Purification and characterization of myrosinase from the cabbage aphid (Brevicoryne brassicae), a brassica herbivore

Transcription

1 Eur. J. Biochem. 268, 1041±1048 (2001) q FEBS 2001 Purification and characterization of myrosinase from the cabbage aphid (Brevicoryne brassicae), a brassica herbivore Bo Pontoppidan 1, Barbara Ekbom 2, Susanna Eriksson 1 and Johan Meijer 1 1 Department of Plant Biology, Uppsala Genetic Centre, and 2 Department of Entomology, Swedish University of Agricultural Sciences, Uppsala, Sweden Aphids are among the most serious insect pests of agricultural crops in the world. They often have specific hosts, and the cabbage aphid (Brevicoryne brassicae) isa specialist on Cruciferae. It has previously been described that certain insects contain the enzyme myrosinase (EC ), which is considered an important defence enzyme of crucifers. Myrosinase was purified to homogeneity from cabbage aphid soluble extracts using anionexchange and phenyl-sepharose chromatography. The protein has an apparent subunit molecular mass of 57±58 kda and is a dimer. The isoelectric point is 4.9 and the enzyme has a temperature optimum around 40 8C. The enzyme was active towards the glucosinolates tested, sinigrin and glucotropaeolin, but was inhibited by ascorbate at concentrations that normally activate plant myrosinases. Using sinigrin as the substrate K m was determined as 0.41 mm, and the k cat as 36 s 21. With glucotropaeolin the K m and k cat values were determined as 0.52 mm and 22.8 s 21, respectively. The enzyme was stable upon storage at 4 8C for many months, but lost some activity upon freezing. The insect myrosinase did not cross-react with antibodies raised to plant myrosinase. Peptide sequencing of a tryptic digest of the protein showed homology to b-glucosidases. The presence of myrosinase in an insect pest specialist may be an example of a coevolution process that facilitates host specialization. Keywords: brassica; cabbage aphid (Brevicoryne brassicae, Homoptera, Aphididae); glucosinolate; insect pest; myrosinase (S-thioglucosidase, EC ). Plants have evolved different defence systems in order to escape herbivory. Such systems include constitutive physical and chemical barriers as well as inducible systems [1]. Herbivores may also undergo reciprocal evolution in response to plant defence to avoid the negative effects of such defences. An example of a constitutive barrier is the myrosinase-glucosinolate system present in capparales plants [2]. The secondary compounds, glucosinolates, are degraded to various toxic products through the action of the enzyme myrosinase upon tissue damage. These toxic compounds have been suggested to play a role in plant defence against generalist insects and pathogens as well as serving as cues for organisms that are specialists on glucosinolate-containing plants [2]. The system is highly complex and plants contain many different myrosinase genes [3] in addition to a number of structurally different glucosinolates with aliphatic, aromatic or indolylic side chains [4]. In addition, plant glucosinolate composition changes during development and there also seems to be tissue-dependent glucosinolate profiles as well as diurnal effects on glucosinolates [5]. A large number of O-glycosyl hydrolases or glycosidases (EC n) have been described from various sources. This large superfamily Correspondence to B. Pontoppidan, Department of Plant Biology, Uppsala Genetic Centre, Box 7080, Swedish University of Agricultural Sciences, SE Uppsala, Sweden. Fax: , Tel.: , Bo.Pontoppidan@vbiol.slu.se Enzymes: myrosinase (S-thioglucosidase, EC ). (Received 25 September 2000, revised 17 November 2000, accepted 14 December 2000) has been divided into different families [6] and clans [7] where myrosinases have been assigned to family 1 and clan GH-A. This division of glycosidases into families is based on amino-acid sequence similarities, which are thought to reflect structural and catalytic properties common to the enzymes. One feature that seems to distinguish myrosinases from O-glycosidases is the substitution of glutamate with glutamine in a critical position of the active site [8], which also explains the different substrate specificities observed for O- versus S-glycosidases. A number of insect pests have coevolved with members of the capparales and are able to use them as hosts for feeding or oviposition [9]. Insects that have circumvented the detrimental effects of secondary metabolites contained in plants may even use these compounds in their own defence against natural enemies. For instance, one way aphids may escape their enemies is through use of alarm pheromones, olfactory signals produced by disturbed aphids. Aphids are known to produce alarm pheromone, for example, when an individual is attacked by predators or parasitoids. Neighbouring aphids respond to alarm pheromone with defensive or avoidance behaviour. The main chemical component of the alarm pheromone of most aphid species is (E)-b-farnesene [9a]. Examples of deleterious insect pests on oilseed rape (Brassica napus) are cabbage aphids (Brevicoryne brassicae) and mustard aphids (Lipaphis erysimi), diamond back moth (Plutella xylostella), flea beetles (Phyllotreta spp.), and pollen beetles (Meligethes aeneus). What enables these insects to successfully colonize Brassica plants is not known. We have recently started a survey of some Brassica specialists to gain insight as to how these pests have adapted to the glucosinolate-myrosinase system. During

2 1042 B. Pontoppidan et al. (Eur. J. Biochem. 268) q FEBS 2001 Fig. 1. Structures of the substrates tested. this survey we could confirm that certain aphid species actually seemed to possess their own myrosinase activity. Earlier studies detected myrosinase in both the mustard aphid and the cabbage aphid [10,11]. These aphid myrosinases hydrolysed a glucosinolate in vitro and exhibited electrophoretic mobilities different from myrosinase from the host plants. The purpose of this study is to characterize the cabbage aphid myrosinase and compare it with plant myrosinases. MATERIALS AND METHODS Materials Sinigrin (Sigma Chemical Co., St Louis, MI, USA), glucotropaeolin (Merck, Darmstadt, Germany), p-nitrophenyl-b-glucopyranoside (Sigma), p-nitrophenol (ICN Biochemicals Inc., Irvine, CA, USA) and ascorbate (Sigma) were used. Insects Cabbage aphids were collected from test plots with oilseed rape and white mustard (Sinapis alba) close to the Swedish University of Agricultural Sciences in Uppsala. The insects were then maintained in cages on oilseed rape or mustard plants under standard greenhouse conditions. Before experiments, the aphids were removed from the plants and kept in containers with access to water only for 24 h before analysis. Purification of aphid myrosinase Aphids were stored at 280 8C before use. The insects were transferred to petri dishes on dry ice and any remaining plant material was carefully removed. After measurement of the mass the insects were ground in a mortar under liquid nitrogen. The fine powder was transferred to test tubes and extraction buffer (20 mm Bistris/HCl, ph 6.1, 1 mm EDTA, 5% glycerol, 1 mm phenylmethanesulfonyl fluoride, 1 mm benzamidin, 1 mm dithiothreitol) was added to make a 10% homogenate. The solution was kept overnight at 4 8C onan end-over-end mixer. The extract was centrifuged at g for 20 min at 4 8C and the supernatant centrifuged at g for 45 min in a refrigerated RP55T rotor. The final supernatant was diluted 1 : 1 with 50 mm Bistris/HCl, ph 6.1 and applied to a Q-Sepharose (Pharmacia, Uppsala, Sweden) column equilibrated with 50 mm Bistris/HCl, ph 6.1. The column was washed with the same buffer and retained myrosinase activity was eluted with 50 mm Bistris/ HCl, ph 6.1, 1 m ammonium sulfate. The fractions containing myrosinase activity were pooled and applied to a phenyl-sepharose FPLC column (Pharmacia) equilibrated with 50 mm Bistris/HCl, ph 6.1, 1 m sodium chloride. The myrosinase was found in the flow-through fraction and running a gradient down to 50 mm Bistris/HCl, ph 6.1, could elute no further activity. The myrosinase-containing fractions were dialysed against a 500-fold excess of 20 mm Tris/HCl, ph 8 or 50 mm Bistris/HCl, ph 6.1 at 4 8C overnight. The dialysate was applied to a small Q-Sepharose or MonoQ (Pharmacia) column equilibrated with 20 mm Tris/HCl, ph 8 or 50 mm Bistris/HCl, ph 6.1, respectively. The myrosinase was retained and eluted with a linear salt gradient. Myrosinase-containing fractions were pooled and stored at 4 8C. Catalytic characterization of aphid myrosinase The catalytic activity of the myrosinase was routinely determined using sinigrin as the substrate at 8.5 mm final concentration in 50 mm citrate buffer, ph 4.5 at 37 8C [12]. The reaction was stopped by heating at 95 8C and levels of liberated glucose were measured using a glucose oxidase reagent kit (Randox Laboratory Ltd, Ardmore, UK) according to the manufacturer's instructions. The turnover was kept well below 10% to ensure a linear reaction. Enzyme activity against linamarin and glucotropaeolin was assayed using the same conditions as for sinigrin. Ascorbic acid from a newly prepared solution was included in certain assays, but was not used routinely. Catalytic activity towards p-nitrophenyl-b-glucopyranoside was investigated using 3 mm final substrate concentration in 50 mm citrate buffer, ph 4.5 at 37 8C and overnight incubation. The reaction was stopped by addition of 1 m Na 2 CO 3. The amount of released p-nitrophenol was determined from absorbance measurements at 405 nm using p-nitrophenol as a standard. The structures of the substrates used are shown in Fig. 1. Kinetic properties were determined using 3 nm enzyme concentration under the same assay conditions, but with substrate concentrations ranging from 0.2 to 8.5 mm for sinigrin and 0.3 to 8.5 mm for glucotropaeolin. The incubations were stopped at different time points to assure low substrate conversion. Physico-chemical characterization of aphid myrosinase The subunit molecular mass of the aphid myrosinase was analysed by 10% SDS/PAGE according to Laemmli [13]. Samples were diluted in sample buffer containing 2-mercaptoethanol and heated to 95 8C for 5 min before analysis. Staining was achieved in Coomassie brilliant blue R-250. Western blot analysis of purified cabbage aphid myrosinase was performed using the mouse monoclonal antibody 3D7 raised against the oilseed rape myrosinases [12]. IEF was performed using precast IEF gels from Pharmacia on a Phast System (Pharmacia). The gels contained ampholines in the 3±9 ph range and the samples were applied as 1 ml drops on the middle of the gel after prefocusing. The gel was stained with Coomassie Fast R

3 q FEBS 2001 Myrosinase from cabbage aphid (Eur. J. Biochem. 268) 1043 Fig. 2. SDS/PAGE analysis of purified cabbage aphid myrosinase. Samples contained molecular mass standards with the masses indicated in kda (lane 1), 10 ml of homogenate supernatant (lane 2), 10 ml of Q-Sepharose pool 1 (lane 3), 10 ml of phenyl-sepharose flow-through fractions (lane 4), and 0.5 mg purified aphid myrosinase (lane 5). (Pharmacia) and the isoelectric point was calculated using a set of standard proteins (Pharmacia). Protein was determined according to Peterson [14] using BSA as the standard protein. The oligomerization state was determined using sucrose density gradient centrifugation analysis. Linear gradients were prepared from 5.5 ml each of 5% and 20% sucrose containing 50 mm Tris/HCl, ph 7.6. Sample diluted in NaCl/Pi was applied on top of the gradient. The standard proteins were: catalase, 240 kda (bovine liver, Sigma); IgG (swine, Dakopatts), 150 kda; BSA, 67 kda (RIA grade, USB); and cytochrome c, 13.7 kda (horse heart, Sigma). They were run on a separate gradient. The centrifugation was performed at 20 8C for 17 h at g in a RPS40T swing-out rotor. Subsequently fractions of four drops were collected from the bottom of the gradients using a peristaltic pump. The sample fractions were analysed for myrosinase activity and protein content by SDS/PAGE. As the sucrose was found to give a background activity in the myrosinase assay, corresponding fractions from the standard protein gradient were analysed to compensate for background activity. The standard proteins were localized Fig. 3. Western blot analysis using the monoclonal antibody 3D7 raised against oilseed rape myrosinases. The antibody was used as a hybridoma supernatant and diluted 1 : 100 before use and detection was accomplished using alkaline phosphatase-conjugated rabbit anti- (mouse IgG) (Dako) and 5-bromo-4-chloro-3-indolyl phosphate/ nitroblue tetrazolium (Sigma). Samples contained molecular mass standards with the masses indicated in kda (lane 1), oilseed rape crude extracts from seed (lane 2), cotyledon (lane 3) and leaf (lane 4), and purified cabbage aphid myrosinase (lane 5). The position of the aphid myrosinase on the Ponceau red-stained membrane is indicated with arrows. by assay of enzyme activity for catalase [15], absorbance at 280 nm for IgG and BSA, and absorbance at 409 nm for cytochrome c. The purified protein was subjected to amino-acid sequence analysis of internal peptides. The purified aphid enzyme was cleaved with trypsin and the digest was separated by a SMART system (Pharmacia) using a C4-reversed phase column and a water acetonitrile gradient containing 0.1% trifluoroacetic acid. The peaks were collected and several of the peaks were sequenced using Table 1 Purification scheme of cabbage aphid myrosinase. Volume (ml) Protein (mg) Specific activity (mmol min 21 mg 21 ) Purification factor Yield (%) Homogenate Extract Q-Sepharose Phenyl-Sepharose Q-Sepharose

4 1044 B. Pontoppidan et al. (Eur. J. Biochem. 268) q FEBS 2001 Fig. 4. Determination of isoelectric point for purified cabbage aphid myrosinase by IEF. Lanes 2 and 3 contain aphid myrosinase from two different preparations. A protein standard mix with the indicated isoelectric points was applied to lane 1 (amyloglucosidase 3.5; soybean trypsin inhibitor 4.55, b-lactoglobulin A 5.2; bovine carbonic anhydrase B 5.85; human carbonic anhydrase B 6.55; horse myoglobin 6.85, 7.35; lentil lectin 8.15, 8.45, 8.65; trypsinogen 9.3) Electrospray mass spectroscopy on a QTOF MS-MS instrument (Micromass, Manchester, UK). This sequence analysis cannot distinguish between leucine and isoleucine and often it is difficult to discriminate between lysine and glutamine residues. Leucine was arbitrarily chosen while glutamine was chosen in internal positions because tryptic cleavage should, in general, render a lysine to become a C-terminal residue. The resulting peptide sequences were analysed by BLAST. RESULTS In the initial screening of several generalist and specialist insect pests found on oilseed rape we could consistently record myrosinase activity in cabbage and mustard aphids using sinigrin as substrate. Aphids were deprived of plant material 24 h before analysis to minimize the possibility that the aphid enzyme activity might be due to contaminating plant myrosinase ingested by the insect. We could indeed purify the aphid myrosinase using three purification steps including anion-exchange and hydrophobicinteraction chromatography. The purification resulted in a homogeneous enzyme (Fig. 2). All myrosinase activity comigrated on the chromatographic columns indicating that Fig. 5. Effects of ascorbate (A) and temperature (B) on cabbage aphid myrosinase activity. (A) Different concentrations of ascorbate were included in the assay using 8.5 mm sinigrin and incubated at 37 8C for 35 min. The effects are shown as percentage activity of the aphid enzyme in the absence of ascorbate. (B) Effects of temperature on myrosinase enzyme activity. Approximately 30 ng of purified cabbage aphid myrosinase was incubated with 8.5 mm sinigrin at the temperatures indicated for 35 min only minor differences occur if more than one myrosinase isoform is present in the aphid. A typical purification (Table 1) resulted in a homogeneous preparation with. 95% pure protein after three chromatographic steps with a yield of 17% and a 160-fold purification factor. The purified enzyme was found to be very stable and could be stored for more than six months at 4 8C without any change in specific activity. However, freezing of the enzyme was found to cause loss of activity. The apparent subunit molecular mass was estimated as 57±58 kda using SDS/PAGE (Fig. 2). For one enzyme preparation the mass was estimated as 58.2 ^ 1.4 kda (mean ^ SD, n ˆ 3 on two different gels) and for another preparation as 57 ^ 0.5 kda (mean ^ SD, n ˆ 5 on two different gels). Western blot analysis of insect extracts and of the pure enzyme showed no reaction when probed with the monoclonal antibody 3D7 reactive with plant myrosinases (Fig. 3). The isoelectric point for the cabbage aphid myrosinase was determined using IEF on slab gels and was estimated to 4.9 ^ 0.1 (mean and SD for 11 measurements of two different enzyme preparations on four gels) (Fig. 4). No charge heterogeneity was observed upon the IEF suggesting only one myrosinase to be present in the final preparation. The bands were sharp indicating no or little glycosylation of the protein.

5 q FEBS 2001 Myrosinase from cabbage aphid (Eur. J. Biochem. 268) 1045 Fig. 6. Kinetic analysis of purified cabbage aphid myrosinase. Purified cabbage aphid myrosinase was incubated with varying concentrations of sinigrin (A) and glucotropaeolin (W). Approximately 30 ng of enzyme was used and the different incubations were stopped at different time points to ensure a low conversion of substrate in order to get an accurate estimate of initial velocities. The data are plotted according to Lineweaver±Burk. Oligomerization analysis using sedimentation velocity centrifugation in sucrose gradients showed a mass of approximately 126 kda when the enzyme activity was assayed in the gradient fractions (results not shown). Fractions from the gradient were subjected to SDS/PAGE analysis and silver staining to see whether any inactive oligomerization states were present. The results from densitometric analysis of the protein bands were superimposable on the activity curve. Accordingly the native myrosinase seems to be a dimer. No bands were present in the gradient except for the myrosinase defined by catalytic activity measurement further supporting the homogeneity of the final preparation. Effects of ascorbate on the purified aphid myrosinase were analysed. Ascorbate is known to activate plant myrosinases but the mechanism behind this effect is unknown [16,17]. The effects of ascorbate were tested in the concentration range 10 nm to 1 mm on the activity towards sinigrin. Concentrations around 0.3 mm ascorbate that strongly activate plant myrosinases were inhibitory to the cabbage aphid myrosinase (Fig. 5A). A slight, but insignificant, activation of the aphid myrosinase was observed with 10 nm ascorbate. In the routine measurements no ascorbate was added to the incubations. The effects of temperatures between 4 8C and 75 8C on the myrosinase activity were also studied (Fig. 5B). The Fig. 7. Alignment of six obtained peptide sequences with the Sinapis alba myrosinase `1MYR' and the Trifolium repens cyanogenic b-glucosidase `1CBG'. The catalytic nucleophile is marked with 1, residues involved in binding of the glucose ring with *, residues involved in aglycone binding in myrosinases with $ and the residue serving as general acid/base in O-b-glucosidases is marked with #.

6 1046 B. Pontoppidan et al. (Eur. J. Biochem. 268) q FEBS 2001 enzyme was found to have a temperature optimum with maximum activity around 40 8C using sinigrin as substrate (Fig. 5B). This is a common observation for hydrolytic enzymes due to increased reactivity of water and aminoacid side chains at an elevated temperature. The routine measurements were performed at 37 8C. The kinetic parameters of the purified aphid myrosinase were determined using sinigrin and glucotropaeolin as substrates. The K m and k cat values were estimated as 0.41 mm and 36 s 21 for sinigrin (Fig. 6) using the nonlinear regression analysis program SIMFIT [18]. Using glucotropaeolin as the substrate (Fig. 6) the K m and k cat values were estimated as 0.52 mm and 22.8 s 21, respectively. The cabbage aphid myrosinase was tested for O-bglucosidase activity towards p-nitrophenyl-b-glucopyranoside and linamarin. These compounds are standard substrates for assays of general b-glucosidases and cyanogenic b-glucosidases such as linamarase. The substrate concentrations used are at least ten-fold higher than the K m values reported for p-nitrophenyl-b-d-glucopyranosidase and maize b-glucosidase [19] or linamarin and cassava linamarase [20]. The enzyme was found to convert these substrates with specific activities of 0.24 and 0.02 mmol min 21 mg 21 protein, respectively. Peptide sequence was determined from selected peptides obtained after trypsin digestion and separation of peptides by HPLC. Electrospray mass spectroscopy gave ten peptide sequences that were analysed in BLAST searches to find homologues and aligned against known plant myrosinases using clustal w. Several peptides are homologous to plant myrosinases, but are more similar to O-b-glucosidases than to myrosinases. Six peptide sequences with homology to family 1 b-glycosidases are shown aligned against the crystallized MA-type myrosinase, `1MYR' from Sinapis alba [8], and the crystallized cyanogenic b-glucosidase, `1CBG' from Trifolium repens [21] (Fig. 7). Four peptides showed no significant homology to known sequences and are listed below the alignment (Fig. 7). DISCUSSION We were able to corroborate earlier reports of myrosinase activity in aphids [10,11] by the successful purification of a homogeneous myrosinase enzyme with catalytic activity towards two glucosinolates. There is always a risk that a herbivore may contain plant material and that the plant enzyme may be mistaken as being endogenous to the insect. In the case of the aphid myrosinase several factors suggest that the protein is normally present in the insect: (a) the insects were deprived of food prior to analysis to minimize the risk of plant enzymes contaminating the guts of the insects; (b) the size of the protein is smaller than that of any known plant myrosinases, which range from 62 to 77 kda; (c) the lack of reaction with the 3D7 antibody raised against oilseed rape myrosinases and with reactivity towards a wide range of other plant myrosinases from, for example, S. alba [22] and Arabidopsis thaliana [23]; (d) ascorbate inhibited the aphid enzyme preparation while plant myrosinases generally are stimulated by ascorbate; (e) the peptide sequences determined for the purified enzyme deviate from all known myrosinase sequences and are more similar to O-b-glucosidases. This is not surprising as many genes within the glucosidase family seem to share a similar gene organization and also use similar catalytic mechanisms suggesting a common origin and conservation of many amino-acid residues and overall structural features [2,8]. Although there is only 46% sequence identity between the crystallized myrosinase and linamarase, more than 90% of the Ca atoms can be superimposed [8]. Given this, significant differences also occur in that all known plant myrosinases share a small number of conserved amino-acid residues that distinguish them from O-b-glucosidases, most notably the substitution of glutamine for glutamate in the active site [8]. Also the residues involved in aglucone recognition are conserved among plant myrosinases and not found in O-b-glucosidases [8]. The myrosinaseglucosinolate system is considered to have evolved late in plant evolution [2] being present in a limited number of families and it is therefore expected that the insect myrosinase has evolved by the convergence of an O-b-glucosidase, perhaps as a consequence of host specialization. The aphid myrosinase showed some features similar to the plant myrosinases, however. We could show that the insect enzyme is also a dimer, has a relatively high temperature optimum, a low isoelectric point, and is active towards at least two structurally different glucosinolates, one aliphatic and one aromatic, which is indicative of low substrate specificity. The k cat of the aphid myrosinase is approximately 10 times higher than that of the unactivated plant enzyme from Raphanus sativus [17] or B. napus [24], but 10 times lower than that of the ascorbate-activated plant enzyme. Activity towards O-glucosides was low indicating that the enzyme is indeed a myrosinase and not a b-glucosidase with low substrate specificity. The sequence similarity of several peptides to hydrolytic enzymes also indicates conserved structural features common to hydrolytic enzymes. What then is the function of the aphid myrosinase? It has been shown that glucosinolates can be transported in the vascular system of B. napus [25]. Myrosinase was originally discovered in idioblastic myrosin cells of crucifers [26]. We have recently discovered that myrosinase can also be found in cells of the phloem in A. thaliana and B. napus [23]. In principle, phloem myrosinases should be able to hydrolyse glucosinolates if both glucosinolate and enzyme are ingested by the insect. We have, however, observed that fractions of free plant myrosinase decrease considerably upon wounding and that the enzyme is recovered in the insoluble cell fraction. It is thus possible that the aphids may themselves degrade ingested glucosinolates to increase the nutritional quality of the phloem. MacGibbon and Beuzenberg [11] considered a nutritional function of aphid myrosinase to be less likely as phloem in itself is rich in sugar. Instead the aphid myrosinase could give the insect access to the nitrogen- and sulfur-containing aglucone part of the glucosinolate. It is also possible that the enzyme may help minimize the deleterious effects of the toxic degradation products produced after plant myrosinase hydrolysis of glucosinolates. The insect myrosinase may give rise to hydrolysis products of lower toxicity. Myrosinase induction in a generalist insect, the desert locust (Schistocerca gregaria), was detected when they were fed on Brassicaceae and was

7 q FEBS 2001 Myrosinase from cabbage aphid (Eur. J. Biochem. 268) 1047 suggested to be an adaptation for short-term tolerance to plant allelochemicals [27]. If the aphid myrosinase is secreted in saliva it may play a role either in host recognition or detoxification [28]. Another possible function may be connected with aphid defence against natural enemies. The alarm pheromone of the mustard aphid contains isothiocyanates, possible degradation products of glucosinolates, together with (E)-b-farnesene [29]. An alarm pheromone with a host plant component might ensure that the defensive reaction would be aphid-species specific. Many aphids are dependent on bacterial symbionts for their survival. The bacteria help the aphids by transforming phloem sap into essential amino acids [30]. It is not known whether the insect myrosinase may actually be produced by and localized to the bacteria or if it is resident in insect cells. Myrosinase activity has been reported in certain bacteria [31,32]. The presence of myrosinase in an insect specialized to plants normally containing myrosinase is an interesting observation that may be interpreted as a coevolutionary process between the insect and the host plant. This knowledge is important when considering the consequences of engineering the myrosinase-glucosinolate system of brassica crops for improved resistance against pests. ACKNOWLEDGEMENTS This study was supported by the Nordic Joint Committee for Agricultural Research (NKJ), Swedish Council for Forestry and Agricultural Research (SJFR), Swedish Foundation for Strategic Research (SSF) and Lamms Stiftelse. We are grateful to B. Ek, I. Schenning and Y. Tillman for technical assistance, to S. Wretblad for the generous gift of glucotropaeolin and to C. HoÈgfeldt and S. Eriksson for taking care of the insects. REFERENCES 1. Karban, R. & Baldwin, I.T. (1997) Induced Responses to Herbivory. The University of Chicago Press. Chicago and London. 2. Rask, L., Andreasson, E., Ekbom, B., Eriksson, S., Pontoppidan, B. & Meijer, J. (2000) Myrosinase: gene family evolution and herbivore defence in Brassicaceae. Plant Mol. Biol. 42, 93± Xue, J., Lenman, M., Falk, A. & Rask, L. (1992) The glucosinolate-degrading enzyme myrosinase in Brassicaceae is encoded by a gene family. Plant Mol. Biol. 18, 387± Fenwick, G.R., Heaney, R.K. & Mullin, W.J. (1983) Glucosinolates and their breakdown products in food and food plants. Crit. Rev. Food Sci. Nutr. 18, 123± Halkier, B.A. (1999) Glucosinolates. In Naturally Occurring Glycosides (Ikan, R., ed.), pp. 193±223. John Wiley & Sons Ltd., Chichester. 6. Henrissat, B. (1991) A classification of glycosyl hydrolases based on amino acid sequence similarities. Biochem. J. 280, 309± Henrissat, B. & Bairoch, A. (1996) Updating the sequencebased classification of glycosyl hydrolases. Biochem. J. 316, 695± Burmeister, W.P., Cottaz, S., Driguez, H., Iori, R., Palmieri, S. & Henrissat, B. (1997) The crystal structures of Sinapis alba myrosinase and a covalent glycosyl-enzyme intermediate provide insights into the substrate recognition and active-site machinery of an S-glycosidase. Structure 5, 663± Ekbom, B. (1995) Insect pests. In Brassica Oilseeds. Production and Utilization (Kimber, D.S. & McGregor, D.I., eds), pp. 141±152. CAB International, Wallingford, UK. 9a. Pickett, J.A., Wadhams, L.J., Woodcock, C.M. & Hardie, J. (1992) The chemical ecology of aphids. Ann. Rev. Entomology 37, 67± MacGibbon, D.B. & Allison, R.M. (1968) A glucosinolase system in the aphid Brevicoryne brassicae. N. Z. J. Sci. 11, 440± MacGibbon, D.B. & Beuzenberg, E.J. (1978) Location of glucosinolase in Brevicoryne brassicae and Lipaphis erysimi (Aphididae). N. Z. J. Sci. 21, 389± Lenman, M., RoÈdin, J., Josefsson, L.-G. & Rask, L. (1990) Immunological characterisation of rapeseed myrosinase. Eur. J. Biochem. 194, 747± Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680± Peterson, G.L. (1977) A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal. Biochem. 83, 346± Aebi, H. (1984) Catalase in vitro. Methods Enzymol. 105, 121± Ohtsuru, M. & Hata, T. (1979) The interaction of l-ascorbic acid with the active center of myrosinase. Biochim. Biophys. Acta 567, 384± Shikita, M., Fahey, J.W., Golden, T.R., Holtzclaw, W.D. & Talalay, P. (1999) An unusual case of `uncompetitive activation' by ascorbic acid: purification and kinetic properties of a myrosinase from Raphanus sativus seedlings. Biochem. J. 341, 725± Bardsley, W.G. (1993) SIMFIT. A computer package for simulation, curve fitting and statistical analysis using life science models. In Modern Trends in Biothermokinetics (Schuster, S., Rigoulet, M., Ouhabi, R. & Mazat, J.P., eds), pp. 455±458. Plenum Publishing Corporation, New York, USA. 19. Cicek, M., Blanchard, D., Bevan, D.R. & Esen, A. (2000) The aglycone specificity-determining sites are different in 2, 4-dihydroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA)- glucosidase (Maize beta-glucosidase) and dhurrinase (Sorghum beta-glucosidase). J. Biol. Chem. 275, 20002± Eksittikul, T. & Chulavatnatol, M. (1988) Characterization of cyanogenic beta-glucosidase (linamarase) from cassava (Manihot esculenta Crantz). Arch. Biochem. Biophys. 266, 263± Barrett, T., Suresh, C.G., Tolley, S.P., Dodson, E.J. & Hughes, M.A. (1995) The crystal structure of a cyanogenic b-glucosidase from white clover, a family 1 glycosyl hydrolase. Structure 3, 951± HoÈglund, A.-S., Lenman, M. & Rask, L. (1992) Myrosinase is localised to the interior of myrosin grains and is not associated to the surrounding tonoplast membrane. Plant Sci. 85, 165± AndreÂasson, E. (2000) Structural and Functional Studies of the Myrosinase-Glucosinolate System in Arabidopsis thaliana and Brassica napus. PhD Thesis, Swedish University of Agricultural Sciences, Uppsala, Sweden. 24. James, D.C. & Rossiter, J.T. (1991) Development and characteristics of myrosinase in Brassica napus during early seedling growth. Physiol. Plant 82, 163± Brudenell, A.J.P., Griffiths, H., Rossiter, J.T. & Baker, D.A. (1999) The phloem mobility of glucosinolates. J. Exp. Bot. 50, 745± Bones, A. & Rossiter, J.T. (1996) The myrosinase-glucosinolate system, its organisation and biochemistry. Physiol. Plant 97, 194± Mainguet, A.M., Louveaux, A., El Sayed, G. & Rollin, P. (2000) Ability of a generalist insect, Schistocerca gregaria, to overcome thioglucoside defence in desert plants: tolerance or adaptation? Entomol. Exp. Appl. 94, 309±317.

8 1048 B. Pontoppidan et al. (Eur. J. Biochem. 268) q FEBS Urbanska, A., Tjallingii, W.F., Dixon, A.F.G. & Leszczynski, B. (1998) Phenol oxidising enzymes in the grain aphid's saliva. Entomol. Exp. Appl. 86, 197± Dawson, G.W., Griffiths, D.C., Pickett, J.A., Wadhams, L.J. & Woodcock, C.M. (1987) Plant-derived synergists of alarm pheromone from turnip aphid Lipaphis (Hyadaphis) erysimi (Homoptera, Aphididae). J. Chem. Ecol. 13, 166± Douglas, A.E. (1998) Nutritional interactions in insect±microbial symbioses: aphids and their symbiotic bacteria Buchnera. Annu. Rev. Entomol. 43, 17± Oginsky, E.L., Stein, A.E. & Greer, M.A. (1965) Myrosinase activity in bacteria as demonstrated by the conversion of progoitrin to goitrin. Soc. Exp. Biol. Med. Proc. 119, 360± Tani, N., Ohtsuru, M. & Hata, T. (1974) Purification and general characteristics of bacterial myrosinase produced by Enterobacter cloacae. Agr. Biol. Chem. 38, 1623±1630.