| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Crop and Soil Sciences, Center for Applied Genetic Technologies, 111 Riverbend Road, Univ. of Georgia, Athens, GA 30602-6810
b Monsanto Co., 3302 S.E. Convenience Blvd., Ankeny, IA 50021
c Monsanto Co., 2229 Avenida Militar, Isabela, PR 00662
d Monsanto Co., P.O. Box 410, State Route 48– Bldg. 2, Stonington, IL 62567
e Monsanto Co., 703 E. Benton Street, Oxford, IN 47971
f Monsanto Co., 9631 Hedden Road, Evansville, IN 47725
* Corresponding author (rboerma{at}uga.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
The GT soybean cultivars currently available were derived from the 40-3-2 line, which was produced by introducing the CP4 EPSPS coding sequence from Agrobacterium sp. strain CP4 into the cultivar A5403 (Padgette et al., 1995). The transgene sequence encodes a naturally occurring EPSPS enzyme and is fused to a chloroplast transit peptide from Petunia x hybrida Vilm. to induce delivery of the bacterial EPSPS to chloroplasts. Multienvironment trials showed that applications of up to 1.68 kg a.e. ha–1 glyphosate to line 40-3-2 and related lines at any time between early vegetative growth and pod filling did not adversely affect yield (Delannay et al., 1995; Elmore et al., 2001). However, some other independently transformed lines with the same construct as 40-3-2 showed signs of pollen sterility and reduced pod set when treated with glyphosate near the onset of flowering (Delannay et al., 1995). These lines exhibited tolerance to glyphosate applied during vegetative stages, but microscopic inspection of stained anthers from plants treated with glyphosate around the time of flowering revealed empty pollen grains.
A backcross program was initiated at the Univ. of Georgia in May 1996 to transfer glyphosate tolerance from a GT donor line to ‘Benning’ [Maturity Group (MG) VII; Boerma et al., 1997], ‘Boggs’ (MG VI; Boerma et al., 2000), ‘Haskell’ (MG VII; Boerma et al., 1994), and ‘Prichard’ (MG VIII; Boerma et al., 2001). The GT donor line was derived from a F2 plant from the cross ‘Resnik’(2) x 40-3-2. Resnik is a MG III cultivar (McBlain et al., 1990). Atypical GT:glyphosate sensitive (GS) segregation ratios were observed in some lines during the backcross transfer of the CP4 EPSPS construct from the GT donor line (H. R. Boerma, unpublished data). In a November 1998 greenhouse planting of BC3F2 plants, the portion of GS plants was less than the 3:1 GT to GS ratio expected for segregation of a single transgene. In February 1999, we found fewer than expected segregating (GT/GS) BC3F2:3 lines in a Puerto Rican nursery, and in June 1999, unexpected segregation ratios were observed in yield tests of BC3F2:4 lines. Several lines had fewer GS plants than expected for a 5:1 GT to GS ratio for progeny derived from a segregating BC3F2:3 line treated with glyphosate. Some BC3F2 populations showed the expected 3 GT:1GS segregation ratio, whereas others exhibited segregation closer to the 15 GT:1 GS ratio that would be expected if two independent GT transgenes were segregating.
Independent segregation of two GT transgenes was an unlikely explanation for the atypical segregation ratios since extensive molecular characterization and testing of 40-3-2 and its progenies had shown that they carry a single transgene (Delannay et al., 1995; Padgette et al., 1995). We therefore suspected that nontransgenic gametes produced in plants hemizygous for the CP4 EPSPS transgene were sensitive to glyphosate applied during the late vegetative stages of plant development, and we designed a series of experiments to test this hypothesis. The specific objectives of this study were (i) to determine whether gametic selection by glyphosate was the cause of the atypical segregation ratios observed, (ii) to investigate the effects of two different glyphosate application rates at three different stages of F2 plant development on tolerance in F2:3 lines, and (iii) to determine whether the nontransgenic (non-CP4 EPSPS) male and/or female gametes were sensitive to glyphosate.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
Rate and Timing Experiment
Four different populations were developed by making reciprocal crosses within two sets of parents (Table 2). Set 1 included ‘HIG36930’ (GT) and ‘HI36920’ (GS), and Set 2 included ‘DCP4100A0R’ (GT) and CX400 (GS). The crosses were made in January 2000 between the GT and GS parents at the Monsanto Puerto Rican Nursery near Isabela, Puerto Rico. The F1 seed from these crosses were planted on 27 Apr. 2000 at the Monsanto nursery and were not treated with glyphosate. The F1 plants from each cross were harvested individually. The F1:2 families were planted 13 Sept. 2000 at the Monsanto Puerto Rican Nursery (one row per F1:2 family; 20–50 seed per row). Individual F1:2 rows received one of the seven following treatments: Treatment 1 = no glyphosate; Treatment 2 = 0.84 kg a.e. ha–1 of glyphosate at the V3 stage of development (Fehr et al., 1971); Treatment 3 = 1.68 kg a.e. ha–1 of glyphosate at the V3 stage of development; Treatment 4 = 0.84 kg a.e. ha–1 of glyphosate at the V3 + 2 to 3 wk (onset of flowering); Treatment 5 = 1.68 kg a.e. ha–1 of glyphosate at the V3 + 2 to 3 wk; Treatment 6 = 0.84 kg a.e. ha–1 of glyphosate at the V3 + 3 to 4 wk (mid-flowering); and Treatment 7 = 1.68 kg a.e. ha–1 of glyphosate at the V3 + 3 to 4 wk. These treatments included a no-glyphosate control (Treatment 1), and factorial combinations of two rates of glyphosate applied at one of three different stages of plant development (Treatments 2 to 7). The lower rate of 0.84 kg a.e. ha–1 (32 oz acre–1) falls within the range of rates of glyphosate recommended by Monsanto for control of annual and perennial weeds in soybean fields, and the higher rate of 1.68 kg a.e. ha–1 (64 oz acre–1) is the maximum rate that was used in the evaluation of line 40-3-2 (Delannay et al., 1995). The original plan was to plant 35 F1:2 rows for each cross (140 rows total), with each of the seven treatments applied to five F1:2 rows, but there were insufficient F1:2 families available for two of the crosses. For the HIG36930 x HI36920 cross, each of the seven treatments was applied to two rows (total of 14 F1:2 rows). For the HI36920 x HIG36930 cross, Treatments 1, 3, and 4 were applied to three F1:2 rows, Treatments 2 and 7 to a single F1:2 row each, and Treatments 5 and 6 to four F1:2 rows each. Glyphosate was applied as nearly as possible to the targeted stage of development for each population.
|
The F2:3 lines from the two sets of reciprocal crosses were evaluated in the field for their reaction to glyphosate. These lines were planted 19 June 2001 at the Univ. of Georgia Plant Sciences Farm near Athens, GA. Eighteen individual F2:3 lines were planted for each of the 24 cross x treatment combinations that involved glyphosate application (four crosses x six treatments), and 36 individual F2:3 lines were planted for the no-glyphosate control (Treatment 1) of each cross. Each line was represented by 40 to 50 seed planted in a 2.44-m row. For each group of 18 F2:3 rows (each group was derived from the same cross and received the same treatment in Puerto Rico), one row each of the female and male parents of the cross was also planted. Glyphosate was applied in the form of Roundup Ultra Max (50.2% glyphosate) on 10 July 2001 (plants at the V2 to V3 stage) at a rate of 0.91 kg a.e. ha–1 (28 fluid oz. acre–1), and again on 11 July 2001 at a rate of 1.30 kg a.e. ha–1 (40 fluid oz. acre–1). A total of 5107 plants were rated for glyphosate sensitivity between 17 and 20 July 2001. The F2:3 rows in which all plants appeared GT were given a score of "1"; rows with both GS and GT plants were scored "2"; and rows in which all plants were GS were scored "3." In the F2:3 rows with plants expressing both reactions, the numbers of GT and GS plants were determined to estimate the segregation ratio. A 
2 test (P 
0.05) was then used to test for significant differences between the observed GT:GS ratios and the ratio expected for segregation of a single transgene in the absence of gametic selection (3 GT:1 GS).
Gamete Sex Sensitivity Experiment
To investigate further the effect of glyphosate applications on male and female gametes, remnant F2:3 seed from the populations of DCP4100A0R x CX400 and CX400 x DCP4100A0R that had received Treatment 5 in the F2 generation (1.68 kg a.e. ha–1 glyphosate applied at V3 + 2 to 3 wk) were planted in the Monsanto Puerto Rican Nursery in August 2001. These F2:3 lines were not treated with glyphosate. Eight individual F3 plants were harvested from each of nine F2:3 lines per cross. The F3:4 lines derived from these 18 F2:3 lines were then tested for glyphosate tolerance in a greenhouse experiment. Twenty-four seed per F3:4 line were planted 10 Jan. 2002 in a 450-mL polystyrene cup (24 seed/cup). After emergence the plants in each cup were thinned to 18 plants per cup. Six cups of each of the two parents (nine plants per cup) were included in the experiment as controls. All plants received an application of a 0.84 kg a.e. ha–1 of glyphosate (Roundup Ultra) on 25 Jan. 2002 and again on 28 Jan. 2002. The numbers of live and dead plants were recorded 4 Feb. 2002 to determine which of the F2:3 lines were homozygous GT, and which ones were segregating for GT and GS.
Sensitive Gamete Sex Confirmation Experiment
A greenhouse experiment was conducted to test the hypothesis that nontransgenic male gametes produced in hemizygous CP4 EPSPS plants were sensitive to glyphosate. To produce the F1 plants to test this hypothesis, six F1 plants from the cross [G96-2272(2) x GT donor line], representing BC1F1 plants hemizygous for the CP4 EPSPS transgene, were treated with 0.8% glyphosate approximately 1 wk before the onset of flowering. Pollen collected from these plants was then used to pollinate breeding line G92-2167. G92-2167 and G96-2272 are both GS, nontransgenic breeding lines developed at Univ. of Georgia. On 6 Feb. 2003 the eight F1 seed from the cross of G92-2167 x [G96-2272(2) x GT donor line]F1 were planted in two 450-mL polystyrene cups at the Univ. of Georgia greenhouse at Athens, GA. In addition, 10 cups each of G92-2167, ‘USG 7732nRR’, ‘AGS Prichard RR’, and G96-2272 were planted as checks. Each cup was over-seeded, and later thinned to six plants per cup. On 24 Feb. 2003 all plants in the experiment were treated with 1.5% glyphosate. The numbers of GT and GS plants were determined on 7 March 2003.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
2 = 116.01, P < 0.01). In contrast, the numbers of GT (353 observed vs. 362 expected) and GS (130 observed vs. 121 expected) F2 plants derived from untreated F1 plants were consistent with a 3:1 segregation ratio (2 = 0.67, P > 0.10). These data show that the application of 0.8% glyphosate 15 and 3 d before the onset of flowering had a highly lethal effect on male and/or female gametes in the F1 plants, resulting in a near-total elimination of GS plants in the F2 generation.
Rate and Timing Experiment
The 2001 field experiment was conducted to evaluate the effects of the rate and timing of glyphosate application on developing gametes. The hypotheses tested for each treatment were no gametic selection in hemizygous plants vs. gametic selection in hemizygous plants. The 1:2:1 segregation ratio of all-GT: segregating: all-GS lines observed for the F2:3 progeny of untreated F2 plants (Treatment 1) from all four populations is consistent with segregation of a single transgene in the absence of gametic selection (Table 2). In contrast to the no-glyphosate control, glyphosate application to F1:2 rows at both rates (0.84 and 1.68 kg a.e. ha–1; Treatments 2–7) eliminated the all-GS phenotypic class in the F2:3 generation, regardless of the stage of plant development (Table 2).
Glyphosate application rate and time of application did affect the proportion of lines that were all-GT or segregating for tolerance (Table 2). There was no difference between the glyphosate application rates at the V3 stage (Treatments 2 and 3) within a population. In three of the populations, the ratio of all-GT rows to segregating rows was approximately 1:2 for the V3 treatments, but in the CX400 x DCP4100A0R population the ratio was closer to 1:1. We do not have a good explanation for why this ratio was different. In contrast to the lack of a difference at the V3 stage, differences between the two glyphosate rates were evident in applications made closer to the onset of flowering. The lower rate (0.84 g a.e. ha–1) resulted in an approximate 1:1 ratio of all-GT rows to segregating rows when applications were made at the post-V3 stages of development (Treatments 4 and 6; Table 2). In comparison, application of the higher glyphosate rate (Treatments 5 and 7) to post-V3 plants generally reduced the proportion of segregating lines, particularly when the application occurred at the onset of flowering (i.e., V3 + 2 to 3 wk). At this stage of plant development, the differences in the effects of the two application rates were generally greater than when glyphosate was applied a week later. The effects of each dosage at the V3 + 3 to 4 wk stage of development (Treatments 6 and 7) were similar in the populations developed from HI36920 and HIG36930, but appeared to differ between the populations developed from DCP4100A0R and CX400 (Table 2). Application of 1.68 kg a.e. ha–1 at the onset of flowering (Treatment 5) completely eliminated the segregating row class, except for in the HIG36930 x HI36920 population, where only one out of 18 rows was segregating for GT. In this row, there were a total of 37 plants, two of which were GS and 35 which were GT (Table 2). Even though the row was segregating, there were significantly fewer GS plants in the row than would be expected (about 9 GS plants) for a segregating F2:3 line in the absence of glyphosate-induced gametic selection (2 = 7.57, P = 0.006).
The effects of glyphosate application rates and timing were also evident in the percentage of GS plants in the segregating F 2:3 rows (Table 2). Glyphosate applications at either rate during the V3 stage (Treatments 2 and 3) had little effect compared with the no-glyphosate treatment (Treatment 1). The percentage of GS plants dropped substantially, however, at the V3 + 2 to 3 wk stage (4.4–7.5%; Treatments 4 and 5) and at the V3 + 3 to 4 wk stage (2.8–10.4%; Treatments 6 and 7), especially with the higher application rate (2.8–5.4%; Treatments 5 and 7). The 1.68 kg a.e. ha–1 rate also resulted in an increase in the number of F2:3 rows containing all GT plants (130 of 144 rows) when compared with the 0.84 kg a.e. ha–1 rate (67 of 114 rows). In contrast, the effect of the lower glyphosate rate was similar at all three stages of development.
The results of this experiment provided evidence for the elimination of nontransgenic gametes in the hemizygous transgenic F2 plants that were treated with glyphosate near the onset of flowering. Since the CP4 EPSPS transgene acts as a dominant gene, it was not possible to determine if gametic selection was acting on both male and female gametes or only on the male or the female gametes.
Gamete Sex Sensitivity Experiment
The objective of this experiment was to determine whether male and female gametes were equally or differentially sensitive to glyphosate. If both male and female nontransgenic gametes in hemizygous F2 plants treated with glyphosate were nonviable, the treatment of hemizygous F2 plants with glyphosate would result in all GT plants in the F3 generation. On the basis of previous research in cotton, Gossypium hirsutum L., (Pline et al., 2002), and evidence that male reproductive organ development is more sensitive to environmental factors than female reproductive organ development in several crop species (Frankel and Galun, 1977), we expected the male gametes to be more sensitive than the female gametes. If nontransgenic female gametes in hemizygous F2 plants treated with glyphosate were viable and nontransgenic male gametes were nonviable, then half of the F3 progeny from these plants should be hemizygous and half homozygous for the transgene (all F3 progeny GT). The hemizygous F3 plants would produce 25% GS progeny in the F4 generation. Thus, the existence of GS plants in F3:4 lines would suggest that female wild-type gametes in hemizygous F2 plants survived the glyphosate treatment. The F3:4 individuals evaluated in this study were descended from F2:3 remnant seed of F2 plants which had received the 1.68 kg a.e. ha–1 glyphosate application at V3 + 2 to 3 wk (Treatment 5) in the previous rate and timing experiment. The F3 plants in these F2:3 lines were either all homozygous for the CP4 EPSPS transgene or a combination of homozygotes and hemizygotes.
All the plants of the GS cultivar CX400 were sensitive to the glyphosate treatment, while all the plants of the GT cultivar DCP4100A0R were tolerant (Table 3). Six F2:3 families from each of the crosses were found to be segregating for GT in the F4 generation, and the ratio of segregating to all-GT F3:4 families was approximately 1 segregating to 2 homozygous GT (Table 3). The observed ratio of 6 homozygous GT F2:3 lines to 12 segregating F2:3 lines supports the hypothesis that glyphosate was toxic to male or female gametes, but not to both types of gametes. The segregation ratio among the F3:4 lines derived from the 12 segregating F2:3 lines was 46 all-GT F3:4 lines to 44 segregating F3:4 lines (Table 3). Without gametic selection, one would expect a ratio of 1 (all-GT) to 2 (segregating) among the F3:4 lines. These data also suggest that glyphosate was lethal to either the male or female gametes, but not both types of gametes. It is possible that some wild-type male gametes also survived, but research has shown that female gametes are better protected than male gametes in soybean and cotton plants hemizygous for the same CP4 EPSPS transgene construct (Delannay et al., 1995; Pline et al., 2002). Thus, it appeared that there was differential sensitivity of either the male or female gametes to the glyphosate. The next task was to confirm the glyphosate sensitivity of the nontransgenic male gametes.
|
We produced eight F1 seed from the cross of G92-2167 x [G96-2272(2) x GT line]. All of the G92-2167 and G96-2272 plants included as controls were GS, whereas all of the Haskell-RR and Prichard-RR control plants were GT. All eight F1 seeds of G92-2167 x [G96-2272(2) x GT line] germinated and all were GT. The low probability of not recovering a GS plant from these eight F1 plants if both transgenic and nontransgenic male gametes are produced in equal frequency, (1/2)8 or 1/256, supports the hypothesis that the application of glyphosate was lethal to the nontransgenic male gametes, and that few, if any, survived the rate of glyphosate administered. These results are in accord with those from the earlier experiments, and with evidence from cotton for detrimental glyphosate effects on the development of pollen (Pline et al., 2001, 2002).
Padgette et al. (1995) reported that when line 40-3-2 was crossed with 17 nontransgenic cultivars, the GT:GS ratios among the F2 progeny of each cross were never significantly different from the 3 GT:1 GS segregation ratio expected for segregation of a single transgene conditioning GT. Extensive molecular analysis of 40-3-2 also indicates that the line carries a single copy of the CP4 EPSPS transgene (Padgette et al., 1995). The deleterious effects of glyphosate on the pollen of some other soybean lines carrying this construct were not observed in 40-3-2 or GT lines derived from it. Although Elmore et al. (2001) observed a slight yield reduction in GT soybean lines, they concluded that this was because of unrelated genetic differences between GT and conventional near isolines.
Delannay et al. (1995) suggested that herbicide tolerance can be divided into vegetative and reproductive tolerance, and observed that vegetative tolerance is easier to achieve in soybean. Studies with transgenic GT tobacco, Nicotiana tabacum L., also showed that reproductive tissue was more sensitive to glyphosate than was vegetative tissue (Ye et al., 2001). Evaluations of several soybean lines carrying the CP4 EPSPS construct revealed that some lines with vegetative tolerance were subject to pollen sterility and reduced pod set when treated with glyphosate 10 to a few days before flowering commenced (Delannay et al., 1995). Glyphosate applications just before meiosis appeared to have the greatest effect on pollen development in the semisensitive lines. This may be the result of several factors: (i) a shift in glyphosate transport away from vegetative tissues and toward reproductive tissues (Pline et al., 2001); (ii) the greater sensitivity of reproductive tissues relative to vegetative tissues in plants carrying GT constructs (Pline et al., 2002); and (iii) the high susceptibility of microsporocytes to glyphosate during microsporogenesis (Pline et al., 2002). Applications of glyphosate to the same lines at or past the onset of flowering have comparatively little effect on fertility (Delannay et al., 1995). Glyphosate-induced boll abscission in cotton is also influenced by application method and environmental factors (Viator et al., 2004).
Studies of glyphosate translocation and activity in cotton may provide clues about what occurs in soybean plants that are hemizygous for the GT transgene. Pline et al. (2002) described morphological abnormalities in pollen from cotton treated with glyphosate. These included the presence of large vacuoles, numerous starch grains, and aberrations in the endoplasmic reticulum. Pollen development in glyphosate-treated plants was apparently inhibited or aborted at the vacuolate microspore and valcuolate microgamete stages of microgametogenesis and resulted in immature pollen at anthesis. If soybean reproductive tissues serve as a sink for glyphosate, as do cotton bolls and squares (Pline et al., 2001; Viator et al., 2003), then this would increase the likelihood of glyphosate toxicity to immature nontransgenic pollen. The observation that glyphosate toxicity in hemizygous soybean plants was apparently limited to nontransgenic pollen suggests that the CP4 EPSPS transgene is already expressed in developing microspores at a level adequate to protect them.
In summary, the series of experiments confirmed that glyphosate-induced gametic selection was occurring in soybean plants that were hemizygous for the CP4 EPSPS transgene and that this involved nontransgenic male gametes. In soybean breeding programs utilizing populations created by crossing GT and GS parents, glyphosate-induced gametic selection can be viewed from two perspectives. First, the application of glyphosate near the onset of flowering can be used to hasten the recovery of homogeneous GT lines. Second, the application of glyphosate to hemizygous plants and the resultant selection against non-CP4 EPSPS male gametes will distort linkage in the genomic region that harbors the CP4 EPSPS transgene. In backcrosses to transfer the transgene to other genetic backgrounds, the survival of the key recombinant EPSPS gametes should remain unaffected by glyphosate, thereby increasing the frequency of recombinant progeny compared with what would be observed with no glyphosate treatment.
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 22, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
