Kabaeva, Zhyldyz: Genetic analysis in hypertrophic cardiomyopathy: missense mutations in the ventricular myosin regulatory light chain gene

54

Chapter 4. Discussion

The ventricular myosin essential (MYL3) and regulatory light chain (MYL2) genes were analysed in a group of 71 unrelated clinically well-characterized HCM patients. Systematic analysis revealed two missense mutations in MYL2 associated with either benign or malignant HCM phenotype. Additionally, one silent mutation, two single nucleotide polymorphisms (SNPs), and a number of sequence differences were detected while screening the MYL2 and MYL3 genes.

4.1 Patient cohort and screening approach

The patients enrolled in the present study revealed typical features of HCM and were well representative of the overall HCM population.56 Similarly to other studies, the age of the patients varied widely, however, most of them were already at midlife at the time of diagnosis.5,44,66 The majority of the patients had mild or no symptoms and were diagnosed after the third decade of their life. In most cases, LV hypertrophy involved the entire IVS or both IVS and the free wall. These two patterns of hypertrophy distribution have also been previously described as the most common in HCM.45 Despite the high number of patients with obstructive HCM in the present cohort, operative management of obstruction was performed only in few of them. This suggests that the number of patients with massively increased pressure gradient and in the need of operative treatment in the overall population of HCM patients is small in comparison with the number of individuals, who can be treated by drugs.

Among a variety of available techniques,67 the PCR-SSCP method used in the present work has been shown to be a reliable and informative method for detecting unknown mutations in DNA fragments of interest.64,67,68 Although it has been argued


55

that the 80-90% sensitivity of SSCP analysis does not exclude a possibility of some mutations remaining undetected, there is a common agreement that this method is the most suitable for mutation screening in a large patient group, because it is economical, rapid and simple in carrying out. Moreover, the sensitivity of SSCP analysis can be significantly increased by adjusting the running conditions.63,69 In the present work, two SSCP runs under different conditions were used to screen each of coding exons in order to achieve a high detection rate and to decrease the number of missed genetic variants. Furthermore, since the SSCP analysis is more sensitive for shorter DNA fragments,64 most of the PCR fragments were 150-400 base pairs in length. The efficiency of SSCP analysis in the present study is supported by the identification of such minor DNA changes as SNPs and point mutations. Notably, results of SSCP analysis were always consistent with the results of DNA sequencing: each of the observed aberrant SSCP patterns had an underlying sequence variation. By contrast, sequence differences did not show any aberrancy on SSCP. Observation of high number of sequence differences also underlines the possibility of frequent errors in reference sequences, and how careful one should be in using them.

It is necessary to note that the present work did not aim to study genetic polymorphisms but disease-causing mutations. SNPs were determined in the course of this study, because they mimic disease-causing mutations on the SSCP analysis. SNPs are common single nucleotide allelic variations, which are present at least in 1% of a population.70 According to recent studies, SNPs occur on average every 1,000-2,000 nucleotides.70-72 It is supposed that they account for much of the functional heterogeneity in gene expression and protein activity exhibited in the human population.72,73 In contrast to disease mutations, SNPs do not directly cause any disorder, however, recent studies showed that SNPs or particular combinations of them might be associated with individual susceptibility to common polygenic disorders (diabetes, cancer, cardiovascular and neurological diseases, and others).57,71,74,75

In the present study, the exact distribution of the observed SNPs among the 71 examined patients was not estimated, because, firstly, it was not consistent with the study purposes, and, secondly, it would have required sequencing of corresponding DNA fragments in all patients. However, recurrent observation of the SSCP patterns characteristic of g.8393 G>A and g.8580 C>T/A indicates that these polymorphisms are common sequence variations. According to SSCP analysis, the frequency of these


56

variants in this study population was much more than 1%. Interestingly, the g.8393G>A polymorphism was also observed in South African as well as Danish HCM patients and controls.16 Thus, further studies are needed to determine the frequency of the MYL2 g.8393G>A and g.8580C>T/A polymorphisms and whether they have any implication in HCM or other diseases.

4.2 The Glu22Lys and Arg58Gln mutations in MYL2

HCM is caused by mutations in nine genes encoding for sarcomeric proteins including those for ventricular myosin light chains, although the frequency of mutations in each of these genes is variable.3 In the present study, screening of 71 unrelated HCM patients revealed only two mutations in MYL2 and no mutation in MYL3. These data underline the rarity of ELC/RLC mutations in HCM. The absence of mutations in MYL3 confirms that the contribution of ELC mutations to the HCM causes is significantly less than that of RLC mutations.15,16 In a study, which was previously conducted in our lab, an independent group of 85 HCM patients was screened, and neither MYL2 nor MYL3 mutations were found (unpublished data). Collectively, 186 unrelated HCM patients (71 patients from the present study and 85 from the previous study) underwent genetic analysis in two independent studies in our lab: the frequency of MYL2 mutations in this relatively large patient cohort is approximately 1%. These data are consistent with data obtained by Potter et al.10 but not with data from later studies, which estimated the frequency of MYL2 mutations as 4.4%15 and 7%.16

So far, only three disease-associated mutations have been identified in the ELC gene.10,59 In the RLC gene, seven point and one splice site mutations have been detected.10,15,16 These mutations and available information on associated phenotypes are listed in table 4.1. In contrast to some other genes, limited number of families confounds generally applicable conclusions regarding the disease course and prognosis in HCM caused by myosin light chain gene mutations. Therefore, in rare HCM forms, every single family, identified, becomes valuable.


57

Table 4.1. The known HCM associated mutations in the RLC and ELC genes

Gene

Exon

Mutation

Number of families

Associated phenotype

Reference

MYL2

2

Ala13Thr

1 individual

midventricular hypertrophy

Poetter et al,199610

 

 

 

 

 

 

 

2

Ala13Thr

1 family

septal hypertrophy,

Andersen et al, 200116

good prognosis and survival

 

2

Phe18Leu

1 family

septal hypertrophy

Flavigny et al, 199815

 

2

Glu22Lys

2 brothers and 1

midventricular hypertrophy

Poetter et al,199610

unrelated individual

 

2

Glu22Lys

1 family

septal hypertrophy, benign course and prognosis

present study

 

3

Asn47Lys

1 individual

midventricular hypertrophy

Andersen et al,200116

 

4

Arg58Gln

2 families

septal hypertrophy,

Flavigny et al, 199815

2 sudden deaths

 

4

Arg58Gln

1 family

septal hypertrophy,

present study

malignant course/prognosis

 

5

Pro94Arg

1 individual

no phenotype is described

Poetter et al, 199610

 

5 and

Lys103Glu and IVS6-1*

1 family

septal hypertrophy,

Andersen et al, 200116

Intr. 6

good prognosis and survival

MYL3

3

Ala57Gly

2 families and 1

septal hypertrophy, 2 sudden deaths

Lee W-H et al, 200159

unrelated individual

 

4

Met149Val

1 family

midventricular hypertrophy

Poetter et al,199610

 

4

Arg154His

1 individual

midventricular hypertrophy

Poetter et al,199610

Note: the mutations identified in the present work are shown in bold; Intr. 6, intron 6; *IVS6-1, a splice site mutation in intron 6.

This study presents two German families with the Glu22Lys and Arg58Gln mutations in the ventricular myosin regulatory light chain gene. The Glu22Lys and Arg58Gln variants have been previously observed in American and French HCM population, respectively.10,15 The Glu22Lys mutation was identified in two brothers and one unrelated individual by Potter et al.10, whereas the Arg58Gln mutation was detected in two unrelated families by Flavigny et al.15 In contrast to the previous study, this work presents a larger family spanning across three generations (family K) bearing the Glu22Lys mutation. However, the identified family with the Arg58Gln mutation (family B) was smaller than the two families described by Flavigny et al.15 Two individuals in family B, who died suddenly, probably did had the Arg58Gln mutation, because, similarly to proband II-2, they suffered from HCM. Moreover, the proband and her deceased sister eventually inherited the Arg58Gln mutation from their deceased father, since their mother was genetically and clinically healthy.


58

That these two mutations cause HCM is supported by several observations. First, there was a clear cosegregation of the Glu22Lys and Arg58Gln mutation with HCM in the present and previous studies: the mutations were present in all clinically affected family members. Second, neither of these mutations was observed in control individuals (105 controls were screened in the present study) indicating that they are not common polymorphisms. Third, the altered residues as well as flanking sequences show strong evolutionary conservation across vertebrate species suggesting an important role for the RLC function (figure 3.7). Furthermore, the RLC carrying these mutations is a protein of the cardiac sarcomere: the causality of mutant sarcomeric proteins in HCM is well established.2,3

The identification of the Glu22Lys and Arg58Gln mutations in the different study populations suggests that the codons 22 and 58 are highly susceptible to mutations. Such hot spots have also been observed in the beta--myosin heavy chain gene: mutations often affected codons encoding amino acids 403, 719 and 741.50 Interestingly, different mutations were observed at these sites. Associated HCM phenotypes were generally different for distinct mutations but showed high similarity for the same mutations.50 As discussed further, the HCM phenotypes observed in the present and previous studies were similar for the Arg58Gln but not for the Glu22Lys mutation.

4.3 Genotype-phenotype correlations

In the previous study by Potter et at.10, the Glu22Lys mutation was associated with a particular phenotype with massive hypertrophy of the cardiac papillary muscles and adjacent ventricular tissue causing midcavity obstruction, whereas the Arg58Gln mutation caused typical asymmetric septal hypertrophy in the study by Flavigny et at.15 In the present study, the individuals bearing the Glu22Lys mutation had no massive midventricular hypertrophy with midcavity obstruction but asymmetric hypertrophy of interventricular septum. The individuals of family B with the Arg58Gln mutation also had asymmetric septal hypertrophy similar to that described by Flavigny et at.15 These observations suggest that identical mutations in myosin light chains can cause diverse patterns of LV hypertrophy as mutations in other genes do. Furthermore, the midventricular hypertrophy was also observed in HCM cases caused by mutations in beta--


59

myosin heavy chain and cardiac myosin binding protein C and is, therefore, not an unique feature of myosin light chain mutations.16,59

Concerning the disease course and prognosis, a recent study suggested that myosin light chain mutations cause only benign HCM phenotypes,16 however, the present work shows that clinical course and prognosis associated with myosin regulatory light chain mutations can differ markedly. The Glu22Lys mutation was associated with a benign phenotype. Two affected individuals had reached an advanced age. There was no case of sudden cardiac death in the family, and most of the family members had only mild hypertrophy, a late onset of symptoms or no symptoms at all. In contrast, the Arg58Gln mutation identified in family B was associated with two cases of premature sudden cardiac death. Furthermore, myocardial hypertrophy developed in early childhood and was accompanied by disease symptoms as premature fatigue and later on as arrhythmias. Flavigny et at.15 also reported two sudden deaths at young age in one of the identified families with the Arg58Gln mutation. Thus, pooled data from the study by Flavigny et at.15 and the present work suggest that the Arg58Gln mutation indeed may cause a malignant HCM phenotype with a high risk of sudden death and, therefore, could be added to the panel of mutations associated with a poor prognosis (see table 1.1). Genotyping for such mutations could be recommended in order to improve risk stratification in HCM patients and early diagnosis of individuals in the need for prophylactic therapy. Apparently, the identification of more families with the Arg58Gln mutation will be of value in proving these observations.

The penetrance of HCM was demonstrated to vary widely.76 Complete penetrance of familial HCM was shown to be a feature of some malignant mutations in beta--myosin heavy chain.54 Low disease penetrance is characteristic of mutations in cardiac troponin T77 and myosin binding protein C.52 Furthermore, the penetrance of cardiac myosin binding protein C mutations were shown to be age related: generally HCM develops after midlife.78 Variable disease penetrance was also described for some previously identified RLC/ELC mutations.15,59 Similarly, in family B and K the respective mutations penetrated to the HCM phenotypes differently. Although the Arg58Gln mutation was associated with complete disease penetrance, family B is too small to draw final conclusions. The Glu22Lys mutation, in contrast, showed reduced HCM penetrance (57%): among seven genetically affected individuals only four had apparent HCM. However, given the age-related penetrance of cardiac myosin binding


60

protein C mutations, it can be assumed that "healthy carriers" of family K may still develop HCM later on in their life and therefore need to be followed up.

It was also shown that not all HCM phenotypes, when expressed, manifest to the same degree, for instance, with regard to magnitude of LV hypertrophy and severity of clinical course.50,52 Such variable expressivity is a typical feature of HCM and on certain extent, depends on the nature of the causative gene/mutation. Thus, malignant mutations in the beta--myosin heavy chain gene, apart from complete penetrance, showed significant myocardial hypertrophy.54,79 Cardiac troponin T gene mutations were generally associated with mild hypertrophy.53,77 The extent of hypertrophy caused by mutations in the myosin light chain genes varied form mild to massive.10,15,59 The Arg58Gln was associated with moderate hypertrophy in the previous and present studies, whereas the Glu22Lys mutation caused massive hypertrophy in the earlier study but not in this work.

The expressivity of HCM phenotypes was further shown to vary within the family members carrying exactly the same mutation.50,59 This was also observed in family K and B. The proband of family B (II-2) had non-obstructive HCM, whereas her sister (II-1) exhibited obstruction of LV outflow tract with increased pressure gradient. In family K, the older individual showed moderate hypertrophy of the entire septum, while younger patients had mild mid- and basal septal hypertrophy.

Several mechanisms are implicated in the variability of the penetrance and expressivity in HCM.3 The differences in the phenotypic expression and penetrance among causal genes could be explained by the functional role of the respective protein. Thus, in the case of beta--myosin heavy chain mutations, more abundant phenotypes are expected considering the primary function of this protein in cardiac contraction. The diversity of phenotypes associated with mutations in the same disease gene may be due to the localization in differently important protein regions as well as to the kind of the mutation. Another contributing factor is the individual genetic background (i.e. modifier genes), which may have a modulatory role. Thus, in the case of family B and K, the distinct localization of the mutations within RLC may be responsible for the differences in phenotypic expression of HCM. Furthermore, the observed variable expressivity of mutations within the same families suggests the presence of specific modifier genes. Further studies of large numbers of families carrying the Glu22Lys and Arg58Gln mutations are necessary to clarify whether the observed malignant and


61

benign phenotypes are unique to these particular mutations and generally applicable, or whether they are due to the differences in genetic backgrounds in families B and K.

4.4 Possible functional implications of the Glu22Lys and Arg58Gln mutations

The high evolutionary conservation of the amino acid residues affected by the Glu22Lys and Arg58Gln mutations suggests their essentiality for the RLC function. The full-length amino acid alignment (partly presented in figure 3.14) also showed that the highest sequence homology is shared by the RLCs from the same muscle type and, hence, with similar functions. This underlines that the amino acid sequence determines functional properties of the protein and, consequently, any alteration in this sequence will affect its function.

Possible functional implications of the Glu22Lys and Arg58Gln mutations have been studied by Szsesna et al.,29 who investigated the effects of the Glu22Lys and Arg58Gln mutations on the RLC calcium-binding and phosphorylation properties. In that study, the Glu22Lys mutant could not be phosphorylated and had decreased Ca2+ affinity, whereas the Arg58Gln mutant did not bind Ca2+ at all.

In an earlier study, Levine et al.80 investigated functional and structural consequences of the Glu22Lys mutation in deltoid muscle fibers obtained from a HCM patient carrying this mutation, since the same RLC isoform is expressed in both cardiac ventricle and slow twitch fibers of the deltoid muscle. The study revealed that the biopsied fibers show loss of the normal arrangement of myosin heads associated with the relaxed state. This eventually accounted for a local change in electrical charge caused by the Glu22Lys mutation; charge alteration subsequently may affect the normal RLC conformation and weaken the RLC structural support to the myosin neck. That the normal net charge of the RLC N-terminus is important for the RLC conformation and RLC-myosin interaction has been also shown in a study by Sweeney et al.81

Thus, the Glu22Lys and Arg58Gln mutations could alter the function of the molecular motor myosin by either eliminating the normal effects of RLC phosphorylation and calcium binding or by affecting the allosteric interaction of the myosin heavy


62

chain/RLC complex. This, consequently, may disturb the normal manner of the force generating myosin-actin interaction and lead to a contractile deficit. According to the current hypothesis on HCM pathogenesis, the impaired sarcomeric contractility induces increased expression of trophic factors in the heart, which leads to clinical and pathological phenotypes characteristic of HCM.24 Further genetic and functional studies will hopefully help us to complete our understanding of the mechanisms that underlie the development of HCM, because the establishment of new efficient management strategies can be based only on accurate knowledge of both the etiologies and pathogenic mechanism of a disorder.

In conclusion, two mutations were identified in the ventricular myosin regulatory light chain gene and associated with either benign or malignant HCM phenotypes. The Glu22Lys mutation was associated with a late onset of clinical symptoms, benign course and good prognosis, whereas the Arg58Gln mutation was associated with an early onset of clinical manifestation and premature sudden cardiac death. These findings show that genotyping could give valuable information for the risk stratification, genetic counselling and treatment strategies in hypertrophic cardiomyopathy


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.

DiML DTD Version 2.0
Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML - Version erstellt am:
Mon Jan 6 12:37:59 2003