characterization of development, behavior and

Comparative Biochemistry and Physiology Part A 136(2003) 427–439

1095-6433/03/$ - see front matter� 2003 Elsevier Science Inc. All rights reserved.doi:10.1016/S1095-6433(03)00200-9

Characterization of development, behavior and neuromuscularphysiology in the phorid fly,Megaselia scalaris

Douglas A. Harrison, Robin L. Cooper*

Department of Biology, University of Kentucky, Lexington, KY 40506-0225, USA

Received 30 July 2002; received in revised form 24 June 2003; accepted 25 June 2003

Abstract

The Phoridae is known as ‘scuttle flies’ because they walk in rapid bursts of movement with short pauses between.In this study, larval locomotive behavior and development was characterized in the phorid,Megaselia scalaris.Comparison was made with the well-characterized fruit fly model,Drosophila melanogaster. Developmentally, the rateof maturation was consistently slower forMegaselia than Drosophila. This disparity was exaggerated at lowertemperatures, particularly during larval development. In addition to slower growth, movements inMegaselia were alsoslower, as evidenced by reduced rates of larval body wall contractions and mouth hook movements.Megaselia larvaealso displayed a unique behavior of swallowing air when exposed to a small pool of liquid. This permitted floating uponimmersion and, therefore, might prevent drowning in the natural environment. The anatomical and physiologicalproperties of a neuromuscular junction in the phorid larvae were also examined. The innervation of the motor nerveterminals on the ventral abdominal muscle(m6) is innervated by Type Ib and Is axons, similar toDrosophila. As inDrosophila, the Is terminals produce larger excitatory postsynaptic potentials(EPSPs) than the Ib. The amplitudes ofthe EPSPs inM. scalaris were reduced compared to those ofD. melanogaster, but unlikeD. melanogaster the EPSPsshowed marked facilitation when stimulated with a 20 Hz train. We conclude that there may be differences in synapticstructure of the nerve terminals that could account for the different electrophysiological behaviors.� 2003 Elsevier Science Inc. All rights reserved.

Keywords: Neurotransmission; Neuromuscular junction; Behavior; Development

1. Introduction

The genetics, development, learning, behaviorand synaptic physiology of the neuromuscularjunctions have been examined for decades in thefruit fly model, Drosophila melanogaster. How-ever, other fly families show interesting character-istics that are not seen inDrosophila. Includedamong these are particular behaviors that mayoffer insights into different sensory, motor and

*Corresponding author. Tel.:q1-859-257-5950; fax:q1-859-257-1717.

E-mail address: [email protected](R.L. Cooper).

behavioral mechanisms and their underlying neuraldevelopment and function. For example, the phoridfly, Megaselia scalaris, shows an unusual adultlocomotory behavior. This species walks in shortbursts with periods of rest between, hence givingphorids, the common name of ‘scuttle flies’(Mil-ler, 1979). To date, the physiological basis of thisphenomenon remains to be determined.

Phorids are a diverse family in which somespecies have distinctive wing venation, while oth-ers may have reduced wings or no wings at all(Disney, 1994). Basic development of the commonphorid speciesM. scalaris has been studied with

428 D.A. Harrison, R.L. Cooper / Comparative Biochemistry and Physiology Part A 136 (2003) 427–439

regard to sex determination(Traut, 1994), oogen-esis (Disney, 1994; Greenberg and Wells, 1998),fecundity (Benner, 1985) and spermatogenesis(Wolf et al., 1993). Also, descriptions of growthrates in the wild have proven useful in forensicinvestigations(Greenberg and Wells, 1998). How-ever, because of the specific interest in forensicapplications, these growth rates for larvalM. sca-laris have previously been determined usingground meat as a dietary source(Greenberg andWells, 1998). Because phorids can thrive on atremendous variety of food sources, we examinedgrowth on a standard cornmeal-based diet com-monly used for rearingDrosophila. Investigationof developmental rates showed that although, theinitial egg and final adult sizes of males are verysimilar between the two species, the life cycle issignificantly longer and the larvae and pupae aremuch larger forM. scalaris.

The characteristic locomotion of the adultM.scalaris has been well documented. Using high-speed video recording, it has been shown that bothsexes have the distinctive scuttling behavior,though the males have a greater tendency forcontinuous movements(Miller, 1979). However,the mechanisms underlying this behavior and theability to modulate this locomotive pattern byendogenous or external environmental conditionshave not been determined. In contrast to the adults,phorid larvae show fluid and continuous move-ments that are typical of other fly species. Todetermine whether neuroanatomy and neurophy-siology of phorid larvae are also similar to otherfly species, a representative neuromuscular junc-tion was examined. We investigated innervationprofiles of the motor neurons on the larval bodywall muscles(m6 and m7) and characterized thesynaptic properties of the neuromuscular junctionsof these muscles. The m6 and m7 abdominalmuscles and their associated neuromuscular junc-tions are the most studied muscles of larvalD.melanogaster for neural development, target rec-ognition, synaptic plasticity and synaptic physiol-ogy. We conclude that although, there are strongsimilarities in neuroanatomy of these junctions,there are several differences in the physiology ofM. scalaris, including decreased excitatory post-synaptic potentials(EPSPs) and facilitation ofEPSPs in response to a train of stimulatory pulses.

Lastly, in the course of behavioral analysis, wenoted a unique behavior inM. scalaris larvae.When the phorid larvae are placed in shallow

liquid, they rapidly swallow air, producing bubblesin the gut. We suggest that such behavior mayhave adaptational value for the species in dampenvironments, providing buoyancy that may allowthe larvae to float in shallow water.

2. Methods

2.1. Fly stock and staging of larvae

A common wild-type laboratory strain ofD.melanogaster, Canton S, was compared with anative strain ofM. scalaris, which was collectedon the University of Kentucky campus(Lexington,KY). Synchronous egg collections were made over2 h periods on apple juice–agar plates with yeastpaste. The eggs were transferred to a standardcornmeal–agar medium(Lewis, 1960) andallowed to hatch and develop with a 12:12 dark–light cycle. The developmental studies were carriedout at 18 and 258C. These two temperatures werechosen as representative of the diverse conditionsin which these two species thrive. Also, a substan-tial amount of data is known for larval develop-ment of Drosophila at these temperatures indefined conditions, which provides a means foradditional comparisons(Li et al., 2002). Themethods used to stage fly larvae have beendescribed previously(Campos-Ortega and Harten-stein, 1985). All the animals were reared in vialswith the same cornmeal–agar-dextrose-yeast medi-um used to rearDrosophila (Lewis, 1960).

2.2. Behavioral assays

Early 3rd instar larvae were used for behavioralassays. Feeding and locomotory behavior wasassessed in both species as described in Necka-meyer(1996). In brief, single animals were placedon a 2% agar surface and the number of bodywall contractions was counted for 2 min, afterwhich an animal was placed in a 2% yeast solutionoverlaid on an agar plate. In this condition,Dro-sophila larvae immediately fed, initiating a patternof repetitive mouth hook movements. At least 10animals were assayed in three independent exper-iments for the number of full mouth hook contrac-tions in 2 min (Sewell et al., 1975). Behavioraldata had previously been gathered at 258C forDrosophila, which allows additional comparisonsto be made withM. scalaris (Neckameyer, 1996).

429D.A. Harrison, R.L. Cooper / Comparative Biochemistry and Physiology Part A 136 (2003) 427–439

2.3. Anatomy of the neuromuscular junctions(NMJs)

Morphology of the motor nerve terminals onmuscle m6 was examined using fluorescently-labeled anti-HRP viewed by confocal microscopy(Johansen et al., 1989). Fluorescent images of thenerve terminals were viewed with a Leica DM REupright fluorescent microscope using a 40= waterimmersion objective. The composite images of theZ-series were collected with a Leica TCS NTySPconfocal microscope for illustration. The Leicasoftware was used to digitize the images.

2.4. Electrophysiology

The preparations were taken from early and late3rd instar larvae staged as described above. Thelarval dissections were performed as described inCooper et al.(1995b). The recording arrangementand solutions are as previously described(Stewartet al., 1994; Neckameyer and Cooper, 1998). Thephysiological saline contained(in mM): 1.0CaCl , 70 NaCl, 5 KCl, 10 NaHCO , 5 trehalose,2 3

115 sucrose, 5 BES(N,N-bisw2-Hydoxyethylx-2-aminoethanesulfonic acid). All experiments wereperformed at room temperature(19–228C). Intra-cellular recordings were made with the use ofmicroelectrodes filled with 3 M KCl having aresistance of 30–60 MV. The electrophysiologicalvalues of interest were the resting membranepotential (Rp), the EPSP amplitude for both Isand Ib motor nerve terminals in segment four ofmuscle six. To compare EPSP amplitudes betweenspecies, a short train of four stimuli(20 Hz) wasgiven. A facilitation index(FI) was calculated bytaking the ratio of the 4th EPSP amplitude overthe 1st EPSP amplitude and subtracting one. Thus,if no facilitation occurred, a value of zero wouldbe obtained. The peak amplitude values for theEPSPs were measured from the base, precedingthe response, to the peak. If the responsesdecreased in amplitude throughout the stimulustrain, a negative value would result. The mean andstandard error of the mean(S.E.M.) was calculatedfor the resting membrane potential and EPSPresponses. The individual EPSP amplitudes of theIb, Is as well as the mean value for the compositeof recruiting Ib and Is together were comparedbetween species. The input resistance of the musclefibers was determined by plotting the slope of thecurrent–voltage relationship based on Ohm’s law

for current injections(250 ms in duration) overthe range of"5 nA.

Intracellular potentials were recorded with a 1X LU head stage and an Axoclamp 2A amplifier.Field recorded excitatory postsynaptic potentials(fEPSPs) were measured with focal macropatchelectrodes in conjunction with a 0.1 X LU headstage and an Axoclamp 2A amplifier in a bridgemode configuration. The fEPSP synaptic potentialswere obtained using the loose patch technique bylightly placing a 10–20mm fire polished glasselectrode directly over a spatially isolated varicos-ity along the nerve terminal that were viewedunder a 40= water immersion lens(Nikon, NA0.55). The macropatch electrode is specific forrecordings within the region of the electrodelumen. The seal resistance was in the range of 100KV to 1 MV. The evoked fEPSPs and fieldminiature excitatory postsynaptic potentials(fm-EPSPs) were recorded to VHS tape and directlyto MacLab Scope software(ver. 3.5.4) for calibra-tion and measurement.

3. Results

3.1. Growth

In D. melanogaster, overall larval growth isgradual, without large punctuate jumps betweeninstars (Alpatov, 1929; Morgan, 1926; Plunkett,1926). This gradual growth also holds true specif-ically for discrete muscle groups(Li et al., 2002).Development of theM. scalaris larvae also showsthis progressive increase in size, however,Mega-selia have longer bodies as larvae and pupae ascompared to identically rearedDrosophila (Fig.1). This size difference is particularly dramatic bythe pupal stage(Fig. 1c). Interestingly, the eggsof Megaselia are only slightly larger thanD.melanogaster, with distinctly different morphologyof the chorion(Fig. 1a). Furthermore, the adultmales of the species are also very similar in size.Therefore, despite great disparity in sizes of theintermediate stage animals, the beginning embryosand resulting adult males are quite similar inD.melanogaster and M. scalaris. In contrast, thephorid females are much larger than either themales or theD. melanogaster females. AlthoughD. melanogaster females are larger than males,this sexual dimorphism is much more pronouncedin M. scalaris. It should be noted thatM. scalarisis a scavenger species and can grow in the wild


Fig. 1. Comparison of morphology and size ofM. scalaris andD. melanogaster throughout development. Though the eggs ofthe two flies are similar in size,(a) all larval instars,(b) pupae,(c) and adult females,(d bottom row) of M. scalaris are sig-nificantly larger thanD. melanogaster. M. scalaris is on theleft side of each panel. The adult males of both species areshown in the upper row of d. Scale bar is 0.1 mm for panel Aand 1 m for B, C and D.

on a variety of media, including rotting meat. Inthe laboratory, this species is also reared on variousfood sources, but is frequently grown on a modi-fied Drosophila medium consisting of a cornmeal-based food supplemented with a small piece ofmeat (Willhoeft and Traut, 1990). However, forthe purposes of this study, both species were rearedon an identical diet, comprised of a standardcornmeal–agarDrosophila medium without anymeat.

Consistent with their greater size, the phoridsdeveloped more slowly thanDrosophila. Compar-ison was made in the developmental rate throughembryonic hatching and through to pupation forM. scalaris and D. melanogaster at both 18(Fig.2a and b) and 25 8C (Fig. 2a and b). Both1 1 2 2

embryonic and larval growth is approximately 1.5times longer in the phorids than inDrosophilareared at 258C (Fig. 2c). The relative rate alsoholds for embryonic development of the two spe-cies at 18 8C. However, larval development isdramatically slower(three-fold) for phorids thanfor Drosophila at 18 8C (Fig. 2b ). The extended1

larval stages ofM. scalaris as compared withD.melanogaster likely reflect the substantially greateramount of growth required in the phorids to reachpupariation. Both species tend to pupariate out ofthe food in a dry location, such as on the lid orwalls of the growth chambers.

3.2. General behaviors

Since the adult locomotive behavior is so strik-ingly different between these two species, wepredicted that there may be some difference inlarval movement. To test this prediction, two stan-dard behavioral assays of body movements wereused: the rate of body wall contraction and therate of mouth hook movements(Neckameyer,1996). Animals in the early wandering 3rd instarwere used for both species. In contrast to adultlocomotion, the movements of phorid larvae arefluid and continuous. Yet, in comparison withDrosophila larvae, the locomotor behavior of thebody wall contractions was substantially reducedfor M. scalaris at 18 8C (Fig. 3a ;P-0.05, t-test,1

ns20), though not significantly different at 258C(Fig. 3a). Even more notable differences were2

observed in the number of mouth hook movementswhen the larvae were exposed to a 2% yeastsolution overlaid on an agar plate.M. scalaris areless active in eating the food, and thus they have


Fig. 2. A rank sum of the number of embryos hatched and animals pupated over time for bothM. scalaris andD. melanogaster. M.scalaris have longer embryonic(a ) and dramatically longer larval developmental periods at 188C (b ). The differences between the1 1

two species are not as dramatic at 258C (a and b). The time required for 50% of the embryos to develop from egg deposition to2 2

hatching(H ) and for 50% of the larvae to develop from egg deposition to pupariation(P ) is indicated for each species at each50 50

temperature(c).

fewer mouth hook movements thanD. melanogas-ter (P-0.05, t-test,ns20; Fig. 3b and b). The1 2

differences are greater at 258C than at 188C. Thereduced feeding rate inM. scalaris is becauselarvae expended effort in swallowing air whenplaced on a yeast solution that was nearly able toimmerse the animal.D. melanogaster did not elicitthis type of behavior, yet both species were ableto breathe under these conditions by raising thespiracles out of the solution to maintain access toair. This air swallowing inM. scalaris caused thegut to fill with air bubbles(Fig. 4). Because this

behavior was only observed upon exposure to aliquid environment, we hypothesized that thismight be a survival strategy to allow the animalto float if immersed. To test this idea, we exposedboth M. scalaris and D. melanogaster to the 2%yeast solution used for the behavioral assay for10, 20 or 60 min. This afforded the larvae anopportunity to swallow air. Afterwards, we placedthe larvae into a 100 ml beaker containing roomtemperature water(80 ml). All of the D. melano-gaster (ns60) sank to the bottom and died aftera short time, whereas many of theM. scalaris


Fig. 3. The locomotor behavior assessed by the number of bodywall contractions per minute(averaged over 2 min) for early3rd instar is reduced forM. scalaris as compared toD. melan-ogaster at 18 8C (a ) but not at 258C (a ). The number of1 2

mouth hook movements per minute(averaged over 2 min)when exposed to a 2% yeast solution was substantially lowerfor M. scalaris thanD. melanogaster at both temperatures(b).

Fig. 4. When partly immersed in aqueous solution,M. scalaris swallowed air, filling the gut with bubbles(arrowheads) that aidedlarvae in floating(see text). D. melanogaster did not show this behavior.

larvae fared much better. After only a 10 min pre-exposure to yeast paste, eight out of 10 larvaesank rapidly, while the remaining two were unableto float and bobbed within the water column. After

a 20 min pre-exposure, 11 out of 20 larvae floated(10 entirely on the surface and one within thewater column), and even after 30 min in the water,only 50% sank and died. When given a 60 minpre-exposure in yeast paste, nine out of 10 larvaefloated and survived for 30 min in water. When aflat wooden ramp(i.e. a clinical type of tonguedepressor) is placed in the beakers, the floatinglarvae ofM. scalaris crawled up on the wood andwere able to fully escape the water. We concludethat the air swallowing response can be a success-ful means of preventing drowning, depending uponthe amount of time permitted before completeimmersion.

3.3. Anatomy of larval muscle and NMJs

The innervation pattern for the neuromuscularjunctions on muscle six(m6) and muscle seven(m7) within abdominal segment four is differentin late 3rd instarM. scalaris than for D. melano-gaster. In Drosophila, the terminals are primarilyType I endings from the two major axons(Is andIb) (Atwood et al., 1993). The two neurons inner-vate m6 and m7 in the region, where the twomuscles are in close apposition to one another.The terminals of the axons branch from thisapposition point on to the muscle(Fig. 5a). In M.scalaris, a nerve branch innervating m6 frequentlyarises from a lateral edge of the muscle(Fig. 5b,double arrow). In addition, it is unusual to observein D. melanogaster such separation in the two


Fig. 5. Neuromuscular junctions on abdominal muscles six and seven within abdominal segment four at the late 3rd instar forM.scalaris. The larvae were raised at 188C. Motor terminals were labeled with fluorescein-tagged anti-HRP and imaged with a confocalmicroscope. Muscle six is the muscle on the right side of both panels. Note the lateral innervation on m6 in panel B(double arrowhead).The Ib and Is terminals are readily identifiable with the Ib terminals containing bigger varicosities as compared to the smaller ones forthe Is terminal. Scale bar: 34mm for A and 51mm for B.

nerve terminals as seen often inM. scalaris (Fig.5b). As for D. melanogaster, the anatomic profilesof the terminals match the Is and Ib terminalclassification forM. scalaris. The type Ib (big)terminals have large varicosities as compared tothe Type Is (small) terminals with smallervaricosities.

3.4. Synaptic physiology at the NMJ

We assessed the synaptic transmission of theneuromuscular junctions ofM. scalaris in compar-ison to D. melanogaster for two well establishedmotor neurons. The terminals of the Ib and Ismotor neurons are well known in structure andfunction in D. melanogaster and for a species offlesh fly, Sarcophaga bullata (Feeney et al., 1998),so we chose to investigate these two motor nerveson muscle six within the 4th abdominal segmentin M. scalaris for comparative physiological pur-poses. The Ib axon produces smaller EPSPs thanthe Is axon in bothM. scalaris andD. melanogas-ter. Fig. 6a shows representative examples of theEPSPs of a larva in the late 3rd instar. To comparedifferences and variability between species, the

mean and S.E.M. for the composite EPSPs(IbqIs) are shown(Fig. 6c). The amplitudes of thecomposite responses were significantly lower inM. scalaris as compared toD. melanogaster (P-0.05, t-test, ns13). The individual responses ofthe Ib (5.6 mV mean"0.5 S.E.M.) and Is (11.2mV mean"0.9 S.E.M.) responses were also lowerin M. scalaris as compared to the Ib(14.6 mVmean"0.8 S.E.M.) and Is (25.8 mV mean"1.3S.E.M.) in D. melanogaster (P-0.05, ns13).The same physiological saline(HL3) was used toexamine physiological profiles in both species.The saline was developed forD. melanogaster butalso appeared sufficient forM. scalaris. The aver-age values for the resting membrane potentialswere less negative inM. scalaris as compared toD. melanogaster (Fig. 6d). Additional characteri-zation of the synaptic efficacy was carried out inthe presence of short stimulus trains. Trains of fourpulses delivered at 20 Hz are a commonly usedparadigm in physiologic studies ofD. melanogas-ter. With this stimulation paradigm inM. scalaris,an unexpected result occurred. The amplitude ofthe EPSPs within the response train facilitated(Fig. 6b). In D. melanogaster, the EPSP responses


Fig. 6. In abdominal muscle six, the Ib axon produces a smaller EPSPs than the Is axon for both species. Representative examples areshown(a) of EPSPs recorded within the late 3rd instar while either the Ib or Is axons were selectively stimulated. Short stimulus trainsof four pulses delivered at 20 Hz indicated that the responses facilitated forM. scalaris as compared toD. melanogaster (b). Theaverage responses for the Ib and Is as well as the composite IbqIs EPSP for both species are shown(c) along with the average restingmembrane potential(d). The facilitation index(FI) shows that in the same physiological environment, synaptic transmission is differentbetweenM. scalaris andD. melanogaster with Megaselia having more pronounced facilitation(e). Error bars are for standard error ofthe mean. All recordings were made in HL3 saline(see Section 2). The scale bar in a is for bothM. scalaris (M.s.) andD. melanogaster(D.m.). The same time scale in b is also for both species.

depress in size, thus producing a negative facili-tation index(Fig. 6b). The facilitation index forthe two species is shown in Fig. 6e. In all themeasures indicated, there is a significant difference(P-0.05) between the two species.

To determine if a difference in the muscle fiberinput resistance might contribute to the differencesof the EPSP amplitudes between the species, meas-ures were taken in the same conditions for bothspecies. The input resistance forM. scalaris wason average 7.6 MV ("1.1 MV, S.E.M.; ns6)and forD. melanogaster, the input resistance under

the same conditions has been reported to be onaverage 7.5 MV ("0.3 MV, S.E.M.;ns6; Stew-art et al., 1996). Thus, there is no significantdifference in muscle fiber input resistance betweenthe two species.

To quantify the field recorded excitatory post-synaptic potentials(fEPSPs), focal macropatchrecordings were made over selective regions of theIb and Is nerve terminals. To insure that the fEPSPsand EPSPs corresponded with the selective stimu-lation, the Ib or Is axons, EPSPs were recorded inmuscle seven(m7), while fEPSPs were recording


Fig. 7. In order to match synaptic transmission with terminal types, simultaneous fEPSPs and intracellular obtained EPSPs recordingsare made while recruiting selectively the axons of the segmental nerve(sn). The EPSPs were monitored in m7, while the fEPSPs weremonitored over the two terminal strings of small(Is) or large(Ib) varicosities on m6(a). The fEPSPs obtained from small varicositieswere only observed when the small EPSP or compound EPSP was recorded, but not the single large EPSP(b). In contrast, the fEPSPsobtained from large varicosities were only observed when the large EPSP or compound EPSP was recorded, but not the small largeEPSP(c). Scale bars are in mV for the EPSPs and the fEPSPs(in parentheses).

from selective varicosities of the two types ofterminals on m6. The axons of Ib and Is bifurcateto innervate both m6 and m7 muscles within eachhemi-segment. This stimulation and selectiverecording procedure was similar to that describedby Kurdyak et al. (1994) in differentiating theresponses from Ib and Is terminals inDrosophilaexcept that we recorded EPSPs in m6. The modi-fied procedure allowed more varicosities to be

sampled on m6 without producing damage at thesite of recordings (Fig. 7). We consistentlyobserved that the Is terminals produced largerfEPSPs with more multiples of quantal contentthan the Ib terminals(Fig. 8a and b). This isobserved for both species and has previously beenwell documented inD. melanogaster (Kurdyak etal., 1994; Stewart et al., 1994; Cooper et al.,1995b). The field potentials recorded directly over


Fig. 8. To quantify the field excitatory postsynaptic potentials(fEPSPs), focal macropatch recordings were made over visu-alized varicosities on the nerve terminals. The Ib terminals(a)produces smaller fEPSPs with fewer multiples of quantal con-tent than the Ib terminals(b). This is observed for both speciesand has been well documented inD. melanogaster (Kurdyaket al., 1994; Stewart et al., 1994; Cooper et al., 1995b). Thestimulus artifact is followed by the evoked fEPSP and in thesome instances, a spontaneous quantal event or field miniatureEPSP(fmEPSP) is recorded. An ensemble average of 400evoked responses while recording from a single Ib or a Is var-icosity is shown. The Is varicosities on average have a greaterquantal content per evoked event(c). The scales are the samefor a and b.

the Ib varicosities would depict release of onlysingle quanta or failures when evoked forM.scalaris. This is not a common occurrence for Ibterminals inD. melanogaster, since multiple quan-tal are revealed during an evoked stimulus andrarely are failures in release observed. The singlequantal size is observed by recording spontaneousor field miniature recorded excitatory postsynapticpotentials(fmEPSPs) as seen later in the tracesfor both the Ib and Is responses(Fig. 8a and b).Averages of 400 evoked trials depict the overall

difference in synaptic efficacy for the varicositieson the Ib and Is terminals. The average of evokedfEPSPs for the Ib is markedly smaller than thatfor the Is terminals(Fig. 8c).

4. Discussion

Although, there are distinct morphological dif-ferences betweenM. scalaris andD. melanogaster,both develop from eggs of very similar size thatgrow to be adults that are comparable in size, atleast for males(M. scalaris adult females aresubstantially larger than males orD. melanogasterof either sex). Yet along the way, the larvae andpupae ofMegaselia are much larger than compa-rably stagedDrosophila. Consistent with the largerlarval and pupal sizes, the larval development ofM. scalaris is strikingly slower at 188C andsomewhat slower at 258C than that ofD. melan-ogaster under the same environmental conditions.The exaggerated disparity between the species at18 8C may indicate thatM. scalaris is morestressed at that lower temperature thanD.melanogaster.

In an earlier study(Greenberg and Wells, 1998),in which M. scalaris was raised on ground beef,the average minimum times to pupariation atvarious temperatures were approximately 100 h(29 8C), 155 h(22 8C) and 200 h(19 8C). In ourstudy, pupariation started approximately 250 h at25 8C and 420 h at 188C, when larvae weremaintained on a cornmeal diet. Though it is impos-sible to make quantitative comparisons betweenthese separate experiments, partly because therearing temperatures used were different, we canmake estimates of the growth by extrapolationfrom these data. Based on the times to pupariationat 29 and 228C, we know that the time topupariation at 258C when reared on ground meatmust be between 100 and 155 h, and most likelywould be approximately 130 h. When comparedwith the 250 h required to reach pupariation on acornmeal diet, we must conclude thatM. scalarisgrows substantially faster on meat than the vege-table-based food used in this study. Thus, in spiteof the capability of this phorid to survive on awide variety of media, there is a clear differencein the quality of these two types to support optimalgrowth of M. scalaris. However, this does notindicate that M. scalaris grows poorly on thecornmeal medium. Although, they develop moreslowly than animals raised on meat, even the


cornmeal-based food supports robust proliferationof these phorids, producing large numbers of viableprogeny.

Significant differences in specific behaviors ofM. scalaris andD. melanogaster were also noted.The rates of body wall contractions and mouthhook movements are both slower inM. scalaris.However, the substantially slower mouth hookmovements inM. scalaris are due, at least in part,to replacement of feeding activities with a noveland unexpected behavior: larvae swallow air whenexposed to an aqueous environment. This behaviorpermits the larvae to float, even in deep pools ofliquid, if permitted sufficient time to gulp air priorto being exposed to deeper water. This wouldappear to be a successful survival strategy in theevent that larvae become immersed in liquid intheir environment. This might be particularly effec-tive in situations where a larva may be trapped ina gradually increasing volume of liquid, such asduring a rainstorm. However, based on our testsusing only brief exposure prior to total immersion,sudden complete immersion of the larva wouldlikely not permit sufficient swallowing of air tobuild buoyancy before sinking. Thus, the survivalutility of this behavior is limited only to situationsin which immersion would be gradual. As for bodywall contractions between the two examined spe-cies, no significant difference was obtainedbetween the two temperatures tested. Alteration intemperature has substantial effects on synaptictransmission at NMJs for insects. Such slowingdown of synaptic transmission is likely to accountfor the slightly slower movements in body wallcontractions and reduction in mouth hook move-ments at the lower temperatures, since they arenot able to maintain the same degree of synaptictransmission.

In addition to differences in growth rate, sizeand some behaviors, the architecture of the neu-romuscular junction and electrophysiology of thelarvae are also different. The Ib and Is motor nerveterminals and the corresponding muscle six(m6)in segment four of the 3rd instarM. scalaris larvaeshow differences in their properties as comparedto D. melanogaster. The motor nerve terminalinnervation on muscle six is erratic inM. scalarisas compared to the more consistent patternsobserved inD. melanogaster (Li et al., 2002). Thephysiological differences were that the EPSPs forboth the Ib and Is terminals are smaller and theresting membrane potential reduced(less negative)

in M. scalaris. The time constants of the rise anddecay are similar. However, the EPSPs show facil-itation in M. scalaris, whereas inD. melanogasterthe responses depress in the same physiologicalsaline. Similarities exist between the two speciesin that they both have Type Ib and Is terminalsinnervating the m6 and m7 abdominal muscles.Also, the Is terminals in both species produce alarger EPSP than do the Ib terminals, and the fieldpotentials for the Is varicosities are larger thanthose for the varicosities of the Ib. These differ-ences are likely due to the increased number ofvesicular events associated with the synapses con-tained in the Is terminal as compared to the Ibterminal.

Since the m6 muscle is substantially larger inM. scalaris in the wandering 3rd instar as com-pared to D. melanogaster, one may expect theamount of innervation on the muscle from themotor terminals to be proportional to the surfacearea or volume. We found it difficult to quantifythe length of the Ib and Is motor nerve terminalsof M. scalaris as comparedD. melanogaster, sinceinnervation occurred on the ventral and lateralsides on the m6 muscle as compared to the dorsalsurface innervation inD. melanogaster. In growthstudies ofD. melanogaster from 1st instar to late3rd instar stages, both m6 and its motor axons Iband Is were characterized by second order regres-sions after the data was transformed with a naturallog. The number of varicosities and terminal lengthof Is axon increases more rapidly than for the Ibaxon throughout development(Li et al., 2002).One might expect a reduced growth rate forM.scalaris, since their larval develop is reduced at18 8C as compared toD. melanogaster.

The reduced amplitude of the EPSPs for the Iband Is in M. scalaris might be a result of lessinnervation, but there could also be biophysicaldifferences in the postsynaptic muscle fiber toaccount for this, since the events are measured inthe muscle. However, the time constants for therise and decay of the compound EPSPs were notsignificantly different between the two species(data not shown). Since there is no significantdifference in the input resistance of the musclefibers of the two species, which could have impact-ed the rising phase of the EPSPs, one would thenpredict that less transmitter is released from eachsynaptic varicosity or the receptivity of the musclefiber is decreased inM. scalaris. Synaptic fieldpotential recordings obtained directly over discrete-


ly isolated varicosities of the Ib and Is terminalsdid reveal smaller evoked field potential for bothterminals inM. scalaris. This is likely due to thefact that the mean quantal content for varicositiesof the Ib and Is terminals forM. scalaris is lessthan that previously reported forD. melanogaster(Cooper et al., 1995b; Stewart et al., 1996; Kur-dyak et al., 1994).

The homoeostatic regulation of synaptic efficacythroughout development inD. melanogaster is notcompletely elucidated, but many studies note astrong correlation between the length of motornerve terminals and synaptic strength of the vari-cosities with development of the muscle. Severalstudies have used genetic manipulations to addressthese issues inD. melanogaster (Petersen et al.,1997; Davis and Goodman, 1998) and mice(San-drock et al., 1997). Much of the structural regu-lation is at the level of the synapse. It has beendemonstrated that homoeostatic regulation innerve–muscle matching is correlative to the com-plexity of the synapses in containing greater orfewer active zones(Stewart et al., 1996). Activezones are structural entities on the presynapticmembrane where vesicles dock to fuse with themembrane. Such differences in synaptic complex-ity account for the differences between the Ib andIs varicosities inD. melanogaster (Atwood et al.,1993; Kurdyak et al., 1994) and for high and lowoutput synapses in crustaceans(Atwood and Coo-per, 1995, 1996a,b; Cooper et al., 1995a, 1996a,b;Govind and Chiang, 1979; King et al., 1996). Toresolve this issue in relation to synaptic efficacyin M. scalaris, electron micrographic studies ofthe motor nerve terminals will need to be under-taken. To obtain an accurate assessment of thesynaptic area and numbers of active zones persynapse, one would need serial, cross-sections andreconstruction via electron micrographs. At thispoint in time, we were not able to conduct suchstudies.

Future comparative investigations with otherflies regarding the air gulping behavior inM.scalaris would be of interest to determine howbroadly throughout insect evolution this potentiallyadvantageous behavior is employed. In addition,further investigation into the composition of thelarval hemolymph is needed inM. scalaris, sincethe synaptic transmission is quite different(i.e.low EPSP amplitude and high facilitation) fromthat of D. melanogaster. This could be accountedfor, in part, by a low Ca2yMg2 ratio in theq q

HL3 media for M. scalaris. The difference we

report may be due to non-physiologic based ioniccontent for this species. Slight alterations in ioniccomposition of the designed hemolymph-likesalines for D. melanogaster show substantialchanges in the characteristics of the EPSPs(Ballet al., 2003; Stewart et al., 1994). Lastly, it wouldalso be beneficial to obtain comparative informa-tion on the synaptic structure of the Ib and Isterminals.

Acknowledgments

We thank Mr Ryan Ball(University of KY) andMs Racennah Braxton(Talladega college, AL) forhelping to conduct behavioral assays used in thisstudy. Dr Susan Harrison assisted with photogra-phy. Funding was provided by NSF grants IBN-0091535 (DH), IBN-9808631 (RLC) andNSF-ILI-DUE 9850907 (RLC). We thank MrMike Sharkey(Department of Entomology, Uni-versity of Kentucky) and Dr Brian Brown(Asso-ciate Curator of Entomology, Natural HistoryMuseum of Los Angeles County, CA) for speciesidentification.

References

Alpatov, W.W., 1929. Growth and variation of the larvae ofDrosophila melanogaster. J. Exp. Zool. 52, 407–437.

Atwood, H.L., Cooper, R.L., 1995. Functional and structuralparallels in crustaceans andDrosophila neuromuscular sys-tems. Am. Zool. 35, 556–565.

Atwood, H.L., Cooper, R.L., 1996a. Assessing ultrastructureof crustacean and insect neuromuscular junctions. J. Neu-rosci. Methods 69, 51–58.

Atwood, H.L., Cooper, R.L., 1996b. Synaptic diversity anddifferentiation: crustacean neuromuscular junctions. Invert.Neurosci. 1, 291–307.

Atwood, H.L., Govind, C.K., Wu, C.-F., 1993. Differentialultrastructure of synaptic terminals on ventral longitudinalabdominal muscles inDrosophila larvae. J. Neurobiol. 24,1008–1024.

Ball, R., Xing, B., Bonner, P., Shearer, J., Cooper, R.L., 2003.Long-term maintenance of neuromuscular junction activityin culturedDrosophila larvae. Comp. Biochem. Physiol. A134, 247–255.

Benner, D.B., 1985. Oocyte development and fecundity inMegaselia scalaris (Phoridae: Diptera). Int. J. Entomol.(Honolulu) 27, 280–288.

Campos-Ortega, J.A., Hartenstein, V., 1985. The embryonicdevelopment of Drosophila melanogaster. Springer-Verlag,Berlin.

Cooper, R.L., Marin, L., Atwood, H.L., 1995a. Synapticdifferentiation of a single motor neuron: conjoint definitionof transmitter release, presynaptic calcium signals, andultrastructure. J. Neurosci. 15, 4209–4222.

Cooper, R.L., Stewart, B.A., Wojtowicz, J.M., Wang, S.,Atwood, H.L., 1995b. Quantal measurement and analysis


methods compared for crayfish andDrosophila neuromus-cular junctions and rat hippocampus. J. Neurosci. Methods61, 67–78.

Cooper, R.L., Harrington, C., Marin, L., Atwood, H.L., 1996a.Quantal release at visualized terminals of crayfish motoraxon: intraterminal and regional differences. J. Comp. Neu-rol. 375, 583–600.

Cooper, R.L., Winslow, J., Govind, C.K., Atwood, H.L., 1996b.Synaptic structural complexity as a factor enhancing prob-ability of calcium-mediated transmitter release. J. Neurophy-siol. 75, 2451–2466.

Davis, G.W., Goodman, C.S., 1998. Synapse-specific controlof synaptic efficacy at the terminals of a single neuron.Nature. 392, 82–86.

Disney, R.H.L., 1994. Scuttle Flies: The Phoridae. Chapmanand Hall, London, Glasgow.

Feeney, C.J., Karunanithi, S., Pearce, J., Govind, C.K.,Atwood, H.L., 1998. Motor nerve terminals on abdominalmuscles in larval flesh flies,Sarcophaga Bullata: compari-son withDrosophila. J. Comp. Neurobiol. 402, 197–209.

Govind, C.K., Chiang, R.G., 1979. Correlation between pre-synaptic dense bodies and transmitter output at lobsterneuromuscular terminals by serial section electron micro-scopy. Brain Res. 161, 377–388.

Greenberg, B., Wells, J.D., 1998. Forensic use ofMegaseliaabdita and M. scalaris (Phoridae: Diptera): case studies,development rates, and egg structure. J. Med. Entomol. 35,205–209.

Johansen, J., Halpern, M.E., Johansen, K.M., Keshishian, H.,1989. Stereotypic morphology of glutamatergic synapses onidentified muscle cells ofDrosophila larvae. J. Neurosci. 9,710–725.

King, M.J.R., Atwood, H.L., Govind, C.K., 1996. Structuralfeatures of crayfish phasic and tonic neuromuscular junc-tions. J. Comp. Neurol. 372, 618–626.

Kurdyak, P., Atwood, H.L., Stewart, B.A., Wu, C.-F., 1994.Differential physiology and morphology of motor axons toventral longitudinal muscle in larvalDrosophila. J. Comp.Neurol. 350, 463–472.

Lewis, E.B., 1960. A new standard food medium. Drosoph.Inf. Serv. 34, 117–118.

Li, H., Peng, X., Cooper, R.L., 2002. Development ofDro-sophila larval neuromuscular junctions: maintaining synapticstrength. Neuroscience 115, 505–513.

Miller, P.L., 1979. A possible sensory function for the stop-gopatterns of running in phorid flies. Physiol. Entomol. 4,361–370.

Morgan, T.H., 1926. Genetics and the physiology of develop-ment. Am. Nat. LX 1, 489–515.

Neckameyer, W., 1996. Multiple roles for dopamine inDro-sophila development. Dev. Biol. 176, 209–219.

Neckameyer, W.S., Cooper, R.L., 1998. GABA transporters inDrosophila melanogaster: developmental expression, behav-ior, and physiology. Invert. Neurosci. 3, 279–294.

Petersen, S.A., Fetter, R.D., Noordemeer, J.N., Goodman, C.S.,DiAntonio, A., 1997. Genetic analysis of glutamate receptorsin Drosophila reveals a retrograde signal regulating presy-naptic transmitter release. Neuron 18, 1237–1248.

Plunkett, C.R., 1926. The interaction of genetic and environ-mental factors in development. J. Exp. Zool. 46, 181–244.

Sandrock, A.W., Dryer, S.E., Rosen, K.M., Gozani, S.N.,Kramer, R., Theill, L.E., et al., 1997. Maintenance ofacetylcholine receptor number by neuregulin at the neuro-muscular junction in vivo. Science 276, 599–604.

Sewell, D., Burnet, B., Connolly, K., 1975. Genetic analysisof larval feeding behavior inDrosophila melanogaster.Genet. Res. Cambridge 24, 163–173.

Stewart, B.A., Atwood, H.L., Renger, J.J., Wang, J., Wu, C.-F., 1994. Improved stability of Drosophila larval neuromus-cular preparation in haemolymph-like physiologicalsolutions. J. Comp. Physiol. A 175, 179–191.

Stewart, B.A., Schuster, C.M., Goodman, C.S., Atwood, H.L.,1996. Homeostasis of synaptic transmission inDrosophilawith genetically altered nerve terminal morphology. J. Neu-rosci. 16, 3877–3886.

Traut, W., 1994. Sex determination in the flyMegaseliascalaris, a model system for primary steps of sex chromo-some evolution. Genetics 136, 1097–1104.

Willhoeft, U., Traut, W., 1990. Molecular differentiation of thehomomorphic sex chromosomes inMegaselia scalaris (Dip-tera) detected by random DNA probes. Chromosoma 99,237–242.

Wolf, K.W., Jeppesen, P., Mitchell, A., 1993. Spermatid nucle-us of Megaselia scaris Loew (Insecta, Diptera, Phoridae): astudy using anti-histone antibodies, scanning electronmicroscopy, and a centromere-specific oligonucleotide. Mol.Reprod. Dev. 35, 272–276.

characterization of development, behavior and

Documents