Studies on the initiation of growth and moulting in Locusta migratoria migratorioides R. & F.—IV. The relationship between the stomatogastric nervous system and neurosecretion

J. Ins. Phvsiol., 1963, Vol. 9, pp. 423 to 430. Pngamon Press Ltd.

Printed

in

Great Britain

STUDIES ON THE INITIATION OF GROWTH AND MOULTING IN LOCUSTA MIGRATORIA MIGRATORIOIDES R. & F.-IV. THE RELATIONSHIP BETWEEN THE STOMATOGASTRIC NERVOUS SYSTEM AND NEUROSECRETION KENNETH

U. CLARKE and PETER A. LANGLEY

Department of Zoology, University (Reckved

15 November

Nottingham

1962 ; revised 2 February

1963)

Abstract-Studies were made on the neurosecretory system in normal locusts from the end of the second instar to the middle of the fifth instar. Samples were taken at 24-hr intervals, fixed in Bouin’s fluid, and the whole heads sectioned and stained with Gomori’s chrome haematoxylin and phloxine. Special attention was paid to the medial neurosecretory cells, their chiasma, the nervi corporis cardiaci I, and the corpora cardiacum. Histologically no trace of a secretory cycle of any kind could be found; the picture was one of constant production and secretion of neurosecretory material. In locusts from which the frontal ganglion had been removed there was little difkence from the normal animal during the first 200 hr following the opemtion. After this time there was considerable accumulation of material in the nervi corpork cardiaci I, and the corpora cardiaca was shrunken and abnormal in appearance; the cells of the pars intercerebralis differed little from normal. In starved animals after 61.hr the corpora cardiacs was also shrunken and abnormal in appearance. There was, however, no material in the nervi wrporis cardiaci I. Based on evidence presented in this and the previous papers, and on facts published by other workers, the following hypothesis of the control of growth and moulting in locusts is proposed. There is a constant production of neurosecretory material in the medial neurosecretory cells and constant release of it from the corpora cardiaca. Its production and/or release are dependent upon impulses arising in the stretch receptors of the pharynx which are conducted via the posterior pharyngeal nerves, frontal ganglion, and frontal connectives to the brain and via the recurrent nerve and hypocerebral ganglion to the corpora cardiaca. Surgical interference in this pathway prevents the secretion of material and leads to the failure of the animal to grow and moult. During an instar the movements of the pharynx associated with the intake of food provide the necessary stimuli, whereas at an ecdysis the swallowing of air into the gut which continues throughout ecdysis will provide the necessary stimuli at this time. This can give rise to a cyclical stimulation of the prothoracic glands in the following mer. During an instar the neurosecretory material is used in the ordinary processes of growth and metabolism; at an ecdyais, when food absorption is absent, these processes will be less. As the material is still being produced, there wuld be a rim in titre of the hormone in the blood which could result in the stimulation of the prothoracic gland, and consequent upon this the start of a moulting cycle. 423

424

KENNETH

u.

CkARUE AND

PETJZR

A.

LANGLEY

INTRODUCTION

previous publications (CLARKEand LANGLM, 1963b, c) it has been shown that the severing of the frontal connectives, the removal of the frontal ganglion, and the separation of the frontal ganglion from the surface of the oesophagus resulted in the immediate cessation of growth and moulting in Locu.rta migrutoriu L. THOIVBEN and MBLLER(1959) and STRANGEWAYS-DIXON (1961) have provided strong evidence for the endocrine control of protein metabolism in Calliphora erythrocephala Meig. THOMSENand MBLLERhave quite specifically implicated the medial neurosecretory cells of the pars intercerebralis in the control of protein digestion. It was therefore suggested that these ‘growth arresting’ operations mediated their effect via the neurosecretory system of the animal on protein metabolism. If this is so, then changes in the neurosecretory system might be expected to follow surgical interference with the relevant parts of the stomatogastric system. The present paper establishes the normal activity of the neuroendocrine system throughout the development of the animal and the changes that result from ‘growth arresting’ operations. IN

MATERIAL

AND TECHNIQUES

Insects were reared under crowded conditions at 28 + 05°C and 70 per cent + 5 per cent relative humidity. The techniques described in our previous paper (CLAFUCE and LANGLEY,1963b) were used in the ‘growth arresting’ operations. BRAIN Modion

’

Differontiol

orotocrtbrol

/

\

‘*l*\niforrn

Nerve tracts protocorrbral Chiasma

’

staining sying\ A cells (Qrk staining)

vis?‘e \ Stained. Unstoinrd

’ 9 Cllll (phloxinophil staining)

A mat&

CORPORA Anterior

/

B\motwiol

CARDIAC

I

lobe

Posterior rxiensron of anterior lobe

\

Malrriol scoitered

from cells \ Chiosmo not visible

Nwrosccretory molrrial present (A or B motewall

Postrrior

lobe I

N~uros~cretory motrriol absent \ M&OJiOl

/ \ L&al conc*ntrotion Unitor& donsit throughout e. l! (poriphory of gland closr 10 aorta Wall)

FIG. 1. Criteria used for the estimation of neurosecretory activity. More than 200 insects were used in the preparation of serial sections to study the histological appearance of the pars intercerebralis of the protocerebral region of the brain, the nervi corporis cardiaci I, and the corpora cardiaca. In some cases the organs were dissected from material frxed in Bouin’s fluid before embedding.

GROWTH AND MOULTING

IN LOCUST.4 MIGRATORlkt

M~tXX.4TO~CWX%’

R. & F.-W

425

However, the most valuable i~o~a~on was obtained by sectioning the whole head of each insect, thus leaving the neuroendocrine system intact. To facilitate this, the heads were embedded in ester wax according to the technique developed by STEEDMAN (1947), and sections were cut at 10 p thickness. All sections were stained with Gomori’s chrome haematoxylin and phloxine (C.H.P.) according to the technique used by BARGMANN (1949) and DAWSON (1953). HIGHNAM(1961) has described the anatomy and histology of the neurosecretory system in Schistocercagregaria Forsk., in relation to the maturation of adult females. The system in Locusta mipatoria m&atorioides R. & F. has proved to be almost identical to that of Sctitocerca, and Fig. 1 shows the criteria which have been used in the present investigation for the assessment of neurosecretory activity. The lzeuIocndoctitK system of achy

d~e~~.~

k.zrvae

A study of serial sections of whole heads of insects sampled at intervals of 24 hr from the end of the second instar until the middle of the fifth instar revealed no histological signs of a cycle of secretion which could be correlated with the progress of the growth and moulting cycles. The secretory cells of the pars intercerebralis presented a constant picture. There were always numerous A cells present whose cytoplasm stained pale blue. In most of the sections, some of these cells stained blue-black and were full of neurosecretory material (C.H. positive). In other A cells, scattered granules of material could be seen in the cytoplasm and in the tracts of the nervi corporis cardiaci I but always in small amounts (Fig. 2(A)), At the point where the nervi corporis cardiaci I emerged from the brain to enter the corpora cardiaca, neurosecretory material was undetectable in the nerves except in a few instances where it was present in very minute quantities (Fig. 2(B)). The corpora cardiaca also presented a constant picture. There was always a large amount of neurosecretory material in the anterior lobes concentrated in the area of the glands closest to the aorta wall. Material was also present in the base of the glands but not in such high concentration (Fig. 2(C)). When the tracts of the nervi corporis cardiaci I and II were visible within the corpora cardiaca they contained little or no stainable material (Fig. 2(D)). The nervi corporis cardiaci II were not observed either inside or outside the brain in these sections, since their course from the lateral groups of neurosecretory cells in the protocerebrum to the corpora cardiaca lay in a different plane from that of the nervi corporis cardiaci I and sections were cut to include as much of the course of the latter as possible. An examination of the B cells and B material (phloxinophil) throughout the neuroendocrine system produced no evidence of cyclical secretory activity. Numerous cells were present in the brain which had phloxinophil inclusions, mainly in the nucleus. The anterior lobes of the corpora cardiaca contained numerous scattered p~o~nop~ granules in addition to the abundant A material, while the posterior lobes contained only phloxinophil B material. This material was not observed in the tracts of the nervi corporis cardiaci I, although HIGHNAM (1961)reports its presence in adult female Schisfocerca gregaria Forsk.

426

KENNETHU. CLARKEAND PETER A. LANGLEY

Serial sections through the head of a mature adult female insect demonstrated large amounts of A material in the neurosecretory cells of the pars intercerebralis and in the nervi corporis cardiaci I within the brain (Fig. 3(A)). This is in agreement with the observations made by HIGHNAM(1961) on adult female Sckistocerca gregaria Forsk. The corpora cardiaca of the adult presented the same aspect as those of the larva. The neuroendocrine system of starving insects Insects were deprived of food from the beginning of the third instar and sampled at regular intervals. Serial sections demonstrated that the appearance of the neuroendocrine system was the same as for normal insects during the first 30 hr of the instar. However, observations on insects which had survived to 61 hr showed some differences. The neurosecretory cells of the brain resembled those of the normal insect, there being a few A cells containing C.H.-positive material (Fig. 4(A)). Small amounts of C.H.-positive material were also found within the tracts of the nervi corporis cardiaci I and the corpora cardiaca were full of this dark-staining A material. However, the corpora cardiaca were obviously abnormal, being considerably shrunken and with little visible cytological detail (Fig. 4(B)). An investigation of the general tissues in the head at this time revealed a total absence of fat body and a reduction in the size of muscle bands. The air sacs were inflated and filled the spaces in the head cavity caused by the reduction in tissue content. This can also be seen in Fig. 4(B). The nemoendocrine system of insects deprived of tkeir frontal ganglia The insects used for this investigation were sampled from two groups. Group (a) had their frontal ganglia removed immediately after the second ecdysis before they had commenced to feed and group (b) had their frontal ganglia removed 12 hr later after feeding had been resumed. Serial sections of the heads of these insects revealed the same pattern of events in the neuroendocrine system of both groups. The neurosecretory cells of the pars intercerebralis, the nervi corporis cardiaci I, and the corpora cardiaca resembled those of normal insects for some time after the operations. However, 200 hr after the removal of the frontal ganglion the tracts of the nervi corporis cardiaci I contained large amounts of stainable A material (Fig. 3(C)). Wh ere the tracts emerged from the brain they were loaded with material and stained very darkly with chrome haematoxylin (C.H. positive) (Fig. 3(C)). At this time the corpora cardiaca were abnormal and had a similar aspect to those of starved insects (Fig. 2(D)). DISCUSSION

Signs of a secretory cycle in the neuroendocrine system have been correlated with moulting in the larva of the flour moth Anugastu kiikniella Zell. (&HM, 1951). &BAS (1957) has demonstrated a neurosecretory cycle in the cells of the pars intercerebralis of larvae of Lmusta mzjptoria L. which is correlated with moulting,

tic

A i"j

nr

B '

~

X

C

D

FIG. 2. Sections through the brain and corpora cardiaca of a normal, 9-hr-old, third instar larva. All sections stained with Gomori's chrome haematoxylin and phloxine. A, T.S. showing medial neurosecretory cells (nc) and chiasma (c). Note by comparison with Fig. 3(A) the relative absence of neurosecretory material in the me&al cells and in the chiasma. B, T.S. showing the nervi corpori cardiaci 1 (x). Note the absence of stainable material from these nerves. Compare with Fig. 3(C). C, T.S. through the anterior lobe of the corpora cardiacs. D, T.S. through the anterior lobe of the corpora cardiacs showing the tract, clear of neurosecretory material, of the nervi corpori cardiaci 1 (x) within the gland.

nC

tic

A

8

,i C FIG. 3. Sections through the brain and corpori cardiaca of Locusta. All sections stained with GomoWs chrome haematoxylin and phloxine. A, T.S. showing medial neurosecretory cells and chiasma of a normal adult immature female locust. Note the relative abundance of neurosecretory matertal compared with that found in the larva. B, C, and D all relate to a third instar larva 211 hr old from which the frontal ganglion had been removed immediately after ecdysls. B, T.S. showing the medial neurosecretory cells (nc) and the chiasma (c). T h e axons in the ehiasma are rather more darkly stained than m the normal larva (Fig. 2(A)) but still much less than in the normal adult (Fig. 3(A)). C, T . S . showing nervi corpori cardiaci 1 (x). Note the obvious abundance of material in the nerve. Compare with Fig. 2(B). D, T.S. through the anterior lobes of the corpora cardiaca. Note the obvious abnormality of the gland (compare with Fig. 2(C)) and of the surrounding tissues; food (f) is present in oesophagus.

• ,~t",

l

f'

/

X

A

J

B

FIc. 4. Sections through the brain and corpora cardiaca of a third instar larva, 61 hr old, that has been starved from immediately after the second ecdysis. A, T.S. through the brain showing the medial neurosecretory cells and the chiasrna. B, T.S. through the anterior lobes of the corpora cardiaca. For further information see text.

CRcwTH

AND ~V~W..TING IN L0cu.W~

MIGRATORIA

nIru_4ToiwoxDEs~.& F.-IV

427

although she found no signs of a cycle of secretion in the corpora cardiaca. On the other hand, HERL~T-LEWIS and PAQUET(1956) could find no correlation between the secretory activity of the neuroendocrine system and moulting in Caratius morosus Brunner. Since it is generally accepted that the stainable neurosecretory material in the insect neuroendocrine system is not a hormone itself but a high molecular lipid or lipo-protein carrier to which the active substance is bound (SCHARRERand SCHARRER,1944), the extent to which cytological changes provide an indication of hormone release may vary from one insect to another. The observations made in this investigation upon normal insects of the phase gregaria are not in agreement with the findings of C)ZBAS(1957), but since she used insects of the phase solitaria for her studies, these conflicting results may be justified. It has been a widely accepted view since endocrine studies on insects began that the presence of large amounts of neurosecretory material in the neuroendocrine system is an indication of secretory activity. When little or no material is present then the system is less active (DUPONT-RAABE, 1952; ARW and GABE, 1952; H~T-~~Is and PAQUET, 1956). Recently, however, HIGHNAM (1961) has proposed the concept that an active neurosecretory cell is one which appears to be empty or almost empty. The material is synthesized in the cell body and transported rapidly down the axon. Thus, histological examination at this time reveals very little stainable material in the neurosecretory cells or their axons. If the release mechanism for the liberation of this material at the axon endings is not operating, the material will accumulate in the axons and the neurosecretory cell bodies until they are filled. At this time the synthesis of neur OsecretoTy material by the cells will cease and the system will become inactive. His~lo~~ preparations will then demonstrate large amounts of stainable material in the neuroendocrine system. HIGHNAM (personal communication) has recognized dispersed and concentrated aggregates of neurosecretory material within the cell bodies of the neurosecretory system in relation to this concept, which he has applied to the reproductive cycle in Schistocerca gregaria Forsk. The results of the present investigation can only be explained in terms of this new concept, since the liberation of neurosecretory material from the neuroendocrine system is essential for the maintenance of the processes of development throughout the larval life of an insect. The absence of such material from the tracts of the nervi corporis cardiaci I in normal tic&a larvae must indicate a continuous synthesis and transport of this material. The constant presence of neurosecretory material in the corpora cardiaca, however, does not preclude the possibility of a periodic release of its associated hormone into the blood, though this is difficult to envisage, since variations in the amount of stainable material in the corpora cardiaca were not detected. The similarity between the corpora cardiaca of insects deprived of their frontal ganglia and those of starving insects is understandable when the respective ages of the two groups are considered. If removing the frontal ganglion has caused a cessation of protein utilization, then 200 hr after the performance of the operation

428

%NNETH

u.

&UWE

AND PJ3lXR

A. LANGLEY

the structure of the body tissues is bound to be breaking down. This results in the appearance of pathological symptoms in the organs of the body. It is significant that these symptoms occur after only 60 hr in insects which have been deliberately starved. Convincing evidence for an accumulation of material in the nervi corporis cardiaci I of insects in which both frontal connectives had been severed was not obtained. This was due to the high mortality among insects subjected to this operation, which prevented the examination of the neuroendocrine system after 128 hr of the third instar had passed. When only one frontal connective was severed, however, subsequent development of the insects concerned was unaffected and this was reflected in an entirely normal appearance of the neuroendocrine system. The same was true for insects in which only the recurrent nerve was severed. From the results of this investigation it is concluded that the cessation of growth, induced by destroying the nervous connexions between the oesophagus and the brain, is accompanied by a cessation of liberation of neurosecretory material from the neuroendocrine system, which results in an accumulation of this material in the corpora cardiaca and the axons of the nervi corporis cardiaci I. Histological studies have indicated that in a normally developing larva of htu the synthesis of neurosecretory material in the cells of the pars intercerebralis, and its liberation from the axon endings in the corpora cardiaca, is a continuous process. This poses the problem of how such a continuous process can result in a cyclical effect upon the insect as manifested by the regular cycles of growth and moulting and in the cyclical mitotic activity of the prothoracic glands which is considered to be under the control of an activating factor from the brain. Observations made on the prothoracic glands of numerous locusts have shown that the brain factor takes effect some hours before the moment of ecdysis, and it was thought that this factor was liberated in a cyclical manner during each instar at this time (CLARI(Eand LANGLEY,1963a). This hypothesis no longer seems to be tenable, since the secretory products of the pars intercerebralis have been implicated in metabolic processes which are continuous throughout an instar (reviews by VAN DER KLOOT, 1960, and SCHEER, 1960), and the present investigation has revealed no histological evidence of a cycle of secretion. A possible explanation for the mode of operation of the neuroendocrine system in relation to the growth and moulting of Locwta is offered here but must be regarded only as an hypothesis which best fits the known facts. Throughout a single instar the insect feeds, digests, absorbs, and assimilates its food and increases its body tissue content. During this time the foregut is continually being exercised with the passage of food. These distortions of the foregut stimulate the stretch receptors on the surface of the foregut (CLARKEand LANGLEY, 1963b). Nervous impulses are passed to the frontal ganglion and thence to the brain through the frontal connectives. Throughout the instar food material is being metabolized to meet the normal growth requirements of the insect. During this time the brain hormone is being fully utilized in the control of these processes.

GROWTH AND MOULTING IN LOCUSTA

MIGRATORIA

MIGRATOIUOIDES

R. & F.--IV

429

It is thus proposed that the medial neurosecretory cells of the pars intercerebralis are stimulated to synthesize, transport, and release their secretions into the blood. Towards the end of the instar it has been shown that the insect ceases to feed. It also ceases to grow. Thus, the digestion and assimilation of food materials are halted in preparation for ecdysis. However, movements of the foregut do not cease, in fact they are probably accentuated, firstly in the process of emptying the foregut of food and secondly in the swallowing of air, which is a necessary preliminary to the shedding of the cuticle. Therefore, the transmission of nervous impulses to the brain continues and the synthesis of the brain hormones is maintained, the amino acids available in the blood from previous digestion and absorption being utilized for this purpose. Since these hormones are not required for the metabo~sm of food material at this time their titre in the blood is raised to a critical point at which the prothoracic glands are triggered into activity. A similar series of events occurs at all temperatures, but at 25°C the response of the prothoracic glands to the high titre of brain hormone is slower, hence the delay in the appearance of mitotic divisions within the glands. At 28°C and above, the prothoracic glands are responding as rapidly as possible to the increased titre of hormone. After ecdysis, feeding is resumed, the hormone titre again falls as the metabdic demands of the insect dictate and protboracic gland activity subsides. The response of the epidermal cells to ecdysone, which is considered to be the hormone secreted by the activated prothoracic glands, follows in a temperature-dependent manner. A mechanism such as that suggested above would ensure that the regular cycles of growth and moulting are in phase with cycles of feeding and it is postulated that a similar mechanism exists in all continuously feeding insects whose development is not arrested by other phenomena during the course of their life cycles. It is, perhaps, sign&ant that the mechanisms known to initiate cycles of moulting in other insects may be explained in terms of this postulated control mechanism, Thus, complete starvation which initiates moulting in Tineok (TIBKXUCK, 1926) and GuZk& (MELUINKOV, 1908) may be accompanied by movements of the foregut. This would ensure the continued release of the brain hormone into the blood, where, due to the depressed metabolism of the insect, it would he under-utilized and its titre would therefore rise. When this reaches a critical level the prothoracic glands would be activated and moulting would follow. Intermittent starvation is known to delay moulting in these insects and this may be due to the partial utilization of the brain hormones for metabolism which, under these conditions, just prevents the hormone titre from reaching its critical level in the blood. In cases where tissue regeneration is known to elicit cycles of moulting, such as in Sphrodomntis (PRIZIBRAMand MEGWAR, 1912) it may be that the neuroendocrine system is stimulated directly by wounding. HODGSON and GELDIAY (1959) have shown that wounding causes a depletion of neurosecretory materials in the corpora cardiaca of a cockroach; or perhaps the increased demand upon protein reserves under such conditions stimulates the neuroendocrine system to become hyperactive.

430

KENNETH

U. CLARKE

AND PETER A.LANCLEY

REFERENCES Anvv L. and GABE M. (1952) Don&es histophysiologiques sur la neuro-s&&ion chez quelques ephemeropteres. Cellule 55, 203-212. BARGMANNW. (1949) Uber die neurosekretorische Verkniipfung von Hypothalamus und Neurohypophyse. 2. Zellforsch. 34, 610-634. CLARKE K. U. and LANGLEYP. A. (1963a) Studies on the initiation of growth and moulting in Locusta migratoria migratorioades R. & F.-I. The time and nature of the initiating stimulus. J. Ins. Physiol. 9, 287-292. CLARKE K, U. and LANGLEYP. A. (1963b) Studies on the initiation of growth and moulting in Locusta migratoria migratorioides R. & F. -11. The role of the stomatogastric nervous system. J. Ins. Physiol. 9, 363-373. CLARKE K. U. and LANGLEYP. A. (1963~) Studies on the initiation of growth and moulting in Locusta migratoria migratorioides R. & F .-I I I. The role of the frontal ganglion. J. Ins. Physiol. 9, 411-421. DALTON A. B. (1953) Evidence for the termination of neurosecretory fibres within the pars intermedia of the hypophysis of the frog, Rana pip&s. Anat. Rec. 115, 63. DUPONT-RMBB M. (1952) Contribution a l’etude du r6le endocrine du cerveau et notamment de la pars intercerebralis chex les phasmides. Arch. ZooI. exp. gdn. 89, 128-138. HERLANT-Mxxwxs H. and PAQUETL. (1956) Neuros&xCtion et mue chex Carat&s morosus. Ann. Sci. nat. (Zool.) (11) 18, 163-168. HICHNAM K. C. (1961) The histology of the neurosecretory system of the adult female locust Schistocerca gregaria Forsk. Quart. J. mice. Sci. 102, 27-38. HIGHNAM K. C. (1962) Neurosecretory control of ovarian development in the desert locust. Neurosecretion. Mem. Sot. Endocrin. 12, 379-390. HODGSONE. S. and GELDIAY S. (1959) Experimentally induced release of neurosecretory materials from roach corpora cardiaca. Biol. Bull., Woods Hole 117, 275-283. MBTALNIKOVS. (1908) Recherches expirimentales sur les chenilles de Gall& mellonella. Arch. Zool. 8, 489-588. &BAs S. (1957) Two kinds of secretions in the corpora cardiaca of Locusta migratoria (Solitaria). Commun. Fat. Sci., Ankara 8,45-57. PRIZIBRAMH. and MEGUSARF. (1912) Wachstumsmessungen an Sphodromantis bioculata Burm.-I. Lange und Masse. Arch. EntwMech. Org. 34, 68&741. F&IM M. (1951) Du zeitliche Folge der Tatigkeitsrhythmen insekretorischer Organe von Ephestia ktiniella wiihrend der Metamorphose und des Imaginellebens. Arch. EntwMech. Org. 145, 205-248. SCHARRERB. and SCHARRERE. (1944) Neurosecretion .-VI. A comparison between the intercerebralis cardiacum allatum system of insects and the hypothalamo-hypophyseal system of vertebrates. Biol. Bull., Woods Hole 87, 242-251. SCHBERB. T. (1960) The neuroendocrine system of insects. Vitamins and Hormones 18, 141-195. STEEDMANH. F. (1947) Ester wax: a new embedding medium. Quart.J. micr. Sci. 85, 123. STRANGEWAYS-DIXON J. (1961) The relationships between nutrition, hormones and reproduction in the blowfly Calliphora erythrocephala Meig.-II. The effect of removing the ovaries, the corpus allatum and the median neurosecretory cells upon selective feeding, and the demonstration of the corpus allatum cycle. r. exp. Biol. 38, 637-646. THOMSENE. and MILLER I. (1959) Neurosecretion and intestinal proteinase activity in an insect, Calliphora erythrocephala Meig. Nature, Lond. 183, 1401-1042. TITSCHACKE. (1926) Untersuchungen tiber das Wachstum den Nahrungsverbrauch und die Einsaugung.-II. Tin&a biseiliella. Z. wiss. Zool. 128, 509-569. VAN DER KLOOT N. G. (1960) Neurosecretion in insects. Annu. Rev. Ent. 5, 35-52.