2016 Volume 85 Issue 4 Pages 315-321
We aimed to characterize the inheritance of HEPX (heptachlor exo-epoxide) uptake ability in summer squash (Cucurbita pepo L.). Crosses between ‘Patty Green’, a cultivar that cannot take up HEPX, and ‘Toyohira 2’, a cultivar that can take up high levels of HEPX, were evaluated in this study. The pattern of inheritance for F1 progeny indicated partial dominance since the measured amount of accumulated HEPX was close to that in ‘Toyohira 2’. In the F2 generation, plants segregated into those that did not take up HEPX and those that did take up HEPX at approximately a ratio of 1:5. This segregation pattern was similar to that for the inhibiting gene (dominant suppression of a dominant allele) in the dihybrid; the expected segregation ratio of 3:13 was supported by a chi-square test. Indeed, the I gene suppresses the N gene (non-transporting gene), but the i gene cannot suppress N (II or Ii suppression of NN or Nn). In this case, the genotype of ‘Patty Green’ is proposed to be iiNN and that of ‘Toyohira 2’ to be IInn. Additionally, we proposed three gene models to explain quantitative variation in HEPX transport. The genotypes of ‘Patty Green’ and ‘Toyohira 2’ are presumed to be ABC and abc, respectively. HEPX cannot be taken up unless two or more different dominant genes are present in a plant. Thus, the genotypes can be divided into HEPX non-transporting (Abc:aBc:abC:abc) and HEPX transporting (ABC:ABc:AbC:aBC) classes. Two or three different dominant genes, irrespective of the gene combination, work together to take up HEPX. In this model, the expected segregation ratio of 10 HEPX non-transporting:54 HEPX transporting was supported by a chi-square test. This pattern of inheritance was also supported by the segregation ratio of self-propagated plants (BC1-s) derived from a backcross. Although both of these inheritance models were correct phenotypically, the function of these genes should be clarified to explain the quantitative differences in HEPX uptake.
Harmful agricultural chemicals that are extremely persistent in the environment (Nash and Woolson, 1967) and that accumulate in crops pose a potential threat to human health (Jorgenson, 2001). Heptachlor (1,4,5,6,7,8,8-heptachloro-3a,4,7,7a-tetrahydro-4,7-methano-1H-indene) was used for insect control worldwide from 1953 until its use was banned in the 1970s. Heptachlor is a cyclodiene-type insecticide, and is categorized as a persistent organic pollutant (POP) (United Nations Environment Programme, 2001). In Japan, heptachlor was registered in 1957, and its use as a pesticide was prohibited in 1975. Heptachlor was not manufactured in Japan, but 1500 t was imported between 1958 and 1972 (Japan Plant Protection Association, 1959–1973). Although precise records of heptachlor use in agriculture are not available, large amounts appear to have been used on cultivated land in Japan.
In soils, heptachlor is oxidized to the epoxide form, heptachlor exo-epoxide (1,4,5,6,7,8,8-heptachloro-2,3-epoxy-3a,4,7,7a-tetrahydro-4,7-mathanoindan) (Miles et al., 1969) that is also persistent, bio-accumulative and toxic (Zhu et al., 1995). Heptachlor exo-epoxide (hereafter called HEPX) is more stable than heptachlor in ecosystems (Lu et al., 1975), and can still be detected in soils (Aigner et al., 1998; Bidleman et al. 2006; Leone et al., 2001; Shegunova et al., 2007) in countries where the use of heptachlor has been banned.
Cucurbits crops can take up weathered POPs such as p,p′-DDE, dieldrin (a cyclodiene-type insecticide) or HEPX. HEPX has been detected in some commercially produced squash in the US (Schafer and Kegley, 2002). Also, HEPX at concentrations exceeding the maximum residue limit set by the Food Sanitation Act of Japan (sum of heptachlor and HEPX: < 0.03 mg·kg−1 FW) were detected in fruits of winter squash produced in some areas in 2007 (Murano et al., 2009). In these areas, the local administrative organization was required to quantify HEPX contamination in crops and soils, and farmers were obligated to stop cultivating squash if levels of HEPX residue in the fruits were above a certain threshold. In these former production areas, immediate action is necessary to address the problem of HEPX contamination in squash. Inter- and intraspecific variations in HEPX transport were previously recognized in Cucurbita sp. (Otani et al., 2007; Sugiyama et al., 2013b; White et al., 2003), and present solutions for this problem. One of the solutions for reducing the risks of hazardous chemicals in crops is a method of grafting using rootstock cultivars that accumulate lower amounts of the harmful chemicals. For example, concentrations of dieldrin can be reduced in cucumber fruit by using low-uptake rootstock cultivars (Otani and Seike, 2006, 2007). An alternative approach to solving the problem without grafting is to produce a new squash cultivar with an impaired ability to take up HEPX.
Although not an organic pollutant, cadmium uptake in rice and soybean was tested to identify cultivars and lines with low uptake (Arao et al., 2003; Ishikawa et al., 2005). The inheritance of cadmium concentration in durum wheat was ascribed to a single gene with the low cadmium accumulation trait being dominant (Clarke, 1997; Grant et al., 2008). Ishikawa et al. (2012) found three mutants in which a favorable low-cadmium trait was expressed by different mutations in the same gene, and the transporter encoded by the mutant gene was defective in root cadmium influx. To date, there have been no attempts to breed squash impaired in HEPX transport. Recently, ‘Patty Green’, breeding material with a very low ability to transport HEPX, was found (Sugiyama et al., 2013a). Also, individuals with the low HEPX transporting trait are able to be selected by evaluating their HEPX transport ability. We think this strategy to reduce pesticide exposure by developing squash cultivars with low HEPX transport ability should be successful, and propose to produce new squash cultivars that do not take up HEPX in the future.
Previously, we reported (Sugiyama et al., 2013b) that the low HEPX transport of ‘Hokkai 1’ (Cucurbita maxima Duchesne ex Lam.) did not have complete dominance in the F1 resulting from crosses between ‘Hokkai 1’ and the slightly high HEPX transporter ‘BHA’, and the inheritance was found to be quantitative after examination of the F2. The inheritance of this trait is, therefore, somewhat ambiguous since the difference in HEPX transport between parents was small. Therefore, in order to evaluate the inheritance of HEPX transport more clearly, more distantly related varieties should be used. In this study, we sought to characterize the inheritance of HEPX uptake genes by using ‘Patty Green’ that has an extremely low ability to transport HEPX and ‘Toyohira 2’ that has a very high HEPX transport ability. We used the HEPX concentrations in aerial tissues (leaves and stems) of young plants to evaluate the inheritance because HEPX uptake ability in the aerial tissues of young plants is a good indicator for evaluating HEPX uptake ability in the fruit (Sugiyama et al., 2013b).
Summer squash cultivars ‘Patty Green’ and ‘Toyohira 2’ were used as pure genetic lines for investigating the inheritance of HEPX uptake. Twelve ‘Patty Green’, 11 ‘Toyohira 2’, 10 F1 progeny resulting from crosses using ‘Patty Green’ as the seed parent, 10 F1 progeny resulting from crosses using ‘Toyohira 2’ as the seed parent, 104 F2 progeny, and 57 backcrossed progeny (BC1: F1 (‘Patty Green’ × ‘Toyohira 2’) × ‘Patty Green’) were used in the experiments.
Plastic pots (400 mL) were filled with 363 g of the contaminated soil. The soil in each pot was fertilized with a chemical fertilizer (0.40 g N as (NH4)2SO4, 0.17 g P as Ca(H2PO4)2·H2O, and 0.33 g K as KCl) and 2.5 g of dolomite. Seeds of the test plants (parents, F1 and F2) were germinated in perlite in several batches from October 2011 to June 2012. Seeds of BC1 were germinated in perlite on October 1 and 15, 2012. Nine days after sowing, the germinated seedlings were transferred individually to contaminated-soil-containing pots, and the plantlets were grown at 25°C with a 14 h day length and 10 h night length in a growth chamber. Fifteen days after planting, the aerial parts (shoots and leaves) of the plants were harvested for measurement of HEPX content.
2. Inheritance of HEPX transport in BC1-s progenyBC1 progeny of 12 plants were used in this experiment. Seeds were sown in plastic pots (750 mL) filled with 680 g of the contaminated soil on May 7, 2013. The plantlets were grown at about 25°C to 30°C (maximum) and at about 10°C to 15°C (minimum) in a glasshouse under natural light conditions. The sixth true leaf from each BC1 progeny was harvested from a 39-day-old plant (on June 14, 2013), and the entire leaf was used for the analysis of HEPX content. The 12 plants from which the sixth true leaf was removed were transplanted to a field on June 15, 2013 into beds 60 cm apart (1.5 m × 60 cm) at NARO Hokkaido Agricultural Research Center (Sapporo, Hokkaido, Japan), a field not contaminated with heptachlor or HEPX. After the HEPX content of all plants was measured, the plant N-9 (0.0268 mg·kg−1 FW) which showed the greatest ability to transport HEPX was self-pollinated to produce BC1-s.
Seeds (BC1-s) obtained from the N-9 plant were used in the following experiment. Two seeds were sown in each of 50 plastic pots (400 mL) filled with 363 g of the contaminated soil on May 20, 2014. Eleven days after sowing, the pots were thinned to one plant per pot. Ten days later, aerial parts of the surviving plant were harvested, and used for HEPX content analysis. Plant growing conditions were the same as those described above in section 1.
Preparation of contaminated soilContaminated soil, an andosol (FAO/UNESCO), was collected from the plowed layer (0–15 cm depth) of farmland in Japan in 2007. Precise records were not available, but this field had received regular applications of heptachlor for insect control from the 1960s to the early 1970s. The collected soil was air-dried, fully mixed, and passed through a 2-mm sieve. The resultant preparation was used as the source of contaminated soil for experiments in which plants were grown in pots. The soil concentration of heptachlor was 3.4 μg·kg−1, and the soil concentration of HEPX was 57.5 μg·kg−1.
Soil and plant analysesAnalytical methods (extraction, purification, and measurement) and quality control for heptachlor and HEPX in the soil and plant samples were described in our previous report (Murano et al., 2009). In brief, some the soil and plant samples were spiked with 50 ng of 13C10-labeled heptachlor and HEPX (Cambridge Isotope Laboratories, Andover, MA, USA) as internal standards. Soil samples were Soxhlet-extracted with acetone (Wako Pure Chemical Industries Ltd., Osaka, Japan).
Samples of aerial tissues were finely cut, mixed, and divided into two subsamples. One subsample was dried at 70°C to measure the moisture content, and the other sample was frozen and stored at −20°C until extracted for HEPX content analysis. Plant samples were homogenized for 5 min in acetone. The extract was purified on InertSep K-solute (GL Science, Tokyo, Japan) and ENVI-Carb II/PSA (Supelco, Bellefonte, PA, USA) columns by elution with n-hexane (Wako Pure Chemical Industries Ltd.). The sample was then spiked with 13C12-labeled 2,2',4,4',5,5'-HxCB (Wellington Laboratories, Guelph, Ontario, Canada) as a syringe spike, and concentrated to 50 μL under a gentle stream of nitrogen. The purified samples were analyzed using a gas chromatograph-mass spectrometer (GC/MS) (HP6890-5973N; Agilent Technologies, Santa Clara, CA, USA) equipped with an ENV-8MS capillary column (30 m × 0.25 mm i.d. × 0.25 μm film thickness, Kanto Kagaku, Tokyo, Japan). Unless otherwise noted, all heptachlor and HEPX concentrations are expressed on a dry weight basis of the particular matrix. The limits of quantitation (LOQs) were calculated according to the Japan Industrial Standard (JIS) K0312 (Japan Industrial Standard 1999). The LOQs for heptachlor in the soil and aerial tissues (shoots, leaves) were 0.001 μg·kg−1, 0.005 μg·kg−1, and 0.005 μg·kg−1, respectively. The LOQs for HEPX in the soil and shoots were 0.0004 μg·kg−1 and 0.002 μg·kg−1, respectively. The concentrations of heptachlor in all of the plant samples were under the LOQs; therefore, we described only HEPX in this report.
Statistical analysesThe goodness of fit for different segregation ratios was calculated when the inheritance of HEPX transport was determined by one to three genes. Statistical analyses were performed using SPSS 12.0J (SPSS, Chicago, IL, USA) for Windows®. A chi-square test was used to determine the probability in the segregation rate of F2, BC1, and BC1-s.
Plant resourcesThe ‘Patty Green’, ‘Toyohira 2’, F1, F2 BC1, and BC1-s materials are preserved at the NARO Hokkaido Agricultural Research Center, Hitsujigaoka, Sapporo, Hokkaido 062-8555, Japan.
We investigated the inheritance of the low HEPX transport in F1, F2, and BC1 progeny resulting from crosses between ‘Patty Green’ and ‘Toyohira 2’. The average HEPX concentration in shoots of ‘Patty Green’ was 0.028 mg·kg−1 and that of ‘Toyohira 2’ was 1.094 mg·kg−1 (Fig. 1A). The average of the parents was 0.561 mg·kg−1. The average HEPX concentration in the F1 progeny resulting from crosses between ‘Patty Green’ and ‘Toyohira 2’ was 0.851 mg·kg−1 (seed parent: ‘Patty Green’) and 0.834 mg·kg−1 (seed parent: ‘Toyohira 2’). The average HEPX concentrations in the F1 derived from a cross between ‘Patty Green’ and ‘Toyohira 2’ were closer to that in ‘Toyohira 2’, and the inheritance of HEPX transport indicated a partially dominant trait in the reciprocal F1 progeny. In the F2 population, HEPX concentrations ranged widely with the lowest value at 0.015 mg·kg−1 and the highest value at 1.377 mg·kg−1 (Table 1; Fig. 1B). Eighteen of the F2 progeny were classified as HEPX non-transporting plants (very low HEPX concentrations), whereas 86 of the F2 progeny were classified as plants capable of transporting HEPX. This segregation ratio was approximately 1 HEPX non-transporting:5 HEPX transporting.
Frequency distribution of the HEPX concentration. A: Parents (‘Patty Green’ and ‘Toyohira 2’) and F1 progeny. B: F2 progeny resulting from crosses between ‘Patty Green’ and ‘Toyohira 2’. C: BC1 progeny resulting from crosses between ‘Patty Green’ and ‘Toyohira 2’.
Chi-square values for three models of gene inheritance in F2 progeny.
If monogenic inheritance (one gene) is associated with this trait, the expected ratio would be 1 HEPX non-transporting:3 HEPX transporting as predicted by Mendelian inheritance. The P-value for the chi-square test was about 0.070 (Table 1), showing that the difference between the observed and expected ratios was not high. When two genes were associated with the inheritance of this trait and epistasis was absent, the segregation ratio (9:3:3:1) was not fitted to the results. Therefore, we needed to consider epistasis for the inheritance of HEPX transporting. As for the segregation of HEPX non-transporting and HEPX transporting, 1:15, 6:10, or 7:9 deviated to a great extent from the expected ratio (data not shown). The closest expected ratio would be 3 HEPX non-transporting:13 HEPX transporting. This type of epistasis in a dihybrid is known as an inhibiting gene (dominant suppression of a dominant allele). The P-value for the chi-square test was about 0.706 (Table 1), showing that there was high fitness between the observed and expected ratios. If three genes were associated with the inheritance of this trait, the closest expected ratio would be 10 HEPX non-transporting:54 HEPX transporting, the inheritance pattern that indicates this ratio is unknown in the trihybrid. In the chi-square test, the P-value was about 0.636.
Backcrossed progeny (BC1) derived from the F1 population (‘Patty Green’ and ‘Toyohira 2’) and ‘Patty Green’ separated into HEPX non-transporting plants and HEPX transporting plants (Fig. 1C). The number of plants incapable of transporting HEPX (0.022–0.100 mg·kg−1) compared with the number of plants capable of transporting HEPX (0.474–1.161 mg·kg−1) was 29 to 28, respectively (Table 2). This segregation ratio was almost 1:1 (χ2 = 0.018, P = 0.895).
HEPX concentration and chi-square values in BC1 progeny.
Self-propagated plants (BC1-s) derived from the backcross that were capable of transporting HEPX were classified into HEPX non-transporting plants and HEPX transporting plants (Table 3). The number of plants incapable of transporting HEPX (0.007–0.012 mg·kg−1) compared with the number of plants capable of transporting HEPX (0.174–0.638 mg·kg−1) was 4:46, respectively. When goodness of fit was calculated for the genetic hypothesis by using a chi-square test, the 1:3 ratio (χ2 = 7.707, P = 0.006) was rejected. The 10:54 ratio (χ2 = 2.205, P = 0.138) showed a fit better than that of the 3:13 ratio (χ2 = 3.793, P = 0.051).
Distribution of HEPX concentrations in BC1-s progeny and chi-square values for three models of gene inheritance.
In a previous paper, we investigated the inheritance of low HEPX transport in progeny resulting from crosses between ‘Hokkai 1’ and ‘BHA’, cultivars of C. maxima (Sugiyama et al., 2013b). In the F2 progeny, the HEPX concentration in the aerial tissues of ‘Hokkai 1’ and ‘BHA’ showed a wide range of values. The results suggested that the low HEPX transport ability was inherited as an incomplete dominant in the F1 progeny and as polygenes in the F2 progeny. However, the low HEPX uptake cultivar ‘Hokkai 1’ was not as extreme in its low ability to transport HEPX as ‘Patty Green’. The HEPX concentration of ‘Patty Green’ was about one-tenth that of ‘Hokkai 1’. ‘Toyohira 2’ accumulated HEPX at about twice the level of ‘BHA’. Therefore, experiments using ‘Patty Green’ and ‘Toyohira 2’ as genetic materials should provide a more accurate evaluation of the inheritance of this trait.
There was no effect of maternal inheritance because reciprocal crosses in the F1 yielded almost the same results (Fig. 1A). The inheritance of HEPX transport in F1 progeny derived from crosses between ‘Toyohira 2’ and ‘Patty Green’ was partially dominant for progeny that could take up HEPX. The HEPX concentrations in F2 progeny that resulted from crosses between ‘Patty Green’ and ‘Toyohira 2’ were divided into two classes: those that could take up HEPX (86 progeny) and those that could not (18 progeny) (Table 1).
If HEPX transport is determined by one gene (A: dominant, a: recessive), the segregation ratio from crosses between ‘Patty Green’ (aa) and ‘Toyohira 2’ (AA) should have been 1 HEPX non-transporting:3 HEPX transporting in the F2. The segregation ratio of HEPX transport in BC1 progeny derived from a cross between F1 (Aa) and ‘Patty Green’ (aa) was 1:1. Self-propagated plants of BC1-s derived from plants capable of transporting HEPX (AA or Aa) were all HEPX transporting plants (AA) or segregated into 1 HEPX non-transporting:3 HEPX transporting. In the F2 segregation ratio, goodness of fit to the expected value was not very high (Table 1). Monogenic inheritance was rejected on the basis of results from the BC1-s (Table 3).
In case of HEPX transport determined by two genes, we considered dihybrid ratios caused by epistasis. This segregation ratio, 3 HEPX non-transporting:13 HEPX transporting, can be explained by two genes, as exemplified by the inheritance of an inhibiting gene responsible for silk worm color (Tanaka, 1913). Indeed, the I gene suppresses the N gene, but the i gene cannot suppress N (II or Ii suppression of NN or Nn). The N gene can inhibit HEPX transport, but the n gene cannot inhibit HEPX transport. Therefore, the genotype of ‘Patty Green’ is proposed to be iiNN and that of ‘Toyohira 2’ could be IInn. The F2 generation is expected to have the following composition of genotypes: 9 I_N_ HEPX transporting because the dominant I allele is present:3 iiN_ HEPX non-transporting because the dominant N allele is present:3 I_nn HEPX transporting because the recessive n and dominant I allele are present:1 iinn HEPX transporting because the recessive n allele is present. Therefore, the ratio is 13 HEPX transporting:3 HEPX non-transporting. If this inheritance pattern was due to the action of an inhibiting gene, where the genotype of BC1-s with HEPX transport ability is IiNN or IiNn, because BC1 plants (IiN_, iiN_) were derived from a cross between F1 (IiNn) and ‘Patty Green’ (iiNN), the segregation ratio in self-propagated plants of BC1 would be as follows: 1) 3 HEPX transporting:1 HEPX non-transporting when the BC1-s genotype is IiNN, or 2) 13 HEPX transporting:3 HEPX non-transporting when the genotype of BC1-s is IiNn. The probability of the 13:3 ratio for the existence of an inhibiting gene had low fitness but could not be rejected (Table 3).
However, since it is difficult to explain the behavior as incomplete dominance of the F1 and the quantitative character of HEPX transport in the F2 progeny by the action of only two genes or Mendelian inheritance (Hartl and Jones, 1998), it is necessary to consider a new inheritance model. Thus, we examined a trihybrid inheritance model that matched with the phenotypic ratios in the F2 progeny. The segregation ratio of the F2 progeny was approximately 1 HEPX non-transporting:5 HEPX transporting. We hypothesized that there might be three genes for HEPX transport. In this model, the genotypes of ‘Patty Green’ and ‘Toyohira 2’ would be abc and ABC, respectively. The genotype of the F1 progeny would be AaBbCc. Thus, the F2 genotype segregation-ratio for three gene is ABC:ABc:AbC: aBC:Abc:aBc:abC:abc = 27:9:9:9:3:3:3:1. If HEPX cannot be taken up unless two or more dominant genes are present in a plant, the F2 progeny would segregate into HEPX non-transporting (Abc, aBc, abC, abc) and HEPX transporting (ABC, ABc, AbC, aBC) plants. In other words, two or more different dominant genes (AB, AC, BC, ABC) work together cooperatively to take up HEPX. We call this proposed model the “cooperative gene model”. The theoretical segregation-ratio should be 10 HEPX non-transporting:54 HEPX transporting, a value very close to the observed segregation ratio (Table 1). The P-value of this new segregation ratio was almost the same as the segregation ratio for the model for an inhibitor gene.
To verify this segregation pattern, we first investigated the segregation ratio of the progeny resulting from a backcross of F1 × ‘Patty Green’. The recessive gene in ‘Patty Green’ has a homozygous genotype aabbcc because the ratio of HEPX non-transporting and HEPX transporting phenotypes was approximately 1:1. Therefore, ‘Toyohira 2’ genotype is AABBCC and the F1 has genotype AaBbCc. Next, we selected a BC1 plant that could take up HEPX. BC1 has one of the following genotypes: AaBbCc, AaBbcc, AabbCc, or aaBbCc. If the inheritance is due to the “cooperative gene model”, the segregation ratio for BC1-s should be: 1) 54 HEPX transporting:10 HEPX non-transporting if the BC1-s genotype is AaBbCc, or 2) 9 HEPX transporting:7 HEPX non-transporting if BC1-s is another genotype. The experimentally measured segregation ratio was 46:4. In the chi-square test, this segregation pattern was fit to the probability for a cooperative gene 54:10 (Table 3).
We hypothesized that the continuous variation for HEPX transport in the F2 progeny is related to the function of multiple genes (Shull, 1931). If A, B, and C are multiple genes that similarly function to take up HEPX, and there is no dominance like kernel color in the F2 (Matsuo, 1981), then these genes act additively to take up HEPX, and the amount of transported HEPX increases as the number of dominant genes increases. However, this hypothesis must await further examination about the contributions of genetic variation, environmental variation or another contributory factor in a polygene system.
Our results suggested that the inhibiting gene model was consistent with the inheritance of HEPX transport in the case of a dihybrid. On the other hand, if a trihybrid model were involved in transporting HEPX, then two or more dominant genes worked together. Both of these models for the inheritance of HEPX transport were correct phenotypically, but we must first determine the function of these genes to explain the quantitative differences in HEPX uptake. Although these inheritance models still need to show molecular evidence (QTL analysis or gene mapping), the inheritance information is expected to be useful for breeding new squash cultivars that do not take up HEPX.
