Introduction
Acetyl-coenzyme A carboxylase (ACCase or ACCs; EC.6.4.1.2) is an essential enzyme that catalyzes the formation of malonyl-CoA. The reaction product, malonyl-CoA, is involved in the biosynthesis of de novo fatty acids in plastids and the elongation of long-chain fatty and secondary metabolites that are crucial for energy storage, cell or organelle biomembrane structure composition, and hormonal regulation (Délye et al. Reference Délye, Pernin and Michel2011; Harwood, Reference Harwood1988; Keereetaweep et al. Reference Keereetaweep, Liu, Zhai and Shanklin2018; Konishi et al. Reference Konishi, Shinohara, Yamada and Sasaki1996; Ohlrogge and Browse Reference Ohlrogge and Browse1995; Petit et al. Reference Petit, Bay, Pernin and Délye2010; Podkowinski et al. Reference Podkowinski, Jelenska, Sirikhachornkit, Zuther, Haselkorn and Gornicki2003; Yang et al. Reference Yang, Guschina and Hurst2018; Ye et al. Reference Ye, Ma, Zhang, Li, Yang and Fu2018; Yu et al. Reference Yu, Collavo, Zheng, Owen, Sattin and Powles2007). ACCase-targeting herbicides inhibit chloroplastic ACCase activity in grasses of the Poaceae family, resulting in a decrease in fatty-acid production (Lancaster et al. Reference Lancaster, Norsworthy and Scott2018; Powles Reference Powles2005). These herbicides are commonly applied postemergence to control grass weeds in both crop and turf systems.
Herbicide resistance to ACCase-targeting herbicides has developed via target-site (TSR) and non–target-site resistance (NTSR) mechanisms. The TSR amino acid substitutions associated with ACCase-targeting herbicides occurring in the carboxyl transferase domain have been reported as follows: Gln-1756-Glu, Ile-1781-Leu, Thr-1805-Ser, Lys-1930-Arg, Trp-1999-Cys, Trp-2027-Cys, Ile-2041-Asn or Val, Asp-2078-Gly, Cys-2088-Arg, and Gly-2096-Ala (Beckie and Tardif Reference Beckie and Tardif2012; Collavo et al. Reference Collavo, Panozzo, Lucchesi, Scarabel and Sattin2011; Délye Reference Délye2005; Kaundun Reference Kaundun2010; Kaundun et al. Reference Kaundun, Hutchings, Dale and McIndoe2012, Reference Kaundun, Hutchings, Dale and McIndoe2013; Powles and Yu Reference Powles and Yu2010). NTSR encompasses a range of processes, including increased enzyme expression, increased enzyme abundance, enhanced metabolism of the herbicides, herbicide detoxification, reduced herbicide uptake, penetration, and impaired translocation. Different enzymes are involved in the development of NTSR, including, but not limited to, cytochrome P450 monooxygenases, glutathione-S-transferases, glycosyl-transferases, hydrolases, oxidases, and peroxidases (Cocker et al. Reference Cocker, Moss and Coleman1999; Kaundun Reference Kaundun2014; Kukorelli et al. Reference Kukorelli, Reisinger and Pinke2013; Powles and Yu Reference Powles and Yu2010; Preston Reference Preston2003; Yuan et al. Reference Yuan, Tranel and Stewart2007).
A research component in herbicide-resistance discovery is a functional assay evaluating the interaction of the target site with herbicide. Functional assays measure the target-site activity of suspected resistance biotypes compared with the known susceptible biotypes. Functional assays assess enzymatic activity in the presence of herbicides to determine whether amino acid changes affect the enzyme–herbicide interaction. The most commonly used functional assay to assess ACCase activity is a 14C-based radiometric assay (Cocker et al. Reference Cocker, Moss and Coleman1999; Cruz-Hipolito et al. Reference Cruz-Hipolito, Osuna, Dominguez-Valenzuela, Espinoza and De Prado2011; De Prado et al. Reference De Prado, Osuna and Fischer2004; Secor and Cséke Reference Secor and Cséke1988; Seefeldt et al. Reference Seefeldt, Fuerst, Gealy, Shukla, Irzyk and Devine1996; Yang et al. Reference Yang, Dong, Li and Moss2007). The 14C-based radiometric assay, however, is expensive and requires special 14C-detection equipment and handling of radioactive materials, and the enzyme can be insoluble in scintillation mixtures. Howard and Ridley (Reference Howard and Ridley1990) found a similar inhibition concentration value comparing the 14C-based radiometric assay and a malachite green colorimetric assay in the maize (Zea mays L.) ACCase–fluazifop-p-butyl interaction test. Compared with the 14C-based assay, the malachite green assay offers several advantages, including simplicity, specificity for inorganic orthophosphate, accuracy, high sensitivity, stability of the reagents, and lower cost, because nonradioactive materials are used for the labeling of the enzyme-substrate (Baykov et al. Reference Baykov, Evtushenko and Avaeva1988; Carter and Karl Reference Carter and Karl1982; Geladopoulos et al. Reference Geladopoulos, Sotiroudis and Evangelopoulos1991; Van Veldhoven and Mannaerts Reference Van Veldhoven and Mannaerts1987).
Our research objective was to develop the methodology and evaluate the malachite green colorimetric assay as a functional assay for evaluating ACCase–herbicide interaction. To our knowledge, the malachite green functional assay has not been utilized to assess ACCase herbicide resistance, as the majority of studies have used the 14C functional assay. Furthermore, our research will be the first to report ACCase activity for all three families of ACCase-targeting herbicides. Two resistant biotypes with the Ile-1781-Leu amino acid substitution and a known susceptible biotype of southern crabgrass [Digitaria ciliaris (Retz.) Koeler] were used as a model for evaluating the effectiveness of the malachite green assay in ACCase-resistance detection. Previous research quantified the TSR mechanism and rate response screen to these biotypes (Basak et al. Reference Basak, McElroy, Brown, Gonçalves, Patel and McCullough2019; Yu et al. Reference Yu, McCullough and Czarnota2017).
Materials and Methods
Seed Sample Collection and Growth Conditions
Previously collected seeds of two resistant biotypes (R1 and R2) of D. ciliaris with confirmed resistance to select ACCase-targeting herbicides and one susceptible (S) biotype were included in this study (Basak et al. Reference Basak, McElroy, Brown, Gonçalves, Patel and McCullough2019; Yu et al. Reference Yu, McCullough and Czarnota2017). To generate green plant material for enzyme extraction, seeds of resistant and susceptible D. ciliaris biotypes were sown in separate plastic flats containing commercial potting soil and peat moss (2:1 v/v). The plastic flats were placed for 2 wk in a greenhouse set for 32/25 C (day/night). Ambient lighting was used throughout the experiment with no supplemental light added. Relative humidity levels alternated between 65% during the day and 75% during the night. No supplementary fertility was provided because of the quick turnaround and no signs of nutrient stress were observed. Plastic flats were irrigated three times daily (around 0.2 cm cycle−1) to provide adequate moisture.
Malachite Green Colorimetric Assay
Research was conducted in the Department of Chemistry and Biochemistry, Auburn University, Auburn, AL, USA. ACCase extraction and activity bioassay were performed as described by Howard and Ridley (Reference Howard and Ridley1990) with some modifications (Supplementary Data 1). Enzymes were extracted in the cold chamber at 4 C from healthy plants of three D. ciliaris biotypes: S, R1, and R2. Approximately 10 g of fresh leaf tissues were harvested and ground in liquid nitrogen with a mortar and pestle and then suspended in 40 ml of ice-cold enzyme extraction buffer (100 mM Tricine, pH 8.0, 5mM dithiothreitol, 10 mM MgCl2.6H2O, 1 mM Na2EDTA, 0.5% [w/w] polyvinylpyrrolidone, 20% glycerol, and 1 mM phenylmethylsulfonyl fluoride). The homogenate was stirred for 30 min on ice and then filtered through four layers of cheesecloth. The solution was kept on ice until being centrifuged at 22,000 × g (Optima XE-90 Ultracentrifuge, Beckman Coulter, Inc. Brea, CA, USA) for 30 min to remove cell debris. The pellet was discarded, and the supernatant was collected and adjusted to 30% ammonium sulfate saturation with solid ammonium sulfate. After being stirred for 20 min, the solution was centrifuged at 22,000 × g for 30 min. The supernatant was decanted, adjusted to 60% ammonium sulfate saturation, and centrifuged to allow protein precipitation as previously described. The final pellet after the 60% precipitation was resuspended in 2 ml elution buffer I (10 mM Tris, 20 mM mercaptoethanol, 1 mM Na2EDTA, 1 mM benzamidine, 10 mM MgCl2.6H2O, and 20% glycerol). The enzyme extract was desalted on a Sephadex G-25 column (SIGMA@ Chemical Company, St. Louis, MO, USA) equilibrated with elution buffer II solution (10 mM Tris, 20 mM mercaptoethanol, 1 mM EDTA, 1 mM benzamidine, 10 mM MgCl2.6H2O, and 10% glycerol). The enzyme extracts were frozen at −80oC and assayed within a week of extraction.
The enzyme concentration in the enzyme extracts was measured using a Bradford assay (Bradford Reference Bradford1976) with bovine serum albumin (BSA) as a standard. The concentration of enzyme extract used for all biotypes was 5.3 µM after standardization with BSA. Using SDS-PAGE (superior protein separation−polyacrylamide gel electrophoresis) analysis, the enzyme extracts from S, R1, and R2 were separated and compared with maize. The assay was performed as three independent extractions, and each treatment was replicated three times per ACCase-targeting herbicide dose such as sethoxydim (Segment®, BASF, Research Triangle Park, NC), clethodim (Envoy®, Valent, Walnut Creek, CA), fluazifop-p-butyl (Fusillade®, Syngenta, Greensboro, NC), and pinoxaden (Axial®, Syngenta, Greensboro, NC). The assay was carried out using 96-well plates (TECAN®, Morrisville, NC), where each well contained a total of 250 µl of the reaction mixture. Each reaction mixture contained 25 µl of enzyme extract (final concentration of enzyme for each biotype in each tube was maintained to 0.53 µM by adding 25 µl of enzyme extracts), 25 µl of ACCase-targeting herbicide at a series of concentrations (0, 0.6, 1.2, 2.5, 5.0, 10, 20, and 40 µM), and 150 µl of the enzyme assay buffer (0.1 M Tricine, pH 8.0, 15 mM KCl, 3 mM MgCl2.6H2O 1 mM dithiothreitol, 0.01 BSA, 120 mM NaHCO3, and 25 mM ATP). Then 25 µl of acetyl CoA (lithium salt, final concentration 4.5 mM) was added to start the reaction. All the reaction mixtures were incubated at 30 C for 20 min before the addition of malachite green to stop the reaction. The reaction was terminated by the addition of 25 μl of malachite green termination solution. The malachite green stock solution was prepared using 72.9 mg of malachite green dissolved in 3.31 ml of 12.1 M HCl with molecular-grade deionized water added for a final volume of 200 ml. The solution was filtered through a 0.45-µm PTFE filter. The malachite green termination solution was prepared with a 5 ml malachite green stock solution mixed with 1.44 ml of 8.5 mM ammonium molybdate and 0.104 ml of 10% Triton-X.
Standard curves were generated with inorganic phosphate-containing non-treated control at 1.2, 2.5, 0.5, 1.0, 2.0, and 4.0 µM concentrations dissolved in water and added to the wells before the addition of 25 µl of malachite green solution. The absorbance of ACCase enzyme activity was taken at 630 nm colorimetrically on a microplate photometer (TECAN®) and was expressed as a percentage of the nonherbicidal control (Supplementary Data 1).
The design of the experiment was replicated twice in time as a completely randomized design with three replications. An ANOVA using PROC GLM in SAS v. 9.4 (SAS Institute Inc. Cary, NC, USA) was performed on all data to detect the significant differences among the herbicide concentrations and biotypes. The linear model was developed with herbicide treatment, herbicide rates, biotypes, replication, and experiments repeated in time as main effects. Experimental run by herbicide treatment by biotype was evaluated as an indicator of differences over experimental runs. The data were pooled overruns for subsequent analysis, as differences between the data of the two experimental runs were not detected in the ANOVA at the 0.05 probability level. Regression models were developed using Prism v. 5.0 (GraphPad Software, La Jolla, CA). ACCase-targeting herbicide concentrations causing 50% inhibition of the ACCase activity (IC50) values were estimated using nonlinear regression models. The following nonlinear regression analysis was used to calculate the IC50 value in the enzymatic experiment:
$Y = bottom + (top - bottom)/(1+10^{(X -log IC50)})$
where Y is the enzyme response (%), X is log-transformed ACCase-targeting herbicide concentration (µM), top and bottom are the plateaus in the units of the y axis, and logIC50 is the log-transformed ACCase-targeting herbicide concentration (µM). The 95% confidence intervals (α = 0.05) for the estimates were calculated for nonlinear regression model parameters. Regression equations were used to calculate inhibition concentration values at 50% (referred to as IC50 values) compared with that of the non-treated for each biotype and each ACCase-targeting herbicide. The IC50 and R/S values (ratio of R to S IC50 values) were determined for each resistant biotype versus the susceptible biotype. Percent ACCase activity relative to the non-treated response to ACCase-targeting herbicide was modeled for all three biotypes using the least-squares fit model to allow for calculation of IC50 values (Figure 1) presented in Table 1.
Response curves for percent acetyl-coenzyme A carboxylase (ACCase) activities of resistant and susceptible Digitaria ciliaris biotypes in response to the increasing concentrations of the ACCase-targeting herbicides, sethoxydim, clethodim, fluazifop-p-butyl, and pinoxaden. The response was modeled based on the log rate of ACCase-targeting herbicides to create equal spacing between rates using least-squares fit regression of ACCase activity to the non-treated check. Means are represented by differing symbols for each biotype, and regression equation models are represented by differing line types for each biotype. Vertical bars represent the standard errors of the means (n = 6). Digitaria ciliaris biotypes: R1 and R2, resistant; S, susceptible. The concentration of ACCase-targeting herbicides required to cause 50% inhibition of ACCase activity (IC50) was calculated from concentration-response curves. CI, confidence interval.
Comparison of resistant and susceptible Digitaria ciliaris biotypes for percent of acetyl-coenzyme A carboxylase (ACCase) activity to increasing ACCase-targeting herbicide concentration relative to the non-treated control measured with the least-squares fit model.
a ACCase-targeting herbicides: sethoxydim, clethodim, fluazifop-p-butyl, and pinoxaden.
b D. ciliaris biotypes: R1 and R2, resistant biotypes; S, susceptible biotype.
c In the least-squares fit equation, X represents the concentration of ACCase-targeting herbicide, Y represents the response variable of ACCase activity.
d Parameter estimates and parameter estimate 95% confidence intervals (CI) are presented as means of model comparison.
Results and Discussion
Herbicide treatment by biotype by experimental run interactions was nonsignificant (P > 0.05) for ACCase enzyme activity; therefore, data pooled over the experimental runs. Herbicide treatment by biotype was significant (P < 0.05) for all four herbicides tested. Data presented will focus on biotype response to increasing concentrations of the four herbicides tested. ACCase activity in the presence of a given herbicide concentration was expressed as a percentage of enzyme activity reduction relative to no herbicide. In general, 40 µM of all four herbicides resulted in completely diminished ACCase activity for all three biotypes. IC50 values were, however, consistently higher for ACCase activity from R1 and R2 biotypes relative to the S biotype (Figure 1). Depending on the herbicide under evaluation, IC50 for R1 ACCase was 7.6- to 21.9-fold higher than for S ACCase, and IC50 for R2 ACCase was 16.3- to 58.7-fold higher than for S ACCase (Table 1).
R1 and R2 had higher ACCase activity in response to sethoxydim compared with S (Figure 1). For example, sethoxydim at 0.63 µM inhibited S ACCase activity 41%, while R1 and R2 ACCase activity was inhibited 7.5% and 3.7%, respectively. Sethoxydim at 1.3, 2.5, and 5 µM inhibited S ACCase enzyme activity 59.5%, 73.2%, and 81.1%, respectively, while R1 activity was inhibited 11.2%, 19.4%, and 34.9%, respectively, and the R2 biotype activity was inhibited 6.9%, 12.1%, 23.4%, respectively. Sethoxydim at 10 and 20 µM inhibited S biotype ACCase activity 88.3% and 91.2%, respectively, whereas R1 was inhibited 68% and 78.8%, respectively, and R2 was inhibited 38.1% and 54.3%, respectively. R1 and R2 IC50 values were 15.3 and 41.1 µM, respectively, compared with 0.7 µM for S, which was 21.9-fold higher than S for R1 and 58.7-fold higher than S for R2 (Figure 1).
Previous research reported R1 and R2 were less resistant to clethodim when foliar applied compared with other ACCase-targeting herbicides (Yu et al. Reference Yu, McCullough and Czarnota2017), which was confirmed with the assay. S was relatively more sensitive to clethodim than R1 and R2 (Figure 1). For example, clethodim at 0.63, 1.3, and 2.5 µM inhibited S ACCase activity 48.3%, 64.8%, and 75.9%, respectively, while R1 was inhibited 14.9%, 27.4%, and 44.5%, respectively, and R2 10.7%, 19.7%, and 30.6%, respectively. R1 and R2 IC50 values for clethodim were 3.5 and 7.5 µM, respectively, compared with 0.46 µM for S, which was 7.6- and 16.3-fold higher for R1 and R2, respectively, than S. The assay was sensitive enough to detect a difference in the inhibition of R1 and R2 by clethodim, which was unexpected based on previous postemergence applications (Yu et al. Reference Yu, McCullough and Czarnota2017). R1 and R2 ACCase activities were lower in the presence of clethodim relative to sethoxydim, fluazifop-p-butyl, and pinoxaden, which may explain the difference in whole-plant response observed previously.
Similar to sethoxydim, fluazifop-p-butyl and pinoxaden inhibited ACCase activity of S more than that of R1 and R2. Fluazifop-p-butyl at 0.63 to 10 µM inhibited S biotype ACCase activity 47.1% to 94.2%, while the enzyme activity of the R1 and R2 biotypes was inhibited 10.8% to 64.4% and 7.1% to 52%, respectively. R1 and R2 IC50 values for fluazifop-p-butyl were 8.9 and 17.1 μM compared with 0.5 μM for S, which was 17.8- and 34.2-fold higher for R1 and R2, respectively, than S. Pinoxaden at 0.63 to 10 µM inhibited S ACCase activity 26.4% to 87%, respectively, while R1 activity was inhibited 8.6% to 55.4%, respectively, and R2 activity was inhibited 3.2% to 45.9%, respectively. R1 and R2 IC50 values for pinoxaden were 12.7 and 28.4 µM, respectively, compared with 1.5 µM for S, which was 8.5- and 18.9-fold higher for R1 and R2, respectively, than S (Figure 1).
Research Implications
While the ultimate research purpose was to develop the malachite green assay and test its effectiveness, unique results were acquired from this research. The development of the malachite green functional assay was conducted using previously researched D. ciliaris populations. Understanding previous research is necessary to interpret the assay results. Both R1 and R2 D. ciliaris were first identified as sethoxydim resistant in centipede grass [Eremochloa ophiuroides (Munro) Hack.] sod production fields in Georgia, USA, by Yu et al. (Reference Yu, McCullough and Czarnota2017). R1 and R2 were determined to be >64 times more resistant than the susceptible population. R1 and R2 had similar cross-resistance to fenoxaprop and fluazifop, but less resistance to clethodim. Clethodim at 290 g ha−1 controlled D. ciliaris 83% at 4 wk after treatment. Clethodim reduced susceptible shoot dry biomass to 21% of the non-treated, while R1 and R2 were reduced to 47% and 46%, respectively, of the non-treated. Additional research by Basak et al. (Reference Basak, McElroy, Brown, Gonçalves, Patel and McCullough2019) identified R1 and R2 as resistant to pinoxaden and a known target-site mutation, Ile-1781-Leu.
We theorized that R1 and R2 would respond similarly to herbicides using the malachite green assay. However, one unanticipated result was observed for this assay. ACCase activity of R1 and R2 biotypes was inhibited to different extents for each herbicide. This result can only be because of differential interaction with the ACCase substrate and the tested ACCase-targeting herbicide, as no absorption, translocation, or metabolism is at play as would be the case when screening whole plants. The ACCase carboxyl transferase domains were sequenced and reported for S, R1, and R2 previously (Basak et al. Reference Basak, McElroy, Brown, Gonçalves, Patel and McCullough2019). No other amino acid substitutions except Ile-1781-Leu were observed between R1 or R2 that would explain the difference between these biotypes (Supplementary Data 2). Therefore, the specific mechanisms behind lower herbicide inhibition of ACCase from the R2 biotype compared with R1 and S remain unknown. We theorize that R2 expresses more resistant chloroplastic ACCase homologues containing Ile-1781-Leu compared with non-resistant homologues; however, such is only a theory at this time and was not a focus of our research. Digitaria ciliaris is a polyploid species with ACCase encoded on separate progenitor genomes (Adoukonou-Sagbadja et al. Reference Adoukonou-Sagbadja, Schubert, Dansi, Jovtchev, Meister, Pistrick, Akpagana and Friedt2007; Bennett et al. Reference Bennett, Bhandol and Leitch2000). Enzyme extraction for this procedure provides a bulk sample of all translated ACCase, resulting in a mixture of resistant and susceptible chloroplastic ACCase isoforms in the extract. While the total amount of enzyme is the same in each sample, it is unknown what the ratio of resistant to susceptible isoforms is in an extracted enzyme sample. This unexpected result is seen as a positive outcome, as it indicates the sensitivity of the assay in detecting subtle differences in ACCase to ACCase-inhibiting herbicide interactions that will be beneficial for determining mechanisms of resistance in the future.
The primary purpose of this research was to evaluate the effectiveness of the malachite green assay to assess plastidic ACCase–herbicide interactions. Based on these results, we conclude that the malachite green assay is a highly sensitive assay for measuring ACCase activity as well as a functional assay for ACCase-targeting herbicide resistance. While only the Ile-1781-Leu amino acid substitution was evaluated, we see no reason that other known mutations could not be evaluated in the same system. Utilization of the malachite green assay in the future will eliminate the need for a 14C-based radiometric assay and may uncover other unknown subtle differences in ACCase to herbicide interactions that are still unknown.
Acknowledgments
This publication was supported by the Alabama Agricultural Experiment Station and the Hatch Program of the National Institute of Food and Agriculture, U.S. Department of Agriculture. The authors declare that the research was conducted without any commercial or financial interactions that could be interpreted as likely conflicts of interest.
Open access
