Purification and characterization of β-glucosidase from honey bees (Apis mellifera)

Insect Biochemistry and Molecular Biology 32 (2002) 679–690 www.elsevier.com/locate/ibmb

Purification and characterization of β-glucosidase from honey bees (Apis mellifera) J. Pontoh, N.H. Low

*

Department of Applied Microbiology and Food Science, University of Saskatchewan, Saskatoon, Canada Received 9 July 2001; received in revised form 15 August 2001; accepted 27 September 2001

Abstract β-glucosidase has been purified from the ventriculus and honey sac of Apis mellifera using a combination of anion- and cationexchange, hydroxyapatite and gel-permeation chromatography. In addition, β-glucosidase from the hypopharyngeal glands has been partially purified using anion-exchange and gel-permeation chromatography. The purified β-glucosidase gave a positive result by glycoprotein staining. This β-glucosidase consists of only one subunit and has Mr of 72 kDa as determined by SDS–PAGE. IEF– PAGE showed several bands with pIs ranging from 4.5 to 4.8. These multiform proteins have been proposed as having different degrees of glycosylation. The pH optimum of the purified β-glucosidase from the ventriculus and honey sac are 5.0. These enzymes were stable at temperatures up to 50 °C and have a relatively wide pH stability range of 4.0 to 9.0. MALDI-TOF-MS peptide mass maps of purified β-glucosidase from the ventriculus, honey sac and hypopharyngeal glands showed six matching masses. These results indicate that the β-glucosidase isolated from the hypopharyngeal glands, honey sac and ventriculus is the same. It is proposed that β-glucosidase is produced in the hypopharyngeal glands, secreted into the mouth during feeding and then passes to the honey sac. From the honey sac, this enzyme is transferred into honeycomb cells and the ventriculus.  2002 Elsevier Science Ltd. All rights reserved. Keywords: β-Glucosidase; Purification; Characterization; Honey bee; Ventriculus; Honey sac; Hypopharyngeal glands; MALDI-TOF-MS

1. Introduction β-Glucosidase (EC 3.2.1.21; β-d-glucoside glucohydrolase) hydrolyses terminal, non-reducing 1,4linked-β-glucose residues releasing β-d-glucose from oligo- or polysaccharides (Webb, 1992). Other names for this enzyme are gentiobiase and cellobiase. β-Glucosidase occurs ubiquitously in plants, animals, fungi and bacteria (Esen, 1993). In insects, β-glucosidase has been reported from various orders and families, with the majority having been isolated from the intestinal tract (Terra and Ferreira, 1994). β-Glucosidase has been subdivided into three classes based on substrate specificity. Class 1 includes enzymes with glycosyl β-glycosidase and aryl β-glycosidase activity; these enzymes have the ability to hydrolyze cellobiose, lactose, β-p-nitrophenylglucoside (β-PNPG), β-p-nitrophenylgalactoside (βPNPGal), β-p-nitrophenylfructoside (β-PNPFru) and

* Corresponding author.

other similar substrates. Class 2 includes those with only glycosyl β-glucosidase activity; therefore, they can only hydrolyze substrates such as cellobiose and lactose. Class 3 includes enzymes with only aryl (or alkyl) βglucosidase activity; these enzymes would have significant activity towards β-PNPG and similar substrates (Terra and Ferreira, 1994). Purification of digestive enzymes is one of the main obstacles in the study of insect digestion (Applebaum, 1985). Digestive β-glucosidases from insects have been studied in their crude form because of difficulties in purification and include those from Locusta migratoria (Morgan, 1975), Rhodnius prolixus (Terra et al., 1988), Pheropsophus aequinoctialis (Ferreira and Terra, 1989), Tenebrio molitor (Terra et al., 1985), Pyrearinus termitilluminans (Colepicolo et al., 1986), Sophrorhinus insperatus (Adedire and Balogun, 1995), and Scaptotrigona bipunctata (Schumaker et al., 1993). Partial purification of β-glucosidase from the digestive tracts of Abracris flavolineata (Marana et al., 1995), Rhagium inquisitor (Chipoulet and Chararas, 1985), Sitophiulus

0965-1748/02/$ - see front matter  2002 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 5 - 1 7 4 8 ( 0 1 ) 0 0 1 4 7 - 3

680

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

oryzae (Baker and Woo, 1992), Rhynchosciara americana (Ferreira and Terra, 1983), Dysdercus peruvianus (Silva et al., 1996), and Erinnyis ello (Santos and Terra, 1985) has been accomplished using ammonium sulfate precipitation, hydroxyapatite treatment and gel electrophoresis. Chararas and Chipoulet (1982), Pratviel et al. (1987) and Marana et al. (2000) have purified β-glucosidase from Phoracantha semipunctata, Thaumetopoea pityocampa and Spodoptera frugiperda, respectively, employing ammonium sulfate precipitation, hydroxyapatite and gel-permeation chromatography and chromatofocusing. Several forms of digestive β-glucosidase have been isolated from the midgut of insects. Baker and Woo (1992) purified two β-glucosidases from S. oryzae with Mrs of 140 and 170 kDa which showed different substrate specificities. Marana et al. (1995) isolated two βglucosidases from A. flavolineata having different pIs and substrate specificities but the same Mr (82 kDa). Terra and co-workers found several forms of β-glucosidase from insects studied based on their electrophoretic mobility and substrate specificity (Ferreira and Terra, 1983; Ferreira et al. 1997, 1998). β-Glucosidase activity has been detected in honey (Low et al., 1986), and this activity was correlated with the formation of β-O-glycosidic linked oligosaccharides in this food (Low et al., 1988). Gilliam et al. (1988) detected the presence of β-glucosidase activity in the midgut and hindgut of honey bee (Apis mellifera) and Ferreira et al. (1998) isolated β-glucosidase from the midgut of Scaptotrigona bipunctata, the same family (Apidae) as the honey bee. The origin of β-glucosidase activity in honey has not been identified, but it could be from the honey bee. If the origin of β-glucosidase activity in honey is from the bee, then β-glucosidase should be present in the honey sac and in the organs secreting digestive enzymes in the mouth. Several glands discharging secretions in the mouthparts of the honey bee have been identified including thoracic, head and hypopharyngeal glands (Snodgrass and Erickson, 1992). The presence of α-glucosidase in the hypopharyngeal glands of the honey bee has been confirmed and has been related to α-glucosidase activity in honey (Simpson et al., 1968). In this study, we report the isolation and purification of β-glucosidase from the hypopharyngeal glands, honey sac and ventriculus of the honey bee (A. mellifera). In addition, a number of physicochemical and kinetics properties of this enzyme have been determined. 2. Materials and methods 2.1. Honey bee dissection and raw extract preparation Honey bee (Apis mellifera) was collected from several hives within a 20 km radius of the city of Saskatoon,

Canada, during three consecutive summers from 1995 to 1997. Honey bees were killed on site by treatment with dry-ice and were immediately frozen (⫺30 °C). In general, the collected honey bees used for this experiment were adult workers based on fore-wing size (苲9 mm). Each frozen (⫺30 °C) honey bee was cut to remove the head and abdomen from the discarded thorax. After thawing (5 °C), the abdomen was dissected by peeling off both dorsal and ventral plates with two pairs of #5 Dumont forceps (A. Dumont & Fils, Autils, Switzerland) to expose the honey sac and ventriculus. The honey sac was then removed and placed in a glass test-tube (13×100 mm). The exposed ventriculus was then disconnected from the hindgut and placed in a separate testtube. The hypopharyngeal glands were isolated by first breaking the occiput at the back of the head, and then the glands were removed by forceps. All dissections were performed using a stereomicroscope (WILD, Heerbrugg, Switzerland). Approximately 250 honey bees were required for βglucosidase isolation from the ventriculus, approximately 1000 honey bees were required for β-glucosidase isolation from the honey sac, and approximately 25 honey bees were required for partially purification of βglucosidase from the hypopharyngeal glands. Collected ventriculus, honey sacs or hypopharyngeal glands were mashed with a glass rod (7×17 mm) in the presence of glass beads (acid washed, 150–212 µm; Sigma Chemical Co., St Louis, MO, USA) in a test-tube (16×100 mm) containing approximately 5 ml phosphate buffer (0.05 M, pH 7.5) and then diluted into approximately 15 ml of the same buffer. After extraction the mixture was centrifuged (Sorvall SS-1 Superspeed Angle Centrifuge; Ivan Sorvall Inc., Norwalk, CT, USA) for 30 min at 6000g, and filtered through glass-wool (Sigma Chemical Co.) to remove suspended particulates. This extraction was repeated and the combined extracts were made up to a total volume of 45 ml for ventriculus, 90 ml for honey sac and 15 ml for hypopharyngeal glands with the same buffer. This solution is designated as the raw extract in glucosidase purification. 2.2. Purification In addition to β-glucosidase, two α-glucosidases were also purified. All chromatographic purification techniques were performed using an FPLC system (Pharmacia, Sweden) with UV detector (Pharmacia) at 280 nm in the cold room (4 °C). The purification of these α-glucosidases will be presented and discussed elsewhere (Pontoh and Low, in preparation). A 20×1.0 cm column (Bio-Rad Laboratories, Hercules, CA, USA) packed with approximately 8 ml of DEAE Sepharose FF (Pharmacia Biotech, Uppsala, Sweden; anion exchange) was used. Approximately 45 ml of raw extract was added to the column employing

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

a mobile phase of 0.05 M phosphate buffer at pH 7.5 at a flow rate of 1 ml/min. The absorbed proteins were eluted from the column employing the same mobile phase with a linear gradient of NaCl from 0.0 to 0.5 M. Fractions (7 to 42 ml) containing both α- and β-glucosidase activities (Fig. 1A) were collected and transferred into an Amicon Stirred Cell fitted with a YM-30 membrane (Amicon Inc., Beverly, MA, USA) for concentration to approximately 3 ml for gel-permeation chromatography. A 500 µl aliquot of the concentrated sample was injected into a 50×1.0 cm column (Bio-Rad Laboratories) packed with Preparative Grade Superose 12 (Pharmacia Biotech) and proteins were eluted with a mobile phase of 0.01 M buffer phosphate at pH 6.8 at a flow rate of 0.4 ml/min. The fractions (11 to 15 ml) containing both α- and β-glucosidase activities were collected and transferred into an Amicon Stirred Cell and the buffer was changed to 0.05 M acetate at pH 4.5 containing 10% glycerol. Approximately 30 ml of the fractions containing both α- and β-glucosidase activities from the previous purification was added to a 20×1.0 cm column (Bio-Rad Laboratories) packed with CM Sepharose FF (Pharmacia Biotech; cation exchange) employing a mobile phase of 0.05 M acetate buffer at pH 4.5 containing 10% glycerol at a flow rate of 1 ml/min. Proteins were eluted from the column employing a mobile phase of the same buffer with a stepwise linear gradient from 0.05 to 0.5 M. Fractions (110 to 126 ml) containing βglucosidase activity (Fig. 1B) were collected for further purification.

681

Pooled fractions containing β-glucosidase activity from CM Sepharose FF purification were diluted 1:3 with 10% glycerol to reduce buffer molarity and were then applied to a pre-packed Mono-S (0.5×5 cm; Pharmacia Biotech) column employing a mobile phase of 0.05 M acetate buffer at pH 4.5 containing 10% glycerol at a flow rate of 0.4 ml/min. Proteins were eluted from the column employing a mobile phase of the same buffer with a stepwise linear gradient of the same buffer from 0.05 to 0.5 M at pH 4.7 containing 10% glycerol. Fractions (54 to 57 ml) containing β-glucosidase activity were collected for further purification. Fractions containing β-glucosidase activity from Mono-S purification were quantitatively transferred into an Amicon stirred cell to change the buffer to 0.05 M phosphate at pH 6.8. Approximately 25 ml of this solution were applied to a Macro-Prep Ceramic Hydroxyapatite (Type I, 40 µm; Bio-Rad Laboratories) column (1.0×8.0 cm) that had been equilibrated with 0.01 M phosphate buffer at pH 6.8 at a flow rate of 0.4 ml/min. Proteins were eluted from the column employing a mobile phase of the same buffer with a linear gradient of the same buffer from 0.01 to 0.3 M at pH 6.8. Fractions (30 to 35 ml) containing β-glucosidase activity (Fig. 1C) were collected for further purification. Fractions containing β-glucosidase activity from hydroxyapatite chromatography were concentrated with a Centricon to a final volume of 200 µl. This sample was then injected into the Superose 12 (Pharmacia Biotech) column employing a mobile phase of 0.01 M phosphate buffer at pH 6.8 at a flow rate of 0.4 ml/min. Fractions (12.5 to 15.5 ml) containing β-glucosidase activity (Fig. 1D) were collected for further analysis.

Fig. 1. Selected chromatograms of β-glucosidase purification from the ventriculus of A. mellifera: (A) DEAE Sepharose FF; (B) CM Sepharose FF; (C) hydroxyapatite; (D) Superose 12.

682

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

A partially purified β-glucosidase sample from the hypopharyngeal glands was obtained by subjecting the raw extract to DEAE Sepharose FF and Superose 12 chromatographic treatments as previously outlined. The fractions containing β-glucosidase activity were collected and concentrated to 苲200 µl employing a Centricon. The concentrated sample was further purified using gel electrophoresis (Section 2.6).

buffers used for these particular experiments were 0.05 M glycine–hydrochloric acid (HCl) for pH 2.0–3.5, 0.05 M sodium acetate for pH 3.5–5.5, 0.05 M sodium phosphate for pH 5.5–8.5, 0.05 M glycine–sodium hydroxide (NaOH) for pH 8.5–11.0, and 0.05 M sodium phosphate–NaOH for pH 11.0–12.5. The resulting solutions were incubated at 25 °C for 20 min. Residual enzyme activity was determined with β-PNPG following standard assay conditions.

2.3. Enzyme assay β-Glucosidase activity was assayed employing a modification of the Low method (Low et al., 1986). A selected volume of sample was diluted with its sample buffer to a total volume of 0.1 ml and was added to 1 ml of 0.02 M β-PNPG in 0.1 M acetate buffer at pH 5.0 in a test-tube. The mixture was then incubated in a water bath (35 °C) for 20 min. To this solution was added 0.1 ml of Tris buffer (3 M, pH 10.0). After cooling to room temperature, the solution was transferred into a spectrophotometer cuvette and the absorbance was measured at 400 nm using a spectrophotometer (Spectronic 1201; Milton Roy Inc., Rochester, NY, USA). One unit of β-glucosidase activity is defined as the amount of enzyme required to produce 1 µmol of pnitrophenol per minute at the aforementioned assay conditions. 2.4. Protein assay Protein concentration was determined using the micro assay procedure of Bio-Rad (Bio-Rad Protein Assay; Bio-Rad Laboratories) which is based on the Bradford method (Bradford, 1976). A standard curve was established using bovine serum albumin with a concentration range from 1.2 to 10.0 µg/ml.

2.5.2. Glucosidase activity Effects of temperature and pH on the activity of purified β-glucosidase were examined employing β-PNPG as the substrate. Samples which were very high in enzyme activity, such as those obtained from the ventriculus, were diluted 10 times with 0.01 M phosphate buffer at pH 6.8 prior to use in these experiments. Sample volumes used varied from 2.5 to 75 µl in order to produce absorbance readings between 0.100 and 0.900 under standard assay conditions. Regardless of the initial sample volume used, each was made up to 100 µl with 0.01 M phosphate buffer at pH 6.8. Samples were added to 1.0 ml of β-PNPG solutions (0.02 M) at various pHs ranging from 3.0 to 10.0 at 0.5 unit increments. The substrate solutions were preheated in a water bath at temperatures ranging from 20 to 75 °C at 5 °C increments. Therefore, there were 180 temperature–pH treatment combinations. After incubation for 20 min, 100 µl of 3.0 M Tris buffer at pH 10.0 was added to each temperature– pH treatment and the absorbance was measured at 400 nm. Enzyme activities for each combination condition were plotted as a response surface, using Microsoft Excel 97 (Microsoft Corporation, Redmond, WA, USA). This methodology was based on protocols outlined by Bergmeyer (1983) and Rautela et al. (1979). 2.6. Gel electrophoresis

2.5. Effects of pH and temperature 2.5.1. Glucosidase stability. The temperature stability of β-glucosidase was determined as follows. Five µl of purified β-glucosidase (previously concentrated so as to give absorbance readings between 0.100 to 0.900 with β-PNPG) was diluted to a final volume of 100 µl with preheated 0.05 M phosphate buffer at pH 6.8. Samples were incubated in a preheated water bath for 20 min at a series of temperatures ranging from 20 to 75 °C at 5 °C intervals. Residual enzyme activity was determined with β-PNPG following standard assay conditions. The pH stability of β-glucosidase was determined as follows. Five µl of purified β-glucosidase (previously concentrated so as to give absorbance readings between 0.100 and 0.900 with β-PNPG) was diluted to a final volume of 100 µl with the appropriate buffer so that the pH ranged from 2.0 to 12.5 at 0.5 unit increments. The

All PAGE (polyacrylamide gel electrophoresis) analyses were performed using a Mini-gel system (Bio-Rad Laboratories) equipped with a power supply, Model EPS 500/400 (Pharmacia Biotech) at 4±1 °C. The gels were cast with 1.0 mm spacers (Bio-Rad Laboratories) and a 0.75 mm 10-well comb (Bio-Rad Laboratories). For SDS (sodium dodecyl sulfate)–PAGE, the purified β-glucosidase samples were transferred into microcentrifuge tubes (500 µl) and vacuum centrifuge freeze-dried (Heto-Holten, Allerod, Denmark). Sample buffer (10 µl) of 125 mM Tris–HCl buffer at pH 6.8 containing glycerol (10% v/v), SDS (2% w/v), mercaptoethanol (ME; 5% v/v) and bromphenol blue (0.01% w/v) was added to each microcentrifuge tube (500 µl) and the resulting protein solution was denatured by heating in boiling water for 4 min. For discontinuous SDS–PAGE, a 4% polyacrylamide stacking gel (pH 6.8) with a 7.5% polyacrylamide separ-

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

ating gel (pH 8.3) was placed in a 200 mV electric field for 35 min; and a gradient (4–20%) polyacrylamide gel (Ready Gels, Bio-Rad Laboratories; pH 8.3) was run in a 200 mV electric field for 50 min. Proteins were visualized with Coomassie blue, silver nitrate or GelCode glycoprotein staining kit (Pierce Chemical Company, Rockford, Il). The Coomassie blue staining was performed following the method described by Bollag and Edelstein (1991). Silver staining was performed as outlined by Rosenberg (1996). Glycoproteins were stained using a GelCode Glycoprotein Staining Kit following the suggested protocol of the manufacturer. Native IEF (isoelectric focusing)–PAGE was used following the method outlined by Bollag and Edelstein (1991) employing a Mini-Protean II Cell (Bio-Rad Laboratories). Samples from each final purification step were individually vacuum centrifuge freeze-dried for 90 min employing a vacuum centrifuge freeze-dryer. The resulting powder was dissolved in 15 µl of sample buffer prepared by mixing 3 ml glycerol, 200 µl ampholyte solution and 1.8 ml ultrapure water. Approximately 5 µg of purified or standard protein mixture was loaded per well. An IEF protein standard (0.5 µl; Bio-Rad Laboratories) was run concurrently. IEF–PAGE was initially run at 100 mV for 1 h, then at 250 mV for 1 h, and finally at 500 mV for 30 min. Prior to staining, the ampholyte was removed while simultaneously fixing the proteins by placing the gel in 10% trichloroacetic acid for 20 min, followed by overnight agitation in a 1% trichloroacetic acid solution. Proteins were stained with Coomassie blue and/or by the glycoprotein method. Glucosidases were localized on the gel by excising into 2 mm bands which were then assayed for β-glucosidase activity using β-PNPG as the substrate. The pH profile of the gel was determined by suspending the band in 10 mM KCl followed by pH measurement (accumet, Model 15, Fisher Scientific, Pittsburgh, PN, USA). The pH gradient of the gel was also estimated from IEF protein standards. 2.7. Enzyme kinetics Kinetic properties (Km and Vmax) of purified β-glucosidase were estimated following the procedure outlined by Stauffer (1989). Substrates for β-glucosidase included β-PNPG and cellobiose. Substrate concentrations were dependent on Km values. For β-PNPG and cellobiose the concentration ranges were 1–20 mM and 5–100 mM, respectively. The hydrolytic reaction conditions for βglucosidase with each of these substrates was 0.10 M acetate buffer at pH 5.0 and 35 °C. Sample aliquots (1.1 ml) were withdrawn from the hydrolysis reaction mixture at 2.5-min intervals for a total reaction time of 15 min. Glucosidase activity was terminated after aliquot removal either by the addition

683

of 100 µl of 3.0 M Tris buffer at pH 10 (β-PNPG) or by heating at 90 °C for 2 min (cellobiose). Products from enzymatic hydrolysis were analyzed either spectrophotometrically (p-nitrophenol) or by highperformance anion-exchange liquid chromatography (HPAE; glucose). The HPAE analysis of hydrolyzed cellobiose was performed on a Dionex Bio LC 4000 gradient liquid chromatographic system (Dionex, Sunnyvale, CA, USA) with a Dionex CarboPac PA1 column (4×250 mm) in series with a CarboPac PA1 guard column (3×25 mm). All separations were carried out isocratically with a mobile phase of 40 mM NaOH. The flow rate was 1 ml/min and the sample injection volume was 50 µl. Non-hydrolyzed cellobiose was eluted using a mobile phase of 300 mM NaOH, followed by re-equilibration of the column with 40 mM NaOH. Total analysis time for each sample was 20 min. Carbohydrate detection was accomplished using a pulsed amperometric detector (PAD, Dionex) with a gold electrode (2.5 mm diameter) at a sensitivity of 10 µA. The electrode was maintained at the following potentials and durations: E1=0.05 V (t1=120 ms); E2=0.60 V (t2=120 ms); E3=⫺0.80 V (t3=420 ms) (Swallow and Low, 1990). The initial hydrolytic reaction rate at each substrate concentration was calculated by plotting product concentration versus sampling time. A plot of these initial reaction rates versus their corresponding substrate concentrations generated the Michaelis–Menten curve. The maximum velocity (Vmax) and Michaelis–Menten constant (Km) were calculated graphically by plotting the ratio of substrate concentration and initial rate versus substrate concentration (Hanes Plot). All calculations and graphs were done using Microsoft Excel 5.2 (Microsoft Corporation). 2.8. Peptide mass mapping A blank gel (2×6 mm) from SDS–PAGE (Section 2.6) and treated with Coomassie blue was cut into 苲1 mm pieces and placed in a microcentrifuge tube (500 µl; Eppendorf, Hamburg, Germany). The gel was destained by washing three times with 100 µl of 25 mM NH4HCO3 in 50% acetonitrile (CH3CN) for approximately 10 min each. The gel was then dehydrated by the addition of 50 µl of 100% CH3CN for 10 min followed by vacuum centrifuge freeze-drying for 25 min. The gel was rehydrated with 40 µl of trypsin solution (20 µg of trypsin [Promega, Madison, WI] in 800 µl of 25 mM NH4HCO3) and placed in a cold room (4 °C) for 45 min. The tube was then transferred into a water bath (37 °C) and incubated for 15 h. After incubation, 100 µl of ultrapure water was added into the microcentrifuge tube and sonicated for 10 min. The supernatant was removed and transferred into a fresh microcentrifuge tube. The residual peptides in the gel were further extracted with

684

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

50 µl of 50% acetonitrile in 5% trifluoroacetic acid and pooled with the aforementioned supernatant. Bovine serum albumin was used as the external standard in this research. Approximately 15 µg of BSA in 10 µl sample buffer (125 mM Tris–HCl at pH 6.8) containing glycerol (10% v/v), SDS (2% w/v), mercaptoethanol (ME; 5% v/v) and bromphenol blue (0.01% w/v) was transferred into a microcentrifuge tube (500 µl) and denatured by heating in a boiling water bath for 4 min. The denatured BSA was subjected to SDS–PAGE, stained with Coomassie blue, and the BSA band was dissected into 苲1 mm pieces and treated using the same protocol as the blank. Purified β-glucosidase samples (苲0.1 µg) from the ventriculus and honey sac were transferred individually into microcentrifuge tubes (500 µl) and vacuum centrifuge freeze-dried. Following the addition of 15 µl of SDS–PAGE sample buffer (50 mM Tris–HCl at pH 6.8) containing SDS (2% v/v) and mercaptoethanol (ME; 15 mM), the samples were transferred to a boiling water bath for 4 min. The denatured samples were then analyzed by SDS–PAGE. The resulting gels for these samples were stained with a modified silver staining protocol (Shevchenko et al., 1996). The β-glucosidase bands were cut into 苲1 mm pieces and transferred into microcentrifuge tubes (500 µl). The gels were destained by washing three times with 100 µl of 15 mM potassium ferricyanide in 50 mM sodium thiosulfate (Gharahdaghi et al., 1999) for 10 min each. The gels were dehydrated with 50 µl of 100% CH3CN for 10 min followed by vacuum centrifuge freeze-drying for 25 min. Samples were then tryptically digested and processed as outlined for the blank gel. A concentrated sample of partially purified β-glucosidase from the hypopharyngeal glands was transferred into a microcentrifuge tube (500 µl), vacuum centrifuge freeze-dried and denatured following the same sample preparation protocol followed for β-glucosidase from the ventriculus. The denatured proteins were subjected to SDS–PAGE analysis and modified silver stained. The band correlating to the Mr of β-glucosidase from the ventriculus and honey sac (苲72 kDa) was excised and treated following the outlined protocol for purified βglucosidase from the ventriculus. Tryptic peptides from each of the aforementioned samples were individually desalted and concentrated using a C18 ZipTip (0.6 µl bed of C18; Milipore Corporation, Bedford, MA, USA). Bound peptides were eluted directly onto the MALDI (matrix-assisted laser desorption/ionization) target plate using 75:25 acetonitrile:water containing 0.1% TFA and 5 mg/ml α-cyano-4hydroxycinnamic acid (UV-absorbing matrix). Positive-ion MALDI mass spectra were acquired on a Voyager-DE STR (PerSeptive Biosystems, Framingham, MA, USA) equipped with a N2 laser (337 nm). The instrument was operated in the reflectron mode using

delayed extraction, with an accelerating voltage of 20 kV and a grid voltage of 72.5%. Data were acquired using 200 laser shots per spectrum, and close external calibration was done using des-Arg Bradykinin (m/z 904.4681) and ACTH (adrenocorticotropic hormone) fragment 1–17 (m/z 2093.0861). Peptide matching for sample mass spectra was based on an accuracy of ±1 Da. Peptide masses were matched to the SWISSPROT (released 39.8, November 2000) and TrEMBL (released 15.4, October 2000) databases using ProteinProspector v 3.2.1 (http://prospector.ucsf.edu/).

3. Results 3.1. b-Glucosidase localization In order to determine the best organ in A. mellifera for protocol development for β-glucosidase isolation and purification, the β-glucosidase activity in the raw extract from the hypopharyngeal glands, honey sac, ventriculus and hindgut was determined. β-Glucosidase activity for each of these organs is shown in Table 1. The ventriculus had the highest, followed by the hypopharyngeal glands and hindgut, with the honey sac showing the lowest βglucosidase activity. Based on these results, the ventriculus was used as the starting organ for β-glucosidase purification. 3.2. b-Glucosidase purification The effects of each purification step on total protein, activity, specific activity and purification values from the ventriculus and honey sac are shown in Tables 2 and 3. Purification value is defined as the ratio of specific activity achieved at a specific purification step by the specific activity of the raw extract. In addition to β-glucosidase, α-glucosidase was also purified following the same protocol, and these results will be reported elsewhere (Pontoh and Low, in preparation). The majority of β-glucosidase did not bind to the anion exchanger (DEAE Sepharose FF) at pH 7.5 (Fig. 1A); however, this step afforded the removal of a sigTable 1 Mean and standard deviation of total activity (units) for β-glucosidase (β-glucosidase as substrate) in selected honey bee organs of A. mellifera Organs

β-Glucosidase activity (milliunits)a

Hypopharyngeal glands Honey sac Ventriculus Hindgut

19±5 30±6 116±17 55±2

a

Based on duplicate studies of 10 honey bees for each study.

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

685

Table 2 Effect of purification steps on the total protein, activity, specific activity and purification value, and yield of β-glucosidase from the ventriculus (250 bees) of A. mellifera Purification method

Raw extract DEAE Sepharose FF Superose prep. grade CM Sepharose FF Mono S Hydroxyapatite Superose 12

Protein (mg)

84.5 48.4 11.9 1.7 0.91 0.24 0.05

β-Glucosidase Activity (units)

Specific activity (unit/mg)

Purification value

Yield (%)

8.01 5.15 1.62 1.05 0.72 0.28 0.26

0.10 0.11 0.14 0.63 0.79 1.14 5.66

1.00 1.12 1.44 6.67 8.38 12.0 59.7

100 64.3 20.2 13.1 9.0 3.4 3.3

Table 3 Effect of purification steps on the total protein, activity, specific activity and purification value, and yield of β-glucosidase from the honey sac (1000 bees) of A. mellifera Purification method

Raw extract DEAE Sepharose FF Superose prep. grade CM Sepharose FF Mono S Hydroxyapatite Superose 12

Protein (mg)

126.0 83.5 16.4 1.54 0.21 0.10 0.01

β-Glucosidase Activity (units)

Specific activity (unit/mg)

Purification value

Yield (%)

1.42 1.04 0.30 0.18 0.07 0.05 0.03

0.01 0.01 0.02 0.12 0.34 0.53 3.49

1.00 1.11 1.63 10.4 30.4 47.2 309

100 71.8 21.3 12.7 5.1 3.7 2.2

nificant (苲40%) amount of non-glucosidase protein with a small loss in β-glucosidase activity (Table 2). Purification with preparative gel-permeation chromatography (Superose 12) resulted in β-glucosidase coeluting with α-glucosidase due to their similarity in relative molecular masses. This purification step resulted in an increase in specific activity of β-glucosidase from 0.11 to 0.14 units/mg (Table 2). Cation-exchange chromatography (CM-Sepharose FF) was able to resolve β- from α-glucosidase (Fig. 1B), and this step coupled with hydroxyapatite (Fig. 1C) and high-performance gel-permeation (Prepacked Superose 12; Fig. 1D) chromatography resulted in a highly purified β-glucosidase. Approximately 46 µg (3.3% yield) of pure β-glucosidase with a specific activity of 5.66 units/mg was obtained from the ventriculus of 250 honey bees. The purification of β-glucosidase from the honey sac followed the same protocol developed for that of the ventriculus. The total protein, activity, specific activity and purification value, and yield of β-glucosidase after each purification step are shown in Table 3. Approximately 9.0 µg of β-glucosidase with a specific activity of 3.49 units/mg was obtained from the honey sacs of 1000 honey bees.

Although β-glucosidase from the hypopharyngeal glands was not purified to the same extent using this protocol, a partially purified β-glucosidase was obtained for SDS–PAGE and MALDI-TOF-MS analysis. 3.3. Electrophoresis The relative molecular mass of β-glucosidase from the ventriculus and honey sac of A. mellifera was estimated from SDS–PAGE comparison to standard proteins (Fig. 2A) to be approximately 72 kDa. The purified β-glucosidase from the ventriculus and the honey sac consisted of only one subunit. Purified β-glucosidase from the ventriculus and the honey sac gave a positive result to glycoprotein staining indicating that these isolated enzymes are glycoproteins. Isoelectric focusing polyacrylamide gel electrophoresis (IEF–PAGE) showed that the purified β-glucosidase from the ventriculus and honey sac consists of several bands (Fig. 2B). The isoelectric point (pI) was estimated from IEF standard proteins as 4.5–4.8. The fact that multiform proteins were observed by IEF– PAGE indicates that β-glucosidase from these organs may be present in different degrees of glycosylation.

686

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

Fig. 2. SDS–PAGE (A) and IEF–PAGE (B) of purified β-glucosidase from the ventriculus of A. mellifera stained with Coomassie blue and silver, respectively.

3.4. Effects of pH and temperature The effects of pH and temperature on β-glucosidase stability are shown in Fig. 3. The temperature stability of β-glucosidase from the ventriculus and honey sac ranged from 20 to 50 °C. The pH stability range of β-glucosidase from the ventriculus was 3.5 to 9.5, and from the honey sac was 4.5 to 9.0. The pH optimum for β-glucosidase from the ventric-

ulus and honey sac was 5.0 for each organ (Fig. 4). The β-glucosidase activity slowly increased as a function of temperature from 20 to 60 °C, and dropped significantly after 60 °C due to heat inactivation. β-Glucosidase activity increased as a function of pH from 3.0 to 5.5 and significantly decreased after pH 6.0, especially in combination with higher temperatures (⬎50 °C).

Fig. 3. Stability of purified β-glucosidase from the ventriculus and honey sac of A. mellifera after incubation (20 min, pH 6.8) at selected temperatures (A) and pHs (20 min at 25 °C) (B) using β-PNPG as substrate.

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

687

Fig. 4. Response surface plots of β-glucosidase activity (β-PNPG as substrate) for purified enzymes from the ventriculus and honey sac of A. mellifera at a pH range of 3 to 10 and a temperature range of 20 to 75 °C.

3.5. Enzyme kinetics The Vmax and Km for β-glucosidase from the ventriculus of A. mellifera using β-PNPG as substrate were 1.1 µmol nitrophenol/min/mg glucosidase and 12.0 mM, respectively (Table 4). The Vmax and Km for β-glucosidase from the honey sac were 0.76 µmol nitrophenol/min/mg glucosidase and 11.3 mM, respectively. The initial reaction rate versus substrate concentration plot for β-glucosidase isolated from each organ followed Michaelis–Menten kinetics. β-Glucosidase from the ventriculus and honey sac of A. mellifera was shown to have no activity towards cellobiose (0.1 M in 0.1 M acetate buffer at pH 5.0). 3.6. MALDI-TOF peptide mass mapping Two separately purified samples of β-glucosidase from the ventriculus were denatured and analyzed by SDS–PAGE and subjected to MALDI-TOF-MS analysis. The MALDI-TOF mass spectrum of these samples showed 12 matching peptide masses (Fig. 5) with 22% coverage (based on the Mr of 72 kDa). Peptides fragments assigned as originating from β-glucosidase based on two replicate experiments are shown in Table 5.

The MALDI-TOF mass spectrum of purified and denatured β-glucosidase from the honey sac after SDS– PAGE showed six peptide masses (Table 5) which matched those of β-glucosidase from the ventriculus with 10.1% coverage. The MALDI-TOF mass spectrum of partially purified and denatured β-glucosidase from the hypopharyngeal glands after SDS–PAGE showed seven peptide masses which matched those of β-glucosidase from the ventriculus with 11.7% coverage (Table 5).

4. Discussion β-Glucosidase has been purified from the ventriculus and honey sac of the honey bee, A. mellifera. This represents the first time that this enzyme has been isolated and purified from any honey bee or honey bee organ. This finding shows that β-glucosidase is a distinct enzyme and explains the presence of β-linked oligosaccharides (e.g. gentiobiose) in honey (Low et al., 1988). β-Glucosidase activity was found in the hypopharyngeal glands, honey sac, ventriculus and hindgut of the honey bee, A. mellifera. These results suggest that β-

Table 4 Selected enzyme kinetic parameters for purified β-glucosidase from the ventriculus and honey sac of A. mellifera employing various substrates Organ

Ventriculus

Honey sac

a b

Substrate

Vmaxa (µmol product/min/mg glucosidase)

Kma (mM)

Vmax/Km

β-PNPG Cellobiose

1.1±0.2 NAb

12.0±1.6

0.092

β-PNPG Cellobiose

0.76±0.02 NA

11.3±4.2

0.069

Based on triplicate studies. No activity (no glucose detected by HPAE-PAD).

688

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

Fig. 5. MALDI-TOF-MS of purified undeglycosylated β-glucosidase from the ventriculus A. mellifera. Peptides originating from β-glucosidase (*) and trypsin (★). Table 5 Peptide masses (Da) from MALDI-TOF-MS analysis of trypsin-treated undeglycosylated β-glucosidase from the ventriculus, honey sac and hypopharyngeal glands of A. mellifera Ventriculus

Mass (Da)

Honey sac

1st sample

2nd sample

958.5 970.5 978.5 995.6 1017.6 1060.1 1193.6 1223.6 1251.6 1320.7 1700.9 2818.4

958.6 970.6 978.6 995.6 1017.6 1060.1 1193.7 1223.7 1251.7 1320.8 1700.9 2818.5

glucosidase is produced in the hypopharyngeal glands, secreted into the mouth during feeding and passes to the honey sac. From the honey sac, β-glucosidase is transferred into honeycomb cells and the ventriculus. The observed high β-glucosidase activity found in the ventriculus and some activity in the hindgut indicate that this enzyme has resistance to protease activity. The wide temperature stability (20–60 °C) for β-glucosidase from the ventriculus and honey sac agree with that reported by Low et al. (1986) for β-glucosidase in honey. The temperature stability for β-glucosidase was found to be higher than that observed for α-glucosidase-

Hypopharyngeal glands

958.5 970.5

958.5 970.5

1193.7 1223.7 1251.7

1193.6 1223.6 1251.6

1699.9

1700.8 2818.4

1 and -2 isolated from these organs (Pontoh and Low, in preparation). This temperature stability is most likely not related to the environmental temperatures (36 °C) to which honey bees are normally exposed (Southwick, 1992). The pH stability of β-glucosidase from the ventriculus and honey sac of A. mellifera was slightly different; however, the pH optimum was identical (pH 5.0). This pH optimum was higher than that reported by Low et al. (1986) of 4.2 for unpurified β-glucosidase from honey. The reason for this difference could be due to the purity of the β-glucosidase sample and/or the optim-

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

ization methodology employed. β-Glucosidase has low activity at high pH (Fig. 4), but is stable over a wide pH range (Fig. 3). This behaviour indicates that the hydrolytic activity of β-glucosidase is favoured in an acidic environment. The Km values for β-glucosidase from the ventriculus and honey sac for β-PNPG were almost identical (11 and 12 mM) and were similar to that reported by Low et al. (1986) for unpurified β-glucosidase (13 mM) from honey. β-Glucosidase from the ventriculus and honey sac was shown to have no activity towards cellobiose. Therefore, this enzyme belongs to the alkyl- and/or arylβ-glucoside hydrolyzing group (Group 3), according to the classification of Terra and Ferreira (1994). This result indicates that the role of β-glucosidase in the alimentary canal of honey bees is not for the digestion of cellulosic material, such as pollen, but is most likely present to hydrolyze glycoside toxins ingested by the honey bee (Terra and Ferreira, 1994). The peptide masses from MALDI-TOF-MS of denatured β-glucosidase from the ventriculus, honey sac and hypopharyngeal glands showed at least six matching masses with 10.1% coverage. Although the percent coverage was low, this coverage and the number of identified peptide masses fall well within the range that has been used to unambiguously identify proteins by MALDI-TOF-MS, which is ⱖ3 matching peptide masses (Henzel et al., 1993). The similarities of β-glucosidase properties from both the ventriculus and honey sac in Mr, temperature stability, temperature and pH optimum, Km and MALDITOF-MS peptide maps indicate that these enzymes are the same. The similarities of MALDI-TOF-MS peptide maps of β-glucosidase from the hypopharyngeal glands and those from the honey sac and ventriculus indicate that this enzyme is produced in the hypopharyngeal glands, secreted into the mouth, and then passes through the digestive system in the same manner as α-glucosidase (Simpson, 1960). The hypopharyngeal glands as the source of β-glucosidase is quite different from other insects, with this enzyme being produced in the midgut or ventriculus (Terra and Ferreira, 1994).

Acknowledgements This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC). We thank Dr. M. Pato for her assistance in collecting and interpreting the kinetic data contained in this article.

References Adedire, C.O., Balogun, R.A., 1995. Digestive enzymes and regional localization of proteolytic endopeptidases in the alimentary canal

689

of the kola weevil, Sophrorhinus insperatus Faust (Coleoptera: Curculionidae). Entomon 20, 183–189. Applebaum, S.W., 1985. Biochemistry of digestion. In: Kerkut, G.A., Gilbert, L.A. (Eds.) Comprehensive Insect Physiology, Biochemistry and Pharmacology. Pergamon Press, New York, pp. 279–311. Baker, J.E., Woo, S.M., 1992. Beta-glucosidases in the rice weevil, Sitophiulus oryzae: Purification, properties and activity levels in wheat- and legume-feeding strains. Insect Biochemistry and Molecular Biology 22, 495–504. Bergmeyer, H.U., 1983. Establishing optimum measurement conditions. In: Bergmeyer, H.U., Bergmeyer, J., Grabl, M. (Eds.) Methods of Enzymatic Analysis, 3rd edn. Fundamentals, Vol. I. Verlag Chemie, Weinheim, Germany, pp. 114–142. Bollag, D.M., Edelstein, S.J., 1991. Protein Methods. Wiley-Liss, New York. Bradford, M., 1976. Rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Analytical Biochemistry 72, 248–254. Chararas, C., Chipoulet, J.M., 1982. Purification by chromatography and properties of a β-glucosidase from the larvae of Phoracantha semipunctata. Comparatives in Biochemistry and Physiology 72B, 559–564. Chipoulet, J., Chararas, C., 1985. Purification and partial characterization of a cellobiase from the larvae of Rhagium inquisitor. Comparative Biochemistry and Physiology 82B, 327–332. Colepicolo, N.P., Bechara, E.J.H., Ferreira, C., Terra, W.R., 1986. Evolutionary considerations of the spatial organization of digestion in the luminescent predaceous larvae of Pyrearinus termitilluminans (Coleoptera: Elateridae). Insect Biochemistry 16, 811–817. Esen, A., 1993. β-glucosidases: overview. In: Esen, A. (Ed.) β-Glucosidases. Biochemistry and Molecular Biology. ACS Symposium Series 533. American Chemical Society, Washington, DC, pp. 1– 14. Ferreira, C., Terra, W.R., 1983. Physical and kinetics properties of a plasma membrane bound β-d-glucosidase (cellobiose) from midgut cells of an insect (Rhynchosciara americana larvae). Biochemistry Journal 213, 43–51. Ferreira, C., Terra, W.R., 1989. Spatial organization of digestion, secretory mechanism and digestive enzyme properties in Pheropsophus aequinoctialis (Coleoptera: Carabidae). Insect Biochemistry 19, 383–391. Ferreira, C., Parra, J.R.P., Terra, W.R., 1997. The effect of dietary plant glycosidases on larval midgut β-glucosidases from Spodoptera frugiperda and Diatraea saccharalis. Insect Biochemistry and Molecular Biology 27, 55–59. Ferreira, C., Torres, B.B., Terra, W.R., 1998. Substrate specificities of midgut β-glucosidases from insects of different orders. Comparative Biochemistry and Physiology 119B, 219–225. Gharahdaghi, F., Weinberg, C.R., Meagher, D.A., Imai, B.S., Mische, S.M., 1999. Mass spectrometric identification of proteins from silver-stained polyacrylamide gel: a method for the removal of silver ions to enhance sensitivity. Electrophoresis 20, 601–605. Gilliam, M., Lorenz, B.J., Richardson, G.V., 1988. Digestive enzymes and microorganisms in honey bees, Apis mellifera: influence of streptomycin, age, season and pollen. Microbios 55, 95–114. Henzel, W.J., Billeci, T.M., Stults, J.T., Wong, S.C., Grimley, C., Watanabe, C., 1993. Identifying proteins from two-dimensional gels by molecular mass searching of peptide fragments in protein sequence databases. Proceedings of the National Academy of Sciences of the United States of America 90, 5011–5015. Low, N.H., Vong, K.V., Sporns, P., 1986. A new enzyme, β-glucosidase, in honey. Journal of Apicultural Research 25, 178–181. Low, N.H., Nelson, D.L., Sporns, P., 1988. Carbohydrate analysis of Western Canadian honeys and their nectar sources to determine the origin of honey oligosaccharides. Journal of Apicultural Research 27, 245–251. Marana, S.R., Terra, W.R., Ferreira, C., 1995. Midgut β-d-glucosidases

690

J. Pontoh, N.H. Low / Insect Biochemistry and Molecular Biology 32 (2002) 679–690

from Abracris flavolineata (Orthoptera: Acrididae). Physical properties, substrate specificities and function. Insect Biochemistry and Molecular Biology 25, 835–843. Marana, S.R., Terra, W.R., Ferreira, C., 2000. Purification and properties of a beta-glucosidase purified from midgut cells of Spodoptera frugiperda (lepidoptera) larvae. Insect Biochemistry and Molecular Biology 30, 1139–1146. Morgan, M.R.J., 1975. Relationship between gut cellobiase, lactase, aryl β-glucosidase, and aryl β-galactosidase activities of Locusta migratoria. Insect Biochemistry 5, 609–617. Pontoh, J., Low, N.H. Purification and characterization of α-glucosidases from two honey bee species (Apis mellifera and A. cerana) (in preparation). Pratviel, F., Clermont, S., Percheron, F., Chararas, C., 1987. Studies on glycosidases and glucanases in Thaumetopoea pityocampa larvae. II. Purification and some properties of a broad specificity βd-glucosidase. Comparative Biochemistry and Physiology 86B, 173–178. Rautela, G.S., Snee, R.D., Miller, W.K., 1979. Response surface cooptimization of reaction conditions in clinical chemical methods. Clinical Chemistry 25, 1954–1964. Rosenberg, I.M., 1996. Protein Analysis and Purification. Benchtop Techniques. Birkhauser, Boston. Santos, C.D., Terra, W.R., 1985. Physical properties, substrate specificities and a probable mechanism for a β-d-glucosidase (cellobiase) from midgut cells of the cassava hornworm (Erinnyis ello). Biochimica et Biophysica Acta 831, 179–185. Schumaker, T.T.S., Christofoletti, P.T., Terra, W.R., 1993. Properties and compartmentalization of digestive carbohydrases and proteases in Scaptotrigona bipunctata (Apidae: Meliponinae) larvae. Apidologie 24, 2–17. Shevchenko, A., Wilm, M., Vorm, O., Mann, M., 1996. Mass spectrometric sequencing of proteins from silver-stained polyacrylamide gels. Analytical Chemistry 68, 850–858.

Silva, C.P., Ribeiro, A.F., Terra, W.R., 1996. Enzyme markers and isolation of the microvillar and perimicrovillar membranes of Dysdercus peruvianus (Hemiptera: Pyrrhocoridae) midgut cells. Insect Biochemistry and Molecular Biology 26, 1011–1018. Simpson, J., 1960. The functions of the salivary glands of Apis mellifera. Journal of Insect Physiology 4, 107–121. Simpson, J., Riedel, I.B.M., Wilding, N., 1968. Invertase in the hypopharyngeal glands of the honeybee. Journal of Apicultural Research 7, 29–36. Snodgrass, R.E., Erickson, E.H., 1992. The anatomy of the honey bee. In: Graham, J.M. (Ed.) The Hive and the Honey Bee. Dadant and Sons, Hamilton, IL, pp. 103–169. Southwick, E.E., 1992. Physiology and social physiology of the honey bee. In: Graham, J.M. (Ed.) The Hive and the Honey Bee. Dadant and Sons, Hamilton, IL, pp. 171–196. Stauffer, C.E., 1989. Enzyme Assays for Food Scientists. Van Nostrand Reinhold, New York, NY. Swallow, K.W., Low, N.H., 1990. Analysis and quantitation of the carbohydrates in honey using high performance liquid chromatography. Journal of Agricultural and Food Chemistry 38, 1828– 1832. Terra, W.R., Ferreira, C., Bastos, F., 1985. Phylogenic considerations of insect digestion. Disaccharidases and the spatial organization of digestion in the Tenebrio molitor larvae. Insect Biochemistry 13, 143–150. Terra, W.R., Ferreira, C., Garcia, E.S., 1988. Origin, distribution, properties and functions of the major Rhodnius prolixus midgut hydrolases. Insect Biochemistry 18, 423–434. Terra, W.R., Ferreira, C., 1994. Insect digestive enzymes: properties, compartmentalization and function. Comprehensive Biochemistry and Physiology 109B, 1–62. Webb, E.C., 1992. Enzyme nomenclature. Recommendation (1992) of the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology. Academic Press, San Diego, CA.