Journal of Applied Genetics (2018) 59:253–268

https://doi.org/10.1007/s13353-018-0444-7

HUMAN GENETICS • REVIEW

Splicing mutations in human genetic disorders: examples,

detection, and confirmation
Abramowicz Anna 1 & Gos Monika 1

Received: 18 February 2018 / Revised: 8 April 2018 / Accepted: 10 April 2018 / Published online: 21 April 2018
# The Author(s) 2018

Abstract
Precise pre-mRNA splicing, essential for appropriate protein translation, depends on the presence of consensus Bcis^ sequences
that define exon-intron boundaries and regulatory sequences recognized by splicing machinery. Point mutations at these con-
sensus sequences can cause improper exon and intron recognition and may result in the formation of an aberrant transcript of the
mutated gene. The splicing mutation may occur in both introns and exons and disrupt existing splice sites or splicing regulatory
sequences (intronic and exonic splicing silencers and enhancers), create new ones, or activate the cryptic ones. Usually such
mutations result in errors during the splicing process and may lead to improper intron removal and thus cause alterations of the
open reading frame. Recent research has underlined the abundance and importance of splicing mutations in the etiology of
inherited diseases. The application of modern techniques allowed to identify synonymous and nonsynonymous variants as well
as deep intronic mutations that affected pre-mRNA splicing. The bioinformatic algorithms can be applied as a tool to assess the
possible effect of the identified changes. However, it should be underlined that the results of such tests are only predictive, and the
exact effect of the specific mutation should be verified in functional studies. This article summarizes the current knowledge about
the Bsplicing mutations^ and methods that help to identify such changes in clinical diagnosis.

Keywords Splicing mutation . Spliceosome . Pre-mRNA splicing . Splicing enhancers and silencers

Introduction In brief, the splicing process is catalyzed by the

spliceosome, a protein-RNA complex containing five small
The mature mRNA contains only coding sequences—the nuclear ribonucleoproteins (snRNPs, U1, U2, U4–U6) and
intronic ones are removed from the transcript during the splic- over 300 different proteins. The presence of small nuclear
ing process (Rahman et al. 2015). This process is held in RNAs (snRNAs) in snRNPs allows to form complementary
nucleus and is dependent from the presence and interaction RNA-RNA complexes and thus identify the specific se-
between the so-called cis and trans elements. The cis elements quences of the splicing sites by the spliceosome (Fig. 2,
are the DNA sequences that define exons, introns, and other Faustino and Cooper 2003).
regulatory sequences necessary for proper splicing. They are In most of cases (98.7%), the exon/intron boundary se-
termed consensus splice site sequences and include donor (5′) quences contain GT and AG motifs at the 5′ and 3′ ends of
and acceptor (3′) splice sites, branch point and polypyrimidine the intron, respectively. Noncanonical GC-AG and AT-AC
tract sequences, and auxiliary cis elements including splicing sequences at the splice sites occur in 0.56 and 0.09% of the
silencers and enhancers (Fig. 1). splice site pairs. The noncanonical GC-AG splice site, like the
GT-AG splice pair, is processed by major spliceosome com-
plex. The AT-AC splicing motifs, that are present in U-12 type
introns, are recognized by functionally relevant U12-
Communicated by: Michal Witt dependent spliceosome called minor spliceosome (Parada
et al. 2014). This complex is similar to the major spliceosome
* Gos Monika with the exception of several snRNPs. The U1, U2, U4, and
[email protected] U6 are replaced with U11, U12, U4atac, and U6atac, respec-
1
tively. The U11/U12 proteins bind to their target sequences as
Department of Medical Genetics, Institute of Mother and Child,
Kasprzaka 17a, 01-211 Warsaw, Poland
a preformed di-snRNP complex (Turunen et al. 2013).
254 J Appl Genetics (2018) 59:253–268

Fig. 1 The schematic localization of the cis and trans splicing elements. and the polypyrimidine tract sequences are highly degenerated and
The cis elements are the DNA sequences that include donor (5′) and together with donor and acceptor sites are recognized by the elements
acceptor (3′) splice sites, branch point and polypyrimidine tract of the splicing complex called spliceosome. Spliceosome proteins
sequences, and splicing silencers and enhancers. Donor and acceptor together with splicing repressors and activators recognize cis splicing
sites are evolutionary conserved and are usually defined by GT and AG elements and are called trans-acting elements
nucleotides at the 5′ and 3′ ends of the intron, respectively. The branch site

Fig. 2 The scheme of the splicing process. The splicing process is snRNP, and the ATP-dependent (A) complex is formed. The interaction
performed in two steps. The first step is the recognition of the splicing between the branch point and the U2snRNP protein is stabilized by
sites at intron/exon junctions, and the second one is the intron removal specific RNA helicases (Prp5 and Sub2), and this is a signal for the
and exon ends joining. During the splicing process, four complexes recruitment of U4/5/6 tri-snRNP and formation of the B complex (pre-
between the pre-mRNA and spliceosome are formed. The first one is catalytic spliceosome). Further action of additional RNA helicases leads
the early complex (E). The U1 snRNP recognizes and binds to the to change of spliceosome conformation that leads to the release of U1 and
complementary AG-GU sequence at the donor splice site (5′ end of the U4 snRNPs, the interaction between U6 with U2 snRNP, and the
intron). In the same time, the SF1 protein binds to the branch point. The formation of a pre-mRNA loop and the C complex. In this complex,
SF1 is recognized and bound by the U2AF65 protein that also binds to the two transesterification reactions take place, intron is removed, and ends
polypyrimidine sequence located between the branch point and 3′ end of of exon are joined (Fredericks et al. 2015; Tazi et al. 2009)
the intron. Then, the SF1 is displaced from the branch point by the U2
J Appl Genetics (2018) 59:253–268 255

Any errors during the splicing process may lead to improp- function (Chabot and Shkreta 2016). The aberrant splicing of
er intron removal and thus cause alterations of the open read- pre-mRNA due to the presence of point mutations, e.g., nucle-
ing frame. Therefore, the spliceosome complex has to correct- otide substitutions, that alter the consensus splicing regulatory
ly recognize and cut out the intronic sequences from the pre- sequences in a specific gene, may lead to the specific heredi-
mRNA molecule. The proper identification of the splice site is tary monogenic disorders. This article summarizes the current
demanding as the consensus sequences are very short and knowledge about the Bsplicing mutations^ and methods that
there are many other sequences similar to the consensus mo- help to identify such changes in clinical diagnosis.
tifs of the canonical splice sites. These sequences are known
as cryptic, noncanonical, or pseudo splice site sequences.
Sometimes, a pseudo splice site matches the consensus se- Cis-element splicing mutations
quences better than the natural one, but because of the lack
of cis-acting regulatory elements, necessary for the proper In general, the term splicing mutations usually refers to the
exon identification, it is not used in the splicing process point mutation at the cis consensus sequences that effects in
(Cartegni et al. 2002). The cis-regulatory elements include improper exon and intron recognition in messenger RNA and
exonic and intronic splicing enhancers (ESE and ISE, respec- results in the generation of an aberrant transcript of the mutat-
tively) and exonic and intronic splicing silencers (ESS and ed gene. The splicing mutation may occur in both introns and
ISS, respectively) (Glisovic et al. 2008). These elements are exons and disrupt existing splice sites, create new ones, or
recognized by specific splicing repressors and activators activate the cryptic ones. They also can influence splicing
(trans-acting elements) that help to properly carry out the enhancers and silencers binding or alter the secondary struc-
splicing process. ture of messenger RNA and therefore prevent the binding of
Splicing enhancers are conserved nucleotide sequences, the spliceosome elements. Usually such mutations result in
specifically recognized by the serine and arginine-rich pro- exon/exon fragment skipping during the pre-mRNA splicing.
teins (SR proteins). These proteins bind to the specific If the resulting deletion is an in-frame one, the shorter protein
intronic/exonic splicing enhancers via RNA recognition mo- will be synthetized. When the deletion results in the shift of
tifs and interact with other splicing factors such as snRNP the open reading frame, a premature stop codon (PTC) can be
proteins. This interaction is mediated by the C-terminal RS introduced and shorter protein can be produced. However, the
domain of the SR proteins. The SR proteins establish an presence of the PTC in the transcript usually leads to faster
RNA/RNA interaction at the 5′ splice site and the branch mRNA degradation during a protective process called non-
point, assist the formation of an early spliceosomal complex sense mediated decay (NMD). The degradation of the defec-
E, and seem to promote the inclusion of exons they are bound tive messenger RNA prevents aberrant protein synthesis and
with (Caceres and Kornblihtt 2002; Lee and Rio 2015). has the same effect as gene deletion or nonsense mutation
Also, negative regulators are involved in the regulation of (Sterne-Weiler and Sanford 2014). According to Wimmer
the splicing process. These proteins, mainly heterogeneous et al. (2007), splicing mutations can be briefly divided into
nuclear ribonucleoproteins (hnRNP), bind to exonic and five categories that are further briefly discussed (Fig. 3).
intronic splicing silencers (Wagner and Garcia-Blanco
2001). The hnRNP protein family includes different proteins Mutations in the canonical acceptor and donor sites
with molecular weight between 34 and 120 kDa, called alpha- (type I and IV mutations)
betically from hnRNP A to hnRNP U (Han et al. 2010). The
hnRNP proteins differ with RNA-binding domain (RBD) that In general, mutations in the canonical acceptor and donor sites
is responsible for the interaction specificity (Geuens et al. affect strongly conserved sequences that define exon-intron
2016). Their regulatory action may involve the Blooping-out^ boundaries. The 5′ splice site (CAG/GUAAGU sequence)
mechanism—the hnRNP proteins bind to both ends of the and 3′ splice site (NYAG/G sequence) are recognized by the
exon, and because of their interaction, the exon is Blooped elements of the spliceosome. Therefore, any variants in these
out^ and becomes blocked to the spliceosome complex. canonical sequences might alter interaction between pre-
Another proposed mechanism suggests that hnRNPs coat the mRNA and proteins involved in the intron removal. The most
exon and act as an antagonist of splicing enhancers thus classical mutations affect + 1 and + 2 residues at the 5′ donor
preventing their binding (Caceres and Kornblihtt splice site and − 1 and − 2 residues at the 3′ acceptor splice
2002; Zhang et al. 2008a, b; Kolovos et al. 2012). site. The meta-analysis of splicing mutations (478 mutations
To summarize, the splicing process is a complex event that in 38 selected genes) indicated that the donor splice site mu-
is important for proper protein synthesis. Any alterations of tations were more prevalent than the acceptor splice site var-
this process might lead to the decrease of the level of the iants (ratio 1.5:1) (Krawczak et al. 2007). Similar results were
specific messenger RNA and thus deprivation of the protein obtained if a single gene was taken into consideration. Pros
level that can result in aberrant cellular metabolism and/or et al. (2008) had found that the splicing mutations in the NF1
256 J Appl Genetics (2018) 59:253–268

Fig. 3 The types of splicing mutations. According to Wimmer et al. fragment, (4) variants in the canonical splice sites resulting in the usage
(2007), splicing mutations can be briefly divided into five categories: of the cryptic exonic or intronic splice site that leads to the inclusion of an
(1) splicing mutations within the canonical splice sites leading to whole intron fragment or exon fragment skipping, and (5) mutations within the
exon skipping, (2) deep intronic variants creating new splice sites exon, usually leading to ESE disruption, resulting in the entire exon
resulting in the inclusion of cryptic exons, (3) exonic single nucleotide skipping
variants creating new splice sites that result in the loss of an exon

gene more frequently affected 5′ splice site (65%) than the 3′ structures formed by the pre-mRNA. The functional study of
one (35%). According to HGMD® Professional database (re- selected splicing mutations in the DMD gene revealed that the
lease 2017.3), this difference is lower and equals to 56% and same substitution at the same position from the exon can lead
44% for 5′ and 3′ splice site, respectively. to different effect. The NM_004006.2: c.3277+1G>A muta-
Mutations at the canonical splice sequences usually lead to tion in intron 25 leads to the removal of exon 25 from the
single exon skipping (type I mutation, see Table 1). However, transcript, while the NM_004006.2: c.6439+1G>A mutation
if the splice site is weak and the presence of mutation uncovers in intron 45 activates the cryptic exonic splicing site resulting
the cryptic splice site in a neighboring exon or intron, this in the inclusion of exon 45 shorter by 32 bp at the 3′ end. The
alternative site can be used in the splicing process (type IV experimental studies with minigene assays have confirmed
mutation). This can lead to the inclusion of the intron fragment this observation and shown that in the case of the c.6439+
or the removal of an exon fragment if the cryptic splice site is 1G>A mutation, three distinct splicing isoforms that use dif-
present in intron or exon, respectively. In such situation, sev- ferent alternative cryptic sites were observed. The authors
eral distinct transcripts can be generated like in the case of the suggested that the splicing mutation in introns with strong
c.1525-1G>A variant in intron 9 of the CFTR gene. Detailed (highly similar to the consensus motifs, e.g., with high com-
analysis revealed the presence of three different mRNA iso- plementarity to the 5′ end of the U1 snRNA) splicing motifs,
forms that utilized distinct alternative splice sites located with- like the donor site of the intron 45 of the DMD gene, stimu-
in intron 10 and exon 10 at positions c.1610–1611 and lates the use of cryptic splice sites in exons and introns. The
c.1678–1679 and were lacking whole exon 10 or its frag- splicing complex specifically recognizes strong splice site,
ments, respectively (Ramalho et al. 2003). and if the canonical splicing site is mutated, there is a higher
The effect of mutation at the canonical splice site might probability of the activation of the cryptic splice site. In the
also depend on the strength of the splicing site, localization case of weak splice sites, the probability of whole exon skip-
of cryptic splice sites, density of ESE and ESS, or secondary ping is higher than the use of alternative splicing motifs. The
Table 1 Examples of different splicing mutations

Disease Gene Mutation Type of splicing Effect on pre-mRNA splicing Commentary Reference
mutation

Type I
Autosomal dominant MIP c.606+1G>A Donor splice site Exon 3 skipping Point mutation at the canonical splice site Zeng et al. (2013)
congenital cataract mutation leads to the whole exon skipping
Neurofibromatosis type 1 NF1 c.1845+1G>A Exon 15 and 16 skipping Multi-exon deletion, mutation, and weak Fang et al. (2001)
splicing sites force 2 exon skipping,
J Appl Genetics (2018) 59:253–268

description in the main text

Ehlers-Danlos syndrome COL5A2 c.925-2A>G Acceptor splice site Several transcripts: (1) exon 7 and 6 Spliceosome utilizes the donor-splice site of Symoens et al. (2011)
mutation skipping and (2) exon 7 skipping exon 5 and the acceptor-splice site of
exon 8 leads to skipping of both exons
6 and 7
Succinyl-CoA:3-ketoacid OXCT1 c.1248+5G>A Donor splice site Skipping of exons 12 and 13 Mutation leads to Bsplicing paralysis^ and Hori et al. (2013)
CoA transferase (SCOT) mutation the removal of whole intron 11-mutated
deficiency intron 13 fragment, including exons
12 and 13, description in the main
text
Becker muscular dystrophy DMD c.3432+1G>A Exon 25 skipping Point mutation at the canonical splice site Habara et al. (2009)
(BMD) leads to the whole exon skipping
Familial dysautonomia ELP1 c.2204+6T>C Exon 20 skipping (tissue specific) Shortened transcript has the highest Ibrahim et al. (2007),
(FD) (IKBKAP) expression in nervous cells, in lymphoid Axelrod et al. (2011)
cells, the normal transcript is present at the
highest level
Type II
Neurofibromatosis type 1 NF1 c.888+651T>A Deep intronic Cryptic exon inclusion (132 bp) Mutation creates strong 3′ splice site; this Messiaen and Wimmer
(NF1) variants matches with strong, already existing 5′ (2008)
cryptic splice site. This lead to the inclusion
of the cryptic exon containing PTC.
Possible admixture of normally spliced
transcript
c.288+1137C>T Cryptic exon inclusion (118 bp) Mutation creates a strong donor splice site Svaasand et al. (2015)
which interacts with strong, already
existing acceptor cryptic splice site. This
lead to the inclusion of the cryptic exon
containing PTC
Cystic fibrosis (CF) CFTR c.3718-2477C>T Cryptic exon inclusion (84 bp) Mutation creates a novel donor site and results Sanz et al. (2017)
in the inclusion of the cryptic exon with
PTC
Androgen insensitivity AR c.2450-118A>G Two transcripts: (1) cryptic exon Mutation creates a novel binding site for Känsäkoski et al. (2016)
syndrome inclusion (85 bp) and (2) cryptic splicing enhancer, protein SRSF1
exon inclusion (202 bp) (SF2/ASF)—leading to inclusion of a
pseudo-exon containing a PTC
Fabry disease GLA c. 639+919 G>A In-frame inclusion of cryptic exon Mutation abolishes motif that binds splicing Palhais et al. (2016)
(57 bp) silencer protein (hnRNPA1/A2). This leads
to the recognition and inclusion of the
257

pseudoexon sequence
 258

Table 1 (continued)

Disease Gene Mutation Type of splicing Effect on pre-mRNA splicing Commentary Reference
mutation

Type IV
Cystic fibrosis (CF) CFTR c.1525-1G>A Acceptor splice site Several transcripts: (1) deletion of In all three mRNA isoforms, the alternative Ramalho et al. (2003)
mutation exon 10, (2) partial deletion acceptor sites are used. This lead to the
(85 bp) of exon 10, and (3) partial skipping of an exon fragment. Although the
deletion (153 bp) of exon 10 bioinformatic analysis showed that the
alternative acceptor sites of much greater
strength were present in intron 9 than these
actually used, there are ESE sites in exons
10 and 11 that drive the splicing process
towards the use of particular acceptor sites
Ehlers-Danlos syndrome COL5A1 c.655-2A>G Several transcripts: (1) major product The removal of intron 4 is delayed because of Takahara et al. (2002)
exon 5 and 6 skipping, (2) exon 5 the acceptor-site mutation and formation of
skipping, (3) partial deletion SECRIs of exons 5 and 6, description in the
(12 bp) of exon 5, and (4) partial main text
deletion (15 bp) of exon 5
Duchenne muscular DMD c.6614+1G>A Donor splice site Partial deletion (32 bp) of exon 45 Mutation at the canonical splice site activates Habara et al. (2009)
dystrophy (DMD) mutation donor cryptic splice site within exon 45 and
results in 32 bp deletion at 3′ end of exon
45. Intron 45 has strong acceptor splice site
and after its identification, spliceosome is
forced to use the nearest donor splicing site.
The probability of the use of cryptic donor
site is higher
X-linked TRAPPC2 c.238+1A>G Seven different splicing transcripts Very rare AT splicing donor site is changed Xiong et al. (2009)
spondyloepiphyseal into canonical GT splice site that uses major
dysplasia tarda splicing complex instead of minor splicing
complex. The activation of cryptic acceptor
splice sites is possible and leads to the
generation of several alternative transcripts.
In this case, bioinformatic software is not
effective in identifying the cryptic splice
acceptor sites because of the noncanonical
ends
Type V
Medium-chain acyl-CoA ACADM c.362C>T Missense mutation, Exon 5 skipping Mutation causes the loss of ESE site and Ward and Cooper (2010)
dehydrogenase (MCAD) (MCAD) (p.Thr96Ile) abolish ESE abolishes SF2/ASF protein binding motif
deficiency thus leading to exon skipping
Neurofibromatosis type 1 NF1 c. 3362A>G Missense mutation, Two transcripts: (1) containing Mutation results in the presence of two Xu et al. (2014)
(p.Glu1121Gly) decreased the ratio substitution and (2) exon 20 mRNA isoforms: one properly spliced
of the ESE/ESS skipping contains missense change (p.Glu1121Gly)
and the other one lacks exon 20
Stickler syndrome COL2A1 192G>A Nonsense mutation, Exon 2 skipping This is an example of class I-NAS. This McAlinden et al. (2008)
(p.Cys64Ter) abolish ESE nonsense mutation (p.Cys64Stop) causes
J Appl Genetics (2018) 59:253–268

exon 2 skipping by the disruption of ESE.

Table 1 (continued)

Disease Gene Mutation Type of splicing Effect on pre-mRNA splicing Commentary Reference
mutation

This decreases the level of proper

transcript, although shorter mRNA
isoforms do not undergo NMD
Mutations of the branch point sequence
Neurofibromatosis type 1 NF1 2410-18C>G Branch point Partial retention (17 bp) of intron 15 Mutation abolishes the original branch point Xu et al. (2014)
J Appl Genetics (2018) 59:253–268

sequence and creates a putative ESE. Other

splicing mutations near this position were
reported: 2410-16A>G, 2410-15A>G, and
2410-12T>G. It points out that this intron
fragment is critical for proper splicing of
exons 15 and 16.
Xeroderma pigmentosum XPC c.413-9T>A Exon 4 skipping Mutation creates binding site for some factors Khan et al. (2010)
that prevent the recognition of the weak 3′
splice site at intron 3′ exon 4 junction thus
leading to the decrease of normal XPC
mRNA level
XPC c.413-24A>G Exon 4 skipping, 3% normal Mutation abolishes the interaction with U2 Khan et al. (2010)
transcript snRNP and weak c.413-9 site is used
instead of it the normal one resulting in the
expression of minimal amounts of
full-length XPC transcript
Mutations within polypirymidine tract
Hemophilia B F9 c.253-19_ Polypyrimidine tract Exon 3 skipping Mutation shortens the polypyrimidine tract Van de Water et al.
253-16del from 24 nucleotides to 20. This results in (2004)
inefficient splicing and exon 3 skipping
Mitochondrial ACAT1 c.121-13T>A Exon 3 skipping Mutation disrupts polypyrimidine tract and Aoyama et al. (2017)
acetoacetyl-CoA thiolase causes exon 3 skipping. Moreover, other
(T2) deficiency substitutions in this position (T>A,C,G)
also cause exon 3 skipping.

PTC premature termination codon, NMD nonsense-mediated decay

259
260 J Appl Genetics (2018) 59:253–268

authors also checked other 14 mutations affecting + 1 position intron 13, 12, and 11. The presence of mutation affecting
of the DMD gene and found that this observation was true for donor site of the intron 13 caused the delay in this intron
all variants that were localized in exons longer than 170 bp removal thus also affecting the time of intron 12 and 11 splic-
(Habara et al. 2009). ing. Authors suggested that this caused so-called splicing pa-
In certain circumstances, several exons might be deleted ralysis that could be solved by the removal of whole intron 11-
like in the case of the c.1845+1G>A variant in the NF1 gene mutated intron 13 fragment, including exons 12 and 13 (Hori
that leads not only to exon 16 but also to the upstream exon 15 et al. 2013).
removal. During the splicing process of the mutated pre- The effect of a specific splicing mutation may also depend
mRNA, the donor site in the intron 14 and acceptor site in on the tissue type in which the primary transcript is expressed
intron 16 are utilized. The proposed model for double exon and thus the availability of the specific factors. A good exam-
skipping emphasizes the role of strong consensus splice sites. ple is the c.2204+6T>C mutation in the IKBKAP gene, that is
The exon 15 is surrounded by weak splicing sites, and strong found in homozygous state in about 99% of patients with
splicing sites of the exon 16 are needed for proper intron 15 familial dysautonomia inherited in autosomal recessive man-
and 16 removal. If the strong donor site of the intron 16 is ner (Cuajungco et al. 2003; Slaugenhaupt et al. 2001). The
weakened due to the presence of the c.1845+1G>A mutation, bioinformatic analysis with Human Splicing Finder (HSF) did
the stronger splicing sites at the intron 14 are recognized by not predict that the c.2204+6T>C variant affects splicing. The
the splicing machinery and exon 15 is removed from the tran- functional studies had shown, however, that this substitution
script (Fang et al. 2001). caused skipping of the exon 20. Functional analysis of the
The deletion of several exons and generation of distinct IKBKAP RNA level in distinct tissue types revealed that the
mRNA isoforms due to the presence of a splicing mutation wild-type transcript was present at different levels in all ex-
might be also related to the order of intron removal. This was amined samples. Its level was the highest in lymphocyte-
analyzed in detail in the case of the c.655-2A>G mutation in derived cell lines, while the transcript lacking exon 20 had
the COL5A1 gene that mutations were found in Ehlers-Danlos the highest expression in nervous cells. This can explain
syndrome (EDS) patients. In the cells taken from the patient why the nervous system deficits are the most prominent in
with this mutation, several transcripts were found. The major familial dysautonomia patients (Ibrahim et al. 2007; Axelrod
one was lacking exons 5 and 6, the other one was lacking only et al. 2011).
exon 5, while in two other transcripts 12 and 15, nucleotides
of exon 5 were missing due to the use of the cryptic splicing Deep intronic variants as a cause of cryptic exon
sites. It was suggested that the major transcript is created be- inclusion (type II mutations)
cause of the delayed removal of the intron 4 due to the pres-
ence of acceptor site mutation. As the removal of the intron 5 Another category of splicing mutations includes deep intronic
is not affected, a large multiexon structure called Bspliced mutations—usually substitutions localized within large in-
exon clusters in RNA intermediates^ (SECRI) consisting of trons that result in the inclusion of an intron fragment—so-
exons 5 and 6 is formed. Subsequent splicing events can lead called cryptic exon or pseudoexon, into the mature transcript.
to the removal of both exons 5 and 6 and the formation of the Functionally, such variants create novel acceptor or donor
major transcript. It can occur, however, that only intron 6 is sites that are recognized by the splicing complex and are used
removed from the SECRI and the transcript that contains in combination with the existing intronic cryptic splice sites. It
spliced exons 5, 6, and 7 can be a substrate for the splicing is also possible that deep intronic mutations result in the cre-
the used cryptic acceptor sites located in the exon 5. In the ation of novel regulatory elements (e.g., splicing enhancers)
case of the transcript lacking only exon 5, it was suggested and the recognition of the specific intronic sequences as an
that the first intron being removed is the intron 6 thus exonic ones (detailed review in Vaz-Drago et al. 2017).
preventing the removal of both exons 5 and 6 (Takahara One of the most common and well-known deep intronic
et al. 2002). change is a c.3718-2477C>T (legacy name: c. 3849+10 kb
The third mechanism that was proposed to explain the C>T) variant being one of the most frequent mutations in
skipping of several exons due to the splicing mutation was CFTR gene responsible for cystic fibrosis in Polish population
the Bsplicing paralysis^ model. It was described in detail for (Sobczyńska-Tomaszewska et al. 2013). This mutation is lo-
the c.1248+5G>A mutation in OXCT1 gene that was found in cated within the intron 19 and creates a novel donor site that
patient with succinyl-CoA:3-ketoacid CoA transferase defi- results in the inclusion of an 84-bp cryptic exon into the ma-
ciency (Hori et al. 2013). This mutation leads to the skipping ture mRNA. This cryptic exon contains an in-frame STOP
of exons 12 and 13 from the transcript and first studies aimed codon, and thus, the translated protein is shorter and nonfunc-
to identify appropriate SECRI intermediates, but without any tional (Sanz et al. 2017). The CF patients with c.3718-
success. Further functional analysis had shown that introns in 2477C>T mutation often have a relatively mild phenotype
the normal transcript were spliced in the following order: with a variable disease expression. It was found that the
J Appl Genetics (2018) 59:253–268 261

severity of the disease is inversely correlated with the level of fragment due to the nonspecific activity of the splicing com-
correctly spliced transcripts that suggest that the splicing reg- plex (Nissim-Rafinia and Kerem 2002). For example, the
ulation might be an important modifier of the CF clinical presence of c.3362A>G variant in the NF1 gene results in
course in the presence of intronic mutations. Functional stud- two mRNA isoforms: one properly spliced contains the sub-
ies had shown that the overexpression of the HTRA2-β1, stitution that can lead to missense change at the protein level
SC35 splicing factors in the presence of the c.3718- (p.Glu1121Gly), and the other one lacks exon 20. The tran-
2477C>T mutation promotes proper CFTR pre-mRNA splic- script mosaicism was confirmed with in vitro quantitative
ing and restored the function of the chloride channel (Nissim- studies in two patients harboring this mutation. One of them
Rafinia et al. 2004). had higher level of the transcript lacking exon 20, although it
Deep intronic mutations are not common, but their effect did not seem to correlate with the disease severity as both
on transcript splicing and further protein synthesis is signifi- patients presented similar phenotype. It was suggested that
cant. The analysis for the presence of such mutations should the level of different transcript isoforms might be related to
be considered when the identification of potentially pathogen- the individual genetic variability (Xu et al. 2014).
ic variants in the coding regions and exon/intron boundaries It is worth to mention in this section the process called
was not effective, and the patient phenotype is specific for a nonsense-associated altered splicing (NAS) that was de-
mutation is a specific gene. The techniques that can be helpful scribed as a mechanism that should protect transcripts contain-
in the identification of deep intronic mutations that include ing premature STOP codons from the nonsense-mediated de-
RNA/cDNA sequencing and genome or targeted whole gene cay (Cartegni et al. 2002). During this process, the transcripts
next-generation sequencing. However, the mutations identi- containing PTC are alternatively spliced and fragments with
fied on genomic level should be confirmed with functional premature nonsense codon are removed from mature RNA.
RNA testing. The exact mechanism of this process has not been described in
detail, but it is obvious that in certain circumstances, it is
Exonic mutations affecting splicing (type III and V) dependent from the frameshift of the reading frame and not
mutations from the ESE disruption (so-called class II-NAS). This pro-
cess was described in the context of the regulation of TCRβ
Apart from the intronic mutations affecting canonical splicing transcripts (Wang et al. 2002). More common are examples of
sites or activating cryptic exons, also the changes in the exonic class I-NAS—altered splicing due to the presence of the non-
sequences may affect the pattern of pre-mRNA splicing. Such sense mutation that affects ESE site and results in whole exon
exonic mutations might have double effect. First, they can skipping (Bühler and Mühlemann 2005). Such variants have
introduce a new 5′ or 3′ splice site or activate the cryptic one been described in the FBN1, BRCA1, or COL2A1 genes (see
that would be stronger than the original one, thus leading to Table 1).
changes in pre-mRNA processing and the loss of an exon
fragment (so-called type III splicing mutation). Second, the Mutations affecting branch point and polypyrimidine
presence of exonic changes that cause the disruption of exonic tract
splicing enhancers may also lead to the entire exon skipping
(so-called type V splicing mutation) (Wimmer et al. 2007). As mentioned, the splicing process is also dependent from the
The application of the RNA/cDNA sequencing in the di- presence of specific sequences: branch site and the
agnosis of genetic diseases helps to identify type III and V polypyrimidine tract sequences that bind specific proteins in-
splicing mutations. In genes that are analyzed with this ap- volved in the formation of splicing complexes. The branch
proach, many different variants were identified and shown to point motif, localized between − 9 and − 400 bp downstream
influence proper pre-mRNA processing. According to LOVD from the acceptor site with the consensus sequence YUNAY
mutation database (accessed on 10.10.2017), 26 exonic splic- in humans, is essential for early spliceosome complex forma-
ing mutations were identified in the NF1 gene. Eighteen of tion. As the sequences of the branch point are highly
them (69%) affected regulatory splicing sequences, especially degenerated, their exact localization is difficult to determine.
ESE motifs, resulting in the specific exon skipping. The other It seems, however, that mutations localized in the branch point
mutations (8.31%) created a new cryptic splice site thus lead- sequence might lead to an exon skipping due to improper
ing to the deletion of the exon fragment. binding of the SF1 and U2 snRNP splicing proteins and dis-
The exonic mutation causing splicing alterations can be ruption of the natural acceptor splicing site. Mutations in
easily misclassified as synonymous, missense, or nonsense branch point sequence can also cause intron retention (whole
variant. Usually, the presence of such variants results in the or its fragment) if they create new 3′ splice site (Caminsky
generation of two different transcripts from one mutated al- et al. 2014).
lele: one has a proper length and has a modified nucleotide, The polypyrimidine tract with sequence enriched in pyrim-
and the other one is shorter and lack whole exon or its idine nucleotides /(Y)12–17/ is located between 5 and 40 bp
262 J Appl Genetics (2018) 59:253–268

from the acceptor splice site, upstream from the branch learning approaches. Most of the tools focus on the analysis
point sequence. This sequence binds the U2AF65 of the consensus splicing sites and require the sequence input
spliceosome subunit and polypyrimidine tract-binding pro- at least including positions − 3_ + 6 or − 20_ + 3 in the case of
tein that is involved in alternative splicing regulation. Any 5′ donor and 3′ acceptor sites, respectively. These tools are
mutations in this sequence probably lead to splicing alter- based on position weight matrix model (Analyzer Splice
ations, although the list of such variants is limited (Ward Tool, SpliceView), probabilistic maximum dependence de-
and Cooper 2010). composition model (GENSCAN, GeneSplicer), machine
Point mutations at the branch point and polypyrimidine learning techniques (NetGene2, NNSplice—artificial neural
tract are very rare and hard to identify when the genomic networks, SplicePort—support vector machine,
DNA, mainly coding sequences, are analyzed. As the consen- SplicePredictor—no web interface, Bayesian model), or max-
sus sequences of these motifs are degenerated, it is hard to imum entropy distribution model (MaxEntScan). There are
predict their exact localization and therefore conclude about also tools developed to detect how distant mutation may in-
the possible effect of a specific variant in these regions. Many fluence splicing (Spliceman) or to predict exon skipping,
of mutations affecting branch point or polypyrimidine tract cryptic site activation, or generation of aberrant transcripts
described so far have been identified by the RNA/cDNA se- from primary sequence (CRYP-SKIP logistic regression mod-
quencing, or their effect was assessed in functional studies el). To predict whether a single nucleotide variant can affect
with minigene assay (see BFunctional analysis of splicing branch site sequence or polypirymidine tract, specific algo-
mutation^ section). Such analysis should be performed to con- rithms were developed to identify these sites (e.g., Branch
firm the splicing effect of an intronic variant identified near the Site Analyzer, SVM-BPfinder) or to predict variant pathoge-
acceptor splice site (Lewandowska 2013). nicity (e.g., IntSplice—support only vcf files; Jian et al. 2014;
Ohno et al. 2018).
Also, the effect of a specific variant on the possible ESE or
How to detect and confirm splicing mutation? ESS alterations should be examined in the case of exonic
mutations. There are several algorithms that can be used to
About 9% of all mutations reported in the Human Gene assess these changes like ESE Finder, based on functional
Mutation Database (HGMD) are splicing mutations (18761/ SELEX method or RESCUE-ESE Web Server, RESCUE-
208368) (HGMD database, accessed on October 10, 2017), based MODEL HEXplorer score, and ESRsearch, all three
although it is obvious th at this number may be based on relative enhancer and silencer classification by unan-
underestimated. Most of the reported mutations were identi- imous enrichment approach using frequencies of hexameric
fied by genomic DNA sequencing, and it cannot be ruled out sequences. Some models are based on the results of the func-
that the number of missense or nonsense substitutions is in tional analyses of random sequences for enhancer or silencer
fact splicing mutations especially when occur within cis-act- properties with minigene assays (FAS-ESS) or the direct in-
ing elements. Only in several monogenic disorders (e.g., neu- teraction between selected splicing factors and their RNA tar-
rofibromatosis type I), the RNA sequencing was implemented get motifs (SpliceAid2). Only SPANR tool uses splicing code
into routine molecular diagnosis, mainly due to the problems modeling-based approach and was trained on data from dif-
with RNA stability and availability. Nevertheless, the applica- ferent human tissues. The EX-SKIP and HOT-SKIP tools in-
tion of this method allowed to identify synonymous (silent) tegrate several approaches to analyze potential ESE/ESS se-
and nonsynonymous/nonsense mutations as well as deep quences (Grodecká et al. 2017). Other bioinformatic pro-
intronic mutations that affected pre-mRNA splicing. This grams, such as mFold or pFold, can be used to predict whether
underlined the fact that the analysis of the DNA sequences a given mutation may affect mRNA secondary structure
(exome or gene) should also include the identification of pos- (Caminsky et al. 2014).
sible splicing mutations. As the functional testing is challeng- For the maximum comfort of the user, some programs that
ing, the in silico algorithms were developed to test for possible use different algorithms were developed and made available
splicing alterations (Table 2). via the website. The most known are Human Splicing Finder
(HSF) and SROOGLE, that predict the presence of cis-
In silico analysis of potential slicing mutation splicing elements in the uploaded sequence or generate pre-
diction for a particular variant in a specific gene. Another SNV
These tools were developed for research purposes, although in prediction online tool is MutPredSplice that can analyze single
certain circumstances can be implemented into routine diag- variant or a set of variants uploaded as a vcf file. Multiple
nostics. Algorithms proposed for splicing analysis differ be- algorithms were also implemented in a commercially avail-
tween each other with the database containing information able software like Alamut Visual Software (Interactive
about the consensus sequences, the statistical model used for Biosoftware) or MutationForecaster (includes a tool previous-
the analysis, or training methods that are used in machine ly known as ASSEDA; CytoGnomix® Inc).
J Appl Genetics (2018) 59:253–268 263

Table 2 Prediction algorithms for the analysis of splicing effect

Prediction of splice sites

NetGene2 http://www.cbs.dtu.dk/services/NetGene2/ Hebsgaard et al. (1996), Brunak et al. (1991)

Splice Site Prediction by www.fruitfly.org/seq_tools/splice.html Reese et al. (1997)
Neural Network
SplicePredictor http://brendelgroup.org/bioinformatics2go/SplicePredictor.php Brendel et al. (2004)
Splice port http://spliceport.cbcb.umd.edu/ Dogan et al. (2007)
SpliceView http://bioinfo.itb.cnr.it/oriel/splice-view.html Shapiro and Senapathy (1987),
Rogozin and Milanesi (1997)
Analyzer Splice Tool http://ibis.tau.ac.il/ssat/SpliceSiteFrame.htm Carmel et al. (2004)
GENSCAN http://genes.mit.edu/GENSCAN.html Burge and Karlin (1997)
GeneSplicer http://www.cbcb.umd.edu/software/GeneSplicer/gene_spl.shtml Pertea et al. (2001)
MaxEntScan http://genes.mit.edu/burgelab/maxent/Xmaxentscan_scoreseq.html Yeo and Burge (2004)
Spliceman http://fairbrother.biomed.brown.edu/spliceman/ Lim et al. (2011), Lim and Fairbrother (2012)
CRYP-SKIP http://cryp-skip.img.cas.cz/ Divina et al. (2009)
SROOGLE http://sroogle.tau.ac.il/ Schwartz et al. (2009)
Human Splicing Finder www.umd.be/HSF/ Desmet et al. (2009)
MutPredSplice http://www.mutdb.org/mutpredsplice/submit.htm Mort et al. (2014)
Alamut Visual Software http://www.interactive-biosoftware.com Interactive Biosoftware
MutationForecaster https://mutationforecaster.com/index.php CytoGnomix® Inc.
Prediction of branch site and polypyrimidine tract
Branch Site Analyzer http://ibis.tau.ac.il/ssat/BranchSite.htm Kol et al. (2005)
SVM-BPfinder http://regulatorygenomics.upf.edu/Software/SVM_BP/ Corvelo et al. (2010)
IntSplice https://www.med.nagoya-u.ac.jp/neurogenetics/IntSplice/ Shibata et al. (2016)
Variant annotations
Variant Effect Predictor tool https://www.ensembl.org/info/docs/tools/vep/index.html Ensembl release 91—December 2017©
EMBL-EBI
Alamute Batch software http://www.interactive-biosoftware.com/alamut-batch/
Prediction of ESE or ESS
ESEfinder http://exon.cshl.org/ESE Cartegni et al. (2003), Smith et al. (2006)
RESCUE-ESE programs http://genes.mit.edu/burgelab/rescue-ese/ Fairbrother et al. (2002)
HEXplorer score https://www2.hhu.de/rna/html/hexplorer_score.php Erkelenz et al. (2014)
ESRsearch http://esrsearch.tau.ac.il/ Goren et al. (2006), Fairbrother et al. (2002),
Zhang and Chasin (2004)
FAS-ESS http://genes.mit.edu/fas-ess/ Wang et al. (2004)
SpliceAid2 http://193.206.120.249/splicing_tissue.html Piva et al. (2012)
SPANR tool http://tools.genes.toronto.edu/ Xiong et al. (2015)
EX-SKIP http://ex-skip.img.cas.cz/ Raponi et al. (2011)
HOT-SKIP http://hot-skip.img.cas.cz/ Raponi et al. (2011)
Prediction of mRNA secondary structure
mFold http://unafold.rna.albany.edu/?q=mfold Zuker (2003)
pFold http://daimi.au.dk/~compbio/pfold/ Knudsen and Hein (2003)

Also, advanced tools used for variant annotations (e.g., software for variant annotation, like Alamut Batch soft-
from next generation sequencing) can utilize splicing predic- ware (Interactive Biosoftware), also includes analysis of
tion algorithms. For example, the Variant Effect Predictor tool, splicing mutation. The Alamut Batch uses MaxEntScan,
that is also available online, includes special plugins that per- HSF, NNSplice, GeneSplicer, and other tools for splicing
form splicing analysis with MaxEntScan model and dbscSNV analysis (user manual available at http://www.interactive-
matrix (a part of the dbNSFP database). The commercial biosoftware.com).
264 J Appl Genetics (2018) 59:253–268

Functional analysis of splicing mutation and resulting in shorter proteins. The test consists of four
steps: (1) generation of a template for protein synthesis by
The bioinformatic algorithms are a useful tool in the assessing amplification of the DNA fragment or RT-PCR of the
the possible effect of the identified changes, although it should mRNA, (2) in vitro template transcription and translation in
be underlined that the result of such test is only predictive, and the presence of labeled methionine or leucine amino acids, (3)
the exact effect of the mutation should be verified in functional electrophoresis (SDS-PAGE) of the synthetized proteins, and
studies. An additional test that can be used to confirm that the (4) visualization and assessment of the protein size. When the
specific splicing variant has a pathogenic effect is the analysis splicing, nonsense, or frameshift mutation that results in PTC
of a variant segregation with the disease in affected and unaf- is present, the shorter than expected protein is synthetized.
fected family members performed at the DNA level, but still Using this method, the protein abundance and its size are easy
the exact splicing effect should be tested in laboratory (Théry to assess (Hauss and Müller 2007). Nowadays, this method
et al. 2011; Fredericks et al. 2015). There are several methods was replaced by Sanger or next-generation sequencing, but it
that can be used to analyze the functional effect of a particular remains an option to test for possible splicing defects resulting
variant. The effect of the splicing mutations can be tested on in shortened protein.
different levels: DNA, RNA, RNA-protein interactions, or
protein level itself.
The simplest and the most effective method to determine Summary
whether the selected variant affects splicing is to analyze the
RNA extracted from a relevant patient tissue or cell line de- It is clear that mutations affecting splicing pattern may be the
rived from a patient cells. The sequencing of RNA/cDNA cause of genetic disorders, although their frequency might be
after reverse transcription PCR (RT-PCR) allows to verify underestimated. During the genomic DNA analysis, they can
whether the identified variant influence the mRNA sequence. be easily overlooked and erroneously classified as synony-
The main problem with such approach is the possibility of mous changes or benign amino acid substitutions. However,
nonsense-mediated decay. In such situation, the effect of the the analysis of RNA/cDNA clearly shows that such mutations
potential splicing mutation can be easily missed out. To over- have a significant impact on the pre-mRNA splicing. It was
come this disadvantage, the patient cells can be treated with assumed that larger genes with long introns were more prone
NMD inhibitors such as puromycin that blocks RNA degra- to splicing defects, but it is now more obvious that also a
dation (Baralle and Baralle 2005). The most common mate- significant number of mutations in smaller genes also cause
rials used for such functional testing are patient fibroblasts or abnormal mRNA splicing (Chen and Manley 2009). In addi-
short-term cultures of blood mononuclear cells (leukocytes). tion, the analysis of NF1 and ATM genes showed that many of
If the appropriate material for functional RNA sequenc- the identified splicing mutations were located outside the ca-
ing is not available, an alternative possibility is a minigene nonical splicing sites and could be easily missed during geno-
assay—an in vitro hybrid system that allows for Bexon mic DNA analysis. There is a growing evidence that misclas-
trapping.^ This system is especially useful to the analysis of sification of mutation is a common error and the overall num-
genes with low expression level in leukocytes or fibroblasts ber of splicing defects is probably underestimated (Xiong
(Singh and Cooper 2006). In minigene assay, the amplified et al. 2015).
fragment of the analyzed gene, e.g., specific exon with sur- The importance of splicing mutations in the pathogenesis
rounding intronic sequences with and without mutations, is of genetic diseases gave rise to numerous experimental and
cloned into a special expression plasmid enabling the analysis clinical studies that focus on the development of drugs that
the pre-mRNA splicing (Fig. 4). This approach can be used to can reverse the effect of splicing mutations. The most prom-
confirm that the potential splicing variant affects splicing ef- ising approach is the use of antisense oligonucleotides
ficiency or causes the activation of the alternative cryptic (AONs; Pros et al. 2009; Bergsma et al. 2018). The AONs
splicing sites, and to test the role of a cis-acting elements on are short synthetic DNA or RNA molecules that recognize
splicing regulation (Sharma et al. 2014). complementary target pre-mRNA fragments and can modu-
These techniques are limited because they do not assess the late its splicing by blocking of binding sites for proteins in-
relative level of transcript isoforms. The most valuable meth- volved in the regulation of the splicing process (Siva et al.
od to resolve this problem is real-time PCR (RT-PCR) or 2014). Synthetic AONs can be used to prevent aberrant splic-
quantitative PCR (qPCR) which enables to measure the quan- ing or to induce additional exon exclusion to restore the proper
tity of each mutated transcript and compare it to the level of reading frame. The first approach is used in the therapy of
the nonmutated one. This method also allows to test whether spinal muscular atrophy (SMA), autosomal recessive disorder
the mutated transcript undergo NMD (Xu et al. 2014). in 95% of cases caused by homozygous deletion of exon 7 of
The protein truncation test (PTT) was designed as a screen- the SMN1 gene (Jedrzejowska et al. 2010). In the 5q13.2
ing tool to identify mutations that lead to the formation of PTC locus, there is a paralogue of the SMN1 gene—the SMN2 gene
J Appl Genetics (2018) 59:253–268 265

Fig. 4 The principles of minigene

assay. In the minigene assay, the
amplified fragment of the
analyzed gene, e.g., specific exon
with surrounding intronic
sequences with and without
mutations, is cloned into a special
expression plasmid enabling the
analysis the pre-mRNA splicing
(see description in text)

that differs with five nucleotides from the original one. It en- Acknowledgments We would like to thank Prof. Jerzy Bal for the critical
review of the article and his insightful comments about the manuscript.
codes the identical protein, although the SMN2 transcript pro-
cessing is ineffective due to the presence of c.840C>T substi-
Funding The studies on neurofibromatosis type I and other RASopathies
tution in the exon 7. Its presence results in exon 7 skipping in are possible thanks to the Institute of Mother and Child Intramural Grants
about 80% of SMN2 transcripts, and a truncated, quickly de- and Polish National Science Centre grant UMO-2013/09/B/NZ2/03164.
graded nonfunctional protein is synthetized. Two models have
been proposed to explain exon 7 skipping. First assumes that Compliance with ethical standards
the presence of c.830C>T substitution causes ESE disruption
and inhibits the binding of SF2/ASF splicing activators Conflict of interest The authors declare that they have no conflict of
interest.
(Cartegni and Krainer 2002). The second one proposes that
the substitution leads to the formation of the exonic splicing
Ethical approval This article does not contain any studies with human
silencer that binds hnRNPA1 and represses exon 7 inclusion participants or animals performed by any of the authors.
to the transcript (Kashima et al. 2007; Cartegni et al. 2006).
Nevertheless, the Spiranza—an ASO drug that was developed Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://
for SMA patients—blocks splicing silencers near the exon 7 creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
of the SMN2 gene thus generating properly spliced (exon 7 distribution, and reproduction in any medium, provided you give appro-
included) transcripts that encode functional SMN protein. The priate credit to the original author(s) and the source, provide a link to the
clinical testing has shown high efficacy of Spiranza treatment Creative Commons license, and indicate if changes were made.
in patients with all types of SMA (Chiriboga 2017).
The approach that induces exon skipping to restore reading
frame was developed for Duchenne muscular dystrophy pa-
tients with frameshift mutations in the DMD gene. In these
patients, it was proved that large in-frame deletions are less References
deleterious than mutations causing loss of reading frame, and
this triggered the idea of the restoration of a reading frame Aoyama Y, Sasai H, Abdelkreem E, Otsuka H, Nakama M, Kumar S,
Fukao T (2017) A novel mutation (c.121-13T>A) in the
with ASO. First studies focused on the exon 51 skipping, as polypyrimidine tract of the splice acceptor site of intron 2 causes
the deletion of this exon is frequently seen in DMD patients, exon 3 skipping in mitochondrial acetoacetyl-CoA thiolase gene.
and the developed drugs (drisapersen and eteplirsen) were Mol Med Rep 15:3879–3884
successfully used in clinical trials (Veltrop and Aartsma-Rus Axelrod FB, Liebes L, Gold-von Simson G et al (2011) Kinetin improves
IKBKAP mRNA splicing in patients with familial dysautonomia.
2014). The promising results of the ASO-based therapies
Pediatr Res 70(5):480–483
demonstrate that the treatment of genetic disorders caused Baralle D, Baralle M (2005) Splicing in action: assessing disease causing
by splicing defects is possible and this points out the impor- sequence changes. J Med Genet 42(10):737–748
tance of the identification of such mutations in clinical Bergsma AJ, van der Wal E, Broeders M, van der Ploeg AT, Pim
samples. Pijnappel WWM (2018) Alternative splicing in genetic diseases:
266 J Appl Genetics (2018) 59:253–268

improved diagnosis and novel treatment options. Int Rev Cell Mol Faustino NA, Cooper TA (2003) Pre-mRNA splicing and human disease.
Biol 335:85–141 Genes Dev 17(4):419–437
Brendel V, Xing L, Zhu W (2004) Gene structure prediction from con- Fredericks AM, Cygan KJ, Brown BA, Fairbrother WG (2015) RNA-
sensus spliced alignment of multiple ESTs matching the same geno- binding proteins: splicing factors and disease. Biomol Ther 5(2):
mic locus. Bioinformatics 20:1157–1169 893–909
Brunak S, Engelbrech J, Knudsen S (1991) Prediction of human mRNA Geuens T, Bouhy D, Timmerman V (2016) The hnRNP family: insights
donor and acceptor sites from the DNA sequence. J Mol Biol 220: into their role in health and disease. Hum Genet 135(8):851–867
49–65 Glisovic T, Bachorik JL, Yong J, Dreyfuss G (2008) RNA-binding pro-
Bühler M, Mühlemann O (2005) Alternative splicing induced by non- teins and post-transcriptional gene regulation. FEBS Lett 582(14):
sense mutations in the immunoglobulin μ VDJ exon is independent 1977–1986
of truncation of the open reading frame. RNA 11(2):139–146 Goren A, Ram O, Amit M, Keren H, Lev-Maor G, Vig I, Pupko T, Ast G
Burge C, Karlin S (1997) Prediction of complete gene structures in hu- (2006) Comparative analysis identifies exonic splicing regulatory
man genomic DNA. J Mol Biol 268:78–94 sequences—the complex definition of enhancers and silencers.
Caceres JF, Kornblihtt AR (2002) Alternative splicing: multiple control Mol Cell 22(6):769–781
mechanisms and involvement in human disease. Trends Genet Grodecká L, Buratti E, Freiberger T (2017) Mutations of pre-mRNA
18(4):186–193 splicing regulatory elements: are predictions moving forward to
Caminsky N, Mucaki EJ, Rogan PK (2014) Interpretation of mRNA clinical diagnostics? Int J Mol Sci 18(8):1668
splicing mutations in genetic disease: review of the literature and Habara Y, Takeshima Y, Awano H, Okizuka Y, Zhang Z, Saiki K, Yagi M,
guidelines for information-theoretical analysis. F1000Res 3:282 Matsuo M (2009) In vitro splicing analysis showed that availability
Carmel I, Tal S, Vig I, Ast G (2004) Comparative analysis detects depen- of a cryptic splice site is not a determinant for alternative splicing
dencies among the 5′ splice-site positions. RNA 10(5):828–840 patterns caused by +1G–>A mutations in introns of the dystrophin
Cartegni L, Krainer AR (2002) Disruption of an SF2/ASF-dependent gene. J Med Genet 46(8):542–547
exonic splicing enhancer in SMN2 causes spinal muscular atrophy Han SP, Tang YH, Smith R (2010) Functional diversity of the hnRNPs:
in the absence of SMN1. Nat Genet 30(4):377–384 past, present and perspectives. Biochem J 430(3):379–392
Cartegni L, Chew SL, Krainer AR (2002) Listening to silence and under- Hauss O, Müller O (2007) The protein truncation test in mutation detec-
standing nonsense: exonic mutations that affect splicing. Nat Rev
tion and molecular diagnosis. Methods Mol Biol 375:151–64
Genet 3(4):285–298
Hebsgaard SM, Korning PG, Tolstrup N, Engelbrecht J, Rouze P, Brunak
Cartegni L, Wang J, Zhu Z, Zhang MQ, Krainer AR (2003) ESEfinder: a
S (1996) Splice site prediction in Arabidopsis thaliana DNA by
web resource to identify exonic splicing enhancers. Nucleic Acid
combining local and global sequence information. Nucleic Acids
Res 31(13):3568–3571
Res 24(17):3439–3452
Cartegni L, Hastings ML, Calarco JA, de Stanchina E, Krainer AR (2006)
Hori T, Fukao T, Murase K, Sakaguchi N, Harding CO, Kondo N (2013)
Determinants of exon 7 splicing in the spinal muscular atrophy
Molecular basis of two-exon skipping (exons 12 and 13) by c.1248+
genes, SMN1 and SMN2. Am J Hum Genet 78(1):63–77
5g>a in OXCT1 gene: study on intermediates of OXCT1 transcripts
Chabot B, Shkreta L (2016) Defective control of pre-messenger RNA
in fibroblasts. Hum Mutat 34(3):473–480
splicing in human disease. J Cell Biol 212(1):13–27
Ibrahim EC, Hims MM, Shomron N, Burge CB, Slaugenhaupt SA, Reed
Chen M, Manley JL (2009) Mechanisms of alternative splicing regula-
tion: insights from molecular and genomics approaches. Nat Rev R (2007) Weak definition of IKBKAP exon 20 leads to aberrant
Mol Cell Biol 10(11):741–754 splicing in familial dysautonomia. Hum Mutat 28:41–53
Chiriboga CA (2017) Nusinersen for the treatment of spinal muscular Jedrzejowska M, Milewski M, Zimowski J, Zagozdzon P, Kostera-
atrophy. Expert Rev Neurother 17(10):955–962 Pruszczyk A, Borkowska J, Sielska D, Jurek M, Hausmanowa-
Corvelo A, Hallegger M, Smith CWJ, Eyras E (2010) Genome-wide Petrusewicz I (2010) Incidence of spinal muscular atrophy in
association between branch point properties and alternative splicing. Poland—more frequent than predicted? Neuroepidemiology 34(3):
Meyer IM, ed. PLoS Comput Biol 6(11):e1001016 152–157
Cuajungco MP, Leyne M, Mull J, Gill SP, Lu W, Zagzag D et al (2003) Jian X, Boerwinkle E, Liu X (2014) In silico prediction of splice-altering
Tissue-specific reduction in splicing efficiency of IKBKAP due to single nucleotide variants in the human genome. Nucleic Acids Res
the major mutation associated with familial dysautonomia. Am J 42(22):13534–13544
Hum Genet 72:749–758 Känsäkoski J, Jääskeläinen J, Jääskeläinen T et al (2016) Complete an-
Desmet FO, Hamroun D, Lalande M, Collod-Béroud G, Claustres M, drogen insensitivity syndrome caused by a deep intronic
Béroud C (2009) Human Splicing Finder: an online bioinformatics pseudoexon-activating mutation in the androgen receptor gene. Sci
tool to predict splicing signals. Nucleic Acids Res 37(9):e67 Rep 6:32819
Divina P, Kvitkovicova A, Buratti E, Vorechovsky I (2009) Ab initio Kashima T, Rao N, David CJ, Manley JL (2007) hnRNP A1 functions
prediction of mutation-induced cryptic splice-site activation and ex- with specificity in repression of SMN2 exon 7 splicing. Hum Mol
on skipping. Eur J Hum Genet 17(6):759–765 Genet 16(24):3149–3159
Dogan RI, Getoor L, Wilbur WJ, Mount SM (2007) SplicePort—an in- Khan SG, Yamanegi K, Zheng Z-M et al (2010) XPC branch-point se-
teractive splice-site analysis tool. Nucleic Acids Res 35(Web Server quence mutations disrupt U2 snRNP binding resulting in abnormal
issue):W285–W291 pre-mRNA splicing in xeroderma pigmentosum patients. Hum
Erkelenz S, Theiss S, Otte M, Widera M, Peter JO, Schaal H (2014) Mutat 31(2):167–175
Genomic HEXploring allows landscaping of novel potential splic- Knudsen B, Hein J (2003) Pfold: RNA secondary structure prediction
ing regulatory elements. Nucleic Acids Res 42(16):10681–10697 using stochastic context-free grammars. Nucleic Acids Res 31(13):
Fairbrother WG, Yeh RF, Sharp PA, Burge CB (2002) Predictive identi- 3423–3428
fication of exonic splicing enhancers in human genes. Science Kol G, Lev-Maor G, Ast G (2005) Human-mouse comparative analysis
297(5583):1007–1013 reveals thatbranch-site plasticity contributes to splicing regulation.
Fang LJ, Simard MJ, Vidaud D, Assouline B, Lemieux B, Vidaud M, Hum Mol Genet 14(11):1559–1568
Chabot B, Thirion JP (2001) A novel mutation in the neurofibroma- Kolovos P, Knoch TA, Grosveld FG, Cook PR, Papantonis A (2012)
tosis type 1 (NF1) gene promotes skipping of two exons by Enhancers and silencers: an integrated and simple model for their
preventing exon definition. J Mol Biol 307(5):1261–1270 function. Epigenetics Chromatin 5(1):1
J Appl Genetics (2018) 59:253–268 267

Krawczak M, Thomas NS, Hundrieser B, Mort M, Wittig M, Hampe J, Raponi M, Kralovicova J, Copson E, Divina P, Eccles D, Johnson P,
Cooper DN (2007) Single base-pair substitutions in exon-intron Baralle D, Vorechovsky I (2011) Prediction of single-nucleotide
junctions of human genes: nature, distribution, and consequences substitutions that result in exon skipping: identification of a splicing
for mRNA splicing. Hum Mutat 28(2):150–158 silencer in BRCA1 exon 6. Hum Mutat 32:436–444
Lee Y, Rio DC (2015) Mechanisms and regulation of alternative pre- Reese MG, Eeckman FH, Kulp D, Haussler D (1997) Improved splice
mRNA splicing. Annu Rev Biochem 84:291–323 site detection in genie. J Comp Biol 4(3):311–323
Lewandowska M (2013) The missing puzzle piece: splicing mutations. Rogozin IB, Milanesi L (1997) Analysis of donor splice sites in different
Int J Clin Exp Pathol 6(12):2675–2682 eukaryotic organisms. J Mol Evol 45(1):50–59
Lim KH, Fairbrother WG (2012) Spliceman—a computational web serv- Sanz DJ, Hollywood JA, Scallan MF, Harrison PT (2017) Cas9/gRNA
er that predicts sequence variations in pre-mRNA splicing. targeted excision of cystic fibrosis-causing deep-intronic splicing
Bioinformatics 28(7):1031–1032 mutations restores normal splicing of CFTR mRNA. PLoS One
Lim KH, Ferraris L, Filloux ME, Raphael BJ, Fairbrother WG (2011) 12(9):e0184009
Using positional distribution to identify splicing elements and pre- Schwartz S, Hall E, Ast G (2009) SROOGLE: webserver for integrative,
dict pre-mRNA processing defects in human genes. Proc Natl Acad user-friendly visualization of splicing signals. Nucleic Acids Res
Sci U S A 108(27):11093–11098 37(Web Server issue):W189–W192
McAlinden A, Majava M, Bishop PN, Perveen R, Black GC, Pierpont Shapiro MB, Senapathy P (1987) RNA splice junctions of different clas-
ME, Ala-Kokko L, Männikkö M (2008) Missense and nonsense ses of eukaryotes: sequence statistics and functional implications in
mutations in the alternatively-spliced exon 2 of COL2A1 cause the gene expression. Nucleic Acids Res 15(17):7155–7174
ocular variant of Stickler syndrome. Hum Mutat 29(1):83–90 Sharma N, Sosnay PR, Ramalho AS et al (2014) Experimental assess-
Messiaen LM, Wimmer K (2008) NF1 mutational spectrum. In: ment of splicing variants using expression minigenes and compari-
Kaufmann D (ed) Neurofibromatoses, vol 16. Karger, Basel, pp son with in silico predictions. Hum Mutat 35(10):1249–1259
63–77 Shibata A, Okuno T, Rahman MA, Azuma Y, Takeda J, Masuda A,
Mort M, Sterne-Weiler T, Li B et al (2014) MutPred Splice: machine Selcen D, Engel AG, Ohno K (2016) IntSplice: prediction of the
learning-based prediction of exonic variants that disrupt splicing. splicing consequences of intronic single-nucleotide variations in the
Genome Biology 15(1): R19. Nissim-Rafinia M, Aviram M, human genome. J Hum Genet 61(7):633–640
Randell SH, et al (2004) Restoration of the cystic fibrosis transmem- Singh G, Cooper TA (2006) Minigene reporter for identification and
brane conductance regulator function by splicing modulation. analysis of cis elements and trans factors affecting pre-mRNA splic-
EMBO Rep 5(11):1071–1077 ing. BioTechniques 41(2):177–181
Nissim-Rafinia M, Kerem B (2002) Splicing regulation as a potential Siva K, Covello G, Denti MA (2014) Exon-skipping antisense oligonu-
genetic modifier. Trends Genet 18(3):123–127 cleotides to correct missplicing in neurogenetic diseases. Nucleic
Nissim-Rafinia M, Aviram M, Randell SH, Shushi L, Ozeri E, Chiba- Acid Ther 24(1):69–86
Falek O, Eidelman O, Pollard HB, Yankaskas JR, Kerem B (2004) Slaugenhaupt SA, Blumenfeld A, Gill SP, Leyne M, Mull J, Cuajungco
Restoration of the cystic fibrosis transmembrane conductance regu- MP, Liebert CB, Chadwick B, Idelson M, Reznik L, Robbins C,
lator function by splicing modulation. EMBO Rep 5(11):1071–1077 Makalowska I, Brownstein M, Krappmann D, Scheidereit C,
Ohno K, Takeda JI, Masuda A (2018) Rules and tools to predict the Maayan C, Axelrod FB, Gusella JF (2001) Tissue-specific expres-
splicing effects of exonic and intronic mutations. WIREs RNA 9: sion of a splicing mutation in the IKBKAP gene causes familial
e1451. https://doi.org/10.1002/wrna.1451 dysautonomia. Am J Hum Genet 68:598–605
Palhais B, Dembic M, Sabaratnam R, Nielsen KS, Doktor TK, Bruun Smith PJ, Zhang C, Wang J, Chew SL, Zhang MQ, Krainer AR (2006)
GH, Andresen BS (2016) The prevalent deep intronic c. 639+919 An increased specificity score matrix for the prediction of SF2/ASF-
G>A GLA mutation causes pseudoexon activation and Fabry dis- specific exonic splicing enhancers. Hum Mol Genet 15(16):2490–
ease by abolishing the binding of hnRNPA1 and hnRNP A2/B1to a 2508
splicing silencer. Mol Genet Metab 119(3):258–269 Sobczyńska-Tomaszewska A, Ołtarzewski M, Czerska K, Wertheim-
Parada GE, Munita R, Cerda CA, Gysling K (2014) A comprehensive Tysarowska K, Sands D, Walkowiak J, Bal J, Mazurczak T, NBS
survey of non-canonical splice sites in the human transcriptome. CF working group (2013) Newborn screening for cystic fibrosis:
Nucleic Acids Res 42(16):10564–10578 Polish 4 years’ experience with CFTR sequencing strategy. Eur J
Pertea M, Lin X, Salzberg SL (2001) GeneSplicer: a new computational Hum Genet 21(4):391–396
method for splice site prediction. Nucleic Acids Res 29(5):1185– Sterne-Weiler T, Sanford JR (2014) Exon identity crisis: disease-causing
1190 mutations that disrupt the splicing code. Genome Biol 15(1):201
Piva F, Giulietti M, Burini AB, Principato G (2012) SpliceAid 2: a data- Svaasand EK, Engebretsen LF, Ludvigsen T, Brechan W, Sjursen W
base of human splicing factors expression data and RNA target (2015) A novel deep intronic mutation introducing a cryptic exon
motifs. Hum Mutat 33(1):81–85 causing neurofibromatosis type 1 in a family with highly variable
Pros E, Gómez C, Martín T, Fábregas P, Serra E, Lázaro C (2008) Nature phenotypes: a case study. Hereditary Genet 4:152
and mRNA effect of 282 different NF1 point mutations: focus on Symoens S, Malfait F, Vlummens P, Hermanns-Lê T, Syx D, De Paepe A
splicing alterations. Hum Mutat 29(9):E173–E193 (2011) A novel splice variant in the N-propeptide of COL5A1
Pros E, Fernández-Rodrígue J, Canet B, Benit L, Sánchez A, Benavides causes an EDS phenotype with severe kyphoscoliosis and eye in-
A, Ramos FJ, López-Ariztegui MA, Capellá G, Blanco I, Serra E, volvement. PLoS One 6(5):e20121
Lázaro C (2009) Antisense therapeutics for neurofibromatosis type Takahara K, Schwarze U, Imamura Y, Hoffman GG, Toriello H et al
1 caused by deep intronic mutations. Hum Mutat 30:454–462 (2002) Order of intron removal influences multiple splice outcomes,
Rahman M, Nasrin F, Masuda A, Kinji O (2015) Decoding abnormal including a two-exon skip, in a COL5A1 acceptor-site mutation that
splicing code in human diseases. J Investig Genomics 2(1):00016 results in abnormal pro-alpha1(V) N-propeptides and Ehlers-Danlos
Ramalho AS, Beck S, Penque D, Gonska T, Seydewitz HH, Mall M, syndrome type I. Am J Hum Genet 71:451–465
Amaral MD (2003) Transcript analysis of the cystic fibrosis splicing Tazi J, Bakkour N, Stamm S (2009) Alternative splicing and disease.
mutation 1525-1G>A shows use of multiple alternative splicing Biochim Biophys Acta 1792(1):14–26
sites and suggests a putative role of exonic splicing enhancers. J Théry JC, Krieger S, Gaildrat P et al (2011) Contribution of bioinformat-
Med Genet 40(7):e88 ics predictions and functional splicing assays to the interpretation of
268 J Appl Genetics (2018) 59:253–268

unclassified variants of the BRCA genes. Eur J Hum Genet 19(10): Xiong F, Gao J, Li J et al (2009) Noncanonical and canonical splice sites:
1052–1058 a novel mutation at the rare noncanonical splice-donor cut site
Turunen JJ, Niemelä EH, Verma B, Frilander MJ (2013) The significant (IVS4+1A>G) of SEDL causes variable splicing isoforms in X-
other: splicing by the minor spliceosome. Wiley Interdisc Rev RNA linked spondyloepiphyseal dysplasia tarda. Eur J Hum Genet
4(1):61–76 17(4):510–516
Van de Water NS, Tan T, May S, Browett PJ, Harper P (2004) Factor IX Xiong HY, Alipanahi B, Lee LJ et al (2015) The human splicing code
polypyrimidine tract mutation analysis using mRNA from peripher- reveals new insights into the genetic determinants of disease.
al blood leukocytes. J Thromb Haemost 2:2073–2075 Science (New York, NY) 347(6218):1254806
Vaz-Drago R, Custódio N, Carmo-Fonseca M (2017) Deep intronic mu- Xu W, Yang X, Hu X, Li S (2014) Fifty-four novel mutations in the NF1
tations and human disease. Hum Genet 136(9):1093–1111 gene and integrated analyses of the mutations that modulate splicing.
Veltrop M, Aartsma-Rus A (2014) Antisense-mediated exon skipping: Int J Mol Med 34(1):53–60
taking advantage of a trick from mother nature to treat rare genetic Yeo G, Burge CB (2004) Maximum entropy modeling of short sequence
diseases. Exp Cell Res 325(1):50–55 motifs with applications to RNA splicing signals. J Comput Biol
11(2–3):377–394
Wagner EJ, Garcia-Blanco MA (2001) Polypyrimidine tract binding pro-
Zeng L, Liu W, Feng W et al (2013) A novel donor splice-site mutation of
tein antagonizes exon definition. Mol Cell Biol 21(10):3281–3288
major intrinsic protein gene associated with congenital cataract in a
Wang J, Chang YF, Hamilton JI, Wilkinson MF (2002) Nonsense- Chinese family. Mol Vis 19:2244–2249
associated altered splicing: a frame-dependent response distinct Zhang XH, Chasin LA (2004) Computational definition of sequence
from nonsense-mediated decay. Mol Cell 10(4):951–957 motifs governing constitutive exon splicing. Genes Dev 18(11):
Wang Z, Rolish ME, Yeo G, Tung V, Mawson M, Burge CB (2004) 1241–1250
Systematic identification and analysis of exonic splicing silencers. Zhang C, Li WH, Krainer AR, Zhang MQ (2008a) RNA landscape of
Cell 119:831–845 evolution for optimal exon and intron discrimination. Proc Natl
Ward AJ, Cooper TA (2010) The pathobiology of splicing. J Pathol Acad Sci U S A 105:5797–5802
220(2):152–163 Zhang Z, Lotti F, Dittmar K et al (2008b) SMN deficiency causes tissue-
Wimmer K, Roca X, Beiglböck H, Callens T, Etzler J, Rao AR, Krainer specific perturbations in the repertoire of snRNAs and widespread
AR, Fonatsch C, Messiaen L (2007) Extensive in silico analysis of defects in splicing. Cell 133(4):585–600
NF1 splicing defects uncovers determinants for splicing outcome Zuker M (2003) Mfold web server for nucleic acid folding and hybridi-
upon 5′ splice-site disruption. Hum Mutat 28(6):599–612 zation prediction. Nucleic Acids Res 31(13):3406–3415