JOURNAL OF EXPERIMENTAL ZOOLOGY 289:81–89 (2001)

© 2001 WILEY-LISS, INC.

Dynamin, Encoded by shibire, Is Central toCardiac Function

ERIK JOHNSON,1 JOHN RINGO,1 AND HAROLD DOWSE1,2*1Department of Biological Sciences, University of Maine, Orono,Maine 04469

2Department of Mathematics and Statistics, University of Maine, Orono,Maine 04469

ABSTRACT Employing the Drosophila heart, a model system for genetic and molecular inves-tigation of cardiac physiology, we demonstrate here an essential role for the protein dynamin,encoded by the Drosophila gene shibirets (shits), in maintaining normal heart function. In fliesbearing two temperature-sensitive alleles of shi, shits1 and shits2, heartbeat is both slower and lessrhythmic than in wild-type animals. Serotonin and norepinephrine, normally cardioacceleratoryin wild type, are without effect in flies bearing the shi mutation. Electrocardiogram (EKG) analy-sis reveals a bigeminal beat in mutant hearts, unlike the single electrical pulse in wild-type. Thegene no action potential temperature sensitive, with previously-described cardiac aberrations similar tothose of shi, interacts with shi: shi/shi;nap/nap mutants have almost wild-type heart function. J.Exp. Zool. 289:81–89, 2001. © 2001 Wiley-Liss, Inc.

Grant sponsor: American Heart Association.*Correspondence to: Harold B. Dowse, Department of Biological

Sciences, 5751 Murray Hall, University of Maine, Orono, ME 04469-5751. E-mail: dowse@maine.edu

Received 2 February 2000; Accepted 18 August 2000

Despite the wide range of genetic and molecu-lar tools available in Drosophila for the investi-gation of physiology, very little has been done withthis organism’s heart until recently. The Droso-phila cardiac system offers a useful model for thestudy of pacemaker function at the level of con-stituent ion channels, control of cardiac activityby neurotransmitters, and the genetics of heartdisease. We have been developing the Drosophilacardiac system as a model for the study of verte-brate heart pacemakers and their control systems.

The dorsal vessel in D. melanogaster is a tubu-lar organ located medially and dorsally in thehemocoel and transports hemolymph cranially(Rizki, ’78; Curtis et al., ’99). The caudal regionof the dorsal vessel contains openings called os-tia, and forms the heart proper, the most poste-rior region containing at least one putativepacemaker (Rizki, ’78). The anterior portion is de-void of ostia, and is called the aorta (Rizki, ’78;Miller, ’85). Heartbeat is myogenic (Rizki, ’78;Dowse et al., ’95; Gu and Singh, ’95; Johnson etal., ’98), originating in a pacemaker in the mostcaudal region of the heart, although the beat di-rection may reverse during late pupal stages, in-dicating the presence of a satellite pacemaker inthe anterior regions (Rizki, ’78).

Employing genetics and pharmacology, we havedescribed an ensemble of ion channels central tothe Drosophila pacemaker (Dowse et al., ’95;

Johnson et al., ’98). Ca2+ and K+, but not Na+ chan-nels contribute to pacemaking (Dowse et al., ’95;Johnson et al., ’98). The heart is sensitive to neu-rotransmitters, being accelerated by octopamine,serotonin, acetylcholine, and norepinephrine (John-son et al., ’97). Neuropeptides native to Drosophila,and several isolated from Limulus, are also effec-tive in altering rate (Johnson et al., 2000).

In our effort to uncover genes involved in car-diac function, we have been screening mutationsknown to affect the nervous system. The muta-tion shibirets (shits) reversibly induces paralysis attemperatures exceeding 29° (Grigliatti et al., ’73),a result of cessation of endocytic recovery of vesiclemembrane at synapses (Masur et al., ’90); theyinterdict endocytosis in other regions as well, no-tably the garland cells (Kosaka and Ikeda, ’83).The shits1 and shits2 alleles are missense alter-ations in the GTPase domain of dynamin that in-terfere with its function in a temperature-sensitivemanner (Scaife and Margolis, ’97). shi encodes theDrosophila homologue of the protein dynamin(Chen et al., ’91; van der Bliek and Myerowitz,’91), essential for scission of clathrin-coated pitsduring vesicle formation (Gagnon et al., ’98). En-

82 E. JOHNSON ET AL.

docytosis regulates the effects of certain neu-rotransmitters by sequestering membrane-boundreceptors (Gagnon et al., ’98). We determine herethat mutations at the shi locus interfere withheart function, and test the hypothesis that thephenotype is a result of interference with the se-questration of receptors.

We had demonstrated that the regulatory geneno action potential temperature sensitive (napts) (Wu et al.,’78), causes bradycardia and arrhythmia at alltemperatures (Dowse et al., ’95), similar to thatcaused by shits (this report). We had earlier hy-pothesized that nap’s effect on heart function wascaused by downregulation of the Na+-channel en-coding gene paralyticts (parats) (Kernan et al., ’91),but animals homozygous and hemizygous forparats have normal hearts (Dowse et al., ’95).Given the similarity in in the cardiac phenotypebetween nap and shi, we hypothesize that napdownregulates shi to produce heart defects andinvestigate this possibility by testing a series ofdouble mutants bearing varying numbers of cop-ies of shits1, and napts.

MATERIALS AND METHODSFly culture

Drosophila melanogaster were cultured at 25°in LD 12:12 on malt/yeast/molasses/agar mediumin glass vials or bottles. Propionic acid was addedto combat mold.

Optical recording of heartbeatAnimals were tested as pupariation began (P1,

Ashburner, ’89). The early pupal heart is similarto that of the adult, but there are a number ofdifferences in structure and histology (Curtis etal., ’99). The pupa is nearly transparent, render-ing the heart visible. Pupae were placed in anOlympus BH2 microscope fitted with an opticaldata acquisition system. Temperature at the stagewas controlled with a Sensortek TS-4. The signalproduced by the beating heart was amplified, digi-tized with a Metrabyte DAS-8 AD converter at100 Hz, and stored on a microcomputer (see Dowseet al., ’95, for details). At each temperature, theanimal was allowed to equilibrate for 90 sec, af-ter which 30 sec of data were taken and the tem-perature immediately raised to the next step. Thisunvarying protocol eliminated subjectivity in se-lection of data segments to be analyzed.

Analysis of heartbeat dataHeart rate was estimated by Maximum Entropy

Spectral Analysis (MESA) (Ulrych and Bishop, ’75;

Dowse and Ringo, ’89) (Fig. 1). Regularity of theheartbeat was quantified by autocorrelation(Chatfield, ’80). Rhythmicity in the signal resultsin recurring peaks of positive and negative corre-lation (Fig. 1A), and the decay envelope reflectsregularity in the signal (Chatfield, ’80). The heightof the second peak was expressed as a fraction ofthe height of the zeroth lag, (i.e., a coefficient ofcorrelation; see Dowse et al., ’95, for details). Werefer to this coefficient as “rhythmicity” (r). Wecommonly observe that as r decreases, there aremore skipped beats, irregular intervals betweenbeats, periods of heartbeat cessation, and ratevariations with time (Johnson et al., 2000).

Injection of neurotransmittersWe tested norepinephrine and serotonin at con-

centrations of 1, 10, 100, and 1000 µM in Canton-S (wild-type), shits1, and napts. Five animals weretested at each concentration, and animals wereused only once. Glass electrodes, drawn out on aNarashige PB-7 unit and polished to a 3-µm ap-erture on a Narashige microforge, were used. Amicromanipulator positioned the electrode in thedorsal posterior region near the caudal region ofthe heart. A WPI Nanoliter Injector was used toinject 50 nL of solution. Injected substances thor-oughly perfused the myocardium rapidly. Volumewas estimated by measuring the diameter of a testdroplet extruded in air on a video monitor cali-brated with a stage micrometer. The size did notvary measurably over many test extrusions oncethe instrument was set. We estimate a rapid (lessthan 1 min) dilution of the injected substance ofapproximately 200:1 by the fly’s hemolymph basedon our own measurements and published valuesof insect body composition (Jones, ’77). Larvaewere allowed to acclimate to 25° for 90 sec. Werecorded data for 30 sec prior to injection, andfor a further 30 sec after a delay of 120 sec. Inthe development of this protocol, we performeda series of tests, finding that if the heart reactsat all to a treatment, it does so almost immedi-ately, and the response continues unchanged forat least 10 min, eliminating concern that the in-jection in the region of the heart causes tran-sient effects. After injection, electrode tips wereexamined under the compound microscope forbreakage that might have altered injection vol-ume. All neurotransmitters were dissolved inCa2+-free Drosophila Ringer’s solution at pH 7.Ca2+ in the Ringer’s solution causes increased ratewhen the solution is injected alone (E. Johnson,unpublished observations).

DYNAMIN IS CENTRAL TO CARDIAC FUNCTION 83

Electrocardiogram analysisWe used the technique of Rizki (’78) to record

from intact pupae. Sharpened tungsten electrodeswere inserted laterally into the abdomen near thecaudal end of the heart, likely site of the pace-maker, and near the anterior end of the dorsalvessel (system ground). Voltage was pre-amplifiedwith a World Precision Instruments DAM-50 am-plifier and boosted with a Grass Low Level DCamplifier. The signal was digitized by a DASH8A/D converter and recorded onto a microcomputerdigitizing at a rate of 100 Hz. Five specimens ofthe following genotypes were tested: Canton-S

(wild-type), shits1, and napts. After recording, ani-mals were placed singly into food vials and wereallowed to recover. All subjects survived to adult-hood, indicating no serious injury was done by theelectrodes.

RESULTSThe hearts of flies bearing the shibire

mutation beat more slowly and lessrhythmically than those of wild-type

animalsTwo temperature-sensitive alleles of the gene

shi, shits1, and shits2, produce substantially aber-

Fig. 1. The mutation shits1 renders heartbeat weaker andless rhythmic. (A) Top, Spectral analysis of 30 sec of datafrom a wild-type heart at 25°C. The spectrum shows one pe-riodicity, at 2.6 Hz. Middle, autcorrelation analysis of the samedata. The decay envelope of the autocorrelation function in-dicates long-range coherence, the result of a regular beat.Rhythmicity is 0.73. Bottom, the original 30-sec optical sig-nal. Time scale bar = 5 sec. (B) Raw data only for the samefly at 35°C. The rate has increased to 3.1 Hz with a rhyth-

micity of 0.77. (C) A shits1 fly tested at 25° and (D) 35°. Thisheart is slower than wild type at both temperatures, 25°: 1.9Hz; 35°: 2.2 Hz. Rhythmicity is weaker, 25°: 0.34; 35°: 0.06.(E), Summary graph of heart rate for all animals at all tem-peratures; nap data have been included for comparison (Can-ton-S, solid line; shi, dotted line; nap, dashed line). Note thatwild-type heart rate responds nearly linearly to increases oftemperature. Both shi and nap heart rates rise slowly andthen saturate, unable to accelerate further.

84 E. JOHNSON ET AL.

rant heartbeat. Hearts of flies homozygous orhemizygous for the alleles shits1and shits2 are in-creasingly slower and more arrhythmic than wildtype with increasing temperature (Fig. 1). Unlikethe paralytic phenotype, which shows rapid onsetwhen the restrictive temperature is reached(Grigliatti et al., ’73), the heart defects show nosharp transition at 29°; rather, severity increasesuniformly as temperature rises (Fig. 1, Table 1).The cardiac phenotype was mapped to the locususing deletions and duplications of shi. shi/(Df1)sd72a behaves like shi hemi- and homozy-gotes and a duplication of shi, shits1/Dp(1:Y)y+shi+,rescued the wild-type phenotype (Table 1). As withpassing out, the heart phenotype is recessive.

The electrocardiograms of flies bearing theshi and nap mutations are irregular

In the wild-type heart, a single peak of extra-cellular voltage rises and falls smoothly, with ashort recovery period (Fig. 2) This is a result ofthe simple peristaltic beat that moves craniallyafter initiation by the pacemaker (Rizki, ’78;Curtis et al., ’99). In contrast, the shits1 heart gen-erates a signature double-peak, suggesting a se-rious compromise in electrical communicationthroughout the myocardium (Fig. 2). The EKG ofthe napts1 heart is irregular as well, but differsqualitatively from that of shits. There are usuallysmall extra peaks scattered throughout the record.These are never as large or as regular as the sec-ondary peaks in shi, which constitute genuine big-eminal beats. The voltages in the EKGs areuniversally much lower in nap hearts, but we havenot quantified this rigorously owing to the con-siderable variation introduced by the recordingapparatus. The napts1–induced heart defect has

been previously mapped to the locus with dele-tions (Dowse et al., ’95).

The normally cardioacceleratoryneurotransmitters norepinephrine and

serotonin fail to stimulate the hearts of fliesbearing shi and nap mutations

To test further our hypothesis that the regula-tion of receptors is defective in shits mutants, weinjected two cardioacceleratory neurotransmitters,norepinephrine and serotonin, in varying concen-trations into wild-type and shits1 pupae. As previ-ously described, the response of the wild-typepupal heart to both these neurotransmitters is arate increase without loss of rhythmic strength(Johnson et al., ’97). The heart of the shits1 fly doesnot accelerate and becomes even more arrhyth-mic in response (Fig. 3). Like those of shits, heartsof napts individuals do not respond to either neu-rohormone (Fig. 3).

shi and nap double mutants have near-normal heart function

Flies homozygous for shits1 alone or naptsl alonehave bradycardia and arrhythmia. Animals hem-izygous for shits1 and homozygous for naptsl, shits1/Y;naptsl/naptsl, have near-normal heartbeat at allbut one temperature, 37°. The double mutants,shits1/shits1; naptsl/naptsl have significantly slowerhearts at all temperatures, but the beat is nor-mally rhythmic. shits1/Y;naptsl/+ animals havenear normal rate except at one temperature.Heartbeat in shi/shi;nap/+ animals was signifi-cantly slower and less rhythmic than in wild-typeat a number of temperatures. Data are summa-rized in Table 2.

TABLE 1. Defects in heart rate and rhythmicity in shi and nap mutantsa

20°C 25°C 30°C 35°C 37°C

Canton-S 2.28 ± 0.05† 2.41 ± 0.04 2.72 ± 0.04 3.08 ± 0.05 3.33 ± 0.05(n=60) 0.62 ± 0.04† 0.54 ± 0.04 0.45 ± 0.04 0.45 ± 0.05 0.40 ± 0.04shits1/ + 2.17 ± 0.08 2.47 ± 0.08 2.84 ± 0.09 3.22 ± 0.13 3.44 ± 0.15(N=5) 0.50 ± 0.08 0.45 ± 0.11 0.45 ± 0.11 0.41 ± 0.07 0.39 ± 0.08shits1/Dp(1:Y)y+shi+ 2.13 ± 0.12 2.58 ± 0.11 2.88 ± 0.14 2.47 ± 0.64 2.96 ± 0.38(N=5) 0.30 ± 0.11 0.41 ± 0.13 0.43 ± 0.12 0.33 ± 0.13 0.43 ± 0.14shits1/Df(1)sd72a 1.82 ± 0.04b 2.20 ± 0.07 2.51 ± 0.10 1.52 ± 0.52b 1.61 ± 0.60b

(N=5) 0.37 ± 0.07b 0.33 ± 0.17 0.14 ± 0.09b 0.11 ± 0.03b 0.05 ± 0.05b

shits1 1.76 ± 0.11b 1.79 ± 0.21b 1.84 ± 0.29b 2.07 ± 0.25b 1.59 ± 0.32b

(N=10) 0.45 ± 0.07 0.43 ± 0.09 0.42 ± 0.08b 0.28 ± 0.07b 0.16 ± 0.04b

shits2 1.64 ± 0.09b 2.03 ± 0.09 2.00 ± 0.24b 1.08 ± 0.44b 0.93 ± 0.48b

(N=10) 0.15 ± 0.03b 0.20 ± 0.05b 0.11 ± 0.03b 0.07 ± 0.01b 0.10 ± 0.02b

aTop lines are heart rates, bottom rhythmicity scores (both mean ± S.E.M.). Data are ordered by heart rate at 20°C. bIndicates the mean issignificantly lower than that of wild type (ANOVA, Tukey’s test, α = 0.05).

DYNAMIN IS CENTRAL TO CARDIAC FUNCTION 85

DISCUSSIONThe protein dynamin, encoded by the wild-type

shi gene, is essential for normal cardiac functionin Drosophila. Mutations at the locus cause brady-cardia and arrhythmia. The cardiac phenotype, likethe pass-out phenotype, is recessive. The pass-outphenotype is a result of cessation of synaptic trans-mission owing to failure of endocytosis (Masur etal., ’90) and it might be that the cardiac phenotypeoriginates in or is mediated by the nervous sys-tem. However, we have previously demonstratedthat neither Na+ channels nor a functional nervoussystem are essential for cardiac function in Droso-phila (Dowse et al., ’95). Animals homozygous andhemizygous for paralyticts (Suzuki et al. ’71) have

normal heartbeat at all temperatures, includingthose above the point at which Na+-dependent ac-tion potentials cease (Wu and Ganetzky, ’80;Kauvar, ’82; Jackson et al., ’84), and the Na+-chan-nel blocker, tetrodotoxin (Barchi, ’88), has no ef-fect on heartbeat even at very high doses (Johnsonet al., ’98). Thus this possibility is ruled out.

Endocytosis regulates neurotransmitter sensi-tivity by sequestering membrane-bound proteinsof the G-Protein-Coupled Receptor (GPCR) fam-ily, prototypically the β2 adrenergic receptor(Gagnon et al., ’98). β2 adrenergic receptors aredownregulated, e.g., when agonists are presentchronically, by being encapsulated in clathrin-coated pits. During endocytosis, dynamin is re-

Fig. 2. Electrocardiogram (EKG) analysis of heartbeat re-veals defects in the generation and conduction of heartbeatin shits1, and napts mutants. (A) The Canton-S heart producesa single voltage excursion per beat as peristalsis moves ante-riorly. (B) In contrast, the shits1 heartbeat is bigeminal. Be-fore the main beat is completed, a second weak peak ofvarying height and timing occurs. (C) napts hearts behavemuch like shits1, but the main peak is weaker overall and thesecond peak is less pronounced. Voltage is in arbitrary units,time in sec.

86 E. JOHNSON ET AL.

quired for clathrin-coated pits to pinch off fromthe plasma membrane (Gagnon et al., ’98). Thereceptors are either transported to lysosomes tobe broken down or reactivated in endosomes andreturned to the cell surface (Zhang et al., ’97). M1,M3, and M4 muscarinic cholinergic receptors aresimilarly sequestered (Lee et al., ’98).

We have previously demonstrated that ambientlevels of neurotransmitters are necessary for nor-mal heartbeat in Drosophila (Johnson et al., ’97).

Heart rate is reduced in isolated-perfused prepa-rations compared to intact preparations in whichthe organ is bathed in hemolymph, and the mu-tation DOPA Decarboxylasets (Ddcts) (Wright, ’77),which encodes an enzyme necessary for the syn-thesis of dopamine, serotonin, and (by inference)norepinephrine (Livingstone and Tempel, ’83;Johnson et al., ’97). The number of receptors is adynamic steady state resulting from feedbackkeyed to the level of neurotransmitters present.

Fig. 3. The mutations shits1 and napts interdict the nor-mal cardioacceleratory effects of norepinephrine. (A) Before,30 s of heartbeat recorded optically from a wild-type fly. Theheart beat regularly at a rate of 2.7 Hz with a rhythmicityof 0.73. After, the rate of the heart accelerated to 4.3 Hz,and the regularity dropped slightly to 0.45. (B) Before, theheart of an uninjected shits1 fly beat at 2.2 Hz, with a rhyth-micity of 0.16. After, the application of norepinephrine caused

the rate to drop to 1.3 Hz, while rhythmicity remained es-sentially unchanged at 0.19. (C) Before, the heart of a napfly beat with a rate of 1.8 Hz and a rhythmicity of 0.17.After, the injection caused the rate and rhythmicity to dropto 0.5 Hz and 0.07, respectively. 30 s of data were takenand injections were done after a delay of 120 sec. The injec-tion concentration of norepinephrine was100 mM for all pu-pae shown here.

DYNAMIN IS CENTRAL TO CARDIAC FUNCTION 87

This steady state is implemented by downregu-lation, which depends on endocytosis (Gagnon etal., ’98). The shi heart, lacking the potential fornormal downregulation in response to ambientlevels of neurotransmitters, would be chronicallyoverstimulated in this scenario.

Initial evidence in support of this hypothesiscomes from our finding that normal responses tothe neurohormones norepinephrine and serotoninare interdicted by the shits mutation. If the systemis being chronically overstimulated, additional neu-rotransmitter offered by injection would be unableto stimulate an already overtaxed system further.The observed bradycardia would be the overt re-sult of a system pushed beyond its oscillatory lim-its, responding with chaos. We cannot, however,rule out the possibility that a very slow, irregu-larly beating heart simply cannot accelerate.

EKG analysis of heartbeat in shits1 and wild-type pupae supports the first of the above hypoth-eses. Under normal conditions in the heart,whether in Drosophila or a mammal, the impulsedies out at the end of its passage through the myo-cardium, surrounded by refractile tissue (Janse,’92). If a second impulse enters the myocardiumbefore the first is completely extinguished, a con-dition called reentrant excitation occurs, whichcan result in arrythmias (Janse, ’92). If the shipacemaker is cycling at a frequency more rapidthan the myocardium can accommodate, it willpass an impulse into the caudal portion of themyocardium, which has already passed throughits refractory period, before the more anteriorwave of the beat has terminated. This is, strictlyspeaking, a reentrant arrhythmia (Janse, ’92).Drosophila now potentially offers a model system

to study the phenomenon. The characteristic big-eminal beat of shi hearts is consistent with reen-trant excitation, the additional irregular peakindicating that an abortive second beat hasstarted. Faulty coordination between the pace-maker and the myocardium would result in anovertly slower, less rhythmic beat. Ultimate reso-lution of this question awaits our precise locationof the pacemaker and investigation of its electro-physiology.

The action of napts to affect the heart remainsunknown. EKGs of nap hearts are abnormal.Small voltage spikes occur frequently at irregu-lar intervals, apparently failing to lead to a gen-eral depolarization. We tested the sensitivity ofnap mutants to norepinephrine and serotonin, andlike shits, hearts of napts individuals do not respondnormally. nap could be interacting with a geneencoding an ion channel central to the pacemaker.napts is an allele of the male lethal gene maleless(mle) (Kernan et al., ’91). The mle gene product isnecessary for dosage compensation of the X chro-mosome in males (Kuroda et al., ’91) as well asnormal levels of the Na+ channel subunit encodedby the gene parats (Kernan et al., ’91; Reenan etal., 2000). The molecular mechanism for the re-duction in numbers of Na+ channels is a splicingcatastrophe in which one or more exons may beskipped in the splicing process (Reenan et al.,2000). There is behavioral evidence indicating thatnapts interacts with the cacophony (cac) gene, amutation of which alters the male mating song andresults in convulsion-like reactions to elevated tem-peratures (Peixoto and Hall, ’98). The lesion is pu-tatively in a gene encoding a Ca2+ channel subunit(Peixoto and Hall, ’98). If the molecular interac-

TABLE 2. Effect on heart rate and rhythmicity by varying copies of nap and shia

20°C 25°C 30°C 35°C 37°C

Canton-S 2.28 ± 0.05 2.41 ± 0.04 2.72 ± 0.04 3.08 ± 0.05 3.33 ± 0.05(N=60) 0.62 ± 0.04 0.54 ± 0.04 0.45 ± 0.04 0.45 ± 0.05 0.40 ± 0.04shits1/Y;nap/nap 2.13 ± 0.06 2.25 ± 0.08 2.39 ± 0.16 2.69 ± 0.15 2.63 ± 0.12b

(N=10) 0.47 ± 0.08 0.40 ± 0.08 0.43 ± 0.11 0.41 ± 0.08 0.26 ± 0.06shi/shi;nap/nap 1.97 ± 0.12b 1.99 ± 0.20b 2.32 ± 0.20b 2.23 ± 0.18b 2.17 ± 0.24b

(N=12) 0.40 ± 0.06 0.34 ± 0.05 0.33 ± 0.06 0.33 ± 0.07 0.27 ± 0.05shi/Y;nap/+ 1.91 ± 0.11b 2.27 ± 0.12 2.59 ± 0.13 2.92 ± 0.13 3.04 ± 0.13(N=9) 0.46 ± 0.10 0.43 ± 0.10 0.42 ± 0.13 0.40 ± 0.11 0.39 ± 0.12shits1 1.76 ± 0.11b 1.79 ± 0.21b 1.84 ± 0.29b 2.07 ± 0.25b 1.59 ± 0.32b

(N=10) 0.45 ± 0.07 0.43 ± 0.09 0.42 ± 0.08b 0.28 ± 0.07b 0.16 ± 0.04b

napts1 1.73 ± 0.10b 1.82 ± 0.11 1.66 ± 0.18b 1.58 ± 0.21b 1.50 ± 0.24b

(N=20) 0.32 ± 0.05b 0.38 ± 0.06b 0.33 ± 0.06b 0.30 ± 0.06b 0.21 ± 0.05b

shi/shi;nap/+ 1.72 ± 0.13b 1.87 ± 0.17b 2.11 ± 0.21b 2.52 ± 0.13 2.43 ± 0.42b

(N=11) 0.11 ± 0.09b 0.34 ± 0.22 0.51 ± 0.06 0.12 ± 0.08b 0.29 ± 0.06aTop lines are heart rates, bottom rhythmicity scores. Data are ordered by heart rate at 20°C, ± S.E.M. bIndicates the mean is significantlylower than that of wild type (ANOVA, Tukey’s test, α = 0.05).

88 E. JOHNSON ET AL.

tion between napts and cac is the same as describedfor napts and parats, it would presumably result ina reduction of Ca2+ channels. Cardiac excitabilitywould be reduced, with a concomitant lowered re-sponse to neurotransmitters.

In this scenario, the actions of shi and napwould be opposite—overstimulation vs. loweringof sensitivity, respectively, with a resulting wild-type phenotype. There is precedent for this inter-pretation. The mutations ether-a-go-go (eag)(Kaplan and Trout, ’68) and Shaker (Sh) (Catsch,’44; Kaplan and Trout, ’68) encode K+ channels.Double mutants comprising nap and either ofthese mutations show wild-type behavior, owingto a restored balance between the two antagonis-tic channels (Ganetzky and Wu, ’82). Given theknown molecular interaction of napts with parats,we presume that napts would act similarly on shits.Double mutants could produce only a more pro-foundly aberrant phenotype by this mechanism,thus the interaction that produces normal heart-beat in the double mutants likely occurs at thephysiological level.

Dynamins form a large and diverse family ofproteins. In mammals, there are at least threegenes that encode subfamilies, and the numberof potential protein types is multiplied further bysplice variations (Urrutia et al., ’97). The prod-ucts of one of the mammalian genes, DYNII, isubiquitous and functions in clathrin-mediated en-docytosis wherever found (Urrutia et al., ’97). InDrosophila, multiple splice variants of shi encodemultiple dynamins (Chen et al., ’91). We suggestthat non-lethal defects in human dynamins in-volved in endocytic sequestration of cardioactiveneurotransmitter receptors will turn out to un-derlie some congenital cardiac defects.

CONCLUSIONSThe Drosophila cardiac system offers a useful

invertebrate model for the study of myogenic pace-maker function at the level of constituent ionchannels, modulation of cardiac activity by neu-rotransmitters, and the genetics of heart disease.To date, two genes originally discovered in Droso-phila have proven to have human homologues thatunderlie the etiology of cardiac pathology. The firstis tinman (tin), an NK-2 type homeobox-contain-ing gene essential for cardiac development inDrosophila (Kim and Nirenberg, ’89). NKX2-5, amammalian homologue of tin, has been implicatedin a developmental atrial septal defect and atrio-ventricular nodal dysfunction (Schott et al., ’98).The second is ether a-go-go (eag) (Bruggemann et

al., ’93), which produces heart arrhythmia inDrosophila (Johnson et al., ’98); HERG, a mam-malian homologue of eag, underlies a form of LongQT syndrome when defective (Keating and San-guinetti, ’96). We expect further parallels to beuncovered as we delineate the molecular processesof the pacemaker.

LITERATURE CITEDAshburner M. 1989. Drosophila, a laboratory manual. Cold

Spring Harbor, NY: Cold Spring Harbor Press.Barchi R. 1988. Probing the molecular structure of the

voltage-dependent sodium channel. Annu Rev Neurosci11:455–495.

Bruggemann A, Pardo LA, Stuhmer W, Pongs O. 1993. ethera-go-go encodes a voltage-gated channel permeable to K+

and Ca2+ and modulated by cAMP. Nature 365:445–448.Catsch A. 1944. Eine erbliche Störung des Bewegung-

mechanismus bei Drosophila melanogaster. Z Ind AbstVererb 82:64–66.

Chatfield C. 1980. The analysis of time series. London:Chapman and Hall.

Chen MS, Obar RA, Schroeder CC, Austin TW, Poodry CA,Wadsworth SC, Vallee RB. 1991. Multiple forms of dynaminare encoded by shibire, a Drosophila gene involved in en-docytosis. Nature 351:583–586.

Curtis NJ, Ringo JM, Dowse H. 1999. Morphology of the pu-pal heart, adult heart, and associated tissues in the fruitfly, Drosophila melanogaster. J Morphol 240:225–235.

Dowse HB, Ringo J. 1989. The search for hidden periodicitiesin biological time series revisited. J Theor Biol 139:487–515.

Dowse HB, Ringo JM, Power J, Johnson E, Kinney K, WhiteL. 1995. A congenital heart defect in Drosophila caused byan action potential mutation. J Neurogenet 10:153.

Gagnon AW, Kallal L, Benovic J. 1998. Role of clathrin-medi-ated endocytosis in agonist-induced downregulation of theβ2-adrenergic receptor. J Biol Chem 273:6976–6981.

Ganetzky B, Wu C-F. 1982. Indirect suppression involvingbehavioral mutants with altered nerve excitability in Droso-phila melanogaster. Genetics 100:597–614.

Grigliatti T, Hall L, Rosenbluth R, Suzuki DT. 1973. Tem-perature-sensitive mutations in Drosophila melanogasterXIV A selection of immobile adults. Molec Gen Genet120:107–114.

Jackson FR, Wilson S, Stricharz G, Hall L. 1984. Two typesof mutants affectin voltage-sensitive sodium channels inDrosophila melanogaster. Nature 308:189–191.

Janse MJ. 1992. Reentrant arrhythmias. In: Fozzard HA etal., editors. The heart and cardiovascular system. 2nd edi-tion. New York: Raven Press, p 2055–2094.

Johnson E, Ringo J, Dowse H. 1997. Modulation of Droso-phila heartbeat by neurotransmitters. J Comp Physiol167:89–97.

Johnson E, Ringo J, Bray N. Dowse H. 1998. Genetic andpharmacological identification of ion channels central to theDrosophila cardiac pacemaker. J Neurogenet 12:1–24.

Johnson E, Ringo J, Dowse H. 2000. Native and heterolo-gous neuropetides are cardioactive in Drosophila melano-gaster. J Insect Physiol 46:1229–1236.

Jones JC. 1977. The circulatory system of insects. Spring-field, IL: Charles C. Thomas. p 7–32.

Kaplan W, Trout W. 1968. The behavior of four neurologicalmutants of Drosophila. Genetics 61:399–409.

DYNAMIN IS CENTRAL TO CARDIAC FUNCTION 89

Kauvar L. 1982. Reduced [H3]-tetrodotoxin binding in the naptsparalytic mutant of Drosophila. Genetics 187:172–173.

Keating MT, Sanguinetti M. 1996. Molecular genetic insightsinto cardiovascular disease. Science 272:681–685.

Kernan M, Kuroda M, Kreber R, Baker B, Ganetzky B. 1991.napts, a mutation sodium channel activity in Drosophila,is an allele of mle, a regulator of dosage compensation. Cell66:949–959.

Kim Y, Nirenberg M. 1989. Drosophila homeobox genes. ProcNat Acad Sci USA 86:7716–7720.

Kosaka T, Ikeda K. 1983. Reversible blockage of membraneretrieval and endocytosis in the garland cell of the tem-perature-sensitive mutant of Drosophila melanogaster,shibirets. J Cell Biol 97:499–507.

Kuroda M, Kernan M, Kreber R, Baker B, Ganetzky B. 1991.The maleless protein associates with the X-chromosome toregulate dosage compensation in Drosophila. Cell 66:935–947.

Lee K, Pals-Rylaarsdam R, Benovic J, Hosey M. 1998.Arrestin-independent internalization of the m1, m3, and m4subtypes of muscarinic cholinergic receptors. J Biol Chem273:12967–12972.

Livingstone MS, Tempel B. 1983. Genetic dissection ofmonoamine neurotransmitter synthesis in Drosophila. Na-ture 303:67–70.

Marsh M, McMahon HT. 1999. The structural era of endocy-tosis. Science 285:215–220.

Masur, S, Kim, Y-T, Wu, C-F. 1990. Reversible inhibition ofendocytosis in cultured neurons from the temperature-sen-sitive mutation shibirets. J Neurogenet 6:191–206.

Miller TA. 1985. Structure and physiology of the circulatorysystem. In: Kerkut GA, Gilbert L, editors. Comprehensiveinsect physiology, biochemistry, and pharmacology. vol III.New York: Pergammon Press. p 289–353.

Peixoto A, Hall J. 1998. Analysis of temperature-sensitivemutants reveals new genes involved in the courtship songof Drosophila. Genetics 148:827–838.

Reenan R, Hanrahan CJ, Ganetzky B. 2000. The mlenaptsRNA helicase mutation if Drosophila results in a splicingcatastrophe of the para Na+ channel transcript in a regionof RNA editing. Neuron 25:139–149.

Rizki T. 1978. The circulatory system and associated cellsand tissues. In: Ashburner M, Wright TRF, editors. The ge-netics and biology of Drosophila. Cold Spring Harbor, NY:Cold Spring Harbor Press, p 1839–1845.

Scaife RM, Margolis RL. 1997. The role of the PH domainand SH3 binding domains in dynamin function. Cell Sig-naling 9:395–401.

Schott J-J, Benson D, Basson C, Pease W, Silberbach G, MoakJ, Maron B, Seidman C, Seidman J. 1998. Congenital heartdisease caused by mutations in the transcription factorNKX2-5. Science 281:108–111.

Suzuki DT, Grigliatti T, Williamson R. 1971. Temperature-sensitive mutations in Drosophila melanogaster VII. A mu-tation (parats) causing a reversible adult paralysis. Proc NatlAcad Sci USA 68:890–893.

Ulrych T, Bishop T. 1975. Maximum entropy spectral analy-sis and autoregressive decomposition. Rev Geophys SpacePhys 13:183–200.

Urrutia R, Henley JR, Cook T, McNiven M. 1997. Thedynamins: Redundant or distinct functions for an expand-ing family of related GTPases? Proc Nat Acad Sci USA94:377–384.

van der Bliek A, Myerowitz E. 1991. Dynamin-like proteinencoded by the Drosophila shibire gene associated with ve-sicular traffic. Nature 351:411–414.

Wright T. 1977. The genetics of dopa decarboxylase and α-methyl dopa sensitivity in Drosophila melanogaster. J ComplPhysiol B 162:278–283.

Wu C-F, Ganetzky B. 1980. Genetic alteration of nerve mem-brane excitability in temperature-sensitive paraltic muta-tions of Drosophila melanogaster. Nature 286:814–816.

Wu C-F, Ganetzky B, Jan Y, Jan L, Benzer S. 1978. A Droso-phila mutant with a temperature-sensitive block in nerveconduction. Proc Natl Acad Sci USA 75:4047–4051.

Zhang J, Barak L, Winkler K, Caron M, Ferguson S. 1997. Acentral role for β arrestins and clathrin-coated vesicle-me-diated endocytosis in β2-adrenergic receptor resensitization.J Biol Chem 272:27005–27014.