Ea ly a e depola iza ions p omo e ansmu al een y in ischemic
human en icles wi h educed epola iza ion ese e
Sa a Du a
a
,
1
, Ana Minchol
e
a
,
1
, E nes o Zacu
b
, T. Alexande Quinn
c
, Pe e Tagga
d
,
Blanca Rod iguez
a
,
*
a
Depa men o Compu e Science, BHF Cen e o Resea ch Excellence, Uni e si y o Ox o d, Ox o d, Uni ed Kingdom
b
Ins i u e o Biomedical Enginee ing, Uni e si y o Ox o d, Ox o d, Uni ed Kingdom
c
Depa men o Physiology and Biophysics, Dalhousie Uni e si y, Canada
d
Ins i u e o Ca dio ascula Science, Uni e si y College London, Ba s Hea Cen e, Uni ed Kingdom
a icle in o
A icle his o y:
Recei ed 1 Oc obe 2015
Recei ed in e ised o m
25 Janua y 2016
Accep ed 29 Janua y 2016
A ailable online 2 Feb ua y 2016
Keywo ds:
Ischemic hea disease
Compu e -based model
Ven icula a hy hmia
Po assium channels
Repola iza ion
abs ac
Aims: Acu e ischemia is a majo cause o sudden a hy hmic dea h, u he p omo ed by po assium
cu en blocke s. Mac o- een y a ound he ischemic egion and ea ly a e depola iza ions (EADs) caused
by elec o onic cu en ha e been sugges ed as po en ial mechanisms in animal and isola ed cell s udies.
Howe e , en icula and human-specific a hy hmia mechanisms and hei modula ion by epola iza-
ion ese e emain unclea . The goal o his pape is o un a el mul iscale mechanisms unde lying he
modula ion o a hy hmic isk by po assium cu en (I
K
) block in human en icles wi h acu e egional
ischemia.
Me hods and esul s: A human en icula biophysically-de ailed model, wi h acu e egional ischemia is
cons uc ed by in eg a ing expe imen al knowledge on he elec ophysiological ionic al e a ions caused
by co ona y occlusion. A hy hmic isk is e alua ed by de e mining he ulne able window (VW) o
een y ollowing ec opy a he ischemic bo de zone. Mac o- een y a ound he ischemic egion is he
main een an mechanism in he ischemic human en icle wi h inc eased epola iza ion ese e due o
he ATP-sensi i e po assium cu en (I
K(ATP)
) ac i a ion. P olonga ion o e ac o iness by 4% caused by
30% I
K
educ ion coun e ac s he es ablishmen o mac o- een y and educes he VW o een y (by
23.5%). Howe e , a u he dec ease in epola iza ion ese e (50% I
K
educ ion) is less an i-a hy hmic
despi e u he p olonga ion o e ac o iness. This is due o he es ablishmen o ansmu al een y
enabled by elec o onically- igge ed EADs in he ischemic bo de zone. EADs a e p oduced by L- ype
calcium cu en (I
CaL
) eac i a ion due o p olonged low ampli ude elec o onic cu en injec ed du ing
he epola iza ion phase.
Conclusions: Elec o onically- igge ed EADs a e iden ified as a po en ial mechanism acili a ing in a-
mu al een y in a egionally-ischemic human en icles model wi h educed epola iza ion ese e.
©2016 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY license
(h p://c ea i ecommons.o g/licenses/by/4.0/).
1. In oduc ion
Acu e ischemia associa ed wi h co ona y hea disease is a
leading cause o sudden a hy hmic dea h. P o-a hy hmic mech-
anisms o acu e ischemia ha e been ex ensi ely in es iga ed,
al hough o en in animal models a he single cell o issue le el
a he han in human en icles. Seminal s udies by Janse e al. in
pig and dog hea s highligh he complexi y o he p o-a hy hmic
and spa io- empo ally dynamic subs a e in acu e ischemia (Janse
and Wi , 1989; Janse e al., 1980, 1979). He e ogenei y in exci -
abili y and epola iza ion p ope ies ac oss he bo de be ween he
no mal and ischemic a eas leads o he es ablishmen o een y
a ound he ischemic egion ollowing ec opic exci a ion. The same
s udies also showed in amu al een y in ce ain cases (high-
ligh ing he po en ial a iabili y in he mechanisms), eco ded wi h
mic oelec odes on and 4 and 8 mm below he epica dial su ace in
pig hea s a ec ed by le an e io descending co ona y a e y oc-
clusion. Howe e , he mechanisms ha de e mine een y o ma-
ion and in amu al pa e ns in acu e ischemia in he h ee-
*Co esponding au ho . Depa men o Compu e Science, Uni e si y o Ox o d,
Pa ks Road, OX1 3QD, Ox o d, Uni ed Kingdom.
E-mail add ess: [email p o ec ed]x.ac.uk (B. Rod iguez).
1
Equal con ibu ion fi s au ho s.
Con en s lis s a ailable a ScienceDi ec
P og ess in Biophysics and Molecula Biology
jou nal homepage: www.else ie .com/loca e/pbiomolbio
h p://dx.doi.o g/10.1016/j.pbiomolbio.2016.01.008
0079-6107/©2016 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/).
P og ess in Biophysics and Molecula Biology 120 (2016) 236e248
dimensional human hea emain unclea , due o low esolu ion o
in amu al eco dings.
Elec o onic cu en flowing due o g adien s in memb ane
po en ial in he e ogeneous issue has been hypo hesized as a
po en ially-impo an p o-a hy hmic ac o in acu e ischemia
(Co onel e al., 1991; Janse e al., 1980). One po en ial mechanism is
ha elec o onic cu en flow could p omo e he occu ence o
ea ly a e depola iza ions (EADs) in acu e ischemia (Kuma and
Joyne , 1994; Ve ke k e al., 2000). Single cell s udies showed
EADs igge ed by elec o onic cu en in human ca diomyocy es
om ailing hea s, bu only when cells in phase 3 epola iza ion
we e coupled o e y high po en ials co esponding o pla eau le el
(such as 0 mV) (Ve ke k e al., 2000). Neighbo ing cells would
he e o e ha e o expe ience e y di e en memb ane po en ial
le els o his o occu . Fu he mo e, elec o onic in e ac ions in
well coupled issue (as is s ill he case in he ea ly phase o acu e
ischemia) a e known o dec ease he p opensi y o EAD o ma ion
in issue (Weiss e al., 2010; Xie e al., 2010), e en i hey can be
igge ed in single cells as in Ve ke k e al. (2000). The e o e, i
emains unclea wha condi ions a e needed o EADs o be ig-
ge ed by elec o onic cu en , as well as he ole o EAD o ma ion
in he egionally ischemic human hea .
P o-a hy hmic mechanisms including bo h een y and EAD
o ma ion s ongly depend on he balance o cu en s du ing
epola iza ion, and he specific deg ee o edundancy in epola-
izing cu en s, e med epola iza ion ese e, a ailable. This is
a ec ed by a a ie y o causes including in e -subjec di e ences in
ion channel densi y, d ug e ec s and mu a ions. Indi iduals wi h
educed epola iza ion ese e a e known o be a highe
a hy hmic isk, pa icula ly in he p esence o a p o-a hy hmic
subs a e such as ischemia due o co ona y hea disease as
shown in clinical s udies (S einbe g e al., 2014; Waldo e al., 1996).
Fu he mo e, clinical s udies show ha in e en ions ha educe
epola izing cu en s such as class III d ugs esul in inc eased isk
o longe - e m hospi aliza ion and dea h in pa ien s wi h co ona y
hea disease (S einbe g e al., 2014). S udies, such as he Su i al
Wi h ORal D-So alol (SWORD) ial, also demons a e ha admin-
is a ion o he hERG (human E he -
a-go-go Rela ed Gene) blocke
so alol is associa ed wi h inc eased mo ali y in pa ien s su e ing
om ecen myoca dial in a c ion, ela ed o ischemic hea disease
(Waldo e al., 1996). How changes in he deg ee o po assium cu -
en a ailable du ing epola iza ion al e s he ischemic human
hea o inc ease a hy hmic isk emains unclea .
In es iga ing he p o-a hy hmic subs a e o acu e ischemia
and i s al e a ion by po assium block expe imen ally is e y chal-
lenging. O pa icula di ficul y is mapping he h ee-dimensional
olume o he hea wi h enough spa io- empo al esolu ion o
in es iga e complex and i egula in amu al p opaga ion pa e ns
h ough he en icula wall. Mic oelec ode eco dings and op ical
mapping ha e bo h been used in animal s udies and hey bo h
su e om a di e en se o limi a ions. The o me exhibi low
esolu ion, whe eas he la e p o ides eco dings only o he issue
su ace and wi h pho on sca e ing a e aging e ec s (Bishop e al.,
2007, p. 2007). Acu e ischemia is u he complica ed by he ac
ha he elec ophysiological changes induced a e dynamic in ime
and ep oducibili y o expe imen al findings is comp omised.
In his s udy, we in es iga e he mechanisms o inc eased
a hy hmic isk in he he e ogeneous subs a e caused by acu e
myoca dial ischemia o di e en deg ees o educed epola iza ion
ese e using an ana omically-based biophysically-de ailed human
en icles model. We hypo hesize ha in human en icles wi h low
epola iza ion ese e, g adien s in memb ane po en ial ac oss he
ischemic bo de zone p omo e he occu ence o elec o onically-
igge ed EADs ha con ibu e o he es ablishmen o ansmu al
mic o- een y. By analyzing high spa io- esolu ion simula ion da a,
we un a el he h ee-dimensional mechanisms o EAD o ma ion
and p opaga ion dynamics in he acu ely-ischemic human en i-
cles, hei modula ion by epola iza ion ese e and hei po en ial
con ibu ion o inc eased a hy hmic isk.
2. Me hods
2.1. Human en icula model in acu e egional ischemia
An ana omically-based mul iscale model o egionally-ischemic
human en icles was de eloped based on ex ensi e expe imen al
eco dings om (Ca melie , 1999; Co onel e al., 1991, 1988; Du e
e al., 1970; Ed a dsson e al., 1980; Glukho e al., 2010; Lee e al.,
1981; Spec o e al., 1996; Su on e al., 2000; Tagga e al., 2001;
Wilensky e al., 1986), including biophysically-de ailed memb ane
kine ics, ischemia induced he e ogenei ies and fibe o ien a ion
(based on he S ee e me hod (S ee e e al., 1969)). Human
en icle memb ane kine ics we e simula ed wi h he en Tussche
2006 ac ion po en ial (AP) model (TP06) (Tussche and Panfilo ,
2006) wi h an added ATP-sensi i e po assium cu en (I
K(ATP)
)
(Michailo a e al., 2005). The TP06 has been ex ensi ely used and
shown o be sui able o simula e een an a hy hmias as well as
ischemic elec ophysiological al e a ions in human (Du a e al.,
2013; Kazbano e al., 2014). An expanded e sion o he
me hods is a ailable as Supplemen al Ma e ial.
2.2. He e ogeneous elec ophysiological subs a e in he ischemic
egion
Size and loca ion o ischemic egions exhibi g ea a iabili y.
We ep esen ed he ischemic egion in ou model as a ec ing ~40%
o he le en icle, which is wi hin he ange o 13e72% epo ed
by Lee e al. (1981).(Fig. 1A) and co esponds o g ea es likelihood
o en icula fib illa ion (Cu is, 1998). The ischemic egion was
modeled in ag eemen wi h expe imen al findings by Co onel e al.
(1991); Rod íguez e al. (2006); Wilensky e al. (1986) including: (i)
he ischemic cen al zone (ICZ); (ii) he bo de zone (BZ), (iii) and
he endoca dial BZ (a laye o endoca dial issue wi h access o
oxygen and nu ien s om blood in he en icula ca i y) (Co onel
e al., 1988; Wilensky e al., 1986).
In he ICZ, ionic al e a ions caused by acu e ischemia we e
in oduced (Co onel e al., 1991; Rod íguez e al., 2006; Wilensky
e al., 1986), wi h a deg ee o ischemic se e i y co esponding o
highes a hy hmic isk (Ca melie , 1999; Janse and Wi , 1989;
Kazbano e al., 2014; Rod íguez e al., 2006) du ing ea ly s ages
o ischemia (up o 15 min). The BZ included a linea g adien in
elec ophysiological p ope ies om he ICZ o he no mal zone
(NZ) issue as shown in expe imen s (Co onel e al.,1988; Wilensky
e al., 1986). Calcium o e load o gap junc ion changes we e no
included as hey occu a la e s ages o ischemia (Ca melie , 1999).
The human model also included ansmu al he e ogenei ies in AP
du a ion (APD) o ep oduce he physiological in e se APD - ac i-
a ion ime ela ionship (Fig. 1B and C) esul ing in a posi i e T
wa e in he pseudo-ECG, as in da a ob ained in in i o human
hea s (F anz e al., 1987). Mo e de ails can be ound in he
Supplemen al Ma e ial.
2.3. S imula ion p o ocol
Pu kinje-like ac i a ion was simula ed wice wi h a cycle leng h
(CL) o 800 ms, by s imula ing he endoca dium (S1) o mimic he
ac i a ion sequence in Du e e al. (1970). This was ollowed by a
p ema u e exci a ion (S2), applied in a egion close o he BZ p one
o p ema u e bea s ha mimics findings by Janse e al. (1980). The
coupling in e al (CI) o he p ema u e s imulus (e.g., ime in e al
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248 237
di e ence be ween S1 and S2) was a ied o quan i y he ulne -
abili y window (VW) o een y. The la e was compu ed as he
ange o CIs ha esul ed in een y in he human en icles.
As illus a ed in Fig. 1, he human en icles model ep oduced
he epo ed elec ophysiological he e ogenei y caused by egional
ischemia, including ele a ed es ing ansmemb ane po en ial (V
m
)
and sho APD (Rod íguez e al., 2006) in he ICZ wi h espec o NZ,
in ag eemen wi h in i o human and ex si u animal s udies
(Ca melie , 1999; Su on e al., 2000,p.200;Tagga e al., 2001;
Wilensky e al., 1986). Res ing po en ial in he ICZ is 70 mV
compa ed o 86 mV in no mal issue, and APD is 30% sho e in
he ICZ han in he NZ, close o he 35% di e ence epo ed in hu-
man in i o measu emen s (Su on e al., 2000). The simula ed
pseudo-ECG was compu ed as in Gima and Rudy (2002), as he
ex acellula unipola po en ial a a posi ion 3.6 cm away om he
epica dial su ace (Fig. 1A), and yielded a physiological QT in e al
o ~400 ms and posi i e T wa e, as desc ibed in in i o human
hea s (F anz e al., 1987).
2.4. Reduced epola iza ion ese e
In o de o in es iga e mechanis ic implica ions o educed
Fig. 1. Elec ophysiological he e ogenei y in he ana omically-based model o he human en icles in acu e egional ischemia. A. Res ing po en ial dis ibu ion wi h ele a ed
po en ials in ischemic issue (g een solid line), loca ion o he pseudo-ECG p obe (black as e isk) and ec opic s imula ion si e (magen a solid line). B. Ac ion po en ial du a ion (APD)
dis ibu ion highligh ing APD sho ening in he ischemic cen al zone (ICZ) ( op) and examples o ac ion po en ials in no mal zone (NZ), bo de zone (BZ) and ICZ om he loca ions
ma ked wi h g een do s in he APD map. C. Ac i a ion imes dis ibu ion ( op) and pseudo-ECG (bo om) exhibi ing fi s he QRS complex, ollowed by a posi i e T wa e (due o he
in e se ela ionship be ween APD and ac i a ion imes) and ischemia-induced ST ele a ion.
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248238
epola iza ion ese e in acu e ischemia, we conside ed h ee
scena ios ep esen ed by a 0, 30 and 50% dec ease o he apidly
ac i a ing delayed ec ifie po assium cu en (I
K
) conduc ance in
he TP06. The simula ions yield a 5% and 9% p olonga ion in QT in
he pseudo-ECG, which is simila o he 7% and 12% inc ease ound
by Fossa e al. in heal hy pa ien s, 2e4 h a e being gi en 160 mg
and 320 mg o so alol espec i ely (Fossa e al., 2007).
One-dimensional simula ions we e conduc ed o cha ac e ize
he e ec o educed epola iza ion ese e in NZ and ICZ issue o
a ying deg ees o I
K
educ ion. APD was calcula ed a 90% epo-
la iza ion a s eady s a e o each CL. E ec i e e ac o y pe iod
(ERP) was e alua ed a e eaching s eady s a e a each CL and
calcula ed as he sho es CI o induce sus ained p opaga ion in he
issue. Conduc ion eloci y (CV) was calcula ed a s eady s a e om
he ime in e al sepa a ing wo ac i a ion imes a wo poin s in
he cen e o he issue sepa a ed by 2 mm. Fig. 2 illus a es he
known elec ophysiological e ec s o acu e ischemia, including
APD sho ening, ERP p olonga ion and slow CV o all deg ees o
epola iza ion ese e. No e ha no EADs occu ed in any o he
single cell and one-dimensional simula ions p esen ed in Fig. 2.
3. Resul s
3.1. Mac o- een an pa e ns in he acu ely-ischemic human
en icle model suppo expe imen al e idence in animal models
Simula ions show he es ablishmen o mac o- een y a ound
he ischemic egion in he human en icles ollowing p ema u e
exci a ion du ing he VW o all le els o epola iza ion ese e. As
illus a ed in Fig. 3, in all cases, he mac o- een an pa e n was
simila , including 3 main phases consis en wi h expe imen al
pa e ns (Janse e al., 1980). Fi s ly, unidi ec ional conduc ion block
occu s due o p olonged e ac o iness in he ICZ, wi h p opaga ion
a ound his egion ia he NZ and BZ a ound he ICZ (fi s column in
Fig. 3). Then, e og ade p opaga ion occu s h ough he ICZ once
issue dis al o p ema u e exci a ion has eco e ed (middle column
in Fig. 3). Finally, wa e on p opaga ion een e s h ough he NZ
( hi d column in Fig. 3). In all cases, he mac o- een an pa e n
was simila and las ed o up o 3 een y cycles a e which i sel -
e mina ed. This may indica e ha he es ablishmen o sus ained
en icula achyca dia and i s ansi ion o en icula fib illa ion
may equi e he con ibu ion o mo e han one ec opic bea and/o
p o-a hy hmic subs a e ac o s.
Dec eased epola iza ion ese e by 30% I
K
educ ion leads o a
dec ease o he VW o een y (65 ms wi h 30% I
K
educ ion s.
85 ms wi hou I
K
dec ease; CI ¼355e420 ms o 30% I
K
educ ion
s. CI ¼335e420 ms o no I
K
dec ease). The main an i-a hy hmic
mechanism o dec easing I
K
is ha p olonged ERP a o s conduc-
ion block bo h in he NZ and ICZ a sho CIs, he e o e leading o
ei he bidi ec ional block jus a e p ema u e exci a ion o ailu e
o sus ain e og ade conduc ion in he ICZ. While he pa e ns o
mac o- een y we e simila when I
K
educ ion was inc eased o
50%, he an i-a hy hmic ac i i y o I
K
diminished, as shown by an
inc eased wid h o he VW (80 ms; CI ¼355e435 ms). The causes o
his inc ease in he VW wi h a u he educ ion in epola iza ion
ese e a e in es iga ed below.
3.2. Elec o onically- igge ed EADs acili a e ansmu al een y
nea he ischemic bo de zone unde condi ions o educed
epola iza ion ese e
De ailed analysis o he h ee-dimensional ac i a ion pa e ns in
he simula ions e ealed ha he inc ease in he VW unde con-
di ions o educed epola iza ion ese e wi h 50% I
K
block was
caused by ansmu al een y suppo ed by elec o onically-
igge ed EADs nea he ischemic BZ, as illus a ed in Fig. 4.
Fi s ly, wo mac o- een an cycles a e es ablished a ound he ICZ
s a ing a 1200 ms and 1460 ms (as in Fig. 3). Then, p olonged
e ac o iness in he ICZ p e en s e og ade p opaga ion, and he
mac o- een an p opaga ion dies ou (Fig. 4, 1381 ms and 1775 ms
panels). Howe e , p ese ed exci abili y in he endoca dial BZ
sus ains p opaga ion o a small wa e on (Fig. 4,1775 ms snapsho ,
ci cle and a ow), which finds a non e ac o y unnel o p opa-
ga ion ia he la e al BZ, om endoca dium o epica dium (Fig. 4,
1831 ms, snapsho , ci cle and a ow). Following b eak h ough on
he epica dium, he wa e on een e s back ansmu ally owa ds
he endoca dium and a ound he ischemic BZ (Fig. 4, 1870 ms,
1920 ms and 1950 ms snapsho s).
This pa e n o ansmu al p opaga ion is only seen o 50% I
K
block and he cellula -le el mechanisms we e in es iga ed by
analyzing he ime cou se o V
m
a di e en loca ions om endo-
ca dium o epica dium in he egion sus aining he ansmu al
pa e ns o een y, as indica ed in Fig. 5A. The AP aces in Fig. 5B
show he de elopmen o la ge ansmu al V
m
g adien s (e.g. a
1831 ms) wi h ully depola ized epica dial cells (black aces) close
o epola izing endoca dial issue (g een aces). The ol age g a-
dien s esul in elec o onic cu en flowing owa ds he endoca -
dial cells, and his acili a es he de elopmen o elec o onically-
igge ed EADs (Fig. 5B, g een and blue aces om 1920 ms).
This is a consequence o a weak epola iza ion ese e due o 50%
I
K
block and he elec o onic flow o cu en ha allows o he
inc ease in V
m
du ing he pla eau o he AP. The elec o onic cu en
ha acili a es EAD o ma ion was es ima ed o be be ween 1 and
Fig. 2. Elec ophysiological p ope ies in a one-dimensional homogeneous fibe o
human en icula issue unde con ol and acu e ischemic condi ions o a ying
deg ees o he apidly ac i a ing delayed ec i ying po assium cu en (I
K
) educ ion.
Resul s show s eady-s a e alues o APD, e ec i e e ac o y pe iod (ERP) and con-
duc ion eloci y (CV) o cycle leng hs (CL) anging om 350 o 1500 ms. No mal and
ischemic cells we e assigned ex acellula po assium concen a ion ([K
þ
]
o
) o 5.4 and
8.5 mmol/L, ATP-sensi i e po assium cu en (I
K(ATP)
) ac i a ion o 0 and 5%, and peak
conduc ance o as sodium cu en (I
Na
) and L- ype calcium cu en (I
CaL
) o 100 and
75%, espec i ely, o hei o iginal alues in he TP06 model.
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248 239
1.5 pA/pF in ampli ude and 100e150 ms in du a ion by compu ing
he di usi e pa o he cu en in he monodomain equa ion.
In he absence o EAD o ma ion igge ed by elec o onic cu -
en , epica dial and endoca dial cells would be e ac o y a he
same ime, and his would lead o conduc ion ailu e. Howe e , as
shown in Fig. 5B, EAD o ma ion p olongs he ime du ing which
he endoca dial cells a e depola ized (g een aces), inc easing he
ime a ailable o achie e he ull eco e y o neighbou ing issue
(black and blue aces) and he con inua ion o he p opaga ion, as
was sugges ed by Kuma and Joyne (1994); Pogwizd and Co
(1990); Yan e al. (2001). This mechanism explains he inc eased
p opensi y o een y and VW ex ension o 50% compa ed o 30%
I
K
educ ion, and is u he illus a ed in Fig. 6 h ough a com-
pa ison o simula ions o s ong I
K
(0% educ ion) and 30% and
50% educ ion ( op, middle and bo om panels). In all h ee cases,
p opaga ion fi s p oceeds a ound he ICZ, and hen h ough he ICZ
once i s issue has eco e ed. Fo 0% and 30% I
K
educ ion, p op-
aga ion comple ely engul s he en icles and een y ails o be
es ablished. Howe e , in he case o 50% I
K
educ ion, p olonged
e ac o iness in he ICZ leads o conduc ion block close o he BZ,
ollowed by endoca dial p opaga ion and EAD o ma ion (1810 ms
and Fig. 6A and B), leading o ansmu al een y as shown in Figs. 4
and 5. Mo e de ails can be ound in he supplemen al mo ies.
3.3. P olonged low ampli ude cu en may igge EADs due o
eac i a ion o he L- ype calcium cu en ga es
In o de o in es iga e he mechanisms unde lying he o ma-
ion o EADs in he acu ely-ischemic human en icles, we simu-
la ed he e ec o a cu en injec ed du ing he epola iza ion phase
on he human en icula ionic dynamics and AP unde no mal and
ischemic condi ions wi h a ious deg ees o epola iza ion ese e.
Based on he condi ions shown in Figs. 3, 4 and 6, ou hypo hesis
was ha a cu en o low ampli ude and p olonged du a ion could
igge L- ype calcium cu en (I
CaL
) eac i a ion and EADs i
injec ed du ing he epola iza ion phase. Resul s p esen ed in
Figs. 7 and 8 show he ollowing EAD mechanisms. Fi s ly, ha
EADs can be igge ed when a cu en o low ampli ude (less han
1/20 h o he h eshold o AP igge ing) is injec ed in he NZ, BZ
and ICZ human en icle ca diomyocy es wi h no mal and educed
epola iza ion ese e (Fig. 7). This is consis en wi h esul s in ou
en icula simula ions showing EAD o ma ion ac oss he BZ o
50% I
K
block (Figs. 5 and 6). Secondly, analysis o he ionic mech-
anisms unde lying EAD o ma ion shows ha eac i a ion o he
I
CaL
cu en and i s ac i a ion ga e a e he main d i ing o ce (Fig. 8,
second and hi d ow), while calcium elease om he sa coplasmic
e iculum (as specula ed by Ve ke k e al. (2000)) is no in ol ed
(Fig. 8, ou h ow). Thi dly, key o EAD o ma ion is ha he
ampli ude o he cu en injec ed is abo e a ce ain h eshold. This
is illus a ed o example in he BZ and ICZ ows in Fig. 7A: a
con inuous 1.1 pA/pF cu en ails o de elop EADs o no mal and
educed epola iza ion ese e in he BZ and ICZ espec i ely, bu
hey a e igge ed wi h an inc eased ampli ude o 1.2 pA/pF.
Fu he mo e, educed epola iza ion ese e a o s EAD occu -
ence, pa icula ly when he p olonged cu en is applied du ing
phase 3 o epola iza ion (i.e. o V
m
20 mV) (Fig. 6B). Indeed,
he las wo columns show ha he occu ence o EADs is acili a ed
by I
K
block. Low ampli ude cu en applied a ea ly s ages o
epola iza ion a o s he occu ence o EAD o ma ion compa ed o
la e epola iza ion s ages (Fig. 8B). Fou hly, he window o EAD
o ma ion du ing he epola iza ion phase is less in he BZ and ICZ
compa ed o NZ (Fig. 7B). The e o e, EADs occu o low ampli ude
cu en applied a memb ane po en ials om 0 o 35 mV in NZ
ca diomyocy es, 0 o 25 mV in BZ ca diomyocy es and
0 o20 mV in ICZ ca diomyocy es. Finally o e all, ischemia e-
duces he p opensi y o EAD de elopmen (Fig. 7). This is due bo h
o he s onge epola iza ion ese e in ischemia caused by I
K(ATP)
ac i a ion and ele a ed ex acellula po assium concen a ion
([K
þ
]
o
). The la e dec eases he exci abili y o he cell and he e o e
he likelihood o egene a i e EAD-like beha io .
Fig. 3. Mac o- een an pa e n o p opaga ion in he human en icles a ound he acu e ischemic zone in con ol ( op) and o 30% (middle) and 50% (bo om) I
K
educ ion
ollowing p ema u e exci a ion applied wi h coupling in e als (CIs) o 355 ms, 360 and 361 ms, espec i ely. The limi s o he BZ a e ma ked wi h g een lines and he di ec ion o
p opaga ion wi h whi e a ows.
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248240

4. Discussion
High- esolu ion simula ion da a ob ained wi h an ana omically-
based model o he egionally-ischemic human en icles e eal,
o he fi s ime, he elec o onic igge o EADs and hei
con ibu ion o ansmu al een y and inc eased a hy hmic isk
unde condi ions o educed epola iza ion ese e. In ou human
en icula model, EADs de elop as a consequence o complex
pa e ns o p opaga ion, epola iza ion and cu en flow in he
he e ogeneous ischemic a ea, whe eas hey do no de elop spon-
aneously in single cell o in homogeneous issue wi h he same
memb ane kine ics. Simula ions show ha educed I
K
causes an
an i-a hy hmic p olonga ion o e ac o iness, which coun e ac s
mac o- een y o ma ion. Howe e , i simul aneously causes a p o-
a hy hmic educ ion in epola iza ion ese e, which p omo es
EAD o ma ion and ansmu al een y. A hy hmic isk in egional
ischemia is he e o e de e mined by a fine balance be ween mac o-
een an mechanisms d i en by he e ogenei y in e ac o iness
a ound he ischemic egion, and mic o- een an pa e ns,
elec o onically- igge ed EAD o ma ion and ansmu al een y.
Reac i a ion o he L- ype calcium cu en is he p ima y mecha-
nism unde lying EAD o ma ion by p olonged low ampli ude cu -
en injec ed du ing epola iza ion. Ou in es iga ions we e
p ima ily mechanis ic, and hus conduc ed using a single human
en icles model gi en he compu a ional demands. We iden i y
key ac o s ha will a y be ween indi iduals, and ha may explain
Fig. 4. Dis ibu ion o ansmemb ane ol age (V
m
) h oughou he en icles a di e en imes ollowing ec opic exci a ion esul ing ini ially in mac o- een y, bu ailu e o
suppo e og ade p opaga ion (1381 ms), ollowed by p opaga ion h ough he endoca dial BZ (1775 ms) and ansmu ally (1831 ms), leads o in amu al een y (1831e1950 ms).
Fo each ime ins an , h ee di e en iews a e shown, om le o igh : he epica dium, he endoca dium and he depola ized issue wi h V
m
abo e 20 mV (cells wi h
V
m
<20 mV a e anspa en ). Isopo en ial lines a e shown in g ey linking same po en ial le els. Mo e de ails can be ound in he supplemen al mo ies.
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248 241
in e -subjec di e ences in a hy hmic isk in acu e ischemia
modula ed by epola iza ion ese e.
4.1. Compu a ional model o he egionally-ischemic human
en icles and a ied epola iza ion ese e
Fo his s udy, a no el human model was de eloped, which in-
eg a es he biophysical in o ma ion on memb ane kine ics and he
p o-a hy hmic al e a ions induced by acu e egional ischemia, as
well as fibe o ien a ion and ealis ic ana omy. Simula ion esul s
a e in ag eemen wi h in i o and ex si u human expe imen al
s udies including: human en icula ac i a ion ime (Du e e al.,
1970); in e se APD-ac i a ion ime ela ionship (Hanson e al.,
2009); ischemia-induced he e ogenei y in AP, CV and ERP (Figs. 1
and 2)(Ca melie , 1999; Rod íguez e al., 2006); size o he
ischemic egion (Fig. 1)(Lee e al., 1981); pseudo-ECG, and i s co -
esponding QT in e al (414 ms), posi i e T wa e, and T peak- o-
Tend alue (~100 ms) (Fig. 1)(B owne e al., 1983). The ag ee-
men be ween expe imen s and simula ions lends c edibili y o ou
new findings.
4.2. Mac o and mic o- een y in he egionally-ischemic human
en icles
The mac o- een an pa e ns o ac i a ion in he egionally-
ischemic human en icles model a e also consis en wi h hose
epo ed in pig and canine hea s in he seminal wo k by Janse e al.
(1980) and include h ee necessa y condi ions (Fig. 3). Fi s ly, he
p esence o conduc ion block is necessa y and occu s when he -
e ogenei y in e ac o iness is such ha a he ime o ec opic
exci a ion, NZ issue is exci able and ICZ issue e ac o y (ERP in
ICZ >ERP in NZ). Secondly, e og ade conduc ion occu s when
he e ogenei y in e ac o iness, ischemic a ea size and CV a e such
ha p opaga ion p oceeds a ound he ICZ, while i emains blocked
in he ICZ un il ischemic issue dis al o he loca ion o ec opic
exci a ion eco e s. A p olonged ERP in ischemic issue would
equi e inc eased ischemic egion size o dec eased CV o allow o
e og ade conduc ion o occu . Finally, he een y is comple e
once p opaga ion has e og adely a e sed he ICZ and he issue
in he NZ has eco e ed. This condi ion depends on he ERP in he
NZ, CV in he ICZ and he size o he ischemic egion. A p olonged
ERP in he NZ, as wi h I
K
educ ion, would be expec ed o dec ease
he likelihood o een y.
Ou simula ion esul s show ha I
K
educ ion sligh ly lowe s
he chances o mee ing all h ee condi ions o een y by educing
he he e ogenei y in ERP be ween he ICZ and NZ ( om a ange o
62e84 ms in con ol o 57e78 ms o 50% I
K
educ ion), while
p olonging he ERP in bo h egions, he e o e educing he chances
o e og ade p opaga ion and een y. This leads o a dec ease in
he VW o 30% I
K
educ ion bu he end becomes di e en wi h
50% I
K
dec ease. The main eason is he eme gence o a newly
iden ified p o-a hy hmic mechanism e ealed by ou simula ions:
ansmu al een y sus ained by elec o onically- igge ed EAD
de elopmen .
E en hough Janse e al. also epo ed he occu ence o in a-
mu al een y in some o hei acu e ischemia expe imen s in pig
and dog, hey we e unable o esol e he unde lying mechanisms
due o he low esolu ion o hei mic oelec ode eco dings (Janse
e al., 1980). Lukas and An zele i ch also poin ed a he possibili y
o phase 2 een y in simula ed ischemic issue using hin
Fig. 5. Elec o onically- igge ed EADs a e eco ded nea he BZ in he acu ely-ischemic human en icles wi h 50% I
K
educ ion acili a ing ansmu al pa e ns o een y. A.
Snapsho s o a ansmu al c oss-sec ion illus a ing p opaga ion o elec ical exci a ion om endoca dium o epica dium a he imes indica ed. B. Time cou se o V
m
a he poin s
indica ed in panel A om endoca dium o epica dium. Legend indica es he ansmu al loca ion co esponding o each AP ace wi h % indica ing ansmu al dis ance om he
endoca dium. Ec opic exci a ion is applied a ¼1181 ms (CI ¼361 ms as in Fig. 3).
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248242
epica dial shee s in acido ic “ischemic”solu ion and highligh ed
he need o alida ion o hei hypo hesis in “ u u e s udies using
high- esolu ion echniques”(Lukas and An zele i ch, 1996). Ou
simula ions indeed yield high- esolu ion da ase s consis en wi h a
wide ange o es ablished knowledge, and e eal o he fi s ime
he de elopmen o elec o onically- igge ed EADs implica ed in
in amu al een y caused by he he e ogeneous disease subs a e
in he h ee-dimensional olume o he human en icles.
4.3. Mechanisms o EAD de elopmen in issue: elec o onic kick o
in insic ins abili y?
EADs ha e long been iden ified as a po en ially impo an p o-
a hy hmic playe (Le y and Wiseman, 1991; Ma ban e al., 1986)
and hey ha e been demons a ed p ima ily in single cell s udies
wi h highly comp omised epola iza ion ese e (caused by dis-
ease, mu a ions and/o pha macological challenge) (Vande sickel
e al., 2014; Zeng and Rudy, 1995). Despi e issue simula ion
s udies showing hei occu ence a he cen e o spi al wa e o o s
(Zemlin and Pe so , 2007; Zemlin e al., 2009), hei con ibu ion
o p o-a hy hmic issue dynamics is s ill con o e sial, as elec-
o onic coupling has also been shown o supp ess EAD o ma ion
and p opaga ion (Huelsing e al., 2000; Mendez e al., 1969; Pueyo
e al., 2011). Theo e ical s udies ha e demons a ed ha EAD o -
ma ion in issue equi es a huge educ ion o epola iza ion ese e
by simul aneously applying a 4- old inc ease in I
CaL
and 80%
educ ion o I
K
,asinVande sickel e al. (2014). This may occu in
ce ain ex eme condi ions, bu i does no explain inc eased
a hy hmic isk in disease condi ions caused by mildly educed
epola iza ion ese e as in clinical ials in pa ien s wi h co ona y
a e y disease, bu wi hou gene ic diso de s (S einbe g e al.,
2014).
Ou simula ions iden i y a mo e likely p o-a hy hmic mecha-
nism o EAD o ma ion and een y, sus ained by he
Fig. 6. In amu al een y is acili a ed by p olonged APD due o EAD o ma ion in he acu ely-ischemic human en icles wi h educed epola iza ion ese e o long CI ¼430 ms.
A. Ec opic exci a ion a ¼1253 ms (co esponding o CI ¼430 ms) ails o induce een y o 0 and 30% I
K
educ ion (fi s and second ow, espec i ely), bu leads o es ablishmen
o ansmu al een y o 50% I
K
educ ion ( hi d and ou h ow). B. Time cou se o he ac ion po en ial in he a ea o ansmu al een y o 50% I
K
educ ion ma ked wi h a blue
ci cle in panel A. Legends indica e he ansmu al loca ion o each ace (pe cen age o ansmu al dis ance om he endoca dium). Mo e de ails can be ound in he supplemen al
figu e.
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248 243
he e ogeneous en icula subs a e (Figs. 4e6). Reduced epola -
iza ion ese e acili a es EAD o ma ion, bu is no he p ima y
igge o epola iza ion ins abili y, as spon aneous EADs do no
occu in single cell o homogeneous issue p epa a ions wi h he
same memb ane kine ics. Ou esul s show, o he fi s ime, ha
EADs may be igge ed in he ischemic human hea by he cell's
en i onmen in a he e ogeneous subs a e, which c ea es complex
elec o onic flow o cu en ac oss he BZ. Repola iza ion ese e
and elec o onically- igge ed EAD o ma ion a e he e o e iden i-
fied as po en ially impo an de e minan s o p o-a hy hmic isk
in he he e ogeneous subs a e o acu e ischemia in human.
4.4. Implica ions
The mechanisms highligh ed in ou s udy e eal he complex
balance o elec ophysiological p ope ies ha de e mine p o-
Fig. 7. Cha ac e iza ion o EADs induced by low ampli ude cu en injec ed du ing he epola iza ion phase o a ying cu en ampli udes (A) and epola iza ion le els (B) o
human en icula ca diomyocy es ep esen a i e o NZ ( op), BZ (middle) and ICZ (bo om) o 0, 30 and 50% I
K
educ ion (sho dash, long dash and con inuous lines, espec i ely)
(BZ: [K
þ
]
o
¼7 mmol/L, I
K(ATP)
ac i a ion ¼3%, I
Na
and I
CaL
peak conduc ance ¼85% and ICZ: [K
þ
]
o
¼8.5 mmol/L, I
K(ATP)
ac i a ion ¼5%, I
Na
and I
CaL
peak conduc ance ¼75%). In panel
A, cu en is applied a 20 mV wi h ampli ude 0.7, 0.9, 1.1 and 1.2 pA/pF ( om le o igh ) and 170 ms du a ion. In panel B, cu en o 1.1 pA/pF ampli ude and 170 ms du a ion is
applied a di e en le els o epola iza ion om 0 o 40 mV ansmemb ane po en ial. The s imula ion p o ocol consis ed o a ain o 100 s imuli wi h ampli ude 30 pA/pF
du ing 1 ms (1.3* h eshold o AP igge ), a a CL o 500 ms.
S. Du a e al. / P og ess in Biophysics and Molecula Biology 120 (2016) 236e248244