Download DNA level results in a phenotype of the patient

DNA level results in a phenotype of the patient
Myotonic dystrophy (DM) is a complex, slowly progressing, highly variable,
mutisystemic disorder which occurs in patients of any age. It is an autosomal
dominant clinical syndrome affecting approximately 1 in 8000 individuals
worldwide and is characterised by myotonia in skeletal muscles, progressive
myopathy, cateracts, gonadal dysfunction, mental retardation, and cardiac
conduction defects (Machuca-Tzili et al. 2005; Harper 2001). DM is caused
by nucleotide expansion repeats above a critical size. There are two main
types of the disease, myotonic dystrophy type-1 (DM1) and myotonic
dystrophy type-2 (DM2), caused by tandem repeat expansions in different
genes. DM1, the most common form of the disease, is caused by cytosinethymine-guanine (CTG) trinucleotide repeats, whereas DM2 is caused by
cytosine-cytosine-thymine-guanine (CCTG) tetranucleotide repeats.
The DM1 mutation was identified in 1992 to be expanded CTG repeats
located within the 3’-untranslated region (3’-UTR) of the DMPK (dystrophia
myotonica protein kinase) gene located on chromosome 19 at position 13.3
(J. David Brook et al. 1992). In healthy, unaffected individuals there are
between 5-38 CTG repeats here. Symptoms can occur in DM1 patients with
as few as 50 repeats, with larger expansions (up tp >4000) correlating with an
increased severity of the disease and also a lower age of onset (Tsilfidis et al.
1992). DM1 also shows an increased severity in phenotype through
inheritance in successive generations, due to the repeats in the DNA
becoming larger through meiosis and replication. This is termed “anticipation”,
and can occur in many nucleotide repeat disorders when DNA polymerase
“slips” on the repeat sequence during replication, forming a hairpin structure.
This is incorporated into DNA and over each replication cycle the repeat
sequence grows larger. DM1 is considered the more severe of the two
disease types, exhibiting adult-onset, child-onset, and congenital (the most
severe) forms of the disease. Only the adult-onset form has been reported in
DM2.
Unlike DM1, the repeat size does not appear to affect the severity or age of
onset for DM2. The CCTG expansions in DM2 are generally much larger than
in DM1, ranging from 75 – 11,000 repeats in length, with a mean size of 5000
repeats (Liquori et al. 2001). The fact that increased repeat length does not
affect disease severity in DM2 may be because of a “ceiling” effect, where
any further increase to repeat size will have no effect on disease severity
because the repeat is already so long it results in maximal effects. DM1
repeats are generally much shorter, and below this theoretical ceiling
threshold, so that an increase in length will still have an effect on severity.
There have been several hypotheses proposed to explain DM pathogenesis.
Initially, when only one disease type (DM1) was known, the DMPK gene was
investigated. The idea being that the CTG repeats inhibited DMPK mRNA or
protein production, which resulted in DMPK haploinsufficiency. Expression of
DMPK mRNA and protein was shown to be reduced in DM1 muscle (Fu et al.
1993). DMPK knockout mice were also shown to exhibit cardiac conduction
defects (Berul et al. 1999), a characteristic symptom of DM. (insert diagram
from this cardiac conduction defects paper?) However, this study only
focused on cardiac conduction problems, and no findings of other DM
symptoms were reported. Mice only exhibited mild myopathy in another
DMPK-knockout study by Jansen et al. (1996), and still lacked many of the
multisystemic symptoms of the disease. Neither of these studies produced
mice that exhibited myotonia, one of the characteristic features of the
disease, and in fact to date, no DMPK loss-of-function mutation has been
reported in DM1 individuals. The multisystemic features of the disease cannot
therefore be explained by DMPK haploinsufficiency alone.
Another model behind the pathogenesis of DM1 involves the CTG repeats in
the 3’UTR of the DMPK gene affecting the expression of adjacent genes.
Otten & Tapscott (1995) demonstrated that a DNase I hypersensitive site is
found 3’ of the CTG repeat in wild-type alleles. DM1 individuals with large
expansions exhibited alleles with DNase I resistance, as well as
inaccessibility of nucleases to the adjacent hypersensitive site. This
suggested the CTG repeat expansion alters the chromatin structure, causing
a region of condensed chromatin, which could affect expression of adjacent
genes. A homeodomain-encoding gene, SIX5, is located in this region and its
mRNA level has been shown to be decreased in DM1 individuals (Charles A.
Thornton et al. 1997). SIX5-knockout mice have also been shown develop
cateracts, another symptom of DM, however no muscle pathology was
observed in these mice (Klesert et al. 2000), which means decreased SIX5
expression cannot be the primary cause of DM1. A second gene, DMWD, is
also located next to the DMPK gene and is has been shown to have
decreased levels of RNA expression in some DM1 cell lines (Alwazzan et al.
1999). However, no specific link was determined between reduced DMWD
expression and phenotype, or between repeat length and DMWD expression,
probably due to the small sample population used and the lack of variation in
repeat lengths within it. With regards to the effect on adjacent genes, more
has been reported and is known about SIX5 than DMWD.
A RNA gain-of-function hypothesis was proposed in which the mutant RNA
transcribed from the expanded repeat-containing allele is sufficient to induce
disease symptoms. This hypothesis was based on several observations. The
fact that the loss of function of DMPK or surrounding genes did not reproduce
the major symptoms of the disease, which meant another main mechanism
must be present. Also, it was shown that expression of only the DMPK 3’UTR region with 200 CTG repeats was sufficient to inhibit myogenesis
(Amack et al. 1999). It is also known that the expanded CTG repeats are
transcribed into CUG repeats, which accumulate in discrete nuclear foci
(Davis et al. 1997), a hallmark of dominant RNA-mediated diseases. The
repeats are thought to form imperfect hairpins, becoming more stable as the
repeat length increases. The association of RNA binding proteins leads to the
accumulation of these repeat containing RNAs within nuclear foci. The result
of this is ineffective metabolism of mutant mRNA, and the bound proteins can
no longer perform their normal functions. A study by (A Mankodi et al. 2000)
gave direct experimental support for a RNA gain-of-function mechanism
behind DM. In this study they used a mouse model containing 250 CTG
repeats in the 3’UTR regions of the human skeletal α-actin gene. This gene is
not linked with DM, and yet the mice developed myotonia and showed muscle
histology similar to DM1 individuals. This demonstrated that on their own, the
repeats were enough to induce pathogenic features of DM1, irrelevant of the
gene context.
A second type of DM, termed DM2, was identified in 1998 (Ranum et al.
1998) and in 2001 it was attributed to CCTG repeats within intron 1 of the
zinc finger 9 (ZNF9) gene on chromosome 3 (Liquori et al. 2001). Patients
with DM2 showed similar symptoms to those with DM1, although subtle
differences existed and no congenital form has been reported in DM2. The
fact these two repeat sequences are located in completely different genes
and on different chromosomes, yet can still cause such similar symptoms
supports the idea of RNA gain-of-function as a common pathogenic
mechanism.
Repeat containing RNA can cause pathogenic disease features through
interaction with RNA-binding proteins. Two important proteins have been
identified by their ability to bind CUG repeats in RNA: CUGBP (CUG-binding
protein) and MBNL1 (Muscleblind-like 1). CUG repeat expansions can fold
into hairpin-like structures (Napierala & Krzyzosiak 1997) and particular
proteins can be sequestered to these hairpins, altering the levels at which
they are present in the nucleoplasm.
MBNL1, was identified in a study by Miller et al. (2000) when looking for
proteins that bind CUG repeat expansions. In DM1 cells, MBNL1 is
sequestered to the nuclear foci in DM1 cells containing CUG repeat RNA,
thereby depleting it from the nucleoplasm (Jiang et al. 2004). Mice with a
modified MBNL1 gene that had a deletion of exon 3 (containing a RNAbinding motif) developed some of the characteristic DM1 features, including
cateracts and myotonia, as well as exhibiting alternative splicing defects
(Kanadia et al. 2003). Another study by Kanadia et al. (2006) showed that
overexpression of MBNL1 in a mouse model for DM was shown to restore
normal splicing and reverse myotonia. Disrupted alternative splicing is one of
the most important molecular features of DM and the mis-regulated splicing
events are believed to be behind many of the symptoms of DM. Misregulation of splicing of the skeletal muscle-specific chloride channel 1 (ClC1) leads to the myotonia seen in DM1, for example (Mankodi et al. 2002;
Charlet-B et al. 2002). The two studies by Kanadia et al. suggested that
depletion of the MBNL1 protein, a regulator of alternative splicing, resulted in
many of the downstream splicing problems that gave rise to multiple DM
phenotypes. Another study by (Ho et al. 2004) showed decreased expression
of MBNL1 in cultured cells caused aberrant splicing in both the cardiac
troponin T (cTNT) and insulin receptor (IR) genes, which is consistent with
DM1 individuals.
While there is much evidence supporting MBNL1 sequestration as a cause of
DM1 pathogenesis, it is not the only mechanism involved. It should be noted
that in the study by Kanadia (2003), while mice did display symptoms such as
cateracts and myotonia, no signs of muscle wastage appeared. This therefore
suggests some features of the disease are due to other causes and not
MBNL1 loss-of-function. Supporting this idea is a study by (Ho et al. 2005)
that showed MBNL1 was sequestered to both CUG- and CAG-containing
nuclear foci with similar affinity, however only the CUG repeats caused misregulation of splicing in the cTNT and IR genes. This therefore suggests
sequestration of MBNL1 alone is not enough to cause all the splicing changes
involved in DM, and other factors are in force.
Another protein involved in the modulation of splicing events is CUGBP. It
was identified by (Timchenko et al. 1996) by its ability to bind short singlestranded CUG repeats. However, CUGBP does not in fact bind doublestranded repeats (such as those present in the hairpin structures), or colocalise with the foci of expanded repeat transcripts (Fardaei et al. 2001), and
so is not sequestered like MBNL1. Instead, CUGBP levels are increased in
DM1 muscle cells and heart tissue (Philips et al. 1998; Timchenko et al.
2001). A mouse model overexpressing CUGBP in muscle exhibited splicing
defects and muscle problems similar to DM1 (Ho et al. 2005), which further
supported the idea that CUGBP is involved in abnormal splicing and disease
pathogenesis. Many other studies have reported similar findings, providing
further evidence that increased CUGBP expression (and the subsequent
downstream splicing problems) is one of the main effects of the DMPK repeat
expansion (see Wang et al. 2007 and Orengo et al. 2008, for example).
A study by Timchenko et al. (2001) revealed that DM myoblasts become
incapable of cell cycle withdrawal during differentiation. Their findings
suggested that alterations in CUGBP levels disrupted a CUGBP translational
target, p21, resulting in impaired withdrawal from the cell cycle in DM
patients. Another mouse model study by Timchenko et al. (2004) revealed
CUGBP to have a role in the regulation of translation. It interacts with the
mRNA transcript of MEF2A (myocyte enhancer factor 2A), increasing its
translation. MEF2A is a transcription factor involved in myogenesis, and its
levels are found to be increased in DM1 skeletal muscle, which is believed to
contribute towards delayed myogenesis.
The entire mechanism through which CUGBP levels are increased because
of the repeat containing RNAs is not fully known, although it has been shown
that PKC (protein kinase C) mediates phosphorlyation of CUGBP (KuyumcuMartinez et al. 2007). The expression of expanded CUG repeats in DMPK
RNA induces hyperphosphorylation of CUGBP, which causes an increased
protein half-life and steady-state levels. This study used a phorbol ester to
activate PKC, and the direct phosphorylation of CUGBP by both PKCα and
PKCβII was observed in vitro. PKCα and PKCβII were also found to be
activated in cells expressing DMPK-CUG-repeat containing RNA in DM1 cell
cultures, DM1 tissues, and also a DM1 mouse model. The mechanism
through which the repeat containing RNA activates the PKC pathway remains
unknown, although this will surely be an area for future work and it could lead
to some potential treatment areas for the disease.
There is a wide range of effects the aberrant splicing, caused by altered
levels of CUGBP and MBNL1, can have (figure #). For example, DM patients
are predisposed to diabetes type II, due to an unusual form of insulin
resistance caused by a defect in skeletal muscle. DM patients express
inappropriate levels of the IR-A (insulin receptor-A) isoform, due to misregulation of splicing resulting in increased skipping of the IR exon 11. This
abnormal splicing causes a switch from IR-B to IR-A (Savkur et al. 2001). The
IR-A isoform has a twofold higher affinity for insulin, as well as a faster
internalisation and recycling time than the B isoform. The mis-regulation of IR
alternative splicing can be reproduced in normal cells by the expression of
CUG-repeat RNA, and also the expression of CUGBP in normal cells induces
the change to IR-A as well (Savkur et al. 2001).
Aberrant splicing events, such as that observed for IR, cause a variety of
different symptoms. The exact mechanism behind some of the phenotypes
still remain relatively unknown (figure #) and are areas of future work. The
aberrant splicing events involved in the cTNT, ClC-1, and IR are among the
better understood pathways behind DM phenotypes (see Ho et al. 2004,
Mankodi et al. 2002, and Savkur et al. 2001, for example). For these three
examples (cTNT, CLC-1, and IR), when CUGBP induces inclusion of an exon,
MBNL1 increases exclusion of that exon, and vice-versa.
Interestingly, in these three examples (cTNT, IR and ClC-1), when CUGBP1
induces inclusion of an exon, MBNL1 increases exclusion of the same exon
and vice versa. Kalsotra et al. [42] performed a large screen to identify
alternative splicing events regulated during mouse heart development and
identified CUGBP1 and MBNL1 as regulators of more than half of the splicing
transitions tested. Many of these developmentally regulated splicing events
are modulated exclusively by CUGBP1 or MBNL1, but all events regulated by
both proteins exhibit antagonistic responses [42]. In addition, while MBNL1
nuclear levels increase during development, CUGBP1 nuclear levels
decrease [22,42], supporting the hypothesis that the level and localization of
these two proteins control a fetal to adult splicing transition, which is reversed
in DM1 tissues (Figure 2).
Potential causes of clinical distinctions: One possible mechanism for
modulating the toxic RNA effect could be that CUG-BP and MBNL affinity for
CUG differs from their affinity for transcripts containing CCUG.
Clinical differences could reflect variation in the RNA effects produced by
dissimilarities in temporal and cell specific expression patterns of DMPK and
ZNF9.
Alternatively, there could be a separate mechanism accounting for phenotypic
differences differences between DM1 and DM2 such as alterations in the
expression of locus specific genes including DMPK and SIX5 in DM1.
Mention the DM3 thing and criticise it.