Herbicide-Resistant Plants Carrying Mutated Acetolactate Synthase

sequences obtained from herbicide-resistant yeast mutants, two patterns have become clear. First, the ALS sequences that can be mutated to cause resis...
0 downloads 0 Views 2MB Size
Chapter 31

Herbicide-Resistant Plants Carrying Mutated Acetolactate Synthase Genes

Downloaded by UNIV OF NORTH CAROLINA on June 20, 2013 | http://pubs.acs.org Publication Date: February 23, 1990 | doi: 10.1021/bk-1990-0421.ch031

Mary E. Hartnett, Chok-Fun Chui, C. Jeffry Mauvais, Raymond E. McDevitt, Susan Knowlton, Julie K. Smith, S. Carl Falco, and Barbara J. Mazur Agricultural Products Department, E. I. du Pont de Nemours and Company, Experimental Station, P.O. Box 80402, Wilmington, DE 19880-0402

Acetolactate synthase (ALS) is the target enzyme for three unrelated classes of herbicides, the sulfonylureas, the imidazolinones, and the triazolopyrimidines. We have cloned the genes which specify acetolactate synthase from a variety of wild type plants, as well as from plants which are resistant to these herbicides. The molecular basis of herbicide resistance in these plants has been deduced by comparing the nucleotide sequences of the cloned sensitive and resistant ALS genes. By further comparing these sequences to ALS sequences obtained from herbicide-resistant yeast mutants, two patterns have become clear. First, the ALS sequences that can be mutated to cause resistance are in domains that are conserved between plants, yeast and bacteria. Second, identical molecular substitutions in ALS can confer herbicide resistance in both yeast and plants. These findings have been extended by oligonucleotide directed in vitro mutagenesis of plant ALS genes, followed by introduction of the mutated genes into sensitive plants. The herbicide-resistant transgenic plants so produced provide additional evidence for the commonality of mutations which specify herbicide resistance in ALS genes. Some implications of this work for predicting and addressing the problem of herbicide-resistant weeds are discussed. Commercial herbicides are traditionally discovered by screening chemical compounds for toxicity to weeds and secondarily for lack of toxicity to particular crop species. Selective activity to weeds but not to crops is often due to metabolism of the herbicide to a non-toxic derivative by the crop. Development of one new herbicide often requires the expensive screening of thousands of compounds and generally is carried out only for the major crop species. To increase grower options by enabling the use of desirable herbicides on additional crop species, to increase the margin of safety in the use of selective herbicides, and to increase options for crop rotations, genetic modification of crops to express resistance to herbicides can be used. Herbicide resistance can be introduced into plants by three different strategies. One method is mutation breeding, in which mutations are introduced into the germ line through chemical mutagenesis of the seeds or pollen. A second method employs plant cell culture to select individual resistant cells in vitro followed by regeneration of plants from the resistant cells. A third method, which will be discussed in this paper, is genetic transformation. In this approach genes that code for altered 0097-6156/90/0421-0459$06.00/0 © 1990 American Chemical Society

In Managing Resistance to Agrochemicals; Green, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

460

MANAGING

RESISTANCE TO AGROCHEMICALS

herbicide target proteins that are not inhibited by the herbicide, or genes that code for herbicide detoxifying enzymes, are transferred to plant cells, and resistant plants are regenerated. We have successfully used the genetic transformation technique to produce crop plants that are resistant to the sulfonylurea herbicides.

Downloaded by UNIV OF NORTH CAROLINA on June 20, 2013 | http://pubs.acs.org Publication Date: February 23, 1990 | doi: 10.1021/bk-1990-0421.ch031

BACKGROUND The sulfonylurea herbicides are a new family of chemical compounds, some of which are selectively toxic to weeds but not to crops. The selectivity of the sulfonylureas results from their metabolism to non-toxic compounds by particular crops, but not by weeds. In addition to efficient weed control, the sulfonylurea herbicides provide environmentally desirable properties such as field use rates as low as two grams/hectare and very low toxicity to mammals. The high specificity of the herbicides for their molecular target contributes to both of these properties. In addition, the low toxicity to mammals results from their lack of the target enzyme for the herbicides. Sulfonylureas inhibit the enzyme acetolactate synthase (ALS), also known as acetohydroxyacid synthase (AHAS), which catalyzes the first common step in the biosynthesis of the branched chain amino acids leucine, isoleucine and valine. In mammals these are three of the essential amino acids which must be obtained through dietary intake because the biosynthetic pathway for the branched chain amino acids is not present. The prototype structure of a sulfonylurea herbicide is shown in Figure 1. Prior to the identification of the site of action of the sulfonylurea herbicides, herbicide-resistant tobacco plants were selected in vitro using cell culture techniques. Genetic studies of the resistant plants indicated that the resistance phenotype was dominant or semi-dominant and segregated as a single nuclear gene in one of two linkage groups (1). The site of action of the sulfonylurea herbicides was identified through physiological studies in bacteria. Sulfonylurea herbicides inhibited the growth of some bacteria on minimal media, but not on rich media or on minimal media supplemented with the branched chain amino acids. Biochemical studies indicated that bacteria produced an ALS enzyme that was inhibited by the sulfonylurea herbicide sulfometuron methyl at nanomolar concentrations (2). Three isozymes of ALS are present in Escherichia coli and Salmonella typhimuriwn; ALS II

and ALS III are sensitive to inhibition by the herbicide, whereas ALS I is insensitive to the herbicide (2,3). Immediately following the discovery of the site of action of the sulfonylurea herbicides in bacteria, the sensitivity of plant ALS to sulfonylurea herbicides was demonstrated (4) and biochemical, genetic, and physiological studies showed that resistant tobacco plants expressed a herbicide-insensitive form of the ALS enzyme that cosegregated with the herbicide-resistant whole plant phenotype (5). Herbicide-resistant mutants of bacteria, yeast, and algae have been isolated (2,6,7). In genetic studies of these organisms the herbicide-resistance phenotype has been shown to segregate with the herbicide-insensitive ALS enzyme. The genes coding for the resistant form of ALS in bacteria and yeast have been mapped to the ilvG and ILV2 loci, respectively, and both sensitive and resistant forms of the genes have been sequenced. ALS has also been shown to be inhibited by two other structurally unrelated classes of herbicides, the imidazolinones (8,9) and the triazolopyrimidines (10,11). It has been shown that the toxicity of the sulfonylurea herbicides to bacteria is due, in part, to the accumulation of an ALS substrate a-ketobutyrate, which is itself toxic. It has been suggested that the dual effects of the accumulation of a toxic substrate and the inability to synthesize isoleucine, leucine and valine make ALS a particularly good target for herbicides (12).

In Managing Resistance to Agrochemicals; Green, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

31. HARTNETTETAL.

Herbicide-Resistant Plants Carrying ALS Genes

Downloaded by UNIV OF NORTH CAROLINA on June 20, 2013 | http://pubs.acs.org Publication Date: February 23, 1990 | doi: 10.1021/bk-1990-0421.ch031

PLANT ALS GENES The ALS gene from the yeast Saccharomyces cerevisiae was cloned using a high copy number plasmid library to rescue wild type yeast from minimally inhibitory concentrations of herbicide. The resistance was due to overproduction of the enzyme from the plasmid, which allowed cells to overcome the inhibition by the herbicide (6). Localization of the ALS gene on the plasmid was carried out by deletion analysis, by transposon mutagenesis (13), and by heterologous hybridization to a cloned Salmonella typhimurium ALS gene. The hybridization observed between the yeast and bacterial genes indicated sequence conservation between ALS genes from distantly related organisms Subsequent comparisons of the deduced amino acid sequences from the yeast and bacterial genes,revealed three regions of extensive homology separated by four nonconserved regions (15). These results suggested that hybridization probes could be designed to isolate ALS genes from other organisms (14). To explore this possibilty, fragments of the yeast ALS gene were hybridized to genomic DNAfromAnabaena 7120 to determine the optimum fragment giving cross-hybridization. Afragmentencompassing almost the entire ALS coding region of the yeast clone was chosen and used to screen genomic libraries of Anabaena 7120, Arabidopsis thaliana and Nicotiana tabacum (tobacco), under low

stringency conditions. Hybridization was detected to genomic clones from all three species and phage carrying the presumptive ALS genes were isolated (16). Complete sequence analyses of the Arabidopsis and tobacco ALS genes indicated that they code for proteins of 670 (2013 bp) and 667 (2004 bp) amino acids, respectively, with predicted molecular weights of about 73 kilodaltons. Neither of the two genes contain introns, as had been inferred from hybridization experiments with the yeast gene probe which indicated that the plant genes were coded by contiguous 2 Kb DNA fragments. The plant genes are highly conserved, with 75% sequence identity at the nucleotide level and 85% at the amino acid level (16). When compared to the deduced amino acid sequences of ALS from yeast and bacteria, both plant ALS enzymes maintain the three conserved regions previously noted in the microbial enzymes. The sequence similarity between the plant genes, however, extends into the four non-conserved regions from bacteria and yeast (Figure 2). One region of the plant genes, the 5' end of the coding sequence, is not conserved between Arabidopsis and tobacco. This portion of each gene is believed to code for a chloroplast transit sequence, since ALS is nuclear encoded (1,5) but the enzyme functions in the chloroplast (17,18). Because the degree of homology in this region is greater at the nucleotide level than at the amino acid level and the hydrophobicity plots of these transit peptides reveal regions of similar profiles, it was suggested that the hydrophobicity characteristics are more important for function than is the primary amino acid sequence (16). Southern blot analyses of genomic DNA from Arabidopsis and tobacco were used to determine the number of ALS genes in these two species. When the genomic blots were hybridized with the isolated ALS genes as probes, two fragments of digested tobacco DNA, and one fragment of digested Arabidopsis DNA hybridized (16). This suggested that there are two ALS genes in tobacco and one in Arabidopsis. The molecular data correlated well with genetic data which had identified two distinct loci that could mutate to yield herbicide-resistant ALS enzymes in tobacco (1,5), which is an allotetraploid. Similar DNA hybridization analyses of soybean and maize indicated that each contains multiple ALS genes. HERBICIDE-RESISTANT ACETOLACTATE SYNTHASE Genomic DNA libraries were prepared from two herbicide-resistant lines of tobacco, the C3 and Hra lines, and probed with the cloned tobacco ALS gene. The C3 line is 100-fold more resistant to the sulfonylurea herbicide chlorsulfuron than is wild type

In Managing Resistance to Agrochemicals; Green, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

461

Downloaded by UNIV OF NORTH CAROLINA on June 20, 2013 | http://pubs.acs.org Publication Date: February 23, 1990 | doi: 10.1021/bk-1990-0421.ch031

462

MANAGING RESISTANCE TO AGROCHEMICALS

,R

S0NHCONH —( 2

R Sulfonylurea

FIGURE 1

Prototype structure of the sulfonylurea herbicides.

1

100

200

300

400

500

600

I

I

I

I

I

I

I

Plant E. c o l l

Ml •!

I n i g

B||

H

I

FIGURE 2 Amino acid conservation between acetolactate synthase genes. The numbers indicate amino acid residues. The first bar represents a comparison of the deduced amino acid sequences of ALS from tobacco and Arabidopsis; the second bar represents a comparison between the amino acid sequences of the three E. coli ALS isozymes. Regions of conservation are shown in white.

In Managing Resistance to Agrochemicals; Green, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

Downloaded by UNIV OF NORTH CAROLINA on June 20, 2013 | http://pubs.acs.org Publication Date: February 23, 1990 | doi: 10.1021/bk-1990-0421.ch031

31. HARTNETTETAL.

Herbicide-Resistant Plants Carrying ALS Genes

463

tobacco and has been shown genetically to be mutated at the SuRA locus (1). The Hra tobacco line, which was isolated by two successive rounds of selection to obtain a highly resistant mutant, is 1000-fold more resistant and is mutated at the SuRB locus (19). Two different genes, representing the two genetic loci, SuRA and SuRB, were isolatedfromboth mutant lines. Comparison of the DNA sequences of the genes from both loci indicated 97% identity at the nucleotide level and 99% identity at the amino acid level in the mature protein regions. The putative transit peptide regions differ considerably more, as a result of in-frame nucleotide duplications or deletions and 23 nucleotide substitutions (20). The complete DNA sequence of the four genes revealed the identity of the mutations at each of the loci that confer herbicide resistance. The SuRA gene from the C3 line codes for a substitution of glutamine for proline at amino acid position 194. A mutation at the analogous site in the yeast ALS gene (amino acid position 192), which is within a conserved region of all ALS genes, had previously been shown to confer resistance (21). Interestingly, the SuRB gene from the Hra line codes for an alanine for proline substitution at the analogous position (amino acid 191), as well as a leucine for tryptophan substitution at amino acid position 568 (20). The two mutations found in this gene are consistent with the two cycles of selection used to isolate the Hra line. The tryptophan residue at position 568 is within another highly conserved region of ALS and an analogous mutation was found in the yeast ALS gene (see below). Similarly, the mutant ALS gene from a herbicide-resistant Arabidopsis line was isolated using the wild type Arabidopsis ALS gene as a probe and sequenced. The mutation in the resistance gene encoded a serine substitution for the analogous proline (amino acid position 197) in the Arabidopsis protein (22). The identification of mutations in diverse ALS genes coding for substitutions of the equivalent proline or tryptophan residues lead us to postulate that mutations resulting in herbicide resistance might generally be conserved. The resultsfromthree different experimental approaches are consistent with this postulate. The first approach employed selection of mutations in the yeast ALS gene to uncover additional sites of herbicide-resistance mutations. Forty-one independently isolated spontaneous mutations in the yeast ALS gene were characterized by DNA sequencing. These mutations revealed 24 different amino acid substitutions that occur at 10 different sites ranging from the amino to the carboxy ends of the protein. The amino acids at these 10 sites are highly conserved among natural herbicidesensitive ALS enzymes; the amino acid residues present in the wild type herbicidesensitive yeast enzyme have been found in all wild type plant ALS enzymes that have been sequenced. Site-directed mutagenesis was used to make additional amino acid substitutions at these sites in yeast ALS. At some of the sites, e.g. alall7, prol92, or trp586, nearly any substitution for the wild type amino acid that was tested resulted in a herbicide-resistant enzyme (Table I). Each of the mutant enzymes was characterized by enzyme assays to compare its activity, and its sensitivity to the sulfonylurea herbicide chlorimuron ethyl, to the wild type enzyme. These analyses have indicated that some of the mutations have little adverse effect on the activity of the enzyme, while decreasing sensitivity to the herbicidefromthree to greater than one thousandfold. The characteristics of these mutant enzymes were further evaluated in vivo in order to investigate the utility of particular herbicide/mutant enzyme combinations (Falco et al., manuscript in preparation). In the second approach, herbicide-resistance mutations in the Arabidopsis ALS gene were studied in E. coli. To do this, wild type and mutant Arabidopsis genes were functionally expressed in E. coli, such that the plant genes complemented a branched chain amino acid auxotrophy in the bacteria (Smith et al. 1989, PNAS in press). ALS enzyme assays on extracts prepared from E. coli expressing the mutant Arabidopsis gene indicated that the mutant enzyme is resistant to sulfonylurea herbicides but is sensitive to the imidazolinone herbicide imazaquin. This selective

In Managing Resistance to Agrochemicals; Green, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

Downloaded by UNIV OF NORTH CAROLINA on June 20, 2013 | http://pubs.acs.org Publication Date: February 23, 1990 | doi: 10.1021/bk-1990-0421.ch031

464

MANAGING

RESISTANCE TO AGROCHEMICALS

resistance of the Arabidopsis mutation when expressed in E. coli is consistent with the activity measurements determined in plants (23). A simple in vivo assay employing filter paper disks impregnated with herbicides has allowed the screening of the mutant genes against a large number of herbicides. These two types of assays thus provide a rapid and facile means for determining the sensitivity of mutant plant enzymes to a variety of herbicides. The third approach used oligonucleotide site-directed mutagenesis of the tobacco ALS gene to analyze mutations that confer herbicide resistance. Among the three mutant genes isolated from the herbicide-resistant tobacco and Arabidopsis plants, two mutation sites were identified. Mutations were made in the wild type genefromthe SuRA locus at these sites and at additional sites predicted to confer resistancefromthe yeast experiments. DNAfragmentscarrying the mutations were subcloned into the wild type gene from the SuRB locus via a common restriction enzyme fragment creating chimeric ALS genes (Figure 3). This resulted in expression of the genefromthe more active SuRB promoter and permitted evaluation of the mutations in a novel ALS enzyme to further test the generality of resistance mutations. Tobacco protoplasts were transformed with mutant and wild type chimeric tobacco ALS genes by direct DNA uptake and analyzed for the ability to produce calli on minimally inhibitory concentrations of the herbicide. In the experiment shown in Table II, four novel herbicide-resistance mutations in the tobacco ALS gene were identified. When introduced into the chimeric tobacco ALS, a proline to serine change at amino acid residue 191, which had been shown to produce herbicideresistant enzymes from yeast and Arabidopsis, allowed the formation of herbicideresistant tobacco calli. The protoplast transformation experiment also revealed that substitution of leucine for tryptophan at position 568 or alanine for proline at position 191 resulted in herbicide resistance. While these substitutions had been shown to confer resistance individually in yeast ALS, they had previously been tested only in combination in a plant gene isolated from the tobacco Hra mutant. Finally, when alanine 199 was substituted with aspartic acid, resistant calli formed, revealing a site that had been previously identified in yeast ALS, but not plant ALS, to confer resistance (Hartnett et al. manuscript in preparation.). Tests of mutations coding for substitutions at two other sites, which were shown to produce herbicide-resistant yeast enzymes, were negative, but inconclusive (not shown). These results confirmed the utility of the microbial systems for predicting and evaluating herbicideresistance mutations and support the postulate that mutations resulting in herbicide resistance are conserved. ENGINEERING HERBICIDE-RESISTANT CROPS The mutant genes isolatedfromthe Hra and C3 lines of tobacco were reintroduced into tobacco and tested for the ability to confer herbicide resistance. These experiments indicated that a mutant gene from either the SuRA or SuRB locus can confer herbicide resistance in transgenic tobacco plants (20, 24). Since the introduction of the mutant gene from the Hra line resulted in higher levels of resistance, this gene was transferred to commercial lines of tobacco, and regenerated plants were assayed for levels of sulfonylurea resistance. Several different methods were used to determine the levels of resistance in the transformed plants, and to assess the feasibility of their use in thefield.Resistance was measured by assaying leaf ALS activity, callus growth, and seed germination and growth in the presence of herbicide, and by monitoring plant phytotoxicity after foliar spray applications of the herbicide. The results of these tests were consistent, yet indicated the need to monitor resistance by several methods in order to identify those lines most suitable for crop breeding.

In Managing Resistance to Agrochemicals; Green, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

31.

HARTNETTETAL.

Herbicide-Resistant Plants Carrying ALS Genes

TABLE I . H e r b i c i d e R e s i s t a n t Y e a s t ALS

Downloaded by UNIV OF NORTH CAROLINA on June 20, 2013 | http://pubs.acs.org Publication Date: February 23, 1990 | doi: 10.1021/bk-1990-0421.ch031

W i l d Type Amino A c i d Residue

465

Mutants

Amino A c i d Substitutions Resulting i n Resistance

116G

>

S N

117A

>

P S T I L V N Q D E K R H W F Y M

192P

>

A S V Q E R W Y

200A

>

T V D E R W Y C

251K

>

P T N D E

354M

>

V K C

379D

>

G P S V N E W

583V

>

A N Y C

586W

>

G A S I L V N E K R H Y C

590F

>

G L N R C

TABLE I I . H e r b i c i d e R e s i s t a n t C a l l u s F o l l o w i n g C o - U p t a k e o f K a n a m y c i n R e s i s t a n c e a n d S u l f o n y l u r e a R e s i s t a n c e Genes

ALS Gene

r

I

0,.3

none wild

Protoplast Transformants %Chlorsulfuron /Kanamycin

typel

0,.0

prol91—>ala

11,.8

prol91—>ser

39,.0

trp568—>leu

5,.5

alal99—>asp

29,.5

SuRB-Hra

29,.3

prol91—>ala trp568—>leu

50,.0

prol91—>ser trp568—>leu

27,.2

1

T h e w i l d t y p e a n d mutant g e n e s ( e x c e p t f o r SuRB-Hra) were c h i m e r i c t o b a c c o ALS g e n e s a s d e s c r i b e d i n t h e t e x t a n d shown i n F i g u r e 3.

In Managing Resistance to Agrochemicals; Green, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

466

MANAGING

< CJ CJ CJ