- Email: [email protected]
Aspects of cuticular sclerotization in the locust, Scistocerca gregaria, and the beetle, Tenebrio molitor
ARTICLE IN PRESS Insect Biochemistry and Molecular Biology Insect Biochemistry and Molecular Biology 37 (2007) 223–234 www.elsevier.com/locate/ibmb
Aspects of cuticular sclerotization in the locust, Scistocerca gregaria, and the beetle, Tenebrio molitor Svend Olav Andersena,, Peter Roepstorffb a
Institute of Molecular Biology and Physiology, University of Copenhagen, The August Krogh Building, Universitetsparken 13, DK-2100 Copenhagen O, Denmark b Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark Received 16 September 2006; received in revised form 15 November 2006; accepted 16 November 2006
Abstract The number of reactive amino groups in cuticular proteins decreases during the early period of insect cuticular sclerotization, presumably due to reaction with oxidation products of N-acetyldopamine (NADA) and N-b-alanyldopamine (NBAD). We have quantitated the decrease in cuticular N-terminal amino groups and lysine e-amino groups during the first 24 h of sclerotization in adult locusts, Schistocerca gregaria, and in larval and adult beetles, Tenebrio molitor, as well as the increase in b-alanine amino groups in Tenebrio cuticle. The results indicate that nearly all glycine N-terminal groups and a significant part of the e-amino groups from lysine residues are involved in the sclerotization process in both locusts and Tenebrio. A pronounced increase in the amount of free b-alanine amino groups was observed in cuticle from adult Tenebrio and to a lesser extent also in Tenebrio larval cuticle, but from locust cuticle no b-alanine was obtained. Hydrolysis of sclerotized cuticles from locusts and Tenebrio by dilute hydrochloric acid released a large number of compounds containing amino acids linked to catecholic moieties. Products have been identified which contain histidine residues linked via their imidazole group to the b-position of various catechols, such as dopamine, 3,4-dihydroxyphenyl-ethanol (DOPET), and 3,4dihydroxyphenyl-acetaldehyde (DOPALD), and a ketocatecholic compound has also been identified composed of lysine linked via its e-amino group to the a-carbon atom of 3,4-dihydroxyacetophenone. Some of the hydrolysis products have previously been obtained from sclerotized pupal cuticle of Manduca sexta [Xu, R., Huang, X., Hopkins, T.L., Kramer, K.J., 1997. Catecholamine and histidyl protein cross-linked structures in sclerotized insect cuticle. Insect Biochemistry and Molecular Biology 27, 101–108; Kerwin, J.L., Turecek, F., Xu, R., Kramer, K.J., Hopkins, T.L., Gatlin, C.L., Yates, J.R., 1999. Mass spectrometric analysis of catechol-histidine adducts from insect cuticle. Analytical Biochemistry 268, 229–237; Kramer, K.J., Kanost, M.R., Hopkins, T.L., Jiang, H., Zhu, Y.C., Xu, R., Kerwin, J.L., Turecek, F., 2001. Oxidative conjugation of catechols with proteins in insect skeletal systems. Tetrahedron 57, 385–392], but the lysine-dihydroxyacetophenone compound and the histidine–DOPALD adduct have not been reported before. It is suggested that the compounds are derived from NADA and NBAD residues which were incorporated into the cuticle during sclerotization, and that the lysine-dihydroxyacetophenone as well as the DOPET and DOPALD containing adducts are degradation products derived from cross-links between the cuticular proteins, whereas the dopamine-containing adducts are derived from a noncrosslinking reaction product. r 2006 Elsevier Ltd. All rights reserved. Keywords: Insect; Cuticle; Sclerotization; Histidine; Lysine; N-terminal amino acids; Adduct formation; N-acetyldopamine; N-b-alanyldopamine
1. Introduction Soon after ecdysis the cuticles of insects undergo a number of changes and modifications, which together Corresponding author. Tel.: +45 3532 1703; fax: +45 3532 1567.
E-mail address: [email protected] (S.O. Andersen). 0965-1748/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.ibmb.2006.11.006
optimize the local physical properties of the cuticular regions. The modifications can be deposition of chitin and proteins to form an endocuticle, secretion of waxes unto the surface of the epicuticle (reviewed by Blomquist and Dillwith, 1985), incorporation of ortho-diphenols, such as N-acetyldopamine (NADA) and N-b-alanyldopamine
ARTICLE IN PRESS 224
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
(NBAD), into the cuticular matrix (sclerotization) (reviewed by Sugumaran, 1998; Andersen, 2005), formation and deposition of cuticular melanin (reviewed by Kayser, 1985), dehydration of the cuticular matrix (Hillerton and Vincent, 1979; Vincent and Ablett, 1987), and chlorination of tyrosine residues in proteins (Andersen, 2004). Not all of the changes occur in all types of cuticle; cuticles from different insect species and from different developmental stages and different regions of the same animal vary widely both in appearance and physical properties. The precise properties of a given piece of cuticle will depend upon the exact balance between the various modifications, and it is possible that several yet undescribed modifications also contribute to the final outcome. To obtain a more detailed understanding of the factors influencing cuticular mechanical properties, it is necessary to study the modifications in a given cuticular region, and to quantitatively determine how rapidly and to what extent they occur. Our present understanding of the sclerotization process gives a fairly coherent picture of the major chemical reactions involved (reviewed in Andersen, 2005), although many details are still controversial, and more detailed studies of the quantitative aspects are needed. Fig. 1 illustrates the pathway for forming reactive intermediates by oxidation of the sclerotization precursors, NADA and NBAD. During sclerotization NADA and NBAD are enzymatically oxidized to the corresponding orthoquinones, which can be isomerized to para-quinone methides. Both the o-quinones and the p-quinone methides can react with nucleophilic groups on the cuticular proteins, and the p-quinone methides may also be rearranged to 1,2-dehydro-N-acyldopamines. Oxidation products of the 1,2-dehydro-acyldopamines can form dihydroxyphenyl-dihydrobenzodioxine derivatives by
reaction with catechols (Andersen et al., 1996), but they may also react with other nucleophilic groups in the cuticles. As cuticular proteins are nearly devoid of the sulphurcontaining amino acids, cysteine and cystine, the imidazole group in histidine residues is the most important nucleophile for reaction with o-quinones and p-quinone methides during sclerotization, but free amino groups, such as N-terminal groups and e-amino groups of lysine, are also potential reactants. Adducts between NADA and N-acetylhistidine have been produced by in vitro incubation of the two compounds together with cell-free pieces of cuticle (Andersen et al., 1991, 1992a,b), and adduct IV between histidine and 3,4-dihydroxyphenyl-ethanol (DOPET, I, Fig. 2) and adduct VI (Fig. 3) between dopamine and histidine have been isolated from acid hydrolysates of sclerotized pupal cuticle of Manduca sexta (Kerwin et al., 1999; Kramer et al., 2001). Lysine e-amino groups and free N-terminal amino groups in the cuticular proteins have been proposed to take part in sclerotization (Hackman, 1971; Neville, 1975), but the evidence is less convincing than that for histidine involvement. The number of free amino groups which can be chemically modified decreases during the initial phase of sclerotization of locust cuticle (Andersen, 1972), indicating that they may be involved in the process. More precise and sensitive analytical methods are now available to follow the gradual disappearance of reactive amino groups, and it is possible to distinguish lysine amino groups from amino groups from N-termini and from b-alanine. It would be an advantage also to follow the disappearance of unmodified histidine residues during sclerotization. As we have not succeeded in this the investigation was extended to include isolation of adducts between amino acids and ortho-diphenols from hydrolysates of sclerotized cuticle, and the results confirm the involvement of both histidine and lysine residues in sclerotization. Two different
Fig. 1. The current scheme for sclerotization of insect cuticles. The acyldopamines NADA and NBAD are during sclerotization oxidized to o-quinones, which are rearranged to p-quinone methides. Both types of quinones can react with nucleophilic groups (HX) present in the cuticular proteins to give ringsubstituted adducts and sidechain-substituted adducts, respectively. The p-quinone methide can also be rearranged to a dehydro-derivative, which after oxidation to the quinone stage can react with available catechols to give dihydroxyphenyl–dihydrobenzodioxine.
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
225
frozen immediately at 18 1C or kept for one or more hours at 23 1C before being frozen. 2.2. Dinitrophenylation analysis Fig. 2. Some monomeric catechols produced by acid hydrolysis of sclerotized cuticle. I: 3,4-dihydroxyphenyl-ethanol (DOPET); II: 3,4dihydroxyphenyl-acetaldehyde (DOPALD); III: arterenone.
Fig. 3. Catecholic adducts obtained by acid hydrolysis of sclerotized cuticle. IV: histidine-DOPET adduct; V: histidine-DOPALD adduct; VI: histidine-dopamine adduct; VII: lysine-3,4-dihydroxyacetophenone adduct.
insect species were studied: a locust, Schistocerca gregaria, which as adult has a nearly colorless cuticle and uses NADA for sclerotization, and a beetle, Tenebrio molitor, which uses both NADA and NBAD for sclerotization and the adults have a dark brown cuticle. 2. Materials and methods 2.1. Rearing of Schistocerca and Tenebrio Desert locusts, S. gregaria, were reared in crowded stock cultures held at about 30 1C and light during daytime (8 h duration) and ambient temperature (about 23 1C) and darkness during night conditions (16 h duration). They were fed on wheat bran and fresh lettuce. Animals were collected when halfway through ecdysis from 5th stage nymphs to adults; they were then either immediately frozen at 18 1C or kept for one or more hours at 30 1C before being frozen. T. molitor larvae were kept at 23 1C and fed on wheat bran, carrots and apples. Large larvae were collected when halfway through ecdysis to the next larval instar, and pharate adults were collected before they had completely finished shedding the pupal cuticle; they were then either
Four cuticular regions in adult locusts were chosen for analysis: hindleg femur cuticle, hindleg tibial cuticle, ventral abdominal segments nos. 3–5, and dorsal abdominal segments nos. 3–5. The abdominal samples included the intersegmental cuticle connecting the three segments. The cuticular pieces were immersed in 1% potassium tetraborate (K-borate) to soften adhering tissues and facilitate cleaning. The cleaned pieces were washed briefly in distilled water and ethanol, air dried and weighed on a Sartorius microbalance, model MC 5. The samples were separately placed in 2 ml 4% NaHCO3 containing 1% dinitrofluorobenzene (DNFB) and stored in darkness overnight at ambient temperature. Next day they were washed briefly in distilled water, placed in 20 ml 40% ethanol for 24 h to wash out any unreacted DNFB present, air dried, reweighed, and hydrolyzed in 2.0 ml 6 M ultrapure HCl in sealed evacuated tubes at 105 1C for 20 h. The hydrolysates were taken to dryness in vacuo over solid NaOH, redissolved in 1.00 ml distilled water, centrifuged for 30 min at 6000 rpm, and 10 ml aliquots were taken for reversed phase high-performance liquid chromatography (RP-HPLC) analysis. Ventral abdominal cuticle from adults and dorsal plus ventral abdominal cuticle from last instar larvae of T. molitor were analyzed in a similar way. Zero hour cuticle was obtained from animals, which were halfway through the process of ecdysis, and the ages of later cuticular samples were measured from the moment the animals escaped fully from the old cuticle. The Tenebrio samples were cleaned, washed, dried and weighed as described above for locust cuticle. Dinitrophenylation and hydrolysis of samples were also performed as described above with the single difference that only half the volumes were used of buffer, DNFB, and hydrochloric acid. The DNP-derivatives were separated on a Lichrospher 100 RP-18 column, 4.6 250 mm (Merck, Darmstadt, Germany). The A-buffer was 0.1% trifluoroacetic acid (TFA) and the B-buffer was 90% acetonitrile. The column was equilibrated with 25% B-buffer and eluted with a linear gradient from 25% to 50% B-buffer in 36 min. The absorbancy of the eluate was measured from 200 to 400 nm by means of a Dionex diode array system (UVD340U UV/ VIS detector), and the resulting peaks were integrated at 350 nm by means of the Chromeleon Management System. The elution scheme does not separate all DNP-amino acids, but gives a satisfactory separation of the DNPderivatives relevant in this study. The identities of the observed peaks (e-DNP-lysine, DNP-glycine, DNPb-alanine, O-DNP-tyrosine, and dinitroaniline) were confirmed by co-chromatography with authentic standards both in the above system and on a polystyrene column (Source 5RPC, Amersham Biosciences, Uppsala, Sweden).
ARTICLE IN PRESS 226
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
2.3. Purification of modified dopamine residues from hydrolysis of sclerotized cuticle Sclerotized cuticles were purified from light-brown adult T. molitor collected less than 8 h after ecdysis and from adult S. gregaria collected either less than 24 h or 5–6 weeks after ecdysis. The frozen animals were homogenized in 1% K-borate, and the insoluble cuticular residues were washed repeatedly in copious amounts of K-borate, distilled water, and 96% ethanol. Cuticular samples (250–500 mg) were hydrolyzed in 10–20 ml 1 M HCl at 100 1C for 8 h, and after removing the insoluble residue by filtration, the soluble fraction was taken to dryness in vacuo over solid NaOH, dissolved in 3 ml 0.1% TFA, and applied to a cation exchange column (SP-Sepharose, 7 1.0 cm, Amersham Biosciences) in equilibrium with 0.1% TFA. The column was washed with 0.1% TFA until the ultraviolet absorbance at 280 nm of the eluate had decreased to near the baseline, and a linear gradient of NaCl in 0.1% TFA was then applied to elute the material retained on the column. The UV-absorbance of the eluate at 280 nm was recorded, and the fractions which were eluted later than arterenone (III, Fig. 2) were collected and screened by RP-HPLC for compounds of interest. 2.4. RP-HPLC-conditions The eluates collected after cation chromatography were fractionated on a Source 5RPC column (Amersham Biosciences) and eluted with a linear gradient in acetonitrile (from 2% to 70% acetonitrile in 0.1% TFA for 35 min). The peaks, which eluted early and had UV-absorbtion maxima near 280 nm were collected and subjected to further purification on a Lichrospher 100 RP-18 column (4.6 250 mm, Merck, Darmstadt, Germany). The ultraviolet absorbtion was recorded by the Dionex diode array system, and the major peaks showing UV-absorbtion spectra similar to that of dopamine were collected for further characterization. 2.5. Synthesis of 3,4-dihydroxyacetophenone-lysine adduct 3,4-dihydroxyphenacyl chloride (1 mmole, 186 mg) and a-N-acetyllysine (1 mmole, 188 mg) were dissolved in 10 ml 15% K-borate and left over night at ambient temperature. Acetic acid was added to adjust pH to 5, and after removal of the precipitated boric acid by centrifugation the reaction products were purified by RP-HPLC and cation exchange as described above for purification of the hydrolysis poducts from sclerotized cuticle. The main product was then hydrolyzed in 1 M HCl at 100 1C for 5 h and repurified by RP-HPLC on a Lichrospher 100 RP-18 column. 2.6. Mass spectrometry Molecular weight determination of the components in the different fractions collected after RP-HPLC was
Fig. 4. Hydrolysate of dinitrophenylated 0-h locust femur cuticle separated by RP-HPLC on a Lichrospher RP-18 column. 1: e-DNPlysine; 2: DNP-glycine; 3: dinitroaniline; 4: O-DNP-tyrosine.
performed by matrix assisted laser desorption/ionization mass spectrometry (MALDI-MS) using an ABI STR MALDI mass spectrometer (Applied Biosystems, Framingham, Ma) in reflecting mode. 0.5 ml of the eluate was mixed with 0.5 ml matrix solution (2,5-dihydroxybenzoic acid (DHB) 20 mg/ml in 70% acetonitrile, 0.1% TFA) on the target and dried and spectra recorded for an appropriate number of laser shots to obtain a good S/N. Structural information on the compounds was obtained by collision induced dissociation on a 4700 TOF/TOF mass spectrometer (Applied Biosystems, Farmingham, Ma) after isolation of the appropriate m/z range. Sample preparation was as above except that the DHB matrix solution was replaced with a solution of a-cyano-4-hydroxy cinnamic acid (5 mg/ml in 70% acetonitrile, 0.1% TFA). 3. Results 3.1. Dinitrophenylation studies Fig. 4 shows separation on a Lichrospher column of the hydrolysis products from a dinitrophenylated 0-h femur from the locust S. gregaria. The chromatogram is dominated by the peak of e-DNP-lysine, and the smaller later-eluting peaks have been identified as DNP-glycine, dinitroaniline and O-DNP-tyrosine, respectively. Fig. 5 shows the decrease in amounts of e-DNP-lysine and DNPglycine during the early phase of locust cuticular sclerotization; DNP-glycine decreased to nearly zero during the first 4 h after ecdysis, and DNP-lysine decreased during the same period to about 50% of its initial value and stayed thereafter nearly constant. Similar results were obtained for cuticle from tibia and dorsal and ventral abdomen of locusts. Fig. 6 shows that the amount of DNP-glycine recovered from dinitrophenylated larval abdominal cuticle of T. molitor decreased to low values during the first few hours after ecdysis, that a somewhat slower decrease to
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
227
Fig. 5. Yields of e-DNP-lysine ( ) and DNP-glycine (K) obtained from locust femur cuticle during the first 24 h after ecdysis. Fig. 7. Yields of e-DNP-lysine ( ), DNP-glycine (K), and DNP-balanine (m) obtained from ventral abdominal cuticle from adult Tenebrio during the first 24 h after ecdysis. Table 1 Amino acid analyses of adult S. gregaria femur cuticle sclerotized for 0 and 3 h after ecdysis (the results with standard errors are shown as residues per 100 residues; 5 analyses per sample)
Fig. 6. Yields of e-DNP-lysine ( ), DNP-glycine (K), and DNP-balanine (m) obtained from larval Tenebrio cuticle during the first 24 h after ecdysis.
about 70% of the initial value was observed for the e-amino group of lysine, and that DNP-b-alanine was obtained in trace amounts of from 0-h cuticle and in slightly larger amounts from older cuticle. Fig. 7 shows that the decrease in amounts of e-DNP-lysine in the ventral abdominal cuticle of adult T. molitor was pronounced and continued for a longer period (at least 24 h) than observed for larval cuticle. A low, but significant yield of dinitrophenylated b-alanine was obtained from 0-h beetles, and
Amino acid
0 h femur
3 h femur
Aspartic acid Threonine Serine Glutamic acid Glycine Alanine Cysteine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Proline
4.1070.18 2.5570.10 4.7770.04 3.8870.15 7.8370.15 33.8270.44 0 7.5270.16 0 3.7070.05 4.6670.06 8.3370.22 0.7870.03 2.0670.03 1.7470.12 3.8370.02 10.4170.09
3.8870.06 2.4170.05 4.6570.06 3.5670.06 7.8270.12 35.2870.33 0 7.6970.06 0 3.9270.08 4.8670.05 7.2370.09 0.7170.01 2.1270.03 1.4970.05 3.7070.05 10.6870.10
after a delay of 2 h a considerable increase in yield of DNPb-alanine was observed. 3.2. Amino acid analysis The amino acid compositions of 0- and 3-h femur cuticle from S. gregaria were determined and compared (Table 1).
ARTICLE IN PRESS 228
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
No significant changes were observed in the content of most amino acids, indicating that deposition of endocuticle does not yet play a role; the most significant change was in the content of lysine, which decreased by 14% from 1.74 to 1.49 residue percent, which is markedly lower than the decrease of ca. 50% in reactive lysine residues observed in the dinitrophenylation experiments. 3.3. Purification of catecholic adducts Sclerotized cuticles from S. gregaria and T. molitor adults were subjected to partial hydrolysis in 1 M HCl, and a number of compounds containing catechol moieties and carrying more than one positive charge were purified from the hydrolysates. Figs. 8A and B show the elution patterns obtained by cation-exchange of hydrolysates of cuticles from 5–6 week old adult S. gregaria and adult Tenebrio, which was sampled less than 8 h after ecdysis, respectively. The peak at 135 ml was identified as arterenone (III), and the later eluting fractions (a and b) indicated by horizontal bars were collected and purified by RP-HPLC. The compounds eluted between 155 and 190 ml (fraction a) in Fig. 8A can be assumed to carry two positive charges at pH 2; they were purified on a Source 5RPC column (Fig. 9A) and the peak (Sg-a1) eluted at 5.5–6.0 min was further purified on a Lichrospher column (Fig. 9B). The ultraviolet spectrum of the compound eluted at 4.4 min in Fig. 9B (Sg-a1-1) corresponded to that of dopamine, indicating that the compound is a catechol, and mass spectrometry showed that the compound has an m/zvalue of 306.1. After 10–20 min exposure to alkaline
conditions (1% sodium carbonate) in the presence of oxygen, the compound was degraded to 3,4-dihydroxyphenyl-glyoxal and histidine. Reduction of the compound with sodium borohydride resulted in a product which eluted later from the HPLC-columns than did the unreduced compound, and the reduction resulted in an increase in m/z-value to 308.1 mass units. When subjected to MS/MS both the reduced and the unreduced compound gave a peak at m/z 156.4, corresponding to histidine, and the unreduced compound gave also a peak at m/z 260, corresponding to a loss of H2O+CO. Such a loss of 46 mass units indicates that histidine is connected to the sidechain of the catechol moiety via its Nt (N 1) position (Turecek et al., 1998). The combined results indicate that compound Sg-a1-1 is an adduct (V) between histidine and 3,4-dihydroxyphenyl-acetaldehyde (DOPALD, II), and that the reduced compound is the corresponding adduct (IV) between histidine and DOPET (I). A compound with an ultraviolet spectrum similar to that reported for ketocatecholic compounds (Andersen and Barrett, 1971) was purified from the peak Sg-a2 eluted at 8 min in Fig. 9A. Treating the compound with sodium borohydride gave a reduction product having a typical catecholic ultraviolet spectrum, and the reduction caused an increase in molecular mass from an m/z-value of 297.1 to 299.1. The compound was subjected to MS/MS fragmentation, and fragment peaks were observed at m/z-values of 84.1 and 130.1, indicating the presence of a lysine residue, and at 137.1, indicating the presence of a keto-group joined to the catechol ring (Andersen and Roepstorff, 1978; Roepstorff and Andersen, 1980).
Fig. 8. Cation exchange chromatography of acid hydrolysates a (A) 500 mg cuticle from mature locusts and (B) 500 mg cuticle from adult Tenebrio less than 8 h after ecdysis. The peaks containing arterenone are marked with *, and the fractions marked by horizontal bars were collected for further purification.
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
229
Fig. 9. (A) Reverse phase chromatography on Source 5RPC of the material eluted between 155 and 190 ml (Sg-a) in Fig. 8A (locust cuticular hydrolysate). The eluates marked with bars (Sg-a1 and Sg-a2) were collected for further purification. (B) Purification of peak at 5.5–6.0 min (Sg-a1) in Fig. 9A on a Lichrospher column; the marked peak was collected for characterization.
The results suggest that the compound is an adduct between lysine and 3,4-dihydroxyacetophenone (structure VII). The compound was also obtained from hydrolysates of sclerotized cuticle from adult T. molitor, although in lesser amounts than from locust cuticle. Compound VII was synthesized by reaction between 3,4-dihydroxyphenacyl chloride and a-N-acetyllysine followed by acid hydrolysis of the main product. The purified hydrolysis product was indistinguishable from the compound obtained from hydrolysates of locust sclerotized cuticle both with regard to migration on RP-HPLC columns, ultraviolet absorbtion spectrum and mass spectrum, indicating the correctness of structure VII. Figs. 10A and B show purification of the Tenebrio compounds (Tm-a) which eluted between 155 and 190 ml in Fig. 8B. When subjected to chromatography on a Lichrospher column the broad peak (Tm-a1) at about 5 min in Fig. 10A was resolved in a number of peaks as shown in Fig. 10B; they all contained catecholic compounds according to their ultraviolet spectra, but so far only a few have been identified. The mass spectra obtained from the peaks indicated that most of them contain several compounds; the mass spectrum of the peak at 5.5 min (Tm-a1-1) in Fig. 10B thus gave several major peaks corresponding to m/z-values of 156.4, 213.4, 308.4, and 395.3 mass units. The m/z-values of 156.4 and 213.4 correspond to those expected for histidine and glycyl-histidine (or histidylglycine), respectively, while a mass of 308.4 indicates the presence of an adduct (IV) between DOPET and histidine. The composition of the adduct was confirmed by MS/MS, which showed a fragmentation pattern dominated by a peak of m/z 156.4, corresponding to histidine. The m/z-value of 395.3 could indicate an adduct between
DOPET and seryl-histidine, but its identity has not yet been confirmed. The mass spectrum of the double peak at 6.1 min (Tm-a1-2) in Fig. 10B showed the presence of a compound having an m/z-value of 365.3 mass units, corresponding to an adduct between DOPET and glycyl-histidine. This was confirmed by MS/MS, where a major fragmentation ion was present at m/z 213.4, and a loss of 46 mass units (H2O+CO) was observed, indicating that the dipeptide is glycylhistidine and not histidylglycine. The compound eluted at 5.5 min (Tm-a2) in Fig. 10A was identified as the same adduct (V) between histidine and DOPALD as found in hydrolysates of locust cuticle. The identification was based upon mass spectrum, degradation by treatment with alkali, reduction to a histidine–DOPET adduct with sodium borohydride, and co-chromatography with the locust compound. Fig. 11A and B show purification of the Tenebrio compounds (Tm-b) eluted between 220 and 250 ml in Fig. 8B. The resolution of the double peak (Tm-b1) in Fig. 11A was only slightly improved on the Lichrospher column (Fig. 11B). Mass spectrometry showed that the peak Tm-b1-1 at 4.2 min in Fig. 11B contained a compound having an m/z-value of 307.1 and that the peak Tm-b1-2 at 4.6 min contained a compound of m/z-value 364.2. The m/z-value of 307.1 agrees with an adduct between dopamine and histidine (VI), and the fragmentation pattern observed in MS/MS indicates that an imidazole nitrogen (Nt) from histidine is connected to the b-position of the dopamine sidechain. The m/z-value of 364.2 agrees with an adduct between dopamine and histidine plus glycine. MS/MS of the m/z 364.2 component indicated that a glycylhistidine moiety is connected to the b-position in dopamine side chain.
ARTICLE IN PRESS 230
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
Fig. 10. (A) Reverse phase chromatography on Source 5RPC of the material eluted between 155 and 190 ml (Tm-a) in Fig. 8B (Tenebrio cuticular hydrolysate). The eluate marked by a bar was collected for further purification. (B) Purification of peak at 4.5–5.5 min (Tm-a1) in Fig. 10A on a Lichrospher column; the marked peaks were collected for characterization.
Fig. 11. (A) Reverse phase chromatography on Source 5RPC of the material eluted between 220 and 250 ml (Tm-b) in Fig. 8B (Tenebrio cuticular hydrolysates). (B) Purification of peak at 4.0–4.5 (Tm-b1) in Fig. 11A on a Lichrospher column; the marked peaks were collected for characterization.
The dopamine–histidine adduct was also isolated from sclerotized locust cuticle, but in much smaller amounts than were obtained from Tenebrio cuticle. 4. Discussion An important step in the sclerotization of insect cuticle is the reaction of quinoid derivatives of various catechols with nucleophilic groups on the cuticular matrix proteins. The most important catecols involved in the process are
NADA and NBAD, but also DOPET has been proposed as a probable sclerotization compound (Kerwin et al., 1999; Kramer et al., 2001). The imidazole ring in proteinbound histidine residues seems to be the most important nucleophilic group, although also the free amino groups present in lysine residues and the N-terminal amino groups are likely to be involved. Amino acid sequence determination on cuticular proteins from pharate adult migratory locusts (Locusta migratoria) showed that the majority of the proteins have glycine as the N-terminal amino acid
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
(Andersen et al., 1995); until now 31 pharate locust cuticular proteins have been sequenced, and glycine is the N-terminal residue in 30 of them and the remaining protein has isoleucine as N-terminus (Klarskov et al., 1989). The amino acid sequences of the few cuticular proteins sequenced from S. gregaria are nearly identical to those from L. migratoria, and accordingly it can be expected that the majority of pharate cuticular proteins from S. gregaria also have glycine as N-terminal residue. The results obtained by dinitrophenylation of locust cuticle during the early period of the sclerotization process demonstrate that besides e-N-DNP-lysine only DNP-glycine is obtained in significant amounts, confirming the assumption that glycine is the dominating N-terminal amino acid in S. gregaria pharate cuticular proteins. The amino groups from lysine as well as from N-terminal glycine are rapidly rendered unavailable for dinitrophenylation: after 4 h sclerotization about 90% of the N-terminal glycine residues and about 50% of the lysyl e-amino groups have become unreactive. The contents of free N-terminal and lysine amino groups also decrease in larval and adult Tenebrio cuticles during the first day of sclerotization, but the decrease is significantly slower than in locust cuticle. Another difference between the cuticles of the two insect species is the presence of free amino groups of b-alanine in both larval and adult Tenebrio cuticles and the absence of this amino acid from locust cuticle. Small amounts of b-alanine are present in 0-h cuticle from Tenebrio larvae, and the amounts of increase slowly for at least 24 h, presumably due to incorporation of NBAD. Larger amounts of balanine are present in 0-h cuticle from adult Tenebrio, no increase in content of free b-alanine amino groups is observed during the first 2 h of sclerotization, then the content increases for some hours to decrease again between 9 and 24 h after ecdysis, presumably due to utilization of more amino groups for sclerotization than are provided by incorporation of NBAD. The disappearance of reactive amino groups is probably due to their reaction with sclerotization agents, but it could also be explained by assuming that the dinitrophenylation reagent has gradually increasing difficulties in getting access to the amino groups, for instance due to closer packing of the protein chains caused by dehydration of the cuticular matrix. During the first few hours after ecdysis the water content of locust cuticle decreases from 43% to 32% of the wet weight (Andersen, 1981), probably contributing to the increase in cuticular stiffness observed during the early phase of sclerotization (Vincent, 1980; Vincent and Ablett, 1987), but it is unlikely to contribute to the decreased reactivity of amino groups. Before analysis by dinitrophenylation all the cuticular samples were dried to constant weight and had experienced a water loss much greater than the small natural loss occurring soon after ecdysis, resulting in tightly packing of the protein chains both in the unsclerotized and the sclerotized samples.
231
Amino acid composition of unsclerotized locust femur cuticle compared to that of femur cuticle allowed to sclerotize for 3 h showed a decrease of 14% in lysine content and no significant decrease in content of histidine, indicating that if covalent bonds are formed between these two amino acids and catechol derivatives then the majority of the bonds must be acid labile, which is in contrast to the isolation of significant amounts of histidine–dopamine adducts from sclerotized pupal cuticle of M. sexta hydrolyzed in 6 M HCl (Kerwin et al., 1999; Kramer et al., 2001). The disappearance of reactive amino groups was determined only for the first 24 h after ecdysis, although cuticular sclerotization in both locusts and Tenebrio continues for a couple of weeks. Deposition of a considerable amount of endocuticle occurs during the whole sclerotization period, and much of this endocuticle is also subject to sclerotization but to a lesser degree than the exocuticle. During the early hours after ecdysis only little endocuticle is deposited and its contribution to the weight of the cuticular samples and to their lysine content will be insignificant. The results shown in Figs. 3–5 will, therefore, mainly represent modifications of exocuticular amino groups. As histidine has been shown to be involved in adduct formation with dopamine derivatives during sclerotization of Manduca pupal cuticle (Schaefer et al., 1987; Christensen et al., 1991; Xu et al., 1997; Kerwin et al., 1999; Kramer et al., 2001), it would be of interest to determine to what extent this amino acid is consumed during the sclerotization process. But so far it has not been possible to find a useful method for determining changes in the content of reactive histidine residues. Since histidine–dopamine and histidine–DOPET adducts have been obtained from acid hydrolysates of sclerotized Manduca pupal cuticle (Kerwin et al., 1999; Kramer et al., 2001), it is of interest to determine whether such acid-stable adducts also can be obtained from sclerotized locust and Tenebrio cuticle. To avoid the presence of low-molecular weight carbohydrates in the hydrolysates 1.0 M HCl was chosen for cuticular hydrolysis, as chitin is not significantly degraded at this acid strength. The cuticular proteins are probably not completely degraded to free amino acids, and a large number of small peptides will be formed when dilute acids are used for hydrolysis. Most of the amino acids and peptides will only have a single positive charge at pH 2, while histidine–dopamine adducts and lysine–dopamine adducts will carry three positive charges at low pH-values, and an adduct between dopamine and glycine will carry two positive charges. Cation exchange was accordingly chosen as the first step in the purification procedure, and components eluted later than arterenone were selected for characterization. Many compounds with ultraviolet spectra similar to those of dopamine or ketocatechols were present in this late region of the eluates, and most of them have not yet been identified. So far an adduct between dopamine and histidine (VI) has been identified from hydrolysates of
ARTICLE IN PRESS 232
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
Tenebrio cuticle and obtained in smaller amounts from locust cuticle. From Tenebrio cuticular hydrolysates two other compounds were isolated, which eluted from the cation exchange column shortly after arterenone, suggesting the presence of two positive charges; by mass spectrometry they were identified as adducts between histidine and DOPET (IV) and histidine and DOPALD (V), respectively. From hydrolysates of locust cuticle only V was obtained. Compound V is very unstable in alkaline solutions, and treatment with dilute Tris-base or sodium carbonate resulted in rapid degradation of the compound to 3,4-dihydroxyphenyl-glyoxal and histidine. Reduction of the aldehyde group in V gave an adduct between DOPET and histidine, identical to compound IV. Compounds IV and VI have previously been isolated from Manduca pupal cuticle after hydrolysis in 6 M HCl (Kerwin et al., 1999; Kramer et al., 2001), whereas the unstable compound V has not been reported before. Compounds IV and V can readily be converted to each other and it appears likely that they are degradation products of identical or very similar structures in the sclerotized cuticles. It was suggested that the histidine– DOPET adduct obtained from pupal cuticle of Manduca is a reaction product between histidine and the p-quinone methide of DOPET, indicating that DOPET as well as NADA and NBAD is utilized as a sclerotizing compound in this insect (Kerwin et al., 1999; Kramer et al., 2001).
Fig. 12. Suggested structures for two interchain cross-links in insect sclerotized cuticle.
The histidine–DOPET compound (IV) obtained from Tenebrio cuticle might have a similar origin, and the aldehyde adduct (V) could then be formed by oxidation of the DOPET adduct. We are in favour of explaining the aldehyde adduct (V) as a direct degradation product of a NADA-derived cuticular cross-link; the histidine–DOPET compound (IV) could then be formed by reduction of the aldehyde compound during hydrolysis of Tenebrio cuticle. The quinone formed by oxidation of dehydro-NADA reacts readily with the two phenolic groups in catechols to form benzodioxine derivatives (Fig. 1), and it will probably also react spontaneously with two closely positioned nucleophilic groups. If these groups are present in separate protein chains the result will be a covalent cross-link between these chains, as suggested previously (Andersen, 1990). Structures VIII and IX (Fig. 12) show two of several such possible cross-links. In structure VIII an imidazole group from histidine is linked to the b-position of the NADA side-chain and an amino group from lysine is linked to the a-position of the same NADA-unit, and lysine residues are linked to both the a- and the b-position of a NADA-unit in structure IX. Fig. 13 suggests that acid hydrolysis of compound VIII will remove the acetyl group, resulting in an intermediate, carrying a primary and a secondary amino group on the same carbon atom. This configuration is unstable, and the intermediate is likely to be further hydrolyzed to the aldehyde-containing histidineadduct (V), ammonia and lysine. Compound VII could possibly be produced by hydrolytic degradation of a cross-link, where lysine residues are linked to both a- and b-positions of a NADA residue (structure IX). Fig. 14 suggests that hydrolytic deacetylation of the NADA moiety in IX might be followed by loss of ammonia and formation of a Schiff base which by hydrolysis will give compound VII. The adducts between catechols and amino acids which have been obtained by acid hydrolysis of sclerotized insect cuticles are probably all derived from N-acyldopamines (NADA and NBAD), and they can be divided into two groups: those carrying an amino group on the a-carbon atom of the side-chain and those lacking this amino group. The adducts between dopamine and either histidine or glycyl-histidine belong to the former group, and they are assumed to be produced by reaction of protein-bound
Fig. 13. Suggested pathway for formation of adduct V by acid-catalyzed degradation of a cuticular cross-link in which a histidine and a lysine are linked to a NADA residue.
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
233
Fig. 14. Suggested pathway for formation of adduct VII by acid-catalyzed degradation of a cuticular cross-link in which two lysine residues are linked to a NADA residue.
histidine residues with the p-quinone imide of NADA. They are probably not derived from cross-links in the cuticles, whereas the adducts belonging to the latter group (IV and VII, and possibly V) are suggested to be derived from cuticular cross-links formed by reaction of oxidized dehydro-NADA with imidazole and/or amino groups in the cuticular proteins. In fully sclerotized cuticle both cross-linking and noncross-linking adducts will be involved in polymer formation due to continued addition of oxidized dehydro-NADA to available catecholic groups. It is our hope that further studies of the still unidentified compounds present in the cuticular hydrolysates will contribute to a more detailed understanding of the chemistry of sclerotization, and either confirm or modify the suggestions proposed here. Acknowledgments We are grateful for economic support from the Carlsberg Foundation, and we thank Lene Skou and Kate Rafn for measuring the mass spectra. The Danish Research Agency is acknowledged for funding the Danish Biotechnology Instrument Centre for purchase of mass spectrometers. References Andersen, S.O., 1972. An enzyme from locust cuticle involved in the formation of crosslinks from N-acetyldopamine. J. Insect Physiol. 18, 527–540. Andersen, S.O., 1981. The stabilization of locust cuticle. J. Insect Physiol. 27, 393–396. Andersen, S.O., 1990. Sclerotization of insect cuticle. In: Ohnishi, E., Ishizaki, H. (Eds.), Molting and Metamorphosis. Japan Scientific Societies Press, Tokyo, pp. 133–155. Andersen, S.O., 2004. Chlorinated tyrosine derivatives in insect cuticle. Insect Biochem. Mol. Biol. 34, 1079–1087. Andersen, S.O., 2005. Cuticular sclerotization and tanning. In: Gilbert, L.I., Iatrou, K., Gill, S.S. (Eds.), Comprehensive Molecular Insect Science, vol. 4. Elsevier Pergamon Press, pp. 145–170. Andersen, S.O., Barrett, F.M., 1971. The isolation of ketocatecholsfrom insect cuticle and their possible role in sclerotization. Insect Biochem. 17, 69–83. Andersen, S.O., Højrup, P., Roepstorff, P., 1995. Insect cuticular proteins. Insect Biochem. Mol. Biol. 25, 153–176.
Andersen, S.O., Jacobsen, J.P., Roepstorff, P., Peter, M.G., 1991. Catecholamine-protein conjugates: isolation of an adduct of N-acetylhistidine to the side chain of N-acetyldopamine from an insect-enzyme catalyzed reaction. Tetrahedron Lett. 32, 4287–4290. Andersen, S.O., Jacobsen, J.P., Roepstorff, P., 1992a. Coupling reactions between amino compounds and N-acetyldopamine catalyzed by cuticular enzymes. Insect Biochem. Mol. Biol. 22, 517–527. Andersen, S.O., Peter, M.G., Roepstorff, P., 1992b. Cuticle-catalyzed coupling between N-acetylhistidine and N-acetyldopamine. Insect Biochem. Mol. Biol. 22, 459–469. Andersen, S.O., Peter, M.G., Roepstorff, P., 1996. Cuticular sclerotization in insects. Comp. Biochem. Physiol. 113B, 689–705. Andersen, S.O., Roepstorff, P., 1978. Phenolic compounds released by mild acid hydrolysis from sclerotized cuticle: purification, structure, and possible origin from cross-links. Insect Biochem. 8, 99–104. Blomquist, G.J., Dillwith, J.W., 1985. Cuticular lipids. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Comprehensive Insect Physiology, Biochemistry and Pharmacology, vol. 3. Pergamon Press, Oxford, pp. 117–154. Christensen, A.M., Schaefer, J., Kramer, K.J., Morgan, T.D., Hopkins, T.L., 1991. Detection of cross-links in insect cuticle by REDOR NMR spectroscopy. J. Am. Chem. Soc. 113, 6799–6802. Hackman, R.H., 1971. The integument of Arthropoda. In: Florkin, M., Gilbert, L.I. (Eds.), Chemical Zoology, vol. 6. Academic Press, New York, pp. 1–62. Hillerton, J.E., Vincent, J.F.V., 1979. The stabilization of insect cuticles. J. Insect Physiol. 25, 957–963. Kayser, H., 1985. Pigments. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Comprehensive Insect Physiology, Biochemistry and Pharmacology, vol. 10. Pergamon Press, Oxford, pp. 367–415. Kerwin, J.L., Turecek, F., Xu, R., Kramer, K.J., Hopkins, T.L., Gatlin, C.L., Yates, J.R., 1999. Mass spectrometric analysis of catechol-histidine adducts from insect cuticle. Anal. Biochem. 268, 229–237. Klarskov, K., Højrup, P., Andersen, S.O., Roepstorff, P., 1989. Plasmadesorption mass spectrometry as an aid in protein sequence determination. Application of the method on a cuticular protein from the migratory locust (Locusta migratoria). Biochem. J. 262, 923–930. Kramer, K.J., Kanost, M.R., Hopkins, T.L., Jiang, H., Zhu, Y.C., Xu, R., Kerwin, J.L., Turecek, F., 2001. Oxidative conjugation of catechols with proteins in insect skeletal systems. Tetrahedron 57, 385–392. Neville, A.C., 1975. Biology of the Arthropod Cuticle. Springer, Berlin. Roepstorff, P., Andersen, S.O., 1980. Electron impact and chemical ionization mass spectra of catechol derivatives from insect cuticle. Biomed. Mass Spectrom. 7, 317–320. Schaefer, J., Kramer, K.J., Garbow, J.R., Jacob, G.S., Stejskal, E.O., Hopkins, T.L., Speirs, R.D., 1987. Aromatic cross-links in insect cuticle: detection by solid-state 13C and 15N NMR. Science 235, 1200–1204. Sugumaran, M., 1998. Unified mechanism for sclerotization of insect cuticle. Adv. Insect Physiol. 27, 229–334.
ARTICLE IN PRESS 234
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
Turecek, F., Kerwin, J.L., Xu, R., Kramer, K.J., 1998. Distinction of Nsubstituted histidines by electrospray ionization mass spectrometry. J. Mass Spectrom. 33, 392–396. Vincent, J.F.V., 1980. Insect cuticle: a paradigm for natural composites. Soc. Exp. Biol. Symp. 34, 183–209.
Vincent, J.F.V., Ablett, S., 1987. Hydration and tanning in insect cuticle. J. Insect Physiol. 33, 973–979. Xu, R., Huang, X., Hopkins, T.L., Kramer, K.J., 1997. Catecholamine and histidyl protein cross-linked structures in sclerotized insect cuticle. Insect Biochem. Mol. Biol. 27, 101–108.
Aspects of cuticular sclerotization in the locust, Scistocerca gregaria, and the beetle, Tenebrio molitor Svend Olav Andersena,, Peter Roepstorffb a
Institute of Molecular Biology and Physiology, University of Copenhagen, The August Krogh Building, Universitetsparken 13, DK-2100 Copenhagen O, Denmark b Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark Received 16 September 2006; received in revised form 15 November 2006; accepted 16 November 2006
Abstract The number of reactive amino groups in cuticular proteins decreases during the early period of insect cuticular sclerotization, presumably due to reaction with oxidation products of N-acetyldopamine (NADA) and N-b-alanyldopamine (NBAD). We have quantitated the decrease in cuticular N-terminal amino groups and lysine e-amino groups during the first 24 h of sclerotization in adult locusts, Schistocerca gregaria, and in larval and adult beetles, Tenebrio molitor, as well as the increase in b-alanine amino groups in Tenebrio cuticle. The results indicate that nearly all glycine N-terminal groups and a significant part of the e-amino groups from lysine residues are involved in the sclerotization process in both locusts and Tenebrio. A pronounced increase in the amount of free b-alanine amino groups was observed in cuticle from adult Tenebrio and to a lesser extent also in Tenebrio larval cuticle, but from locust cuticle no b-alanine was obtained. Hydrolysis of sclerotized cuticles from locusts and Tenebrio by dilute hydrochloric acid released a large number of compounds containing amino acids linked to catecholic moieties. Products have been identified which contain histidine residues linked via their imidazole group to the b-position of various catechols, such as dopamine, 3,4-dihydroxyphenyl-ethanol (DOPET), and 3,4dihydroxyphenyl-acetaldehyde (DOPALD), and a ketocatecholic compound has also been identified composed of lysine linked via its e-amino group to the a-carbon atom of 3,4-dihydroxyacetophenone. Some of the hydrolysis products have previously been obtained from sclerotized pupal cuticle of Manduca sexta [Xu, R., Huang, X., Hopkins, T.L., Kramer, K.J., 1997. Catecholamine and histidyl protein cross-linked structures in sclerotized insect cuticle. Insect Biochemistry and Molecular Biology 27, 101–108; Kerwin, J.L., Turecek, F., Xu, R., Kramer, K.J., Hopkins, T.L., Gatlin, C.L., Yates, J.R., 1999. Mass spectrometric analysis of catechol-histidine adducts from insect cuticle. Analytical Biochemistry 268, 229–237; Kramer, K.J., Kanost, M.R., Hopkins, T.L., Jiang, H., Zhu, Y.C., Xu, R., Kerwin, J.L., Turecek, F., 2001. Oxidative conjugation of catechols with proteins in insect skeletal systems. Tetrahedron 57, 385–392], but the lysine-dihydroxyacetophenone compound and the histidine–DOPALD adduct have not been reported before. It is suggested that the compounds are derived from NADA and NBAD residues which were incorporated into the cuticle during sclerotization, and that the lysine-dihydroxyacetophenone as well as the DOPET and DOPALD containing adducts are degradation products derived from cross-links between the cuticular proteins, whereas the dopamine-containing adducts are derived from a noncrosslinking reaction product. r 2006 Elsevier Ltd. All rights reserved. Keywords: Insect; Cuticle; Sclerotization; Histidine; Lysine; N-terminal amino acids; Adduct formation; N-acetyldopamine; N-b-alanyldopamine
1. Introduction Soon after ecdysis the cuticles of insects undergo a number of changes and modifications, which together Corresponding author. Tel.: +45 3532 1703; fax: +45 3532 1567.
E-mail address: [email protected] (S.O. Andersen). 0965-1748/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.ibmb.2006.11.006
optimize the local physical properties of the cuticular regions. The modifications can be deposition of chitin and proteins to form an endocuticle, secretion of waxes unto the surface of the epicuticle (reviewed by Blomquist and Dillwith, 1985), incorporation of ortho-diphenols, such as N-acetyldopamine (NADA) and N-b-alanyldopamine
ARTICLE IN PRESS 224
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
(NBAD), into the cuticular matrix (sclerotization) (reviewed by Sugumaran, 1998; Andersen, 2005), formation and deposition of cuticular melanin (reviewed by Kayser, 1985), dehydration of the cuticular matrix (Hillerton and Vincent, 1979; Vincent and Ablett, 1987), and chlorination of tyrosine residues in proteins (Andersen, 2004). Not all of the changes occur in all types of cuticle; cuticles from different insect species and from different developmental stages and different regions of the same animal vary widely both in appearance and physical properties. The precise properties of a given piece of cuticle will depend upon the exact balance between the various modifications, and it is possible that several yet undescribed modifications also contribute to the final outcome. To obtain a more detailed understanding of the factors influencing cuticular mechanical properties, it is necessary to study the modifications in a given cuticular region, and to quantitatively determine how rapidly and to what extent they occur. Our present understanding of the sclerotization process gives a fairly coherent picture of the major chemical reactions involved (reviewed in Andersen, 2005), although many details are still controversial, and more detailed studies of the quantitative aspects are needed. Fig. 1 illustrates the pathway for forming reactive intermediates by oxidation of the sclerotization precursors, NADA and NBAD. During sclerotization NADA and NBAD are enzymatically oxidized to the corresponding orthoquinones, which can be isomerized to para-quinone methides. Both the o-quinones and the p-quinone methides can react with nucleophilic groups on the cuticular proteins, and the p-quinone methides may also be rearranged to 1,2-dehydro-N-acyldopamines. Oxidation products of the 1,2-dehydro-acyldopamines can form dihydroxyphenyl-dihydrobenzodioxine derivatives by
reaction with catechols (Andersen et al., 1996), but they may also react with other nucleophilic groups in the cuticles. As cuticular proteins are nearly devoid of the sulphurcontaining amino acids, cysteine and cystine, the imidazole group in histidine residues is the most important nucleophile for reaction with o-quinones and p-quinone methides during sclerotization, but free amino groups, such as N-terminal groups and e-amino groups of lysine, are also potential reactants. Adducts between NADA and N-acetylhistidine have been produced by in vitro incubation of the two compounds together with cell-free pieces of cuticle (Andersen et al., 1991, 1992a,b), and adduct IV between histidine and 3,4-dihydroxyphenyl-ethanol (DOPET, I, Fig. 2) and adduct VI (Fig. 3) between dopamine and histidine have been isolated from acid hydrolysates of sclerotized pupal cuticle of Manduca sexta (Kerwin et al., 1999; Kramer et al., 2001). Lysine e-amino groups and free N-terminal amino groups in the cuticular proteins have been proposed to take part in sclerotization (Hackman, 1971; Neville, 1975), but the evidence is less convincing than that for histidine involvement. The number of free amino groups which can be chemically modified decreases during the initial phase of sclerotization of locust cuticle (Andersen, 1972), indicating that they may be involved in the process. More precise and sensitive analytical methods are now available to follow the gradual disappearance of reactive amino groups, and it is possible to distinguish lysine amino groups from amino groups from N-termini and from b-alanine. It would be an advantage also to follow the disappearance of unmodified histidine residues during sclerotization. As we have not succeeded in this the investigation was extended to include isolation of adducts between amino acids and ortho-diphenols from hydrolysates of sclerotized cuticle, and the results confirm the involvement of both histidine and lysine residues in sclerotization. Two different
Fig. 1. The current scheme for sclerotization of insect cuticles. The acyldopamines NADA and NBAD are during sclerotization oxidized to o-quinones, which are rearranged to p-quinone methides. Both types of quinones can react with nucleophilic groups (HX) present in the cuticular proteins to give ringsubstituted adducts and sidechain-substituted adducts, respectively. The p-quinone methide can also be rearranged to a dehydro-derivative, which after oxidation to the quinone stage can react with available catechols to give dihydroxyphenyl–dihydrobenzodioxine.
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
225
frozen immediately at 18 1C or kept for one or more hours at 23 1C before being frozen. 2.2. Dinitrophenylation analysis Fig. 2. Some monomeric catechols produced by acid hydrolysis of sclerotized cuticle. I: 3,4-dihydroxyphenyl-ethanol (DOPET); II: 3,4dihydroxyphenyl-acetaldehyde (DOPALD); III: arterenone.
Fig. 3. Catecholic adducts obtained by acid hydrolysis of sclerotized cuticle. IV: histidine-DOPET adduct; V: histidine-DOPALD adduct; VI: histidine-dopamine adduct; VII: lysine-3,4-dihydroxyacetophenone adduct.
insect species were studied: a locust, Schistocerca gregaria, which as adult has a nearly colorless cuticle and uses NADA for sclerotization, and a beetle, Tenebrio molitor, which uses both NADA and NBAD for sclerotization and the adults have a dark brown cuticle. 2. Materials and methods 2.1. Rearing of Schistocerca and Tenebrio Desert locusts, S. gregaria, were reared in crowded stock cultures held at about 30 1C and light during daytime (8 h duration) and ambient temperature (about 23 1C) and darkness during night conditions (16 h duration). They were fed on wheat bran and fresh lettuce. Animals were collected when halfway through ecdysis from 5th stage nymphs to adults; they were then either immediately frozen at 18 1C or kept for one or more hours at 30 1C before being frozen. T. molitor larvae were kept at 23 1C and fed on wheat bran, carrots and apples. Large larvae were collected when halfway through ecdysis to the next larval instar, and pharate adults were collected before they had completely finished shedding the pupal cuticle; they were then either
Four cuticular regions in adult locusts were chosen for analysis: hindleg femur cuticle, hindleg tibial cuticle, ventral abdominal segments nos. 3–5, and dorsal abdominal segments nos. 3–5. The abdominal samples included the intersegmental cuticle connecting the three segments. The cuticular pieces were immersed in 1% potassium tetraborate (K-borate) to soften adhering tissues and facilitate cleaning. The cleaned pieces were washed briefly in distilled water and ethanol, air dried and weighed on a Sartorius microbalance, model MC 5. The samples were separately placed in 2 ml 4% NaHCO3 containing 1% dinitrofluorobenzene (DNFB) and stored in darkness overnight at ambient temperature. Next day they were washed briefly in distilled water, placed in 20 ml 40% ethanol for 24 h to wash out any unreacted DNFB present, air dried, reweighed, and hydrolyzed in 2.0 ml 6 M ultrapure HCl in sealed evacuated tubes at 105 1C for 20 h. The hydrolysates were taken to dryness in vacuo over solid NaOH, redissolved in 1.00 ml distilled water, centrifuged for 30 min at 6000 rpm, and 10 ml aliquots were taken for reversed phase high-performance liquid chromatography (RP-HPLC) analysis. Ventral abdominal cuticle from adults and dorsal plus ventral abdominal cuticle from last instar larvae of T. molitor were analyzed in a similar way. Zero hour cuticle was obtained from animals, which were halfway through the process of ecdysis, and the ages of later cuticular samples were measured from the moment the animals escaped fully from the old cuticle. The Tenebrio samples were cleaned, washed, dried and weighed as described above for locust cuticle. Dinitrophenylation and hydrolysis of samples were also performed as described above with the single difference that only half the volumes were used of buffer, DNFB, and hydrochloric acid. The DNP-derivatives were separated on a Lichrospher 100 RP-18 column, 4.6 250 mm (Merck, Darmstadt, Germany). The A-buffer was 0.1% trifluoroacetic acid (TFA) and the B-buffer was 90% acetonitrile. The column was equilibrated with 25% B-buffer and eluted with a linear gradient from 25% to 50% B-buffer in 36 min. The absorbancy of the eluate was measured from 200 to 400 nm by means of a Dionex diode array system (UVD340U UV/ VIS detector), and the resulting peaks were integrated at 350 nm by means of the Chromeleon Management System. The elution scheme does not separate all DNP-amino acids, but gives a satisfactory separation of the DNPderivatives relevant in this study. The identities of the observed peaks (e-DNP-lysine, DNP-glycine, DNPb-alanine, O-DNP-tyrosine, and dinitroaniline) were confirmed by co-chromatography with authentic standards both in the above system and on a polystyrene column (Source 5RPC, Amersham Biosciences, Uppsala, Sweden).
ARTICLE IN PRESS 226
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
2.3. Purification of modified dopamine residues from hydrolysis of sclerotized cuticle Sclerotized cuticles were purified from light-brown adult T. molitor collected less than 8 h after ecdysis and from adult S. gregaria collected either less than 24 h or 5–6 weeks after ecdysis. The frozen animals were homogenized in 1% K-borate, and the insoluble cuticular residues were washed repeatedly in copious amounts of K-borate, distilled water, and 96% ethanol. Cuticular samples (250–500 mg) were hydrolyzed in 10–20 ml 1 M HCl at 100 1C for 8 h, and after removing the insoluble residue by filtration, the soluble fraction was taken to dryness in vacuo over solid NaOH, dissolved in 3 ml 0.1% TFA, and applied to a cation exchange column (SP-Sepharose, 7 1.0 cm, Amersham Biosciences) in equilibrium with 0.1% TFA. The column was washed with 0.1% TFA until the ultraviolet absorbance at 280 nm of the eluate had decreased to near the baseline, and a linear gradient of NaCl in 0.1% TFA was then applied to elute the material retained on the column. The UV-absorbance of the eluate at 280 nm was recorded, and the fractions which were eluted later than arterenone (III, Fig. 2) were collected and screened by RP-HPLC for compounds of interest. 2.4. RP-HPLC-conditions The eluates collected after cation chromatography were fractionated on a Source 5RPC column (Amersham Biosciences) and eluted with a linear gradient in acetonitrile (from 2% to 70% acetonitrile in 0.1% TFA for 35 min). The peaks, which eluted early and had UV-absorbtion maxima near 280 nm were collected and subjected to further purification on a Lichrospher 100 RP-18 column (4.6 250 mm, Merck, Darmstadt, Germany). The ultraviolet absorbtion was recorded by the Dionex diode array system, and the major peaks showing UV-absorbtion spectra similar to that of dopamine were collected for further characterization. 2.5. Synthesis of 3,4-dihydroxyacetophenone-lysine adduct 3,4-dihydroxyphenacyl chloride (1 mmole, 186 mg) and a-N-acetyllysine (1 mmole, 188 mg) were dissolved in 10 ml 15% K-borate and left over night at ambient temperature. Acetic acid was added to adjust pH to 5, and after removal of the precipitated boric acid by centrifugation the reaction products were purified by RP-HPLC and cation exchange as described above for purification of the hydrolysis poducts from sclerotized cuticle. The main product was then hydrolyzed in 1 M HCl at 100 1C for 5 h and repurified by RP-HPLC on a Lichrospher 100 RP-18 column. 2.6. Mass spectrometry Molecular weight determination of the components in the different fractions collected after RP-HPLC was
Fig. 4. Hydrolysate of dinitrophenylated 0-h locust femur cuticle separated by RP-HPLC on a Lichrospher RP-18 column. 1: e-DNPlysine; 2: DNP-glycine; 3: dinitroaniline; 4: O-DNP-tyrosine.
performed by matrix assisted laser desorption/ionization mass spectrometry (MALDI-MS) using an ABI STR MALDI mass spectrometer (Applied Biosystems, Framingham, Ma) in reflecting mode. 0.5 ml of the eluate was mixed with 0.5 ml matrix solution (2,5-dihydroxybenzoic acid (DHB) 20 mg/ml in 70% acetonitrile, 0.1% TFA) on the target and dried and spectra recorded for an appropriate number of laser shots to obtain a good S/N. Structural information on the compounds was obtained by collision induced dissociation on a 4700 TOF/TOF mass spectrometer (Applied Biosystems, Farmingham, Ma) after isolation of the appropriate m/z range. Sample preparation was as above except that the DHB matrix solution was replaced with a solution of a-cyano-4-hydroxy cinnamic acid (5 mg/ml in 70% acetonitrile, 0.1% TFA). 3. Results 3.1. Dinitrophenylation studies Fig. 4 shows separation on a Lichrospher column of the hydrolysis products from a dinitrophenylated 0-h femur from the locust S. gregaria. The chromatogram is dominated by the peak of e-DNP-lysine, and the smaller later-eluting peaks have been identified as DNP-glycine, dinitroaniline and O-DNP-tyrosine, respectively. Fig. 5 shows the decrease in amounts of e-DNP-lysine and DNPglycine during the early phase of locust cuticular sclerotization; DNP-glycine decreased to nearly zero during the first 4 h after ecdysis, and DNP-lysine decreased during the same period to about 50% of its initial value and stayed thereafter nearly constant. Similar results were obtained for cuticle from tibia and dorsal and ventral abdomen of locusts. Fig. 6 shows that the amount of DNP-glycine recovered from dinitrophenylated larval abdominal cuticle of T. molitor decreased to low values during the first few hours after ecdysis, that a somewhat slower decrease to
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
227
Fig. 5. Yields of e-DNP-lysine ( ) and DNP-glycine (K) obtained from locust femur cuticle during the first 24 h after ecdysis. Fig. 7. Yields of e-DNP-lysine ( ), DNP-glycine (K), and DNP-balanine (m) obtained from ventral abdominal cuticle from adult Tenebrio during the first 24 h after ecdysis. Table 1 Amino acid analyses of adult S. gregaria femur cuticle sclerotized for 0 and 3 h after ecdysis (the results with standard errors are shown as residues per 100 residues; 5 analyses per sample)
Fig. 6. Yields of e-DNP-lysine ( ), DNP-glycine (K), and DNP-balanine (m) obtained from larval Tenebrio cuticle during the first 24 h after ecdysis.
about 70% of the initial value was observed for the e-amino group of lysine, and that DNP-b-alanine was obtained in trace amounts of from 0-h cuticle and in slightly larger amounts from older cuticle. Fig. 7 shows that the decrease in amounts of e-DNP-lysine in the ventral abdominal cuticle of adult T. molitor was pronounced and continued for a longer period (at least 24 h) than observed for larval cuticle. A low, but significant yield of dinitrophenylated b-alanine was obtained from 0-h beetles, and
Amino acid
0 h femur
3 h femur
Aspartic acid Threonine Serine Glutamic acid Glycine Alanine Cysteine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Proline
4.1070.18 2.5570.10 4.7770.04 3.8870.15 7.8370.15 33.8270.44 0 7.5270.16 0 3.7070.05 4.6670.06 8.3370.22 0.7870.03 2.0670.03 1.7470.12 3.8370.02 10.4170.09
3.8870.06 2.4170.05 4.6570.06 3.5670.06 7.8270.12 35.2870.33 0 7.6970.06 0 3.9270.08 4.8670.05 7.2370.09 0.7170.01 2.1270.03 1.4970.05 3.7070.05 10.6870.10
after a delay of 2 h a considerable increase in yield of DNPb-alanine was observed. 3.2. Amino acid analysis The amino acid compositions of 0- and 3-h femur cuticle from S. gregaria were determined and compared (Table 1).
ARTICLE IN PRESS 228
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
No significant changes were observed in the content of most amino acids, indicating that deposition of endocuticle does not yet play a role; the most significant change was in the content of lysine, which decreased by 14% from 1.74 to 1.49 residue percent, which is markedly lower than the decrease of ca. 50% in reactive lysine residues observed in the dinitrophenylation experiments. 3.3. Purification of catecholic adducts Sclerotized cuticles from S. gregaria and T. molitor adults were subjected to partial hydrolysis in 1 M HCl, and a number of compounds containing catechol moieties and carrying more than one positive charge were purified from the hydrolysates. Figs. 8A and B show the elution patterns obtained by cation-exchange of hydrolysates of cuticles from 5–6 week old adult S. gregaria and adult Tenebrio, which was sampled less than 8 h after ecdysis, respectively. The peak at 135 ml was identified as arterenone (III), and the later eluting fractions (a and b) indicated by horizontal bars were collected and purified by RP-HPLC. The compounds eluted between 155 and 190 ml (fraction a) in Fig. 8A can be assumed to carry two positive charges at pH 2; they were purified on a Source 5RPC column (Fig. 9A) and the peak (Sg-a1) eluted at 5.5–6.0 min was further purified on a Lichrospher column (Fig. 9B). The ultraviolet spectrum of the compound eluted at 4.4 min in Fig. 9B (Sg-a1-1) corresponded to that of dopamine, indicating that the compound is a catechol, and mass spectrometry showed that the compound has an m/zvalue of 306.1. After 10–20 min exposure to alkaline
conditions (1% sodium carbonate) in the presence of oxygen, the compound was degraded to 3,4-dihydroxyphenyl-glyoxal and histidine. Reduction of the compound with sodium borohydride resulted in a product which eluted later from the HPLC-columns than did the unreduced compound, and the reduction resulted in an increase in m/z-value to 308.1 mass units. When subjected to MS/MS both the reduced and the unreduced compound gave a peak at m/z 156.4, corresponding to histidine, and the unreduced compound gave also a peak at m/z 260, corresponding to a loss of H2O+CO. Such a loss of 46 mass units indicates that histidine is connected to the sidechain of the catechol moiety via its Nt (N 1) position (Turecek et al., 1998). The combined results indicate that compound Sg-a1-1 is an adduct (V) between histidine and 3,4-dihydroxyphenyl-acetaldehyde (DOPALD, II), and that the reduced compound is the corresponding adduct (IV) between histidine and DOPET (I). A compound with an ultraviolet spectrum similar to that reported for ketocatecholic compounds (Andersen and Barrett, 1971) was purified from the peak Sg-a2 eluted at 8 min in Fig. 9A. Treating the compound with sodium borohydride gave a reduction product having a typical catecholic ultraviolet spectrum, and the reduction caused an increase in molecular mass from an m/z-value of 297.1 to 299.1. The compound was subjected to MS/MS fragmentation, and fragment peaks were observed at m/z-values of 84.1 and 130.1, indicating the presence of a lysine residue, and at 137.1, indicating the presence of a keto-group joined to the catechol ring (Andersen and Roepstorff, 1978; Roepstorff and Andersen, 1980).
Fig. 8. Cation exchange chromatography of acid hydrolysates a (A) 500 mg cuticle from mature locusts and (B) 500 mg cuticle from adult Tenebrio less than 8 h after ecdysis. The peaks containing arterenone are marked with *, and the fractions marked by horizontal bars were collected for further purification.
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
229
Fig. 9. (A) Reverse phase chromatography on Source 5RPC of the material eluted between 155 and 190 ml (Sg-a) in Fig. 8A (locust cuticular hydrolysate). The eluates marked with bars (Sg-a1 and Sg-a2) were collected for further purification. (B) Purification of peak at 5.5–6.0 min (Sg-a1) in Fig. 9A on a Lichrospher column; the marked peak was collected for characterization.
The results suggest that the compound is an adduct between lysine and 3,4-dihydroxyacetophenone (structure VII). The compound was also obtained from hydrolysates of sclerotized cuticle from adult T. molitor, although in lesser amounts than from locust cuticle. Compound VII was synthesized by reaction between 3,4-dihydroxyphenacyl chloride and a-N-acetyllysine followed by acid hydrolysis of the main product. The purified hydrolysis product was indistinguishable from the compound obtained from hydrolysates of locust sclerotized cuticle both with regard to migration on RP-HPLC columns, ultraviolet absorbtion spectrum and mass spectrum, indicating the correctness of structure VII. Figs. 10A and B show purification of the Tenebrio compounds (Tm-a) which eluted between 155 and 190 ml in Fig. 8B. When subjected to chromatography on a Lichrospher column the broad peak (Tm-a1) at about 5 min in Fig. 10A was resolved in a number of peaks as shown in Fig. 10B; they all contained catecholic compounds according to their ultraviolet spectra, but so far only a few have been identified. The mass spectra obtained from the peaks indicated that most of them contain several compounds; the mass spectrum of the peak at 5.5 min (Tm-a1-1) in Fig. 10B thus gave several major peaks corresponding to m/z-values of 156.4, 213.4, 308.4, and 395.3 mass units. The m/z-values of 156.4 and 213.4 correspond to those expected for histidine and glycyl-histidine (or histidylglycine), respectively, while a mass of 308.4 indicates the presence of an adduct (IV) between DOPET and histidine. The composition of the adduct was confirmed by MS/MS, which showed a fragmentation pattern dominated by a peak of m/z 156.4, corresponding to histidine. The m/z-value of 395.3 could indicate an adduct between
DOPET and seryl-histidine, but its identity has not yet been confirmed. The mass spectrum of the double peak at 6.1 min (Tm-a1-2) in Fig. 10B showed the presence of a compound having an m/z-value of 365.3 mass units, corresponding to an adduct between DOPET and glycyl-histidine. This was confirmed by MS/MS, where a major fragmentation ion was present at m/z 213.4, and a loss of 46 mass units (H2O+CO) was observed, indicating that the dipeptide is glycylhistidine and not histidylglycine. The compound eluted at 5.5 min (Tm-a2) in Fig. 10A was identified as the same adduct (V) between histidine and DOPALD as found in hydrolysates of locust cuticle. The identification was based upon mass spectrum, degradation by treatment with alkali, reduction to a histidine–DOPET adduct with sodium borohydride, and co-chromatography with the locust compound. Fig. 11A and B show purification of the Tenebrio compounds (Tm-b) eluted between 220 and 250 ml in Fig. 8B. The resolution of the double peak (Tm-b1) in Fig. 11A was only slightly improved on the Lichrospher column (Fig. 11B). Mass spectrometry showed that the peak Tm-b1-1 at 4.2 min in Fig. 11B contained a compound having an m/z-value of 307.1 and that the peak Tm-b1-2 at 4.6 min contained a compound of m/z-value 364.2. The m/z-value of 307.1 agrees with an adduct between dopamine and histidine (VI), and the fragmentation pattern observed in MS/MS indicates that an imidazole nitrogen (Nt) from histidine is connected to the b-position of the dopamine sidechain. The m/z-value of 364.2 agrees with an adduct between dopamine and histidine plus glycine. MS/MS of the m/z 364.2 component indicated that a glycylhistidine moiety is connected to the b-position in dopamine side chain.
ARTICLE IN PRESS 230
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
Fig. 10. (A) Reverse phase chromatography on Source 5RPC of the material eluted between 155 and 190 ml (Tm-a) in Fig. 8B (Tenebrio cuticular hydrolysate). The eluate marked by a bar was collected for further purification. (B) Purification of peak at 4.5–5.5 min (Tm-a1) in Fig. 10A on a Lichrospher column; the marked peaks were collected for characterization.
Fig. 11. (A) Reverse phase chromatography on Source 5RPC of the material eluted between 220 and 250 ml (Tm-b) in Fig. 8B (Tenebrio cuticular hydrolysates). (B) Purification of peak at 4.0–4.5 (Tm-b1) in Fig. 11A on a Lichrospher column; the marked peaks were collected for characterization.
The dopamine–histidine adduct was also isolated from sclerotized locust cuticle, but in much smaller amounts than were obtained from Tenebrio cuticle. 4. Discussion An important step in the sclerotization of insect cuticle is the reaction of quinoid derivatives of various catechols with nucleophilic groups on the cuticular matrix proteins. The most important catecols involved in the process are
NADA and NBAD, but also DOPET has been proposed as a probable sclerotization compound (Kerwin et al., 1999; Kramer et al., 2001). The imidazole ring in proteinbound histidine residues seems to be the most important nucleophilic group, although also the free amino groups present in lysine residues and the N-terminal amino groups are likely to be involved. Amino acid sequence determination on cuticular proteins from pharate adult migratory locusts (Locusta migratoria) showed that the majority of the proteins have glycine as the N-terminal amino acid
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
(Andersen et al., 1995); until now 31 pharate locust cuticular proteins have been sequenced, and glycine is the N-terminal residue in 30 of them and the remaining protein has isoleucine as N-terminus (Klarskov et al., 1989). The amino acid sequences of the few cuticular proteins sequenced from S. gregaria are nearly identical to those from L. migratoria, and accordingly it can be expected that the majority of pharate cuticular proteins from S. gregaria also have glycine as N-terminal residue. The results obtained by dinitrophenylation of locust cuticle during the early period of the sclerotization process demonstrate that besides e-N-DNP-lysine only DNP-glycine is obtained in significant amounts, confirming the assumption that glycine is the dominating N-terminal amino acid in S. gregaria pharate cuticular proteins. The amino groups from lysine as well as from N-terminal glycine are rapidly rendered unavailable for dinitrophenylation: after 4 h sclerotization about 90% of the N-terminal glycine residues and about 50% of the lysyl e-amino groups have become unreactive. The contents of free N-terminal and lysine amino groups also decrease in larval and adult Tenebrio cuticles during the first day of sclerotization, but the decrease is significantly slower than in locust cuticle. Another difference between the cuticles of the two insect species is the presence of free amino groups of b-alanine in both larval and adult Tenebrio cuticles and the absence of this amino acid from locust cuticle. Small amounts of b-alanine are present in 0-h cuticle from Tenebrio larvae, and the amounts of increase slowly for at least 24 h, presumably due to incorporation of NBAD. Larger amounts of balanine are present in 0-h cuticle from adult Tenebrio, no increase in content of free b-alanine amino groups is observed during the first 2 h of sclerotization, then the content increases for some hours to decrease again between 9 and 24 h after ecdysis, presumably due to utilization of more amino groups for sclerotization than are provided by incorporation of NBAD. The disappearance of reactive amino groups is probably due to their reaction with sclerotization agents, but it could also be explained by assuming that the dinitrophenylation reagent has gradually increasing difficulties in getting access to the amino groups, for instance due to closer packing of the protein chains caused by dehydration of the cuticular matrix. During the first few hours after ecdysis the water content of locust cuticle decreases from 43% to 32% of the wet weight (Andersen, 1981), probably contributing to the increase in cuticular stiffness observed during the early phase of sclerotization (Vincent, 1980; Vincent and Ablett, 1987), but it is unlikely to contribute to the decreased reactivity of amino groups. Before analysis by dinitrophenylation all the cuticular samples were dried to constant weight and had experienced a water loss much greater than the small natural loss occurring soon after ecdysis, resulting in tightly packing of the protein chains both in the unsclerotized and the sclerotized samples.
231
Amino acid composition of unsclerotized locust femur cuticle compared to that of femur cuticle allowed to sclerotize for 3 h showed a decrease of 14% in lysine content and no significant decrease in content of histidine, indicating that if covalent bonds are formed between these two amino acids and catechol derivatives then the majority of the bonds must be acid labile, which is in contrast to the isolation of significant amounts of histidine–dopamine adducts from sclerotized pupal cuticle of M. sexta hydrolyzed in 6 M HCl (Kerwin et al., 1999; Kramer et al., 2001). The disappearance of reactive amino groups was determined only for the first 24 h after ecdysis, although cuticular sclerotization in both locusts and Tenebrio continues for a couple of weeks. Deposition of a considerable amount of endocuticle occurs during the whole sclerotization period, and much of this endocuticle is also subject to sclerotization but to a lesser degree than the exocuticle. During the early hours after ecdysis only little endocuticle is deposited and its contribution to the weight of the cuticular samples and to their lysine content will be insignificant. The results shown in Figs. 3–5 will, therefore, mainly represent modifications of exocuticular amino groups. As histidine has been shown to be involved in adduct formation with dopamine derivatives during sclerotization of Manduca pupal cuticle (Schaefer et al., 1987; Christensen et al., 1991; Xu et al., 1997; Kerwin et al., 1999; Kramer et al., 2001), it would be of interest to determine to what extent this amino acid is consumed during the sclerotization process. But so far it has not been possible to find a useful method for determining changes in the content of reactive histidine residues. Since histidine–dopamine and histidine–DOPET adducts have been obtained from acid hydrolysates of sclerotized Manduca pupal cuticle (Kerwin et al., 1999; Kramer et al., 2001), it is of interest to determine whether such acid-stable adducts also can be obtained from sclerotized locust and Tenebrio cuticle. To avoid the presence of low-molecular weight carbohydrates in the hydrolysates 1.0 M HCl was chosen for cuticular hydrolysis, as chitin is not significantly degraded at this acid strength. The cuticular proteins are probably not completely degraded to free amino acids, and a large number of small peptides will be formed when dilute acids are used for hydrolysis. Most of the amino acids and peptides will only have a single positive charge at pH 2, while histidine–dopamine adducts and lysine–dopamine adducts will carry three positive charges at low pH-values, and an adduct between dopamine and glycine will carry two positive charges. Cation exchange was accordingly chosen as the first step in the purification procedure, and components eluted later than arterenone were selected for characterization. Many compounds with ultraviolet spectra similar to those of dopamine or ketocatechols were present in this late region of the eluates, and most of them have not yet been identified. So far an adduct between dopamine and histidine (VI) has been identified from hydrolysates of
ARTICLE IN PRESS 232
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
Tenebrio cuticle and obtained in smaller amounts from locust cuticle. From Tenebrio cuticular hydrolysates two other compounds were isolated, which eluted from the cation exchange column shortly after arterenone, suggesting the presence of two positive charges; by mass spectrometry they were identified as adducts between histidine and DOPET (IV) and histidine and DOPALD (V), respectively. From hydrolysates of locust cuticle only V was obtained. Compound V is very unstable in alkaline solutions, and treatment with dilute Tris-base or sodium carbonate resulted in rapid degradation of the compound to 3,4-dihydroxyphenyl-glyoxal and histidine. Reduction of the aldehyde group in V gave an adduct between DOPET and histidine, identical to compound IV. Compounds IV and VI have previously been isolated from Manduca pupal cuticle after hydrolysis in 6 M HCl (Kerwin et al., 1999; Kramer et al., 2001), whereas the unstable compound V has not been reported before. Compounds IV and V can readily be converted to each other and it appears likely that they are degradation products of identical or very similar structures in the sclerotized cuticles. It was suggested that the histidine– DOPET adduct obtained from pupal cuticle of Manduca is a reaction product between histidine and the p-quinone methide of DOPET, indicating that DOPET as well as NADA and NBAD is utilized as a sclerotizing compound in this insect (Kerwin et al., 1999; Kramer et al., 2001).
Fig. 12. Suggested structures for two interchain cross-links in insect sclerotized cuticle.
The histidine–DOPET compound (IV) obtained from Tenebrio cuticle might have a similar origin, and the aldehyde adduct (V) could then be formed by oxidation of the DOPET adduct. We are in favour of explaining the aldehyde adduct (V) as a direct degradation product of a NADA-derived cuticular cross-link; the histidine–DOPET compound (IV) could then be formed by reduction of the aldehyde compound during hydrolysis of Tenebrio cuticle. The quinone formed by oxidation of dehydro-NADA reacts readily with the two phenolic groups in catechols to form benzodioxine derivatives (Fig. 1), and it will probably also react spontaneously with two closely positioned nucleophilic groups. If these groups are present in separate protein chains the result will be a covalent cross-link between these chains, as suggested previously (Andersen, 1990). Structures VIII and IX (Fig. 12) show two of several such possible cross-links. In structure VIII an imidazole group from histidine is linked to the b-position of the NADA side-chain and an amino group from lysine is linked to the a-position of the same NADA-unit, and lysine residues are linked to both the a- and the b-position of a NADA-unit in structure IX. Fig. 13 suggests that acid hydrolysis of compound VIII will remove the acetyl group, resulting in an intermediate, carrying a primary and a secondary amino group on the same carbon atom. This configuration is unstable, and the intermediate is likely to be further hydrolyzed to the aldehyde-containing histidineadduct (V), ammonia and lysine. Compound VII could possibly be produced by hydrolytic degradation of a cross-link, where lysine residues are linked to both a- and b-positions of a NADA residue (structure IX). Fig. 14 suggests that hydrolytic deacetylation of the NADA moiety in IX might be followed by loss of ammonia and formation of a Schiff base which by hydrolysis will give compound VII. The adducts between catechols and amino acids which have been obtained by acid hydrolysis of sclerotized insect cuticles are probably all derived from N-acyldopamines (NADA and NBAD), and they can be divided into two groups: those carrying an amino group on the a-carbon atom of the side-chain and those lacking this amino group. The adducts between dopamine and either histidine or glycyl-histidine belong to the former group, and they are assumed to be produced by reaction of protein-bound
Fig. 13. Suggested pathway for formation of adduct V by acid-catalyzed degradation of a cuticular cross-link in which a histidine and a lysine are linked to a NADA residue.
ARTICLE IN PRESS S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
233
Fig. 14. Suggested pathway for formation of adduct VII by acid-catalyzed degradation of a cuticular cross-link in which two lysine residues are linked to a NADA residue.
histidine residues with the p-quinone imide of NADA. They are probably not derived from cross-links in the cuticles, whereas the adducts belonging to the latter group (IV and VII, and possibly V) are suggested to be derived from cuticular cross-links formed by reaction of oxidized dehydro-NADA with imidazole and/or amino groups in the cuticular proteins. In fully sclerotized cuticle both cross-linking and noncross-linking adducts will be involved in polymer formation due to continued addition of oxidized dehydro-NADA to available catecholic groups. It is our hope that further studies of the still unidentified compounds present in the cuticular hydrolysates will contribute to a more detailed understanding of the chemistry of sclerotization, and either confirm or modify the suggestions proposed here. Acknowledgments We are grateful for economic support from the Carlsberg Foundation, and we thank Lene Skou and Kate Rafn for measuring the mass spectra. The Danish Research Agency is acknowledged for funding the Danish Biotechnology Instrument Centre for purchase of mass spectrometers. References Andersen, S.O., 1972. An enzyme from locust cuticle involved in the formation of crosslinks from N-acetyldopamine. J. Insect Physiol. 18, 527–540. Andersen, S.O., 1981. The stabilization of locust cuticle. J. Insect Physiol. 27, 393–396. Andersen, S.O., 1990. Sclerotization of insect cuticle. In: Ohnishi, E., Ishizaki, H. (Eds.), Molting and Metamorphosis. Japan Scientific Societies Press, Tokyo, pp. 133–155. Andersen, S.O., 2004. Chlorinated tyrosine derivatives in insect cuticle. Insect Biochem. Mol. Biol. 34, 1079–1087. Andersen, S.O., 2005. Cuticular sclerotization and tanning. In: Gilbert, L.I., Iatrou, K., Gill, S.S. (Eds.), Comprehensive Molecular Insect Science, vol. 4. Elsevier Pergamon Press, pp. 145–170. Andersen, S.O., Barrett, F.M., 1971. The isolation of ketocatecholsfrom insect cuticle and their possible role in sclerotization. Insect Biochem. 17, 69–83. Andersen, S.O., Højrup, P., Roepstorff, P., 1995. Insect cuticular proteins. Insect Biochem. Mol. Biol. 25, 153–176.
Andersen, S.O., Jacobsen, J.P., Roepstorff, P., Peter, M.G., 1991. Catecholamine-protein conjugates: isolation of an adduct of N-acetylhistidine to the side chain of N-acetyldopamine from an insect-enzyme catalyzed reaction. Tetrahedron Lett. 32, 4287–4290. Andersen, S.O., Jacobsen, J.P., Roepstorff, P., 1992a. Coupling reactions between amino compounds and N-acetyldopamine catalyzed by cuticular enzymes. Insect Biochem. Mol. Biol. 22, 517–527. Andersen, S.O., Peter, M.G., Roepstorff, P., 1992b. Cuticle-catalyzed coupling between N-acetylhistidine and N-acetyldopamine. Insect Biochem. Mol. Biol. 22, 459–469. Andersen, S.O., Peter, M.G., Roepstorff, P., 1996. Cuticular sclerotization in insects. Comp. Biochem. Physiol. 113B, 689–705. Andersen, S.O., Roepstorff, P., 1978. Phenolic compounds released by mild acid hydrolysis from sclerotized cuticle: purification, structure, and possible origin from cross-links. Insect Biochem. 8, 99–104. Blomquist, G.J., Dillwith, J.W., 1985. Cuticular lipids. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Comprehensive Insect Physiology, Biochemistry and Pharmacology, vol. 3. Pergamon Press, Oxford, pp. 117–154. Christensen, A.M., Schaefer, J., Kramer, K.J., Morgan, T.D., Hopkins, T.L., 1991. Detection of cross-links in insect cuticle by REDOR NMR spectroscopy. J. Am. Chem. Soc. 113, 6799–6802. Hackman, R.H., 1971. The integument of Arthropoda. In: Florkin, M., Gilbert, L.I. (Eds.), Chemical Zoology, vol. 6. Academic Press, New York, pp. 1–62. Hillerton, J.E., Vincent, J.F.V., 1979. The stabilization of insect cuticles. J. Insect Physiol. 25, 957–963. Kayser, H., 1985. Pigments. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Comprehensive Insect Physiology, Biochemistry and Pharmacology, vol. 10. Pergamon Press, Oxford, pp. 367–415. Kerwin, J.L., Turecek, F., Xu, R., Kramer, K.J., Hopkins, T.L., Gatlin, C.L., Yates, J.R., 1999. Mass spectrometric analysis of catechol-histidine adducts from insect cuticle. Anal. Biochem. 268, 229–237. Klarskov, K., Højrup, P., Andersen, S.O., Roepstorff, P., 1989. Plasmadesorption mass spectrometry as an aid in protein sequence determination. Application of the method on a cuticular protein from the migratory locust (Locusta migratoria). Biochem. J. 262, 923–930. Kramer, K.J., Kanost, M.R., Hopkins, T.L., Jiang, H., Zhu, Y.C., Xu, R., Kerwin, J.L., Turecek, F., 2001. Oxidative conjugation of catechols with proteins in insect skeletal systems. Tetrahedron 57, 385–392. Neville, A.C., 1975. Biology of the Arthropod Cuticle. Springer, Berlin. Roepstorff, P., Andersen, S.O., 1980. Electron impact and chemical ionization mass spectra of catechol derivatives from insect cuticle. Biomed. Mass Spectrom. 7, 317–320. Schaefer, J., Kramer, K.J., Garbow, J.R., Jacob, G.S., Stejskal, E.O., Hopkins, T.L., Speirs, R.D., 1987. Aromatic cross-links in insect cuticle: detection by solid-state 13C and 15N NMR. Science 235, 1200–1204. Sugumaran, M., 1998. Unified mechanism for sclerotization of insect cuticle. Adv. Insect Physiol. 27, 229–334.
ARTICLE IN PRESS 234
S.O. Andersen, P. Roepstorff / Insect Biochemistry and Molecular Biology 37 (2007) 223–234
Turecek, F., Kerwin, J.L., Xu, R., Kramer, K.J., 1998. Distinction of Nsubstituted histidines by electrospray ionization mass spectrometry. J. Mass Spectrom. 33, 392–396. Vincent, J.F.V., 1980. Insect cuticle: a paradigm for natural composites. Soc. Exp. Biol. Symp. 34, 183–209.
Vincent, J.F.V., Ablett, S., 1987. Hydration and tanning in insect cuticle. J. Insect Physiol. 33, 973–979. Xu, R., Huang, X., Hopkins, T.L., Kramer, K.J., 1997. Catecholamine and histidyl protein cross-linked structures in sclerotized insect cuticle. Insect Biochem. Mol. Biol. 27, 101–108.