Figures
Abstract
The lesser grain borer Rhyzopertha dominica (F.) is one of the most destructive insect pests of stored grain. This pest has been controlled successfully by fumigation with phosphine for the last several decades, though strong resistance to phosphine in many countries has raised concern about the long term usefulness of this control method. Previous genetic analysis of strongly resistant (SR) R. dominica from three widely geographically dispersed regions of Australia, Queensland (SRQLD), New South Wales (SRNSW) and South Australia (SRSA), revealed a resistance allele in the rph1 gene in all three strains. The present study confirms that the rph1 gene contributes to resistance in a fourth strongly resistant strain, SR2QLD, also from Queensland. The previously described rph2 gene, which interacts synergistically with rph1 gene, confers strong resistance on SRQLD and SRNSW. We now provide strong circumstantial evidence that weak alleles of rph2, together with rph1, contribute to the strong resistance phenotypes of SRSA and SR2QLD. To test the notion that rph1 and rph2 are solely responsible for the strong resistance phenotype of all resistant R. dominica, we created a strain derived by hybridising the four strongly resistant lines. Following repeated selection for survival at extreme rates of phosphine exposure, we found only slightly enhanced resistance. This suggests that a single sequence of genetic changes was responsible for the development of resistance in these insects.
Citation: Mau YS, Collins PJ, Daglish GJ, Nayak MK, Ebert PR (2012) The rph2 Gene Is Responsible for High Level Resistance to Phosphine in Independent Field Strains of Rhyzopertha dominica. PLoS ONE 7(3): e34027. https://doi.org/10.1371/journal.pone.0034027
Editor: Christian Schönbach, Kyushu Institute of Technology, Japan
Received: January 18, 2012; Accepted: February 20, 2012; Published: March 26, 2012
Copyright: © 2012 Mau et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Paul Ebert acknowledges financial support from the Grain Research Development Corporation (grant UQ00010) and the Australian Research Council (grant C10027038) and an AusAID scholarship for Yosep Mau. Patrick Collins, Gregory Daglish and Manoj Nayak acknowledge the support from the Australian government's Cooperative Research Centres Program. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Phosphine (PH3) is the most economically viable fumigant for the control of insect pests of stored grain, making it the major method of control worldwide [1], [2], [3], [4], [5]. Low level resistance to phosphine was first reported in an FAO global survey undertaken in the 1970s [6]. Widespread high level resistance to phosphine in recent years now threatens the continued use of this chemical [7]. The lesser grain borer Rhyzopertha dominica (F.) is one of the most destructive pests of stored grains and high levels of phosphine resistance have been reported from several countries, such as Bangladesh [8], Brazil [9], [10], India [11], [12], China [13], [14], and the Philippines [15]. In Australia, strongly resistant R. dominica was first detected in southern Queensland in 1997 [16], followed by detection of another strongly resistant strain 300 km to the north in 1998 (Collins, unpublished). Emergence of strongly phosphine resistant strains of R. dominica was recently reported in New South Wales [17] and in South Australia (Wallbank, personal communication).
Genetic analysis of the initial strongly resistant strain from southern Queensland SRQLD, (elsewhere referred to as QRD569), identified two major genes responsible for resistance [3], [18]. Molecular genetic analysis demonstrated that the strong resistance of SRQLD was provided by synergistic interaction between the two genes, rph1 and rph2 [19]. Schlipalius et al. [19] also showed that rph1 is the major gene responsible for the weak resistance phenotype of the strain WRQLD, previously referred to as QRD369 [3].
Comparative genetic analysis [20] revealed that two major genes are responsible for the strong resistance phenotype of insects from more recent resistance outbreaks in New South Wales (SRNSW) and South Australia (SRSA). Results of complementation analysis involving the weakly resistant strain WRQLD suggest that the gene responsible for the weak resistance phenotype, rph1, contributes to resistance in both SRNSW and SRSA. Interaction between rph1 and a second gene(s) is likely responsible for the strong resistance phenotype exhibited by these strains. However, it was not determined whether or not any genes, other than rph1, were shared between the two strains. To date, the genetics of resistance in the second strong resistance strain collected from southern Queensland (referred to as SR2QLD in the remainder of this report) has not been determined. Thus, it was not known whether the rph1 and rph2 genes originally identified in SRQLD also were responsible for resistance in SR2QLD.
This paper reports the genetic analysis of SR2QLD as well as three additional strains that had been found previously to contain a resistance allele at the rph1 locus. These four strongly resistant strains, SRQLD, SR2QLD, SRSA, and SRNSW, originate from four widely separated geographical regions of Australia. They have not previously been analysed to determine whether the rph2 gene originally identified in SRQLD also contributes to their resistance phenotype. Knowing the genetic basis of resistance in these strains will allow us to predict how resistance will develop in the field. We also have created a laboratory strain that contains the resistance alleles from each of the four strongly resistant strains. Our results are consistent with the same two genes contributing to resistance in each of the four resistance outbreaks. Not surprisingly, crossing these strains and reselecting for high level resistance does not result in a major increase in the resistance level.
Materials and Methods
Insect Strains
Four strongly resistant strains (SRNSW, SRSA, SRQLD and SR2QLD) and a weakly resistant strain (WRQLD) were employed in this study. Two strains (SRNSW, SRSA) were collected from Merriwagga in south-western New South Wales (1999) and Port Adelaide in South Australia (2000), respectively, and were assigned the following collection reference numbers: NNRD2864 and NSRD3075 [17]. Strains SRQLD and SR2QLD were collected from Millmerran (1997) and Wandoan (1998) in southern Queensland and were assigned collection reference numbers QRD569 and QRD676, respectively. The weakly resistant strain (WRQLD) was also collected from Millmerran (1997) and was assigned reference number QRD369. The approximate distance between the geographic origins of any two strongly resistant strains was 250–1500 km. All resistant strains were selected with phosphine for at least five generations to promote homozygosity at resistance loci. All strains were cultured on whole wheat and maintained at 30°C and 55% relative humidity.
Complementation and allelic relationship analysis
Complementation analysis was carried out to determine whether the rph1 gene that controls weak resistance to phosphine in WRQLD and contributes to strong resistance in SRQLD [19], also contributes to strong resistance in the second strongly resistant strain from Queensland, SR2QLD. The test involved crossing WRQLD and SR2QLD and determining the resistance phenotype of the F1 and F2 progeny as previously described for SRQLD [3]. In addition, crosses between all four strongly resistant strains were made to determine the relationships between their respective resistance alleles. Crosses representing all six pairwise strain combinations were produced from the four strongly resistant strains: SRNSWxSRSA, SRQLDxSRNSW, SRQLDxSRSA, SRQLDxSR2QLD, SRNSWxSR2QLD and SR2QLDxSRSA. F1 and F2 progenies were generated from each cross and the response to phosphine exposure of individuals within these progenies were examined to determine the relationships between the resistance genes of the parental strains.
Interaction between resistance genes when combined in a single strain
We also determined whether a strain could be selected with enhanced resistance to phosphine from a population containing all resistance genes from each of the four independently derived strongly resistant strains employed in this study. To combine all the resistance genes from the four strains in a single population, we employed two different crossing strategies; combined crosses and double crosses.
Combined crosses
Combined crosses provided the easiest way to introduce all resistance genes into a single population. Initially, two pairwise crosses were made from the four parental strains, SRQLDxSRSA and SRNSWxSR2QLD, by mating an adult virgin female with an adult male on kibbled grain inside a plastic capsule. Five identical matings were set up for each of the two crosses. After two weeks, the adults were removed to fresh kibbled grain and the old grain containing the progeny from each set of five identical matings was pooled in a single plastic cup. This cup was topped up with fresh kibbled grain and the F1 progeny were allowed to mature. Fifty mature (1–2 weeks post eclosion) F1 progeny from each of the two crosses were combined and allowed to mate freely for two weeks to produce the combined progeny. This is referred to as the combined crosses (CC). Collectively, the progeny of the combined crosses would carry all resistance alleles from the four parental strains. The resulting F1 progeny of the combined crosses were, again, allowed to mate freely for two weeks to produce the F2 generation. Approximately 600 adult F2 progeny were allowed to mate freely to produce an F3 generation. Selections for phosphine resistance commenced in the F3 generation at a concentration of 0.5 mg/L phosphine. The survivors of this selection were retained to produce subsequent generations. The second and third selections were performed on the F5 and F7 generations, both at 1.0 mg/L phosphine. A parallel population of F7 individuals that had previously been subjected to two rounds of selection were also tested for their resistance phenotype. A second mortality response curve was also performed in the F9 generation, by which time the line had already been subjected to three rounds of phosphine selection.
Double Hybrid Crosses
The second approach was to set up defined crosses in each of two generations to establish a doubly hybrid strain. Initially, two single crosses were made, SRQLD×SRNSW and SR2QLD×SRSA. Subsequently, F1 individuals of the two single crosses were mated to produce double cross progeny. This ensured that the parental strains contributed equally to the genotypes present in the resulting progeny. These crosses involved the same four parental strains used for the combined crosses (CC) but in different pairwise combinations in case this influenced the likelihood of selecting specific resistance genotypes.
The initial crosses were carried out as reciprocal single pair crosses between virgin individuals, for a total of ten crosses for each pair of strains. The parents from the initial crosses were pooled on fresh kibbled grain to establish the two single cross strains (SRQLD×SRNSW) and (SR2QLD×SRSA). Reciprocal crosses were then set up between progeny of the initial crosses to create 20 double crosses, each of which produced progeny derived from all four parental strains. The adults were removed and the grain containing the eggs of 20 double cross mating pairs was combined in a bottle containing ∼200 g whole grain. The resulting adult (1–3 week post eclosion) F1 double cross progeny were transferred to new grain and left to mass cross for two weeks to produce an F2 generation. The F3 generation was produced through mass crosses of approximately 600 F2 adults. Successive generations were produced by mass crossing progeny of the previous generation. Selection for phosphine resistance was carried out on adults of the F3, F5 and F7 generations. Progeny of both the initial single crosses and subsequent double crosses were selected at 0.5 mg/L phosphine in the F3 generation, but 1.0 mg/L phosphine in the F5 and F7 generations. Phosphine resistance was quantified in the F7 and F9 generations.
Phosphine fumigation
Phosphine resistance of the parental strains and the progeny of the crosses was measured by exposing insects to phosphine fumigation for 48 hours [3] at a range of phosphine concentrations (0.01–2.0 mg/L). Phosphine was generated in a collection tube containing aluminium phosphide introduced into a 5% sulphuric acid solution [21]. Phosphine concentration was determined by gas chromatography [Varian (Mount Waverley, Victoria, Australia) aerograph model 90-P] utilising dichlorofluoromethane (Refrigerant F24; Lovelock Luke, Mayne, Queensland, Australia) as the carrier gas and a gas density balance detector.
Adult beetles (1–3 weeks post eclosion) were confined within small plastic cups (50 beetles per cup) containing 5 g whole grain. The cups were placed inside gas-tight desiccators and phosphine was injected into the desiccators through a septum. The insects were exposed to phosphine for 48 hours at 25°C and 70% r.h. then held for 14 days at 25°C and 55% r.h. when end-point mortality was assessed. A minimum of 100 insects was fumigated at each phosphine concentration.
Data analysis
Mortality data for each strain or hybrid line were subjected to log-concentration/probit-regression analysis [22]. Mortality data were first corrected for control mortality (≤10%) based on Abbott's formula [23]. The probit analysis was carried out using the GenStat7 statistical package [24]. The goodness-of-fit to the log-dose/probit mortality line was determined by a chi-square test. In the goodness-of-fit calculation, at doses where the expected response was less than one, the number of observed responses was combined with the value for an adjacent dose and the degrees of freedom for the chi-square analysis were adjusted accordingly. The resistance factors for both single cross and double cross progenies were calculated by dividing the LC50 of each progeny generation by the mean LC50 of the parental strains.
Results
Complementation and allelic relationship analysis
The resistance phenotype of the four strongly resistant R. dominica strains that have been characterised in Australia differ by at most three fold. The rank order of resistance is SRSA≅SR2QLD<SRQLD≅SRNSW. Previous work suggested that evolution of resistance was constrained in the order in which genetic changes could occur [19]; that two genes contributed to resistance in multiple instances and that at least one resistance gene was common to multiple highly resistant strains [20]. The following comparative genetic experiments use complementation tests and gene stacking to reveal previously unknown resistance mechanisms or previously unidentified resistance genes.
The rph1 gene contributes to resistance in SR2QLD
The second strongly resistant strain from Queensland, SR2QLD, is distinct from SRQLD that was analysed previously [18] (Schlipalius et al. 2002). SR2QLD was crossed with the weakly resistant strain, WRQLD, which is homozygous for the resistance allele of the rph1 gene. Probit analysis revealed that the responses of parental strains SR2QLD and WRQLD as well as their F1 progeny were linear (Figure 1). The relevant chi-square values of the response data were 1.407 (df = 7, p = 0.985) and 4.95 (df = 7, p = 0.666) for SR2QLD and WRQLD respectively, and 2.717 (df = 5, p = 0.744) for the F1.
Results are presented as log-dose mortality of the F1 hybrids and the F2 progeny with reference curves of the parental strains, WRQLD (Weak R-Strain) and SR2QLD (R-Strain). Phosphine exposure was for 48 hours at 25°C and 70% r.h.
The F1 progeny of this cross would be expected to exhibit a resistance phenotype at least as strong as the weakly resistant strain, WRQLD, if the rph1 gene contributes to resistance in SR2QLD. If this gene does not contribute to resistance in SR2QLD, the hybrid progeny would be heterozygous for the incompletely recessive resistance alleles and nearly completely sensitive to phosphine. The F1 hybrids are more resistant than WRQLD, as expected if the rph1 gene contributes to resistance in both parental strains (Figure 1).
Genetic complementation analysis of SRNSW×SRSA
A resistance allele of the rph1 gene was previously found to contribute to the phosphine resistance phenotype of both SRNSW and SRSA via complementation analysis with the weakly resistant strain WRQLD [20]. In the present study, SRNSW and SRSA were crossed and their hybrid progeny analysed to confirm lack of complementation at rph1, as well as to determine relationships between resistance alleles at other loci. Probit analysis of both parental strains and their F1 progeny revealed linear responses (Figure 2, Table 1), indicating that the strains are homogenous with respect to their resistance phenotypes.
Results are presented as log-dose mortality of the F1 hybrids and the F2 progeny, together with reference curves of the parental strains, SRSA and SRNSW, and the weakly resistant strain from Queensland (WRQLD). Phosphine exposure was for 48 hours at 25°C and 70% r.h.
The previous report that both strains contain resistance alleles of rph1 suggests that the hybrid progeny should at least show a level of resistance equivalent to that of the reference strain, WRQLD, which is homozygous resistant at rph1. The F1 hybrid progeny actually show a much higher level of resistance than seen for WRQLD (Figure 2), a level of resistance equivalent to that of the strongly resistant parental strain SRSA. As both strains are known to carry a semi-dominant resistance allele at a second locus [20], the results could be explained as an additive semi-dominant effect of two distinct genes (in addition to homozygosity at rph1). However, the resistance phenotype due to the second gene in each strain differs by about two fold. Thus, the results could also be explained as two distinct alleles in the same gene that, in combination, provide a level of resistance dictated by the weaker of the two alleles.
Interpretation of the F2 results is complicated by the fact that there is only a two fold difference in the LC50 between the parental strains, SRSA and SRNSW. The narrow distance between the parental mortality curves makes it difficult to identify a plateau in the F2 response curve as an indicator of monogenic control of resistance (beyond that caused by the rph1 locus). Because the level of resistance due to the second gene differs between the strains three distinct F2 genotypes would result, even if the second resistance factor in each strain was actually an allele of the same gene. This increases the difficulty of distinguishing a plateau in the curve. One thing that is clear is that none of the F2 individuals approach the sensitivity of WRQLD as would be expected of 6.25% of the individuals if the second resistance factors were not allelic. The most reasonable conclusion is that the strong resistance phenotype exhibited by these strains is controlled by rph1 together with a second gene that is common to the two strains, though the strengths of the alleles of the second gene differ considerably.
Genetic complementation analysis of SRNSW×SRQLD
SRNSW was also crossed to SRQLD, a well-characterised strongly resistant strain and the first strongly resistant strain collected in Australia. Responses of SRNSW and SRQLD, as well as their F1 progeny were linear (Figure 3, Table 1). Response curves of the parental strains are almost overlapping, indicating a very similar level of resistance between the two strains, with SRNSW only 1.2 times the resistance of SRQLD. The hybrid of this cross is slightly more sensitive to phosphine than is either parent, though it is still much more resistant (>10-fold) than the weakly resistant strain WRQLD. Both parents are known to be homozygous for a recessive resistance allele at rph1 [3], [20] as well as homozygous for a second gene that is weakly semi-dominant. The semi-dominance can not explain the very high level of resistance in the F1 progeny rather, the resistance is equivalent to the synergistic action between the rph1 and rph2 genes previously reported for SRQLD [19]. Thus, the strong resistance exhibited by the F1 progeny of this cross suggests that, as with SRQLD, rph1 and rph2 are responsible for resistance in SRNSW. The lack of a minor effect resistance factor that is unique to one or the other parental strain could completely explain the very minor decrease in resistance of the hybrid progeny relative to their parents.
Results are presented as log-dose mortality of the F1 hybrids and the F2 progeny with reference curves of the parental strains, SRQLD and SRNSW, and the weakly resistant strain from Queensland (WRQLD). Phosphine exposure was for 48 hours at 25°C and 70% r.h.
As with the F1 progeny, a small proportion (5%–25%) of the F2 progeny are more sensitive than the parental strains. Most of the F2 progeny, on the other hand, show a resistance level between those of the parental strains (Figure 3). The F2 progeny would be expected to form a plateau at ∼75% mortality if a single gene in addition to rph1, controls resistance in the stronger resistant parent SRNSW. However, as the response curves of both parental strains are overlapping, it is impossible to determine whether a plateau occurs at this point. Interpretation of the F2 results is further complicated by the weak incomplete recessivity of rph2 in SRQLD [3], [18] and a second resistance gene (possibly rph2) in SRNSW [20]. On balance, the F2 results are completely consistent with the F1 data, and indicate that a resistance allele of rph2 does indeed contribute to resistance in SRNSW. The results also suggest that the previously noted minor, dominant resistance factor from one of the parents is likely the product of a single gene.
Genetic complementation analysis of SRQLD×SRSA
As was done previously with the strongly resistant strain from New South Wales, the strongly resistant strain from South Australia, SRSA, was also crossed with its Queensland counterpart, SRQLD, which is homozygous for both phosphine resistance genes rph1 and rph2. Probit analysis of the response to phosphine of the parental strains and F1 progeny of this cross indicate a homogeneous response (Figure 4). Unlike the previous SRQLD×SRNSW cross in which the parental strains were nearly equally resistant, the resistance phenotype of the SRSA strain is less than half that of the strain from Queensland, SRQLD. The F1 progeny exhibited a slightly more resistant phenotype than the parental strain SRSA. The response of the F1 is as would be expected from non-complementation at the rph1 locus, indicating that each of the two parental strains is homozygous for a resistance allele of the rph1 gene. This is consistent with a previous detailed analysis in which the rph1 gene was found to contribute to resistance in each of SRSA, SRNSW [20] and SRQLD [18]. The additional resistance is presumably due to additional factors from the parental strains. Because of the comparative weakness of the strong resistance phenotype of SRSA, it is not possible to determine whether the resistance phenotype of the F1 is due to non-complementation at the rph2 locus or an additive effect of the incompletely recessive allele previously attributed to the second locus [20].
Results are presented as log-dose mortality of the F1 hybrids and the F2 progeny with reference curves of the parental strains, SRSA and SRQLD, and the weakly resistant strain from Queensland (WRQLD). Phosphine exposure was for 48 hours at 25°C and 70% r.h.
A small proportion of the F2 progeny (15%–40%) are more sensitive than each of the parental strains as well as the F1 progeny. This is similar to the observation of a minor effect resistance allele in the previous SRQLD×SRNSW cross. The F2 response curve demonstrates no clear indication of a plateau at ∼75% mortality. As rph1 and rph2 genes are responsible for the strong resistance of SRQLD, the lack of a plateau may either indicate that an rph2 resistance allele is not present in SRSA or that interpretation of the resistance phenotype is clouded by incomplete recessivity of rph2 together with the phenotype of the minor effect gene. The data indicate that strong resistance of SRSA is controlled by rph1 plus a second major effect gene that may simply be a weak allele of rph2. While the data are consistent with this explanation, they do not rule out alternative explanations for the second major effect gene.
Genetic complementation analysis of SRQLD×SR2QLD
As with SRNSW and SRSA, we also crossed SRQLD with SR2QLD, a second strongly resistant strain collected from Queensland. Probit analysis revealed a linear response curve to phosphine exposure for each parental strain as well as for the F1 progeny (Figure 5), suggesting genetic homogeneity of each strain as well as the hybrid progeny. The resistance of the two parental strains differs by only 1.6 fold, with SRQLD being more resistant.
Results are presented as log-dose mortality of the F1 hybrids and the F2 progeny with reference curves of the parental strains, SR2QLD (R-Strain 2) and SRQLD (R-Strain 1), and the weakly resistant strain from Queensland (WRQLD). Phosphine exposure was for 48 hours at 25°C and 70% r.h.
The F1 progeny are highly resistant to phosphine exposure - slightly more than SR2QLD. This indicates that the second Queensland strain, SR2QLD, is not able to complement the original strongly resistant strain from Queensland, SRQLD, at either rph1 or rph2. Thus, rph1 and rph2 are responsible for the strong resistance phenotypes of each strain. As with the preceding analyses, a small fraction of the progeny is unusually sensitive to phosphine exposure. This would appear to result from the lack of a minor dominant resistance factor contributed by one of the two parental strains. The fact that this minor, dominant factor has been apparent in crosses between SRQLD and each of three independent strains, strongly implicates SRQLD as the source of this additional resistance factor, a possibility that was first noted in Collins [3].
Genetic complementation analysis of SR2QLD×SRNSW
A cross between SR2QLD and SRNSW was also made to determine allelic relationships between phosphine resistance genes in the two strains. Probit analysis revealed that the parental strains and the F1 progeny all show a homogeneous response to phosphine, indicated by linear phosphine resistance response curves (Figure 6). As with the cross with the F1 progeny of SR2QLD×SRQLD, the F1 progeny of SR2QLD×SRNSW have a phenotype that is intermediate between the two strains. This indicates that resistance alleles of rph1 and rph2 are present in both strains.
Results are presented as log-dose mortality of the F1 hybrids and the F2 progeny with reference curves of the parental strains, SR2QLD and SRNSW, and the weakly resistant strain from Queensland (WRQLD). Phosphine exposure was for 48 hours at 25°C and 70% r.h.
The F2 progeny are mostly more resistant than the parental strain SR2QLD, though none is as resistant as SRNSW (Figure 6). The F1 and F2 data together with results from crosses with the reference strain, SRQLD suggest that the strong resistance phenotypes of SR2QLD and SRNSW are due to a resistance alleles of the rph1 and rph2 gene with the possibility of additional genes of minor effect.
Genetic complementation analysis of SRSA×SR2QLD
We also crossed the second resistant strain from Queensland, SR2QLD with the strain from South Australia, SRSA. Each of the parental strains and the F1 hybrid progeny showed linear response curves (Figure 7, Table 1). The resistance level of strain SR2QLD is only 1.26 fold higher than strain SRSA, demonstrated by the very close proximity of their response curves. The resistance phenotype of the F1 progeny of this cross was similar to that of the parental strain SRSA. The F2 progeny were intermediate between the two parental strains (Figure 7). This indicates that alleles at the same loci, rph1 and probably rph2, contribute to resistance in both SRSA and SR2QLD strains.
Results are presented as log-dose mortality of the F1 hybrids and the F2 progeny with reference curves of the parental strains, SR2QLD and SRSA, and the weakly resistant strain from Queensland (WRQLD). Phosphine exposure was for 48 hours at 25°C and 70% r.h.
Complex genetic crosses
The most reasonable explanation of the preceding genetic results is that all four highly resistant strains simply carry alternative alleles of the same two resistance genes. As demonstrated with the pairwise crosses, hybrids between such strains in the field should not lead to a dramatic increase in resistance to phosphine. The crosses that follow were designed to test the effect of combining all four resistance genotypes in a single strain and subjecting it to strong selection. The goal was to determine the resistance phenotype that would result from genetic combinations that had not been tested in the pairwise crosses. This includes testing the effects of resistance factors other than rph1 and rph2 after repeated selection for homozygosity of recessive alleles. The first of two strategies that were used consisted of setting up two pairwise crosses between strains that differed most strongly in their resistance phenotypes. The F1 progeny were then pooled to establish a “combined cross” strain. The second strategy consisted of setting up two pairwise crosses between strains that were most similar in their resistance phenotypes. This was followed by setting up a cross between hybrid F1s from the initial crosses to establish a “double hybrid” strain.
Combined crosses (CC)
The mortality response data of each of the four parental strains used to establish the line with the combined genotype were subjected to probit analysis as presented in Table 1. The LC50 values of the parental strains are non-overlapping according to their 95% fiducial limits, indicating that the response phenotypes were unique. The strains that were initially crossed exhibited an approximately 2-fold difference in resistance levels (Table 1).
The combined strain was produced in two steps. Initially two single crosses were produced, SRSA×SRQLD and SR2QLD×SRNSW. The progeny of these two crosses were then combined to establish an F2 generation. Selection for resistance was carried out at 0.5 mg/L phosphine in the F3 generation, 1.0 mg/L in the F5 generation and 1.0 mg/L in the F7 generation. Phosphine resistance of the combined cross was tested at the F7 and F9 generations, after having been selected for phosphine resistance two and three times, respectively. The F7 generation had a level of resistance 1.4 fold that of SRNSW and in the F9 generation resistance had increased to 1.9 fold. Taken together, after 9 cycles of breeding and three selections with phosphine, the progeny of the combined crosses showed an increase in resistance less than two fold higher than that of the most strongly resistant parental strain SRNSW. This level of resistance is no more that previously observed for the minor effect gene contributed by SRQLD.
Each of the response curves of the F7 and F9 progeny was linear suggesting that the line was quickly driven to genetic homogeneity by as few as two rounds of selection (Figure 8). The slopes of the response curves of the three weakest parental strains were nearly parallel (Figure 8), whereas the slope of SRNSW more closely matched those of the F7 and F9 progenies of the combined cross (Figure 8). This suggests that a genetic factor from SRNSW has been selected in the progeny.
Results are presented as log-dose mortality of the F7 and the F9 progenies with reference curves of the parental strains, SRQLD, SR2QLD, SRNSW and SRSA. Phosphine exposure was for 48 hours at 25°C and 70% r.h.
Double Crosses (DC)
We also combined all four strong resistance genotypes using an alternative strategy in which the two most strongly resistant strains, SRNSW and SRQLD, were crossed as were SRSA and SR2QLD. A double hybrid line was then produced by crossing individual offspring from each of the two single crosses. Selection for resistance was carried out in the F3, F5 and F7 generations as described for the combined crosses except that the single cross lines were also subjected to selection. The parental, single-cross and double-cross strains all exhibited homogeneous response curves with the exception of the single cross (SRSA×SR2QLD) at the F9 generation which gave a significant χ2 result (P<0.05) with a heterogeneity factor of 2.39 (Table 2).
All progenies of the single crosses exhibited a significantly higher level of resistance than the parental strains from which they were derived. Similarly, the F9 generation of the double cross was significantly more resistant than any of the parental lines, including the single cross lines from which the double cross lines were derived. Even the resistance level of the F7 progeny of the least resistant of the two single crosses SRSA×SR2QLD is essentially equivalent to that of the strongest resistant strain SRNSW given the overlapping LC50 (95% Fiducial Limit) (Table 2, Figure 9). These results are tabulated as resistance factors (ratios of LC50 values) in Table 3.
Results are presented as log-dose mortality of the F7 and the F9 progenies with reference curves of the F7 and F9 progenies the single crosses, SRQLD×SRNSW and SRSA×SR2QLD, and the parental strains, SRQLD, SR2QLD, SRNSW and SRSA. Phosphine exposure was for 48 hours at 25°C and 70% r.h.
The F9 (selected 3 times) progeny of each single and double cross had a resistance factor from 1.5 to 2.1 times that of the most resistant parental strain from which it was derived (shown in bold in Table 3). The resistance factor of the double cross strain increased between the F7 and F9 generations by 30%. Interestingly, this was simply the sum of the increases of the two single cross strains (12% and 18%). The small changes and similarity between the strains indicates that the genetic interactions are simply additive as would be expected if sensitive alleles of minor effect genes were being progressively eliminated.
Discussion
Allelic relationships between phosphine resistance genes
Due to the growing threat of resistance across the world, understanding the genetics of phosphine resistance will provide globally significant insight into effective phosphine resistance management strategies. The present work is a continuation of previous research in which we found that the rph1 gene contributed to the strong resistance phenotype in three of the strains that have been re-examined in this study [20]. We also demonstrate that an allele of rph1 contributes to resistance in a fourth strongly resistant strain, SR2QLD. In addition to rph1, a second genetic factor that is semi-dominant contributes to the strong resistance phenotype of all four strains SR2QLD, SRQLD, SRNSW and SRSA. This second gene was previously characterised in SRQLD and is named rph2.
Crosses between the strongly resistant strains provided additional insight into relationships between phosphine resistance alleles from the four resistant strains. These results confirm that the rph1 gene contributes to resistance in each of SRNSW and SRSA [20], SRQLD [18], [19] and SR2QLD. The results also suggest that the incompletely recessive synergistic factor first noted in SRQLD [3] is a synergistic resistance factor in each of the four strains. Schlipalius et al. [19] proposed that the evolution of resistance was constrained by the fact that the rph2 gene is relatively insignificant as a resistance locus in the absence of the resistance allele at rph1. We now provide evidence that the evolution of resistance in independent outbreaks is constrained to just two major genes in R. dominica. In addition, a minor effect, dominant resistance factor was identified that contributes about 2-fold resistance. It is interesting to note that strong resistance toward phosphine in Tribolium castaneum is also due to two synergistically interacting genetic factors, suggesting that elements of resistance may be shared between species as well [25].
Combining resistance genotypes to establish maximal resistance levels
We also combined resistance genotypes from all four strongly resistant strains to determine if enhanced resistance to phosphine could be produced. We employed both mass-combined and defined double-hybrid crossing strategies to combine all resistance alleles in a single population. The progenies of both mass-combined and double-hybrid crosses exhibited increased levels of resistance relative to the parental strains after either two or three rounds of selection for phosphine resistance. The highest resistance levels obtained from the selected progenies of both mass-combined and double-hybrid crosses were 1.8 and 1.9-fold higher than the most resistant parental strain, SRNSW. This level of increase in resistance can most likely be attributed to genes of minor effect contributed by the genetic backgrounds of the parental strains.
The present study confirms and extends our previous understanding that the rph1 gene is a common contributor to resistance. We also present strong evidence that rph2 together with rph1 explains nearly all of the strong resistance phenotype. A few additional resistance factors appear to contribute to the resistance of the strains that we have investigated in this paper. The effect of such minor genes is to increase resistance about 2-fold beyond the level of the most strongly resistant parental strain. The overarching hypothesis from this work is that limited genetic mechanisms are responsible for all strong resistance outbreaks in R. dominica and possibly other species as well. These results, however, do not rule out the possibility that novel resistance genes may eventually be isolated from R. dominica or other species in Australia or elsewhere. In this regard, the extremely high level of phosphine resistance recently observed in Cryptolestes ferrugineus is a prime candidate for further study [26].
We have now demonstrated that four strongly phosphine resistant strains of R. dominica each carry alleles of rph1 and rph2 that are responsible for nearly the entire resistance phenotype. This finding is quite remarkable given that phosphine is a very small and reactive molecule that could potentially have many target sites within a cell. Nevertheless, our findings suggest that the job of monitoring and managing resistance will be much more manageable than could have been the case if the genetic basis of resistance was more complex. It remains to be determined whether the same genetic basis of resistance extends to other species. If this is the case, it will allow development of a universal marker that will be useful in efficient monitoring and management of phosphine resistance.
The fact that homozygosity of rph1 alone confers only weak resistance to each of the four strains, suggests that as with SRQLD, the strong resistance phenotype is due to a synergistic interaction between rph1 and rph2. This means that the phosphine resistance problem can be alleviated by strategies that disrupt resistance caused by either rph1 or rph2, as such strategies need not target both mechanisms to be effective.
Development of diagnostic markers for monitoring resistance will be greatly facilitated by cloning of the resistance gene. Cloning of the gene will also allow comparative genetic analysis of resistance between species. Identification of the gene will also facilitate detailed genetic studies into the mode of action and mechanisms of resistance toward phosphine. This type of work is made more valuable by the outcome of the current study, which suggests that the resistance genes that we have identified may define the extent of the problem that will be faced by the grains industry.
Acknowledgments
We thank Dr. Barry Wallbank who provided the beetle strains collected from New South Wales and South Australia. Thanks to Dr. David Schlipalius for valuable discussion and comments on the manuscript. We also acknowledge and thank Linda Bond and Phillip Taylor for their assistance with maintenance of beetle cultures and Nicholas Valmas, Steven Zuryn, Jujiao Kuang, Emily Daniels for their assistance and valuable discussion.
Author Contributions
Conceived and designed the experiments: YSM PJC PRE. Performed the experiments: YSM. Analyzed the data: YSM PJC GJD PRE. Contributed reagents/materials/analysis tools: PJC. Wrote the paper: YSM PJC GJD MKN PRE. Carried out several confirming experiments that support the conclusions of the paper: MKN.
References
- 1. Chaudhry MQ (1997) A review of the mechanisms involved in the action of phosphine as an insecticide and phosphine resistance in stored-product insects. Pesticide Sci 49: 213–228.
- 2. Bengston M, Collins PJ, Daglish GJ, Hallman VL, Kopittke R, et al. (1999) Inheritance of phosphine resistance in Tribolium castaneum (Coleoptera: Tenebrionidae). J Econ Ent 92: 17–20.
- 3. Collins PJ, Daglish GJ, Bengston M, Lambkin TM, Pavic H (2002) Genetics of resistance to phosphine in Rhyzopertha dominica (F.) (Coleoptera: Bostrichidae). J Econ Ent 95: 862–869.
- 4. Rajendran S, Gunasekaran N (2002) The response of phosphine-resistant lesser grain borer Rhyzopertha dominica in mixed-aged cultures to varying concentrations of phosphine. Pest Manag Sci 58: 277–281.
- 5. Collins PJ, Daglish GJ, Pavic H, Kopittke RA (2005) Response of mixed-age cultures of phosphine-resistant and susceptible strains of lesser grain borer, Rhyzopertha dominica, to phosphine at a range of concentrations and exposure periods. J Stored Prod Res 41: 373–385.
- 6. Taylor RWD (1989) Phosphine a major grain fumigant at risk. Int Pest Control 31: 10–14.
- 7.
Mills KA (2001) Phosphine resistance: where to now? In: Donahaye EJ, Navarro S, Leesch JG, editors. 128 p. Proc Int Conf on Controlled Atmospheres and Fumigation in Stored Products, Fresno, California.
- 8. Tyler PS, Taylor RW, Rees DP (1983) Insect resistance to phosphine fumigation in food warehouses in Bangladesh. Int Pest Control 25: 10–13, 21.
- 9.
Ansell MR, Dyte CE, Smith RH (1990) The Inheritance of Phosphine Resistance in Rhyzopertha dominica and Tribolium castaneum. In: Fleurat-Lessard F, Ducom P, editors. pp. 961–969. Proc 5th Int Working Conf on Stored Product Protection, Bordeaux, France. Institut National de la Recherche Agronomique.
- 10. Lorini I, Collins PJ, Daglish GJ, Nayak MJ, Pavic H (2007) Detection and characterisation of strong resistance to phosphine in Brazilian Rhyzopertha dominica (F.) (Coleoptera: Bostrychidae). Pest Manag Sci 63: 358–364.
- 11. Rajendran S (1992) Selection for resistance to phosphine or methyl bromide in Tribolium castaneum (Coleoptera: Tenebrionidae). Bull Ent Res 82: 119–124.
- 12. Rajendran S, Narasimhan KS (1994) Phosphine resistance in the cigarette beetle Lasioderma serricorne (Coleoptera: Anobiidae) and overcoming control failures during fumigation of stored tobacco. Int J Pest Manag 40: 207–210.
- 13.
Ren YL, O'Brien IG, Whittle GP (1994) Studies on the effect of carbon dioxide in insect treatment with phosphine. In: Highley E, Wright EJ, Banks HJ, Champ BR, editors. Proc 6th Int Conf on Stored Product Protection, Canberra. pp. 173–177. CAB, Wallingford, UK.
- 14.
Zeng L (1999) Development and countermeasures of resistance in stored grain insects in Guangdong of China. In: Jin Z, Liang Q, Liang Y, Tan X, Guan L, editors. Proc 7th Int Working Conf on Stored-product Protection, Beijing, China, 14–19 October 1998. Chengdu, Sichuan Publishing House of Science and Technology. pp. 642–647.
- 15.
Acda MA (2000) Phosphine resistance in Rhyzopertha dominica (F.) (Coleoptera: Bostrychidae) from the Philippines. MSc. Thesis. St. Lucia, Australia: The University of Queensland. 79 p.
- 16.
Collins PJ (1998) Resistance to grain protectants and fumigants in insect pests of stored products in Australia. In: Banks HJ, Wright EJ, Damcevski KA, editors. Stored grain in Australia: Proc Australian Post-harvest Technical Conference, Canberra, Australia. CSIRO. pp. 55–57.
- 17.
Wallbank BE, Farrell JF (2002) Development of phosphine resistance in Rhyzopertha dominica in New South Wales, and some strategies for containment. In: Wright EJ, Banks HJ, Highley E, editors. Proc Stored Grain in Australia 2000, Canberra. pp. 105–108. Stored Grain Research Laboratory, CSIRO Entomolgy, Canberra, Australia.
- 18. Schlipalius DI, Cheng Q, Reilly PEB, Collins PJ, Ebert PR (2002) Genetic Linkage analysis of the lesser grain borer Rhyzopertha dominica identifies two loci that confer high-level resistance to the fumigant phosphine. Genetics 161: 773–782.
- 19. Schlipalius DI, Chen W, Collins PJ, Nguyen T, Reilly PEB, et al. (2008) Gene interactions constrain the course of evolution of phosphine resistance in the lesser grain borer, Rhyzopertha dominica. Heredity 100: 506–516.
- 20. Mau YS, Collins PJ, Daglish GJ, Nayak MK, Pavic H, et al. (2012) The rph1 gene is a common contributor to the evolution of phosphine resistance in independent field isolates of Rhyzopertha dominica. PLoS ONE 7(2): e31541.
- 21. Valmas N, Ebert PR (2006) Comparative Toxicity of Fumigants and a Phosphine Synergist Using a Novel Containment Chamber for the Safe Generation of Concentrated Phosphine Gas. PLoS ONE 1(1): e130.
- 22.
Finney DJ (1971) Probit Analysis 3rd ed. London: Cambridge University Press.
- 23. Abbott WS (1925) A method of computing the effectiveness of an insecticide. J Econ Ent 18: 265–267.
- 24.
Payne RW (2004) GenStat for Windows Release 7.2. VSN International. Oxford, UK.
- 25. Jagadeesan R, Collins PJ, Daglish GJ, Ebert PR, Schlipalius DI (2012) Phosphine resistance in the Rust Red Flour Beetle, Tribolium castaneum (Coleoptera: Tenebrionidae): inheritance, gene interactions and fitness costs. PLoS ONE 7(2): e31582.
- 26.
Nayak M, Holloway J, Pavic H, Head M, Reid R, et al. (2011) Developing strategies to manage highly phosphine resistant populations of flat grain beetles in large bulk storages in Australia. In: Carvalho MO, Fields PG, Adler CS, Arthur FH, Athanassiou CG, et al., editors. Proc 10th Int Working Conf on Stored Product Protection, Estoril, Portugal. pp. 396–401.