Genetically modified animal models in cardiovascular research

INTRODUCTION

In spite of major advances in cardiology, cardiovascular disease remains the principle cause of death and morbidity in industrial societies. The 20^th century has been characterized by rapid growth in medical technology, rapid dissemination of the information related to research and idea development, and relatively rapid incorporation of new information into practice. Improvements in the methods of diagnosis and therapy in clinical cardiology have led to real progress in the treatment, and in some cases, amelioration or even prevention of cardiovascular disease. However, while techniques have grown increasingly sophisticated, our ability to treat the root causes of cardiovascular disease remains limited, That is, we can treat many of the symptoms, but, because of limitations in understanding the primary etiologies that underlie the onset and development of the pathogenic processes, the root causes remain unaffected by clinical practice.

Until recently, approaches for studying diseases used the tools of pharmacology, surgery, biochemistry or physiology. It is humbling to realize that we are now attempting to describe complex physiological systems without a complete understanding or appreciation of the players' identities. Estimates for the number of genes in the mammalian genome range from 50,000-100,000 but currently, only about 10% have been described in any sort of useful way. Thus, we may be missing approximately 90% of the most basic information needed for understanding genetic output- information that almost certainly underlies much of cardiovascular disease, either directly or indirectly by dictating the heart' s response to environmental cues. As this review is being written, announcements are being made concerning the completion of the human genome project. Within the next 1-3 years comparative information on other useful mammalian systems such as the rat and mouse will follow. While providing the structural basis for a complete understanding of mammalian physiology, in their raw state these data will do little to help the clinician. Their value will rest largely upon how efficiently the data bases are "mined' for informational content by the basic scientists. Understanding the raw data by placing them in their biological contexts should lead to a new kind of medicine. These data will both reveal the primary genetic etiologies responsible for disease development as well as provide new management strategies for the treatment of cardiovascular disease.

If, in the past 25 years, molecular biology has taught us anything about clinical medicine, it is that specific .protein complements underlie cell and organ function, Protein production is, in turn, controlled by cell-specific defined transcriptional patterns of the common gene set Since cell and ultimately organ function depend upon the polypeptides that are present, it is not surprising that when function is altered, for example, during the development of disease, changes in the protein pools occur. In the heart, numerous examples of contractile protein changes correlate with functional alterations, both during normal development and during the development of numerous pathologies. Similarly, different congenital heart diseases are characterized by certain shifts in the motor proteins, For example, in heart failure the up- or down-regulation of different (hypothesized) effectors/modulators has been well documented. To understand these relationships, and to establish models in which the pathogenic processes can be studied longitudinally, it is necessary to direct the heart to stably synthesize, in the absence of other pleiotropic changes, the candidate protein, Subsequently, one can determine whether or not the protein' s presence causes the effects directly or indirectly with the goal being to define potential therapeutic targets. By affecting the heart' s protein complement in a defined manner, one has the means to establish both mechanism and the function of the different mutated. proteins or protein isoforms. Transgenesis and gene targeting, and the subsequent creation of stable animal models provides a means to modify the mammalian genome and study the consequences in the cardiovascular system. By directing expression of an engineered protein to the heart, one is now able to effectively remodel the cardiac protein profile and study the consequences of a single genetic manipulation at the molecular, biochemical, cytological and physiologic levels, both under normal and stress stimuli.

Our discussion will focus on the cardiomyopathies, and how genetically engineered models have been productively used to both confirm the primary etiology and to begin to explore the pathogenic processes that eventually result in heart failure. The advantages and weaknesses of current technology will be explored, as well as developments that are on the horizon; developments that promise to make these approaches even more valuable. Rather than being encyclopedic, the review will focus on how current concepts are being applied to the field, so that the future developments, which will inevitably astound us, can be appreciated in context.

CREATING STABLE ANIMAL MODELS

A number of techniques are now available, which allow the heart's contractile apparatus to be altered in a defined manner: we will focus on those approaches that result in germline transmission of the remodeling event(s). Thus, the desired modifications can be propagated throughout multiple generations and result in the creation of stable, new animal models. Necessarily, such stable changes need to be performed at the level of the genome and two distinct but .complementary approaches have been developed: transgenesis and gene targeting. Each modifies the mammalian genome but they are very different, both technically and in their genetic outcomes. Using transgenesis, the heart' s contractile apparatus can be significantly remodeled, but this method relies exclusively on the transgene's ability to produce a phenotype against the endogenous context, as the transgene does not affect endogenous gene tic output. Like transgenesis, gene targeting ( often referred to as gene ablation, or gene knockout) directly changes the gene tic structure, and the heart' s contractile apparatus can also be altered dramatically. Ablating a gene may lead to a loss of function, which can help establish a candidate sequence' s function. Gene targeting can also be used to make changes in the sequences encoding a contractile protein' s functional domain, or at a single amino acid residue, resulting in the establishment of precise structure-function relationships. These approaches, which are compared and contrasted in Table I, are rapidly creating a group of animals whose altered contractile protein complements can lead to a fundamental understanding of the structure/function relationships which underlie the heart's function at the molecular, biochemical, whole organ and whole animal levels.

Transgenesis

Transgenesis allows one to stably alter the mammalian genome such that the modifications can be transmitted through the germline. Although the procedures are technically possible in larger animals¹, because of cost, time and animal husbandry considerations, basic research has been largely limited to the mouse and, in rare cases, the rabbit^2-3. Development of the technique, which consists of injecting DNA directly into the fertilized embryo, arose from technical advances in the embryo manipulation and recombinant DNA fields⁴. Insertion of the exogenous DNA is random and multiple copies are placed into the genome, usually at one site in a head to tail or head to head arrangement. The insertion point cannot be controlled, nor can the investigator predetermine the number of copies placed into the recipient genome. This multi-step process is described in Table 1. Thus, unlike gene targeting, in which the homologous recombination event can be selected (see below, next section), transgenesis is a random process and the DNA is placed at sites other than where the homologous sequences are located. This means that the transgene's expression will be superimposed upon that of the endogenous genes. Thus, for a phenotype to present, the transgene must yield a dominant effect. The ultimate transcriptional activity of the transgenic DNA can be copy number dependent or independent and is often influenced by the particular chromosomal location into which it is placed ("position effects"), rendering it more or less transcriptionally active. Additionally, DNA insertion is not a benign process and can lead to significant disruption, rearrangement or deletion of the flanking DNA, resulting in insertional mutagenic effects⁵. Fortunately, these effects are most often recessive, and since analyses of the transgenic animals are usually restricted to the heterozygotes, this does not present a problem. The position-dependent and copy number effects necessitate the analysis of multiple lines, an expensive and labor intensive process. Nevertheless, multiple mutations and chimeric constructs can be generated and numerous lines of mice made relatively rapidly as compared to gene targeting. These points recommend careful consideration of the transgenic approach for cardiac remodeling studies.

Systemic expression of a transgene can seriously complicate any resultant cardiac phenotype. The development of promoters or sequences that control and drive gene expression, whose expression is largely cardiac-restricted, has enhanced the utility of the transgenic approach for studying cardiovascular disease- in the absence of other confounding effects on other organ or muscle systems. Not surprisingly, the transcriptional sequences of the cardiac contractile genes themselves have been used effectively to drive cardiac transgenic expression. These include the actin, myosin light chain 2v, and α - and β-myosin heavy chain promoters among others⁶. A large number of transgenes (numbering in the hundreds) affecting the heart have been expressed using cardiac-specific promoters and models for general remodeling events such as hypertrophy or specific cardiac diseases have been made⁷. It is interesting that the outcomes, in terms of the overall protein levels, can vary widely in the different transgenic experiments, depending upon the protein that is being made. For example, as we will see below, many of the hypertrophic cardiomyopathies are caused by mutations in the contractile proteins⁸. A detailed exanimation of the transgenic paradigm has been carried out using many of the other contractile proteins. These studies show that, despite significant over-expression at the RNA level, the protein levels remain unaffected. As the transgenic RNA is being made at much higher levels than the endogenous RNA, the net effect in the cardiomyocyte is to dilute out the endogenous protein with increasing amounts of the transgenic polypeptide. The end result is that replacement of the endogenous protein with the transgenic-encoded species occurs⁹.

It is important to understand the limitations of the transgenic approach, some of which have been alluded to above. First, it must be emphasized, it relies upon the fact that any transgene that is expressed must "dominate' over the endogenous gene product. Second is the potential for insertional mutagenesis, even in the heterozygous state. Third, the question of the physical act of over-expression must be considered. As' noted above, for the contractile proteins, this does not appear to be a problem. In addition, as a control, the wild type protein can always be expressed at similar levels to ensure that no phenotype presents. However, the promoters that are normally used in these experiments derive from the contractile proteins. These polypeptides make up the majority of the cardiomyocytes' protein mass and are expressed at very high levels with steady state RNA transcripts numbering in the tens of thousands. Thus, a very active promoter is part of the transgenic construct. In many experiments where potentially toxic substances, powerful biological signaling molecules or amplifiers are being expressed transgenically, the outcomes are very difficult to interpret. For example, in a recent report, the marker gene, Green Fluorescent Protein (GFP) was expressed using a cardiac specific promoter¹⁰. The resultant hypertrophy was interpreted to mean that over-expression of any molecule could result in hypertrophy. Perusal of the general literature indicates that, clearly, this is not the case. Just as likely an explanation is that expression of GFP in the context of the heart compromised function and led to the resulting hypertrophy. That is, GFP is not a harmless protein but rather, is capable of causing a hypertrophic response. Thus, in an experiment where gain-of-function is being explored, and there is no control in which the wild type protein can be expressed in a parallel cohort with the mutation under study, interpretation of the data becomes difficult or impossible.

Gene Targeting

Gene targeting, in contrast to transgenesis, replaces the endogenous DNA sequence at a chosen site with exogenously prepared (and, presumably, altered) DNA. Gene targeting uses homologous recombination to insert the DNA at the specific gene site leading to the inactivation (or mutation) of the gene of interest, generating animals deficient (or with a mutation) in that gene product. This technology is typically used to inactivate or "knock out" specific genes in the genome and hence is referred to as a "loss-of-function" experiment. Technically, it is very different than transgenesis, and much more time consuming. Conventional gene targeting approaches depend upon homologous recombination of the electroporated DNA into totipotent embryonic stem cells¹¹. The electroporated DNA usually contains a selectable marker and, after subsequent selection procedures, those cells which have undergone homologous recombination can be identified via suitable molecular screening procedures such as Southern analysis or PCR¹². There are now many variations on the basic theme, but the salient points remain more or less invariant. The DNA to be inserted consists of extensive regions of homology, which flan k the target site and contains at least one, and usually two selectable markers. The targeting event is planned such that the endogenous sequences are disrupted, with DNA elements critical for either or both the transcription/translation of the gene being targeted and replaced by either the selectable marker or an innocuous, non-coding sequence. This results in a null allele, that is, a gene locus that cannot produce either a translatable transcript and/or a functional polypeptide. The gene ablation or targeting allows one to study the consequences of recessive alterations if the subsequent animals are bred to homozygosity.

This strategy is most often used to determine if a particular gene has a critical function, or to explore the potential gene dosage effect by ablating a single allele. If the targeting results in a translatable, but altered transcript, a poi son peptide can result which interferes with the endogenous protein's function. resulting in a "dominant negative" targeting. Additionally, homologous recombination in embryonic stem cells can be used to insert, at a defined site, a large fragment of DNA incorporating a single entire transgene¹³ such that its activity and function(s) can be studied free of the confounding effects of copy number, and varying chromosomal contexts.

The literature now contains many examples of genes that have been ablated, and are important for general cardiac development and function. Frequently, the targeted genes encode polypeptides that are involved pleiotropically in cardiac development or basic structure. While the ablation can have a major impact on the contractile apparatus, the deficit/remodeling is not a primary consequence of the null mutation but rather reflects a more basic deficit that occurred in normal cardiac development. Examples include the hox-l.5 knockout, which results in widespread cardiovascular abnormalities reminiscent of DiGeorge Syndrome¹⁴ or ablation of the muscle-specific LIM protein¹⁵ which leads to a dilated cardiomyopathy, as well the endothelin-1¹⁶ and connexin 43 knockouts¹⁷. Ablation of these genes, whose functions are apparently needed for normal cardiac development and establishment of the myocardium's inherent structure have been tremendously informative, but are not directly germane to understanding the structure/function relationships which underlie normal and abnormal cardiac function in the adult.

The advantages of gene targeting over transgenesis can be substantial. First, the point of insertion and number of inserted sequences are both known. Second, the mutated DNA is inserted only at the defined site. Third, if a mutation, not an ablation is made, that mutation will be expressed under the control of the endogenous promoter, resulting in physiologically relevant expression levels, and mimicking exactly the endogenous transcriptional patterns of the gene under study. However, for a number of reasons, data dealing directly with remodeling the heart via ablation or modification of genes, particularly those encoding elements of the contractile apparatus, are limited.

Importantly, one might expect that ablation of a critical component of the contractile apparatus would be lethal. A functional heart is critical to fetal development by embryonic day (ED) 11-12 in the mouse. If the heart is seriously compromised, the fetus rarely survives past ED 12.5. Thus, homozygous animals carrying the null mutation at both alleles would not come to term but rather would die during embryogenesis and be resorbed¹⁸. Before one goes to the trouble of making a targeted animal it is also prudent to have the analytical infrastructure in place for the necessary analyses and these have been largely lacking for an assessment of cardiac function in utero. Our abilities to discern cardiac contractile function in the intact, mid-gestation embryo, as reflected by both stroke volume and heart rate have been quite limited but are developing rapidly¹⁹. A second concern is that gene targeting is very time consuming, and it is not unusual for an experiment to take years, (versus months if a transgenic approach is used). This is particularly true if one is trying to make a mutation in the gene, and not simply ablate it, as those experiments require two separate targeting events²⁰. Nevertheless, gene targeting remains the more elegant and intellectually satisfying of the two approaches as it is more precise.

GENETICALLY ENGINEERED MODELS OF CARDIOVASCULAR DlSEASE

Familial Hypertrophic Cardiomyopathy

The familial hypertrophic cardiomyopathies (FHC) include a group of primary cardiac muscle disorders characterized by a high incidence of morbidity and mortality. Their classification is based on their pathophysiology and etiology. Cardiomyopathy was originally defined as a non-coronary disease of heart muscle²¹. In 1980, the World Health Organization redefined the term to mean heart disease of undefined etiology. Although a number of different categories of cardiomyopathies are recognized (dilated, hypertrophic, restrictive/obliterative), our discussion will be restricted largely to FHC. These cardiomyopathies are inherited, rather than acquired as a secondary consequence to altered cardiac load²², are autosomal dominant, and cause hypertrophic cardiomyopathy, as measured by an increase in mass. FHC is characterized by unexplained cardiac hypertrophy without increased cardiac load, or in the absence of other systemic abnormalities. Clinically, the autosomal dominant diseases show variable penetrance with hypertrophy occurring in either ventricle; usually (>95% ), there is involvement of the intraventricular septum²³. Histologically, there is well-characterized disarray at both the myocyte and myofiber levels. Bizarre nuclear morphology and karyomegaly are also often present although these characteristics can be age dependent and may not present in children²⁴. The histopathology also often includes extensive fibrosis that can be either focal or interstitial in nature²⁵. As a group of diseases, the familial hypertrophic cardiomyopathies are somewhat common, with independent studies yielding estimates that 1:500 individuals are affected^26,27. However, FHC is a major cause of sudden death in otherwise healthy appearing young adults²⁸. A recent study estimated that, in the sudden cardiac death of 35 young adults, 12 were due to FHC²⁹. Even within a family in which the disease is due to a single genetic defect, the severity, onset and penetrance of the pathology is highly variable (presumably due to the existence of modifier loci, although this has not yet been formally demonstrated). Penetrance is age-related, with onset usually occurring in adolescence^25,30,31.

Beginning with the seminal paper of Geisterfer-Lowrance et al³², a genetic basis for the primary etiology of the disease began to be established. Using gene linkage techniques coupled with molecular biological approaches, and taking advantage of a large pedigree, the Seidman's established that FHC mapped to chromosome 14q 11. They demonstrated that a point mutation in exon 13 of the beta cardiac myosin heavy chain (MyHC) gene (β-MyHC) was present in all affected individuals from the kindred. The gene encodes the major motor protein responsible for pumping in the adult ventricle. The mutation converted a highly conserved arginine residue (Arg403) to a glutamine (Arg403Gln). Since that initial observation, over 50 different point mutations leading to various amino acid substitutions have now been mapped to β-MyHC mutations in this gene account for ~30-40% of FHC in the general population. The large number of different mutations, and their various locations in the MyHC molecule helps, in part to explain the disease's widely varying severity: different mutations have different effects on MyHC's ability to carry out its function. For example, the Arg403Gln mutation is highly penetrant and has severe effects on morbidity and mortality of the affected patient population. The residue is located near the nucleotide-binding pocket, and is at a site in the myosin' s head that participates in actin-myosin interactions. Thus, the mutation lies at and near critical sites for the motor' s functions³³.

The genetics of dilated and arrhythmogenic cardiomyopathies are not so well known/understood as FHC, but several genetic loci have been described, and a few genes identified ^34,35. These disorders, like FHC, may present heterogeneous clinical features and morbidity and other modifier genes undoubtedly influence mortality.

We now know that FHC is an autosomal dominant genetic disorder resulting from mutations in different genes encoding sarcomeric proteins that make up the thick, thin and titin filaments. Multiple mutations in β-MyHC, the essential and regulatory myosin light chains (ELC and RLC), myosin-binding protein C (MyBP-C), tropomyosin ( α- Tm), cardiac troponin T (cTnT), troponin I (cTnI), actin and titin have all been defined^8,36. Clinically, FHC typically induces hyperdynamic ejection, impaired relaxation, delayed early filling, myocyte disarray and fibrosis, and increased chamber end-systolic stiffness. Diagnosis is usually made based .on the physical examination and an echocardiogram but making an accurate diagnosis can be particularly difficult in children, who may not have cardiac hypertrophy until adulthood.

Although the molecular genetics of understanding the disease's etiology are quite advanced, the pathogenic processes that actually are responsible for morbidity and mortality have remained obscure. This is because their description must necessarily take place over prolonged periods of time, and in a population whose behavior and environment is uncontrolled. To mitigate these difficulties, a substantial amount of effort has been devoted towards generating relevant animal models in which disease progression can be studied closely in a controlled manner. Development of animal models has shed significant light into the pathogenesis of FHC, linking the genetic mutations to particular functional deficits. In theory , a mutant sarcomeric protein that is produced and presumably causes disease development can alter sarcomere function via at least two different mechanisms:

• It can function as a "poison peptide," exerting a dominant negative effect. The altered protein is incorporated in the sarcomere, leading to structural changes and the development of compensatory hypertrophy.

• It can act as a "null allele," potentially leading to haplo-insufficiency. The production of insufficient quantities of normal protein produces an imbalance in sarcomere stoichiometry , impairing sarcomeric structure and function.

Below, we will briefly consider relevant examples that illustrate the power of animal models in uncovering the primary disease deficits.

The Myosin Heavy Chain MyHC, by mass, is the major component of the thick filament in the sarcomere and provides the motor function for cardiac contraction. Myosin is a hexameric protein consisting of two heavy chains (Mr ≈ 220 kDa) and two pairs of non-identical light chains (Mr ≈18-27 kDa). The heavy chains are composed of two separate domains: a globular head joined to a α-helical rod by a hinge region. The heads, localized at the amino-terminal end of the molecule, contain the catalytic ATPase site, and an actin- binding site. Additionally, the light chains are bound to each head at the head-rod junction region, also called "neck", "hinge region" or referred as "subfragment-1" (S1).

Underlying the different motor functions of the different muscle types, the myosin heavy chains exist as a gene family³⁷ and, in the heart, two isoforms, termed alpha and beta are expressed. α and β, are encoded respectively by the MYH6 and MYH7 genes, and are organized in tandem in a cluster on chromosome 14q 11.2-q 13, with MHY7 being located 4 kb upstream from MHY6. MYH7 Is composed of 40 exons, and encodes a protein of 1935 amino acid residues (Figure 1). α and β isoforms are organized as homodimers V1 and V3 respectively. Under some conditions, an intermediate heterodimeric form, V2, can be present, made up of α and β gene products. V 1 and V3 display different intrinsic A TPase activities; V 1 being the most active (by about three to four-fold). These different isoforms are present in varying amounts, and there is a good correlation between the speed of contraction and the relative V1/V3 ratios. Thus, the human ventricle consists mostly of V3^38,39 while the adult mouse ventricle, which has a rate of 500-800 beats per minute, contains V1.

FIGURE 1: The FHC ß-MHC mutations. (A) The human mutations previously described are indicated, the red color referring to the more malignant ones. The intron-exon organization of the gene is represented and the affected exons are shown in blue. The mutations that have been engineered in animal models are shown below the intron-exon diagram. (B) Schematic representation of the ß-MHC protein encoded by the MYH7 gene, with its domains and binding sites. (C) Principle characteristics of animal models reproducing some of the human mutations, comparison with the human disease and disease pathogenesis. Missense mutations are represented by the letter encoding the amino acid (aa) residue that is mutated followed by the aa position number, followed by the letter of the new aa. Truncation mutations can arise as a result of splice site mutations (Splx; with x=position of the nucleotide mutation), deletion mutations (dely or ?1-y; with y=position number of the last amino acid residue present in the truncated protein), and frameshift mutations (Framz; with z=position of the nucleotide mutation). * indicates a complex substitution mutation involving MYH6 and MYH7 genes. Ex=exon, In=intron.

In the human population, multiple FHC mutations in β-MyHC exist. To date, more than 63 distinct mutations have been defined. These are spread throughout the molecule but, for the most part, are loca>ular head regio, in the head-rod junction, as well as a few sites in the rod region. Missense and nonsense mutations and deletions have all been reported, and this diversity is responsible, at least in part, for the widely varying clinical picture of FHC. Thus, the clinician sees tremendous differences in terms of the degree of hypertrophy, onset and evolution of the symptoms of the disease, occurrence of sudden cardiac death (SCD), and the patient' s prognosis (classified as benign or malignant). The effects of the different mutations on the different functions of the myosin heavy chain lead to different prognoses. The principle clinical mutations associated with a malignant prognosis include Arg403Gln^32,40-42, Arg453Cys^33,43,44, Gly716Arg^33,45,46, Arg719Trp^45,47,48, and, depending on the ethnic origin of the patients, Val606Met^42,44,49-51. The Arg403Gln substitution has dramatic consequences, with 50% of the affected individuals dying by 40 years of age^41,52.

In order to study the consequences of this mutation, the missense mutation, Arg403Gln (R403Q) was generated in the mouse using gene targeting⁵³. This mouse has subsequently been analyzed extensively over its lifetime and has yielded valuable information concerning the natural history of the disease. In the human, the mutation only presents in heterozygous form and, consistent with this observation, the homozygote mice do not live for more than a few days after birth⁵⁴. Sedentary heterozygotes survived, demonstrating that the presence of Arg403Gln myosin in the sarcomere leads to abnormal cardiac function but is compatible with life. The heterozygotes exhibited the same cardiac histopathology and pathophysiology^55-58 observed in human FHC. Cardiac dysfunction preceded histopathologic changes, and myocyte disarray, hypertrophy, and fozygotes showed a more severe phenotype than females, as assessed by gender-specific electrophysiologic abnormalities^{5easons for these gender differences remain obscure. Exercise capacity is compromised in the heterozygotes}53, In the future, this mouse model for FHC will continue to allow us to study the impact of background genotype, diet and physical activity on phenotype.

The functional consequences of the Arg403Gln mutation at the molecular level⁶¹ have been assessed: actin-activated cycling is accelerated, and this could account for the faster rate of rise in pressure observed in vivo⁵⁵, as well as the faster rate of force development in muscle strips⁶². Such a gain of function is consistent with the hemodynamic phenotype observed in humans, and is likely to stimulate the development of a compensatory hypertrophy that is not due to reduced power-generating capacity but instead is related to chronically increased energetic demands on the myocardium. In this model, compensatory hypertrophy would increase tissue mass to reduce wall stress and energy utilization per unit volume of myocardium. Ultimately, validation of this mouse model will come from experiments measuring force and sliding velocities of human β-MyHC and regulated thin filaments from the heart. Further investigations using these models should help to direct the clinical management of FHC in humans, and to define therapeutic targets and/or interventions designed to slow or reverse development of the hypertrophic phenotype for FHC and other cardiomyopathies (Figure 1).

How valid is it to extrapolate the mouse data to the human? To approach this question, Roopnarine et al⁶³ placed 3 human mutations in rodent α-MyHC myosin and examined their A TPase activities. Interestingly, the degree of enzymatic impairment of the mutant myosins correlated with the clinical phenotype of patients carrying the corresponding mutation. These data suggest that, despite specie differences, the animal models may well be predictive for the human. Another approach to answering this question is to generate larger animals carrying the disease locus. To this end, the Arg403Gln mutation was also generated in rabbit, which, like human (and unlike the mouse), expresses the same cardiac β-MyHC isoform (the two proteins share 98% of homology). Transgenic rabbits that over expressed the human mutant β-MyHC in the heart and had partial replacement of the normal protein with the mutant polypeptide, exhibited cardiac hypertrophy, myocyte and myofibrillar disarray and increased interstitial collagen content, but had normal systolic function³. There was a high incidence of premature death, and the observed phenotype was very similar to that of the human. The mutant protein exerted a dominant-negative effect, consistent with the genetics of the human population. Thus, the transgenic rabbit expressing the Arg403Gln mutant protein is a desirable model to study the pathogenesis and search for therapeutic targets for human HCM. The rabbit is also more suitable for noninvasive functional studies using echocardiography or electrophysiology. Although these techniques have been developed for the mouse⁶⁴, they remain restricted to a small number of specialized academic centers, and the rabbit model will certainly be more accessible to the general cardiology community for detailed study. In particular, extensive electrophysiological studies have already been performed in rabbit⁶⁵, providing useful data on the arrhythmogenic dysfunction that is common in FHC, especially the occurrence of sudden cardiac death.

The Myosin Essential (ELC) and Regulatory (RLC) Light Chains

Two unique species of myosin light chain, the essential (MLC1 or ELC) and regulatory (MLC2 or RLC) bind to the elongated neck region (regulatory region) of the myosin head. These small proteins appear to have both structural and regulatory roles in myosin function, stabilizing the long alpha helical neck of the myosin head and affecting its rigidity or stiffness⁶⁶. They belong to the superfamily of EF-hand proteins, which includes calmodulin and troponin C. Cardiac compartment-specific isoforms are expressed in most vertebrate species, including man. The ventricular myosin essential ( or alkali) LC contains 195 amino-acid residues (Mr ≈ 21-25 kDa), encoded by the MYLJ gene located on chromosome 3p21.2-p21.3 and is made of 7 exons (Figure 2). The regulatory LC is composed of 166 amino acid residues (Mr ≈ 25-37 kDa), encoded by the MYL2 gene located on chromosome 12q23-q24.3 and consisting of 7 exons (Figure 2). MLC2 is phosphorylatable through the PKA pathway.

FIGURE 2. The FHC Myosin Light Chain Mutations. (A) The human mutations previously described are indicated. The intron-exon organization of the gene is represented and the affected exons are shown in blue. The mutati ons that have been engineered in animal models are shown below the intron-exon diagram. (B) Schematic representation of the ELC and RLC proteins encoded by the respective MYL3 and MYL2 genes, with important binding sites. (C) Principle characteristics of animal models reproducing some of the human mutations, comparison with the human disease and disease pathogenesis. Missense mutations are represented by the letter encoding the amino acid (aa) residue that is mutated followed by the aa position number, followed by the letter of the new aa. Truncation mutations as splice site mutations are noted Splx; with x=position of the nucleotide mutation. Ex=exon, In=intron.

The concept that mutations in the myosin light chains might be causative for FHC was a logical outgrowth of the accumulating data that alterations in other sarcomeric proteins could lead to the disease. Additionally, a large cluster of MyHC mutations mapped to the domains that bound the light chains, implicating this region as being critically important to the normal function of the motor. Point mutations in both ELC and RLC have now been reported. Two missense mutations in the conserved amino acid residues: Met149Val and Arg154His in ELC^67,68 are associated with the disease.

Structure-function studies show that there is an increase in actin translocation velocity in an in vitro motility assay^67,68. Seven missense mutations plus one truncation mutation in RLC have also be en documented^67,69. More than half of these (Ala13Thr, Phe18Leu, Glu22Lys, Pro94Arg) occurred at highly conserved amino-acid residues, and modeling indicated that disruption of the phosphorylation site might result. This, in turn, could affect flexibility of the myosin neck. As is the case for the MyHC mutations, RLC mutations displayed variable expression and incomplete penetrance.

Both light chain mutations are associated with a rare and striking cardiac phenotype, which involves massive hypertrophy of the cardiac papillary muscles and adjacent ventricular tissue, causing a mid cavity obstruction⁶⁷. MLC mutations may interfere with the stretch-activation response of papillary muscle and adjacent ventricular tissue, a property only found in portions of the heart that increase power output⁷⁰. These changes might translate into impairment in the elasti city of the neck region of myosin, leading to a reduction in the oscillatory power of the papillary muscles, which is normally augmented by a strong str etch-activation response. The authors hypothesized that over time, a compensatory hypertrophic response takes place to increase power, but eventually obstructs the ventricular cavity.

In an attempt to validate the hypothesis that the papillary muscle hypertrophy associated with mutations in either light chain reflected a common effect on myosin function, a large ~ 12 kilobase human genomic fragment containing the mutated locus for the Met149V al essential myosin light chain was used to generate transgenic mice. The phenotype was recapitulated⁷⁰. This model provided the opportunity to study the stretch-activation response before the hearts were distorted by the hypertrophic process and confirmed that the stretch-activation response might playa role in the mammalian heart, providing a new way to modulate human cardiac function. Moreover, a novel transgenic mouse in which the phosphorylatable site was ablated showed the importance of this post-translational modification⁷¹. Indeed, pharmaceutical modulation of cardiac RLC phosphorylation may provide a new target to therapeutic intervention in heart failure.

As noted above, the ELC transgenic mice were made with an intact human genomic fragment. Thus, although the phenotype was accurately recapitulated, the experiment was subject to some ambiguity in terms of the Met149Val mutation being causative, as there were many other sequences present. In order to unambiguously establish a causal relationship for the regulatory and essential light chain mutations in cardiac hypertrophy, multiple lines of mice that expressed either the wild type or mutated forms, utilizing cDNA clones encompassing only the gene loci's coding regions were made. Surprisingly, when robust levels of expression resulted in 50% or greater replacement of endogenous protein with the mutated form, no hypertrophy was detected, even in senescent animals⁷² (in press). Although changes occurred at the myofilament and cellular levels, with the myofibrils showing increased Ca²⁺ sensitivity and significant deficits in relaxation in a transgene-dose-dependent manner, no overt hypertrophy at either the cardiomyocyte or chamber levels occurred. Because of the discordance of these data with the data obtained in the transgenic mice containing the human genomic fragment, the current view, that these point mutations by themselves cause significant cardiac hypertrophy, should be reassessed. The possibility remains that these point mutations may simply be closely linked to the real, yet undiscovered lesion in the human locus.

These data show the usefulness of transgenesis and animal models for establishing the unique structure-function relationships in the mutated proteins. Genetic remodeling allows the scientific community to test the causality of particular mutations. Such experiments have significant implications for clinical practice as, before wide spread screenings are contemplated, let alone implemented, establishing a direct cause-an-effect relationship for the particular mutation is imperative.

The Myosin Binding Protein C (MyBP-C)

MyBP-C is a large protein (Mr≈130 k Da) of 1173 amino acid residues. Considering its abundance in the cardiomyocyte (it can make up as much as 4% of the cardiomyocyte protein mass) and the 25 years that have passed since its discovery, its function(s) have remained surprisingly obscure. MyBP-C, a major component of the thick filament, binds to both the myosin (thick) and titin filament systems. Discovered over twenty-seven years ago⁷³, interest in the protein' s role( s) intensified after multiple mutations in the polypeptide were linked to familial hypertrophic cardiomyopathy³¹. MyBP-C is localized in the C region of the A band, and has an unique organization, with 7-9 axial bands in each half-sarcomere. Like other myosin binding proteins and titin, MyBP-C belongs to the intracellular immunoglobulin (Ig) superfamily and is composed of repeated Ig and fibronectin domains⁷⁴. In vitro modeling and reconstitution experiments, as well as experiments carried out using cell transfections, indicate that the protein probably plays an important role in assembling and maintaining the overall architecture of the sarcomere^75,76. The cardiac isoform, consisting of 35 exons (Figure 3), is encoded by the human MYBPC3 gene, which is localized on chromosome 11p11.2. Cardiac MyBP-C contains three isoform-unique domains, some of which may modulate contraction via phosphorylation by the protein kinase A or/and a calmodulin-dependent protein kinase.

FIGURE 3. The FHC Myosin Binding Protein C Mutations. (A) The human mutations previously described are indicated. The intron-exon organization of the gene is represented and the affected exons are shown in blue. The mutations that have been engineered in animal models are shown below the intron-exon diagram. (B) Schematic representation of the MyBP-C protein encoded by the MYBPC3 gene, with its domains and binding sites. (C) Principle characteristics of animal models reproducing some of the human mutations, comparison with the human disease and disease pathogenesis. Missense mutations are represented by the letter encoding the amino acid (aa) residue that is mutated followed by the aa position number, followed by the letter of the new aa. Truncation mutations can arise because of splice site mutations (Splx; with x=position of the nucleotide mutation); deletion mutations (?1-y; with y=position number of the last amino acid residue present in the truncated protein) or frameshift mutations (Framz; with z=position of the nucleotide mutation).

Mutations in the gene for cardiac MyBP-C account for approximately 15% of cases of FHC. A large number of mutations (currently >31) have been described, with most resulting in splice-site mutations, insertions and deletions^31,77-80, which produce truncated proteins. The truncations most frequently occur at the COOH-terminus, which contains a myosin-binding site, with the titin-binding site sometimes still present, sometimes not. Interestingly, presentation of the disease is often delayed until the fifth-sixth decades, and is characterized by late-onset HCM, incomplete penetrance, and a relatively favorable clinical profile⁸¹, contrasting with other, more malignant sarcomeric gene mutations. Because of these characteristics, longitudinal pathological changes linked to the disease state in the MYBP-C patients are hard to obtain.

The transgenic approach should provide a powerful tool for studying the longitudinal pathological processes involved in the disease's onset and progression late in life. This approach can also help determine basic structure-function relationships, and establish dose-response curves in which the onset, severity, and progression of the disease can be correlated with levels of the mutated protein. Indeed, transgenesis provides the researcher with the technology to distinguish between two fundamentally different causes for the disease: the "poison peptide" effect or functional haploinsufficiency⁷⁹. In this disease, haploinsufficiency was implied by the finding that the mutated protein was not found in human biopsies⁸⁰. This raised the possibility that the truncated protein was unstable and the disease developed, not because of the presence of mutated polypeptide, but because there were insufficient amounts of normal protein.

To assess these possibilities, as well as to create models in which pathogenic processes could be discerned over the animal's lifetime, two transgenic mouse models were generated. The first model replaced approximately half of the cardiac MyBP-C with a protein lacking the myosin and titin binding domains⁸². The encoded truncated protein was stable, but was not incorporated efficiently in the sarcomere, suggesting that the mutant protein may not act as a "poison peptide." The pathophysiological and ultrastructural abnormalities in the mice were dose-related and illustrated the structural consequences caused by the lack of enough functioning MyBP-C incorporation in the sarcomeric assembly and/or incorporation of the aberrant polypeptide. Sarcomere dysgenesis was prevalent, indicating that normal amounts of wild type protein were needed for the continued structural integrity of the cardiomyocyte's sarcomeres. The phenotype of these mice was surprisingly mild during development and early adulthood, reproducing the relatively benign clinical phenotype. Whole organ function was unaltered in young adult mice, while fiber mechanics indicated subtle alterations in force production. Moreover, the hypertrophic process was very low, in agreement with clinical data.

However, the pathophysiological mechanisms relevant to this phenotype may vary considerably, depending upon the particular MyBP-C mutation. To test this hypothesis, a second model of transgenic mice expressing in the heart a mutant MyBP-C lacking only the myosin-binding site was generated⁸³. As hypothesized, this mutation exhibited very different features. Only modest levels of protein were found in the heart, consistent with the human biopsy data⁸⁰. Despite normal levels of endogenous MyBP-C, sarcomere dysgenesis was also present at the ultrastructural level in this model, confirming that for this particular mutation, the mutated protein rather acts as a "poison peptide." Consistent with clinical observations, the phenotype at the whole organ and animal levels was quite subtle and difficult to detect before 1 year , although mild hypertrophy was present, and fiber mechanics impaired at early stages. However, transgenic mice displayed compromised exercise capacities associated with bradycardia, reminiscent of the hypotensive response found in exercised FHC patients⁸⁴. These two models are compared and contrasted in Figure 3.

The MyBP-C transgenic models illustrate some important points concerning the utility of the general approach. First, a small animal rodent model can accurately reproduce elements of the human pathology. Second, the relatively rapid rate with which these models can be created allows one to demonstrate the diversity of the pathophysiological mechanisms that can result in FHC, in which different mutations of the same gene (MYBPC3) are involved. Because of the limited investigations that can be done in human beings, mouse models provide a unique opportunity for uncovering pathogenic processes that evolve gradually over the lifespan of the animal.

Alpha-Tropomyosin ( α- Tm)

Tropomyosin (Tm) is a rigid rod-shaped protein that binds along the length of the actin filament and to the troponin complex. It regulates the calcium sensitive interaction of actin and myosin. In the adult heart, the predominant isoform is the striated muscle α-Tm isoform, composed of 284 amino acid residues (Mr ≈ 34-36 kDa). Alternative splicing from the gene, TPM1, generates α- Tm. The human gene is located on chromosome 15q22 and consists of 14 exons (Figure 4). α-Tm contains two TnT binding domains, one calcium dependent and one calcium-insensitive at the COOH-terminal region of the molecule that attach α-Tm to the troponin complex. α- Tm stabilizes and stiffens the filament, and in the absence of calcium, blocks the myosin-binding site of actin. It may also playa role in determining the degree of cooperativity and calcium sensitivity.

FIGURE 4. The FHC a-tropomyosin Mutations. (A) The human mutations previously described are indicated. The intron-exon organization of the gene is represented and the affected exons are shown in blue. The mutations that have been engineered in animal models are shown below the intron-exon diagram. (B) Schematic representation of the a-Tm protein encoded by the TPM 1 gene, with its binding sites. (C) Principle characteristics of animal models reproducing some of the human mutations, comparison with the human disease and disease pathogenesis. Missense mutations are represented by the letter encoding the amino acid (aa) residue that is mutated followed by the aa position number, followed by the letter of the new aa. Ex=exon.

Mutations in the TPM1 gene represent <5% of those responsible for FHC. Four missense mutations have been identified, two in exon 2 (Ala63Val, Lys70Thr) that could alter α-Tm binding to actin^85,86, and 2 in exon 5 (Asp175Asn, Glu180Gly), that may affect the calcium dependent binding to TnT^87,88. The clinical presentation associated with the different mutations is quite similar⁸⁵, but the phenotypic severity is not well established. Depending on the ethnic population, the importance of left ventricle hypertrophy, which can lead progressively to left ventricle dilatation and SCD varies substantially. Until recently, the Asp 175Asn mutation was the mutation most extensively studied: proof of mutant protein expression and incorporation in the sarcomere were performed using patient muscle and confirmed that the mutation acts as a dominant negative allele rather than a null allele⁸⁸.

To investigate the functional consequences of Tm mutations, transgenic mice expressing the missense mutation Asp 175Asn in the cardiac compartment were generated⁸⁹. Expression of the mutant protein was associated with a reciprocal decrease in endogenous α-Tm levels. With replacement exceeding 50% of endogenous α-Tm, contractility and relaxation, as evaluated in working heart preparations, were decreased. Using echocardiography, transgenic mice exhibited normal ventricular function, but responded less vigorously to exercise and β-adrenergic stimulation. However, life expectancy was normal, as is the case for some patients. Histologically, variable occurrence of myocyte disarray, hypertrophy and fibrosis were apparent. The mutation, which is located in a TnT binding area of α- Tm, may disrupt the already weak TnT –α-Tm interactions in this region, subsequently altering myofilament calcium sensitivity. To test this hypothesis, skinned fibers obtained from papillary muscles were isolated. The mechanics and kinetics of these fibers, both from mice mutant hearts, as well as from human muscle, showed that calcium sensitivity was increased⁸⁸.

Essential complementary data to help to understand the pathogenesis of FHC related to α- Tm mutations came from a knock-out mouse model in which the α-Tm gene was ablated by gene targeting⁹⁰. As expected, homozygous "knockout" mice were not viable.

However, heterozygous mice exhibited little or no change in cardiac function or structure, demonstrating that the mouse can easily deal with haploinsufficiency of the α-Tm gene. These data strongly suggest, assuming that human and murine cardiac muscles are similar, that cardiac abnormalities due to α-Tm mutations are not associated with haploinsufficiency, as can occur with α-MyHC gene ablations¹⁸.

Interestingly, total mRNA was decreased by 50% in heterozygous mice. However, mRNA bound to polysomes (that is, being actively translated into protein), as well as protein levels, were both similar to wild type littermates, suggesting that the relative amount of mRNA translated is not affected, and that translational regulation plays a major role in the control of Tm expression⁹¹. These findings imply that in heterozygotes, regulatory mechanisms maintain the level of myofibrillar Tm despite the reduction in α- Tm mRNA, and explain why inactivation of one α- Tm allele does not cause the pathology observed in FHC (Figure 4).

Cardiac Troponin T

Cardiac TnT (cTNT: Mr ≈ 37 kDa), the tropomyosin-binding subunit of the troponin complex, is an asymmetric molecule of 294 amino acid residues. In the human heart, four isoforms have been characterized: cTnT1-cTnT4^92,93 their expression depends on the developmental stage and pathophysiological conditions. All cardiac human isoforms derive from alternative splicing of the single gene, TNNT2, located on chromosome 1q32⁸⁷ and composed of 17 exons (Figure 5). Cardiac TnT has several domains: the NH₂-terminal region contains a phosphorylation site that is the target of PKC, as well as a binding site to the C-terminus of α-tropomyosin. The COOH-terminal region contains a calcium-dependent binding site to α-tropomyosin and binding sites for TnC, Tnl, and possibly actin. There are 3 potential phosphorylation sites, which may regulate crossbridge kinetics by decreasing the maximum A TPase rate⁹⁴. The protein plays an important structural role by positioning the troponin complex along the thin filament. In addition, TnT confers calcium sensitivity to the inhibitory activity of the cTnl-cTnC complex on the actomyosin ATPase⁹⁵.

FIGURE 5. The FHC Cardiac Troponin T Mutations. (A) The human mutations previously described are indicated. The intron-exon organization of the gene is represented and the affected exons are shown in blue. The mutations that have been engineered in animal models are shown below the intron-exon diagram. (B) Schematic representation of the cTnT protein encoded by the TNNT2 gene with its binding sites. (C) Principle characteristics of animal models reproducing some of the human mutations, comparison with the human disease and disease pathogenesis. Missense mutations are represented by the letter encoding the amino acid (aa) residue that is mutated followed by the aa position number, followed by the letter of the new aa. Truncation mutations can arise because of splice site mutations (Splx; with x=position of the nucleotide mutaor deletion mutations (?1-y; with y=position number of the last amino acid residue present in the truncated protein). Ex=exon, In=intron.

To date 14 mutations have been described; 12 are missense mutations distributed throughout the diversity , the clinical presentation is usually very similar , characterized by undetectable, mild or moderate hypertrophy, incomplete penetrance, a poor prognosis and a high incidence of SCD in adolescence or early adulthood. This highly malignant class emphasizes that development of cardiac hypertrophy can occur independently of sudden death, distinguishing this phenotype from mutations in the motor protein, myosin.

Two different human cTnT mutations have been modeled: a splice site donor mutation found in intron 15 of the genomic DNA^{87< /SUP>, and a missense mutation Arg92Gln}87. Transgenic mice that exp ress the splice site mutation produced, as expected, a carboxy-terminal truncated protein cTnT¹⁰⁰. They developed cardiomyopathy, exhibited myocellular disarray, diastolic and milder systolic dysfunction. The mutant hearts were smaller, because of decreased hyperplasia and smaller cardiomyocyte size. A better understanding of the pathogenesis of the disease arose from this model, as it was apparent that the truncated protein was incorporated in the sarcomere. As was the case for the unstable MyBP-C, the mutated cTnT was present only at very 10W levels ( <5% of the total cTnT), arguing for a "poison peptide" mechanism rather than a "null allele." Moreover, higher levels o( mutant protein (> 10% ) in the heart was lethal within 24 hours of birth, highlighting a dose-effect relationship between the amount of mutant cTnT and early mortality. Thus, the severity of the phenotype in animals Should be considered in comparison with the high incidence of SCD observed in patients, which may be related to the level of truncated protein expressed in the disease.

Faced with the diversity of mechanisms leading to the human disease, the effects of a different cTnT mutation, corresponding to the missense mutation Arg92Gln has been reproduced in 3 mouse models. These models, which expressed varying levels of mutated protein, exhibited different phenotypes, but disclosed a Common feature: the absence of hypertrophy. Consequently, the amount of mutated protein may play an important role in determining the phenotype of the disease, and may be involved in the heterogeneous presentation observed in a patient population which carries an identical mutation (alternatively, other, modifier genes may be responsible for the variable presentation). When only low levels of mutant protein were expressed, the mouse hearts displayed cardiac myocyte disarray, increased interstitial collagen content, and diastolic dysfunction ¹⁰¹, phenotypes similar to those found in human HCM. In such cases, human disease probably occurs via a dominant-negative effect in which altered cardiomyocyte function and disarray are the primary abnormalities and hypertrophy a compensatory process. When higher expression of the mutant protein was obtained in 2 separate laboratories, transgenic mice displayed a different phenotype, in which impairment in global cardiac function arose as a primary effect and occurred independently of, and prior to the development of histological abnormalities of HCM. Besides this crucial and common mechanism of action, each of the two transgenic models exhibited its own distinctive features, which might reflect the hetedriven by the strong mouse α-MyHC promoter ¹⁰², showed decreased left ventricle ejection fraction (L VEF), in contrast to the normal. LVEF found in human patients. This type of discordance may also represent one of the limits in the use of genetically engineered animal models, and in extrapolating results from mouse to human disease. In another transgenic mouse model, which used a mouse mutant cDNA driven by a rat α-MyHC promoter ¹⁰³, high levels of the mutant protein (30, 67 and 92% of the total cTnT) were associated with dose-dependent alterations.

These included smaller left ventricles, significant induction of atrial natriuretic factor and β-MyHC transcripts in the hearts (a transcriptional pattern characteristic of a hypertrophic response), and mitochondrial pathology. However, when high levels of mutant cTnT are expressed, the mutation might act as a "change in function" rather than a "poison peptide," explaining the phenotypic differences observed between models exhibiting various levels of mutant protein. In each of these models, it is important to emphasize that the overall protein stoichiometry was unchanged. That is, the amount of mutant cTnT + endogenous cTnT, equalled the levels ofs show that different cTnT alleles are associated with distinct phenotypes, su models will be complemented by studies using gene therapy. This will allow temporary expression of a mutated protein in the heart. Recently, Yu and colleagues have expressed mutant Arg92Gln cTnT in th studies, the mutant protein was incorporated in the sarcomere without inducing early myocyte and sarcomere disarray, suggesting that prolonged myofibrillar incorporation of the mutant cTnT is essential for its dominant-negative effect on cardiac myocyte structure in intact myocardium.

Cardiac Troponin I

TnInhibitory is the inhibitory subunit of the troponin complex. The heart contains a specific cardiac isoform that is characterized by an NH₂-terminal extension of 33 aminoacid residues as compared to the skeletal isoforms. TnI binds to actin, preventing myosin ATPase activity. Inhibition by this subunit becomes reversible in the presence of calcium that binds to cTnC and induces a conformational change in the troponin complex leading to the release of the inhibitory activity of the cTnl, allowing contraction. The cardiac isoform exhibits unique properties, including cooperative binding to actin-tropomyosin¹⁰⁵. The 210 amino acid polypeptide (Mr ≈ 30 kDa) is encoded by the TNNI3 gene, which is located on chromosome 19pI3.2-q13.2, and is composed of 8 exons (Figure 6).

FIGURE 6. The FHC Cardiac Troponin I Mutations. (A) The human mutations previously described are indicated. The intron-exon organization of the gene is represented and the affected exons are shown in blue. The mutations that have been engineered in animal models are shown below the intron-exon diagram. (B) Schematic representation of the cTnI protein encoded by the TNNI3 gene with its binding sites. (C) Principle characteristics of animal models reproducing some of the human mutations, comparison with the human disease and disease pathogenesis. Missense mutations are represented by the letter encoding the amino acid (aa) residue that is mutated followed by the aa position number, followed by the letter of the new aa. Ex=exon.

Cardiac TnI contains several functional domains. At the amino terminus is an extension that contains 2 sites, which can be phosphorylated by PKA (serines at position 23 and 24 ). Phosphorylation decreases the calcium sensitivity of the myofilament, and inhibits the cooperative binding to actin. The serines at positions 42 and 44 can be phosphorylated by PKC: phosphorylation reduces the maximum A TPase rate. An inhibitory region, which binds to actin and cTnC, causes relaxation through inhibition of the actomyosin interaction. Finally, a COOH-terminal region appears to be essential for the calcium sensitivity of the myofilaments^106.

Mutations in cTnl are also associated with HCM¹⁰⁷. To date 8 mutations have been described, 7 missense mutations and one deletion, all located between exons 5 and 8. Three mutations (Arg145Gly, Arg145Gln, and Arg162Trp) are in the inhibitory region, the other ones lying in the COOH-region of the molecule. However, the functional consequences of all these mutations remain unknown. They result in ventricular hypertrophy that can be, in some cases, restricted to the apex¹⁰⁷.

In an attempt to elucidate the mechanisms by which mutations in cTnI can lead to the development of HCM, in vitro analyses using reconstituted human cardiac troponin complexes expressing the mutant cTnl's, Arg145Gly and Arg136Trp have been performed. These two mutations decreased A TPase inhibition under relaxing conditions, and increased the enzyme's calcium sensitivity. In vivo, this may translate into impairment of relaxation and altered contractility, which may provide the hypertrophic stimulus leading to HCM¹⁰⁸. James and colleagues¹⁰⁹(submitted) also assessed the pathogenesis of the disease, in vivo, by generating transgenic mouse carrying the missense mutation Arg145Gly. The animals developed pathology similar to human FHC, exhibiting cardiomyocyte hypertrophy and interstitial fibrosis. Functional alterations at the whole organ level included hypercontractility with diastolic dysfunction, and were characterized at the fiber level by increased sensitivity to calcium. However, the phenotype in the animals was much more severe, with frequent SCD and a dramatically shortened lifespan. This difference may be attributed to the characteristics of mouse heart, which must rapidly cycle through systole and diastole, and thus might be more sensitive (than the human's) to perturbations in the calcium handling apparatus (Figure 6).

Sensitivity of the mouse heart to abnormal troponin proteins is thus supported by mutations in multiple cardiac troponin subunits ( cTnT, cTnl). The mouse models provide an important tool to understand the resultant pathogenic structure-function relationships and highlight the differences in phenotype severity of the troponin mutations between human and mouse heart.

Familial Dilated Cardiomyopathy

Familial dilated cardiomyopathy (FDC) is an autosomal dominant genetic disease that represents 20-30% of idiopathic dilated cardiomyopathies DCM), accounting for 60% of all cardiomyopathies. Idiopathic dilated cardiomyopathies are an important cause of morbidity and mortality, and a common cause of heart failure leading to cardiac transplantation. FDC is characterized by ventricular dilation and decreased systolic function; the diagnosis is usually based on echocardiography. SDC is frequent, brought about from either ventricular arrhythmias or ischemia. As is the case for FHC, FDC is both clinically and genetically heterogeneous. S ix different genetic loci have been described: lpl-lql¹¹⁰, 1q32¹¹¹, 2q31¹¹², 3p22-25¹¹³, 9q13-q22¹¹⁴, and 10q21-23¹¹⁵, but only actin has been unambiguously identified as causing FDC³⁴. Chromosomal mapping has been hampered by the small sizes of the available families. Mapping of potential disease loci to chromosomes lp1-1ql and 3p25 also is made more difficult by age-related penetrance. The locus 1q32 is rich in candidate genes, including muscle transcription factor MEF-2, renin and the helix-loop-helix DNA binding protein MYF-4. The locus 1 p 1-1 q 1 contains the gene encoding the gap junction protein connexin 40, but this has not yet been shown causative. Olson et al ³⁴ have recently identified 2 missense mutations in actin that cause DCM; this represents the first autosomal disease-related gene for DCM.

Mutations in the cytoskeletal protein desmin can also cause a dilated cardiomyopathy³⁵ and recently one of the mutations has been modeled in transgenic mice¹¹⁶.

Desmin is a muscle-specific member of the intermediate filament family of cytoskeletal proteins, and is expressed in striated and smooth muscle. The protein consists of 470 amino acids (Mr ≈ 53 kDa) and is encoded by a single gene located on chromosome 2q33-q35. It contains nine exons. It is a fundamental component of the Z disc and contains three domains: a non-helical NH₂-terminal head, a helical rod domain, which is critical for filament assembly, and a COOH-terminal tail, whose function remains unknown. In mature striated muscle fibers, the desmin filament lattice surrounds the Z-discs, interconnects them to each other and links the entire contractile apparatus to the sarcolemmal cytoskeleton, cytoplasmic organelles and the nucleus¹¹⁷.

Seven missense mutations in the coding region of the desmin gene have been found, six clustered in the COOH-terminal rod^118,119 and one in the COOH-tail domain³⁵: the latter may reflect a distinct functional domain as only disease in the heart was noted. Two additional deletions involving either 7¹²⁰ or 32¹²¹ amino acid residues in the rod region have also been reported. The desmin-related cardiomyopathies are a heterogenous group of diseases that are characterized by intracytoplasmic accumulations of desmin^122-125. This uncommon, often familial, disorder affects skeletal and/or cardiac muscle, and is sometimes associated with intestinal or vascular smooth muscle^125,126 defects. Skeletal muscle abnormalities are characterized by muscle weakness and atrophy, while cardiac disease often presents as conduction block, arrhythmia and restrictive heart failure. Desmincardiomyopathies can consist either of dilated or restrictive cardiomyopathy with a poor prognosis after early onset in adulthood¹²⁷.

In an attempt to investigate the structure-function relationships of the desmin mutations, and their ability to cause desmin-related myopathy, two mouse models have been created using either gene targeting through homologous recombination, or transgenesis. Generation of desmin null mice resulted in viable and fertile animals, demonstrating that desmin is not needed either for the formation of the heart or the alignment of functioning myofibrils¹²⁸. However, desmin plays an essential role in the maintenance of myofibril, myofiber and whole muscle tissue structural and functional integrity, as confirmed by cardiac, skeletal and smooth muscle abnormalities early after birth^129,130. The consequences were most severe in the heart, which exhibited progressive degeneration and necrosis with extensive calcification and fibrosis. Cardiac alterations included the development of a concentric cardiomyocyte hypertrophy, with induction of embryonic gene expression. Ventricular dilatation and compromised systolic function appeared later in life¹³¹. However, the relationship of the desmin null mutation to desmin related myopathy is not clear, as the total desmin complement in human cardiac disease is significantly increased.

The genetics of the disease in the human patient population were confusing, as it appeared that at least some of the mutations did not behave as dominant negatives. To determine whether desmin mutations were directly causative of desmin related myopathy, transgenic mice overexpressing the 7 amino acid residue deletion¹²⁰ were generated¹¹⁶. Mice expressing the mutant desmin displayed aberrant intrasarcoplasmic, desminpositive aggregates characteristic of human desmin related cardiomyopathy (Figure 7). The overall morphology of the desmin network was significantly affected. In cardiomyocytes, myofibril alignment was perturbed, and nuclear shape distorted. Both contractile and relaxation functions of left ventricle myocytes were significantly compromised, as well as whole organ function. The mutant mice also developed significant concentric hypertrophy. Thus, this model demonstrated that the mutation acts in a dominant negative way, and the mutant protein is causative for the resultant cardiomyopathy. In addition, new insights in the progression of the disease, the cause and effect relationships underlying the pathogenic processes should arise from further analyses.

FIGURE 7. Electron Microscopy of Desmin Aggregates in DCM Mouse Model. Sections of hearts from 8 week control (A), wild type desmin overexpression (B) and a transgenic animal that is over-expressing a desmin found in a DCM patient (C). All samples are derived from the left ventricle. In both (A) and (B), sarcomere organization is normal with clearly distinguishable I- and A-bands, and M-lines as well as regularly aligned Z-bands. (C), Accumulation of electron dense aggregates (*) surrounding the nucleus affects the regular organization of the sarcomeres. Z-bands alignment is perturbed. At higher magnification (D), desmin aggregates contain electron dense material (*) and intermediate filaments (arrow). Arrow head: actin and myosin filaments. N = nucleus.

Future progress will depend on the discovery of new families whose disease loci can be accurately mapped so that candidate genes can be identified and tested in the animal models but, at the present time, the molecular mechanisms leading to autosomal dominant FDC remain largely unknown. Moreover, the understanding of the pathogenesis of autosomal dominant inheritance can be even more complicated, considering that a least a subset of FDC cases arise as secondary diseases, resulting from skeletal myopathies, conduction defects or other systemic disorders. FDC occurrence can also be associated with autosomal recessive, X-linked and mitochondrial inheritance^132-134. Although the molecular mechanisms of some of these associated pathologies have been clarified by experimentation and animal systems, including transgenic mouse models^135-137, progress in applying the data to the mechanisms underlying autosomal dominant inheritance s extremely limited.

Molecular genetics of inherited DCM appear much more diverse than FHC. Other cell structures and organelles in the cardiomyocytes such as the cytoskeletal system and the extracellular matrix are probably involved. Genes for matrix components, growth factors, receptor signaling Proteins, transcription factors, may all be candidates, each presenting new possible mechanisms for causing autosomal dominant DCM. In fact, over-expression of a large number of these different proteins via cardiac-specific expression in transgenic animals, can lead to dilated cardiomyopathy^7,15,138-142. Future identification of the role and targets of these proteins using transgenic and gene targeted animal models will provide an explanation for many inherited and acquired DCM's.

Congenital Heart Disease

Although not the focus of this review, this class of disease bears mention simply because of its prevalence in the pediatric population. Congenital heart disease (CHD) includes a number of different disorders, classified as myocardial, vascular and arrhythmogenic defects, and accounts for the most common birth defects, affecting nearly 1% of live births¹⁴³. Although many human cardiac malformations are well characterized anatomically and physiologically, the genetic bases of these abnormalities, which arise during cardiac morphogenesis, are still poorly understood. With the recent advances in genetics and the development of molecular technologies, especially linkage analysis and positional cloning, disease-related genes have been and/or are in the process of being identified. Difficulties persist because the proteins involved in this syndrome are mostly unknown. As these genes are conserved during evolution in vertebrates, many candidates have first been identified by studying cardiac regulation and development in simple animal systems that are amenable to detailed genetic analyses, such as flies, nematodes and zebrafish. The candidates are then subsequently validated in the mouse. The ability to genetically manipulate the simple organism s and perform saturation mutagenic studies offers an opportunity to understand the potential players' roles in heart development and their contribution to heart defects. These studies have led to the identification of numerous proteins, both structural and regulatory , which can play major roles in the normal and abnormal formation of the heart^138,144-146.

FUTURE DIRECTIONS

In the past 15 years, transgenesis and gene targeting have proven to be the methods of choice for carrying out genetically based experiments in mammals. Improvements in the technologies should continue to provide finely tuned methods to explore cardiovascular diseases using animal models. With respect to transgenesis, the most pressing issue concerns an ability to express transgenes in a reversible manner; that is, to be able to control transgene expression at any moment in the animal's lifetime. A technique that offers great promise is the use of tetracycline to reversibly control gene expression^147,148.

The paradigm as described is a binary system, requiring two transgenes: 1) a cardiacspecific promoter driving the tetracycline transactivator (tTA) sequence coupled to the transcription activator virion protein 16 (VPI6), and 2) a cytomeglovirus (CMV) minimal promoter coupled to multimers (5- 7 copies) of the tetracycline operon (tetO), This chimeric promoter is then ligated to the target transgene that is to be conditionally controlled. In the presence of tetracycline, the tT A-VP16 binds to the drug. Transcription of the reporter gene, dependent upon the very low basal activity of the minimal promoter is "off." If tetracycline is not present in the system, tTA-VPI6 binds to the tetO allowing the VPl6 transactivator to increase expression driven by the CMV minimal promoter. Expression of the reporter gene can thus be turned on and off by removing or supplying tetracycline. Conditional transgenesis would be a particularly informative tool for remodeling the cardiovascular system since a significant portion of human cardiac pathology is related to compensatory mechanisms inherent to the plastic nature of the heart. Alternatively, in a different tet system, the tet operator sequences have been fused to a promoter, and the tetR converted into a transactivator protein that activates transcription continuously. In the presence of tetracycline, tTA can not bind to the tet operator sequences, resulting in repression of transcription. Conditional transgenesis in the heart has been demonstrated^149-152 although it is not yet clear if the current systems, in which transgene expression is controlled by the administration of tetracycline or other effectors, is robust enough to be used to control large protein pools necessary to influence sarcomere protein content. The inducible system clearly offers the potential for a tremendously powerful tool in exploring cardiovascular disease as it offers the researcher a chance not only to observe whether a particular transgene can cause a disease, but whether, when the transgene is turned off, the heart is able to recover. Different treatments can be tried in order to determine the most effective methods to reverse the disease.

Advances in gene targeting will be centered on making the ablation event organspecific. A significant concern with gene targeting as it is presently done is the lack of target organ specificity. That is, the targeting event is propagated in all somatic cells and can, if the target gene is critical to the function of other organ or muscle systems, complicate or even confound one's ability to dissect out the direct effects on cardiac function. To overcome this difficulty, attempts have been made to develop organ specific gene modification using the DNA recombinase enzyme family and significant progress with the Cre recombinase has been reported. The approach is disarmingly straightforward. The DNA recombinase, Cre, can excise DNA in the mammalian genome with high efficiency by binding to its target sequence: a 34 base pair DNA substrate termed loxP. By flanking a target sequence with loxP sites on each side, and driving expression of Cre specifically in the heart using cardiac specific promoters, it should be possible to restrict a targeting event only to the heart by breeding a mouse carrying the loxP-flanked DNA to another line which expresses Cre in the cardiac compartment. A number of laboratories have made substantial progress in the creation of the requisite mouse lines, and positive results of a cardiac-specific gene targeting have recently been reported^141,153,154. In fact, the system appears to work quite well; when the mice are bred to one another, in the animals containing both loci, the endonuclease is specifically expressed in the heart. The enzyme then recognizes the targeted locus by binding to the lox sites flanking the region and excises the piece of DNA. The result is a gene ablation-but only in the heart (or cardiac compartment where the endonuclease is specifically expressed). The procedure is quite effective and opens up even more possibilities for carrying out precise structure-function and loss-of-function studies only in the heart.

Variations on these themes, not yet foreseen, should allow us to modulate heart function by reversible interventions at the gene and protein levels such that the molecular bases for normal and abnormal cardiac function can be determined. These, in turn, will be used to explore the normal and abnormal cardiac physiology. In all of these studies, as major and minor effects are noted, it is important to keep in mind that a mouse is not a man. Extension of some of the most valuable models to large animal transgenics will provide both a measure of the relevance of the conclusions drawn from the murine models and more suitable models for testing potential treatment modalities.

Send proofs to:

Jeffrey Robbins Division of Molecular Cardiovascular Biology Children´s Hospital Research Foundation Cincinnati, OH 45229-8098 Phone: [phone] Fax: [phone] E-mail: [e-mail]