Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis from soils in Spain

ARTICLE IN PRESS Microbiological Research 159 (2004) 59–71

www.elsevier-deutschland.de/micres

Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis from soils in Spain ! E. Quesada-Moraga, E. Garc!ıa-Tovar, P. Valverde-Garc!ıa, ! C. Santiago-Alvarez* Departamento de Ciencias y Recursos Agr!ıcolas y Forestales, Ca! tedra de Entomolog!ıa Agr!ıcola, ETSIAM, ! ! Universidad de Cordoba, Apartado 3048, Cordoba 14080, Spain Accepted 16 January 2004

KEYWORDS Bacillus thuringiensis; Insecticidal activity; Geographical diversity; Cockroach; Locust

Abstract Bacillus thuringiensis is a spore-forming bacterium showing the unusual ability to produce endogenous crystals during sporulation that are toxic for some pest insects. This work was performed to study the composition, ecological distribution and insecticidal activity of isolates of this entomopathogenic bacterium from the Spanish territory. Using a standard isolation method, B. thuringiensis was isolated from 115 out of 493 soil samples collected in the Iberian Peninsula and the Canary and Balearic Archipelagos. The percentages of samples with B. thuringiensis were 31.7, 27.6 and 18.5 and the B. thuringiensis index 0.065, 0.067 and 0.11 for the Iberian Peninsula, Canary and Balearic Archipelagos, respectively. The prairies were shown to be the worst source of B. thuringiensis while forests, urban and agricultural habitats showed similar percentages. Strain classification based on H-antigen agglutination showed a great diversity among the Spanish isolates, which were distributed among 24 subspecies, including three new ones andaluciensis, asturiensis and palmanyolensis. We differentiated 65 different protein profiles of spore–crystal mixtures by sulfatepolyacrylamide gel electrophoresis and we selected 109 isolates representative of these profiles to evaluate their insecticidal activity against insects from the Orders Orthoptera, Dictyoptera, Coleoptera, Lepidoptera and Diptera. We found variable percentages of isolates active against Coleoptera and Lepidoptera, one isolate highly active against mosquito larvae and for the first time, three isolates active against cockroaches and locusts. & 2004 Elsevier GmbH. All rights reserved.

*Corresponding author. ! E-mail address: [email protected] (C. Santiago-Alvarez). 0944-5013/$ - see front matter & 2004 Elsevier GmbH. All rights reserved. doi:10.1016/j.micres.2004.01.011

ARTICLE IN PRESS 60

Introduction Bacillus thuringiensis (Berliner) is a gram-positive spore-forming bacterium which has the unusual property of producing a parasporal protein crystal (delta-endotoxin) which is toxic (Cry proteins) for some pest insects (Lecadet et al., 1999). Because of the insecticidal activity, which represents an alternative strategy to chemical insecticides, a number of strains has been isolated and used to control pest of agricultural and medical importance. To this end, there has been much effort in many countries to isolate new strains with increased potency against target pest insects and a wider host range. A simple and reliable tool for classifying B. thuringiensis strains is the H serotyping, a phenotypic method, which allows the differentiation of these strains into serotypes and subspecies on the basis of flagellar antigens (De Barjac and Bonnefoi, 1962; De Barjac and Franchon, 1990). At least 69 serotypes and 80 subspecies are numbered and registered at the International Entomopathogenic Bacillus centre (IEBC) collection at Institute Pasteur, Paris, France (Lecadet et al., 1999). Although there is some correlation between subspecies and pathogenicity, at the moment it is known that strains of different subspecies may be toxic to insect of various orders, and even that strains from the same subspecies show differences in toxicity. The toxicity spectrum of B. thuringiensis subspecies is determined by the different delta-endotoxin genes (cry genes) carried by their strains, and by the encoded proteins (Cry proteins) (Ho. fte and Whiteley, 1989); thus assessment of Cry proteins is a good basis to study insecticidal activity of B. thuringiensis and is an important component of studying B. thuringiensis resources. It has been suggested by some authors that even if B. thuringiensis is a ubiquitous distributed organism, the normal habitat of this bacterium is in the soil, and many efforts for isolating novel B. thuringiensis from soils of several countries from the five continents has been performed (Dulmage and Aizawa, 1982; Martin and Travers, 1989; Zhang et al., 2000; Uribe et al., 2003). Today, several tens of thousands of isolates, probably more than 50,000 are distributed among various private and public collections, and are considered to be potential ‘‘reservoirs’’ for novel insecticidal toxins (Sanchis et al., 1996; Lecadet et al., 1999). To be able to estimate the risk of releasing any microorganism into the environment it is important to understand the way that it interacts with its surroundings and the other biota. In the case of B. thuringiensis, there has been extensive study of its toxicology and

E. Quesada-Moraga et al.

safety, but there has been only limited research on its role in the environment. This is the case for Spain, where surveying new B. thuringiensis strains and their Cry proteins has been reported (Aldebis et al., 1994a, b; Iriarte et al., 1998), indicating a rich diversity of serotypes (H-serotypes) and insecticidal activities. But the ecological distribution of this bacterium in the soil remained unexplored. Spain has a heterogeneous territory that contains two Archipelagos, the Canary Archipelago in the Atlantic Ocean, and the Balearic Archipelago in the Mediterranean Sea, which gives unique geographical features and abundant biological resources to ecologically study the distribution of this organism in the soil. The objective of this work was to study the composition and ecological distribution of serotypes and Cry proteins in B. thuringiensis isolates from the Spanish territory using SDS-PAGE together with bioassays of insecticidal activity against insect species of different orders.

Materials and methods Sample collection and isolation Samples were collected from different geographical sites distributed through 41 Spanish provinces (out of a total of 50 provinces) from the Iberian Peninsula, the Canary and the Balearic Archipelagos (Fig. 1). They were taken by scraping off surface material with a sterile spatula and then obtaining a 10 g sample 2–5 cm below the surface. These samples were stored in sterile plastic bags at ambient temperature. B. thuringiensis was isolated according to the method of Ohba et al. (1987). One gram of soil sample was suspended in 9 ml of sterile distilled water and shaken for 5 min. Ten-fold serial dilutions with sterile distilled water of the heattreated suspension were plated on Nutrient Agar (Oxoid), pH 7.2. Microscopic observation and subsequent isolation of crystalliferous spore-forming Bacillus colonies were made after incubation at 281C for 4–7 d. Morphological features of vegetative cells, spores, and parasporal inclusions were examined under a phase-contrast microscope. After identification of parasporal inclusions the colonies were inoculated onto nutrient agar and stored as stock cultures for serological and insecticidal activity test.

Strain classification B. thuringiensis strain classification based on Hantigen agglutination (De Barjac, 1981; De Barjac

ARTICLE IN PRESS Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis

61

Figure 1. Distribution of soil samples through the Spanish territory (Iberian Peninsula and the Canary and Balearic Archipelagos).

and Franchon, 1990) was carried out in Unite! de ! Bacteries Entomopathogenes, WHO Collaborating Centre, Institut Pasteur, Paris.

Preparation of B: thuringiensis spore–crystal mixtures Sporulating cultures of B. thuringiensis strains were produced in the standard B. thuringiensis medium (UG) containing Bactopeptone (7 g l 1), glucose and salts as previously reported (Lecadet and Dedonder, 1971), at 28–301C for 72 h, or until more than 90% of the cells had lysed, releasing spores and crystals. Spore crystal mixtures were washed (10,000g for 10 min at 41C) once in 0.5 M NaCl and then twice in cold sterile water. Two millilitres samples of lysed cultures were washed by centrifugation and resuspended once in 2 ml of 0.5 M NaCl, and then twice in cold sterilised water containing 1 mM of the protease inhibitor phenyl-methane-sulfonylfluoride (PMSF).

SDS-PAGE analysis The protein content of spore–crystal mixtures was determined by SDS-PAGE analysis, as described by Laemmli (1970), using 10% or 12.5% acrylamide separating gels. Samples (5–15 mg) of washed spore–crystal mixtures, prepared as described by Thomas and Ellar (1983), were placed in 2  concentrated sample buffer and heated at 801C for 10 min, as previously described (Lecadet et al., 1992) and loaded onto the gel immediately

before electrophoresis. Gels were stained in a solution containing 50% (v/v) ethanol, 10% (v/v) acetic acid and 0.1% (w/v) Coomasie brilliant blue R250 for 40 min, and then destained in a solution containing 6.75% (v/v) glacial acetic acid and 9.45% (v/v) ethanol.

Toxicity assays We use for the toxicity assays insect species belonging to the orders Lepidoptera (Spodoptera . littoralis (Boisduval), Spodoptera exigua (Hubner), . Agrotis segetum (Denis & Schiffermuller) and Ocnogyna baetica (Rambur)), Coleoptera (Tenebrio molitor (L.)), Diptera (Culex pipiens L.), Orthoptera (Dociostaurus maroccanus (Thunberg)) and Dictyoptera (Blatella germanica (L.). Sporulating cultures of B. thuringiensis strains were produced in the standard B. thuringiensis medium (UG) containing Bactopeptone (7 g l 1), glucose and salts as previously reported (Lecadet and Dedonder, 1971), at 28–301C for 72 h, or until more than 90% of the cells had lysed, releasing spores and crystals. Spore crystal mixtures were washed (10,000g for 10 min at 41C) twice in cold sterile water. The resulting pellet was resuspended in sterile water and was supplemented with 0.1% of the wetting agent (AGRALs). As a control, insect were treated with water plus the same concentration of AGRALs. The lepidopteran species S. littoralis, S. exigua, A. segetum were obtained from a stock colony maintained at insectary conditions (26721C,

ARTICLE IN PRESS 62

6075% RH and 16 h day length) on an artificial diet (Santiago-Alvarez and Vargas-Osuna, 1988), while the O. baetica larvae were taken from field populations of the species in the province of Co! rdoba (Spain). All the tested lepidopteran species were treated according to the method of Santiago-Alvarez and Vargas-Osuna (1988). Newly moulted third instar larvae were placed singly in plastic cups (30 mm diameter) and fed on alfalfa leaf disk (5 mm diameter). Each disc was treated by means of a microapplicator with a 3 ml aliquot containing the B. thuringiensis spore–crystal suspension supplemented with 0.1% of the wetting agent (AGRALs). For each strain and for control, 90 larvae in three replicates (3  30) were used. The assay was conducted at insectary conditions. The mortality of larvae was recorded daily until pupation. The larvae of T. molitor were taken from a commercial stock. Batches of 10 newly moulted third and four instar larvae were placed into Petri dishes and were fed on 0.3 g of an artificial sterilised diet (46.5% white flour, 46.5% white cornmeal and 7% dried yeast) supplemented with 300 ml of the B. thuringiensis stock suspensions. As control, 0.3 g of the artificial diet were supplemented with 300 ml of sterile distilled water. We used four replicates for each B. thuringiensis strain and for the controls. Mortality was recorded until pupation. C. pipiens were obtained from a laboratory population originally started with females collected in the field and reared in accordance with the method described by Singh and Moore (1985). Batches of 10 mosquito larvae in glass assay containers (120 mm deep, 38 mm diameter) holding 50 ml of water were added 200 ml of the B. thuringiensis stock suspensions. Four replications (containers) per strain and control were used. The mosquito larvae were incubated at 25–271C for 24 h as recommended by Lacey (1997) and then mortality was recorded. The D. maroccanus nymphs used in this assay come from a stock colony maintained under controlled conditions L13:D11 photoperiod, 26741C T and 40–60% RH. Wooden cages (50  50  50 cm3) were used to maintain populations of nymphs, with a 60 W bulb inside that supplied extra heat during the light period. The locusts were fed with dry wheat bran and wheat (Triticum sp.) seedlings (Quesada-Moraga and Santiago-Alvarez, 2000). Newly moulted IV instar nymphs were placed in individual plastic cups (40 mm +; 20 mm depth) and were fed with wheat leaf pieces (1.5–2 cm) immersed in the B. thuringiensis stock suspension, or in water for the

E. Quesada-Moraga et al. controls, both supplemented with 0.1% of AGRALs. Eighty nymphs, four replicates of 20 nymphs (4  20) were used for the treated and control insects. Nymphs that had not tasted the treated leaf piece after 24 h were discarded; the others were transferred to wooden cages (30  30  30 cm3) with a 60 W bulb. The bioassay was conducted at the same conditions as the stock colony. Mortality was recorded every 24 h until moult to fifth instar, which takes approximately 6–9 days. Laboratory cultures of the German cockroach (B. germanica) were reared as previously described (Singh and Moore, 1985). Batches of 10 III and IV instar nymphs were placed into 0.7 l jars covered with filter paper and containing a source of water and slanted pieces of papers as harbourage. Approximately 2 g of mouse pellets per batch were immersed in the stock B. thuringiensis suspensions. As controls, the mouse pellets were immersed in water plus 0.1% AGRALs. Three replicates were made at each strain including the control. The toxicity of the B. thuringiensis strains was evaluated as percentage of dead nymphs during 30 d of observations.

Results We collected samples that represent a variety of soils; some samples were from agricultural soils, while others were from fruit orchards, urban soils, forest soils and prairies. A total of 115 out of 493 samples collected from the Iberian Peninsula and the Canary and Balearic Archipelagos, yielded B. thuringiensis (Table 1). Of these, the higher percentage of samples with B. thuringiensis was observed in the Canary Archipelago (31.7), very similar to the one observed in the Balearic Archipelago (27.6), while the lower was found in the Iberian Peninsula (18.5) (Table 1). The percentage of samples with B. thuringiensis did not show significant differences between sample habitats, except the case of the prairies, which were the poorest in B. thuringiensis. In total, we identified 273 B thuringiensis colonies upon examining 3716 Bacillus like-colonies. The B. thuringiensis index was variable depending on the territory where the samples were taken, the highest value being found in the samples from the Balearic Archipelago (an index of 0.11), while the Canary Archipelago and the Iberian Peninsula showed similar values, 0.067 and 0.065, respectively (Table 1). In order to study the phenotype of our isolates, crystalliferous sporeformers were identified by H

ARTICLE IN PRESS Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis

63

Table 1. Distribution of B. thuringiensis (Bt) isolates in samples collected from Spain Geographical area and sample habitat

Bt indexa

Number of samples

Number of colonies

Examined

With Bt (%)

Examined

With Bt

Peninsula Agricultural Fruit orchards Forest Urban Prairies Total

135 37 68 15 31 286

29 (21.5) 6 (16.2) 12 (17.6) 3 (20.0) 3 (9.6) 53 (18.5)

1583

103

0.065

Canary Archipelago Agricultural Fruit orchards Forest Urban Total

51 6 24 39 120

18 (35.3) 2 (33.3) 8 (33.3) 10 (25.6) 38 (31.7)

1473

99

0.067

Balearic Archipelago Agricultural Fruit orchards Forest Urban Total

27 6 24 9 66

5 (18.5) 2 (33.3) 8 (33.3) 3 (33.3) 18 (27.3)

660

71

0.110

Total Agricultural Fruit orchards Forest Urban Prairies Total

213 49 116 63 31 472

52 (24.4) 10 (20.4) 28 (24.1) 16 (25.4) 3 (9.6) 109 (23.10)

3716

273

0.073

a

Bt index was calculated as: no. of colonies divided by the total number of bacterial colonies examined.

agglutination test and 259 isolates appeared to belong to 24 subspecies, including three new ones, andaluciensis, asturiensis and palmanyolensis (new serotypes H37, H53 and H55, respectively) (Table 2). Among them, subspecies aizawai (47.5%), darmstadiensis (12.4%), mexicanensis (10.8%) kim (9.3%), and konkukian (5.4%) were predominantly isolated (Table 2). There were three subspecies isolated from the three geographical areas, darmstadiensis, mexicanensis and konkukian, seven from two areas, kurstaki, sotto, aizawai, tochigiensis, neoleonensis, sooncheon, andaluciensis, and nine subspecies collected from only one area galleriae, entomocidus, morrisoni, nigeriensis, israelensis, shandongiensis, kim and palmanyolensis (Fig. 2). In most cases, isolates from a given soil sample belonged to the same subspecies. However, we recovered B. thuringiensis of two or three subspecies from 12 single samples, being subspecies mexicanensis the most frequently detected in these combinations (41.6%).

For the study of the genotype, the protein profiles of spore–crystal mixtures were resolved by SDS-PAGE. We found diversity of protein profiles among isolates of each subspecies, regardless their origin (Table 3). The higher protein profile diversity was directly related to the number of isolates per subspecies, although the 24 isolates of subspecies kim were similar. Of the 65 different protein profiles that we have found in this work, we represent in Table 3 the most frequently isolated, being detected in two or more subspecies. It can be seen that all the polypeptides were distributed in the three territories in strains from the same or different subspecies as observed for the 160 and 80 kDa polypeptides in the Iberian Peninsula, the 160 and 90–100 kDa polypeptides in the Canary Archipelago, and the 150, 130–140 and 80 kDa ones in the Balearic Archipelago. Within them, there are usual protein profiles as the ‘‘lepidopteran-like ones’’, 130–140; 66, 130–140 kDa; and the ‘‘coleopteran-like’’ ones, 69–75 kDa, while there are

ARTICLE IN PRESS 64

E. Quesada-Moraga et al.

Table 2. Distribution of B. thuringiensis subspecies in Spanish soils Subspeciesa

kurstaki sotto galleriae entomocidus aizawai morrisoni nigeriensis tolworthi darmstadiensis pakistani israelensis * indiana kumamotoensis * tochigiensis shandongiensis * neoleonensis mexicanensis konkukian * seoulensis N andaluciensis sooncheon * kim N asturiensis N palmanyolensis Total

Number of isolates per sampled area

Total

Peninsula

Canary Archipelago

Balearic Archipelago

n

%

0 1 0 0 53 0 0 1 12 1 0 0 1 1 0 2 22 8 0 3 2 0 1 0 108

1 2 0 0 70 0 0 0 4 0 0 1 0 4 0 0 5 1 1 1 0 0 0 0 90

1 0 1 1 0 1 1 0 16 0 1 0 0 0 2 3 1 5 0 0 3 24 0 1 61

2 3 1 1 123 1 1 1 32 1 1 1 1 5 2 5 28 14 1 4 5 24 1 1 259

0.8 1.2 0.4 0.4 47.5 0.4 0.4 0.4 12.4 0.4 0.4 0.4 0.4 1.9 0.8 1.9 10.8 5.4 0.4 1.5 1.9 9.3 0.4 0.4

N a

New subspecies. Asterisks indicate subspecies that were not previously described in Spain.

also some unusual protein profiles as those including bands of 160, 150, 120, 90–100 and 40–45 kDa (Fig. 3). As it can be observed in Table 4, there were protein profiles of subspecies sotto, aizawai, darmstadiensis, tochigiensis, mexicanensis, konkukian and andaluciensis either common to different geographical areas or specific to each one (Fig. 4 and Table 4). A protein-profile of subspecies konkukian showing a single polypeptide of 90– 100 kDa was found to be common to the three geographical regions. Isolates of subspecies neoleonensis and sooncheon from different geographical origin did not show common protein profiles (Fig. 4). The isolates of each subspecies were also classified in accordance with the number of bands of the protein profile (Table 5). Most isolates of the subspecies aizawai produced only one band (43.1%), followed by three bands (25.2%), more than three bands (17.9%) and two bands (13.8%). Of the single banded isolates, most produced a polypeptide of 130–140 kDa (94.3%), while in the double banded isolates, the same polypeptide

appeared with a component of 120 kDa (70.6%) or a component of 80 kDa (29.4%). The rest of isolates of this subspecies showed a great diversity of electrophoretic protein patterns with three or more bands. Isolates of subspecies darmstadiensis produced a single band of 130–140 kDa (70.8%), 150–160 kDa (25%) or 70–80 kDa (4.2%). Other isolates of this subspecies showed three or more bands, in most cases showing typical ‘‘lepidopteran-like’’ protein profiles, with polypeptides of 130–140 and 60– 70 kDa. The higher protein profile diversity was found within the subspecies mexicanensis while the lower one was found within the subspecies kim, in which all isolates showed the same protein-profile. Given the high number of B. thuringiensis isolates obtained, we selected isolates representative of the different protein profiles of the spore– crystal mixtures for evaluating the insecticidal activity against different insect orders. In a first trial, we selected 55 strains of the subspecies aizawai representative of the different protein profiles for the toxicity assays against the lepidopteran S. littoralis. Most of these strains (67%)

ARTICLE IN PRESS Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis

65

PENINSULA

tolworthi pakistani kumamotoensis asturiensis

sotto aizawai tochigiensis andaluciensis

neoleonensis sooncheon darmstadiensis mexicanensis konkukian

indiana seoulensis kurstaki

CANARY ARCHIPELAGO

galleriae entomocidus morrisoni nigeriensis israelensis shandongiensis kim palmanyolensis

BALEARIC ARCHIPELAGO

Figure 2. Distribution of B. thuringiensis subspecies in the three geographical areas of study. Intersection between circles indicates subspecies distributed in two or three areas.

showed insecticidal activity higher than 90%, 29% of them showed moderate–high mortality 75–90%, while two of isolates were not active. SDS-PAGE analysis revealed that the spore–crystal mixtures produced by these isolates of the subspecies aizawai toxic for S. littoralis were composed of polypeptides of 130–140 kDa with or without a component of 60–70 kDa. The two non-active isolates presented protein profiles composed of single polypeptides of 137 and 158 kDa, respectively. In a second trial, 30 strains from 15 subspecies representative of the different protein profiles of the spore–crystal mixtures were selected for the toxicity assays against four lepidopterans, S. littoralis, S. exigua, A. segetum and 26 strains against O. baetica. As it can be seen in Table 6, there were isolates with different degrees of activity, but curiously, the best three isolates, which belong to the subspecies morrisoni, indiana and mexicanensis, were active against the four lepidopterans. These isolates produced a polypeptide of 130–140 kDa either single (morrisoni) or in addition to one polypeptide of 110 kDa (indiana) or 60–70 kDa (mexicanensis). In a third trial, 25 strains from 15 subspecies, with unusual crystal proteins were selected for the toxicity assays against the coleopteran, T. molitor,

the dipteran C. pipiens, the orthopteran, D. maroccanus, and the dictyopteran, B. germanica. The isolates active for T. molitor belonged to subspecies andaluciensis, aizawai, tochigiensis, kim, palmanyolensis and sooncheon. Of these, only the isolates of subspecies kim and palmanyolensis showed usual ‘‘tenebrionis type’’ protein profiles, while the isolates of the other subspecies were composed of polypeptides that are not usually toxic against coleopterans. Only one isolate of the subspecies aizawai, which showed a typical ‘‘israelensis type’’ protein profile, was highly active against C. pipiens. Three isolates belonging to the subspecies sooncheon, asturiensis and kim were moderate toxic against the orthopteran D. maroccanus and the dictyopteran B. germanica. The protein profiles of these isolates were composed of a single polypeptide of 80 kDa (subspecies kim), of two polypeptides of 130–140 and 60–70 kDa (subspecies soocheon) and of three polypeptides of 203, 130–140 and 45 kDa (subspecies asturiensis).

Discussion Our results indicate that the Spanish soils are very rich in diversity of B. thuringiensis subspecies and

ARTICLE IN PRESS 66

E. Quesada-Moraga et al.

Table 3. Distribution of subspecies in accordance with the territory and the protein-profile of the crystal–spore mixtures resolved by SDS-PAGE Composition and frequency of B. thuringiensis protein profiles

Distribution

Composition (kDa)

Peninsula

160

Frequency (%) 1.9

150 130–140

1.5 25.9

90–100

3.5

80

10.4

aizawai aizawai darmstadiensis konkukian tochigiensis entomocidus

69–75

3.1

40–45

3.5

60–70, 130–140

1.9

80, 130–140

2.7

sotto kumamotoensis mexicanensis andaluciensis aizawai darmstadiensis aizawai

120, 130–140 90–100, 160

5.0 4.2

aizawai konkukian

30–40, 60, 100, 150

5.4

mexicanensis andaluciensis

Othersa a

Canary Archipelago

Balearic Archipelago

aizawai

darmstadisnesis sooncheon darmstadiensis morrisoni darmstadiensis konkukian

aizawai aizawai konkukian sotto tochigiensis seoulensis

sotto darmstadiensis tochigiensis

galleriae entomocidus kim palmanyolensis darmstadiensis neoleonensis neoleonensis sooncheon

aizawai kurstaki aizawai indiana

darmstadiensis nigeriensis shandongiensis konkukian

aizawai mexicanensis

30.8

Protein profiles present only in single subspecies.

Cry protein profiles. This founding is not exclusive of the soils but also of dust from stored products and of the olive oil storage facilities and the olive mills from various parts of Spain (Bel et al., 1997; Iriarte et al., 1998). We have found variable percentages of samples with B. thuringiensis depending on their origin, 31.7% and 27.6% in the Canary and Balearic Archipelagos, respectively, which are lower than the general percentages obtained from samples of Asia and Central and South Africa (94%), Europe (84%), USA (60%) and New Zealand (56%) (Meadows, 1993). Nevertheless, these percentages were obtained not only from soil samples but also from samples from other sources as insects, silos ands mills, which may be more successful sources of B. thuringiensis than soil samples (Chaufaux et al., 1997; Zhang et al., 2000). We have found no correlation between the percentage of soil with B. thuringiensis and the B. thuringiensis index. In fact, the higher B. thuringiensis index was found in the Balearic Archipe-

lago, 0.1, and the lower ones in the Canary Archipelago and the Iberian Peninsula. But the average B. thuringiensis positive colonies per positive sample is in fact higher in the Balearic Archipelago (3.9) than in the Canary Archipelago (2.6) and the Iberian Peninsula (1.9); thus difference in this way seems not to be very important. In general, these values were similar to those reported by other authors (De Lucca et al., 1981; Martin and Travers, 1989). Interestingly, the higher percentage of B. thuringiensis has been found in the Canary Archipelago, which presents a clear volcanic origin, in agreement with the findings of Iriarte et al. (1998) who also found the higher percentage of B. thuringiensis in volcanic soils. In contrast, it is difficult to explain the difference between the percentages of samples with B. thuringiensis in the Iberian Peninsula, 18.5%, and the Balearic Archipelago, 27.6%, since under a geological point of view, this Archipelago was part of the Iberian Peninsula millions of years ago.

ARTICLE IN PRESS Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis

67

Figure 3. SDS-PAGE analysis of protein profiles of crystal proteins among some B. thuringiensis isolates showing unusual crystal proteins. M: molecular weight markers. Lane 1: B. thuringiensis (Bt) ser. asturiensis (H53); Lane 2: Bt ser. konkukian (H34); Lane 3: Bt ser. sooncheon (H41); Lane 4: Bt ser. kumamotoensis (H18a18b); Lane 5: Bt. ser. kim (H52); Lane 6: Bt. ser. nigeriensis (H8b8d).

Furthermore, of the subspecies that have been detected only in one of the three territories, the higher number, nine subspecies, were isolated from the Balearic Archipelago. These results could be explained as a combination of a local evolution to some extent related to the distribution of the specific Balearic entomofauna, and a movement across the world by such factors as international trade (Meadows, 1993). In addition, the tourism, the main economical activity of this Archipelago, could give B. thuringiensis important physical possibilities to spread from another country to the Balearic land. The H-serotyping of the strains revealed that B. thuringiensis flora in Spanish soils is heterogeneous consisting of at least 24 subspecies out of a total of 80 known subspecies (Lecadet et al., 1999). We have found three new subspecies in the Spanish territory, andaluciensis (serovar H37), asturiensis (serovar H66) and palmanyolensis (serovar H55), which together with subspecies balearica (serovar H48), navarrensis (serovar H50) and pirenaica (serovar H57), also isolated in Spain, and with subspecies azorensis (serovar H64), graciosensis (serovar H66) and vazensis (serovar H67), isolated in the Azores Archipelago, makes Spain and Azores main sources of new subspecies in the last years

(Lecadet et al., 1999). Of particular interest is that we have predominantly isolated subspecies aizawai, darmstadiensis, mexicanensis, kim and konkukian. This result could be expected as subspecies aizawai has been obtained from soils of at least 24 countries from the five continents, being one of most widely distributed in the world (Anonymous, 1998). Subspecies darmstadiensis is also commonly isolated from the world soils, at least 16 countries, although it has not been found in Australia and Africa (Anonymous, 1998). In contrast, to our knowledge, we have described for the first time subspecies mexicanensis in Spain and Europe; there were isolates of this subspecies from Mexico, Korea, Japan, China and Thailand (Anonymous, 1998). Our results indicate that Spanish soils were very rich in subspecies mexicanensis as indicated by the fact that it has been obtained from the three geographical areas of our study, and even by the fact that we have detected it together with some others subspecies in single soil samples. The environmental adaptation of this subspecies to Spain has been confirmed by other works, which have obtained it from dust and olive oil storage facilities (Bel et al., 1997) and from dead insects (Aldebis et al., 1994b). Interestingly, subspecies kim representing 9.3% of the samples and being

ARTICLE IN PRESS 68

E. Quesada-Moraga et al.

Table 4. Number of different SDS-PAGE protein profiles of spore–crystal mixtures within each subspecies and its distribution Subspecies

kurstaki sotto galleriae entomocidus aizawai morrisoni nigeriensis tolworthi darmstadiensis pakistani israelensis n indiana kumamotoensis n tochigiensis shandongiensis n neoleonensis n mexicanensis konkukian n seoulensis andaluciensis sooncheon n kim asturiensis palmanyolensis

Number of different profilesa Total

Commonb

Peninsula

Canary Archipelago

Balearic Archipelago

2 2 1 1 16 1 1 1 9 1 1 1 1 3 1 3 6 2 1 3 5 1 1 1

0 1 (PC) F F 5 (PC) F F F 1 (PB), 1(CB), 1(PC) F F F F 1 (PC) F 0 1(PC) 1(PCB), 1(PB) F 1 (PC) 0 F F F

NP 0 NP NP 5 NP NP 1 2 1 NP NP 1 0 NP 1 4 0 NP 2 2 NP 1 NP

1 1 NP NP 6 NP NP NP 1 NP NP 1 NP 2 NP NP 0 0 1 0 NP NP NP NP

1 NP 1 1 NP 1 1 NP 3 NP 1 NP NP NP 1 2 1 0 NP NP 3 1 NP 1

a

NP, not present. Common profiles are indicated when subspecies are present in more than one region. PCB (three regions), Peninsula and Balearic Archipelago (PB), Peninsula and Canary Archipelago (PC) and Canary and Balearic Archipelago (CB). n Subspecies that were not previously described in Spain. b

Figure 4. Variation of protein profiles of spore–crystal mixtures among strains of B. thuringiensis of the same subspecies isolated either from the same or different geographical regions. Lanes 1–2: Bt ser. andaluciensis (H37) from the same geographical region; Lanes 3–4: Bt ser. neoleonensis (H24a24b) isolated from the Iberian Peninsula (3) and from the Balearic Archipelago (4); Lanes 5–7: Bt ser. darmstadiensis (H10a10b) isolated from the Iberian Peninsula (5 and 6) and from the Canary Archipelago (7).

ARTICLE IN PRESS Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis

69

Table 5. Distribution of the total number (%) of isolates of each subspecies following the number of bands of the SDSPAGE protein profile Subspecies

Number of bands of the SDS-PAGE protein profile

kurstaki sotto galleriae entomocidus aizawai morrisoni nigeriensis tolworthi darmstadiensis pakistani israelensis indiana kumamotoensis tochigiensis shandongiensis neoleonensis mexicanensis konkukian seoulensis andaluciensis sooncheon kim asturiensis palmanyolensis Total

1

2

3

43

0 (0.0) 3 (0.0) 1 (100.0) 0 (0.0) 53(43.1) 1 (100.0) 0 (0.0) 0 (0.0) 24 (75.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (100.0) 4 (80.0) 0 (0.0) 3 (60.0) 5 (17.8) 1 (6.6) 1 (100.0) 0 (0.0) 2 (20.0) 24 (100.0) 0 (0.0) 1 (100.0) 124 (47.7)

0 (0.0) 0 (0.0) 0 (0.0) 1 (100.0) 17(13.8) 0 (0.0) 1 (100.0) 1 (100.0) 3 (9.4) 0 (0.0) 1 (100.0) 1 (100.0) 0 (0.0) 1 (20.0) 0 (0.0) 0 (0.0) 1 (3.6) 6 (40.0) 0 (0.0) 1 (25.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 34 (13.1)

2 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 31(25.2) 0 (0.0) 0 (0.0) 0 (0.0) 5 (15.6) 1 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 2 (100.0) 0 (0.0) 5 (17.9) 8 (53.4) 0 (0.0) 2 (50.0) 3 (60.0) 0 (0.0) 1 (100.0) 0 (0.0) 60 (23.1)

0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 22(17.9) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 2 (40.0) 17 (60.7) 0 (0.0) 0 (0.0) 1 (25.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 42 (16.2)

Table 6. Distribution of toxicity of B. thuringiensis isolates to some insect pest Target insect

Number of isolatesa A

B

C

D

Non-toxic

Total

Lepidoptera S. littoralis S. exigua A. segetum O. baetica

19 10 20 18

0 0 5 2

6 0 3 0

49 3 0 3

8 17 2 3

82 30 30 26

Coleoptera T. molitor

6

0

0

0

19

25

Diptera C. pipiens

0

0

0

1

24

25

Orthoptera D. maroccanus

0

0

2

1

22

25

Dictyoptera B. germanica

0

0

2

1

22

25

a

Isolates with insecticidal activity between: (A) 0% and 25%, (B) 25% and 50%, (C) 50% and 75% and (D) 75% and 100%.

isolated only from the Balearic archipelago. This is the first report of subspecies kim in Spain and the second in Europe; it has been isolated from soils of Korea and China and from leaves of coniferous trees of Denmark (Anonymous, 1998). Is the presence of this subspecies in the Balearic archipelago a clear example of the above world movements? We have isolated subspecies konkukian from 5.4% of the soil samples and from the three geographical areas, which it is in agreement with other works, which indicated that this subspecies is also widely distributed in Spain (Iriarte et al., 1998). The bacterial formulations registered as bioinsecticides in Spain (Santiago-Alvarez and QuesadaMoraga, 2001) are based mainly (around 80%) on subspecies kurstaki, of which we have isolated only two strains from the Canary and the Balearic Archipelago, respectively. Although these strains could represent residues from commercial formulations, their SDS-PAGE protein profiles showed crystal mixtures different from the commercial formulations, which clearly indicates that they were not commercial residues. This result is

ARTICLE IN PRESS 70

supported by those obtained with subspecies israelensis and aizawai, which represent the rest (around 20% and 10%, respectively) of the B. thuringiensis based commercial formulations registered in Spain. We have isolated only one strain of the subspecies israelensis in a non-cultivated area in the Balearic Archipelago while Iriarte et al. (1998) did not isolate this subspecies from the soil. The commercial use of subspecies israelensis is more or less restricted to pest of human importance, as mosquito species. Since this type of pest is not common in Spain, it is possible to conclude that our strain does not represent a residue from a commercial formulation. Concerning subspecies aizawai, it is the base of very few commercial formulations in Spain, whereas it was the predominantly isolated in our work. Furthermore, some of our strains showed different protein profiles from the commercial ones, which again indicates that they did not represent residues of the commercial formulation. An important feature of the distribution of the protein profiles between different geographical areas was that the main insecticidal crystal proteins, which are normally considered to range between 27 and 140 kDa, were distributed in the three territories with no exceptions. That is the 80 kDa crystal protein was found in protein profiles of strains of subspecies seoulensis in the Canary Archipelago, of subspecies galleriae, kim and palmanyolensis in the Balearic one, and of subspecies aizawai in the Iberian Peninsula. It seems that the presence of the full range of insecticidal crystal proteins would gives B. thuringiensis the guarantee of killing a susceptible insect host. But some of the tested isolates were not toxic to any of the insect hosts, which could be explained by the fact that some crystals may be toxic to insects that are not normally susceptible to B. thuringiensis as locusts and cockroaches. The search for B. thuringiensis strains producing endotoxins active against Orthoptera and Dictyoptera has and is being pursued by a number of research institutes and commercial companies, but the results of such large-scale screenings if any remain unknown. Particularly, in intensive screening trials specifically directed at finding B. thuringiensis strains pathogenic to locusts, neither Zelazny et al. (1997) nor Chaufaux et al. (1997) found any isolates active against Locusta migratoria (L.) and Schistocerca gregaria (Forska( l). Our results strongly support the insecticidal activity of three strains against the Mediterranean or Moroccan locust D. maroccanus. Furthermore, one of these strains has been shown to cause strong cytopathic effects in the midgut of this locust

E. Quesada-Moraga et al. ! (Quesada-Moraga and Santiago-Alvarez, 2001) and is actually being evaluated under field conditions for its possible development as a commercial product. This work shows that the soil is a very important source of B. thuringiensis strains providing a large genetic resource for its use in the development of bioinsecticides to control insect pests that have not previously reported to be susceptible to B. thuringiensis. The ecological aspects of the distribution of subspecies and protein profiles in the soil could contribute to a better understanding of the role of B. thuringiensis in the environment.

Acknowledgements This research was partially supported by the collaborative research program between the University of Cordoba and Newbiotechnic S.A.

References Aldebis, H.K., Cobo, F., Vargas-Osuna, E., Santia! go-Alvarez, C., 1994a. Analysis of the d-endotoxin of Bacillus thuringiensis spanish isolates by polyacrylamide gel electrophoresis. Proceedings of the Sixth International Colloquium on Invertebrate Pathology and Microbial Control, Montpellier, France. ! Aldebis, H.K., Vargas-Osuna, E., Santiago-Alvarez, C., ! serologica ! 1994b. Caracterizacion de cepas de Bacillus * thuringiensis Berliner aisladas de insectos espanoles. Bol. Sanid. Vegetal. Plagas 20, 765–769. Anonymous, 1998. Collection of Bacillus thuringiensis and Bacillus sphaericus (Classified by H Serotypes), Catalogue no. 1, Institut Pateur. Bel, Y., Granero, F., Alberola, T.M., Martinez-Sebastian, M.J., Ferre, J., 1997. Distribution, frequency and diversity of Bacillus thuringiensis in olive tree environments in Spain. Syst. Appl. Microbiol. 20, 652–658. Chaufaux, J., Marchal, M., Gilois, N., Jehanno, I., Buisson, C., 1997. Recherche de souches naturelles du Bacillus thuringiensis dans differents biotopes, a travers le monde. Can. J. Microbiol. 43, 337–343. De Barjac, H., 1981. Identification of H-serotypes of Bacillus thuringiensis. In: Burges, H.D. (Ed.), Microbial Control of Pest and Plant Diseases 1970–1980. Academic Press, London, pp. 35–44. De Barjac, H., Bonnefoi, A., 1962. Essai de classification biochimique et serologique de 24-souches de Bacillus du type B. thuringiensis. Entomophaga 7, 5–31. De Barjac, H., Franchon, E., 1990. Classification of Bacillus thuringiensis strains. Entomophaga 35, 233–240.

ARTICLE IN PRESS Isolation, geographical diversity and insecticidal activity of Bacillus thuringiensis

De Lucca, A.J., Simonson, J.G., Larson, A.D., 1981. Bacillus thuringiensis distribution in soils of the United States. Can. J. Microbiol. 27, 865–870. Dulmage, H.T., Aizawa, K., 1982. Distribution of Bacillus thuringiensis in nature. In: Kurstak, Z. (Ed.), Microbial and Viral Pesticides. Dekker, New York, pp. 209–327. . Hofte, H., Whiteley, H.R., 1989. Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53, 242–255. Iriarte, J., Bel, Y., Ferrandis, M.D., Andrew, R., Murillo, ! J., Caballero, P., 1998. Environmental J., Ferre, Distribution and Diversity of Bacillus thuringiensis in Spain. Syst. Appl. Microbiol. 21, 97–106. Lacey, L.A., 1997. Bacteria: laboratory bioassay of bacteria against aquatic insects with emphasis on larvae of mosquitoes and black flies. In: Lacey, L. (Ed.), Manual of Techniques in Insect Pathology. Academic Press, New York, pp. 79–90. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of basterophage T4. Nature 227, 680–685. Lecadet, M.M., Dedonder, R., 1971. Biogenesis of the crystalline inclusion of Bacillus thuringiensis during sporulation. Eur. J. Biochem. 23, 282–294. Lecadet, M.M., Chaufaux, J., Ribier, J., Lereclus, D., 1992. Construction of novel Bacillus thuringiensis strains with different insecticidal activities by transduction and transformation. Appl. Environ. Microbiol. 58, 840–849. Lecadet, M.M., Franchon, E., Cosmao Dumanoir, C., ! Ripoteau, H., Hamon, S., Laurent, P., Thiery, I., 1999. Updating the H-antigen classification of Bacillus thutingiensis. J. Appl. Microbiol. 86, 660–672. Martin, P.A.W., Travers, R.S., 1989. Worldwide abundance and distribution of Bacillus thuringiensis isolates. Appl. Environ. Microbiol. 55, 2437–2442. Meadows, M.P., 1993. Bacillus thuringiensis in the environment: ecology and risk assessment. In: Entwistle, P.F., Cory, J.S., Bailey, M.J., Higgs, S. (Eds.), Bacillus thuringiensis, An Environmental Biopesticide: Theory and Practice. Willey, New York, pp. 193–220.

71

Ohba, M., Yu, Y.M., Aizawa, K., 1987. Insect toxicity of Bacillus thuringiensis isolates from soils in Japan. J. Invertebr. Pathol. 47, 12–20. Quesada-Moraga, E., Santiago-Alvarez, C., 2000. Rearing and breeding of the mediterranean locust Dociostaurus maroccanus under laboratory conditions. J. Appl. Entomol. 125, 121–124. ! Quesada-Moraga, E., Santiago-Alvarez, C., 2001. Histopathological Effects of Bacillus thuringiensis on the midgut of the Mediterranean locust Dociostaurus maroccanus. J. Invertebr. Pathol. 78, 183–186. Sanchis, V., Chaufaux, J., Lereclus, D., 1996. Am! elioration biotechnologique de Bacillus thuringiensis: ! Les enjeux et les risques. Ann. Inst. Pasteur/Actualites 7, 271–284. Santiago-Alvarez, C., Vargas-Osuna, E., 1988. Reduction of reproductive capacity of Spodoptera littoralis males by a Nuclear Polyhedrosis Virus (NPV). J. Invertebr. Pathol. 52, 142–146. Santiago-Alvarez, C., Quesada-Moraga, E., 2001. Empleo de Bacillus thuringiensis en los sistemas agr!ıcolas. In: Caballero, P., Ferre! J, . (Eds.), Desarrollo y aplicaciones de Bacillus thuringiensis en el Control Integrado de Plagas. Phytoma Press, Valencia, pp. 191–217. Singh, P., Moore, R.F., 1985. Handbook of Insect Rearing, Vol. I. Elsevier, Amsterdam, 488pp. Thomas, W.E, Ellar, D., 1983. Mechanism of action of Bacillus thuringiensis var. israelensis insecticidal dendotoxin. FEBS Lett. 154, 362–368. Uribe, D., Mat!ınez, W., Ceron, J., 2003. Distribution and diversity of cry genes in native strains of Bacillus thuringiensis obtained from different ecosystems from Colombia. J. Invertebr. Pathol. 82, 119–127. Zhang, H.Y., Yu, Z.N., Deng, W.X., 2000. Isolation, distribution and toxicity of Bacillus thuringiensis from warehouses in China. Crop Prot. 19, 449–454. Zelazny, B., Goettel, M.S., Keller, B., 1997. The potential of bacteria for the microbial control of grasshoppers and locusts. Mem. Entomol. Soc. Can. 171, 147–156.