Upload
duongnguyet
View
223
Download
2
Embed Size (px)
Citation preview
DOI: 10.1161/CIRCGENETICS.113.000580
4
AKAP9 is a Genetic Modifier of Congenital Long-QT Syndrome Type 1
Running title: de Villiers et al.; AKAP9: a LQTS modifier
Carin P de Villiers, PhD1; Lize van der Merwe, PhD1,2; Lia Crotti, MD, PhD3-5;
Althea Goosen, RHN6; Alfred L. George, Jr., MD7-9; Peter J. Schwartz, MD3;
Paul A. Brink, MD6; Johanna C. Moolman-Smook, PhD1; Valerie A. Corfield, PhD1
1Medical Research Council (MRC) Centre for Molecular and Cellular Biology, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, 6Department of Internal
Medicine, Stellenbosch University, Stellenbosch; 2Department of Statistics, University of Western Cape, Bellville, South Africa; 3IRCCS Istituto Auxologico Italiano, Center for Cardiac Arrhythmias of Genetic
Origin and Laboratory of Cardiovascular Genetics, Milan; 4Department of Molecular Medicine, University of Pavia, Pavia, Italy; 5Institute of Human Genetics, Helmholtz Zentrum München,
Neuherberg, Germany; 7Department of Medicine, 8Department of Pharmacology, 9Institute for Integrative Genomics, Vanderbilt University, Nashville, TN
Correspondence:
Carin de Villiers, PhD
Medical Research Council (MRC) Centre for Molecular and Cellular Biology
Division of Molecular Biology and Human Genetics
Faculty of Medicine and Health Sciences
Room 4036, Teaching Block, University of Stellenbosch
Tygerberg campus, Francie van Zyl Drive
PO Box 19063, Tygerberg 7505, South Africa
Tel: +27 72 456 7369
Fax: +27 21 938 9863
E-mail: [email protected]
Journal Subject Codes: [5] Arrhythmias, clinical electrophysiology, drugs; [89] Genetics of cardiovascular disease
logy, DiDiDiDivivivivisisisis onononon ooooffff MoMoMoMoley and Human Genetics, F lt f M di i d H lth S i 6Department of InternS no G
nr n, e
d
y and d d HuHuHumamamam n nnn GeGeGenennn tics, Faculty of Medicine ee anananand Health Scienceees,s,s,s 6Department of InternStelelelleeeenboschchhh Uniiiversity, Stellenbosch; 2Departmment of Statttisisi tics, University of Westernouuuthhhh Africa; 3IRIRIRCCCCCCSSS IsIsIsIstitititutututototo AAAAuxuxu oolololoogicco o Italliaanoo,, CeCeCentnnter fffoor CCCCara dddiacac AAAArrrhyhyhyhythhhmimimimiasas oof fff Gandndndd Laboratoryy ooof CCaarrdiovovovo ascular GGeneeetiics, MMilannn;;; 4444DDDeDepaaarttmennt of ff MMMoM lecccuc lllarrr MMMedddiccinrsiiiitytytyy ooof f Paaavivivivia,a,a, PPPaaava iaa, IItI aly;y;y;; 5555InInnnstststs itititutttte ofoffof HHHumumummannn GGGGenennneteteteticiccs,sss HHHelelelmhmhmm oooltzz ZZZZeneee trtrtrtruuumu MMMMünnnncchhc en, Germany; 7Depapapartr ment ooof f f MeMM dicine, 8Depaaartr ment of f f f PhPP armacoololoologygg , 999Institute for Inte
GeGeGeG nonononomimimicscscss, VaVaVaVandndndn ererererbibibib ltltltlt UUUUnininiveveversrssrsitititityyy, NNNNasasasa hvhvhvhviiilillelele, , , , TNTNTNTN
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
by guest on M
ay 14, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
5
Abstract:
Background – Long-QT syndrome (LQTS), a cardiac arrhythmia disorder with variable
phenotype, often results in devastating outcomes including sudden cardiac death. Variable
expression, independently from the primary disease-causing mutation, can partly be explained by
genetic modifiers. This study investigates variants in a known LQTS-causative gene, AKAP9, for
potential LQTS-type 1 (LQT1) modifying effects.
Methods and Results – Members of a South African LQT1 founder population (181 non-
carriers and 168 mutation carriers) carrying the identical-by-descent KCNQ1 p.Ala341Val
(A341V) mutation were evaluated for modifying effects of AKAP9 variants on heart rate-
corrected QT interval (QTc), cardiac events and disease severity. Tag single nucleotide
polymorphisms in AKAP9 rs11772585, rs7808587, rs2282972 and rs2961024 (order: 5’-
3’positive strand) were genotyped. Associations between phenotypic traits and alleles, genotypes
and haplotypes were statistically assessed, adjusting for the degree of relatedness and
confounding variables. The rs2961024 GG genotype, always represented by CGCG haplotype
homozygotes, revealed an age-dependent QTc increase (1% per additional 10 years) irrespective
of A341V mutation status (P=0.006). The rs11772585 T allele, found uniquely in the TACT
haplotype, more than doubled (218%) the risk of cardiac events (P=0.002), in the presence of
A341V; additionally, it increased disease severity (P=0.025). The rs7808587 GG genotype was
associated with a 74% increase in cardiac event risk (P=0.046), while the rs2282972 T allele,
predominantly represented by the CATT haplotype, decreased risk by 53% (P=0.001).
Conclusions – AKAP9 has been identified as a LQT1-modifying gene. Variants investigated
altered QTc irrespective of mutation status, as well as cardiac event risk, and disease severity, in
mutation carriers.
Key words: arrhythmia, long QT syndrome, AKAP9, KCNQ1
sisisisingngngnglelelele nnnnucucucucleleleleototototidididide
961022224 4 4 4 (o(o(o(ordrdrdrderererer: 5’555 --
s o
y
n y
e
m
more than doubled (218 ) ( =0.002), in the presence
strananand)d)d) wwwererere ee gegegennnotyped. Associations betetetweww en phenotyypic trtrtraits and alleles, geno
ypppesss were stattatisisi ttticacallllllly yy asasasseses ssssededede , addjustss iing ffor ththththe dededdegrrreee ooofff reelaaatetetedndndnd esesess ss ananand dd d
ng vavavarirr ables.s.s. Thehee rss296101010102422 GGGG GG gegenoootyyypee, alwaww ysysyy rrepeppreesenntededed bbbby yyy CGGGGCGCGCGC happploty
es, revealed annn aaagege-dededepepepep ndndndenee t QTQQQ cc increaeeasese (((1%%1% pper addditiionononal 100 years) irrespe
mutation status ((((PPP((( =000.0060606).)).) TTThehhh rs111 777772525258585 TTT allellle,,, ffffounddd uniiiquqq ely yy in the TAC
mmmororee thththanan ddddououblblblbledededd (((2121212 8888%)%)%)) tttheheheh rrisisisi k kk ofofof ccarardididiiacac eeveventnttss (((PPPP((( =0=0=00.0.0.000202020 ),),), iiiinn ththththee prpresesenencece
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
6
The long-QT syndrome (LQTS), a hereditary cardiac arrhythmia disorder resulting from
abnormal ventricular repolarization, is characterized by a prolonged QT interval on the surface
electrocardiogram (ECG) and the occurrence of cardiac events including syncope, cardiac arrest
and sudden death.1 This syndrome is genetically heterogeneous and hundreds of mutations have
been identified in different LQTS-susceptibility genes.2,3 The particular gene harboring the
disease-causing mutation influences the clinical course of the syndrome and may affect specific
triggers of cardiac events.4,5
However, despite much progress having been made in identifying LQTS disease-causing
genes and risk factors, the syndrome is associated with a high degree of unexplained phenotypic
variability, often in families with the same primary disease-causing mutation.6-8 This is seen in
the wide range of heart rate-corrected QT interval duration (QTc) , the incidence of cardiac
events (syncope, cardiac arrest and sudden death), age at which the first event occurs, as well as
the number and severity of the events experienced, with the most feared being sudden cardiac
death. Furthermore, reports of low penetrance in certain LQTS mutation carriers make it difficult
to predict clinical outcomes.7 This variability suggests that additional genetic and/or
environmental factors are at play. Consequently, recent focus has shifted towards identifying
genes, or more specifically genetic variants, that, although not directly disease-causative, may
contribute to the resulting phenotype.9-12
The current study investigates potential modifying effects of the AKAP9 gene. This gene
encodes, amongst other isoforms, the A-kinase anchor protein (AKAP), yotiao. Yotiao forms a
macromolecular complex with voltage- -subunits, Kv7.1 (also known
as KCNQ1), and -subunits (KCNE1), that are responsible for the slowly
activating delayed-rectifier K+ current, IKs.13-15 This AKAP isoform interacts with and enables
ing LQQTS diseaseeee----caccc
of unexpllllaiiiinedddd ppphehehehennnno
6 8 n
n c
n e
r i
hermore reports of lo penetrance in certain LQTS m tation carriers make it di
ofteteten nn inininin fffamamamilllieieieies with the same primary yy dididd sease-causing mumumutation.6-8 This is seen
nngegegee of heart raateee-cooorrrrectetetet d QTQTQTQ intnntervavval duurratiiiononn (((QQTQ ccc) , theh inccciiidi encececec off ccardrdddiiiai c
cope,ee ccccararara dddiacc aarrrresesttt anand susuudddddddden ddddeaeaeathththth))), aagegege at tt whwhwhwhicicichhhh thththeeee fififif rsrstttt evenenenent tt occucursrs, asas we
r and severity yy offff thehh events expepp iirienced,d,dd wiiithhh h the most ffffeareddd d bebb iniii ggg sudden cardi
hehh rt fof ll et iin taiin LLQTQTSS ta iti iri kke iit didi by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
7
the phosphorylation of KCNQ1.13 However, not only does it directly bind with and assist the
phosphorylation of KCNQ1, it is also phosphorylated itself and facilitates the conversion of
phosphorylation-induced changes into altered channel activity.16,17 It is therefore clear that yotiao
is crucial in the regulation of the KCNQ1-KCNE1 channel and it comes as no surprise that a
mutation (S1579L) discovered in this AKAP isoform results in a LQTS phenotype (LQT11).18
It is likely that AKAP9, in addition to its identified role in disease causation, acts as a
disease modifier. Here we test this hypothesis in members of a South African LQTS founder
population of Western European origin.6,19 Mutation carriers in these families harbor an
identical-by-descent KCNQ1 disease-causing mutation, p.Ala341Val
(NM_000218.2:c.1022C>T) hereafter referred to as A341V, yet their disease expression varies
considerably, providing an ideal population in which to evaluate additional modifying factors.
Materials and Method
Founder population subjects
The study population consisted of 23 South African LQTS founder families analyzed in the
present investigation, who carry the same KCNQ1 LQTS-causative mutation, A341V.6,19 All
subjects in the study provided written informed consent and the Stellenbosch University’s
Faculty of Medicine and Health Sciences Institutional Review Board approved this project
(2000/C077). Participants provided demographic information and history of disease (personal
and family). Further clinical data collected from clinical records included the timing and type of
symptoms experienced, treatment and ECG recordings. LQTS A341V-mutation status was
determined for all participants who provided blood. Exon screening of the KCNQ1, KCNH2,
SCN5A, KCNE1 and KCNE2 genes in the A341V founder family probands revealed a second
mutation/variant in only three of the probands (KCNE1: D91E and KCNH2: R328C 20 and
families harbor annnn
a
l , providing an ideal population in which to evaluate additional modi ing facto
a
opulati j
18.2.22:c:c.11110202020 2C222 >T>T>T>T) hereafter referred to asass AAA341V, yet their r diddd sease expression va
lyyy, ppprp oviding anann ideeaaala popopopo ulatatatiioi nnn iin wwwhhichh tto evevalalaluauau teee aaddittiooonnan lll l moodidididifyyyinnng fafafaf cto
and Method
ooopupulalalatititionon susubjbjbjjecectsts
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
8
KCNE1:D85N unpublished data) . No statistical correction for the presence of these variants was
made, as a small number of individuals carried them and/or there is a lack of functional and
clinical information on their effects in the study founder population and LQTS families in
general.20 Genomic DNA was extracted from peripheral blood leucocytes using a previously
described method.21 Clinical information available on the study subjects and the number of
individuals genotyped per category is summarized in Table 1. The number of individuals
included in different stages of the analyses is shown in Figure 1.
ECG analysis
Of the 349 participants for whom at least one single nucleotide polymorphism (SNP) was
successfully genotyped in this study, baseline resting ECGs recorded in the absence of beta-
blocker therapy were available for 273 individuals, of whom 137 were mutation carriers (Figure
1). The 12-lead ECG was analyzed by an investigator experienced with LQTS to ascertain
baseline heart rate (HR), RR intervals and QT interval duration. The QT interval, recorded in the
absence of beta-blocker therapy, was corrected for heart rate using the Bazett formula and ranged
between 397ms and 687ms for the mutation carrier group. 22
Cardiac events
Cardiac events recorded included syncope, aborted cardiac arrest (requiring resuscitation), or
sudden cardiac death. Mutation carriers were considered symptomatic if they had experienced at
least one of the above mentioned cardiac events irrespective of age. Mutation carriers were
classified as asymptomatic if they were older than 20 years and had never experienced one of
these events. This cut off value was chosen because a first cardiac event in this LQTS founder
population almost invariably occurred before the age of 20 years and very rarely after age 20.6
The time to first event analysis was performed on 114 mutation carriers of whom 88 were
orphihiihism ((((SNSNSNSNP)P)P)P) wwwasasasas
y a
r F
l
art rate (HR), RR intervals and QT interval duration. The QT interval, recorded
beta blocker therap as corrected for heart rate sing the Ba ett form la and r
y gegegenononotyttytypepeped ddd innn ttthis study, baseline restingngng ECGs recorded iinnn the absence of beta
raapapy yyy were avaailabllle for 27222 3 ininini diiividuuualls, ooff whwhwhomommm 1113777 werre mumumutttationnnn caaarrrrierrs (F
leadddd EEECGCGCGCG wasas aananallylyzyy ed bbbyyy aan invvvesesestitititigagatotott rr d exexpepepee iirienenenceeddd iwiwithhthth LLLLQTQTQTQTSSS to aascscerertattaiinin
art rate (H( R)), ,, RRRRRR iiiintervallsl anddd QQQQT T ini tervalll dddduratiioii n. ThThThhe QTQTQTQT iiinterval,, recorded
bbbet bbllo kck thhe tedd ffo hhe t te isi thhe BB ett ffo lla dd by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
9
symptomatic and 26 were asymptomatic. Mutation carriers who had never experienced cardiac
events but were on beta-blocker therapy prior to 20 years of age were excluded (n=17), as the
treatment could have prevented cardiac events from occurring. Furthermore, symptomatic
individuals who were on beta-blocker therapy prior to their first cardiac event were excluded
(n=7), as the age for their first cardiac event may have been different in the absence of beta-
blocker therapy. Finally, mutation carriers lacking required information for statistical modeling
(e.g. ECG and cardiac event information) did not form part of the analysis (n=30).
The severity of the disorder was evaluated using the score given in Table 2. This analysis
included 97 mutation carriers who had never been on beta-blocker therapy, or only started taking
beta-blocker treatment after the age of 20 years. In this “off” beta-blocker therapy group, all
individuals, except for two, were above the age of 20 years. One could already be classified in
the most severe category by the age of 9 years and the other, last followed up at the age of 16
years, was already symptomatic (had experienced one syncopal episode).The number of
individuals present in each severity category 0 to 3 was 27, 31, 25 and 14, respectively.
Individuals on beta-blocker therapy before the age of 20 years were separately investigated. That
group consisted of 43 individuals, with 26 below the age of 20 years (individuals per category 0
to 3: 13, 9, 7 and 14).
AKAP9 genotyping
TaqMan Validated SNP Genotyping Assays (Applied Biosystems, Foster City CA, USA) were
used to genotype all DNA samples for the intronic variants rs11772585, rs7808587, rs2282972
and rs2961024 (URL:http://www.ncbi.nlm.nih.gov/SNP). These SNPs were selected to ensure
entire gene coverage, including 2kb on either side, using Tagger analysis (URL:
http://www.broadinstitute.org/mpg/tagger), applying default settings with an r2 cut off value of
en in Table 2. Thissss aaaan
rapy, or onlllly tttstartetetetedddd t
e l
, d
v
a
present in each se erit categor 0 to 3 as 27 31 25 and 14 respecti el
er trrreaeaeatmtmtmenenent t tt afffteteter the age of 20 years. In ttthihihh s “off” beta-blooockckckcker therapy group, al
, exxexe cept for twowoo, wwwerrre aaaabobobob ve tthehehe aage oof 200 yeaarsrsss.... OOOneee ccouldl alrrreeeadyy bbbbe ee ccclassisiiifffif ed
vere cccatatatategegege ory bybyby tttheheh aagegg oooffff 9999 yearararsss aaaandndnd tttthehehh oothththeeeer, lalalastt fffooolollololl wewed upupupup aaat thhhhee agagee ffofof
already yy syyymppptoma iitiic (((hhhah d dd expepp iiriiencedd d one syyyncopppallll epipiiisodeddd ).)).) ThThThThe number of
t iin hh irit at 00 to 33 2277 3311 2255 dd 1414 iti ll by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
10
0.8 and a minor allele frequency (MAF) of 0.2 (common SNPs).23 The HapMap Genome
Browser release#24 (Phase 1 & 2) full dataset was used and information pertaining to the CEU
HapMap population (Utah residents with ancestry from northern and western Europe) was
selected.
Polymerase chain reactions (PCR) were set up in 384-well plates, each reaction consisted
of 20ng of genomic DNA, 2.5
primer and probe dye mix, as well as 1.25 -free sterile water. Amplifications took place
using the following cycle: 2min at 50°C, 10min at 95°C, followed by 40 cycles of 15s at 92°C
and 1min 30s at 60°C. Allele discrimination was achieved by analyzing PCR amplified products
with the ABI Prism 7900HT Sequence Detection System SDS v. 2.3 software (autocaller
confidence level=95%). All allelic and genotype data given for the chosen SNPs corresponded to
the positive DNA strand relative to the reference sequence.
Statistical Analysis
SNP analysis
Pedstats v 6.11 (Wigginton and Abecasis, 2005) was used to check Mendelian inheritance, as
well as to perform exact tests Hardy-Weinberg Equilibrium (HWE) amongst selected maximally
informative but unrelated individuals from the founder population.24 Haploview v.4.2 (Barrett et
al. 2005) was used to estimate the MAFs, as well as the linkage disequilibrium (LD) measure
D’.25 As Haploview does not always select the same group of maximally informative but
unrelated individuals for estimating LD, the analysis was repeated 100 times, and median values
are reported . R, a language and environment for statistical computing, freely available from
www.R-project.org and R packages Coxme (Terry Therneau, 2012. coxme: Mixed Effects Cox
Models. R package version 2.2-3. http://CRAN.R-project.org/package=coxme) and kinship2
40 cycles of 15s at t t t 92999
ng PCRCRCRCR amplilililififififi ddedd pppprrrro
B
n
e
A
i
BI PPPriririr smmsm 77790999 0H0H0HHT Sequence Detection SSSyyysttem SDS v. 2.3 ssofooo tware (autocaller
levevevvel=95%). AlAlAll alllleeeeliic aana d geggenottypeee datta giveven n n n fofofor thtthe chhhoosennn SNPPPPss cooorrrresppoon
e DNANANANA sssstrtrtrt and dd rerelalall tititiveve to thththeeee referererennnce eee seequququeencececee.
Analysyy is
ii by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
11
(Terry Therneau, Elizabeth Atkinson, Jason Sinnwell, Martha Matsumoto, Daniel Schaid and
Shannon McDonnell, 2012. kinship2: Pedigree functions) were used for statistical analysis.
Traits
The traits investigated in this study were QTc duration, a severity score (Table 2) and the time
until the first cardiac event (age). As this founder population contains many affected individuals,
the QTc duration data was not normally distributed; it was positively skewed due to some very
large values. Because the mean and standard deviation are not appropriate statistical measures in
this case, terms of the median and interquantile range (IQR) were described. Furthermore, linear
statistical modeling required a log-transformation to approximate symmetry prior to analysis
ensuring the validity of results and conclusions. Kaplan-Meier curves are used to illustrate the
age at first cardiac event.
Linear mixed-effects models were used for QTc duration and the severity score, while
mixed-effects Cox regression models were used for the analysis of time to first cardiac event. All
models were adjusted for gender as a fixed effect, and for the degree of relationship between
each pair of individuals in the study (using a per individual random effect with kinship
coefficients), as well as a per family uncorrelated random effect. The linear models were further
adjusted for age and the QTc model also for the presence or absence of the A341V mutation as a
fixed effect. Therefore, significant QTc analysis associations were representative of the change
in QTc irrespective of mutation status. The mixed-effects Cox regression models for the time to
first cardiac event analysis were additionally adjusted for QTc. Effect estimates from the models
are reported as percentage change (not milliseconds) for QTc, due to the back-transformation
(exponentiation) of the logarithms, change in score for severity and hazard ratio (HR) for
experiencing a cardiac event.
cribed. Furthermorrrre,e,e,e l
metry yy priiiior ttott anaaaalylylylys
e
c
e h
c n
re adj sted for gender as a fi ed effect and for the degree of relationship bet ee
e valalalalidididditititityyy ofofofof resesesults and conclusions. Kaaaplplpp aan-Meier curvess aaare used to illustrate
caarardddiac event.
ear mmmixxixixededede -effffefef ttctctss momodedd ls wwwweere ussseddeded ffffor QQQQTTcTcT dddururururatatatioioion annanand ddd thhththee sevevevveriririity sscocorere, wwh
cts Cox regrgg essiiiion modddd llels were used ddd foff r hthhe anallyyy isiiis offff tiimii e to ffffiirii st cardiac even
dadjj st ded ff dde ffii ded ffff t dd ffo hth dde ff lla iti hshiip bbet by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
12
As an indication of precision, 95% confidence intervals (CI) are also provided.
Genetic association
For single SNPs, the genotype effect was tested by splitting it into an additive term (number of
minor alleles) and a heterozygote flag (yes/no) in all models. This is equivalent to testing a
genotype effect on two degrees of freedom. If either of the components was significant, the
corresponding genetic model, additive, dominant or recessive on the minor allele, was tested.
The result from the best fitting model is reported.
Simwalk v.2.91 (Sobel and Lange, 1996) was used to infer the most probable maternal
and paternal haplotype configuration for individuals with the necessary genotype and/or phase
information in the pedigree.26 Haplotypes with an estimated frequencies larger than 0.01 were
analysed. An additive (without heterozygote flag) model of association was fitted for individual
haplotypes, as described for SNP evaluations, thereby comparing each haplotype to all others in
each model. Additionally, the interaction of each possible confounder (age, gender and presence
or absence of A341V) was tested with all genetic variables (alleles and haplotypes) on QTc.
Furthermore, the interaction of gender was tested with each genetic variable on the risk of first
cardiac event.
In this study, no correction for multiple testing was performed. Correcting for multiple
testing can be overly stringent in family-based association studies. Furthermore, a Bonferroni
correction assumes independence between the different tests performed and is therefore not
appropriate when analyzing different SNPs in LD.27
Results corresponding to P-values lower than 5% are described as significant and
reported. A flow diagram of the statistical methods applied is provided in the supplementary
information.
most probable matatatatereee
y genotttyt pe a dddnd/or rrr pphphph
n 26
An additive ithout heterozygote flag model of association was fitted for indiv
, e
l s
of A341V) as tested ith all genetic ariables (alleles and haplot pes) on QTc
n innnn ttthehehee pppedededdigiii rerereeee.26 Haplotypes with an esesestitimated frequenccieieiees larger than 0.01 w
AnAnAnA aaadditive (wwittthoututt hhetttteere ozygyygoootee flagaag) momodeeelll ofofof aassocoociationonn wwwaaas fitttttetetet d fooor innndddid v
, as ddddesesesscrcrcrcribibibeddd ffforor SSSNPNNPNP evaaaluululuationsnsns, ththththerebebbby cocompmpmpmpararariniining gg eaeeaeachchh hhhhaplolololotytytype totot aallll oothththe
l. Additionally,y,y, thhehh iiiinteractiiiion offf f eachhhh popp ssibibibibllel confofff unddded r (a((a( gegg , ,, gegg nder and pppres
ff A3A34141V)V) t tedd ii hth llll etiic iiablbl (( lalllelle dnd hh llot )s) QTQT by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
13
Results
Of the 350 individuals, for whom DNA was available, only one individual was not successfully
genotyped for at least one of the SNPs (Figure 1 and Table 1). Furthermore, all four SNPs
analyzed were in HWE for the selected subset of unrelated individuals. Minor allele frequencies
varied from 0.10 (T allele) for rs11772585 to 0.46 (G allele) for rs7808587. The number of
individuals successfully genotyped for the different SNPs and allele and haplotype frequencies
are given in Table 3. Due to strong linkage disequilibrium between the SNPs, and the fact that
this study was conducted with related individuals, haplotype frequencies varied from the
expected if assuming linkage equilibrium. The LD analysis, in the unrelated subgroups, revealed
a D’ value of 1 for all SNP pairs investigated, for all 100 runs. A D’ value of 1 indicates that the
two markers being tested are in complete (not perfect) linkage disequilibrium28, with at least one
two-marker haplotype having a frequency of 0, as was the case for all SNP pairs investigated in
the current study.
Each SNP was evaluated for association with all traits and the significant results are
tabulated in Table 4, as well as discussed briefly below.
QTc analysis
Initial investigation of the data revealed a highly significant correlation of gender (P<0.001) with
QTc. Females had an estimated 5.6% longer QTc than males. Additionally, the presence of
A341V was correlated with QTc (P<0.001). Mutation carriers had an estimated 21.4% longer
QTc than non-carriers. QTc heritability was estimated at 31% (55% unadjusted for A341V).
A significant interaction between age and rs2961024 on QTc (P=0.006), after adjusting
for gender and mutation status, was detected. Individuals with the TT genotype for this SNP
showed a 1% (95% CI, 0.2 to 1.3%) shorter QTc with every additional 10 years of life, while, for
ies varied from theeee
elatedddd s bbbubgroupspp , rerererev
o a
r s
r e
h SNP as e al ated for association ith all traits and the significant res lts are
of 111 fofoforrr alalala l l l SNSNSNP PPP pairs investigated, for alalalall l ll 1011 0 runs. A D’ vvalaaa ue of 1 indicates tha
rs beeeing testedd aaare iiinnn compmmm letette (nnnoot pperrfecct)) linnnkakakagegegg disiisequiliiibrririuuum28282828, wiwiww thhhh attt lelel as
r haplplplotototo ypypypy e hahah iivivingngng aa fffrequququenenenency offff 0000, aaaas wasas thehehe ccasasaseee fofofoforrrr lalallll SNSNSNSNPPPP pappapaiririri s inini veveststtiigigaate
study.yy
hhh SSNPNP ll at ded ff iia iti iithh lalll tr iaits dnd thhe iig inififi nt llts by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
14
each minor G allele carried in genotypes GT and GG, this age effect increased by 1% (95% CI,
0.3 to 1.6%) over the same time period compared to the TT genotype. The net result was no
significant effect for GT, and a 1% increase for the GG genotype. Figure 2 indicates this increase
compounded over a number of years. The curves commenced at the highest QTc for a female
mutation carrier (Figure 2A) but dropped for male mutation carrier (Figure 2B) and non-carrier
estimates (Figure 2C and D) due to the difference in baseline QTc values. However, the relative
position of the genotype curves to each other remained the same.
As an example, a 10 year old individual with the GG genotype and a QTc of 450ms
would have an increase of 4.5ms over the next 10 years. The 1% increase over the following 10
years would then be calculated from a QTc of 454.5ms. Continuing this process, the effect of this
risk genotype from 10 to 40 years of age would be an increase of 13.6ms (QTc 40
years 463.6ms). However, as the millisecond change depends on the QTc of the individual, and
the QTc values for mutation carriers in this founder population ranged between 397ms and
687ms (mean age=34.9), the actual change in milliseconds, increase or decrease, between
mutation carriers is expected to be extremely variable.
Interestingly, all individuals in the founder population who had the GG genotype at
rs2961024 were also homozyotes for each of the other three SNPs investigated, that is they all
had two copies of the CGCG (order: rs11772585, rs7808587, rs2282972 and rs2961024;
frequency=0.282) haplotype. Consistent with this observation, and with the above results,
individuals with two copies of the CGCG haplotype had an average QTc 1% (95% CI, 0.2 to
1.4%) longer per 10 year increase in age (P=0.012), while carrying one copy of CGCG showed
no significant effect on QTc duration and in individuals without a copy of this haplotype there
was an average 1% (95% CI, 0.2 to 1.3%) shorter QTc duration per 10 year increase in age.
and a QTc of 450000msmmm
ase over tttthehhh ffff llollooowiwiwiwin
d then be calculated from a QTc of 454.5ms. Continuing this process, the effect
p
6 l
l
an age 34 9) the act al change in milliseconds increase or decrease bet een
d thhhhenenen bbbbeee cacacalcululululated from a QTc of 454.555msm . Continuing thihihiis process, the effect
ppe fffrom 10 to 40 yyeaaars ofofofo ageee wwouuld bbe aan inccrereeeaasaa ee of 13.6m6m6ms (QQQQTcc 44440
6ms)))). HoHoHoHoweveerr, asas ttthhhehe milililliililissecondndnd chahahah ngngngee dededd pepepeendndndsss onnn tttthhhehe QQQQTcTTT oooofff f thththhe inini diddidiviviiidduduala
lues for mutation carriei rs iin thhhis fffounddder pppopppulatioiii n ranggg deddd bbbbetween 397ms and
3434 99)) thhe ct ll hch ii imilllliis dds iin dde bbet n by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
15
However, given the small number of CATG observations it is impossible to discern whether the
effect of rs2961024 is restricted to the CGCG haplotype.
Interestingly, after adjusting for A341V and gender, rs2961024 was also significantly
associated with QTc (P=0.014) independent of age, with the GG genotype associated with longer
QTc relative to the TT and GT genotypes. Furthermore, including an adjustment for age yielded
the same result, with the presence of the GG genotype associated with a 5% longer QTc (95%
CI, 1.0 to 9.1%) compared to TT and GT. However, no statistically significant additive allelic or
haplotype associations with QTc were observed given these adjustments.
Therefore, after adjusting for age, gender and A341V, both rs2961024, and the
interaction between age and rs2961024, are significant predictors of QTc. The QTc increases
with age and this observation is progressively more pronounced with each additional G allele
compared to the TT genotype; even ignoring age, the average QTc is higher. Therefore, the age-
dependent effect can be considered the best description of the observed association between
rs2961024 and QTc.
Cardiac events analysis
The risk of experiencing a cardiac event was significantly greater in subjects with a longer QTc
duration for all models investigated (P=0.000035), consistent with previously reported
associations in this founder population.6 The significant effects described below are independent
of QTc.
The AKAP9 SNP, rs11772585, was associated with an increased risk of cardiac events
(P=0.002). For this SNP, no TT genotypes were observed, yet having the T allele within the CT
genotype more than doubled the risk of having a cardiac event (HR=2.18; 95% CI, 1.33 to 3.60).
Figure 3A illustrates the 118% greater probability of a cardiac event. Haplotype analysis
nts.
961000024242424, andd dd ththththe
n e
o e
effect can be considered the best description of the observed association between
and QTc
betwtwtweeeeeenn n agagage ananand rs2961024, are significcccanaant predictors of QTQTQTQTc. The QTc increas
ndddd thhhis observaatiiion issss prooooggrg esssisisiveeely mmmoore prpronnououuunncnn eed wwith eaaachhh addiiiti ioiioi nananalll l GGGG aala le
o thhhe TTTTTTTT gennotottypypype;e; eveven iiignngngnoring gg agagagageeee, ththhhee avaverereragaaage ee QTQQTQTcccc iiisis hhhigiigigher.rrr TTTThhhherereffoforere, thththe
effect can be considididderedddd thheh bbbest dedd sc iriiptptptiioi n of thehhh obbbsb erved ddd association between
dd QTQT by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
16
supported this result (Table 4).
Another AKAP9 SNP, rs7808587, was associated with a recessive model (P=0.046), with
the GG genotype increasing the cardiac event risk by 74% (HR=1.74, 95% CI, 1.01 to 3.01)
relative to the AG and AA genotypes. Figure 3B illustrates this increased probability of a cardiac
event.
Finally, rs2282972 was associated with a dominant model (P=0.001) with the CT and TT
genotypes reducing the cardiac event risk by 53% (HR=0.47, 95% CI, 0.29 to 0.74) relative to
the CC genotype. Figure 3C illustrates this reduced probability of a cardiac event.
Consistent with the above result, CATT, the predominant haplotype containing the
rs2282972 T allele, was associated with a decrease of 35% (HR=0.65, 95% CI, 0.45 to 0.94) per
haplotype copy (P=0.022).
Severity score
The effect of any of these SNPs on the severity of the disease was investigated using the score
given in Table 2. In agreement with the other results, rs11772585 also affected disease severity
for the “off” beta-blocker therapy group. The T allele was significantly associated (P=0.025)
with a higher score by a value of 0.58 (95% CI, 0.07 to 1.08). Haplotype analysis supported this
result (Table 4). The separately investigated smaller group consisting of individuals on treatment
did not show any statistically significant results.
Discussion
The present study demonstrates that yotiao, the protein encoded by the AKAP9 gene, is a
modifier of the clinical phenotype of LQTS. This finding is important given that individuals with
LQTS present with symptoms that can vary considerably, irrespective of their carrying the same
disease-causal mutation, as demonstrated by the families in this study.6-8 In this founder
rdiac event.
otype co tttnt iiaiiniiiinggg tttthehehehe
T 4
c
c
of any of these SNPs on the severity of the disease was investigated using the sc
able 2 In agreement ith the other res lts rs11772585 also affected disease se e
T alalalallelelel lelelee, wawawas asasassociated with a decreaseee ooof f 35% (HR=0.65,55, 9995% CI, 0.45 to 0.94
coooopyyyy (P( =0.02222)...
core
of any yy of these SSSSNPNPNPs on thheh severitii y yy fofff thehhh ddddisiii ease was iiiinvestigagg ted usinggg the sc
abbblle 22 IIn nt iithh hth othhe llt 1s117177272585855 lal ffff tedd didi by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
17
population, KCNQ1 A341V mutation carriers have a broad range of QT intervals and they may
be symptomatic or asymptomatic, depending on the occurrence of cardiac events.6 Worldwide,
the A341V mutation is associated with a very severe form of the disease, with the likelihood of
cardiac events occurring more frequently and earlier in affected individuals relative to other
KCNQ1 mutations.29 However, LQT1 individuals older than 20 years of age who, despite not
having treatment, are asymptomatic, seldom become symptomatic afterwards.5,29 Accordingly,
mutation carriers identified after this age rarely started treatment, explaining the low beta-
blocker treatment percentage for the study population in Table 1.
Despite the severe phenotype of the A341V mutation, it has been shown that it only has a
modest dominant negative effect on the wild-type channel.6 The channel subunits encoded by the
mutated allele, due to their association with other subunits to form the complete ion channel
complex, negatively influence the function of the wild-type subunits and thereby the
performance of the channel. Heijman et al. showed that the severity conferred by this mutation
can, in part, be explained by the reduced phosphorylation of KCNQ1.30 Yotiao plays an
important role in this phosphorylation process, as well as in downstream changes in channel
activity as a consequence of this phosphorylation.13,17 Therefore, it could be proposed that
variants in the AKAP9 gene that result in altered yotiao expression, structure and/or function are
likely to influence channel properties and could explain some of the severity and phenotypic
differences observed. As a number of KCNQ1 mutations, including A341V, suppress cyclic
adenosine monophosphate (cAMP) or protein kinase A (PKA) mediated channel regulation30,31,
it is possible that modifying effects by AKAP9 variants are acting on the wild-type channels.
This investigation strengthens the notion that yotiao is indeed involved in conferring
some of the severity and phenotypic variability. AKAP9 variants in this study were shown to alter
een shohhh wn tttthhhhat ttt iiit oooonlnlnlnly
m 6
e e
e
c t
be e plained b the red ced phosphor lation of KCNQ1 30 Yotiao pla s an
minananantntnt nnnegegegatata ivvve e e effect on the wild-type ccchahahh nnel.6 The channnnnneleee subunits encoded
ellel ,,,, due to theeirrr asssooociaaiatititit on wwwiiti hhh ootheeerr suubuuniiitsts ttttoo oo foff rmrrm thee cccompmm letetetee iiionnn chaaannne
egattiviivivelelelelyyy y iiinflflflueuencncee thhththe funcncncctititition ooofff ththththee wiwiiildldldd-typypypeeee sususubbbbunininiitttsts aa ddndnd theheheherererer bbbby tttheheh
ce of the channellll. HeHH ijiji man et al. shohh wedddd thahh t the severiiitytyty confffef rred by yy this mutatd
bbe llaiin ded bb hth dd dd hph hph llatiio fof KKCNCNQ1Q1 3030 YYotiia lpl by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
18
the QTc interval, as well as the risk and severity of cardiac events, in the South African A341V
founder population. As these findings were observed in a population with a very specific genetic
background, further work is required to verify that the same effect is observed in other
population groups, as well as for different types of LQTS. Furthermore, as these variants are
located in the intronic regions of the gene, they themselves are unlikely to be the direct cause of
the modifying effects and variants in LD with them require further investigation. However, this
work clearly indicates involvement of the AKAP9 gene region as a LQTS modifier.
Most other genetic modifying effects reported have been associated with already known
LQTS genes. Variants associated with QTc and/or sudden cardiac death have been described for
KCNQ1, KCNH2, SCN5A, ANK2, KCNE1, KCNE2 and KCNJ2 (LQT1-7).11,32-35 Additionally,
recent genome-wide association studies have revealed a few novel QTc-associated loci,
including the nitric oxide synthase 1 adaptor protein gene (NOS1AP).34-36 NOS1AP has since
been repeatedly associated with QTc interval, LQTS disease severity and sudden cardiac death
risk37-40, with significant modifying effects reported for the same founder family described in the
current study.9
AKAP9 variant associated with QTc duration
The heritability of the QT interval is reported to range between 25% and 55% (latter value from
present study), with many common variants examined to date not explaining much more than 7%
of the total QTc variation.34,41 Furthermore, 31% of the estimated heritability for the South
African A341V founder population could not be explained by the A341V mutation, age or
gender (this study).
The current study identified an age-dependent additive allelic association between the
rs2961024 SNP and QTc (Table 4 and Figure 2). This SNP was identified as a modifier for QTc
iated with alreadydydyy kkkkn
th have bbbbeen deddd scccriririribbbbe
C 11 32 35 a
o
h c
t e
ith significant modif ing effects reported for the same fo nder famil describedd
CNHNHNH222, SCSCSCSCN5NNN AAA,,, ANK2, KCNE1, KCNE2 22 anand KCNJ2 (LQTTT1111-7).11,32-35 Additiona
ommmmeeee-wide assoociiiationnn stuuuuddid es hhhavvvee revveealeded a fffewewww nnovvveel QTcTcTc-aasa sssociatatata ededede loooci,,,
he niiiitrtrtricicicic oooxideded ssynynynthththasase 111 addadadaptor rr pprprotototo ieieiinn gegeneee ((((NONONONOS1S1S1S APAPAPP((((( ).))) 343444-3-3-3-36 NONONONOS1S1S1S1APPP hhahah ss isisincn
tedly yy associatedddd wiiiti hhhh QTQTQTQ c iniii terv lall,,, LQLQLQL TSTSTSS ddddiiisi ease severiiiti y yy anddd d suddddd en cardiac de
ithh isi ififiic t didiff iin fefffect rt ded ffo hth ffo dde ffa imill dde iribbedddd by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
19
irrespective of mutation status in this specific founder population. Older individuals, with two
copies of the minor allele (GG genotype), displayed a longer QTc. The opposite was true for
individual carrying two copies of the major allele. Pfeufer et al. also report age-specific effects
on QT interval when looking at the gene encoding the RING finger protein 207, RNF207, as well
as that encoding the cardiac phospholamban protein, PLN.35 However, in their investigation,
each minor allele of the particular SNP investigated contributed towards a larger QT-prolonging
effect (rs846111) or a lower QT-shortening effect (rs12210810) in younger individuals.
Although, in this study, an association between rs2961024 and QTc was observed both
with and without considering the age-dependent effect, other investigations, not evaluating
interactions with specific confounding variables, could be missing specific variable-dependent
QTc associations. Additionally, studies using additive allelic models only could miss dominant
and recessive effects. This highlights the importance of evaluating different genetic inheritance
models, as well as interactions of genotypes with confounding variables on clinical outcomes.
Many previously reported common variants associated with QT interval only alter the variability
by small percentages, emphasizing the relevance of considering multiple models for detection.41
AKAP9 variants associated with modifying risk and severity of cardiac events
The AKAP9 gene not only alters QTc duration but also influences the risk and severity of cardiac
events, after controlling for QTc. However, different variants were associated with cardiac event
risk and severity than the one shown to alter QTc interval duration (Table 4). The extent of the
phenotypic variability in LQTS, as well as the existence of mutation carriers with normal QTc
intervals that experience cardiac events, suggests that there are at least one or more different
variants involved in modifying cardiac events and severity to QT interval duration.6-8,42
Furthermore, as yotiao-dependent regulation of the KCNQ1-KCNE1 channel is so versatile, in
QQTc was observeeeed d d d b
ations, n tttot evallull atatttininining
s d
a n
v a
well as interactions of genotypes with confounding variables on clinical outcom
io sl reported common ariants associated ith QT inter al onl alter the aria
s witititith h h spspspecececifififificcc ccconfounding variables, coooululuu d be missingg speeciccc fic variable-depend
attttiooons. Additiononnallylylyy, studududu ies uusinnngg addddditiveve allllllelelelicicic mmodoodels ononnlyyy cccouldlddld mmmisisiss dooommin
ve efeffffeefefectctctctsss. Thihihiss hihihih ghghhhlilillightstss tttthehehehe impmppooortatatat nccee fofoff eeevaavavalulluluattatatinnnggg diddidiffffffererentttt gegegegenetititicc iiininhehehh iririta
well as interactiiiions offf gggenotypypypes withh h co fnfffounding gg va iriii babbbles on clinical outcom
i lsl rt ded iri ts iiatedd ii hth QQTT iinte lal lnl lalte hth iria by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
20
targeting a number of proteins to the channel complex and in ensuring not only efficient KCNQ1
phosphorylation but also the subsequent channel response, it is not hard to imagine that different
variants in its encoding gene could result in slightly different phenotypic expression.13,17,43,44
Interestingly, Kao et al. also report a few NOS1AP variants affecting the QT interval but not
sudden cardiac death.45 Work dedicated to understanding the different mechanisms responsible
for these observations requires further attention.
Three AKAP9 SNPs were significantly associated with cardiac event risk, and one of
these was also associated with disease severity. The rs11772585 minor allele (T) was
significantly associated with both the risk of having a cardiac event and increasing the severity of
the disease. The presence of this allele more than doubled the risk of a cardiac event. These
results were supported by the haplotype analysis (Table 4). The rs7808587 GG genotype was
associated with a 74% increase in cardiac event risk, while the rs2282972 minor allele (T)
revealed a protective effect, reducing the risk of an event by 53%. A protective effect associated
with another LQTS-causative gene variant, KCNQ1 rs2074238, has also recently been reported,
with the relevant allele (T allele) modifying both the risk of symptoms and QTc duration.46
Conclusions
Variants in a number of arrhythmia related genes have been reviewed and their contributions
towards QTc interval and sudden cardiac death described in different populations.11,12,41
However, the LQTS gene, AKAP9, encoding a prominent protein in the regulation of the
KCNQ1-KCNE1 channel, has not previously been described as a potential modifier. Results
evaluated here clearly demonstrate that this gene contributes to LQTS phenotypic variability.
However, as these SNPs are located in the intronic regions of this gene, it remains to be seen
which functional and/or regulatory variants in LD with these SNPs are responsible for these
r allele (T)) was
nd increasiniii g ththththe sesesesevvvve
. e
e w
w
p c
er LQTS ca sati e gene ariant KCNQ1 rs2074238 has also recentl been repo
. Thehehehe ppprereresesesencnn e e ofooo this allele more than ddddoououblb ed the risk of aaa cardiac event. These
e suuupu ported byy ttthe hahahahapllotototo ype ee aanalaalysis (Taabble 4)4)4).. ThThThe rsrss780085888777 GGGG gggennnoootypppee w
with hhh aaa 74747474%%% inncrcreaeasese iiin n cardrdrddiaiaiaiac evvvenenent iririskskkk, hwhwhililileeee thththeee rs2s2s22828282829292977277 mimimiminononor alallllelelelel (((T)T)T)
prpp otective effect, ,, redddud ciiiinggg thhheh riiisi kk k offff an event byby 553%3%3%3%. AAA A prpp otectiiiive effect assoc
LQLQTSTS atii iiant KCKCNQNQ11 2s2070742423838 hha lal ltl bbe by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
21
modifying effects. Furthermore, it would be interesting to see if similar effects can be replicated
in other populations. Importantly, these findings provide insight into the role that AKAP9 plays,
not only as an LQTS-causal gene, but also as a phenotypic modifier.The identification of the
mutation-specific risk associated with a growing number of modifier genes epitomizes the
evolution in the understanding of LQTS, a disease which increasingly represents the clearest
example of how tightly connected the relationship between genotype and phenotype can be, and
how this is progressively impacting on clinical management.29,47
Acknowledgments: An affiliation with the MRC Biostatistics Unit is hereby acknowledged for part of the study. The authors thank Ina le Roux and Glenda Durrheim for technical assistance.
Funding Sources: Genotyping the selected AKAP9 SNPs was enabled by funding provided by the National Research Foundation grant FA2006040400017 (V.A.C.). The clinical work was supported by NIH grants HL068880 (A.L.G and P.J.S.), and by financial support from Telethon – Italy (Grant no. GGP07016). The financial assistance of the National Research Foundation (DAAD-NRF) towards this research is hereby acknowledged. Opinions expressed and conclusions arrived at are those of the authors and are not necessarily to be attributed to the DAAD-NRF.
Conflict of Interest Disclosures: None.
References:
1. Schwartz PJ, Crotti L, Insolia R. Long QT syndrome: from genetics to management. Circ Arrhythm Electrophysiol. 2012;5:868–877.
2. Hedley PL, Jørgensen P, Schlamowitz S, Wangari R, Moolman-Smook J, Brink PA, et al. The genetic basis of long QT and short QT syndromes: a mutation update. Hum Mutat.2009;30:1486–1511.
3. Schwartz PJ, Ackerman MJ, George Jr. AL, Wilde AAM. Impact of Genetics on the Clinical Management of Channelopathies. J. Am. Coll. Cardiol. 2013;62:169–180.
4. Priori SG, Schwartz PJ, Napolitano C, Bloise R, Ronchetti E, Grillo M, et al. Risk stratification in the long-QT syndrome. N Engl J Med. 2003;348:1866–1874.
herebebebby y y y acacacacknknknknowowowowleleleledgdgdgdgeeededm for tetetetechchchchninininicacacacal l l l asasasassisisisistss a
o da ab ea oRs arrived at are those of the authors and are not necessarily to be attributed to the
R
ourururrces: Geeenononon tytyytypingngngng ttthehehee ssselellelececececteeed d d d AKAKAKAPAPAPA 9 SSNPPsss wawawwas enenenenababableeedd dd byy ffffununuundididid nggg pppprororovivivividedal RRResearch Foouunndaatiiiion grgg ant tt FAFFAF 2200660040404000010 7 7 7 (VVV.AAA.C.)). ThThThe clinnnnicicicical wwworkrkkk wabyyy NNNNIHI grantnn s HHHL0606688808080 ((A.L.L GGG anddd PP.J.SS.),,, aandnd bby y fiffinnancciiial suusuppororortt t frfrfrooom TTTeleant no.oo GGGGGGGP0707070110101666).)) TThTT e fififinanananancialll aaassisisis stanancece ooofff ththththe ee NaNNaNatitiitiononalall RRRReseaeaeaearrcrch FoFoF unundadadd tititioRF) towards thhhhisisisis rrreseseseseaeaeae rcrcrcrch h h isisiss hhhherererrebebebeby y y y acacaca knknnowowowwleleleledgdd ededede . OpOpOpOpinininnioioioionsnsns eeexpxpxpxprererressed and s arrived at are thhohh se offf f hhthe au hthhhors and ddd are not necessarilililily yy to bbbe attributed to the
RFF.
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
22
5. Schwartz PJ, Priori SG, Spazzolini C, Moss AJ, Vincent GM, Napolitano C, et al. Genotype-phenotype correlation in the long-QT syndrome: gene-specific triggers for life-threatening arrhythmias. Circulation. 2001;103:89–95.
6. Brink PA, Crotti L, Corfield V, Goosen A, Durrheim G, Hedley P, et al. Phenotypic variability and unusual clinical severity of congenital long-QT syndrome in a founder population. Circulation. 2005;112:2602–2610.
7. Priori SG, Napolitano C, Schwartz PJ. Low penetrance in the long-QT syndrome: clinical impact. Circulation. 1999;99:529–533.
8. Vincent GM, Timothy KW, Leppert M, Keating M. The spectrum of symptoms and QT intervals in carriers of the gene for the long-QT syndrome. N Engl J Med. 1992;327:846–852.
9. Crotti L, Monti MC, Insolia R, Peljto A, Goosen A, Brink PA, et al. NOS1AP is a genetic modifier of the long-QT syndrome. Circulation. 2009;120:1657–1663.
10. Giudicessi JR, Ackerman MJ. Determinants of incomplete penetrance and variable expressivity in heritable cardiac arrhythmia syndromes. Transl Res. 2013;161:1–14.
11. George Jr. AL. Common genetic variants in sudden cardiac death. Heart Rhythm. 2009;6:S3–9.
12. Kolder ICRM, Tanck MWT, Bezzina CR. Common genetic variation modulating cardiac ECG parameters and susceptibility to sudden cardiac death. J Mol Cell Cardiol. 2012;52:620–629.
13. Marx SO, Kurokawa J, Reiken S, Motoike H, D’Armiento J, Marks AR, et al. Requirement of a macromolecular signaling complex for beta adrenergic receptor modulation of the KCNQ1-KCNE1 potassium channel. Science. 2002;295:496–499.
14. Barhanin J, Lesage F, Guillemare E, Fink M, Lazdunski M, Romey G. KvLQT1 and IsK (minK) proteins associate to form the IKS cardiac potassium current. Nature. 1996;384:78–80.
15. Sanguinetti MC, Curran ME, Zou A, Shen J, Spector PS, Atkinson DL, et al. Coassembly of K(V)LQT1 and minK (IsK) proteins to form cardiac I(Ks) potassium channel. Nature.1996;384:80–83.
16. Chen L, Kurokawa J, Kass RS. Phosphorylation of the A-kinase-anchoring protein Yotiao contributes to protein kinase A regulation of a heart potassium channel. J Biol Chem.2005;280:31347–31352.
17. Kurokawa J, Motoike HK, Rao J, Kass RS. Regulatory actions of the A-kinase anchoring protein Yotiao on a heart potassium channel downstream of PKA phosphorylation. Proc Natl Acad Sci U S A. 2004;101:16374–16378.
NOS1AP is a geneneneneti.
nce anananandddd vavavavaririririabababableleleley
ICRM, Tanck MWT, Bezzina CR. Common genetic variation modulating cardiam 2
y in n n heheheheriririritatatat blblblble cacacardiac arrhythmia syndromomomees. Transl Res. 202020013;161:1–14.
JJJrJ . AL. Commmooon ggeeenetttticicici varrriiai nnntss innn ssuddeden cacardrdrdiaac deddeath. HeHeHeaart Rhhhhyyythmhmhm. 2000009
ICRM, Tanck MMMMWTWTWTWT, BeBeBeB zzzzzzinininina a a CRCRCRCR. CoCoCoC mmmmmmonononon gggennnneteteticicicic vvvarararariaiaiaiattitiononon mmmmodooo ulating cardiameters and susceptptptiiibibibibillil tytyty to suddddddden cardididiiac ddddeath. J JJ MMMollll CCCC lellllllMM CCCCarddidd ol.llll 2012;;52:62
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
23
18. Chen L, Marquardt ML, Tester DJ, Sampson KJ, Ackerman MJ, Kass RS. Mutation of an A-kinase-anchoring protein causes long-QT syndrome. Proc Natl Acad Sci U S A.2007;104:20990–20995.
19. De Jager T, Corbett CH, Badenhorst JC, Brink PA, Corfield VA. Evidence of a long QT founder gene with varying phenotypic expression in South African families. J Med Genet.1996;33:567–573.
20. Hedley PL, Durrheim GA, Hendricks F, Goosen A, Jespersgaard C, Stovring B, et al. Long QT syndrome in South Afr Cardiovasc. J. Afr. 2013; 24:231–237.
21. Corfield VA, Moolman JC, Martell R, Brink PA. Polymerase chain reaction-based detection of MN blood group-specific sequences in the human genome. Transfusion. 1993;33:119–124.
22. Bazett HC. An analysis of time relations of electrocardiograms. Heart. 1920;7:353–370.
23. De Bakker PIW, Yelensky R, Pe’er I, Gabriel SB, Daly MJ, Altshuler D. Efficiency and power in genetic association studies. Nat Genet. 2005;37:1217–1223.
24. Wigginton JE, Abecasis GR. PEDSTATS: descriptive statistics, graphics and quality assessment for gene mapping data. Bioinformatics. 2005;21:3445–3447.
25. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265.
26. Sobel E, Lange K. Descent graphs in pedigree analysis: applications to haplotyping, location scores, and marker-sharing statistics. Am J Hum Genet. 1996;58:1323–1337.
27. Campbell H, Rudan I. Interpretation of genetic association studies in complex disease. Pharmacogenomics J. 2002;2:349–360.
28. Balding DJ, Bishop M, Cannings C, eds. Handbook of Statistical Genetics. New York, NY: John Wiley & Sons; 2007: 927.
29. Crotti L, Spazzolini C, Schwartz PJ, Shimizu W, Denjoy I, Schulze-Bahr E, et al. The common long-QT syndrome mutation KCNQ1/A341V causes unusually severe clinical manifestations in patients with different ethnic backgrounds: toward a mutation-specific risk stratification. Circulation. 2007;116:2366–2375.
30. Heijman J, Spätjens RLHMG, Seyen SRM, Lentink V, Kuijpers HJH, Boulet IR, et al.Dominant-negative control of cAMP-dependent IKs upregulation in human long-QT syndrome type 1. Circ Res. 2012;110:211–219.
31. Barsheshet A, Goldenberg I, O-Uchi J, Moss AJ, Jons C, Shimizu W, et al. Mutations in Cytoplasmic Loops of the KCNQ1 Channel and the Risk of Life-Threatening Events:
eeeearararartttt. 1919191920202020;7;7;7;7:3:3:3:353535353–3–3–3–370707070
uler DDDD EfEfEfEffififificicicicienenenencycycycy aaande
t
m
E Lange K Descent graphs in pedigree anal sis: applications to haplot ping loc
eneteteticicici aaassssssococociatititiooon studies. Nat Genet. 2000050505;37:1217–1223..
toooonn n n JE, Abecaasiiis GGGRRR.R PPEDEDEE STTTTATATATSS: dddeescrripptiveve sssstatatat ttit sttticcs, ggraaaphhhiicici s annnnddd d quququalitttty yfffooro gggene mamamapppiiing dddata. BBBiB oinfnfforrrmmattticcs. 202005555;2221:::34445––3444777.
JC, Fry B, Malllllellel r r r J,JJJ DDDalalala y y MJMJMJMJ. .. HaHaHaHaplplpllovovovovieieew:w:w: aaaanananan lyyyysisisis s ss anananand d d vivivivisususuualalalizizizzatatatatioii n of LD and mapspp . Bioinfformatiics. 2020202 0505055;2;21:11 2626262 33–3 2626262 55.5
E LL KK DDe t hhs ii didi ll isi lili iti to hh llot ipi llo by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
24
Implications for Mutation- -Blocker Therapy in Type 1 Long-QT Syndrome. Circulation. 2012;125:1988–1996.
32. Aydin A, Bähring S, Dahm S, Guenther UP, Uhlmann R, Busjahn A, et al. Single nucleotide polymorphism map of five long-QT genes. J. Mol. Med. 2005;83:159–165.
33. Sedlacek K, Stark K, Cunha SR, Pfeufer A, Weber S, Berger I, et al. Common Genetic Variants in ANK2 Modulate QT Interval: Results From the KORA Study. Circ. Cardiovasc. Genet. 2008;1:93–99.
34. Newton-Cheh C, Eijgelsheim M, Rice KM, de Bakker PIW, Yin X, Estrada K, et al.Common variants at ten loci influence QT interval duration in the QTGEN Study. Nat Genet.2009;41:399–406.
35. Pfeufer A, Sanna S, Arking DE, Müller M, Gateva V, Fuchsberger C, et al. Common variants at ten loci modulate the QT interval duration in the QTSCD Study. Nat Genet. 2009;41:407–414.
36. Arking DE, Pfeufer A, Post W, Kao WHL, Newton-Cheh C, Ikeda M, et al. A common genetic variant in the NOS1 regulator NOS1AP modulates cardiac repolarization. Nat Genet.2006;38:644–651.
37. Eijgelsheim M, Newton-Cheh C, Aarnoudse ALHJ, van Noord C, Witteman JCM, Hofman A, et al. Genetic variation in NOS1AP is associated with sudden cardiac death: evidence from the Rotterdam Study. Hum. Mol. Genet. 2009;18:4213–4218.
38. Tobin MD, Kähönen M, Braund P, Nieminen T, Hajat C, Tomaszewski M, et al. Gender and effects of a common genetic variant in the NOS1 regulator NOS1AP on cardiac repolarization in 3761 individuals from two independent populations. Int. J. Epidemiol. 2008;37:1132–1141.
39. Arking DE, Khera A, Xing C, Kao WHL, Post W, Boerwinkle E, et al. Multiple Independent Genetic Factors at NOS1AP Modulate the QT Interval in a Multi-Ethnic Population. Plos One.2009;4:e4333.
40. Tomás M, Napolitano C, De Giuli L, Bloise R, Subirana I, Malovini A, et al. Polymorphisms in the NOS1AP Gene Modulate QT Interval Duration and Risk of Arrhythmias in the Long QT Syndrome. J. Am. Coll. Cardiol. 2010;55:2745–2752.
41. Jamshidi Y, Nolte IM, Spector TD, Snieder H. Novel genes for QTc interval. How much heritability is explained, and how much is left to find? Genome Med. 2010;2:35.
42. Goldenberg I, Horr S, Moss AJ, Lopes CM, Barsheshet A, McNitt S, et al. Risk for life-threatening cardiac events in patients with genotype-confirmed long-QT syndrome and normal-range corrected QT intervals. J Am Coll Cardiol. 2011;57:51–59.
r C, et al. Common n n n vvavvaaat ttt GeGeGeGenenenenett.tt 2222000000009;9;9;9;44441:4:4:4:407070707
a M etetetet aaaallll AAAA ccccomomomommomommoni e
4
h fenetic variation in NOS1AP is associated with sudden cardiac death: evidence fra
M ecommon genetic ariant in the NOS1 reg lator NOS1AP on cardiac repolari at
iant tt ininini tttheheheh NNNNOSOSOSS1 regulator NOS1AP mododododulu ates cardiac reppolooo arization. Nat Gene44–4–4–66656 1.
heieieim mm M, Newewewttonnn-CCChhheh C,CC, Aararnoonouudseee AALHLHJ, vvan NNoooordd C,, WiWWitttttemee ananan JCMCMCM, HoHH fenetiiiicc vavavavariririr atioonn iiinin NNNOOOOS1AAAPPPP iiiis assssococociaiaaattted ddd iwiwiiththth ssssuduududdededen caacacardrddiiaiacc deatatatath:h:h: eviviiddedencncee frffam Study. Hum.mmm MMMMolololo . GeGeGeG neneneet.t.t.t 222000000009;9;9;;18181818:4:44212213–3–3–3–4244 18181818..
MMMD, Kähöönenenenen nn n M,M,MM, BBBBrarraraununnund ddd P,P,PP, NNNNieieeiemimmiminenenenen nn T,T,TT, HHHHajaaa atatat CCCC,, ToToToTomamamamaszsszszewewewewskskkski i ii M,M,MM, etetetet al. Gendeeetiic iiant ii hth NONOS1S1 llat NNOSOS1A1APP didi lol ii at by guest on M
ay 14, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
25
43. Terrenoire C, Houslay MD, Baillie GS, Kass RS. The cardiac IKs potassium channel macromolecular complex includes the phosphodiesterase PDE4D3. J. Biol. Chem.2009;284:9140–9146.
44. Li Y, Chen L, Kass RS, Dessauer CW. The A-kinase anchoring protein Yotiao facilitates complex formation between adenylyl cyclase type 9 and the IKs potassium channel in heart. J. Biol. Chem. 2012;287:29815–29824.
45. Kao WHL, Arking DE, Post W, Rea TD, Sotoodehnia N, Prineas RJ, et al. Genetic Variations in Nitric Oxide Synthase 1 Adaptor Protein Are Associated With Sudden Cardiac Death in US White Community-Based Populations. Circulation. 2009;119:940–951.
46. Duchatelet S, Crotti L, Peat RA, Denjoy I, Itoh H, Berthet M, et al. Identification of a KCNQ1 Polymorphism Acting as a Protective Modifier Against Arrhythmic Risk in Long-QT Syndrome. Circ Cardiovasc Genet. 2013;6:354–361.
47. Schwartz PJ. Sudden cardiac death, founder populations, and mushrooms: What is the link with gold mines and modifier genes? Heart Rhythm. 2011;8:548–550.
Table 1. LQTS founder population data summary for individuals in whom at least one SNP was
genotyped
At least one SNP genotyped * N Non Carriers(n=181)
Mutation Carriers(n=168)
Gender, male (%) 349 95 (52) 75 (45)
BB, yes/total (%) 234 5/68 (7) 97/166 (58)
Age in years at which ECG was analyzed, mean (SD) 273 34 (20.2) 34.9 (22.9)
BB starting age in years, mean (SD) 95† 32.3 (22.5) 17.2 (17.7)
QTc, median (IQR) 273 401 (383-418) 483 (463-512)
Cardiac events, yes (%) 349 0 (0) 122 (73)
Age first cardiac event, median (range) 115 N/A 6 (2-22)
Age last followed up about CE, median (IQR) 161 N/A 40 (20-57)
BB indicates beta-blocker therapy; SD, standard deviation; IQR, interquantile range; CE, cardiac events*All SNPs were genotyped in 176 non carriers and 162 mutation carriers† There were two additional individuals for whom it was known that beta-blocker therapy was started after the age of 20 years. However, the age at which the treatment started was unknown.
hrooms: WWWWhahhh t ttt iiiis tttthehehehe l
Q PQTS founder popopopopupupulalalalatititionononon dddatatatataaa a sususuummmmmmmmararara y yy fofofor r r inininindidd vivivividududuualalalalss inininn wwwhohohoh m mmm atatata least one SNP
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
26
Table 2. LQTS severity score used to evaluate effects on disease severity
Score value Satisfying condition
0 No cardiac events
1 1-3 syncope episodes
2 More than 3 syncope episodes; no cardiac arrest or sudden cardiac death
3 Cardiac arrest and/or sudden cardiac death
Table 3. Minor allele and haplotype frequencies in the South African LQTS founder population
rs11772585 rs7808587 rs2282972 rs2961024
Number genotyped 346 345 345 342
Minor Allele T G T G
MAF 0.10 0.46 0.44 0.29
Haplotypes
HF=0.443 C A T T
HF=0.282 C G C G
HF=0.171 C G C T
HF=0.098 T A C T
MAF indicates minor allele frequency; LD, linkage disequilibrium; CI, confidence interval; HF, haplotype frequency
t or sudden cardiaacccc d
diac dedededeatatatathhhh
M lMiiinor allelele anananand dd hahahhaplplplplotoo ypypypypeee e frffrfreqeqequeueeuencncnncieieieiessss ininnin ttttheheehe SSouououthtthth AAAAfrfrfricciccanananan LLLLQTQTQTQTS SSS fofofoounuunundedd r popull
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
27
Table 4. LQTS modifying effects of AKAP9 SNPs and haplotypes
Modifying effect SNP/Haplotype* P value Effect size Effect description
QTc
% change per 10 years
rs2961024 0.006 1% per G vs. TT
CGCG 0.012 1% per CGCG vs. 0:CGCG
CE % change in risk
rs11772585 0.002 118% CT CC
rs7808587 0.046 74%
rs2282972 0.001 53%
TACT 0.006 100%
CATT 0.022 35% per CATT
Severity change in score
rs11772585 0.025 0.58
TACT 0.032 0.54 ACT
haplotype absent; 1:, one copy of the haplotype present; CE, cardiac events* Haplotype order: rs11772585, rs7808587, rs2282972, rs2961024
Figure Legends:
Figure 1. Number of individuals entered into analyses based on available information. NC=non-
carriers; MC=mutation-carriers; CE=cardiac event; BB=beta-blocker treatment
Figure 2. Age-dependent change in QTc for rs2961024 genotypes by gender and A341V
mutation status. Modelled curves indicating the 1% per 10 year increase with each G allele
rn A
nnts
CATT 0.020200 2 35% per CATT
rityyyy n score
rs11117777725252585 0.0025 0.5888
TATATAACTCTCTC 0.0.0.03030322 22 0.0.0.0 545454 A
hhhhapapaplolololotytytypepepepe aaabsbsbsbsenenentttt;;;; 1:1:1:,,, onoonone e ee coccopypypypy oooof ff f thththt e eee hahahhaplplplplotototo ypyy e preseenntsts by guest on M
ay 14, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000580
28
relative to the TT genotype for a female mutation carrier (A), male mutation carrier (B), female
non-carrier (C) and male non-carrier (D). The net result was no significant effect for GT, and a
1% increase for the GG genotype. The curves were identical for the different categories;
however, they start at different baseline QTc values.
Figure 3. Genotype specific risk of cardiac events for AKAP9 . Kaplan-Meier cumulative event-
free survival of rs11772585, rs7808587 and rs2282972 genotypes.
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
Schwartz, Paul A. Brink, Johanna C. Moolman-Smook and Valerie A. CorfieldCarin P. de Villiers, Lize van der Merwe, Lia Crotti, Althea Goosen, Alfred L. George, Jr., Peter J.
AKAP9 is a Genetic Modifier of Congenital Long-QT Syndrome Type 1
Print ISSN: 1942-325X. Online ISSN: 1942-3268 Copyright © 2014 American Heart Association, Inc. All rights reserved.
TX 75231is published by the American Heart Association, 7272 Greenville Avenue, Dallas,Circulation: Cardiovascular Genetics
published online August 2, 2014;Circ Cardiovasc Genet.
http://circgenetics.ahajournals.org/content/early/2014/08/02/CIRCGENETICS.113.000580World Wide Web at:
The online version of this article, along with updated information and services, is located on the
http://circgenetics.ahajournals.org/content/suppl/2014/08/02/CIRCGENETICS.113.000580.DC1Data Supplement (unedited) at:
http://circgenetics.ahajournals.org//subscriptions/
is online at: Circulation: Cardiovascular Genetics Information about subscribing to Subscriptions:
http://www.lww.com/reprints Information about reprints can be found online at: Reprints:
document. Permissions and Rights Question and Answer this process is available in the
located, click Request Permissions in the middle column of the Web page under Services. Further information aboutnot the Editorial Office. Once the online version of the published article for which permission is being requested is
can be obtained via RightsLink, a service of the Copyright Clearance Center,Circulation: Cardiovascular Genetics Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:
by guest on May 14, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
SUPPLEMENTAL MATERIAL
Statistical analysis overview: Each outcome was modeled as a function of two genetic terms.
The first was additive allelic (counting the number of minor alleles) and the second was a
heterozygote flag. After detecting significant associations, we tested whether a dominant or
recessive model provided a better fit. We also tested genetic interactions with each of A341V
mutation, age and gender on QTc and with gender on cardiac events. Each model was adjusted
for appropriate confounders.