IDENTIFICATION OF NEURONES CONTAINING CARDIO- … · manner immediately after adult eclosion, producing a marked rise in heartbeat frequency and a facilitation of wing inflation (Tublitz

J. exp. Biol. 116,395-410 (1985) . 3 9 5Printed in Great Britain © The Company of Biologists Limited 1985

IDENTIFICATION OF NEURONES CONTAINING CARDIO-ACCELERATORY PEPTIDES (CAPs) IN THE VENTRAL NERVE

CORD OF THE TOBACCO HAWKMOTH, MANDUCA SEXTA

BY NATHAN J. TUBLITZ* AND JAMES W. TRUMANDepartment of Zoology, University of Washington, Seattle, WA 98195, U.SA.

Accepted 12 October 1984

SUMMARY

1. The abdominal ganglion neurosecretory cells responsible for the syn-thesis and release of two insect neurohormones, cardioacceleratory peptides 1and 2 (CAPi and CAP2), from the perivisceral organs (PVOs) have beenidentified in the tobacco hawkmoth, Manduca sexta.

2. Previous work established the existence of two groups of abdominalganglion cell bodies with axons projecting to the PVO: four laterally-situatedpairs and five pairs lying on the midline (Taghert & Truman, 1982A). Micro-dissection and bioassay of various parts of an abdominal ganglion revealedthat CAP activity was greatest in the medial portion of the ganglion, theportion containing the 10 midline neurones.

3. Six of the 10 midline neurosecretory cells, the new midline bilateral(MB) cells, appeared to differentiate post-embryonically, commencing dif-ferentiation late in the last larval instar and reaching maturity midwaythrough adult development. The development of the new MB cells wasmirrored by the accumulation of CAP activity in the abdominal nerve cord.Not present in measurable amounts in larvae, CAP activity was firstdetectable a few days after pupation and reached maximal levels midwaythrough adult development.

4. CAP-like bioactivity was collected from the PVO in response to anti-dromic stimulation of the nerve containing the new MB axons. No CAP-likebioactivity was detected in those preparations in which the new MB axonswere severed or in which other nerves were stimulated.

5. Intracellular stimulation of a new MB neurone evoked the release fromthe PVO of measurable levels of CAP bioactivity. It was shown that thisstimulation-evoked, cardioacceleratory activity was sensitive to proteasetreatment, and was released only from the cell that was stimulated. On thebasis of these experiments, it was concluded that the CAPs are synthesizedand secreted from the new MB cells.

• Present address: Department of Zoology, University of Cambridge, Downing Street, Cambridge CB2 3EJ,England.

Key words: Peptides, identified neurones, insect cardioregulation, insect neuroendocrinology.

396 N. J. TUBLHZ AND J. W. TRUMAN

INTRODUCTION

Peptide hormones, produced by the central nervous system, have been implicatedin various types of behaviour in insects, and also perform important roles in theregulation of many physiological processes (Truman & Taghert, 1983; Gainer &Frontali, 1979). Responsible for the synthesis and release of these neuropeptides, theinsect neuroendocrine system has been well-characterized at the cellular level utilizingcytological and ultrastructural techniques (Raabe, 1982; Miller, 1980; Maddrell,1974). Unfortunately, information on the cellular physiology of insect neuro-hormones has not been as forthcoming, primarily because of the relative difficulty inunambiguously identifying the individual neurosecretory cells that produce thesehumoral factors. Identification of the peptide-containing neurones is clearly essentialto our understanding of the control mechanisms which govern neuropeptide synthesisand release.

There are two major neurohaemal organs serving the insect nervous system.Neurosecretory products synthesized in the brain are primarily released from thecorpora cardiaca-corpora allata complex at the base of the insect brain. A parallelneurosecretory system exists in the ventral nerve cord. The primary neurohaemalrelease sites for each of the ventral ganglia are the paired perivisceral organs (PVOs),located on the segmentally-arrayed transverse nerves (Raabe, Cazal, Chalaye & deBesse, 1966; Taghert & Truman, 19826). Although the PVOs appear to contain andrelease several neurohormones, to date only one has been associated with identified,PVO-projecting neurones. Bursicon, the hormone responsible for cuticle sclero-tization in insects, was localized in four pairs of neurones in abdominal ganglia ofManduca sexta (Taghert & Truman, 19826).

In addition to bursicon, two other peptide hormones have been isolated from thePVOs of pharate adult Manduca sexta. Known as Cardioacceleratory Peptides 1 and2 (CAPj and CAP2), they each elicit a dose-dependent increase in heart rate whenpulse applied on an in vitro portion of the Manduca heart (Tublitz & Truman,1985a). They are co-released into the haemolymph from the PVOs in a pulsatilemanner immediately after adult eclosion, producing a marked rise in heartbeatfrequency and a facilitation of wing inflation (Tublitz & Truman, 19856). The presentpaper describes experiments designed to identify the peptidergic neurones respon-sible for the synthesis and release of the two CAPs found in the tobacco hawkmoth,Manduca sexta.

MATERIALS AND METHODS

The methods for rearing of animals, the preparation of tissue for biochemicalanalysis, and gel nitration procedures have been previously described (Tublitz &Truman, 1985a,6).

Identified insect peptidergic neurones 397

In vitro Manduca heart bioassay

The detection of CAP bioactivity was accomplished using an in vitro Manducaheart bioassay described in detail in a previous paper (Tublitz & Truman, 1985a). Tosummarize briefly, a portion of the abdominal heart from a pharate adult male wasdissected free from all adjacent tissue and pinned into a small, horizontally-orientatedperfusion chamber. The in vitro heart was subjected to a constant perfusion of saline(100 ml h"1) and many samples (> 50) could be sequentially bioassayed on a singleheart preparation.

The beat frequency of the isolated heart was monitored using a small isotonic forcetransducer (B ionix Corp., U.S.A.). The electrical signal generated by the transducerwas amplified and fed into a frequency converter, which converted the intervalbetween successive contractions into instantaneous frequency.

CAP bioactivity was usually expressed by taking the percentage increase in heartrate and converting to ANC units. One ANC unit is defined as that amount ofcardioactivity present in a single, pharate adult abdominal nerve cord (Tublitz &Truman, 1985a). When very small amounts of CAP activity were bioassayed, the datawere best expressed in raw form, i.e. as percentage increase in heart rate.

Ganglion microdissection

Individual unfused, abdominal ganglia from pharate adult males were frozen andlongitudinally sectioned with a razor blade chip into three strips of approximatelyequal width (Fig. 1A). For each datum point, pieces from the medial and lateralportions of six abdominal ganglia were pooled, heat-treated at 80 °C for 5 min, andcentrifuged in a Beckmann microfuge for 3 min. The supernatant was drawn off andbioassayed for CAP activity on the isolated.Afana'Mca heart.

Tissue preparation

Tissues were prepared according to the procedure described by Tublitz & Truman(1985a). In short, tissues were homogenized in 0-1 moll"1 acetic acid after heattreatment for 5 min at 80 °C. The homogenate was centrifuged and the supernatantdrawn off for subsequent use.

Surgery

For the extirpation experiments, the fifth abdominal ganglion was removed fromday 1 male pupae. Following anaesthetization with CO2, a small, rectangular windowof cuticle was removed from the ventral portion of the fifth abdominal segment. Allperipheral nerves and connectives leading from ganglion A5 were transected closeenough to the ganglion so as to leave intact the anterior and posterior transversenerves, which contain the perivisceral organs (PVOs). After the tracheal supply hadbeen cut, the ganglion was removed. A small amount of crystalline phenylthioureawas applied to the opening to inhibit tyrosinase activity (Williams, 1959) and thewound was sealed with the excised piece of cuticle using melted beeswax.

398 N. J. TUBLITZ AND J. W. TRUMAN

Operated animals were allowed to recover and proceeded normally through adultdevelopment. Seventeen days later at the pharate adult stage, these animals were againanaesthetized to facilitate the removal of the PVOs anterior and posterior to theextirpated ganglion. The respective PVOs were pooled from five animals for eachassay point and assessed for CAP activity on the isolated Manduca heart bioassay.

Sham operations were conducted in an identical manner to the experimentaloperations, except that the ventral nerve cord remained intact.

Extracellular recording and stimulation

A semi-intact abdominal preparation was employed for these experiments.Individual pharate adult males were chilled on ice for 30min prior to dissection.Following isolation from the rest of the body, each abdomen was opened with a mid-dorsal incision and pinned over a hole in a waxed recording chamber. The preparationwas continuously aerated through this hole, which supplied air to the spiracles. Thecuticular side of the preparation was kept dry by a liberal application of vacuum grease(Dow Corning, U.S.A.) to the pinned margin of the abdomen. The abdominalportion of the ventral nerve cord was exposed by removal of the gut, reproductive tractand some fat body. The tracheal supply to the ventral nerve cord, the central con-nectives and peripheral nerves were all left intact throughout the experiments exceptwhere noted. A standard lepidopteran saline was utilized for all experiments (Tublitz& Truman, 19856).

For those experiments requiring extracellular recordings, nerve activity wasmonitored by standard glass suction electrodes. Recordings were collected en passanton intact nerves and data were stored on tape.

CAP activity released from the PVO was collected by means of a Vaseline well(volume = approx. 0-1 ml) erected around the transverse nerve at a point immediatelydistal to the transverse nerve-ventral nerve anastomosis. In the case of extracellularstimulation experiments, this collection site was always contralateral to the point ofelectrical stimulation. Current for electrical stimulation was applied through glasssuction electrodes at a frequency of 0-1—1-0 Hz for 5 min. Immediately after stimu-lation, the saline drop within the Vaseline well was removed and frozen on dry ice forlater assay of CAP activity on the in vitro Manduca heart.

Intracellular recording and stimulation

For the intracellular recording and stimulation experiments, abdomens frompharate adult males were isolated, dissected and placed in the recording chamber asdescribed above. The connectives were transected anterior to the ganglion, and a wax-covered stainless steel platform placed underneath ganglion A4 or A5 to stabilize it forintracellular recordings. Glass microelectrodes, filled with either 2 mol I"1 potassiumacetate or 3 mol 1~ potassium chloride and having resistances in the range of 25 to50 MQ, were used in these experiments. Intracellular stimulation was accomplishedby passing current into cells using a bridge circuit built into a high impedanceamplifier (Getting, Model 5). Direct current pulses were applied at frequencies nogreater than 1 Hz for up to 15 min.


A Vaseline well was placed around a portion of the PVO of these preparations asdescribed above. The contents of the well were collected at various times, frozen ondry ice and stored at — 20°C, usually for less than 24 h. Each sample was thawed justprior to assay. All of the samples from a given experiment were sequentially tested onthe same heart bioassay preparation, that had been pre-calibrated with 5-hydroxy-tryptamine (5-HT). Samples were bioassayed only when there was no more than 1 %variability in the frequency of contraction of the isolated heart over a 10-min period.

RESULTS

Ganglionic localization of CAP activity

Previous experiments involving cobalt backfilling of nerve routes leading to thePVO identified nine pairs of neurones whose axons project to this neurohaemal site(Taghert & Truman, 19826). As a first step in discerning which of these neuronesproduce the CAPs, individual pharate adult abdominal ganglia were microdissectedinto medial and lateral strips, pooled and bioassayed for CAP activity as described inMaterials and Methods. This procedure divided the cells projecting to the PVO intotwo groups: each lateral strip contained the cell bodies of the four bursicon neurones(Taghert & Truman, 1982a), whereas the medial portion retained the somata of all ofthe remaining transverse nerve projecting cells (Fig. 1A). The medial section of theganglion contained at least an order of magnitude more cardioacceleratory activitythan that present in the lateral portion (medial = 7-75 % vs lateral = 2-50 % increasein heart rate; Fig. IB), taking into account the logarithmic nature of the CAP dose-response curve (Tublitz & Truman, 1985a). These results implied that some or all ofthe medial cells might contain the CAPs.

Accumulation of CAP activity during adult development

The results of the ganglion microdissection experiments suggested that the CAP-containing cells might be located along the midline of the segmental ganglia. Fivepairs of neurosecretory cells whose axons project to the PVO reside in this area. Of thisgroup, four pairs have a branched axon that projects bilaterally out of the ventralnerve, and will subsequently be referred to as the Midline Bilateral (MB) neurones.Three out of four pairs of MB neurones were originally observed in the pupal stage byTaylor & Truman (1974). Since these cells differentiate at the onset of metamorphosisand attain their mature size by the pharate adult stage (Taghert, 1981), they willsubsequently be called the new MB neurones. The appearance of the three pairs ofnew MB neurones in the adult as compared to larvae was of interest because CAPactivity has not been found in prepupal nerve cords (Tublitz & Truman, 1985a;Tublitz, 1983).

The relationship between the maturation of the new MB cells and the appearance ofCAP activity was first examined by measuring the levels of CAP in the abdominalCNS through development. At selected times, abdominal nerve cords (ANCs) wereremoved from 20 animals, pooled, and treated as described in Materials and Methods.The resultant supernatant was loaded onto a C18 Sep-pak (Waters, U.S.A.), followed

400 N. J. TUBLITZ AND J. W. TRUMANby sequential washes using 20% and 80% acetonitrile solutions, respectively.Cardioacceleratory activity in each wash was monitored using the isolated Manducaheart bioassay. This procedure effectively separated the biogenic amines from the twoCAPs, the former recovered in the 20% acetonitrile fraction while the 80%acetonitrile fraction contained all of the CAP activity.

Fig. 2 shows the appearance of CAP activity in the abdominal CNS throughoutadult development. Negligible during the first 2 days after pupation, CAP activity wasfirst detected on day 4 of adult development. Activity increased rapidly in theabdominal nerve cord during the next few days, reaching a plateau at day 8 followingpupal ecdysis. This CAP level was subsequently maintained for the duration of adultdevelopment up to and including the day of adult eclosion.

Medial

Lateral

10 i -

1sS 5o

0

-!£•£•:?

•:•:•:•:•:[> T

Medial Lateral

Fig. 1. CAP activity in areas of the pharate adult abdominal ganglion. (A) Diagrammatic repre-sentation of an abdominal ganglion showing the position of cells which project to the PVO. Cellswhose axons leave the ventral nerve and project to the posterior PVO, filled and open circles; the newMB cells which arise post-embryonically during adult development, open circles; cells whose axons goto the anterior PVO, filled triangles; cells whose axons project from the connectives to the nextposterior ganglion, leave the dorsal nerve, and terminate in the posterior PVO, filled squares. Theletters a, d, v andp refer to the anterior, dorsal, ventral and posterior pair of MB cells, respectively.(B) Results from ganglion microdissection experiments. For each determination, tissues from fivepharate adult ganglia were pooled and bioassayed for CAP activity. Each histogram represents themean ± S.E.M. of six separate determinations.


The kinetics of CAP accumulation during adult development was paralleled by thegrowth of the new MB neurones. Sectioned material for the analysis of the growth ofthe new MB cells was prepared by Dr S. Reiss. Individual abdominal ganglia weredissected from staged animals at various times after pupation, fixed, seriallysectioned, and stained with haemotoxylin and eosin. The new MBs were individuallyidentified in the serial sections by soma position within the ganglia, and the diametersof their cell bodies and nuclei measured. Each pair of the new MB neurones remainedundifferentiated until midway through the fifth larval instar (P. H. Taghert & J. W.Truman, unpublished observations). Shortly after the prepupal peak of ecdysone(Wielgus, Bollenbacher & Gilbert, 1979) the new MB cells began to enlarge (Fig. 3).This somatic growth continued through the prepupal period and into the first week ofadult development, increasing from a larval diameter of 5 /im to 17 fim at day 5 of adultdevelopment. Growth of the new MB neurones was completed by day 8, with a finalsoma diameter of 22-24/xm.

1-00

.2 0-75c

<

ft.

u"Sc3O

0-50

0-25

4 6 8 10 12

Days of adult development

14 16 18

Fig. 2. The accumulation of CAP in the abdominal nerve cord during adult development. One ANCunit is equivalent to the amount of CAP activity present in a single abdominal nerve cord in the pharateadult Manduca sexta. Each point represents the mean ± s.E.M. from 10 determinations.


Extirpation experiments

Additional evidence to support the hypothesis that the new MB neurones containthe CAPs was provided by experiments in which CAP accumulation in the PVOduring adult development was prevented by surgical intervention. A single abdominalganglion was removed from animals early in adult development and the CAP activityin the PVO anterior and posterior to the extirpated ganglion was subsequentlymeasured after the insect had reached the pharate adult stage. There was a markedreduction in the amount of cardioacceleratory activity in the PVO posterior to themissing ganglion, whereas normal levels were seen in the anterior PVO (Fig. 4).There was no significant difference in CAP levels between the two PVOs in sham-operated animals (data not shown). These data suggest that the CAPs reach the PVOfrom the next anterior ganglion, which is consistent with the known projection patternof axons from the new MB cells (Taghert, 1981).

Release of CAP activity by extracellular stimulation of the transverse nerve

The MB cells are unique among abdominal ganglion neurones leaving the ventralnerve in that each MB neurone has a branched axon that leaves both left and rightnerves. This unique anatomical arrangement was exploited by stimulating the ipsi-

30

25

20

03

510

Soma

Nucleus

I I I I I I 1 I I I I I I I I I I I I I IVth EG EG P P P P P P

+ 1 +3 +2 +5 +8 +11 +14Days of development

I I I

Fig. 3. The somatic and nuclear growth of the new MB cells during metamorphosis. Open squares,diameter of soma; filled circles, diameter of nucleus. Each point represents the mean ± S.D. from atleast five separate MB cells. Vth, fifth instar larva; EG, wandering larva; P, pupa; A, adult.


lateral transverse nerve, and measuring the amount of CAP activity released from thecontralateral transverse nerve (Fig. 5). This treatment induced the release ofdetectable amounts of CAP activity in the presence or absence of intact connectionsbetween the two PVOs via the medial nerve. Under these stimulation conditions,approximately 5 % of the CAP stored in the PVOs was released into the collection siteduring each stimulation experiment. Transverse nerve stimulation did not induceCAP release if the branch connecting the transverse nerve with the ventral nerve wassevered. Extracellular stimulation of the dorsal nerve or the ventral nerve at pointsdistal to the transverse nerve branch also did not evoke the release of detectable CAPbioactivity.

Release of CAP activity by intracellular stimulation of single new MB neurones

An individual new MB cell from either the 'dorsal' or 'anterior' pairs (refer to Fig. 1and see Discussion for details of the different new MB cells) was impaled with a glassmicroelectrode. Although the neurones investigated in this study were not visiblein situ during any developmental stage including the pharate adult, unequivocal

30 I—

20

•2

10

TN4

TN5

TN4 TN5

Fig. 4. The effect of ganglion extirpation on CAP accumulation in the perivisceral organs duringadult development. TN4 and TN5 were removed at the pharate adult stage and separately assayed forCAP activity. Each histogram represents the mean ± S.E.M. from 10 animals, each assayed indi-vidually. TN, transverse nerve containing the perivisceral organs.

404 N. J. TUBLTTZ AND J. W. TRUMAN

identification of these cells during intracellular penetrations was made possible due tothe ganglionic position of their cell bodies and their electrophysiological properties. Inthe pharate adult the somata of the new MB cells lie along the midline in the anteriorhalf of each abdominal ganglion. This region of the ganglion is almost devoid of cellbodies, containing the somata of only 12 neurones, the majority of which are identifiedmotoneurones. The new MB neurones are the only cells in this region whose axonsproject from the ventral nerve to the transverse nerve (Taghert, 1981). In addition,the somata of these cells showed large, overshooting action potentials with 40-50 msdurations, a characteristic of many insect neurosecretory cells (Miyazaki, 1980; Tag-hert, 1981). This property clearly distinguished them from the surrounding moto-neurones and interneurones.

Sequential saline drops were collected from the Vaseline well surrounding a smalllength of the posterior PVO. Saline was collected prior to intracellular stimulation,immediately after stimulation, and after a 60-s washout period, and assayed for CAP

0

% R

ate

incr

ease

2

n

-

-

T•:::x'*x*x

•&:$:-i§&

T

8|$S§x-:-

TN

:'.\nIntact

(6)

TN-MNcut

(3)

TN-VNcut

(6)

DN stim

(3)

VNstim

(3)

Fig. 5. Extracellular stimulation evokes CAP release. Inset: schematic diagram of the preparation.The collection site is symbolized by the dotted square contralateral to the point of stimulation. Eachhistogram represents the amount of CAP activity collected under various stimulation regimes (mean± S.D.). Extracellular stimulation was via the ipsilateral transverse nerve (TN), except where noted.Intact, all peripheral nerves were intact; TN-MN cut, the junction between the ipsilateral andcontralateral TNs was severed; TN-VN cut, the nerve branch between the VN and TN was tran-sected; DN stim, stimulation of a dorsal nerve (DN); VN stim, stimulation of a ventral nerve (VN)distal to the VN-TN anastomosis; AG, abdominal ganglion; MN, median nerve; PVO, perivisceralorgan.


bioactivity (e.g. Fig. 6). Only the sample obtained immediately after intracellularstimulation contained CAP-like bioactivity, producing 5 % increase in heart rate whenapplied to the isolated heart bioassay. The other samples, collected prior to stimu-lation and after a saline rinse, were devoid of cardioacceleratory activity. In five out ofsix cells, intracellular stimulation resulted in the release of measurable amounts ofCAP activity (Table 1). The amount of CAP released during intracellular stimulationof a new MB cell was less than 1 % of the total CAP stored in the PVO.

Table 1. Intracellular stimulation of a new MB neurone evokes peptide releaseExpt no. Pre Stimulation Post

123456Mean±s.E.M.

All figures are expressed t

1%2%0%0%0%0%

0-50 ±0-34%

is percentage increase in

4%5%5%1%4%5%

4-00 ±0-63%

—0%1%0%0%0%

0-20 ±0-08%

heart rate of the isolated Manduca heart bioassay. Pre,Stimulation and Post refer to the time the sample was collected relative to the application of intracellular currentto the new MB neurone.

DN

Bath

Prev

Stimv

Postv

-,50

-U5 beats min '

-140

1 min

Fig. 6. Intracellular stimulation of a single new MB neurone evokes the release of CAP activity. Toppanel: schematic diagram of a pharate adult abdominal ganglion showing the position of the cell bodyand major processes of a new MB neurone. Bath, the collection site on the transverse nerve (TN).Bottom panel: the results from a single experiment. Each sample was sequentially bioassayed on the invitro Manduca heart. Samples were applied when the basal variability of the heart beat frequencyremained at or below 1 % for a 10-min period. Pre, sample collected prior to intracellular stimulation;Stim, sample collected immediately after stimulation; Post, sample collected after a 60-s washoutperiod. DN, dorsal nerve; VN, ventral nerve; MN, median nerve; PVO, perivisceral organ.


To determine whether the cardioacceleratory activity released by intracellularstimulation was a peptide(s), the saline drop collected after stimulation of a new MBcell was treated with a protease. The 100 /il drop was divided into two, approximatelyequal, 50/il samples, one of which was incubated with Pronase (0-5 mgrnl"1). Bothsamples were maintained at 30 °C for 3 h, followed by boiling at 100°C for lOmin toinactivate the protease. The activity in pronase-treated sample was completely des-troyed, while the control sample contained detectable amounts of CAP-likecardioacceleratory activity (Fig. 7).

To resolve whether the CAP activity was actually released by the neurone that wasstimulated rather than by a cell that was postsynaptic to the stimulated cell, extra-cellular recordings were obtained from both ipsilateral and contralateral ventralnerves during intracellular stimulation of a new MB neurone. Stimulation of the newMB cell resulted in the appearance of only one unit in each ventral nerve (Fig. 8). Theunits were temporally phase locked with each other and with the spike recorded in thesoma of the new MB cells (Fig. 8). These extracellular units never appeared spon-taneously in the absence of a depolarizing current pulse to the new MB cell body.Characteristic of insect neurosecretory cells, new MB somata failed to produce actionpotentials when stimulated at frequencies greater than 2-5 Hz (Taghert, 1981;Miyazaki, 1980). Extracellular activity in the ventral roots was eliminated wheneverfrequency-dependent spike failure occurred in the new MB soma.

DISCUSSION

The primary goal of this study was to identify the peptidergic neurones in theabdominal ganglion in Manduca that were responsible for the synthesis and release ofthe two CAPs. Previous work (Tublitz & Truman, 19850,6) showed that the CAPs

Sample+

Pronase Sample

• •

W^\65

beats min~l

60

10s

Fig. 7. The effects of Pronase upon CAP activity collected as a result of intracellular stimulation. Halfof the bath sample collected after intracellular stimulation was incubated in Pronase (0-5 mgml"1).Both the Pronase and control samples were incubated at 30 °C for 3h. Following the incubationperiod, each sample was heat treated at 80°C for 5 min prior to assaying for CAP activity.


were primarily localized to and released from the abdominal perivisceral organs(PVOs), the major neurohaemal organs of the insect ventral nerve cord. Usinghistological and ultrastructural techniques, Taghert (1981) identified nine pairsof neurosecretory neurones that project to the abdominal PVOs in pharate adultManduca. All but one pair have axons that terminate in the posterior transversenerve. Of these, four pairs have cell bodies that are clustered along the lateral marginof the ganglion and have been identified as the bursicon-containing neurones (Taghert& Truman, 19826). The somata of the remaining four pairs, the Midline Bilateral(MB) neurones, lie along the midline (Taylor & Truman, 1974; Taghert, 1981). Alleight MB neurones and six of the eight lateral cells send their axons out of the ventralnerve prior to their termination in the posterior transverse nerve.

Of the eight MB cells, three pairs differentiate post-embryonically during meta-morphosis (Taylor & Truman, 1974; Taghert, 1981) and have been called the newMB neurones (Fig. 1A). Several lines of evidence presented in this study suggestedthat the new MB cells were the CAP-containing neurones. The ganglion micro-dissection experiments showed that the majority of ganglionic CAP activity resided in

(5)

VN(5,L)

VN(5,R)

500 ms

Fig. 8. Intracellular stimulation of a CAP neurone showing the presence of a single, phase-locked,extracellular unit in each of the ventral roots. Top trace: intracellular recording from a CAP neuronein the pharate adult. Middle and bottom traces: extracellular recordings from the paired ventralnerves (VN). R, right; L, left; 5, fifth abdominal ganglion.


the medial portion of the ganglion, the location of the new MB somata (Fig. IB). Theresults of the extracellular stimulation experiments, in which CAP release was evokedby nerve stimulation only when the transverse nerve-ventral nerve branch remainedintact, were consistent with the axonal morphologies of the new MB cells (Fig. 5).These data support the hypothesis that the new MB cells contain and release the twoCAPs.

Direct evidence that the new MB cells contain the CAPs was provided by experi-ments in which intracellular stimulation evoked the release of CAP bioactivity. Thematerial collected from the PVO immediately after stimulation was biologically activeon the isolated heart bioassay (Fig. 6; Table 1) and this bioactivity was destroyedwhen incubated with a protease (Fig. 7). Samples taken prior to stimulation or after awashout period were devoid of CAP bioactivity, indicating that the release of thismaterial was activity dependent. Separate experiments suggested that intracellularstimulation of a new MB cell probably did not result in the driving of any other unitsthat project to the release site (Fig. 8). We therefore conclude that the CAPs werereleased from the new MB cell that was stimulated.

Several diverse techniques have generally been employed in the cellular localizationof neuropeptides. Immunohistochemical techniques, although successful in a varietyof vertebrate and invertebrate preparations (Brownstein, 1980; Duve & Thorpe,1981; Bishop & O'Shea, 1982), are replete with interpretational difficulties. Some ofthese problems have been overcome in those systems in which single neurones aredissected free from adjacent cells and individually assayed for peptidergic activity.This single cell assay technique enabled the identification of the cells that produce anumber of peptides including the small cardioactive peptides in gastropod molluscs(Lloyd, 1978), diuretic hormone (Berlind & Maddrell, 1979), proctolin (Adams& O'Shea, 1983), prothoracicotropic hormone (Agui, Granger, Gilbert & Bollen-bacher, 1979) and bursicon (Taghert & Truman, 19826) in insects. To our know-ledge, however, this is the first identification of a specific peptidergic neurone basedupon evoked release of that peptide by intracellular stimulation.

One unexpected finding was that the onset of differentiation of the new MBs and thekinetics of their subsequent somatic growth during adult development (Fig. 3) wasmirrored by the post-pupation accumulation of CAP activity in the ventral nerve cord(Fig. 2). In most neuronal systems investigated to date, the initial detection of aneurosecretory product, either neurotransmitter or neurohormone, usually occursafter the cell has passed through the final stages of differentiation and somatic growthhas terminated (Goodman & Spitzer, 1978; Potter, Landis & Furshpan, 1980). Thereis at least one instance where a neurosecretory product can be detected in cells thathave not finished differentiating. InAplysia the neurohormone egg-laying hormone isfound in the ELH-containing bag cells while they are still in migration from the bodywall to their final destination (Scheller et al. 1983). Although these results might notbe typical, they do suggest that certain cells have the ability to synthesize andaccumulate neurosecretory material prior to the termination of differentiation.

Taken as a whole, these experiments substantiate the hypothesis that the cells in thenew MB neurone group contain and release CAP bioactivity, but it has not been


rigorously demonstrated that all the new MB cells contain CAP bioactivity. Asmentioned previously, there are three pairs of new MB cells, the so-called 'anterior','dorsal' and 'ventral' pairs (Fig. 1A). All intracellular penetrations attempted in thisstudy focused only upon the 'anterior' or 'dorsal' pairs without distinguishing betweenthem, and no attempt was made directly to stimulate the ventral pair of cells. A fourthMB pair, the 'posterior' cells, was also not investigated. This pair shares similar axonaltrajectories to the new MB cells (i.e. bilateral axons exiting both ventral nerves andterminating in both branches of the transverse nerve), but differs from the new MBsin that it arises during embryogenesis and persists throughout all life stages. It is quitepossible that this pair of neurones also release the cardioactive peptides. However, ifthis were so, the cells presumably release some other product in the larva since theCAPs are absent from this stage (Tublitz & Truman, 1985a). It is also unclearwhether a single new MB cell contains and releases one or both of the CAPs. Futurestudies on the CAP system should investigate these issues.

We wish to thank Dr Janis Weeks for advice and support during the course of theseexperiments. Drs L. M. Riddiford and A.O.D. Willows critically reviewed earlierdrafts of this manuscript. This research was supported by a NSF PredoctoralFellowship and by a NIH training grant.

R E F E R E N C E S

ADAMS, M. E. & O'SHEA, M. (1983). Peptide co-transmitter at a neuromusecular junction. Science, N.Y. 221,286-289.

AGUI, N., GRANGER, N. A., GILBERT, L. I. & BOLLENBACHER, W. E. (1979). Cellular localization of the insectprothoracicotropic hormone: in vitro assay of a single neurosecretory cell. Proc. natn. Acad. Sci. U.SA. 76,5964-5968.

BERUND, A. & MADDRELL, S. H. P. (1979). Changes in hormone activity of single neurosecretory cell bodiesduring a physiological secretion. Brain Res. 161, 459-467.

BISHOP, C. & O'SHEA, M. (1982). Neuropeptide proctolin: immunocytochemical mapping of neurons in theCNS of the cockroach. J. comp. Neurvl. 207, 223-238.

BROWNSTETN, M. J. (1980). Peptidergic pathways in the CNS. Proc. R. Soc. B 210, 79-90.DUVE, H. & THORPE, A. (1981). Gastrin/CCK-like immunoreactive neurons in the brain of the blowfly,

Calliphora erythrocephala. Gen. comp. Endocrinol. 43, 381—391.GAINER, H. &FRONTALI, N. (1979). Peptides in invertebrate nervous systems. In Peptides in Neumbiology, (ed.

H. Gainer), pp. 259-294. New York: Plenum Press.GOODMAN, C. S. & SPITZER, N. C. (1978). Embryonic differentiation of identified neurones: differentiation

from neuroblast to neurone. Nature, Land. 280, 208—213.LLOYD, P. E. (1978). Neurohormonal control of cardiac activity in the snail,Helix aspersa.jf. comp. Physiol. 128,

277-283.MADDRELL, S. H. P. (1974). Neurosecretion. In Insect Neumbiology, (ed. J. E. Treherne), pp. 305-307.

Amsterdam: North Holland.MILLER, T. A. (1980). Neurohormonal Techniques in Insects. New York: Springer-Verlag.MIYAZAH, S. (1980). The ionic mechanism of action potentials in neurosecretory and non-neurosecretory cells

of the silkworm. Bombyx mori.J. comp. Physiol. 140, 43-52.POTTER, D. D., LANDIS, S. C. &FURSHPAN, E. J. (1980). Dual function during development of rat sympathetic

neurones in culture. J. exp. Biol. 89, 57-71.RAABE, M. (1982). Insect Neurohormones. New York: Plenum Press.RAABE, M., CAZAL, C , CHALAYE, D. &DEBESSE, N. (1966). Actioncardioacceleratricedesorgansneurohemaux

perisympathetique ventraux des quelques insectes. CJi. Acad. Sci. Paris 263, 2002—2005.SCHELLER, R. H., JACKSON, J. F., MCALLISTER, L. B., ROTHMAN, B. S., MAYERI, E. & AXEL, R. (1983). A

single gene encodes multiple neuropeptides mediating a stereotyped behaviour. Cell 32, 7-22.


TAGHERT, P. H. (1981). The identification and analysis of specific peptide-containing neurones in the moth,Manduca sexta. Ph.D. dissertation, University of Washington, Seattle.

TAGHERT, P. H. & TRUMAN, J. W. (1982a). The distribution and molecular characteristics of the tanninghormone bursicon in the tobacco hornworm, Manduca sexta. J. exp. Biol. 98, 373—383.

TAGHEBT, P. H. & TRUMAN, J. W. (1982&). Identification of the bursicon containing neurones in abdominalganglion of the tobacco hornworm, Manduca sexta. J. comp. Physiol. 90, 367-388.

TRUMAN, J. W. & TAGHERT, P. H. (1983). Neuropeptides in the insects. In: Brain Peptides, (edsD. Krieger, M.Brownstein & J. Martin), pp. 165-182. New York: Wiley & Sons.

TUBLITZ, N. J. (1983). The biochemical and physiological analyses of two cardioregulatory neuropeptides in thetobacco hawkmoth, Manduca sex'ta. Ph.D. dissertation, University of Washington, Seattle.

TUBLITZ, N. J. & TRUMAN, J. W. (1985a). Insect cardioactive peptides. I. Distribution and molecularcharacteristics of two cardioacceleratory peptides in the tobacco hawkmoth, Manduca sexta. J. exp. Biol. 114,365-379.

TUBLTTZ, N. J. & TRUMAN, J. W. (19856). Insect cardioactive peptides. II . Neurohormonal control of heartactivity by two cardioacceleratory peptides in the tobacco hawkmoth, Manduca sexta. J. exp. Biol. 114,381-395.

WlELGUS, J. J., BOLLENBACHER, W. E. & GILBERT, L. I. (1979). Correlations beteen epidermal DNA synthesisand haemolymph ecdysteroid titre during the last larval instar of the tobacco hornworm, Manduca sexta.J. Insect Physiol. 25, 9-16.

WILLIAMS, C. M. (1959). The juvenile hormone. I. Endocrine activity of the corpora allata of the adult cecropiasilkworm. Biol. Bull. mar. biol. Lab., Woods Hole 116, 323-328.

IDENTIFICATION OF NEURONES CONTAINING CARDIO- … · manner immediately after adult eclosion, producing a marked rise in heartbeat frequency and a facilitation of wing inflation (Tublitz

Documents