IIT-BREF Biotek, Indian Institute of Technology, Kharagpur 721 302, India
*Plant Molecular and Cellular Genetics, Bose Institute, P1/12, CIT Scheme VII-M, Calcutta 700 054, India
^§Present address: Department of Microbiology and Cell Biology, Indian Institute of Science, Bangalore 560 012, India

†Corresponding author (Fax, 91-3222-77980; Email, soumitra@hijli.iitkgp.ernet.in).

Helicoverpa armigera is a major pest of many tropical crop plants. Soybean trypsin inhibitor (SBTI) was highly effective against the proteolytic activity of gut extract of the insect. SBTI was also inhibitory to insect growth when present in artificial diet. The gene coding for SBTI was cloned from soybean (Glycine max, CV Birsa) and transferred to tobacco plants for constitutive expression. Young larvae of H. armigera, fed on the leaves of the transgenic tobacco plants expressing high level of SBTI, however, maintained normal growth and development. The results suggest that in certain cases the trypsin inhibitor gene(s) may not be suitable candidates for developing insect resistant transgenic plants.

1. Introduction

Helicoverpa armigera is a lepidopteran, polyphagous insect pest that feeds on important major crop plants like cotton, tomato, tobacco, chickpea, and pigeon pea (Jotwani and Butani 1984). Protease inhibitors are polypeptides present in many plants, which in addition to their physiological function, provide a natural defense against insect attack (Green and Rayan 1972; Laskowski and Sealock 1978; Malehorn et al 1994). Possibility of utilizing trypsin inhibitor gene(s) of plant origin as bio-insecticide for developing insect resistant transgenic crop plant has been a fairly accepted line of approach in crop biotechnological programme. Trypsin inhibitors present in different plants show variable levels of activity towards the proteolytic enzymes of the target insects (Broadway and Duffey 1986; Purcell et al 1992). It has been reported earlier that transgenic plants with suitable trypsin inhibitor gene can resist against insect damage (Hilder et al 1986; Johnson et al 1989; Xu et al 1996).

In the present study, trypsin inhibitors were purified from seeds of soybean (Glycine max) and cowpea (Vigna unguiculata). Soybean trypsin inhibitor (SBTI) was found to be more effective than cowpea trypsin inhibitor (CPTI) in reducing the proteolytic activity of gut extracts obtained from full-grown larvae of H. armigera. Thus SBTI was selected as the candidate gene for developing the insect resistant transgenic plant. The gene coding for the SBTI peptide was cloned. Suitable chimeric construct was developed for over expression of the SBTI gene when transferred to tobacco. Through Agrobacterium tumefaciens mediated gene transfer protocol, transgenic tobacco plants showing high expression of the transferred SBTI gene were generated. Feeding tests of young larvae of H. armigera on the leaves of the highly expressive transgenic tobacco plants did not show significant level of impairment of insect growth in comparison to untransformed plants.

2. Materials and methods

2.1 Materials

Soybean seeds (G. max, CV Birsa) were obtained from Birsa Agricultural University, Ranchi and cowpea seeds (V. unguiculata, CV Black eye) were obtained from a local nursery. H. armigera insects were collected from Plant Protection Division, Central Institute for Cotton Research, Indian Council of Agricultural Research, Nagpur.

2.2 Purification of trypsin inhibitors

The trypsin inhibitors were purified from mature seeds of soybean and cowpea following the method of Gatehouse et al (1980). About 40 g dry seeds were ground into powder and extracted in 200 ml extraction buffer (0·1 M Na-acetate, 0·3 M NaCl, 0·01 M CaCl₂, pH 4·0) for 16 h at 4° C. The extracts were centrifuged at 10,000 g for 30 min. The supernatant was collected and ammonium sulphate was added to the final concentration of 30%, 60% and 90% (w/v) sequentially. Most of the trypsin inhibitor activity was retained in the 60% ammonium sulphate precipitate, which was further dissolved in 50 mM Tris-Cl (pH 8·0) and passed through trypsin ligated CNBr⁺-sepharose column. Bound trypsin inhibitor was eluted with 0·2 M HCl. Purified trypsin inhibitors were run in SDS-PAGE following the protocol of Shagger and Jagow (1987).

For purifying SBTI from transgenic tobacco plants, 1 g leaf tissue from different transgenic tobacco plants was extracted with 10 ml extraction buffer, and centrifuged
at 4° C for 15 min at 48,000 g. To the supernatant, 1 M Tris-Cl (pH 8·0, 1·0 ml) and trypsin ligated CNBr⁺-sepharose beads (0·5 ml) were added, mixed thoroughly for 3–4 h at 4° C. SBTI was eluted from the beads by 0·2 M HCl. Total protein in each fraction was determined by Bradford (1976) assay, and the amount of SBTI present was determined by enzyme linked immunosorbant assay (ELISA).

2.3 Extraction of H. armigera larval midgut and proteolytic activity assay of the gut extract

Midguts of 20 full grown H. armigera larvae were extracted in 1 ml buffer containing 0·1% NaCl and 50 mM Tris-HCl (pH 8·0). The extract was centrifuged at 10,000 g for 30 min at 4° C. The supernatant was diluted 50-fold with the same buffer. To 100 m l of the diluted stock, 100 m l of 5 mM CaCl₂ in 50 mM Tris-HCl (pH 8·0) was added along with different concentrations of the purified inhibitors. The volume was made up to 450 m l with water and incubated at 37° C for 30 min. Subsequently, 250 m l of azocasein (2·5%) was added to each tube and incubated at 37° C for 10 min. The reaction was stopped by adding 700 m l of 10% TCA. The reaction mixtures were centrifuged, supernatants were mixed with equal volume of 0·5 N NaOH and the absorbance at 428 nm was recorded.

2.4 Cloning of SBTI gene and construction of plant expression vector

Total genomic DNA from the germinated seeds of soybean was isolated following the method of Rogers and Benedich (1988). SBTI gene was amplified by polymerase chain reaction (PCR) by using 100 ng of DNA, 2 m M of each primer, 0·2 mM of dNTPs, 1 × Taq buffer, 2 mM MgCl₂ and 10 U Taq DNA polymerase in 100 m l reaction volume. Thermo-cycling was carried out at 94° C 1 min, 55° C 1 min and 72° C 1 min for 30 cycles with a pre-heating at 94° C for 4 min and extension at the end for 7 min at 72° C. The amplified product was cloned into pUC18 vector at BamHI and SmaI restriction sites and subsequently subcloned into the plant expression vector, pBI121 (Jefferson et al 1987) at BamHI and SacI restriction sites. All molecular biological methods adopted were as described in Sambrook et al (1989).

2.5 Development of transgenic tobacco plants

A. tumefaciens, strain LBA4404 was transformed with the plant expression vector by freeze-thaw method of Hofgen and Willmitzer (1988). Tobacco leaf discs were co-cultivated with overnight grown A. tumefaciens culture and transformed plants were regenerated (Horsch et al 1985) on plates containing 50 mg/l kanamycin.

2.6 Analysis of transgenic tobacco plants

Transgenic tobacco plants were analysed in search for the presence of the SBTI gene following Southern blot analysis (Southern 1975). Total genomic DNA was isolated from several independent transgenic lines and digested with restriction enzyme EcoRV, run in 0·9% agarose gel and transferred to nylon membrane. The presence of SBTI gene was monitored through the use of radiolabelled probe. Expression of SBTI gene in transgenic plants was detected by Western blot analysis (Towbin et al 1979). Total protein was extracted by trypsin extraction buffer (pH 4·0) and separated in a 15% SDS-PAGE (Laemmli 1970) and transferred to nitrocellulose membrane by electro-transfer method. SBTI band was detected through the use of anti-SBTI rabbit polyclonal antibody and HRP linked goat anti-rabbit IgG antibody.

2.7 Feeding assay of H. armigera larvae

Adults of H. armigera were collected and reared for two generations in the laboratory on artificial diet before using them for feeding assay. About 250 g of redgram flour was boiled for 15 min in a pressure cooker with 250 ml of water and homogenized. To the warm mixture, methyl-p-benzoate (1·65 g), sorbic acid (0·85 g), ascorbic acid (2·65 g), streptomycin sulphate (0·12 g) 10% formaldehyde (6·75 ml) and multivitamin complex (800 mg) were added with the addition of SBTI and/or casaminoacid, whenever needed. The mixture was finally distributed into individual vials. Thirty larvae (three days old) were fed in each experiment set. Parallel feeding assay of the larvae (three days old) was conducted on the fresh leaves of transgenic and untransformed tobacco plants. In all the cases body weight of individual larva was recorded at regular interval.

2.8 Determination of free amino acids and amide contents

Fresh leaf tissues of tobacco (200 mg) suspended in 1 ml 0·1 M Tris-Cl (pH 8·0) were homogenized in liquid nitrogen. The extracts were then centrifuged at 10,000 g for 10 min at 4° C. The supernatant was mixed with equal volume of 10% TCA, incubated on ice for 10 min and centrifuged. Each aliquot (0·1 ml) was mixed with 1·0 ml of ninhydrin reagent (Spies 1957) and boiled in water bath for 20 min. Absorption values of the developed purple colour were recorded at 570 nm. Casein acid hydrolysate (casaminoacid) dissolved in same buffer and treated in the similar manner, represented for the standard line of the experiment.

3. Results

3.1 Purification of trypsin inhibitors

Trypsin inhibitors were purified from the extracts of mature seeds of soybean and cowpea by ammonium sulphate precipitation followed by affinity column chromatography and analysed by SDS-PAGE (figure 1). Very often, inhibitors like CPTI, belonging to Bowman–Birk family (Birk 1961) are present in multimeric forms due to the presence of multiple S–S linkages (Gennis and Cantor 1976). Boiling with 2% SDS and 2% b-mercaptoethanol was not sufficient to convert them fully to monomeric forms and a faint band of dimeric CPTI (>> mark in lane 1 figure 1) could also be found along with a major 8 kDa band (arrow). Soybean seeds contain both Kunitz type (KTi, Kunitz 1947) and Bowman–Birk type (SBBI, Kay 1976; Wei 1983) trypsin inhibitors. Affinity purified SBTI (20 kDa, Kunitz type) was obtained with a small amount of co-purified proteins (<< mark in lane 3 and 4, figure 1). The molecular weights of these co-purified proteins and their ability to bind trypsin ligated sepharose column indicate that they represent Bowman–Birk type inhibitor. The Kunitz type SBTI corresponds to about 99% of the purified protein.

3.2 Functional activity assay of purified trypsin inhibitors

The efficacy of CPTI and SBTI in reducing the proteolytic activity of gut extracts of H. armigera was determined. To the larval gut extract, aliquots of purified inhibitors were added (1 to 6 m g) and incubated at 37° C for 30 min. This allowed the trypsin to bind with the inhibitor. The activity of trypsin inhibitors was determined in terms of inhibition of proteolytic activity, as revealed through reduction of absorbance values of the TCA extract of the reaction mixture at 428 nm (figure 2). Similar treatments were carried out on bovine trypsin, as standard. Results indicated (figure 2) that both the purified inhibitors were effective against the proteolytic activity of bovine trypsin and H. armigera gut extract. However, SBTI was selectively more effective than CPTI against the gut extract. Hence, SBTI was selected as the target trypsin inhibitor for the present study.

3.3 Cloning of SBTI gene

Functional SBTI is coded by KTi3 gene in soybean (Jofuku et al 1989). Two primers were designed from the available information on the nucleotide sequences of KTi3 gene (Jofuku and Goldberg 1989); primer-1 (5¢ atcccgggatccATGAAGAGCACCATCTTCTT3¢ ) and primer-2 (5¢ tagggatcccgggtcACTCCATGCGAGAAAGG-CCATATTTTCT3¢ ) for amplification of the complete coding region of the gene with additional flanking sequences (lower case letters) for SmaI and BamHI restriction enzyme sites. With the help of these primers, PCR amplification was carried out by using total soybean genome, as template (figure 3A).

The amplified product was cloned into a pUC18 vector and transformed into E. coli, DH5a . It is known that soybean genome contains other non-functional members of KTi gene family (Jofuku et al 1989). By comparing the available KTi gene family nucleotide sequences, a third primer, primer-3 (5¢ AGGTCCTTCTGGTAGATC3¢) was designed from the KTi3 specific region. Plasmid DNA isolated from the independent colonies were further subjected to PCR based screening with primer-1 and primer-3 to exclude the possibility of inclusion of any such non-functional KTi3 homologue. The clones that showed amplification of 380 bp in second PCR (figure 3B), were finally selected. The cloned SBTI gene was sequenced with the help of gene specific primers by chain termination method of Sanger (1977) using [a ³²P]dATP. Possibility of generating any error from mis-incorporation by Taq DNA polymerase in PCR step was eliminated by repeating the experiment several times. Nucleotide sequences were found to be identical to that of the known KTi3 gene except for three nucleotides. These three nucleotide mismatches resulted in the alteration of two amino acids, glu¹² and asn¹³ to asp¹² and ser¹³, respectively. Three different types (Ti^a, Ti^b and Ti^c) of Kunitz trypsin inhibitors with altered amino acid sequence in soybean have been reported earlier (Kim et al 1985). The functional KTi3 gene encodes for the polypeptide sequence of Ti^a type (Jofuku et al 1989). Ti^a and Ti^c are similar except that gly⁵⁵ of Ti^a is replaced by glu⁵⁵ in Ti^c. However, Ti^b is quite different from others with as many as 8 different amino acids, which includes conservative as well as non-conservative replacements. In the present case, glu¹² of Ti^a is replaced by asp¹² that is common in the homologous protein series but asn¹³ to ser¹³ is a rare substitution. Thus, the cloned gene may represent for a mutant KTi3 gene. It may also be likely that it may represent for a new member gene of KTi family. However, the functional properties of the cloned SBTI remained unaltered as revealed through trypsin inhibition activity of SBTI purified from the transgenic tobacco plants.

3.4 Construction of expression vectors and transformation of tobacco plant

The SBTI gene was cloned into a plant expression vector, pBI121 (Jefferson et al 1987) at BamHI and SacI restriction sites to generate pSBIN. The chimeric gene construct was composed of the SBTI gene fused with CaMV 35S promoter and the nos polyadenylation terminator sequences. The recombinant vector, pSBIN was then transformed into A. tumefaciens (strain LBA4404) and used for plant transformation. Tobacco leaf discs were transfected with A. tumefaciens strain containing the pSBIN. Plantlets were regenerated on MS medium (Murashige and Skoog 1962) containing 2 mg/l IAA, 8 mg/l kinetin and 50 mg/l kanamycin. Regenerated plantlets were transferred to 1/2 MS for rooting and finally transferred to soil.

3.5 Analysis of transgenic plants for the presence and expression of SBTI gene

The presence of the transferred SBTI gene in the tobacco genome was detected by Southern blot analysis. Genomic DNA (5 m g) from different transgenic plants were digested by restriction enzyme EcoRV and probed with [a ³²P]dATP labelled SBTI DNA. There exists only one restriction site for EcoRV in the pSBIN vector construct outside the coding sequences of SBTI. Thus the number of bands in the autoradiogram (figure 4) indicated the number of independent insertion sites of the transferred SBTI gene within the tobacco genome, which varied from one to five.

Expression of the transferred SBTI gene was determined by Western blot analysis. Fresh leaves of transgenic tobacco plants were extracted with trypsin inhibitor extraction buffer. The extracted leaf protein (5·0 m g) was run on a 15% SDS-PAGE. The proteins separated in the gel were transferred to nitrocellulose membrane through electro-transfer. Detection of SBTI specific band was carried out with the help of polyclonal anti-SBTI rabbit antibody and anti-rabbit IgG goat antibody linked with horse radish peroxidase, used as primary and secondary antibodies, respectively. Western blots of five individual transgenic plants showed (figure 5) variations in expression level of the transgene.

3.6 Purification, quantitation and functional activity assay of SBTI from transgenic plants

Level of expression of SBTI in the transgenic tobacco plants was estimated through indirect ELISA, after purifying through trypsin ligated CNBr⁺ sepharose beads. The amount of SBTI recovered from different transgenic plants varied from undetectable to 1% (w/w) of the total protein (table 1). However, tests carried out with known amount of pure SBTI when mixed with untransformed tobacco plant extracts, revealed that 30–45% of the inhibitor could only be recovered following the above mentioned protocol. This reflected that the level of SBTI expression in the transgenic plants might be in certain cases as high as 3 to 5% of the total soluble protein.

Functional activity assay of the trypsin inhibitor recovered from transgenic plants was estimated on bovine trypsin as well as on H. armigera gut extract. Results (data not shown) indicated that SBTI recovered from soybean seeds and from leaves of transgenic plants were equally effective in reducing the proteolytic activity of bovine trypsin and the insect gut extracts.

3.7 Determination of free amino acids and amides content of tobacco leaves

The feeding assay experiment (described in the next section) with the larvae of H. armigera on the leaves of transgenic tobacco plants indicated that the insect growth is favoured by a certain component of the growing plants. To determine if the free amino acids and amides of the growing plants play any role in combating antimetabolite effect of trypsin inhibitor, the level of these compounds present in the tobacco leaves was estimated. Leaves were collected from top, middle and lower parts of young tobacco plants. Amount of total ninhydrin positive compound (NPC) present in the leaves was estimated by ninhydrin reaction method (Spies 1957). The level of NPC present was estimated to be 0·8 ± 0·04% (w/w) in the top leaves. The middle and lower leaves showed 0·4 ± 0·03% and 0·3 ± 0·03% (w/w) NPC contents, respectively. NPC present in the artificial diet of H. armigera was found to be 0·15%.

3.8 Insect feeding assay

Purified SBTI was added at concentrations of 0·05% and 0·1% (w/w) to the artificial diet of H. armigera. In each set of diet, three days old 30 larvae were reared. Similarly, three days old 30 larvae were fed on leaves of transgenic plants. Changes in body-weight of the larvae were monitored at three, five and seven days of incubation. Feeding assay on the leaves of highly expressive SBTI transgenic lines showed that there was hardly any difference in the body weight of the larvae from the ones grown on leaves of untransformed plants (figure 6). Analysis of variance showed (calculated F = 0·004 against the tabulated F_{(0·05, 2,12)} = 3·38) that the observed differences in body weights were not significant. Development and fecundity of these insects remained unaffected. On the other hand, SBTI recovered from the transgenic plants when mixed at 0·05% (w/w) with artificial diet, caused (figure 6) significant reduction in body-weight (t = 12·75, against t_{(0·05, 58)} = 2·0).

Insect feeding assay was also carried out with the addition of SBTI and casaminoacids. Purified SBTI from transgenic tobacco plants was added to the artificial diet at 0·05% and 0·1% (w/w), along with 0·4%, 0·8% and 1·6% (w/w) of casaminoacids. The results of the insect feeding assay have been summarized in figure 6 by showing the effect of SBTI at 0·05% with or without 0·4% and 0·8% of casaminoacids, as any further higher dosages of SBTI (0·1%) or casaminoacids (1·6%) were similar to that observed with lower dosages. The results demonstrated that toxicity of SBTI to insects could be reversed by supplementing casaminoacids in the artificial diet at an equivalent level (0·8%) that is present in the young tobacco leaves.

4. Discussion

Insecticidal properties of protease inhibitors are well established. Transgenic plants with constitutive expression of protease inhibitors have shown in the past to be resistant to insect pests (Hilder et al 1986; Johnson et al 1989, 1991; Duan et al 1996; Irie et al 1996). On the contrary, our findings indicated that the growth and development of H. armigera larvae were unaffected when fed on the leaves of transgenic tobacco plants expressing very high levels of SBTI. Purified SBTI extracted
from the transgenic plants when mixed with artificial diet caused significant reduction in growth of H. armi-gera larvae. The SBTI produced in transgenic plants
was functionally active, as evident from it’s effective-ness in reducing the proteolytic activity of insect gut extract.

The mode of action of the protease inhibitor as anti-metabolite is not fully understood. Direct inhibition of digestive enzymes is not considered to be the main effect. It has been conjectured (Pusztai et al 1992) that a more important factor might be depletion of essential amino acids caused due to the presence of the inhibitors. However, parallel suggestions like insects grown on transgenic plants can produce inhibitor resistant proteolytic enzymes in the body to overcome the effect of the anti-metabolite has also been put forward (Jongsma et al 1995). Transgenic tobacco plants with giant taro proteinase inhibitor (GTPI) gene failed to cause mortality to feeding larvae of H. armigera and this resistance was associated with the production of increased level of chymotrypsin, elastage and GPTI insensitive protease (Wu et al 1997) by the insects. In the context of the present study, this however, cannot explain the differential behaviour of the insects when fed on artificial diet and on transgenic plant leaves. It is observed that insect growth is somehow favoured when feeding on growing plant parts. Free amino acids and amides present in the growing plants perhaps help the insects to overcome the stress imposed by the presence of the anti-metabolite. Our results suggest that the high levels of free amino acids and amides present in the young tobacco leaves may be sufficient to sustain growth and development of the insects, even in the presence of SBTI. The artificial diet contains much less amount of free amino acids and amides that are not sufficient for the sustenance of the insects. Insects, belonging to the order hemiptera, are known to nourish on free amino acids and amides of plants (Sogowa 1970).

There is no report on the effectivity of SBTI towards the insect H. armigera. However, transgenic tobacco plants expressing CPTI have been shown to be resistant against this insect (Hilder et al 1986). The SBTI gene has been reported to be inhibitory to the brown plant hopper in transgenic rice plants (Lee et al 1999). In contrast to our findings, a recent report (Schuler et al 1998) indicates that the serine-proteinase inhibitor KTi3 from soybean resulted up to 100% mortality of first instar cotton leafworms (Spodoptera littoralis) when expressed in transgenic tobacco. However, the fact remains that no crop expressing proteinase inhibitor transgene(s) has yet been commercialized. Thus, the present findings and the earlier report (Jongsma et al 1995) points to the possibility that the trypsin inhi-bitors may not be suitable insecticidal agents for a transgenic approach for crop protection against insect damage.

Acknowledgement

Financial assistance in the form of a fellowship by the Council of Scientific and Industrial Research, New Delhi to AKN is gratefully acknowledged.

MS received 15 October 1998; accepted 16 August 1999

Corresponding editor: Rakesh Tuli