An inhibitor from Lupinus bogotensis seeds effective against aspartic proteases from Hypothenemus hampei

Phytochemistry 71 (2010) 923–929

Contents lists available at ScienceDirect

Phytochemistry

journal homepage: www.elsevier .com/locate /phytochem

An inhibitor from Lupinus bogotensis seeds effective against aspartic proteasesfrom Hypothenemus hampei

Diana Molina a, Humberto Zamora b, Alejandro Blanco-Labra c,*

a National Coffee Research Center (Cenicafé), Plant Breeding Department, Plan Alto Km 4 – vía antigua a Manizales, Chinchiná, Caldas, Colombiab Department of Chemistry, National University of Colombia, Carrera 30 #45-03, Ciudad Universitaria, Bogotá, Cundinamarca, Colombiac Department of Biotechnology and Biochemistry, Centro de Investigación y de Estudios Avanzados, Unidad Irapuato, Km 9.6 Libramiento Norte, Carretera Irapuato-León,C.P. 36821 Irapuato, Guanajuato, Mexico

a r t i c l e i n f o a b s t r a c t

Article history:Received 25 August 2009Received in revised form 23 November 2009Available online 26 March 2010

Keywords:Lupinus bogotensisLeguminosaePlant defenseAspartic protease inhibitorVicilinsHypothenemus hampeiAspartic protease

0031-9422/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.phytochem.2010.03.006

* Corresponding author. Tel.: +52 462 6239600; faxE-mail addresses: [email protected]

unal.edu.co (H. Zamora), [email protected] (A

The coffee berry borer, Hypothenemus hampei (Ferrari), is one of the most devastating coffee pests (Coffeaarabica L.) worldwide. Digestion in the midgut of H. hampei is facilitated by aspartic proteases. This is thefirst report of an aspartic protease inhibitor from Lupinus bogotensis. The L. bogotensis aspartic proteaseinhibitor (LbAPI) exhibited a molecular mass of 12.84 kDa, as determined by MALDI-TOF, and consistsof a single polypeptide chain with an isoelectric point of 4.5. In thermal activity experiments, stabilitywas retained at pH 2.5 after heating the protein at 70 �C for 30 min, but was unstable at 100 �C. The pro-tein was also stable over a broad range of pH, from 2 to 11, at 30 �C. In in vitro assays, LbAPI was highlyeffective against aspartic proteases from H. hampei guts with a half maximal inhibitory concentration(IC50) of 2.9 lg. LbAPI inhibits pepsin in a stoichiometric ratio of 1:1. LbAPI inhibition of pepsin was com-petitive, with a Ki of 3.1 lM, using hemoglobin as substrate. Its amino-terminal sequence had 76% homol-ogy with the seed storage proteins vicilin and b-conglutin. The homology of LbAPI to vicilins from Lupinusalbus L. suggests that they may also serve as storage proteins in the seed. LbAPI could be a promising toolto make genetically modified coffee with resistance to H. hampei.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

In 70 countries in the humid tropics, coffee (Coffea spp., Rubia-ceae) is the most important agricultural commodity (Jaramilloet al., 2006). In Colombia, coffee (Coffea arabica L.) accounts for12% of the agricultural gross domestic product (MADR, 2006).The coffee berry borer, Hypothenemus hampei (Ferrari) (Coleoptera:Curculionidae: Scolytinae) (Le Pelley, 1968), is the insect pest thatcauses the highest economic losses in the world’s coffee crop (Jara-millo et al., 2006).

Many coffee farmers currently rely on the application of syn-thetic insecticides, such as endosulfan and chlorpyrifos, both ofwhich are toxic to insects (Jaramillo et al., 2006). However, thegrowing concern of environmentalists and the increased resistanceof this pest to insecticides (Brun et al., 1989) have prompted re-search to be conducted on environmentally friendly control strate-gies (Jaramillo et al., 2006). Among these, several plant proteinshave the potential to control the insect, including lectins, proteaseinhibitors (PIs), and storage proteins such as vicilins and arcelins(Chrispeels and Raikhel, 1991; Ryan, 1990; Sales et al., 2000). Also,

ll rights reserved.

: +52 462 6245996.m (D. Molina), Hmzamorae@. Blanco-Labra).

a-amylase inhibitors, in particular an a-amylase inhibitor ex-tracted from Phaseolus vulgaris seeds, was active on the a-amylasefrom H. hampei (Valencia et al., 2000).

The ability of PIs to interfere with insect growth and develop-ment has been attributed to their ability to bind the active sitesof digestive proteases and block, alter, or impede access to the sub-strate, thus reducing the availability of essential amino acids de-rived from ingested protein (Ryan, 1990). Previous studies haveinvestigated the effect of PIs on the Coleoptera, such as those re-ported by Aguirre et al. (2004), Macedo et al. (2004), and Araújoet al. (2005), among others.

Aspartic proteases have been found in Coleoptera species, suchas Callosobruchus maculatus (Silva and Xavier-Filho, 1991), Zabrotessubfasciatus (Silva and Xavier-Filho, 1991), and H. hampei (Preciadoet al., 2000), in which the acidic midgut pH provides a favorableenvironment for these proteases (optimal pH �3–5) (Ryan,1990). Considering that the aspartic proteases of H. hampei digestfood protein in the insect’s midgut (Preciado et al., 2000), the inhi-bition of these proteases provides a promising defense mechanismfor the control of this insect.

Aspartic inhibitors are characterized by the presence of twoaspartic acid residues in their active site and their reactivity at alow pH (Ryan, 1990). Reports on these inhibitors are scarce; someof these studies have been reported in wheat (Galleschi et al.,

Table 1Purification yield of LbAPI from Lupinus bogotensis seeds.

Step Total activity (IU) Protein (mg/ml) Specific activity (IU/mg) Purification (fold) Yield (%)

Crude extract 20,450 250.70 81.6 1.0 100Econo-Pac desalting 16,520 167.60 98.9 1.2 80.8(NH4)2SO4 fractionation 14,900 89.00 167.4 2.0 72.9Ultrafiltration 12,420 30.00 414.0 5.1 60.7Q Sepharose 5775 0.37 15608.0 191.3 28.2Elution from gel SDS–PAGE 790 0.05 15800.0 193.7 3.9

Fig. 1. (A) Inhibitor zymogram of LbAPI using 16% SDS–PAGE. Lanes 1 and 2represent an inhibitor zymogram that shows the bands corresponding to PIs of L.bogotensis resistant to digestion by the aspartic proteases of H. hampei. Lane 1:standard proteins; lane 2: 20 lg LbAPI. (B) Inhibition zymogram of aspartic-likeproteases present in the intestinal tract of adult H. hampei using a 10% native gel.Lane 1: 20 lg extract from adult H. hampei (positive control); lane 2: 20 lg extractfrom adult H. hampei + 5 lg LbAPI.

924 D. Molina et al. / Phytochemistry 71 (2010) 923–929

1997), tomato (Cater et al., 2002), a novel potato cathepsin D inhib-itor (NID) (Ritonja et al., 1990), and squash aspartic proteinaseinhibitors (SQAPI) (Christeller et al., 1998). This scarcity of reportscontrasts with the large number of serine and cysteine PIs discov-ered to date (Richardson, 1991; Bhattacharyya et al., 2006; Macedoet al., 2007; Ramos et al., 2009).

Lupinus bogotensis, one of 200 species of Lupinus native to theAmericas subgenus Platycarpos (Watson, Kurl.), belongs to the Fab-aceae family (Turner, 1995). It is found in the Andean region and isalso present in Colombia. Lupin proteins belong to the families ofthe a, b, d, and c globulins; the b-conglutins or vicilins are storageproteins that belong to the 7S globulin family present in legumeseeds (Derbyshire et al., 1976). Vicilins are the main globulin com-ponent, formed by 10–12 major types of subunits with molecularmasses ranging from 15 to 72 kDa. This globulin contains no disul-fide bonds (Melo et al., 1994).

Despite considerable interest in the use of PIs to create insectresistance in plants through genetic engineering (Vila et al.,2005), no studies have explored the use of aspartic protease inhib-itors for the control of H. hampei. The present study describes puri-fication and characterization of an inhibitor, which represents thefirst PI isolated from L. bogotensis Benth, with inhibitory activityagainst the aspartic proteases of H. hampei. The results suggest anew mechanism of action for vicilins against Coleoptera.

2. Results and discussion

2.1. Purification and molecular characterization

A new protease inhibitor was purified using a combination ofchromatographic methods, which were monitored by measuringthe inhibitory activity of the fractions against the aspartic prote-ases of H. hampei. Most aspartic inhibitory activity from crude seedextracts was obtained by precipitation at 80% (w/v) saturation(NH4)2SO4. The precipitate was further separated by ultrafiltration.The proteins eluted from a 30 kDa membrane inhibited the aspar-tic proteases of the H. hampei. After further purification by anionexchange chromatography on a Q Sepharose column, five out ofeight major protein peaks were eluted using a linear NaCl gradientdemonstrated aspartic inhibitory activity (data not shown). Thefractions were pooled, desalinated, and lyophilized. Final purifica-tion was obtained after separating the proteins by electrophoresisand eluting the pure proteins directly from the gel. The inhibitorthat presented the highest inhibition of the insect’s aspartic prote-ase was selected for further study characterization and named L.bogotensis aspartic protease inhibitor (LbAPI).

The specific activity after protein purification was 15,800 aspar-tic protease inhibition units (AIU/mg), with a purification factor193.7 times that of the crude extract and with a final recovery of3.9% of the original activity. Up to 96% of the total inhibitory unitsand �99% of protein content were lost during purification (Table1). In the present study, the crude extract of L. bogotensis contains20,450 AIU/g seed, a value that corresponds to 34.5 AIU/mg of pro-tein in the dry seed, considering a seed protein content of 42% ofdry matter (Egaña et al., 1992). This level is similar to that found

for the Bowman-Birk inhibitor content found in L. albus seeds (LaB-BI) (Scarafoni et al., 2008). The LbAPI yield was 50 lg/g seed, rep-resenting 0.02% seed protein content of dry matter. These resultsconfirm that Lupinus spp. seeds contain PIs, contradicting the re-sults of Gallardo et al. (1974) who found them absent in the seedsof species of this genus.

Inhibition zymograms in polyacrylamide gels containing 1%hemoglobin were used to detect the aspartic protease inhibitoryactivity of the band corresponding to LbAPI. The presence of theinhibitor, which is resistant to digestion by the aspartic proteasesof H. hampei, was visualized as a clear zone against a dark bluebackground (Fig. 1A, lane 2). In a second gel (Fig. 1B), the asparticproteolytic activity (control) was visualized in lane 1, whereascomplete inhibition of the proteolytic activity of the aspartic prote-ases of H. hampei by LbAPI was displayed in lane 2.

LbAPI consists of a single polypeptide chain with an apparentmolecular mass of 12 kDa based on SDS–PAGE analysis under bothnon-reducing (Fig. 2A) and reducing (data not shown) conditions.This molecular mass was later confirmed by mass spectrometryunder non-reducing conditions (12.84 kDa), where LbAPI appearedas a single peak (Fig. 2B). Similar results were obtained for the

Fig. 2. (A) Electrophoretic analysis (16% SDS–PAGE, under non-reducing conditions) of LbAPI fractions obtained during purification. Lane 1: standard proteins; lane 2: crudeextract; lane 3: pigment-free crude extract; lane 4: salted-out fraction with ammonium sulfate; lane 5: LbAPI-Q Sepharose fraction; lane 6: gel-eluted LbAPI fraction. (B)MALDI-TOF mass spectrum of LbAPI. See text for details. (C) LbAPI isoelectric focusing. Lane 1: standard proteins; lane 2: LbAPI.

D. Molina et al. / Phytochemistry 71 (2010) 923–929 925

10.55 and 10.52 kDa SQAPI inhibitors reported for Cucurbita max-ima seeds (Christeller et al., 1998).

The purified inhibitor appeared as a single band in SDS–PAGE,with an acidic isoelectric point of 4.5 (Fig. 2C), indicating the pos-sible absence of isoforms. The acidic isoelectric point of 4.5 couldbe explained based on its relatively high content of acidic aminoacids (glutamic acid 35.5% and aspartic acid 8.7%). Acidity is a com-mon characteristic of the Kunitz trypsin inhibitor from Archiden-dron ellipticum (AeTI) (Bhattacharyya et al., 2006), the Hyptissuaveolens trypsin inhibitor (HSTI) (Aguirre et al., 2004), and theisoinhibitors of Hevea brasiliensis (Sritanyarat et al., 2006).

2.2. Amino-terminal sequencing

Amino-terminal sequence analyses of LbAPI for the first 22 ami-no acids established a 76% sequence similarity with L. albus vicilinand L. albus b-conglutin and 71% similarity with Lupinus angustifo-lius conglutin beta (Table 2). The high concentration of acidic resi-dues (8 of 22) was similar to the sequence found for thechymotrypsin inhibitor of Schizolobium parahyba (SPC) (Souzaet al., 1995).

The amino-terminal sequence of LbAPI did not resemble any ofthe known sequences of aspartic protease inhibitors. The amino-terminal sequence of SQAPI also was not homologous to other pro-teins, suggesting the inhibitors to constitute a novel family (Chris-teller et al., 1998).

To date, few studies have investigated the inhibitors of aspar-tic proteases. Therefore, only a limited amount of information onthe primary structures of these inhibitors is available in the dat-abases, making it difficult to detect similarity with other inhibi-tors of this type (Rawlings et al., 2004). Serine proteinaseinhibitors, in contrast, have been extensively studied and areclassified into 16 families based on sequence similarity, topolog-ical similarity, and binding mechanism (Laskowski and Kato,1980).

Other PIs also have homology with seed storage proteins; thetrypsin inhibitors of buckwheat (BWI-2b) display significanthomology with proteins of vicilin family (Park et al., 1997). In addi-tion, the Kunitz-type trypsin inhibitor of Inga laurina (ILTI) displayssignificant homology with sporamin and miraculin (Macedo et al.,2007). Similarly, the inhibitor of cathepsin D isolated from potatopossesses many residues that are identical to those of miraculin(Ritonja et al., 1990).

2.3. Effect of LbAPI on the aspartic proteases of H. hampei

The extract containing aspartic proteases of H. hampei wasinhibited by LbAPI with a half maximal inhibitory concentration(IC50) of 2.9 lg (Fig. 3). This inhibitor did not affect trypsin and chy-motrypsin activities.

Considering that the amino-terminal sequence of LbAPI has anhigh homology with vicilin of L. albus, it would be possible that

y = -81.141x + 88.252R2 = 0.9635

0

20

40

60

80

100

120

0 0.2 0.4 0.6 0.8 1 1.2Log LbAPI (µg)

Res

idua

l act

ivity

(%)

Fig. 3. LbAPI inhibition of the midgut aspartic-like proteolytic activity (hemoglobinassay) from adult H. hampei (N).

0102030405060708090

100110

0 0.5 1 1.5 2 2.5Molar ratio of inhibitor/enzyme

Res

idua

l act

ivity

(%)

Fig. 4. Titration curve of aspartic inhibition by LbAPI. Increasing concentrations ofinhibitor were added to a fixed concentration of pepsin (3.1 nM). Residual enzymeactivity was determined using hemoglobin as the substrate. Each point representsthe mean value of three independent assays.

0

5

10

15

20

25

-4 -3 -2 -1 0 1 2 3 4 5 6 7IAPLb [µM]

1/v

[OD

280/1

0min

/ml]-

1

Hemoglobin ( 0.075 mM) Hemoglobin ( 0.15 mM)

Fig. 5. Kinetic analysis of pepsin inhibition by LbAPI: Dixon plot for the determi-nation of the inhibition constant (Ki) of LbAPI against pepsin. Enzyme assays werecarried out with pepsin in the presence of increasing concentrations of LbAPI at twodifferent concentrations of hemoglobin. The reciprocals of velocity were plottedagainst the LbAPI concentration, and the Ki value was obtained from the interceptsof two lines at two concentrations of substrate.

Table 2Comparative amino-terminal sequence alignment of the first 22 amino acids of Lupinus bogotensis LbAPI with vicilin of L. albus, b-conglutin precursor of L. albus, and conglutinbeta of L. angustifolius. White letters on a black background indicate identical amino acid residues.

Sequence Position %Similarity

Accessionnumber

LbAPI 1:22 H E S P E E R E Q E E Y D P R Q Q S H H V EVicilin L. albus 38:54 76% CAI84850.2 H E R P E E R E Q E E W Q P R R Q

b-conglutinprecursor L. albus

38:54 76% AAS97865.1 H E R P E E R E Q E E W Q P R R Q

Conglutin beta L.angustifolius

41:54 71% ABR21772.1 P K E R E E E E H E P R Q Q

The last five amino acids showed no similarity with vicilin and conglutin beta of the genus Lupinus.

926 D. Molina et al. / Phytochemistry 71 (2010) 923–929

the defense role reported for vicilins could be associated with theircapacity to inhibit aspartic proteases in the midgut of Coleopterasuch as H. hampei.

In another study it was reported that the resistance of seeds ofthe cowpea variety Tvu 2027 to C. maculatus was due to the pres-ence of high levels of Bowman-Birk trypsin inhibitors (Gatehouseand Boulter, 1983); however, it was later demonstrated that suchresistance was not related either to serine or cysteine proteinaseinhibitors (Xavier-Filho et al., 1989), or to a-amylase inhibitors(Reis et al., 1997). Subsequent studies demonstrated that such aresistance to C. maculatus was probably due to the presence of vic-ilins (Sales et al., 2000) in its globulin fractions. Later on it was re-ported that the vicilins bind to the peritrophic membrane ofDiatraea saccharalis larvae, significantly reducing their survival rate(Mota et al., 2003). Uchoa et al. (2006) also demonstrated the bind-ing of vicilins from cowpea seeds to the internal organs and tissuesof C. maculatus, such as body fat and Malpighian tubules. However,the exact antibiosis mechanism caused by those proteins remainsunknown (Sales et al., 2000), since the vicilins of cowpea lines thatare both resistant and susceptible to C. maculatus were found tobind to the chitin in the midgut of larvae without any obvious signsof cell damage (Sales et al., 2000).

The results obtained indicate that PIs are multifunctional pro-teins that not only inhibit aspartic proteases, but are also capableof binding to chitin-containing structures in the midgut. It hasbeen postulated that plant protease inhibitors could be involvedin a complex interaction between the plant’s nutritional valueand the insect’s digestive physiology, i.e., by considering it highlyimprobable that inhibition of proteases would be the sole functionof these proteins in the midgut of insects (Broadway et al., 1986;Gatehouse et al., 1992).

2.4. Inhibitory properties and Ki determination

Inhibition of pepsin at pH 2.5 by increasing amounts of LbAPI isshown in Fig. 4. The titration curve for pepsin was linear up to 90%inhibition, fitting a slow-tight binding mechanism with 1:1 stoichi-ometry (Fig. 4). The titration curve shows that inhibition did notreach 100%. Dixon plot analysis indicated a competitive inhibition,with an inhibition constant of 3.1 lM for pepsin, when hemoglobinwas used as the substrate (Fig. 5). Several other inhibitors have alsobeen shown to act as competitive inhibitors (Bhattacharyya et al.,2006; Chaudhary et al., 2008).

2.5. Stability studies

LbAPI was found to be stable when treated for 30 min at tem-peratures ranging from 30 to 70 �C (data not shown); however,

D. Molina et al. / Phytochemistry 71 (2010) 923–929 927

above 80 �C its activity fell sharply. The high thermostability ofLbAPI has also been observed for others PIs, such as the inhibitorin AeTI (Bhattacharyya et al., 2006), the inhibitor of LaBBI (Scaraf-oni et al., 2008) and the isoinhibitors of H. brasiliensis (Sritanyaratet al., 2006).

Analysis of the pH stability of LbAPI indicated that it was verystable within a range of conditions from highly acidic to highlyalkaline, with a maximum inhibition observed at pH 2.5 (datanot shown). Similar results were reported for the ILTI (Macedoet al., 2007) and the trypsin inhibitor from Putranjiva roxburghii(PRTI) (Chaudhary et al., 2008).

3. Conclusions

The search of novel PIs as a possible alternative pest controlstrategy was studied. This research led to the discovery of an inhib-itor present in the seeds of L. bogotensis, a plant species of Andeanregion, which is highly effective against H. hampei, a Coleopteranthat is considered to be the most serious pest of coffee crops. Thisfinding contradicts previous studies that reported the lack of inhib-itory activity in the genus Lupinus, but supports those of Scarafoniet al. (2008) who found trypsin inhibitors in L. albus seeds.

The PI amino-terminal sequence showed similarity with thevicilins of L. albus, suggesting that vicilins could also play a rolein defense by acting as inhibitors of the aspartic proteases of Cole-opteran insects. However, it is necessary to obtain the whole se-quence of the protein, as well as to conduct studies in vivo toprove the effect of LbAPI on the growth and development of H.hampei.

4. Experimental

4.1. Materials

Dry seeds of L. bogotensis were purchased from Semicol (Bogotá,Cundinamarca, Colombia). The breeding unit of the National Centerfor Coffee Research (Cenicafé), located in Chinchiná (Caldas,Colombia), supplied recently emerged adults of H. hampei thathad been raised on dry parchment coffee at 45% moisture undercontrolled conditions at 27 �C and 65–75% relative humidity.

The reagents used as well as the bovine hemoglobin, bovine ser-um albumin (BSA), pepsin (EC 3.4.23.1), trypsin (EC 3.4 21.4), chy-motrypsin (EC.3.4.21.1), pepstatin A, N a-benzoyl-L-arginine ethylester (BAEE), and N-benzoyl-L-tyrosine ethyl ester (BTEE) were ob-tained from Sigma–Aldrich (St. Louis, MO, USA). The Bio-Gel� P-6DG exclusion resin, packed in Econo-Pac� 10DG columns, waspurchased from Bio-Rad (Hercules, CA, USA). The high performanceionic exchange resin for both the Q Sepharose and the K9/30 col-umn were obtained from Amersham Pharmacia (Uppsala, Sweden).The ultrafiltration membranes (30 kDa) and the Centriprep� werepurchased from Amicon (Beverly, MA, USA). The 3.5 kDa dialysismembranes were obtained from Spectra/Por� (Broadwick Street,CA, USA).

4.2. Purification of LbAPI

The dry seeds of L. bogotensis were ground in a Retsch cryogenicmill and sieved in a standard separator using two meshes: 1.4 and0.85 mm. The powder obtained (150 g) was defatted by stirring for30 min in a mixture of CHCl3:MeOH (750 ml, 4:1, v/v), with theextraction process repeated three times. The dry powder (100 g)was extracted at 4 �C using deionized H2O (500 ml) over a 6 h per-iod. The suspension was clarified by centrifugation at 7500 g for30 min, the pellet was discarded and the suspension was lyophi-lized. The lyophilized crude extract (3 mg/ml) was added to an

Econo-Pac� 10DG exclusion size column (1.5 � 12 cm) packed withBio-Gel P-6DG to eliminate pigments. The crude protein in the ex-tract was precipitated with (NH4)2SO4 at 80% saturation, and theprecipitate was then dissolved in 20 mM Tris–HCl buffer, pH 8, dia-lyzed against deionized water in a 3.5 kDa membrane (Spectra/Por�), and then lyophilized.

Lyophilized proteins were subsequently dissolved in 10 ml of20 mM Tris–HCl buffer, pH 8, for separation by ultrafiltration witha 30 kDa membrane. The eluted fractions from the 30 kDa mem-brane exhibiting inhibitory activity against aspartic proteases ofH. hampei were concentrated using a Centriprep� with a 3 kDacut-off limit (Amicon, Beverly, MA). This material was subse-quently purified on a Q Sepharose anionic exchange column(9 mm i.d. � 30 cm) (Amersham Pharmacia�) equilibrated with20 mM Tris–HCl buffer, pH 8. After washing the column with buffersolution, elution was performed with 400 ml of a 0.0–0.4 M NaCllinear gradient in 20 mM Tris–HCl buffer (pH 8) at a flow rate of0.8 ml/min.

Fractions (1.5 ml) were collected, and those presenting inhibi-tory activity were combined and dialyzed against deionized waterin a 3 kDa cut-off Centriprep� (Amicon, Beverly, MA) and thenlyophilized. Proteins displaying inhibitory activity were separatedin 10% polyacrylamide gels (20 cm long, 0.75 mm diameter) undersemi-native conditions according to Schägger and von Jagow(1987) using the Bio-Rad Protean� II xi cell, 20 cm electrophoresissystem with a 180-V Bio-Rad power supply system for 16 h at 4 �C.The active protein was eluted from the gel using an electro-elutor(Bio-Rad Model 422) for 6 h at 4 �C.

After elution, the protein was dialyzed against deionized H2O(50 ml) using the Centriprep� of 3 kDa cut-off (Amicon, Beverly,MA) and then lyophilized. Active protein was collected and ana-lyzed for purity. SDS–PAGE was performed as previously described(Schägger and von Jagow, 1987).

4.3. Protein determination

The protein concentration was determined by the Bradford pro-tein assay (Bradford, 1976). Bovine serum albumin (BSA) was usedas the standard.

4.4. Polyacrylamide gel electrophoresis

Sodium dodecyl sulfate polyacrylamide gel electrophoresis(SDS–PAGE) was carried out as described by Schägger and von Ja-gow (1987). The proteins used as molecular weight standardswere: myosin (210 kDa), b-galactosidase (118 kDa), bovine serumalbumin (82 kDa), carbonic anhydrase (40 kDa), soybean trypsininhibitor (31 kDa), lysozyme (17 kDa), and aprotinin (6 kDa). Theproteins were detected by staining with 0.1% Coomassie brilliantblue R-250 for 30 min and distaining with 10% acetic acid for 1 h.

4.5. LbAPI zymogram

For the inhibitor zymogram, the LbAPI obtained by Q Sepharoseanionic exchange chromatography was separated by SDS–PAGE.The gel was incubated in a solution of 1% bovine hemoglobin in0.2 M citrate buffer (pH 2.5) for 1 h at 4 �C, washed with deionizedH2O, and then incubated in a solution of 10 mg/ml of the H. hampeiprotein extract in 0.2 M citrate buffer (pH 2.5) for 4 h at 30 �C. Fi-nally, the gel was stained with Coomassie blue for 30 min and thendistained with AcOH:H2O (1:9, v/v) for 1 h. Aspartic protease inhib-itors were observed as dark bands against a light background be-cause the aspartic protease digests the hemoglobin present in thegel, except at the site where the band (inhibitor) is located in thegel.

928 D. Molina et al. / Phytochemistry 71 (2010) 923–929

An inhibition zymogram was used to evaluate the inhibition ofaspartic proteases during gel electrophoresis. To do so, protein ex-tract (20 lg) obtained from the midgut of H. hampei was pre-incu-bated with LbAPI (5 lg) for 30 min at 30 �C, and then loaded into a10% native gel. The gel was incubated in a solution of 1% bovinehemoglobin in 0.2 M citrate buffer (pH 2.5) for 1 h at 4 �C. Finally,the gel was stained with Coomassie blue R-250 for 30 min and thendistained with 10% acetic acid for 1 h.

4.6. Mass spectrometry

Matrix-assisted laser desorption/ionization time-of-flight(MALDI-TOF) mass spectrometric analyses were performed usingan ABI PerSeptive DE-Pro MALDI-TOF mass spectrometer (AppliedBiosystems, Framingham, MA, USA). One microliter was spotted ona plate and mixed with a matrix (sinapic acid, CH3CN–H2O contain-ing (4:6, v/v)/1% CF3CO2H). Spectra were obtained in positive linearmode. The MALDI-TOF was calibrated using a Saquazyme calibra-tion mixture (Applied Biosystems) consisting of bovine insulin(5 kDa), E. coli thioredoxin (11 kDa), and horse apomyoglobin(16 kDa), prior to sample analysis.

4.7. Isoelectric focusing

A Phast System analyzer was used to measure the isoelectricpoint of the purified inhibitor using Phast Gel Media IEF 3–9 at apH range of 3–9. Samples were run in the gel as described by Phar-macia (Separation Technique File No. 100). After isoelectric focus-ing (IEF), the bands were detected by the Fast Coomassie StainingTechnique for the Fast System (Development Technique File No.200). Isoelectric point markers included the lentil lectin-middleband (pI 8.45), lentil lectin-acid band (pI 8.15), myoglobin-basicband (pI 7.35), myoglobin-acid band (pI 6.85), human carbonicanhydrase B (pI 6.55), bovine carbonic anhydrase B (pI 5.85), andb-lactoglobulin A (pI 5.20).

4.8. Amino-terminal determination

The first 22 amino acids from the amino-terminal end of thepurified protein were determined using repeated cycles of the Ed-man degradation method on an Applied Biosystems Procise� cLCProtein Sequencer, Model 491 cLC, at the protein sequencing facil-ity of Commonwealth Biotechnologies Inc. (Richmond, VA, USA).The sequences obtained were used to search for similarities withproteins reported in the databases of the National Center for Bio-technology Information (NCBI) (National Institutes of Health,USA) using the protein–protein BLAST (Blastp) program with thealgorithm for non-redundant protein sequences (nr).

4.9. Inhibitory activity assay

Inhibitor activities against proteases of H. hampei and pepsinwere determined by monitoring the hydrolysis of hemoglobinat 280 nm in 0.2 M citric acid, 0.1 M NaCl, pH 2.5 (Lenney,1975). The inhibitor LbAPI was incubated with either the asparticprotease of H. hampei in 0.2 M citrate buffer (pH 2.5) or 3.1 nMpepsin in 0.2 M citrate buffer (pH 2.5), respectively. After15 min incubation at 30 �C, 1.3 ml of 1% hemoglobin was added,and then after 2 h incubation at 30 �C, reactions were stopped byaddition of 2.5 ml of 0.34 M Cl3CO2H (TCA). Inhibitor activityagainst trypsin was assayed by monitoring the initial rate ofhydrolysis of BAEE at 253 nm in 0.15 M Tris–HCl, CaCl2 0.05 M,pH 8.1 after 3 min of incubation (Schwertz and Takenaka,1955). Inhibitor activity against chymotrypsin was assayed bymonitoring the initial rate of BTEE hydrolysis at 256 nm in0.15 M Tris–HCl, CaCl2 0.05 M, pH 8.1 after 3 min of incubation

(Birk, 1976). One unit of proteolytic activity was defined as theamount of enzyme that catalyzed an increase of 0.01 absorbanceunits under the assay conditions employed. Inhibitor activity wasdefined as the difference between the proteolytic activities mea-sured in the absence and presence of inhibitor.

4.10. Enzyme extraction

Adult H. hampei were homogenized in 0.2 M succinic acid buffer(pH 6.0) at a 1:5 w/v ratio and a temperature of 4 �C. The suspen-sion was centrifuged at 7500 g for 15 min at 4 �C. The supernatantwas dialyzed against deionized H2O in a 3.5 kDa membrane (Spec-tra/Por�). The supernatant was then lyophilized and used as thesource of aspartic proteases.

4.11. Ki determination

Enzyme inhibition vs. pepsin was characterized at two differentsubstrate hemoglobin concentrations: 0.075 and 0.150 mM. Sam-ples were prepared to reach inhibitor concentrations of: 0, 3, 4.5,and 6 lM. The initial slope v was determined for each inhibitorconcentration. The reciprocal velocity (1/v) vs. [PI] for each sub-strate concentration was plotted (Dixon plots). A single regressionline for each substrate concentration was obtained, and the Ki wascalculated from the intersection of the two lines.

4.12. Thermal and pH stability

To measure the stability of LbAPI, a solution of the inhibitor(1 mg/ml) was heated for 30 min at temperatures ranging from30 to 100 �C and then cooled to 0 �C before testing for residualinhibitory activity against the aspartic proteases of H. hampei. Allexperiments were carried out in triplicate, and the results reportedrepresent the mean value of three independent tests.

To measure the stability of LbAPI at different pH values, a solu-tion of the inhibitor (1 mg/ml) was diluted with an equal volume ofdifferent buffers (at 100 mM): sodium citrate (pH 2–4), sodiumacetate (pH 5), sodium phosphate (pH 6–7), Tris–HCl (pH 8), andsodium bicarbonate (pH 9–10). After incubation in each bufferfor 30 min at 37 �C, the pH was adjusted to 2.5 to evaluate theresidual inhibitory activity against the aspartic proteases of H.hampei. All experiments were carried out in triplicate and the re-sults presented represent the mean of the three independent tests.

Acknowledgements

We sincerely appreciate the Ministry of Agriculture and RuralDevelopment of Colombia for the funds provided to carry out thisproject, to Colciencias for sponsoring Diana Molina’s Ph.D. fellow-ship. We also thank Aristofeles Ortiz, and Oscar Hernandez fortheir assistance and to Dr. John Délano F., Dr. Alvaro Gaitán B.,Dr. Marco Cristancho., Dr. Hernando Cortina G. and J. L. Castro-Guillén for their invaluable help in revising the manuscript.

References

Aguirre, C., Váldes, S., Mendoza, G., Rojo, A., Blanco, A., 2004. A novel 8.7 kDaprotease inhibitor from chan seeds (Hyptis suaveolens L.) inhibits proteases fromlarger grain borer Prostephanus truncatus (Coleoptera: Bostrichidae). Comp.Biochem. Physiol., Part B: Biochem. Mol. Biol. 138, 81–89.

Araújo, C., Bezerra, I., Oliveira, A., Moura, F., Macedo, L., Gomes, C., Barbosa, A.,Macedo, F., Souza, T., Franc, O., Bloch, C., Sales, M., 2005. In vivo bioinsecticidalactivity toward Ceratitis capitata (fruit fly) and Callosobruchus maculatus(cowpea weevil) and in vitro bioinsecticidal activity toward different orders ofinsect pests of a trypsin inhibitor purified from tamarind tree (Tamarindusindica). J. Agric. Food Chem. 53, 4381–4387.

Bhattacharyya, A., Mazumdar, S., Mazumdar, S., Babu, C., 2006. A Kunitz proteinaseinhibitor from Archidendron ellipticum seeds: purification, characterization, andkinetics properties. Phytochemistry 67, 232–241.

D. Molina et al. / Phytochemistry 71 (2010) 923–929 929

Birk, Y., 1976. Trypsin and chymotrypsin inhibitors from soybeans. MethodsEnzymol. 45, 700–707.

Bradford, M.M., 1976. A rapid and sensitive method for the quantification ofmicrogram quantities of protein using the principle of protein–dye binding.Anal. Biochem. 72, 248–254.

Broadway, R.M., Duffey, S.S., Pearce, G., Ryan, C.A., 1986. Plant proteinase inhibitors:a defense against herbivorous insects? Entomol. Exp. Appl. 41, 33–38.

Brun, L.O., Marcillaud, C., Gaudichon, V., Suckling, D.M., 1989. Endosulfan resistancein Hypothenemus hampei (Coleoptera: Scolytidae) in New Caledonia. J. Econ.Entomol. 82, 1311–1316.

Cater, S.A., Less, W.E., Hill, J., Brzin, J., Kay, J., Phylip, L.H., 2002. Aspartic proteinaseinhibitors from tomato and potato are more potent against yeast proteinase Athan cathepsin D. Biochim. Biophys. Acta 596, 76–82.

Chaudhary, N.S., Shee, C., Islam, A., Ahman, F., Yernool, D., Kumar, P., Sharma, A.K.,2008. Purification and charaterization of a tryspin inhibitor from Putranjivaroxburghii seeds. Phytochemistry 69, 2120–2126.

Chrispeels, M.J., Raikhel, N.V., 1991. Lectins, lectin genes and their role in plantdefense. Plant Cell 3, 1–9.

Christeller, J.T., Farley, P.C., Ramsay, R.J., Sullivan, P.A., Laing, W.A., 1998.Purification, characterization and cloning of an aspartic proteinase inhibitorfrom squash phloem exudate. Eur. J. Biochem. 254, 160–167.

Derbyshire, E., Wright, D.J., Boulter, D., 1976. Legumin and vicilin, storage proteinsof legume seeds. Phytochemistry 15, 3–24.

Egaña, J.I., Uauy, R., Cassorla, X., Barrera, G., Yañez, E., 1992. Sweet lupin proteinquality in young men. J. Nutr. 122, 2341–2347.

Gallardo, F., Araya, H., Pak, N., Tagle, M.A., 1974. Toxic factors in Chilean legumes. II.Trypsin inhibitor activity. Arch. Latinoam. Nutr. 24, 183–189.

Galleschi, L., Bottari, A., Capocchi, R.R., Saviozzi, F., 1997. Partial characterization of apepsin inhibitor from soft wheat bran. Sci. Des. Aliments 17, 173–182.

Gatehouse, A., Boulter, D., 1983. Assessment of the antimetabolic effects of trypsininhibitors from cowpea (Vigna unguiculata) and other legumes in developmentof the bruchid beetle Callosobruchus maculatus. J. Sci. Food Agric. 34, 345–350.

Gatehouse, A.M.R., Hilder, V.A., Powell, K., Boulter, D., Gatehouse, J.A., 1992.Potential of plant-derived genes in the genectic manipulation of crops for insectresistance. In: Gatehouse, A.M.R., Hilder, V.A., Boulter, D. (Eds.), Plant GeneticManipulation for Crop Protection, Biotechnology in Agriculture No. 7. C.A.B.International, Wallingford, pp. 221–234.

Jaramillo, J., Borgemeister, C., Baker, P., 2006. Coffee berry borer Hypothenemushampei (Coleoptera: Curculionidae): searching for sustainable controlstrategies. Bull. Entomol. Res. 96, 1–12.

Laskowski, M., Kato, I., 1980. Protein inhibitors of proteinases. Annu. Rev. Biochem.49, 593–626.

Le Pelley, R.H., 1968. Pests of Coffee. Longmans Green and Co., London. p. 590.Lenney, J.F., 1975. Three yeast proteins that specifically inhibit yeast proteases A, B

and C. J. Bacteriol. 122, 1265–1273.Macedo, M.L.R., de Sá, C.M., Freire, M.G.M., Parra, J.R.P., 2004. A Kunitz-type

inhibitor of Coleopteran proteases, isolated from Adenanthera pavonica L. seedsand its effect on Callosobruchus maculatus. J. Agric. Food Chem. 52, 2533–2540.

Macedo, M.L., Garcia, V.A., Freire, M.G., Richardson, M., 2007. Characterization of aKunitz trypsin inhibitor with a single disulfide bridge from seeds of Inga laurina(S.W.) Willd. Phytochemistry 68, 1104–1111.

MADR-IICA-OAC (Ministry of Agriculture and Rural Development, InteramericanInstitute of Cooperation on Agriculture, and Colombian Agrochain Observatory)2006. Información estadística. <http://www.agrocadenas.gov.co/home.htm>(accessed January 2009).

Melo, T.S., Ferreira, R.B., Teixeira, A.N., 1994. The seed storage proteins from Lupinusalbus. Phytochemistry 37, 641–648.

Mota, A.C., DaMatta, R.A., Lima Kilho, M., Silva, C.P., Xavier-Filho, J., 2003. Cowpea(Vigna unguiculata) vicilins bind to the peritrophic membrane of larvalsugarcane stalk borer (Diatrea saccharalis). J. Insect Physiol. 49, 873–880.

Park, S.S., Abe, K., Kimura, M., Urisu, A., Yamasaki, N., 1997. Primary structure andallergenic activity of trypsin inhibitors from the seeds of buckwheat (Fagopyrumesculentum Moench). FEBS Lett. 400, 103–107.

Preciado, D.P., Bustillo, A.E., Valencia, A., 2000. Caracterización parcial de unaproteinasa digestiva proveniente de la broca del café (Coleoptera: Scolytidae).Cenicafé 51, 20–27.

Ramos, V., Freire, M.G.M., Parra, J.R.P., Macedo, M.L.R., 2009. Regulatory effects of aninhibitor from Plathymenia foliosa seeds on the larval development of Anagastakuehniella (Lepidoptera). Comp. Biochem. Physiol., Part A: Integr. Physiol. 152,255–261.

Rawlings, N.D., Tolle, D.P., Barrett, A.J., 2004. Evolutionary families of peptidaseinhibitors. Biochem. J. 378, 705–716.

Reis, C.M., Calvet, M.M., Sales, M.P., Fernandes, K.V.S., Gomes, V.M., Xavier-Filho, J.,1997. a-Amylase inhibitors of legume seeds and their involvement in theresistance to the bruchid beetles. Arquivos de Biologia e Tecnología 40, 413–418.

Richardson, M., 1991. Seed storage proteins: the enzyme inhibitors. In: Dey, P.M.,Harborne, J.B. (Eds.), Method in Plant Biochemistry, Amino Acids, Proteins andNucleic Acids, vol. 5. Academic Press, New York, pp. 259–305.

Ritonja, A., Krizaj, I., Mesko, P., Kopitar, M., Lucovnik, P., Strukelj, B., Pungercar, J.,Buttle, D., Barrett, A., Turk, V., 1990. The amino acid sequence of a novelinhibitor of cathepsin D from potato. FEBS Lett. 267, 13–15.

Ryan, C.A., 1990. Proteinase inhibitors in plants: genes for improving defensesagainst insects and pathogens. Annu. Rev. Phytopathol. 28, 425–449.

Sales, M.P., Gerhardt, I.R., Grossi-de-Sá, M.F., Xavier-Filho, J., 2000. Do legumestorage proteins play a role in defending seeds against bruchids? Plant Physiol.124, 515–522.

Scarafoni, A., Consonni, A., Galbusera, V., Negri, A., Tedeschi, G., Rasmussen, P.,Magni, C., Duranti, M., 2008. Identification and characterization of a Bowman-Birk inhibitor active towards trypsin but not chymotrypsin in Lupinus albusseeds. Phytochemistry 69, 1820–1825.

Schägger, H., Von Jagow, G., 1987. Tricine–sodium dodecyl sulfate–polyacrylamidegel electrophoresis for separation of protein in the range of 1 to 100 kDa. Anal.Biochem. 166, 368–379.

Schwertz, G.W., Takenaka, Y., 1955. A spectrophotometric determination of trypsinand chymotripsin activity. Biochim. Biophys. Acta 16, 571–575.

Silva, C.P., Xavier-Filho, J., 1991. Comparison between the levels of aspartic andcysteine proteinases of the larval midguts of Callosobruchus maculatus (F.) andZabrotes subfasciatus (Boh.) (Coleoptera: Bruchidae). Comp. Biochem. Physiol.Part B: Biochem. Mol. Biol. 99, 529–533.

Souza, E.M., Mizuta, K., Sampaio, M.U., Sampaio, C.A., 1995. Purification and partialcharacterization of a Schizolobium parahyba chymotripsin inhibitor.Phytochemistry 39, 521–525.

Sritanyarat, W., Pearce, G., Siems, W., Ryan, C., Wititsuwannakul, R.,Wititsuwannakul, D., 2006. Isolation and characterization of isoinhibitors ofthe potato protease inhibitor I family from the latex of the rubber trees, Heveabrasiliensis. Phytochemistry 67, 1644–1650.

Turner, B.L., 1995. A new species of lupinus (Fabaceae) from Oaxaca, México: ashrub or tree mostly three to eight meters high. Phytologia 79, 102–107.

Uchoa, A.F., DaMatta, R.A., Retamal, C.A., Alburquerque-Cunha, J.M., Souza, S.M.,Samuels, R.I., Silva, C.P., Xavier-Filho, J., 2006. Presence of the storage seedprotein vicilin in internal organs of larval Callosobruchus maculatus (Coleoptera:Bruchidae). J. Insect Physiol. 52, 169–178.

Valencia, A., Bustillo, A.E., Ossa, G.E., Chrispeels, M.J., 2000. a-Amylases of the coffeeberry borer (Hypothenemus hampei) and their inhibition by two amylaseinhibitors. Insect Biochem. Mol. Biol. 30, 207–213.

Vila, L., Quilis, J., Meynard, D., Breitler, J.C., Marfa, V., Murillo, I., 2005. Expression ofthe maize proteinase inhibitor (mpi) gene in rice plants enhances resistanceagainst the striped stem borer (Chilo supressalis): effects on larval growth andinsect gut proteinase. Plant Biotechnol. J. 3, 187–202.

Xavier-Filho, J., Campos, F.A.P., Ary, M.B., Peres, C., Carvalho, M.M., Macedo, M.L.,Lemos, F.J., Grant, G., 1989. Poor correlation between the levels of proteinaseinhibitors found in seeds of different cultivars of cowpea (Vigna unguiculata)and the resistance/susceptibility to predation by Callosobruchus maculatus. J.Agric. Food Chem. 37, 1139–1143.