1. Abstract
Myofibrillar Myopathy (MFM) is a muscular disorder involving myofibrillar disarrangement commencing at the Z-discs. This condition displays great variety in clinical features and several genes have been found to be associated. Due to this genetic heterogeneity, MFM has been divided into subtypes based on the gene affected (MFM1-MFM8).
Current diagnosis of MFM is based predominantly on clinical examination, which is inadequate for distinguishing among MFM subtypes. This is related to the fact that different subtypes display overlap and inconsistency in their features.
Research into the mechanics of the disease employs case studies and animal models. The aim is to determine the cause of MFM phenotypes at molecular level. A deeper understanding of the protein-protein interaction at the z-disc and the genes involved is necessary for the development of improved diagnosis and treatment.
2. Introduction
It is the repetitive assembly of sarcomeres into myofibrils which gives striated muscle its characteristic appearance through the microscope (Gautel and Djinovic-Carugo, 2016). These contractile units are made up of various intermediate protein filaments including actin and myosin, and are flanked by z-discs (fig. 2.1). Although figure 2.1 gives a simplified representation of a sarcomere, there are many more proteins at work to ensure contraction and maintenance in striated muscle fibres (fig 2.2).
It has been found that mutations leading to abnormalities in some of these proteins can be involved in a range of muscular disorders (Clark et al., 2002) amongst which myofibrillar myopathy (MFM).
Fig. 2.1 Schematic overview of sarcomeres, featuring intermediate protein filaments, Z-discs, I-band, A-band and M-line. Linkage of the sarcoplasmic reticulum (SAR) to the sarcolemma is mediated by T-tubules (Tt). Wavy lines represent other interacting proteins. (Gautel and Djinovic-Carugo, 2016)
Fig. 2.2. The molecular architecture of a myocyte,
featuring proteins involved in skeletal and cardiac myopathies.
(Dalakas et al., 2000)
MFM is a group of muscle disorders with varying clinical symptoms and underlying genetics. Common to all types of MFM is myofibrillar disarrangement beginning at the Z-discs, and the presence of abnormal protein aggregates (Fichna et al., 2018). Since different MFM types can have overlapping symptoms, clinical examination and muscle biopsies do not give conclusive results, leading to inadequacy of current treatment.
This review aims to compare the genetics and phenotypes of the most common types of MFM, with an eye to potential future treatment based on genetic screening.
3. Disease characteristics
Current diagnosis is based primarily on clinical symptoms and muscle biopsies (Selcen and Engel, 2011)
Clinical aspects of MFM vary greatly, even within subtypes. The age of onset ranges from the 1st to the 8th decade of life (table 4.1.2).
Progressive weakness of skeletal muscle is typical. Although distal muscles are commonly affected first, some MFMs originate at the limb-girdle area. Cardiac muscle often becomes involved at a later stage, but can also occur in initial stages of the disease. Sometimes, development of neuropathy is observed (Selcen and Engel, 2011) (Fichna et al., 2018).
The pathology of MFM shows a clearer pattern at muscle fibre level than clinical level. General disorganisation of the sarcomere, Z-disc widening, and presence of abnormal protein aggregates, are common histopathological features. An increased variability in fibre shape and size is also observed, and oxidative enzyme activity is reduced (Fichna et al., 2014) (Selcen and Engel, 2011).
Trichrome staining often reveals presence of amorphous, hyaline or granular structures which vary in shape, size and colour in the abnormal fibres. (Selcen and Engel, 2004). Furthermore, presence of (rimmed) vacuoles containing membranous materials from degraded organelles is not uncommon (Selcen and Engel, 2011).
Research shows that the occurrence of ‘rubbed-out’ fibres (fig 3.1) is relatively frequent in MFM1 and MFM2, while increased presence of vacuoles is more typical of MFM3 and MFM4 (Claeys et al., 2009).
Fig. 3.1 Appearance of ‘rubbed-out’ fibres (1) versus ‘core-like lesions’ (2) stained with NADH-TR for MFM1 and MFM2.
Adapted with modifications from Claeys et al. (2009)
Although certain clinical and morphological features are more prominent in some MFMs than others, the pattern is not consistent enough to conclusively distinguish among the subtypes (Fischer et al., 2008)
4. Genetics
4.1 Overview
To date, a total of 17 genes were found to be associated with forms of MFM, as reviewed by Fichna et al. (2018) (table 4.1.1). The first eight genes listed in table 4.1.1 are most commonly associated with MFM and its classical phenotype.
The remaining nine genes have occasionally been associated with characteristics of MFM and are more commonly causative of other muscular diseases.
Gene
Phenotypes besides MFM
DES
LGMD1, dilated cardiomyopathy, Kaiser-type neurogenic scapuloperoneal syndrome
CRYAB
Dilated cardiomyopathy, cataract
MYOT
LGMD1, spheroid body myopathy
LDB3/ZASP
Dilated cardiomyopathy, hypertrophic cardiomyopathy, left ventricular non-compaction
FLNC
Restrictive cardiomyopathy, hypertrophic cardiomyopathy
BAG3
Dilated cardiomyopathy, polyneuropathy
KY
–
PYROXD1
–
PLEC
LGMD1, epidermolysis bullosa simplex
FHL1
LGMD1, Salih myopathy, tibial muscular dystrophy, dilated cardiomyopathy, hypertrophic cardiomyopathy
ACTA1
Nemaline myopathy, scapulohumeroperoneal myopathy, congenital myopathies
TTN
HSPB8
Charcot-Marie-Tooth disease, distal motor neuropathy
LMNA
Emery-Dreifuss muscular dystrophy, dilated cardiomyopathy, restricted cardiomyopathy, Charcot–Marie–Tooth disease, progeria, lipodystrophy, dermatopathy
DNAJB6
LGMD1
SQSTM + TIA1
Hypercapnic respiratory insufficiency
1 LGMD abbreviates Limb Girdle Muscular Dystrophy
Table 4.1.1 MFM associated genes and their other phenotypes.
Adapted with modifications from Fichna et al. (2018)
The inheritance pattern of the disease is often autosomal dominant, although it can be autosomal recessive in some CRYAB variants (table 3.1.2) or X-linked in FHL1 (D’Avila et al., 2016, Selcen and Engel, 1993).
This review will focus on the genes associated with MFM1-MFM8, as they are the most common subtypes of MFM.
Type
Gene
Protein
Primary Muscle Type
Inheritance Pattern
Age of Onset
MFM1
DES
Desmin
Cardiac/ Skeletal
Autosomal Dominant
Autosomal Recessive
3rd-5th decade
1st-2nd decade
MFM2
CRYAB
-B-crystallin
Cardiac/ Skeletal
Autosomal Dominant
Autosomal Recessive
4th-6th decade
1st decade
MFM3
MYOT
Myotilin
Mild Cardiac/Skeletal
Autosomal Dominant
Autosomal Recessive
4th–8th decade
3rd-5th decade
MFM4
LDB3/ZASP
Z-band alternatively spliced PDZ-containing protein (ZASP)
Cardiac/ Skeletal
Autosomal Dominant
5th-6th decade
MFM5
FLNC
Filamin C
Cardiac/ Skeletal
Autosomal Dominant
3rd-7th decade
MFM6
BAG3
Bcl2-associated athanogene-3
Cardiac/ Skeletal
Autosomal Dominant
1st-2nd decade
MFM7
KY
Kyphoscoliosis peptidase
Skeletal
Autosomal Recessive
1st decade
MFM8
PYROXD1
Pyridine Nucleotide-Disulphide Oxidoreductase Domain-containing protein 1
Skeletal
Autosomal Recessive
1st decade
Table 4.1.2 Key features of major MFM subtypes.
Adapted with modifications from Kley et al. (2016).
4.2 MFM1 (OMIM #601419)
Desmin, which is associated with MFM1, is a type III intermediate filament protein which interlinks z-discs and other structural and contractile organelles, creating a 3D structure extending over the myofibril (Kouloumenta et al., 2007). Using in situ hybridisation, the DES gene encoding the desmin protein, was assigned to human chromosome 2q35(Viegas-Pequignot et al., 1989).
Expression of DES varies greatly in different tissues and is most prominent in skeletal, cardiac and certain types of smooth muscle (Lazarides, 1980) (Capetanaki et al., 1984). Furthermore, desmin is highly abundant in Purkinje fibers and thus likely involved in electric cardiac function (Schrickel et al., 2010). MFM1 often affects cardiac and skeletal muscle and involves conduction disorders and arrhythmias. Other features include progressive weakness of lower limb muscles, spreading from distal to proximal muscles or vice versa (Goldfarb et al., 2008). Some patients present with smooth muscle myopathy and respiratory distress (Goldfarb et al., 1998) (Fichna et al., 2014). The variable involvement of different muscle types is thought to be associated with the protein domain wherein the mutation lies (Goldfarb et al., 2008).
Most commonly associated with MFM1 are mutations disrupting the desmin rod domain in the 2B helix of its C-terminal end (fig. 4.2.1), as observed by Goldfarb et al. (1998) in two affected American families, and by Fichna et al. (2014) in three Polish families. The latter study found that structural defects associated with a heterozygous Q348P (CAG>CCG) mutation were more severe than those associated with a A357_E359 deletion. A suggested explanation for this is that Q348P involves substitution with a proline residue, while the A357_E359del mutation does not. Proline’s R-group is a specific ring structure which prevents the amino acid from participating normally in hydrogen bonding within -helices, resulting in kinks in the secondary protein structure of desmin. (Fichna et al., 2014).
Other mutations associated with desminopathy localise to the 1A and 1B helical segments and the head and tail domain of desmin (Goldfarb and Dalakas, 2009) (fig. 4.2.1). The latter has recently been gaining interest. Mutations in the tail domain, such as the heterozygous c.1297C>A which was identified in a 25-year-old Romanian female (Jurcu et al., 2017), may affect the interaction of desmin with other cytoskeletal proteins. The construction
of a cytoplasmic intermediate filament network is hereby obstructed.
Figure 4.2.1. Structural characteristics of desmin and mutations identified in DES.
Top: some of the identified mutations in different segments of the desmin protein. Bottom: molecular model of desmin coiled-coil segment based on structural knowledge of the similar human vimentin protein. (Goldfarb and Dalakas, 2009)
Animal models have been constructed to study the effects of desmin mutations in more detail. A study by Joanne et al. (2013) employed adeno-associated virus vectors to compare the structural effects of R406W and E413K DES mutations to the wild type in tibialis anterior muscles. (fig 4.2.2)
Fig. 4.2.2 Morphological effects of desmin mutations in tibialis anterior muscles.
Serial transversal (A-C) and longitudinal (D) sections of tibialis anterior muscle in different stains. Nuclei were stained blue with DAPI (C, D). Note abnormal mitochondria accumulation (B) for both R406W and E413K (blue arrows). Accumulation of desmin in the perinuclear region is observed for R406W and E413K (yellow arrows). Desmin accumulation is less severe for E413K mutants.
Adapted from (Joanne et al., 2013)
4.3 MFM2 (OMIM #608810)
B-crystallin, also known as HSPB5 (Mitzelfelt et al., 2016), is a small heat shock protein (HSP), found in several tissues and organs. HSPs are expressed in response to stress and HSPB5 has been shown to have an essential role in the heart, skeletal muscle, lens, lung and kidney. (Arrigo et al., 2007, Dubin et al., 1989). It appears to interact with intermediate filament proteins such as desmin, titin and actin, preventing unfavourable non-covalent interactions amongst them (Perng et al., 1999). Accordingly, mutations in the CRYAB gene affecting B-crystallin structure can lead to a variant of MFM with phenotypes similar to those of MFM1.
Loss-of-function CRYAB mutations, such as a 343delT mutation (Mitzelfelt et al., 2016), are a suggested cause for the formation of desmin and B-crystallin protein aggregates in MFM2. A heterozygous R120G missense mutation in a French family (Vicart et al., 1998) and a D109A missense mutation in a Polish family (Fichna et al., 2017) are other examples of MFM2-associated CRYAB mutations. A putative pathological explanation is the weakening of intramolecular interactions within the core of CRYAB, causing -pleated sheets to unfold and the protein stability to decrease. The formation of ectopic protein clumps could be due to loss of chaperone function of B-crystallin, allowing other IF proteins to interact and aggregate.
Alternatively, if CRYAB mutations result in gain-of-function, B-crystallin can form aggregates (Fichna et al., 2017). Moving on from the protein core, mutations such as 464delCT and 451C>T (Q151Q), resulting in C-terminal truncation of B-crystallin, may also have MFM-related deleterious effects (Selcen and Engel, 2003).
4.4 MFM3 (OMIM #609200)
The myotilin gene (MYOT), associated with MFM3, encodes a 57 kDa structural protein expressed in cardiac muscle and more so in skeletal muscle (Salmikangas et al., 1999). This protein is part of the I-line of the sarcomere and interacts with -actinin.
Limb Girdle Muscular Dystrophy type 1A (LGMD1A) was reclassified as MFM3 based on a workshop by Straub et al.(2018).
Clinical features of MYOT-MFM include distal myopathy of the lower limbs or limb girdle involvement, and signs of neuropathy. Development of cardiomyopathy is observed regularly. A peculiar feature observed in MFM3 is an abnormal abundance of ubiquitin and gelsolin, which is not observed in MFM1 and MFM2 (Selcen and Engel, 2004) (Olive et al., 2005).
Notably, MFM-related mutations in the myotilin gene (MYOT) are frequently located in exon 2 and involve serine substitutions (Selcen and Engel, 2004) (Olive et al., 2005). Although one of these heterozygous missense mutations affects an -actinin binding domain of myotilin (S95I), other mutations (S60C, S60F, S55F) do not and the pathological mechanism remains vague.
4.5 MFM4 (OMIM #609452)
The Z-band alternatively spliced PDZ-motif containing protein (ZASP), encoded by the LDB3/ZASP gene, is a skeletal and cardiac Z-disc protein which plays a role in structural sarcomere integrity. (Faulkner et al., 1999) (Au et al., 2004). PDZ motifs at the N-terminal of ZASP are important for protein-protein communication as they can recognise certain C-terminal or internal sequences in other proteins (Harris and Lim, 2001). Furthermore, exons 4 and 6 in ZASP contain a ZASP-like motif (ZM). This ZM is necessary for interaction with -actinin, which is a protein facilitating attachment of actin filaments to the Z-disc (Klaavuniemi et al., 2004).
First associated with MFM4 was the Caucasian A165V mutation in exon 6 of LDB3 (Selcen and Engel, 2005), and it was later identified as the cause of Markesbery disease (Griggs et al., 2007). This mutation lies in the ZM and likely affects cooperation with -actinin, similar to the A147T mutation, located near the ZM (Selcen and Engel, 2005). A recently discovered c. 463 A>C (p.N115H) missense mutation within a PDZ-domain of ZASP has similar pathological features. (Zheng et al., 2017).
Symptoms include skeletal myopathy commencing distally in the lower limbs, and cardiomyopathy and neuropathy can play a role.
Six different ZASP isoforms are expressed dependent on the muscle type. Huang et al. (2003) used knock-in/knock-out methods to study splicing activity of Cypher, the murine homologue of ZASP. Alternative splicing of exons 9 and 10 result in presence of three different isoforms in skeletal muscle, two of which contain a PDZ domain at the N-terminal, an internal ZM and three C-terminal LIM domains (Lin et al., 2014). The remaining short isoform does not contain the LIM domains, thought to be associated with protein kinase C interaction. Depending on the exon affected, MFM4 could have a range of effects in either cardiac or skeletal muscle, or both.
4.6 MFM5 (OMIM #609524)
Filamins are a group of related proteins which are involved in actin-binding and dimerization based on structural components such as 24 immunoglobulin-like (Ig-like) domains following the N-terminal actin binding regions (Himmel et al., 2003). The FLNC gene, associated with MFM5, encodes the protein filamin c, which is known to be expressed in cardiac and skeletal muscle (Maestrini et al., 1993).
MFM5 typically involves proximal myopathy of the lower limbs with limb girdle involvement. Respiratory weakness can occur and cardiac issues occur in approximately one-third of patients (Furst et al., 2013) (Shatunov et al., 2009).
Vorgerd et al. (2005) described the first mutation (8130G>A W2710X) in the FLNC gene related to an autosomal dominant form of MFM in a German family. A truncation of immunoglobulin-like domain 24, necessary for dimer formation, results in aggregation of filamin c in skeletal muscle fibres. A subsequent study by Ruparelia et al. (2016), suggests that this mutation in FLNC affects the interaction of filamin C with binding proteins at the Z-disc. This leads to formation of small protein aggregates in which mutant filamin C gets entangled, leaving an inadequate amount available to function at the Z-disc. Bag3, which acts as a chaperone in autophagy pathways also gets caught in these aggregates. This inhibits the clearance of protein aggregates which worsens the effect.
Other mutations including a c.2997_3008del resulting in a p.Val930_Thr933del residue deletion in a German family (Shatunov et al., 2009) and a K899-V904del and V899-C900ins in a Chinese family (Luan et al., 2010), both localizing to the 7th Ig-like domain of filamin c, were reported to be associated with a similar form of MFM. Although the pathology involved in the p.Val930_Thr933del remains uncertain, the phenotypic effects of this mutation resemble those observed in W2710X patients.
4.7 MFM6 (OMIM #612954)
Bag3 (Bcl-2-associated athanogene-3) is a muscle protein encoded by BAG3, which acts as a co-chaperone to the Hsp70 family, and is concentrated near the Z-disc (Selcen et al., 2009) (Homma et al., 2006). Experiments on mice demonstrate that BAG3 is predominantly expressed in cardiac and skeletal (striated) muscle (Homma et al., 2006).
Typical symptoms of the rare BAG3-MFM (MFM6) include progressive muscle weakness in limbs and axial musculature, along with development of cardiomyopathy and respiratory insufficiency. Frequent occurrence of neuropathy in MFM6 is notable (Jaffer et al., 2012).
Commonly associated with autosomal dominant forms of Bag3-opathy is the heterozygous missense mutation p.Pro209Leu (c.626C>T) in exon 3 of BAG3 (Selcen et al., 2009) (Odgerel et al., 2010) (Lee et al., 2012).
4.8 MFM7 (OMIM #617114)
Blanco et al. ((2001) showed that the Kyphoscoliosis peptidase (ky) gene in mice encodes a protein expressed in heart and skeletal muscle.The homologous human KY gene is only mildly expressed in cardiac muscle (Hedberg-Oldfors et al., 2016). In kyphoscoliosis mice, effects of a recessive loss-of-function mutation of the ky gene were used to demonstrate the importance of the KY protein in muscle growth and strength. Beatham et al. (2004) showed that the KY protein has a transglutaminase-like domain and a role in regulation of filamin c. Besides filamin c, other sarcomeric proteins such as KY-interacting protein 1, titin and myosin binding protein C slow-type interact with KY.
KY-MFM (MFM7) typically begins with myopathy of the lower limbs. Axial and upper limbs can also become involved, and in severe cases tongue muscle atrophy and impaired cognition were involved. Cardiac issues have not been observed.
A homozygous loss-of-function mutation c.1071delG (p.Thr358Leufs*3) in the KY gene was found in a Kurdish female, born of consanguineous parents and suffering from MFM7 (Hedberg-Oldfors et al., 2016). Subsequently, a homozygous nonsense c.405C>A mutation was identified in two MFM Arab-Israeli brothers, again with related parents (Straussberg et al., 2016). Misplacement of filamin c became apparent after studying the muscle pathology. However, more cases must be analysed to clarify the disease mechanism.
4.9 MFM8 (OMIM #617258)
Pyridine nucleotide-disulphide oxidoreductase domain-containing protein 1 (PYROXD1), is classified a nuclear-cytoplasmic oxidoreductase 1 encoded by PYROXD1, and belongs to the pyridine nucleotide-disulphide reductases (PNDRs) (O’Grady et al., 2016). Two enzymatic domains this protein appears to contain are an N-terminal pyridine nucleotide-disulphide oxidoreductase and a C-terminal NADH-dependent nitrate reductase.
PYROXD1-MFM typically entails progressive weakening of the proximal and distal muscles. Commonly, creatine kinase levels are elevated and the disease affects facial muscles as well as swallowing and respiratory functioning. Out of the nine patients from five unrelated families, only one individual displayed cardiac restriction.
Five mutations have been found, including 1 frameshift mutation, 2 splice site mutations, and two missense mutations. The two missense mutations, c.464A>G (p.Asn155Ser) and c.116G>C (p.Gln372His), seem to cause a partial loss-of-function in PYROXD1.
5. Discussion
As becomes apparent from this review that a range of mutations in different genes can cause MFM. There is often a fine line between different muscle disorders, taking the reclassification of LGDM1A as MFM3 (Straub et al., 2018) as an example. This observation and the fact that disease symptoms can vary significantly in different cases, raises the question as to whether MFM can be considered one disease.
However, the collection of genes linked to the MFM phenotype is not endless. They are genes encoding proteins located in or near the Z-disc, and have role in structure, chaperoning or autophagy in the myofibril (Fichna et al., 2018). Genes involved in MFM are often heavily associated with other (muscular) disorders (table 4.1.1), explaining the overlap between distinct muscle diseases. Proteins, being involved in multiple pathways, tend to interact with many other proteins. Mutations in any of the genes involved can disrupt pathways.
The trademark which distinguishes MFM from other muscular diseases and supports linkage of the different types of MFM, is the role of the Z-disc as starting point of myofibrillar disorganisation (Fichna et al., 2018).
For adequate diagnosis and prognosis, noting the differences amongst the subtypes is essential. As this is not currently achieved by clinical examination, improved genetic approach is necessary. For example, consider that MFM7 and MFM8 are predominantly restricted to skeletal muscle, while other subtypes involve cardiac muscle, necessitating close cardiac monitoring. Genetic examination of affected individuals will identify genes concerned, giving an indication of expected pathology and whether to expect cardiac issues.
When considering genetic screening of family, the ages and inheritance patterns typical of different MFMs should be taken into account (table 4.1.2).
The late disease onset common to some MFMs implies that asymptomatic younger members of the family may still succumb at a later stage. The inheritance pattern indicates the likelihood of other family members being affected. If the mutation is autosomal dominant, one mutated allele is enough to affect an individual, whereas patients of autosomal recessive subtypes must have two malfunctioning alleles. This does not mean that anyone suffering from an autosomal recessive subtype is a homozygote. Two recessive alleles with different loss-of-function mutations can also cause disease in a heterozygote.
Regarding correlation between ethnicity and known mutations, finding meaningful connections is difficult. Noticeably, most of the work published is based on studies done in the Western world. This creates a bias, making it seem like certain mutations are more common in Western countries, while few studies are carried out elsewhere. Of course, there are exceptions: most cases of MFM4 are associated with p.A147T or p.A165V mutations, originating from Caucasians (Zheng et al., 2017).
Whilst there have been advances in genetic research of MFM, the underlying genetics have not yet been established for each patient, which could be related to inadequate screening techniques. It is also presumable that more genes are involved in MFM than have been identified so far. Identification of mutations in genes SQSTM1 and TIA1 was only recently linked to the MFM phenotype (Niu et al., 2018). A clearer overview of the proteins involved in myofibrillar mechanisms could help identify other genes involved in MFM. Establishing whether pathway dysfunction is caused by loss-of-function or gain-of-function mutations would assist treatment choice.
Although one must be realistic regarding time and cost of research, an improved genetic approach towards MFM diagnosis has potential. Knowledge of the more frequent mutations leads to more efficient screening. With targeted genome editing techniques such as CRISPR-Cas9 advancing, new treatments could be tested on the well-studied forms of MFM.
The fact that each MFM subtype presents with varying severity and age of onset in different cases, prompts one to wonder what other factors besides type and location (exon) of the gene mutation play a role. Some hypothesize that epigenetics and environmental factors influence the way in which the disease is expressed (Griggs and Udd, 2011). The observation that skeletal muscles of Cypher-I-mice must be put under strain before myopathy is observed could also suggest correlation between lifestyle and disease symptoms (Selcen and Engel, 2005).
6. Conclusion
MFM subtypes are associated with a variety of genetic causes and an even wider range of clinical symptoms. Nevertheless, it becomes apparent that the associated genes are all involved in maintaining structure and integrity at the Z-disc.
Establishing which dysfunctional mechanisms act as the cause of MFM, will guide research towards identifying all the genes involved. This prospect holds potential for future MFM research, aiming for a clearer interpretation of the development of this disease. Eventually, a deeper understanding of the genes and proteins involved, paired with an efficient approach towards genetic screening of patients, could lead to advances in treatment.
2018-11-21-1542809062