Biochemical evidence of phenoloxidase activity (pro-PO system) in larvae of Allogamus auricollis (Insecta, Trichoptera)

Comp. Biochem. Physiol. Vol. 102B,No. 4, pp. 867-871, 1992 Printed in Great Britain

0305-0491/92$5.00+ 0.00 © 1992PergamonPress Ltd

BIOCHEMICAL EVIDENCE OF PHENOLOXIDASE ACTIVITY (PRO-PO SYSTEM) IN LARVAE OF ALLOGAMUS AURICOLLIS (INSECTA, TRICHOPTERA) M. F. Bglvm,* M. PAGANI and G. SCARi Department of Biology, Section of Zoology and Cytology, University of Milan, via Celoria 26, 20133 Milan, Italy (Received 3 December 1991)

Abstract--1. Allogamus auricollis cell-free hemolymph proteins were analyzed by SDS-PAGE. Under reducing conditions the gel pattern showed two main components (83 and 76 kDa) and some lesser bands. 2. After native PAGE, a single band showed phenoloxidase activity by /n situ enzymatic staining. 3. Spectrophotometric analysis of the cell-free plasma fraction was carried out with substrate and PTU as inhibitor.

INTRODUCTION Insect defense reactions against microorganisms or parasites include cellular and humoral responses involving several hemocyte types and inducible/not-inducible molecular factors, normally present or newly synthesized in the hemolymph (for a review, see G6tz and Boman, 1985). Cellular defences mainly consist of phagocytosis (Stairs, 1964; Anderson et al., 1973; Ratner and Vinson, 1983) or encapsulation (Grimstone et al., 1967; Rowley and Ratcliffe, 1981; Amirante, 1986; G6tz, 1986), both carried out by freely circulating hemocytes. Unfortunately, it is hard to comprehend these functions because of the confused terminology used for hemocyte classification (for reviews, see Gupta, 1979, and Breh61in and Zachary, 1986). Humoral immunity in arthropods involves a number of not-inducible molecular factors, such as lectins or phenoloxidase, and infection-inducible factors, such as cecropins or lysozyme (Chadwick and Aston, 1979; Amirante and Basso, 1984; S6derh/ill and Smith, 1986a; Boman and Hultmark, 1981, 1987). As reported by Faye (1978) and Boman (1981), one of the first steps after infection is RNA synthesis, followed by protein synthesis. Changes in the protein pattern can be observed after SDS-PAGE separation; these newly synthesized proteins, revealed by isotope labelling, are conventionally called PI-P9. These protein families show their antibacterial activity by effective destruction of bacterial cell walls (Pye and Boman, 1977; Rasmuson and Boman, 1979; Hultmark et al., 1980). Phenoloxidase (PO), a defense factor normally present in the hemolymph of many arthropods, is the active form from the precursor prophenoloxidase (pro) in the so-called pro-PO system (Ashida and *To whom all correspondence should be addressed. Abbreviations--SDS, sodium dodecyl sulphate; PAGE, polyacrylamide gel electrophoresis; DTT, dithiothreitol; kDa, kilodaltons; L-dopa, L-dihydroxyphenylalanine, rm, relative mobility; PTU, phenylthiourea.

Ohnishi, 1967; Ashida, 1971; Zlotkin et al., 1973; Ashida and S6derhiill, 1984; Leonard et al., 1985; Christensen and Tracy, 1989, Johansson and S6derh/ill, 1989). This enzyme is involved in three main immunity-related processes: wound healing, humoral encapsulations and melanization of nodules in cellmediated encapsulation of foreign particles (for reviews see Nappi, 1975 and Lackie, 1980). The prophenoloxidase-activating system is a complex cascade of enzymes and co-factors inducing melanin synthesis by the host (S6derh/ill and Smith, 1986a). In many arthropods the first activation step of the pro-PO system is a limited proteolysis of the precursor prophenoloxidase. The system seems to be triggered physiologically by various factors, including lipopolysaccharides (LPS), fl-l,3-glucans and low Ca 2+ levels (S6derh/ill and Smith, 1986a, Beckage et al., 1987) or by organic solvents, proteolytic enzymes or temperature changes (Ashida and S6derh~ill, 1984). Preliminary evidence for the presence of a pro-PO system in A. auricollis larvae was provided by melanization of the insect body wall after injury (Scariet al., 1990). In this study we analyzed the cell-free hemolymph protein patterns of Allogamus auricollis larvae biochemically. In detail, we looked for a pro-PO system by non-denaturing electrophoresis and spectrophotometric analysis of phenoloxidase activity. MATERIALS AND METHODS

Chemicals All chemicals were of analytical grade and they were purchased from Sigma Chemical Co. (St Louis, MO); Bio Pad Laboratories (Richmond, CA) or Boehringer Mannheim Biochemica GmbH (Germany). Animals and sample collection Caddis fly larvae were collected from underneath large stones in the Serio fiver at the third collecting station (Valvassori et al., 1988). No morphometric analysis of the larvae was done and they were not assigned to any specific stage of development at that time (Scariet aL, manuscript in preparation). Identification of caddis fly larvae as Alloga-

867

868

M.F. BRIVIOet al.

mus auricollis was kindly done by G. P. Moretti. A. auricollis larvae were kept in tanks at 17°C and fed lettuce leaves. Hemolymph samples from a laboratory strain of A. auricollis larvae were collected under CO 2 anesthesia by .aspiration with siliconized sterile glass capillaries. From each larva about 5 #1 hemolymph was collected. Collected samples were maintained at 0-4°C to avoid auto-oxidation. Whole hemolymph was centrifuged in a microfuge (HERMLE Z231 M, setting No. 3) for 5min to obtain cell-free fractions. Samples were finally stored at - 2 0 ° C or immediately analyzed.

---)

SDS-PAGE Monodimensional electrophoresis was carried out as described by Laemmli (1970) with slight modifications. The resolving gel (1 mm thick) for analytical purposes was 10% acrylamide gel. Hemolymph samples were denatured by heating (100°C) with 1 vol of 2 × sample buffer (0.02 M sodium phosphate buffer, pH 7, 2% SDS, 0.2 M DTT, 20% glycerol, 0.002% Bromophenol Blue) for 5 min, loaded onto the gels and run overnight at 50 V (constant voltage) in a PROTEAN II CELL (Bio Rad). Molecular weights were determined by concurrently running molecular weight standards (from Bio Rad) and by estimation according to Weber and Osborne (1969). After separation, the gels were stained with Coomassie Blue R-250. Gel patterns were scanned by an HP ScanJet plus scanner and the images were analyzed by a custom program on a 80386 computer (Olivetti). Non-denaturing PAGE Native electrophoresis was carried out by a slight modification of the technique of Bidochka et al. (1989). The gel was polymerized in two layers: stacking gel---4% acrylamide in 0.5 M Tris-HC1, pH 8.8; separating gel--7.5% acrylamide in the same buffer. Aliquots of cell-free hemolymph in I vol of non-denaturing sample buffer (0.25 M Tris-HCl, pH 6.8, 1 M sucrose, 0.001% Bromophenol Blue) were applied directly to the gels. Electrophoresis was carried out at 50V (costant voltage) until the marker dye reached the bottom of the gel. Staining for enzymatic activity After the native PAGE, gels were stained for in situ phenoloxidase activity. The proenzyme was activated by soaking the gels in 50% 2-propanol in 0.1 M potassium phosphate buffer, pH 6.3, at 4°C for 2 hr. After several rinses in bidistilled water, gels were incubated overnight in 0.002 M L-dihydroxyphenylalanine (L-dopa) in 0.1 M potassium phosphate buffer, pH 6.3, at 37°C; and also in L-dopa buffer, pH 7.3. Assay of phenoloxidase activity Aliquots of cell-free hemolymph (5-10#1) were added to 1.5ml 0.004M L-dopa in 0.01 M potassium phosphate buffer, pH 7.3. The reaction was initiated by the addition of the sample to the L-dopa buffer and the changes in absorbance at 490 nm (A490nm/min)at 20°C were recorded for 90min by a KONTRON UVIKON 810P spectrophotometer. Phenylthiourea (PTU) was tested as phenoloxidase activity inhibitor; 100 #1 of 0.1% PTU in potassium phosphate buffer (0.01 M, pH 7.3) were added to 1.5 ml of the reaction mixture a few minutes after the start of the reaction and the absorbance was read as described above.

w A

.31 B

Fig. I. SDS-PAGE (10% acrylamide) of Allogamus auricollis cell-free hemolymph. Samples were denatured by addition of an equal volume of 2 × sample buffer (see Materials and Methods) followed by heating at 100°C for 5-10 min. Thirty microlitres (15 pl of cell-free hemolymph) was loaded, separated overnight at 50 V (constant voltage) and finally stained with Coomassie Blue R250, lane A. Two main bands are detectable (full arrowheads) of about 83 and 76kDa; other-lesser components, ranging from 50 to 170kDa, are also present. Lane B: calibration molecular weight markers (from Bio Rad).

showed the presence of two m a i n bands, with relative mobilities o f 83 a n d 76 kDa. M i n o r c o m p o n e n t s with molecular weights ranging from 50 k D a to 170 k D a were also seen (Fig. 1, lane A). The two main c o m p o n e n t s (Fig. 2, densitometric scan, peaks a a n d b) seem to account quantitatively

RESULTS

SDS-PAGE The h u m o r a l response involves a pool of molecular c o m p o n e n t s present in the h e m o l y m p h . However, analysis o f the cell-free insect h e m o l y m p h fluid by analytical S D S - P A G E a n d Coomassie Blue staining

rrn ~

all~

kDa

Fig. 2. Densitometric scan of A. auricollis hemolymph SDS-PAGE pattern. Peaks a and b represent the two main components of the cell-free plasma.

869

Allogamus auricollis phenoloxidase Phenoloxidase activity

I 1B

!ii!ii~~¸¸¸¸¸~i

Fig. 3. Non-denaturing electrophoresis (native PAGE) of A. auricollis cell-free hemolymph. Samples were diluted with a

volume of non-denaturing sample buffer and loaded directly onto the gel. Electrophoresis was performed at constant voltage and stained either for total protein content (Coomassie Blue) or enzymatic activity (L-dopa). Lane A: Coomassie blue staining, a single band (arrow) is visible; plus high molecular weight components at the top of the separating gel (full arrowhead), Lane B: only one band (arrow) was positive for phenoloxidase activity after enzymatic staining.

0.~................. iI1 l A (490nm)/5 min

Ol~.

I

I

I

0

10

20

30

L

I

I

40 50 60 Time (min)

I

I

I

70

80

90

Fig. 4. Time course of L-dopa oxidation activity in cell-free hemolymph of A. auricollis. Five microlitres of sample were incubated with 0.04 M L-dopa (in 0.01 M K-phosphate buffer, pH 7.3). Absorbance (490 nm) was recorded every 5 min. prophenoloxidase shows "self-activation" phenomena. After incubation of phenoloxidase containing supernatant from centrifuged Allogamus hemolymph with L-dopa (0.04 M), there was a typical progressive curve. Up to 5#1 of sample, the relationship for substrate conversion vs amount of enzyme was linear. The spectrophotometric analysis at 490 nm showed markedly increasing absorbance up to 60min of reaction, after which the curve became a plateau (Fig. 4). Pbenylthiourea (PTU), a common phenoloxidase inhibitor, in the reaction mixture immediately inhibited substrate (L-dopa) oxidation: the increase in absorbance stopped and the curve plateaued. (Fig. 5; PTU, t = 7 min). When PTU was added at the start of the reaction, no increase in absorbance was detectable (Fig. 5; PTU, t = 0).

DISCUSSION

for about 70-80 per cent of the total protein content in the cell-free supernatant of Allogamus hemolymph. Non-denaturing P A G E

We performed non-denaturing electrophoresis on polyacrylamide gels to localize the phenoloxidase activity in Trichoptera cell-free hemolymph. Electrophoresis of samples under native conditions shows a simple pattern with a single broad band (Fig. 3, lane A, arrow). After overnight treatment of the gels in L-dopa (in 100mM K-phosphate buffer, pH 6.3), the in situ phenoloxidase activity was localized as a single band (Fig. 3, lane B, arrow). No activity was observed when staining was done at a more basic pH (L-dopa in K-phosphate buffer, pH 7.3). High molecular weight species, or aggregates, can also be seen at the top of the resolving gel (Fig. 3, lane A, full arrowhead), but they were not positive for phenoloxidase activity. Phenoloxidase activity

When collected as described above (low ionic strength without protease inhibitors), hemolymph

In insects, the pro-PO system is one of the most important immunity-related factors, normally present in the hemolymph and hemocytes (Gftz and Boman, 1985). Usually, invertebrates possess this PTU inhibition 0.5 A (490 nm)/5 rain

0.25

~

0

~

10

Xll

IX

20

~

X

l

~

l

~

l

~

J

X

l

~

l

x J

I ~

~ 1

~l

X ~

~

30

40 50 60 70 80 90 -time (min) Fig. 5. Inhibition of .4. auricollis phenoloxidase activity by

PTU (phenylthiourea). 0.1% PTU (final concentration) was added 7 rain after the start of the reaction (I-1)and at to (*).

870

M.F. Bmvlo et al.

enzymatic system, although some arthropod groups seem to lack it (Sfderh/ill and Smith, 1986b). Phenoloxidase and endogenous lectins together are the "key components" responsible for the agglutination and encapsulation of invading organisms and microorganisms (Amirante and Basso, 1984, Boman and Hultmark, 1987), although the enzyme itself is able to induce a humoral response by melanization (Taylor, 1969). In arthropods, phenoloxidase is the final component of the pro-PO system and it is present as a zymogen (inactive precursor) mainly stored in cytoplasmic hemocyte vesicles (Leonard et al., 1985). Activation seems to be mediated by serine proteases that cleave the precursor to generate the lower molecular weight-activated phenoloxidase (Ashida et al., 1974; Ashida and Dohke, 1980). The natural prophenoloxidase cascade activators, such as LPS (lipopolysaccharides from bacterial cell walls) and fl-l,3-glucans (from fungal cells), act on the serine proteases (Ashida and Yoshida, 1988). Other components that act as physiological inhibitors, presumably on the proteases, prevent undesired activation of the pro-PO cascade (Sugumaran et al., 1985; Saul and Sugumaran, 1986; Hergenhahn et al., 1988). Ashida and Dohke (1980) showed that the Bombyx mori prophenoloxidase has a molecular weight of 80 kDa, is dimeric, and loses a 5 kDa peptide from each subunit after the proteolytic activation. Biochemical analysis of the cell-free Allogamus hemolymph showed an electrophoretic pattern under reducing conditions with two large bands (83 and 76 kDa). This S D S - P A G E pattern is similar to that described by Ashida and Dohke (1980) for the silkworm prophenoloxidase system. To confirm that there is a pro-PO system in A. auricollis, we carried out some enzymatic assays. Evidence of a phenoloxidase activity in the plasma supernatant was obtained by in situ localization after L-dopa native PAGE or by spectrophotometric analysis. As further confirmation, the colorimetric reaction was completely stopped by addition of the specific inhibitor phenylthiourea. In contrast with what has been reported for other insects, Allogamus auricollis hemolymph extracts required no further activation, other than the standard procedures of collection by abdominal injury and bleeding. Under these conditions, both the hemocytes (data not shown) and cell-free fractions of the hemolymph showed phenoloxidase activity, oxidizing the chromogenic substrate L-dopa. The present data, particularly the absence of any induced activation, suggest that the A. auricollis pro-PO cascade is modulated by an alternative pathway, perhaps similar to that described by Srderh/ill and Smith (1986a) in crustaceans. In crustaceans, more than one enzyme is involved in prophenoloxidase activation. A serine protease called "serine protease s" seems to be activated spontaneously by low salt concentration, especially by low Ca 2÷ ion concentrations ( < 5 mM). In a previous work (Valvassori et al., 1988) it was observed that A. auricollis larvae are naturally parasitized by the horsehair worm Gordius villoti. After entry, the parasitoids develop in the hemocoelic

cavity and finally kill the host by perforating the body wall. Our data about the presence a the pro-PO system in the hemocoel of caddis fly larvae suggest that Nematomorpha have developed an immuno-evasion system which makes the parasite able to inactivate the host defences. During the earliest Gordius development stages, a PAS-positive envelope entirely surrounds the worm body. This acellular coat could act as a mechanical and/or molecular barrier against the host recognition processes, protecting the worm by a self-simulation strategy. In order to clarify whether or not parasitization depresses the pro-PO system in vivo, we are currently doing induction studies giving activator compounds directly to the host after infestation. Future studies, with isolation and biochemical characterization of the Gordius envelope, will provide tools to further investigate the modulation of the immuno-response in this parasite-host model. Acknowledgements--We thank G. BottS. for his technical

assistance in computing the densitometric scan and M. Bondi and G. Amirante for reviewing the manuscript. This work was supported by a grant of MURST.

REFERENCES

Amirante G. A. (1986) Cellular immune responses in crustaceans. In Hemocytic and Humoral Immunity in Arthropods (Edited by Gupta A. P.), pp. 61-75. Wiley, New York. Amirante G. A. and Basso V. (1984) Analytical study of lectins in Squilla mantis L. (Crustacea, Stomatopoda) using monoclonal antibodies. Devl comp. Immunol. 8, 721-726. Anderson R. S., Holmes B. and Good R. A. (1973) In vitro bactericidal capacity of Blaberus craniifer hemocytes. J. Invertebr. PathoL 22, 127-135. Ashida M. (1971) Purification and characterization of prephenoloxidase from hemolymph of the silkworm Bombix mori. Archs Biochem. Biophys. 144, 749-762. Ashida M. and Ohnishi E. (1967) Activation of prophenoloxidase in hemolymph of the silkworm Bombix mori. Archs Biochem. Biophys. 122, 411-416. Ashida M. and Dohke K. (1980) Activation of the pro-phenoloxidase by the activating enzyme of the silkworm Bombix mori. Insect Bochem. 10, 37-47. Ashida M. and S6derh/ill K. (1984) The prophenoloxidase activating system in crayfish. Comp. Biochem. Physiol. 77B, 21-26.

Ashida M. and Yoshida H. (1988) Limited proteolysis of prophenoloxidase during activation by microbial products in insect plasma and effect of phenoloxidase on electrophoretic mobilities of plasma proteins. Insect Biochem. 18, 11-19. Ashida M,, Dohke K and Ohnishi E. (1974) Activation of pre-phenoloxidase. III. Release of a peptide from pre-phenoloxidase by the activating enzyme. Biochem. biophys. Res. Commun. 57, 1089-1095. Beckage N. E., Templeton T. J., Nielsen B., Cook D. I. and Stoltz D. B. (1987) Parasitism-induced haemolymph polypeptides in Manduca sexta (L.) larvae parasitized by the braconid wasp Cotesia congregata (Say). Insect Biochem. 17, 439-454. Bidochka M. J., Gillespie J. P. and Khachatourians G. C. (1989). Phenoloxidase activity of acridid grasshoppers from the subfamiles Melanoplinae and Oedipodinae. Comp. Biochem. Physiol. 9413, 117-124.

Allogamus auricollis phenoloxidase Boman H. G. (1981) Insect responses to microbial infection. In Microbial Control of lnsects, Mites and Plant Diseases (Edited by Burges D.), pp. 769-784. Academic Press, New York. Boman H. G, and Hultmark D. (1981) Cell-free immunity in insects. Trends Biochem. Sci. 6, 306-309. Boman H. G. and Hultmark D. (1987) Cell-free immunity in insects. A. Rev. Microbiol. 41, 103-126. Breh61in M. and Zachary D. (1986) Insect haemocytes: a new classification to rule out the controversy, In Immunity in Invertebrates (Edited by Breh61in M.), pp. 36~8. Springer, Berlin. Chadwick J. S. and Aston P. (1979) An overview of insect immunity. In Animal Models of Comparative and Developmental Aspects of lmmunity and Disease (Edited by Gcrshwin M. E. and Cooper E. L.), pp. 1-14. Pergamon Press, New York. Christensen B. M. and Tracy J. W. (1989) Arthropod-transmitted parasites: mechanisms of immune interaction. Am. Zool. 29, 387-398. Faye I. (1978) Insect immunity. IV. Early fate of bacteria injected in Saturniid pupae. J. Invertebr. Pathol. 31, 19 -26.

G6tz P. (1986) Encapsulation in arthropods. In Immunity in Invertebrates (Edited by Breh61in M.), pp. 153-170. Springer, Berlin. G6tz P. and Boman H. G. (1985) Insect immunity. In Comprehensive Insect Physiology, Biochemistry and Pharmacology (Edited by Kerkut G. A. and Gilbert L. I.), Vol. 3, pp. 453-485. Pergamon Press, Oxford. Grimstonc S. R., Rotheram S. and Salt G. (1967) An electron-microscope study of capsule formation by insect blood cells. J. Cell Sci. 2, 281-292. Gupta A. P. (1979) Hemocyte types: their structures, synonymies, inter-relationships and taxonomic significance. In Insect Hemocytes (Edited by Gupta A. P.), pp. 85-128. Cambridge University Press, Cambridge. Hergenhahn H. G., Hall M. and S6derh/ill K. (1988) Purification and characterization of an ~2macroglobulin-like proteinase inhibitor from plasma of the crayfish Pacifastacus leniusculus. Biochem. J. 255, 801-806. Hultmark D., Steiner H., Rasmuson T. and Boman H. G. (1980) Insect immunity. Purification and properties of three inducible bactericidal proteins from hemolymph of immunized pupae of Hyalophora cecropia. Eur. J. Biochem. 106, 7-16. Johansson M. W. and S6derh/ill K. (1989) Cellular immunity in crustaceans and the pro-PO system. Parasitol. Today, 5, 171-176. Lackie A. M. (1980) Invertebrate immunity. Parasitology 80, 393-412. Laemmli U. K. (1970) Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature, Lond. 227, 680~585.

871

Leonard C., S6derh/ill K. and Ratcliffe N. A. (1985) Studies on prophenoloxidase and protease activity of Blaberus craniifer haemocytes. Insect Biochem. 15, 803-810. Nappi A. J. (1975) Parasite encapsulation in insects. In Invertebrate Immunity (Edited by Maramorosch K. and Shope R. E.), pp. 293-326. Academic Press, New York. Pye A. E. and Boman H. G. (1977) Insect immunity. III. Purification and partial characterization of immune protein P5 from hemolymph of Hyalophora cecropia pupae. Infect. Immun. 17, 408-414. Rasmuson T. and Boman H. G. (1979) Insect immunity. V. Purification and some properties of immune protein P4 and haemolymph of Hyalophora cecropia pupae. Insect Biochem. 9, 259-264. Rather S. and Vinson S. B. 0983) Phagocytosis and encapsulation: cellular immune responses in Arthropoda. Am. Zool. 23, 185-194. Rowley A. F. and Ratcliffe N. A. (1981) Insects. In Invertebrate Blood Cells (Edited by Ratcliffe N. A. and Rowley A. F.), pp. 421-488. Academic Press, London. Saul S. J. and Sugumaran M. (1986) Protease inhibitor controls prophenoloxidase activation in Manduca sexta. FEBS 208, I13-I16. Scari G., Brivio M. F., Vailati G., Magnetti P., Pagani M., Paganoni M. and Valvassori R. (1990) Cellular defense reaction in larvae of Allogamus auricollis Pictet (Trichoptera), p. 251. Atti 53, Congresso UZI. S6derh/ill K. and Smith V. J. (1986a) The prophenoloxidase activating system: biochemistry of its activation and role in arthropod cellular immunity, with special reference to crustaceans. In Immunity in Invertebrates (Edited by Breh61in M.), pp. 208-223. Springer, Berlin. S6derh/ill K. and Smith V. J. (1986b) Prophenoloxidaseactivating cascade as a recognition and defense system in arthropods. In Hemocytic and Humoral Immunity in Arthropods (Edited by Gupta A. P.), pp. 251-285. Wiley, New York. Stairs G. R. (1964) Changes in susceptibility of Galleria melonella (Linnaeus) larvae to nuclear-polyhedrosis virus following blockage of the phagocytes with India ink. J. Insect Pathol. 6, 373-376. Sugnmaran M., Saul S. J. and Ramesh N. 0985) Endogenous protease inhibitor prevent undesired activation of prophenolase in insect hemolymph. Biochem. Biophys. Res. Commun. 132, 1124-1129. Taylor R. (1969) A suggested role for the polyphenol-phenoloxidase system in invertebrate immunity. J. Invertebr. Pathol. 14, 427-428. Valvassori R.; Scari G., De Eguileor M., Di Lernia L., Magnetti P. and Melone G. (1988) Gordius villoti (Nematomorpha) life cycle in relation with caddis fly larvae. Boll. Zool. 55, 269-278. Zlotkin E., Gurevitz M. and Shulov A. (1973) The toxic effect of phenoloxidase from the haemolymph of tenebrionid beetles. J. Insect Physiol. 19, 1057-1065.