Differential regulation of two types of
intracellular calcium release channels during
end-stage heart failure.
L O Go, … , B S Fyfe, A R Marks
J Clin Invest.
1995;
95(2)
:888-894.
https://doi.org/10.1172/JCI117739
.
The molecular basis of human heart failure is unknown. Alterations in calcium homeostasis
have been observed in failing human heart muscles. Intracellular calcium-release channels
regulate the calcium flux required for muscle contraction. Two forms of intracellular
calcium-release channels are expressed in the heart: the ryanodine receptor (RyR) and the inositol
1,4,5-trisphosphate receptor (IP3R). In the present study we showed that these two cardiac
intracellular calcium release channels were regulated in opposite directions in failing
human hearts. In the left ventricle, RyR mRNA levels were decreased by 31% (P < 0.025)
whereas IP3R mRNA levels were increased by 123% (P < 0.005). In situ hybridization
localized both RyR and IP3R mRNAs to human cardiac myocytes. The relative amounts of
IP3 binding sites increased approximately 40% compared with ryanodine binding sites in
the failing heart. RyR down-regulation could contribute to impaired contractility; IP3R up
regulation may be a compensatory response providing an alternative pathway for mobilizing
intracellular calcium release, possibly contributing to the increased diastolic tone
associated with heart failure and the hypertrophic response of failing myocardium.
Research ArticleFind the latest version:
Rapid
Publication
Differential Regulation of Two Types of Intracellular Calcium
Release
Channels
during End-Stage
Heart
Failure
Loewe 0. Go, Maria C.Moschella,James Watras,* Kushal K. Handa, Billie S.
Fyfe,*
and Andrew R. MarksMolecular Medicine Program and Division of Cardiology, Department of Medicine, and tDepartment of Pathology, Mount Sinai Schoolof Medicine, New York 10029; and *Departments of Medicine and Physiology, University of Connecticut, Farmington, Connecticut 06032
Abstract
The molecular basis of human heart failure is unknown. Alterations in calcium homeostasis have been observed in
failinghuman heart muscles. Intracellular calcium-release
channels regulatethe calciumfluxrequiredfor muscle
con-traction. Two formsof intracellular calcium-release chan-nels are expressed in the heart: the ryanodine receptor
(RyR)and the inositol 1,4,5-trisphosphate receptor
(LP3R).
Inthe presentstudy we showed that these two cardiac intra-cellular calcium release channelswereregulatedinopposite directionsinfailinghumanhearts.In theleftventricle,
RyR
mRNAlevels were decreasedby 31% (P < 0.025) whereasIP3R
mRNA levels were increasedby123% (P<0.005).In situ hybridization localized bothRyRandIP3R mRNAs to humancardiac myocytes. Therelativeamounts of IP3 bind-ing sites increased - 40%comparedwithryanodinebindingsites in thefailing heart. RyR down-regulation could
con-tribute to impaired contractility; IP3R up regulation may beacompensatory'response providingan
alternative'path-way for mobilizing intracellular calcium release, possibly contributingto theincreased.diastolictoneassociated with heartfailureandthehypertrophic responseoffailing myo-cardium. (J. Clin. Invest. 1995.
95:888-894.)
Key words:ryanodine receptor * inositol 1,4,5-trisphosphate receptor-calcium channel *excitation-contraction
coupling,
conges-tiveheartfailureIntroduction
Congestiveheart failure
(CHF)'
affects morethan2.3 millionpeoplein theUnited States.Approximately 400,000newcases
Addresscorrespondence toAndrew R.Marks, M.D.,Molecular Medi-cineProgram,Box1269,MountSinai School ofMedicine,One Gustave L.LevyPlace, NewYork, NY 10029. Phone: 212-241-0309; FAX:
212-996-4498.
Receivedforpublication28July1994and in revisedform31 Octo-ber1994.
1. Abbreviationsusedinthispaper:CHF, congestiveheartfailure;1CM,
ischemic cardiomyopathy; IDCM, idiopathic dilated cardiomyopathy; IP3R,inositol 1,4,5-trisphosphatereceptor; LV, left ventricle; RV, right
ventricle;RyR,ryanodinereceptor.
are diagnosed each year, and mortality is substantial with 200,000-400,000 deaths per year (1). The risk of death
in-creases with worsening clinical class; patients in New York HeartAssociation class IV CHF havea50-60% one-year mor-tality (2). The magnitude of these numbers underscores the importance of elucidating the mechanisms underlying myocar-dial dysfunction in end-stage CHF in order to provide a frame-work for future diagnostic and therapeutic approaches.
Excitation-contraction coupling in the heart requires the activation of the calcium-release channel in the sarcoplasmic reticulum(SR) (3). Calcium influx via the voltagegatedcalcium channel of the transverse tubuletriggers' the release of calcium from the SR into the cytosol via the ryanodine receptor (RyR)/ calcium-release channel. This phenomenon is referred to as cal-cium-induced calcium release(4, 5). Cytosolic calciumbinding
totroponinCactivates the contractile apparatus.
A second type ofcalcium-release channelhas been identi-fiedontheendoplasmic reticulumof rodent brain and smooth muscle(6, 7), andmostrecently in rodent myocardium (8). This channel is activated by EP3, aubiquitous second messenger (9, 10). I[P3 generatedby phosphoinositide hydrolysis binds-toits intracellular receptor, a calcium release channel on the endo-plasmic reticulum that is structurally related tothe RyR. This type of activation of intracellular calcium release, which stimu-lates among otherthings smooth muscle contraction, is referred
to aspharmacomechanical coupling(11-13).
The structures-of the cardiac RyR and the type 1 inositol
1,4,5-trisphosphate receptor (IP3R) have been determined by cDNAcloning.TheRyR and IP3Rare tetramerscomprisedof four 565-kD subunits (14-16), and four 313-kD (6, 7, 17),
respectively.Theyaremembersofagenefamilythatcomprises
thelargestchannel structures identifiedtodate.
Abnormal calcium homeostasis has been demonstrated in
failing human myocardium (18). Previous studies found that
mRNAlevels of the SRcalcium-ATPaseandDHPR were de-creased in end-stage human CHF (19, 20). Recent work has demonstrated that mRNA levels of theRyRwerealso reduced in the'myopathic left ventriclecomparedwithnormal (21, 22).
To date, IP3R mRNA levels in normal andfailing human hearts havenotbeendescribed. Indeed it isonly recently that IP3R expression in the heart has been appreciated. The de-creased mRNA levels of the excitation-contraction coupling
calcium channelsduringCHF(19-22) and thepossibility that neurohormonal activation could regulate cardiac contractility
(23), ledustocompareRyR and IP3R mRNA levels infailing
human hearts.Wereasoned thatpharmacomechanical coupling
mightassume a moreimportant rolein'regulatingcalcium ho-meostasis
during
CHF.Thegoalsof the presentstudy, therefore,888 Goetal.
J. Clin. Invest.
X The AmericanSocietyfor ClinicalInvestigation,Inc.
were to determine: (a) whether the IP3R is expressed in the human myocardium, and (b) whether the levels of IP3R are regulatedduring CHFinthesame manner as the RyR.
Methods
Clinical data. For each of the 32 patients, the pretransplant medical
record was reviewed to determine the clinicaldiagnosis, ejection fraction
(based on echocardiography or radionuclide angiography),
hemody-namicparametersobtained as part oftransplantevaluation (cardiac in-dex,pulmonarycapillary wedge pressure,pulmonaryvascularresistance index), and medications taken within the 2-moperiodbefore
transplanta-tion.
Harvesting oftissue and RNA preparation. Thisstudywasconducted
with the approval of the institutional review board at MountSinai Medi-cal Center. Explanted hearts wereobtained prospectivelyintheoperating
room from 32 patients who underwent cardiac transplantation at our center between April 28, 1992 and January 22, 1994.Tissue sections weighing 2 gramswereharvested fromtheleft anterolateral ventricle (LV), right ventricular free wall (RV), andventricularseptum ofeach heart. Scartissue was avoidedas much as possible during sampling. Sections fromsixnormalLVandfivenormal RV and septa were also
collectedas controltissue. (Sources of normal tissuewere trauma
vic-tims who had not consented for organ donationand potential donors rejected due tounderlyingsystemicdiseases.) Samples.-were
immedi-atelyflashfrozen inliquidnitrogen and stored at-800Cuntil use. Total
RNA'wasisolated from thesesamples afterhomogenization,using the standard guanidiniumisothiocyanatelysismethod followed by
centrifu-gation over a cesium chloride cushion(14).
Northern and slot blot analysis. Purified RNA (20pug) was size-fractionated onformaldehyde/agarosegelsand transferred onto nitrocel-lulosefiltersovernight usinglOX SSC{lXSSC=0.15 MNaCJIO.OlSM sodiumcitrate,pH 7.0). Total RNA (12.5,5, and 2.5 sg)wasloaded
onto a slot blot apparatus (GEBCO BRL, Gaithersburg, MD). Filters
were then baked at800Cfor 2 h in a vacuum oven andprehybridized
at420C overnightin buffer with 1XDenhardt'ssolution (0.02%
polyvi-nylpyrrolidone/0.02% FicolJIO.02% bovine serum albumin), SX SSC, 0.025M sodiumphosphate (pH7.4), sonicated calf thymus DNA (50
mg/ml), 0.1% SDS, and 50% (vol/vol) formamide. The filters were
hybridizedto cDNA probesovernightat420Cin the same buffer mixture
followedbywashing in 0.2X SSCat550Cfor 15 min (22).Filterswere exposed on aStoragePhosphorScreenand on x-ray film at-800Cwith asingle intensifyingscreen. The presence of a single band on northern blot analysis for each of the mRNAs examined permitted further quanti-fication byslot blot analysis.Relativeamounts ofmRNA were deter-mined with the Phosphorimager using Imagequant software.
cDNAprobes. Two cDNA probes were used for Northern and slot blotanalyses: HCRCl-a580bpEcoRI and BamHI fragment corre-sponding to nucleotides 5027-5647 of the rabbit cardiac RyR cDNA
(22); and R-IP3RI-a 1.7-kb EcoRI fragment of rat aortic smooth muscle cDNAcorrespondingtonucleotides987-2705of rat brainIP3R
cDNA sequence (8). All cDNA probes were uniformly labeled with random primers using Klenow anda-p32dCTP to a specific activity of > I09
cpm/Ag.
Tissuedistributionandspecificityof these cDNA probes were assessed using rabbit tissue Northern blots. For internal control andnormalization ofmRNAPhosphorimagerscores, a synthetic 28 S cDNA oligonucleotide based on the human28 S ribosomal RNA se-quence was used after exchange labelingwith polynucleotide kinase andy-p32ATP.In situ hybridization. Tissuesamplesfromhuman hearts were fixed in4%paraformaldehydein 0.1 M PBS overnight,cryoprotected in0.5 and 1 M sucrose in 0.1 M PBS for 30-45 min, and then frozen in Cryo-Embed compound in liquid nitrogen andstoredat-80°C until use.
10-pm
thick sections were obtained by using anIBHIcryostat andcollectedontocoatedslides. Sections'wererefixedwithparaformaldehydefor20 min,incubated for 30 min withproteinase K(3 mg/mlin0.1 MTris, pH7.5; 0.01 MEDTA), in 0.2MHCO for 20 min, and treated with
aceticanhydride (0.25% in 0.1 Mtriethanolamine buffer, pH 8.0) for 10 minat roomtemperature. Threemillioncpmof sense and antisense cRNA [35 ]CTP-labeled probes, transcribedfromlinearized HCRC1and
R-IP3R1 cDNA, wereapplied toindividual slides. Hybridizationwas
performedin 50%formamide,0.3 MNaCl,10% dextransulfate,2 mM
EDTA, IX Denhardt's,0.01 M TrisHCl, pH 8,0.05M DTT for 16-20 h. Washingconditions included 2X SSPE in presence of 10 mM DTT for 1 hatroomtemperature followedby treatment with 20mg/ ml RNasein 4X SSPE and 10 mM DTT for 30 minat 370C.Slides
werefurther washedat60Cfor 1 h in50% formamide,2XSSPE and 10 mMDTT,transferredto0.3 M ammonium acetateand1%glycerol beforedippinginNT`B2 autoradiographicemulsion. Slides were
devel-oped in Kodak D-19 (Eastman Kodak Co., Rochester, NY) after 2-wkexposures. Specimenswerephotographed using aZeiss Axiophot microscopeequippedwith a darkfield condenser(8).
Radioligandbindingassays. 50/ugofmembraneswereincubated
inappropriateconditions:(a) ryanodine binding 37°Cfor 1 h with 10 mM[3H]ryanodine in buffer(0.5M KCl,5 mMHepes, 0.1 mMCaCI2,
10 mMATP,pH 7.4) (24); (b)1P3 binding-onicefor 10 min with 10 mM[3H]IP3inbuffer(0.1MNaCI,5 mMHepes, 1mMEGTA, 1 mMDTT,2 mMEDTA, pH 8.3) (25). Bound radioactivitywaspelleted by ultracentrifugationand measuredusingaliquidscintillationcounter.
Nonspecific binding was measured by incubating with a >100-fold excess of therespective unlabeled ligand, and specific bindingcalculated
astotalminusnonspecificbinding.
Statistical analysis.Alldataareexpressed asmean±standarderror
(SE). Differences were assessed with the use of unpaired two-tailed
Student'sttest and consideredsignificant ifP< 0.05.
Results
Clinical and hemodynamic characteristics ofpatients. The clini-caldiagnosis of the 32cardiac transplant patients were as fol-lows: 15 had ischemic cardiomyopathy (ICM),14had idiopathic dilated cardiomyopathy (IDCM), 2 had hypersensitivity
myo-carditis, and 1 had sarcoidosis. The clinical and pretransplant hemodynamic characteristics of the patients arelistedin Table
I.There were22 malesand 10females, and themean age was
47±2.3yr.The degree of heart failure in this group of patients
was severe as gauged by ejection fraction (14±1.1%), cardiac index (2.2±0.1 lier/min per
m2),
pulmonary capillary wedgepressure (23±1.7 mm/Hg), and pulmonary vascular resistance
index(5.5±0.5 Wood units-M2). Within thetwomonthsbefore
transplantation, all patients were taking digoxin, diuretics, angiotensin-convertingenzyme inhibitors, and antithrombotic/ anticoagulant medications.22patientswere onintravenous pos-itiveinotropes and'sixpatients were on,6-adrenergic blockers
atthe time of transplantation. None of the patients weretaking calciumchannelblockers.
Tissuedistribution and specificity ofcDNA probes. Northern
blot analysis of rodent tissue RNA hybridized to either the HCRC-1 orR-IP3R1 cDNAprobe is shown in Fig. 1. HCRCl identifiedasingle - 16-kb mRNA species representing the
car-diac isoform of the RyR in both rabbit heart and brain; no
signalsweredetected inskeletal muscle, uterus, and kidney (the
latter two rich in smooth muscletissue). On the other hand,
R-IP3R1 identified asingle -10-kb mRNA species representing
the type 1 isoform of the IP3R in all rodent tissues except
skeletal muscle. Thesignalwasstrongest in rat heart,
intermedi-ate in rabbit brain, and'weakest in rabbit uterus and kidney. Nonspecific hybridization (e.g., to ribosomal RNAs) was not observed.
Calcium-release channel mRNA levels in human hearts.
TableL Clinical Characteristics of Patients with End-Stage Heart Failure
Patient Age Sex EF CI PCWP PVRI
% L'min/m2 mm Hg Wood U/r2
A
Ht
Br Sk Ut
KiB
Ht Ki Ut Sk Br
64 45 52 58 42 48 51 59 45 46 62 52 56 39 45 45 60 58 63 59 14 13 54 63 43 48 45 42 18 40 43 45 47±2.3 M M M M M M F F M F M M F M M M F F F M M M M M F M M M M F M F 18 2.1 2.1 15 2.6 2.2 4 2.1 6 2 12 1.8 20 1.6 9 1.9 15 1.6 2.1 11 2.1 10 2.6 14 3.7 20 1.9 22 2.4 14 2.2 15 1.9 15 2.6 2.6 10 1 5 2 15 1.8 1.6 16 3 1.9 16 2.4 9 1.8 24 1.4 23 3.6 6 2.4 15 2.5 14±1.1 2.2±0.1 26 14 18 26 26 29 29 18 30 23 42 33 6 10 9 9 23 38 26 27 30 28 38 30 20 18 8 23 34 16 16 8 23±1.7 4.7 8.1 8.3 6.6 6.2 5.5 6 S 5.6 7.5 4.8 4.8 3.1 1.9 5.3 3.3 5.4 3.6 3.5 3.5 S 5.6 8.7 3.8 3.3 4.7 16.9 6.8 9 2.2 3.3 3.2 5.5±0.5
EF, ejectionfraction; CI,cardiac index; PCWP, pulmonarycapillary
wedge pressure;PVRI,pulmonaryvascular resistanceindex;ICM,
isch-emiccardiomyopathy; IDCM, idiopathic dilatedcardiomyopathy; SE, standarderror.
analysiswith thesamecDNAprobes,HCRC1 detecteda
single
16-kb RyR mRNA while R-IP3R1 recognized a distinct
- 10-kb IP3R mRNA; no other signals were detected. Fig. 2
showsrepresentative lanes from Northern blotanalysisof RNA prepared from the differentregionsof normal andfailinghuman ventricles, hybridized to probes for the cardiac RyR, aortic smoothmuscle IP3R,and28S. Inboth normal and
cardiomyo-pathic tissue, the RyR mRNAwas much more abundant than IP3RmRNAinallregions. However,theintensityof themRNA
signals variedaccordingtotheregionfrom which itwas
taken,
theamountof RNA presentasindicatedbythe28 S ribosomal
RNA signal,andwhetheror notthetissuewascardiomyopathic.
Toquantifythesesignalsmoreprecisely,slot blot analysiswas
performed.
Resultsofquantitativeslot blotanalysisof human myocar-dial RNA using HCRC1 probe are summarized in Fig. 3 A.
IP3R *.
wo28S
RyR
.1
28Si
28S
E
S
28S
II
Figure 1. Northernblotanalysisdemonstratingspecificity and tissue
distribution ofcDNAprobes.20pgoftotalRNAisolatedfrom rabbit brain (Br), mixed skeletal muscle(Sk),uterus (Ut), kidney (Ki), and either rabbit or ratheart(Ht) tissue was subjected toformaldehyde/
agarosegel electrophoresis, transferredtonitrocellulose filters,and hy-bridizedtorandom-labeled cDNA probes. The 28 S RNA is shown as
size marker. (A) HCRC1 identifiedasingle 16-kb mRNA species representing thecardiac isoformof the RyR in both rabbit heart and
brain; nosignalsweredetectedinskeletalmuscle,uterus,andkidney. (B)On thisblot,Ht represents ratheart,while the rest representrabbit tissues. R-IP3RI identified asingle - 10-kb mRNAspecies representing
the type 1 isoform oftheIP3Rinalltissuesexceptskeletalmuscle. Thesignal was strongest in heart, intermediateinbrain, and weakest in uterusandkidney.
Consistent with previous findings (22), cardiac RyR mRNA
levels werereduced in the LV ofcardiomyopathicpatients by
31% (P< 0.025) compared with normals. CardiacRyR mRNA levels were also decreased in the septum ofcardiomyopathic
patients by37% (P<0.05)comparedwithnormals. The
differ-ences weresignificantinsubgroup comparisons of normal pa-tients with ICM andIDCM patients (datanot shown). In the
RyR
IP3R
28S
LV RV Septum Figure 2. Representative
Northernblotanalysisof
NL CM NLCM NL CM calcium-release channel
mRNAexpressionin
W1*
*W
- normal andmyopathichuman hearttissue.20
utgof total RNAprepared
_
_U,, from different regions of
son
the heart obtained from normal(NL)patientsandpatientswithcardiomyopathy(CM)wassize-fractionatedon formalde-hyde/agarose gelsandblottedtonitrocellulosefilters.Foreachregion,
thesameblotwasusedtohybridizewitheach random-labeled cDNA
probeatseparatetimes. Blotswereexposedtox-ray films foreither 1
wk(R-IP3R1)or3 d(HCRC1).The 28S ribosomal RNA is shownas a measureoftotalRNAloading.Inallsamples,theHCRCI probe
detected the 16-kbRyRmRNAwhile theR-IP3R1 probe recognized the - 10-kb IP3R mRNA.
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Mean+SE
A
0
c
0
6
0
.4
z E
B
0
E
0
0
4
z
E
C',
160
140 120 100
80 60
40 20
0
250
200
150 100 50
Left Ventricle RightVentricle Septum
p40.005 p40.001
[
NLEcm
O NL
Ecm
0
Left Ventricle Right Ventricle Septum
Figure 3. Relative normalizedRyR andIP3RmRNAlevels in different
regionsof normal andmyopathichumanhearttissue. TotalRNAisolated fromhearts of normal (NL) andcardiomyopathy (CM)patients was blotted to nitrocellulose filtersin serial dilutionsby loading onto slot blot apparatus, andhybridizedtothe(A)HCRC1 and (B)R-IP3R1
cDNAprobe.Calcium-release channel mRNA levelswerequantified with aPhosphorimager and normalized to 28 S ribosomal RNA levels. (A) Cardiac RyR mRNA levelswerereduced in the myopathicleft
ventricle by 31% (P<0.025) and septum by 37% (p < 0.05) compared with NL. In contrast, theRyR mRNA levels were nonsignificantly ele-vated in themyopathic rightventriclecomparedwith NL. (B) TheIP3R
mRNAlevels wereincreased in themyopathic left ventricle by 123% (P<0.005), right ventricle by96%(P<0.005), and septum by 123% (P< 0.001) compared with NL.
myopathicRV, RyRmRNAexpression was increased by 27% compared with normals, although the difference was not statisti-cally significant (P = 0.38).
Fig. 3Bshows the results ofquantitative slot blot analysis
usingR-IP3R-1 probe.Unlike RyR mRNA levels which were downregulated, IP3RmRNAlevels were increased inthe LV ofcardiomyopathic patients by 123% (P < 0.005) compared with normals. IP3RmRNAlevels ofcardiomyopathicpatients
werealsoincreased in theRVby96% (P < 0.005) and in the septumby 123% (P < 0.001) compared with normals. These differences remained significant even in subgroup comparisons of normal patients with ICM and IDCM patients (data not
shown).
Radioligand binding assays were performed on selected
patientsamplestodetermine whether changes in mRNA levels
were paralleled by changes in high affinity binding sites for
[3H]ryanodine
and[3H]IP3.High affinity[3H]ryanodine
bindingsites were decreased by 30% in the myopathic LV (n = 4) compared with normals (n=3), but the level of [3H]IP3 binding sites (n= 4) was unchanged compared to normals (n = 3).
Localization of calcium-release channelmRNA incardiac myocytes. The myocardium is comprised of multiple cell types (e.g., cardiac myocytes, vascular smooth muscle cells, endothe-lial cells), all of which could express and potentially regulate RyRand IP3R channels. To determine the cellular location of the mRNA signals, in situ hybridization was performed on a
subset of normal and myopathic tissue sections. Fig. 4 shows
representative results of in situ hybridization using specific cRNA probes on the myopathic LV. Hybridization with the IP3R antisense probe demonstrated that the IP3R mRNA is present not only in the vascular smooth muscle surrounding intramyocardial arteries, but also within the cardiac myocytes in a patchydistribution (Fig.4C);higher magnification revealed that the patchy appearance was due to clustering of granules within pleomorphic nuclei. Adjacent sections hybridized with senseprobe didnotshowsignals above background distribution
asexpected (data not shown).In situhybridization using cardiac RyR antisense probe revealed abundant RyR mRNA signals within cardiac myocytes, but not in vascular smooth muscle cells (Fig.4D).
Discussion
The present study demonstrates that: (a)twodifferent types of calcium-release channelsareexpressed inthenormal and failing humanmyocardium-the RyR and the IP3R, and (b) distinct andopposite regulation of each type of calcium-release channel
mRNAoccurs during end-stage CHF.
Although the dominant mechanism of excitation-contrac-tion coupling in cardiac muscle is calcium-induced calcium release viathe RyR (4, 5), evidence has accumulatedsupporting
arolefor IP3-induced calcium releaseaswell. Functional stud-ies on isolated animal ventricular fibers have shown that IP3
triggers calcium release from the SR (26)and potentiates the effects of caffeine-induced calcium release(27). Northern blot analyses using the cloned murine IP3RcDNA as aprobe have identified the - 10-kb IP3R mRNAin various mouse tissues,
including the heart (7, 28). Recently, our laboratory showed thatthe IP3R isexpressed inratcardiac myocytesby immuno-histochemistry and in situ hybridization (8). Using Northern blotanalysis andin situhybridizationin the presentstudy, we
have now conclusively demonstrated that the IP3R mRNA is alsoexpressed in both normal and diseased human cardiac myo-cytes, albeit in relatively low amounts compared with the RyR (Fig. 2 and 4).
Previous studieson humanCHF have demonstrated down regulation of the mRNAs encodingtheexcitation-contraction
couplingcalcium channels inmyopathicLVtissue. ThemRNA levelof SRcalcium-ATPase has beenshownto be reducedby 35-55% in end-stage CHF (19-22). Likewise, mRNA levels of thevoltagegated calcium channelDHPR weredecreased in myopathic LV by 47% (20). Our group and others have also shown that RyRmRNAexpressionin theLV wasdecreased in
cardiomyopathy patients (21, 22).
Thecurrentstudyextends these observations.Notonlywas
Figure 4. In situhybridization demonstrating specificity ofthe cRNAprobes andlocalizingcalcium-release channel mRNA incardiocytes.Antisense
andsense35S-labeledcRNAprobeswereusedonparaformaldehyde-fixedcardiac tissuespecimens obtained from patientswithcardiomyopathy.
(A) Left ventriclesection stained withhematoxylin and eosincontainingboth cardiac tissue(asterisk) andvascular smooth muscle tissue around
bloodvessels (arrow). (B) Theareasurrounding the asterisk inAismagnified furtherunderphasecontrast, toshow thestriations whicharetypical
of cardiac fibers. (C)Darkfieldview ofthe samesectionasinA,hybridizedtothe IP3R antisenseprobe, showingthatIP3RmRNA is present in vascular smoothmuscle(arrow) and moreimportantlyin cardiac myocytes. (D)Darkfieldviewofanadjacent sectionhybridizedtothe RyR
antisense probe, showingabundantRyRmRNAincardiac fibersbut not in vascular smoothmuscle (arrow).Bars: (A, C,andD)80
pHm;
(B)40ILm.
whether the tissuewas obtained from 1CM orIDCMpatients (Fig. 3 A).Interestingly, therewas anopposite trend of increased RyR mRNAlevels in myopathic RV compared with normals, albeit the increasewasstatistically nonsignificant. The absence ofasignificantdifference in RyR mRNA levels between normal andmyopathicRVmay bepartly duetothevaryingdegree of RVfailure present among thepatients.
Moreimportantly,thisstudyreportsforthefirsttimeanup regulation of the IP3R mRNA levels in the ventricles of patients withend-stageCHFcomparedwith normalpatients,incontrast toother calciumchannels whicharedown regulated (Fig. 3 B). Levels of IP3R mRNA were markedly increased in the myo-pathic LVby 98% in ICM andby143% in IDCM (P < 0.005 for both), and in the myopathic RV by 80% in ICM and by 110% in IDCM (P < 0.005 for both), compared with normal tissue. IP3RmRNAlevelswerealsosignificantlyupregulated in the myopathic septum by 122-124% (P< 0.001), whether
1CMorIDCM.
Recent studies examining protein levels of SR
calcium-ATPase in human CHF have produced conflicting results: one
reported reduction of SR calcium-ATPase protein levels by 36% infailing myocardium (29) while another didnotfind any
differ-encebetween nonfailing andfailing human myocardium (30). With regard to the present study, we sought to determine whether levels of ryanodine and IP3 binding sites were regu-lated during CHF similartomRNA levels. The level of high
affinity [3H]ryanodinebinding sites was decreased by 30%in
the myopathic LVcomparedwith normals whereas the levelof
[3H]IP3 binding sites remained unchanged, consistent withthe
dataonmRNAlevels of RyR, butnotIP3R.Bindingsiteswere
calculated asthemoles ofradioligandbinding per mg of total protein in each sample, which in effect expresses the absolute number of ryanodine and IP3 binding sites relative to much
more abundant proteins that may be regulated during CHF. Directcomparison of ryanodinetoIP3binding sites reveals that therelative amount of IP3 binding sites increased 43%
com-892 Goetal.
.1
pared with ryanodine binding sites in the failing heart. Thus although the absolute level of IP3binding siteswas notaltered during CHF, the ratio of LP3R to RyR quantifiedby binding studies was higherinfailingversus normal humanmyocardium. In the tissues comprising the myopathic LV chamber, the simultaneously reduced RyR and elevated IP3R mRNA levels argueagainstauniversal decreaseorincrease ineither cardiac geneexpression or number of myocytes astheexplanation for these findings. Recent animal studies havereproduced this find-ing of decreased RyR calcium-release channel expression in CHFusing various animal models suchaschronic doxorubicin cardiomyopathy in rabbits (24) and rapid ventricular pacing
induced or spontaneous CHF in dogs (31). Our preliminary studies using rats with acute doxorubicin-induced
cardiomyopa-thy also showeddown-regulation of RyRmRNAlevels,aswell
asupregulation ofLP3RmRNAlevels(32). That various
pro-cesses leading to CHF manifest similar changes in calcium-release channelregulation suggest that thesechangesare funda-mental to thepathophysiology of CHF andnotmerely dueto
theeffects of individualpharmacologic agentsor entities. Itis evidentthatregulation of calcium channelmRNAlevels vary depending on the channel type (Fig. 3, A and B). The decreasedexpression in myopathic hearts oftwochannels in-volved in excitation-contraction coupling, theDHPR and the RyR,implies that intracellular calcium regulationby this path-way may be impaired and could contribute to compromised cardiac contractility. Support for this concept isprovidedbya
study which reportedprolonged contractions and calcium
tran-sients infailinghearts, then proceededtodemonstrate that
vera-pamil and ryanodine togetherabolished the abnormal calcium transients(18). Clinical trials have documented further deterio-ration of CHF inpatients treated with organic calcium channel blockers including diltiazem (33) and nifedipine (34), sug-gesting that the excitation-contraction coupling mechanism may bealready jeopardized in themyopathic heart.
The up regulation of IP3R mRNA levels and increased IP3R/RyR ratio are novel findings and may provide further insights into the molecular basis underlying different aspects of chronic CHF. For example, hypertrophy is an early response of the myocardium to the increased wall stress produced by ventriculardilatation and peripheral vasoconstriction (23), and thisenduringresponse manifestedby the gross biventricular and microscopic myocyte hypertrophy and enlarged pleomorphic nuclei whicharecharacteristic of explanted hearts from patients with end-stage CHF (35). Multiple growth factors have been showntocausecardiac myocyte hypertrophy with DNA replica-tion (polyploidy) in in vitro studies (36), and some of these factorsexerttheir effects via theIP3/calciumsignalingpathway, suchasangiotensin II (37, 38)orendothelin (36, 39). Increased IP3Rexpressioninventricular myocytesduring CHF is
consis-tentwith the up regulation ofa signaling pathway mediating thehypertrophic response.
The IP3R could participateinregulatingcontractility in the failing heart by analogytoits roleincontrolling smooth muscle contraction via pharmacomechanical coupling (10, 12). CHF is
adisorder of thecirculation characterized by elevated neurohor-monal activity (23). Studies using intracoronary infusions of
adrenergic agonists in patients with CHF have demonstrated both
P-adrenergic
receptor-mediated lusitropy and a-adrenergicreceptor-mediatedinotropy in such patients (40). Such modula-tion of contractility by adrenergic agents can be carried out
through
theLP3R-mediated
intracellular calcium release(10,
41), with up regulation of IP3R as a mechanism to increase sensitivity in the failing myocardium.
Inaddition toregulating contractility andfacilitating hyper-trophy in the failing myocardium, the IP3/calcium signaling pathway mayplay apathological role in cardiac arrhythmogen-esis (4). Delayed afterpotentials in hypertrophied and failing myocardium have been attributedtoabnormaltransient inward currents which occur during diastole; if large enough, these afterdepolarizations can reach threshold andtriggerarrhythmias
(42). Development ofhigh myoplasmic calcium level during
diastole (> 500 nM) due to intracellular calcium oscillations appeared to induce this inward current (43, 44). Recent data from skinned cardiac fibers demonstrated that IP3 enhanced suchcalcium oscillations from the SR(45). Thus,the increased EP3Rexpression in myocardium ofpatients withend-stage CHF couldpotentially explainwhy these patients are so susceptible
toarrhythmias.
In view of the decreased expression in cardiomyopathic
tissue of the calcium channels involvedinexcitation-contraction coupling, the alternativepathway ofregulatingintracellular cal-cium via the EP3R calcal-cium-release channel may provetobe a
significant compensatory mechanism in thefailinghuman heart. Therapydesigned to enhancepharmacomechanicalcoupling by targetingtheIP3Rmightprove efficacious.However,this spec-ulation should be tempered as we did not examine inositol
polyphosphate hydrolysispersein oursamples by membrane bound phospholipase C, which in turn is dependent upon G-protein coupled receptors, tyrosine kinase coupled receptors, and mechanical deformation for its activation (9, 10, 37, 38).
Insummary, differentialregulationof calcium-release chan-nel mRNA levels occurs during end-stage CHF according to
channel type. RyR mRNA levels in the myopathic LV and septum is downregulatedcomparedtonormal. Incontrast,there
is a consistent trendof increased IP3R mRNA levels infailing
humanmyocardium comparedtonormal. Suchanalteration in calcium handling may be a molecular mechanism underlying
impaired excitation-contraction coupling, enhanced pharmaco-mechanicalcoupling, compensatory hypertrophy, and increased
arrhythmia susceptibility which are observed in progressive
myocardial dysfunction. Theseinsights into the molecular basis of CHF should providenewdirections forfuturediagnosticand
therapeutic approaches.
Acknowledgments
Wethank Drs.Valentin Fuster and Richard Gorlin for helpful comments andforsupport, Dr. Alan Gass andMaryCourtney,R.N., forproviding
clinicaldata,Dr.StevenLansmanforprovidingaccess totheexplanted hearts,and Dr.LuyiSen forprovidingthree of the normal heartsamples.
This work was supported in partbygrants from the National
Insti-tutes of Health (NS-29814) and the American Heart Association (to A. R.Marks). A. R.Marks is aBristol-Meyers/Squibb Established
In-vestigator of the American Heart Association. L.0.Go is an American
CollegeofCardiology/MerckResearch Fellow.
References
1. Parmley,W., K. Chatterjee, G.Francis, B. Firth, and R. Kloner. 1991.
Congestiveheart failure-new frontiers. West. J.Med. 154:427-441. 2.McKee, P.,W.Castelli,P.McNamara, and W. Kannel. 1971. The natural
historyof congestive heart failure: theFramingham Study. N. Engl. J. Med. 285:1441-1446.
4.Callewaert, G. 1992.Excitation-contractioncoupling in mammalian cardiac cells. Cardiovasc. Res. 26:923-932.
5. Fabiato, A. 1983.Calcium-inducedrelease of calcium from the cardiac sarcoplasmic reticulum. Am. J. Physiol. 245:Cl -C14.
6. Furuichi, T., S. Yoshikawa, A. Miyawaki, K. Wada, N. Maeda, and K. Mikoshiba. 1989. Primary structure andfunctional expressionof the inositol
1,4,5-trisphosphate-binding proteinP400.Nature(Lond.).342:32-38.
7. Marks, A., P. Tempst, C. Chadwick, L. Riviere, S. Fleischer, and B. Nadal-Ginard. 1990. Smooth muscle and brain inositol1,4,5-trisphosphatereceptors are structurally and functionally similar. J. Biol. Chem. 265:1-4.
8. Moschella, M., and A. Marks. 1993. Inositol1,4,5-trisphosphatereceptor expression in cardiacmyocytes.J. Cell Biol. 120:1137-1146.
9.Berridge,M., and R. Irvine. 1989. Inositol phosphates and cell signalling. Nature(Lond.). 341:197-205.
10.Berridge,M. 1993. Inositol trisphosphate and calcium signalling. Nature (Lond.).361:315-325.
11.Ehrlich, B., and J. Watras. 1988. Inositol 1,4,5-trisphosphate activatesa channelfrom smooth musclesarcoplasmic reticulum.Nature (Lond.).
336:583-586.
12. Hathaway, D., K. March, J. Lash, L. Adam, and R. Wilensky. 1991. Vascularsmooth muscle-a review of the molecular basis of contractility. Circu-lation. 83:382-390.
13. Somlyo, A., M. Bond, A. Somlyo, and A. Scarpa. 1985. Inositol trisphos-phate-induced calcium release and contraction in vascular smooth muscle. Proc. Natl. Acad. Sci. USA.82:5231-5235.
14.Marks, A., P. Tempst, K. Hwang, M. Taubman, M. Inui, C. Chadwick, S.Fleischer,and B.Nadal-Ginard. 1989.Molecularcloningandcharacterization
of theryanodine receptor/junctionalchannelcomplexcDNA from skeletal muscle
sarcoplasmic reticulum.Proc. Natl. Acad. Sci. USA.86:8683-8687.
15.Otsu, K., H. Willard, V. Khanna, F. Zorzato, N. M. Green, and P.-H. MacLennan. 1990. Molecular cloning of cDNA encoding the Ca2+ release chan-nel(ryanodinereceptor) of rabbit cardiac musclesarcoplasmicreticulum. J. Biol. Chem.265:13472-13483.
16. Takeshima, H., S. Nishimura, T. Matsumoto, H. Ishida, K. Kangawa, N. Minamino, H. Matsuo, M. Ueda, M.Hanaoka,T.Hirose,andS. Numa. 1989.
Primarystructure andexpressionfromcomplementaryDNA ofskeletal muscle
ryanodinereceptor. Nature(Lond.).339:439-445.
17.Chadwick, C.,A.Saito,and F. S. 1990. Isolation and characterization of the inositoltrisphosphatereceptor from smooth muscle. Proc. Natl. Acad. Sci. USA.87:2132-2136.
18. Gwathmey, J., L.Copelas, R. McKinnon, F. Schoen, M. Feldman, W. Grossman, and J. Morgan. 1987. Abnormal intracellular calcium handling in
myocardiumfrompatientswithend-stageheartfailure. Circ. Res. 61:70-76. 19.Mercadier,J., A. Lompre, P. Duc, K. R.Boheler,J. B.Fraysse,C.
Wisnew-sky,P. D. Allen, M.Komajda, and K. Schwartz. 1990.Altered sarcoplasmic
reticulum Ca2+-ATPase geneexpressionin the human ventricleduring end-stage
heart failure. J. Clin. Invest.85:305-309.
20. Takahashi, T., P. Allen, A. Marks, A.Denniss,F.Schoen, W. Grossman, J.Marsh, and S. Izumo. 1992. Alteredexpressionof genesencodingtheCa2+
regulatory proteinsin themyocardiumofpatients withend-stageheart failure: Correlation withexpressionof the Ca2+-ATPase gene. Circ. Res.71:1357-1365.
21.Arai, M., N. Alpert, D. MacLennan, P. Barton, and M.Periasamy. 1993. Alterations insarcoplasmicreticulum geneexpressionin humanheart failure: a possiblemechanism for alterations insystolicanddiastolicpropertiesof thefailing myocardium.Circ. Res.72:463-469.
22.Brilliantes, A.,P.Allen,T.Takahashi,S.Izumo,and A. Marks. 1992. Differences in cardiac calcium release channel(ryanodine receptor) expressionin
myocardiumfrompatientswithend-stage heart failure causedbyischemicversus
dilatedcardiomyopathy.Circ. Res.71:18-26.
23.Packer,M.1992.Pathophysiologyand treatment of chronic heart failure. Lancet. 340:88-95.
24.Dodd, D.,J.Atkinson,R. Olson,S.Buck,B.Cusack,S.Fleischer, and R.Boucek Jr. 1993.Doxorubicincardiomyopathyisassociated withadecrease
in calcium release channel of the sarcoplasmic reticulum in a chronic rabbit model. J. Clin.Invest.91:1697-1705.
25. Benevolensky, D.,I. Moraru, and J. Watras. 1994. Micromolar calcium decreases affinity of inositol trisphosphate receptor in vascular smooth muscle. Biochem. J. 299:631-636.
26. Kentish, J., R. Barsotti, T. Lea,I.Mulligan, J. Patel, and M. Ferenczi. 1990. Calcium release fromcardiac sarcoplasmic reticuluminduced by photorelease of calcium orIns(1,4,5)P3.Am. J. Physiol.258:H610-H615.
27.Nosek, T., M. Williams, S. Ziegler, and R. Godt. 1986. Inositol trisphos-phate enhances calcium release in skinned cardiac and skeletal muscle. Am J Physiol.250:C807-C811.
28.Nakagawa, T., H. Okano, T. Furuichi, J. Aruga, and K. Mikoshiba. 1991. The subtypes of the mouse inositol1,4,5-trisphosphatereceptor are expressed in atissue-specificanddevelopmentally specificmanner. Proc.Nati.Acad. Sci. USA. 88:6244-6248.
29. Hasenfuss, G., H. Reinecke, R. Studer, M. Meyer, B. Pieske, J. Holtz, C. Holubarsch, H. Posival, H. Just, and H. Drexler. 1994. Relation between myocar-dialfunction and expression of sarcoplasmicreticulumCa2+-ATPase in failing andnonfailing humanmyocardium. Circ. Res. 75:434-442.
30.Movsesian,M., M. Karimi, K. Green, and L. Jones. 1994.
Ca2+-trans-porting ATPase,phospholambanandcalsequestrin levels innonfailingand failing humanmyocardium.Circulation.90:653-657.
31. Cory, R., L. McCutcheon, M. O'Grady, A. Pang, J. Geiger, and P. O'Brien.
1993.Compensatorydownregulationofmyocardial Cachannelin SR from dogs with heart failure.Am.J. Physiol.264:H926-H927.
32. Handa, K., M. Moschella, B. Fyfe, and A. Marks. 1993. Regulation of calciumchannel expression during heart failure. Circulation. 88(supplI):I-624.
(Abstr.)
33. Goldstein, R., S. Boccuzzi, D. Cruess, and S. Nattel. 1991.Diltiazem
increaseslate-onset congestive heart failure inpost-infarctionpatients with early reduction inejectionfraction.Circulation. 83:52-60.
34.Elkayum,V., J. Amin, A. Mehra, J. Vasquez, L. Weber, and S. Rahimtoola. 1990. Aprospective,randomizeddouble-blindcross-overstudytocompare the
efficacyandsafetyof chronicnifedipinetherapy with that of isosorbide dinitrate and their combination in the treatment of chronic congestive heart failure. Circula-tion.82:1954-1961.
35. Tazelaar,H.,and W. Edwards. 1992. Pathology of cardiactransplantation:
recipient hearts(chronicheartfailure) and donor hearts (acute and chronic
rejec-tion).Mayo Clin. Proc.67:685-696.
36. Chien, K.,K. Knowlton, H. Zhu, and A. Chien. 1991. Regulation of cardiacgeneexpression during myocardialgrowthand hypertrophy: molecular studies ofanadaptive physiologicresponse. FASEB(FedAm.Soc. Exp.Biol.)
J.5:3037-3046.
37.Sadoshima,J., and S. Izumo. 1993.Signaltransductionpathwaysof
angio-tensin II induced c-fos geneexpression incardiac myocytes in vitro: roles of
phospholipid-derivedsecond messengers. Circ. Res. 73:424-438.
38.Sadoshima,J.,Y.Xu,H.Slayter,and S. Izumo. 1993. Autocrine release ofangiotensinH mediates stretch-inducedhypertrophyof cardiac myocytes in vitro.Cell.75:977-984.
39.Lovenberg, W., and R. Miller. 1990. Endothelin:areview of its effects andpossiblemechanisms of action. Neurochem. Res. 15:407-417.
40.Colucci,W.1993. In situassessmentofa-and/3-adrenergicresponses in
failinghumanmyocardium. Circulation. 87(supplVII):VII63-VII67.
41.Otani,H., H.Otani,and D. Das. 1988.al-Adrenoreceptor-mediated phos-phoinositidebreakdown andinotropic response inrat left ventricularpapillary
muscles. Circ. Res. 62:8-17.
42.Aronson,R.,and Z.Ming.1993. Cellular mechanisms ofarrhythmiasin
hypertrophiedandfailing myocardium.Circulation.87(supplVII):V1176-VII83.
43.Colquhoun, D.,E.Neher,H.Reuter,andC. Stevens. 1981. Inwardcurrent
channelsactivated byintracellular Ca in cultured cardiac cells. Nature(Lond.).
294:752-754.
44.Orchard, C.,D.Eisner,and D. Allen. 1983.Oscillationsof intracellular
Ca2'inmammaliancardiac muscle. Nature(Lond.). 304:735-738.
45.Zhu, Y.,andT. Nosek. 1991. InositoltrisphosphateenhancesCa2' oscilla-tions butnot Ca2+-induced Ca2' release from cardiac sarcoplasmic reticulum.
PflugersArch. 418:1-6.