Identification of PBAN-like immunoreactivity in the neuroendocrine system and midgut of Dysdercus cingulatus ( Heteroptera : Pyrrhocoridae

Polyclonal antibodies against PBAN were used to map the distribution of PBAN-like antigenic sites in the brainsuboesophageal ganglion (Br-SOG) complex, associated neurohaemal structures, ventral nerve cord ganglia and in the alimentary canal. A pair of lateral neurosecretory cells immunopositive to the antiserum were found in each half of the deutocerebrum. PBANlike immunoreactivity (PLI) was also noticed in the tritocerebral region of the brain and in the aorta. Two groups of immunopositive cells of four and two cells respectively, were found in the SOG. Small and weakly immunoreactive neurons were identified in the prothoracic ganglion, whereas in the pterothoracic ganglion a pair of cells reacted positively to the antibody. Immunoreactive cells were also identified in the corpora cardiaca. Some of the epithelial endocrine cells of the midgut also showed immunoreactivity to PBAN antiserum.

PBANs are members of the pyrokinin family of neuro peptides, which are engaged in multiple physiological functions in insects, characterised by a common C-terminal sequence FXPRL amide (Abernathy et al., 1995).The C-terminal pentapeptide motif has been shown to be the shortest pheromonotropic fragment (Raina & Kempe, 1990) required for biological activity.PBAN, diapause hormone and other neuropeptides with the FXPRL-NH2 C-terminus result from the proteolytic processing of a protein precursor in both Bombyx mori and Helicoverpa zea (Sato et al., 1993;Ma et al., 1994).According to Matsumoto et al. (1990), the same neuro peptide controls cuticular melanization in larvae and pheromone production in adults of lepidoptera.PBAN or peptide(s) that belong to the PBAN peptide family induce embryonic diapause in Bombyx mori (Imai et al., 1991) and have myotropic functions in cockroaches and locusts (Holmanetal., 1986;Schoofs etal., 1991).
The neuroendocrine tissues of insects react positively to several vertebrate and invertebrate bioactive peptides (Kramer, 1984).The gut is also a very important endo crine organ in insects as a number of bioactive peptides have been immunochemically demonstrated there (Sehnal & Zitnan, 1990).Midgut and hindgut of insects receive granule-filled axons from the last abdominal ganglion (Nayar, 1954;Brown, 1967) and from corpora cardiaca directly (Johnson & Bowers, 1963).
The presence of immunoreactivity in several species of lepidopteran insects suggests a wide distribution of PBAN-like peptides within this order.However, phero monotropic activity has also been found in the heads of other non-lepidopteran insects, such as cockroaches and locusts (Sreng & Roelofs, 1988;Sreng et al., 1990).This activity could be due to the presence of PBAN or neuro peptides that share the active C-terminal part of the mole cule with PBAN (Fonagy et al., 1992;Kuniyoshi et al., 1992).Dysdercus cingulatus is a serious pest of cotton and there is no evidence of this family of neuropeptides in heteropteran insects.Hence it was thought reasonable to look for PBAN-like peptides in this particular insect, which might motivate further research leading to the con trol of this pest insect.

Antisera production
Polyclonal antibodies to synthetic H. zea PBAN (Peninsula Lab Inc., Belmont.CA) were raised in New Zealand strain of white rabbit.The rabbit was immunised with 250 pg of PBAN in Freund's complete adjuvant injected subcutaneously.Ten days later another booster injection was given of the same amount of PBAN in Freund's incomplete adjuvant.On 20th day, an additional booster dosage of 250 pg of antigen in 500 pl PBS was injected.Seven days later the rabbit was bled from the mar ginal ear vein.The serum was separated by centrifugation, aliquoted and stored at -40°C.The antiserum was further subjected to affinity purification to improve its specificity.For this the method followed was that ofKingan et al. (1992).

Dissection and fixation
Female insects of different ages were used.The brain-SOG with associated retrocerebral complex (CC-CA and aorta) and prothoracic and pterothoracic ganglia were dissected under icecold insect saline (Bindokas & Adams, 1988).The gut was fixed with a droplet of Bouin's fixative containing saturated picric acid, formaldehyde and glacial acetic acid and gut contents were removed.For paraffin sections, the tissues were embedded in paraffin wax with ceresin (MP.60°C).Sections were cut at 6pm thickness and mounted on albumin-coated glass slides.For wholemount immunostaining, tissues were fixed in Bouin's fixative for 2 hrs at room temperature.

Immunohistochemistry
For immunostaining of paraffin sections, the two step immunoperoxidase method of Schooneveld & Veenstra (1988) was followed.The sections were deparaffinised in xylene, then rehydrated and the activity of endogenous peroxidase blocked by immersing for 20 min in 0.05% hydrogen peroxide in PBS and treated with 10% normal goat serum for 10 min.Sections were incubated overnight at 4°C in antiserum against Hez-PBAN at different dilutions such as 1 : 500, 1 : 1000, 1 : 1500, 1 : 2000 and 1 : 2500.Slides were again rinsed in PBS, then treated with the secondary antibody (peroxidase labelled goatanti rabbit Ig from Bangalore Genei, India) at a dilution of 1 : 100 for 1 hr at room temperature.The sections were then rinsed again in PBS, stained with a mixture of PBS, H2O2 and 3-3/-diaminobenzidene tetra hydrochloride (DAB-HCl), dehy drated in ethanol series, cleared in xylene and mounted in DPX.The same experiments were conducted using the Hez-PBAN antiserum (a gift from Dr. A.K. Raina, USA) at a dilution of 1 : 2500.

Wholemount immunostaining
After fixation, tissues were washed with cold 70% ethanol overnight, followed by washes in 70% ethanol, 95% ethanol with 0.1% H2O2 and 70% ethanol, and then rinsed several times in cold PBS containing 1% Triton X-100 (PBST), until the yellow colour was completely removed.Tissues were then pre incubated in blocking serum, washed and incubated with pri mary antiserum for 24 hrs at 4°C.After washing with cold PBST, the tissues were incubated with secondary antibody for 1 hr at room temperature.After the final washing in cold PBS and once in 0.05 M Tris-HCl (pH 7.6), the tissues were stained with Tris-HCl containing 0.05% DAB and 0.03% H2O2 for 10 min utes.The staining was stopped by rinsing in PBS, and then the tissues were dehydrated, cleared and mounted.Control slides were prepared by: i. Omission of primary antibody to check nonspecific binding of secondary antibodies and other compounds applied in the procedure.ii.Replacing specific antisera with pre-immunoserum.iii.Pre-adsorption of antiserum overnight with 0.001 mg/ml Hez-PBAN.

RESULTS
These studies utilized two different polyclonal antisera.Different dilutions of our primary antisera were checked and the best results were obtained at a dilution of 1 : 500.When the antiserum provided by Dr. A.K. Raina was used the dilution was 1 : 2500, as recommended.Both antisera revealed the same PLI in all these structures.
Immunostaining of the brain revealed one pair of PBAN-positive lateral neurosecretory cells in the deutocerebrum (Fig. 1).PBAN-like immunoreactivity could be localised in the tritocerebral region of the brain and con siderable amount of PLI was also detected in the wall of the aorta (Fig. 2).Serial sections of the Br-SOG complex revealed two clusters of PBAN-like immunoreactive cells within the SOG with four and two cells in each group (Fig. 3, 4, 5).Horizontal sections revealed the presence of immunoreactive neurons in the CC (Fig. 6).Corpus allatum (CA) showed no detectable amount of immunoreactivity.
A few stained neurons were found to be present in the prothoracic ganglion.However staining of these neurons was generally weak and disappeared when a higher dilu tion of the antiserum (1 : 2500) was used.In this group of insects, the mesothoracic, metathoracic and abdominal neuromeres fuse to form a structure called the 'pterotho racic ganglion'.One pair of lateral immunopositive somata was observed in this ganglion (Fig. 9).
There was very little PBAN-like immunoreactivity in newly emerged adults.But after 48 hours, i.e. before they started mating, the immunoreactive cells contained large quantities of neurosecretion.However, after mating, PLI could not be detected in these tissues (Fig. 7, 8).
In this insect the midgut is composed of four regions, the first, second, third and fourth ventriculi.PLI was detected in the first and second ventriculus of the midgut (Fig. 10, 11).Some of the epithelial endocrine cells showed an immunopositive reaction to the antiserum (Fig. 11).
No immunostaining was observed when pre-immune serum was used and PBAN-like immunoreactivity was absent when the antiserum was preadsorbed overnight with Hez-PBAN.

DISCUSSION
Pheromone production in lepidopteran insects is known to be regulated by the neuropeptide PBAN.We used an antiserum to Hez-PBAN to map the distribution of immu noreactive somata in the neuroendocrine system as well as the alimentary canal of the heteropteran insect Dysdercus cingulatus.
Immunohistochemical studies revealed a pair of PBAN positive lateral neurosecretory cells in each half of the deutocerebrum.The presence of PLI in the tritocerebral region and the aorta suggests neurohaemal function of aotra.This finding contradicts the results obtained with other lepidopteran insects, where PLI was not detected in the brain (Blackburn et al., 1992;Ichikawa et al., 1995;Davis et al., 1996).However, Kingan et al. (1992) identi fied a fine perhaps 'fuzzy' immunoreactivity in the tritocerebrum of Helicoverpa zea.They suggested that this tritocerebral staining is of branches of the maxillary neurites near their dorsal extension in the SOG.Two clusters of immunoreactive somata were detected in the SOG of Dysdercus cingulatus.This also differs from that reported for several lepidopteran insects like Helicoverpa zea (Blackburn et al., 1992;Kingan et al., 1992), Bombyx mori (Ichikawa et al., 1995), Manduca sexta (Davis et al., 1996), Lymantria dispar (Golubeva et al., 1997) and recently for Agrotis ipsilon (Duportets et al., 1998) where there are three clusters of cells -the mandibular, maxillary and labial.The insect SOG is formed by the fusion of the neuromeres of the mandibular, maxillary and labial segments.In this study maxillary and labial neuroendocrine cells showed PBANlike immunoreactivity.Golubeva et al. (1997) have shown that many of the SOG cells in L. dispar are neu rosecretory, in that they have varicose endings in identi fied neurohaemal organs.In D. cingulatus the immuno reactivity found in the CC may indicate its functions as a neurohaemal structure for the peptide.
In this study PLI was weak in the prothoracic ganglion but a pair of large neurosecretory cells was located in the pterothoracic ganglion.In the gypsy moth, the prothoracic ganglion and pterothoracic ganglion each contain a pair of ir-somata (Golubeva et al., 1997).A pair of ventral midline cells in each of the abdominal ganglion including the terminal one (TAG) are immunopositive in the Euro pean corn borer (Ma & Roelofs, 1995).
There is now considerable evidence of both central and peripheral sites of action for various peptidergic neurons in insects (Nassel, 1995).The patterns of staining revealed by both immunocytochemistry and ELISA bio assays indicated that there are sites of release not only into the haemolymph, but also within the CNS itself (Rafaeli et al., 1991;Kingan et al., 1992;Davis et al., 1996).Thus activation of the pheromone glands in lepidopterans may involve both neural and hormonal mecha nisms.In contrast to the pheromone glands in lepidoptera, the function is carried out by the scent glands in heteroptera (Staddon, 1979).
Cross biological activities are often observed between the FXPRL amide peptides of which PBAN is a member.Further research is therefore needed to confirm that these immunoreactive cells are engaged only in pheromone production in this heteropteran insect.The high immuno reactivity observed before mating and lack of immunoreaction after mating suggest a possible role in eliciting mating behaviour.
PBAN-like material was also observed in different regions of the alimentary canal, especially in the midgut of D. cingulatus.This distribution of PBAN-like immu noreactivity suggests a multiplicity of functions besides the regulation of pheromone production.Yu et al. (1995) observed an allatostatin immunoreactivity in the proctodeal innervation and endocrine cells of the midgut, which was previously demonstrated in neurosecretory cells of the brain and ventral nerve cord ganglia of D. punctata.These peptides were identified originally on the basis of their ability to inhibit JH biosynthesis by CA.They were subsequently found to inhibit myogenic contractions of the hindgut in cockroaches.FMRF-amide positive cells of the abdominal ganglion send axons via the median abdominal nerve to the hindgut where they form an inner vation pattern very similar to that of proctolin-containing neurons (An et al., 1998).In M. sexta, eclosion hormone containing neurons have varicose axons running through the entire length of cephalic, thoracic and abdominal gan glia before they reach their release sites on the hindgut (Truman & Copenhaver, 1989).
Bom-PBAN gene codes for four other insect neuropep tides, which regulate pheromone production, induction of embryonic diapause, melanization of the integument and also stimulate muscle contraction in insect gut (Kawano et al., 1992).These peptides share with PBAN the C-terminal pentapeptide FXPRL-NH2.Bom-PBAN I can produce a strong melanization of the cuticle of Pseudaletia separata (Matsumoto et al., 1990).
Thus, it seems that PBAN-like immunoreactivity exhib ited by epithelial endocrine cells of the gut could result from cross-reactivity of the PBAN-antiserum with other unidentified myotropic peptides of the same family, which share the active C-terminus of the peptide.There is evidence that some gut endocrine cell products enter the body circulation and exert hormonal effects on distal tar gets.Due to the various routes of action, a single hor mone can exert a variety of effects in insects (Sehnal & Zitnan, 1996).Further research is however needed to con firm the functions of PBAN-like peptides in this group of insects.

Figs
Figs 1-6:1-sagittal section of the brain showing PBAN-like immunoreactive cells in the lateral region of the deutocerebrum (arrow mark); 2 -tritocerebral immunopositive neurons and their extension to the aorta; 3 -two groups of maxillary and labial immunopositive somata in the SOG; 4 -5 -enlarged view of the two sets of somata in the SOG.Scale 25 pm; 6 -immunoreactivity exhibited by CC.Scale 40 pm.