Introduction
Weed resistance to herbicides limits the number of effective herbicide options and has become a major obstacle to weed control worldwide. To date, weeds have evolved resistance to 23 of the 26 known active herbicide sites and 167 different herbicides (Heap Reference Heap2020). The most common and strongest type of weed resistance is resistance to acetolactate synthase (ALS) inhibitors (HRAC Group B), followed by resistance to photosystem II (HRAC Group C1), 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) (HRAC Group G), and acetyl-CoA carboxylase (ACCase) inhibitors (HRAC Group A) (Heap Reference Heap2020). Compared with weed resistance to the abovementioned types of herbicides, which are associated with a higher risk of resistance than other herbicides (Collavo et al. Reference Collavo, Panozzo, Lucchesi, Scarabel and Sattin2011; Tranel Reference Tranel2002), weed resistance to protoporphyrinogen oxidase (PPO) inhibitors seems to have evolved over a longer period of time. PPO-inhibiting herbicides have been used since the 1960s, but the first case of resistance to PPO inhibitors was reported in 2001 and involved waterhemp [Amaranthus tuberculatus (Moq.) Sauer] in soybean [Glycine max (L.) Merr.] fields (Dayan et al. Reference Dayan, Owens, Tranel, Preston and Duke2014; Shoup et al. Reference Shoup, Al-Khatib and Peterson2003). Since then, resistance to PPO inhibitors has evolved in wild poinsettia (Euphorbia heterophylla L.), common ragweed (Ambrosia artemisiifolia L.), Palmer amaranth (Amaranthus palmeri S. Watson), and redroot pigweed (Amaranthus retroflexus L.) (Rousonelos et al. Reference Rousonelos, Lee, Moreira, Vangessel and Tranel2012; Salas et al. Reference Salas, Burgos, Tranel, Singh, Glasgow, Scott and Nichols2016; Trezzi et al. Reference Trezzi, Felippi, Mattei, Silva, Nunes, Debastiani, Vidal and Marques2005; Wang et al. Reference Wang, Guo, Zhang, Zhao, Ge, Lv, Liu and Wang2017). The widespread cultivation of glyphosate-resistant soybean has reduced dependence on PPO-inhibiting herbicides, which has reduced the selection pressure for resistance to PPO inhibitors (Young Reference Young2006). However, in the past two decades, growers’ dependence on PPO-inhibiting herbicides has increased as weeds have developed resistance to glyphosate and ALS inhibitors (Chen et al. Reference Chen, Huang, Zhang, Huang, Wei, Chen and Wang2015; Legleiter et al. Reference Legleiter, Bradley and Massey2009; Salas et al. Reference Salas, Burgos, Tranel, Singh, Glasgow, Scott and Nichols2016). Consequently, resistance to PPO inhibitors has evolved over the past two decades.
Plant PPO catalyzes the conversion of protoporphyrinogen IX to protoporphyrin IX in the tetrapyrrole biosynthethic pathway, which produces heme and chlorophyll (Deybach et al. Reference Deybach, Silva, Grandchamp and Nordmann1985; Poulson and Polglase Reference Poulson and Polglase1975). There are two isoforms of PPO, plastid-targeted and mitochondrial-targeted PPO (namely, PPO1 and PPO2), which are encoded by the PPX1 and PPX2 genes, respectively (Lermontova et al. Reference Lermontova, Kruse, Mock and Grimm1997). In some plant species, PPO2 is dual-targeted in plastids and mitochondria (Watanabe et al. Reference Watanabe, Che, Iwano, Takayama, Yoshida and Isogai2001). PPO-inhibiting herbicides reduce the conversion of protoporphyrinogen IX to protoporphyrin IX, and the accumulated protoporphyrinogen IX is transported to the cytoplasm and spontaneously undergoes peroxidation to form protoporphyrin IX. Under light conditions, the accumulated protoporphyrin IX produces singlet oxygen, which can cause lipid membrane oxidation and plant injury (Jacobs and Jacobs Reference Jacobs and Jacobs1993; Jacobs et al. Reference Jacobs, Jacobs, Sherman and Duke1991; Lee and Duke Reference Lee and Duke1994). These two isoforms of PPO are the targets of PPO-inhibiting herbicides.
To date, a total of 13 species of weeds have developed resistance to PPO-inhibiting herbicides (Heap Reference Heap2020). In most of the reported cases of field-evolved resistance, resistance is caused by target-based mutations. Five amino acid mutations in PPX2 located at three sites have been reported to cause herbicide resistance to PPO-inhibiting herbicides in weeds. One mutation is a deletion of glycine at position 210 (Gly-210) in PPX2 (Lee et al. Reference Lee, Hager and Tranel2008; Patzoldt et al. Reference Patzoldt, Hager, McCormick and Tranel2006; Salas et al. Reference Salas, Burgos, Tranel, Singh, Glasgow, Scott and Nichols2016). This position serves an important stabilizing role in the capped region of the α-8 helix, and its absence unwinds the last turn of the helix, thus causing the cavity of the active site to expand by approximately 50%; this expansion reduces the binding of herbicides and PPO (Dayan et al. Reference Dayan, Daga, Duke, Lee, Tranel and Doerksen2010; Yang et al. Reference Yang, Hao, Dayan, Tranel and Yang2013). Another mutation is an amino acid substitution of arginine at position 128 (or 98), present as Arg-98-Leu in A. artemisiifolia, Arg-128-Gly and Arg-128-Met in A. palmeri, Arg-128-Gly in A. retroflexus, and Arg-128-Ile and Arg-128-Lys in A. tuberculatus (Giacomini et al. Reference Giacomini, Umphres, Nie, Mueller, Steckel, Young, Scott and Tranel2017; Huang et al. Reference Huang, Cui, Wang, Wu, Zhang, Huang and Wei2020; Nie et al. Reference Nie, Mansfield, Harre, Young, Steppig and Young2019; Varanasi et al. Reference Varanasi, Brabham, Norsworthy, Nie, Young, Houston, Barber and Scott2018b). The Arg-128 and Arg-98 positions are homologous, and the difference in the code number results because there is a 30-amino-acid signal peptide in A. palmeri but not in A. artemisiifolia. Position 128 is highly conserved among multiple plant species, and this site plays an important role in substrate positioning. Therefore, an amino acid substitution at this site is generally associated with herbicide resistance (Dayan et al. Reference Dayan, Barker and Tranel2018; Nie et al. Reference Nie, Mansfield, Harre, Young, Steppig and Young2019). Recently, a novel substitution of glycine with alanine at position 399 (Gly-399-Ala), which is located in the catalytic domain of PPO, was reported in A. palmeri (Rangani et al. Reference Rangani, Salas-Perez, Aponte, Knapp, Craig, Mietzner, Langaro, Noguera, Porri and Roma-Burgos2019). The steric hindrance caused by Gly-399-Ala results in a reduced affinity for PPO-inhibiting herbicides (Rangani et al. Reference Rangani, Salas-Perez, Aponte, Knapp, Craig, Mietzner, Langaro, Noguera, Porri and Roma-Burgos2019). In addition, although non–target site based resistance to this class of herbicides has been reported in only a few cases, involving species such as A. tuberculatus and A. palmeri, it cannot be ignored (Obenland et al. Reference Obenland, Ma, O’Brien, Lygin and Riechers2017; Varanasi et al. Reference Varanasi, Brabham and Norsworthy2018a, 2019).
Amaranthus retroflexus occurs in autumn crops, such as soybean, cotton (Gossypium hirsutum L.), peanuts (Arachis hypogaea L.), and corn (Zea mays L.), in fields in most areas of China. Additionally, it is one of the dominant weeds in soybean fields in Heilongjiang Province. In the past few years, with the evolution of resistance to ALS-inhibiting herbicides, PPO-inhibiting herbicides, especially fomesafen, have become the main herbicides for controlling this weed (Chen et al. Reference Chen, Huang, Zhang, Huang, Wei, Chen and Wang2015). Under the selective pressure of PPO-inhibiting herbicides, A. retroflexus has evolved resistance to fomesafen (Wang et al. Reference Wang, Guo, Zhang, Zhao, Ge, Lv, Liu and Wang2017). Several studies have indicated that the Arg-128-Gly mutation in PPO2 is one of the target site–based resistance mechanisms in A. retroflexus, but more comprehensive and detailed information has thus far not been discovered (Huang et al. Reference Huang, Cui, Wang, Wu, Zhang, Huang and Wei2020; Wang et al. Reference Wang, Wang, Zhao, Zhu, Sun, Liu and Wang2019). Therefore, the objectives of this study were to characterize the fomesafen sensitivity of susceptible and resistant A. retroflexus and to systematically illustrate the target site–based resistance mechanism, including variations in the PPO gene sequence, PPO enzyme activity, enzyme kinetics, and PPO expression.
Materials and Methods
Plant Material
Fifty-six suspected resistant A. retroflexus populations were collected from areas where fomesafen control failed. The sampling area included approximately 7,000 km2 in Heilongjiang Province, China. The central coordinates of the collection area are approximately 125.26°N and 49.18°E. Two susceptible populations (referred to as S1 and S2) were collected from noncultivated land areas (124.03°N, 47.36°E) with a geographic distance of approximately 200 km from where resistance evolved. Approximately 10 plants were collected from each population, and then the seeds were air-dried and stored in paper bags at room temperature until use.
Whole-Plant Bioassay
Seeds were germinated for 5 d in 9-cm-diameter petri dishes containing two layers of filter paper and 5 ml of deionized water at 25 C. Seedlings were transplanted to plastic pots (5 seedlings per pot) containing a commercial nutrient matrix and grown in phytotrons set at 25 C/16-h day, 18 C/8-h night, 65% relative humidity, and a light intensity of 12,000 lx. A single-dose bioassay with 250 g ai ha−1 fomesafen was conducted first to screen the resistant populations, with the S1 and S2 populations serving as susceptible controls. After that, two resistant populations (referred to as R1 and R2) approximately 50 km apart were selected to determine their sensitivity to fomesafen with S1 and S2 as controls. In addition, all plants from both resistant populations survived a dose of 250 g ai ha−1. The treatment doses of fomesafen were 0, 0.514, 1.54, 4.63, 13.89, 41.67, and 125 g ai ha−1 (0-, 1/729-, 1/243-, 1/81-, 1/27-, 1/9- and 1/3-fold the recommended dose, respectively) for the susceptible populations and 0, 4.63, 13.89, 41.67, 125, 375, and 1,125 g ai ha−1 (0-, 1/81-, 1/27-, 1/9-, 1/3-, 1-, and 3-fold the recommended dose, respectively) for the resistant populations. All tested populations were directly cultivated with seeds collected in the field. Seedlings were treated with fomesafen once they reached a 10-cm height using an automatic spraying system equipped with a TeeJet® 9503EVS flat-fan spray nozzle (Spraying Systems, Wheaton, IL, USA). The spray volume was 300 L ha−1 at 0.28 MPa. The aboveground biomass of each treatment was collected after 3 wk of treatment and dried at 70 C for 72 h, and then the dry weights were recorded. Each treatment consisted of four replicates, and the entire experiment was repeated twice.
PPO Gene Amplification and Sequencing
Total RNA was extracted from 10 individual plants within each resistant and susceptible population using an EasyPure Plant RNA kit (TransGen Biotech, Beijing, China). Then, cDNA was synthesized using Transcript a First Strand cDNA Synthesis Kit with DNA removal (TransGen Biotech). Two nearly full-length PPX1 and PPX2 genes were amplified with the primers PPX1F/PPX1R and PPX2F/PPX2R (Table 1). The primers for amplifying nearly full-length PPX1 were designed by Primer Premier 5.0 (PREMIER Biosoft, San Francisco, CA, USA) according to the PPX1 sequence of A. tuberculatus (GenBank accession no. DQ386112) and synthesized by a commercial sequencing company (Personal Biotech, Shanghai, China). The primers for amplifying PPX2 were synthesized according to a previous study (Giacomini et al. Reference Giacomini, Umphres, Nie, Mueller, Steckel, Young, Scott and Tranel2017). PCR was conducted in a 25-μl volume that consisted of 1 μl of cDNA (60 ng), 1 μl of each primer (1 μM), and 22 μl of Taq Master Mix (US Everbright, Suzhou, China). PCR for amplifying PPX1 and PPX2 was performed on a thermal cycler (Bio-Rad, Hercules, CA, USA) with the following profile: 94 C for 5 min; 33 cycles at 94 C for 30 s, 58 C for 30 s, and 72 C for 80 s; and a final extension at 72 C for 5 min. The PCR products were sequenced by Personal Biotech. Sequence alignment was conducted using DNAMAN 9.0 (Lynnon Biosoft, San Ramon, CA, USA), and amino acid numbering for PPX1 and PPX2 was performed based on that of A. tuberculatus (GenBank accession nos. DQ386112, DQ386114).
Table 1. Primers used in this study.
In Vitro PPO Assay
The PPO activity assay was conducted by measuring the formation rate of protoporphyrin IX at 30 C according to the fluorometric method (Ishida et al. Reference Ishida, Miller-Sulger, Kohno, Böger and Wakabayashi2000; Nicolaus et al. Reference Nicolaus, Sandmann and Böger1993). Seedling preparation was conducted as described earlier. The leaves (1 g) were harvested after exposure to light for 12 h and homogenized in 5 ml of an extraction buffer with a pH of 7.3 containing 0.05 M HEPES buffer, 0.5 M sucrose, 1 mM 1,4-dithiothreitol (DTT), 1 mM MgCl2, 1 mM ethylenediaminetetraacetic acid (EDTA), and 0.2% bovine serum albumin (BSA). After two steps of centrifugation (800 × g for 2 min and then 17,000 × g for 6 min), the precipitate was dissolved again in a lysosomal buffer with a pH of 7.3 containing 0.05 M Tris and 2 mM EDTA. PPO inhibition was determined after adding 100 μl of fomesafen (0 to 100 μM) to a final assay volume of 3 ml of reaction buffer containing 0.1 mM Tris, 1 mM EDTA, 5 mM DTT, 1% (v/v) Tween-80, 100 μl of crude PPO, and approximately 8 μM protoporphyrinogen IX. The reaction system was placed in a 30 C water bath and shaken in the dark for 30 min. The protoporphyrinogen IX was freshly obtained by reducing protoporphyrin in the dark with a sodium amalgam in a nitrogen atmosphere. The protein content of the crude extraction enzyme was measured using the Bradford method (Bradford Reference Bradford1976). The amount of protoporphyrin produced was measured with a Hitachi F-2700 fluorescence spectrophotometer (Hitachi, Tokyo, Japan) using an excitation wavelength of 405 nm and an emission wavelength of 633 nm (Rangani et al. Reference Rangani, Salas-Perez, Aponte, Knapp, Craig, Mietzner, Langaro, Noguera, Porri and Roma-Burgos2019). The PPO activity was calculated by the amount of protoporphyrin produced per unit of time (nmol Proto-IX mg protein−1 h−1). For the PPO enzyme kinetic assay, protoporphyrinogen IX was omitted from the reaction mixture, and serial protoporphyrinogen IX concentrations of 0.5, 1, 2, 3, 4, 5, 7, 9, and 10 μM were used in the final reaction system. There were three technical replicates for each sample and three biological replicates for each treatment. The entire experiment was repeated two times.
PPO Expression Assay
A PPO expression assay was conducted at the transcription and translation levels. Amaranthus retroflexus seedlings were prepared as described earlier and treated with fomesafen at a dose of 37.5 g ai ha−1 after growing to a height of 10 cm. The foliage of the resistant and susceptible populations was harvested at 0 (untreated control), 6, 12, 24, and 48 h after treatment (HAT). Total RNA extraction and cDNA synthesis were conducted as described earlier.
PPO gene expression, including PPX1 and PPX2, was determined by real-time quantitative reverse transcriptase PCR (qRT-PCR) in an AB 7500 Fast Real-Time PCR system (Applied Biosystems, Dun Laoghaire, County Dublin, Ireland) using TB Green Fast qPCR Mix (Takara, Dalian, China). The actin gene was selected as the internal reference gene (Huang et al. Reference Huang, Chen, Zhang, Huang, Wei, Zhou, Chen and Wang2016). The primers used in the qRT-PCR assay were PPX1-qF/PPX1-qR, PPX2-qF/PPX2-qR, and actin-qF/actin-qR (Table 1). The primers were designed by Primer Premier 5.0, and the primers used for amplifying actin were designed according to the actin sequence of silver cock’s comb (Celosia argentea L.) (GenBank accession no. HQ844002.1). To determine the specificity and amplification efficiency of primers, melting and standard curves were generated. A three-step profile was used in thermal cycling: 94 C for 5 s, 59 C for 30 s, and 72 C for 5 s. The Ct values of each sample were recorded, and the relative expression levels of PPX1 and PPX2 were calculated with the 2−∆∆Ct method (∆Ct = Ctppx − CtActin, ∆∆Ct = ∆Cttreated − ∆Ctuntreated). The actin gene served as the endogenous control, and the untreated sample served as the reference sample. Each sample was set up with three technical repeats and two biological repeats. The entire experiment was repeated two times.
The determination of the PPO enzyme content was carried out using an enzyme-linked immunosorbent assay (ELISA) kit (Jianglai Bio, Shanghai, China). The kit was only used to determine the PPO content without distinguishing between plastid PPO and mitochondrial PPO, not to measure PPO activity. Foliage material (0.5 g) from each harvested plant material sample was homogenized in 5 ml of extraction buffer at 0 C. Subsequent procedures were carried out following the instructions in the product manual. A standard curve was made according to the standard products provided by the ELISA kit, and then the PPO content was calculated according to the standard curve obtained. Each sample was set up with three technical repeats and three biological repeats. The entire experiment was repeated two times.
Statistical Analysis
The dose–response data (including plant growth and enzyme activity inhibition) were analyzed by ANOVA using SPSS v. 20.0 (IBM SPSS, Chicago, IL, USA). The data were pooled if there was no significant difference between two experimental trials. Means were analyzed using Fisher’s protected LSD test at the 5% level of probability. Dose–response data were fit to the four-parameter logistic equation (Equation 1) proposed by Seefeld et al. (Reference Seefeld, Jensen and Fuerst1995) using SigmaPlot v. 12.5 (Systat, San Jose, CA, USA) to calculate the GR50 (or I50).
$$y = C + {{D - C} \over {1 + {{\left( {{x \over {{\rm{G}}{{\rm{R}}_{50}}}}} \right)}^b}}}$$
where x represents the herbicide concentration, y represents % of control, C represents the lower limit, D represents the upper limit and GR50 (or I50) represents the herbicide dose causing 50% growth inhibition (or enzyme activity).
For enzyme kinetics, the Km values were calculated by fitting the data to the Michaelis-Menten equation (Equation 2) using SigmaPlot 12.5 (Systat).
$$v = {{{V_{max}}S} \over {{K_m} + S}}$$
where v represents the catalytic velocity, S represents the substrate concentration and V max represents the maximum catalytic velocity.
Results and Discussion
Whole-Plant Bioassay
The dose–response experiments at the whole-plant level indicated that the two resistant populations showed significantly different responses to fomesafen than did the two susceptible populations. Fomesafen (recommended dose 375 g ai ha−1) at a dose of 41.7 g ai ha−1 caused clear evidence of injury in the susceptible plants within 2 d after treatment (DAT), and these plants died within 7 DAT. Fomesafen did not kill the susceptible plants at doses below 13.9 g ai ha−1, but their growth was severely inhibited as the dose increased from 4.6 to 13.9 g ai ha−1, with chlorosis, stunting, and crinkling leaves providing evidence of injury. However, fomesafen at doses from 4.6 to 13.9 g ha−1 had little effect on the growth of resistant plants, which exhibited no evidence of injury such as chlorosis or stunting. All plants from both the R1 and R2 populations survived the recommended dose of 375 g ha−1 (Figure 1). Although the growth of the resistant plants was significantly inhibited, it should be noted that this result was obtained under phytotron conditions, which usually causes weeds to be more sensitive to herbicides than under natural conditions. The high resistance factors (27.0 and 27.9) of the R1 and R2 populations indicated a high level of resistance to fomesafen. At doses of 41.7 and 125 g ha−1, the resistant plant foliage was injured, but the growing point and damaged leaves recovered and resumed growth at 10 DAT. The average GR50 value for the two susceptible populations was 5.77 g ha−1 (5.35 for S1 and 6.18 for S2, respectively). The GR50 values for the two resistant populations (R1 and R2) were 143.6 and 155.7 g ha−1, respectively, and the calculated resistance factors were 27.9 and 27.0, respectively (Table 2).
Figure 1. Dose–response curve fit to a four-parameter nonlinear regression model for aboveground dry weights of the Amaranthus retroflexus populations treated with fomesafen. Mean values and their SEs are shown.
Table 2. Parameter estimates from the logistic analysis of growth inhibition by fomesafen for the susceptible and resistant Amaranthus retroflexus populations. a
a C, lower limit; D, upper limit; b, the slope at GR50; Rf, resistance factor.
b Rf values were calculated as the mean values of the two susceptible populations. Fitting parameters and their SEs are shown.
We investigated the resistance of A. retroflexus in an area of 7,000 km2 in Nenjiang County, Heilongjiang Province. The results showed that almost all populations developed the same resistance as observed in the R1 and R2 populations tested in this study (data not shown). In practice, some growers use a 5-fold dose of fomesafen to improve the control efficacy of A. retroflexus. The poor fomesafen efficacy against A. retroflexus indicated the evolution of herbicide resistance, which could be attributed to the pervasive practice of soybean monoculture in this area along with the repeated use of fomesafen over a 30-yr period (Huang et al. Reference Huang, Cui, Wang, Wu, Zhang, Huang and Wei2020; Wang et al. Reference Wang, Wang, Zhao, Zhu, Sun, Liu and Wang2019). Herbicides with specific targets nearly always seem to induce resistance in plants after years of continuous application. In addition, A. retroflexus in this area has evolved high resistance to ALS inhibitors in the past few years, and thus, the overreliance on PPO-inhibiting herbicides in this area has provided greater selection pressure for PPO-inhibitor resistance (Chen et al. Reference Chen, Huang, Zhang, Huang, Wei, Chen and Wang2015; Huang et al. Reference Huang, Chen, Zhang, Huang, Wei, Zhou, Chen and Wang2016; Wang et al. Reference Wang, Guo, Zhang, Zhao, Ge, Lv, Liu and Wang2017, Reference Wang, Wang, Zhao, Zhu, Sun, Liu and Wang2019).
PPX1 and PPX2 Gene Amplification and Sequencing
Two nearly full-length PPX1 and PPX2 transcripts of 10 individual plants from each population (R1, R2, S1, and S2) were obtained by amplifying cDNA with the two pairs of primers PPX1-F/PPX1-R and PPX2-F/PPX2-R (Table 1). The lengths of the two amplified genes were 1,529 bp and 1,549 bp, respectively. Three known point mutations endowing herbicide resistance were encompassed in the PPX2 transcripts, including Gly-210, Arg-128, and Gly-399. Sequence alignment in PPX1 revealed that one homozygous codon changed from GCA to GGA at site 292 in all tested resistant plants, which led to an amino acid change from Gly to Ala. However, this codon change did not seem to provide herbicide resistance, because the Ala-292 codon was detected in the susceptible A. tuberculatus plants (GenBank accession no. DQ386112). All amplified PPX2 transcripts from the resistant plants included a substitution of the AGG codon at site 128 with GGG, which resulted in an amino acid change from Arg to Gly. The sequencing chromatograms indicated that all Arg-128-Gly mutations were homozygous in the 10 resistant plants of each resistant population (Supplemental Figure). No other single-nucleotide polymorphisms were found in the tested plants.
Target gene mutations endowing weed resistance to herbicides are prevalent in agricultural production (Powles and Yu Reference Powles and Yu2010). A crystal structure assay of PPO indicated that Arg-128 was highly conserved and was located at a substrate-binding site (Heinemann et al. Reference Heinemann, Diekmann, Masoumi, Koch, Messerschmidt, Jahn and Jahn2007; Koch et al. Reference Koch, Breithaupt, Kiefersauer, Freigang, Huber and Messerschmidt2004). Therefore, alterations at this site may result in changes in functional proteins. To date, it has been clarified that four types of Arg-128 substitutions in PPX2 can cause weed resistance to herbicides, including Arg-128-Gly in A. palmeri and A. retroflexus, Arg-128-Met in A. palmeri, Arg-128-Ile in A. tuberculatus, and Arg-128-Leu in A. artemisiifolia (Giacomini et al. Reference Giacomini, Umphres, Nie, Mueller, Steckel, Young, Scott and Tranel2017; Nie et al. Reference Nie, Mansfield, Harre, Young, Steppig and Young2019; Rousonelos et al. Reference Rousonelos, Lee, Moreira, Vangessel and Tranel2012; Wang et al. Reference Wang, Wang, Zhao, Zhu, Sun, Liu and Wang2019). In this study, two resistant populations containing the Arg-128-Gly mutation showed highs level of resistance to fomesafen with resistance ratios of 27.0 to 27.9. Generally, the PPO Arg-128-Gly mutation confers upon weeds high levels of resistance (resistance ratio >10) to diphenyl ethers (e.g., fomesafen, lactofen, fluoroglycofen-ethyl, and acifluorfen). The published literature indicated that the greatest level of resistance to diphenyl ether herbicides was with fomesafen, with resistance levels of 80.2-fold in A. artemisiifolia and 41.8-fold in A. retroflexus (Rousonelos et al. Reference Rousonelos, Lee, Moreira, Vangessel and Tranel2012; Wang et al. Reference Wang, Wang, Zhao, Zhu, Sun, Liu and Wang2019). However, not all Arg-128 substitutions endow weed resistance to herbicides; for example, Arg-128-Lys retains sensitivity to fomesafen because of a similarly sized and positively charged side chain (Nie et al. Reference Nie, Mansfield, Harre, Young, Steppig and Young2019). In addition, two types of artificially altered Arg-128 can occur after changing one nucleotide, with Arg-128-Ser and Arg-128-Thr endowing fomesafen resistance (Nie et al. Reference Nie, Mansfield, Harre, Young, Steppig and Young2019). However, neither mutation has yet been found in the field.
In plants, there are two types of PPO, plastid-targeted PPO (PPO1) encoded by the PPX1 gene and mitochondria-targeted PPO (PPO2) encoded by the PPX2 gene (Lermontova et al. Reference Lermontova, Kruse, Mock and Grimm1997), and both types of PPO are targets of PPO inhibitors. Consistent with all previous reports of weeds resistant to PPO inhibitors, no mutations suspected of endowing resistance were identified in the PPX1 gene of the 20 plants from the two resistant populations. Therefore, PPO2 played a more important role in the evolution of PPO-inhibitor resistance. One plausible explanation is the distribution of PPO2 in organelles (Dayan et al. Reference Dayan, Owens, Tranel, Preston and Duke2014). In Amaranthus species, PPX2 encodes dual-targeted PPO2 due to a signaling sequence for plastid import (Emanuelesson et al. Reference Emanuelesson, Nielsen, Brunak and Heijne2000; Patzoldt et al. Reference Patzoldt, Hager, McCormick and Tranel2006). Therefore, PPO activity in both plastids and mitochondria will be protected when the mutation endowing resistance occurs in PPO2 in the presence of PPO inhibitors. If the resistance mutation is present in PPO1, the susceptible mitochondria-targeted PPO is still inhibited by PPO inhibitors, and thus the plants are still sensitive to PPO inhibitors.
PPO Expression Assay by qRT-PCR and ELISA
In this study, the expression level of PPO was explored at the transcription level and translation level by qRT-PCR and ELISA, respectively. To verify the specificity and amplification efficiency of primers used for qRT-PCR, the melting and standard curves were obtained. The results showed that the melting curves were all single peaks with melting temperatures >82 C and amplification efficiencies of 102.3% to 106.8%. A commercial ELISA kit for the determination of the PPO content was used in this experiment.
Twenty-three pots (with 5 plants per pot) of each population were treated with fomesafen at a dose of 37.5 g ha−1, which is higher than the GR50 value of S1 plants. Fomesafen at this rate caused susceptible plants to stunt within 6 h and undergo necrosis within 24 HAT. The quality of RNA extracted from the S1 plant leaves sampled at 24 HAT was too low to perform qRT-PCR. Therefore, the PPO gene expression of S1 plants 24 HAT is not shown in the results. During the observation time, the PPX1 and PPX2 genes were affected by fomesafen and exhibited reduced expression in both the susceptible and resistant plants over time (Figure 2). However, the degree of reduction is different. PPX1 gene expression was severely suppressed, with reductions of approximately 90% in the resistant plants at 24 HAT and more than 90% in the susceptible plants at 12 HAT. Compared with the PPX1 gene expression, the PPX2 gene expression was less affected by fomesafen, especially in the resistant plants. The variation trends of the PPX1 and PPX2 gene expression in the two resistant populations were the same (Figure 2). No significant difference in basal PPO gene expression was observed between the susceptible and resistant populations (Figure 2). The expression level of the PPX1 gene in S1 plants was higher than that of the resistant plants after 6 h of treatment, but after 12 h, the PPX1 expression level quickly decreased to a level significantly lower than that of the resistant plants (Figure 2A). The expression of the PPX2 gene in the S1 plants continuously decreased over time, dropping from 0.55 to 0.13 (relative to the expression of the internal reference gene) within 12 HAT, whereas the expression of the PPX2 gene in the resistant plants remained at approximately 0.3 (relative to the expression of the internal reference gene). At 24 HAT, the expression of the PPX2 gene in resistant plants had not changed significantly from that at 6 or 12 HAT (Figure 2B).
Figure 2. Gene expression levels of PPX1 (A) and PPX2 (B) over time in the susceptible and resistant plants after treatment with fomesafen. The gene expression level was calculated with the 2−∆∆ method. The actin gene served as an internal reference gene for normalizing the variations in cDNA amounts, and the untreated susceptible populations served as a control. Mean values and their SEs are shown. Different lowercase letters indicate significant differences at the 5% level as determined by the LSD test within treatment times.
ELISA revealed a similar basal PPO content between the susceptible and resistant plants without fomesafen treatment (Figure 3). After fomesafen treatment, the content of PPO continuously decreased within 24 HAT, from 36.4 to 15.5 ng/g in the R1 plants and from 37.9 to 18.3 ng/g in the R2 plants, before slowly increasing. However, the PPO content of S1 plants decreased over time in association with the intensification of injury symptoms and was no longer detectable at 24 HAT. In the R1 and R2 plants, the trend of the PPO content was the same (Figure 3).
Figure 3. Basal and induced PPO content variation in the susceptible and resistant plants after the fomesafen treatment using an enzyme-linked immunosorbent assay. Mean values and their SEs are shown. Different lowercase letters indicate significant differences at the 5% level as determined by the LSD test within treatment times.
Overexpression of the target enzyme is another mechanism of weed target-site resistance, including gene amplification and changes in the gene promoter (Odell et al. Reference Odell, Caimi, Yadav and Mauvais1990; Powles Reference Powles2010). The most common case is glyphosate resistance caused by EPSPS gene amplification (Gaines et al. Reference Gaines, Zhang, Wang, Bukun, Chisholm, Shaner, Nissen, Patzoldt, Tranel, Culpepper, Grey, Webster, Vencill, Sammons and Jiang2010). The expression level of PPO in the resistant and susceptible plants was the same at transcription and translation levels without the fomesafen treatment (Figures 2 and 3). After fomesafen treatment, both the PPO enzyme protein and PPO gene expression (PPX1 and PPX2) significantly decreased within 24 h, but the resistant plants showed weaker decreases than the susceptible plants, especially with respect to PPX2 gene expression and enzyme protein content (Figures 2 and 3). Although the PPO transcription and content of the resistant plants were significantly suppressed, the plants were free of significant visual damage and grew normally. After 48 h of fomesafen treatment, a slight increase in the PPO enzyme content indicated that resistant plants would gradually return to normal growth without being affected by the agent (Figure 3). It should be noted that the PPX1 and PPX2 genes seem to play different roles in herbicide resistance. In resistant plants, the expression of PPX1 may be suppressed due to an unknown feedback regulation mechanism (Hao et al. Reference Hao, Tan, Yang, Wang, Zhan, Xi and Yang2013), but PPX2 is maintained at a high level relative to PPX1. We suspect that PPX2 is less susceptible to some type of feedback regulation because of the dual-targeted peculiarity of PPO2. In this regard, PPX2 seems to play a more important role in herbicide resistance than PPX1.
In Vitro Assay of PPO Activity
In vitro assays indicated that the basal PPO activity of resistant plants was significantly higher than that of susceptible plants (Figure 4). In the absence of fomesafen, the extractable PPO activity was 3.42 nmol Proto-IX mg protein−1 h−1 in the S1 plants, and 5.45 and 5.58 nmol Proto-IX mg protein−1 h−1 in the R1 and R2 plants, respectively (Figure 4). Due to the similar expression levels of PPO between the resistant and susceptible plants without fomesafen treatment, it can be concluded that the increase in extractable PPO activity in resistant plants was not related to the expression of PPO but was related to the Arg-128-Gly mutation, which might contribute to fomesafen resistance. The results of the fomesafen sensitivity analysis of PPO are shown in Table 3 and Figure 5. A wild-type PPO isolated from the S1 plants was observably inhibited by fomesafen, with an I50 value of 0.091 μM (Table 3; Figure 5). However, PPO isolated from the R1 and R2 plants was significantly less inhibited by fomesafen than was wild-type PPO, with I50 values of 11.36 and 19.32 μM, respectively (Table 3; Figure 5). The fomesafen I50 values for R1 and R2 were 124.7- and 212.3-fold greater than those of the S1 plants, respectively (Table 3). Thus, as expected, the Arg-128-Gly mutation resulted in insensitive PPO, thus conferring high-level resistance to fomesafen, which might be one of the direct causes of resistance. Previous crystal structure and computational assays have revealed that Arg-128 is located in the pocket between the active site and the entryway of the substrate and that this site is crucial for stabilizing the substrate (Hao et al. Reference Hao, Tan, Xu, Cao, Xi and Yang2014; Koch et al. Reference Koch, Breithaupt, Kiefersauer, Freigang, Huber and Messerschmidt2004). Conformational analysis indicated that the protoporphyrinogen IX binding process should be very fast, but the protoporphyrin IX dissociation process should be slower (Hao et al. Reference Hao, Tan, Yang, Wang, Zhan, Xi and Yang2013). The substitution of Arg-128 with a nonpolar amino acid could enlarge the mouth of the pocket, which would facilitate proto egress and thus improve the enzyme’s activity. Usually, the Arg-128 mutations, including Ala-128, Ly-s128, and Glu-128, reduce the affinity between PPO and the substrate but greatly increase the catalytic speed (Hao et al. Reference Hao, Tan, Yang, Wang, Zhan, Xi and Yang2013; Heinemann et al. Reference Heinemann, Diekmann, Masoumi, Koch, Messerschmidt, Jahn and Jahn2007).
Figure 4. Extractable PPO activity in the absence of fomesafen in the resistant and susceptible Amaranthus retroflexus populations. Different lowercase letters indicate significant differences at the 5% level as determined by the LSD test.
Table 3. Parameter estimates from the logistic analysis of in vitro inhibition of PPO activity by fomesafen for the susceptible and resistant Amaranthus retroflexus populations. a
a C, lower limit; D, upper limit; b, the slope at I50. Fitting parameters and their SEs are shown.
Figure 5. In vitro inhibition of the PPO activity of susceptible and resistant Amaranthus retroflexus populations by fomesafen. Mean values and their SEs are shown.
PPO Enzyme Kinetic Analysis
The PPO Km and V max values for each population were evaluated. Km is the substrate concentration at which the reaction rate is at half-maximum and is an inverse measure of the substrate’s affinity for the enzyme. A small Km indicates high affinity. There was no difference in PPO enzyme kinetics between the two resistant populations, but differences between the resistant and susceptible plants were significant (Table 4; Figure 5). The Km value for the susceptible plants was 1.36 μM, and the resistant plants showed a higher Km (approximately 2.5-fold) than the susceptible plants (Table 4). The increased Km in the resistant plants indicates lower substrate affinity. The PPO V max values for the resistant plants were 8.24 to 7.89 Proto-IX mg protein−1 h−1, approximately 2-fold higher than those of the susceptible plants (Table 4). These results confirmed the high extractable PPO activity in the resistant plants. Although the Arg-128-Gly mutation exerted a negative impact on substrate binding, A. retroflexus plants containing Arg-128-Gly PPO grew normally in the absence of fomesafen. We speculate that the higher basal PPO activity in the resistant plants may have had a beneficial effect on plant growth (Figure 4). This mechanism partly explains the widespread prevalence of the Arg-128-Gly mutation. Although the enzyme still maintains its normal catalytic function, the change in enzyme conformation removes the important hydrogen-bonding interactions between enzyme and herbicide, which may lead to PPO insensitivity and, subsequently, fomesafen resistance (Hao et al. Reference Hao, Tan, Yang, Wang, Zhan, Xi and Yang2013, Reference Hao, Tan, Xu, Cao, Xi and Yang2014). Although PPO containing Arg-128-Gly can maintain normal function, compared with that of wild-type PPO, the Km of mutant PPO is significantly lower (Table 4). Therefore, it may exhibit obvious functional abnormalities in the case of insufficient substrate. In this regard, resistant plants containing the Arg-128-Gly mutation may reduce their fitness under certain conditions that result in a lack of substrate. However, from the actual situation, we did not observe such a fitness cost in the field or indoors. In contrast, the wide distribution of fomesafen-resistant A. retroflexus in soybean fields in Heilongjiang Province may imply that the Arg-128-Gly mutation may not cause a significant fitness penalty.
Table 4. Enzyme PPO Km (μM) and V max (nmol Proto-IX mg protein−1 h−1) values of the susceptible and resistant Amaranthus retroflexus populations. a
a Fitting parameters and their SEs are shown.
In summary, the higher extractable PPO activity in the resistant A. retroflexus plants was conferred by an Arg-128-Gly substitution in PPX2 rather than increased expression of PPO. The changes in PPO enzyme kinetics and the reduction in PPO sensitivity to fomesafen caused by the Arg-128-Gly mutation in PPX2 are the target-site resistance mechanisms in resistant A. retroflexus. In addition, PPX2 plays a more important role in resistance because of its more active expression than PPX1 after fomesafen treatment.
Acknowledgments
The authors thank all workers for their assistance in conducting this research and AJE for their professional English language editing services. This work was financially supported by the National Natural Science Foundation of China (32001940), the Natural Science Foundation of Shandong Province (ZR2020QC135), the Earmarked Fund for the China Agriculture Research System (CARS-13), and the Agriculture Science and Technological Innovation Project of Shandong Academy of Agricultural Science (CXGC2018E21). The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be considered as a conflict of interest.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/wsc.2021.14
