The Bacillus cereus sensu lato group encompasses several ubiquitous, endospore-forming Gram-positive bacteria including B. cereus, B. anthracis, and Bacillus thuringiensis (Bt). Despite their close phylogenetic relationship, their pathogenic properties are contrasting, since B. anthracis causes anthrax disease, B. cereus produces food poisoning in mammals, and Bt is a well-known pathogen of invertebrates (Helgason et al., 2000). Phylogenetic analysis showed that among the species of B. cereus sensu lato group, B. cereus and Bt are the closest bacteria. The main differences between them are found in their mobile genetic elements (MGE), such as insertion sequences, prophages, and plasmids. It was proposed that the acquisition of such MGE in B. cereus and Bt was facilitated by the absence of a functional CRISPR-Cas systems in many of the B. cereus sensu lato strains, allowing incorporation of novel genetic elements that may help for their better adaptation to diverse ecological niches (Zheng et al., 2020).

Due to their insecticidal properties, Bt have been exploited worldwide as biological pesticides for the control of multiple insect pests (Sanahuja et al., 2011). These bacteria are characterized by the production of pesticidal proteins (PPs) during their sporulation phase of growth. Most of these PPs accumulate as parasporal inclusions within the mother cell compartment and have been recently renamed in the Bacterial Pesticidal Protein Resource Center (BPPRC¹) as follows: PPs belonging to the three domain crystal proteins retained the mnemonic name of Cry, while proteins related to Cyt family retained the mnemonic name of Cyt. PPs with homology to Etx-Mtx2 family are now named Mpp. The PPs with homology to aegerolysin are Gpp. Those with homology to the Toxin10/Bin family received the name of Tpp and the PPs with a predominantly alpha helical structure are named App (Crickmore et al., 2020). In addition, some Bt bacteria also produce and secrete additional PPs during their vegetative growth, such as proteins related to the vegetative insecticidal protein Vip3 that retained the mnemonic name of Vip, while the proteins related to the catalytic component of Vip2 are now named Vpa, and the PPs related to the Vip1 binding partner from the binary toxin are now recognized as Vpb (Crickmore et al., 2020). All these proteins are encoded in large megaplasmids as individual genes or grouped in pathogenic islands (PAI), accompanied by repeat sequences, insertion elements, and transposases, which may allow a higher recombination rate among diverse Bt strains (Mahillon and Chandler, 1998; Fiedoruk et al., 2017).

In searching novel PPs produced by Bt, that could be useful for pest control in agriculture or farming activities, a large number of Bt strains have been isolated from insect corpses, soil, and phylloplane. Characterization of such strains is not an easy task. Probably, the faster and low-cost approach to screen insecticidal toxins from a Bt strain collection is the traditional PCR strategy, either using a set of universal primers or designed specific primers (Cerón et al., 1994; Cerón et al., 1995; Porcar and Juárez-Pérez, 2003). In addition, also DNA-hybridization and DNA-microarray techniques were shown to be useful for the identification of specific cry genes (Beard et al., 2001; Letowski et al., 2005). However, these DNA-based techniques are limited to detect genes that were previously identified and lack information regarding to their expression. Proteomic analysis of parasporal crystal is more accurate to determine the presence and abundance of PPs expressed in specific Bt crystal inclusions, and such studies have been performed to identify the PPs produced by Bt subsp galleriae VKPMB-1757 and wuhanensis VKPMB-1226 strains (Chestukhina et al., 1994). Recently the advances in sequence strategies have shown that compete genome sequencing is the most efficient strategy to characterize interesting Bt strains and to discover novel PPs. Multiple Bt strains have already been characterized by using Illumina or PacBio strategies in order to obtain their complete genome sequence (Doggett et al., 2013; Liu et al., 2013; Palma et al., 2014; Gao et al., 2015; Cao et al., 2018).

Spodoptera frugiperda is a maize and rice insect pest that has become important worldwide, since this pest has recently migrated from America to Africa and Asia (Sun et al., 2021). This insect shows low susceptibility to Cry1Ac toxin from Bt but it is efficiently controlled with Cry1Fa toxin (Blanco et al., 2010). However, some populations of S. frugiperda have already evolved resistance to Cry1Fa-maize in different countries (Storer et al., 2012; Banerjee et al., 2017; Boaventura et al., 2020). Thus, the identification of additional toxins with high toxicity and no cross-resistance to Cry1Fa is needed to provide effective alternative tools for efficient control of this insect pest. Bt strain GR007 is a strain from our Bt strain collection toxic for lepidopteran insects including S. frugiperda. However, the identity of the PP in GR007 remains unknown. In this work, we have combined third-generation sequencing to obtain the complete genome of GR007 strain and used MS/MS proteomic analysis of its parasporal crystal characterization, in order to identify the expressed proteins. The Bt GR007 genome consists of four replicons: a circular chromosome and three extrachromosomal megaplasmids. Our analysis showed that the two large plasmids harbor multiple PPs arranged in PAIs. We have identified 20 genes encoding for PPs: 10 Cry proteins, two Vip proteins, two proteins that act as binary toxins (Vpa-Vpb), one Mpp protein and a cluster of five genes for insecticidal toxin components (Tc’s) that were originally described in Photorhabdus luminescens (Waterfield et al., 2001).

According to our proteomic analysis, the crystal inclusion of this strain is confirmed by eight Cry proteins. The genes for these Cry proteins were cloned to analyze their individual expression in Bt cells. Only seven Cry proteins were able to form parasporal crystals when expressed in Bt and all of them displayed insecticidal activity against to at least one lepidopteran species. Our work shows that combining these two approaches, whole genome sequence and LC-MS proteomic analysis of crystal inclusions, allows a complete characterization of novel Bt strains containing PPs with potential biotechnological applications.

Whole genome sequence screening of Bt strains has been one of the most successful approaches for PP gene discovery in recent years. In addition, it has provided data for the analysis of the evolution of the Bt genome and analysis of horizontal gene transfer events that have contributed to the evolution of Bt strains (Zheng et al., 2017). Bt bacteria are characterized by the presence of insecticidal proteins that accumulate in crystal inclusions during their sporulation phase of growth and such proteins are codified in different plasmids. It was shown that the B. cereus group selectively inactivated the CRISPR Cas system, which correlates with acquisition of mobile elements, such as different plasmids containing genetic information that could help in the adaptation of these bacteria to diverse environments (Zheng et al., 2020). Bt GR007 strain did not contain a functional CRISPR Cas system, which correlated with its high PP gene content, encoded in different plasmids. Phylogenetic analysis of whole genomes showed that Bt GR007 strain (BioProject: PRJNA736034) is closely related with Bt HD12 strain (BioProject: PRJNA302106). Interestingly, the genome sequence comparison between both strains suggested that HD12 might had a genome rearrangement where information contained in plasmids pGR157 and pGR55, that are present in Bt GR007 strain, recombined to form a single large plasmid named pHD12016 in HD12 strain, loosing cry1Fb and vip3Ag genes during this recombination process.

It is now recognized that different Bt strains harbor several PP that in theory may expand their host specificity by improving its toxic activity against different insect targets (Wang et al., 2020). Bt GR007 strain showed insecticidal activity against lepidopteran insects. This insect specificity correlated with the identification of Cry1 proteins expressed in the crystal inclusions that are specific against lepidopteran insects (Crickmore et al., 2020). However, other PP genes found in PAI-1.3 island of this strain, such as vpa2Ac and vpb1Ca genes and the putative novel mpp gene, suggest that it may also display also coleopteran toxicity, but this hypothesis remains to be studied in the future.

The genomic analysis allowed the identification of multiple PAI islands. The PAI-1.1 is the longest PAI region in this strain, containing seven different cry genes. All of them codify for lepidopteran specific Cry proteins (Cry1Da, Cry1Id, Cry1Ja, Cry1Nb, Cry1Ab, Cry1Bb, and Cry1Hb) (Crickmore et al., 2020). This PAI-1.1 also codifies for a cluster of five insecticidal toxin components (Tcs). The Tcs proteins were originally identified in enterobacteria P. luminescens and Xenorhabdus nematophila, which are symbiont of nematodes (Waterfield et al., 2001). These Tcs proteins are pore-forming toxins that kill lepidopteran and dipteran insects. It was previously shown that tcs genes may be found in other bacteria such as Serratia entomophila that is also an insect pathogen or Yersinia pestis, that is mammalian pathogen transmitted by an insect vector (Waterfield et al., 2001). It was also reported that some Bt strains may harbor these genes and that tcaA and tcaB genes were expressed during the infection of gypsy moth larvae by these Bt strains (Blackburn et al., 2011). In the case of GR007 strain, it still remains to be demonstrated if these Tcs proteins participate in toxicity.

The PAI-1.2 contains only two genes encoding for PPs (Vip3Af and Cry2Ad), which are also recognized for their toxicity against lepidopteran insects (Estruch et al., 1996; Liao et al., 2015). In contrast the PAI-1.3 codifies for proteins that have been associated with toxicity against different insect orders, like the PP that shows 34.6% identity with Mpp4Aa from Lysinibacillus sphaericus that belongs to the Mtx2-protein family active against mosquito larvae (Rey et al., 2016). In Bt the Mtx2 like-toxins (now named as Mpp or Gpp) are toxic to other insect orders, such as the Mpp64Ba and Mpp64Ca toxins (previously known as Cry64Ba and Cry64Ca) that are highly active against hemipteran pests (Liu et al., 2018), or Gpp34Aa/Tpp35Aa proteins (previously known as Cry34/Cry35) that showed toxicity against coleopteran pests (Kelker et al., 2014). However, the new PP found in GR007 strain was not classified in the BPPRC web site (see text footnote 1), since we lack information of its pesticidal activity and its identity with other Bt PPs is lower than 35%. This PAI-1.3 island also contains genes that codify for the binary Vpa2Ac and Vpb1Ca toxins, that were previously shown to be toxic to coleopteran larvae (Bi et al., 2015). However, all these PP proteins were not detected in the parasporal crystal inclusion of Bt GR007 so their expression and insect specificity remains to be analyzed. Finally The PAI2.0 island contains three genes encoding for PP that have been shown to be toxic to lepidopteran larvae (Cry1Fb, Cry1Ka, and Vip3Ag) (Koo et al., 1995; Estruch et al., 1996; DaSilva et al., 2016). The Vip3A proteins are expressed in the vegetative phase of growth and are secreted into the medium. The expression of Vip3Ag and Vip3Af in GR007 strain remains to be characterized.

All PAIs from GR007 strain showed multiple Tnp’s sequences (blue arrows in Figure 1B). It has been proposed that Tnp sequences have contributed to the transfer of PP genes and their recombination among differed Bt strains to generate the great variability of these proteins (Mahillon and Chandler, 1998).

Analysis of non-toxin virulence factors produced by Bt GR007 strain showed that most of them are codified in the chromosome of this bacterium (Table 3). It has been proposed that Chitinases, collagenases, cell wall hydrolases, phospholipases, and proteases such as metalloproteases, camelysin (a surface metalloproteinase), bacillolysins (a neutral proteinase), and subtilisin (a serine protease) may be important for efficient larval body utilization during the infection process (Malovichko et al., 2019). Enhancin and collagenase proteins may also participate in the utilization of insect tissues during the last stages of saprophytic colonization (Malovichko et al., 2019). The cytolysins such as hblL2, hblL1, and hblB, a three-component hemolytic complex, was shown to be produced by B. cereus sensu lato strains and has been implicated as a cause of diarrhea associated with food poisoning (Worthy et al., 2021). These cytolysin genes and other cytolysins such as hly II and hly III, xhlA, tlyA were found in the chromosome of GR007 strain. Finally, the internalin proteins were initially described in Listeria monocytogenes, playing an important role in cell invasion (Dramsi et al., 1997). In the case of B. cereus it was proposed that internalins may play a role during pathogenesis and that they are induced in insects after oral ingestion (Fedhila et al., 2006). All these virulence factors may potentially add competitive advantages to the Bt strains, improving their toxicity to the target insect as previously suggested (Malovichko et al., 2019).

Here, we cloned the cry genes found in GR007 strain and analyzed their expression and toxicity against two lepidopteran insects, M. sexta and S. frugiperda. The toxicity of the Bt GR007 strain to these insect pests was relatively low compared to the toxicity of some Cry proteins expressed individually (Table 5). Seven proteins were successfully expressed in Bt cells. The toxicity of Cry1Ab against M. sexta (LC₅₀ 8.6 ng/cm²) was relatively low when compared to previous reports that have showed LC₅₀ values two or threefold lower (Gómez et al., 2018; Pacheco et al., 2018). In contrast, the toxicity of this protein against S. frugiperda showed similar values to other reports (Gómez et al., 2018). However, it is important to mention that it was shown that different S. frugiperda populations showed great variability in their susceptibility to Cry1Ab, ranging from not susceptible at all, up to highly susceptible (Graser et al., 2017; Gómez et al., 2018; Figueiredo et al., 2019). The Cry1Bb protein showed to be highly active against M. sexta with LC₅₀ value of 5.4 ng/cm² and this protein was the most active toxin against S. frugiperda among all tested proteins in this work, with a LC₅₀ value of 128.5 ng/cm² (Table 5), showing two times higher toxicity to S. frugiperda than previously reported (Luo et al., 1999). The toxicity of Cry1Da against M. sexta and S. frugiperda was similar to previous reports (Höfte and Whiteley, 1989; Wang et al., 2019). In the case of Cry1Fb, Cry1Hb, and Cry1Ka there are no reports previous in the literature about the toxicity of these proteins against these target insects. It was only shown that Cry1Fb was active against Agrotis ípsilon and was not toxic against Heliothis virescens (de Maagd et al., 2003; Karlova et al., 2005); and that Cry1Ka was active against Arfogeia rupae, but not active against Plutella xvlostella, Spodoptera exigua, or Bombyx mori (Koo et al., 1995). Finally the toxicity of Cry1Ja was previously analyzed against S. exigua showing high toxicity (Herrero et al., 2001; Ibargutxi et al., 2006) and the reported toxicity of Cry1Ja against M. sexta was much higher than the LC₅₀ value reported in this work (Herrero et al., 2001).

Spodoptera frugiperda has been shown to be highly susceptible to Cry1Fa and transgenic maize expressing Cry1Fa has been shown to be effective in controlling this insect pest. However, S. frugiperda has evolved resistance to Cry1Fa-maize in Puerto Rico, United States, Brazil, and Argentina (Storer et al., 2012; Farias et al., 2014; Chandrasena et al., 2018). Thus, additional Cry proteins that show no cross-resistance to Cry1Fa are likely to provide tools to counter resistance of this pest to transgenic maize. In the case of Cry1Bb, it was shown that this toxin and Cry1Fa share a binding site in the brush border membranes of S. frugiperda suggesting that Cry1Bb and Cry1Fa should show cross-resistance in this insect species (Luo et al., 1999). However, the Vip3Aa has been shown to be effective against the Cry1Fa resistant populations (Welch et al., 2015). The toxicity of the other PP proteins codified in Bt GR007 strain such as Cry1Id, Cry1Nb, Cry2Ad, the new Mpp, Vpa2Ac, Vpb1Ca Vip3Af, and Vip3Ag remains to be analyzed against different target insects including insect populations that have evolved resistance to other Cry proteins in the field.

Finally, we analyzed the promotor region of GR007 PP genes. We found that cry1Da, cry1Ja, cry1Bb, cry1Hb, cry1Fb, and cry1Ka genes have promotor regions highly similar to cry1Ab gene suggesting that all of them may be regulated by Sig^E and Sig^K factors. Previously the regulation of some cry genes from HD12 strain was analyzed showing that the expression of cry1Ae, cry1Bb, cry1Fb, and cry1Ja genes is regulated by both sigma factors, while cry1Da was only regulated by Sig^E. However, the problem was that the upstream sequence of cry1Da gene was not fully analyzed (Song et al., 2017). They did not included the −35 region of Sig^K promoter, resulting in the loss of Sig^K function when fused to LacZ gene (Song et al., 2017). Here we show that the cry1Da gene that is present in GR007 strain has the complete promoter region including −35 and −10 DNA sequence for binding of both sigma factors, suggesting that its expression is determined by both sigma factors. Our analysis also showed that promotor regions from cry1Nb, cry1Id, and cry2Ad showed different sequences, suggesting that they do not share the same regulation as the other cry genes present in this strain (Supplementary Figure 3). Blast analysis showed that the promotor region from cry2Ad is similar to the upstream sequence of cry2Ab gene. The promotor region of cry1Id is similar to the upstream sequence of cry1Ia, while the promotor region of cry1Nb has no similarities to any other sequence including all known upstream sequences from other cry genes.

Overall, our data indicate that in nature certain Bt strains contain and express multiple PP genes codifying for proteins that display activities against different target pest. These strains could be excellent candidates to generate novel formulations. The information provided here regarding the genome of Bt GR007 strain will help with the understanding Bt toxin gene diversity and regulation of its PPs.

The data for this study can be found in BioProject: PRJNA736034 and Gene Bank Accession numbers CP076539, CP076541, CP076540, and CP076542.

JS collected the sample. SP, IG, and JS conducted the experiments. SP and MC analyzed the data. AB and MS conceived of the idea and designed the study. SP, MS, and AB participated in drafting the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.