Resistance mutations of Pro197, Asp376 and Trp574 in the acetohydroxyacid synthase (AHAS) affect pigments, growths, and competitiveness of Descurainia sophia L

D. Sophia is one of the most problematic weed species infesting winter wheat in China, and has evolved high resistance to tribenuron-methyl. Amino acid substitutions at site of Pro197, Asp376 and Trp574 in acetohydroxyacid synthase (AHAS) were mainly responsible for D. sophia resistance to tribenuron-methyl. In this study, D. sophia plant individually homozygous for specific AHAS mutation (Pro197Leu, Pro197His, Pro197Ser, Pro197Thr, Asp376Glu and Trp574Leu) were generated. In addition, the effects of resistance mutations on pigments, growths and competitiveness of susceptible (S) and resistant (R) plants of D. sophia were investigated. The results indicated the R plants carrying Pro197Leu or Pro197His or Asp376Glu or Trp574Leu displayed stronger competitiveness than S plants. The adverse effects on R plants aggravated with the increase of R plants proportion, which made the R plants against domination the weed community in absent of herbicide selection. Therefore, these resistance mutation have no obvious adverse effects on the pigments (chlorophyll a, chlorophyll b and carotenoid), relative growth rates (RGR), leaf area ratio (LAR) and net assimilation rate (NAR) of R plants.

Descurainia sophia L. is an annual and notorious broad-leaf weed infesting winter wheat, which has evolved extremely high resistance to SU herbicide tribenuron-methyl across China [20][21][22] . Our research confirmed the TSR and NTSR mechanisms conferred D. sophia resistance to tribenuron-methyl. Resistance mutations were identified at site of Pro197 (substituted by Ala, Leu, Thr, Ser, and His) or Asp376 (by Glu) or Trp574 (by Leu) in AHAS1 or (and) AHAS2 in tribenuron-methyl resistant D. sophia [22][23][24] . In addition, one or more cytochrome P450s mediated D. sophia resistance to tribenuron-methyl 25 . Notwithstanding this, the pleiotropic effects of resistance mutations on the growths of D. sophia plants were not reported. The objectives of this study were to investigate the impacts of resistance mutations on: (1) the pigments contents in S and R plants; (2) classic growth of D. Sophia growth, such as relative growth rates (RGR), leaf area ratio (LAR), net assimilation rate (NAR) and pigment contents; (3) relative competitive ability of susceptible (S) and resistant (R) plants under condition of monoculture or admixture.

Materials and Methods
Plant materials. The S (SD8) D. sophia was collected at Linyi city of Shandong province in China that had never been treated with herbicides, which was confirmed to be individually homozygous by genotyping before. The original D. sophia population of each purified subpopulation was harvested from winter wheat fields in China, where tribenuron-methyl had been used continuously for at least fifteen years. In order to minimize the genetic background differences of different R plants, plants individually homozygous for specific AHAS mutation (Pro197Leu, Pro197His, Pro197Ser, Pro197Thr, Asp376Glu and Trp574Leu) were grown to generate seeds (Fig. 1). Seeds from single plants were collected separately and were pooled after genotyping. By this way, purified subpopulations individually homozygous for specific AHAS mutation (Pro197Leu, Pro197His, Pro197Ser, Pro197Thr, Asp376Glu and Trp574Leu) were obtained and used in this study ( Table 1).
Seeds of purified S and R subpopulations were immersed in 30% hydrogen peroxide solution for 35 min, and then soaked in 0.03% gibberellins solution for 24 h after rinsing with distilled water. Then, seeds were placed in Petri dishes to germinate for 96 h. Germinating seedlings with similar size were selected and transplanted into square plastic pots (7.5 cm slides) containing moist loam soil, and were brought up in an artificial climate chamber under conditions of 25 °C/15 °C (light/dark), 14 h photoperiod with luminous intensity of 15,000 lx. D. sophia plants were watered and rearranged regularly to minimize the environmental effects on the plants growth.
Whole-plant response experiments to tribenuron-methyl. In order to confirm the resistance to tribenuron-methyl of each purified R subpopulation, seedlings at 14 days after transplant (DAT) were used for whole-plant response experiment. Plants were returned to artificial climate chamber after herbicide treatment, and the above-ground shoots were harvested 21 days later. The above-ground shoots were oven dried for 96 h at 65 °C. The experiment was conducted with three replicates per herbicide dose and repeated twice.
Determination of the pigment content (chlorophyll a, chlorophyll b and carotenoid) in S and R plants. The extraction and determination of chlorophyll a, chlorophyll b and carotenoid were conducted according the methods described by Wellburn (1994) 26 . Above-ground seedling of individual plant at 35 or 50 DAT was ground to fine powder with a mortar and pestle in liquid nitrogen, and then homogenized in 3 mL water solutions containing 80% acetone. Six plants of S or each R subpopulation were used for pigment test. The homogenate was filtered into a brown bottle through filter paper wetting with 80% acetone. The chloroplast pigment on the filter paper was washed with 80% acetone solutions, and combined with the filtrate. The filtrate was adjusted to 25 mL by adding 80% acetone, and used for following test. Six plants were selected from each subpopulation and used for pigment extraction.
The absorbance value of each filtrate at 470, 646 and 663 nm was determined with Lambda 35 spectrophotometer (Perkins-Elmer). The concentration of chlorophyll a, chlorophyll b and carotenoid was calculated respectively by formulas (1), (2) and (3). The total chlorophyll is the sums chlorophyll a and b.

= .
− . The contents of various pigments in a unit of fresh weight of the tissue were calculated by following formula.
where A is the content of a pigments; C is the concentration of pigments; n-is the volume of extraction solution; N is the dilution ratio; W is the fresh weight of sample.

Determination of RGR, LAR and NAR of S and R plants.
The above-ground shoots were harvested respectively at 23, 29, 34, 39, 45, 55 and 63-day stage. The above-ground shoots were oven dried for 96 h at 65 °C, and dry weight was measured. The leaf area of each plant was measured immediately after harvest.
The unbiased RGR was estimated by the formula proposed by Hoffmann and Poorter (2002) 27 . RGR = (ln W 2 − ln W 1 )/(t 2 − t 1 ). Where W 1 and W 2 are means of dry weight per plant at times t 1 and t 2 . The ln W 1 and ln W 2 are the natural logarithm-transformed means of dry weight per plant.
Leaf area per plant was measured with Photoshop CS3 extended (Adobe Systems Inc., USA). LAR was calculated by the formula proposed by Hunt (1982) 28 Where W 1 and W 2 are means of dry weight per plant at times t 1 and t 2 , L A1 and L A2 are means of leaf area per plant at t 1 and t 2 . The ln W 1 and ln W 2 are the natural logarithm-transformed means of dry weight per plant, ln L A2 and ln L A2 are the natural logarithm-transformed means of leaf area per plant.
NAR was estimated by the formula proposed by Hunt (1982). 28 . Where W 1 and W 2 are means of dry weight per plant at times t 1 and t 2 , L A1 and L A2 are means of leaf area per plant at t 1 and t 2 . The ln W 1 and ln W 2 are the natural logarithm-transformed means of dry weight per plant.

Relative competition ability of S and R plants under condition of monoculture. Individual seed-
ling was transplanted into a square plastic pot (7.5 cm slides), and harvested at 35 and 50 DAT respectively. The leaf area of each plant was measured immediately after collection. The above-ground shoots were oven dried for 96 h at 65 °C, and dry weight was measured. Total 20 plants from each S or R subpopulation were selected for test.  29 . The experiment was conducted in a randomized complete block design with four replications. The above-ground shoots of S or R plants in the same tray were harvested separately at 50 DAT. The leaf area of each plant was measured immediately after harvest. The above-ground shoots were oven dried for 96 h at 65 °C, and dry weight of above-ground biomass was measured. The relative crowding coefficient (RCC) was calculated according to the following formula 17,30 . where, DB S n:n and DB R n:n is the mean dry biomass per S and R plant respectively at ratio of n:n, N is the number of mixed plantings; here N = 3. Based on this definition, an RCC value greater than 1.0 indicate S is superior competitiveness over R; an RCC value lower than 1.0 shows R is outcompeting S; While, an RCC value around 1.0 indicates S and R have similar competition ability.

Statistical analyses.
The data of whole-plant response experiments was converted into the percentage of control and subjected to a non-linear regression analysis using GraphPad Software (v.5.0) 31,32 . In this study, two AHAS isozymes with different sensitivities to tribenuron-methyl were confirmed to coexist in all R (pHB8, pHB22, pHB23, pHB24, pHB25 and pHB42) subpopulations. The GR 50 (herbicide concentrations causing 50% plant growth reduction) values for all S and R subpopulations were calculated using single-sigmoid equation f(x).
One-way analysis of variance (ANOVA) with Dunnett's post-test (α = 5%) was performed to assess pairwise differences in plant growth (RGR, NAR, LAR, BCAAs content and relative competition ability). Data Availability. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Results
Whole-plant response experiments to confirm resistance to tribenuron-methyl in each R subpopulation. The whole-plant response experiments established that all R subpopulations exhibited extremely high resistance levels to tribenuron-methyl (Fig. 2). The pHB25 subpopulation carrying Asp376Glu mutation exhibited the highest resistance level to tribenuron-methyl, with the resistance index (RI) of 815. The RI value of pHB42 subpopulation (with Trp574Leu) was 366.3, which was higher than all plants carrying Pro197 mutations (Pro197Leu, His, Thr and Ser). In contrast, the resistance level of R subpopulations carrying different Pro197 mutations displayed no significant differences.
Effects of resistance mutations on the pigments (chlorophyll a, chlorophyll b and carotenoid) in S and R plants. At   (carotenoid, chlorophyll a, chlorophyll b and total chlorophyll) in D. sophia plants, while the other resistance mutations have no obvious effects on corresponding R plants. In contrast to results of 35 DAT, the pigment contents in S and all R subpopulations at stage of 50 DAT exhibited no significant differences ( Table 2).

Effects of resistance mutations on RGR, LAR and NAR of S and R plants. The Pro197Leu and
Pro197Thr reduced plant growth during all tested periods, and decreased RGR about 13% and 23% during period of 23-63 DAT respectively. The other mutations decreased RGR only at partial stages. In addition, all mutations had adverse pleiotropic effects on plant growth during periods of 23-34 and 23-39 DAT. While, only partial mutations had negative impacts on RGR during other periods (Table 3).
For the LAR and NAR during 23-63 DAT, only Trp574Leu increased the LAR (about 29%), while the other mutations had no obvious pleiotropic effects on LAR and NAR. In addition, the Trp574Leu increased the LAR during all tested periods, all mutations decreased the NAR during period of 23-39 DAT (Tables 4 and 5).

Relative competition ability of S and R plants under monoculture (one plant per pot). At stage
of 50 DAT, the dry weight of S plants increased about 56%, 46%, 42%, 18% and 70% than R plants carrying mutation of Pro197Leu, Pro197His, Pro197Thr, Asp376Glu and Trp574Leu, respectively. While the S and R plants (with Pro197Ser) displayed no significant differences in terms of dry weight. By contrast to plants at 50 DAT, the effects of resistance mutation on S and R plants at 35 DAT were complex. For example, the Pro197His and Trp574Leu significantly decreased the dry weight of R plants; Pro197Leu, Pro197Thr and Asp376Glu did not change the dry weight of R plants comparing with S plants. Therefore, the Pro197Ser mutation significantly increased the dry weight of R plants (Table 6).
In contrast to dry weights, the individual leaf area of S at 50 DAT was about 39%, 26%, 42% and 57% larger than R plants carrying Pro197Leu, Pro197His, Pro197Thr and Trp574Leu, respectively. In addition, the individual leaf area of S (35 DAT) was about 45% and 53% larger than R plants with Pro197His and Trp574Leu, respectively. Therefore, the individual leaf area of S (35 DAT) displayed no significant differences with R plants carrying Pro197Leu, Pro197Thr and Asp376Glu (Table 6).    All the R (carrying Pro197Leu, Pro197His, Asp376Glu or Trp574Leu) displayed greater competitiveness over S plants according the RCC values (by far less than 1.0) in terms of dry weight and leaf area. While, the pHB23 (with Pro197Ser) and S plants displayed similar competition ability in terms of dry weight (with RCC of 0.92) and leaf area (RCC of 0.96) respectively. By contrast, pHB24 plants with Pro197Thr mutation displayed stronger competitiveness than S plants in term of dry weights (with RCC of 0.62), and exhibited less competition ability than S plants in term of individual leaf area (with RCC of 1.28) (Tables 7 and 8).

Discussion
Effects of resistance mutation on resistance in D. sophia. At present, total seven resistance mutations were identified in AHAS isozymes at sites of Pro197 (Pro197Leu, His, Ser, Thr, Tyr), Asp376 (Asp376Glu) and Trp574 (Trp574Leu) in D. sophia populations [22][23][24] . In recent years, we have identified six out of seven mutations from more than sixty D. sophia populations. In order to minimize the differences in genetic background, individual plant with homozygous AHAS mutation (Pro197His, Pro197Ser, Pro197Thr, Pro197Leu, Asp376Glu and Trp574Leu) was purified and used in this study. As expected, all resistance mutations caused D. sophia subpopulations evolve extremely high resistance to tribenuron-methyl, which were confirmed by previous experiments of dose-response, AHAS sensitivity inhibition and AHAS mutation identification [21][22][23][24] . In addition, the resistance mutations on the characterization of AHAS isozymes were conducted. The results (not include in this manuscript) also confirmed these resistance mutations played very important roles in the resistance evolution of D. sophia to tribenuron-methyl. Among these resistance mutations, Asp376Glu conferred the highest resistance to tribenuron-methyl. The D. sophia plants with Asp376Glu survived in tribenuron-methyl at dose of 37.5 g a.i. ha −1 , which killed completely the plants carrying mutations of Pro197 and Trp574Leu (Fig. 1). Therefore, the Asp376Glu mutation is such a weak AHAS resistance mutation (in terms of growth inhibited by chlorsulfuron) in Raphanus raphanistrum populations comparing with the Ala122Tyr, Pro197Ser and Trp574Leu mutations 33 . Obviously, the resistance or cross-resistance patterns conferred by a given resistance mutation were depended not only on the site of mutation, but also on specific amino acid substitution. While it is not clear whether this difference is related with weed species or (and) specific herbicide. Hence, the impact of resistance-endowing mutations on resistance should be evaluated on a case-by-case basis, and generalizations should be avoided.

Effects of resistance mutation on the classic growth of S and R plants.
The RGR is the product of NAR and LAR, which depended on genetic background and environmental condition. Where, NAR is largely the net result of carbon gain (photosynthesis) and losses (respiration, exudation and volatilization) expressed per     The resistance costs are considered as a basic tenet of evolutionary genetic, therefore the resistance costs are not necessary and universal in all resistance cases. The expression of resistance cost was strongly influenced by resistance mechanisms, specific resistance alleles, characteristics of target enzyme, genetic background, weed species and growth environment 2 . Hence, it is difficult to accurately assess the resistance costs in R weeds. Although a great deal of research effort has been invested to measure the resistance costs in R weed species, numerous studies were flawed. Vila-Aiub et al. (2009) reported only 25% of studies assessing resistance costs explicitly met the criteria of controlling genetic background 2 . To date, only a few resistance cost cases were confirmed in AHAS herbicide-resistant weed species. For example, Trp574Leu mutation of Amaranthus powellii can cause pleiotropic effects on the early growth and development in competitive conditions. Not only the leaves of the resistant plants were distorted and much smaller than those of S plants, one S population outperformed on R population by 7~15 times under competitive conditions 17 . In addition, the Ala205Val mutation reduced the reproductive output and fitness in resistant S. ptychanthum comparing with S biotype. This would likely cause S individuals to dominate in the absence of herbicide selection pressure 35 .