1. Introduction
Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae), commonly referred to as Old World bollworm, is a major crop pest. Native to Europe, Asia, Africa, and Australia, it was once confined to the Eastern Hemisphere. In 2013,
H. armigera was reported in Brazil [
1], but it arrived earlier and established between 2006 and 2008 [
2,
3] and spread throughout South America. There is concern that
H. armigera could become established in North America as well, based on its migratory ability and wide host range [
4]. The first record of
H. armigera in Puerto Rico occurred in 2014 [
5] and the species has since become established there. In 2015,
H. armigera was detected in Florida but due to management and subsequent monitoring is considered eradicated [
5]. Larvae of
H. armigera are highly polyphagous and have been recorded on 68 plant families worldwide [
6]. These host plants include economically important crops like maize, tomato, soybean, cotton, and many specialty crops [
7].
H. armigera has rapidly evolved resistance to many classes of pesticides and insecticidal proteins, making management increasingly difficult [
6,
8].
In the western hemisphere,
H. zea (Boddie), the sister species of
H. armigera, is an insect pest of major economic importance.
Helicoverpa zea has independently evolved resistance to pesticides as well as
Bt via strong selection pressure from widespread use [
9,
10]. The introduction of
H. armigera to North America poses unique challenges because the two species are morphologically similar, requiring genitalic dissection to differentiate them, and have been shown to hybridize in both laboratory and natural settings [
8]. Because dissection of individual moths is both difficult and time-consuming, molecular methods provide the most efficient means to identify
H. armigera. In recent years, several molecular identification methods have been developed for detecting
H. armigera both from interceptions and in surveys (e.g., [
11,
12,
13,
14,
15]).
Continual large-scale screening for
H. armigera is essential for management where recent introductions have been noted or the risk of introduction is high. Currently, the preferred method for molecular screening of bulk trap samples from domestic
H. armigera surveys is using the droplet digital PCR methods described by Zink et al. [
13] following bulk DNA extractions of 25–50 individual moths per reaction. This sample size allows for the identification of
H. armigera regardless of sample quality and is efficient for additional screening of individual specimens in the case of a positive ddPCR result, which is carried out using real-time PCR following Gilligan et al. [
11].
The benefits of screening samples with ddPCR are accuracy and sensitivity as well as high tolerance to the presence of PCR inhibitors, all of which are due to the droplet partitioning resulting in small reaction volumes [
13,
16,
17,
18]. Nevertheless, ddPCR is not widely available for insect screening due to the initial equipment expense and technical expertise required for meaningful implementation. Alternatively, real-time PCR is widely available and simple to implement and validate across multiple labs [
19,
20]. Real-time PCR is a sensitive and accurate molecular method with potential for use in screening individual specimens and bulk survey samples for
H. armigera [
11,
14,
21], but a bulk screening assay has not yet been developed for widespread use. The utility of real-time PCR for accurate detection of a single specimen in a bulk sample depends on developing a time and cost-efficient DNA extraction method of sufficient quality to minimize PCR inhibitors. Oliveira et al. [
14] optimized a squish buffer [
22] bulk extraction method by increasing the NaCl concentration in the buffer, determining the ideal centrifugation speed, and adding a post-centrifugation purification step using paramagnetic beads. Despite the clear benefits of reducing false negatives and significantly increasing end Relative Fluorescent Units (RFUs), bead purification increases sample handling, is more costly and time-consuming, and requires specialized equipment and training, all of which will complicate validation and inhibit widespread implementation of this protocol.
Here, we modify a bulk DNA extraction method by simplifying the protocol to minimize bench time and maximize extraction efficiency with the goal of equalizing our method with the real-time PCR and ddPCR efficiency of Oliveira et al. [
14] using the real-time PCR assay as designed and optimized therein. This assay targets the 18S ribosomal DNA (rDNA) and the internal transcribed spacer 1 (ITS1) region to differentiate
H. armigera from
H. zea. The non-coding ITS regions of the rDNA array are highly effective for molecular species identification because the sequences are generally conserved within species with a high amount of variability between species and occur in high copy numbers in the genome [
11,
23,
24]. Using the DNA extraction method detailed here, the bulk DNA extraction portion of the real-time PCR assay created by Oliveira et al. [
14] is simplified so that
H. armigera screening efforts can be accessible to and comparable across multiple labs.
4. Discussion
Due to the threat the introduction of H. armigera to the U.S. poses to food security, a rapid diagnostic method is critical and must be widely adopted to track its spread. Currently, ddPCR is used for bulk screening due to its accuracy and tolerance to PCR inhibitors; however, real-time PCR is a more common, accessible, and less expensive technology that can be adapted to bulk sample screening for H. armigera. The newly optimized DNA bulk extraction method presented here includes the addition of caffeine and RNase A to a single-step extraction technique that significantly improves upon recently published methods by increasing extraction efficiency and reducing processing time and cost per specimen for use with real-time PCR.
The squish buffer DNA extraction method has been shown to be successful for ddPCR [
13] and was used previously to screen bulk samples by high-resolution melt (HRM; [
12]). Using HRM, Perera et al. [
12] were able to successfully screen samples of one
H. armigera in 24
H. zea but results were inconsistent and not easily replicated in other labs, with different equipment, or with different reagents, making validation of this protocol difficult (see [
13]). To develop an effective screening method based on real-time PCR, Oliveira et al. [
14] designed a new assay based on the same locus (ITS1) as Perera et al. [
12] and modified the squish buffer method by increasing the concentration of NaCl, using a slower centrifugation speed, and adding a bead purification step. Combined, these changes led to the consistent detection of a single
H. armigera leg among 50
H. zea legs, a ratio consistent with what is typically used for bulk extractions from survey samples. Bead purification significantly increased detection in bulk samples, but also increased processing time, sample handling, and cost per sample. In order to eliminate the bead purification and increase assay simplicity and speed, in this study we further modified the squish buffer formulation with the goal of increasing bulk DNA extraction efficiency for improved target detection. We found that adding freeze–thaw steps, Proteinase K, Triton X, Tween 20, and substitution of NaCl for KCl did not improve extraction efficiency as measured by Cq value when compared to the control samples. Proteinase K is effective for removing proteins but will act as an inhibitor by interfering with Taq polymerase during PCR if it is not effectively removed from extracted DNA [
29,
30]. The addition of ascorbic acid, CTAB, and SDS produced false negative results (no Cq value) for the three samples (
Table 4), likely because both CTAB and SDS can delay amplification and inhibit PCR even at low concentrations [
31]. Because our goal was to simplify the extraction, we chose not to introduce additional purification steps to remove surfactants or enzymes. The addition of RNase A and caffeine each improved bulk DNA extraction by producing lower Cq values than the O-buffer extraction.
Co-extraction of RNA in DNA extraction protocols leads to a less pure DNA product and the RNA becomes a contaminant in spectrofluorometry, leading to problems when quantifying DNA concentration [
32]. In our case, where PCR is the end goal, RNA remaining in the DNA extract can bind to template DNA and inhibit amplification in PCR [
33]. By adding RNase A to the squish buffer, RNA impurities are removed without introducing PCR inhibitors. Caffeine has been shown to slow protein aggregation [
34], making DNA more available for extraction. Tsabar et al. [
35] showed that caffeine treatment impairs DNA repair and degrades restriction enzymes. As such, caffeine’s ability to degrade restriction enzymes may be one reason why it is effective in our bulk DNA extraction by blocking proteins that would otherwise interfere with template DNA and possibly Taq, thereby inhibiting PCR. Combining caffeine and RNase A in the squish buffer extraction produced significantly lower Cq values than the O-buffer with and without BP for the diagnostic
H. armigera probe, decreased processing time and sample handling compared to bead purification, did not require special equipment, and added only minimal bench time (approximately 45 min more than O-buffer without BP).
Along with increased salt concentration and secondary bead purification, Oliveira et al. [
14] found that running the diagnostic assay without the control probe significantly increased the accuracy and precision of their real-time PCR results both with and without bead purification. We rejected the idea of running the assay without the control probe despite improvements to the DNA extraction process because it is necessary to ensure that PCR is occurring in each well. The RC-buffer extraction results detailed here consistently matched the efficiency of Oliveira et al. [
14] real-time PCR results with no control probe when run as a duplex real-time PCR assay.
The RC-buffer extraction is highly repeatable and was tested with 70 biological replicates at a ratio of one
H. armigera leg to 50
H. zea legs to represent a realistic bulk sample size, which was run with three technical replicates each. This sample size was chosen because the concentration of
H. armigera DNA relative to
H. zea would be sufficient in positive samples to reduce false negatives due to low DNA quality or concentration. If a positive is found, each specimen must be extracted individually and screened using real-time PCR with the protocol from Gilligan et al. [
11] to confirm the total number of
H. armigera present in the sample and then further confirmed by CO1 barcoding and possibly dissection for regulatory purposes. Because each of the 70 replicates contains legs from different
H. armigera individuals, variation in the
H. armigera probe results is expected due to slight differences in leg size and quality resulting in differences in target (
H. armigera) DNA concentrations relative to non-target (
H. zea) DNA. Such variation is expected in real-world trap samples as well and needs to be accounted for in the interpretation of results. To determine if a sample is positive, we developed the following interpretation guidelines based on the Cq values of the 70 samples extracted with RC-buffer with the ratio of one
H. armigera leg to 50
H. zea legs and based on findings from Oliveira et al. [
14]. With a baseline threshold of 100 RFU for each probe, the control probe Cq must be >5 and ≤31 to ensure the presence of lepidopteran DNA, while the diagnostic probe Cq value must be >10 and ≤37. The difference between the control probe and the diagnostic probe must be >0 and ≤8. Additionally, all samples tested on ddPCR were positive for
H. armigera (
Figure 2). The correlation between positive results in our samples for both ddPCR and real-time PCR confirms the conclusions of Oliveira et al. [
14] that this real-time PCR assay equalizes the success rate of ddPCR in detecting
H. armigera from bulk samples of up to 1:100 individuals.
The findings from our study are important because successful mass screening efforts for H. armigera depend on a rapid and efficient bulk DNA extraction method that produces DNA of sufficient quality to be used for real-time PCR. This process needs to be repeatable and accessible to promote use across multiple labs. The addition of RNase A and caffeine in this modified squish buffer DNA extraction results in a DNA extraction of sufficient quality for use in detecting H. armigera in bulk samples using real-time PCR without adding bench time or specialized methods so it can be widely implemented across labs. The findings presented here regarding the addition of RNase A and caffeine might also be applicable to similar bulk extractions of other arthropod species.