Resistance of Lepidopteran Pests to Bacillus thuringiensis Toxins: Evidence of Field and Laboratory Evolved Resistance and Cross-Resistance, Mode of Resistance Inheritance, Fitness Costs, Mechanisms Involved and Management Options

Bacillus thuringiensis (Bt) toxins are potential alternatives to synthetic insecticides for the control of lepidopteran pests. However, the evolution of resistance in some insect pest populations is a threat and can reduce the effectiveness of Bt toxins. In this review, we summarize the results of 161 studies from 20 countries reporting field and laboratory-evolved resistance, cross-resistance, and inheritance, mechanisms, and fitness costs of resistance to different Bt toxins. The studies refer mainly to insects from the United States of America (70), followed by China (31), Brazil (19), India (12), Malaysia (9), Spain (3), and Australia (3). The majority of the studies revealed that most of the pest populations showed susceptibility and a lack of cross-resistance to Bt toxins. Factors that delay resistance include recessive inheritance of resistance, the low initial frequency of resistant alleles, increased fitness costs, abundant refuges of non-Bt, and pyramided Bt crops. The results of field and laboratory resistance, cross-resistance, and inheritance, mechanisms, and fitness cost of resistance are advantageous for predicting the threat of future resistance and making effective strategies to sustain the effectiveness of Bt crops.


Introduction
Bacillus thuringiensis (Bt)-transformed plants are globally used as cultivated crops [1,2].Bt toxins are environmentally benign and more selective on target pests; moreover, crop dependency upon the application of conventional insecticides is also reduced, thus they have revolutionized the agriculture sector [3,4].The Bt toxins ingested by insects are proteolytically activated in the insect gut lumen and the activate toxins interact with membrane proteins in lepidopteran midgut cells, causing the formation of ionic pores in the cell membrane and inducing cell death [5].As Bt crops help to reduce the incidence of attack by lepidopteran pests, the yield potential of cotton and maize increases many fold and higher profits may incur [6][7][8].The Bt crops have been cultivated on >560 million hectares area worldwide since 1996 [2,9].Bt cotton has been planted on 69 million hectares of area globally, with 79-95% area in Australia, China, India, Pakistan, and the United States [10], while 67% of the total area of Bt maize is planted in the United States [1].
Some lepidopteran insect pests present tolerance to Bt toxin effects and develop resistance, which may be a risk to the sustained success of Bt cotton and maize.The Bt resistance in lepidopteran pests of cotton and maize has been reported globally and is summarized in this review.We described the observed pattern of field-evolved and laboratory-selected resistance, the mechanism of resistance, and the fitness costs of resistance reported against Bt toxins during the period of 2000-2018.Due to time constraints, we cannot extend this review further at present.However, we plan to write another review paper in the future, focusing on the resistance scenario from 2019 to 2024 in various pests against Bt toxins.In this review, 24 cases of field-evolved resistance to 12 Bt toxins and 37 instances of crossresistance to 28 Bt toxins in nine lepidopteran species were reported.Additionally, the review described 63 cases of laboratory-selected resistance and 58 cases of inheritance of resistance to 15 toxins in 11 species, 57 cases of the mechanism of resistance to 13 toxins in 11 species, and 36 reports of the fitness cost of resistance to 11 toxins in 12 species.These results may provide insights to develop and improve proactive strategies for the resistance management of various pest species to Bt toxins.

Bt Resistance Management
After the first report of Bt resistance by McGaughey [11] researchers worldwide have attempted to elucidate the cross-resistance, genetics, fitness costs and mechanisms triggering the Bt resistance in some Lepidoptera pests.These efforts are vital if novel strategies are to be developed to use currently available Bt toxins effectively and efficiently for the prevention of resistance development.
Many strategies for Bt resistance management have been proposed, including high toxin dose in combination with refugia, pyramiding/stacking (use of multiple toxins simultaneously), and crop rotations.High dose-refuges that contain toxin-free host plants near the Bt crops might play an important role in insect resistance management (IRM) plans.Using a 20% refuge of non-transgenic crops treated with non-Bt foliar insecticides or a 4% non-transgenic crop, delays the development of resistance in cotton and maize pests [12,13].These refuges allow the influx of susceptible individuals that dilute the occurrence of resistance genes.The success of this approach occurs in the recessive mode of inheritance in combination with pest biology [14].Ultra-high Bt toxin doses in combination with the refugia may also kill the heterozygote insects carrying the resistant alleles.So, the individuals would be more likely to mate with non-resistant insects, with a possible decrease in homozygote individuals [14].Data about cross-resistance, mode of inheritance, fitness costs, and mechanisms of resistance involved provide assistance for the detection, monitoring, modeling, and future risk assessment of resistance [15].
Pyramiding Bt toxins is a recent technology used to delay the inception of resistance for plants having transgenic genes of two different toxins, when cross resistance is absent [2,16].Other promising factors that favor the pyramiding crop are low initial resistance alleles frequency and fitness costs, associated with a recessive inheritance [16,17].Cross-resistance between Cry 1Ac and Cry 2Ab is unlikely to occur due to the different amino acids and binding sites in the insect midgut [18,19].So, pyramiding the most feasible Bt toxins that are expected to have no cross-resistance is a good approach to IRM practices.This technology provides a broader spectrum of activity against lepidopterans, enhances control of caterpillars that are susceptible to single Bt toxin transgenic plants, and provides better opportunities for managing insect resistance to Cry proteins [20].
Reducing the selection pressure by restricting the expression of Bt proteins to susceptible plant tissue or at a certain threshold level may also be an important management strategy.Expression of Bt toxins could be induced by wound inducible promoters after insect feeding or by the application of chemicals [21].For example, it has been proposed that cotton plants undergo low yield loss when Bt toxin expression is restricted to only young bolls, which stimulates Heliothis larvae to feed on the other plant parts [22].
Resistance development can be mitigated by using comprehensive and systematic resistance management strategies.With pyramiding and high dose-refuge, especially in the case of a recessive genetic basis of resistance, non-insecticidal control (pheromone mating disruption, crop rotation, and suppression of overwintering stages of insects though soil or stalk disruption) may be added in the transgenic and non-transgenic cropping systems to manage and monitor pest population growth.Thus, crops that express Bt toxins are ensured to be a valuable tool in IRM programs.

Field-Selected Resistance to Bt Toxins
Field-selected resistance is a genetic-based decrease in susceptibility of a pest population due to its exposure to the toxin in the field [23].Insecticide resistance has been suggested to be a pre-adaptive phenomenon in the sense that, before an organism is exposed to the toxin, there are few individuals in the population with one or more resistance genes that allow them to survive the toxin exposure [24,25].Then, field resistance occurs after continuous exposure of the insect population to a toxin, which may increase the number of resistant alleles in the next generations.Insect pest control concerns due to field-selected resistance vary from none to severe according to the number of resistant individuals, the extent of the resistance increase, the density and geographical distribution of resistant populations, and the accessibility of alternative control methods [26].
Monitoring of Bt field-selected resistant populations may be conducted in laboratory tests to define baseline data and to ensure effective proactive management of insecticide resistance in target pests [26].For this purpose, the field populations of different lepidopteran pests of cotton and corn are collected and exposed to different Bt crystals (Cry) and vegetative insecticidal proteins (Vip).These field populations are sampled from Bt and non-Bt crops to detect the field-selected resistance.The resistance was classified into categories, and the resistance ratios (RRs) were estimated as [27]: RR = LC 50 of field strain/LC 50 of reference susceptible strain where LC 50 is the lethal concentration of toxin, killing 50% of insects tested under a controlled environment.
Based on the RR, we defined field resistance into three categories: (i) no to very low resistance, if the RR of the tested population is 1-10, (ii) low to moderate resistance, if the RR is 11-50, and (iii) high to very high resistance if the RR > 51.In the very low resistance, the amount of resistant insects in the population is not sufficient to reduce the efficacy of the Bt toxins for pest control.On the other hand, in the low-moderate resistance, there is a decrease in the efficacy of the Bt toxins to control field populations, indicating the necessity of resistance management.Finally, if the field population reaches very high resistance, the efficacy of the Bt toxins in the field fails and the resistance management strategy must be improved.
In this study, we reviewed 24 cases of resistance monitoring in nine species of main lepidopteran pests that were targeted by 12 Bt toxins in nine countries during the years 2000-2018 (Table 1).
We found reports in China of very high resistance to Cry1C for Plutella xylostella [28], no-very low resistance to Cry1Ac for Pectinophora gossypiella [29][30][31] and very low-moderate resistance to Vip3Aa11 for Helicoverpa armigera [32].In India, P. gossypiella presented nomoderate resistance to Cry1Ac and Cry1Ac+Cry2Ab2 [33] and H. armigera showed no-very high resistance to Cry1Ac [34,35] and no-very low resistance to Cry2Ab [36], whereas in Pakistan, H. armigera showed very high resistance to Cry1Ac [37].In West Africa, this pest presented very low-moderate resistance to Cry1Ac and no-very low resistance to Cry2Ab2 toxins [38].In the USA, moderate resistance to Cry2Ab, very high resistance to Cry1Ac, no-high resistance to Vip3A, very low resistance to Cry1Fa, and moderate resistance to MVP II were reported for H. zea [39][40][41][42].Very low resistance to Cry1Ac and very low-very high resistance to Vip3A were reported for H. virescens [40,41].The Spodoptera frugiperda field populations have been reported to develop high-very high resistance to Cry1F [43][44][45] and low resistance to Cry2Ab2 [46] in the USA.In contrast, no-very low resistance to Cry1F was found in different S. frugiperda populations from the USA [47].For Diatraea saccharalis, very low resistance to Cry1F, moderate-very high resistance to Cry1Ac, and high-very high resistance to Cry1Ab and Cry1Aa have been reported [48].In Brazil, no-very low resistance to Vip3Aa20 and Cry1Ac for H.zea and H. armigera [49][50][51], very low-very high resistance to Cry1F for S. frugiperda [52][53][54], and no-very low resistance to Vip3Aa20 for S. frugiperda and D. saccharalis have been reported.Ostrinia furnacalis and O. nubilalis have shown no-low resistance to Cry1Ab maize in the Philippines, Germany, and Europe [55][56][57].
Those different levels of resistance to the same Bt toxin by different populations of the same species, to different toxins, and among different pests reinforce the necessity of laboratory monitoring of field populations to the bio-insecticide to predict possible selection of field resistance and the use of pesticide management to avoid very high resistance.RR (resistance ratio) = LC 50 of field strain ÷ LC 50 of susceptible strain.

Laboratory-Selected Resistance to Bt Toxins
Genetic-based increases in resistance to an insecticide in a population due to its continuous exposure in the laboratory is termed laboratory-selected resistance [23].The laboratory-selected resistant strains are useful for evaluating pesticide risk assessment, cross-resistance, stability, fitness costs, and modes of resistance, including genetics, biochemical, and molecular mechanisms.
We reviewed 63 studies of laboratory resistance published in peer-reviewed journals for 11 species of major lepidopteran pests on cotton and corn that were targeted by 15 Bt toxins during the years 2000-2018 (Table 2).We searched the literature by using PubMed and Google Scholar, checking the bibliographies of all articles found with the keywords of laboratory-selection, Bt resistance, Bt toxins, and lepidopteran pests.
High resistance to Cry1Ac after eight generations [105], very high resistance to Cry1Ac after 15 generations [106] in P. gossypiella; low resistance to Cry1Ac after five generations [107], high resistance after 10-19 generations [108,109], and very high resistance after 15 generations [110] in H. armigera have been reported for Indian populations.
These studies indicate that the proportion of individuals carrying the resistance alleles increases according to Bt toxin exposure.These studies also describe that laboratoryselected strains exhibited higher level of resistance than those from the field, which may be due to population strains under continuous selection pressure.So, selection pressure may increase the rate and frequency of resistance development [119].All non-behavioral resistance mechanisms involve changes in physiology that are the result of selection for resistance alleles [120].

Cross-Resistance to Other Bt Toxins
Cross-resistance is defined as resistance due to a single mechanism and/or mode of action that provides resistance to a different insecticide.We found 37 cases of crossresistance for nine species of major lepidopteran pests of cotton and corn that were targeted by 28 Bt toxins from 2000 to 2018.
All together, these findings reveal that different pest strains resistant to a Bt toxin may develop cross-resistance to different toxins, indicating possible multiple resistances.

Moderate and Low Level of Cross-Resistance to Other Bt Toxins
The moderate cross-resistance to Cry1Ac was reported in a Cry1Ab-R strain of P. xylostella from Malaysia, to Cry1Ac in a Cry1Ab-R strain of D. saccharalis [99], to Cry1Ab, Cry1Ac in a Cry1F-R strain of S. frugiperda [102], and Cry1Ab in a Cry1Ac-R strain of H. zea [42] from the USA.Moderate cross-resistance was conferred in Cry1Ab-R, Cry1F-R, and Cry1Ah-R strains of O. furnacalis to Cry1Ac [70], Cry1F, Cry1Ab, and Cry1Ac [70][71][72][73], and a Cry1Ac-R strain of H. armigera to Cry1Ab, and Cry1Aa [74,75,82] in China (Table 4).A low level of cross-resistance was observed in China for Cry1Aa, Cry1Ab and Cry1F in a Cry1Ac-R strain of P. xylostella [68], to Cry1F in a Cry1Ac-R strain of O. furnacalis [73], to Cry1Aa, and Cry1Ab in a Cry2Ab-R strain of H. armigera [82].A Bt kurstaki-R strain of P. xylostella from Malaysia showed a low level of cross-resistance toCry1Ab [65].Similarly, Cry1F-R and Cry2Ab2-R strains of S. frugiperda also exhibited low level of cross-resistance to Cry2Ab2 and Cry2Ae in Brazil and the USA, respectively [101,115] (Table 4).
Ostrinia furnacalis Helicoverpa armigera a Resistance ratio, calculated as LC 50 of tested strain/LC 50 of susceptible strain.

Mechanisms of Bt Resistance
Variation in any step of the toxin mode of action may result in reduced susceptibility to Bt toxins.This also contributes to conferring cross-resistance among toxins [125].Therefore, the success of resistance management is also dependent upon the biochemical mechanisms of resistance to detect the resistance genes in the field populations.There are different mechanisms involved in the evolution of Bt resistance, including reduced binding sites, mutated binding sites, altered proteolysis, or even behavioral changes.These multiple mechanisms of resistance correspond to genetic changes for insect survival after insecticide exposure.Thus, resistance to the insecticide can be expressed in the next generation [126].Fifty-seven studies of 13 Bt resistant lepidopteran species have demonstrated different mechanisms of resistance to 11 Bt toxins (Table 7).
In China, cadherin gene and aminopeptidase-N mutations were linked to resistance to Cry1Ac [74,81,142,143], multiple mechanisms were linked to resistance to Cry1Ac [76,77,144], and reduced binding sites were linked to resistance to Cry1Ac [145] for H. armigera.Cadherin-like mutations were linked to resistance to Cry1Ac, and Cry2Aa for P. gossypiella [5], S. exigua [146,147], altered expression of ALP and ABCC genes were linked to resistance [148] to Cry1Ac for P. xylostella.
These findings reveal that different insects have different mechanisms, including multiple mechanisms of Bt toxins, indicating that further studies are necessary for the comprehension of the molecular resistance mechanisms in different species and populations.

Inheritance of Resistance to Bt Toxins
The expression of resistant genes in heterozygotic individuals can confer dominance of resistance to insecticides [23].If resistance is completely dominant, only one parental individual needs to possess the trait for it to be fully expressed in the offspring.Thus, completely dominant resistance alleles rapidly become established in the populations, which are hard to manage.If resistance is incompletely dominant, the trait will be partially expressed in heterozygous offspring.This type of resistance can be managed with higher expression of the toxin or rotational uses of different Bt toxins where there is no cross-resistance [163].If the resistance is completely recessive, only offspring that are homozygous for the resistance allele will be resistant and the resistance cannot rapidly be established in the population if appropriate management practices are followed, because the persistence of heterozygotes may be promoted (e.g., by alternative hosts or refuge) and they may be easily killed under field conditions.

Conclusions
In 22 studies, a high cross-resistance, 17 studies, a moderate cross-resistance, 21 studies, a very low cross-resistance, and 22 studies, the absence of cross-resistance were observed.Most of the resistant populations showed recessive to incompletely dominant types of resistance.The absence of cross-resistance to different Bt toxins in different populations is compatible with the idea of multiple resistance mechanisms involved.In the case of the absence of cross-resistance and a recessive mode of inheritance, the pyramiding of two different toxin genes, refuges, or high doses helps delay the development of Bt resistance.The inheritance and cross-resistance patterns extremely affect the strategies for managing Bt resistance.A high-dose strategy only can slow resistance when the resistance is recessive; in another case, if the resistance is dominant, then this strategy accelerates the development of resistance.
Concisely, many Bt toxins show no/very low levels of resistance, a lack of crossresistance, recessive inheritance, and increased fitness costs, hence they are still a good alternative to synthetic insecticides for the control of lepidopteran pests.Knowledge of different management practices in combination with the rate of resistance development is essential to develop successful resistance management strategies.In addition, evaluation of already implemented management practices and awareness of pest biology will surely lead to beneficial decisions for the management of Bt resistance.The midgut is an important site for insecticidal action; therefore, it is likely that future attempts to develop insecticidal compounds will increasingly use the insect midgut as a target organ.Consequently, the selective toxicity of different Bt toxins is determined by the receptors in the midgut.Most studies of the fitness costs of Bt resistance have been conducted on moths.More studies on economically important beetles are needed.
Although this review provides substantial data to understand the selection of strains resistant to Bt toxins, some factors with the potential to drive resistance remain poorly understood, including the different modes of application of the toxin in the field, how

Table 1 .
Field-evolved resistance to Bacillus thuringiensis toxins in different lepidopteran pests from 2000-2018.

Table 2 .
Laboratory-selected resistance to Bacillus thuringiensis toxins in different lepidopteran pests from 2000-2018.

Table 3 .
High to very high levels of cross-resistance to Bacillus thuringiensis toxins in resistant strains of different lepidopteran pests from 2000-2018.Very high levels of cross-resistance to other Bt toxins in the different Bt resistant strains Resistance ratio, calculated as LC 50 of tested strain/LC 50 of susceptible strain.
High levels of cross-resistance to other Bt toxins in the different Bt resistant strains a

Table 4 .
Low to moderate levels of cross-resistance to Bacillus thuringiensis toxins in resistant strains of different lepidopteran pests from 2000-2018.Moderate levels of cross-resistance to other Bt toxins in the different Bt resistant strains Low levels of cross-resistance to other Bt toxins in the different Bt resistant strains a Resistance ratio, calculated as LC 50 of tested strain/LC 50 of susceptible strain.

Table 5 .
Very low cross-resistance to Bacillus thuringiensis toxins in resistant strains of different lepidopteran pests from 2000-2018.

Table 6 .
Absence of cross-resistance to Bacillus thuringiensis toxins in resistant strains of different lepidopteran pests from 2000-2018.

Table 7 .
Mechanisms conferring resistance to Bacillus thuringiensis toxins in different lepidopteran pests from 2000-2018.

Table 8 .
Inheritance and type of resistance to Bacillus thuringiensis in different lepidopteran pests from 2000-2018.

Table 9 .
Fitness costs associated with Bacillus thuringiensis resistance in different lepidopteran pests from 2000-2018.