IMmunocytochemical localization of human growth hormone- and prolactin-like antigenic determinants in the insects, Locusta migratoria and Sarcophaga bullata

Camp.

Biochem.

Physiol.

Vol. 95A,

No.

3, pp.

373-318,

1990

0300-9629/90

Printed in Great Britain

$3.00

+ 0.00

0 1990Pergamon Press plc

IMMUNOCYTOCHEMICAL LOCALIZATION OF HUMAN GROWTH HORMONE- AND PROLACTIN-LIKE ANTIGENIC DETERMINANTS IN THE INSECTS, LOCUSTA MIGRATORIA AND SARCOPHAGA BULLATA K. SWINNEN, J. VANDEN BROECK,* P. VERHAERT and A. DE LOOF Zoological Institute of the University, Naamsestraat

59, B-3000 Leuven, Belgium

(Received 24 August 1989)

Abstract-l. By use of the peroxidase-antiperoxidase immunocytochemical method, substances immunoreactive to antisera directed against human growth hormone (hGH) and prolactin (hPr1) were localized in the nervous system of larval and adult Locusfa migratoria and of adult Sarcophaga bullata belonging to different age groups. 2. No major differences in the distribution of cerebral immunoreactive materials were observed between males and females or between juvenile and adult insects. 3. Differential immuno-labeling of alternating tissue sections demonstrated that materials resembling hGH or hPr1 are present in distinct neurons in the locust, whereas neurons immunoreactive to both antisera were detected in the fleshfly (Sarcophaga).

INTRODUCIION In recent years, an increasing number of substances immunochemically resembling different vertebrate neuropeptides and gastrointestinal peptide hormones have been found in insects and other invertebrates; a number which, at this moment, already exceeds 50 (e.g. RCmy et al., 1979; Hansen et al., 1982; El-Salhy et al., 1983; Duve and Thorpe, 1983; Veenstra et al., 1985; Verhaert et al., 1984 and Verhaert, 1987; De Loof et al., 1988). In this paper, we present immunocytochemical data demonstrating the distribution of substances antigenically similar to human growth hormone (hGH) and prolactin (hPr1) in the central nervous system of two insect species, the African migratory locust and the grey fleshfly. MATERIALS AND METHODS

at 1: 2000. The second layer reagent (a generous gift from Dr F. Vandesande) was a goat anti-rabbit IgG antiserum (GAR) diluted 1: 30 (incubation time of 30 min). Subsequently, PAP complex (Dr F. Vandesande) was added as a third layer reagent (working dilution 1: 300) for 30 min. All incubations were performed at room temperature and between all steps thorough washing of the tissue sections was performed. Peroxide and diaminobenzidine were used to localize antigens present in the tissue resulting in a brown stain at those sites where the peroxidase enzyme was retained during the above immunocytochemical procedure. Double stainings were performed according to Vandesande (1983). Method specificity was controlled by the systematic omission of different steps in the immunocytochemical procedure. These treatments resulted in a complete absence of staining. Serum specificity was tested by the use of a pre-immune rabbit serum instead of the primary antisera and by solid phase immuno-absorption of both antisera using bovine GH (Monsanto) coupled to Sepharose 4B beads.

Imaginal and all five larval instars of Locusta migratoriu migratoroides R. and F. and freshly hatched (day 0), 3(day 3) and 5- (day 5) day-old adult Sarcophaga bullata Parker were used. Locust central nervous system (CNS) i.e. brain, suboesophageal ganglion and the ventral nerve cord (three thoracic ganglia and five abdominal ganglia), together with the adjacent corpora cardiacs/corpora allata complexes and fly CNS, i.e. brain, suboesophageal ganglion and ventral ganglion, were microdissected and fixed in situ in Bouin-Hollande’s 10% sublimate solution. Tissues were dehydrated in ethanols, cleared in xylene and embedded in Paraplast. Complete series of 4pm tissue sections were processed in the peroxidase-antiperoxidase (PAP) method (Vandesande and Dierickx, 1976). In this procedure, rehydrated tissue sections were incubated (overnight) with rabbit primary antisera against human growth hormone (UCB Bioproducts, lot no. 1515jOO2)and human prolactin (Eurogenetics, lot no. RP2448L). The optimal working dilution of both antisera was determined *Author to whom all correspondence

should be addressed.

RESULTS In neuronal tissues of both insects, the hGH and hPr1 antisera revealed clear bilaterally symmetrical immunoreactivities. The staining results described in the following sections are, therefore, restricted to only one half of the neuronal ganglia. No major differences were observed between males and females or between larvae and imagines of either insect species. Locusta

migratoria

(Fig.

1)

hGH Antiserum. Immunoreactivity to hGH was found mainly in the optic lobe (Fig. la). A first group of immunoreactive cell bodies, consisting of four to six neurons, is located between the lamina ganglionaris (LG) and the compound eye retina (Fig. 2). Some GH-immunoreactive nerve fibers in this region enter the LG and the medulla externa (ME).

374

K. SWINNENet al.

,

TRITO

,CA

Fig. I. Diagram of the brain-suboesophageal complex of Locusta migratoriu showing hGH- (a) and hPrl- (b) immunoreactive cell bodies (0) and nerve fibers (-----). CA, corpora allata; CC, corpora cardiaca; DEUTERO, deuterocerebrum; OC, ocellus; OL, optic lobe; PROTO, protocerebrum; SOG, suboesophageal ganglion; TRITO. tritocerebrum.

A second group (three to five immunoreactive cell bodies) is situated in a small area at the base of the eyestalk, ventrally to the ME. Their immunopositive axons enter a thick axon bundle, lying more distally in the optic lobe (Fig. 3). Immunoreactive nerve fibers, probably originating from this bundle, run through the brain and branch in the protocerebral neuropil. Numerous terminal arborizations of nerve fibres, coming from the brain, are observed in the extrinsic (storage) part of the corpora cardiaca (CC). Only a few branches of the immunopositive fibres reach the corpora allata (CA) where they are present among the CA gland cells. None of the intrinsic glandular cells of the CC and CA exhibit hGH-immunoreactivity. Immunoreactivity in the suboesophageal ganglion (SOG) is confined to a few very fine nerve fibres in the neuropil. No hGH-resembling antigens are observed in cell bodies in any of the eight ganglia of the ventral nerve cord. Only a few nerve cell processes are positively labeled in the neuropil. hPr1 Antiserum. Positive reaction with the hPr1 antiserum (Fig. lb) was obtained in the protocerebrum, where 20-30 small (lo-15 pm) perikarya of the anterior pars intercerebralis (PI) are intermingled with non-immunoreactive neurons of similar size (Fig. 4). Many of their immunoreactive axons join together to form a median nerve tract which descends and divides into two further tracts crossing one another along the median axis of the brain. They leave the brain ventro-caudally in the nervi corporis cardiaci I (NCC I), via which the immunoreactive material reaches the CC. Here it is stored in numerous axonal ramifications in the extrinsic part of the organ (Fig. 5).

As with the hGH antiserum, none of the intrinsic glandular cells of the CC or the CA contain immunoreactive material. In the deuterocerebrum, intensely stained nerve fibres are present, arranged in a circle around the bulbus olfactorius. Immunoreactivity to hPrl in the ganglia of the ventral nerve cord is confined to some sparse nerve fibres. Double labeling. Double staining with both antisera revealed that none of the hPrl-immunoreactive neurons or nerve fibres (e.g. in P.I., CC, CA, ventral nerve cord) are hGH-immunopositive. Conversely, hPrl-like material is not found in hGH-immunoreactive neurons (e.g. in the optic lobe). Sarcophaga

bullata (Figs

6 and 7)

hGH Antiserum. Material resembling hGH is present in seven to eight large (2&25pm) median neurosecretory cells which appear to be arranged in a triangle (Figs 6a, 8a and 9a); their axonal nerve fibres also show hGH-immunoreactivity. About six other hGH-immunoreactive neurons are observed between the PI and the pars lateralis (PL): three of them (relatively small cell bodies, 12-18 pm) are located at the dorsal surface. (Fig. 10a) The other three are located lower down. Via the pons protocerebralis, the immunoreactive axonal fibres of the neurosecretory cells reach the corpus centrale where they frequently branch. They then fuse to form one axon bundle in each brain hemisphere. These bundles go downward to the ventral side of the brain and leave the brain as the NCC I. Nerve fibres in the neuropil and two neurons in the ventral cortex of the SOG are also labeled by the hGH antiserum.

Insect somatotropin- and prolactin-like substances

Fig. 2. Section through the optic lobe of Locusra migrate-in, showing hGH-immunoreactive

perikarya

distally in the optic lobe. (x 590). Fig. 3. Section through the optic lobe of Locusta migraforia, showing hGH-immunoreactive perikarya in the basal part of the eyestalk. Axonal fibres enter a thick axon bundle (Fig. 3, arrows) (x 400). Fig. 4. Cell bodies and axonal fibres in the pars intercerebralis of Locusta migratoria, labeled by the hPrl antiserum. (x 360). Fig. 5. Terminal arborizations

in the corpora cardiaca of Locusta migratoriu, labeled by the hPrl antiserum ( x 300).

Immunopositive material in the ventral ganglion (VG) is present in three to six neurons, all located in the dorsal cortex (Fig. 7a). hPr1 Antiserum. HPrl-like material is present in median neurosecretory cell bodies (similar to the distribution of hGH-like material) (Fig. 6b): about nine to ten large perikarya and axonal fibres show a strong hPrl-immunoreactivity (Figs 8b and 9b). Immunopositive hPrl-like antigens are also stored in 1I to I5 neurons, spread in small groups of three to four cells between the PI and PL (Fig. lob). Anteriorly in the optic lobe, in the cortex of the medulla interna (MI), another group of three hPrl-immunopositive cell bodies is evident. HPrlimmunoreactive axonal fibres enter the MI, between the MI and ME and between the ME and LG. A last group of hPrl-like material containing cell bodies in the brain consists of 6-10 very small neurons situated posteriorly in the PI.

Immunopositive neuronal processes leave the brain through the NCC I, following a pathway similar to the one described above for hGH-immunoreactive axon bundles. In the VG hPrl-immunoreactivity is detected in six to ten neurons of the ventral cortex while the hGHlike material is present in neurons of the dorsal cortex (Fig. 7b). Corn~~r~ti~~~use of both u~tiseru. Labelings of alternating tissue sections with the hGH and hPr1 antiserum, respectively, demonstrated that five to six of the median neurosecretory cells which are arranged in a triangle and contain hGH-like antigens are also immunoreactive to the anti-hPr1 serum and vice versa (Figs 6c, 8 and 9). However, other median neurosecretory cells which are located among these “double” stained neurons, are only recognized by one of the antisera (two and four “single” stained cells for the hGH and the hPr1 antiserum, respectively;

316

K.

SWINNEN

et al.

Fig. 6. Diagram of the brain-suboesophageal complex of Sarcophaga bulluta showing neurosecretory cells (0) and nerve fibres (-----), labeled by the hGH (a) or hPr1 (b) antiserum. Most of the immunoreactive cell bodies

and nerve fibres are labeled by both antisera PI, pars intercerebralis; PL, pars lateralis;

Figs 8 and 9, arrows). This is confirmed by double stainings: the structures, stained with the hGH antiserum and those labeled by the hPr1 antiserum partly overlap. Six other “double” immunoreactive neurons are present among the immunoreactive cells located between the PI and PL (Figs 6c and 10). In other parts of the brain and VG, no “double” labeled immunoreactive neuronal cell bodies occur, although the axonal pathway of the hGH- and hhl-like material is very similar (Figs. 6a,b): many nerve fibres appear “double” immunoreactive (Fig. 6~). However, the hPrl antiserum revealed more immunoreactive nerve fibres than did the hGH antiserum. Controls

Anti-hGH absorbed with bovine GH did not result in any positive reactions in the insect tissue sections. Cross-absorbed anti-hPr1, however, still

Fig. 7. Diagram

of fly ventral ganglion showing hGHand hPrl- (b) immunoreactive cell bodies.

(a)

(c). OF, oesophageal SOG, suboesophageal

foramen; OL, optic lobe; ganglion.

stained the same neurons, as did the original hPr1 antiserum. The hGH and hPr1 antisera specifically stained relevant vertebrate tissues (intense immunoreaction in the acidophilic cells of the rat anterior pituitary). DlSCtiSSlON

During the past decade, many immunocytochemical studies have demonstrated reactions of anti-vertebrate peptide or protein hormone antisera with insect nervous tissue antigens. These data suggest a high complexity of the insect nervous system, containing several substances antigenically similar to vertebrate counterparts. These substances probably function as transmitters, modulators or hormones. In this paper, results obtained with antihGH and anti-hPrl antisera are described in the nervous system of two insect species belonging to different orders. GH or somatotropin and prolactin (Prl) are vertebrate pituitary protein hormones that are structurally related. Their genes are supposed to have originated from a duplication of a common ancestral gene sequence (e.g. Miller and Eberhardt, 1983). Therefore, data on the presence of GH- or Prl-like substances in insects may be interesting from an evolutionary point of view. Unfortunately, only a few data exist on the occurrence of GH- or Prl-like material in insects. In 1977 Patton and Kuo found a substance in the haemolymph and in the brain of Acheta, Adalia, Galleria, Prodenia and Apis, having several characteristics in common with GH (the substance had the approximate molecular weight

Insect somatotropin-

and prolactin-like

substances

317

9a

c i

IOa Figs 8-10.

Alternating

Fig. 8. Five neurons

tissue sections

in PI stained

of Sarcophuga bullata brain, antiserum, respectively.

by both antisera

and two neurons

stained

with hGH

recognized

(a) or hPr1 (b)

by the hGH

antiserum

only (a, arrows) (x 510). Fig. 9. One single neuron between PI and PL, recognized by both antisera and other containing only hGH- (a, single arrows) or hPrl (b, double arrows)-like material. Fig. 10. Neurons

between

PI and PL, labelled by both antisera, and a single neuron immunoreactivity (b, double arrows) ( x 700).

of a GH-dimer). GH and Prl immunoreactivities have been described in the central nervous system of Leptinotarsa (Veenstra et al., 1985) Periplaneta (Verhaert et al., 1986; Verhaert, 1987; De Loof et al., 1988) and Leucophaea (Hansen et al., 1988). Bhaktan and Gilbert (1968) reported an effect of vertebrate GH treatment on insects. They found a release of diglycerids and free fatty acids from the fat body of Hyafophora after GH administration.

neurons in PI (x 510). containing

hPrl-

In agreement with those earlier reports, the present data suggest the occurrence of substances in the nervous system of insects which have some characteristics in common with hGH and/or hPr1. At least two distinct neurochemicals, detected by anti-hGH and anti-hPrl antisera, are present in the central nervous system of Locusta migratoria and Sarcophaga bullata. Differences in the localization of the hGH and the hPr1 immunoreactive tissue antigens

K. SWINNEN et al.

378

have been demonstrated. In several neurons of the fleshfly, colocalization of the hGH- and hPrlimmunoreactivities has been found. This observation may be explained by (1) the occurrence of a substance antigenically similar to both hGH and hPr1 or (2) the presence of two distinct substances within the same neuron, one being recognized by the anti-hGH and the other by the anti-hPr1 antiserum. No major differences between the distribution patterns of the neurochemicals in larval and adult locusts or in young and old adult fleshflies have been observed. This permanent occurrence suggests a permanent functional need for those neuronal substances. Furthermore, the immunopositive neurochemicals do not seem to be male- or female-specific, neither do their distribution patterns seem to be sex dependent. Bovine GH (bGH) has been used to pre-absorb the hPr1 and the hGH antisera. Apparently, all relevant anti-hGH antibodies bind to bovine GH, since after pre-absorption of the hGH antiserum with bGH no reaction could further be obtained in any of the insect neurons. However, cross-absorption of the hPr1 antiserum did not have any effect on the immunoreactivity of this antiserum in either insect species. The cross-absorbed hPr1 antiserum reacted with the same types of neurons as the non-absorbed one. These results suggest that the “hGH-like” insect subtance is also “bGH-like” and that the “hPrllike” substance most likely is not. Futher data on the molecular structure of the immunoreactive compounds will be needed in order to determine to what extent these insect neuronal substances are similar to the vertebrate pituitary hormones, somatotropin and prolactin. Acknowledgements-The authors wish to thank the NFWO and the IWONL of Belgium for financial support, Prof. Dr F. Vandesande for providing the necessary secondary antisera, Mr R. Smisdom for technical assistance, Mrs J. Puttemans for photography and Mr J. Evans for text correction. J. Vanden Broeck and P. Verhaert are research assistants of the NFWO of Belgium.

REFERENCES

Bhaktan N. M. G. and Gilbert L. I. (1968) Effects of some vertebrate hormones on lipid mobilization in the insect fat body. Gen camp. Endocrinol. 11, 186197. De Loof A., Verhaert P., Schoofs L., Huybrechts R., Frankenne F., Hennen G., Vaudry H. and Jegou S. (1988) Vertebrate-type hormones in insects. In Endocrinological Frontiers in Physiological Insect Ecology (Edited by Sehnal F., Zabza A. and Denlinger D. L.) pp. 813-831. Wroclaw Technical University Press, Wroclaw, Poland.

Duve H. and Thorpe A. (1983) Immunocytochemical identification of a-endorphin-like material in neurones of the brain and corpus cardiacum of the blowfly, Calliphora vomiforia (Diptera). Cell Tiss. Res. 233, 415426. El-Salhy M., Falkmer S., Kramer K. J. and Speirs R. D. (1983) Immunohistochemical investigations of neuropeptides in the brain, corpora cardiaca and corpora allata of an adult lepidopteran insect, Manduca sexfa (L). Cell Tiss. Res. 232, 295-217. Hansen H. L., Hansen G. N. and Scharrer B. (1982) Immunoreactive material resembling vertebrate neuropeptides in the corpus cardiacum and corpus allatum of the insect Leucophaea tnaderae. Cell Tiss. Res 225, 3 19-329. Hansen G. N., Hansen B. L. and Scharrer B. (1988) Diversity of prolactin systems in the insect Leucophaea maderae: use of antiserum polyclonality for immunocytochemical detection of neuropeptide heterogeneity. Cell Tiss. Res 252, 557-563. Miller W. L. and Eberhardt N. L. (1983) Structure and evolution of the growth hormone gene family. Endocrine Ret). 97-130. Patton R. L. and Kuo C.-C. (1977) The distribution of the STH-like protein in insects. insect Biochem. 7, 487489. R&my C.. Girardie J., Dubois M. P. (1979) Vertebrate neuropeptide-like substances in the suboesophageal ganglion of two insects: Locusfa migraforia R. and F. (Orthoptera) and Bombyx mori (Lepidoptera). Gen. camp. Endocrinol. 37, 93-100. Scharrer B. (1978) Peptidergic neurons: Facts and trends. Gen. camp. Endocrinol. 34, 50-62. Vandesande F. (1983) Immunohistochemical double staining techniques. In Neuroimmunohisfochemisfry (Edited by Cue110 E.), pp. 257.-272. John Wiley, New York. Vandesande F. and Dierickx K. (1976) Immunocytochemical demonstration of separate vasotocinergic and mesotocinergic neurons in the amphibian magnocellular neurosecretory system. Cell Tiss. Res. 179, 289-296. Veenstra J. A., Romberg-Privee H. M., Schooneveld H. and Polak J. M. (1985) Immunocytochemical localization of peptidergic neurons and neurosecretory cells in the neuro-endocrine system of the Colorado potato beetle with antisera to vertebrate regulatory peptides. Hisfochemistry 82, 9-18. Verhaert P. (1987) Onderzoek naar het voorkomen van stoffen met kenmerken van peptidehormonen van gewervelde dieren bij insekten. Een studie op de Amerikaanse kakkerlak Periplanefa americana L. PHD thesis, K. U. Leuven, Belgium. 241 pp. Verhaert P.. Geysen J., De Loof A. and Vandesande F. (1984) Immunoreactive material resembling vertebrate neuropeptides and neurophysins in the brain, suboesophageal ganglion, corpus cardiacum and corpus allatum of the dictyopteran Periplanefa americana L. Cell Tiss. Res. 238, 55-59. Verhaert P., Van Mellaert H.. Swinnen K. and De Loof A. (1986) Vertebrate somatotropin-like peptides in insects. In Progress in Developmental Biology, Parf A (Edited by Slavkin H. C.), pp. 359-364. A. R. Liss, Inc., New York.