Peripheral Neurosecretory Cells of Insects Contain a Neuropeptide with Bursicon-Like Activity

We have discovered a group of three, hitherto undescribed, peripheral neurosecretory cells (PNC) in the neck region of crickets (Gryllus birnaculatus) with the aid of cobalt chloride backfills (Tyrer & Altman, 1974) of prothoracic and suboesophageal nerves, followed by silver intensification (Bacon & Altman, 1977) (Fig. 1). In addition, a network of fibres with superficial varicosities is revealed on all anterior nerves of the prothoracic ganglion and one nerve of the suboesophageal ganglion when PNC are labelled anterogradely (Figs 1, 2A,B). Electron micrographs of the PNC (Fig. 2C) reveal a cytoplasm containing electron-dense vesicles, mitochondria, Golgi complexes and rough-surfaced endoplasmic reticulum, indicating that these cells are in an active metabolic phase. Sections of the nerves covered with PNC projections show processes embedded in the neural sheath of the nerve, filled with electron-dense vesicles of the same type as those in the PNC cytoplasm (Fig. 2D). Many PNC fibres enter the median nerve, an unpaired nerve between the neck connectives, via its branches, the transverse nerves (Fig. 1). In insects, median and transverse nerves are neurohaemal organs (NHO) where neurosecretory material is stored and released into the haemolymph (Gupta, 1983).

The median nerve (MN) between the suboesophageal and prothoracic ganglia of adult crickets with its two branches, the transverse nerves (TN) (Fig. 1), contains bursicon-like activity (Table 1; Honegger et al. 1988). Bursicon is a protein with a relative molecular mass of 30–60 (Mills & Lake, 1966; Fraenkel et al. 1966; Reynolds, 1977, 1983) that induces tanning of the cuticle in freshly moulted insects (Fraenkel & Hsiao, 1965; Cottrell, 1962). Its function can be tested using the ‘newly emerged ligated fly bioassay’ (Fraenkel & Hsiao, 1965; Seligman, 1980), in which a saline homogenate of the tissue of interest is injected into newly emerged and ligated flies. A tanning score is calculated for each test fly, the maximum score being 6 points for a fully tanned fly (Honegger et al. 1988). The score depends on the bursicon content of the injected homogenate. In the present study, all homogenates were made from tissue preparations of adult male and female crickets at least 14 days after eclosion.

The prothoracic MN of crickets contains fibres of two cell clusters in the posteriolateral region of the prothoracic ganglion (Honegger et al. 1988), axons of four motoneurones in the suboesophageal ganglion innervating the neck muscles M62 (Honegger et al. 1984) and fibres of the PNC.

No tanning activity could be detected in the saline homogenates of ganglion sections containing the two prothoracic cell clusters (mean score 0·26, N= 4 tests with 20 flies per test; mean score of saline controls 0·1; Honegger et al. 1988). In locusts, homologous cell populations have been shown to be FMRFamide- and bovine pancreatic polypeptide (BPP)-immunoreactive (Myers & Evans, 1985a,b). The negative result with homogenates of the posterior lateral cell clusters corresponds with our observation that FMRFamide (10⁻⁴ to 10⁻⁸mol l⁻¹, Sigma) has no tanning activity in the bioassay (score equals saline controls). Thus, the posterior lateral cells of the prothoracic ganglion do not contribute to the bursicon-like activity of the MN.

The axons of two of the four suboesophageal M62 motoneurones with their cell bodies in the posteriolateral labial neuromere (Fig. 1) descend through the neck connectives to the prothoracic ganglion, where they turn through 180°, without branching, and leave through the prothoracic anterior median nerve (I.aMN; Fig. 1) to proceed through both transverse nerves towards muscles M62. The axons of the other two motoneurones with their somata at the border between the labial and maxillary neuromeres project to the same muscles via the suboesophageal posterior median nerve (S.pMN; Fig. 1) and the transverse nerves (Honegger et al. 1984). We tested for bursicon activity in homogenates of the I.aMN, the neck connectives, the muscles M62 and in sections of both sides of the posterior suboesophageal ganglion which contain the somata of the two M62 motoneurones. Since the cell body positions are variable, the ganglion sections had to be fairly large. They were dissected with one cut, made slightly anterior and parallel to nerve N5 and a second cut perpendicular to the first extending to the lateral edge of the neck connective, and contained roughly 600–650 somata each (numbers calculated from 10 μm transverse sections stained with ethyl gallate).

Homogenates of I.aMN caused tanning in test flies but homogenates of the neck connectives and muscles M62 did not (Table 1). These results indicate that the two M62 motoneurones with the more posterior somata and axons projecting through the neck connectives can be ruled out as a source of bursicon-like material since homogenates of these as well the neck connectives of the I.aMN and of muscles M62 should then have caused tanning. Suboesophageal ganglion sections contain bursicon-like activity (Table 1). Therefore, the other two motoneurones with more anterior somata may contribute to the bursicon-like activity found in the median-transverse nerve complex.

Saline homogenates of the prothoracic PNC of adult crickets show bursicon-like activity (Table 1). In addition, homogenates of all the nerves covered with varicose fibres from the PNC (Fig. 1) show bursicon-like activity (Table 1). Since the cobalt histology shows the PNC net to be of comparable density on all nerve surfaces, nerves with a small diameter, such as N8(SOG) and Nl(PTG), must carry less bursicon-like material than nerves with a large diameter, such as N3(PTG). In these tests, therefore, concentrations of nerve homogenates were normalized with respect to the surface area of N3(PTG) (Table 1). Homogenates of the prothoracic nerves N4 and N5, which do not carry PNC fibres, had no detectable tanning activity even when doubly concentrated homogenates were injected (Table 1; N5 is the prothoracic nerve with the largest diameter, N4 has an equal diameter to N3).

Although the structure in which the PNC are embedded is tiny, homogenates of the PNC of standard concentration had a high tanning score (Table 1). This indicates strongly that the bursicon activity is intrinsic to the PNC and does not originate from fibres of unknown central neurones innervating the PNC. Our anatomical results support this conclusion. First, cobalt stainings of prothoracic nerves never revealed any axon collaterals projecting to the PNC. Second, electron micrographs of the PNC cluster showed only few fibres embedded in the thin connective tissue sheath surrounding them. These few profiles, derived from unknown efferent neurones, are very unlikely to account for the high bursicon activity. In addition, these fibres contain vesicles of the same type as the PNC cytoplasm, suggesting that they originate from the PNC.

The bursicon-like activity of the PNC and of the MN is completely abolished after 3h of incubation of homogenates with proteinase K (5 mg ml⁻¹, Sigma) at 37 °C (Table 1). Thus, the tanning activity of the prothoracic PNC in crickets is associated with at least one peptide.

Our results show for the first time that a distinct group of prothoracic peripheral neurosecretory cells contains a bursicon-like neuropeptide. Bioassays combined with cobalt stains indicate that it is also present in fibres of homologous PNC of the mesothoracic segment. Bursicon-like activity, has been demonstrated in ganglia of the central nervous system of various insects (Mills & Lake, 1966; Taghert & Truman, 1982a; Vincent, 1972), including crickets (Honegger et al. 1988; and this report). In Manduca sexta, bursicon released shortly after eclosion seems to be produced by four central neurones of the abdominal ganglia and released via abdominal NHO (Taghert & Truman, 1982b). It remains to be investigated whether the centrally and peripherally derived bursicon-like proteins are identical.

Considerable amounts of a bursicon-like substance must be synthesized in the PNC of adult crickets, although its function in adult insects is unknown. It is possible, however, that it regulates the tanning of growing cuticle in adults in the same way as in larvae. In adult locusts, the apodemes and other parts of the endocuticle show daily growth layers (Neville, 1967), and the endocuticle continues to grow for up to 3 weeks after adult emergence (Andersen, 1973). In addition, bursicon could be used in adult insects during the process of wound repair when epidermal cells secrete new cuticle (Neville, 1975).

We have shown that a bursicon-like peptide is produced by nerve cells lying in the periphery of the nervous system. The activity of nerve cells in the periphery of the nervous system, which are poorly insulated from the haemolymph, may be directly controlled by circulating levels of hormones and ions. The resulting secretion of bursicon from the fibres of these nerve cells may be directed not only into the exterior haemolymph but also into the interior medium, i.e. the central nervous system, thereby modulating its activity.

We thank F. P. Fischer for help and advice during the electron microscopic studies, A. Goxers for help in the bioassays, and P. Braunig and J. Kien for critical comments on the manuscript. Supported by DFG, HO 463/17-1.