Cross–Institutional co–operation in the development of new Bacillus thuringiensis products in vietnam O. Skovmand1 ,N. Thi Hoai Tram2, D. Thi Thanh Huyen2, N. Thanh Ha2, N. Giang Lien3 , N. Thi Hoai Ha3, and Le Van Trinh4 1) Intelligent Insect Control, Montpellier, France. 2) Food Industries Research Institute, 301 Nguyen Trai Rd., Thanhxuan, Hanoi, Vietnam. 3) Hanoi University of Science -Vietnam National University, 334 Nguyen Trai Rd., Thanhxuan, Hanoi, Vietnam 4) National Institute of Plant Protection, Dongngac, Tuliem, Hanoi, Vietnam. ABSTRACT The entomopathogen Bacillus thuringiensis (Bt) has been used in experimental production, screening programs, field tests, and sold commercially in Vietnam. 3 Vietnamese participants that now co-operate in a product development project have been involved in several phase of such programs: FIRI (Food Industries Research Institute), VNU (Hanoi University of Science, Vietnam National University) and NIPP (National Institute of Plant Protection). The present program profits from the experience in these institutions to make a coordinated effort to develop affordable and efficient Bt products to Vietnamese farmers for the control of diamondback moth (Plutella xylostella), and Helicoverpa armigera, Spodoptera litura, Spodoptera exigua and rice leaf folder (Cnapalocrocis medinalis). These insects are important leaf-eating, lepidopteran pests in Vietnam, and the latter are badly or not controlled with existing, commercial Bt products (without beta-exotoxin). For the selection strains for fermentation and product development, the specific activity for all screened strains are calculated. Contrary to LC-50, this figure is not dependent on fermentation yield, which is a function of suitability of the substrate for the strain. The specific activity can be calculated from LC-50 and fermentation yield measured as amount of cry-protein in the tested sample or product. It is a fast method for optimizing strain selection for insect species. The program includes the purchase of new equipment for the project as well as updating existing equipment used in the screening, bioassay and pilot production. Concept and strategy was developed with the European company Intelligent Insect Control. The Vietnamese Government and the Danish Aid Program under Danida support the project. Keywords: Bacillus thuringiensis, Plutella xylostella, Helicoverpa armigera, specific activity. INTRODUCTION Bacillus thuringiensis (Bt) is a spore-forming, rod-shaped, gram-positive bacterium. During sporulation, Bt produces intracellular protein crystalline bodies, -endotoxins, which are pro-toxins. Upon activation, they are highly toxic to specific insects. When insects digest the crystals, these are solubilized by the alkaline digestive juice in the mid-gut, then activated by proteases, and the insects die while some stomach lining cells are destroyed. Among all microbial insecticides, products based on Bt are the most successful. Several basic research programs in Vietnam have tested the effect of various strains of Bt on important pest species. The quality of the generated products was a major problem, not only due to low efficiency rate of the used Bt strain, but also due to poor quality of formulation and storage ability. Consequently, these products were not economically viable on the Vietnamese market. Three Vietnamese institutions belonging to the Ministry of Industry, the Ministry of Agriculture and Rural Development and the Vietnam National University have in co-operation with Intelligent Insect Control formulated a project where each institution has a role in the Bt product development based on it’s core experience. The project obtains support from Danish Aid Agricultural Program under Danida aiming at developing cheap and high quality Bt products for control of insect pests on rice and vegetables. In this paper, we describe the preliminary results of our Danida project. MATERIALS AND METHODS - Shake flask cultures were produced on a rotary shaker at 200 rpm and 30oC. Spore-crystal proteins complexes were recovered after 72 hr. - Batch culture experiments were carried out in a 14 L fermentor (Bioflo 110) supported by Danida project. The impeller speed was maintained at 350 rpm, airflow at 5-7 liters/min at 0.5 bar. The spore-crystal proteins complexes were recovered by continuous flow centrifugation (Heraeus) and then eventually formulated as wettable powder by drying or the ferment was formulated directly for bioassays. - The number of viable cells and free-spores was counted separately using a Neubauer bacterial counting chamber under phase contrast light microscope (Nikon E400). - Crystal proteins were isolated, purified by method of Yamamoto (1990), quantitative analyzed by Bradford (1976) and qualitative analyzed by SDS-PAGE , Laemmili (1970). - PCR technique was used to identify cry genes of Bt strains using primers specific to cryI (Lep1A, Lep1B, Lep 2A, Lep 2B and K5un2, K3uni2) by method of Asano (1996). - Bioassay: Second-instar larvae of Plutella xylostella or Heliothis armigera were used for bioassays: three to five dilutions per test sample and 10 larvae per concentration. Indicative LC50 was obtained with 3 concentrations, 10 times dilutions, and the value confirmed in bioassay with 5 concentrations, 3 times dilution. The trials were implemented in climate chamber, 82% RH, 18oC, 14/10 light on/off. Bioassays in this phase were carried out using leaf dipping technique for P. xylostella and H. armigera. Observations 2 and 4 days after treatment. RESULTS AND DISCUSSION Our strategies of Bacillus thuringiensis product development are as following: Strategy I is a more basic one that involves identifying strains active against P. xylostella, H. armigera, and Cnapalocrocis medinalis (rice leaf folder) based on the identification of specific activity of strains to target insects. Strategy II aims at developing products based on Bt strains with known and sufficient activity against target insects identified by strategy I or known from existing products. Selected strains are fermented on cheap substrates, activity quantified in bioassays and finally tested in field plots. In the first period of the project (from Oct. 1, 2001 to March 31, 2002), 38 strains Bt strains of the Type Culture Collection of FIRI and VNU - including several strains isolated from commercial products - were studied for growth and sporulation rate and sporal lysis. 44 samples of ferments were tested for activity on diamond back moth (DBM, Plutella xylostella) and H. armigera in National Institute of Plant Protection (NIPP). Strains with high specific activity were kept and sent to VNU that characterized strains by morphology of cry toxin, and toxin profile by PCR. Meanwhile, FIRI worked on further fermentation development on industrial substrates. Results of fermentations are shown in table 1A, and of bioassay activity in table 2. These tables show that several strains have high activity against P xylostella, whereas fewer have high activity against H armigera, among these two commercial strains (Btkcp and kurstaki) and three Vietnamese isolates (Table 2). For selected samples, the activities as measured by LC50 are compared to the amount of cry protein produced (Table 3). When the ratio of specific activity is calculated (1000 /(LC50 x Cry-Protein ug/ml), the strain with the highest activities are found disregarding the strain ability to grow on the tested substrates. This ratio is constant for a given strain and independent on substrates unless the fermentation substrate influences the composition of the cry toxin or the amount of other metabolites that may contribute to the toxicity. The data seems to support that various substrates that may differ in yields, do not differ in the specific ratio, but this has to be further investigated. Yields of fermentation are measured as spores/ml after 38-40 h cultivation and mostly arrived to about 1.2 x109spores/ml, depending on substrate (Fig 1). A mixture of spores, crystals and debris was harvested by centrifugation and formulated. In a few cases, spray dried powder was made and tested. The project continues screening Bt strains from collections to get access to and characterize novel strains type for control of H. armigera and C. medinalis and ferment and develop formulations of Bt kurstaki strains toward these insects and P. xylostella. The pilot plant of FIRI will be renewed in 2003 and produce the first products for field test on P xylostella in cabbage. These tests will be carried out in co-operation between NIPP and the Plant Protection Department of the Ministry of Agriculture. Table 1: Selected Bt strains and their quantitative characterization Spore count Rate of sporal Protein by Bt strain Media (x109/ml) lysis (%) Product form Bradford (g/ml) Btk CF B2A 1,12 83 Liquid-Flask 138 (Bt kurstaki) HPT 1,00 95 Liquid-Ferm 488 HPT 0,84 100 Liquid Flask 740 HP 0,90 93 Liquid-Flask 1735 HP 0,94 92 Liquid-Ferm 3225 Dendrobacilin CYS 1,4 99 Liquid-flask 176 (Bt dendrolimus) HP 1,4 98 Liquid-Flask 143 HPT 1,3 98 Liquid-Flask 350 Bt kurstaki CYS 0,79 100 Liquid-Flask 120 (Bt kurstaki) HP 0,93 100 Liquid-Flask 176 B2A 0,72 94 Liquid-Flask 330 Bt AMS CYS 0,75 100 Liquid-Flask 164 (Bt kurstaki) Bt HD12 HPT 1,07 100 Liquid-Flask 168 (Bt entomocidus) B3A 1,03 100 Liquid-Flask 156 CYS: contains glucose, peptone, yeast extract, mineral salts (Tram et al. 02); HP: contains glucose, peptone, and mineral salts (Giang et al. 1996); HPT: contains soybean, maize meals and mineral salts (Giang et al. 1996); B2A: contains 1% Vietnamese fish meals + HPT; B3A: contains 1% Malaysia fish meal + HPT. For all Bt strains, the highest spore counts are obtained with the HPT substrate. The product origin refers to fermentation equipment. Flask is shake flask, and Ferm is a Bioflo fermentor. Table 2: Selected Bt strains and their activity on two pest insects LC 50 after 4 days of treatment Protein Gene detected by Bt strain Media P.xylostella H.armigera detected by PCR (% ferment dilution (% ferment dilution) SDS-PAGE (kDa) Btk CP HP 0,010 1,30 (Bt kurstaki) 130 Cry I HPT 0,065 3,18 Bt kurstaki HP 0,50 1,92 130 Cry I (Bt kurstaki) CYS 0,50 2,22 Bt H_H1A VNU 0,07 4,15 130 Cry I Bt AM-4 VNU 0,87 3,44 130 Cry I (Bt kurstaki) Bt AMS CYS 0,28 36,7 130 Cry I (Bt kurstaki) CYSHP 0,58 10,3 Bt HD-12 B3A 50,9 (Bt entomocidus) 130 Cry I HPT 0,67 7,02 Bt 8_2 VNU 0,42 1,09 130 Cry I Bt G_3 VNU 0,23 2,30 130 Cry I Bt T_11 VNU 0,23 33,0 130 Cry I The table shows LC-50 values for some of the strains tested on P xylostella and H armigera. The last columns indicate the presence of 130 kDa toxins (by SDS-page) and Cry1 toxins (by PCR). The activity is measured as mortality of dilutions of ferment on dipped cabbage leafs. The ratio of activity against P xylostella and H armigera is not constant, not even for various kurstaki strains. The PCR method used for the detection of Cry genes does not discriminate between the types of Cry-I genes and is thus too insensitive to explain these variations. Further analysis discriminating between more crystal types will be made on the most interesting strains. Remark that for Btk-cp better activities (lower LC-50) are obtained on both lepidopteran species with a sample fermented on HP compared to the results obtained with that of HPT, even the latter gave a higher spore count (Table 1). Table 3 : Selected Bt strains and their specific activity on two pest insects Bt strain sample Fermentation ug protein LC-50 P Specific LC-50 H Spec. substrate /ml sample xylostella ratio armigera ratio Btk-CP HP 1735 1,30 0,44 HPT 740 0,65 2,08 3,18 0,43 B2A 138 1,50 4,83 Bt kurstaki HP 176 0,81 7,01 1,98 2,86 CYS 120 0,95 8,77 2,22 3,75 B2A 330 1,50 2,02 Dendrobacillin HP 143 0,57 12,3 11,32 0,62 Bt AMS CYS 164 0,28 21,8 36,7 0,17 The table shows the calculation of specific activity, which is a fast and substrate independent measure for insect species activity. The specific activity is calculated as 1000 x the reciprocal of the product of LC-50 (Table 2) and amount of cry protein (Table 1) per ml ferment. The higher the value of the specific activity, the more effective is the strain on the insect. A strain with high specific activity is a good strain, and if the yields are low (and thus the LC-50 high), the strain should be selected for fermentation development. In the examples shown, the highest specific activity on P xylostella is obtained with Bt kurstaki and Dendrobacillin, whereas the two other strains have lower specific activity. But the yields of Bt kurstaki and dendrobacillus measured in ug protein pr ml sample are smaller than those of the others, showing that the fermentation substrate should be better adapted to these strain. When the specific activity for H armigera is also considered, Bt kurstaki is obviously the choice of strain, being far more effective than the others. The specific activity is thus an effective tool for selecting Bt strains instead of LC-50 that inevitably is influenced on fermentation yield on a substrate that is not optimized for the strain. ACKNOWLEDGEMENTS The authors gratefully acknowledge the financial support from the Royal Danish Embassy in Hanoi. We thank Ms. Phuong Lien, Programme Officer of RDE for her helpful assistance, and thank Prof. Toshihiko Iizuka and Prof. Shin- ichiro Asano from Hokkaido University, Japan for PCR primers. The support and assistance of Ministry of Industries and other colleagues of FIRI is also gratefully acknowledged. REFERENCES Asano, S. 1996. Identification of cry gene from Bacillus thuringiensis by PCR and isolation of unique insecticidal bacteria. Memories Fac. Agr. Hokkaido Univ. 19: 529-563. Bradford, M. 1976. A rapid and sensitive method for the quantification of microgram quantificaties of protein utilizing the principle of protein-dye binding. Annal. Biochem. 72:248-254. Giang D.T., Tram N.T.H., Trang P.T. 1996. Study on Experimental Production and Application of Bt insecticide from Bt to control pests in Vietnam. In Proceedings of the Second Pacific Rim Conference on Biotechnology of Bacillus thuringiensis and Its Impact to the Environment. Nov. 4-8. Chiang Mai, Thailand: 265-283. Laemmli U.K., 1970. Cleavage of structural protein during the assembly of the head of bacteriophage T4. Nature (London) 277: 680-685. Nguyen T.H. Tram, Shin-ichro Asano and Toshihiko Iizuka. 2000. Identification and analysis of mosquitocidal toxin gene of a novel Bacillus thuringiensis strain. In Fellowship Research Reports in 1999 of The Matsumae International Foundation: 159-173.