Pest Management Science Pest Manag Sci 56:209±217 (2000) psb Poa A Mutation (valine219 to isoleucine) in annua resistant to metribuzin and diuron Lemma W Mengistu,1† George W Mueller-Warrant,2* Aaron Liston3 and Reed E Barker2 1Crop and Soil Science Department, Oregon State University, Corvallis, OR 97330 USA 2USDA-ARS, National Forage Seed Production Research Center, 3450 SW Campus Way, Corvallis, OR 97331-7102 USA 3Botany and Plant Pathology Department, Oregon State University, Corvallis, OR 97330 USA

Abstract: The -binding region of the chloroplast psbA gene from a total of 20 biotypes of Poa annua L resistant and susceptible to metribuzin and diuron was selectively ampli®ed using PCR. Sequence analysis of the fragment from six herbicide-resistant biotypes of P annua exhibited a substitution from valine to isoleucine at position 219 of the D1 protein encoded by the psbA gene. This is the same mutation as reported for Chlamydomonas and Synechococcus through site-directed mutagenesis and in cell cultures of Chenopodium rubrum L. To our knowledge this is the ®rst report of a higher plant exhibiting resistance in the ®eld to photosystem II inhibitors due to a psbA mutation other than at position 264. The existence of additional biotypes of P annua resistant to diuron or metribuzin but lacking mutation in the herbicide-binding region indicates that resistance to these can also be attained by other mechanisms. # 2000 Society of Chemical Industry

Keywords: herbicide resistance; grass-seed production; ®eld selection; diuron; metribuzin

1 INTRODUCTION of some weed species can rapidly metabolize A range of herbicides, including s-triazines, substituted via hydroxylation reactions in the plant's metab- , and phenolic derivatives, block the photosyn- olism.12 Triazine-resistant weeds, however, degrade thetic electron transport chain on the reducing side of atrazine at a much lower rate than maize (Zea mays photosystem II (PS II).1,2 PS II herbicides act by L).13 displacing the secondary plastoquinone QB from its Triazine resistance is generally target-site based, binding site in PS II.3 The protein target of the PS II resulting from a modi®cation at the herbicide binding herbicides has been identi®ed as the D1 polypeptide site, the D1 protein of PS II.14 The substituted subunit of PS II, also called the herbicide or QB herbicides, although chemically distinct from triazines, binding protein.4 The D1 polypeptide is encoded by are also potent PS II inhibitors14 and bind at the chloroplast psbA gene,5 and is highly conserved in overlapping but not identical sites with the triazines.15 plants, algae and cyanobacteria.6 In higher plants the Biotypes highly resistant to triazine herbicides as a chloroplast psbA gene is located in the large single copy result of a modi®ed D1 protein fail to show resistance region close to the left junction of the inverted repeat.7 to substituted urea herbicides14 because mutation

The ®rst report of herbicide resistance to a photo- Ser264 to Gly providing resistance to triazine herbicides synthesis inhibitor involved a triazine herbicide,8 and does not affect binding of substituted urea herbi- since then biotypes of more than 60 species have been cides.2,15 Triazine-resistant plants, however, have documented to possess triazine resistance.9 The 1995/ been reported to show resistance to other PS II 96 International Survey of Herbicide Resistant Weeds inhibitors such as triazinones, uracils and pyridazi- revealed 183 herbicide-resistant weed biotypes (124 nones.16,17 different species) in 42 countries, and 32% of these The psbA gene from several organisms has been were triazine-resistant.10 Resistance to triazine herbi- sequenced, and, quite frequently, resistance by photo- cides has been attributed either to a rapid metabolic synthetic organisms to PS II herbicides, like triazi- detoxi®cation of the herbicide11 or to a modi®ed site of nones, triazines, and DCMU (diuron), results from action that prevents herbicide binding within the mutations in the psbA gene, leading to an amino acid 12 chloroplast. Susceptible as well as resistant biotypes exchange in the QB binding protein. In higher plants a

* Correspondence to: George W Mueller-Warrant, USDA-ARS, National Forage Seed Production Research Center, 3450 SW Campus Way, Corvallis, OR 97331-7102 USA E-mail: [email protected] † Current address: Department of Plant Sciences, North Dakota State University, Fargo, ND 58105 USA Contract/grant sponsor: Oregon Seed Council (Received 21 October 1999; accepted 1 November 1999)

# 2000 Society of Chemical Industry. Pest Manag Sci 1526±498X/2000/$17.50 209 LW Mengistu et al mutation affecting amino acid 264 (serine) leads to between the primary QA and secondary QB quinones resistance against triazine herbicides, and in algae, of PS II reaction centers,33±35 leading to an increased mutations in any of positions 215, 219, 255, 256 or susceptibility to photoinhibition in the resistant 275 are known to confer resistance to several classes of biotypes.36 The triazine-resistant mutants display a herbicides.18 marked decrease in growth rate at normal light Double and triple mutants of Chenopodium rubrum L intensities37 and a reduction in competitive ®tness.38 cell cultures were characterized where a valine to However, mutations of the D1 at positions 219 and isoleucine change at position 219 was decisive for 275 have been demonstrated to have no role in metribuzin resistance, although there were some other electron transfer system of PS II.6 A study conducted changes between amino acid residues 209 and 291. on the molecular basis of resistance to atrazine and For example, metribuzin-resistant cell cultures pos- cross-resistance to diuron used uniparental mutants of sessed psbA mutations in most cases at 219 and 251, or Chlamydomonas reinhardi Dang resistant to diuron at 219, 251 and 256, or at 219, 229 and 270, or at 219, (strain Dr2) and to atrazine (strain Ar207) that

220 and 270, or at 219, 251 and 272. All these changes possessed Val219 to Ile and Phe255 to Tyr, respec- resulted in resistance to metribuzin. None of the tively.6 Strain Dr2 showed a two-fold resistance to mutant cell cultures possessed the classical ®eld atrazine and a 15-fold resistance to diuron, while strain mutation at residue 264.18 Most of the mutations in Ar207 had a 15-fold resistance to atrazine and a 0.5- the D1 protein leading to herbicide resistance in fold resistance (two-fold increased susceptibility) to cyanobacteria and green algae were identi®ed by diuron. Electron transport for these two mutants was randomly induced mutagenesis or by site-directed normal, as opposed to Ser264 to Ala or Ser264 to Gly gene technology methods for the psbA gene.19 mutations that alter the electron transport system.6 An A detailed model of the herbicide-binding region in extended study of the mutant strain Dr-18 of C PS II has been developed.20 Based on this model, 17 reinhardi 39 revealed that the mutant required 10-fold amino acids of the psbA gene are in contact with the QB higher concentrations of DCMU and CMU to show binding site. These include: Phe211,Met214, His215, the same degree of inhibition as the wild type. This Leu218, Val219, Thr237, Ile248, Ala251, His252, Phe255, mutant was later sequenced, and determined to 6 Gly256, Ala263, Ser264, Phe265, Asn266, Ser268 and consist of a Val219 to Ile exchange in the psbA gene. Leu275. The eight site mutations including Val219 that The resistance to PS II inhibitors due to alteration of have been reported to cause resistance to herbicides all the psbA gene is maternally inherited. However, involve residues that have contact with QB. The hybrids with resistance from paternal plastids were relative importance of each residue in herbicide found at a frequency of 0.2 to 2%.40 binding has been further characterized.20 For exam- Poa annua is one of the worst weed problems ple, mutation Ser264 had greatest impact on binding of confronting grass-seed growers in the Willamette most PS-II-inhibiting herbicides. Val219 and Leu275 Valley of western Oregon. Its presence in harvested were characterized as being located peripherally to the seeds complicates seed cleaning and reduces crop 41 QB binding site, as compared to the other amino acids value if seed cannot be cleaned to `Poa-free' status. found nearer the center. Mutations to the more bulky Over 20 populations of P annua in grass ®elds of amino acids Ile219 or Phe275 will only touch those Oregon have been shown to be ethofumesate-resistant, inhibitors extending into this peripheral space. As a triazine-resistant and urea herbicide-resistant.10 Sev- result, the Val219 to Ile mutant showed R/S I50 ratios of eral of the P annua populations have shown multiple 20, 2, 2, and 3 for metribuzin, atrazine, s-triazine and resistance to atrazine, diuron, terbacil and ethofume- s-thiazole, respectively.20 sate,41±43 with greenhouse experiments indicating a High levels of resistance to atrazine have been 10-fold increase in resistance to both atrazine and correlated with a single amino acid substitution (Ser to diuron.42 Managing herbicide-resistant weeds requires Gly) in higher plants at position 264 in the D1 a good understanding of the genetics of resistance. protein.21±23 A change of Ser to Ala at the same Understanding the mechanisms of resistance involved position in this protein is correlated with lower levels of in these populations of P annua for various herbicides resistance to atrazine and diuron in algal mutants24 is vital. Wise management decisions will require full and a cyanobacterial transformant.25 In higher plants understanding of the mechanisms of resistance, the spontaneous herbicide resistance has been found in gene or genes involved, inheritance of resistance and the ®eld in various weed plants.21,26,27 In all reported gene ¯ow, as well as the details of the biology of the cases involving the psbA gene, a Ser264 to Gly change is species. This study therefore was designed to ful®l the responsible for the acquired resistance.28 This muta- objective of understanding the mechanism of resis- tion is reported for dicotyledenous species such as tance involved in populations of P annua resistant to Amaranthus hybridus L,21,29Solanum nigrum L,23 Che- diuron and metribuzin in western Oregon grass-seed nopodium album L,30,31 and Senecio vulgaris L.28 The crops. only monocotyledenous species with a psbA gene mutation at residue 264 giving resistance to atrazine 32 is Poa annua L. This mutation in higher plants is 2 MATERIALS AND METHODS accompanied by a slowing down of electron transfer Seeds for these experiments were obtained from

210 Pest Manag Sci 56:209±217 (2000) psbA mutation in Poa annua resistant to metribuzin and diuron greenhouse-grown P annua plants collected in 1994 a rapid one-step extraction of DNA (ROSE) buffer and 1995 from grass-seed ®elds of western Oregon for [Tris HCl (pH 8; 10mM), EDTA (pH 8; 312mM), studies of genetic diversity.44 Seeds of several addi- lauryl sarkocyl (10g litre1), PVPP (1% per 50ml ®nal tional herbicide-resistant types and one susceptible volume) and b-mercaptoethanol (2g litre1)] as ex- type were obtained from Dr Carol Mallory-Smith plained by Steiner et al. 45 The following steps were (Dept Crop & Soil Sci, Oregon State Univ, Corvallis, added to clean DNA samples further and to quantify OR). the DNA yield. The supernatant was placed in clean tubes, two-thirds volume (330ml) cold isopropanol was added, each sample was then well mixed and held 2.1 Greenhouse experiments In the greenhouse experiments, we tested the tolerance at 20°C for 10min. The samples were then 1 to ®eld rates of diuron and metribuzin using 225 centrifuged for 10min at 12000rev min and pellets accessions of P annua. Seeds from single plants of the were carefully saved in the tube while the supernatant 225 accessions were planted in rows using ¯ats in the was decanted. The pellets were washed with etha- greenhouse. Every ¯at consisted of ten rows, and each nol‡water (80‡20 by volume, 300ml) by spinning for row was a single accession of P annua. Eighty of 225 5min, then air-dried. The dried pellets were resus- accessions were tested for their reactions both to pended in Tris base‡EDTA pH 8(TE; 400ml). 1 metribuzin and diuron, while the other 145 accessions RNAse (0.5mgml ,8ml; ®nal concentration 1 were only tested for diuron resistance because quan- 10mgml ) was added to the samples and incubated tities of seed were limited. for 30min at 37°C. The DNA was precipitated by Every accession's seedlings were counted after rocking the tubes after adding ammonium acetate emergence and recorded before herbicide treatment. (7.5 M, 200ml) absolute ethanol (1.2 ml), and then On 30 May 1997, seedlings in the two-leaf stage were held for 10min at 20°C. Samples were ®nally spun treated with diuron at 1.1kg AI ha1 for all 225 for 10min, decanted, washed, and air-dried. The accessions and metribuzin at 0.37kg AI ha1 for 80 pellets were then resuspended in TE (80ml) and kept accessions. Two ¯ats from the metribuzin treatments in storage at 20°C until needed for ampli®cation. were divided into halves, and one half was covered DNA was quanti®ed for every sample using a with a plastic sheet while spraying the herbicide, to ¯uorometer, and the DNA samples were diluted to 1 serve as an untreated control. Other untreated controls ®nal concentrations of 10ngml . from earlier plantings were also grown in the green- house for comparisons. The treated seedlings were 2.3 PCR conditions and DNA amplification kept in the greenhouse at 12h supplemental light, Two primers were designed and synthesized based on 21°C day, 15°C night and watered daily. Seedlings homologous regions of sequences available from were counted again at two and three weeks after GenBank for the chloroplast psbA gene of Arabidopsis herbicide treatment, and the progeny classes and thaliana (GI:515373) and Nicotiana plumbaginifolia accession classes were determined based on their (GI:11754). These primers should amplify a 933 base reaction to the herbicides as (R) resistant, (I) inter- pair fragment of the chloroplast psbA gene including mediate and (S) susceptible. Progeny were scored as the herbicide-binding region. The primer sequence 5' resistant if uninjured by herbicide, intermediate if GGATGGTTTGGTGTTTTG 3' corresponded to injured but alive, and susceptible if killed. Accessions bases 91±108 of the psbA5'region. The second primer were scored as resistant if any progeny were resistant, 5' TAGAGGGAAGTTGTGAGC 3' corresponded to susceptible if all progeny were killed, and otherwise bases 1006±1023 of the 3' region of the DNA intermediate. After identifying the response of P annua sequence. Both primers were used to PCR-amplify accessions to diuron and metribuzin, the mechanism the target region from total DNA (Fig 1). The same of resistance was studied through DNA sequence two primers were also used for sequencing the analysis of the psbA gene. fragment in separate reactions, and results were aligned during analysis. Another primer was synthe- 2.2 DNA extraction sized based on the speci®c sequence we obtained from We extracted total DNA using a modi®ed procedure of our P annua, and this primer was only used for

Figure 1. A polyacrylamide gel electrophoresis indicating the amplified 933 base pair fragment DNA of the herbicide-binding region of the psbA gene from Poa annua accessions. The two primers used to amplify this DNA fragment were 5' GGATGGTTTGGTGTTTTG 3' and 5' TAGAGGGAAGTTGTGAGC 3'. The PCR reaction conditions are explained in the text. Numbers (1–7) are P annua samples and M is the marker.

Pest Manag Sci 56:209±217 (2000) 211 LW Mengistu et al sequencing the fragment. The sequence of the third populations were still segregating for resistance to primer was (5' CTCCTGTTGCAGCTGCTACT diuron. In general, these accessions were more 3'), which starts just before the region of amino acid sensitive to metribuzin than to diuron. However, three residue 150 and corresponds to bases 446±465 of the accessions from the Bowers farm near Harrisburg, psb A sequence. This primer was designed to give a Oregon, possessed individuals highly resistant to very clear sequence of the herbicide-binding region. metribuzin, and several accessions at two other sites PCR was performed in a 100ml reaction volume with were intermediate in their response to metribuzin. 90ml of reaction mix [1Â Stoffel buffer, 3.75mM Only one accession (#145) out of 80 tested was

MgCl2,50mM of each dNTP, 1.0pM of each primer, resistant to both diuron and metribuzin. Its reaction and 1 unit Taq Stoffel fragment] aliquotted into a 96- may have been maternally inherited because all the well assay plate and then overlaid with three drops of progeny from the single plant parent showed uniform mineral oil. Finally, 10mlof10ngml1 diluted DNA reaction to the herbicides. Since all 225 of the sample was added to the reaction mixture under the accessions were tested for diuron while only 80 were oil. The thermal cycling program started with dena- also tested for metribuzin, we presume some of the turation for 7min at 94°C followed by 43 cycles of 30- diuron-resistant accessions were also resistant to s denaturation at 47C, a 1°C per 3-s ramp, and metribuzin. The uniformity of their response to diuron extension at 72°C for 2min. A ®nal extension at 72°C across the rows of progenies was similar to that of for 5min concluded the DNA ampli®cation. After accession #145, indicating that these other accessions ampli®cation the samples were kept at 4°C for a short were similar to accession #145. period (<24h) or at 20°C for longer periods until electrophoresis. 3.2 Sequence analysis Initial attempts to sequence the 933 base pair fragment 2.4 Electrophoresis, DNA extraction from gels and of the psbA gene from our accessions used either one or purification the other of the two 18-mer primers from PCR The PCR product was run in both TBE (Tris ampli®cation as their sequencing template. The base‡boric acid‡EDTA) agarose gel electrophoresis DNA sequencer used (ABI Prism2 Model 377) gave and polyacrylamide gel electrophoresis. Polyacryla- no results in the template region. Nucleotide identi- mide gels were made from Tris-HCl (pH 8.8), ®cation became ambiguous somewhere between 300 acrylamide (7.5% ®nal), ammonium persulfate, and and 600 bases after the primer in any individual run, TEMED (N,N,N',N'-tetramethylethylenediamine), sometimes making it dif®cult to align the 5' and 3' and resulted in the approximately 933 base pair single sequences and be con®dent of the results in the fragment seen in Fig 1, along with several very faint herbicide-binding region (amino acid residues 211 to bands of lower molecular mass. DNA was extracted 275, bases 541 to 735 from the 5' end or bases 199 to from the 933 base pair fragment on the agarose gel, 393 from the 3' end of our fragment). In order to and puri®ed for sequencing by following the QIAquick obtain clearer results in this critical region, we Gel Extraction protocol (QIAGEN Inc, Valencia, generated a new (nested) primer for use as a California 1997) and Kit (buffers, tubes, ®lters). The sequencing template based on the sequence already puri®ed product for each sample was submitted to the obtained for bases 446 to 465 of the psbA gene (bases Central Services Laboratory at Oregon State Uni- 356 to 375 from the 5' end of our fragment). This third versity. Multiple sequences were obtained for most primer usually gave good readings all the way out to accessions. Analyses of the sequences were made ®rst the 3' end of the fragment (psbA codon 1023). by editing multiple sequences, and once a consensus The only detectable difference in the DNA se- sequence was obtained homology tests were made with quences of the 20 accessions occurred at codon 655, known psbA sequences. Translation was made using which changed from wild-type guanine to mutant Expasy, an Internet translation tool [http://www. adenine. This mutation was present in samples expasy.ch/tools/dna.html]. extracted from three separate plants grown out from seed produced by accession #145. The mutation was detectable in DNA sequence runs using each of the 3 RESULTS three primer templates for four separate accessions. 3.1 Poa annua accessions reaction to metribuzin The mutation was detected using the nested primer in and diuron in greenhouse two other accessions. Wild-type in the herbicide- A large number of P annua seedlings were susceptible binding region was con®rmed in nine accessions with to diuron and metribuzin, with treatments causing 46 the nested primer, two accessions with the 3' primer, to 94% and 75 to 96% lethality, respectively (Table 1). and three accessions with the 5' primer. For the 15 However, a large percentage of the accessions con- accessions sequenced with the nested primer, the high tained at least one individual (out of approximately 50 quality of the data strongly implies that there were no tested) highly resistant to diuron. Some individual psbA differences among these accessions from codon accessions were relatively uniform in their response to 469 to 1023, except for the mutation at 655. The seven diuron, while others possessed resistant, intermediate, accessions sequenced using the 5' PCR primer were and susceptible types, suggesting that many of these reliably sequenced (and showed no differences) from

212 Pest Manag Sci 56:209±217 (2000) psbA mutation in Poa annua resistant to metribuzin and diuron

Table 1. Progency testing herbicide tolerance classification Progency tolerance Accession tolerance class a class b RISR I S Accession source and herbicide treatment Accessions tested (number) (% of progeny) (% of accessions)

Diuron, 1.1kg ha1 Bowers farm perennial ryegrass 76 1.4 4.9 93.7 19.1 6.4 74.5 McLagan farm, Pugh Rd 33 6.4 35.6 57.9 66.7 15.2 18.2 McLagan farm, Bell Plain Rd 28 28.7 25.0 46.2 82.1 10.7 7.1 Glaser farm perennial ryegrass 50 16.6 33.9 49.5 66.0 22.6 11.3 Glaser farm meadowfoam 33 6.8 28.2 65.0 51.5 12.1 36.4 Pugh farm perennial ryegrass 1 5.1 41.0 53.8 100.0 0.0 0.0 Hyslop farm perennial ryegrass 4 1.9 14.7 83.3 25.0 25.0 50.0 Metribuzin, 0.37kg ha1 Bowers farm perennial ryegrass 28 4.4 2.6 93.0 10.7 3.6 85.7 McLagan farm, Pugh Rd 16 3.7 5.0 91.3 0.0 0.0 100.0 McLagan farm, Bell Plain Rd 10 15.8 8.7 75.5 0.0 30.0 70.0 Glaser farm perennial ryegrass 5 0.9 3.1 96.0 0.0 40.0 60.0 Glaser farm meadowfoam 21 0.1 5.4 94.5 0.0 4.8 95.2 a Progeny were scored as resistant if uninjured by herbicide, intermediate if injured but alive, and susceptible if killed. b Accessions were scored as resistant if any progeny were resistant, susceptible if all progeny were killed, and otherwise intermediate. Data shown are from ®rst rating for progeny class (two weeks after treatment) and second rating for accession class (three weeks after treatment).

codon 136 to 548. From codon 549 to 670, these sis thaliana (L) Heynh, 30 with maize, 26 with rice sequencing runs included up to 15 unidenti®ed bases, (Oryza sativa L), and 13 with barley (Hordeum vulgare although there were no identi®ed differences except at L). The deduced amino acid sequence for our 655. fragment was identical to those from corresponding Six accessions (including accession #145) out of the regions of the psbA genes of barley and maize except

20 that were sequenced exhibited a Val219 to Ile for the Val219 to Ile change in the mutant. Amino acid substitution in the psbA gene (Table 2). Several sequences for Arabidopsis, rice, and tobacco differed accessions were resistant to diuron or metribuzin but from those for wild-type Poa annua only at residue did not possess the psbA mutation. Accessions 238, coding for arginine instead of lysine. The 13 silent susceptible to diuron and metribuzin also did not mutations between Poa annua and barley occurred at exhibit this mutation. None of the six mutants showed nucleotides 153, 192, 198, 258, 409, 441, 504, 528, the classical mutation of Ser264 to Gly, but rather 624, 633, 732, 846 and 975. Homology at the nucleic remained wild-type at that position and at all other acid level is important for primer binding and sites within the herbicide-binding region except for successful PCR ampli®cation. While our second (3' residue 219 (Fig 2). GenBank accession numbers are PCR) primer exactly matched the published nucleo- AF131886 (GI:4972677) for the herbicide-resistant tide sequences of tobacco, maize, rice, barley, and mutant possessing the Val219 to Ile change, and Arabidopsis, the ninth position of our ®rst (5' PCR) AF131887 (GI:4972678) for the wild-type. The primer matched only the two dicotyledons, differing 888-base-pair consensus nucleotide sequence we from the three grasses (thymine in the primer vs obtained for wild-type Poa annua was compared with cytosine in the grasses). Because our 3' primer did not published psbA sequences for two dicotyledons and give usable readings at a distance of 925 bases, we do three grasses. There were 76 base pair differences with not know whether codon 99 of Poa annua psbA gene is tobacco (Nicotiana plumbaginifolia), 72 with Arabidop- thymine or cytosine. Based on our fragment's overall

Number of Number with psbA

accessions mutation (valine219 Number with Herbicide reaction of progeny a sequenced to isoleucine) wild-type psbA Diuron S and metribuzin S 3 0 3 Diuron S and metribuzin R 2 0 2 Diuron R and metribuzin unknown 14 5 9 Diuron R and metribuzin R 1 1 0 Table 2. Herbicide reaction and psbA Diuron R and metribuzin S 0 ± ± mutation status of 20 Poa annua a biotypes R=some or all progeny resistant, S=all progeny susceptible.

Pest Manag Sci 56:209±217 (2000) 213 LW Mengistu et al

Figure 2. Consensus nucleotide sequence for psbA gene fragments of 14 wild-type Poa annua covering the herbicide-binding region of the D1 protein (middle lines), nucleotide sequence for the barley psbA gene where it differs from Poa annua (upper lines), and deduced amino acids (lower lines). The metribuzin- and diuron-resistant mutant psbA sequence possessed a change of Val219 to Ile at the amino acid residue and nucleotide position (in the box) as a result of a change from GTA to ATA. Numbers at the right refer to the residue number of amino acids from the entire psbA gene. The underlined bases correspond to the third primer (5' CTCCTGTTGCAGCTGCTACT 3') we used for sequencing close to the herbicide binding region. We present here only the region that resulted in consistent sequences for all accessions, covering from amino acid residues 46 to 341 of the entire chloroplast psbA gene. Single letter amino acid code: A: alanine, C: cysteine, D: aspartate, E: glutamate, F: phenylalanine, G: glycine, H: histidine, I: isoleucine, K: lysine, L: leucine, M: methionine, N: asparagine, P: proline, Q: glutamine, R: arginine, S: serine, T: threonine, V: valine, Y: tyrosine, W: tryptophan.

closer homology to the grasses, however, and particu- ryegrass. These sites were well known for herbicide- larly to barley, it seems probable that codon 99 in the resistant P annua. Using randomly ampli®ed poly- Poa annua psbA gene is cytosine, and that we had a morphic DNA (RAPDs), we found that herbicide mismatch at the ninth position of our 5' PCR primer. selection pressure at these ®elds has reduced the 44 The six accessions that showed the Val219 to Ile among population genetic variation in P annua. The mutation came from two ®eld sites located 10 km six accessions possessing the Val219 to Ile mutation apart: two accessions from Glaser perennial ryegrass showed ®ve different RAPD banding patterns, with and four from McLagan Bell Plain Rd perennial the two accessions from Glaser perennial ryegrass

214 Pest Manag Sci 56:209±217 (2000) psbA mutation in Poa annua resistant to metribuzin and diuron having the same banding pattern.44 Interestingly, one Oregon grass-seed crops will complicate the problem of the RAPD banding patterns was shared by a psbA of addressing herbicide resistance. Gressel et al 46 mutant at McLagan Bell Plain Rd perennial ryegrass emphasized that `the tendency to cut dose rate is and psbA wild-types at Glaser perennial ryegrass and increasing resistance due to multiple cumulative McLagan Pugh Rd perennial ryegrass. RAPD banding events (polygenic, ampli®cation, or sequential muta- patterns of the six psbA mutants occurred in an tions within a gene).' The dilemma here is whether average of 14 individuals over the entire collection, crop rotation or rate of herbicide use should target the ®ve time more frequently than those of the average major gene or minor genes with cumulative effects, accession.44 since both types exist in the same populations. The mutation of the herbicide-binding D1 protein at residue 264 has been characterized to cause 4 DISCUSSION inhibition of the electron transport system, leading Failure to ®nd psbA mutants of higher plants in the the mutant to be less ®t than susceptible biotypes,

®eld other than Ser264 to Gly has long puzzled especially without herbicide use. However, the Val219 researchers. Our sequence analysis on herbicide- to Ile mutation we report now has been characterized resistant P annua accessions revealed that there are as causing resistance to herbicides but having no effect biotypes of higher plants with site mutations other on electron transfer. Because the mutant may be than residue 264, and these biotypes resist PS II equally ®t as the wild-type, use of no herbicide or low inhibitors at ®eld rates. In addition to Poa annua's selection pressure of herbicide may fail to reduce the short life cycle and signi®cant amount of outcrossing, frequency of this mutant in the populations. Another herbicide rates in western Oregon grass-seed crops aspect of the mutation of Val219 to Ile is its capability of may have been nearly ideal for selection of this psbA resisting more than one kind of herbicide, as it is mutation. Grass-seed growers have applied 1 to reported to result in resistance to metribuzin, atrazine, 1.5kgha1 atrazine, 2.2kgha1 and 1.1 to diuron and ioxynil in various organisms. Unlike the 1 2.2kg ha diuron for approximately 30 years. These Ser264 to Gly change in other weed species that exhibit rates are limited by the tolerance of crops such as negative cross-resistance to diuron, the absence of this perennial ryegrass to these herbicides. We do not yet in Val219 to Ile mutants is a problem to growers, as it know whether the Val219 to Ile mutant accessions also eliminates several alternative herbicides. possessed additional resistance from the nuclear- The degree of resistance by site mutation of Val219 controlled mechanism of resistance apparently present to Ile has been reported to range from 10- to 20-fold in other accessions. Selection in the ®eld may have which is close to the level of resistance found via played a role in combining multiple resistant mechan- detoxi®cation of herbicides by enzyme action through isms into single individuals. polygenic effects. This would be an advantage if high The number of mutant types within the accessions doses of herbicides could be used to delay herbicide we sequenced is far higher than expected by mutation resistance due to minor genes with cumulative effects rates alone of about 10 5 per generation. For example, and mutants of the kind we found. However, grass- out of 20 P annua accessions we sequenced for psbA, seed growers of western Oregon usually grow per- three were susceptible to metribuzin and diuron, 14 ennial ryegrass or tall fescue that would be susceptible were resistant to diuron with reaction to metribuzin to herbicides themselves should rates be increased untested, two were resistant to metribuzin but substantially above those currently used. One option susceptible to diuron, and one was resistant to both may be to rotate to crops like orchardgrass that better metribuzin and diuron. Of the 14 diuron-resistant, tolerate those herbicides or tolerate other herbicides metribuzin-untested biotypes, ®ve exhibited a Val219 such as pronamide, and use higher doses of herbicides to Ile substitution in their psbA gene. Sequences from to eliminate resistant weeds from their ®elds. Alter- the remaining nine diuron-resistant biotypes and the native crops like meadowfoam (Limnanthes species) or two metribuzin-resistant biotypes were wild-type in clovers (Trifolium species) are another option, allowing their psbA gene. The metribuzin-and diuron-resistant use of herbicides with different modes of action. The biotype possessed the mutation at the position we have alternative is that continued increases in populations indicated. It is likely that more psbA mutants would of P annua biotypes with multiple herbicide resistance have been found had we sequenced all diuron-resistant in western Oregon grass-seed crops may affect seed accessions within our population. We suspect that the producers and companies, leading to shifts in produc- rate and pattern of herbicide use by western Oregon tion, increased seed cleaning costs, and decreased grass-seed growers must have provided suitable selec- pro®tability. tion pressure to select these mutants. Higher rates We believe that more populations of P annua must might have killed the mutants, while lower rates might be studied to determine the level and type of resistance have allowed the susceptible types to survive. in the speci®c ®elds. Monitoring is needed to avoid The existence of both maternally inherited, single- contamination of non-infested ®elds. The procedure gene, site-mutation herbicide resistance and appar- we presented here could be further modi®ed and ently nuclear controlled resistance for the same simpli®ed for the ef®cient screening of a large number herbicides within populations of P annua in western of samples for speci®c mutations, such as Val219 to Ile.

Pest Manag Sci 56:209±217 (2000) 215 LW Mengistu et al

ACKNOWLEDGEMENTS and Naber D, Metribuzin resistance in photoautotrophic The authors thank the Oregon Seed Council for partial Chenopodium rubrum cell cultures. Characterization of double funding of this project. Contribution of the Agricul- and triple mutations in the psbA gene. FEBS Lett 329:43±46 (1993). tural Research Service, USDA, in cooperation with the 19 Kless H, Shamir MO, Malkin S, McIntosh L and Edelman M, Agricultural Experimental Station, Oregon State Uni- The D-E region of the D1 protein is involved in multiple versity Tech. Paper No. 11504 of the latter. quinone and herbicide interactions in photosystem II. Bio- chemistry 33:10501±10507 (1994). 20 Tietjen KG, Kluth JF, Andree R, Haug M, Lindig M, MuÈller KH, Wroblowsky HJ and Trebst A, The herbicide binding REFERENCES niche of photosystem II ± A model. Pestic Sci 31:65±72 (1991). 1 Trebst A, Inhibitors in electron ¯ow: Tools for the functional and 21 Hirschberg J and McIntosh L, Molecular basis of herbicide structural localization of carriers and energy conservation sites. resistance in Amaranthus hybridus. Science (Washington). Methods Enzymol 69:675±715 (1980). 222:1346±1349 (1983). 2 Arntzen CJ, P®ster K and Stienback KE, The mechanism of 22 Hirschberg J, Bleecker A, Kyle DJ and McIntosh L, The chloroplast triazine resistance: Alterations in the site herbicide molecular basis of triazine herbicide resistance in higher plant action, in Herbicide Resistance in Plants, ed by LeBaron HM and chloroplasts. Z Naturforsch 39:412±420 (1984). Gressel J, John Wiley & Sons, New York. pp 185±214 (1982). 23 GolouÈbinoff P, Edelman M and Hallick RB, Chloroplast coded 3 Velthuys BR, Electron-dependent competition between plasto- atrazine-resistance in Solanum nigrum: psbA loci from suscep- quinone and inhibitors for binding to photosystem II. FEBS tible and resistant biotypes are isogenic except for a single Lett 126:277±281 (1981). codon change. Nucleic Acids Res 12:9489±9496 (1984). 4 P®ster K, Steinback KE, Gardner G and Arntzen CJ, Photo- 24 Erickson JM, Rahire M, Bennoun MP, Delepelaire P, Dines B af®nity labelling of a herbicide receptor protein in chloroplast and Rochaix JD, Herbicide resistance in C reinhardii results membranes. Proc Natl Acad Sci USA 78:981±985 (1981). from a mutation in the chloroplast gene for the 32 kiloDalton 5 Zurawski G, Bohnet H, Whitfeld P and Bottomley W, Nucleotide protein in photosystem II. Proc Natl Acad Sci USA 81:3617± sequence of the gene for the Mr 32000 thylakoid membrane 3621 (1984). protein from Spinacia oleracea and Nicotinia debneyi predicts a 25 Golden SS and Heselkorn R, Mutation to herbicide resistance totally conserved primary translocation protein of Mr 38 950. maps within the psbA gene of Anacystis nidulans R2. Science Proc Natl Acad Sci USA 79:7699±7703 (1982). (Washington) 229:1104±1107 (1985). 6 Erickson JM, Rahire M and Rochaix JD, Herbicide resistance 26 Van Oorschot JLP, Chloroplastic resistance of weeds to triazines and cross-resistance: Changes at three distinct sites in the in Europe, in Herbicide Resistance in Weeds and Crops,edby herbicide binding protein. Science (Washington) 228:204±207 Caseley JC, Cussans GW and Atkin RK, Butterworth- (1985). Heinemann, Oxford. pp 87±101 (1991). 7 Driesel AJ, Speirs J and Bohnert HJ, Spinach chloroplast mRNA 27 Rubin B, Herbicide resistance in weeds and crops, progress and for a 32000 Dalton polypeptide. Size and localization on the prospects, in Herbicide Resistance in Weeds and Crops, Vol II, ed physical map of the chloroplast DNA. Biochem Biophys Acta by Caseley JC, Cussans GW and Atkin RK, Butterworth- 610:287±310 (1980). Heinemann, Oxford. pp 387±414 (1991). 8 Ryan GF, Resistance of common groundsel to simazine and 28 Trebst A, The molecular basis of plant resistance to photosystem atrazine. Weed Sci 18:614±616 (1970). II herbicides, in Molecular Genetics and Evolution of Pesticide 9 LeBaron HM, Distribution and seriousness of herbicide-resistant Resistance, ed by Brown TM, ACS symposium series 645, weed infestations worldwide, in Herbicide Resistance in Weeds Washington, DC. pp 44±51 (1996). and Crops, ed by Caseley JC, Cussans GW and Atkin RK, Butterworth-Heinemann, Oxford. pp 27±43 (1991). 29 McNally S, Bettini P, Sevignac M, Darmency H, Gasquez J and 10 Heap IM, The occurrence of herbicide-resistant weeds world- Dron M, A rapid method to test for chloroplast DNA wide. Pestic Sci 51:235±243 (1997). involvement in atrazine resistance. Plant Physiol 83:248±250 11 Shimabukuro RH, Atrazine metabolism in resistant corn and (1987). sorghum. Plant Physiol 43:1925±1930 (1968). 30 Gasquez J, Mouemar AAL and Darmency H, Triazine herbicide 12 Ali A and Machado VS, A comparative analysis of leaf resistance in Chenopodium album L. Occurrence and charac- chlorophyll ¯uorescence, Hill reaction activity and 14C- teristics of an intermediate biotype. Pestic Sci 16:390±395 atrazine tracer studies to explain differential triazine suscept- (1985). ibility in wild turnip rape (Brassica campestris) biotypes. Can J 31 Bettini P, Sevignac M, McNally S, Darmency H, Gasquez J and Plant Sci 644:707±713 (1984). Dron M, Molecular mechanisms involved in the appearance of 13 Gressel J, Herbicide tolerance and resistance: Alteration of site of atrazine resistance in Chenopodium album. J Cell Biochem 10:44 activity, in Weed Physiology, Vol II Herbicide Physiology,edby (1986). Duke SO, CRC Press, Boca Raton, Florida. pp 159±189 32 Barros MDC and Dyer TA, Atrazine resistance in grass Poa (1985). annua is due to a single base change in the chloroplast gene for 14 Gronwald JW, Resistance to photosystem II inhibiting herbi- the D1 protein of photosystem II. Theor Appl Genet 75:610± cides, in Herbicide Resistance in Plants: Biology and Biochemistry, 616 (1988). ed by Powles SB and Holtum JAM, Lewis Publishers, Boca 33 Arntzen CJ, Ditto CL and Brewer PE, Chloroplast membrane in Raton, Florida. pp 27±60 (1994). triazine-resistant Amaranthus retro¯exus biotypes. Proc Natl 15 Trebst A, The molecular basis of resistance to photosystem II Acad Sci USA 76:278±282 (1979). herbicides, in Herbicide Resistance in Weeds and Crops,edby 34 Bowes J, Crofts AR and Arntzen CJ, Redox reactions on the Caseley JC, Cussans GW and Atkin RK, Butterworth- reducing side of photosystem II in chloroplasts with altered Heinemann, Oxford. pp 145±164 (1991). herbicide-binding properties. Arch Biochem Biophys 200:303± 16 Oettmeier W, Masson K, Fedtke C, Konze J and Schmidt RR, 308 (1980). Effect of photosystem II inhibitors on chloroplasts from species 35 Ort OR, Ahrens WH, Martin B and Stoller EW, Comparison of either susceptible or resistant towards s-triazine herbicides. photosynthetic performance in triazine-resistant and suscep- Pestic Biochem Physiol 18:357±367 (1982). tible biotypes of Amaranthus hybridus. Plant Physiol 72:925±930 17 Fuerst EP, Arntzen CJ, P®ster K and Penner D, Herbicide cross- (1983). resistance in triazine-resistant biotypes of four species. Weed 36 Sundby C, Chow WS and Anderson JM, Effects on photosystem Sci 34:344±353 (1986). II function, photoinhibition, and plant performance of the 18 Schwenger-Erger C, Thiemann J, Barz W, Johanningmeier U spontaneous mutation of serine 264 in photosystem II reaction

216 Pest Manag Sci 56:209±217 (2000) psbA mutation in Poa annua resistant to metribuzin and diuron

center D1 protein in triazine-resistant Brassica napus L. Plant Cooperating, ed by Young III, WC, Department Crop and Soil Physiol 103:105±113 (1993). Science Ext/CrS 110. pp 29±32 (1997). 37 Hart JJ and Stemler A, Similar photosynthetic performance in 42 Heap IM, Multiple herbicide resistance in annual bluegrass (Poa low light-grown isonuclear triazine-resistant and susceptible annua). WSSA Abstracts 42:56 (1995). Brassica napus L. Plant Physiol 94:1295±1300 (1990). 43 Gamroth DM, Brewster BD and Mallory-Smith CA, Herbicide 38 Holt JS and Thill DC, Growth and productivity of resistant banding in grass-seed crops, in 1996 Seed Production Research at plants, in Herbicide Resistance in Plants: Biology and Biochem- Oregon State University, USDA-ARS Cooperating, ed by Young istry, ed by Powles SB and Holtum JAM, Lewis Publishers, III, WC, Department Crop and Soil Science Ext/CrS 110. p 33 Boca Raton, Florida. pp 293±316 (1994). (1997). 39 Lien S, McBride JC, McBride AC, Togasaki RK and Pietro AS, 44 Mengistu LW, Genetic diversity and herbicide resistance in A comparative study of photosystem II speci®c inhibitors: The annual bluegrass (Poa annua L). PhD Thesis, Oregon State differential action on a DCMU resistant mutant strain of C University, Corvallis, Oregon pp 1±122 (1999). Dissertation reinhardi, in Photosynthetic Organelles, Plant & Cell Physiol (Special Issue) 243±256 (1977). abstract no. AAT 9842465. 40 Darmency H, Genetics of herbicide resistance in weeds and 45 Steiner JJ, Poklemba CJ, Fjellstrom RG, and Elliott LF, A rapid crops, in Herbicide Resistance in Plants: Biology and Biochemistry, one-step DNA extraction technique for PCR and RAPD ed by Powles SB and Holtum JAM, Lewis Publishers, Boca analysis. Nucleic Acids Res 23:2569±2570 (1995). Raton, Florida. pp 263±297 (1994). 46 Gressel J, Gardner SN and Mangel M, Prevention versus 41 Mueller-Warrant GW, Treatment of diuron-resistant annual remediation in resistance management, in Molecular Genetics bluegrass in a perennial ryegrass seed crop, in 1996 Seed and Evolution of Pesticide Resistance, ed by Brown TM, ACS Production Research at Oregon State University, USDA-ARS symposium series 645, Washington, DC. pp 169±186 (1996).

Pest Manag Sci 56:209±217 (2000) 217