SEARCH

SEARCH BY CITATION

Keywords:

  • Biopesticide;
  • cry gene;
  • cyt gene;
  • δ-Endotoxin;
  • Polymerase chain reaction;
  • Bacillus thuringiensis

Abstract

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References

The polymerase chain reaction (PCR) is a molecular tool widely used to characterize the insecticidal bacterium Bacillus thuringiensis. This technique can be used to amplify specific DNA fragments and thus to determine the presence or absence of a target gene. The identification of B. thuringiensis toxin genes by PCR can partially predict the insecticidal activity of a given strain. PCR has proven to be a rapid and reliable method and it has largely substituted bioassays in preliminary classification of B. thuringiensis collections. In this work, we compare the largest B. thuringiensis PCR-based screenings, and we review the natural occurrence of cry genes among native strains. We also discuss the use of PCR for the identification of novel cry genes, as well as the potential of novel technologies for the characterization of B. thuringiensis strains.


1Introduction

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References

The entomopathogenic bacterium Bacillus thuringiensis was first isolated by the Japanese scientist S. Ishiwata, in 1901, from silkworm larvae exhibiting the sotto disease [1]. Ten years later, E. Berliner [2] formally described the species from an isolate originating from Anagasta kuehniella, collected in the German region of Thuringia, which gave the name to the species. Due to its insecticidal properties, B. thuringiensis has been used commercially in the biological control of insect pests for the last four decades. Its toxicity was first supposed to be limited to Lepidoptera, but interest increased after discovering a dipteran-active strain belonging to serovar israelensis[3] and a morrisoni strain active against Chrysomelidae (Coleoptera) which was named as subspecies tenebrionis[4]. Currently, bioinsecticides based on B. thuringiensis are used world-wide for the control of many insects within the orders Lepidoptera, Diptera and Coleoptera. Novel toxicities against Hymenoptera as well as non-insect organisms, such as mites, nematodes, protozoa, and plathelmintes have also been reported [5].

The insecticidal activity of B. thuringiensis is due mainly to its ability to synthesize, during the sporulation phase, large amounts of proteins that form a parasporal crystal. When a susceptible insect ingests these crystalline proteins, known as δ-endotoxins, they are solubilized and proteolytically digested to yield the active toxic form. Toxins specifically bind to protein receptors in the epithelial insect midgut [6,7] and produce pores, leading to the loss of normal membrane function [8]. As a result of membrane permeability, epithelial cells lyse and feeding activity is paralyzed. Finally, insects die of starvation, septicemia or a combination of both.

A δ-endotoxin can be defined as a major protein component of a parasporal crystal showing ‘significant sequence similarity to one or more toxins within the established nomenclature or as a B. thuringiensis parasporal inclusion protein that exhibits pesticide activity or some experimentally verifiable toxic effect to a target organism’ (see B. thuringiensis toxin nomenclature (N. Crickmore et al.) at http://www.biols.susx.ac.uk/Home/Neil_Crickmore/Bt/index.html). There are two types of δ-endotoxins: the highly specific Cry (from crystal) toxins which act via specific receptors and the non-specific Cyt (cytolytic) toxins, with no known receptors. Both families of toxins are classified exclusively on the basis of their amino acid sequence identity. Four ranks have been defined, and the boundaries are 45, 78 and 95%. The degree of similarity among Cry proteins is highly variable and some Cry proteins as well as the Cyt toxins do not cluster with the main Cry lineage. Some Cry proteins are even more closely related to Bacillus sphaericus mosquitocidal toxins (Mtx and Bin) than to other B. thuringiensis toxins. cry genes, often located in plasmids, encode the Cry proteins. B. thuringiensis strains can harbor several cry genes, and some isolates can contain up to eight different cry genes [9]. In general, the type of cry and cyt genes present in a strain correlates to some extent with its insecticidal activity. Thus, the identification of the gene content in a B. thuringiensis strain can be used to predict its insecticidal potential. To simplify the diversity of terms used to define each rank of the current classification, we will use two designations throughout: (i) subfamily, including all the members of the first rank of genes belonging to the cry or cyt families (cry1 subfamily, cry2 subfamily, cyt1 subfamily, etc.) and (ii) group, including all those genes belonging to the second rank of the classification (cry1C group, cyt2B group, etc.).

After many years of successful use in the field, the first cases of resistance to B. thuringiensis appeared, renewing interest in the search for novel toxins to delay or overcome this phenomenon. However, the search for novel toxin-encoding genes is difficult and time-consuming. Traditionally, it was approached by testing B. thuringiensis collections against different insect species. If a new toxic activity, in terms of host range or insecticidal potency, was detected, the gene (or genes) involved was identified, cloned, and expressed in vitro. Because this strategy is time-consuming, other procedures were developed to detect putative new genes present in B. thuringiensis strains. The advent of molecular techniques facilitated, initially, the identification of the known cry genes. DNA hybridization as well as the use of monoclonal and polyclonal antibodies were the first molecular approaches used to identify previously unreported toxins. These methods made it possible to test a greater number of strains and reduced the time and cost of the analysis. However, they are fairly insensitive.

A milestone in the analysis of B. thuringiensis collections was the advent of the polymerase chain reaction (PCR). This technique, originally proposed by Kleppe et al. [10] and independently conceived and developed by Mullis and coworkers [11] can be used to amplify specific DNA fragments and thus to determine the presence or absence of a target gene. The identification of B. thuringiensisδ-endotoxin genes by PCR has proven to be a very useful method for strain characterization and its use as a preliminary selection step offers many advantages in terms of rapidity and reproducibility. Although it is likely that only a fraction of the PCR-based identifications of cry and cyt genes have been published, the available reports provide a considerable amount of data that we have reviewed and analyzed. The most important PCR-based screenings of B. thuringiensis collections (in terms of the number of strains and genes identified) are compared in this work; the use of PCR for the identification of both known and unknown genes is described, and a comparative analysis of the geographical diversity of cry gene occurrence is shown. We also suggest the use of novel techniques obtained from recent developments in molecular analysis, to further the study of B. thuringiensis collections.

2Use of PCR for the prediction of insecticidal activity

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References

2.1Identification of δ-endotoxins by PCR

The PCR-based identification of B. thuringiensis cry genes was first developed by Carozzi et al. [12], who introduced this technique as a tool to predict insecticidal activity. These authors designed 12 oligonucleotides from cry1Ab, cry3A and cry4A genes and used them as primers in PCRs directed toward the identification of Lepidoptera-, Coleoptera-, and Diptera-active strains, respectively. Using bioassays, they determined the biological activity of 28 strains and found correspondence with the toxicity predicted on the basis of the amplification product profiles. Carozzi et al. proposed PCR as an accurate, fast methodology for the identification of novel strains and the prediction of insecticidal activity of new isolates, and they also forecast the possible use of PCR for the discovery of previously unknown cry genes. They suggested that strains yielding unusual PCR profiles could be selected for further analyses leading to the identification and characterization of hypothetical novel cry genes.

Over the last decade, PCR has been widely exploited to determine the cry gene content of many B. thuringiensis strain collections [13–18]. The screening has also been performed to identify cyt genes, although to a lesser extent [17,19–21]. The PCR screening programs have increased our knowledge concerning the natural occurrence of single cry and cyt genes among strains, their combinations within the strains and their occurrence in different geographic locations. Since PCR allows quick and simultaneous screening of many strains, it has partially substituted bioassays in preliminary characterization of B. thuringiensis collections. Today, PCR has become a routine screening step for large strain collections, in both public and private research laboratories. However, prediction of insecticidal activity by PCR must always be corroborated by bioassays, in order to assess the potential of promising isolates as biopesticides [22].

The efficacy of PCR in identifying the large family of cry genes, with amino acid identities ranging from less than 45% to more than 95%, is based on the presence of conserved regions. Most of the B. thuringiensis protoxin crystal genes share conserved nucleotide blocks, in a number that varies from five, for naturally truncated genes such as cry1I or cry3A, to eight for the largest genes with encoded proteins of 1000 residues or more [23]. The efficacy of PCR for cry gene identification relies on the alternation of conserved and variable nucleotide regions. By designing oligonucleotides to be used as primers either from conserved blocks or from variable regions, it is possible to recognize either entire gene subfamilies or specific, individual genes. An intermediate approach is the selection of regions that are highly conserved among several genes from different subfamilies but usually exhibiting the same host range [12].

The easiest strategy to identify cry or cyt genes by PCR is the use of a primer pair that specifically recognizes a single cry gene that will yield an amplicon as large as the distance separating the primer annealing sites. This can be done with a pair of specific primers or by combining a universal primer selected from a conserved block and thus able to anneal to the entire gene family, and a specific primer selected from a variable region. However, the high number of cry genes known so far makes this gene-by-gene strategy inapplicable for large-scale screening purposes. For practical reasons, primer pairs designed from highly conserved regions and recognizing entire cry gene subfamilies are often used in a preliminary screening prior to performing a second PCR with specific primers. The first step saves much time and effort by avoiding the need to test all the specific primers, as only those corresponding to the group of genes amplified in the first PCR are used in the second reaction [16].

Another strategy to expedite the screening is based on the use of a mixture of more than two primers (multiplex PCR) in the same reaction [14,24,25]. Usually, a single universal primer is combined with several specific oligonucleotides that recognize individual genes. In this situation, the PCR amplification yields as many amplified gene fragments as can be recognized by the primers, which can be easily identified on the basis of their size, as determined from their electrophoretic mobility. To date, more than 80 primer pairs specifically recognizing both entire groups (i.e. cry1 genes) and individual cry genes have been designed. Primer size varies from 17 to more than 30 nucleotides, and some ‘universal’ oligonucleotides, directed to the identification of a group or subfamily of genes, are degenerate. An updated list of primers used in the identification of cry and cyt genes is shown in Table 1.

Table 1.  Specific primer pairs directed to the identification of cry and cyt genes
  1. acry and cyt genes recognized by the specific primers shown to the right. Listing up to the secondary rank is given, except for cry1A genes, which are listed up to the tertiary rank. Primers may not always perfectly match all alleles of each gene, and thus some alleles may not be identified. Cross-amplification of related genes may also occur.

  2. bPrimer names are those given in the original report. Superfluous designations such as ‘d’ or ‘r’ have been omitted unless necessary to distinguish between forward and reverse primers. The sequences of primers listed several times are given in the first citation. Degenerate bases are designated as follows: B=C, G or T; D=A, G or T; H=A, C or T; K=G or T; M=A or C; R=A or G; and Y=T or C.

  3. cPrimers sharing the sequence of spe-cry8A(d).

  4. dPrimers sharing the sequence of spe-cry9A(r).

  5. eUnnamed cloning primers.

GeneaForward primer (5′–3′)bReverse primer (5′–3′)bSource
cry1AaIAa (TTCCCTTTATTTGGGAATGC)I(−) (MDATYTCTAKRTCTTGACTA)[24]
 SB-1 (TGCATAGAGGCTTTAAT)U815c (CAGGATTCCATTCAAGG)[26]
 TYIAA (GAGCCAAGCAGCTGGAGCAGTTTACACC)TYIUN12 (ATCACTGAGTCGCTTCGCATGTTTGACTTTCTC)[29]
 CJ1 (TTATACTTGGTTCAGGCCC)CJ2 (TTGGAGCTCTCAAGGTGTAA)[28]
cry1AbIAb (CGGATGCTCATAGAGGAGAA)I(−)[24]
 SB-2 (TCGGAAAATGTGCCCAT)U3-18c (AATTGCTTTCATAGGCT)[26]
 CJ4 (AACAACTATCTGTTCTTGAC)CJ5 (CTCTTATTATACTTACACTAC)[28]
 TY6 (GGTCGTGGCTATATCCTTCGTGTCACAGC)TY14 (GAATTGCTTTCATAGGCTCCGTC)[29]
cry1AcRB-19 (GGGACTGCAGGAGTGAT)U8-15C[26]
 IAc (GGAAACTTTCTTTTTAATGG)I(−)[24]
 CJ6 (GTTAGATTAAATAGTAGTGG)CJ7 (TGTAGCTGGTACTGTATTG)[28]
 CJ4CJ5[28]
 TYIAC (TCACTTCCCATCGACATCTACC)TYIUN12[29]
cry1AdIAd (ACCCGTACTGATCTCAACTA)I(−)[24]
 CJ1CJ2[28]
 CJ3 (CAGCCGATTTACCTTCTA)CJ2[28]
cry1AeIAe (CTCTACTTTTTATAGAAACC)I(−)[22]
cry1BCJ8 (CTTCATCACGATGGAGTAA)CJ9 (CATAATTTGGTCGTTCTGTT)[28]
 IB (GGCTACCAATACTTCTATTA)I(−)[24]
 TYIB (GTCAACCTTATGAGTCACCTGGGCTTC)TYIUN12[29]
cry1CCJ10 (AAAGATCTGGAACACCTTT)CJ11(CAAACTCTAAATCCTTTCAC)[28]
 IC (ATTTAATTTACGTGGTGTTG)I(−)[24]
 TYIC (CAACCTCTATTTGGTGCAGGTTC)TYIUN12[29]
cry1DCJ12 (CTGCAGCAAGCTATCCAA)CJ13 (ATTTGAATTGTCAAGGCCTG)[28]
 ID (CAGGCCTTGACAATTCAAAT)I(−)[24]
 TYID (GGTACATTTAGATATTCACAGCCAC)TYIUN12[29]
cry1ECJ14 (GGAACCAAGACGAACTATTGC)CJ15 (GGTTGAATGAACCCTACTCCC)[15]
 IE (TAGGGATAAATGTAGTACAG)I(−)[24]
 TYIE (CTTAGGGATAAATGTAGTACAG)TYIUN12[29]
cry1FCJ16 (TGAGGATTCTCCAGTTTCTGC)CJ17 (CGGTTACCAGCCGTATTTCG)[15]
 IF (GATTTCAGGAAGTGATTCAT)I(−)[24]
 TYIF (CCGGTGACCCATTAACATTCCAATC)TYIUN12[29]
cry1GG (GCTTCTCTCCAAACAACG)I(−)Porcar et al., unpubl.
cry1HH (ACTCTTTTCACACCAATAAC)I(-)Porcar et al., unpubl.
cry1IV(+) (ATGAAACTAAAGAATCCAGA)V(−) (AGGATCCTTGTGTTGAGATA)[22]
 I-FW (ACAATTTACAGCTTATTAAG)I-RV (CTACATGTTACGCTCAATAT)Porcar et al., unpubl.
cry1JJ (GCGCTTAATAATATTTCACC)I(−)Porcar et al., unpubl.
cry1KK (TGATATGATATTTCGTAACC)I(−)Porcar et al., unpubl.
cry2AUn2(d) (GTTATTCTTAATGCAGATGAATGGG)Un2(r) (CGGATAAAATAATCTGGGAAATAGT)[16]
 II(+) (TAAAGAAAGTGGGGAGTCTT)II(−) (AACTCCATCGTTATTTGTAG)[22]
 2-FW (CGATATGTTAGAATTTAGAAC)2-RV (TACCGTTTATAGTAACTCG)Porcar et al., unpubl.
cry3ACol1A (GTCCGCTGTATATTCAGGTG)Col1B (CACTTAATCCTGTGACGCCT)[12]
 Un3 (CGTTATCGCAGAGAGATGACATTAAC)EE-3Aa (TGGTGCCCCGTCTAAACTGAGTGT)[16]
 CJIIIcte22 (CAATCCCAGTGTTTACTTGGAC)CJIIIA23 (CCCCGTCTAAACTGAGTGT)[15]
cry3BUn3EE-3Ba (ACGAAAGATTCTGCTCCTAT)[16]
cry3CUn3EE-3C (ATTTTGGTACCTCCTGTACCCACC)[16]
 CJIIIcte22CJIIID27 (CGAAATACGAAATACTATGAG)[15]
cry4ADip2A (GGTGCTTCCTATTCTTTGGC)Dip2B (TGACCAGGTCCCTTGATTAC)[12]
 EE-4A (GGGTATGGCACTCAACCCCACTT)Un4 (GCGTGACATACCCATTTCCAGGTCC)[16]
cry4BEE-4B (GAGAACACACCTAATCAACCAACT)Un4[16]
cry5(TAAGCAAAGCGCGTAACCTC)(GCTCCCCTCGATGTCAATG)[21]
cry6VI(+) (TAYGGTTTTAAAKKTGCTGG)VI(−) (TRAATYCTATTRAACAATCCTA)[22]
 (TGGCGTAGAGGCTGTTCAAGTA)(TGTCGAGTTCATCATTAGCAGTGT)[21]
cr7AB1-7A (CATCTAGCTTTATTAAGAGATTC)B5-7A (GATAAATTCGATTGAATCTAC)[66]
 B2-7A (GCTGTATTTCCTATTTATGACCC)B5-7A[66]
 B3-7A (GGGCCTGGATTTACAGGTGG)B5-7A[66]
 B4-7A (GTTAGAGTTCGATACGCTAC)B5-7A[66]
 EE-7Aa (GCGGAGTATTACAATAGAATCTATCC)Un7,8 (CTTCTAAACCTTGACTACTT)[16]
cry8AEE-8A (GAATTTACTCTATACCTTGGCGAC)Un7,8[16]
 spe-cry8A(d) (ATGAGTCCAAATAATCTAAATG)spe-cry8A(r) (TCTCCCCATATATCTACGCTC)[17]
 CJIIIE28 (TGACAAGTACTGGATTCTGCAA)CJIIIE29 (GTTGTTGATGAGGTTCCCCTT)[15]
 B3-8A (GGTCCTGGATTTACAGGAGGAGAT)B5-8A (GATGAATTCGATTCGGTCTAT)[66]
cry8BEE-8B (GACCGCATCGGAAGTTGTGAG)Un7,8[16]
 spe-cry8B(d)cspe-cry8B(r) (GAACATCTCGTAAGGCTC)[17]
 B3-8B (GGGCGTGGTTATACAGGGGGAGAC)B5-8B (GATGAATTCGATTCGGTCTAA)[66]
cry8Cspe-cry8C(d)cspe-cry8C(r) (GGTACTCGATTGTCCAGT)[17]
 EE-8C (GGTGCTGCTAACCTTTATATTGATAG)Un7,8[16]
 B3-8C (GAAGGTCTATATAATGGAGGAC)B5-8C (AATAAATTCAATTCTATCAAT)[66]
cry9ACJ18 (ATATGGAGTGAATAGGGCG)CJ19 (TGAACGGCGATTACATGC)[15]
 spe-cry9A(d) (GTTGATACCCGAGGCACA)spe-cry9A(r) (CCGCTTCCAATAACATCTTTT)[17]
 CJ18 (ATATGGAGTGAATAGGGCG)CJ19 (TGAACGGCGATTACATGC)[15]
 IG (GGTTCTCAAAGATCCGTGTA)I(−)[24]
cry9Bspe-cry9B(d) (TCATTGGTATAAGAGTTGGTGATAGAC)spe-cry9B(r)d[17]
cry9Cspe-cry9C(d) (CTGGTCCGTTCAATCC)spe-cry9C(r)d[17]
cry9DCCGAGCTCTATGAATCGAAATAATCAAAATGAATeCCTCCTAGACACAGGGATGATTTCAATTCe[67]
cry10A10A5 (ATATGAAATATTCAATGCTC)10A3 (ATAAATTCAAGTGCCAAGTA)[68]
cry11AEE11A(d) (CCGAACCTACTATTGCGCCA)EE11A(r) (CTCCCTGCTAGGATTCCGTC)[16]
 11A5 (CCAGCATTAATAGCAGTAGC)11A3 (TGTACACATTTGAGTAAAAA)[68]
cry12(CTCCCCCAACATTCCATCC)(AATTACTTACACGTGCCATACCTG)[21]
cry13Aspe-cry13(d) (CTTTGATTATTTAGGTTTAGTTCAA)spe-cry13(r) (TTGTAGTACAGGCTTGTGATTC)[17]
cry14(ATAATGCGCGACCTACTGTTGT)(TGCCGTTATCGCCGTTATT)[21]
cry15A34C (AGATATCATGGCAATTATGAATGATAT)34D (ACCCGGGTTATTCTTTATCATAATCGC)[69]
cry16ADA5c (TCAAAAGGTGTGGCAAG)CR3c (ATAAGCCCAATATCATG)[70]
cry17ACR8 (AAGTAAAGATTTCTGGG)OX7as (CTGAGGTATTTTGTGGA)[70]
cry18(CCGAGGCGATTTGGATAGAT)(TGCCGGTGTAAACAAAGAAGG)[21]
cry19(AGGGGAGTCCAGGTTATGAGTTAC)(ATTTCCCTAGTTAGTTCGGTTTTT)[21]
cry20(CAATCCCTGGCTTCACTCGT)(CCGCGGGCATTAGGATT)[21]
cry21(ATACAGGGATAGGATTTCAAG)(ATCCCCATTTTCTATAAGTGTCT)[21]
cry22(CAGATGAGATAGATGGGGATTTGA)(ATTCGCTTCTATACTTGGCTGTC)[21]
cry24(AGGGGGCGATGGATACGAC)(GGCCCTGCTACAACCGAAACTA)[21]
cry25(CGTTTTCCGCATTATCATTAGG)(ACGCCCCGGCTGTCTTA)[21]
cry26(CGCGCTGTTCAATTATCAAGTGC)(ATATGGAAAGAAAAGGCGTGTGGA)[21]
cry27(GTGGCATATAGACTAAGGGAGGAA)(TTGCAGGCCATATAAGAGGTGTT)[21]
cry28(GTATTGGACCGAGGAGATGAAAGT)(GTACGGCAAAGCGACAGAACA)[21]
cyt1AGral-cyt(d) (AACCCCTCAATCAACAGCAAGG)Gral-cyt(r) (GGTACACAATACATAACGCCACC)[17]
cyt2upper (AATACATTTCAAGGAGCTA)lower (TTTCATTTTAACTTCATATC)[19]
 (ATCCGCCCATAATACAAG)(GATACGGTTCACAGACG)[21]

The identification of B. thuringiensis cry and cyt genes by PCR is performed under different conditions, as reported by the different authors. The concentration of the PCR mixture components varies as follows: primers, 0.1–1 μM; MgCl2, 1.5–3 mM; dNTPs, about 0.2 mM; and 0.25–2.5 U of DNA polymerase per reaction [14,26–28]. The DNA used as template is obtained either by the typical lysis extraction and purification methods [14,18,29], or by simple time-saving lysis of a cell suspension using physical methods. In the latter case, DNA is usually released by boiling the suspension for 2–10 min [13,15,30], or by alternate freezing (−70°C) and boiling steps [17,24,26]. Template is added to the reaction mixture (final volume 50–100 μl) and the three-step amplification cycles are performed 25–35 times. Typically, a 10–30% aliquot of the reaction mixture is analyzed by agarose gel electrophoresis.

Due to the extreme sensitivity of PCR, false positive amplifications may result from contamination of the PCR reactions. Also, false negative amplifications may be expected, as the amount of target DNA might be at the limit of detection, or due to the presence of inhibitors in the reaction mixture. The theoretical limit of sensitivity in a PCR using bacterial cells with a single-copy target gene is as low as one cell. Although a single bacterial colony from an overnight culture (on solid media) may contain up to several million cells, the release of DNA by boiling methods may be very inefficient. Studies with other bacteria, such as Staphylococcus aureus, have shown that the sensitivity of PCR is in practice much reduced by a set of complex factors, and experimental values of the detection limit are about 1000-fold higher than the theoretical limit [31]. PCR with boiled B. thuringiensis vegetative cells as template needs a minimum of 102–103 cells, depending on the strain and primers, for reproducible results (M. Porcar, unpublished observation). With regard to reaction inhibitors, many different substances usually present in a laboratory, such as laboratory plasticware, cellulose or glove powder are known to inhibit the PCR reaction. Additionally, non-target DNA and cell compounds may also inhibit polymerase activity [32], especially when lysed cells are used as the source of template. As for the contamination that may yield false positives, the exhaustive use of controls and standardization of the PCR conditions are imperative to guarantee reproducibility [33].

2.2Limitations of insecticidal activity prediction by PCR

The ability of PCR-mediated δ-endotoxin gene identification to predict insecticidal activity depends largely on several factors that can make the prediction erroneous. Very often, strains sharing the same cry and cyt gene content differ greatly in their insecticidal potency. The main causes of this lack of correspondence between cry/cyt genotype (as determined by PCR), and biological activity (as determined by bioassays) are described below.

2.2.1Gene identity

Primer design is a key factor in PCR. The 3′-end of primer is critical in the amplification procedure. Indeed, if a mismatch between the primer and the template occurs at this region, the amplification efficacy could be drastically diminished and may even result in the absence of an amplicon [34]. In consequence, any variation in one or two bases at this region may lead to closely related genes not being detected. In contrast, a substitution at the 5′-end or on the middle of the primer will not significantly affect the amplification and thus different genes (i.e. alleles differing in the quaternary rank of the classification but also other cry genes less closely related) can be equally amplified. Due to the high diversity of the cry gene family as well as the increasing number of novel genes described, primer design often requires a compromise between recognition of the whole pool of alleles and the absence of undesired cross-amplifications. For example, the specific primer pair SB-1 and U8-15C, designed in 1993 for the identification of the cry1Aa gene [26], would amplify only nine out of the 11 cry1Aa alleles known at that time. Alleles cry1Aa6 and cry1Aa8, reported in 1994 and 1998, respectively, may not be amplified because the 3′-end of one of the primer annealing sites is not conserved. By contrast, the gene cry1Ag (NCBI entry AFO81248) will be recognized, as it shares the sequences of the cry1Aa SB-1 and U8-15C primers.

Minor amino acid changes can influence insecticidal activity. Reports show that a few changes of certain residues within a loop region of the protein can affect both specific binding to membrane receptors and toxicity [35–37].

Such limitations hamper the use of conventional PCR to identify specific cry genes, and particularly to search for novel, previously undescribed cry genes. Theoretically, novel genes would only be detected if amplicons significantly differed in size from those expected. However, to date, no new cry genes have been described in this way.

2.2.2Expression level

Crystal proteins are synthesized in large amounts during stationary phase and accumulate in one or several parasporal crystals, accounting for up to 30% of the dry weight of sporulated cells. However, the expression level of individual cry genes present in any one strain can vary greatly. B. thuringiensis subsp. aizawai strain HD-133 is a good example of such variation. This strain is known to contain six cry genes, although only three proteins (Cry1Ab, Cry1C, and Cry1D) are expressed in detectable amounts in HD-133 parasporal crystals, at relative ratios of 60:37:3, respectively [22]. The three remaining genes are either silent due to an insertion within the coding sequence (cry1Aa), or expressed at undetectable levels if at all (cry2 and cry1I). This poor expression is probably due to the presence of a weak promoter upstream from the cry2 operon and, as suggested by Kostichka et al. [38], and to the secretion of the Cry1I protein during the early sporulation phase. In addition to the relative expression of each cry gene borne by a strain, a serovar-dependent regulation system of individual cry genes has been described. Cheng et al. [39] found a 3- to 4-fold decrease in expression of a cry1Ab-lacZ fusion within strains of serovar aizawai, as compared to those of serovars kurstaki or tolworthi. Regardless of the cause, the diverse expression levels of individual cry genes obviously weaken the correlation between cry gene content and toxicity.

2.2.3Protein interactions

Another cause of interference in the prediction of insecticidal activity from the identification of toxic genes is a series of protein interactions among δ-endotoxins. The toxicity of B. thuringiensis parasporal crystals depends not only on the activity of their individual components but also on the interactions between such proteins. These interactions were first described by Wu and Chang [40] and Ibarra and Federici [41] who reported that some individual protein components from the inclusion body of the dipteran-active B. thuringiensis subsp. israelensis showed lower activity against Aedes aegypti larvae, than did the native composite crystal suspensions.

The synergistic role of the Cyt toxins in the overall toxicity of B. thuringiensis israelensis has been largely demonstrated [42]. cyt genes have been used to transform another dipteran-active bacterium, B. sphaericus, resulting in a recombinant strain with expanded host range as a consequence of the synergism between Cyt1A and the B. sphaericus binary toxin [43]. Until now, Cyt proteins have only been identified in strains active on Diptera, with a unique exception: EA10192, belonging to serovar andaluciensis, which is PCR-positive in a screening with cyt2 primers, but with no known toxicity [20]. The synergistic effect between Cry and Cyt toxins may not be limited to Diptera since recombinant bacteria expressing Cyt1Aa together with the Coleoptera-active Cry3Aa suppress resistance to the latter toxin [44].

Another case of Cry protein interaction was described when three Cry1A proteins were tested against Lymantria dispar[45]. In this species, synergism was found between proteins Cry1Aa and Cry1Ac, whereas Cry1Aa and Cry1Ab interacted antagonistically. Other protein interactions involve the differential solubility of crystal components. Aronson et al. [46] demonstrated the importance of Cry1Ab in the solubilization and toxicity of HD-133 crystals, when crystals of HD-133 mutants, containing only Cry1C and Cry1D, required higher pH conditions to dissolve, causing a decrease in toxicity against Plodia interpunctella and other lepidopteran species. Finally, antagonistic interactions can also occur with Cry and Cyt toxins expressed together in recombinant strains. One example is the antagonism between Cry1Ac and Cyt1A when tested against Trichoplusia ni[47].

2.2.4Other virulence factors

Even if the detection of the cry and cyt content of a B. thuringiensis strain may allow, to certain extent (see above), the prediction of the insecticidal activity of its purified parasporal crystals, the complete pathogenic effect of a strain may involve other factors. It is known that a series of extracellular compounds synthesized by B. thuringiensis, such as β-exotoxins, phospholipases, proteases, chitinases and the secreted VIPs (vegetative insecticidal proteins) contribute to virulence. Additionally, spores are also known to synergize with the toxic effect of crystals when tested against some insect species, probably due to the invasion of hemocele through the ulcerated midgut, and the subsequent development of septicemia. The highest contribution of the spore to the pathogenicity of B. thuringiensis was reported in wax moth (Galleria mellonella) larvae. The addition of as low as 0.001% spores to a crystal suspension from a serovar aizawai strain greatly increased larval mortality [48]. Other studies have reported spore–crystal synergism against other Lepidoptera, such as Plutella xylostella[49] and P. interpunctella[50]. In summary, only bioassays with purified crystals are expected to correlate with the Cry and Cyt protein content of each strain, and partially with the cry and cyt gene content.

3Natural occurrence of cry genes among B. thuringiensis strains

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References

Ten years of PCR-based identification of cry genes from hundreds of B. thuringiensis strains isolated from samples collected world-wide, have resulted in a large collection of data. We have reviewed the frequency, regional distribution and combinations of individual cry genes from the six largest PCR screenings of natural occurring strains (Table 2). These screenings analyzed from 58 to almost 500 strains, and screened a large number of genes (at least eight different cry1 genes).

Table 2.  Diversity of cry gene content in several B. thuringiensis isolates collected and characterized in screening programs carried out world-wide
  1. aNote that profiles are set on the basis of combinations of only eight genes (cry1Aa, cry1Ab, cry1Ac, cry1B, cry1C, cry1D, cry1E, and cry1F).

  2. bRefers to the total number of strains analyzed, except in the studies in which this value is not shown in the publication (NS).

  3. cIncludes Israel, Kazakhstan and Uzbekistan.

  4. dThe detection of cry genes was by PCR and ELISA.

  5. eBoth reports refer to the same collection. The percentage of strains is that of Juárez-Pérez [25]; the number of cry1 profiles is based on Ferrandis et al. [18]. In this latter report, cry1A-containing profiles were grouped regardless of the type of gene present (cry1Aa, cry1Ab, or cry1Ac); the number of cry1 profiles is thus probably artificially diminished.

Geographic area surveyedIsolates with cry genes/total isolatesNumber of cry genes analyzedNumber of cry1 profilesaPercentage of strains bearing a given cry1 genebReference
    cry1Aacry1Abcry1Accry1Bcry1Ccry1Dcry1Ecry1F 
Taiwan225/NS847859780182000[14]
Asiac126/2152110171710071800[16]
Korea49/581913284592571700[30]
Mexico423/4962427221513961712[17]d
Spain171/22313≥9261831917224<1[18,25]e
China122/NS101831596911925440[54]

3.1Relative frequency of cry genes

The largest collection of B. thuringiensis subjected to PCR screening of cry genes published to date contained isolates from around 500 soil samples [17]. These samples were obtained from five very different macroecological regions of Mexico. A total of 496 strains were subjected to PCR and enzyme-linked immunosorbent assay (ELISA) to identify cry1, cry3, cry5, cry7, cry8, cry9, cry11, cry12, cry13, cry14, cry21, and cyt genes. None of these genes were detected in 14% of the strains. In other studies based solely on PCR analysis, the percentage of strains that failed to produce amplicons of the expected sizes ranged from 16%[30] to more than 40%[16]. These strains may contain cry genes not recognized by the set of primers used, or contain unknown genes.

The most common cry genes found in nature are those within the cry1 subfamily, with about half of the strains or more bearing these genes. The cry2 genes are also very frequent, especially among cry1-containing strains [16]. The occurrence of cry1 groups varies greatly. Some, such as the cry1A genes, are very frequent, being present usually in more than half of the strains; whereas other genes, such as cry1Fs, are rare. As mentioned, the most common cry1 genes are those belonging to the cry1A group, followed by the cry1C and cry1D groups. On the other hand, cry1B, cry1E and cry1F are usually found at low frequencies, although there are some exceptions, such as the high frequency with which cry1E genes were found in a Chinese collection (Table 2). Also, cry1B genes have been found at a relatively high frequency (30% of the strains) in some screenings [15]. The occurrence of cry1I genes among B. thuringiensis strains, as deduced from PCR-based studies, is very high [13,18,51]. Hybridization with specific probes and reverse transcription (RT)-PCR analysis [27] confirm that these genes are widely distributed among B. thuringiensis strains.

Several reports show a high frequency of certain combinations of cry1 genes. For example, cry1I genes frequently occur when other cry1-type genes are present. This observation is consistent with the physical location of cry1I genes, which have been reported to be in close vicinity to other genes of the cry1 subfamily [13,27,52,53]. Another frequent combination found almost world-wide is the linkage cry1Ccry1D[15,17,18,51,54]. In fact, only Kim et al. [30] reported the common presence of cry1C alone. However, in the rest of the studies, cry1D was found alone at a relatively high frequency, but cry1C was almost always associated with cry1D. This cry1Ccry1D linkage may be explained by their location on the same replicon, as described by Sanchis et al. [53] for a B. thuringiensis subsp. aizawai strain in which the genes were separated by only 3 kb. Regarding the cry1D-containing strains that lack cry1C, an evolutionary event could be involved, as proposed by Ferrandis et al. [18], who suggested that the absence of cry1C may be the consequence of a deletion or the negative selection of cry1C from an ancestral cry1Ccry1D linkage. The fact that the cry1C gene is located downstream of an IS sequence [55] may account for this hypothetical mobility.

3.2Genetic diversity and geographic variation

Although a great collection of data is available from the PCR-mediated cry gene screening studies reported to date, there are several factors that limit their comparison: (i) the use of different primer pairs; (ii) the variation in PCR conditions; and (iii) the number of strains and genes analyzed. Another important limitation is the lack of preliminary characterization of isolates. The selection of replicas during the isolation procedure is a frequent occurrence if basic characterization procedures (such as sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE), serotyping crystal morphology, etc.) are not followed, especially when several isolates come from the same sample. However, some collections do not follow this preliminary step, and this may have a strong influence on the apparent gene diversity of the collection.

An attempt to compare the six main PCR screening programs published to date is summarized in Table 2. The number of cry genes analyzed in these programs ranges from eight to 24. Most of them are cry1 genes but some other subfamilies are also represented. Since most of the information concerns cry1 genes, we focused on this subfamily. In order to understand the diversity of cry1 genes within and between collections, the profiles (combinations of genes) of eight cry1 genes are compared. If other genes were analyzed, they have not been taken into account to calculate the number of profiles. Diversity in terms of cry gene content of a B. thuringiensis collection is influenced by the pre-selection procedure and may only partially reflect the real genetic diversity of naturally occurring strains. The environmental diversity of the geographic area surveyed may also account for the high number of different cry gene profiles with respect to the number of samples analyzed. However, and due to the low number of studies combining all these aspects, further studies in this field are needed to obtain stronger conclusions. Those studies should combine a large number of samples analyzed with a high diversity of natural habitats surveyed.

3.3Occurrence within serovars

B. thuringiensis strains can be serologically classified according to the flagellar (H) antigens. To date, a total of 82 different serovars have been described [56] and it is likely that the number of serovars will grow due to the expanding native strain isolation and screening programs. It is generally accepted that serological characterization does not directly reflect the toxicity (potency or host range) of a given strain. However, Dulmage [57] found a certain correlation between toxicity and serovars and their results revealed that serological grouping of B. thuringiensis strains partially describes their toxicity spectra, although a significant variation within serovars occurs (data reviewed by Glare and O'Callaghan [58]). The archetypical example of the correlation between serovar and toxicity may be the B. thuringiensis serovar israelensis strains, which have an almost exclusive toxic spectrum against Nematocera.

Because of the variation in the expression level of cry genes between strains, the combination of PCR with the serological identification of strains appears to be very useful to test the genetic robustness of the serological classification, and to resolve any correlation between cry content and serovar. A few studies have combined PCR-based identification of cry genes and serology [18,30,51,54] and all of them show a high diversity in gene distribution within and among serovars. Hongyu et al. [54] analyzed the relationship between gene composition and serology of 122 strains isolated from natural samples in China, and concluded that although a direct relationship between gene content and serovar was not established, some association was observed. Particularly, some cry1E-containing combinations such as cry1Ab-1Ac-1E and cry1Ac-1E were very frequent among strains within the serotype H4 (sotto-kenyae). However, other cry1E-containing combinations, such as cry1Aa-1Ab-1Ac-1E were absent within serotype H4. In another analysis of B. thuringiensis strains isolated in Spain [51], a serovar-dependent distribution of cry1C and cry1D was suggested, as these genes were very frequent in serovar aizawai. Also, the distribution of cry1B in this collection was restricted to serovar thuringiensis. Despite the apparent correlation shown in these reports, the number of strains analyzed may still be insufficient to determine whether the apparent non-random distribution of some cry genes among serovars corresponds to geographical variation, reflects an inherent serovar-dependent occurrence, or plasmid compatibility/incompatibility interactions.

4Identification of novel cry genes

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References

4.1DNA-based methods

4.1.1PCR

As mentioned above, the PCR amplification of fragments of unexpected size when gene-specific primers are used may lead to the detection of new cry genes. To date, only Cerón et al. [15] have reported an amplified fragment that may correspond to a new cry gene, when using a multiplex PCR with specific primers. Unfortunately, no further information concerning the identity of this putative cry-like gene is available to date.

Due to the intrinsic limitation of ‘classical’ PCR screening of B. thuringiensis collections to identify novel cry genes, several PCR-based methods to detect and characterize unknown cry genes have been developed. In 1993, Kalman et al. [29] proposed a strategy to identify variants of the cry1C group. Their experimental design can be considered as ‘PCR walking’, because a series of primers were designed to anneal throughout the cry1Ca1 sequence. In a single multiplex PCR reaction the entire gene was amplified, obtaining a characteristic amplification profile. If a new cry1C-related gene was present in a strain, at least one of the corresponding PCR products was supposed either to be lacking or to exhibit an unexpected size. In either case, a modification of the predicted PCR profile would indicate the presence of a new cry1C gene. With this method they successfully detected the gene now known as cry1Cb1. Unfortunately, this methodology is restricted to closely related genes within the same group. There are other disadvantages, such as the large number of primers required to analyze each group of genes, and the fact that the putative variant may not be more interesting, in terms of toxicity, than the genes of the same group already described.

Subsequently, two further PCR-based methods were developed to determine the presence of known cry genes and also of putative new cry genes. In both cases, the techniques were adapted to fit a wider spectrum of genes, rather than to fit just one group. Multiple alignment of the DNA coding sequences of members of different subfamilies allows the detection of conserved regions that are unique for each subfamily. This characteristic opened a new scenario in the study and analysis of the cry gene content of a B. thuringiensis strain because ‘universal’ subfamily primers could be designed. Kuo et al. [59] reported the first primer pair able to recognize all the members of the cry1 gene subfamily known at that time, as well as some other genes coding for the actual cry4, cry3 and cry7. Juárez-Pérez et al. [24], Juárez-Pérez [25], and Masson et al. [22] also used the same concept of universal primers to amplify all the members of different subfamilies of the cry genes. Moreover, they introduced the use of degenerate primers in order to increase the probability of amplifying novel group members that did not exactly match the putative conserved region of the subfamilies from which the oligonucleotides were designed.

The first method specifically designed to detect new cry genes was based on the combination of PCR and restriction fragment length polymorphism (RFLP). This is a two-step strategy where group-specific primers are used first, followed by enzymatic digestion of the produced amplicon(s). With this method, a particular RFLP pattern is expected for each gene and, consequently, also for a given combination of genes. If a different profile is obtained due to the absence or presence of a given restriction site, the corresponding fragment(s) may be easily cloned and sequenced, and then used as a probe for the cloning of the entire putative new cry gene. This strategy was used by Kuo and Chak [60] and by Juárez-Pérez [25] who used different universal primers and restriction enzymes. However, when more than four cry genes are present in a strain, as is frequently observed in wild-type strains, the restriction profile is often too complex and the identification of the corresponding genes becomes difficult. Furthermore, due to the high similarity between members of the cry1 subfamily, the PCR-RFLP analysis conducted by Kuo and Chak [60] was unable to detect differences between the cry1Ca, cry1Cb, cry1Ea and cry1Fa genes. To overcome this problem, a second PCR cycle, using alternative forward primers, was proposed to amplify other regions of the genes involved, followed by a second enzymatic reaction. A long electrophoresis run was also required in order to achieve a better resolution of restriction fragments. When this technique was tested with well-known standard strains as well as wild isolates (20 B. thuringiensis isolates) it produced the expected profiles from the former and some few unexpected variants from the latter. Cloning and sequencing of new cry genes corroborated the efficacy of this technique [60].

Another approach to detect new cry genes is based on the use of two sequential PCR reactions, using a multiplex PCR with specific and universal primers [24]. This strategy, called exclusive-PCR (E-PCR), starts with the amplification of already described cry1 genes, followed by a second conditional amplification that will occur only if a new putative cry1 gene(s) is present in the strain. This method is, in fact, a combination of several techniques available at that time: (i) the use of specific primers designed to recognize only one type of gene; (ii) the use of universal primers designed to detect entire groups; and (iii) the use of multiplex PCR. Also, the use of universal degenerate primers was introduced to increase the probability of amplifying sequences with low homology within the subfamily.

The authors [24] used the cry1 subfamily as a model to test this technique because it is the largest cluster of the cry family. For the first amplification reaction, a reverse universal primer for the entire cry1 group and a specific primer for each subgroup were designed. This first step identifies the known cry1 genes contained in the strain. This information is essential for the second PCR step, which is conducted with forward and reverse universal primers designed to amplify a 1.5–1.6 kb fragment of all known and unknown cry1-related genes, combined with each of the specific primers that showed amplification in the first step. It is well known that when a mixture of two forward primers and one reverse primer (or vice versa) is used, the smaller fragment is amplified preferentially over the larger amplicon. It is even possible to exclude the amplification of the larger fragment by modifying the PCR conditions. With this technique, the universal primers used in the second reaction were designed to amplify the larger amplicon, called the ‘family band’, whereas the use of the specific primer and one universal forward primer amplifies a smaller fragment called the ‘type band’. The authors modified the PCR conditions of this second multiplex PCR to eliminate the family band, obtaining only the expected amplicons of the known genes. Therefore, if a new cry gene related to the cry1 group is present in a given strain, it is revealed by the presence of the so-called family band. On the other hand, if no new cry genes are present in the strain, the family band is not expected to be amplified. To experimentally demonstrate the efficacy of this methodology, they used the well-known B. thuringiensis strain HD-133. In the first step, all the genes already described were detected. In the second reaction, and by omitting one primer, they simulated the presence of a new cry gene resulting in the amplification of the family band. These results encouraged the analysis of a large B. thuringiensis collection. The application of the E-PCR methodology resulted in the detection of a new member of the cry1B subgroup [24]. We (Juárez-Pérez, unpublished results) have applied E-PCR to cry gene subfamilies other than cry1. Taking into account that cry1 is the largest cry subfamily, this suggests that E-PCR may be suitable for the entire family of cry genes.

4.1.2Hybridization

Even though the homology between two cry sequences within the same group can be as low as 45%, the homology of some specific regions within the sequences can be as high as 95%. Because of this homology between cry gene sequences, the development of specific probes was used earlier to detect and characterize putative new cry genes, using hybridization techniques [61]. With the advent of faster and more reliable technologies and, especially, with the increasing number of characterized cry genes, this methodology was abandoned due to its low specificity and the high number of probes required. Despite this, a strategy was recently developed to identify novel cry genes using two mixtures of hybridizing probes that are used separately in two different DNA–DNA hybridization reactions [62]. The rationale is also composed of two steps. The conserved sequence mixture was directed towards the identification of all the genes from eight selected cry subfamilies and was used in a first hybridization step. In this way, if a gene belongs to one of the subfamilies tested, the low level of specificity of the hybridization method permits its identification. The second mixture made it possible to identify new variants of the cry groups tested when low stringency hybridization conditions are used. Since the variable regions of the genes can form duplex structures, even if the matching between the probe and the test DNA is not perfect, related genes can be recognized. Using low stringency hybridization conditions and comparing the hybridization profiles produced by the two mixtures with the same membrane (ensuring identical DNA digestion and transfer conditions), they were able to detect different hybridization profiles for the strains tested. Using this method, they reported the discovery of a new gene having 80% identity with the cry family, placing this sequence as a putative new member of an already known group of cry genes. They also mentioned the isolation of several other new genes using the same approach. However, this methodology is not time-saving because it does not, unlike the PCR-based approaches, produce a fragment of the putative new gene that would simplify its cloning. Despite these constraints, this method could be useful to identify new subfamilies of cry genes. The use of entire gene sequences (all those presenting the characteristic conserved blocks of the cry family of genes), a probe mixture of only the conserved regions within a subfamily and in combination with a non-specific reaction (hybridization), could permit the detection of distantly related (unknown) cry subfamilies.

4.2Other analytical methods

Another analytical technique to identify new B. thuringiensis toxins was used by a Russian group [63]. Rather than analyzing the gene content of a strain, they studied in detail the Cry composition of the parasporal body from several B. thuringiensis strains (representing perhaps the first proteomic-type work on B. thuringiensis). By purifying and micro-sequencing the major peptides resulting from trypsin digestion of the solubilized parasporal body, they were able to find that B. thuringiensis subspecies galleriae VKPM B-1757 and wuhanensis VKPM B-1226 parasporal bodies were composed of six and seven proteins, respectively. Comparison of the protein sequences with those of known Cry proteins showed that most of them were already described. However, for each strain two protein sequences lacked complete homology with the known Cry proteins, suggesting possible new toxins. Additionally, this strategy determines the proteins that are actually synthesized and present in the parasporal body, which is an advantage compared with the PCR-based technologies used for toxicity prediction.

5The prediction of toxicity of B. thuringiensis: future prospects

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References

There is no doubt that the use of PCR has greatly improved the screening of the increasing number of B. thuringiensis strains isolated world-wide; however, it is mostly limited to the detection within a strain of previously known genes. Although some strategies have already been developed to detect new genes, they are time-consuming when a great number of isolates are studied. Furthermore, even if a new gene is successfully detected, its toxic activity has to be tested directly by bioassay against a series of insect species. Additionally, the supposition that the expected size of an amplicon implies the detection of a known cry gene is not necessarily true. Furthermore, the detection of a cry gene by PCR is no direct proof of its expression (or level of expression).

The advent of new biological tools to analyze the gene and/or protein expression may help to overcome some of the problems described above. The automation of several routine duties in laboratories and the development of the genomic and proteomic technologies could be used in the analysis of large collections of B. thuringiensis. Macroarrays are the low-cost version of DNA chips and are widely used in gene expression studies. If we take into account the present state of this technology, probes can be designed to specifically identify genes up to the third rank of the cry gene classification, and macroarrays may therefore improve the efficiency of detection of known cry genes present in a given strain. Additionally, the use of cDNA to hybridize the spotted probes enables the detection of only those cry genes that are expressed in the strain. The cDNA can be obtained from a multiplex RT-PCR with universal primers for each cry group or by mixing the RT-PCR products of individual reactions. However, a major inconvenience with this kind of technology may be the non-specific reactions or cross-reactions of the immobilized oligonucleotides in the macroarray membrane with the amplified cDNA. A good compromise between the size of the cDNA and oligonucleotide design will be a major determinant in the successful implementation of this powerful analytical tool.

Additionally, it may be of interest to quantify the relative amount of each cry transcript by real time RT-QPCR (reverse transcription-quantitative PCR) at different stages of spore development. This information may have both practical and basic applications. A better understanding of the proportion of each Cry protein in the crystal can predict more accurately the toxicity spectrum and potency of a strain. Also, this approach may contribute to our understanding of the cry gene transcription and regulation processes under different culture conditions, even if post-transcriptional and translational factors, that play an important role on the Cry protein accumulation in crystals, are not considered by this approach. This information may be important for a basic understanding of the regulation of cry gene expression and for monitoring the industrial production of B. thuringiensis-based products.

It is clear that only the direct study of the Cry protein content of a strain can lead to precise information concerning its toxicity spectrum. The automation and technological improvements of the last few years in the study of protein complexes at a cellular level (proteomics) will contribute to further characterization of the B. thuringiensis parasporal crystal. Chestukhina and coworkers [63] demonstrated that such studies could be successfully carried out. The technical problems with the current state of proteomics lie in the achievement of digested products pure enough to be easily recognized by peptide mapping and/or mass spectrometry [64] and the separation of Cry proteins (Goar, personal communication). If this restriction was solved, we would be able to rapidly and reliably identify the protein content of the parasporal crystal of a given B. thuringiensis strain. Proteomic studies might be the key to determining the relative proportion of Cry proteins forming the crystal, which, to some extent, correlates with the strain's toxic spectrum. This methodology may also provide information concerning the interactions among Cry proteins. Protein ratios can be used to calculate the expected toxicity of the whole crystal with a simple formula [65]. Since the expected LD50 (dose needed to kill 50% of the insects) is

  • image

where ra, rb, and rc are the relative proportions of toxins a, b, and c, respectively, one can easily evaluate synergism among toxins if the exact protein content (three toxins in this example) of the parasporal crystal body is determined.

In summary, the relatively modest success of the PCR-based methods in the identification of novel B. thuringiensis crystal genes is probably related to technical problems concerning either the complexity of PCR-RFLPs of multigenic cry-bearing strains or the adaptation to each laboratory of the E-PCR. Unfortunately, owing to its sensitivity, E-PCR can lead to different results depending on the type of thermocycler and/or consumables used. We suggest that PCR-RFLP may be a useful methodology if it is restricted to the detection of new genes within existing cry groups, since the profiles will be less complex and more restriction enzymes could be used to confirm this difference. E-PCR may be more useful to analyze those strains whose toxicity cannot be explained by the putative gene content of the strain where only known genes were identified. Finally, proteomic technology is a very promising tool that could be adapted to the study of the B. thuringiensis crystal. The presence of each Cry protein, both known and unknown, would be revealed by new peptide profiles or sequences, opening a new age in the understanding of the still fabulous mystery that is the insecticidal parasporal body of B. thuringiensis.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References

We are indebted to Jeroen Van Rie and Jorge E. Ibarra for critical reading of the manuscript and to Cristina Patricio and Patricia Davis for assistance with English language.

References

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Use of PCR for the prediction of insecticidal activity
  5. 3Natural occurrence of cry genes among B. thuringiensis strains
  6. 4Identification of novel cry genes
  7. 5The prediction of toxicity of B. thuringiensis: future prospects
  8. Acknowledgements
  9. References
  • [1]
    Ishiwata, S (1901) On a kind of severe flacherie (sotto disease). Dainihon Sanshi Kaiho 114, 15.
  • [2]
    Berliner, E (1911) Über die Schlaffsucht der Mehlmottenraupe. Z. Gesamte Getreidewes. (Berlin) 3, 6370.
  • [3]
    Goldberg, L.H., Margalit, J (1977) A bacterial spore demonstrating rapid larvicidal activity against Anopheles sergentii, Uranotaenia unguiculata, Culex univitatus, Aedes aegypti and Culex pipiens. Mosq. News 37, 355358.
  • [4]
    Krieg, A, Huger, A.M., Langenbruch, G.A., Schnetter, W (1983) Bacillus thuringiensis var. tenebrionis: ein neuer gegenuber arven von Coleopteren wirksamer athotyp. Z. Ang. Ent. 96, 500508.
  • [5]
    Feitelson, J.S. (1993) The Bacillus thuringiensis family tree. In: Advanced Engineered Pesticides (Kim, L., Ed.), pp. 63–71. Marcel Dekker, New York.
  • [6]
    Höfte, H, Whiteley, H.R (1989) Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53, 242255.
  • [7]
    Lüthy, P. and Wolfersberger, M.G. (2000) Pathogenisis of Bacillus thuringiensis toxins. In: Entomopathogenic Bacteria: From Laboratory to Field Application (Charles, J.F., Delécluse, A. and Nielsen-LeRoux, C., Eds.), pp. 167–180. Kluwer Academic Publishers, Dordrecht.
  • [8]
    Schwart, J.L. and Laprade, R. (2000) Membrane permeabilization by Bacillus thuringiensis toxins: protein formation and pore insertion. In: Entomopathogenic Bacteria: From Laboratory to Field Application (Charles, J.F., Delécluse, A. and Nielsen-LeRoux, C., Eds.), pp. 199–218. Kluwer Academic Publishers, Dordrecht.
  • [9]
    Martínez, C. (2002) Biological and Genetic Characterization of Bacillus thuringiensis Strains Toxic against Helicoverpa armigera, 573 pp. UPNA, Pamplona.
  • [10]
    Kleppe, K, Ohtsuka, E, Kleppe, R, Molineaux, I, Khorana, H.G (1971) Studies on polynucleotides. XCVI. Repair replications of short synthetic DNA's catalyzed by DNA polymerases. J. Mol. Biol. 56, 341361.
  • [11]
    Saiki, R.K., Gelfand, D.H., Stoffel, S, Scharf, S.J., Higuchi, R, Horn, G.T., Mullis, K.B., Erlich, H.A (1988) Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239, 487489.
  • [12]
    Carozzi, N.B., Kramer, V.C., Warren, G.W., Evola, S, Koziel, M.G (1991) Prediction of insecticidal activity of Bacillus thuringiensis strains by polymerase chain reaction product profiles. Appl. Environ. Microbiol. 57, 30573061.
  • [13]
    Gleave, A.P., Williams, R, Hedges, R.J (1993) Screening by polymerase chain reaction of Bacillus thuringiensis serotypes for the presence of cryV-like insecticidal protein genes and characterization of a cryV gene cloned from B. thuringiensis subsp. kurstaki. Appl. Environ. Microbiol. 59, 16831687.
  • [14]
    Chak, K.-F, Chao, D.-C, Tseng, M.-Y, Kao, S.-S, Tuan, S.-J, Feng, T.-Y (1994) Determination and distribution of cry-type genes of Bacillus thuringiensis isolates from Taiwan. Appl. Environ. Microbiol. 60, 24152420.
  • [15]
    Cerón, J, Ortíz, A, Quintero, R, Güereca, L, Bravo, A (1995) Specific PCR primers directed to identify cryI and cryIII genes within a Bacillus thuringiensis strain collection. Appl. Environ. Microbiol. 61, 38263831.
  • [16]
    Ben-Dov, E, Zaritsky, A, Dahan, E, Barak, Z, Sinai, R, Manasherob, R, Khamraev, A, Troitskaya, E, Dubitsky, A, Berezina, N, Margalith, Y (1997) Extended screening by PCR for seven cry-group genes from field-collected strains of Bacillus thuringiensis. Appl. Environ. Microbiol. 63, 48834890.
  • [17]
    Bravo, A, Sarabia, S, Lopez, L, Ontiveros, H, Abarca, C, Ortiz, A, Ortiz, M, Lina, L, Villalobos, F.J., Peña, G, Nuñez-Valdez, M.E., Soberon, M, Quintero, R (1998) Characterization of cry genes in a Mexican Bacillus thuringiensis strain collection. Appl. Environ. Microbiol. 64, 49654972.
  • [18]
    Ferrandis, M.D., Juárez-Pérez, V.M., Frutos, R, Bel, Y, Ferré, J (1999) Distribution of cryI, cryII and cryV genes within Bacillus thuringiensis isolates from Spain. Syst. Appl. Microbiol. 22, 179185.
  • [19]
    Guerchicoff, A, Ugalde, R.A., Rubinstein, C.P (1997) Identification and characterization of a previously undescribed cyt gene in Bacillus thuringiensis subsp. israelensis. Appl. Environ. Microbiol. 63, 27162721.
  • [20]
    Guerchicoff, A, Delécluse, A, Rubinstein, C.P (2001) The Bacillus thuringiensis cyt genes for hemolytic endotoxins constitute a gene family. Appl. Environ. Microbiol. 67, 10901096.
  • [21]
    Ejiofor, A.O., Johnson, T (2002) Physiological and molecular detection of crystalliferous Bacillus thuringiensis strains from habitats in the South Central United States. J. Indian Microbiol. Biotechnol. 28, 284290.
  • [22]
    Masson, L, Erlandson, M, Puzstai-Carey, M, Brousseau, R, Juárez-Pérez, V, Frutos, R (1998) A holistic approach for determining the entomopathogenic potential of Bacillus thuringiensis strains. Appl. Environ. Microbiol. 64, 47824788.
  • [23]
    Schnepf, E, Crickmore, N, Van Rie, J, Lereclus, D, Baum, J, Feitelson, J, Zeigler, D.R., Dean, D.H (1998) Bacillus thuringiensis and its pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62, 775806.
  • [24]
    Juárez-Pérez, V.M., Ferrandis, M.D., Frutos, R (1997) PCR-based approach for detection of novel Bacillus thuringiensis cry genes. Appl. Environ. Microbiol. 63, 29973002.
  • [25]
    Juárez-Pérez, V. (1998) Caractérisation de Nouveaux Gènes d'Endotoxines de Bacillus thuringiensis Berliner, 156 pp. ENSA/AGRO, Montpellier.
  • [26]
    Bourque, S.N., Valero, J.R., Mercier, J, Lavoie, M.C., Levesque, R.C (1993) Multiplex polymerase chain reaction for detection and differentiation of the microbial insecticide Bacillus thuringiensis. Appl. Environ. Microbiol. 59, 523527.
  • [27]
    Shin, B.-S, Park, S.-H, Choi, S.-K, Koo, B.-T, Lee, S.-T, Kim, J.-I (1995) Distribution of cryV-type insecticidal protein genes in Bacillus thuringiensis and cloning of cryV-type genes from Bacillus thuringiensis subsp. kurstaki and Bacillus thuringiensis subsp. entomocidus. Appl. Environ. Microbiol. 61, 24022407.
  • [28]
    Cerón, J, Covarrubias, L, Quintero, R, Ortíz, A, Ortíz, M, Aranda, E, Lina, L, Bravo, A (1994) PCR analysis of the cryI insecticidal family genes from Bacillus thuringiensis. Appl. Environ. Microbiol. 60, 353356.
  • [29]
    Kalman, S, Kiehne, K.L., Libs, J.L., Yamamoto, T (1993) Cloning of a novel cryIC-type gene from a strain of Bacillus thuringiensis subsp. galleriae. Appl. Environ. Microbiol. 59, 11311137.
  • [30]
    Kim, H.S., Lee, D.W., Woo, S.D., Yu, Y.M., Kang, S.K (1998) Distribution, serological identification, and PCR analysis of Bacillus thuringiensis isolated from soils of Korea. Curr. Microbiol. 37, 195200.
  • [31]
    Wilson, I.G., Cooper, J.E., Gilmour, A (1991) Detection of enterotoxigenic Staphylococcus aureus in dried skimmed milk: use of the polymerase chain reaction for amplification and detection of staphylococcal enterotoxin genes entB and entC1 and the thermonuclease gene nuc. Appl. Environ. Microbiol. 57, 17931798.
  • [32]
    Wilson, I.G (1997) Inhibition and facilitation of nucleic acid amplification. Appl. Environ. Microbiol. 63, 37413751.
  • [33]
    Roux, K.H (1995) Optimization and troubleshooting in PCR. PCR Methods Appl. 4S, 185194.
  • [34]
    Kwok, S, Seng-Yung, C, Sninsky, J.J., Wang, A (1994) A guide to design and use of mismatched and degenerated primers. PCR Methods Appl. 4S, 3947.
  • [35]
    Aronson, A.I., Wu, D, Zhang, C (1995) Mutagenesis of specificity and toxicity regions of a Bacillus thuringiensis protoxin gene. J. Bacteriol. 177, 40594065.
  • [36]
    Dean, D.H., Rajamohan, F, Lee, M.K., Wu, S.J., Chen, X.J., Alcantara, E, Hussain, S.R (1996) Probing the mechanism of action of Bacillus thuringiensis insecticidal proteins by site-directed mutagenesis: a minireview. Gene 179, 111117.
  • [37]
    Barboza-Corona, J.E., Lopez-Meza, J.E., Ibarra, J.E (1998) Cloning and expression of the cry1Ea4 gene of Bacillus thuringiensis and the comparative toxicity of its gene product. World J. Microbiol. Biotechnol. 14, 437441.
  • [38]
    Kostichka, K, Warrren, G.W., Mullins, M, Mullins, A.D., Craig, J.A., Koziel, M.G., Estruch, J.J (1996) Cloning of a cryV-type insecticidal protein gene from Bacillus thuringiensis: the cryV-encoded protein is expressed early in stationary phase. J. Bacteriol. 178, 21412144.
  • [39]
    Cheng, P, Wu, L, Ziniu, Y, Aronson, A.I (1999) Subspecies-dependent regulation of Bacillus thuringiensis protoxin genes. Appl. Environ. Microbiol. 65, 18491853.
  • [40]
    Wu, D, Chang, F.N (1985) Synergism in mosquitocidal activity of 26 and 65 kDa proteins from Bacillus thuringiensis subsp. israelensis crystal. FEBS Lett. 190, 232236.
  • [41]
    Ibarra, J.E., Federici, B.A (1986) Parasporal bodies of Bacillus thuringiensis subsp. morrisoni (PG-14) and Bacillus thuringiensis subsp. israelensis are similar in protein composition and toxicity. FEMS Microbiol. Lett. 34, 7984.
  • [42]
    Delécluse, A., Juárez-Pérez, V. and Berry, C. (2000) Vector-active toxins: structure and diversity. In: Entomopathogenic Bacteria: From Laboratory to Field Application (Charles, J.F., Delécluse, A. and Nielsen-LeRoux, C., Eds.), pp. 101–125. Kluwer Academic Publishers, Dordrecht.
  • [43]
    Wirth, M.C., Walton, W.E., Federici, B.A (2000) Cyt1A from Bacillus thuringiensis restores toxicity of Bacillus sphaericus against resistant Culex quinquefasciatus (Diptera: Culicidae). J. Med. Entomol. 37, 401407.
  • [44]
    Federici, B.A., Bauer, L.S (1998) Cyt1Aa protein of Bacillus thuringiensis is toxic to the cottonwood leaf beetle, Chrysomela scripta, and suppresses high levels of resistance to Cry3Aa. Appl. Environ. Microbiol. 64, 43684371.
  • [45]
    Lee, M.K., Curtiss, A, Alcantara, E, Dean, D.H (1996) Synergistic effect of the Bacillus thuringiensis toxins Cry1Aa and Cry1Ac on the gypsy moth, Lymantria dispar. Appl. Environ. Microbiol. 62, 583586.
  • [46]
    Aronson, A.I., Han, E.-S, McGaughey, W, Johnson, D (1991) The solubility of inclusion proteins from Bacillus thuringiensis is dependent upon protoxin composition and is a factor in toxicity to insects. Appl. Environ. Microbiol. 57, 981986.
  • [47]
    Del Rincón-Castro, M.C., Barajas-Huerta, J, Ibarra, J.E (1999) Antagonism between Cry1Ac1 and Cyt1A1 toxins of Bacillus thuringiensis. Appl. Environ. Microbiol. 65, 20492053.
  • [48]
    Li, R.S., Jarrett, P, Burges, H.D (1987) Importance of spores, crystals, and delta-endotoxins in the pathogenicity of different varieties of Bacillus thuringiensis in Galleria mellonella and Pieris brassicae. J. Invertebr. Pathol. 50, 277284.
  • [49]
    Miyasono, M, Inagaaki, S, Yamamoto, M, Ohba, K, Ishiguro, T, Takeda, R, Hayashi, Y (1994) Enhancement of δ-endotoxin activity by toxin-free spore of Bacillus thuringiensis against the Diamondback Moth, Plutella xylostella. J. Invertebr. Pathol. 63, 111112.
  • [50]
    Johnson, D.E., McGaughey, W.H (1996) Contribution of Bacillus thuringiensis spores to toxicity of purified Cry proteins towards indianmeal moth larvae. Curr. Microbiol. 33, 5459.
  • [51]
    Porcar, M, Martínez, C, Caballero, P (2000) Host range and gene contents of Bacillus thuringiensis strains toxic towards Spodoptera exigua. Entomol. Exp. Appl. 97, 339346.
  • [52]
    Chambers, J.A., Jelen, A, Gilbert, M.P., Jany, C.S., Johnson, T.B., Gawron-Burke, C (1991) Isolation and characterization of a novel insecticidal crystal protein gene from Bacillus thuringiensis subsp. aizawai. J. Bacteriol. 173, 39663976.
  • [53]
    Sanchis, V, Lereclus, D, Menou, G, Chaufaux, J, Lecadet, M.-M (1988) Multiplicity of δ-endotoxin genes with different specificities in Bacillus thuringiensis aizawai 7.29. Mol. Microbiol. 2, 393404.
  • [54]
    Hongyu, Z, Ziniu, Y, Wangxi, D (2000) Composition and ecological distribution of Cry proteins and their genotypes of Bacillus thuringiensis isolates from warehouses in China. J. Invertebr. Pathol. 76, 191197.
  • [55]
    Lereclus, D., Delécluse, A. and Lecadet, M.-M. (1993) Diversity of Bacillus thuringiensis toxins and genes. In: Bacillus thuringiensis and Environmental Biopesticide: Theory and Practice (Entwistle, P.S., Cory, J.S., Bailey, M.J. and Higgs, S.R., Eds.), pp. 309–329. Wiley, London.
  • [56]
    Lecadet, M.M., Frachon, E, Dumanoir, V.C., Ripouteau, H, Hamon, S, Laurent, P, Thiéry, I (1999) Updating the H-antigen classification of Bacillus thuringiensis. J. Appl. Microbiol. 86, 660672.
  • [57]
    Dulmage, H.T. (1981) Insectidal activity of isolates of Bacillus thuringiensis and their potential for pest control. In: Microbial Control of Pests and Plant Diseases (Burges, H.D., Ed.), pp. 193–223. Academic Press, London.
  • [58]
    Glare, T.R. and O'Callaghan, M. (2000) A case study: toxicity of Dipal, Bt kurstaki produced by Abbatt Laboratories. In: Bacillus thuringiensis: Biology, Ecology and Safety, pp. 311–316. Wiley, Chichester.
  • [59]
    Kuo, W.S., Lin, J.H., Tzeng, C.C., Kao, S.S., Chak, K.F (2000) Cloning of two new cry genes from Bacillus thuringiensis subsp wuhanensis strain. Curr. Microbiol. 40, 227232.
  • [60]
    Kuo, W.S., Chak, K.F (1996) Identification of novel cry-type genes from Bacillus thuringiensis strains on the basis of restriction fragment length polymorphism of the PCR-amplified DNA. Appl. Environ. Microbiol. 62, 13671377.
  • [61]
    Crickmore, N, Zeigler, D.R., Feitelson, J, Schnepf, E, Van Rie, J, Lereclus, D, Baum, J, Dean, D.H (1998) Revision of the nomenclature for the Bacillus thuringiensis pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62, 807813.
  • [62]
    Beard, C.E., Ranasinghe, C, Akhurst, R.J (2001) Screening for novel cry genes by hybridization. Lett. Appl. Microbiol. 33, 241245.
  • [63]
    Chestukhina, G.C., Kostina, L.I., Mikahilova, S.A., Tyurin, F, Klepikova, S, Stepanov, M (1994) Production of multiple δ-endotoxins by Bacillus thuringiensis: δ-endotoxins produced by strains of the subspecies galleriae and wuhanensis. Can. J. Microbiol. 40, 10261034.
  • [64]
    Rabilloud, T (2002) Two-dimensional gel electrophoresis in proteomics: old, old-fashioned but it still climbs up the mountains. Proteomics 2, 310.
  • [65]
    Tabashnik, B.E (1992) Evaluation of synergism among Bacillus thuringiensis toxins. Appl. Environ. Microbiol. 58, 33433346.
  • [66]
    Ben-Dov, E, Manasherob, R, Zaritsky, A, Barak, Z, Margalith, Y (2001) PCR analysis of cry7 genes in Bacillus thuringiensis by the five conserved blocks of toxins. Curr. Microbiol. 42, 9699.
  • [67]
    Wasano, N, Ohba, M, Miyamoto, K (2001) Two δ-endotoxin genes, cry9Da and a novel related gene, commonly occurring in Lepidoptera-specific Bacillus thuringiensis Japanese isolates that produce spherical parasporal inclusions. Curr. Microbiol. 42, 129133.
  • [68]
    Porcar, M, Iriarte, J, Dumanoir, V.C., Ferrandis, M.D., Lecadet, M.M., Ferré, J, Caballero, P (1999) Identification and characterization of the new Bacillus thuringiensis serovars pirenaica (serotype H57) and iberica (serotype H59). J. Appl. Microbiol. 87, 640648.
  • [69]
    Rang, C. (1996) Etude des relations structure – Fonction chez les endotoxins de Bacillus thuringiensis. University of Montpellier II, Montpellier, 145 pp.
  • [70]
    Barloy, F, Lecadet, M.M., Delécluse, A (1998) Distribution of clostridial cry-like genes among Bacillus thuringiensis and Clostridium strains. Curr. Microbiol. 36, 232237.