PERSPECTIVE

Mitochondrial calcium uptakeGeorge S. B. Williamsa,b,1, Liron Boymana,1, Aristide C. Chikandoa, Ramzi J. Khairallaha,c, and W. J. Lederera,2aCenter for Biomedical Engineering and Technology, and Department of Physiology, University of Maryland School of Medicine, Baltimore,MD 21201; bSchool of Systems Biology, College of Science, George Mason University, Manassas, VA 20110; and cDepartment of Celland Molecular Physiology, Stritch School of Medicine, Loyola University, Chicago, IL 60153

Edited by David E. Clapham, Howard Hughes Medical Institute, Children’s Hospital Boston, Boston, MA, and approved May 15, 2013 (received for review January 14, 2013)

Calcium (Ca2+) uptake into the mitochondrial matrix is critically important to cellular function. As a regulator of matrix Ca2+ levels, this fluxinfluences energy production and can initiate cell death. If large, this flux could potentially alter intracellular Ca2+ ([Ca2+]i) signals. Despiteyears of study, fundamental disagreements on the extent and speed of mitochondrial Ca2+ uptake still exist. Here, we review and quantita-tively analyze mitochondrial Ca2+ uptake fluxes from different tissues and interpret the results with respect to the recently proposed mito-chondrial Ca2+ uniporter (MCU) candidate.This quantitative analysis yields four clear results: (i) under physiological conditions, Ca2+ influx intothe mitochondria via the MCU is small relative to other cytosolic Ca2+ extrusion pathways; (ii) single MCU conductance is ∼6–7 pS (105 mM[Ca2+]), and MCU flux appears to be modulated by [Ca2+]i, suggesting Ca2+ regulation of MCU open probability (PO); (iii) in the heart, twofeatures are clear: the number of MCU channels per mitochondrion can be calculated, and MCU probability is low under normal conditions; and(iv) in skeletal muscle and liver cells, uptake per mitochondrion varies in magnitude but total uptake per cell still appears to be modest. Basedon our analysis of available quantitative data, we conclude that although Ca2+ critically regulates mitochondrial function, the mitochondria donot act as a significant dynamic buffer of cytosolic Ca2+ under physiological conditions. Nevertheless, with prolonged (superphysiological)elevations of [Ca2+]i, mitochondrial Ca2+ uptake can increase 10- to 1,000-fold and begin to shape [Ca2+]i dynamics.

inner mitochondrial membrane | microdomain | NCLX | SERCA | NCX

Early work on mitochondrial biology (1) sug-gested calcium (Ca2+) is an important factorthat regulates mitochondrial function, andthis concept has been broadly supported byothers (2). In 1992, a very influential study(3) showed that mitochondrial Ca2+ uptakewas a genuine cellular mechanism. Recently,investigations in diverse tissues, includingcardiac and skeletal muscle, liver cells, andneurons, have raised the possibility that mi-tochondria also serve as large and dynamicphysiological buffers for Ca2+ (4–15). How-ever, this point has been debated (16), andforces the question: How much Ca2+ movesinto and out of the mitochondria under nor-mal conditions? Here, we seek to answer thisquestion quantitatively using measurementstaken from the literature, and we attempt toplace the numbers in the context of recentinsights into mitochondrial Ca2+ uniporter(MCU) function (17–21).This critical review presents and compares

quantitative mitochondrial Ca2+ influx datafrom eight independent studies of cardiacmitochondrial Ca2+ uptake (20–27), four ofskeletal muscle (21, 28–30), and seven of liverand model cells (18, 21, 31–35). The inves-tigations were carried out in experiments us-ing a wide repertoire of methods that includeCa2+ uptake by mitochondria in suspensionand intact cells, electrophysiological record-ings of Ca2+ currents from mitoplasts, andlipid bilayer experiments conducted on recentMCU candidates. These reports have beenused to argue both in favor of and against

the role of mitochondria as dynamic buffersof intracellular Ca2+ ([Ca2+]i). A surprisingresult from the quantitative analysis is theconsistency of the cardiac uptake results de-spite spanning 40 years of experiments utiliz-ing diverse methods and tissues.

Mitochondrial Ca2+ UptakeCardiac ventricular myocytes seem like theideal environment in which to study Ca2+

flux across the inner mitochondrial mem-brane (IMM) and the possible role(s) of localmicrodomains of elevated [Ca2+]i. In themammalian heart, mitochondria are moreabundant per volume element than in anyother tissue and account for approximatelyone-third of the cell volume (36). These mi-tochondria experience regular, repetitiveelevations in extramitochondrial Ca2+ (i.e.,local elevations of cytosolic [Ca2+]i). Theabundant intermyofibrillar mitochondria(IFM) (37) are in close proximity to thesarcoplasmic reticulum (SR), the primaryintracellular Ca2+ storage organelle, whichreleases Ca2+ with every heartbeat (Fig. 1).Although the cell-wide (global) [Ca2+]iincreases from 100 nM to ∼500 nM (Fig. 1,green arrow) with each heartbeat (38), themicrodomain [Ca2+]i near the ends of theIFM may transiently rise to 10–20 μM duringthe release phase (Fig. 1, red and green lines)(39–41). This high local [Ca2+]i occurs be-cause of the close proximity of the mitochon-drial ends to the SR Ca2+ release units (CRUs)located between the transverse tubule (TT)

and junctional sarcoplasmic reticulum (JSR)membranes. The “diffusional distance” fromthe ryanodine receptor 2 (RyR2) clusters inthe JSR that face the TTs to the ends of themitochondria is approximately 50–100 nm.The actual exposure time to high local [Ca2+]i(10–20 μM) is quite brief, approximately10 ms in the heart (42), which is significantlyfaster than exposures used in some experi-mental conditions (4, 6–10, 13, 14, 43, 44).

Mitochondrial [Ca2+]i Influx in the Heart.The [Ca2+]i transient in heart cells pro-duces repetitive Ca2+ elevations that enve-lope the mitochondria with every heartbeat.This Ca2+ is then removed from the cy-tosol via the sarcoplasmic reticulum/endo-plasmic reticulum Ca2+-ATPase (SERCA)and the sarcolemmal Na+-Ca2+ exchanger(NCX), resulting in cell-wide reduction of[Ca2+]i within 500 ms (36). During this time,mitochondria have the opportunity to

Author contributions: G.S.B.W., L.B., A.C.C., R.J.K., and W.J.L. de-

signed research; G.S.B.W., L.B., A.C.C., and R.J.K. performed re-

search; G.S.B.W. contributed new reagents/analytic tools; G.S.B.W.,

L.B., A.C.C., and R.J.K. analyzed data; and G.S.B.W., L.B., A.C.C.,

R.J.K., and W.J.L. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

1G.S.B.W. and L.B. contributed equally to this work.

2To whom correspondence should be addressed. E-mail: jlederer@umaryland.edu.

This article contains supporting information online at www.

pnas.org/lookup/suppl/doi:10.1073/pnas.1300410110/-/

DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1300410110 PNAS | June 25, 2013 | vol. 110 | no. 26 | 10479–10486

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

July

23,

202

0

sequester Ca2+ from the cytosol through openMCUs. To facilitate the comparison of theseCa2+ fluxes, we scaled all data to a liter ofcytosol (see Table S1). Conventional wisdomis that 70–80 μmol of Ca2+ enters a liter ofcytosol during a [Ca2+]i transient in ven-tricular myocytes and that this Ca2+ issubsequently removed during a contraction–relaxation cycle (25, 45).

A fair comparison of mitochondrial Ca2+

uptake reported by various sources usingdiverse experimental techniques requires thatthe data be converted to the same units (i.e.,micromoles of Ca2+ per liter of cytosol persecond; details are provided in Eqs. S1–S3and Figs. S1–S3). Although the opinions ofdifferent groups are substantially at odds withregard to the role of the ensemble of mito-chondria as a dynamic Ca2+ buffer in theheart, the measurements of mitochondrialCa2+ fluxes appear to be in good agreementwith each other when compared quantita-tively (Fig. 2). Importantly, the whole-cellmitochondrial Ca2+ uptake flux (whole-cell MCU flux) derived from experimentsconducted in cells (Fig. 2, filled circles)does not differ appreciably from uptakemeasured in suspensions of isolated

mitochondria (Fig. 2, open circles) oncescaled appropriately. The solid black linein Fig. 2A (“Cardiac MCU”) represents anempirical best-fit line to the experimentalresults of the form y = mxp, where m = 0.67and p = 1.7, and this line is subsequentlyused for a comparison between the whole-cellMCU flux in other cell types and uptakefrom other Ca2+ transport systems (i.e.,SERCA and NCX). For example, whole-cellMCU flux in the liver (Fig. 2B, colored datapoints) is qualitatively similar to that in theheart (Fig. 2B, black line) and is small overthe physiological range of [Ca2+]i. Notethat when [Ca2+]i is experimentally elevatedbeyond the physiological range (44) to levelsgreater than 200 μM, the measured flux isstill consistent with the cardiac best-fit line(see above). Interestingly, the whole-cellMCU flux in skeletal muscle (Fig. 2C,colored data points) (28) can exceed thatseen in heart (Fig. 2C, solid line) in somecases but is still relatively small over thephysiological range of [Ca2+]i. A detaileddiscussion regarding mitochondrial Ca2+

uptake across mitochondria from differenttissues is provided below.

When the whole-cell MCU flux is com-pared with other major cytosolic Ca2+ fluxes(i.e., SERCA and NCX) present in the heartover the physiological range of [Ca2+]i(Fig. 2D), both theoretical (46, 47) (Fig. 2D,solid lines) and experimental (25) (Fig. 2D,filled circles) measurements suggest thatMCU flux is small by comparison. Thisdramatic difference in magnitude (Fig. 2D,compare red and blue lines with the blackline) suggests that most of the Ca2+ that isadded to the cytosol during the contractionor during a Ca2+ spark is resequestered intothe SR by SERCA or extruded from the cellby NCX. This becomes even more apparentwhen each flux is scaled by all cytosolic re-moval fluxes (Fig. 2E). Although these valuesare for small rodents, the result is similar(Fig. S4) for larger animals (e.g., rabbit) thathave less SERCA and more NCX activity.Importantly, Fig. 2 D and E clearly showsthat fluxes attributed to SERCA and NCX aresignificantly larger than those of MCU overthe physiological range of [Ca2+]i (i.e.,0.1–20 μM). This comparison strongly sug-gests that mitochondrial fluxes have a minoreffect on [Ca2+]i dynamics. Note that thisconclusion is consistent with earlier work byBers and colleagues (48, 49), who showedthat mitochondrial uptake accounted forapproximately 1% of cytosolic Ca2+ “extru-sion.” It is, however, important to note thatthe transport fluxes (SERCA and NCX) dosaturate at high [Ca2+]i levels and that thiswould theoretically allow the whole-cellMCU flux to exceed even SERCA when[Ca2+]i is “superphysiological.” However,fluxes of this magnitude would depolarizethe IMM (18), thereby reducing the drivingforce for Ca2+ entry into the mitochondria.Importantly, under normal conditions,even the microdomain [Ca2+]i that bathesthe ends of an IFM with high [Ca2+]i is un-likely to exceed 20 μM (39–41) and the re-mainder of the IFM experiences much less.We believe this to be true for both heart andskeletal muscle.

Biophysical Properties of MCU Ca2+

Uptake. The cardiac whole-cell MCU flux,measured experimentally over a wide range of[Ca2+]i values, can be visualized (Fig. 3)alongside a theoretical flux (blue line) of theform

Jmcu =PONmitoNmcuimcu

zFVmyo;  

where   

imcu = gmcuðΔΨm −ECa2+ Þ; [1]

where Nmito is the number of mitochondriaper cell (Nmito = 10,000) (50), Nmcu is the

Fig. 1. Schematic diagram shows the spatial distribution of cardiac mitochondrial Ca2+ signaling components. Aspatial representation of a Ca2+ spark (red gradient) initiated at the CRU, which is located between the TT and JSRmembranes, is shown. At the peak of a Ca2+ spark, [Ca2+]i briefly (10 ms) bathes the mitochondrion at levelsindicated by the red line (note that the y axis is log scale). During a [Ca2+]i transient, multiple CRUs release Ca2+,causing the mitochondrion to experience a [Ca2+]i profile similar to the green line. Note that the JSR ends of themitochondria are in high [Ca2+]i microdomains for brief periods during both Ca2+ sparks and Ca2+ transients. Thegreen arrow indicates the approximate global average [Ca2+]i at the peak of a systolic [Ca2+]i transient. LCCs, L-typeCa2+ channels; RyRs, ryanodine receptor 2’s.

10480 | www.pnas.org/cgi/doi/10.1073/pnas.1300410110 Williams et al.

Dow

nloa

ded

by g

uest

on

July

23,

202

0

number ofMCUs permitochondrion (Nmcu=200) [estimated from a single-channel MCUcurrent (imcu) and a whole-mitoplast MCUcurrent (Imcu) by Fieni et al. (21); detailsregarding this estimate are provided inSI Material], ΔΨm is the mitochondrial innermembrane potential, ECa2+ is the Nernst re-versal potential for the IMM, F is Faraday’sconstant, Vmyo is the cytosolic volume of thecell (18 pL), and PO is the open probability(PO = 0.9). Additionally, the single-channelMCU conductance (gmcu) is assumed to fol-low a Michaelis–Menten type relationship,with a Km of 19 mM and a maximal gmcu of8.1 pS (17, 18, 20, 51) (Eqs. S7 and S8 andFig. S7). This relationship between ionchannel conductance and ion concentrationis a known biophysical feature of selectiveionic channels (more information is pro-vided in ref. 52). Additionally, althoughelectrophysiological measurements of MCUcurrents have been carried out by differentinvestigators using different experimentalapproaches, the maximal single-channelgmcu estimated from these measurementsappears to be similar (17, 18, 51). There-fore, we assumed a maximal single-channelgmcu of 8.1 pS in order to estimate the numberof MCU channels per cardiac mitochondrion.We calculate that the Nmcu is approximately200 (details are provided in Eqs. S5–S10 inSI Material). It is important to note that ourwhole-cell MCU flux formulation (Eq. 1) isrobust. Specifically, a change in measuredsingle-channel gmcu would be balanced bya corresponding change in the estimatedNmcu

(Eq. S9). For example, if a twofold increase insingle-channel gmcu were to be observed (i.e.,8.1 to 16.2 pS), a twofold reduction in theNmcu would be implied. In this scenario, noother elements of our formulation (i.e., Eq. 1)require any additional adjustment. In Eq. 1,the imcu (measured in picoamperes, as givenby Ohm’s law and the Nernst reversal po-tential) is converted to a whole-cell MCU flux(Jmcu, in μM s−1) scaled to the cytosolic vol-ume of a ventricularmyocyte.When the IMMis polarized, this flux formulation will yieldresults similar to the Goldman–Hodgkin–Katzformulation. Interestingly, over the physio-logical range of [Ca2+]i (0.1–20 μM), thetheoretical MCU flux (Fig. 3A, blue line)clearly supersedes the measured experimentalfluxes (Fig. 3A, gray circles). Additionally, theexperimental fluxes appear to increase non-linearly (Fig. 2A) with [Ca2+]i, suggesting thepossibility of Ca2+ regulation of MCUuptake. To investigate this issue, we fit theexperimental data with the same flux equationscaled by a PO function given by

A

B C

D E

[Ca2+]i2+]i

Ca2+

-12+

-1

[Ca2+]i2+]i

Ca2+

-1

Heart

Liver

[Ca2+]i

Skeletal

Heart HeartSERCA

NCX

NCX

SERCA SERCA+NCX

0

0.2

0.4

0.6

0.8

1

0 5 10 15 200

50

100

150Wei 2011Fry 1984Crompton 1985Zhou 1998

Scarpa 1973

0 5 10 15 200

100

200

300

400

0 5 10 15 200

50

100

150

0 5 10 15 200

50

100

150

Vinogradov 1973Bygrave 1975Gunter 1998Gunter 1990

Fig. 2. Whole-cell MCU fluxes from heart, liver, and skeletal muscle tissue. (A) Peak cardiac whole-cell MCU fluxes fromdifferent research groups (colored data points). Filled circles indicate measurements from intact mitochondria withincells, and empty circles indicate results from suspensions of isolated mitochondria. The solid black line is a “best-fit”line to all experimental data. (B) Peak liver whole-cell MCU fluxes from several different research groups (colored datapoints) plotted alongside the “Cardiac MCU” line from A. (C) Peak skeletal muscle whole-cell MCU fluxes from several dif-ferent research groups (colored data points) plotted alongside the Cardiac MCU line from A. FT, fast-twitch muscle; ST, slow-twitch muscle. Also see Figs. S1–S3. (D) Cardiac whole-cell MCU flux (black line and black circle) comparedwith other cardiacCa2+ extrusion pathways: SERCA (red line and red circle) and NCX (blue line and blue circle). Solid lines indicate theoreticalfluxes from the studies of Tran et al. (46) and Weber et al. (47), whereas filled circles indicate experimental results fromBassani et al. (25). Also see Fig. S5. in SI Material. (E) Comparison of the relative contribution of each flux when scaled tototal cytosolic extrusion. For example, the bar labeled MCU is calculated as the whole-cell MCU flux divided by the sum ofthe three whole-cell fluxes (i.e., MCU, SERCA, and NCX) shown in Fig. 2D. Conversions and scaling for the heart are 40 mgof mitochondrial protein per milliliter of cell, a cytosolic-to-cellular volume ratio of 0.5, and 10,000 mitochondria per cell(50). Details regarding other unit conversions and scaling are provided in SI Material (see Eqs. S1–S3 and Table S1). Notethat the SERCA and NCX fluxes shown in D and E are for small rodents, but the result is qualitatively similar (Fig. S4) forlarger animals (e.g., rabbit) that have less SERCA andmore NCX activity. Cardiac uptakemeasurements (22–27), liver uptakemeasurements (31–35), and skeletal muscle uptake measurements (28–30) are taken from the literature.

Williams et al. PNAS | June 25, 2013 | vol. 110 | no. 26 | 10481

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

July

23,

202

0

PO = ðPmax −PminÞ� ½Ca2+�ηiKηm;act + ½Ca2+�ηi

�+Pmin;

[2]

where Pmax is the maximal open probability,

Pmin is the minimum open probability, η is

the cooperativity factor, and Km,act is the half-

saturation constant for MCU open probability.

The red line in Fig. 3A shows the resultingbest-fit line when Pmax = 0.9 (18) and η =1.45, Pmin = 0.03, and Km,act = 108 μM (Fig.3B). The nonlinear increase in experimentalfluxes (Fig. 2A, Cardiac MCU line) and theremarkable agreement between the uptakemeasurements (Fig. 2A, gray circles) and thered line in Fig. 3A suggest that [Ca2+]i maybe responsible for modulating (i.e., activating)MCU Ca2+ uptake. Although the details arenot clear, such a mechanism has beenproposed before (53, 54). One possiblemechanism for this Ca2+ regulation is mito-chondrial calcium uptake 1 (MICU1), whichencodes a mitochondrial EF hand protein(55) and appears to regulate MCU uptake. Infact, a recent study showed that MICU1 al-tered mitochondrial Ca2+ uptake at low Ca2+

levels (56). There are, however, other possible

causes of nonlinear MCU uptake, including

the following:

i) Changes in the ΔΨm. Although large Ca2+

uptake fluxes certainly influence ΔΨm (33,53), these fluxes should depolarize theIMM, causing the slope of Ca2+ uptake todecrease with larger [Ca2+]i, rather than in-creasing as seen in Fig. 2A.

ii) Active transport pathways:

a) The mitochondrial NCX transporter(NCLX) that extrudes Ca2+ from themitochondrial matrix (57), is a possi-ble cause of apparent nonlinear up-take. However, NCLX efflux shouldcause the slope of the influx to decline,rather than increase. Note that NCLX-based Ca2+ influx is also possible underrare conditions (58, 59) but is very un-likely in the conditions shown here.

b) Leucine zipper-EF-hand containingtransmembrane protein 1 (Letm1), anelectrogenic Ca2+/H+ exchanger that op-erates at nanomolar [Ca2+]i, is anotherpossible cause of nonlinear uptake. Atelevated matrix Ca2+ ([Ca2+]m) levels

or low cytosolic pH, Letm1 is a Ca2+

extrusion mechanism, whereas atlow [Ca2+]m, it provides additionalCa2+ influx accompanied by matrixalkalinization (60). As such, Letm1reverses transport direction very earlyin the Ca2+ transient and might con-tribute to nonlinearity of Ca2+ uptake.Of note, Letm1 is also inhibited byRuthenium Red (RuR) and is not pres-ent in excitable cells.

iii) Extramitochondrial Ca2+ buffers. In theisolated mitochondria experiments, thepresence of unaccounted Ca2+ buffers(e.g., membrane-bound proteins, remnantCa2+ buffers used during mitochondrialisolation) may account for the nonlinearremoval of Ca2+ from the extramitochon-drial solution, resulting in an overestimationof Ca2+ uptake. However, studies usuallytake special care to avoid such artifacts.

iv) Multiple gmcu levels. Recent work (51) hasindicated the presence of multiple MCUunitary Ca2+ conductance levels. The acti-vation of larger conductance levels couldcreate the nonlinear uptake seen in Fig. 2A.

Although we believe that the nonlinearityof the MCU uptake flux visible in Fig. 2A islikely due to Ca2+ regulation of the MCU,elucidating the cause of this nonlinearity andthe molecular mechanism of Ca2+-regulatedMCU uptake represents a compelling focusfor future experimental investigations.

When experimental conditions push[Ca2+]i beyond the physiological range, theinflux of Ca2+ causes a substantial degrada-tion of the voltage gradient across the IMM(ΔΨm). This lowers the driving force for Ca

2+

influx and is thought to occur because ΔΨm

is sustained in respiring mitochondria pri-marily by proton pumping systems and notby fast conductance K+ channels (53). Ex-perimental measurements under these con-ditions require ΔΨm to be sustained phar-macologically with a K+ ionophore, such asvalinomycin (53), or by voltage-clamping amitoplast (2- to 5-μm vesicles representingthe entire inner membrane of a single mi-tochondrion) (18, 20, 21, 51). A study carriedout with mitoplasts from COS-7 cells showedthat MCU was a highly selective ion channel.The Imcu was observed to have a Michaelis–Menten type saturation with a Km of 19 mM(18). A later study with human cardiacmitoplasts (51) and murine cardiac mito-plasts (20) showed inward Ca2+ currents(termed ImCa1 and Imcu, respectively) con-sistent with the same known features of theMCU as first reported in COS-7 cells (18).

10 100

101

102

103

104

105

100

101

102

103

104

105

106 Cardiac Uptake (see Fig. 2A)

JMCU

, Po=0.9

JMCU

, Po=f(Ca2+)

De Stefani 2011 (Bilayer)

Fieni 2012 (Cardiac Mitoplast)

[Ca2+]i

Ca2+

-1)

A

0 200 400 600 800 1000

0.2

0.4

0.6

0.8

[Ca2+]i

0

MCU Open Probability

B

Fig. 3. Apparent Ca2+-dependent activation of mitochondrial Ca2+ uptake. (A) Measured whole-cell MCU fluxes(gray circles) are replotted from Fig. 2A. The potential influence of MCU PO on cardiac MCU uptake is shown usinga theoretical flux equation with fixed PO = 0.9 (blue line; Eq. 1) or when PO is a function of [Ca2+]i (red line; Eq. 2).The filled blue circle represents the whole-cell MCU flux converted from the study by De Stefani et al. (17), and thefilled green circles represent whole-cell MCU fluxes converted from the study by Fieni et al. (21) (see Eqs. S4–10 inSI Material). (B, Inset) Relationship between PO (black line) and [Ca2+]i over a wide range of [Ca2+]i (0 to 1 mM), whichproduces the red line in Fig. 3A.

10482 | www.pnas.org/cgi/doi/10.1073/pnas.1300410110 Williams et al.

Dow

nloa

ded

by g

uest

on

July

23,

202

0

More recently, MCU activity was measuredin mitochondria from different tissue types,and Imcu in the heart was shown to be signifi-cantly lower than in other tissue types (in-cluding skeletalmuscle, liver, and kidney cells)(21). For comparison, we have converted theheart mitoplast current values to a whole-cellMCU flux and plotted them in Fig. 3A (filledcircles). We have also converted the sin-gle MCU currents from a study by De Ste-fani et al. (17) (Eq. 1 and Fig. 3A, bluecircle). Again, note the agreement betweenthe theoretical Ca2+-activated whole-cellMCU flux values (Fig. 3A, red line) andCa2+ uptake experiments (Fig. 3A, graycircles) conducted at physiological [Ca2+]ilevels, and even mitoplast studies (Fig. 3A,green circles) conducted at high [Ca2+]i levels.

Imcu Magnitude in Cardiac and SkeletalMuscle. Variation in tissue-dependent mi-tochondrial Ca2+ influx is reported in a re-cent paper by Kirichok and colleagues (21).Fieni et al. (21) showed that the Imcu den-sity (measured as picoamperes per pico-farads, pA/pF) in mouse skeletal musclemitoplasts was approximately 28-foldgreater than the Imcu density in cardiacmitoplasts when measured in 100 μM[Ca2+]i at −160 mV. This finding raises an

obvious question: Does the larger skeletalmuscle Imcu density mean that skeletalmuscle mitochondria have a larger effect onthe skeletal muscle [Ca2+]i transient thancardiac mitochondria have on the cardiac[Ca2+]i transient under physiological con-ditions? Because the surface area of the IMMis proportional to its capacitance (1 μF/cm2)and the capacitances of skeletal musclemitoplasts were the same as those of cardiacmitoplasts (0.65 pF and 0.67 pF, re-spectively), one would predict from theresults of the study by Fieni et al. (21) thatskeletal muscle mitochondria would take upmuch more Ca2+ than cardiac mitochondria.However, as shown in Fig. 2C, over thephysiological range of [Ca2+]i (0.1–20 μM),skeletal muscle and cardiac mitochondriatake up similar amounts Ca2+ (at the whole-cell level). Although this finding seems to beat odds with the findings of the study by Fieniet al. (21), the small mitochondrial volumefraction in skeletal muscle (and possibly theCa2+ dependence of Imcu) may provide anexplanation. There is significant recent evi-dence that [Ca2+]i-dependent “activation” ofthe MCU occurs by multiple Ca2+-dependentregulatory mechanisms (the physical locationof some of these proteins is uncertain), in-cluding MICU1 (55, 56), MICU2 (61),

mitochondrial calcium uniporter regulator1 (MCUR1) (62) and Ca2+/calmodulin-dependent protein kinase II (CaMKII) (20).Presumably these (and possibly other MCUregulators) underlie the [Ca2+]i-dependentmodulation of MCU PO shown in Fig. 3B.This regulation is likely to be tissue-specificand time-dependent. The actual Ca2+ fluxwill also depend on the number of functionalMCUs per square micron of IMM; this is alsolikely to be tissue-dependent and a functionof developmental stage and other factors,such as dominant-negative isoforms (63).

Mitoplast Properties. Fieni et al. (21) re-ported a surface area of ∼0.7 pF or 70 μm2

for cardiac mitoplasts prepared using a“French press” methodology. This value is ingood accordance with measurements of theIMM surface area by Page (64) and Smithand Page (65) using EM, as well as with es-timated measurements from “idealized” car-diac mitochondria that are “brick-shaped,”500 nm in width and height, and 1.5 μm long,and have maximally packed cristae (Fig. S6).This suggests that the mitoplasts from thestudy by Fieni et al. (21) are likely the productof a single mitochondrion. In contrast, anotherstudy used mitoplasts with capacitances nearly13-fold larger (9 pF; ref. 20). The membraneorigins of such large “mitoplasts” remain to bedetermined and should be considered wheninterpreting these MCU currents.

Tissue-Dependent MCU Ca2+ Influx UnderPhysiological Conditions. There are manyfactors to consider regarding the influence ofmitochondrial Ca2+ uptake on cytosolic Ca2+

levels under physiological conditions. Al-though small amounts of Ca2+ entering intothe mitochondria may limit the capability ofwhole-cell MCU flux to modulate normal[Ca2+]i transients, these small amounts ofCa2+ play a critical role in regulating mito-chondrial function. In addition to Imcu mag-nitude, one must consider the mitochondrialvolume fraction of the cell when consideringthe likelihood of MCU influencing [Ca2+]i.The relationship between Imcu density andmitochondrial volume fraction (see Table S2)is shown in Fig. 4. One of the highest mito-chondrial volume fractions is in mammalianheart (∼33%), whereas that in skeletal muscleis among the lowest (∼5%) and that in theliver is somewhere in between (∼20%). Tis-sues that display significantly higher MCUactivity levels (i.e., skeletal muscle) also seemto be associated with a lower mitochondrialvolume fraction. In fact, there appears tobe an inverse relationship between Imcu

magnitude and mitochondrial volume frac-tion (Fig. 4, red line). Given the small fraction

Mitochondrial Volume Fraction (%)

MC

U C

urre

nt D

ensi

ty (

pA/p

F)

0 10 20 30 40 500

10

20

30

40

50

60

Cardiac Muscle (Adult)

Skeletal Muscle

Liver

Kidney

Cardiac Muscle (Neonatal)Flight Muscle (Drosophila)

Fig. 4. Mitochondrial Ca2+ uptake plasticity. The whole-mitoplast MCU current density density (Imcu measured aspA/pF) measured by Fieni et al. (21) in different tissues is plotted vs. the fraction of the cell composed of mitochondriain six different tissue types (i.e., skeletal muscle, kidney, liver, neonate heart, adult heart, and flight wing muscle).Additional details of the tissue-specific mitochondrial densities are provided in Table S2. The fit line is a first-orderexponential decay with extra weight applied to skeletal and heart tissues.

Williams et al. PNAS | June 25, 2013 | vol. 110 | no. 26 | 10483

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

July

23,

202

0

of cellular volume comprising mitochondriain skeletal muscle and the small magnitude ofImcu in the heart, it seems that the whole-cellMCU flux will be modest in both tissues andunlikely to influence [Ca2+]i dynamics. It isalso important to consider morphologicaldifferences between tissues. Cardiac mi-tochondria are unique in their abundanceand close proximity to the CRUs (36). Skel-etal muscle mitochondria are a close secondin terms of proximity but are far lessabundant (66). Mitochondria in other tissues(e.g., liver) have more variability but certainlyexperience less [Ca2+]i compared with heartand skeletal muscle, possibly necessitatinga larger Imcu to generate a similar whole-cellflux. Many studies have used a diverse set ofexperimental approaches to measure thisinfluence with varying conclusions. Impor-tantly, under otherwise physiological con-ditions, the acute loss of mitochondrial Ca2+

uptake seems to have a small yet measurableeffect on [Ca2+]i transients in both heartand skeletal muscle (38, 49, 67–70).

Critical Insights on Mitochondrial Ca2+

UptakeThis critical review of the literature focuseson quantitative measurements of mitochon-drial Ca2+ uptake. The data suggest thatunder physiological conditions, the Ca2+ in-flux into mitochondria per liter of cytosol issmall (especially relative to other cytosolicextrusion mechanisms) and unlikely to alter[Ca2+]i. Nevertheless, it also appears thatunder extreme conditions, very large fluxescan occur, as shown in Fig. 3B. This indicatesthat both the PO and single-channel gmcu canassume a wide range of values. Furthermore,mitochondrial Ca2+ uptake displays “self-correcting” behavior, which results in thereduction of large Ca2+ influxes due to thedissipation of ΔΨm (18, 51). Importantly,when cardiac and liver mitochondria are ex-posed to the same conditions, approximatelythe same Ca2+ influx is observed, as shown inFig. 2. We acknowledge that there may bedifferences in mitochondrial size, Nmcu, andMCU activity between tissue types. How-ever, whole-cell MCU flux appears to beconsistent with cardiac uptake (Fig. 2 Band C). Three areas of special interest arediscussed below: (i) How does mitochondrialCa2+ uptake and release affect cellular Ca2+

signaling, (ii) how is gmcu and PO modulated,and (iii) how can computational models beimproved by this knowledge?

MCU Contribution to Cytosolic Ca2+

Signaling. The leading MCU candidate(encoded by NP_001028431) is expressed(at varying levels) in mitochondria across all

tissues (19, 71). The phylogenic distributionof the uniporter’s membrane-spanning poresubunit (MCU) and regulatory partner(MICU1) (55) shows that homologs of bothproteins tend to be coexpressed in all majorbranches of eukaryotic life (71). Evidence sofar suggests that although other Ca2+ influxpathways into the mitochondrial matrix mayexist, MCU is the dominant one. The heartmay well be the ideal place to study MCUuptake, given that cardiac mitochondria areabundant and see regular, repetitive largeelevations in extramitochondrial Ca2+ due tothe cardiac Ca2+ release cycle. This wouldseem to create the ideal set of conditions forlarge whole-cell MCU fluxes. In the heart,however, where fast and abundant non-mitochondrial Ca2+ transporters prevail,MCU fluxes have only a modest effect on[Ca2+]i, consistent with recent experimen-tal results (49, 72). However, when theheart rate increases, each mitochondrion isbathed more frequently in elevated Ca2+

and mitochondria will progressively accu-mulate more Ca2+ even though Ca2+ uptakeper beat is modest. In this sense, [Ca2+]mdynamics could be characterized as a low-pass filter version of [Ca2+]i dynamics. Thus,although cardiac myocytes may be excellentfor examining mitochondrial Ca2+ move-ment, the Ca2+ handling systems in othercells are different and require that quanti-tative experiments be carried out whenevermitochondrial Ca2+ signaling is thought tobe critical. Examples of such systems includedorsal root ganglion neurons (4, 6), liver cells(7), motor nerve terminals (8, 10), gonado-tropes (9), Sertoli cells (13), olfactory sensoryneurons (14), lymphocytes (43), chromaffin(44), and various model cells. Importantly, incells that lack significant Ca2+ extrusionpathways (i.e., SERCA, NCX), mitochondrialCa2+ uptake is more likely to influence cy-tosolic Ca2+ signaling, especially duringrapid and large changes in [Ca2+]i.

Biophysical Properties of MCU: gmcu andPO. The ΔΨm dependence and Ca2+ de-pendence of gmcu has been highlighted as acritical feature (18, 20, 51). Kirichok et al. (18)showed that gmcu saturated at high [Ca2+]i(millimolar range) and was largely open (PO =0.99) when ΔΨm = −200 mV but was gener-ally closed (PO = 0.11) when ΔΨm = −80 mV.Note that these measurements were obtainedat very high [Ca2+] levels (around 105mM). Incontrast, over the physiological range of [Ca2+]i(0.1–20 μM), there appears to be a significantCa2+ dependence of PO (Figs. 2 and 3), whichsuggests a low PO (<0.1). When combined, aconsistent story emerges from the data relatedto the Ca2+ influx into the mitochondria under

physiological conditions: the influx is small. Infact, whole-cell MCU fluxes are largely con-sistent in the heart, skeletal muscle, and livercells based on the findings of 17 independentstudies. Importantly, the findings discussedhere show that whole-cell MCU flux (over thephysiological [Ca2+]i range; Fig. 2) is consistentwith the data from mitoplast studies (18, 20,21) and the latest characterizations of theMCU (17, 19, 51) (Fig. 3) when [Ca2+]i-and ΔΨm-dependent changes in PO (Eq. 2and Fig. 3B) are taken into account.

Quantitative Aspects of MitochondrialCa2+ Fluxes. Computational modeling ofCa2+ signaling at all levels can provide valu-able insights into the interactions withincomplex systems. Such work, for example,has sharpened the discussion on Ca2+ sparks,arrhythmogenic Ca2+ waves, and intra-SRCa2+ movement (41, 73–75). When con-strained by experimental findings, suchmodeling helps formulate new critical ex-periments and can test signaling relationshipsthat otherwise remain beyond current ex-perimental reach. Mitochondrial Ca2+ mod-eling, however, has been less successful atconforming to experimental observations.Our analysis of many models (76–84) sug-gests a wide range in mitochondrial Ca2+

handling kinetics. Some models displayphysiological mitochondrial Ca2+ fluxes (76,77, 81), whereas others (78, 82) display MCUfluxes that are orders of magnitude greaterthan those shown in Fig. 2A and represent acytosolic Ca2+ sequestration mechanism su-perseding even that of SERCA. The un-substantiated magnitude of the whole-cellMCU fluxes in these models calls intoquestion computational results that claimmitochondrial Ca2+ uptake can significantlyinfluence cytosolic Ca2+ signaling (79). It isimperative that futuremodels includemodernMCU formulations [i.e., formulate MCU asa highly selective Ca2+ channel (82)] and beconstrained by the latest experimental obser-vations (17–19, 21, 27, 51).

Future Challenges. There are three clearchallenges for future experiments that aremotivated by the findings of this review:

i) ReexamineCa2+ influx andefflux frommito-chondria in intact cells to provide quantitativeinformation under both physiological andpathophysiological conditions.

ii) Develop mitochondrial computational mod-els of Ca2+movement that take into accountcellular Ca2+ signaling data, metabolic char-acteristics of the mitochondria, subcellular

10484 | www.pnas.org/cgi/doi/10.1073/pnas.1300410110 Williams et al.

Dow

nloa

ded

by g

uest

on

July

23,

202

0

anatomy, and the nanoscopic mitochondrialorganization, including cristae.

iii) Usequantitative experimentalfinding to con-strain the computational models and themodel results to provoke new experiments.

ACKNOWLEDGMENTS. This work was partially sup-ported by USA-Israeli Binational Research Grant 2009-

334 (to W.J.L.). It was also supported by a grant from theLeducq Foundation (European-North American Atrial Fibril-

lation Research Alliance) (to W.J.L.); by National Institutesof Health (NIH) Grants NHLBI R01 HL106059, R01

HL105239-01, and P01 HL67849; by a grant from the

European Union Seventh Framework Program FP7/

2007-2013 under grant agreement No HEALTHF2-

2009-241526, EUTrigTreat, Georg August University,

“Identification and therapeutic targeting of common ar-

rhythmia trigger mechanisms” (to W.J.L.); and by NIH Grant

F32 HL108604 (to G.S.B.W.).

1 Carafoli E, Lehninger AL (1971) A survey of the interaction of

calcium ions with mitochondria from different tissues and species.

Biochem J 122(5):681–690.2 Denton RM, McCormack JG (1980) The role of calcium in the

regulation of mitochondrial metabolism. Biochem Soc Trans 8(3):266–268.3 Rizzuto R, Simpson AW, Brini M, Pozzan T (1992) Rapid changes of

mitochondrial Ca2+ revealed by specifically targeted recombinant

aequorin. Nature 358(6384):325–327.4 Thayer SA, Miller RJ (1990) Regulation of the intracellular free

calcium concentration in single rat dorsal root ganglion neurones in

vitro. J Physiol 425:85–115.5 Werth JL, Thayer SA (1994) Mitochondria buffer physiological

calcium loads in cultured rat dorsal root ganglion neurons. J Neurosci

14(1):348–356.6 Friel DD, Tsien RW (1994) An FCCP-sensitive Ca2+ store in bullfrog

sympathetic neurons and its participation in stimulus-evoked changes

in [Ca2+]i. J Neurosci 14(7):4007–4024.7 Robb-Gaspers LD, et al. (1998) Coupling between cytosolic and

mitochondrial calcium oscillations: Role in the regulation of hepatic

metabolism. Biochim Biophys Acta 1366(1-2):17–32.8 David G, Barrett EF (2000) Stimulation-evoked increases in

cytosolic [Ca(2+)] in mouse motor nerve terminals are limited by

mitochondrial uptake and are temperature-dependent. J Neurosci

20(19):7290–7296.9 Kaftan EJ, Xu T, Abercrombie RF, Hille B (2000) Mitochondria shape

hormonally induced cytoplasmic calcium oscillations and modulate

exocytosis. J Biol Chem 275(33):25465–25470.10 David G, Barrett EF (2003) Mitochondrial Ca2+ uptake prevents

desynchronization of quantal release and minimizes depletion during

repetitive stimulation of mouse motor nerve terminals. J Physiol

548(Pt 2):425–438.11 Isaeva EV, Shirokova N (2003) Metabolic regulation of Ca2+release in permeabilized mammalian skeletal muscle fibres. J Physiol

547(Pt 2):453–462.12 Maack C, et al. (2006) Elevated cytosolic Na+ decreases

mitochondrial Ca2+ uptake during excitation-contraction coupling and

impairs energetic adaptation in cardiac myocytes. Circ Res 99(2):172–182.13 Veitinger S, et al. (2011) Purinergic signalling mobilizes mitochondrial

Ca2+ in mouse Sertoli cells. J Physiol 589(Pt 21):5033–5055.14 Fluegge D, et al. (2012) Mitochondrial Ca(2+) mobilization is

a key element in olfactory signaling. Nat Neurosci 15(5):754–762.15 Drago I, De Stefani D, Rizzuto R, Pozzan T (2012) Mitochondrial

Ca2+ uptake contributes to buffering cytoplasmic Ca2+ peaks in

cardiomyocytes. Proc Natl Acad Sci USA 109(32):12986–12991.16 O’Rourke B, Blatter LA (2009) Mitochondrial Ca2+ uptake:

Tortoise or hare? J Mol Cell Cardiol 46(6):767–774.17 De Stefani D, Raffaello A, Teardo E, Szabò I, Rizzuto R (2011) A

forty-kilodalton protein of the inner membrane is the mitochondrial

calcium uniporter. Nature 476(7360):336–340.18 Kirichok Y, Krapivinsky G, Clapham DE (2004) The mitochondrial

calcium uniporter is a highly selective ion channel. Nature 427(6972):

360–364.19 Baughman JM, et al. (2011) Integrative genomics identifies MCU

as an essential component of the mitochondrial calcium uniporter.

Nature 476(7360):341–345.20 Joiner M-LA, et al. (2012) CaMKII determines mitochondrial

stress responses in heart. Nature 491(7423):269–273.21 Fieni F, Lee SB, Jan YN, Kirichok Y (2012) Activity of the

mitochondrial calcium uniporter varies greatly between tissues. Nat

Commun 3:1317.22 Scarpa A, Graziotti P (1973) Mechanisms for intracellular calcium

regulation in heart. I. Stopped-flow measurements of Ca++ uptake

by cardiac mitochondria. J Gen Physiol 62(6):756–772.23 Fry CH, Powell T, Twist VW, Ward JP (1984) Net calcium

exchange in adult rat ventricular myocytes: An assessment of

mitochondrial calcium accumulating capacity. Proc R Soc Lond B Biol

Sci 223(1231):223–238.24 Crompton M (1985) The Regulation of Mitochondrial Calcium

Transport in Heart. Curr Top Membr Transp 25:231–276.25 Bassani JW, Bassani RA, Bers DM (1994) Relaxation in rabbit and

rat cardiac cells: Species-dependent differences in cellular

mechanisms. J Physiol 476(2):279–293.

26 Zhou Z, Matlib MA, Bers DM (1998) Cytosolic and mitochondrialCa2+ signals in patch clamped mammalian ventricular myocytes.J Physiol 507(Pt 2):379–403.27 Wei A-CA, Liu TT, Cortassa SS, Winslow RLR, O’Rourke BB (2011)Mitochondrial Ca2+ influx and efflux rates in guinea pig cardiacmitochondria: Low and high affinity effects of cyclosporine A.Biochim Biophys Acta 1813(7):1373–1381.28 Sembrowich WL, Quintinskie JJ, Li G (1985) Calcium uptake inmitochondria from different skeletal muscle types. J Appl Physiol59(1):137–141.29 Selles J, Boland RL (1990) In vitro calcium transport propertiesof skeletal muscle mitochondria from vitamin D-deficient and 1,25-dihydroxy-vitamin D3-treated chicks. Calcif Tissue Int 47(1):46–50.30 Allshire AP, Heffron JJ (1984) Uptake, retention, and efflux of Ca2+by mitochondrial preparations from skeletal muscle. Arch Biochem

Biophys 228(1):353–363.31 Vinogradov A, Scarpa A (1973) The initial velocities of calciumuptake by rat liver mitochondria. J Biol Chem 248(15):5527–5531.32 Reed KC, Bygrave FL (1975) A kinetic study of mitochondrialcalcium transport. Eur J Biochem 55(3):497–504.33 Gunter TE, Pfeiffer DR (1990) Mechanisms by which mitochondriatransport calcium. Am J Physiol 258(5 Pt 1):C755–C786.34 Gunter TE, Buntinas L, Sparagna GC, Gunter KK (1998) The Ca2+transport mechanisms of mitochondria and Ca2+ uptake fromphysiological-type Ca2+ transients. Biochim Biophys Acta 1366(1-2):5–15.35 Zoccarato F, Nicholls D (1982) The role of phosphate in theregulation of the independent calcium-efflux pathway of livermitochondria. Eur J Biochem 127(2):333–338.36 Bers DM (2001) Excitation-Contraction Coupling and Cardiac

Contractile Force (Springer, Dordrecht, The Netherlands).37 Lukyanenko V, Chikando A, Lederer WJ (2009) Mitochondria incardiomyocyte Ca2+ signaling. Int J Biochem Cell Biol 41(10):1957–1971.38 Bers DM (2002) Cardiac excitation-contraction coupling. Nature415(6868):198–205.39 Smith GD, Keizer JE, Stern MD, Lederer WJ, Cheng H (1998)A simple numerical model of calcium spark formation and detectionin cardiac myocytes. Biophys J 75(1):15–32.40 Sobie EA, Dilly KW, dos Santos Cruz J, Lederer WJ, Jafri MS (2002)Termination of cardiac Ca(2+) sparks: An investigative mathematicalmodel of calcium-induced calcium release. Biophys J 83(1):59–78.41 Williams GSB, et al. (2011) Dynamics of calcium sparks andcalcium leak in the heart. Biophys J 101(6):1287–1296.42 Cheng H, Lederer WJ (2008) Calcium sparks. Physiol Rev 88(4):1491–1545.43 Hoth M, Button DC, Lewis RS (2000) Mitochondrial control ofcalcium-channel gating: A mechanism for sustained signaling andtranscriptional activation in T lymphocytes. Proc Natl Acad Sci USA

97(19):10607–10612.44 Xu T, Naraghi M, Kang H, Neher E (1997) Kinetic studies of Ca2+binding and Ca2+ clearance in the cytosol of adrenal chromaffincells. Biophys J 73(1):532–545.45 Negretti N, Varro A, Eisner DA (1995) Estimate of net calciumfluxes and sarcoplasmic reticulum calcium content during systole inrat ventricular myocytes. J Physiol 486(Pt 3):581–591.46 Tran K, Smith NP, Loiselle DS, Crampin EJ (2009) Athermodynamic model of the cardiac sarcoplasmic/endoplasmicCa(2+) (SERCA) pump. Biophys J 96(5):2029–2042.47 Weber CR, Ginsburg KS, Philipson KD, Shannon TR, Bers DM(2001) Allosteric regulation of Na/Ca exchange current by cytosolicCa in intact cardiac myocytes. J Gen Physiol 117(2):119–131.48 Bassani JW, Qi M, Samarel AM, Bers DM (1994) Contractile arrestincreases sarcoplasmic reticulum calcium uptake and SERCA2 geneexpression in cultured neonatal rat heart cells. Circ Res 74(5):991–997.49 Andrienko TN, Picht E, Bers DM (2009) Mitochondrial freecalcium regulation during sarcoplasmic reticulum calcium release inrat cardiac myocytes. J Mol Cell Cardiol 46(6):1027–1036.50 Barth E, Stämmler G, Speiser B, Schaper J (1992) Ultrastructuralquantitation of mitochondria and myofilaments in cardiac muscle from 10different animal species including man. J Mol Cell Cardiol 24(7):669–681.51 Michels G, et al. (2009) Regulation of the human cardiacmitochondrial Ca2+ uptake by 2 different voltage-gated Ca2+channels. Circulation 119(18):2435–2443.52 Hille B (1992) Ionic Channels of Excitable Membranes (Sinauer,Sunderland, MA).

53 Bragadin M, Pozzan T, Azzone GF (1979) Kinetics of Ca2+ carrier

in rat liver mitochondria. Biochemistry 18(26):5972–5978.54 Kröner H (1986) “Allosteric regulation” of calcium-uptake in rat

liver mitochondria. Biol Chem Hoppe Seyler 367(6):483–493.55 Perocchi F, et al. (2010) MICU1 encodes a mitochondrial EF hand

protein required for Ca(2+) uptake. Nature 467(7313):291–296.56 Mallilankaraman K, et al. (2012) MICU1 is an essential

gatekeeper for MCU-mediated mitochondrial Ca(2+) uptake that

regulates cell survival. Cell 151(3):630–644.57 Palty R, et al. (2010) NCLX is an essential component of

mitochondrial Na+/Ca2+ exchange. Proc Natl Acad Sci USA

107(1):436–441.58 Griffiths EJ (1999) Reversal of mitochondrial Na/Ca exchange

during metabolic inhibition in rat cardiomyocytes. FEBS Lett 453(3):

400–404.59 Boyman L, Williams GSB, Khananshvili D, Sekler I, Lederer WJ

(2013) NCLX: The mitochondrial sodium calcium exchanger. J Mol

Cell Cardiol 59:205–213.60 Jiang D, Zhao L, Clapham DE (2009) Genome-wide RNAi screen

identifies Letm1 as a mitochondrial Ca2+/H+ antiporter. Science

326(5949):144–147.61 Plovanich M, et al. (2013) MICU2, a paralog of MICU1, resides

within the mitochondrial uniporter complex to regulate calcium

handling. PLoS ONE 8(2):e55785.62 Mallilankaraman K, et al. (2012) MCUR1 is an essential

component of mitochondrial Ca2+ uptake that regulates cellular

metabolism. Nat Cell Biol 14(12):1336–1343.63 Patron M, et al. (2013) The mitochondrial calcium uniporter

(MCU): molecular identity and physiological roles. Journal of

Biological Chemistry 288:10750–10758.64 Page E (1978) Quantitative ultrastructural analysis in cardiac

membrane physiology. Am J Physiol 235(5):C147–C158.65 Smith HE, Page E (1976) Morphometry of rat heart mitochondrial

subcompartments and membranes: Application to myocardial cell

atrophy after hypophysectomy. J Ultrastruct Res 55(1):31–41.66 Rossi AE, Boncompagni S, Wei L, Protasi F, Dirksen RT (2011)

Differential impact of mitochondrial positioning on mitochondrial

Ca(2+) uptake and Ca(2+) spark suppression in skeletal muscle. Am

J Physiol Cell Physiol 301(5):C1128–C1139.67 Baylor SM, Hollingworth S (2007) Simulation of Ca2+movements within the sarcomere of fast-twitch mouse fibers

stimulated by action potentials. J Gen Physiol 130(3):283–302.68 Bruton JD, Dahlstedt AJ, Abbate F, Westerblad H (2003)

Mitochondrial function in intact skeletal muscle fibres of creatine

kinase deficient mice. J Physiol 552(Pt 2):393–402.69 Dedkova EN, Blatter LA (2008) Mitochondrial Ca2+ and the

heart. Cell Calcium 44(1):77–91.70 Lännergren J, Westerblad H, Bruton JD (2001) Changes in

mitochondrial Ca2+ detected with Rhod-2 in single frog and mouse

skeletal muscle fibres during and after repeated tetanic contractions.

J Muscle Res Cell Motil 22(3):265–275.71 Bick AG, Calvo SE, Mootha VK (2012) Evolutionary diversity of

the mitochondrial calcium uniporter. Science 336(6083):886.72 Lu X, et al. (2013) Measuring local gradients of

intramitochondrial [Ca(2+)] in cardiac myocytes during sarcoplasmic

reticulum Ca(2+) release. Circ Res 112(3):424–431.73 Sato D, Despa S, Bers DM (2012) Can the sodium-calcium

exchanger initiate or suppress calcium sparks in cardiac myocytes?

Biophys J 102(8):L31–L33.74 Picht E, et al. (2011) Dynamic calcium movement inside cardiac

sarcoplasmic reticulum during release. Circ Res 108(7):847–856.75 Nivala M, Qu Z (2012) Calcium alternans in a couplon network

model of ventricular myocytes: Role of sarcoplasmic reticulum load.

Am J Physiol Heart Circ Physiol 303(3):H341–H352.76 Magnus G, Keizer J (1997) Minimal model of beta-cell

mitochondrial Ca2+ handling. Am J Physiol 273(2 Pt 1):C717–C733.77 Cortassa S, Aon MA, Marbán E, Winslow RL, O’Rourke B (2003)

An integrated model of cardiac mitochondrial energy metabolism

and calcium dynamics. Biophys J 84(4):2734–2755.78 Cortassa S, et al. (2006) A computational model integrating

electrophysiology, contraction, and mitochondrial bioenergetics in

the ventricular myocyte. Biophys J 91(4):1564–1589.

Williams et al. PNAS | June 25, 2013 | vol. 110 | no. 26 | 10485

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

July

23,

202

0

79 Gauthier LD, Greenstein JL, Winslow RL (2012) Toward anintegrative computational model of the Guinea pig cardiac myocyte.Front Physiol 3:244.80 Dash RK, Qi F, Beard DA (2009) A biophysically basedmathematical model for the kinetics of mitochondrial calciumuniporter. Biophys J 96(4):1318–1332.

81 Pradhan RK, Beard DA, Dash RK (2010) A biophysically basedmathematical model for the kinetics of mitochondrial Na+-Ca2+antiporter. Biophys J 98(2):218–230.82 Nguyen M-HT, Dudycha SJ, Jafri MS (2007) Effect of Ca2+ oncardiac mitochondrial energy production is modulated by Na+ andH+ dynamics. Am J Physiol Cell Physiol 292(6):C2004–C2020.

83 Nguyen M-HT, Jafri MS (2005) Mitochondrial calciumsignaling and energy metabolism. Ann N Y Acad Sci 1047:127–137.84 Saleet Jafri M, Kotulska M (2006) Modeling the mechanism ofmetabolic oscillations in ischemic cardiac myocytes. J Theor Biol242(4):801–817.

10486 | www.pnas.org/cgi/doi/10.1073/pnas.1300410110 Williams et al.

Dow

nloa

ded

by g

uest

on

July

23,

202

0