Printer Friendly

Expression and segregation of genes encoding CryIA insecticidal proteins in cotton.

Cotton is among the first transformed crops to be commercialized, yet epistatic and environmental interaction effects on foreign gene expression in breeding populations have not been reported. Somaclonal effects due to mutations during the tissue culture and regeneration processes and/or the site of gene insertion into the plant genome have been found to influence gene expression in regenerated cotton lines. Benedict et al. (1992, 1993, 1996) found differences in injury from tobacco budworm [Heliothis virescens (F.)] and in the behavior, growth, and survival of tobacco budworm, among cotton somaclones expressing cryIA insecticidal protein genes. They attributed these differences to somaclonal variation and/or positional effects on cryIA gene expression. Researchers also have explained variation in quantitative plant characters, such as plant architecture, fiber length, strength and micronaire, and rates of plant morphogenesis, as somaclonal and/or positional effects (Stelly et al., 1989; Altman et al., 1991; Benedict et al., 1996). Some have argued that these effects on gene expression are important sources of variation for crop improvement (Altman et al., 1991). Two additional sources of phenotypic variation confronting cotton breeders result from epistatic and/or environmental effects on gene expression. These sources of variation can produce desirable or undesirable effects on both native and foreign gene expression. Examination of factors that could influence foreign gene expression in transgenic plants should elucidate some of these factors, and thus facilitate development of improved cultivars with foreign genes.

The insecticidal proteins produced by the cryIA(b) and cryIA(c) genes provide a high level of insect resistance in cotton and many other crops (Gasser and Fraley, 1989). The cryIA (b) and cryIA (c) genes were taken from B. thuringiensis spp. kurstaki, strains HD-1 and HD-73, respectively. These genes code for two [Delta]-endotoxin proteins, CryIA(b) and Cry]A(c), that differ in their toxicity to species of Lepidoptera. The endotoxins are toxic to the larval stage (i.e., caterpillars of moths and butterflies) when ingested (Benedict et al. 1993; Entwistle et al., 1993).

The development of transgenic cottons that express CryIA insecticidal proteins from B. thuringiensis spp. kurstaki has resulted in new lines with improved resistance to key lepidopteran insect pests (Aronson et al., 1986; Hofte and Whiteley, 1989; Perlak et al., 1990; Adang, 1991). Cotton plants expressing modified cryIA gene sequences have demonstrated excellent control of pests, such as tobacco budworm, bollworm [Helicoverpa zea (Boddie)], and pink bollworm [Pectinophora gossypiella (Saunders)], in greenhouse (Benedict et al., 1993) and field experiments (Wilson et al., 1992; Benedict et al., 1996).

The introduction of commercial cotton varieties producing CryIA insecticidal proteins is expected to reduce environmental pollution from synthetic insecticides, increase worker safety, and improve grower profitability (Gould, 1988; Gasser and Fraley, 1989). However, widespread adoption of insect-resistant varieties will impose strong selection on insect populations for more adapted genotypes and may eventually lead to insect populations resistant to the CryIA insecticidal proteins. These important benefits and challenges underscore the need to better understand the genetic basis of plant insect resistance to CryIA proteins and the behavior of these new plant genes in different genetic backgrounds and breeding lines. To date (1994), researchers have reported on the expression and insect control efficacy of cryIA genes only in regenerants of the cultivar `Coker 312' (Perlak et al., 1990, 1991). Characterization of gene expression and efficacy is needed in different cotton backgrounds, and in combination with other insect-resistance traits.

Successful expression of an introduced gene in plants is largely dependent on the promoter, leader sequences, 3' non-coding sequences, the presence of potential volunteer plant regulatory sequences, codon frequency, the structure of the mRNA, and the gene product (Perlak et al., 1990). Perlak et al. (1990, 1991) made modifications to key regions of the cryIA(b) and cryIA(c) structural genes that resulted in a 100-fold increase in expression in cotton plants compared to the wild-type gene sequences. The modifications eliminated ATTTA sequences and almost all potential plant polyadenylation sequences, greatly increased the G + C nucleotide content of the gene, and minimized the use of rare plant codons. They reported the level of CryIA(b) or CryIA(c) proteins in cotton expressing the modified sequences to range from 0.05 to 0.1% of total soluble protein. This range of expression may be explained by differences in the position of gene insertion in the genome, somaclonal variation, interactions with other genes having direct quantitative effects (epistasis), plant-to-plant variation caused by heterogenous environmental conditions, or combinations of these factors. It is not known to what extent introduced genes, such as cryIA, are influenced by these and other factors.

The primary objective of the research reported here was to characterize the expression of two different cryIA gene inserts in various insect-resistant lines derived in three cotton backgrounds. A secondary objective was to examine the segregation of these gene inserts in the [F.sub.2] progeny of crosses between transformed cotton lines and conventional insect-resistant isolines developed in different cotton backgrounds. This research was part of a larger project to improve cotton germplasm by introgression of multiple insect-resistant traits. In a companion experiment, the [F.sub.2:4] isolines reported here were used to study the effect of single and multiple insect-resistant traits on suppression of injury from tobacco budworm, (Sachs et al., 1996).


Genetic Material

Two transformed [R.sub.3] cotton lines, MON 81 and MON 249, homozygous for cryIA(b) and cryIA(c), respectively, were provided by Monsanto Company, St. Louis, MO. These lines were produced by transforming cotton cultivar Coker 312 (C312) with Agrobacterium tumefaciens containing the chimeric cryIA genes driven by a CaMV 35S promoter with a duplicated enhancer region (Perlak et al., 1990). The transformed [R.sub.0] cotton plants were regenerated as described by Trolinder and Goodin (1987).

Insect-resistant cotton isolines were developed by Altman et al. (1989b) using `TAMCOT CAMD-E' (CAMD-E) and `Stoneville 213' (ST213) and by Thompson (1987) using `Deltapine 61'(DP61). The CAMD-E and ST213 isolines included the plant insect-resistance traits glabrous (GB), nectariless (N), and high-terpenoid (HT), and the glandless (GL) trait. In order, the alleles determining these traits were thought to be [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (Lee, 1985), [ne.sub.1] and [ne.sub.2] (Meyer and Meyer, 1961), [Gl'.sub.3] (Wilson and Smith, 1977), and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (Kohel and Lee, 1984). The high-terpenoid allele used in these studies has not been definitively described, and may not be [Gl'.sub.3], although the source is known (Altman et al. 1989a). The GL trait is thought to remove all terpenoid expression from the plant thus it provided a highly susceptible control useful in measuring the level of insect resistance provided by adding single or multiple resistance genes to isolines. The DP61 isolines only included the plant insect-resistance traits, glabrous, nectariless, and glabrous-nectariless (GB + N). The glabrous trait in the DP61 isogenic series was determined by [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (Lee, 1985) rather than the [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] allele. Each set of isolines within a background contained a wild-type (WT) isoline possessing normal traits for that cultivar background. These isolines were selected because of their potential to measure improvement in insect resistance, and expression and segregation of the foreign cryIA genes in cotton. The utility of these characters for improving insect resistance has been reviewed by Thompson (1987).

MON 81 and MON 249 lines were crossed with each isoline and recurrent parent cultivar to form 10 [F.sub.1] populations each in the C312/CAMD-E and C312/ST213 backgrounds, and eight F, populations in the C312/DP61 background, carrying the cryIA (b) or cryIA (c) genes. The crosses were made during the summer of 1990 and the resulting [F.sub.1] plants were selfed during the fall of 1990. Next, 50 to 1700 [F.sub.2] plants from each cross were grown in a field near College Station, TX, during the summer of 1991. These plants were managed with typical agronomic practices and furrow irrigation on Belk clay soil. The number of [F.sub.2] plants was greater for some populations because assuming unlinked loci, the expected frequency of plants homozygous at the cryIA and plant insect-resistance loci ranged from 0.0625 to 0.0039, depending on the number of segregating insect-resistance loci. At least three homozygous plants were recovered from each population.

Plants homozygous at the plant insect-resistance loci were identified from their phenotype, then analyzed immunologically to determine the CryIA protein concentration (Berberich et al., 1990; Fuchs et al., 1990; Sims and Berberich, 1996). The 15 to 20 plants with the highest CryIA protein concentration and five to ten plants with no detectable CryIA protein were selfed within each population. It was assumed that cry1cry plants would be more likely to have higher levels of CryIA protein than cry/- plants. The genotype of each selfed [F.sub.2] plant was determined by progeny testing up to 20 [F.sub.3] seeds. If 20 out of 20 [F.sub.3] seeds tested positive for a CryIA protein, the [F.sub.2] plant was assumed to be homozygous at the cryIA locus. If one or more [F.sub.3] seeds tested negative, then the [F.sub.2] plant was designated hemizygous. The probability that a homozygous plant was correctly identified by the progeny test was 0.994. At maturity, individual [F.sub.2] plants were harvested to form [F.sub.2] derived lines in the [F.sub.3] generation ([F.sub.2:3])

The [F.sub.2:3] seeds from homozygous lines were planted in progeny rows during the fall of 1991 in Waimai, HI, to increase seed number for large-scale field testing. The rows were inspected for phenotypic fidelity of the insect-resistance trait and off-type plants were rogued. At maturity, the individual progeny rows were harvested and bulked to form [F.sub.2:4] lines.

The [F.sub.2:4] lines were planted in fields near College Station, TX and Tivoli, TX on 26 April and 1 May of 1992, respectively. At the Tivoli site plants were managed with typical agronomic practices on Victoria clay soil under dryland conditions. A total of 18 lines, including four CryIA(b)-positive entries [i.e., possessing the cryIA(b) gene and expressing the CryIA(b) insecticidal protein] (WT, GB, N, and HT or GB+N) and one negative entry (i.e., WT, not expressing the insecticidal protein) in each of three recombinant backgrounds, and the three CryIA(b)-negative recurrent parent cultivars were produced. Each plot included four rows 9 m long with 1 m between the rows. The plots were overplanted and thinned to 320 plants [plot.sup.-1] (86 886 plants [ha.sup.-1]) about 1 mo. after planting. Because the potential for outcrossing existed among the [F.sub.2:3] lines grown in Hawaii, the plants in each plot were inspected for phenotypic fidelity and off-type plants were rogued prior to bloom.

Concentration of CryIA Protein

Leaf samples were collected from the youngest expanding terminal leaf (about 4 [cm.sup.2]). One sample was collected from each [F.sub.2] plant in early May when the plants had one to three leaves. To minimize plant-to-plant variation from moisture stress, the [F.sub.2] samples were collected over a 3-d period beginning immediately after an irrigation. A 1-cm-diam. disk was removed from the leaf with a paper punch and directly placed into a 1.5-mL capped tube. The [F.sub.2:4] plants were sampled five times during the season. The plant samples were replicate six times for each line. A replicate consisted of three entire leaves from three randomly selected plants within a plot. All plants were sampled on each sample date. Sampling began before bloom when the plants had five to seven leaves and continued every other week for 10 wk. Sampling dates for College Station were 17 June, 29 June, 17 July, 27 July, and 11 August. Sampling dates for Tivoli were 11 June, 26 June, 9 July, 23 July, and 6 August. All leaf samples were quickly frozen on dry ice, and stored at -30 [degrees] C until analyzed for protein content.

The CryIA(b) or CryIA(c) protein concentrations in cotton leaf extracts were determined by immunological analysis by means of ELISA (Berberich et al., 1990; Fuchs et al., 1990). ELISA methodologies used were similar to those of Sims and Berberich (1996). The frozen leaf disk samples from [F.sub.2] plants were ground while frozen and a protein extract was made from each ground disk. For the frozen [F.sub.2:4] leaf samples, the three leaves from each plant were crushed to a coarse powder and mixed inside the sample bag on dry ice prior to weighing a portion for analysis. Leaf tissue extracts were prepared by grinding the frozen leaf tissue in extraction buffer with a Wheaton Overhead Mixer. The efficiency of extraction of the CryIA proteins by this method ranged from 75 to 81% (results not shown) and was consistent for leaf samples collected over the flowering period. Protein concentrations were presented in two ways: as percent of total soluble protein (%SP) in the extract and as micrograms per gram fresh weight of leaf tissue. Total soluble protein per cotton leaf extract was determined by the microtiter plate application of the Bio-Rad Protein Assay (Richmond, CA). Precision of the ELISA was confirmed by analysis of leaf extract matrix effects and CryIA protein recovery for each cotton background and sampling date. To identify potential matrix effects, leaf extracts from CryIA-positive plants were added to solutions of purified CryIA protein in dilutions of 1:100, 1:50, and 1:25, then analyzed by the ELISA. Also, leaf samples from CryIA-negative plants were spiked with purified CryIA protein and analyzed. We found that the leaf extract did not adversely affect quantification of either CryIA protein when diluted 1:100 or 1:50. With the 1:25 dilution, quantification was more variable and unreliable at concentrations less than 0.2 ng [mL.sup.-1]. Therefore, all leaf extracts were analyzed at the 1:100 dilution first, then reanalyzed at the 1:50 dilution, if necessary, to place the sample concentration within the range of the standard curve. Final CryIA protein concentrations of all leaf samples were adjusted based on recovery efficacy for each sample date.

Statistical Analyses

Analysis of cryIA Gene Expression in CryIA-Positive [F.sub.2] Plants. CryIA-positive plants derived from MON 81 or MON 249, that were either homozygous or hemizygous for a cryIA insert, were analyzed in a completely randomized design. Differences between cryIA(b) and cryIA(c) constructs and insertion effects are denoted hereafter as insert effects. Effects of insert (INS), background (BG), and the insert x background (INS x BG) interactions were considered to be fixed, whereas those from plants within an insert x background [PLANT-(INS x BG)] were considered to be random. To determine the relationship between gene dosage (DOSE) and cryIA gene expression, we analyzed CryIA protein levels of plants identified by progeny tests as homozygous or hemizygous plants. Genotypes representing the dosages were treated as a fixed effect.

Analyses of variance were computed for CryIA protein concentration, expressed as %SP, and adjusted for unequal sample size by a generalized linear model and Type III sums of squares. There were no exact F-tests for the insert, background, and insert x background main effects, so the appropriate error terms were constructed from weighted combinations of the plant-within-insert x background and error mean squares. The appropriate degrees of freedom for each test were estimated by the approximation suggested by Satterthwaite (1946). If the background means were significantly different, they were compared using the Student-Newman-Keuls' multiple comparison procedure.

Analysis of cryIA(b) Gene Expression in [F.sub.2:4] Lines. cryIA gene expression was analyzed only in MON 81-derived [F.sub.2:4] lines. The experimental design was a randomized complete block with six replicates (REP). The treatments were cotton genotypes that included all possible combinations of backgrounds (C312/CAMD-E, C312/ST213, C312/DP61) and traits (GB9 N, HT, GB + N, WT). Effect of environmental location (SITE), genotype (GENE), and the location x genotype (SITExGENE) interactions were treated as fixed, whereas those due to variation among replicates within a location [REP(SITE)l were considered to be random. Time, location x time, and location x background x time effects were added to the model and analyzed as a split-plot in time. Model effects on cryIA(b) gene expression were examined by analyzing the variance of CryIA(b) protein concentration expressed as %SP or microgram per gram fresh weight of leaf tissue.

Correlation of cryIA Gene Expression Between [F.sub.2] Plants and [F.sub.2:4] Lines. The phenotypic relationship between [F.sub.2] plants and [F.sub.2:4] lines is determined by the additive coefficient of relationship between plants, additive genetic variance, nonadditive genetic variance, and permanent and temporary environmental effects (Falconer, 1989). Assuming an additive model and no genotype x environment interaction effects, the phenotype of an individual can be expressed as:

[1] P = [G.sub.A] + [G.sub.NA] + [E.sub.P] + [E.sub.T]

where P is the phenotype, [G.sub.A] is the additive genetic value, [G.sub.NA] is the nonadditive genetic value (inclusive of dominance and epistatic values), [E.sub.P] is the permanent environmental effect, and [E.sub.T] is the temporary environmental effect. The correlation between mean cryIA expression in selected [F.sub.2] parents and mean expression in [F.sub.2:4] progeny is given by:


where r is the correlation coefficient, [bar] [P.sub.2] is the mean expression in selected [F.sub.2] parents, [bar] [P.sub.2:4] is the mean expression in the [F.sub.2:4] Progeny, [Rho] ([bar] [P.sub.2],[bar] [P.sub.2:4]) is the covariance between [F.sub.2] and [F.sub.2:4] means, [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] is the variance of mean expression in selected [F.sub.2] parents, and [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] is the variance of mean expression in the [F.sub.2:4] progeny. From Eq. [1] the variance of the mean of N phenotypes of pooled individuals within a generation is (1 + F) [V.sub.A], + ([V.sub.NA] + [V.sub.PE] + [V.sub.TE])/N, where F is the inbreeding coefficient, VA is the variance of additive genetic effects, [V.sub.NA] is the variance of nonadditive genetic effects, [V.sub.PE] is the variance in permanent environmental effects, and [V.sub.TE] is the variance in temporary environmental effects. The covariance between means from related generations is [a.sub.ij] [V.sub.A], where [a.sub.ij] is the additive coefficient of relationship between generations. Hence, the correlation between mean phenotypes of the [F.sub.2] and the [F.sub.2:4] generations is given by:


where [N.sub.2] is the number of selected [F.sub.2] parents, and [N.sub.2:4] is the number of [F.sub.2:4] progeny. When [V.sub.NA], [V.sub.PE], and [V.sub.TE] are small, or when [N.sub.2] and [N.sub.2:4] are large, the contribution of the nonadditive genetic and environmental components of variance to the correlation is small and the upper bound for the correlation is 0.89.

By means of the above equations, correlations in level of gene expression were calculated between [F.sub.2] plants and [F.sub.2:4] lines to examine components of phenotypic variance (additive genetic and nonadditive nongenetic) (Falconer, 1989).

Analysis of cryIA Segregation. Chi-square goodness-of-fit tests were performed on the [F.sub.2] populations derived from crosses between MON 81 or MON 249 and the insect-resistant cotton isolines to determine if the observed segregation ratios of CryIA-positive plants to negative plants fit the expected Mendelian 3:1 phenotypic ratio. In addition, the [F.sub.2] populations derived from both transformed lines and the recurrent parent cultivars were progeny-tested to determine if the observed numbers of homozygotes, hemizygotes, and wild-type plants fit the expected 1:2:1 genotypic ratio for a single-locus model of inheritance.


Recovery of Cry1A Protein

The CryIA protein recovery, regardless of background, was very high, [is greater than] 90%, over a range of 100 to 400 ng [mL.sup.-1] of diluted leaf extract for [F.sub.2] plants with one to three leaves, or [F.sub.2:4] plants with five to seven leaves. However, as the plants matured, CryIA protein recovery decreased in each background. The recoveries, averaged over background, were 97% at 47 d after planting (DAP), 81% at 60 DAP, 81% at 76 DAP, 59% at 88 DAP, and 35% at 102 DAP.

cryIA Gene Expression in CryIA-Positive [F.sub.2] Plants

The insecticidal protein concentration averaged twofold higher in [F.sub.2] plants with the cryIA(b) gene insert than in [F.sub.2] plants with the cryIA(c) gene insert (Table 1). Mean CryIA protein concentration was lower in CryIA-positive F, plants derived in the C312/ST213 background than in the C312/CAMD-E or C312/DP61 backgrounds. Plant-to-plant CryIA protein variation was large in plants derived from both inserts; CryIA(b) protein content ranged 13-fold, while CryIA(c) protein content ranged 46-fold. Analyses of variance of cryIA gene expression in leaf samples from CryIA-positive [F.sub.2] plants derived from MON 81 or MON 249 and the insect-resistant isolines showed significant INS, BG, and PLANT(INS x BG) effects on cryIA gene expression (Table 2). The INS x BG interaction effects were not significant.

Table 1. Influence of gene insert (INS), background (13G), insert-by-background (INS x BG), and plant [PLANT(INS x BG)] effects on CryIA insecticidal protein concentration (as percent of total soluble protein) in young terminal leaves of [F.sub.2] progeny derived in three cotton backgrounds.
Insert, Background           Percent of total soluble protein
or Insert x
Background                 N     Mean [+ or -] SE([dagger])

   cryIA (b) gene          816  0.277 [+ or -] 0.005a([paragraph])
   cryIA (c) gene          527  0.132 [+ or -] 0.006b

   C312/CAMD-E             351  0.217 [+ or -] 0.007a
   C312/ST213              584  0.177 [+ or -] 0.006b
   C312/DP61               408  0.221 [+ or -] 0.006a

Insert x Background              cryIA (b) gene
   C312/CAMD-E             240  0.301 [+ or -] 0.008
   C312/ST213              330  0.252 [+ or -] 0.008
   C312/DP61               246  0.279 [+ or -] 0.009

                                  cryIA (c) gene
   C312/CAMD-E             111  0.132 [+ or -] 0.012
   C312/ST213              254  0.102 [+ or -] 0.009
   C312/DP61               162  0.162 [+ or -] 0.013

Insert, Background         Percent of total soluble protein
or Insert x
Background                   Range([double dagger])

   cryIA (b) gene               0.069-0.907
   cryIA (c) gene               0.022-1.017

   C312/CAMD-E                  0.039-0.907
   C312/ST213                   0.025-0.694
   C312[DP61                    0.022-1.017

Insert x Background
   C312/CAMD-E                  0.084-0.907
   C312/ST213                   0.069-0.694
   C312/DP61                    0.072-0.800

   C312/CAMD-E                  0.039-0.650
   C312/ST213                   0.025-0.650
   C312/DP61                    0.022-1.017

([dagger]) Least-squares estimates of marginal means.

([double dagger]) Includes plant-to-plant variation observed within each background.

([sections]) C312 = Coker 312, CAMD-E = TAMCOT CAMD-E, ST213 = Stoneville 213, DP61 = Deltapine 61.

([paragraph]) Mean values followed by a different letter are different using the appropriate F-test or Student-Newman-Keuls' multiple comparison test ([Alpha] = 0.01).


Analyses of variance of cryIA gene expression in leaf samples of CryIA-positive plants indicated a significant DOSE effect for the cryIA(b) insert, and PLANT(BG) effects for both inserts (Tables 3 and 4). Homozygotes with the cryIA(b) gene insert averaged 14% more CryIA(b) protein than hemizygotes irrespective of background. In contrast, mean CryIA(c) protein concentration was not affected by dosage of cryIA(c), even though hemizygotes tended to have higher CryIA(c) protein concentration than homozygotes.


Table 4. Influence of background (BG) and gene dosage (DOSE) effects on CryIA(b) or CryIA(c) insecticidal protein concentration (as percent of total soluble protein) in young terminal leaves of [F.sub.2] progeny derived in three cotton backgrounds.
                                  Percent of total soluble protein


Background([double dagger])   N   Mean [+ or -] SE([double dagger])

                                     cryIA (b) gene

C312/CAMD-E                   24  0.411 [+ or -] 0.021
C312/ST213                    23  0.392 [+ or -] 0.021
C312[DP61                     28  0.394 [+ or -] 0.021
Overall                       75  0.399 [+ or -] 0.012a([sections])

                                      cryIA (c) gene

C312/CAMD-E                   17  0.127  [+ or -] 0.033
C312/ST213                    15  0.148  [+ or -] 0.041
C312[DP61                     15  0.259  [+ or -] 0.033
Overall                       47  0.178  [+ or -] 0.021a

                        Percent of total soluble protein


Background([double dagger])      N         Mean [+ or -] SE

                                         cryIA (b) gene

C312/CAMD-E                      21      0.390 [+ or -]  0.024
C312/ST213                       59      0.318 [+ or -]  0.016
C312[DP61                        47      0.341 [+ or -]  0.018
Overall                         127      0.349 [+ or -]  0.011b

                                         cryIA (c) gene

C312/CAMD-E                      52      0.147 [+ or -] 0.020
C312/ST213                       72      0.154 [+ or -] 0.023
C312[DP61                        63      0.253 [+ or -] 0.017
Overall                         187      0.184 [+ or -] 0.012a

([dagger]) C312 = Coker 312, CAMD-E = TAMCOT CAMD-E, ST213 = Stoneville 213, DP61 = Deltapine 61.

([double dagger]) Least-squares estimates of marginal means.

([sections]) Overall mean values followed by a different letter for a given CryIA protein are significantly different using the appropriate F-test.

cryIA(b) Gene Expression in [F.sub.2:4] Lines

CryIA(b) protein and total soluble protein concentrations in terminal leaves of [F.sub.2:4] lines with the cryIA(b) gene insert were determined before bloom at two locations (Tables 5 and 6). They were found to be similar for all line x background (GENE) combinations within a location, but the SITE and REP(SITE) effects were significant (Table 5). The BG and SITE x GENE interaction effects were not significant. The mean CryIA(b) protein content, expressed as %SP, was 50% higher at Tivoli than College Station; but, the mean CryIA(b) protein content, expressed as microgram per gram, was twice as high at College Station. CryIA(b) protein content was more similar between locations when expressed as %SP than microgram per gram; however, both measures showed a significant SITE effect. Total soluble protein content ranged from 0.96% FW at Tivoli to 2.63% FW at College Station.

Table 5. Analysis of variance for CryIA(b) insecticidal protein concentration (GENE) (as percent of total soluble protein or [micro]g [g.sup.-1]) and total soluble protein (as percent of fresh weight) in young terminal leaves of [F.sub.2] progeny derived in three cotton backgrounds in two environments (SITE) in Texas.
                     CryIA(b) protein concentration

                    Percent of total soluble protein

Source              df        MS               F

SITE                 1      0.1284          5.76(*)
REP(SITE)           10      0.0223         11.05(****)
GENE                11      0.0018          0.90 NS
SITE*GENE           11      0.0015          0.77 NS
Error MS           110      0.0020

                        [micro]g [g.sup.-1]

                     MS                F

SITE              8105.40         53.36(****)
REP(SITE)          151.90          5.57(****)
GENE                23.88           0.88 NS
SITE*GENE           34.89           1.28 NS
Error MS            27.26

                        Total soluble protein

                       Percent of fresh weight

                         MS              F

SITE                   101.17   321.64(****)
REP(SITE)               0.31      2.29(*)
GENE                    0.25      1.79 NS
SITE*GENE               0.11      0.82 NS
Error MS                0.14

(*), (****) Significant at the 0.05, and 0.0001 level, respectively; NS, not significant.

Table 6. CryIA(b) insecticidal protein concentration [as percent of total soluble protein (%SP) or [micro]g [g.sup.-1] and total soluble protein concentration (SP) [as percent of fresh weight (%FW)] in young terminal leaves of [F.sub.2:4] recombinant inbred lines possessing insect resistance traits derived in three cotton backgrounds at two Texas locations (sampled College Station, 52 DAP and Tivoli, 42 DAP).
                         College Station, TX

Background([dagger])   Trait([double dagger])
C312/CAMD-E                    WT
C312/ST213                     WT
C312/DP61                      WT
                               GB + N
Overall Mean

                         CryIA(b) protein             SP

                     %SP     [micro]g [g.sup.-1]      %FW

C312/CAMD-E         0.123            32.9             2.72
                    0.138            31.5             2.36
                    0.128            32.1             2.55
                    0.140            33.3             2.41
Mean                0.133            32.5             2.51
C312/ST213          0.137            32.5             2.51
                    0.123            28.8             2.40
                    0.127            32.8             2.69
                    0.120            35.3             2.97
Mean                0.127            32.3             2.64
C312/DP61           0.013            27.6             2.64
                    0.117            31.6             2.85
                    0.122            31.5             2.62
                    0.120            33.1             2.81
Mean                0.115            30.9             2.73
Overall Mean        0.125            31.9             2.63

                                 Tivoli, TX

                        CryIA(b) protein             SP

                     %SP     [micro]g [g.sup.-1]    %FW

C312/CAMD-E         0.162         18.7             1.15
                    0.170         13.7             0.83
                    0.183         17.0             0.90
                    0.168         14.7             0.90
Mean                0.171         16.0             0.95
C312/ST213          0.237         17.2             0.78
                    0.193         17.0             0.95
                    0.175         18.5             1.13
                    0.182         19.7             1.15
Mean                0.197         18.1             1.00
C312/DP61           0.188         20.6             1.14
                    0.185         17.4             0.91
                    0.190         15.9             0.87
                    0.173         12.6             0.74
Mean                0.184         16.6             0.92
Overall Mean        0.184         16.9             0.96

([dagger]) C312 = Coker 312, CAMD-E = TAMCOT CAMD-E, ST213 = Stoneville 213, DP61 = Deltapine 61.

([double dagger]) WT = wild type; GB = glabrous; HT = high terpenoid; N = nectariless; GB + N = glabrous and nectariless.

The CryIA(b) protein and total soluble protein concentrations were also examined twice monthly during bloom in CryIA(b)-positive leaf tissue of the [F.sub.2:4] lines (Fig. 1 and 2). CryIA(b) protein content, expressed as %SP, was influenced by BG (P [is less than] 0.05) and SITE x TIME interaction effects (P [is less than] 0.0001). At Tivoli the CryIA(b) protein content fluctuated over time (P [is less than] 0.001), and BG differences were absent, whereas, at College Station, the CryIA(b) protein content steadily increased (P [is less than] 0.0001) and BG differences were significant (P [is less than] 0.01). Expression of the cryIA(b) gene averaged over time (i.e., over the growing season) was lower in the C312/DP61 background (0.29 %SP) than in C312/ CAMD-E (0.36 %SP) and C312/ST213 (0.39 %SP). In contrast, CryIA(b) protein content, expressed as microgram per gram, was different between locations (P [is less than] 0.0001), across backgrounds (P [is less than] 0.001), and over time (P [is less than] 0.01). Mean CryIA(b) protein concentration was consistently higher at College Station (45.89 [micro]g [g.sup.-1]) than at Tivoli (25.38 [micro]g [g.sup.-1]). At College Station, cryIA(b) gene expression averaged over time was significantly higher in the C312/ST213 background (P [is less than] 0.01) than in the C312/DP61 and C312/CAMD-E backgrounds (56.05 versus 37.55-44.06 [micro]g [g.sup.-1]); whereas, at Tivoli, BG differences were not significant.


Analysis of variance for total soluble protein concentration (percent of fresh weight) during bloom at two locations (Fig. 3) revealed a significant SITE x TIME interaction effect (P [is less than] 0.0001), so SITE and TIME effects were examined separately. BG effects were not significant. Mean total soluble protein content, averaged over backgrounds, varied over time at Tivoli (P [is less than] 0.0001) and at College Station (P [is less than] 0.0001), but the range in total soluble protein content was much greater at College Station (0.70-2.62 %FW) than at Tivoli (0.33-1.22 %FW). Significant differences in mean total soluble protein content were evident at the 47 DAP (P [is less than] 0.0001), 60 DAP (P [is less than] 0.0001), and 76 DAP (P [is less than] 0.01) sampling times, but not at later sampling times. Mean total soluble protein content (%FW) at College Station and at Tivoli was 2.62 versus 1.02 at 47 DAP, 1.85 versus 1.10 at 60 DAP, and 1.23 versus 0.80 at 76 DAP.


Correlation Between Mean cryIA(b) Gene Expression in Selected [F.sub.2] Parents and [F.sub.2:4] Lines

The correlations between mean cryIA(b) gene expression in selected [F.sub.2] parents and their [F.sub.2:4] progenies were similar across backgrounds within a location but differed between locations (Table 7). The values at College Station were consistently high [C312/CAMD-E (r = 0.77), C312/ST213 (r = 0.80), and C312/DP61 (r = 0.85)], while at Tivoli, the values were low [C312/ CAMD-E (r = 0.09), C312/ST213 (r = 0.39), and C312/ DP61 (r = 0.05)].

Table 7. Correlations of CryIA (b) insecticidal protein concentrations (as percent of total soluble protein) between selected [F.sub.2] parental lines and [F.sub.2:4] progeny lines possessing insect resistance traits derived in three cotton backgrounds.
                                 Percent of total soluble protein

Background([dagger])        Trait([double dagger])      [F.sub.2]

C312/CAMD-E                          WT                   0.295
                                     GB                   0.337
                                     HT                   0.345
                                      N                   0.449
Mean                                                      0.357
Correlation (r)
C312/ST213                           WT                   0.446
                                     GB                   0.364
                                     HT                   0.442
                                      N                   0.380
Mean                                                      0.408
Correlation (r)
C312/DP61                            WT                   0.367
                                     GB                   0.468
                                 GB + N                   0.443
                                      N                   0.500
Mean                                                      0.445
Correlation (r)
Pooled correlation (r)

                             Percent of total soluble protein

Background([dagger])           Station, TX      Tivoli, TX

C312/CAMD-E                       0.123           0.162
                                  0.138           0.170
                                  0.128           0.183
                                  0.140           0.168
Mean                              0.132           0.171
Correlation (r)                   0.77            0.09
C312/ST213                        0.137           0.237
                                  0.123           0.193
                                  0.127           0.175
                                  0.120           0.182
Mean                              0.127           0.197
Correlation (r)                   0.80            0.39
C312/DP61                         0.103           0.188
                                  0.117           0.185
                                  0.122           0.190
                                  0.120           0.173
Mean                              0.116           0.184
Correlation (r)                   0.85            0.05
Pooled correlation (r)            0.81            0.18

([dagger]) C312 = Coker 312, CAMD-E = TAMCOT CAMD-E, ST213 = Stoneville 213, DP61 = Deltapine 61.

([double dagger]) WT = wild type; GB = glabrous; HT = high terpenoid; N = nectariless; GB + N = glabrous and nectariless.

Analysis of cryIA Segregation

Phenotypic segregation ratios at the cryIA loci, Chi-square values, and percent seedling emergence for the [F.sub.2] progeny derived from MON 81 [cryIA(b)] or MON 249 [cryIA(c)] and the insect-resistant cotton isolines are in Table 8. Segregation in the [F.sub.2] populations generally fit the Mendelian 3:1 monogenic ratio; however, segregation in the [F.sub.2] populations derived from MON 249 and the CAMD-E isolines deviated from the expected 3:1 ratio (P [is less than] 0.05), wherein the observed 201:100 ratio reflects a deficiency in the number of CryIA-positive plants. Seedling emergence among lines in the C312/CAMD-E background ranged from 10 to 45%, compared with 40 to 90% among lines in the C312/ST213 and C312/DP61 backgrounds.


Genotypic segregation ratios and Chi-square values for the [F.sub.2] progeny derived from MON 81 or MON 249 and the recurrent parent lines are in Table 9. Segregation in all [F.sub.2] populations derived from MON 81 fit the expected 1:2:1 ratio, but segregation in the [F.sub.2] populations derived from MON 249 and CAMD-E deviated. from the expected ratio (P [is less than] 0.05). This population had fewer plants homozygous and hemizygous for cryIA(c) than expected. Analysis of the aggregated segregation data for all the populations derived from MON 249 and the recurrent parent lines also revealed a skewed ratio with fewer homozygotes and hemizygotes (P [is less than] 0.005). Assuming that CryIA-negative wild-type plants are no less fit, the proportion of homozygotes was reduced by 51% and the proportion of hemizygotes was reduced by 12%.

Table 9. Segregation at the cryIA (b) and cryIA (c) loci in [F.sub.2] progeny of crosses between the recurrent parent varieties and modified Coker 312 in three cotton backgrounds.
                                cryIA(b) locus
                              [F.sub.2] progeny

Background      cry/cry   cry/-    -/-    [chi square]   P (df = 2)

C312/CAMD-E       8          8      4        2.40        0.3012
C312/ST213       10         32     15        1.74        0.4196
C312/DP61        20         32     21        1.14        0.5664
Total            40         72     38        0.29        0.8636
Sum (df = 6)                                 5.28        0.5084

                            CryIA(c) locus
                            [F.sub.2] progeny

Background      cry/cry   cry/-    -/-    [chi square]   P (df = 2)

C312/CAMD-E        8       30      23          7.39        0.0248
C312/ST213        14       48      26          2.75        0.2528
C312/DP61          9       33      14          4.00        0.1353
Total             31      111      63         11.40        0.0033
Sum (df = 6)                                  14.14        0.0281

([dagger]) C312 = Coker 312, CAMD-E = TAMCOT CAMD-E, ST213 = Stoneville 213, DP61 = Deltapine 61.


This study was undertaken to characterize the expression of two cryIA insecticidal protein gene inserts in insect-resistant cotton lines derived in three backgrounds. cryIA gene expression in CryIA-positive [F.sub.2] progeny derived from both inserts was influenced by the site-of-gene insertion, background, and plant effects. CryIA concentration (as %SP), averaged over populations and backgrounds, was two-fold higher in plants with the cryIA(b) gene insert from MON 81 than in plants with the cryIA(c) gene insert from MON 249. Perlak et al. (1990) reported two-fold higher expression of a cryIA(c) gene insert compared to a cryIA(b) gene insert in certain Coker 312 lines; however, other transformed plants were recovered with different levels of expression of both genes. Since both genes utilize the same promoter and have a similar nucleotide sequence (Perlak et al., 1991), their findings were explained as insertion effects. Our study provides further evidence for gene insertion effects, such as, site, epistasis, or gene silencing, and shows that these effects are heritable in that they caused similar effects in several different genetic backgrounds of [F.sub.2] families (Table 1).

Significant background effects were also found in CryIA-positive [F.sub.2] populations. The CryIA protein concentration (%SP) was 19% lower in the C312/ST213 background than in the C312/CAMD-E and C312/DP61 backgrounds. It is not known whether background effects of this magnitude could significantly affect plant insect-resistance; however, backgrounds with superior expression might be judged by farmers to be more desirable, and could affect selection pressure for insects to adapt to CryIA proteins.

Plant effects on cryIA gene expression caused by genetic differences unique to each [F.sub.2] plant may have resulted from epistatic effects and/or somaclonal mutations inherited from the Coker 312 regenerant. Epistatic and somaclonal effects should not be underestimated when evaluating new gene inserts for crop improvement. Positive epistatic effects on gene expression could cause breeders to overlook more desirable plants in favor of less desirable plants, and negative epistatic effects could make an otherwise superior plant look inferior. Somaclonal effects could be similarly misleading. To increase the likelihood of developing a superior cultivar, a large number of [R.sub.0] plants should be selected for advanced evaluation. Moreover, our results show that CryIA protein production should be managed as a quantitative trait, despite the fact that CryIA protein is encoded by a single gene. Breeders will need to establish acceptable levels of gene expression, then make certain that these levels are maintained throughout the breeding and improvement process.

Gene dosage at the cryIA locus influenced CryIA protein concentration in [F.sub.2] populations with the cryIA (b) gene insert. In these populations, homozygotes produced 14% more CryIA protein than hemizygotes. No homozygote advantage was found in lines with the cryIA(c) gene insert; instead, there was a tendency for a hemizygote advantage in two out of three backgrounds. From these results, identification of homozygous cry/cry genotypes based solely on CryIA protein production is not recommended, since hemizygotes with superior epistatic or heterotic effects, may be misidentified as homozygotes. Consequently, random sampling from a pool of CryIA-positive plants followed by a progeny test is recommended.

The cryIA(b) gene expression was uniform before bloom in young terminal leaves of [F.sub.2:4] lines planted at the same location. However, the variation in cryIA gene expression due to plant effects present in the [F.sub.2] parental populations, was absent in the [F.sub.2:4] lines. The loss of variation was probably due to the selection scheme and the accumulated inbreeding due to selfing. Selection of [F.sub.2] parental plants from among plants with the highest cryIA gene expression, may have eliminated negative effects caused by epistatic interactions and/or somaclonal mutations. Also, inbreeding would increase homozygosity at all loci, resulting in less genetic variability and epistatic interactions involving dominance. The homogeneity in cryIA(b) gene expression produced by selection and inbreeding suggests that the high variation in expression among the [F.sub.2] populations was caused by epistatic and/or somaclonal effects. Moreover, the results indicate that gene expression in [F.sub.2] plants derived from crosses between regenerated plants and various breeding lines is not a reliable predictor of gene expression in subsequent generations.

Uniform cryIA gene expression among insect-resistant [F.sub.2:4] cotton lines is desirable. Variable gene expression would make it difficult to determine the additive effects of plant insect-resistance traits in lines expressing a cryIA gene. Any advantages afforded by the plant insect-resistance traits would be confounded with varying cryIA gene expression. Uniform gene expression among lines is essential for comparative evaluation of cryIA lines with pyramided plant insect-resistance genes.

cryIA (b) gene expression in the [F.sub.2:4] lines differed dramatically between locations and among replicates within a location. Quantification of CryIA(b) protein as %SP and microgram per gram produced strikingly different results. Using %SP, expression appeared to be higher at Tivoli than College Station; however, using microgram per gram revealed that more CryIA(b) protein was actually present in plants at the College Station location. The differences observed were mostly due to differences in total soluble protein concentration between the locations. Lower total soluble protein synthesis at Tivoli combined with a disproportionate decrease in cryIA(b) gene expression resulted in an apparent increase in cryIA(b) gene expression when calculated as %SP. These two modes of quantification reflect different aspects of gene expression; %SP reflects the magnitude of cryIA(b) expression relative to the level of expression of other soluble protein genes, while microgram per gram of plant tissue reflects the magnitude of cryIA (b) expression independent of the level of expression of other soluble protein genes. These results suggest that cryIA expression and soluble protein synthesis were influenced by similar factors, but they responded to these factors differently. Since total soluble protein concentration may change from location to location and over time within a location, as shown here, the mode of quantification must be chosen and interpreted carefully.

The variation in CryIA(b) protein concentration between locations and among replicates within a location strongly implies an environmental effect on expression. This should be expected, since many factors, such as temperature, moisture, and fertility influence plant gene expression. Any factor that affects soluble protein synthesis or degradation is also likely to influence CryIA(b) protein concentration. Differences in soil moisture content between the locations were unavoidable since equipment failures prevented irrigation at Tivoli. When low rainfall during June and early July resulted in marginal soil moisture at Tivoli, plant moisture stress may have contributed to decreases in total soluble protein and CryIA(b) protein concentrations.

The influence of environmental factors on cryIA(b) gene expression and soluble protein synthesis was also evident in CryIA-positive wild-type [F.sub.2:4] lines examined twice monthly during bloom. CryIA(b) protein, expressed as %SP or micrograms per milliliter, and total soluble protein concentrations varied by location and over time. Since the same [F.sub.2:4] lines were planted at both locations, these differences must have been due to environmental rather than genetic effects.

CryIA(b) protein concentration in [F.sub.2:4] lines, expressed as %SP or micrograms per milliliter, was also influenced by background effects. These background effects were not observed in plant samples collected before bloom, but were evident when samples were collected throughout the bloom period. These results suggest that the physiological concentrations of CryIA(b) protein may differ among cotton cultivars, even if they were derived from the same cryIA(b) gene insertion event.

Research on the relationship between cryIA gene expression and level of plant insect-resistance is needed. If increased gene expression leads to improved plant insect-resistance, it may be prudent for seed companies to select for increased cryIA expression since higher expressing cultivars could have a marketing advantage. However, if increased gene expression has no effect on plant insect-resistance, then development of new varieties expressing cryIA genes would be dramatically simplified.

Environmental factors also influenced parent-off-spring correlations of mean cryIA(b) gene expression between individuals from the [F.sub.2] and the [F.sub.2:4] generations. The correlations were higher at College Station than at Tivoli. Correlation values substantially less than 0.89 reflect a significant contribution of nonadditive or environmental effects to phenotypic values. Values near to 0.89 indicate that these factors are less important, especially when the number of individuals used to estimate the correlations is small. At College Station, the correlation values (0.77-0.85) were near to 0.89; consequently, the small number of individuals used to estimate the correlation was sufficiently large to diminish minor nonadditive genetic and environmental effects. At Tivoli, the correlation values (0.05-0.39) were much less than 0.89, indicating that nonadditive genetic and environmental effects strongly influenced cryIA (b) gene expression. Since expression did not vary before bloom between lines or across backgrounds within a location, it is unlikely that nonadditive genetic effects were an important source of variation. It is more likely that environmental factors, such as moisture and fertility, impacted expression and increased the phenotypic variation. It is recommended that these environmental factors be made as uniform as possible when comparing cryIA(b) gene expression across several locations or years. This will help to ensure that observed differences are genetically based rather than environmentally induced.

The CryIA(b) phenotype segregated as a simple dominant Mendelian trait. This was consistent with the earlier report by Perlak et al. (1990). Non-Mendelian segregation of the CryIA(c) phenotype was observed in [F.sub.2] populations derived from MON 249 and the insect-resistant CAMD-E isolines, but reduced fitness rather than abnormal transmission was suggested. These [F.sub.2] populations that were segregating in a non-Mendelian manner had reduced germination compared with populations derived from MON 249 and the insect-resistant ST213 or DP61 isolines. This suggests that a disproportionately large number of MON 249 x CAMD-E plants inheriting the cryIA(c) gene insert failed to germinate normally. Non-Mendelian segregation also was observed in [F.sub.2] populations derived from MON 249 and each recombinant parent line. The cryIA genotype of each plant within a population was determined by the progeny test at plant maturity. Any association between genotype and reduced germination or survival to maturity could be reflected as a deviation from the expected 1:2:1 genotypic ratio. The aggregated segregation data revealed a skewed ratio with too few homozygotes and hemizygotes. These data suggest that a fitness cost is associated with the inheritance of the cryIA (c) gene insert in MON 249 plants. This hypothesis is strengthened by the fact that fitness was more affected in plants inheriting two copies of this cryIA(c) insert than in plants inheriting only one copy.

The reduced fitness associated with the cryIA(c) gene insert in MON 249 plants may be the result of a direct insertion effect resulting in the silencing of one or more native genes or the result of a linked somaclonal mutation. It was not possible to differentiate between the two without backcrossing to see if normal fitness could be restored. Theoretically, backcrossing could eliminate all but very tightly linked somaclonal mutations, while persistence of the deleterious trait would be evidence for a direct insertion effect.


Our results indicate that cryIA gene expression in terminal leaves is variable and influenced by genetic and environmental factors. The strong influence of environmental factors on cryIA gene expression shows the importance of maintaining uniformity of environmental factors, presumably soil moisture and fertility, in experiments designed to evaluate cryIA expression or plant insect-resistance. Further studies to elucidate the specific environmental factors and to quantify their influence on CryIA gene expression seem justified. This could be particularly important as it relates to the levels of CryIA expression and insect control in production agriculture. Variation in CryIA gene expression from gene insertion and cotton background effects could also provide significant opportunities for increasing cryIA expression using traditional breeding techniques. Our data show that variation from epistatic and/or somaclonal effects caused cryIA gene expression to behave as a quantitative trait, consequently breeders should monitor cryIA expression during the cultivar development process to insure that acceptable expression levels are maintained.

Abnormal phenotypic ratios in [F.sub.2] progeny derived from MON 249 may have resulted from foreign gene insertion and/or somaclonal effects on germination and survival. Understanding the cause of the abnormal segregation ratios could avoid an over commitment of time and resources to development of an unfit line containing deleterious insertion effects or a tightly linked somaclonal mutation. Moreover, simultaneous development of several inserts based on a single gene construct should increase the likelihood that one or more lines are suitable for commercial development. In summary, our results show that pyramiding the foreign cryIA gene with native insect resistant traits, in a single genetic background, can be successfully accomplished by a traditional backcross breeding program (See Altman 1993 for a more complete discussion). Moreover, we have shown, in a separate study with these isolines, that the suppression of insect injury was greatest in the pyramided isoline possessing both the CryIA(b) and high-terpenoid traits in the ST213 background (Sachs et al., 1996).


The authors thank Diane Alertas, Reagan De Spain, Mike Goynes, Melissa Orr, Dennis Ring, Steve Scoggin, and James Green for their assistance in conducting these studies. A special thanks to Carolyn Villanueva for editing and word processing of the manuscript. This research was supported in part by financial assistance from Monsanto Company, Texas Agricultural Experiment Station, USDA-CSRS Grant 90-34103-4986, Texas Higher Education Coordinating Board Grants 999902-144 and 999902-023, and Texas Agricultural Diversification Program Grant 429009-72. This article was approved for publication by the Texas Agricultural Experiment Station, College Station, TX, as Technical Article 31270.

Abbreviations: CaMV, cauliflower mosaic virus; DAP, days after planting; ELISA, enzyme-linked immunosorbance assay; FW, fresh weight; MON, Monsanto.


Adang, M.J. 1991. Bacillus thuringiensis insecticidal crystal proteins: gene structure, action and utilization. p. 3-24. In K. Maramorosch (ed.) Biotechnology for biological control of pests and vectors. CRC Press, Boca Raton, FL.

Altman, D.W. 1993. Gene introgression for cotton improvement: contrast of traditional breeding with biotechnologies, p. 193-216. In J.P. Gustafson et al. (ed.) Gene conservation and exploitation. Plenum Press, New York.

Altman, D.W., D.M. Stelly, and D.M. Mitten. 1991. Quantitative trait variation in phenotypically normal regenerants of cotton. In Vitro Cell. Dev. Biol. 27P:132-138.

Altman, D.W., R.D. Stipanovic, and J.H. Benedict. 1989a. Terpenoid aldehydes in upland cottons. II. Genotype-environment interactions. Crop Sci. 29: 1451-1456.

Altman, D.W., R.D. Stipanovic, J.H. Benedict, M.H. Elissalde, and J.O. Norman. 1989b. Breeding for resistance to Heliothis. Ann. Plant Resist. Newsl. 15:34-35.

Aronson, A.I., W. Beckman and P. Dunn. 1986. Bacillus thuringiensis and related insect pathogens. Microbiol. Rev. 50:1-24.

Benedict, J.H., D.W. Altman, P.F. Umbeck, and D.R. Ring. 1992. Behavior, growth, survival, and plant injury for Heliothis virescens (F.) (Lepidoptera: Noctuidae) on transgenic Bt cottons. J. Econ. Entomol. 85:589-593.

Benedict, J.H., E.S. Sachs, D.W. Altman, D.R. Ring, T.B. Stone, and S.R. Sims. 1993. Impact of [Delta]-endotoxin-producing transgenic cotton on insect-plant interactions with Heliothis virescens and Helicoverpa zea(Lepidoptera: Noctuidae). Environ. Entomol. 22:1-9.

Benedict, J. H., E.S. Sachs, D.W. Altman, W.R. Deaton, R.J. Kohel, D.R. Ring, and S.A. Berberich. 1996. Field performance of cottons expressing transgenic CryIA insecticidal proteins for resistance to Heliothis virescens and Helicoverpa zea (Lepidoptera: Noctuidae). J. Econ. Entomol. 89: 230-238.

Berberich, S.A., S.C. MacIntosh, and R.L. Fuchs. 1990. Quantification of B.t.k. proteins in plants by ELISA. Fourth Symposia of the Protein Society, S28, San Diego, CA.

Entwistle, P.F., J.S. Cory, M.J. Bailey, and S. Higgs (ed.) 1993. Bacillus thuringiensis, An environmental biopesticide: Theory and practice. John Wiley, New York.

Falconer, D.S. 1989. Introduction to quantitative genetics. 3rd. Ed. Longman Scientific & Technical, New York.

Fuchs, R.L., S.C. MacIntosh, D.A. Dean, J.T. Greenplate, F.J. Perlak, J.C. Pershing, P.G. Marrone, and D.A. Fischhoff. 1990. Quantification of Bacillus thuringiensis insect control protein as expressed in transgenic plants. Chapter 12. In Analytical chemistry of Bacillus thuringiensis. ACS Symposium Series No. 432. ACS, Washington, DC.

Gasser, C.S., and R.T. Fraley. 1989. Genetically engineering plants for crop improvement. Science 244:1293-1299.

Gould, F. 1988. Evolutionary biology and genetically engineered crops. BioScience 38:26-33.

Hofte, H., and H.R. Whiteley. 1989. Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53:242-255.

Kohel, R.J., and J.A. Lee. 1984. Genetic analysis of Egyptian glandless cotton. Crop Sci. 24:1119-1121.

Lee, J.A. 1985. Revision of the genetics of the hairiness-smoothness system of Gossypium. J. Hered. 76:123-126.

Meyer, J.R., and V.G. Meyer. 1961. Origin and inheritance of nectariless cotton. Crop Sci. 1:167-169.

Perlak, F.J., R.W. Deaton, T.A. Armstrong, R.L. Fuchs, S.R. Sims, J.T. Greenplate, and D.A. Fischhoff. 1990. Insect resistant cotton plants. Bio/Technology 8:939-943.

Perlak, F.J., R.L. Fuchs, D.A. Dean, S.L. McPherson, and D.A. Fischhoff. 1991. Modification of the coding sequence enhances plant expression of insect control protein genes. Proc. Natl. Acad. Sci. (USA) 88:3324-3328.

Sachs, E.S., J.H. Benedict, J.F. Taylor, D.M. Stelly, S.K. Davis, and D.W. Altman. 1996. Pyramiding CryIA(b) insecticidal protein and terpenoids in cotton to resist tobacco budworm (Lepidoptera: Noctuidae). Environ. Entomol. 25:1257-1266.

Satterthwaite, F.W. 1946. An approximate distribution of estimates of variance components. Biometrics Bull. 2:110-114.

Sims, S.R., and S.A. Berberich. 1996. Bacillus thuringiensis CryIA protein levels in raw and processed seed of transgenic cotton: determination using insect bioassay and ELISA. J. Econ. Entomol. 89:247-251.

Stelly, D.M., D.W. Altman, R. Kobel, T.S. Rangan, and E. Commiskey. 1989. Cytogenetic abnormalities of cotton somaclones from callus cultures. Genome 32:762-770.

Thompson, N.J. 1987. Host plant resistance in cotton. J. Aust. Inst. Agric. Sci. 53:262-270.

Trolinder, N.L., and J.R. Goodin. 1987. Somatic embryogenesis and plant regeneration in cotton (Gossypium hirsutum L.). Plant Cell Reports 6:231-234.

Wilson, F.D., and J.N. Smith. 1977. Variable expressivity and gene action of gland-determining alleles in Gossypium hirsutum L. Crop Sci. 17:539-543.

Wilson, F.D., H.M. Flint, W.R. Deaton, D.A. Fischhoff, F.J. Perlak, T.A. Armstrong, R.L. Fuchs, S.A. Berberich, N.J. Parks, and BR. Stapp. 1992. Resistance of cotton lines containing a Bacillus thuringiensis toxin to pink bollworm (Lepidoptera: Gelechiidae) and other insects. J. Econ. Entomol. 85:1516-1521.

E. S. Sachs, J. H. Benedict,(*) D. M. Stelly, J. F. Taylor, D. W. Altman, S. A. Berberich, and S. K. Davis

E.S. Sachs and S.A. Berberich, Monsanto Company, 700 Chesterfield Parkway North, St. Louis, MO 63198; J.H. Benedict, Texas A&M Univ. Res. & Ext. Center, Route 2, Box 589, Corpus Christi, TX 78406; D.M. Stelly, Texas A&M Univ., Dep. of Soil and Crop Science, College Station, TX 77843; J.F. Taylor and S.K. Davis, Texas A&M Univ., Dep. of Animal Science, College Station, TX 77843; and D.W. Altman, ProfiGen, 800 Harrison St., Nashville, TN 37203. Part of a dissertation submitted by the senior author in partial fulfillment of the requirements for a Ph.D. degree at Texas A&M Univ. Received 24 April 1997. (*)Corresponding author (J-Benedict@TAMU.EDU).

Published in Crop Sci. 38:1-11 (1998).
COPYRIGHT 1998 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Sachs, E.S.; Benedict, J.H.; Stelly, D.M.; Taylor, J.F.; Altman, D.W.; Berberich, S.A.; Davis, S.K.
Publication:Crop Science
Date:Jan 1, 1998
Previous Article:An encoding process for morphological analysis of soybean fruit distribution.
Next Article:Transgressive segregation of tiller angle in rice caused by complementary gene action.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |