lipophilic cations, cellular studies have demonstratedthat the lipophilic properties alone are an importantbut insufficient determinant oftheir myocardial uptakeproperties (2,3). For example, selective synthesis ofvarious 99mTcisonitrile complexes and correlative structure-uptake analysis with cultured heart cells haveshown that agents of increased lipophilicity in generalhave greater myocardial uptake (2). However, a fewhighly lipophilic agents demonstrate poor myocyte uptake suggesting that lipophilicity alone does not fullycharacterize 99mTcisonitnle cellular uptake. Other factors such as the cationic charge, ligand branching pattern, and molecular size ofthe @mTc@isonitrilecomplexalso appear to influence myocardial uptake. Thus, cellular mechanisms of myocardial uptake remain unresolved.

Hexakis (methoxyisobutylisonitrile)technetium-99m([99mTcIMIBI)and hexakis(carbomethoxyisopropylisonitrile)Tc-99m ([99mTclCpI)are two members of thisclass that have undergone the most extensive clinicaltesting to date (4,5,6). Cellular studies have shown[99mTclCpImyocyte net uptake to be linearly proportional to extracellular concentration ofthe complex andno competitive displacement of tracer by carrier [99TcJCPI up to a concentration of iO@ M (7). These dataindicate the absence of a high affinity cell receptor for[99mTc]CpI Myocardial net uptake of [@mTc]MIBI hasbeen determined to be greater than that of [99mTclCpIin cultured chick heart cells (2), as well as in rabbit andguinea pig heart preparations (2,8,9). Myocardial bloodflow is linearly related to [99mTcJMIBIdistribution innormal dog (10) and good correlation between [99mTc]MIBI and thallium (201'fl)distribution in vivo has beenreported in phase II clinical trials (6).

Since uptake mechanisms will affect the appropriateclinical applications of the @mTc@isonitriles,we usedchick heart cells in monolayer culture to characterizemyocardial cellular uptake and washout kinetics of[99mTc]MIBI To determine whether the metabolic stateof the heart may modulate cellular uptake of [@mTc]

Cellular uptake characteristics of hexakis(methoxyisobutylisonitrile)technetium(l)([@―Tc]MlBI),a myocardial perfusion imaging agent, were evaluated in cultured chickembryo heart cells. Myocyte net uptake of @“Tc-MIBlapproached a plateau with a half-time of 9.3 ±1.5 mm(mean ±s.e.m.; n = 10). Tracer [@“Tc]MIBlshowedapparentcompetitivedisplacementbyearner[@Tc]MlBlatrelatively high molar ratios ([@‘TcJMlBl/[@Tc]MlBl)indicating a low affinity cellular retention process (apparent K@-@7 x 1Os). Metabolic inhibition induced by pre-incubationofcellsfor2.5hrinrotenone(10 ,@M),iodoacetate(1mM),or both metabolic inhibitors together reduced 1-mm [@“Tc]MIBIuptaketo 74.1% ±8.0% (p < 0.05),6.2% ±3.4%(p < 0.01), and 10.1% ±3.6% of control (p < 0.01),respectively (n = 11-1 2). Half-maximal inhibitory concentration of iodoacetate was —5 @M.lodoacetate inhibitionof [@“Tc]MIBIuptake kinetics was time-dependent; nosignificant effect on [@‘Tc]MlBluptake was seen duringthefirst60 mmof metabolicinhibitiondespitesignificantdepletion of ATP content determined on the same preparations(controlATP: 40.2 nmoles/mgproteinversus iodoacetate incubation: 2.8 nmoles/mg protein; p < 0.01).However, prolonged metabolic blockade did eventuallydepress 1-mm [@“Tc]MlBluptake. These data indicatethata latecomponentofmyocardialcellinjurycandepress[@“TcJMIBIcellularuptake.

J NucI Med 1990; 31:464-472

exakis(alkylisonitrile)technetium(I) complexes

are a class of potential myocardial perfusion imagingagents that exploit the favorable scintigraphic propertiesof technetium (99mTc)(1). Although these agents are

ReceivedApril13, 1989;revisionaccepted Dec.4, 1989.For repiints contact: DavidPiwnica-Worms,MD,PhD, Departmentof

Radiology,Brighamand Women's Hospital,75 Franas St., Boston, MA02115.

. A Squibb Diagnostics/Radiological Society of North America Research

FundScholar.

464 The Journal of Nuclear Medicine •Vol. 31 •No. 4 •April 1990

Effect of Metabolic Inhibition on Technetium99m-MIBI Kinetics in Cultured ChickMyocardial CellsDavid Piwnica@Worms*, James F. Kronauge, Laurence Delmon, B. Leonard Holman, James D. Marsh,and Alun G. Jones

Departments ofRadiology and Internal Medicine, Cardiovascular Division, Brigham and Women ‘sHospital, HarvardMedical School, Boston, Massachusetts

scribed (2,8). Carrier-added [@Tc]MIBI was synthesized from8—10mg(8—10mmoles) of@Tc (MIBI)6 hexafluorophosphatepowder dissolved in 0.5 ml of tracer [@mTc]MIBIin 95%ethanol solution.

A TP Content. In several series of experiments, cell ATPcontent was assayed fluorometrically by a standard hexokinasereaction (19) on cells which had simultaneously been analyzedfor [@mTc]MIBI uptake. Following incubation in IAA for theappropriate time and subsequent exposure to [99mTc]MIBI@containing buffer for 1 mm also in the presence of inhibitor,cells on coverslips were extracted in 300@ 14% perchloricacid for 10—15mm in 35 mm plastic petri dishes on ice. A250-@@laliquot was then transferred to a microfuge tube,neutralized with 200 Ml1.6 M K2C03 on ice, and the sampleassayed for [99mTc]MIBIcontent in a well-type gamma counterprior to freezing (—20°C).Cell debris remaining in the petrifrom the extraction procedure was also assayed for gammaactivity, then dissolved in 1.9 ml of solution containing 1%sodium dodecyl sulfate and 10 mM sodium borate for proteindetermination by the method of Lowry (15). Appropriategeometric corrections allowed total cell-associated gammaactivity to be calculated from the paired microfuge tubes andpetridishes.Previouslyfrozenextractswerethen thawed,spunfor 5 mm in an air centrifuge, and 150 @ilof the supernatantassayed fluorometrically for ATP content (Kontron Instruments SFM 25, Zurich, Switzerland; excitation 340 nm, emission 460 nm). Reactions were initiated by addition of hexokinase. ATP standards were assayed every third to fourthsample to allow internal calibration. ATP content is expressedas nmol/mg cell protein.

Statistics. Values are presented as mean ±s.c.m. Statisticalsignificance was determined by one-way analysis of variance(20) or the two-tailed Student's t-test (21) as indicated in thetext.

RESULTS

Technetium-99m-MIBI net uptake into spontaneously contractile chick heart cells in monolayer culture approached a plateau by 40 mm with a half-time(t112) for cellular uptake of9.3 ±1.5 mm (n = 10) (Fig.1). Plateau accumulation was maintained for at least90 mm, implying a cellular steady-state or equilibriumcondition had been achieved. Saturable net uptake hasalso been previously observed with [@mTc]CPI,but thelack of competitive displacement of tracer [@mTcJCPIby carrier [@Tc]CPI (concentration range: ‘to l0@Al) mitigated against a high-affinity @mTc@isonitrilereceptor ( 7). Similar net uptake experiments were performed with [99mTc]MIBI including a range with higherconcentrations of [@TcJMIBI (Fig. 2). Half-maximalaccumulation of [99mTc]MIBI (apparent KD) was ob

served at a molar ratio ([99mTc1MIBI/[99@fc]MIBI)of-'@-7x l0@ indicating the presence of a low-affinity retention process.

Cells pre-equilibrated in [99mTcJMIBIload solutionfor 20 mm and then switched to isotope-free solution(37°C)showed [99mTc]MIBIunidirectional washout Idnetics with a mean t112of 8 ±2 mm (n = 4) assuming

MIBI, conditions associated with ischemia were produced by metabolic blockade with rotenone and iodoacetate and found to markedly inhibit [@mTc]MIBIcellular uptake.

MATERIALSAND METHODSCultured Heart Cells andp9mTcJMIBI Kinetic Studies. The

techniques for preparing monolayers of spontaneously contracting chick myocytes from 10- to 11-day-old chick embryoventricles disaggregated with trypsin have been described (11,12,13).Twenty-five-millimetercircularglasscoverslipsservedas substrate for attachment of the cells. Some cultures weregrown to confluence in media containing L-[4,5-3H(N)Jleucine for 3—4days (14). Calibration of cellular 3H counts withprotein determination by the method of Lowry (15) allowednormalization of each coverslip to cell protein (typically 0.1—0.2 mg protein per coverslip). Protocols for cellular uptake

kinetic and washout studies in cultured preparations as wellas in vivo biodistribution studies in newborn chicks have beenpreviously described ( 7). Control experiments showed that[99mTc]MIBI 1-mm binding to blank glass coverslips was14.9% ±2.9% of the total 1-mm activity obtained with cellcontaining coverslips (n = 4); cell uptake was determined bysubtracting this value from total uptake. Net cellular accumulation of radioactivity in Figures 5 and 6 were normalizedto extracellular concentration of the isotope by dividing cellcontent of 201Tlor [99mTc]MIBI by the activity determined inan aliquot of extracellular loading buffer. This allowed directcomparison of net cell uptakes independent of subcellularcompartmentation despite the use of different isotopes ordifferent extracellular concentrations of radioactivity. Thesevalues, when divided by cell-water space (6.6 ed/mg protein;recalculated from Ref. 11), could yield intracellular/extracellular concentration ratios for each agent; however, such analysis would assume homogeneous distribution oftracer throughout the cytosol. Therefore, to avoid confusion regarding thesubcellular distribution of the agents, data were normalizedsimply as cell content divided by extracellular concentrationof radioactivity. Equilibrium calculations were used to estimate total moles of99mTc and 99Tc for some experiments (16,17).

Experimental Solutions. Experiments were performed inHEPES-buffered physiologic salt solution with the followingcomposition (mM): NaCI, 137; KCI, 4.5; MgCI2, 0.5; CaC12,0.9; HEPES, 4.0; dextrose, 5.6; pH 7.35 in air; 37°C.Loadingsolutions were obtained by addition of [99mTc]MIBI (finalconcentration: 50—150@Ci/ml;3—10fmol/@zCi)or 20'fl (finalconcentration: 10—15zCi/ml; @-25fmol/.iCi). Rotenone, 2-deoxyglucose (2DG), and iodoacetic acid (IAA) were dissolvedin dimethylsulfoxide (DMSO) prior to addition to solutions.Final DMSO concentration was typically 0.1 %. DMSO concentrations up to 1% do not significantly alter contractileactivity or action potential configuration of cultured chickheart cells(18). Controlexperimentsalsoshowedno effectofDMSOon [99mTcJMIBIcellularuptake.

Ouabain, amiloride, bumetanide, and verapamil were dissolved in HEPES buffer directly from powder. Thallium-201was obtained from E.I. Dupont, Billerica, MA. Technetium99m-MIBI was synthesized from neat (free) ligand or kitpreparations (Cardiolite, E.I. Dupont) and purified as de

MetabolicInhibitionof [@“Tc]MlBlCellularUptake•Piwnica-Wormset al 465

TIME (mln)

FIGURE30 20 40 60 80 100 Technetium-99m-MlBl unidirectional washout from cultured

TIME (mln) heart cells. All preparations were incubated for 20 mm in aloading solution containing [@“TcJMlBland switched to[@“Tc]MlBl-freecontrolsolutionforwashout. Each pointrepresentsthree determinations.Solid linewas drawn by eye.

mined from four series of experiments involving 11preparations. However, one additional experimentshowed no significant washout. (Other than biologicvariability, the cause of this one aberrant washout experiment could not be determined. This experimentwas one of two washouts performed with [99mTcJMIBIsynthesized from ligand de novo rather than from a kitformulation. However, reverse-phase high performanceliquid chromatography (HPLC) analysis of this [99mTcJMIBI sample was identical to kit formulations [data notshown]). Another series of experiments directly compared [99mTc]MIBI and [99mTc]CPI washout from preparations of the same culture. Total [99mTc]CpIwashoutat 60 mm was greater than [@mTc]MIBIwashout (92%±2.8% washout versus 80% ± 1.8%, respectively; n =3; p= 0.02).

Biodistribution studies on new born chicks in vivodemonstrated significant net [99mTc]MIBImyocardialclearance. Heart activity 30 mm postinjection (2.9% ±0.2% of injected dose/g; n = 5) was less than heartactivity 5 mm postinjection (4.3% ±0.4%; n = 4) (p =0.02).

Technetium-99m-MIBI is a cation. To evaluate if[99mTc]MIBI uptake occurs via defined sarcolemmalcation transport mechanisms, the effect of myocardial

‘I cationic membrane transport inhibitors on [99mTc]MIBI

. I uptake kinetics was determined. Pre-treatment for 1

mm with saturating concentrations of ouabain (100@M;Na/K ATPase inhibitor)(14), amilonde (100 @M;

Na/H exchange blocker at this concentration) (22,23),bumetanide (10 zM; Na+K+2Cl co-transport inhibitor) (24), or verapamil (1 @sM;Ca2@channel blocker)(25) showed no significant effect on subsequent 1-mm

[99mTc]MIBI uptake (p > 0.5) (Fig. 4A). (Control expenments have demonstrated 95% exchange of extra

I

II

FIGURE 1Technetium-99m-MIBInet uptake into cultured chick heartcells. Cells were incubatedin solutioncontaining[es@@@TcJMlBl[typically 50—150 MCi/mI(3—10 fmole total Tc-MlBl/@tCi)]forvarioustimesbeforewashingin ice-coldisotope-freesolutionto clearextracellularspaces.Eachpointrepresentsthemean±s.e.m. of 3—4determinations. Solid line was drawn by eyeassuming a saturable process with T1@of 10 mm. Inset:[@“Tc]MlBlnet uptake in another cultured preparation betterresolvingearlytime points. Sameaxis legendsapply.

a single-compartment for curve fitting (Fig. 3). Twoexperiments showed evidence of two compartments;the faster compartment (65% of activity) had a t@,2of-@-6mm and the slower compartment a t@12of —90mm.

Total washout varied from culture to culture; mean 60-mm washout was 84% ±2% of plateau activity deter

110.

@, 8.

@ 6

IE @.

LOG{(99mTc-MIBI]/(99Tc-MIBII)

FIGURE 2Equilibriumcontent of tracer [es@@@Tc]MlBl(30 mm incubation)as a function of carrier [esTcJMIBI.Load solutionswere preparedby additionof equalaliquotsof [@“Tc]MIBl(1.9 fmoles/XCi) with increasing aliquots of [@Tc]MlBl. [@@@@TcjMlBl/[esTc]MIBI molar ratios are shown on the abscissa. Each pointrepresentsfour preparations.Linewas drawn by eye.

. , -@-J@

-6 -5 -4 -3 -2

466 The Journal of Nuclear Medicine •Vol. 31 •No. 4 •April 1990

I FIGURE4Effect of cationic membranetransportinhibitorson 1-mm [@“Tc]MlBluptake(A)and @°1Tluptake(B).Cellswere preincubatedin solutionscontaininginhibitors for 1 mmprior to determinationoftracer uptake in the continued presence of inhibitors. The following concentrationswereused:amiloride(100pM);ouabain(100 zM);bumetanide(10ILA#f); verapamil (1 SM). Values are

mean ±s.e.m. Numberof determinations are shown in parentheses.Therewere no significant differences between [es@@@Tc]MlBIvalues based onone-way analysis of variance (18). For20111values, + denotes p < 0.05 cornparedto control.

@-

cellular spaces in <30 sec in these cultured preparationsthereby producing rapid onset of action of these trans

port inhibitors secondary to the simple extracellularspaces (12,23)). This contrasted with the transport inhibition profile determined for 201T1uptake in thesepreparations. Unlike [99mTc]MIBI, 1-mm 201Tl uptakewas significantly inhibited by ouabain and bumetanide(Fig. 4B), consistent with the known physiologic similarities between thallium and potassium (K) (26,27).The different transport inhibition profiles of 1-mmuptakes between [@mTc]MIBI and 2OVflcould not beaccounted for by vastly different compartment sizesand uptake rate constants (thereby limiting applicationof the 1-mm uptake assay) since net cell uptakes forboth agents were comparable for 10 mm when normalized to constant extracellular concentration of tracer(Fig. 5). As opposed to unidirectional uptake, steadystate accumulation of [99mTc]MJBI in myocytes couldbe altered by prolonged cationic transport inhibition(Fig. 6). Plateau levels of [@mTc]MIBIwere not significantly different from control in buffers containing bumetanide or amiloride, but were actually increased inbuffers containing ouabain (p = 0.05) or verapamil (p< 0.01).

Since one primary clinical function of scintigraphic

myocardial perfusion imaging agents is to probe ischemic myocardium, the effect on [99mTc]MIBI cellularuptake kinetics of conditions associated with ischemiawere determined (Fig. 7). Site I of mitochondrial electron transport was inhibited with rotenone (10 @M)while glycolysis was inhibited with iodoacetate (IAA; 1mM) or 2-deoxyglucose (2D0; 1 mM) (28). Pre-incubation for 2.5 hr in rotenone alone partially inhibited1-mm [99mTcJMIBI cellular uptake (p < 0.05). Preincubation in IAA, either alone or with rotenone, nearlycompletely inhibited [99mTc]MIBI uptake (p < 0.01).

Pre-incubation in 2D0 alone showed no significanteffect; however, 2DG plus IAA markedly inhibited[99mTc]MIBIuptake (p < 0.01). Control experimentsrevealed no evidence of uptake inhibition (p > 0.5) incells pre-incubated in HEPES-buffered control solutionfor 2.5 hr compared with serum-containing culturemedia prior to determination of [99mTc]MIBIuptake.Thus, the absence of serum in the test solutions couldnot explain the metabolic effect on [@mTc]MIBIuptake.

Correlative ATP contents were determined after 2.5hr of metabolic inhibition (Table 1).Compared to ATPcontent ofcontrol cells pre-incubated in HEPES buffer(30. 1 ±1.4 nmoles ATP/mg protein), rotenone (10@LM)alone partially depressed ATP content and IAA

0C

E

0

FIGURE5Comparative [@“Tc]MlBl() and @°@Tl(0) net uptake intocultured chick myocytes. Data were normalizedto constantextracellular concentration of tracer and expressed as fCicellulartracer content/mg proteinper nCi extracellulartracer/liter (see methods).Each point is the mean ±s.e.m. of 3-4determinations.Solid lineswere drawn by eye.

0 5 10TIME (mln)

Metabolic Inhibition of [@“Tc]MlBlCellular Uptake •Piwnica-Worms et al 467

@O1t%@@@d'@os

80.@

60-@

40.@

C(4)

T

.@ _1 (13)

§@ooLLUi

4

z 20

0/

0 20 40

TIME (mm)FIGURE 6Effect of membrane transport inhibitors on steady-state[@“Tc1MlBlcontent. Cells were incubated in HEPESbuffercontaining inhibitors and tracer for the times indicated andtracer contentdeterminedas describedin legendto Figure1.Dataarenormalizedto constantextracellularconcentrationof[@“Tc]MIBIand expressedas frnolcellulartracer content/mgprotein . nM extracellular[es@@@TcJMlBl.Eachpoint is the mean±s.e.m. of four determinations. Solid lines were drawn byeye.(U)control;(@@)amilonde(100 EM);(@)ouabain(100 zM);(0) bumetanide(10 iM);(•)verapamil(1 SM).

(0. 1mM), either alone or in combination with rotenone( 10 zM), completely depleted cell ATP. 2DG (1 mM)alone had no effect on ATP content.

The concentration-effect curve for IAA inhibition of1-mm [99mTc]MIBIuptake is shown in Figure 8. Halfmaximal inhibition was determined to occur at @-5@M.Onset ofthe inhibitory action ofIAA was time-dependent (Table 2). Pre-incubation in saturating concentrations of IAA (1 mM) for up to 60 mm was withoutsignificant effect on 1-mm [99mTc]MIBI uptake. However, ATP content determined on the same cells wasdepressed to less than 10% of control values after 60mm of exposure to IAA (Table 2). Longer incubationsthen eventually did depress 1-mm [@mTc]MIBIuptakes.

DISCUSSION

To evaluate overall kinetic and cellular processesinvolved in myocardial extraction and retention of[99mTc]MIBI a minimal compartmental analysis wouldrequire a three-component model exchanging in series.The usual physiologic compartments assumed for thisanalysis allows [99mTc]MIBIexchange between intravascular spaces and interstitial(extracellular) spaces andsubsequent exchange between extracellular spaces andmyocytes. More complex compartmental models alsohave been proposed (29). Whole-animal and perfused

468 The Journal of Nuclear Medicine •Vol. 31 •No. 4 •April 1990

(12)

.A@.oo+;@c:%,4 44

( 4f4@60 FIGURE 7

Effect of metabolic inhibitors on 1-mm [@“TcJMlBIuptake.Cells were pre-incubatedfor 2.5 hr in HEPES buffer alone(control) or containing combinations of 2-deoxyglucose (1mM), rotenone (10 MM), and iodoacetate (1 mM) prior todetermination of 1-mm [@“Tc]MlBluptake in HEPES buffer inthe continued presence of inhibitors. Cells maintained inserum-containingculture media were also subsequentlyassayed for [@“Tc]MlBIuptake in HEPES buffer. Number ofdeterminationsare shown in parentheses.* denotesp < 0.01and + denotesp < 0.05 comparedto control uptake.

organ techniques may be best suited to evaluate overallexchange kinetics of [99mTc]MIBIbetween vascular andinterstitial spaces, but these techniques only indirectlyexamine exchange at the cellular level. The complexextracellular spaces and diffusion delays of whole-organmodels is well established in the physiologic literatureto severely limit evaluation of cellular exchange processes (12,30). Cultured chick heart cells are a physiologically stable myocardial preparation that has minimal extracellular diffusion distances (12,31) and lacksendothelial and neural elements that complicate traceranalysis in whole animal or intact tissue preparations.This study therefore examined cellular uptake kineticsand net accumulation of [99mTc]MIBIutilizing this wellcharacterized preparation (31 ). These data can potentially be combined with results from other techniquesthat examine vascular exchange to generate an overallscheme of cardiac extraction processes for 99mTc@isoni@triles.

Steady-State Analysis. Net cellular accumulation of[99mTc]MIBI approached a plateau by 40 mm with ati,2 of 9.3 ± 1.5 mm; mean t112of washout was 8 ±2mm. The presence ofa plateau implied that the cellularmechanisms of unidirectional influx and effiux were ina steady-state at plateau levels. A similar observationhas been made for [@mTc]CPI;however the uptake andwashout kinetics are approximately twice as rapid as[99mTc]MIBI(7). Exposure of cultured myocyte prepa

-7 -6 -5 -4 -3

LOG [IAA]

TABLE 2Effect of Pre-Treatment Time in lodoacetate on 1-Mm

[@‘Tc]MIBlUptakeandATPContentinCultured Heart Cells

Pretreatment1-mm[@“‘TcJMlBluptake ATPcontenttime (mm) (nCi/mg cell protein) (nrnoles/mgcell protein)

0 118 ±29 (n = 40.2 ±7.1 (3)

Cells were pre-incubatedin various inhibitors for 2.5 hr prior toassayfor ATPcontent.Eachvaluewas determinedin triplicateandshownas mean±s.e.m.

•denotesp<005Time-dependenceof IAA inhibition of [@‘TcJMlBluptake and

ATPdepletionin heartcells.Cellswere pre-incubatedin IAA(1mM) for various times prior to determinationof 1-mm[‘@“TcJMlBIuptake and ATP content in the same preparations. Values aremean±s.e.m.(n = numberof determinations).Two additionalseries of experiments on separate cultures confirmed the lack ofeffectof IAA(0.1mM and 1 mM) on [@Tc]MlBl uptakeduringpre-incubationperiodsof upto 60 mm,but inhibitionthereafter.

*p < 0.01.tp<O.05.ND= nonedetectable.

curves of the intravascular pool of [@mTcJMIBI.Therefore, myocytes in vivo would be exposed to a transientbolus of extracellular activity rather than a constantconcentration of [99mTc]MIBIas in this study. Nonethe less, steady-state analysis will yield quantified information regarding cellular exchange processes. Furthermore, these steady-state data indicated that the 1-mmuptake times used in later cellular kinetic analysis wereonly l%—2%of the time required to achieve net equilibration of [99mTc]MIBJ(40—60mm), and therefore,these short times could be used to isolate cellular unidirectional uptake kinetics relatively uncontaminatedby cellular effiux (back flux) (32).

Washout. Total unidirectional washout of [99mTc]MIBI from cultured heart cells (84% at 60 mm) wasgreater than net myocardial clearance in newborn chickin vivo (33% at 30 mm) and significantly greater thannet fractional clearance previously reported for dogs invivo (1 1% at 4 hr) (10). Net myocardial clearance of[99mTc]MIBIin human is 27% at 3 hr (6), intermediatebetween values found in dog and chick. Note that thelarge fractional washout from cultured heart cells represented a unidirectional effiux into isotope-free bufferoriginating from cells previously equilibrated in [@mTc1MIBI. Aside from species differences, these experimentsin vitro therefore isolated myocellular washout kinetics,whereas studies of the slower net myocardial clearancein vivo reflect a complex interaction of washout, retention and uptake between myocytes, interstitial spaces,blood vessels, and other organs.

Displacement. Apparent competitive displacement oftracer by carrier [@Tc]MIBI occurred at high molar

TABLE 1CellATPContentAfter MetabolicInhibition

ATPcontentInhibitor (nmoles/mgcellprotein)

Control2DG(1mM)rotenone (10 @tM)IAA(0.1 mM)IAA + rot

30.1±1.434.5±2.3

*20.6 ±1.4*0.4 ±0.3*21 ±0.3 I 5129 ±26(3)29.4 ±1.8(3)30173

±16 (3)26.7 ±4.6(3)60123 ±41 (4)2.8 ±1.1(4)105t43

± 9.8 (3)ND(3)150t34@45(3)•04±03(3)

rations to constant extracellular concentrations of[99mTc]MIBIcan assist in analyzing cellular mechanisms and kinetics of myocyte uptake and retention,but it is emphasized that the plateau level of net cellularuptake observed in this study was isolating only acomponent of the dynamic behavior of whole-hearttissue in vivo. For example, following an intravenousbolus injection of[99mTc]MIBI, myocardial tissue activity peaks during the first 30 to 120 sec in vivo (2,8,10),in large part secondary to the rapid intravascular clear

ance ofthe agent. Time-activity curves for [@mTc]MIBIin the myocardial interstitial spaces are likely to reflectto some degree the more easily measured time-activity

.5

‘U

4

FIGURE8Concentration-effectcurve for IAA inhibitionof 1-mm[@“Tc]MIBI uptake. Cells were pre-mncubatedfor 2.5 hr in HEPESbuffercontainingIAAof variousconcentrationsprior to determination of [@“Tc]MlBIuptake. Results are expressed aspercent of control uptake. Each point representsfour determinations.Solid linewas drawn by eye assumingan lC@of 5MM.

469MetabolicInhibitionof [es@@@Tc]MlBlCellularUptake•Piwnica-Wormset al

ratios (apparent KD@ 7 x l0@), which implied a lowaffinity cellular retention process. Apparent competitivedisplacement of tracer [99mTc]MIBIby high concentrations of carrier [99TcJMIBIalso has been observed inisolated guinea pig heart slices (3). Previous experiments with [99mTc]CPIin cultured heart cells similarlydemonstrated a low-affinity or nonspecific process ofcellular retention ( 7). Although these data could beconstrued to reflect a low-affinity receptor(s) for 99mTc..isonitriles, several alternative mechanisms are equallyconsistent with the data. Also possible are:

1. The onset of cellular toxicity at high extracellular

concentrations of [@Tc]MIBI, thereby decreasing

net cellular content. (However, note that clinicalimaging is performed far from these saturatingconcentrations of tracer; typical extracellular concentrations of 99mTc@isonitrilesduring clinical imaging are approximately l06-fold lower at ‘tol0°M (7).)

2. The presence of a high capacity subcellular orpericellular compartment that sequestered [99mTc]MIBI. In this regard, preliminary data from cultured cells point to the presence of subcellularcompartmentation of [99mTcJMIBJ within mito

chondria (33). This finding would also furtherjustify the use of the term “uptake―in the contextof cellular tracer studies with [99mTc]MIBI.

Transport Inhibition. The tested cation transportblockers did not inhibit [99mTc]MIBI uptake kineticsand, furthermore, no transport inhibitor decreased plateau accumulation of [99mTc]MIBI. These findingswould be incompatible with direct inward transport of[99mTcJMIBIby a cationic binding site on these membrane transport proteins. However, prolonged exposureof myocytes to ouabain or verapamil actually increased[99mTc]MIBI steady-state accumulation suggesting thatindirect effects of Na/K pump inhibition or Ca2@channel blockade either increased a myocellular compartment for [99mTcJMIBIor decreased [@mTcJMIBIefflux.

Metabolic Inhibition. Controversy exists as towhether distribution of [@mTc]MIBIin vivo is a simplefunction of blood flow or whether it is responsive tomyocardial metabolic function. Preliminary data inisolated guinea pig heart slices indicated that [@mTc]MIBI uptake was inhibited by hypoxia produced bypre-equilibration in buffer gassed with nitrogen/carbondioxide (3). First-pass extraction fractions in isolatedrabbit hearts have been reported to increase after 30mm of reperfusion preceeded by global ischemia (34).The investigators concluded that this reflected in someway myocardial damage and was independent of coronary flow. In addition, [@mTc]MIBI washout from isolated rat heart was increased by ischemic flow rates (3ml/min) relative to control rates (12 ml/min) suggesting

a component of [99mTc]MIBImyocardial retention sensitive to myocardial metabolism (35). However, otherinvestigators report that 15 mm of metabolic blockadewith cyanide and iodoacetate in cultured rat myocyteshad no effect on [99mTc]MIBIuptake (36).

Our data demonstrate strong inhibition of [99mTc]MIBI uptake kinetics by metabolic blockade. Nearlycomplete inhibition of 1-mm [99mTc]MIBIuptake occurred after pre-incubation of heart cells for 2.5 hr inIAA (1 mM) or IAA plus rotenone ( I0 SM). The halfmaximal inhibitory dose ofIAA was @5@zM.However,onset of [99mTc]MIBIuptake inhibition was markedlytime-dependent; pre-incubation periods ofup to 60 mm

([IAA] = 0. 1—1mM) caused no significant inhibition,but thereafter markedly reduced [99mTc]MIBIuptake.In agreement with the early time points, Maublant etal. (36) previously reported no effect of a 15-mm preincubation period in IAA (0. 1 mM) or IAA plus cyanide(5 mM) on subsequent 20-mm [99mTc]MIBIuptake incultured rat myocytes. However, it was likely that their15-mm pre-incubation period was too short to elicit the

marked inhibitory effect of prolonged metabolic blockade observed in this report from cultured chick myocytes. Relatively long incubation times in the presenceof metabolic inhibitors before induction of cell injuryis a characteristic previously reported for cultured ratand chick heart cell models in the study of ischemiarelated conditions (28,37).

Although IAA is generally used as an inhibitor ofglycolysis (glyceraldehyde-3-phosphate dehydrogenase),other mechanisms of action may occur and could involve sulthydral carboxymethylation of other proteinswithin the cytosol, cellular membranes and organelles(38). However, the relatively slow onset of IAA effectswere most consistent with depletion of a substrate orenzyme. The lack of effect of 2DG on [@mTc]MIBIuptake may be due to greater specificity of action onglycolysis or less efficacy as a metabolic inhibitor atconcentrations of 1 mM (39). Note that the lack ofeffect of 2DG or short exposures to IAA on 1-mm[99mTc]MIBI uptake does not exclude a possible effectof the inhibitors on steady-state levels of accumulationof[99mTc]MIBI.Further experiments would be requiredto fully explore changes in unidirectional myocellularuptake rates, washout rates, and cellular compartmentsizes of[@mTc]MIBIduring prolonged membrane transport inhibition or metabolic blockade to clarify potential effects of these manipulations on clinical images.

What is the source of the metabolic effect on [@mTc]MIBI cellular uptake? Myocyte ATP content (alone orin part) does not fully account for the observations. Forexample, [@mTc]MIBIuptake was unaffected by 60 mmof metabolic blockade in IAA (1 mM) at a time whenATP simultaneously determined in the same preparations was <10% of control values (Table 2). However,with more prolonged metabolic inhibition, [99mTc]MIBI

470 The Journal of Nudear Medicine •Vol. 31 •No. 4 •April 1990

uptake eventually declined. In addition, Murphy et al.(28) have examined ATP content, lactate dehydrogen

ase (LDH) release, and intracellular Na, K, and Ca

contents in cultured chick heart cells exposed to IAA(1 mM) and rotenone (0. 1 mM) for up to 2.5 hr. After

60 mm of metabolic inhibition at normal extracellularCa concentrations, ATP content was <2 nmole/mgprotein (control: 40 nmole/mg protein) and intracellular Na and K contents had approached their extracellular values. No further changes in these parametersoccurred up to 2.5 hr. LDH release into the supernatant,on the other hand, was <4% of total cellular LDH at60 mm, but reached nearly 50% of total LDH at 2.5hr. One-minute [99mTc]MIBIuptake values from ourstudy under similar conditions appeared to correlateinversely with LDH release and again correlated poorlywith ATP or intracellular cation content, suggestingthat a component of “lateinjury―or cell death causedinhibition of [99mTc]MIBI myocellular uptake.

The delayed effect of metabolic inhibition on [99mTc]MIBI uptake stands in contrast to the earlier effect ofmetabolic inhibition on TI uptake in similar models.McCall et al. (40) reported that 2°'@Tlinflux into cultured rat myocytes decreased during the first 30—40mmofpre-treatment with 2,4-dinitrophenol (1 mM) or IAA(0. 1 mM) in a time course that paralleled the declinein ATP content during metabolic inhibition determined

in this study and by others (37,39,41). This supportedthe proposal that 60%—70%of 2°―Tluptake was mediated by the plasma membrane Na/K ATPase. Thedivergent responses of 2°@Tland [99mTc]MIBI uptakesto transport and metabolic inhibition add further evidence for differing mechanisms of cellular uptake andretention. Additional experiments are required to explore these complex relationships between perfusionagent uptake and myocardial cell injury.

In summary, E99mTc]MIBInet uptake into culturedheart cells approached a plateau level, which impliedan equilibrium or steady-state process involved in netcellular accumulation of the agent. Conditions associated with ischemia induced by prolonged metabolicblockade with IAA and rotenone markedly inhibited[99mTc]MIBIcellular uptake kinetics. Although [@mTc]MIBI has been shown to be a flow-dependent tracer,our data also indicate that myocellular net accumulation and uptake kinetics can be affected, respectively,by pharmacologic alterations in membrane transportand metabolic status.

ACKNOWLEDGMENTSThe authors thank Mary Chiu and Santwana Mukherjee

for their technical assistance and Joyce Alexander and CrystalPorter for their secretarial assistance.

This work was supported by grants from the NIH (CA34970, HL42966, HL262 15) and the DOE (ER60526).

REFERENCES

1. Jones AG, Abrams MJ, Davison A, et al. Biological studiesof a new class of technetium complexes: the hexakis (alkylisonitrile) technetium (I) cations. mt J Nuci Med Biol 1984:11:225—234.

2. Piwnica-Worms D, Kronauge JF, Holman BL, Davison A,Jones AG. Comparative myocardial uptake characteristics ofhexakis (alkylisonitrile) technetium (I) complexes: effect oflipophilicity.Invest Radiol 1989;24:25—29.

3. Mousa SA, Williams SJ, Sands H. Characterization ofin vivochemistryofcations in the heart. J Nucl Med 1987;28:1351—1357.

4. Holman BL, Jones AG, Lister-James J, et al. A new Tc-99m-labeled myocardial imaging agent, hexakis (t-butylisonitrile)technetium(l) [Tc-99m TBI]: initial experience in thehuman.JNuclMed 1984;25:1350—1355.

5. Sia STB, Holman BL, Campbell S, et al. The utilization oftechnetium-99m CPI as a myocardialperfusion imagingagentin exercise studies. C/in Nuc/ Med 1987; 12:681—687.

6. Wackers FJ, Berman D, Maddahi J, et al. Technetium-99mhexakis 2-methoxyisobutyl isonitrile: human biodistribution,dosimetry, safety, and preliminary comparison to thallium201 for myocardial perfusion imaging. J Nuc/ Med 1989;30:301—311.

7. Piwnica-Worms D, Kronauge JF, Holman BL, Lister-JamesJ, Davison A, Jones AG. Hexakis (carbomethoxyisopropylisonitnle)technetium (I),a newmyocardialperfusionimagingagent: binding characteristics in cultured chick heart cells. JNuc/Med 1988;29:55—61.

8. Kronauge JF. Functional isonitriles for use as myocardialperfusion imaging agent. [Thesis]. Cambridge, MA: Massachusetts Institute ofTechnology, 1987, p 260.

9. Kronauge iF, Davison A, Lister-James J, Noska M, JonesAG. Interspeciescomparison of the distribution of Tc-CPI. JNuclMed 1987; 28:601.

10. Okada RD. Glover D, Gaffney T, Williams S. Myocardialkinetics oftechnetium-99m hexakis-2-methoxy-2-methylpropyl-isonitrile. Circulation 1988; 77:491—498.

11. Horres CR, Lieberman M, Purdy JE. Growth orientation ofheart cells on nylon monofilament. J Membr Biol 1977;34:313—329.

12. Horres CR, Lieberman M. Compartmental analysis of potassium efflux from growth-oriented heart cells. J Membr Biol1977;34:331—350.

13. Barry WH, Biedert S, Miura DS, Smith TW. Changes incellular Na, K, and Ca contents, monovalent cation transportrate, contractile state during washout of cardiac glycosidesfrom culture chick heart cells. Circ Res 1981;49:141—149.

14. Kim D, Smith TW. Effects of amiloride and ouabain oncontractile state, Ca and Na fluxes, and Na content in culturedchick heart cells. Mol Pharmacol 1986; 29:363—371.

15. Lowry OH, Rosebrough NJ, Farr AL, Randell RJ. Proteinmeasurement with the Folin phenol reagent. J Biol Chem1951;193:265—275.

16. Lamson ML Kirschner AS, Hotte CE, Lipsitz EL Ice RD.Generator-produced 99mTc..T@1J_4:carrier free? J NucI Med1975;16:639—641.

17. Deutsch E, Heineman WR, Zodda JP, Gilbert TW, WilliamsCC. Preparationof―no-carrieradded―technetium-99m cornplexes: determination of the total technetium content ofgenerator elements. mt ApplRadiat Isot 1982;33:843—848.

18. Lieberman M, Manasek FJ, Sawanobori T, Johnson EA.Cytochalasin B: its morphological and electrophysiologicalactions on synthetic strands ofcardiac muscle. Dr.'.'Biol 1973;31:380—403.

19. Williamson JR. Corkey BE. Assays of intermediates of the

Metabolic Inhibition of [es@@@Tc]MlBlCellular Uptake •Piwnica-Worms et 81 471

citric acid cycle and related compounds by fluorometric enzymemethods.MethodsEnzymo/1969;13:434—513.

20. Wallenstein S, Zucker CL, Fleiss JL. Some statistical methodsuseful in circulation research. CircRes 1980; 47:1—9.

21. Remington RD. Schork MA. Statistics with applications tothe biologicaland health sciences.Englewood,NJ: PrenticeHall;1970:210—215.

22. BenosDi. Amiloride:a molecular probe ofsodium transportin tissues and cells. Am JPhysiol 1982; 242:C13l—C145.

23. Piwnica-Worms D, Jacob R, Horres CR, Lieberman M. Na/H exchange in cultured chick heart cells: pH1 regulation. JGen Physiol 1985;85:43—64.

24. Liu S, Jacob R, Piwnica-Worms D, Lieberman M. (Na + K+ 2C1)co-transport in cultured chick heartcells.Am JPhysiol(Cell)1987;253:C72l—C730.

25. Barry WH, Smith TW. Mechanisms of transmembrane calcium movements in cultured chick embryo ventricular cells.JPhysiol 1982; 325:243—260.

26. Britten iS, Blank M. Thallium activation of the Na, K activated ATPase of the rabbit kidney. Biochem Biophys Acta1968;159:160—166.

27. Gehring PJ, Hammond P. The interrelationship betweenthallium and potassium in animals. J Pharmacol Exp Ther1968;155:187—201.

28. Murphy E, LeFurgey A, Lieberman M. Biochemical andstructuralchanges in cultured heart cells induced by metabolicinhibition. Am J Physiol (Cell) 1987; 253:C700—C706.

29. BassingthwaighteJB. Overview of the processesof delivery:flow, transmernbrane transport, reaction, and retention. Circulation 1985; 72(supp H):iv39—46.

30. Auwell D, Eisner D, Cohen I. Voltageclamp and tracer fluxdata: effectsofa restrictedextracellularspace. Q Rev Biophys1979;12:213—261.

31. Pinson A, ed. The heart cell in culture, Volumes 1—3.BocaRaton: CRC Press; 1987.

32. McCall D, Zirnrner II, Katz AM. Kinetics of thallium cx

change in cultured rat myocardial cells. Circ Res 1985;56:370—376.

33. Chiu M, Kronauge JF, Piwnica-Worms D. Monitoring mitochondrial and plasma membrane potentials in cultured cellswith a new 99mTcb@d lipophilic cation. J Gen Physiol 1989;94:4lA.

34. Meerdink DJ, Thurber M, SavageS. Leppo JA. Reperfusioneffectson the myocardial extraction ofa technetium-labelledisonitrile analogue (RP3O) and thallium-201. J Nuci Med1988; 29:8 19.

35. Okada RD, Glover DK. Myocardial kinetics of technetium99m-hexakis-2-rnethoxy-2 methylpropylisonitrile (RP-30)under flowand work controlled conditions. JNuclMed 1988;29:792.

36. Maublant JC, Gachon P. Moms N. Hexakis (2-methoxyisobutylisonitrile) technetium-99m and thallium-20l chloride: uptake and release in cultured myocardial cells. J NuclMed 1988; 29:48—54.

37. Buja LM, Hager HK, Parsons D, Chien K, Reynolds RC,Willerson JT. Alterations of ultrastructure and elementalcomposition in cultured neonatal rat cardiac myocytes aftermetabolic inhibition with iodoacetic acid. Lab Invest 1985;53:397—412.

38. Williamson JR. Metabolic control in the perfused rat heart.In: Chance B, Estabrook RW, WilliamsonJR, eds. Controlofenergymetabolism. London: Academic Press; 1965:333.

39. Doorey AJ, Barry WH. The effectsof inhibition of oxidativephosphorylation and glycolysis on contractility and highenergy phosphate content in cultured chick heart cells. CircRes 1983;53:192—201.

40. McCallD, ZimmerLi, KatzAM.Effectof ischemia-relatedmetabolic factors on thallium exchange in cultured rat myocardial cells. Can J Cardiol 1986; 2:176—183.

41. Cheneral JP, Hyde A, Blondel B, Girardier L. Heart cells inculture: metabolism, action potential and transmembraneionic movements. J Physiol (Paris) 1972;64:413—430.

472 The Journal of Nuclear Medicine •Vol. 31 •No. 4 •April 1990