The E ec s o Pha macological Compounds on Bea Ra e
Va ia ions in Human Long QT-Synd ome Ca diomyocy es
Jukka Kuusela
1,2
&Jiyeong Kim
3
&Esa Räsänen
3
&Ka iina Aal o-Se älä
1,2,4,5
Published online: 19 Sep embe 2016
#The Au ho (s) 2016. This a icle is published wi h open access a Sp inge link.com
Abs ac Heal hy human hea a e luc ua es o e ime show-
ing long- ange ac al co ela ions. In con as , a ious ca diac
diseases and no mal aging show he b eakdown o ac al
complexi y. Recen ly, i was shown ha human induced plu-
ipo en s em cell-de i ed ca diomyocy es (hiPSC-CMs) in-
insically exhibi ac al beha io as in humans. He e, we
in es iga ed he ac al complexi y o hiPSC-de i ed long
QT-ca diomyocy es (LQT-CMs). We eco ded ex acellula
ield po en ials om hiPSC-CMs a baseline and unde he
e ec o a ious compounds including β-blocke bisop olol,
ML277, a speci ic and po en I
Ks
cu en ac i a o , as well as
JNJ303, a speci ic I
Ks
blocke . F om he peak- o-peak-in e -
als, we de e mined he long- ange ac al co ela ions by
using de ended luc ua ion analysis. Elec ophysiologically,
he baseline co ec ed ield po en ial du a ions (cFPDs) we e
mo e p olonged in LQT-CMs han in wild ype (WT)-CMs.
Bisop olol did no ha e signi ican e ec s o he cFPD in
any CMs. ML277 sho ened cFPD in a dose-dependen ash-
ionby11%and5–11 % in WT- and LQT-CMs, espec i ely.
JNJ303 p olonged cFPD in a dose-dependen ashion by 22 %
and 7–13 % in WT- and LQT-CMs, espec i ely. A baseline,
all CMs showed ac al co ela ions as de e mined by sho -
e m scaling exponen α. Howe e , in all CMs, he αwas
inc eased when pha macological compounds we e applied in-
dica ing o b eakdown o ac al complexi y. These indings
sugges ha he in insic mechanisms con ibu ing o he ac-
al complexi y a e no al e ed in LQT-CMs. The modula ion o
I
Ks
channel and β1-ad eno ecep o s by pha macological com-
pounds may a ec he ac al complexi y o he hiPSC-CMs.
Keywo ds Induced plu ipo en s em cell .Long QT
synd ome .Ca diomyocy es .Mul ielec ode a ay .
De ended luc ua ion analysis .F ac als .Nonlinea dynamics
In oduc ion
Hea a e dynamics ha e been p e iously analyzed using
con en ional linea and newe nonlinea me hods in heal hy
and diseased s a es ( o e iew see Pe kiömäki [1]). The non-
linea indices o he RR a ia ions include ac als, which a e
geome ically de ined as objec s composed o subuni s (and
sub-subuni s) ha sus ain sel -simila i y on di e en measu e-
men scales [2]. The cha ac e is ic ea u e o ac als is 1/ -like
luc ua ions, which has been shown o be p esen in a heal hy
human hea bea [3–6]. Such luc ua ions possess long- ange
co ela ions indica i e o a memo y e ec , which means ha
he hea a e is no only ela ed o immedia ely p eceding
alue bu also o alues in he emo e pas [2].
Elec onic supplemen a y ma e ial The online e sion o his a icle
(doi:10.1007/s12015-016-9686-0) con ains supplemen a y ma e ial,
which is a ailable o au ho ized use s.
*Ka iina Aal o-Se älä
ka iina.aal o-se ala@u a. i
Jukka Kuusela
jukka.kuusela@u a. i
Esa Räsänen
esa. asanen@ u . i
1
Ins i u e o Biomedical Technology, Uni e si y o Tampe e,
Tampe e, Finland
2
BioMediTech, Tampe e, Finland
3
Depa men o Physics, Tampe e Uni e si y o Technology,
Tampe e, Finland
4
School o Medicine, Uni e si y o Tampe e and Tampe e Uni e si y
Hospi al, Finn-Medi 5, Bioka u 12, FI-33014 Tampe e, Finland
5
Hea Hospi al, Tampe e Uni e si y Hospi al, Tampe e, Finland
S em Cell Re and Rep (2016) 12:698–707
DOI 10.1007/s12015-016-9686-0
In he hea a e ime se ies, a b eakdown o 1/ -like luc-
ua ions may lead in o ei he comple ely unco ela ed andom-
ness (whi e noise) o in o o al co ela ion esembling andom
walk (B ownian mo ion). A numbe o s udies ha e shown
ha in a ious ca diac disease s a es (e.g. conges i e hea
ailu e, myoca dial in a c ion) he ac al 1/ -like long- ange
co ela ions o he hea bea b eakdown p oducing mo e un-
co ela ed andomness [7–13]. Fu he mo e, no mal aging has
been associa ed wi h he b eakdown o ac al complexi y
p oducing mo e o al co ela ion in he hea bea dynamics
[6,14,15]. Thus, i is hough ha he b eakdown o ac al
complexi y in hea a e dynamics may cause he sys em o be
less adap able and less esponsi e o unp edic able s imuli and
s esses inc easing suscep ibili y o inju y and illness [6,16].
The biological o igin o he ac al-like beha io has no
been ye ully es ablished and is somewha con adic o y. In
he hea , he o igin o he ac al-like beha io is hough o
esul om complex in e ac ion be ween agal and sympa he -
ic inpu s o he au onomous ne ous sys em [17,18].
Expe imen al obse a ions in humans ha e suppo ed his no-
ion [19–23], al hough opposing iews ha e been p esen ed
[24]. Howe e , e idence sugges s ha monolaye cul u es o
a en icula ca diomyocy es and human induced plu ipo en
s em cell-de i ed ca diomyocy es (hiPSC-CMs) lacking au o-
nomic con ol exhibi ac al-like complexi y, which indica es
ha he au onomous ne ous sys em inpu is no necessa y o
ac al dynamics [25,26]. I has been u he demons a ed
ha single ca diomyocy es also exhibi ac al-like complexi y
and ha in acellula Ca
2+
-cycling mechanisms con ibu e o
he ac al complexi y o he single ca diomyocy es [27].
Long QT synd ome (LQTS) is a po en ially se e e li e-
h ea ening a hy hmic ca diac disease cha ac e ized by
p olonged QT in e al on he elec oca diog am. LQTS is asso-
cia ed wi h o sades de poin es, a special ype o en icula achy-
ca dia, which may degene a e in o en icula ib illa ion and
cause sudden ca diac dea h [28]. Inhe i ed o ms o LQTS a e a
esul o mu a ions in he ca diac ion channel coding genes. One
o he mos common LQTS genes is KCNQ1, which encodes he
α-subuni o he ol age-ga ed po assium channel esponsible o
he slow delayed ec i ie K
+
cu en (I
Ks
)[29,30]. In Finland, he
p e alence o gene mu a ions associa ed wi h LQTS is high
(0.4 % o Finnish popula ion), which has been explained o be
caused by ou ounde mu a ions wi h one o hem being C-
e minal G589D missense mu a ion in KCNQ1 gene [31,32].
The hiPSCs ep esen excellen esea ch ool o s udy he pa ho-
physiology o inhe i ed ca diac diseases [33–41]. Un il now, he
ac al dynamics o hiPSC-LQT-CMs ha e no been in es iga ed.
He e, we u ilize pa ien -speci ic LQTS disease model [41]
o s udy he ac al dynamics in he symp oma ic and asymp-
oma ic LQTS ype 1 (LQT1)-speci ic CMs ca ying Finnish
ounde mu a ion G589D. The aim o his s udy was o in es-
iga e he ac al dynamics o LQT-CMs in unmedica ed and
medica ed condi ions as compa ed o heal hy con ol and
unde he e ec o a ious compounds a ec ing ca diac
ac ion po en ial.
Ma e ial & Me hods
ECG Reco dings and Human Induced Plu ipo en S em
Cell Gene a ion
The s udy was app o ed by he e hical commi ee o
Pi kanmaa Hospi al Dis ic (R08070). Pa icipan s who
olun ee ed o he s udy ga e hei consen . The ECGs we e
eco ded using MARS-Hol e om a heal hy indi idual,
asymp oma ic LQT-mu a ion ca ie and symp oma ic LQT-
pa ien . The LQT-pa ien s a e on bisop olol medica ion. The
heal hy indi idual has no medica ion. Human iPSCs we e
gene a ed as desc ibed ea lie [42]. The LQT1-speci ic
hiPSCs we e de i ed om pa ien s’skin ib oblas s ca ying
G589D missense mu a ion in KCNQ1 [41,43].
Pa ien Cha ac e is ics
Skin biopsies wi h LQT1 mu a ion we e ob ained om a
symp oma ic 41-yea old emale pa ien (QTc in e al,
456 ms) and om an asymp oma ic 28-yea old emale mu a-
ion ca ie (QTc in e al, 428 ms). Bo h ca y he KCNQ1
G589D mu a ion. The symp oma ic 41-yea old pa ien had
expe ienced seizu es, episodes o unconsciousness and synco-
pe be o e β-blocke (bisop olol) medica ion. The heal hy con-
ol human iPS cells we e de i ed om skin ib oblas s o a
heal hy 55-yea old emale (QTc in e al, 406 ms) [44].
Human Induced Plu ipo en S em Cell Cul u e,
Di e en ia ion and Cha ac e iza ion
Human iPS cells we e cul u ed and di e en ia ed as p e ious-
ly desc ibed [43]. All he hiPSC lines (UTA.04602.WT,
UTA.00208.LQT1, UTA.00211.LQT1, UTA.00303.LQT1
and UTA.00313.LQT1) and he di e en ia ed CMs om hem
ha e been p e iously cha ac e ized elsewhe e [41,43,44].
Mul ielec ode A ay Reco dings and Da a Analysis
In his s udy, 30–45 days old hiPSC-CMs we e used o he
expe imen s. Spon aneously bea ing ca diomyocy e clus e s
we e manually dissec ed and pla ed on 6-well MEAs (6-well
MEA 200/30iR-Ti- c , Mul ichannel Sys ems, Reu lingen,
Ge many), which we e i s coa ed wi h e al bo ine se um
(FBS, In i ogen) o 30 min a oom empe a u e and hen
wi h 0.1 % gela ine (Sigma Ald ich) o 1 h a oom empe -
a u e. The ca diomyocy e clus e s we e cul u ed in EB-medi-
um: KO-DMEM wi h 20 % FBS, NEAA, Glu amax and pen-
icillin/s ep omycin. The expe imen s we e conduc ed in 5 %
S em Cell Re and Rep (2016) 12:698–707 699
FBS con aining EB-medium (5 % EB-medium). Be o e d ug
es s, he ield po en ials o igina ing om he spon aneously
bea ing ca diomyocy es we e eco ded o 30 min (baseline) a
+37 °C wi h he MEA pla o m (MEA2100-2 × 60–2,
Mul ichannel Sys ems, Reu lingen, Ge many) using 10 kHz
sampling equency and MC_Rack (Mul ichannel Sys ems,
Reu lingen, Ge many) so wa e. A e he 30-min baseline
measu emen , he MEA pla e was pu on +37 °C he mal pla e
(Tokai Hi , Japan) o keeping he empe a u e s able while
adding d ugs. The ollowing d ugs we e used in he s udy:
Bisop olol (Sigma-Ald ich), ML277 (Toc is Bioscience) and
JNJ303 (Toc is Bioscience). The d ugs we e dissol ed in di-
me hyl sul oxide (DMSO, Sigma-Ald ich) acco ding o man-
u ac u e ’s ins uc ions. The bisop olol concen a ions we e
chosen based on i s he apeu ic blood se um concen a ion
ange [45]. Fo bisop olol, 260 nM (uppe limi o he he a-
peu ic se um concen a ion) and 520 nM ( wice he uppe limi
o he he apeu ic concen a ion) concen a ions we e used.
ML277 (I
Ks
channel ac i a o ) concen a ions o 1 μMand
2μM we e chosen based on p e ious epo s [46,47]. The
concen a ions o I
Ks
blocke JNJ303 (300 nM and 1000 nM)
we e chosen based on ou p e ious s udy [41]. A e d ug
addi ion, he MEA pla es we e incuba ed o 5 min a
+37 °C he mal pla e be o e he 30-min measu emen ( i s
d ug concen a ion). A e his, we added mo e d ugs o he
cells (second d ug concen a ion) and simila ly as be o e, e-
co ded he ield po en ials o 30 min. We also conduc ed
ehicle con ol expe imen s wi h simila p o ocol as desc ibed
abo e, wi h he excep ion ha no d ugs bu only DMSO
(0.1 %) was added o he cells. The eco ding ime o baseline
and o each d ug concen a ion was 30 min. The da a ob ain-
ed om MEA was analyzed by ou in-house de eloped
Ca dioMDA so wa e, which a e ages ield po en ial signals
using c oss co ela ion algo i hms [48]. F om each eco ding,
he las 2 min om he 30-min eco ding we e chosen o
a e aging he ield po en ial signals. Fo de e mining he ield
po en ial du a ion (FPD), he onse was de e mined as he
beginning o depola izing peak and he o se as T
max
o he
epola izing wa e. The Baze ’s and F ide icia’s o mula we e
used o calcula e he co ec ed ield po en ial du a ion (cFPD).
De ended Fluc ua ion Analysis
We applied de ended luc ua ion analysis (DFA) o he RR-
in e als. DFA is one o he mos used ime-se ies analysis
me hods ha gi es a eliable es ima e o he exis ence and
he cha ac e is ics o long- ange co ela ions in he da a [7].
DFA has been applied in a ious ields o science ang-
ing om physiological signals such as hea bea and
gai [49] o, o example, musical hy hms [50,51],
ain all s a is ics [52], s uc u al p ope ies o DNA
[53], and elec onic quan um anspo [54].
A de ailed desc ip ion o DFA can be ound in he abo e-
lis ed e e ences and he e we only summa ize he main s eps.
Fi s , we ake he peak- o-peak in e als o he hiPSC-CM o
ECG da a se and sub ac he mean alue, so ha we conside
he luc ua ions a ound he mean. Nex , we in eg a e he se ies
by aking a cumula i e sum o he luc ua ions. The ime axis
is hen di ided in o non-o e lapping windows, and in each
window, a leas -squa es line ( end) is i o he da a. The
oo -mean-squa e de ia ions om he end ( esiduals) a e a -
e aged h ough he whole da a se . This p ocedu e is epea ed
o di e en window sizes. As a esul , we ge ela ionship
be ween he window size and he a e age luc ua ion (wi hin
ha window size). The slope in his plo in a log-log scale
co esponds o he DFA exponen α. Whi e noise wi h no
co ela ion be ween consecu i e alues has α=0.5,whe eas
B ownian mo ion wi h s ongly co ela ed alues gene a ed
by unco ela ed consecu i e inc emen s has α=1.5.Ingen-
e al, in e media e p edic abili y be ween hese limi s wi h
0.5 < α≤1.5 indica es long- ange ( ac al) co ela ions.
An i-co ela ions a e cha ac e ized by −0.5 < α< 0.5. The
special case o pink noise α= 1 co esponds o 1/ beha io .
S a is ical Analyses
F om MEA da a, one-way ANOVA ollowed by Dunne ’spos
hoc es was pe o med o es di e ences in baseline alues
be ween con ol and LQT cell lines (IBM e sion 22.0; SPSS
Inc., Chicago, USA). I da ase s did no mee ANOVA equi e-
men s (no mal dis ibu ion, equal a iances), nonpa ame ic es
Mann-Whi ney U ollowed by alpha co ec ion was employed
o compa e con ol cell line and LQT cell lines. In d ug expe -
imen s, he baseline and he e ec o d ug concen a ions we e
compa ed using pai ed sample - es . Simila ly, i da ase s did
no mee he equi emen s o - es , nonpa ame ic Wilcoxon
es was employed. F om DFA da a, pai ed sample - es was
employed o de e mine he s a is ical di e ences be ween base-
line and each d ug concen a ion. The p< 0.05 was conside ed
s a is ically signi ican . The le els o signi icance a e ep esen -
ed as (*) p < 0.05, (**) p<0.01and(***)p< 0.001. The cFPD
p olonga ions o >10 % a e conside ed physiologically signi -
ican . The da a is p esen ed as mean ± s anda d de ia ion (SD).
Resul s
The E ec o Pha macological Compounds
o he hiPSC-CM Clus e s
The 208.LQT1 and 211.LQT1 we e de i ed om symp om-
a ic pa ien whe eas 303.LQT1 and 313.LQT1 we e de i ed
om asymp oma ic mu a ion ca ie . The baseline cFPDs
(F ide icia’s co ec ion) we e signi ican ly mo e p olonged
in hiPSC-LQT-CMs han in heal hy WT-CMs (Fig. 1b, d, ).
700 S em Cell Re and Rep (2016) 12:698–707
Simila esul s we e ob ained when Baze ’s FPD co ec ion
was used (Supplemen al Fig. 1). In con as , we did no ind
any signi ican di e ences in he bea ing a es (BRs) be ween
WT- and LQT-CMs (Fig. 1a, c, e).
β-Blocke Bisop olol The bisop olol did no cause any majo
di e ences in he BRs o he hiPSC-CMs (Fig. 1a). We no ed a
sligh ly inc easing end o BR wi h inc easing bisop olol con-
cen a ion in he hiPSC-CMs om asymp oma ic mu a ion ca -
ie (15 % o 303.LQT1 and 19 % o 313.LQT1). As o he
cFPD, bisop olol caused only mild cFPD p olonga ion in he
WT-CMs (3–7 %) a he concen a ion ange o 260–520 nM.
Simila cFPD p olonga ions we e seen in LQT-CMs om
symp oma ic pa ien (2–6 % o 208.LQT1 and 5–7 % o
211.LQT1) and om asymp oma ic mu a ion ca ie (5–10 %
o 303.LQT1 and 5–8 % o 313.LQT1) a he concen a ion
ange o 260-520 nM (Fig. 1b). The ep esen a i e bisop olol
aces o WT- and LQT-CMs a e illus a ed in Fig. 2a.
I
Ks
Ac i a o ML277 The ML277 caused signi ican inc ease
in he BRs o WT and 211.LQT1, 9 % and 14 %, espec i ely
(Fig. 1c). O e all, he end o inc easing BR wi h inc easing
ML277 concen a ion was obse ed al hough s a is ical sig-
ni icance was no ound o all he cell lines. As expec ed, he
ML277 sho ened he cFPD in a dose-dependen manne , al-
hough he cFPD sho ening was ela i ely mild (Fig. 1d). In
he WT, he sho ening o cFPD was 11 % whe eas in he
LQTs i anged om 5 o 11 % a he concen a ion ange o
1-2 μM. The ep esen a i e ML277 aces o WT- and LQT-
CMs a e illus a ed in Fig. 2b.
I
Ks
Blocke JNJ303 Blocking he I
Ks
channel wi h JNJ303
did no signi ican ly change he BRs o he WT- o LQT-CMs
excep in 313.LQT1, in which he 23 % inc ease in he BR
was obse ed (Fig. 1e). JNJ303 showed a dose-dependen
cFPD p olonga ion in he WT- and LQT-CMs (Fig. 1 ). The
mos inc emen in cFPD was seen in he WT (22 %) whe eas
in he LQTs, he cFPD inc ease was a ound 7–13 % a he
highes concen a ion (1000 nM). The ep esen a i e JNJ303
aces o WT- and LQT-CMs a e illus a ed in Fig. 2c.
De ended Fluc ua ion Analysis (DFA) o Heal hy
Con ol- and LQT-Speci ic Ca diomyocy es and ECG
Da a
We i s de e mined he ac al scaling exponen α om hu-
man subjec s who pa icipa ed o he s udy (n= 3). Resul s a e
shown in Table 1. Nex , we de e mined he scaling exponen α
om hiPSC-CMs de i ed om he same human subjec s pa -
icipa ing o he s udy. In Fig. 3we show he DFA esul s o
he peak- o-peak luc ua ions o heal hy con ol- (WT) and
LQT-speci ic CMs when exposed o bisop olol (Fig. 3a),
Fig. 1 The e ec s o a ious compounds o he human induced
plu ipo en s em cell (hiPSC)-de i ed ca diomyocy es’ ield po en ial
pa ame e s. The uppe ow depic s he bea ing a e (BR) and he lowe
ow F ide icia-co ec ed ield po en ial du a ion (cFPD). The baseline
cFPDs o long QT-speci ic ca diomyocy es we e signi ican ly mo e
p olonged han in heal hy wild ype-ca diomyocy es. The as e isks on
op o he ba s depic he s a is ical signi icance o mean BR o cFPD
change compa ed o baseline alues. Signi icance le els a e indica ed by
(*) p< 0.05, (**) p< 0.01 and (***) p< 0.001, espec i ely
S em Cell Re and Rep (2016) 12:698–707 701
ML277 (Fig. 3b), o JNJ303 (Fig. 3c). The i s ba in all he
subplo s co esponds o he baseline wi h ze o d ug con-
cen a ion. The LQT da a is g ouped acco ding o symp-
oma ic (208.LQT1, 211.LQT1) and asymp oma ic cases
(303.LQT1, 313.LQT1) shown in he middle and in he
igh column, espec i ely.
We ound ha all he a e age baseline alues o he DFA
scaling exponen a e close o one, i.e., α~0.9–1.1. The e was
Fig. 2 The ep esen a i e aces o he human induced plu ipo en s em
cell (hiPSC)-de i ed wild ype- and long QT (LQT)-speci ic
ca diomyocy es unde he e ec o a ious compounds. The LQTs a e
g ouped acco ding o symp oma ic (208.LQT1, 211.LQT1) and
asymp oma ic cases (303.LQT1, 313.LQT1) shown in he middle and
in he igh column, espec i ely. a)β-blocke Bisop olol, b)I
Ks
ac i a o ML277 c)I
Ks
blocke JNJ303. No ice, how he T
max
o he
epola iza ion wa e is ela i ely unchanged be ween baseline and
di e en bisop olol concen a ions depic ing ma ginal e ec o he β-
blocke o he co ec ed ield po en ial du a ion (cFPD). Howe e , I
Ks
ac i a o ML277 shows clea sho ening o cFPD seen by shi in he
T
max
. Simila ly, bu con e sely o ML277, I
Ks
blocke JNJ303 shows
clea p olonga ion o cFPD assessed by he shi in he T
max
Table 1 The pa ien
cha ac e is ics and de ended
luc ua ion analysis om ECG
da a
Heal hy indi idual LQT1-pa ien (symp oma ic) LQT1-pa ien (asymp oma ic)
Age 55 41 28
Medica ion - Bisop olol Bisop olol
QTc (ms) 406 456 428
α1.21 ± 0.03 1.08 ± 0.02 1.07 ± 0.02
702 S em Cell Re and Rep (2016) 12:698–707

no no able di e ence in he baseline α alues be ween
he WT and he LQTs. This is in line wi h ou ECG
da a, which shows αo 1.08, 1.07 and 1.21 o symp-
oma ic and asymp oma ic LQT pa ien as well as o
heal hy indi idual, espec i ely (Table 1). Howe e , all
he pha macological compounds lead o an inc ease in
he DFA scaling exponen α owa d B ownian mo ion
(α= 1.5) when compa ed o he baseline (Fig. 3). The
end is isible bo h in he WT and in he LQTs. In
mos cases a u he inc ease in he concen a ion does
no a ec α. When compa ing all he α alues o WT
and LQTs, we did no ind any s a is ical di e ence
be ween he g oups. We also conduc ed ehicle con ol
expe imen s in which no pha macological compounds
bu only DMSO (0.1 %) was added o he CMs. We
did no obse e any signi ican changes in he αscaling ex-
ponen du ing hese eco dings. The αwas 1.024 ± 0.13 a
baseline, 1.00 ± 0.12 ( i s DMSO addi ion) and 1.03 ± 0.13
(second DMSO addi ion) (n= 6). Thus, he change seen in α
scaling exponen wi h a ious pha macological compounds
e lec he in insic p ope ies o he pha macological com-
pounds hemsel es.
Fig. 3 The de ended luc ua ion analysis (DFA) αscaling exponen s o
he human induced plu ipo en s em cell (hiPSC)-de i ed wild ype- and
long QT (LQT)-speci ic ca diomyocy es. The LQTs a e g ouped
acco ding o symp oma ic (208.LQT1, 211.LQT1) and asymp oma ic
cases (303.LQT1, 313.LQT1) shown in he middle and in he igh
column, espec i ely. a) ( ow) β-blocke Bisop olol, b)I
Ks
ac i a o
ML277 c)I
Ks
blocke JNJ303. A baseline, he αscaling exponen s a e
close o 1 in all hiPSC-CMs. The addi ion o a ious compounds o he
hiPSC-CMs inc eased he αscaling exponen close o B ownian mo ion
(α= 1.5). The as e isks on op o he ba s depic he s a is ical signi icance
o mean αscaling exponen change compa ed o baseline alues.
Signi icance le els a e indica ed by (*) p < 0.05, (**) p < 0.01 and
(***) p < 0.001, espec i ely
S em Cell Re and Rep (2016) 12:698–707 703
Discussion
In human hea s, he ac al dynamics is hough o esul om
complex in e ac ion be ween agal and sympa he ic inpu s o
au onomous ne ous sys em [17,18]. Howe e , e idence sug-
ges ha he hiPSC-CMs also exhibi ac al-like complexi y
indica ing in insic mechanisms o he CMs con ibu ing o
ac al dynamics [26,27]. This s udy u he suppo s he con-
cep ha heal hy hiPSC-CMs lacking au onomous ne ous
sys em inpu exhibi ac al-like complexi y a baseline condi-
ion. We also obse ed ha he un ea ed (wi hou any pha -
macological in e en ion) hiPSC-LQT-CMs also appea o ex-
hibi ac al-like complexi y a baseline condi ion. This esul
is in line wi h he p e ious s udy in un ea ed pa ien s wi h
congeni al LQTS [55]. Thus, ou esul s u he expand his
concep sugges ing ha he in insic mechanisms con ibu ing
o he ac al-like complexi y is no al e ed in hiPSC-LQT-
CMs. Howe e , i is o no e ha he esul s ob ained we e om
2 LQTS pa ien s only and u he s udies on a la ge popula-
ion would be needed o de ini ely conclude he ma e .
The Elec ophysiological P ope ies o he hiPSC-CMs
The baseline beha io o hiPSC-CMs is in line wi h ou p e-
ious s udies [41,43]. Resul s ob ained om his s udy clea ly
show ha he cFPDs, co ec ed wi h ei he Baze ’so
F ide icia’s o mulae, a e mo e p olonged in LQT-CMs han
in heal hy WT-CMs. We also in es iga ed he heal hy and
LQT-speci ic hiPSC-CMs unde he e ec o β-blocke
bisop olol and pha macological compounds a ec ing speci i-
cally o ca diac ion channel I
Ks
. O e all, he d ug e ec s in
LQT-CMs appea ed o be o simila magni ude o hose o
WT-CMs, simila o ou p e ious indings [43].
The E ec o Bisop olol o he F ac al Complexi y o
hiPSC-CMs The β-blocke s a e he s anda d and cu en ly,
he only ea men o choice o LQTS pa ien s [56]. He e, in
o de o alid compa isons be ween cellula and whole hea
da a, we chose o s udy bisop olol in hiPSC-CMs because he
LQTS pa ien s olun ee ed o his s udy we e on bisop olol
medica ion. Bisop olol, a β1-ad eno ecep o selec i e β-
blocke , did no ha e physiological signi icance o he ield
po en ial pa ame e s (>10 % cFPD p olonga ion) in any o he
hiPSC-CMs a clinically ele an concen a ion. I is impo an
o no ice ha we did no ac i a e β- ecep o s wi h β-agonis s
p io o β-blocke applica ion. Howe e , he acu e applica ion
o bisop olol inc eased he αscaling exponen owa d
B ownian mo ion (α= 1.5) in all hiPSC-CMs a he uppe
limi o he apeu ic concen a ion (260 nM). A 520 nM, he
e ec was a enua ed. Al hough he baseline αda a om
hiPSC-LQT-CMs co ela ed wi h he ECG da a (α~ 1.0),
bisop olol αda a did no co ela e wi h he ECG da a. The
disc epancy be ween cellula and hea αda a emains ye
unclea bu i may be a ibu ed o e.g. acu e applica ion o
he bisop olol o in he absence o au onomous ne ous sys-
em in hiPSC-LQT-CMs, which may a ec o ac al complex-
i y upon d ug applica ion. P e ious human s udy has shown
ha he ac al complexi y (α~ 1.0) was unal e ed in pa ien s
wi h congeni al LQTS ea ed o un ea ed wi h β-blocke s
[55]. Thus, he β-blocke s may no ha e signi ican e ec on
he αo he whole LQTS hea s. On he o he hand, he e is
e idence om he human s udies ha he β-blocke ea men
imp o es ac al dynamics o he hea in ad anced conges i e
hea ailu e pa ien s by inc easing he αscaling exponen
du ing 1–3-mon h he apy pe iod [21–23]. I is comple ely
unknown whe he such di e ences esul om he di e en
disease s a es, he ype o β-blocke used o ea men pe iods
among o he hings. Howe e , his s udy clea ly shows ha
al hough he ield po en ial pa ame e s we e no signi ican ly
changed, he in insic mechanisms con ibu ing o ac al-like
complexi y we e al e ed du ing acu e β-blocke ea men . On
he o he hand, u he s udies would be needed o answe
wha he e ec would be in long- e m (ch onic) si ua ion and
wha he mechanism behind o his phenomenon is.
Thesigni icanceo he ac al complexi y esembling
B ownian mo ion in he hea is unclea . The loss o ac ali y
owa d whi e noise (α= 0.5) has been ound o p edispose o
se e e li e- h ea ening a hy hmias and ca diac dea h [9,10,
13]. Howe e , e y li le is known when he ac al dynamics
o he hea become mo e co ela ed esembling B ownian
mo ion (α= 1.5). Such phenomenon has been obse ed in
heal hy elde ly subjec s implying dec ease o ac al complex-
i y wi h age [6,14,15] and, al hough no ye p o en, i has
been sugges ed ha such sys em would be mo e suscep ible o
inju y and illness in he elde ly [6,16].
The E ec o I
Ks
A ec ing Pha macological Compounds
o he F ac al Complexi y o hiPSC-CMs Nex , we in es i-
ga ed he e ec s o he o he wo pha macological compounds
(ML277 and JNJ303) o hiPSC-CMs. The ion channel ac i a-
o ML277 has been shown o augmen speci ically I
Ks
cu en
and sho en he ac ion po en ial du a ion in bo h heal hy and
LQT-CMs [46,47]. Simila ly, in his s udy, he ML277 sho -
ened he cFPD al hough he e ec was ela i ely mild in hese
hiPSC-CM clus e s a 1-2 μM concen a ion. Simila ly, as was
seen wi h bisop olol, also ML277 esul ed in he dys unc ion
o egula o y mechanisms con ibu ing o ac al complexi y
as he αscaling exponen was inc eased owa d B ownian
mo ion in all hiPSC-CMs a 1 μM. Howe e , a 2 μM a small
di e ence was seen ha in he WT-CMs αexponen de-
c eased o baseline le el whe eas he LQT-CMs αexponen s
con inued o inc ease om 1 μM, which was mos p ominen -
ly seen in CMs de i ed om symp oma ic pa ien (208.LQT1
and 00,211.LQT1). Fu he mo e, he e ec o JNJ303, a po-
en and speci ic I
Ks
blocke [57] known o e oke o sades de
poin es, was assessed in hiPSC-CMs. The WT-CMs appea ed
o be mo e sensi i e o JNJ303 han LQT-CMs as measu ed by
704 S em Cell Re and Rep (2016) 12:698–707
cFPD p olonga ion. This esul is in line wi h ou p e ious
indings om single WT- and LQT-CMs measu ed wi h pa ch
clamp using he same cell lines [41]. Al hough no ye p o en
in he expe imen al se ing, his indica es ha he I
Ks
cu en is
diminished in he LQT1-CMs ha bo ing G589D missense
mu a ion compa ed o heal hy WT-CMs. Con adic o y o
ou expec a ions, also he JNJ303 esul ed in simila al e -
a ions o ac al complexi y in hiPSC-CMs as ML277 by in-
c easing he αscaling exponen . Taken oge he , hese e i-
dence imply ha he modula ion o he ion channel gene a ing
I
Ks
cu en wi h pha macological compounds may esul in he
dys unc ion o he in insic mechanisms con ibu ing o ac al
complexi y. Indeed, hy hmic ion channel ac i a ion and inac-
i a ion in pacemake cells has been hough o con ibu e o
he ul adian hy hmici y in addi ion o spon aneous Ca
2+
-cy-
cling [58]. P e ious s udy has shown ha he dis up ion o
in acellula Ca
2+
handling causes al e a ions in he scaling
exponen α, mos ly by dec easing i [27]. He e, we ha e
shown ha a baseline, he hy hmici y o hiPSC-CMs is
sus ained as de e mined by long- ange ac al co ela ions.
Fu he mo e, he modula ion o he ion channel gene a ing
I
Ks
cu en wi h speci ic pha macological compounds dis up s
CM hy hmici y and ac al complexi y.
Po en ial Limi a ion o he S udy
In his s udy, only he acu e e ec s o he a ious compounds
o he ac al-like complexi y we e assessed in hiPSC-CMs.
Fu he s udies would be needed o show he ch onic, long-
e m e ec s o he compounds. Mo eo e , he mechanisms
behind he al e a ions in ac al complexi y we e no s udied.
We did no in es iga e in de ail he e ec o β-ad ene gic
agonis o he long- ange ac al co ela ions o hiPSC-CMs.
Also, wi h his model we could no ake in o accoun he
con inuous exposu e o sympa he ic and pa asympa he ic
s imuli a ec ing CMs in i o. This s udy does no answe o
wha would be he implica ions o al e ed ac ali y owa d
B ownian mo ion in hiPSC-CMs in e ms o long- e m heal h,
adap abili y and esponsi eness o unp edic able s imuli.
Conclusions
In conclusion, he hiPSC-LQT-CMs appea o exhibi ac al-
like complexi y a baseline condi ion sugges ing ha he in in-
sic mechanisms o LQT-CMs con ibu ing o he ac al com-
plexi y a e no al e ed. Al hough he e ec s o a ious com-
pounds o he ield po en ial pa ame e s we e as expec ed, he
ac al-like complexi y o he hiPSC-CMs was signi ican ly al-
e ed in heal hy as well as LQT-speci ic CMs. No signi ican
di e ences in he αscaling exponen we e ound be ween WT-
and LQT-CMs. These indings may sugges ha he ca diac ion
channel gene a ing I
Ks
cu en as well as he modula ion o β1-
ad eno ecep o s by β-blocke bisop olol may con ibu e o he
ac al-like complexi y o he hiPSC-CMs.
Acknowledgmen s We kindly hank Ma kus Haponen and Henna
Venäläinen o hei excellen echnical expe ise wi h he s em cell cul-
u e and ca diac di e en ia ion. We also kindly hank Anna Ki iaho o
gene a ing he iPS cell lines, The END-2 cells we e a kind gi om
Ch is ine Mumme y. On he compu a ional side, we hank Mika
Sa ilah i, and Pe u Luukko o assis ance wi h DFA analysis. This
s udy was unded by g an s om he Finnish Funding Agency o
Technology and Inno a ion (TEKES), Finnish Founda ion o
Ca dio ascula Resea ch, Finnish Cul u al Founda ion and he
Academy o Finland.
Compliance wi h E hical S anda ds
Au ho Con ibu ions JK designed and pe o med MEA expe imen s,
analyzed MEA da a, pe o med s a is ical analyses and w o e he manu-
sc ip , JKim pe o med de ended luc ua ion analysis, analyzed he nu-
me ical esul s and w o e he manusc ip , ER designed expe imen s, pe -
o med de ended luc ua ion analysis and w o e he manusc ip , KA
concei ed he s udy, designed expe imen s and w o e he manusc ip .
All au ho s ead and app o ed he inal manusc ip .
Con lic o In e es The au ho s decla e no po en ial con lic s o
in e es s.
Open Access This a icle is dis ibu ed unde he e ms o he C ea i e
Commons A ibu ion 4.0 In e na ional License (h p://
c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use,
dis ibu ion, and ep oduc ion in any medium, p o ided you gi e
app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link
o he C ea i e Commons license, and indica e i changes we e made.
Re e ences
1. Pe kiömäki, J. S. (2011). Hea a e a iabili y and non-linea dy-
namics in isk s a i ica ion. F on ie s in Physiology, 2,81.
2. Goldbe ge , A. L. (1996). Non-linea dynamics o clinicians: chaos
heo y, ac als, and complexi y a he bedside. Lance , 347(9011),
1312–1314.
3. Kobayashi, M., & Musha, T. (1982). 1/ Fluc ua ion o Hea bea
Pe iod. I.E.E.E. T ansac ions on Bio-Medical Enginee ing, BME-
29(6), 456–457.
4. Peng, C., Mie us, J., Hausdo , J. M., Ha lin, S., S anley, H. E., &
Goldbe ge , A. L. (1993). Long- ange an ico ela ions and non-
Gaussian beha io o he hea bea . Physical Re iew Le e s,
70(9), 1343–1346.
5. Pikkujämsä, S. M., Mäkikallio, T. H., Juhani Ai aksinen, K. E., &
Huiku i, H. V. (2001). De e minan s and in e indi idual a ia ion o
R-R in e al dynamics in heal hy middle-aged subjec s. Ame ican
Jou nal o Physiology. Hea and Ci cula o y Physiology, 280(3
49–3), H1400–H1406.
6. Goldbe ge , A. L., Ama al, L. A. N., Hausdo , J. M., I ano , P. C.,
Peng, C., & S anley, H. E. (2002). F ac al dynamics in physiology:
al e a ions wi h disease and aging. P oceedings o he Na ional
Academy o Sciences o he Uni ed S a es o Ame ica, 99(Suppl.
1), 2466–2472.
S em Cell Re and Rep (2016) 12:698–707 705
7. Peng, C., Ha lin, S., S anley, H. E., & Goldbe ge , A. L. (1995).
Quan i ica ion o scaling exponen s and c osso e phenomena in
nons a iona y hea bea ime se ies. Chaos, 5(1), 82–87.
8. Ho, K. K. L., Moody, G. B., Peng, C., e al. (1997). P edic ing
su i al in hea ailu e case and con ol subjec s by use o ully
au oma ed me hods o de i ing nonlinea and con en ional indices
o hea a e dynamics. Ci cula ion, 96(3), 842–848.
9. Mäkikallio, T. H., Seppänen, T., Ai aksinen, K. E. J., e al. (1997).
Dynamic analysis o hea a e may p edic subsequen en icula
achyca dia a e myoca dial in a c ion. The Ame ican Jou nal o
Ca diology, 80(6), 779–783.
10. Mäkikallio, T. H., Kois inen, J., Jo daens, L., e al. (1999). Hea
a e dynamics be o e spon aneous onse o en icula ib illa ion in
pa ien s wi h healed myoca dial in a c s. The Ame ican Jou nal o
Ca diology, 83(6), 880–884.
11. Vikman, S., Mäkikallio, T. H., Yli-Mäy y, S., e al. (1999). Al e ed
complexi y and co ela ion p ope ies o R-R in e al dynamics
be o e he spon aneous onse o pa oxysmal a ial ib illa ion.
Ci cula ion, 100(20), 2079–2084.
12. Huiku i, H. V., Pou iainen, A., Mäkikallio, T. H., e al. (1999).
Dynamic beha io and au onomic egula ion o ec opic a ial pace-
make s. Ci cula ion, 100(13), 1416–1422.
13. Huiku i, H. V., Mäkikallio, T. H., Peng, C., Goldbe ge , A. L.,
Hin ze, U., & Mølle , M. (2000). F ac al co ela ion p ope ies o
R-R in e al dynamics and mo ali y in pa ien s wi h dep essed le
en icula unc ion a e an acu e myoca dial in a c ion.
Ci cula ion, 101(1), 47–53.
14. Iyenga N, Peng C-, Mo in R, Goldbe ge AL, Lipsi z LA. (1996).
Age- ela ed al e a ions in he ac al scaling o ca diac in e bea
in e al dynamics. Ame ican Jou nal o Physiology. Regula o y,
In eg a i e and Compa a i e Physiology, 271(4 40–4):R1078–84.
15. Pikkujämsä, S. M., Mäkikallio, T. H., Sou ande , L. B., e al.
(1999). Ca diac in e bea in e al dynamics om childhood o se-
nescence: compa ison o con en ional and new measu es based on
ac als and chaos heo y. Ci cula ion, 100(4), 393–399.
16. Lipsi z, L. A., & Goldbe ge , A. L. (1992). Loss o ‘complexi y’
and aging: po en ial applica ions o ac als and chaos heo y o
senescence. JAMA, 267(13), 1806–1809.
17. Goldbe ge , A. L., & Wes , B. J. (1987). F ac als in physiology and
medicine. The Yale Jou nal o Biology and Medicine, 60(5), 421–435.
18. Wes , B. J., & Goldbe ge , A. L. (1987). Physiology in ac al di-
mensions. Ame ican Scien is , 75(4), 354–365.
19. Tulppo, M. P., Mäkikallio, T. H., Seppänen, T., e al. (2001). E ec s
o pha macological ad ene gic and agal modula ion on ac al
hea a e dynamics. Clinical Physiology, 21(5), 515–523.
20. Tulppo, M. P., Ki iniemi, A. M., Hau ala, A. J., e al. (2005).
Physiological backg ound o he loss o ac al hea a e dynamics.
Ci cula ion, 112(3), 314–319.
21. Lin, L., Lin, J., C-, D., Lai, L., Tseng, Y., & Huang, S. K. S. (2001).
Re e sal o de e io a ed ac al beha io o hea a e a iabili y by
be a-blocke he apy in pa ien s wi h ad anced conges i e hea
ailu e. Jou nal o Ca dio ascula Elec ophysiology, 12(1), 26–32.
22. Ridha, M., Makikallio, T. H., Lope a, G., e al. (2002). E ec s o
ca edilol on hea a e dynamics in pa ien s wi h conges i e hea
ailu e. Annals o Nonin asi e Elec oca diology, 7(2), 133–138.
23. Chiu, K., Chan, H., Chu, S., & Lin, T. (2007). Ca edilol can
es o e he mul i ac al p ope ies o hea bea dynamics in pa ien s
wi h ad anced conges i e hea ailu e. Au onomic Neu oscience
Basic Clinical, 132(1–2), 76–80.
24. Tan, C. O., Cohen, M. A., Eckbe g, D. L., & Taylo , J. A. (2009).
F ac al p ope ies o human hea pe iod a iabili y: physiological
and me hodological implica ions. The Jou nal o Physiology,
587(15), 3929–3941.
25. Kuce a, J. P., Heuschkel, M. O., Renaud, P., & Roh , S. (2000).
Powe -law beha io o bea - a e a iabili y in monolaye cul u es o
neona al a en icula myocy es. Ci cula ion Resea ch, 86(11),
1140–1145.
26. Mandel, Y., Weissman, A., Schick, R., e al. (2012). Human emb y-
onic and induced plu ipo en s em cell-de i ed ca diomyocy es ex-
hibi bea a e a iabili y and powe -law beha io . Ci cula ion,
125(7), 883–893.
27. Ben-A i, M., Schick, R., Ba ad, L., e al. (2014). F om bea a e
a iabili y in induced plu ipo en s em cell-de i ed pacemake cells
o hea a e a iabili y in human subjec s. Hea Rhy hm, 11(10),
1808–1818.
28. Schwa z, P. J., C o i, L., & Insolia, R. (2012). Long-QTsynd ome
om gene ics o managemen . Ci cula ion. A hy hmia and
Elec ophysiology, 5(4), 868–877.
29. Ba hanin, J., Lesage, F., Guillema e, E., Fink, M.,
Lazdunski, M., & Romey, G. (1996). K(V)LQT1 and lsK
(minK) p o eins associa e o o m he I(Ks) ca diac po assi-
um cu en . Na u e, 384(6604), 78–80.
30. Sanguine i, M. C., Cu an, M. E., Zou, A., e al. (1996).
Coassembly o K(V)LQT1 and minK (IsK) p o eins o o m ca di-
ac I(Ks) po assium channel. Na u e, 384(6604), 80–83.
31. Ma jamaa, A., Salomaa, V., New on-Cheh, C., e al. (2009). High
p e alence o ou long QT synd ome ounde mu a ions in he
Finnish popula ion. Annals o Medicine, 41(3), 234–240.
32. Piippo, K., Swan, H., Pas e nack, M., e al. (2001). A ounde
mu a ion o he po assium channel KCNQ1 in long QT synd ome:
implica ions o es ima ion o disease p e alence and molecula
diagnos ics. Jou nal o he Ame ican College o Ca diology,
37(2), 562–568.
33. Egashi a, T., Yuasa, S., Suzuki, T., e al. (2012). Disease cha ac e -
iza ion using LQTS-speci ic induced plu ipo en s em cells.
Ca dio ascula Resea ch, 95(4), 419–429.
34. Bellin, M., Casini, S., Da is, R. P., e al. (2013). Isogenic human
plu ipo en s em cell pai s e eal he ole o a KCNH2 mu a ion in
long-QT synd ome. The EMBO Jou nal, 32(24), 3161–3175.
35. Mo e i, A., Bellin, M., Welling, A., e al. (2010). Pa ien -speci ic
induced plu ipo en s em-cell models o long-QT synd ome. New
England Jou nal o Medicine, 363(15), 1397–1409.
36. I zhaki, I., Maizels, L., Hube , I., e al. (2011). Modelling he long
QT synd ome wi h induced plu ipo en s em cells. Na u e,
471(7337), 225–229.
37. Lah i, A. L., Kujala, V. J., Chapman, H., e al. (2012). Model o
long QT synd ome ype 2 using human iPS cells demons a es
a hy hmogenic cha ac e is ics in cell cul u e. Disease Models &
Mechanisms, 5(2), 220–230.
38. Ma sa, E., Rajamohan, D., Dick, E., e al. (2011). D ug e alua ion
in ca diomyocy es de i ed om human induced plu ipo en s em
cells ca ying a long QT synd ome ype 2 mu a ion. Eu opean
Hea Jou nal, 32(8), 952–962.
39. Ma, D., Wei, H., Zhao, Y., e al. (2013). Modeling ype 3 long QT
synd ome wi h ca diomyocy es de i ed om pa ien -speci ic in-
duced plu ipo en s em cells. In e na ional Jou nal o Ca diology,
168(6), 5277–5286.
40. Yazawa, M., Hsueh, B., Jia, X., e al. (2011). Using induced plu ip-
o en s em cells o in es iga e ca diac pheno ypes in Timo hy syn-
d ome. Na u e, 471(7337), 230–236.
41. Ki iaho, A. L., Ahola, A., La sson, K., e al. (2015). Dis inc elec-
ophysiological and mechanical bea ing pheno ypes o long QT
synd ome ype 1-speci ic ca diomyocy es ca ying di e en mu a-
ions. IJC Hea & Vascula u e, 8,19–31.
42. Takahashi, K., Tanabe, K., Ohnuki, M., e al. (2007). Induc ion o
plu ipo en s em cells om adul human ib oblas s by de ined ac-
o s. Cell, 131(5), 861–872.
43. Kuusela, J., Kujala, V. J., Ki iaho, A., e al. (2016). E ec s o
ca dioac i e d ugs on human induced plu ipo en s em cell de i ed
long QT synd ome ca diomyocy es. Sp inge Plus, 5(1), 1–13.
706 S em Cell Re and Rep (2016) 12:698–707