Characterisation of alternative splicing of the cav1 4 calcium channel gene

CHARACTERISATION OF ALTERNATIVE SPLICING OF THE CaV1.4 CALCIUM CHANNEL GENE GREGORY TAN MING YEONG BACHELOR OF SCIENCE HONS, NATIONAL UNIVERSITY OF SINGAPORE A THESIS SUBMITTED FOR T

CHARACTERISATION OF ALTERNATIVE SPLICING OF THE CaV1.4

CALCIUM CHANNEL GENE

GREGORY TAN MING YEONG

BACHELOR OF SCIENCE (HONS), NATIONAL UNIVERSITY OF

SINGAPORE

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF PHYSIOLOGY

NATIONAL UNIVERSITY OF SINGAPORE

ACKNOWLEDGEMENTS

First and foremost, I would like to express my heartfelt gratitude to my

supervisor, Assoc Prof Soong Tuck Wah, for his patience, guidance and support throughout the course of the graduate program I also thank all the members, past and present, of the Ion Channel and Transporter Laboratory for their support,

encouragement and friendship Of special mention is Ms Yu Dejie, who performed a third of the electrophysiological recordings presented here

I express my sincere thanks to my examiners for making time and effort to examine this thesis

I thank the following people for the invaluable gifts of molecular clones: Dr

Roger D Zühlke (University of Bern, Switzerland) for the CaV C WT

pBluescript) Dr John E McRory (University of British Columbia, Canada) for the

CaV C V1.4 pcDNA3.1) Dr Roger Y Tsien (University of California, San

Diego, CA) for the mCherry clone (pRSET-B mCherry)

I thank the following institutions and departments for the support and

opportunities provided in the course of the research: National University of

Singapore, Department of Physiology, Office of Life Sciences (Neurobiology Program) and National Neuroscience Institute

I thank Assoc Prof Tan Chee Hong and Assoc Prof Khoo Hoon Eng for

refereeing my entry into the graduate program

Some preliminary work involving the identification and characterisation of transcription regulatory elements using comparative genomics was performed but are not presented in this thesis I would like to acknowledge the following people who were invaluable in this phase of work: Dr Yap Wai Ho for advice and guidance in the field of comparative genomics, and for gift of cell line Dr Fu Jianlin and

laboratory for work in Xenopus oocyte expression Dr Yu Weiping for guidance and gifts of molecular clones Assoc Prof Gan Yunn Hwen for gifts of various cell lines

Mr Paul Chen Zi Jian (summer student) for the work rendered during the course of the attachment

TABLE OF CONTENTS

CHAPTER 3

CHAPTER 4

4.1.1 Characterising the CaV1.4 c-terminal splice variants using chimeras 644.1.2 43* splice-variant inactivates more rapidly in Ca2+ and activates in more

4.1.7 Recovery from inactivation is modulated by 43* and 45a- splicing 76

4.2.1 Hyperpolarised I-V shift in Ch-43* is caused by loss of CTM 814.2.2 Co-expression with C1878-mCherry suppressed CDI and increased VDI in

CHAPTER 5

5.3 Physiological implications of the exon 43* splice variants in the retina 104

5.3.3 Activation of 43* during rod photoreceptor recovery from light pulse and

CHAPTER 6

LIST OF PUBLICATIONS

Tao J, Lin M, Sha J, Tan G, Soong TW, Li S (2007) Separate locations of urocortin and

its receptors in mouse testis: function in male reproduction and the relevant

mechanisms Cell Physiol Biochem 19(5-6):303-12

Tang ZZ, Liao P, Li G, Jiang FL, Yu D, Hong X, Yong TF, Tan G, Lu S, Wang J, Soong TW

(2008) Differential splicing patterns of L-type calcium channel CaV1.2 subunit in

hearts of Spontaneously Hypertensive Rats and Wistar Kyoto Rats Biochim Biophys

Acta 1783(1):118-30

Tao J, Hildebrand ME, Liao P, Liang MC, Tan G, Li S, Snutch TP, Soong TW (2008)

Activation of corticotropin-releasing factor receptor 1 selectively inhibits CaV3.2

T-type calcium channels Mol Pharmacol 73(6):1596-609

Posters Presented:

Tan G and Soong TW (2004) Alternative Splicing of Human L-Type Voltage Gated

Calcium Channel Gene, Cav1.4 : A Regulation of Retinal Phototransduction 5 th

Combined Scientific Meeting of the 4th Graduate Students' Society-Faculty of

Medicine, the Singapore Society for Biochemistry and Molecular Biology, the

Singapore Society for Microbiology & Biotechnology, and the Biomedical Research and Experimental Therapeutics Society of Singapore, 12-14 May 2004

Tan G, Wong E, Yu W, Yap WH, Venkatash B and Soong TW (2005) Comparative

Genomics Between Human and Fugu Voltage-Gated Calcium Channel Genes The

Society for Neuroscience 35 th Annual Meeting 2005, 12-16 November, Washington

DC, USA

ABSTRACT

CaV1.4 is a member of the L-type family of voltage-gated calcium channels

(LTCC) CaV1.4 is predominantly expressed in the rod photoreceptor synapse and is

etiological in congenital stationary night blindness type-2 (CSNB2) characterised by various visual impairments in addition to night blindness Electroretinography of CNSB2 patients suggest that CaV1.4 mediate neurotransmitter release at the

photoreceptor synapse CaV1.4 was the last among the LTCC to be cloned and

characterised The biophysical properties of this channel display a slow dependent inactivation (VDI) and a unique absence of calcium-dependent

voltage-inactivation (CDI) LTCC properties are extensively diversified by alternative splicing Using the transcript-scanning method, we identified nineteen different splice

variants of CaV1.4 in the human retina Electrophysiological characterisation of the

splice variants at the carboxyl cytosolic tail (c-tail) demonstrated modulations to activation, inactivation and recovery properties Cassette exon 43* negatively shifted

the I-V relationship by -20 mV, hyperpolarised shifted the window current, increased

current density by four-fold, induced robust CDI, suppressed VDI and halved the rate

of recovery from inactivation Exon 45a- was derived from an alternative acceptor site This shortened exon caused an intermediate slowing of the recovery rate A novel c-terminal modulator (CTM) domain was recently described in CaV1.4 that was

responsible for the abolishment of CDI Here, we demonstrated that modulated activation and inactivation properties by exon 43* splicing was a regulation targeted

at the CTM Furthermore, we provide evidence that implicates another c-terminal

domain responsible for the post-inactivation recovery of the channel Splicing of 43* and 45a- both regulate this domain The biophysical properties of the 43* splice variant suggest that it opens early when the rod photoreceptor recovers from a light pulse, and thus serve to initiate neurotransmitter release at the synapse as well as various mechanisms that maintain sustained exocytosis

LIST OF TABLES

TABLE 1.1 Nomenclature for describing alternatively spliced exon variants

TABLE 2.1 Table of primers

TABLE 2.2 PCR programs

TABLE 3.1 Alternatively spliced exons in CaV1.4

TABLE 4.1 Comparision of IBa electrophysiological properties of chimeric channels containing CaV1.4 WT, 42d+, 43* and 45a-

TABLE 4.2 Comparision of ICa electrophysiological properties of chimeric channels containing CaV1.4 WT, 42d+, 43* and 45a-

TABLE 4.3 Comparison of the kinetics of recovery from inactivation in Ba2+

TABLE 4.4 Comparision of IBa electrophysiological properties in co-expression

LIST OF FIGURES

FIGURE 1.1 Alignment of L-type calcium channel cytosolic-termini amino acid

sequences

FIGURE 1.2 Common modes of alternative exon splicing

FIGURE 2.1 Strategy for cloning Ch-WT

FIGURE 2.3 Strategy for cloning

Ch-FIGURE 2.4 Strategy for cloning Ch-42d+

FIGURE 2.5 Strategy for cloning Ch-43*

FIGURE 2.6 Strategy for cloning Ch-45a-

FIGURE 2.7 Strategy for cloning Ch-1718

FIGURE 2.8 Strategy for cloning Ch-1878

FIGURE 2.9 Strategy for cloning C1878-mCherry

FIGURE 3.1 Transcript scanning of CaV1.4 from human retina

FIGURE 3.2 PCR of alternatively spliced clones

FIGURE 3.3 Schematic illustration of CaV1.4 alternative splicing in channel structure FIGURE 4.1 Current-voltage relationship of CaV1.2-1.4 chimera wild type and

alternatively spliced variants

FIGURE 4.2 Activation and steady-state inactivation properties of CaV1.2-1.4 chimera wild-type and alternatively spliced variants

FIGURE 4.3 Calcium-dependent inactivation of current through Ch-WT, 42d+, 43* and 45a-

FIGURE 4.4 Strength of calcium-dependent inactivation in Ch- WT, 42d+, 43* and 45a-

FIGURE 4.5 Voltage-dependent inactivation of current through Ch-WT, 42d+, 43* and 45a-

FIGURE 4.6 Recovery from inactivation in CaV1.2-1.4 chimera wild-type and

alternatively spliced variants

FIGURE 4.7 Density of Ba2+-currents through Ch-WT, 42d+, 43* and 45a-

FIGURE 4.8 Current-voltage relationship of Ch-43* and deletion constructs Ch-1718 and Ch-1878, co-expressed with CTM-containing C1878 peptide

FIGURE 4.9 Current traces and calcium-dependent inactivation of Ch-43* and

deletion constructs in co-expression experiments

FIGURE 4.10 Strength of calcium-dependent inactivation in Ch-43* and deletion constructs in co-expression experiments

FIGURE 4.11 Voltage-dependent inactivation of current through Ch-43* and deletion constructs in co-expression experiments

FIGURE 4.12 Recovery from inactivation in the c-tail deletion constructs

FIGURE 5.1 Alignment of L-type calcium channel cytosolic-termini amino acid

sequences

FIGURE 5.2 Schematic illustration of rod photoreceptor membrane-voltage response

to light pulse

LIST OF ABBREVIATIONS

BLAST basic local alignment search tool

CSNB2 congenital stationary night blindness type-2

c-tail cytosolic carboxyl tail

DCRD distal c-terminus regulatory domain

E.coli Escherichia coli

EDTA ethylenediamine tetraacetic acid

EGTA ethyleneglycol tetraacetic acid

GFP green fluorescence protein

HEPES N-2-hyroxyethylpiperazine-N 2-ethanesulphonic acid

LTCC L-type voltage-gated calcium channel

CHAPTER 1

INTRODUCTION

Voltage-gated calcium channels (VGCC) are critical components to a vast variety of physiological functions Plasma membrane depolarisation of an excitable cell activates VGCC The opening of the channel pore allows an influx of calcium ions into the cytosol that coordinates a plethora of responses like neurotransmitter release, secretion, excitation-contraction coupling, regulation of gene expression and calcium homeostasis (Reviewed in W A Catterall, 2000) Channelopathies like hypokalemic periodic paralysis, Timothy syndrome, congenital stationary night blindness type-2, familial hemiplegic migraine, episodic ataxia type-2 and

spinocerebellar ataxia type-6 all testify to the importance of VGCC to normal

physiology (W A Catterall et al., 2005)

The VGCC is a complex comprising the main pore-forming 1 subunit

coassembling with an intracytosolic  subunit and an extracellular 2 subunit that is disulphide-linked to the membrane-anchoring  subunit A fourth, though non-essential, transmembrane  subunit has also been found with VGCC in the skeletal muscle Functional diversity of VGCC comes primarily from the repertoire of 1

subunit isoforms (F Hofmann et al., 1994) The various auxiliary subunits serve to modulate its function

Ten genes code for the 1 subunit in the human genome These have been functionally classified according to inactivation kinetics, pharmacological sensitivities and tissue distribution: L-type (long-lasting), T-type (transient), R-type (toxin-

resistant), N-type (neuronal) and P/Q-type (Purkinje/granular cell), are ordered according to sequence phylogeny: Cav1.1-1.4, Cav2.1-2.3 and Cav3.1-3.3 Each

member is differentially distributed in various tissues and had been shown to

contribute to various processes (W A Catterall, 2000; and W A Catterall et al., 2005) Cav1.1 in the transverse tubules of skeletal muscles interacts with ryanodine

receptor 1 and mediates direct conformational coupling of membrane depolarization

to Ca2+ release from the sarcoplasmic reticulum Whereas in cardiac muscle Ca2+

entry via Cav1.2 induces ryanodine receptor 2 to release Ca2+ from stores CaV1.3

have been localized to the basal membrane of the outer hair cells of the cochlea that are important for sound amplification Cav2.1 couples action potential to the

exocytosis of synaptic vesicles poised at the pre-synaptic membrane

The 1 subunit structure consists of four homologous domains (I-IV), each having six transmembrane -helices (S1-S6) connected in series by intra- and

extracellular linkers The four domains enclose the Ca2+-conducting pore with the

re-entrant S5-S6 loops of each domain at the innermost, forming the ion-selectivity filter Changes in membrane potential are sensed by the S4 segments Binding sites

for pharmacological agents and various signalling molecules are also found in the 1 subunit

1.2.1 Discovery and night blindness etiology

CaV1.4 belongs to the CaV1 subfamily of VGCC that mediate L-type currents,

typified by being long-lasting, antagonised by dihydropyridines, phenylalkylamines and benzothiazepines and requiring a strong depolarisation to activate Channels under this class are also known as L-type calcium channels (LTCC)

CaV1.4 is the latest among the L-type isoforms to be identified and cloned

The encoding gene (CACNA1F) was discovered by exon prediction on sequences at

human chromosome Xp11.23 (S E Fisher et al., 1997) It resides less than 5 kb adjacent to and upstream of the synaptophysin gene towards the direction of the centromere Using the neighbour-joining method and maximum-parsimony analysis,

it was determined that CaV1.4 shares the highest homology with CaV1.3 among

calcium channels (N T Bech-Hansen et al., 1998)

Genetic crossovers and disease-associated haplotype analysis identified

Cav1.4 to be responsible for incomplete X-linked congenital stationary night

blindness (CSNB2), a recessive non-progressive retinal disorder (N T Bech-Hansen et al., 1998; T M Strom et al., 1998) Although named as a form of night blindness, CSNB2 patients also suffer from varying forms of visual impairments including reduced visual acuity, severe myopia, hypermetropia, nystagmus, strabismus and

cone-rod dystrophy, in addition to impaired night vision (K M Boycott et al., 2001;

R Jalkanen et al., 2006; R Jalkanen et al., 2007) CSNB2 patients displaying vision related clinical manifestation of the disease, like intellectual impairment and autism, have also been reported Studies using linkage analyses have even

non-associated CACNA1F with schizophrenia (J Wei and G P Hemmings, 2006)

Mutation analyses in families with CSNB2 have revealed a multitude of function as well as some gain of function mutations (A Hemara-Wahanui et al.,

loss-of-2005; J C Hoda et al., loss-of-2005; J B Peloquin et al., 2007) in the CACNA1F gene

Electroretinograms (ERG) of CSNB2 patients, under scotopic stimuli, typically

shows a normal a-wave and an extremely reduced b-wave (Y Miyake et al., 1986)

transmission to the second order neurons is largely compromised

1.2.2 Characteristics of the disease mouse model

A CSNB2 mouse model was created by targeting a gene disruption to Cacna1f

The disruption was caused by an inserted pre-mature stop codon that resulted in a loss-of-function of CaV1.4 Loss-of-function mutations represent the majority of

CSNB2-CACNA1F mutations described. ERG analyses of the mutant mouse showed a

loss of scotopic b-wave and cone ERG response together with a concomitant loss of

cortical and collicular visual responses Calcium imaging revealed a marked decrease

in depolarisation-induced Ca2+ entry into the photoreceptor cell bodies and synaptic

terminals Anatomical aberrations of the retinal neurons were also evident These

include irregular sprouting of dendritic appendages from the second order neurons and absence of photoreceptor synaptic ribbons

1.2.3 Restricted tissue distribution and roles

Earlier works using RT-PCR and Northern-blot analysis demonstrated that

Cav1.4 is only expressed predominantly in the retina and at lower levels in skeletal

muscle (N T Bech-Hansen et al., 1998; T M Strom et al., 1998; F Mansergh et al.,

2005) And in situ hybridisation detected CaV1.4 mRNA in the inner and outer nuclear

layers and the ganglion cell layer of the retina (M J Naylor et al., 2000) histochemical studies, using antibodies generated against a peptide within the II-III loop of CaV1.4, displayed strong staining of the outer plexiform layer and light

Immuno-staining of the inner plexiform and outer nuclear layers These correspond to the photoreceptor synapses, the bipolar cell synapses and the photoreceptor cell bodies, respectively (C W Morgans, 2001; C W Morgans et al., 2001) Usage of different antibodies displayed differential labelling of the outer nuclear layer and the outer plexiform layer Thus it was suggested that there are two different isoforms of

CaV1.4, one localised to the photoreceptor cell body and the other localised to the

presynaptic membrane (C W Morgans et al., 2001)

Immuno-labelling of CaV1.4 channels on the outer plexiform layer was

punctate and this corresponds to the rod terminal In contrast, immune-labelling at cone terminals appeared as large patches At higher magnification, the puncta

appeared crescent-shaped that indicates sub-cellular localisation in the active zones Indeed, co-localised staining of CaV1.4 and bassoon, a synaptic marker, could be

demonstrated (C W Morgans et al., 2001; C W Morgans et al., 2005)

Immunoreactivity at the inner plexiform layer also manifested in discrete shaped puncta Further labelling of acutely dissociated retinal neurons indicated that, here, CaV1.4 is localised to the active zones of rod bipolar cell synaptic

crescent-terminals (A Berntson et al., 2003) More recent work using monoclonal antibodies further confirmed localisation of CaV1.4 channels to the active zones of rod

photoreceptors and bipolar cells In addition, detergent-extraction experiments demonstrated that CaV1.4 is anchored to the active zone complex (C W Morgans et

al., 2005) Its polarised distribution to only the photoreceptor synaptic terminals is reflected in the absence of voltage-activated currents in the rod outer segments (D

A Baylor et al., 1979) while voltage-activated calcium currents could be measured in the inner segments (C R Bader et al., 1982) The sub-cellular localisation, the

reduced ERG b-wave in CSNB2 as well as the phenotype exhibited by the

Cacna1f-mutant mouse all support the notion that CaV1.4 functions to trigger glutamate

release upon Ca2+ influx at the synaptic terminals

In situ hybridisation experiments revealed the expression of CaV1.4 mRNA in

rat dorsal root ganglion neurons (S P Yusaf et al., 2001a) that, upon induction of neuropathic pain, exhibited the highest percentage up-regulation compared to the eight other VGCCs investigated (S P Yusaf et al., 2001b) The specific role of CaV1.4

here is unclear However, LTCC activity has been implicated to contribute to pain sensitivity (K A Sluka, 1997; S Gullapalli and P Ramarao, 2002); the slow

inactivation properties of CaV1.4 may make it especially suited to this role

Increased expression of CaV1.4, together with CaV1.2 and 1.3, had also been

shown in cortical neurons after prolonged exposure to nicotine and was suggested

to contribute to drug dependence via abnormal increases of Ca2+ in neurons (H J

Little, 1991; M Katsura et al., 2002)

More recently, using RT-PCR and Western-blot, McRory et al (2004)

demonstrated some levels of CaV1.4 expression in non-neuronal tissue like bone

marrow, thymus, spleen, and adrenal gland In the adrenal gland, L-type currents could be pharmacologically isolated in rodent chromaffin cells (A Perez-Alvarez et al., 2008) and LTCC activity shown to be coupled to exocytosis (V Carabelli et al., 2007); CaV1.4 may be suited to support endocrine secretion due to its slow

inactivating properties (A Marcantoni et al., 2008) In relation to its expression in the lymphoid tissues, CaV1.4 expression was also described in peripheral T-

lymphocytes as well as the Jurkat cell line and was implicated in T-cell activation (M

F Kotturi et al., 2003) In that report, the authors successfully induced T-cell

activation using the LTCC agonist, Bay K 8644, while the antagonist, nifedipine

prevented activation and proliferation

1.2.4 Unique biophysical and pharmacological properties

Electrophysiological and pharmacological characterisation of recombinant

CaV1.4 heterologously expressed in HEK-type cell lines had uncovered several

properties unique to CaV1.4 (A Koschak et al., 2003; L Baumann et al., 2004; J E

McRory et al., 2004) CaV1.4 activate at voltages less negative than CaV1.3 and more

negative than CaV1.2 CaV1.4 displayed slower voltage-dependent inactivation

kinetics than either CaV1.2 or CaV1.3 and have a large window current spanning more

than 40 mV

CaV1.4 also showed lower sensitivity to the dihydropyridine

(DHP)-antagonists isradipine and verapamil than CaV1.2 and has lower binding affinity to

isradipine (W A Catterall et al., 2005) Sensitivity to DHP-blockade was dependent The channel was moderately sensitive to the antagonists L- and D-cis-

voltage-diltiazem but displayed no difference in sensitivities between the two enantiomers Normally, other LTCC are about twenty times more sensitive to D-cis-diltiazem than

L-cis-diltiazem

Perhaps the most striking feature in CaV1.4, compared to other LTCC, is the

absence of the calcium-dependent inactivation (CDI) that was well documented in

CaV1.2 and CaV1.3 and, recently, CaV1.1 (K Stroffekova, 2008) as well CaV1.1 was

previously purported to exhibit no CDI, this is because the CDI was masked by its slow rate of activation In CDI, the presence of Ca2+ significantly accelerates the

inactivation of the channel However for CaV1.4, it inactivates slowly in Ca2+ in a

voltage-dependent manner

1.2.5 The C-terminal Modulator a novel regulatory domain

The proximal domain of the cytosolic tail have been shown to host the key players for CDI in HVA-VGCCs Alignment of the c-tail sequence of CaV1.4 with CaV1.2

and CaV1.3 shows high conservation in the EF-hand and IQ motif region (Figure 1.1)

This may imply that the machinery required for CDI is intact in Cav1.4 Moreover, the

EF-hand and IQ-motif region of Cav1.4 binds CaM in a calcium-dependent manner (A

Singh et al., 2006; C Wahl-Schott et al., 2006) Therefore, in principle, Cav1.4 is

capable of calmodulin-mediated CDI The lack of CDI here could therefore be

mediated by an auxiliary interference on the calcium-sensing and inactivation

coupling mechanism

Further support on the capability of CaV1.4 to undergo CDI is seen in a CSNB2

mutant K159X (A Singh et al., 2006), whereby a stop codon occurs immediately after the IQ motif Functional characterisation of the K1591X revealed that this

Cav1.4 mutant displays strong CDI that has the classical U-shaped dependence on

voltage (B Z Peterson et al., 1999) In a separate investigation, engineered

truncations of the CaV1.4 after the IQ motif also restored CDI to this channel The CDI

of both the K1591X mutant and the truncated Cav1.4 are CaM-dependent as

co-expression with the dominant-negative CaM mutant (CaM1234) abolished CDI (Singh

et al, 2006, Wahl-Schott et al, 2006) These imply that there is a domain downstream

of the IQ motif that abrogates CaM-mediated CDI in full length CaV1.4

Deletional analyses have shown that a carboxyl-terminal modulator (CTM) domain that interferes with CDI in CaV1.4 resides within the last 32 to 100 amino

acids (a.a.) of the c-tail (A Singh et al., 2006; C Wahl-Schott et al., 2006)

Truncations of the CaV1.4 c-terminus at various intervals after the IQ region and up

to 55 a.a from the carboxyl end were successful in restoring CDI Only at removing the last 32 a.a from the c-tail did CDI remain absent The CTM, when co-expressed

as an independent peptide, was able to prevent CDI in the K1591X mutant as well as the truncated CaV1.4 that lacks this region Also in replacing the entire c-tail of

CaV1.2 and CaV1.3 with the c-tail of CaV1.4, or merely by exchanging the distal

portion, the CDI in these chimeric CaV1.2 and CaV1.3 channels were abolished

The CTM functions by interacting with the proximal c-tail of CaV1.4, where

the EF-hand and IQ-motif are located, as shown by co-immunoprecipitation (Co-IP) and GST-pull down assays (A Singh et al., 2006; C Wahl-Schott et al., 2006)

Whether the EF-hand or the IQ motif is the principal binding partner remains

unclear Results from the GST-pull down and co-IP experiments favour the EF-hand

as the main binding target while the FRET assays indicated that IQ region is critical for binding too

Unlike CaV1.2 (M G Erickson et al., 2003), apo-CaM does not preassociate

with the CaV1.4 c-tail (A Singh et al., 2006) FRET analyses show no binding of CaM

to full-length CaV1.4 c-tail at resting Ca2+ levels However when the distal portion of

the c-tail was removed, significant CaM-association was detected at resting Ca2+

concentrations Put together, these findings suggest that in binding to the proximal c-terminus, the CDI-modulatory domain prevents apo-CaM from preassociating with the c-tail

In short, CaV1.4 contains a modulatory domain at the distal end of its c-tail

that binds to the EF-hand and IQ-motif at the proximal region This association interferes with CaM binding and renders the CDI machinery irresponsive to Ca2+

1.3 Alternative splicing diversifies the function of calcium channels

74% of human genes with multiple exons are alternatively spliced (J M Johnson et al., 2003) And this is a common mechanism to generate functional diversity in voltage-gated calcium channels (K Jurkat-Rott and F Lehmann-Horn, 2004; A C Gray et al., 2007) Different splice-forms of various calcium channels have been shown to exhibit diverse electrophysiological properties, and some are expressed in a developmental and tissue-specific manner or are altered in response

to a physiological or pathological condition

There are several ways that exons are alternatively spliced (Figure 1.2) An exon may be excluded while its flanking neighbours are spliced together in exon skipping Either one of two adjacent exons may be spliced in preference to the other

in mutually exclusive exons Splicing of a cryptic exon lying within an intron constitutes a cassette exon Splicing to alternative splice donor and acceptor sites

the splice sites, however, do not occur randomly Splice sites are demarcated, almost

ag at the splice acceptor boundary (P A Sharp and C B Burge, 1997) This

Mutations involving the canonical nucleotide residues defining splice

junctions have been documented in various diseases and were estimated to

contribute to 15% of all point mutation-causing diseases (M Krawczak et al., 1992)

exon lengthening (a+)

exon shortening (a-)

(*)

(i) (∆)

(x)

FIGURE 1.2 Common modes of alternative exon splicing Exons can be spliced in a variety of alternative ways Exon skipping joins two non-adjacent exons while an intermediate one is excluded Introns may be retained and not spliced out In mutually exclusive exons, only one or the other exon get spliced into the final transcript A cassette exon may be spliced when

a cryptic splice-site within an intron is activated Alternative usage of donor and acceptor splice-sites can shorten or lengthen an existing exon A pair of “gt” and “ag” nucleotides reside

at the intronic boundary of splice-sites.

TABLE 1.1 Nomenclature for describing alternatively spliced exon variants

Here, we add the following suffixes or prefix to the exon number to denote the type of

alternative splicing (i.e 3a+, ∆4):

Suffix:

Such mutations usually impose an alternative exon splicing on the gene affected that abrogates its normal function In CSNB2, an A  G substitution at the splice-

acceptor site caused a loss of exon 5 with a resulting frame-shift and appearance of a pre-mature stop codon in exon 7 Another A G mutation in intron 40 led to a loss

of exon 41, frame-shift and stop codon in exon 42 (K M Boycott et al., 2001) A loss

of splice-donor site due to G  C substitution in intron 22 was also reported (M Nakamura et al., 2001)

Conversely, directed alternative splicing is explored as a therapeutic strategy

to evade or undo the aberrations caused by mutations in genetic disorders For example, these include enforced switching of mutually exclusive exons to avoid a mutation and the exploitation of lesser-known cryptic alternative splice junctions to correct for frame-shifts (various strategies are reviewed in S D Wilton and S

Fletcher, 2005) Knowledge of the repertoire and functional characteristics of the alternatively spliced isoforms in disease-causing genes will no doubt define the corrective strategies that can be applied

1.3.1 Effects of alternative splicing in L-type calcium channels

Alternatively spliced exons in CaV1.2 are numerous and have been extensively

described (Z Z Tang et al., 2004; P Liao et al., 2005) The distinctive segregation of some of these alternative exons into two major combinations had enabled two

C V1.2 splice variants, where the alternative-exon

combinations are (1, 8, 9*, 32) and (1a, 8a, ∆9*, 31), respectively Sub-variants of

these two had also been defined (P Liao et al., 2007), as were other forms of tissue and disease combinations (Z Z Tang et al., 2004; Z Z Tang et al., 2008) PKC-

mediated inhibition affects exon 1-containing smooth muscle variant but not the cardiac variant Exon 8a confers a lower sensitivity towards DHP-inhibition to the cardiac CaV1.2 Exon 9* enabled smooth muscle CaV1.2 variants to activate at more

hyperpolarised potentials (reviewed in P Liao et al., 2005)

An alternative splice variant of CaV1.3, displaying an absence of the IQ motif,

is preferentially distributed to the cochlear outer hair cells as appose to the inner hair cells The abolished CDI displayed by this isoform implicates distinct

physiological roles between the CaV1.3 channels in the outer and inner hair cells (Y

Shen et al., 2006) Alternative splicing at exon 42 resulted in a 500 a.a shorter form

of CaV1.3 The short-form channel activates at lower voltages and inactivates at least

twice-more robustly under the influence of calcium (A Singh et al., 2008)

Three alternatively spliced loci in CaV1.4 transcripts from the human retina

were previously described (K M Boycott et al., 2001) These are exons 1, 2 and 9 Alternative use of splice donor site at exon 1 led to a lengthening of the exon by 42 nucleotides (nt) or 14 a.a Exon 2 had four splice variations An alternative splice acceptor site resulted in a 237 nt shorter exon in one variant Two other exon 2 variants, derived from alternative acceptor and donor sites, were frame-shifted and non-productive The fourth variant is a mutually exclusive exon 2 that was also non-

productive The alternatively spliced exon 9 was 33 nt or 11 a.a longer and arose from an alternative donor site

Alternative splicing of exon 2 was also described in murine CaV1.4 transcripts

Interestingly, the shortened exon 2 enabled 10% of the CaV1.4 population to escape

a null mutation in the Cacna1fnob2 mouse Hence, this naturally occurring mutant

mice did not exhibit the full array of disease characteristics as the complete out counterpart (C J Doering et al., 2008)

knock-Northern blot analyses of heart and skeletal muscle RNA, using probes

corresponding to exons 6-12 of CACNA1F, detected 1.4 kb transcripts in both tissues

in addition to ~6 kb transcripts detected only in skeletal muscle The short transcript may correspond to non-productive alternative splice isoforms (S E Fisher et al., 1997) Western blot analyses of CaV1.4 proteins extracted from the rat retina

displayed multiple bands of sizes ranging between 125-190 kD (C W Morgans, 2001;

C W Morgans et al., 2001) It is possible that these reflect different populations of alternatively spliced variants of CaV1.4

Alternative splicing in VGCC is extensive and affects many domains of the calcium channel In CaV1.2, at least twelve exon loci in the gene are alternatively

spliced with up to twelve different variations of splicing occurring in one locus

(compiled in Z Z Tang et al., 2004) Other examples include seven loci in CaV2.1 (T

W Soong et al., 2002) and six loci in CaV3.1 (S Mittman et al., 1999a) Evidence

indicating that CaV1.4 proteins expressed in the retina may consist of splice variants

was discussed above and thus far, only three alternatively spliced loci were

described for CaV1.4

The biophysical properties of CaV1.4 currents measured under heterologous

expression differed from the native calcium current properties of rod photoreceptor, whereby the latter activated at ~30 mV more hyperpolarised potentials (C W

Morgans et al., 2005) This discrepancy suggests that endogenous CaV1.4 is

modulated This can occur by covalent modifications like phosphorylation or by protein interactions Interaction with CaBP4 negatively shifted the voltage of

activation by 10 mV (F Haeseleer et al., 2004) Endogenous modulation of channel biophysical properties can also occur by alternative splicing Indeed a subpopulation

of CaV1.2 splice variant was demonstrated to underlie the biophysical and

pharmacological characteristics of endogenous L-type currents in the smooth muscle (P Liao et al., 2007)

We proposed that the CaV1.4 gene is likely to exhibit extensive alternative

splicing and therefore undertook to systematically screen for these splice variants in human retina In the light of the unique lack of CDI in CaV1.4 and the newly described

CTM, we also hypothesised that splice variation occurring within the c-terminus may serve to modulate the regulatory mechanism of this domain on CDI We therefore characterised the alternatively spliced exons in the c-terminus and showed that one splice variant restored CDI to the channel as well as altered various activation and

recovery from inactivation Given these findings, we next hypothesised that the

activation and inactivation modulations can be attributed solely to the CTM, while channel recovery may rely on a different and perhaps novel domain

performed using Wizard® Plus SV Minipreps DNA Purification System (Promega Corporation, WI, USA) Large-scale plasmid DNA preparations were performed using either of the following: QIAGEN Plasmid Maxi Kit (QIAGEN, Germany), PureLink HiPure Plasmid Filter Maxiprep Kit I Life Technologies, CA, USA) and

Eagle Medium DMEM GIBCO I

Corporation (CA, USA) Foetal bovine serum, FBS, was purchased from HyClone (UT, USA) Human embryonic kidney cell line (HEK-293) was purchased from American Type Culture Collection (VA, USA)

Several of the DNA clones used in this work were kind gifts from various

Snutch (University of British Columbia, Canada), full-length CaV 1 subunit clone

(CaV1.4 pcDNA3.1) was from Dr John E McRory (University of British Columbia,

Roger D Zühlke (University of Bern, Switzerland) and the fluorescent protein

mCherry clone (pRSET-B mCherry) was from Dr Roger Y Tsien (University of

California, San Diego, CA)

The genomic sequence for CaV1.4 was from GenBank Accession No

AJ006216 The reference CaV1.4 cDNA sequence was from GenBank Accession No

AF201304 This sequence was considered as wild-type, WT, in this work CaV1.4 exon

positions were determined by aligning the cDNA sequence against the genomic sequence (using the MegAlign module of the Lasergene® software suite, DNASTAR,

WI, USA), and the exons were number in order from 1 to 48 The reference amino acid sequence used was from GenBank Accession No NP005174 The reference

CaV1.2 cDNA sequences used was from GenBank Accession No NM000719

2.3 Nomenclature for describing alternatively spliced exon variants

Various suffixes and a prefix applied to an exon number are used to describe the type of alternative splicing that occurred at that exon locus (Table 1.1, Figure 1.2) An exon that was skipped in the course of an alternative splicing is prefixed with

occurred at the acceptor or donor site, respectively Retained introns are indicated

To determine the different alternatively spliced exons in the CaV1.4 gene, we

employ the transcript strategy previously described by S Mittman (1999a; 1999b) and Soong et al (2002) In this method, we first designed PCR primer-pairs that span

at least two exons or four splice boundaries along the length of the CaV1.4 gene

(Figure 3.1A) Sufficient pairs of primers were made such that amplicons form

overlapping segments along the entire CaV1.4 sequence The primers, designed using

Oligo Primer Analysis Software (Molecular Biology Insights, Inc., CO, USA), are given

in Table 2.1 PCR was performed using human retina cDNA as template and Taq DNA

polymerase The PCR programs employed depends on the expected product length and the estimated annealing temperatures of the primer pair used, and are generally one of the three programs shown on Table 2.2 Each pair of primer produced

amplicons of varying sizes that corresponded to different alternative splice variants

TABLE 2.1 Table of primers

A1F:2198L20 CATACATGACCACGTTCCAG Reverse primer for transcript

scanning exons 14-15 16 A1F:2447U20 AGAGGAAGAGGAGGGTGCAA Forward primer for transcript

scanning exons 20-23 19 A1F:2567L20 CCTTCTCCTTGGGTACAACT Reverse primer for transcript

scanning exons 16-19 20 A1F:2752U18 ACTCCTTCCGCAACCATA Forward primer for transcript

TABLE 2.1 cont

A1F:4235U17 CAGCCTTCCCGGAAATC Forward primer for transcript

scanning exons 37-40 36 A1F:4409L18 CCCAGGATGGACCAATCT Reverse primer for transcript

A1C:HindIII4636U24 CATAAGCTTCTACATGCTCTGTGC For cloning Ch- 37

A1C:HindIII4636LX GCTCTAGATAGAAGCTTATGAAGTAGAAGACA

A1C:4111NsiIUN ATAAGAATGCGGCCGCATGCATGGAATACATT

A1F:5248stopXbaIL GCTCTAGATCAGTGGGTGTTGGATCCAGC For cloning Ch-1718

A1F:5706stopXba1L GCTCTAGATCACAGACAGGTGAAGGTGCG For cloning Ch-1878

EcoR1kozakMetCT

MU40

CCGGAATTCGCCACCATGTGTCTGCACGTGCCT GGAACCC For cloning C1878-mCherrry

A1F5977L17Xba1 GCTCTAGAGAGGGCGTGGACGCAGG For cloning C1878-mCherrry

mCherryXba1U GCTCTAGAGCCACCATGGTGAGCAAGGGC For cloning C1878-mCherrry

mCherrystopNot1R ATAAGAATGCGGCCGCTCACTTGTACAGCTCGT

TABLE 2.2 PCR programs.

(or non-specific products) These were visualised as multiple bands when separated

by agarose gel electrophoresis Each band was extracted and ligated into pGEM®-T Easy vector These transcript scanning clones were then transformed into DH10B

E.coli For every band cloned, eight to thirty positive transformants (indicated by

blue/white colony selection) were picked and further PCR screened using primers specific to the cloned insert Colonies yielding different sized PCR products were expanded and the plasmid DNAs extracted for DNA sequencing The DNA sequences were analysed, by comparison with the CaV1.4 genomic and cDNA sequences, to

identify the type of alternative splicing that had occurred and to determine the exact location of the alternative exon-intron splice junctions as well as their adherence to

gt ag

In order to transcript scan exon 1, we made use of the Marathon® adaptors that were ligated to the ends of Marathon®-ready cDNAs The adaptor primer was provided by the manufacturer and this was paired with a reverse primer residing in exon 2 or exon 3 To scan exon 48, we used a forward primer from exon 47 or 46 paired with an oligo dT that annealed to the poly-A tail As one primer in each pair was non-specific (i.e adaptor primer, oligo dT), the PCR reactions yielded diffused and multiple bands when separated on agarose gel These were extracted, cloned and rigorously screened in the same manner as described above

Primers that r UT‘ -translated region) of the CaV1.4

gene were designed and long PCR amplification was performed using human retina