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

31

Chapter 3. Results

3.1 Patient characteristics

A total of 71 unrelated HCM patients including 48 males and 52 females aged 22 - 78 years (mean 53.3±14.7) were examined for disease-causing mutations in MYL2 and MYL3. Clinical data on these patients are summarized in table 3.1.

The majority of the patients had either no or mild symptoms (NYHA functional class I and II). Mean interventricular septum (IVS) obtained by echocardiography was 19.6±3.7 mm, while mean left ventricular (LV) posterior wall thickness was 10.5±2.4 mm. LV hypertrophy mostly involved the entire IVS (37% cases of Maron type II) or both IVS and anteriolateral LV free wall (45% of Maron type III).

LV outflow tract obstruction leading to an increased gradient of more than 10 mm Hg between LV and aorta was present in 58 % of the probands. It correlated with the presence of systolic anterior motion of mitral valve (54%). In 8 patients (11%) with the increased outflow tract gradient, an operative management was undertaken. 5 of them (7%) underwent LV myectomy (Morrow procedure), and 3 patients (4%) underwent nonsurgical septal reduction.

At the time of examination, most of the patients (94%) were in sinus rhythm; only few had atrial fibrillation and an implanted pacemaker (1.4% and 4,2%, respectively). ECG findings characteristic of HCM such as Q- and T-wave abnormalities were present in 83% of all cases: 31% of the patients showed abnormal Q waves, while negative T waves were observed in 52% of them.


32

Table 3.1. Clinical features of HCM patients screened in the present study

Number of patients (n)

71

Age, years

53.3±14.7

Sex,% (n)

 

 

Male

48.0

(34)

 

Female

52.0

(37)

Age at diagnosis, years

45.5±16.6

NYHA class, % (n)

 

 

I

46.5

(33)

 

II

42.3

(30)

 

III

9.9

(7)

 

IV

1.4

(1)

IVS thickness, mm

19.6±3.7

PW thickness, mm

10.5±2.4

IVS/PW

1.9±0.5

LVEDD, mm

45.7±6.0

Maron type, % (n)

 

 

I

12.7

(9)

 

II

36.6

(26)

 

III

45.1

(32)

 

IV

5.6

(4)

LVOT gradient increased, % (n)

57.7

(41)

SAM, % (n)

53.5

(38)

Morrow myectomy, % (n)

7

(5)

Nonsurgical septal reduction, % (n)

4.2

(3)

Rhythm, % (n)

 

 

Sinus

94.4

(67)

 

Atrial fibrillation

1.4

(1)

 

Pacemaker

4.2

(3)

Abnormal Q waves, % (n)

31.1

(22)

Negative T waves, % (n)

52.1

(37)

Data are expressed as mean±standard deviation or as relative (%) and absolute (n) values. Abbreviations used in the table: NYHA, New York Heart Association class of heart failure; IVS, interventricular septum; PW, left ventricular posterior wall; LVEDD, left ventricular end-diastolic dimension; LVOT, left ventricular outflow tract; SAM, systolic anterior motion of mitral valve.


33

3.2 Genetic variants in human MYL2 and MYL3

Numbering of identified genetic variants was performed as suggested by Dunnen et al.65 The nucleotide number is preceded by “g.“ when a genomic or by “c“ when a cDNA reference sequence was used. In MYL3, only the cDNA reference sequence was used, because no full-length genomic DNA reference sequence was available. The variants, except those in MYL3 intronic regions, were simply designated by the nucleotide numbers of the respective reference sequences. For instance, g.8353G>A denotes the G-to-A substitution at nucleotide 8353 of the MYL2 genomic reference sequence. The MYL3 intronic variants were designated by the number of nucleotides counted from the first or last nucleotide of an adjacent exon. The negative and positive numbers denote the variant‘s location upstream and downstream of an exon, respectively. For instance, c.158-4_5insGTC denotes an insertion of GTC between nucleotides -4 and -5 upstream of nucleotide 158, which is the first nucleotide of exon 3 according to the MYL3 reference cDNA sequence.

As it has been already noted in the Material and Methods chapter, the reference sequences used in the present work were obtained form GenBank (www.ncbi.nlm.nih.gov). Accession numbers of these reference sequences are listed in table 3.2. In the present study, self-generated sequences of HCM patients or controls consistent with the reference genomic sequences were designated as wild type sequences.

Table 3.2. GenBank accession numbers of the reference sequences used in the present study

 

genomic DNA reference sequence

cDNA reference sequence

MYL2

L01652

X66141

MYL3

J04462

M24122


34

Figure 3.1. Schematic representation of the ventricular myosin regulatory light chain gene (MYL2) and the ventricular myosin essential light chain gene (MYL3). Boxes represent exons, light shaded boxes represent coding DNA of the gene; mutations found in this study are dark shaded; variants with asterisk indicate single nucleotide polymorphisms and the silent mutation; variants without asterisk indicate differences from the reference genomic DNA sequence. Numbering of the genetic variants was performed according to ref. 65. A) Location of two missense mutations, SNPs, and sequence differences in MYL2. B) Location of a silent mutation and sequence differences in MYL3.

Two missense mutations, Glu22Lys and Arg58Gln, were identified in MYL2 and associated with different HCM phenotypes in two families. The Glu22Lys mutation was identified in exon 2, whereas the Arg58Gln mutation was in exon 4. Additionally, one silent mutation and three single nucleotide polymorphisms (SNPs) were detected while screening the MYL2 and MYL3 genes. The c.420C>T (Phe140Phe) silent mutation was identified in exon 4 of MYL3. The g.8393G>A and g.8580C>T/A single nucleotide polymorphisms were observed in introns flanking exons 5 and 6 of MYL2. Finally, a number of sequence differences from the reference genomic DNA sequence were observed in both genes, most of them in intronic regions. An overview of these findings is given in figure 3.1.


35

3.2.1 Identification of the Glu22Lys mutation in family K

The Glu22Lys mutation was detected initially in patient 1853 (or proband II-3 of family K) during screening of exon 2 of MYL2. The sample revealed an aberrant band pattern on SSCP analysis. As shown in figure 3.2 B, the aberrant pattern had three bands instead of two bands as in a normal pattern. Direct automated sequencing of both genomic DNA strands revealed a heterozygous G-to-A (guanine-to-adenine) substitution at nucleotide c.64. On the sequence electropherogram, this was present as two overlapping peaks with a black peak corresponding to G on one allele and a green peak corresponding to A on the other allele (figure 3.2 C). The two overlapping peaks were half the height in comparison with neighbouring peaks and were recognized as "N" by the sequencing analysis software. According to the reference cDNA sequence, this c.64G>A substitution affected the first nucleotide of codon 22 changing it from original GAA to mutated AAA. This subsequently caused a replacement of glutamic acid (Glu) by lysine (Lys) (Glu22lys). In addition to sequencing, the presence of the Glu22Lys mutation was confirmed by RFLP analysis with the Taqalpha I restriction enzyme (see further).


36

Figure 3.2. PCR, SSCP analysis and sequencing of exon 2 of MYL2. A) A 4% agarose gel loaded with 317-bp PCR products of exon 2. M, VIII-DNA ladder; Pr II-3, proband II-3 of family K; 1865 and 1743, other HCM patients; 141 and 142, control individuals; N, negative control. B) Partial SSCP gel. SSCP analysis of proband II-3 of family K (Pr II-3) revealed an aberrant band pattern, which has one additional band (indicated by asterisks) in comparison with patterns shown by neighbouring HCM samples (1438,1434, 1431, and 1427). C) Partial sequence electropherograms of exon 2 of the proband of family K and an individual with the wild type sequence. The proband's electropherogramm showed two typical overlapping peaks at nucleotide c.64: a black peak for guanine on the non-mutated allele and a green peak for adenine on the mutated allele. This G-to-A substitution caused a change of glutamic acid to lysine at codon 22. By contrast, the wild type sequence is homozygous for guanine at position c.64.


37

Figure 3.3. RFLP analysis of exon 2 of MYL2. A) Schematic representation of restriction sites of Taqalpha I. Two normal S1 and S2 restriction sites produce DNA fragments of 174, 78, and 65 bp. The Glu22Lys mutation removes the S2 restriction site resulting in an abnormal fragment of 252 bp. B) The picture of a 4% agarose gel loaded with restriction fragments of the proband of family K (Pr II-3) and controls (221, 222, 224, 225, and 226). The abnormal digestion fragment of 252 bp due to the Glu22Lys mutation is present in the proband of family K but absent in controls. Lane M contains 125-bp DNA ladder. bp, base pairs; U, undigested PCR product of exon 2.

RFLP analysis of proband II-3 and his family members was possible, because the Glu22Lys mutation changed the normal restriction pattern of the Taqalpha I restriction enzyme. The wild type sequence of exon 2 possess two normal Taqalpha I restriction sites, which produce three DNA fragments of 174, 78, and 65 bp each (figure 3.3 A). Taqalpha I recognizes the TCGA sequence. The c.64G>A substitution changes this recognition sequence to the TCAA sequence and, consequently, removes one of the two normal Taqalpha I restriction sites. This will result in the appearance of only two restriction fragments of 65 and 252 bp instead of the three normal fragments (figure 3.3 A). However, the two fragments will be present only when the Glu22Lys mutation is homozygous. In the case of the heterozygous Glu22Lys mutation (as in Family K), the three normal fragments from the non-mutated allele (65, 78 and 174 bp) and the two fragments from the mutated allele (65 and 252 bp) will be observed. On a gel the 65-bp fragments from both alleles will be overlapping each other. Thus, the presence of the Glu22Lys mutation will be recognized by the presence of the additional 252-bp fragment.

Exon 2 of MYL2 was also amplified from the genomic DNA of control individuals and digested with Taqalpha I. 105 controls failed to show the Glu22Lys mutation, because the abnormal 252-bp digestion fragment was observed in none of them (figure 3.3 B).


38

In the family of patient 1853 (designated as family K; figure 3.4), the Glu22Lys mutation was identified in further six individuals by RFLP analysis: the abnormal restriction fragment of 252 bp was observed in family members II-5, III-2, III-3, III-5, IV-1, and IV-2. In addition to proband II-3, three of these individuals (III-5, IV-1, IV-2) had HCM at the time of examination. The pedigree of family K and results of RFPL analysis are presented in figure 3.4. As shown, the pedigree consisted of 12 individuals over 3 generations. Four individuals (I-1, I-2, II-1, II-2) died before this study, and no data on them could be obtained.

Figure 3.4. Pedigree of family K and results of RFLP analysis on available family members. Upper panel: Pedigree. Black symbols represent clinically affected patients; white symbols, clinically unaffected individuals; grey symbol, individuals with uncertain phenotype; symbols with plus sign above, genetically affected individuals; symbols with minus sign above, genetically unaffected individuals; and symbols with diagonal slash, deceased individuals. Proband II-3 (patient 1853) is indicated by arrow. Squares, males; circles, females. Lower panel: Identification of the Glu22Lys mutation in family members by RFLP analysis with Taqalpha I. A picture of a 4% agarose gel loaded with restriction digests. The abnormal 252-bp fragment is present in family members II-3, III-2, III-3, II-5, III-5, IV-1, and IV-2 indicating the presence of the Glu22Lys mutation. Lane U contains undigested amplification product of MYL2 exon 2 of 317 base pairs (bp). Lane M contains 125-bp DNA ladder with sizes of bands shown on the right side of the gel picture. Sizes of the digestion products are shown on the left side of the gel picture.


39

3.2.2 Identification of the Arg58Gln mutation in family B

The Arg58Gln mutation was initially identified in patient 1555 (or proband II-2 of family B) while screening exon 4 of MYL2. The sample showed an aberrant band pattern (with three clear additional bands) on SSCP analysis (figure 3.5 A). Direct automated sequencing of both DNA strands revealed two typical overlapping peaks indicating a heterozygous G-to-A (guanine-to-adenine) substitution at nucleotide c.173 (figure 3.5 B). The c.173G>A substitution resulted in a replacement of arginine (Arg) by glutamine (Gln) at codon 58 (Arg58Gln), because this codon was changed from original CGA to mutated CAA. The Arg58Gln mutation was confirmed by sequencing in two independent runs, because it did not affect any restriction site making RFLP analysis impossible.

Figure 3.5. SSCP analysis and sequencing of exon 2 of MYL2. A) SSCP analysis. The aberrant band pattern observed in proband II-2 of family B (Pr II-2) was absent in controls (34, 33, 31, 23, and 21) and contains three clear additional bands, which are indicated by asterisks. NBP, normal band pattern; ABP, aberrant band pattern. B) Sequence electropherograms of proband II-2 of family B, the proband's mother, and a control individual. Proband II-2 has two overlapping peaks at position c.173 with a black peak corresponding to guanine on the non-mutated allele and a green peak for adenine on the mutated allele. A G-to-A heterozygous substitution resulted in a replacement of arginine by glutamine at codon 58. By contrast, the Arg58Gln mutation was absent in a control and the proband's mother.


40

In the family of patient 1555 (designated as family B; figure 3.6), two more individuals suffered from HCM: proband‘s father I-1 and sister II-1. However, genetic analysis on them could not be performed, because they died before this study, and no DNA could be obtained. Genotyping of only alive proband‘s mother I-2 revealed no Arg58Gln mutation (figure 3.5 B and 3.6). The pedigree of family B with results of SSCP analysis on patient 1555 and her mother is presented in figure 3.6.

Exon 4 of MYL2 was further amplified from DNA of control individuals and subjected to SSCP analysis. 105 controls failed to show the Arg58Gln mutation, because none of them revealed the aberrant band pattern characteristic of this mutation (figure 3.5 A).

Figure 3.6. Pedigree of family B and results of SSCP analysis. Upper panel: Pedigree. Black symbols represent clinically affected patients; white symbols, clinically unaffected individuals; symbols with plus sign above, genetically affected individuals; symbols with minus sign above, genetically unaffected individuals; symbols with diagonal slash, deceased individuals; squares, males; and circles, females. The proband (patient 1555) is indicated by arrow. SCD, sudden cardiac death. Lower panel: SSCP gel. Proband II-2 showed an aberrant mobility pattern absent in her mother (I-2) and control (C).


41

3.2.3 Localization of the mutations in highly conserved RLC regions

The amino acid residues affected by the Glu22Lys and Arg58Gln mutations are strictly conserved throughout evolution: as shown in figure 3.7, glutamic acid at position 22 and arginine at position 58 are invariant in RLC isoforms, which are expressed in the heart of different species (human ventricular/slow skeletal, rat and mouse ventricular, chicken cardiac). Glutamic acid is also preserved among skeletal isoforms of the shown species.

The identified Glu22Lys and Arg58Gln mutations are located in the amino terminal half of the RLC, which contains two putatively important functional regions: the phosphorylation and calcium-binding sites. As shown in figure 3.8, both variants are in alpha-helices flanking the calcium-binding loop. The Glu22Lys is additionally in the region adjacent to the RLC phosphorylation site.

The Glu22Lys and Arg58Gln mutations are further predicted to alter the normal net charge of the RLC N-terminus, because the Glu22Lys variant caused a replacement of negatively charged glutamic acid by positively charged lysine, and the Arg58Gln mutation caused a substitution of positively charged arginine by non-charged glutamine.

Figure 3.7. Amino acid alignment across species and RLC isoforms. The Glu22Lys and Arg58Gln mutations affect the highly conserved amino acids, suggesting the essentiality of these residues for normal protein function. RLCs from the same muscle type show the highest sequence homology. This indicates that functional properties of a protein are determined by its amino acid sequence. Glu and E, glutamic acid; Lys and K, lysine; Arg and R, arginine; Gln and Q, glutamine.


42

Figure 3.8. Localization of the Glu22Lys and Arg58Gln mutations in the RLC sequence (A) and three-dimensional structure (B). The putative phosphorylation site at serine-15 (A) and calcium-binding site at residues 37-48 (A, B) are highlighted. As shown, the Glu22Lys mutation is located close to the RLC phosphorylation site, moreover it is in the alpha-helix flanking the calcium-binding loop. The Arg58Gln mutation is in the alpha-helix, which flanks the calcium-binding loop from the other side. Adapted from ref. 29.


43

3.2.4 Clinical features of family K with the Glu22Lys mutation

Clinical data of the genetically affected members of families K and B are summarized in table 3.3.

In family K, the Glu22Lys mutation was identified in seven individuals. Within them, four individuals had HCM (II-3, III-5, IV-1, and IV-2), one individual had borderline cardiac hypertrophy (II-5), and another one (III-2) was a healthy carrier. Remaining genetically affected individual III-3 (32 years old) had normal ECG, but echocardiographic evaluation could not be performed because of patient‘s unwillingness. All these individuals were asymptomatic apart from proband II-3 and his sister II-5.

75-year old male proband II-3 was referred for clinical evaluation because of episodes of palpitation, chest pain, and dyspnea. His ECG showed sinus rhythm and left bundle branch block. Holter electrocardiography demonstrated an episode of supraventricular tachycardia and polytopic ventricular extrasystoles. Echocardiography revealed asymmetric hypertrophy with IVS of 23 mm. No LV cavity or outflow tract obstruction was observed. Within the following year, the proband was admitted to Franz-Volhard-Klinik twice because of events of atrial fibrillation. During the first visit, he was converted to sinus rhythm by electrical cardioversion. The next time, he underwent a successful high frequency ablation.

Genetically affected proband‘s niece III-5, 42 years old, did not report any symptoms. But her ECG demonstrated abnormal Q waves at leads I and aVL. Echocardiography revealed basal septal hypertrophy of 15 mm. No pressure gradient was present.

20-year old female individual IV-1, who inherited the Glu22Lys mutation from her mother III-5, also did not show any clinical symptoms but had abnormal ECG, and echocardiographic findings characteristic of HCM. Her ECG demonstrated pathologic Q waves at aVL, while echocardiography revealed midseptal hypertrophy related to body surface area with IVS thickness of 12 mm. No LV cavity or outflow tract obstruction was observed.

Similarly, 18 year-old male individual IV-2, who also inherited the mutation from his mother III-5, did not report any clinical symptoms but exhibited ECG and echocardiographic abnormalities characteristic of HCM. ECG showed voltage criteria of


44

LV hypertrophy: Sokolow-Lyon index was 4.8 mV. Echocardiography demonstrated midseptal hypertrophy of 13 mm without any obstruction.

The phenotype of genetically affected proband‘s sister II-5 was defined as “uncertain“. The 62-year old woman reported periodic dyspnea and chest pain. But her ECG was normal. Echocardiography revealed IVS of 13 mm, however, it was in the normal range in relation to her body surface area of 2.45 m2. No pressure gradient was present.

33-year old proband‘s daughter III-2, who inherited the mutation, reported no clinical symptoms. Her ECG revealed pathologic Q waves at lead aVF, but echocardiography showed no myocardial hypertrophy.

Table 3.3. Clinical features of genetically affected individuals of families K and B

Family

 

 

Family K

 

 

Family B

Pedigree number

II-3

II-5

III-2

III-5

IV-1

IV-2

II-2

 

 

 

 

 

 

 

 

Mutation

Glu22Lys

Glu22Lys

Glu22Lys

Glu22Lys

Glu22Lys

Glu22Lys

Arg57Gln

Age (years)

75

62

33

42

20

18

27

Age at diagnosis (years)

75

62

33

42

20

18

7

BSA (sqm)

1.97

2.45

2.08

2.06

1.75

2.01

1.58

Weight (kg)

81

139

98

96

65

80

56

Heart block

LBBB

no

no

no

no

no

no

Negative T

n.a.

no

no

no

no

no

yes

Abnormal Q

n.a.

no

aVF

I, aVL

aVL

no

no

S-L index (mV)

n.a.

1.9

2.4

3.1

2.2

4.8

4.8

IVS (mm)

20

13

10

15

12**

13

21

PW (mm)

13

n.d.

7

8

7

10

7

IVS/PW

1.5

n.a.

1.4

1.8

1.7

1.3

3

LVEDD (mm)

52

47

48

53

44

52

39

Maron type of LVH*

I

I

n.a.

I

IV

IV

III

NYHA class

III

II

I

I

I

I

II

Clinical status

affected

uncertain

unaffected

affected

affected

affected

affected

BSA, body surface area; S-L index, Sokolow-Lyon index; IVS, interventricular septum; PW, left ventricular posterior wall; LVEDD, left ventricular end-diastolic dimension; LBBB, left ventricular bundle branch block; LVH, left ventricular hypertrophy; NYHA, New York Heart Association class of heart failure; n.d., not determined; n.a., not applicable. *According to ref. 28. **In this individual, HCM diagnosis was based on increased IVS thickness for his age, weight and BSA.


45

3.2.5 Clinical features of family B with the Arg58Gln mutation

In family B, three individuals had HCM, two of them died suddenly at young age.

The proband II-2 was 7 years old when HCM was diagnosed during a medical evaluation because of sudden cardiac death of her 28-year old father. When she was 16 years old, therapy with beta-adreno receptor blockers was started due to premature fatigue on exertion. At the age of 25, the patient additionally reported palpitations and presyncopal conditions. She was referred to an electrophysiological examination; during this procedure she developed ventricular tachycardia degenerating into ventricular fibrillation. The proband was converted to sinus rhythm by electrical defibrillation. Afterwards, considering the family history of two sudden cardiac deaths and aggravation of clinical symptoms, a cardioverter-defibrillator (ICD) was implanted. At the age of 27, recurrent events of supraventricular tachycardia (up to 170/min) were registered on the ICD, and she was admitted to Franz-Volhard-Klinik. No shock had been delivered from the ICD by that time. ECG showed voltage signs of LV hypertrophy with T wave inversion. Echocardiography revealed asymmetric septal hypertrophy of 21 mm extending to the LV apex and lateral free wall. No pressure gradient was observed. Electrophysiological investigation demonstrated the common type of atrial flutter with 2:1 conduction ratio. Ablation therapy was considered, but due to the risk of affecting the ICD lead, therapy with sotalol was attempted first. The latter resulted in suppression of the tachycardia and improvement of clinical symptoms.

Proband‘s mother I-2 did not have HCM: she had a normal ECG and LV wall thickness on echocardiography.

Proband‘s father I-1 did not show any symptoms of the disease. He died suddenly at the age of 28. It is known from his wife that HCM was diagnosed on autopsy.

Clinical data on proband‘s younger sister II-1 were kindly sent by Prof. Kienast from the University Clinic in Rostock. She was 5 years old when HCM was diagnosed. The only symptom reported was premature fatigue on exertion. Her ECG showed signs of LV hypertrophy with negative T waves in left chest leads. Echocardiography demonstrated septal hypertrophy of 26 mm, normal thickness of LV posterior wall, and LV outflow tract obstruction. Despite regular medical check-ups and treatment by calcium channel blockers, she died suddenly at home at the age of 21.


46

3.2.6 The c.420C>T (Phe140Phe) silent mutation in MYL3

In MYL3, the c.420C>T (Phe140Phe) silent mutation was detected in one patient (DNA sample 1635) out of the 71 individuals while screening exon 4. The sample showed an aberrant band pattern on a SSCP gel. As shown in figure 3.9 B, the aberrant band pattern possessed an additional band in comparison with normal patterns shown by the neighbouring samples. DNA sequencing revealed a heterozygous C-to-T (cytosine-to-thymine) substitution at nucleotide c.420 (figure 3.9 C). No amino acid change was caused by this substitution, because both codons TTC and TTT encode for phenylalanine.

Figure 3.9. PCR, SSCP analysis and sequencing of exon 4 of MYL3. A) A 4% agarose gel loaded with PCR products of exon 4. The amplified fragments were approximately 400 base pairs (bp) long. Lane M contains VIII-DNA ladder. 1635, a patient carrying the c.420C>T polymorphism; 1781, 1782, and 1783, other HCM patients. B) SSCP analysis of patient 1635 revealed an aberrant band pattern, which has an additional band (indicated by asterisk) in comparison with patterns shown by neighbouring HCM patients 1855, 1853, and 1592. C) Partial sequence of patient 1635 showing the heterozygous c.420C>T silent mutation and of an individual homozygous for the wild type MYL3 allele.


47

3.2.7 Single nucleotide polymorphisms in MYL2

SSCP analysis of PCR products containing exons 5 and 6 and their flanking intronic regions revealed different band patterns, which were unevenly distributed but frequent. This suggested that eventually some common genetic variants underlie the observed SSCP band patterns. Although the relatively high frequency of each band pattern within a group of 71 patients suggested that underlying genetic variants are not disease-causing mutations but rather polymorphisms, in order not to miss any disease-causing mutation, several samples from each subset of samples showing a similar SSCP pattern were selected for sequencing. The latter revealed the g.8393G>A and g.8580C>T/A intronic single nucleotide polymorphisms confirming the initial proposal about common polymorphisms.


48

The g.8393G>A polymorphism in intron 4 of MYL2

After evaluating SSCP gels of the exon 5 fragment, seven samples were sequenced. Two MYL2 allelic variants were observed with regard to nucleotide position g.8393 in the part of intron 4 flanking exon 5. The first one was a g.8393G variant, which designates the presence of guanine at nucleotide position g.8393. This variant corresponded to the reference genomic DNA sequence and was therefore considered the wild type sequence. The second variant was g.8393G>A, which denotes a MYL2 allele possessing adenine at the same g.8393 nucleotide position.

Within the seven sequenced samples, three samples (1595, 1808, and 1744) were homozygous for the g.8393G wild type allele (figure 3.10 A). Three further individuals (1795, 1811, and 1817) were heterozygous for the g.8393G and g.8393G>A alleles (figure 3.10 C). The remaining sample (1819) was homozygous for the g.8393G>A polymorphism (figure 3.10 B).

Figure 3.10. Partial sequence electropherograms of intron 4 of MYL2 showing genetic variants at position g.8393. A) DNA sequence of sample 1595 with guanine at nucleotide position g.8393 on both alleles. B) DNA sequence of patient 1819 showing the presence of the homozygous g.8393G>A polymorphism. C) DNA sequence of patient 1811 showing the presence of the heterozygous g.8393G>A polymorphism.


49

The g.8580C>T/A polymorphism in intron 5 of MYL2

In the case of the MYL2 exon 6 fragment, eight samples underwent DNA sequencing, which revealed three allelic variants with regard to nucleotide position g.8580 in the part of inron 5 flanking exon 6. The first variant was g.8580C, which denotes an allele with cytosine at nucleotide position g.8580. This variant was consistent with the reference genomic DNA sequence and was considered the wild type sequence. The second identified variant was g.8580C>T, which designates a MYL2 allele carrying thymine at nucleotide g.8580. The third variant was g.8580C>A, which denotes an allele possessing adenine at the same g.8580 position.

Among eight sequenced samples, two samples (1584, 1781) were homozygous for the g.8580C wild type allele (figure 3.11 B). Three further samples (1565, 1744, and 1782) were heterozygous for the g.8580C and g.8580C>T alleles (figure 3.11 D). Two other individuals (1707 and 1780) were homozygous for the g.8580C>T polymorphism (figure 3.11 A). The remaining sample (1594) was heterozygous for the g.8580C>T and g.8580C>A alleles (figure 3.11 C).

Figure 3.11. Partial sequence electropherograms of intron 5 of MYL2 showing genetic variants at position g.8580. A) DNA sequence of patient 1707 showing the homozygous g.8580C>T polymorphism. B) DNA sequence of wild type sample 1584 with guanine at position g.8580 on both alleles. C) DNA sequence of patient 1594 showing the presence of the g.8580C>T polymorphism on one allele and the g.8580C>A polymorphism on the other allele. D) DNA sequence of patient 1782 showing the presence of the heterozygous g.8580C>T polymorphism.


50

Both g.8393G>A and g.8580C>T/A MYL2 variants were located in intronic regions and did not cause an amino acid exchange. Furthermore, these polymorphisms are not predicted to have any effect on the splicing process proceeding from their localizations sufficiently far from the splice sites.

After clarifying that no disease-causing mutations but nucleotide polymorphisms did underlie the observed SSCP band patterns, no further sequencing of the exon 5 and exon 6 containing fragments was performed.


51

3.2.8 Genomic sequence differences

While evaluating data obtained in the course of this study, some self-generated genomic DNA sequences had regions (further designated as differences), which mismatched to the reference genomic DNA sequences. These nucleotide differences were considered such rather than polymorphisms or mutations, because they were present in all samples subjected to DNA sequencing. Proceeding from the high quality of self-generated sequences, it was concluded that the observed differences were due to the errors in the reference genomic DNA sequences. This conclusion was confirmed by further analysis of the self-generated sequences in comparison to the reference cDNA sequences. The presence of eventual errors in the reference genomic DNA sequences required careful analysis in order to interpret obtained data accurately.

All observed differences were in intronic regions of both genes, except for c.240A>T and c.169C>G differences, which were detected in the coding part of MYL2 (see overview in figure 3.1). As indicated in figure 3.1, the differences were present as nucleotide substitutions (n=10), nucleotide deletions (n=4) and insertions (n=9).

Figure 3.12 below shows examples of the observed differences in intron 1 (g.1277A>G, g.1278G>A and g.1291T>C) and in exon 4 (c.240A>T) of MYL2. As shown, the c.240A>T difference was observed in comparison to the reference genomic DNA sequence but was in agreement with the reference cDNA sequence.

Figure 3.12. Partial sequence alignment of HCM samples, control and the reference genomic DNA and cDNA sequences. Ref gDNA, the reference genomic DNA sequence; HCM, self-generated DNA sequence of a HCM patient; control, self-generated DNA sequence of a control individual; cDNA, the reference cDNA sequence. A) Partial sequence alignment of intron 1 of MYL2. Three identified differences, g.1277A>G, g.1278G>A and g.1291T>C, are highlighted. B) Partial sequence alignment of exon 4 of MYL2 with the c.240A>T difference highlighted.


52

Figure 3.13. Partial sequence alignment of HCM samples, control and the reference genomic DNA and cDNA. Ref gDNA, the reference genomic DNA sequence; HCM, self-generated DNA sequence of a HCM patient; control, self-generated DNA sequence of a control individual; cDNA, the reference cDNA sequence; RLC, RLC amino acid sequence translated on the basis of the reference cDNA sequence. A) Sequence alignment of exon 3 and a donor splice site of intron 3 of MYL2. The c.169C>G difference is highlighted. B) Sequence alignment of exon 4 of MYL2. The deletion of adenine and guanine (g.7488_7489delAG) at the acceptor splice site of exon 4 is highlighted. The actual splice site is shown three nucleotides upstream. C) Exon 3 and 4 of MYL2 are aligned together to show that the self-generated sequences are in agreement with the reference cDNA but not with the genomic DNA sequences.

The most confusing findings were the c.169C>G and g.7488_7489delAG differences identified in MYL2. The c.169C>G difference denotes the presence of guanine instead of cytosine at nucleotide c.169, which is the last nucleotide of exon 3 (figure 3.13 A). The g.7488_7489delAG difference denotes a deletion of the AG acceptor splice site of exon 4 (figure 3.13 B). But another AG splice site was found three nucleotides upstream and is predicted to be the actual acceptor splice site of exon 4 proceeding from the comparison with the c.DNA reference sequence.

The presence of the c.169C>G difference and shift of the acceptor splice site of exon 4 upstream resulted in three more nucleotides GGC at the beginning of exon 4. This subsequently resulted in two rearranged codons: GGG encoding for glycine and GCA encoding for arginine. These findings were in agreement with the reference cDNA sequence as shown in figure 3.13 C.


53

Figure 3.14. RFLP analysis of exon 3 of MYL2, which confirmed the presence of the c.169C>G difference in all of 71 HCM-patients and 100 control individuals. A) Schematic drawing of the restriction sites of Sty I. According to the MYL2 genomic DNA reference sequence, an amplified product of exon 3 is supposed to have only one restriction site (S1), which gives rise to two digests of 35 and 235 base pairs (bp) each. But in the presence of the c.169C>G difference, it has two restriction sites (S1 and S2), which produce three restriction fragments of 35 bp, 92 bp, and 143 bp each. B) A picture of 4% agarose gel loaded with digests. The presence of three restriction fragments of 35, 92, and 143 bp but not two of 35 and 235 confirms the presence of the c.169C>G difference and of an error in the genomic DNA reference sequence. The same band pattern as shown on this gel was observed in all 71 patients and 100 controls. Lane U contains undigested amplification product of exon 3 of 270 base pairs. Lane M contains 125-bp DNA ladder.

In the case of the c.169C>G difference, it was possible to perform RFLP analysis with the Sty I restriction enzyme. This enzyme recognizes the sequence CCTTGG. According to the MYL2 genomic reference sequence, the amplified fragment containing exon 3 (270 bp long) is supposed to possess a single Sty I restriction site, which produces two DNA fragments of 35 and 235 bp each (figure 3.14). However, the c.169C>G difference introduces an additional Sty I recognition site. Collectively, the two Sty I restriction sites will result in three fragments of 35, 143, and 92 bp. The observation of these three fragments in all 71 HCM probands and 100 controls confirmed the presence of an error in the reference genomic DNA sequence at nucleotide c.169 (figure 3.14).


© 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