Sampling sites

From April 2011 to May 2012, different maize commercial lots, planted with GM and non-transgenic, were localized and classified in four regions: 1) Coahuila State, composed by the municipalities of Torreon, San Pedro de las Colonias, and Saltillo; 2) San Luis Potosi, integrated by the municipality Graciano Sánchez; 3) Chiapas, composed by the municipalities of Uixtla-Mazatan and Zuchate; and 4) Tamaulipas, integrated by the municipality Rio Bravo. In all maize lots (Table 1), lepidopteran larvae of Heliothis and Helicoverpa from different larval instar, and inhabiting the maize cob, were collected by random sampling of plants within plots. Some larvae were preserved in plastic bottles containing ethyl alcohol (70%), and other larvae were allowed to continue their life cycle to corroborate their identification at the species level.

Taxonomic identification of insects

Lepidoptera larvae were identified using the dichotomous key reported by ^{Stehr (2005)}. The larvae, collected and maintained in alcohol, were divided into two groups, according to their morphological characteristics and corresponding to one of the identified species. Then, these larvae were used for DNA isolation. On the other hand, the subgroups allowed to continue their life cycle were placed in plastic containers with moist soil and, thus, promoted its metamorphosis to complete the pupal stage and then to adult under laboratory conditions for taxonomic species level corroboration.

DNA isolation

From each larva, a tissue sample (0.1 g) was collected for DNA isolation. Subsequently, the larval fragment was placed in liquid nitrogen for 12 h. Insect tissue was macerated using a cell disruptor MP (6 m/s by 60 sec) and the MPbio matrix D. DNA isolation was performed using the AxyPrep Blood Genomic DNA Miniprep Kit procedure. DNA integrity was determined by gel electrophoresis in 1% agarose. DNA quantification was carried out in a plate reader Epoch ™ Microplate Spectrophotometer BioTek brand with a Take3 ™ attachment plate, 8 X 2 model, with the program DNA quantification Gen5 1.11.

DNA amplification

Polymerase chain reaction (PCR) was performed to amplify the OnBreGalt5 ((-1, 3-Galactosyiltransferase) gene, using the forward Onb3GalT5-F1 (5'CGTGACAATGATGTCGTTCAA3') and reverse Onb3GalT5-R1 (5'TGCTGCGGCACTAAGCCCAC3') primer, which were previously reported by ^{Coates et al. (2008)}. The PCR was performed with a final volume of 23 µl, integrated as follow: 14.5 ml of sterile deionized water, 2.5 ml of 10X buffer, 1 µl of MgCl2 (50 mM), 0.5 ml of dNTP's (10 mM), 2 ml of each primer (10 pM), 0.5 ml of Taq polymerase (5 U/µl) and 2 ml of the DNA sample (100 ng/µl). The PCR amplification program was: denaturation at 95 °C for 2.30 min, annealing at 69.7 °C for 0.30 min and polymerization at 72 °C for 1 min for 40 cycles using a thermocycler Px2 Thermal cycle. The amplified products were visualized using gel electrophoresis 1.5% agarose. The Marker 100 bp DNA Ladder (Invitrogen ™) was used as reference.

Statistical analysis

The amplified bands in each case were codified as presence (1) and absence (0). The following parameters were estimated with sample data codified with this binary code: allele frequencies, genotype frequencies, genetic balance of the population, and statistical Wright (F_IS, F_IT and F_ST). Then, genetic parameters were used to compare intra-species and inter-species diversity of the insect populations collected on transgenic and non-transgenic maize. All analyses were performed using the Statistical Package Genepop (4.0.10).

Lepidoptera insect species on transgenic and non-transgenic maize

Taxonomical identification of the Lepidoptera population associated to GM and Non-GM maize from different Mexican areas showed that H. zea, and H. virescens were the species damaging the maize ear (Table 2). The larva of Heliothis virescens (Fabricius) was characterized by having a variable body coloration (yellow to green in dark tone), 30 mm to 45 mm long, microspines in the body pinacula. The pinacula are generally large and conical and more pronounced in the abdominal segments I, II and VIII; in addition, in the I and II segments the presence of thin microspines was observed. The mandible had retinaculum. The adult body is light brown with three dark diagonal stripes on the forewings. H. zea was characterized by having a mandible without retinaculum and lack of microspines on the body pinacula. The adult is a straw-yellow moth with a dark spot, although sometimes it is not very well defined. Posterior wings were clearer than the previous ones and with a dark band at the apical end.

The distribution of the collected larvae depending on the maize condition and location is showed in Table 3. As expected, a higher proportion of larvae attacking maize was found in the non-transgenic maize as opposed to the transgenic maize (4:1 times). In both locations, where transgenic maize with Bt genes were planted, Lepidoptera insects attacking ears were observed, which suggests that this insect species had acquiring resistance to Cry toxins from Bt, because some of these larvae surpass the five-instar step. Insect resistance to Bt plants has been reported previously (^{Monnerat et al., 2015}; ^{Tabashnik et al., 2013}), which could reduce the use of these varieties to long-term (^{Carrière, Crowder & Tabashnik, 2010}; ^{Gatehouse, Ferry, Edwards & Bell, 2011}). The main mechanisms by which resistance to insect based on Cry toxins is overcome are mutations in any of four genes encoding glycosyltransferases, such as glycoside specific arthropod glycolipids which only occur in these organisms but not in vertebrates (^{Deist, Raush, Fernandez-Luna, Adang & Bonning, 2014}; ^{Peterson, Bezuidenhout & Van den Berg, 2017}). Post-transduction glycosylation of medium stomach epithelial proteins and lipids receptors in insects may be required prior the binding of activated Bt Cry toxins (^{Coates, Sumerford, Hellmich & Lewis, 2007}). The insect resistance by loss of carbohydrate modification is relevant in resistance to multiple Bt toxins (^{Tay et al., 2015}).

H. virescens was the most frequently observed insect species on maize ears, independently of the transgenic condition (Table 3). The H. virescens frequency on non-transgenic maize was up to 8-fold than H. zea frequency and, in transgenic maize, this frequency deceased up to 2-fold. These differences may be because the transgenic varieties expressing Cry protein are less effective against H. virescens or this insect had acquiring Cry toxin resistance. H. virescens commonly attacks cotton (^{Blanco, Terán-Vargas, López & Abel, 2009}), however, in this study, this species was identified as the most frequent insect pest associated to transgenic and non-transgenic maize ears. In Mexico, transgenic cotton has been cultivated at commercial levels since 1996, whereas experimental transgenic maize plantations were authorized only six years ago. This may explain that H. virescens has been evolving resistance to Bt genes, because it has been exposed to BT toxins for a longer time, although H. zea is the most associated insect on maize ear (^{Zuñiga, Angulo, Rebolledo & Navarro, 2011}). However, in the present study, on non-transgenic maize, H. virescens was found in a higher frequency than H. zea, which suggested that H. virescens, being exposed to transgenic cotton, migrated to maize, expanding its hosts range.

Allelic diversity in Heliothis virescens and Helicoverpa zea

By using the polymerase chain reaction was possible to amplify different bands for genes encoding galactosyltransferases in both H. virescens and H. zea, collected from transgenic and non-transgenic maize. The variation found in the number of amplified bands in both types of maize (Table 4) could be due to mutations that arose with and without selection pressure of the Bt transgenic maize. Moreover, it is interesting that a band of 500 bp was amplified only from DNA isolated from insects collected from GM maize; besides, the frequency of the 250 bp band was higher in these insects. Moreover, bands 300 bp and 600 bp were only obtained from DNA of insects collected in non-GM maize. The main mechanism for resistance to Bt toxins in C. elegans involves a loss of glycolipid carbohydrates. In this case, Bt toxins bind directly and specifically to glycolipids, and this binding is carbohydrate-dependent and relevant to toxin in vivo action (^{Schünemann, Knaak & Fiuza, 2014}). Glycolipids loss has been associated with the loss of at least two genes coding for glycosyltransferases. These enzymes work in the intestine conferring susceptibility to the toxin, given that they are required for interaction of the active toxin with intestinal cells, suggesting that they make an oligosaccharide receptor for toxin (^{Xu, Bi-Cheng, Yu & Sun, 2014}).

In H. zea specimens less bands were amplified (Table 5), But in this case, from DNA of insects collected from GM maize, a different band (450 bp) was amplified, while, from the DNA of H. virescens, collected on non-transgenic maize, some bands (300 bp and 600 bp) were different to those amplified from insects collected on non-transgenic maize. It is important to continue studying the evolution of these bands. All insects were collected feeding on maize ears, so those collected from transgenic maize must have resistance to Cry toxins.

Hardy-Weinberg equilibrium

It was interesting to observe three patterns of the insect species populations, based on Hardy-Weinberg equilibrium (Table 6). Insect populations of H. zea and H. virescens collected in GM maize were found in equilibrium, since the number of homozygotes and heterozygotes observed was like the one expected. This suggests that mating is at random and that gene and genotype frequencies are constant from generation to generation, while natural selection, genetic drift, migration and mutation do not occur (^{Cañon, Cortes, Garcia-Atance, Tupac-Yupanqui & Dunner, 2007}). Furthermore, H. zea and H. viresens populations collected on non-transgenic maize showed a tendency to have an excess of heterozygous and deficiency of homozygous. Different possible causes to this phenomenon have been proposed. One possible cause is that reproductive isolation of these populations may lead to a non-random mating or selective mating, which could cause an increase in homozygotes of the genes involved in the character under the selection for mating (and genes that are in linkage disequilibrium with them) (^{Otto, Servedio & Nuismer, 2008}). Another possible cause is the small population size which causes a random change in genotypic frequencies, and that may be related to sampling (genetic drift) and can be measured by the FIS statistics (^{Cañon et al., 2007}). In addition, the subdivision of the local population into isolated and differentiated reproductive units (Wahlund effect) and the non-random sampling of members from a limited number of families may be other possible causes.

Genetic diversity

Genetic diversity within and between populations of the two species collected on transgenic and non-transgenic maize is shown in Table 7. It was observed that in most of the cases, the diversity within the population was higher than that observed among populations which may be attributed to the different combinations of alleles of individuals of each species. Only in the collected insect attacking the non-transgenic maize in Tamaulipas was an opposite effect observed, where inter-population diversity was higher than that of intra-population, this can be attributed to preference mating among individuals of the same populations. Detection of polymorphism among and within populations can provide information about the genome evolution, the origin of species, and the current state of genetic diversity (^{Maggert, 2012}). Moreover, by determining the intra-population diversity, it is possible to obtain information on the species’ behavior in a specific environment to persist under adverse conditions such as, feeding on a host with Cry toxins (^{Schünemann et al., 2014}).

Wright statistics

The population genetic structure of the two species was inferred by Wright statistics (F_IT, F_ST and F_IS) (^{Wright, 1978}). The F_IS values in the insect populations, collected in both transgenic and non-transgenic, were negatives; the only exception was the insect population collected in non-trangenic maize in Tamaulipas. F_IS estimates the heterozygote deficiency because of the reproductive isolation, which leads to a non-random mating of individuals; in other words, this parameter is a measurement of endogamy coefficient and its values could be between 0 and 1 (^{Wray & Visscher, 2008}). Thus, the data suggest that both insect populations collected in non-transgenic maize from Tamaulipas are non-random mating, indicating a genetic differentiation between these two populations. Negative values found in most of the insect populations indicated that the observed heterozygotes number was higher than the expected, which may be a consequence of the Wahlund effect if the analyzed insects of those populations are the product of recent crossing between animals belonging to genetic different lines, or it could be the result of negative associative mating (^{Cañon et al., 2007}).

The F_ST value, by definition, is the degree of genetic differentiation among populations as a function of allele frequencies (^{Balzarini, Bruno, Peña, Teich & Rienzo, 2010}), and it indicates the proportion of genetic variation in the sub-population relative to the total variation. The observed F_ST values were very low, and this gives an idea of the differences that exist among populations as a result of low genetic drift (^{Jakobsson, Edge & Rosenberg, 2013}).

The F_IT value can be interpreted as a deficiency (1) or excess of heterozygotes (-1) in a population (^{Balzarini et al., 2010}). In this study, it was observed an excess of heterozygotes in the insects’ populations collected from non-transgenic maize at Saltillo, and insects’ populations collected from transgenic maize at Tamaulipas and San Pedro de las Colonias (Table 8).

It has been reported the insect resistance to Bt endotoxins by different mechanisms, which is a challenger to the major bio-pesticide used in traditional formulations and to its genes used in transgenic crops (^{Coates et al., 2008}). Planting of transgenic maize varieties definitively has a high impact on the pest population regulation. A crucial requirement for development of an effective strategy for insect resistance is the knowledge of nature and inheritance mechanisms of this resistance.