AGAGCCAAGG TAAGAAATAG ATTTCATTAA TTTTGGTAAT TCTTGTTTTT 541 CAGAATTCAA AATTACTAAA CTTTTTCAAG AGTTAACCAT GTTTTGTAAA TTTCAGTTCA 601 AAGTTAATAA AATGTGTGAT ATGACAAAAA CTTTAATTTC TAAATAAATG CCTGTTTGGT 661 GACAGTAAGA GCAAGACTTA ACTTTGAAAT AATAAGGGTC GCATCATCTT TGAC Figure 2. Sequence of the canine PLN promoter, exon 1, and part of intron 1. Six annotated PLN promoter elements are boxed. 19 Exon 1 sequence is indicted in bold letters; the two PLN_1 primers used to amplify promoter, exon 1, and part of intron 1 are underlined. 129

130 Chapter 6 The DNA fragments amplified with the PLN_2 and PLN_3 PCR primers were sequenced. High-quality DNA sequences from the 6 dogs covered 130 bp of intron 1, 101 bp of the 5 UTR (exon 2), the complete 159-bp translated region (exon 2), and 279 bp of the 3 UTR (Table 2; Figure 3). No nucleotide changes were found in these sequences, compared with the trace archive and Y00399 sequences. The sequence covering the region starting 130 bp prior to exon 2 and ending 279 bp downstream of the stop codon in exon 2, was deposited in the NCBI GenBank (accession No. AY576872). 1 TGAGAGAAGG AGGCAAAAGA TTGAAGTTAT TACACTCCAA ACACAAGATA GATAAGACCT PLN2_forward 61 CATGGTGACC CTCATCAAAA AGGAATACAG AGATAATTTG ATGTCACAAA TACAAATAAT 121 ATTAAGGAAG ATGAATTAAT ACAAATTGTG TTTTTACCTG TTCAGAGAAT AGGTTACACA 181 CATGATCCTA ACCCAGTCAT TATTTTTATA TTCCAGGCTA CCTAAAAGAA GAGAGTGGTT PLN3_forward 241 GAGCTCACAT TTGGCCGCCA GCTTTTTACC TTTCTCTTCA CCATTTAAAA CTTGAGACTT 301 CCTGCTTTCC TGGGGTCATG GATAAAGTCC AATACCTCAC TCGCTCTGCT ATTAGAAGAG M D K V Q Y L T R S A I R R 361 CTTCAACCAT TGAAATGCCT CAACAAGCAC GTCAAAATCT TCAGAACCTA TTTATAAATT A S T I E M P Q Q A R Q N L Q N L F I N 421 TCTGTCTCAT TTTAATATGT CTCTTGTTGA TCTGCATCAT TGTGATGCTT CTCTGAAGTT F C L I L I C L L L I C I I V M L L GCTGCAAT CTCCAGTGAT GCAACTTGTC ACCATCAACT TAATATCTGC CATCCCATGA PLN2_reverse 541 AGAGGGGAAA ATAATACTAT ATAACAGACC ACTTCTAAGT AGAAGATTTT ACTTGTGAAA 601 AGGTCAAGAT TCAGAACAAA AGAAATTATT AACAAATGTC TTCATCTGTG GGATTTTGTA 661 AACATGAAAA GAGCTTTATT TTCAAAAATT AACTTCAAAA TGACTATAGG TGCGCATAAT 721 GTAATTGCTG AATTCCTCAA CAAAGCTTGT AAAAGTTTCT ATGCCAAATT TTTTCTGAGG 781 GTAAAGTAGG AGTTTAGTTT TAAAACTGCT CTGCTAACCA GTTCACTTC PLN3_reverse Figure 3. Sequence of part of intron 1 and part of exon 2 of the canine PLN. The PLN_2 primers (used to amplify the sequence upstream from exon 2 and 5 untranslated region [UTR] of exon 2) and PLN_3 primers (used to amplify the translated and 3 UTR of exon 2) are underlined. The exon 2 sequence is indicated in bold letters; the coding sequence starts in exon 2 at position 318 with an ATG coding for methionine and ends at position 476 with a TGA stop codon. The 52 amino acids are presented below the nucleotide sequence. 130

131 DISCUSSION Evaluation of the phospholamban gene Because of the striking phenotypic similarity of DCM in humans and dogs, the contribution of genetic mutations to the etiology of DCM in humans has been recognized in research efforts regarding DCM in dogs. It is now clear that DCM is a genetic disease at least in some breeds of dog, although no causal mutation has yet been discovered. Each breed is a genetic isolate; therefore, it is not surprising that the clinical manifestations of DCM are breed specific. Consequently, the disease could be genetically homogeneous within a breed and genetically heterogeneous among different breeds. However, 2 DCM phenotypes have been identified in Dobermanns: sudden death and development of congestive heart failure. Different mutations in the same gene or even different genes might cause these 2 variable disease phenotypes. In humans, most DCM mutations were found in genes coding for proteins of the cell cytoskeleton, and DCM was described as a cytoskeletalopathy (Bowles et al. 2000). The discovery of the PLN R9C mutation as a cause of DCM in humans suggested another pathway to the development of DCM a pathway in which Ca 2+ kinetics are disrupted. It is likely that the pathophysiological mechanisms involved in the development of DCM in dogs are the same as those in humans. Two characteristics of DCM in Dobermanns make PLN an especially attractive candidate gene. These characteristics are ventricular arrhythmias and sudden death as a result of ventricular tachycardia and fibrillation (Calvert et al. 1997; Calvert et al. 2000). Mutations in genes that result in disturbed ion transport in cells have been identified as causes of arrhythmias and sudden death in humans (Roberts and Brugada 2003). Therefore, genes coding for ion channels could well play a role in the development of DCM in Dobermanns and are also good candidate genes An important resource which became recently available from the canine genome project is the 7x redundant dog sequence deposited in the NCBI GenBank trace archive. Together with the canine phospholamban cdna sequence, it enabled rapid reconstruction and evaluation of the PLN in dogs (Uyeda et al. 1987). The sequence of the canine PLN promoter and exon 1 which has been annotated in the NCBI GenBank (accession No. AF037348) had several nucleotide differences, compared with the sequence determined in the present study or the trace archive sequences. In addition, the intron-exon borders in the AF sequence could not be matched with the canine phospholamban cdna sequence Y This was probably due to artifacts in the AF sequence. Therefore, we deposited our sequence of the promoter, exon 1, and part of intron 1 in GenBank under accession No. AY

132 Chapter 6 Another valuable tool in canine genetics is the canine radiation hybrid map (Guyon et al. 2003). Several genetic diseases in dogs have been mapped by use of microsatellite markers and linkage analysis (Acland et al. 1999). The disadvantage of linkage analysis is the requirement of fairly complete family material and informative markers. Because PLN is a small gene, direct sequencing of affected dogs is a more efficient way of evaluation. In the present investigation, the promoter, the complete 5 UTR, the exon1 intron1 and the intron1 exon2 splice sites, the complete coding region (159 bp), and part of 3 UTR of the PLN in Dobermanns, Newfoundlands, and Great Danes with DCM were sequenced. Our data have indicated that there were no mutations in those sequences. Therefore, in our opinion, the PLN is excluded with a high level of confidence as a cause of DCM in these 3 breeds of dog. 132

133 REFERENCES Evaluation of the phospholamban gene Acland, G.M., Ray, K., Mellersh, C.S., et al A novel retinal degeneration locus identified by linkage and comparative mapping of canine early retinal degeneration. Genomics 59: Bowles, N.E., Bowles, K.R., Towbin, J.A The "final common pathway" hypothesis and inherited cardiovascular disease. The role of cytoskeletal proteins in dilated cardiomyopathy. Herz 25: Brownlie, S.E., Cobb, M.A Observations on the development of congestive heart failure in Irish wolfhounds with dilated cardiomyopathy. J Small Anim Pract. 40: Calvert, C.A., Chapman, W.L., Toal, R.L Congestive cardiomyopathy in Doberman Pinscher dogs. J Am Vet Med Assoc. 181: Calvert, C.A., Hall, G., Jacobs, G., et al Clinical and pathologic findings in Dobermanns with occult cardiomyopathy that died suddenly or developed congestive heart failure: 54 cases ( ). J Am Vet Med Assoc. 210: Calvert, C.A., Jacobs, G.J., Smith, D.D., et al Association between results of ambulatory electrocardiography and development of cardiomyopathy during long-term follow-up of Dobermanns. J Am Vet Med Assoc. 216: Dambach, D.M., Lannon, A., Sleeper, M.M., et al Familial dilated cardiomyopathy of young Portuguese water dogs. J Vet Intern Med. 13: Domanjko-Petrič, A., Stabej, P., Žemva, A Dilated cardiomyopathy in Dobermanns, survival, causes of death and pedigree review in a related line. J Vet Cardiol. 4: Dukes-McEwan J Echocardiographic / Doppler criteria of normality, the findings in cardiac disease and the genetics of familial dilated cardiomyopathy in Newfoundland dogs. PhD thesis. Department of Veterinary Clinical Studies. University of Edinburgh. Dukes-McEwan, J., Borgarelli, M., Tidholm, A., et al Proposed guidelines for the diagnosis of canine idiopathic dilated cardiomyopathy. Journal of Veterinary Cardiology. 5: Dukes-McEwan, J., Jackson, I.J The promises and problems of linkage analysis by using the current canine genome map. Mamm Genome. 13: Fatkin, D., Graham, R Molecular mechanisms of inherited cardiomyopathies. Physiol Rev. 82: Goodwin, J.K., Cattiny, G Further characterization of Boxer cardiomyopathy, in Proceedings. 13th ACVIM Forum; Grunig, E., Tasman, J.A., Kucherer, H., et al Frequency and phenotypes of familial dilated cardiomyopathy. Am Coll Cardiol. 31: Guyon, R., Lorentzen, T.D., Hitte, C., et al A 1-Mb resolution radiation hybrid map of the canine genome. Proc Natl Acad Sci U S A. 100: Knöll, R., Hoshijima, M., Hoffman, H.M., et al The cardiac mechanical stretch sensor machinery involves a Z disc complex that is defective in a subset of human dilated cardiomyopathy. Cell 111: Koch, J., Pedersen, H.D., Jensen, A.L., et al M-mode echocardiographic diagnosis of dilated cardiomyopathy in giant breed dogs. Zentralbl Veterinarmed A. 43:

134 Chapter 6 McTiernan, C.F., Frye, C.S., Lemster, B.H., et al. 1999a. The human phospholamban gene: structure and expression. J Mol Cell Cardiol. 31: McTiernan, C.F., Lemster, B.H., Frye, C.S., et al. 1999b. Characterization of proximal transcription regulatory elements in the rat phospholamban promoter. J Mol Cell Cardiol. 31: Meurs, K.M., Miller, M.W., Wright, N.A Clinical features of dilated cardiomyopathy in Great Danes and results of a pedigree analysis: 17 cases ( ). J Am Vet Med Assoc. 218: Miller, S.A., Dykes, D.D., Polesky, H.F A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16: Roberts, R., Brugada, R Genetics and arrhythmias. Annu Rev Med. 54: Rozen, S., Skaletsky, H.J. Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics methods and protocols: methods in molecular biology. Totowa, NJ: Humana Press, 2000: Schmitt, J.P., Kamisago, M., Asahi, M., et al Dilated cardiomyopathy and heart failure caused by a mutation in phospholamban. Science 299: Sisson, D.D., Thomas, W.P. Myocardial diseases In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. 4th ed. Philadelphia: WB Saunders Co, Staaden, R.V Cardiomyopathy of English cocker spaniels. J Am Vet Med Assoc. 178: Tidholm, A., Jonsson, L Dilated cardiomyopathy in the Newfoundland: a study of 37 cases ( ). J Am Anim Hosp Assoc. 32: Towbin, J.A., Bowles, N.E Genetic abnormalities responsible for dilated cardiomyopathy. Curr Cardiol Rep. 2: Uyeda, A., Kitano, K., Fujii, J., et al The cdna sequence of the major phospholamban mrna in canine cardiac ventricular muscle. Nucleic Acids Res. 15: Vollmar, A., Fox, P.R., Meurs, K.M., et al Dilated cardiomyopathy in juvenile Dobermann Pinscher. Journal of Veterinary Cardiology 5:

135

136

137 Chapter 7 Genetic epidemiological studies of DCM in the Dobermann dog point to a crucial role of titin in DCM susceptibility Polona Stabej, Peter A.J. Leegwater, Sandra Imholz, Manon Loohuis, Paul J.J. Mandigers, Arnold A. Stokhof, Aleksandra Domanjko-Petrič, Bernard A. van Oost Manuscript in preparation

138 Chapter 7 ABSTRACT Dilated cardiomyopathy (DCM) is a disease of the myocardium associated with dilatation and impaired contraction of the ventricles. In the dog, it primarily affects large and giant breeds with the Dobermann being one of the most frequently affected. The high prevalence of DCM in specific breeds suggests a genetic background, but causal mutations have not yet been identified in dogs. The gene encoding the giant myofilament protein titin (TTN) was one of the candidate genes that we evaluated as part of our study of the genetic aetiology of DCM in Dobermanns. The diagnosis of DCM was based on clinical and/or radiographic symptoms of congestive heart failure (CHF) and was confirmed in each case by two-dimensional echocardiography. Genotyping of TTN marker REN252E18 revealed a strong partition of TTN alleles between the DCM and DCM non-affected groups. The main difference between DCM and DCM nonaffected Dobermanns is the enrichment for one allele in the DCM non-affected group. The same applies to the haplotypes defined by genotypes of three TTN markers. This result suggests that the titin allele over represented in the DCM non-affected group, is either the allele that is not linked to a pathological mutation or confers protection against DCM in the presence of a mutation elsewhere in the genome. Hypotheses explaining the possible role of TTN in the DCM in the Dobermann dogs will be discussed. 138

139 The titin gene and DCM INTRODUCTION Dilated cardiomyopathy (DCM) is a disorder characterized by ventricular dilatation and systolic contractile dysfunction. In human, inherited gene defects account for approximately 35% of cases (Grunig et al. 1998). DCM in dogs closely resembles human DCM. The genetic component in canine DCM has been warranted because DCM affects mainly dogs of particular breeds and the clinical manifestation of the disease is characteristic for each of these breeds. The Dobermann is reportedly one of the most commonly affected breeds (Calvert et al. 1986). In the Dobermann, the disease has an adult onset and the clinical course is very malignant with death occurring within 12 weeks after diagnosis. The estimates of the DCM frequency in the Dobermann breed are between 30-40% in North-America (Calvert et al. 2000; O Grady et al. 1997) and 20% in Europe. Mutations in several genes encoding proteins of the cell cytoskeleton, ion channels or signalling molecules have been identified as a cause of DCM in human (Liew and Dzau 2004). Linkage analyses and candidate gene studies attempting to elucidate the genetic origin of canine DCM have been fruitless thus far (Dukes-McEwan and Jackson 2002; Jakobs et al. 2004; Stabej et al. 2004). In the case of Dobermann DCM, we have evaluated and excluded the genes encoding alpha-tropomyosin, vinculin, desmin, sarcoglycan-delta and phospholamban (Stabej et al. 2004; Stabej et al. 2005a; Stabej et al. 2005b) and unpublished data), while the actin gene was excluded by other researchers (Meurs et al. 2001). Mutations in the giant muscle filament titin have been shown to cause an autosomal dominant form of DCM in man (Gerull et al. 2002; Itoh-Satoh et al. 2002). We wondered therefore whether polymorphisms in the titin gene are associated with DCM in Dobermanns. The canine titin gene and the polymorphic microsatellite marker REN252E18 have been mapped to the same location on chromosome CFA36 (Guyon et al. 2003). Significant differences of REN252E18 allele frequencies between groups of DCM, DCM-free and randomly selected Dobermanns were found. Additional markers upstream and in the titin gene were identified and tested for association with the DCM. TTN haplotypes across a 239 kb region were defined and significant differences in haplotypes were observed between the three groups. Several hypotheses explaining the association of the TTN haplotypes with DCM in the Dobermann will be discussed. 139

140 Chapter 7 MATERIALS AND METHODS ANIMALS AND DNA SAMPLES Blood samples of 26 Dobermanns diagnosed with DCM, 13 DCM-free and 92 randomly selected Dobermanns were collected. The DCM Dobermanns were patients seen by the Department of Clinical Sciences of Companion Animals (Veterinary Faculty, Utrecht University) and by the Clinic for Small Animals and Surgery (Veterinary Faculty, Ljubljana University) between 1993 and The DCM-free Dobermanns were at least 9 years old and were selected from a larger group of dogs seen by the same clinics at preventive check ups. Ninety-two Dobermanns were randomly selected from the pedigree book of Dutch Dobermanns in year The pedigrees of 11 DCM and 9 DCM-free dogs were collected. Genomic DNA was isolated by the salt extraction method (Miller et al. 1988). DCM diagnosis was based on clinical and/or radiographic symptoms of congestive heart failure (CHF) and echocardiographic evidence of DCM in the absence of other CHF related lesions on two-dimensional echocardiography. Echocardiography was performed in conscious dogs, with the dog in right lateral recumbency according to standard procedures (Thomas et al. 1993). DCM was defined as left ventricular internal diastolic dimension >50 mm, left ventricular internal systolic dimension >38 mm and left ventricular fractional shortening (FS) < 25 % (Calvert et al. 2000). Dobermanns were confirmed as unaffected when they had no signs or symptoms indicative of DCM at an age of nine years or older. CANINE TTN MARKERS MICROSATELLITE MARKERS FROM THE GENOME MAP The REN252E18 microsatellite marker maps to the same position as the titin gene (Guyon et al. 2003). Upon availability of the dog genome trace sequences deposited in the C.familiaris trace archive ( a contiguous sequence of 1,600 kb was assembled around the REN252E18 marker. The 1,6 kb contiguous sequence was aligned with the human human TTN DNA sequence AJ27789 using BLAST2Seqences software at The exact position of the marker in relation to TTN exons could be determined. 140

141 MARKERS SELECTED FROM THE CANINE TTN DNA SEQUENCE The titin gene and DCM At the time of this study, raw trace files of DNA sequences had been deposited in the Canis familiaris Trace archive of the NCBI. We assembled the canine TTN sequences of 333 out of 363 exons from the Canis familiaris trace sequences. Human TTN exons (obtained from the sequence AJ277892) were used as a query for a BLASTN search ( In order to analyze the DNA sequence of the exons and the bordering intronic DNA sequences, primer sets were developed using Primer3 (Rozen and Skaletsky 2000) and PrimerDesigner (version 2.2) software. All primers (Table 1) were tailed with M13 sequencing tags and ordered in 96-wells microtiter plates (Illumina, Inc). Some exons and exon/intron borders were sequenced in four Dobermanns as described below. Three of the markers (two deletions in introns 18 and 345 and one SNP in intron 18) identified in the TTN sequencing process were typed in the DCM and DCM-free Dobermanns. TYPING OF THE MARKERS AND ASSOCIATION ANALYSIS MICROSATELLITE MARKERS AND DELETIONS For each PCR reaction, 25 ng of canine genomic DNA was amplified with 0.33 µm of each primer, 0.6 U of Platinum Taq (Invitrogen), 200 µm dntps, 1x Gibco buffer and 1.5 mm MgCl 2 in a final volume of 15 µl. The PCR program consisted of a denaturation step of 10 min at 94 C, followed by 35 cycles of 30 sec 94 C, 30 sec at Ta, 30 sec 72 C and a final extension at 72 C for 10 min. The primer sequences, annealing temperatures and product lengths are listed in Table 1. The forward primers of the FH2998, REN252E18, Int18-del, and Int345-del were labelled with FAM fluorescent dye. The PCR reactions were diluted 10-20x, 1 µl of the dilution was mixed with 10 µl formamide and µl of size standard Table 1. Oligonucleotide primer sequences used in this study Marker name Forward primer 5' - 3' Reverse primer 5' - 3' Product length REN252E18 1 CAGCATTTCCTCACTTTCCC GGGGAGATTGTGTATCGGAA 256 bp 58 Int18-SNP TGCCTTAGGAACTTGGAAGC 2 TGACCTCTGCTTAGCACGAT bp 55 Int18-del TTACCTGCAGTGCTGTGAAC TGACCTCTGCTTAGCACGAT 223 bp 58 Ex345-del ATGTGAAGGGTATGAGTGAA GTGTCTT CAAGCTCTAGGTCTGGTATT bp 55 1 Guyon et al The primers were M13 tailed. M13_forward (5 -GTTTTCCCAGTCACGAC-3 ); M13_reverse (5 - CAGGAAACAGCTATGAC-3 ) 3 The tail sequence that drives addition of an Adenosine residue at the 3'-end of the complementary DNA strand is indicated in italics. Ta ( C) 141

142 Chapter 7 GS-500-TAMRA (Applied Biosystems, CA) and analysed on the ABI 3100 Genetic Analyzer (Applied Biosystems, CA). Genescan 3.1 software was used for genotype assessment. SNP TYPING The Int18-SNP was typed by DNA sequence analysis. PCR reactions were executed as described above. The reaction was diluted 15x and 1 µl was used in a 10 µl tercycle reaction using 1 µl Big Dye Terminator Ready Reaction Kit (Perkin Elmer ABI), 0.32 µm of the HPLC purified M13 forward primer in 1x sequence buffer (80 mm Tris, 2 mm MgCl 2, ph 9.0). The tercycle consisted of 25 cycles of 30 s at 96 C, 15 s at 55 C, 2 min at 60 C. Tercycle products were purified using multiscreen 96-well filtration plates (Millipore) with Sephadex G-50 (Amersham). The DNA sequences were aligned using Seqman (DNA Star Software) and the SNP was typed by visual examination of electropherographs. The position of markers in relation to the titin gene exons was determined by alignment of the canine TTN contig DNA sequences with the human TTN exons using SeqMan (DNAStar alignment software). The REN252E18 marker was typed in 26 DCM, 13 DCM-free and 92 randomly selected Dobermanns. The markers Int18-SNP, Int18-del and Int345- del and Int354[CA] were typed in 26 DCM and 13 DCM-free Dobermanns. DEFINITION OF TITIN HAPLOTYPES Haplotypes were constructed for the markers Int18-SNP, Int18-del and Int345-del in the DCM and DCM-free Dobermanns. Firstly, the haplotypes in dogs homozygous for all four markers were selected; followed by putative haplotypes derived from the heterozygous dogs in a way that a minimal number of haplotypes was constructed. RESULTS CLINICAL EVALUATION All 26 DCM dogs were symptomatic with evidence of congestive heart failure. Echocardiographic 2D/M mode examination demonstrated marked left ventricular chamber dilatation with reduced shortening fraction as determined by echocardiography. Evaluation of available pedigrees showed that none of the affected dogs shared grandparents and the same applied to the unaffected dogs. Based on the 142

143 The titin gene and DCM birth dates, none of the dogs within a group were full siblings. Two DCM dogs had a sibling in the DCM-free group. In the DCM and DCM-free group 46,15% and 61,54% of dogs were patients of the Ljubljana University, respectively. ASSOCIATION OF TTN MARKERS WITH DCM MARKERS FROM THE DOG GENOME MAP The canine titin gene and the polymorphic microsatellite marker REN252E18 have been mapped to the same location on chromosome CFA36 (Guyon et al. 2003). Upon the availability of the dog genome assembly, we determined the position of the REN252E18 marker in intron 10 of the canine TTN gene, 16.3 kb downstream of exon 1. The nomenclature of exons and introns is according to that of the human reference gene of Genbank accession AJ The REN252E18 marker was typed in 26 DCM and 13 DCM-free Dobermanns (Table 2). A significant difference in allele frequencies of REN252E18 between the groups was observed (Chi-square test P = 0.002). To confirm the apparent association of the marker alleles to DCM, the allele frequencies in the two groups were also compared with those in a large group of 92 randomly selected Dobermanns. While a significant difference in the allele frequency was observed between the affected group and the random group (P = 0.005), no significant difference was detected between the random and the DCM-free group. When analyzing the distribution of REN252E18 alleles in the three groups, two main characteristics were observed: Firstly, allele 256 was found with a high frequency of 52% in the DCM group (Table 2). The frequency of this allele in the random and non-affected groups were 33% and 16%, respectively. Secondly, allele 260 was found with high frequency in the non-affected and random groups with a high number of dogs homozygous for this allele (Table 2). The frequencies of dogs homozygous for the 260 allele in the DCM, DCM-free and random groups were 8%, 61% and 32%, respectively (Figure 1). The differences between DCM/Random and DCM/DCM-free groups were with P values of 0.03 and significant. The difference between the random and DCM-free groups was not significant. A closer look at the genotypes of the DCM dogs nos. 3 and 4 and their siblings numbers 33 and 32, respectively are in accordance with the association of 260 allele with the DCM-free phenotype and the 256 allele with the DCM phenotype (Table 2). 143

144 Chapter 7 Table 2. REN252E18 genotypes in the DCM (n = 26) and DCM-free (n = 13) groups REN252E18 1 Status ID DCM DCM free

145 The titin gene and DCM MARKERS IN THE CANINE TITIN GENE IDENTIFIED BY SEQUENCING To further explore TTN for association with the DCM, additional markers along the titin gene were typed in the DCM, DCM-free and random groups of Dobermanns. The genotyping data of three additional markers in TTN show a similar picture to the one observed for REN252E18 (Table 3). Table 3. The titin gene haplotypes in DCM dogs and DCM-free dogs REN252E18 Ex18-SNP Ex18-del Ex345-del Haplotype 2 Exon 10 Exon 18 Exon 345 Status ID C C L S S L C C S S S S C C L S S S C G L L S S C C S S S S C C S S S S C C S S S S C C S S S S C C S S S L C C S L S L C C S S S L C G S L S L DCM C G S L S L C G S L S L C G S L S L G G L L L L C G S L L L G G L L S L G G L L S L G G L L S L G G L L L L G G L L L L C G S L S L C G S L L L C G S L L L C G S L L L C C S L S S C G S L S L G G L L S L C G L L S L G G L L S L G G L L L L DCM free G G L L L L G G L L L L G G L L L L G G L L L L G G L L L L G G L L L L G G L L L L 1 The location of the marker in TTN in relation to the nearest exon. +/- n indicates the marker lies n bp upstream/downstream of the exon noted in the second row. 2 Haplotype 1 is depicted in gray. In heterozygous dogs, one haplotype is is depicted in bold/italics 145

146 Chapter 7 ASSOCIATION OF HAPLOTYPES IN TTN WITH DCM IN THE DOBERMANN Haplotypes were first derived from dogs homozygous at all three markers. The haplotype 1 was the most frequent haplotype. We were surprised to find that haplotype 1 had a frequency of 76% in the DCM-free group, with 61,5% of these dogs homozygous for the haplotype 1 (Table 3 and Figure 2b). In the DCM group, the frequency of haplotype 1 (31%) as well as the number of dogs homozygous for the haplotype 1 (2 dogs, 8%) is very low in comparison to the DCM-free group. The difference in haplotype frequencies between the DCM and the DCM-free group and the frequency of dogs homozygous for 260 allele were significant with P values of 0,002 and 0,001, respectively. DCM group (n=26) DCM-free group (n=13) 19% 12% 8% 19% 8% 8% 23% 61% 42% Random group (n=92) 8% 9% 39% 32% 12% Legend: Genotypes<5% Figure 1. Frequencies of REN252E18 genotypes in Dobermanns 146

147 The titin gene and DCM Haplotype 2 was the second most frequent haplotype in the two groups of Dobermanns. Its frequency in the DCM group (36%) was relatively high in comparison with the DCM-free group (8%). In addition to haplotypes 1 and 2, three minor haplotypes were defined (Table 3 and Figure 2). DCM-free (n = 13) 8% 8% 8% 76% DCM group (n = 26) 13% 8% 8% 36% 35% Legend: Haplotype 1 Haplotype 2 Haplotype 3 Haplotype 4 Haplotype 5 Figure 2. Haplotype frequencies in Dobermanns 147

148 Chapter 7 DISCUSSION GENOTYPES OF THE TITIN GENE MARKERS IN THE DCM AND DCM-FREE DOBERMANNS Typing of TTN microsatellite marker REN252E18 in the Dobermanns showed a sharp partition of genotypes between the DCM and DCM non-affected Dobermanns. The difference is enrichment of the 260 allele in the DCM-free group and enrichment of all other alleles (especially allele 256) in the DCM group (Table 2, Figure1). The haplotype analysis of two small deletions and one nucleotide substitution showed a similar partition between DCM and DCM-free dogs. The most striking finding of the haplotype analysis was the variance of haplotypes in the group of DCM Dobermanns as opposed to the relative uniformity of the DCM-free group (Table 3). While individual TTN marker alleles showed association with the DCM phenotype (Table 2), they did not combine to a single haplotype that was predominant in the DCM group (Table 3). Next to haplotype 2, which was shared between 12 out of 26 DCM dogs, we defined a total of four different haplotypes in the DCM group (Table 3, Figure 2). The divergent haplotypes in the DCM group and the high frequency of haplotype 1 in the non-affected group and the random group form the basis for the hypotheses that will be discussed. In statistical terms, the two alternative explanations are: 1. All titin alleles other than 260 (or haplotype 1) cause DCM or 2. The 260 allele (haplotype 1) confers protection against DCM. For the first explanation it has to be assumed that the DCM-causing mutation is either very old, so the linkage phase has been lost, even for intragenic markers, or that a DCM-causing mutation has occurred multiple times in the Doberman breed. For the second possibility, it has to be assumed that there is a DCMcausing gene elsewhere in the genome. In favour of the second possibility is the condition that it requires the least number of mutational events. Our data do not allow us to conclude as yet whether the DCM-causing mutation is in a gene different from the titin gene or that the first and the second explanation are not mutually exclusive. There were 2 dogs in the DCM group that were homozygous for the 260 allele and 3 dogshomozygous for haplotype 1. In the DCM-free group, there were five dogs that were not homozygous for the 260 allele or haplotype 1 (Table 3). There are several explanations possible for these exceptions. In the DCM group, it is possible that dogs had developed DCM due to a secondary cause. Conversely, dogs that were classified as healthy may still develop DCM at an age older than 9 years. Another explanation (under the protective allele hypothesis) for the five 148

149 The titin gene and DCM non-affected dogs that are not 260 (haplotype 1) homozygous is that they are DCM non-affected due to the absence of the pathological mutation. In the DCM and DCM non-affected dogs described in this study, several markers for DCM candidate genes and the DLA markers (mentioned below) have previously been typed (Stabej et al. 2004; Stabej et al. 2005a; Stabej et al. 2005b) and showed no association with one or the other phenotype, making it unlikely that the partition of the REN252E18 is due to a selection bias different from the DCM status. SELECTIVE SWEEP A LOCAL REDUCTION OF GENETIC VARIATION The consecutive polymorphic sites in linkage disequilibrium (LD) in haplotype 1 could represent a signature of a recent, local reduction of genetic variation, commonly called selective sweep (Kim and Nielsen 2004). The high frequency of haplotype 1 in the DCM non-affected group is suggestive of a recent positive selection for this haplotype in the European Dobermann population. In the light of our results, this would mean that haplotype 1 confers some kind of protection against development of the prevalent DCM phenotype. Dobermann breeders, recognizing the severity of DCM and sudden death, could have enhanced the positive selection pressure on haplotype 1. An example of a recent variant of a gene that has been subjected to positive selection in a human population is the protective effect of the hemoglobin E variant (HbE; β 26Glu Lys) against Plasmodium falciparum malaria (Ohashi et al. 2004). Similar to the relative long region with LD of haplotype 1 in the Dobermann, an extended LD surrounding the protective variant of hemoglobin E was found (Ohashi et al. 2004). Based on the current knowledge of the properties of titin and DCM, we developed two working hypotheses explaining the emergence of a protective TTN haplotype in the Dobermann breed. In the first hypothesis, DCM would be the result of an autoimmune reaction against titin. In human DCM, a number of observations suggest that autoimmune reactions underlie at least a subset of cases. Ericksson et al. (2003) and Okazaki et al. (2003) demonstrated that an acute immunological reaction to a viral infection can lead to the development of an autoimmune reaction targeted against cardiac antigens. Although cardiotropic viruses are considered to be the most common causative agent, in the majority of cases it has not been possible to isolate a viral genome from the myocardium as proof for this pathway of etiology in natural cases (Feldman and McNamara 2000). Taylor et al. (2004) discovered the crucial role of the HLA system in autoimmune DCM. They demonstrated that transgenic mice expressing the 149

150 Chapter 7 human DLA-DQ8 molecule developed DCM spontaneously, in the absence of an inflammatory or infective trigger. Titin is already known to be involved in an autoimmune disease of human and dog, myasthenia gravis. Myasthenia gravis (MG) occurs spontaneously and is associated with autoantibodies against the nicotinic acetylcholine receptor (AChR). Some of the MG patients (human and dog) also develop antibodies against the titin (Shelton et al. 2001). The functional significance of the titin antibodies is not yet known (Gautel et al. 1993; Shelton et al. 2001). To explore the possibility of Dobermann DCM being an autoimmune disease, we typed the DLA region (DQA1, DQB1 and DRB1) in the Dobermanns. We found no differences in the haplotype frequencies between the DCM and the healthy group. In fact, there was very little variation of alleles of the three genes in the breed; 88.7% of Dobermanns shared one haplotype, 8.75% the other most common haplotype and 1.25% (2 dogs) had an additional two haplotypes (data not presented). In addition, in collaboration with Dr. Skeie (Haukeland University Hospital, Bergen) the sera of seven of our DCM Dobermanns were tested for the presence of antibodies against the main immunogenic region of titin using purified titin antigen MGT-30 (Gautel et al. 1993) and detected no antibodies. The same MGT-30 antigen successfully detected the titin antibodies in dogs with myasthenia gravis (Shelton et al. 2001). Therefore, we suggest that the sera of DCM Dobermanns do not contain the antibodies against the main titin immunogenic region, but we cannot rule out that antibodies against other parts of the giant titin protein are involved. Since there was little variation in DLA haplotypes and no antibodies against the MGT-30 titin antigen in the sera of DCM dogs, we consider the autoimmune hypothesis of Dobermann DCM unlikely. In the second hypothesis, the protective form of TTN would compensate for a DCM causing mutation in another gene. Ventricular premature contractions are often detected prior to the development of DCM and ventricular tachycardia and subsequent sudden death are characteristics of DCM also seen frequently in the Dobermann. These observations suggest that disturbed calcium trafficking in the cardiomyocyt is a possible primary cause of DCM in the breed. In human, genes causing disturbed calcium trafficking (coding for mutated troponin T, titin, SCN5A and ABCC9) were identified as a cause of DCM with heart rhythm disturbances (Chapter 1, Table 1). An interesting experiment supporting a protective role of titin was performed with transgenic mice that develop severe DCM due to cardiac-specific overexpression of the calsequestrin gene (CSQ). A strong mouse strain dependance of the DCM phenotype was observed in mice over-expressing CSQ. Calsequestrin is a sarcoplasmic reticulum Ca 2+ -binding protein, which sequesters the contractile- 150

151 The titin gene and DCM dependent pool of Ca 2+. A genome wide scan in progeny from two mouse strains that showed distinctly different DCM phenotypes identified two loci that were significantly linked to cardiac function and survival (Suzuki et al. 2002). The titin gene maps to the exact same region on chromosome 2 that showed strong linkage to survival. Future studies will determine whether DNA sequence alterations in the titin gene are responsible for the observed differences in the heart failure phenotype (Suzuki et al. 2002). Titin has a pivotal role in the physiology of the heart. In the heart, passive tension primarily developed by the titin is a restorative force that acts in opposition to external stretching forces that lengthen the sarcomere. The force of the titin is also important in stretch sensing and modulation of the actomyosin interaction (Cazorla et al. 2001; Labeit et al. 2003; Miller et al. 2004). Several titin isoforms with different structural and mechanical properties are generated by differential splicing of the exons coding for the I-band region. In the heart, two principal isoforms have been described, a shorter (stiffer) N2B isoform and a longer (more compliant) N2BA isoform (Freiburg et al. 2000). In heart failure patients with DCM, the upregulation of N2BA isoform has been noted. It was suggested that increased N2BA expression negatively impacts systolic function but improves diastolic function by increased chamber compliance (Nagueh et al. 2004). It has also been proven that calcium affects passive myocardial tension in a titin isoform-dependent manner. While calcium significantly increased passive tension during and after stretch in the heart muscle that contained the N2BA titin isoform, passive tension developed by the N2B isoform was calcium insensitive (Fujita et al. 2004). Titin has therefore been proven to have a prominent role not only in the healthy heart, but also in the diseased heart. The association of the titin gene with DCM in the Dobermann described in this article serves as a basis for further studies of DCM in dogs. Firstly, the genotyping of TTN markers presented here in breeds with high incidence of DCM other than the Dobermann will reveal the potential role of TTN in these breeds. Titin gene DNA sequence analysis and studies of physiological properties of the titin variants observed in the Dobermanns using the functional genomics tools, will shed light on the question whether mutations in titin cause DCM or protect against DCM. 151

152 Chapter 7 REFERENCES Calvert, C.A., Brown, J Use of M-mode echocardiography in the diagnosis of congestive cardiomyopathy in Dobermanns. J Am Vet Med Assoc. 189: Calvert, C.A., Jacobs, G.J., Smith, D.D., Rathbun, S.L., Pickus, C.W Association between results of ambulatory electrocardiography and development of cardiomyopathy during long-term follow-up of Dobermanns. J Am Vet Med Assoc. 216: Cazorla, O., Wu, Y., Irving, T.C., Granzier, H Titin-based modulation of calcium sensitivity of active tension in mouse skinned cardiac myocytes. Circ Res. 88: Dukes-McEwan, J., Jackson, I.J The promises and problems of linkage analysis by using the current canine genome map. Mamm Genome. 13: Eriksson, U., Ricci, R., Hunziker, L., Kurrer, M.O., Oudit, G.Y., Watts, T.H., Sonderegger, I., Bachmaier, K., Kopf, M., Penninger, J.M Dendritic cellinduced autoimmune heart failure requires cooperation between adaptive and innate immunity. Nat Med. 9: Feldman, A.M., McNamara, D Myocarditis. N Engl J Med. 343: Freiburg, A., Trombitas, K., Hell, W., Cazorla, O., Fougerousse, F., Centner, T., Kolmerer, B., Witt, C, Beckmann, J.S., et al Series of exon-skipping events in the elastic spring region of titin as the structural basis for myofibrillar elastic diversity. Circ Res. 86: Fujita, H., Labeit, D., Gerull, B., Labeit, S., Granzier, H.L Titin isoform-dependent effect of calcium on passive myocardial tension. Am J Physiol Heart Circ Physiol. 287: H2528-H2534. Gautel, M., Lakey, A., Barlow, D.O., Holmes, Z., Scales, S., Leonard, K., Labeit, S., Mygland, A., Gilhus, N.E., Aarli, J.A Titin antibodies in myasthenia gravis: identification of a major immunogenic region of titin. Neurology. 43: Gerull, B., Gramlich, M., Atherton, J., McNabb, M., Trombitas, K., Sasse-Klaassen, S., Seidman, J.G., Seidman, C., Granzier, H., Labeit, S., et al Mutations of TTN, encoding the giant muscle filament titin, cause familial dilated cardiomyopathy. Nat Genet. 30: Grunig, E., Tasman, J.A., Kucherer, H., Franz, W., Kubler, W., Katus, H.A Frequency and phenotypes of familial dilated cardiomyopathy. J Am Coll Cardiol. 31: Guyon, R., Lorentzen, T.D., Hitte, C., Kim, L., Cadieu, E., Parker, H.G., Quignon, P., Lowe, J.K., Renier, C., Gelfenbeyn, B., et al A 1-Mb resolution radiation hybrid map of the canine genome. Proc Natl Acad Sci U S A. 100: Itoh-Satoh, M., Hayashi, T., Nishi, H., Koga, Y., Arimura, T., Koyanagi, T., Takahashi, M., Hohda, S., Ueda, K., Nouchi, T., et al Titin mutations as the molecular basis for dilated cardiomyopathy. Biochem Biophys Res Commun. 291: Jakobs, P., Bestwick, M.L., Ludwigsen, S.J., Nelsen, S.M., Winther, M.J., Hershberger, R.E., Litt, M Genetic linkage analysis of cardiomyopathy in Irish Wolfhounds. In: 2 nd International Conference Advances in Canine and Feline Genomics, Utrecht, The Netherlands, October 14-17; page 77. Kim, Y., Nielsen, R Linkage disequilibrium as a signature of selective sweeps. Genetics 167:

153 The titin gene and DCM Labeit, D., Watanabe, K., Witt, C., Fujita, H., Wu, Y., Lahmers, S., Funck, T., Labeit, S., Granzier, H Calcium-dependent molecular spring elements in the giant protein titin. Proc Natl Acad Sci U S A. 100: Liew, C.C., Dzau, V.J Molecular genetics and genomics of heart failure. Nat Rev Genet. 5: Liu, P.P., Mason, J.W Advances in the understanding of myocarditis. Circulation 104: Meurs, K.M., Magnon, A.L., Spier, A.W., Miller, M.W., Lehmkuhl, L.B., Towbin, J.A Evaluation of the cardiac actin gene in Dobermanns with dilated cardiomyopathy. Am J Vet Res. 62: Miller, M.K., Granzier, H., Ehler, E., Gregorio, C.C The sensitive giant: the role of titin-based stretch sensing complexes in the heart. Trends Cell Biol. 14(3): Miller, S.A., Dykes, D.D., Polesky, H.F A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16: Nagueh, S.F., Shah, G., Wu, Y., Torre-Amione, G., King, N.M., Lahmers, S., Witt, C.C., Becker, K., Labeit, S., Granzier, H.L Altered titin expression, myocardial stiffness, and left ventricular function in patients with dilated cardiomyopathy. Circulation 110: O Grady, M.R. and Horne, R Occult dilated cardiomyopathy in the Doberman Pinscher. PROC. 13 th ACVIM Forum, Lake Buena Vista, FL, page 298. Ohashi, J., Naka, I., Patarapotikul, J., Hananantachai, H., Brittenham, G., Looareesuwan, S., Clark, A.G., Tokunaga, K Extended linkage disequilibrium surrounding the hemoglobin E variant due to malarial selection. Am J Hum Genet. 74: Okazaki, T., Tanaka, Y., Nishio, R., Mitsuiye, T., Mizoguchi, A., Wang, J., Ishida, M., Hiai, H., Matsumori, A., Minato, N., Honjo, T Autoantibodies against cardiac troponin I are responsible for dilated cardiomyopathy in PD-1-deficient mice. Nat Med. 9: Rozen, S. and Skaletsky, H.J., Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics Methods and Protocols: Methods in Molecular Biology. Humana Press, Totowa, NJ 2000, Shelton, G.D., Skeie, G.O., Kass, P.H., Aarli, J.A Titin and ryanodine receptor autoantibodies in dogs with thymoma and late-onset myasthenia gravis. Vet Immunol Immunopathol. 78: Stabej, P., Imholz, S., Versteeg, S.A., Zijlstra, C., Stokhof, A.A., Domanjko-Petric, A., Leegwater, P.A., van Oost, B.A Characterization of the canine desmin (DES) gene and evaluation as a candidate gene for dilated cardiomyopathy in the Dobermann. Gene 340: Stabej, P., Leegwater, A.J., Imholz, S., Versteeg, S., Zijlstra, C., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A. 2005a. The canine sarcoglycan delta gene: BAC clone contig assembly, chromosome assignment and interrogation as a candidate gene for dilated cardiomyopathy in Dobermann dogs. Cytogenet Genome Res. In press. Stabej, P., Leegwater, P.A.J., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A. 2005b. Evaluation of the phospholamban gene in purebred large-breed dogs with dilated cardiomyopathy. Am J Vet Res. In press. Suzuki, M., Carlson, K.M., Marchuk, D.A., Rockman, H.A Genetic modifier loci affecting survival and cardiac function in murine dilated cardiomyopathy. Circulation 105:

154 Chapter 7 Taylor, J.A., Havari, E., McInerney, M.F., Bronson, R., Wucherpfennig, K.W., Lipes, M.A A spontaneous model for autoimmune myocarditis using the human MHC molecule HLA-DQ8. Immunol. 172: Thomas, W.P., Gaber, C.E., Jacobs, G.J., Kaplan, P.M., Lombard, C.W., Moise, N.S., Moses, B.L Recommendations for standards in transthoracic two-dimensional echocardiography in the dog and cat. Echocardiography Committee of the Specialty of Cardiology, American College of Veterinary Internal Medicine. J Vet Intern Med. 7:

155

156

157 Summary / Samenvatting / Povzetek

158 Summary SUMMARY Dilated cardiomyopathy (DCM) is a disease of the myocardium associated with dilatation of the left or both ventricles. The disease affects almost exclusively certain breeds of dogs and represents a large problem to dogs and their owners. The breed predisposition indicates that genetic factors play an important role in the aetiology of the disease. Canine DCM is closely similar to its human counterpart and the knowledge about human DCM can be applied to canine DCM and vice versa. In chapter 1.1, we evaluate the current knowledge of DCM in humans. Mutations in several genes have been found to cause DCM in man and these have provided valuable insight into the pathogenesis of heart failure and the physiology of the heart. The recent advances in understanding of the molecular background of DCM have illuminated several disease pathways and uncovered the complexity of DCMs. However, the list of DCM genes is still far from complete and a number of DCM genes remain to be discovered together with the modifier genes of which the role is yet poorly understood. Inbred dog breeds replicate several phenotypes of human familial DCMs and have a unique population structure for genetic studies with multiple advantages over the human population. The great leap forward, made in the field of dog genetics during the course of the PhD research of this thesis, is described in chapter 1.2. The amenability of the dog genome for deciphering the molecular genetic background of DCM is discussed in chapter 1.3. Clinical characteristics of DCM in dogs are described in the 2 nd chapter. A review of medical records of DCM dogs seen by the University Clinic for Surgery and Small Animal Medicine (Slovenia) showed that Dobermanns represented 39% of all dogs diagnosed with DCM and were the most frequently affected breed. Reliable clinical diagnosis is essential for genetic studies, especially in phenotypically heterogeneous disorders like DCM. Clinical characteristics typical of Dobermanns with DCM are congestive heart failure and sudden death with significantly shorter survival after diagnosis when compared to dogs of other breeds. The group of Dobermanns described in chapter 2 was further expanded, enriched with the Dutch DCM Dobermanns and used for evaluation of candidate genes described in chapters 3 to 7. At the outset of the DCM research project in 2001, most of the DCM candidate genes had not yet been mapped. Isolation of the bacterial artificial chromosome (BAC) clones carrying the genes of interest, from the canine BAC library was therefore necessary. In chapters 3 to 5, we describe cloning, chromosomal localization and characterization of the canine desmin, α- tropomyosin and δ-sarcoglycan genes. Upon availability of the dog genome 158

159 Summary sequence, the structures of the three genes were determined and the BAC clones were aligned with the dog genome. Microsatellite markers isolated from the BAC clones were typed in the Dobermanns and evaluated for association with the DCM phenotype. None of the three genes was associated with the DCM in Dobermanns. In the process of the evaluation of the α-tropomyosin gene, a remarkable retrotransposition event was observed on the canine X chromosome and this is described in chapter 4. In chapter 6, the evaluation of the phospholamban gene is described. Because phospholamban is a small gene, we decided to use direct DNA sequence analysis, rather than typing of polymorphic markers. In addition, we screened DCM dogs of two other breeds, Great Dane and Newfoundland, for mutations in the phospholamban gene, but no mutations were identified. The gene described in chapter 7 encodes the largest protein known in mammals, titin. Genotyping results of markers for the titin gene in the Dobermann, point to a crucial role of titin in DCM susceptibility. Hypotheses explaining the possible role of the titin gene in Dobermann DCM are discussed in chapter 7. Titin is the first gene found involved in DCM in dogs and represents an important milestone in the genetic research of heart diseases. 159

160 Summary SAMENVATTING Dilaterende cardiomyopathie (DCM) is een aandoening van de hartspier die wordt gekenmerkt door dilatatie van de linker of van beide hartkamers. Deze genetische aandoening komt vrijwel uitsluitend voor bij bepaalde hondenrassen en vormt een groot probleem voor de betreffende honden en hun eigenaar. DCM bij honden vertoont veel overeenkomst aan de menselijke variant en kennis over menselijke DCM kan worden toegepast op DCM bij honden en vice versa. In hoofdstuk 1.1 wordt de huidige kennis over DCM bij de mens besproken. Er zijn mutaties in verscheidene genen gevonden die DCM bij mensen veroorzaken en die waardevolle inzichten verschaffen in de pathogenese van hartklachten en de fysiologie van het hart. Recent is het begrip van de moleculaire basis van DCM sterk toegenomen en werd nieuw licht geworpen op de verschillende manieren waarop de ziekte kan ontstaan. Het heeft ook duidelijk gemaakt hoe complex het ziektebeeld is. De lijst van DCM veroorzakende genen is nog lang niet compleet zodat een aantal DCM genen nog moet worden ontdekt, net als de zogenaamde modifier genes waarvan nog nauwelijks bekend is welke rol ze precies spelen. Hondenrassen, die relatief ingeteeld zijn, vertonen verschillende fenotypes van erfelijke DCM bij mensen en bij de honden bestaat een unieke populatie opbouw voor het uitvoeren van genetisch onderzoek, met meerdere voordelen ten opzichte van de menselijke populatie. De grootste sprong voorwaarts op het gebied van honden genetica die gedurende het onderzoek voor dit proefschrift werd gemaakt, wordt beschreven in hoofdstuk 1.2. Het belang van het genoom van honden voor het ontcijferen van de moleculair genetische achtergrond van DCM wordt behandeld in hoofdstuk 1.3. Klinische kenmerken van DCM bij honden worden beschreven in hoofdstuk 2. Een overzicht van de medische dossiers van patiënten met DCM van de University Clinic for Surgery and Small Animal Medicine (Slovenië) laat zien dat 39% van het totaal aantal honden met DCM bestond uit Dobermanns en dat ze daarmee het meest getroffen ras zijn. Betrouwbare klinische diagnoses zijn essentieel voor genetische studies, vooral bij fenotypisch heterogene aandoeningen zoals DCM. Klinische eigenschappen die typisch zijn voor Dobermanns met DCM zijn plotseling overlijden of congestief hartfalen met een significant kortere overlevingstijd in vergelijking met andere hondenrassen. De groep Dobermanns die beschreven wordt in hoofdstuk 2 is verder uitgebreid met Nederlandse Dobermanns met DCM en gebruikt voor de analyse van kandidaat genen zoals beschreven in de hoofdstukken 3 tot en met 7. Aan het begin van het onderzoek naar DCM in 2001 waren de meeste kandidaat genen nog niet in kaart gebracht. Het was daarom noodzakelijk om Bacterial Artificial Chromosome (BAC) klonen met de desbetreffende genen 160

161 Summary uit de canine BAC library te isoleren. In de hoofdstukken 3 tot en met 5 beschrijven we het kloneren, de chromosomale locatie en het karakteriseren van de desmine, α-tropomyosine and δ-sarcoglycaan genen bij de hond. Toen de informatie uit de honden genoom sequentie beschikbaar was, werd de structuur van de drie genen vastgesteld en zijn de BAC klonen vergeleken met het honden genoom. Microsatelliet markers die geïsoleerd werden uit de BAC klonen zijn getypeerd in de Dobbermanns, en er is bepaald of er een associatie was met het DCM fenotype. Geen van de drie genen was geassocieerd met de DCM bij de Dobbermanns. Gedurende het analyseren van het α-tropomyosine gen, is een opmerkelijke retrotranspositie gebeurtenis waargenomen op het X chromosoom van de hond en dit is beschreven in hoofdstuk 4. In hoofdstuk 6 wordt de analyse van het phospholamban gen beschreven. Omdat phospholamban een klein gen is, hebben we besloten directe DNA sequentie analyse toe te passen, in plaats van gebruik te maken van polymorfe markers. Daarnaast hebben we DCM honden van twee andere rassen, de Duitse Dog en de Newfoundlander, gescreend op mutaties in het phospholamban gen, maar er zijn geen mutaties geïdentificeerd. Het gen dat wordt beschreven in hoofdstuk 7 codeert voor het grootste eiwit in zoogdieren, titine. De resultaten van het genotyperen van titine markers bij de Dobermanns, wijzen erop dat titine een cruciale rol speelt bij de gevoeligheid voor de ontwikkeling van DCM. In hoofdstuk 7 worden verschillende hypotheses besproken voor de rol van titine bij het ontstaan van DCM bij de Dobermann. Titine is het eerste gen waarvan ontdekt is dat het betrokken is bij DCM in honden. Deze ontdekking is een mijlpaal in het genetisch onderzoek naar hartziekten. 161

162 Summary POVZETEK Dilatativna kardiomiopatija (DKM) je bolezen srčne mišice, povezana z dilatacijo levega ali obeh srčnih prekatov. Prizadene skoraj izključno pse velikih in največjih pasem in predstavlja resen problem za pse in za njihove lastnike. Glede na dejstvo, da se DKM skoraj izključno pojavlja pri psih določenih pasem, ima dedni dejavnik zagotovo pomembno vlogo pri nastanku te bolezni. Klinični potek DKM pri psih je zelo podoben kliničnemu poteku DKM pri ljudeh. Zato se izsledki raziskav DKM pri ljudeh nanašajo tudi na DKM pri psih in obratno. V prvem poglavju so opisane značilnosti dedne DKM pri ljudeh. Mutacije številnih genov, ki so bile odkrite v zadnjih letih kot vzrok dednih DKM, so pomembno prispevale k razumevanju razvoja DKM in fiziologije srca. Odkritje molekularnogenetske osnove DKM je omogočilo boljše razumevanje razvoja te bolezni ter razkrilo njeno zapletenost. Navkljub hitremu napredku v poznavanju dedne osnove DKM je seznam mutacij, ki povzročajo ali prispevajo k razvoju DKM, še nepopoln in številna vprašanja ostajajo brez odgovora. Posamične pasme psov posnemajo številne fenotipe, ki so opisani pri ljudeh z DKM. Edinstvena struktura pasje populacije ima številne prednosti pri genetskih raziskavah. Velik korak na področju pasje genetike storjen tekom štiriletne raziskave, predstavljene v doktorski disertaciji je opisan v poglavju 1.2. Uporabnost pasjega genoma za dešifriranje molekularnogenetske osnove DKM je obravnavana v poglavju 1.3. Drugo poglavje opisuje klinične značilnosti DKM pri psih. Pregled registriranih primerov psov z DKM na Kliniki za kirurgijo in male živali (Veterinarska fakulteta, Univerza v Ljubljani) kaže, da je bilo med obolelimi psi največ dobermanov (39 %). Zanesljiva klinična diagnoza je osnova vsake genetske raziskave, še posebno pri heterogenih boleznih, kot je DKM. Tipični klinični znaki DKM pri dobermanih v primerjavi s psi drugih pasem so kongestivno srčno popuščanje in nenadna smrt ter kratka doba preživetja po postavitvi diagnoze. Skupini dobermanov, ki je opisana v drugem poglavju, se je pridružilo še nekaj dobermanov iz Slovenije in Nizozemske. Ta razširjena skupina je predstavljala osnovo za nadaljnje genetske raziskave, opisane v poglavjih, ki sledijo. Ob začetku izvajanja projekta DKM (2001) večina genov, ki so bili kandidati za DKM pri psih, še ni bila uvrščenih na karti pasjega genoma. Zato je bil prvi korak izolacija t. i. BAC (bakterijski umetni kromosom)-klonov kot nosilcev genov, ki smo jih želeli analizirati. Ti BAC-kloni so zbrani v BAC-knjižnici pasjega genoma. Kloniranje in določanje kromosomolne lokacije genov desmina, α-tropomiozina in δ-sarkoglikana je opisano v poglavjih 3 5. Po objavi sekvence pasjega genoma smo določili natančno strukturo teh treh genov in lokacijo BAC- 162

163 Summary klonov. Genetski markerji (mikrosateliti), ki so bili izolirani iz BAC-klonov, so bili tipizirani na genetskem materialu dobermanov. V primeru teh treh genov nismo našli povezave med določeno obliko markerja in DKM. Pri analizi α- tropomiozina smo odkrili zanimivo retrtranspozicijo gena na X kromosom, ki je opisana v četrtem poglavju. V šestem poglavju je opisana analiza gena za fosfolamban. Glede na majhnost gena, ki kodira fosfolamban, smo se odločili za analizo s sekveniranjem. Poleg dobermanov smo fosfolamban analizirali tudi pri novofundlandcih in nemških dogah. Pri nobeni od teh treh pasem nismo našli mutacij. Gen, opisan v sedmem poglavju, nosi zapis za največji protein pri sesalcih titin. Rezultati genotipizacije številnih markerjev pri dobermanih kažejo na pomembno vlogo titina pri dovzetnosti dobermanov za DKM. Hipoteze, ki razlagajo mogoče vloge titina pri DKM, so opisane v zadnjem poglavju. Titin je prvi gen, za katerega je bilo ugotovljeno, da sodeluje pri nastanku DKM pri psih, in predstavlja pomemben mejnik v genetskih raziskavah srčnih obolenj. 163

164

165 Acknowledgments

166 Acknowledgments ACKNOWLEDGMENTS Plans go wrong for lack of advice; many counselors bring success. Proverbs 15:22 (Holy Bible, NLT) This chapter acknowledges those without whom this PhD project would never have been completed. My thanks goes first to my promotor Prof. dr. Bernard van Oost. Dear Bernard, thank you for not only writing a great DCM research plan, but also for giving me the opportunity to execute it. I enjoyed working with you and appreciate your supervision and dedication to the DCM project along with good ideas and willingness to discuss the research at any time (even without an afspraak). I would next like to thank to my co-promotores: Dr. Peter Leegwater, Dr. Arnold Stokhof and Dr Saša Domanjko-Petrič. Dear Peter, Arnold and Saša. Thank you for all the time invested into the DCM research, great ideas, quick corrections of manuscripts and gezellig time we had in the past few years. It has been a pleasure to work with you. Moving to the laboratory work. Arriving as a veterinarian with no laboratory experience, good supervision during my first months in the lab was essential for the safety of the Gezelschapsdieren department. Officially supervised by Serge (Versteeg), non-officially as well by Robin (now Dr. Robin Everts) and occasionally by Bart (another esteemed Dr., Bart van de Sluis), I conducted my first experiments in molecular genetics. Learning from you all was great and thank you for being patient and fun teachers. To Bart, you were a model PhD student and I really admire your research on the anorexic sheep. There are a number of people from the Gezelschapsdieren I would like to thank: Monique, Bas, El Petra, René, Jeannette, Louis, Rosalia, Adri, Ies, Dick, Lisette, Ank and Sara, my fellow Slovene. It was nice to work in the same department and thank you for being such great colleagues. To Jan (Dr. Jan Mol), Walter (Dr. Walter van den Brom) and Bart (Dr. Bart Knol), with whom I share the love for classical music, thank you for the Mozart we played together (Jan), concerts you organized (Bart) and music for violin and organ we will play together (Walter). Frank, thanks for all the useless Dutch tongue twisters you taught me and all the laughs we had. Further I would like to thank Manon for not quitting on the giant gene with 363 exons. I really appreciate your work on the titin. Sandra, many thanks to you as well for the work on the DCM project. It was great to work with you almost from the beginning of the AIO project. I am still trying to learn how to keep a good lab journal from your fine example. Thanks for your friendship in and outside the lab and many thanks for accepting the role 166

167 Acknowledgments of a paranimf at my PhD defense. I would also like to acknowledge my (ex) AIO colleagues: Sacha (now Dr. Sacha), Bart, Ilse (another Dr.), Brigitte, Chen Lee, Eveline, Peter, Niyada, Shahram (a very fresh Dr.), Gaby, Jedee, Lars, Nagesha (thanks for many great badminton matches) and Suzanna for their friendliness and company. To Anje, the not (yet) official AIO; (who would have though that you would end up working on DCM as well!). Thanks for supplying me with plants in the past years as well as entrusting me with your fish, while you are exploring the research life in Manchester/Liverpool. I wish you all best with the DCM and renal disease research. Very special thanks go to my AIO kamergenoten. Dear Linda, Jeanette and Yvette. I could not ask for better kamergenoten! Thank you for your friendship and support in good or bad times. Special thank goes to Linda, the fellow sufferer in molecular genetics. I am really glad we got to work in the same research group and am gratefully indebted to you for your occasional advice in statistics and especially for the translation of the Dutch texts. Furthermore, observing your organizational skills has been very educational, and thanks for all the reminders about various deadlines I might have missed otherwise. Thank you as well for accepting the paranimf role at my PhD defense. With Sandra on one side and you on the other, the defense really does seem like just another research bespreking (as one of the members of my beordelingscommisie puts it). The only difference is that some people wear costumes and we address one another in a rather peculiar way. To the students who worked on the DCM project: Laura Blomer, Dion van der Wee, Debora Poldervaart, Mojca Kobal and Nicole Willems. Thank you for the work you invested into the DCM research. It was nice to work with you and all best in your veterinary carriers. I would also like to thank to Dr. George Voorhout for performing echocardiography and Harry van Engelen for recording the ECGs. Further I thank to Dr. Carla Zijlstra, Lotte de Kreek and Esther van t Veld for carrying out the FISH experiments. Since the grass is always greener on the other side, I occasionally wondered from the Diergeneeskunde, across the canal, to the UMC. Even though the grass was not greener, the sequencers did have more capillaries. I would like to thank to Karen and Alfons from the Divisie Medische Gentica in (UMC, Stratenum), for an ever-welcoming reception at your fine sequencing facilities. Dobermann owners and breeders from Slovenia and The Netherlands played an essential role in the DCM project. Thank you all for your willingness to cooperate, enthusiasm and time invested in the DCM project. Further, I would like to express my gratitude to the people from the academic community whom I asked for advice or collaborated with during my PhD project. To Dr. Bobby Koeleman from the UMC (Universiteit Utrecht), Dr. 167

168 Acknowledgments Jo Dukes-McEwan (University of Liverpool), Prof. dr. van Eden (Universiteit Utrecht), Dr. Sjoerd Repping (AMC), Dr. Geir Olve Skeie (Haukeland University Hospital/Bergen) and to Dr. Kate Meurs (Ohio State University College of Veterinary Medicine). Thanks are also due to friends outside work. To start with those of you directly involved in the DCM project. Dear Caroline, Tijmen and Helena. Thanks for your very special friendship that has no geographical or any other borders. Tijmen, thanks for helping with the Samenvatting of this thesis (on January 1 st!). Caroline, thanks for your exceptional interest in the DCM research and advice on English writing throughout my PhD. Helena, thanks for correcting the Slovene Summary. I would also like to thank to the Eekelder family from Lichtenvoorde, who adopted me for a month back in 1995 and introduced me to the strange Dutch habits for the very first time. A special thanks to friends from the Crossroads International Church. You are a great community. Since it is impossible to list you all, I will only mention the top five. Dear Chris, Esther, José, Robert and Anita. Thanks for a great friendship and support in the past years. Four more friends who just do not fit into any of the above mentioned categories must be acknowledged for their ultimate support of my PhD endeavour: Mojca, Darja, Adrian and Ruža. It is always great fun spending time with you and I miss that we no longer live in the same countries. Finally, I would like to thank to my Grand mom, Mom, Dad and brother Uroš. Thank you for your utter love and support before, during and after my PhD. Mom and Dad, you taught me how to think for myself and work hard, which was essential during my PhD study. Uroš, you are trying to teach me how to cash in my qualities; an area we still need to work on. Mami, mamici, očiju in Urošu se zahvaljujem za nesibično podporo pred, med in po doktoratu. Mamica in oči, naučila sta me misliti s svojo glavo in pridno delati, kar je bilo osnovno za dobre rezultate. Uroš pa me skuša naučiti, kako naj svoje znanje končno dobro vnovčim. To pa mi vsem odkritjem navkljub še vedno ne uspeva najbolje. And ultimately, to my dear fiancé Phil, many thanks for your love, support and belief in me, especially during the last year of my PhD when work was most hectic and free time was a rarity. After the PhD ceremony, we soon need to organize another ceremony. However, this time, the opposition wil only demand two short answers. 168

169

170

171 List of Publications

172 List of Publications LIST OF PUBLISHED AND SUBMITTED ARTICLES Stabej, P., Leegwater, A.J., Imholz, S., Loohuis, M., Mandigers, P.J.J., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A. Genetic epidemiological studies of DCM in the Dobermann point to a crucial role of titin in DCM susceptibility. Manuscript in preparation. Stabej, P., Meurs, K., van Oost, B.A. Molecular genetics of cardiomyopathies in dogs. Invited chapter for The Dog and Its Genome (eds: E. A. Ostrander, U. Giger, and K. Lindblad-Toh); Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, USA. Stabej, P., Leegwater, P.A.J., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A. Duplication of a polymorphic CA-repeat by retrotransposition in the canine genome: implications for the analysis of the association of the α-tropomyosin gene (TPM1) with dilated cardiomyopathy in Dobermanns. Submitted. Stabej, P., Leegwater, A.J., Imholz, S., Versteeg, S., Zijlstra, C., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A The canine sarcoglycan delta gene: BAC clone contig assembly, chromosome assignment and interrogation as a candidate gene for dilated cardiomyopathy in Dobermann dogs. Cytogenet Genome Res. In press. Stabej, P., Leegwater, P.A.J., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A Evaluation of the phospholamban gene in purebred large-breed dogs with dilated cardiomyopathy. Am J Vet Res. In press. Stabej, P., Imholz, S., Versteeg, S.A., Zijlstra, C., Stokhof, A.A., Domanjko- Petrič, A., Leegwater, P.A., van Oost, B.A Characterization of the canine desmin (DES) gene and evaluation as a candidate gene for dilated cardiomyopathy in the Dobermann. Gene 340: Domanjko-Petrič, A., Stabej, P., Žemva, A Dilated cardiomyopathy in Dobermanns, survival, causes of death and pedigree review in a related line. J Vet Cardiol. 4: Stabej, P., Domanjko-Petrič, A Dilated Cardiomyopathy in Dogs. Res. Rep.-Univ. Ljubl. Vet. Fac. 36:

173 List of Publications PROCEEDINGS OF CONGRESSES Stabej, P., Leegwater, A.J., Imholz, S., Loohuis, M., Mandigers, P.J.J., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A. The giant myofilament protein Titin is associated with dilated cardiomyopathy in the Dobermann. Invited speaker at the 2nd International Conference: Advances in Canine and Feline Genomics, Utrecht, The Netherlands, October 14-16, Proceedings p. 28. Stabej, P., Leegwater, P.A., Imholz, S., Loohuis, M., Domanjko-Petrič, A., Stokhof, A.A., van Oost, B.A. Genetics of canine dilated cardiomyopathy protection by the giant. ZonMw Genetica retraite, Rolduc, Kerkrade, The Netherlands, March, Stabej, P., Versteeg, S.A., Stokhof, A.A., Domanjko-Petrič, A., van Oost, B.A. Candidate genes for dilated cardiomyopathy in dogs. 1 st International Conference: Advances in Canine and Feline Genomics, St. Louis, Missouri, May 16-19, Stabej, P., Versteeg, S.A., Zijlstra, C., Stokhof, A., van Oost, B.A. Candidate genes for dilated cardiomyopathy in dogs. ZonMw Genetica retraite, Rolduc, Kerkrade, the Netherlands, November 8-9, Proceedings. Stabej, P., Versteeg, S.A., Zijlstra, C., Domanjko-Petrič, A., Stokhof, A.A., van Oost, B.A. Candidate genes for dilated cardiomyopathy in dogs. Congress of the European Society of Veterinary Internal Medicine (ECVIM), Dublin, Ireland, September 5-1, Proceedings p. 98. van Oost, B.A., Stabej, P. Introduction to molecular genetics and techniques in molecular genetics. Comparative Clinical and Molecular Endocrinology; International Symposium. Utrecht, The Netherlands May, Stabej, P., Domanjko-Petrič. Dilated cardiomyopathy in Doberman Pinschers. Voorjaarsdagen International veterinary Congress, Amsterdam, The Netherlands, April 24-26,

174

175 Curriculum Vitae

176 Curriculum Vitae CURRICULUM VITAE Polona Stabej was born December , in Kranj (Slovenia). After finishing the secondary school, she studied veterinary medicine at the Faculty of Veterinary Medicine (Ljubljana University). During the veterinary study, she did a one-year clinical research on dilated cardiomyopathy in dogs under the supervision of Dr. Aleksandra Domanjko-Petrič. The research was awarded with a Prešern award of the Veterinary Faculty in She received her DVM degree in 1999 and proceeded with a six months veterinary internship in Slovenia. In year 2000, she did a six months research internship at the Utrecht University (Faculty of Veterinary Medicine, Department of Clinical Sciences of Companion Animals) where she worked on genetics of cardiomyopathy. In February 2001, she started her PhD at the same department under the supervision of Prof.dr. Bernard van Oost. The results of the PhD research are presented in this thesis. 176

177 Color figures

178

179 Color figures SERCA2a Phospholamban Titin Myosin Sarcoplasmic reticulum α-tropomyosin nucleus Sarcoglycans Lamin Desmin Dystrophin Troponin-I Troponin-T T-cap Actin Troponin-C Z-disc Figure 1 from page 22. Sarcolemma α γ δ β MLP Cypher β α Dystroglycans KATP channel Na channel 179

180

181 Color figures 1,0,8,6,4 Other breed dogs Survival,2 0,0 0 Days Other breed dogs - censored Dobermanns Dobermann - censored Figure 2 from page 51. 1,0,8,6,4 Survival,2 SD Dobermanns 0, CHF Dobermans Days Figure 3 from page

182

183 Color figures Figure 3 from page 73. b. Figure 3b from page