Genetic Basis of Dinitroaniiine Herbicide Resistance in a Highly ...

Genetic Basis of Dinitroaniiine Herbicide Resistance in a Highly Resistant Biotype of Goosegrass (Eleusine indica) L. Zeng and W. V. Baird

From the Horticulture Department, Box 340375, Poole Agriculture Building, Clemson University, Clemson, SC 29634-0375. We thank Dr. B. J. Gossett (Agronomy and Soils, Clemson University) for help collecting seeds ol the goosegrass blotypes from South Carolina; Dr. H. Skorupska, Dr. B. B. Rhodes, and Dr. D. a Heckel for advising the research and critiquing the manuscript, and Mr. J. K. Wells and Mr. K. Koka for technical assistance. This research was supported by grants from the USDA/SRP1AP (CSRS no. 563-S-SC-H), USDA/NRI (9537315-2152), Clemson University Research Fund, and the South Carolina Agricultural Experiment Station (SCAES) to W.V.B. Technical contribution no. 4182 of theSCAES. Journal of Heredity 1997;8&427-432; 0022-1503/97/J5.00

Goosegrass [Eleusine indica (L.) Gaertn.] is ranked as one of the five most troublesome weeds in the world (Holm et al. 1991). This annual, autogamous weed Is a serious agronomic problem in cotton, soybean, turf, and a number of vegetable crops. Until recently the antimicrotubule, preemergence dinitroaniiine herbicides (DNHs), for example, oryzalin [4-{dipropylamino)-3,5-dinitrobenzenesulfonamide] or trifluralin [2,6-dinitro-N,N-dipropyM(trifluoromethyl) benzen amine], were very effective in controlling goosegrass. However, frequent use of this class of herbicides in the southeastern United States has resulted in the selection of DNH-resistant biotypes (Mudge et al. 1984; Vaughn et al. 1987) Similar arguments have been postulated for the appearance of dinitroaniline-resistant biotypes of green foxtail [Setaria viridis (L.) Beauv] Morrison et al. 1989), Palmer amaranth (Amaranthus palmeri S. Wats) (Gossett et al. 1992), and Johnsongrass [Sorghum halepense (L.) Pers.] (Wills et al. 1992) In addition, a DNH-resistant strain of Chlamydomonas reinhardtii was selected on oryzalin-containing medium following methylmethane sulfonate mutagenesis (James et al. 1988). Among species with

both DNH-resistant and -susceptible biotypes, goosegrass demonstrates the highest level of DNH resistance. The highly resistant (R) biotype (e.g., I;*, = 1.5 x 10~5 M oryzalin) is 35-fold more tolerant to the dinitroanilines than is the susceptible (S) biotype, when root growth is assayed in the presence of the DNHs (Baird et al. 1992). Using a bioassay based on mitotic indices, researchers concluded the R biotype is at least 1000-fold more tolerant than the S biotype (Vaughn 1986a). The mode of inheritance of herbicide resistance varies with the plant species, as well as with herbicide type. For example, triazine resistance in Senecio vulgaris L, Amaranthus retroflexus L, and Chenopodium album L. (Scott and Putwain 1981; Solymosi 1981; Warwick and Black 1980) is maternally inherited, whereas in Abutilon theophrasti Medic, it is inherited as a single, partially dominant nuclear gene (Anderson and Gronwald 1987). Paraquat resistance in Conyza bonariensis (L.) Cronq. is inherited as a single, dominant nuclear gene (Shaaltiel et al. 1988), and siduron resistance in Hordeum jubatum L. is inherited as at least three dominant complementary major genes (Schooler et al. 1972). Most recently dinitroaniiine herbicide re-

427

Downloaded from http://jhered.oxfordjournals.org/ at Michigan State University on October 16, 2013

The Inheritance of dinitroaniiine herbicide resistance was investigated based on the dinitroaniiine response phenotype of F, plants and segregation analysis In F, and F, progenies. Reciprocal crosses were made between two susceptible (S) blotypes and two resistant (R) biotypes. Seven F, hybrids were identified by isoenzyme analysis, and F, and F, populations were derived from these F, plants by selling. The dinitroaniiine response phenotype of F, plants and F2 and F3 seedlings was determined using a root growth bloassay. F, plants were phenotypically susceptible (I.e., Identical to their S parental plants), indicating that resistance Is recessive to susceptibility. Nonparental phenotypes were not observed, and data from reciprocal crosses gave no indication of maternal or paternal inheritance. F2 seedlings were classified as either S or R phenotype. The observed segregation ratio of S to R seedlings fit only a 3:1 (S:R) ratio. Segregation data indicated the dinitroaniiine herbicide-response phenotype is inherited as a single nuclear gene, and it Identified two alleles (i.e., Drp* and Drp') at this single locus. The 3:1 segregation ratio was confirmed by analysis of F, progenies. Taken together, results from F,, F2 and F, plants were consistent with dinitroaniiine herbicide resistance being inherited as a single, recessive nuclear gene.

sistance in 5. viridis was reported to be controlled by a recessive nuclear gene (Jasieniuk et al. 1994). Questions remain as to the mode of inheritance of DNH resistance in goosegrass. Smeda and Vaughn (1994) suggested resistance would show a complex pattern of inheritance, although the results of their preliminary investigation were not published because of the small sample size. In contrast, Jasieniuk et al. (1994) predicted DNH resistance in goosegrass would be dominant because of what was described as the "predominantly outcrossing nature" of the Eleusine mating system, an erroneous conclusion derived from an anatomical study (Graneshaiah and Umashaanker 1982). This assumption regarding reproductive behavior is not strictly supported by studies of E. indica population biology (Werth et al. 1994). Considering the high level of resistance identified in R biotypes of goosegrass, the existence of a second distinct intermediately resistant biotype (Vaughn et al. 1990), and the mode of action of the DNHs (Ashton and Crafts 1981; Murthy et al. 1994) determination of the genetic basis of DNH resistance would make goosegrass a compelling model system in which to study the molecular and biochemical mechanisms of herbicide resistance, and thus improve our knowledge of the plant cytoskeleton and its role in affecting plant morphogenesis. The goal of the research reported here was to determine the inheritance of resistance to the DNHs in the weed species E. indica. Our objectives to reach this goal were to make reciprocal crosses between R and S biotypes, identify F, hybrids, determine the dinltroaniline response phenotype (DRP) of the F, plants, produce F2 and F3 progeny, and analyze segregation data with regard to the Inheritance of DNH resistance and susceptibility. Here we report our findings on the genetic basis of oryzalin tolerance in accessions of the highly resistant biotype of goosegrass. Materials and Methods Plant Material Because of the autogamous nature of reproduction in goosegrass (Baird et al. 1992; Werth et al. 1994), wild plants have self-pollinated for many generations. Thus a high level of homozygosity likely exists in these field populations. Parental lines for controlled crosses were chosen based on their true-breeding DRP and possession of contrasting homozygous isozyme

4 2 8 The Journal of Herecfity 1997.88(5)

phenotypes (see below). Seeds from two populations of the R biotype were collected from Independent locations (MSC-R = Marlboro County, South Carolina; CAL-R = Cherokee County, Alabama), and seeds from two S biotypes were collected from two other independent locations (OSC-S = Orangeburg County, South Carolina; ASC-S = Anderson County, South Carolina). Original collections were made in the fall of 1991. These biotypes were selfed in isolation for two generations in greenhouses to ensure homozygosity, and were checked for the stability of their herbicide response and isozyme phenotypes before they were used in crosses. Crosses Due to the small and tightly compact flowers of goosegrass (Holm et al. 1991), manual emasculation is impractical. Thus the crosses were made without emasculation and plants were pollinated by routine clipping and pollen transfer methods (Fehr and Hadley 1980). Plants were pollinated in the early morning on at least two consecutive days. The pollinated flowers were bagged in glassine paper to prevent outcrossing. After one month, hybrid seeds and seeds from "control" parental self-pollinations were manually harvested, cleaned, and stored dry at room temperature. F, Hybrid Identification The mixture of hybrid and "selfed" seeds produced from each of the controlled crosses were germinated, and individual seedlings were screened for hybridlty using isozyme markers. Isozyme methodology followed Werth et al. (1994). Electrophoresis was carried out in 12% (w/v) starch gels. The enzyme systems were resolved on a morpholine-citrate electrophoretic buffer system (pH 8.3). After electrophoresis, the gel was sliced and separately stained for acid phosphatase (ACP) or isocitrate dehydrogenase (IDH). The ACP locus is polymorphic in goosegrass populations and is represented by two alleles encoding fast (f) and slow (s) allozymes (Werth et al. 1994). Individual MSC-R and CAL-R plants are homozygous ff while individual OSC-S and ASC-S plants are homozygous ss. The ACP enzyme is dimeric (De Cherisey et al. 1985), and therefore the heterozygous phenotype of the F, hybrids will be three-banded. The isozymes of IDH are encoded by three loci in goosegrass—ldh\, Idh2, and idhZ—of which only Idhl is polymorphic with two alleles, Idhl* and Idh2b. Due to gene dupli-

cation, the Idh2 phenotype is three-banded in homozygotes of the Idh2* allele, fourbanded in homozygotes of the Idhb allele, and six-banded in heterozygotes (Werth et al. 1993). Individual ASC-S, OSC-S, and MSC-R plants are homozygous for the Idh2- allele, while individual CAL-R plants are homozygous for the Idh2b allele. Thus F, hybrids between CAL-R and ASC-S will exhibit a six-banded heterozygous pattern (while those between MSC-R and OSC-S will not). All the identified F, hybrids were allowed to grow to maturity and their inflorescence bagged with glassine paper to secure self-pollination. The F2 seeds from each of these F, plants were harvested as Individual seed lots. Bioassays Herbicide response phenotype of F, plants.

Because the dinitroanillnes are preemergence herbicides, a bloassay was required that would determine the DRP of F, plants, while allowing these individuals to survive and mature for advancement to the F2 generation. Current methods for determining the DRP require the susceptible plants to be sacrificed (Beckie et al. 1990; Vaughn 1986a) Therefore we modified a bioassay, which employs clonally propagated material, to screen mature plants for DNH resistance (Murphy TR and Smith A, personal communication). Young tillers of F, plants and their parents were separated from the main crown. With the old roots removed and leaves trimmed to reduce transpiration, the tillers were inserted into trays of solid culture medium (Oasis Wedge System, Kent, Ohio) saturated with l x Hoaglands solution (Hoagland 1948) for adventitious root Initiation. The tillers were treated with herbicide by amending the nutrition solution with specific concentrations of oryzalin (technical grade, 97% purity; Eli Lilly and Company, Indianapolis, Indiana). Oryzalin was used as the model DNH because of its relatively high solubility and low nonspecific binding (Morejohn et al. 1987; Strachan and Hess 1982). The trays were covered with plastic wrap, to further reduce transpiration, and Incubated in growth chambers (14 h daylight, 25°C; 10 h dark, 18°C). On day 7 the tillers were removed from the medium and observed for adventitious root initiation, elongation, and morphology. The S or R phenotype of F, plants was defined by comparison to the rooting response of treated and untreated tillers separated from control parental plants of known DRP (i.e., S or R biotypes). When incubated in medium without oryzalin, the

rooting response of tillers from both S and R biotypes was normal (i.e., newly formed adventitious roots were thin, long, and white In color). When incubated in medium containing oryzalin (i.e., ^ 0.17 ppm), the rooting response of S blotype was abnormal (i.e., newly formed roots were thick, short, or bulbous and brown in color). In contrast, the rooting response of the R blotype incubated at the same oryzalin concentration was normal (i.e., identical to those incubated without oryzalin). The phenotypes of F, plants were classified as either S, R, or nonparental (i.e., intermediate, hyper, or hyporeslstant). Herbicide response phenotype of F2 seedlings. The DRPs of F2 seedlings were determined by a radicle elongation bloassay, that Is, germinate seeds in the presence of specific concentrations of oryzalin and measure radicle length after 8 days of incubation. Seeds were surface sterilized in 2% sodium hypochlorite, rinsed three times with sterilized water, and soaked In nutrient solution (Sommervtlle and Ogren 1982) overnight In the dark at room temperature. The seeds were plated in petri dishes (100 x 15 mm) on two layers of sterile cellulose filter paper (P5; Fisher Scientific, Pittsburgh, Pennsylvania) saturated with the nutrient solution amended with seven different oryzalin concentrations [0 ppm, 0.04 ppm (1 X 10-7M), 0.17 ppm (5 x lO-'M), 0.28 ppm (8 x 10~7 M), 0.69 ppm (2 x 10-« M), 3.46 ppm (1 X 10"5 M), and 17.32 ppm (5 x 10"! M)]. The plates were kept in the dark at room temperature overnight and then at 37°C for 24 h. All seeds germinated (i.e., radicle emerged and elongated to at least 1 mm) within 24 h of the heat treatment. After the seeds germinated, the plates were transferred to a lighted room (25°C) for further seedling growth. The lengths of radicles were measured using a dial caliper on day 8 postincubation. The DRP of F2 plants (i.e., S, R, or nonparental) was defined based on comparison with the radicle length of the selfed seedlings from known S and R biotypes (controls) germinated in the presence and absence of oryzalin, as above. When germinated in the absence of oryzalin, seedlings from both S and R biotypes had radicle lengths greater than 10 mm after 8 days of incubation. When germinated in the presence of a specific concentration of oryzalin (i.e., s 0.17 ppm), the radicle growth of seedlings from the S blotype was inhibited (i.e., discolored and less than 2 mm in length). In contrast, at a certain concentration of oryzalin (i.e., ^0.69

ppm), radicle growth of seedlings from the R biotype was not inhibited (i.e., white In color and at least 10 mm In length) after 8 days of incubation. In this way the phenotypes of F2 plants were classified as either S, R, or nonparental. The DRP bioassay of F2 progenies was designed to reduce experimental error. F2 seeds were treated alongside control seeds (i.e., selfed seeds of their S and R parents). Each plate contained selfed seeds from one of seven F, plants, 25-30 each, and seeds from their (control) S and R parental plants, 15 each. Seeds from each of the F, plants were treated at the seven different oryzalin concentrations mentioned above. There were at least four replicates (i.e., plates) for each treatment. The data were combined from all replications of a given treatment for statistical analysis. The observed segregation ratio of S and R phenotypes was tested for goodness-of-fit to expected Mendellan ratios using Pearson chi-square test (Bliss 1967). Herbicide response phenotypes ofF, seedlings. After taking root length measurements, the treated F2 seedlings were "rescued" in order to determine their genotypes based on phenotype segregation in their F3 families. This data was also used to confirm or deny the original phenotype classification of F2 seedlings. The F2 seedlings treated at 0.17 ppm oryzalin and classified as either S or R phenotype were chosen from a single F, plant. Fifteen R F2 seedlings were transferred directly to pots containing soilless medium, and 30 S F2 seedlings were transferred to petri dishes containing fresh growth medium lacking oryzalin. After 5-7 days, when adventitious roots had grown out, the surviving S F2 seedlings were transferred to soilless medium and grown in the greenhouse with the R F2 seedlings. Thirty seeds (F3 family) were collected from each surviving F2 plant and tested for their DRP using the same radicle elongation bioassay used in the F2 seedling test. The F3 seedling test was designed in a way similar to that for the F2 seedling test. Each plate contained seeds from one of the F3 families, 30 each, as well as their S and R parental control seeds, 15 each, and was treated at 0.17 ppm oryzalin. There was only one replicate for each treatment. The genotypes of the rescued F2 plants were inferred from segregation or nonsegregatlon of the phenotypes In their F3 families.

Figure 1. F, hybrid Identification using Isozyme markers. ACP: Isozyme marker of acid phosphatase; IDH: Isozyme marker of Isocitrate dehydrogenase. 1, 2: female (R) and male (S) parents of CAL-R X ASC-S; 3, 4: F, hybrids of CAL-R X ASC-S; 5, 6: seUed progenies from female plant of CAL-R; 7, 8: female (R) and male (S) parents of MSC-R X OSC-S; 9, 10; F, hybrids of MSC-R X OSC-S; 11, 12. seJfed progenies from female plant of MSC-R.

Results Crosses

A total of 610 crosses were made among four parental combinations between S and R biotypes in a reciprocal fashion (i.e., MSC-R X OSC-S, OSC-S x MSC-R, CAL-R x ASC-S, and ASC-S x CAL-R). A total of 220 seeds were produced from these four crosses. These seeds were germinated and F, hybrid identification was carried out 46 weeks postgermination. F, Hybrid Identification After starch-gel electrophoresis of leaf tissue homogenates from 160 seedlings and staining for the allozymes of ACP, seven seedlings were observed to produce the expected heterozygous isozyme pattern and were identified as F, hybrids (Figure 1). These presumptive F, hybrid plants were confirmed by examination of the second Isozyme marker, DH. All F, hybrid plants identified by their heterozygous ACP marker also produced a heterozygous IDH isozyme pattern (except those individuals from crosses of MSC-R x OSC-S, In which there was no Idhl polymorphism between the parental plants). The seven F, hybrids represented three of the four cross combinations (I.e., four from MSC-R X OSC-S; two from CAL-R x ASC-S; one from ASC-S x CAL-R). F, Hybrids Are Susceptible to Oryzalin The seven F, plants were evaluated for their DRP at four oryzalin concentrations, 0.00 ppm, 0.04 ppm (1 x lO-'M), 0.17 ppm (5 x 10-7 M) and 0.69 ppm (2 x lO"6 M). The rooting response of the tillers from the F, plants (e.g., AC1 in Figure 2) was

Zeng and Bard • Inheritance of Dinitroanaine Herbtade Resistance 4 2 9

Table 1. The dinitroaniline response phenotypes of F, and parental plants determined by rooting of tillers In oryzalln

Oryzalin (ppm) 0.69

Parental plants'

Rooting response F, hy- Tillers Abnorbrids* treated Normal mat

A(S) C(R) AC1 CA1 CA2 M(R) O(S)

Figure 2. The dlnltroanlllne-response phenotypes of the F, and parental planU. ASC-S: treated tiller from female parental plant of ASC-S, showing abnormal rootIng; CAL-R: treated tiller from male parental plant of CAL-R, showing normal rooting; AC1: treated tiller from the F, hybrid of ASC-S x CAL-R, showing abnormal rooting.

MO2 MO3 MO8 MO9 0.17

AC1 CA1 M(R) O(S) 0.04

identical to that of their S parental plants (e.g., ASC-S in Figure 2). When tested at 0.17 ppm or 0.69 ppm, adventitious roots of F, plants were "abnormal," being thick, short, or bulbous and brown in color. When tested at 0.04 ppm or 0.00 ppm, the rooting of tillers from F, plants was "normal," that is, producing thin, long, and white roots. At all oryzalin concentrations the rooting response of the tillers from R parental plants was normal (e.g., CAL-R in Figure 2). Nonparental rooting response was not observed at the concentrations tested. Thus the DRP of reciprocal F, plants was classified as susceptible, exactly like their S parental plants. The DRPs observed in F, plants are summarized in Table 1. Differentiating S and R Phenotypes in Fj Seedlings The DRP of F2 seedlings was evaluated over a range of seven concentrations of oryzalin (see Materials and Methods). When germinated at 0 ppm and 0.04 ppm oryzalin, all F2 seedlings and the selfed S and R control seedlings were normal, having a radicle length greater than 10 mm after 8 days of culture. Radicle growth of all F2 and S and R control seedlings was inhibited to the same extent at 17.32 ppm oryzalin. At these extreme concentrations, the phenotypes of the F2 seedlings could not be distinguished. The radicle growth of S control seedlings was inhibited (i.e., =£2 mm in length) at 0.17 ppm, 0.28 ppm, and 0.69 ppm oryzalin (e.g., OS1 in Figure 3). On the other hand, radicle elongation of R control seedlings was not significantly affected (i.e., greater than 10 mm in length) until the concentration was at

4 3 0 The Journal of Heredity 1997:88(5)

A(S) C(R)

A(S) C(R)

MO9 AC1 CA1

M(R) O(S) 0.00

A(S) C(R)

MO9 AC1 CA1 M(R) O(S) MO2 MO3 MO9

14 16 24 8 12 25 23 14 5 12 8 7 10 14 8 18 12 10 6 7 5 8 7 9 10 8 7 9 10 18 23 5 13 10

0 16 0 0 0 25 0 0 0 0 0 0 10 0 0 18 0 0 6 7 5 8 7 9 10 8 7 9 10 18 23 5 13 10

14 0 24 8 12 0 23 14 5 12 8 7 0 14 8 0 12 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

• A (S), C (R), M (R), and 0 (S) = parental plants of ASC-S, CAL-R, MSC-R, and OCS-S, respectively (see Materials and Methods for complete description). • AC1, CA1, and CA2 = F, plants of ASC-S X CAL-R and CAL-R X ASC-S, repsectlvely; MO2, MO3, M08, and MO9 = F, plants of MSC-R x OSC-S.

2: 3.46 ppm (e.g., MSI in Figure 3). Thus the use of three concentrations (0.17 ppm, 0.28 ppm, and 0.69 ppm) could unambiguously distinguish the S and R phenotypes in F2 seedlings, and these concentrations were used for segregation analysis. The DRP Segregates in a Simple Mendelian Ratio in the F, Generation When F2 seedlings from each of the seven F, plants were tested at 0.17 ppm, 0.28 ppm, and 0.69 ppm oryzalin, radicle growth of a large number of these seedlings was inhibited to the same extent as that of the S control seedlings (e.g., MO3 in Figure 3). Those F2 seedlings were classified as phenotyplcalry susceptible (S). The radicle growth of the remaining seedlings was normal, being the same as that of the R control seedlings (e.g., MO3 in Figure 3). These remaining F2 seedlings were classified as phenotypically resistant (R). Nonparental phenotypes were not ob-

Flgnre 3. The dlnltroanlllne-response phenotype of F, and parental, control seedlings. MSI: selfed seedlings from female parental plant of MSC-R, showing normal radicle growth; OS1: selfed seedlings from male parental plant of OSC-S, showing Inhibition of radicle growth; MO3: F, seedlings derived from the cross of MSC-R X OSC-S, showing segregation of the dinitroaniline response phenotype.

served. Based on the data from phenotype classification and segregation of F2 progenies, the ratio of S to R seedlings in each of seven F2 families was hypothesized and tested for goodness-of-fit to a 3:1 ratio. This statistical analysis showed there was no significant discrepancy between the observed and expected (3:1) ratios for all F2 families tested (Table 2). Segregation of DRP in F, Families One F, plant (I.e., AC1) was chosen for advancement to the F3 generation. All 15 F3 families derived from the 15 R F2 plants did not segregate for the DRP (i.e., all R seedlings). Thus the genotype of those F2 Table 2. Segregation of the dinitroaniline response phenotypes In F, seedling* as determined by radicle growth In oryzalin

F,AC1 AC1 CA1 CA1 CA2 M02 M02 M03 M08 M09

Oryza- F, families' lin (number of (ppm) seedlings) S 0.69 0.17 0.69 0.17 0.69 0.69 0.17 0.69 0.69 0.69

301 355 205 180 185 251 237 255 323 100

221 265 152 136 133 195 182 193 245 76

R

S:R