Investigation of resistance mechanisms to flucarbazone-sodium in wild oat (Avena fatua L.) from China

Abstract

Background

Wild oat (Avena fatua L.) is a self-pollinating, allohexaploid species in the family Gramineae (grasses), which is a malignant weed that mainly harms crops such as wheat. In recent years, a decline in the control efficiency of flucarbazone-sodium against wild oat has occurred in some regions of China.

Results

We identified an ALS-resistant A. fatua population (R population). Whole-plant response assays revealed that the R population exhibited a moderate level of resistance (5.9-fold) to flucarbazone-sodium. Pre-treatment with malathion significantly reduced flucarbazone-sodium resistance in the R population. The known mutation sites and ALS gene relative expression that confer resistance to ALS inhibitor herbicides were not found in R population. Following flucarbazone-sodium treatment, the expression of eight genes related to metabolic enzymes was investigated using quantitative real-time PCR (qRT-PCR). CYP92A6 and the Aldo/keto reductase family were highly expressed in the R population after the application of flucarbazone-sodium.

Conclusions

The mechanism of flucarbazone-sodium resistance in A. fatua is mediated by NTSR, nor TSR. Two genes, CYP92A6 and the Aldo/keto reductase family, were discovered to be possibly related in the metabolism of NTSR in the A. fatua population, justifying more functional studies. The results will serve as a data resource for further studies on the molecular mechanisms of A. fatua to flucarbazone-sodium.

Background

More than 3 million tons of grain were loss each year in China because weeds constantly compete with crops for light, water and nutrients [1]. The use of synthetic herbicides for weed control in China since 1956 has reduced weed control costs and improved weed control efficiency, resulting in higher crop yields [2].

Acetohydroxyacid synthase (AHAS, EC 2.2.1.6), also known as acetolactate synthase (ALS), is a crucial enzyme that catalyzes the production of branched-chain amino acids in plants and microorganisms [3]. Flucarbazone-sodium was registered in China in 2008. It has the advantages of broad-spectrum, low toxicity and wide crop selectivity, and is one of the commonly used ALS inhibitors in wheat fields [4]. However, an excessive dependence on herbicides for weed management invariably results in the emergence of weeds that are resistant to herbicides [5]. It is crucial to comprehend the mechanisms of herbicide resistance for more efficient weed management.

Herbicide resistance mechanisms can be broadly classified into two categories, namely, target site resistance (TSR) mechanisms and non-target site resistance (NTSR) mechanisms [6]. TSR mechanisms include increased expression of the target protein or structural changes to the herbicide binding site, with the latter being the most frequently reported [7]. For example, the ALS gene has been identified to have nine sites where amino acid mutations result in resistance, including Ala-122, Pro-197, Ala-205, Phe-206, Asp-376, Arg-377, Trp-574, Ser-653, and Gly-654 [8, 9]. NTSR mechanisms include reduced herbicide uptake and translocation, increased herbicide sequestration, and enhanced degradation or metabolism of the herbicide to less toxic compounds [10]. Among the NTSRs found in weed species, enhanced metabolism of the herbicides by the enzymes like cytochrome P450s (CYP450s) and glutathione S-transferases (GSTs), is the most widely discussed mechanism [11, 12].

Wild oat (Avena fatua L.) is a self-pollinating, allohexaploid species in the family Gramineae (grasses), which is a malignant weed that mainly harms crops such as wheat [13–15]. Since 1992, resistant wild oat populations have been identified in Western Australia, followed by an increasing number of studies on wild oat resistance [16]. In recent years, A. fatua have also been widely shown to be resistant to five types of herbicides including ACCase (Acetyl-CoA carboxylase) inhibitors, ALS inhibitors, protoporphyrinogen oxidase inhibitors, very long-chain fatty acid synthesis inhibitors and glyphosate [17–19]. In previous studies, A. fatua have evolved TSR and NTSR to sodium flucarbazone-sodium and the effectiveness of flucarbazone-sodium against A. fatua has declined in some parts of China [20, 21]. However, there is no report about the resistance mechanisms for flucarbazone-sodium in A. fatua in China. The aims of this study were to: (1) evaluate the resistance levels to flucarbazone-sodium in different A. fatua populations; (2) determine the underlying mechanism of A. fatua resistance to flucarbazone-sodium.

Results

Impact of P450 and GST inhibitors on resistance to flucarbazone-sodium

Whole-plant dose–response tests revealed GR₅₀ values of 47.91 and 8.18 g ai ha⁻¹ for the R and S populations, respectively, and RI of 5.9 (Table 1). The sensitivity to flucarbazone-sodium was increased in both populations by pretreatment with the malathion and NBD-Cl. Pretreatment of malathion reduced the GR₅₀ of resistant populations to 16.67 g ai ha⁻¹ and pretreatment of NBD-Cl to 36.55 g ai ha⁻¹. This indicates that the R population develops metabolic resistance. Furthermore, the S population was able to survive at 4X (126 g ai ha⁻¹) of flucarbazone-sodium and the ability of metabolic inhibitors to reverse resistance in the S population suggests that the S population used in this study also exhibited metabolic resistance (Fig. 1).

Table 1.

Effect of P450 inhibitor malathion and GST inhibitor NBD-Cl on flucarbazone-sodium resistance in susceptible and resistant A. fatua populations

Treatments	GR50 (g ai ha−1) ± SE a		RIb
	S	R
Flucarbazone-sodium	8.18 ± 1.40	47.91 ± 7.31	5.9
Flucarbazone-sodium + mathion	3.79 ± 0.37	16.67 ± 3.78	4.4
Flucarbazone-sodium + NBD-Cl	3.81 ± 0.51	36.55 ± 10.19	9.6

^aSE standard error

^bRI resistance index defined as the ratio of GR₅₀ value of R population and GR₅₀ value of S population

Fig. 1.

Dose–response curves of susceptible (S) and resistant (R) A. fatua populations to flucarbazone-sodium with ( +) or without P450s inhibitor malathion (M) and GST inhibitor NBD-C1 (N) pre-treatment. Each data point represents the mean fresh weight (percentage of control) ± SE of two repeated experiments

ALS gene sequencing

The target gene amplified in this study was proven to be reliable due to the amino acid sequences produced from the amplification were 97.12% similar to the amino acid sequence of the wild oat ALS gene (GenBank JN175309). In our study, the findings of PCR sequencing revealed signs of heterozygosity; given the species is self-pollinating and allohexaploid, we anticipate that it possesses several copies of ALS, as has recently been discovered in Echinochloa species [22]. We discovered that there was no difference in the ALS gene sequence between the R and S populations (Fig. 2). Therefore, there is no target-site resistance due to mutations in the R population.

Fig. 2.

Alignment of CDS regions encoding ALS gene in S and R populations of A. fatua. The amino acid position of ALS was based on Arabidopsis thaliana

ALS gene expression levels

In the Fig. 3, the level of ALS gene expression was not significant different in the R population than in the S population at 0 day (control), but it was significantly lower in both populations compared to the control at all time periods after flucarbazone-sodium treatment. On the third day of application, there was a difference in expression between the R and S populations, but the rest of the time periods were not significantly different. This indicates that ALS gene overexpression was not present in the R population.

Fig. 3.

Expression level of the ALS gene before and after flucarbazone-sodium treatment at five time periods in R and S plants. Note: *** indicates significant difference at 0.001 level. Bars are means ± SEM

Transcriptome sequencing and analysis

It was decided to use Avena sativa (OT3098) as the reference genome. All sample's clean reads were compared in turn with the chosen reference genome in a ratio that ranged from 90.44% to 95.55%. The transcriptome analysis of 12 samples was completed, and a total of 91.43 Gb clean data was obtained, and the clean data of each sample reached more than 6.69 Gb, and the percentage of Q30 bases was more than 94.14%. The mapping rate was generally higher than 90% (Total Mapped Reads) when the reference genome was fully annotated and there was no contamination in the relevant experiments. Based on the results of expression quantification, the differential gene analysis between groups was performed, and we obtained. The genes that were differentially expressed between the four groups were analyzed. There are 1721 differentially expressed genes in R versus S samples after flucarbazone-sodium, with 869 genes being upregulated and 852 being downregulated.

The sequences were annotated using protein and nucleotide databases [Gene Ontology (GO); Kyoto Encyclopedia of Genes and Genomes (KEGG); orthologous groups and functional annotation (eggNOG); non-redundant protein sequences database (NR); Swiss-Prot, manually annotated and reviewed protein sequence database (Swiss-Prot); Protein family (Pfam)]. To characterize the functional classifications of the annotated unigenes, GO and KEGG analyses were performed to access the distributions of functional categories. The GO database annotated a total of 79,812 unigenes, the KEGG database annotated 52,672 unigenes, and the NR database annotated 9067 unigenes (Table 2). To further characterize the function of the identified DEGs, we performed GO and KEGG pathway enrichment analyses in RT_VS_ST group. According to the GO pathway, differential genes are more involved in cellular component and there were 659 genes related to the metabolic process (Fig. 4). According to the KEGG analysis, it was found that there were more enriched metabolism-related pathways with a greater number of genes enriched in the carbohydrate metabolism and energy metabolism pathways (Fig. 5).

Table 2.

Functional database annotation analysis (GO, KEGG, EggNOG, NR, Swiss-prot and Pfam) for the expressed genes and transcripts

Pathway	Gene number (percent)	Transcript number (percent)	All_Gene number (percent)	All_Transcript number (percent)
GO	79,812(0.8602)	79,445(0.8603)	88,554(0.861)	88,554(0.861)
KEGG	52,672(0.5677)	52,454(0.568)	57,104(0.5552)	57,104(0.5552)
EggNOG	88,991(0.9591)	88,576(0.9591)	98,380(0.9565)	98,380(0.9565)
NR	90,679(0.9773)	90,256(0.9773)	100,405(0.9762)	100,405(0.9762)
Swiss-Prot	80,817(0.871)	80,449(0.8711)	89,600(0.8711)	89,600(0.8711)
Pfam	59,491(0.6412)	59,205(0.6411)	65,384(0.6357)	65,384(0.6357)
Total_annoation	90,711(0.9776)	90,287(0.9777)	100,446(0.9766)	100,446(0.9766)

Fig. 4.

Gene ontology (GO) analysis of unigenes. The unigenes were summarized in biological process, cellular component and molecular function

Fig. 5.

KEGG function classification results of the annotated unigenes in A. fatua population

Identification and screening of differentially expressed genes and the qPCR validation

We focused on metabolism-related gene families associated with NTSR, and among these genes we screened for (1) significant up-regulation of > twofold in RT compared to ST group with P ≤ 0.05 (2) transcriptome predicted CDS fragment > 500 bp (3) the gene annotated to metabolism-related pathways. Eight genes, including one Aldo/keto reductase family, five ABCs, and two P450s, have been investigated for association with NTSR.

To verify the RNA-Seq results of candidate genes, their expression levels in response to flucarbazone-sodium were evaluated with qPCR. Expression differences were compared between the RT and ST group after flucarbazone-sodium treatment. Using RNA-Seq samples, five contigs (Pepsico1_Contig31308.path4, Pepsico1_Contig15890.path1, Pepsico1_Contig15890.path1, Pepsico1_Contig30663.path3, and Pepsico1_Contig9175.path1) were significantly up-regulated with qRT-PCR. Pepsico1_Contig7137.path2 and Pepsico1_Contig10080.path3 were constitutively higher in the R population using additional confirmed samples (Table 3, Fig. 6). The Aldo/keto reductase family and the CYP92A6 annotation appear to be linked to flucarbazone-sodium resistance in the R population.

Table 3.

The candidate differentially expressed genes related to non-target-site resistance of A. fatua to flucarbazone-sodium. Significance by Student’s t tests is indicated by *P < 0.05, **P < 0.01 and *** p < 0.001

Gene ID	Function annotation	Log2 (RT/ST)
		RNA-seq	validation using RNA-Seq samples	validation using individuals from R and S populations
Pepsico1_Contig31308.path4	ABCA2	8.84	3.64***	1.15
Pepsico1_Contig37457.path1	ABCC13	8.26	0.91	1.34
Pepsico1_Contig15890.path1	ABCF1	8.23	1.46*	1.07
Pepsico1_Contig30663.path3	ABC transporter C family MRP4	8.2	0.30*	0.94
Pepsico1_Contig9175.path1	CYP71E1	6.79	0.17**	1.44
Pepsico1_Contig10080.path3	CYP92A6	5.70	2.43*	5.28***
Pepsico1_Contig7137.path2	Aldo/keto reductase family	3.59	3.30**	3.37**
Pepsico2_Contig4209.path1	ABCC1	2.85	1.61	0.69

Fig. 6.

Relative expression of eight candidate genes in S and R populations after flucarbazone-sodium treatment by the RNA-seq and qPCR

Discussion

The yield of crops around the world is seriously threatened by weed resistance to herbicides, and the number of resistant weed biotypes has been rising quickly in recent years. There are currently 269 weed species in existence, including 154 dicotyledonous weeds and 115 monocotyledonous weeds, and they all exhibit varied degrees of resistance to 21 of the 31 known herbicide action sites [9].

Currently, target-site mutations and target gene overexpression are the main reasons for TSR [23]. This study identified a flucarbazone-sodium-resistant population that has evolved 5.9-fold resistance relative to S population. For R population, no mutation at amino acid positions known to confer resistance to flucarbazone-sodium was detected in the target ALS gene coding sequence. There was no apparent difference in expression between the R and S population at 24 h after the application of flucarbazone-sodium. By combining the results of target gene sequencing and gene expression measurements, the presence of TSR was excluded.

Malathion and NBD-Cl are universal indicators of P450-mediated and GST-mediated metabolic resistance in a variety of weeds such as Echinochloa crus-galli L., Amaranthus palmeri and Eriochloa villosa [24–26]. When applied in combination with flucarbazone-sodium, the malathion reversed the observed tolerance to this herbicide of the R population, suggesting that the NTSR may be involved. In addition, metabolic resistance was found to exist in the S population, which is consistent with a prior study. Mechelle et al. [27]investigated resistance of Bromus rigidus populations to ALS-inhibiting herbicides, and the LD₅₀ of the sensitive population declined more than threefold after the plus of malathion. This result may be related to the prolonged and heavy use of herbicides, leading to the evolution of metabolic resistance in sensitive population as well.

Enhanced herbicide metabolism in weeds has been the most extensively studied aspect of the NTSR mechanism. To fully understand the resistance mechanism, NTSR was investigated by using transcriptome sequencing combined with a reference genome. Therefore, we investigated whether elevated levels of expression of resistance-related genes are observed after application. It was found that two genes associated with flucarbazone-sodium resistance in the R population: The Aldo/keto reductase family (AKRs) and the CYP92A6 annotation. Cyt-P450 monooxygenases add a functional group to the herbicide molecule by oxidation, reduction or hydrolysis in the phase I of plant metabolism [12]. More than 30 P450 genes have been shown to metabolize herbicides, with 11 of them capable of metabolizing ALS-inhibiting herbicides [28]. Additionally, we discovered that AKRs might play a role in the metabolism of herbicides. Previous research has demonstrated that by detoxifying glyphosate and mesosulfuron-methyl, the AKRs family enables plants become more resistant to herbicides [29]. Further investigation into transgenic validations is necessary to validate the participation of two genes in the resistance mechanism.

Conclusions

In this study, A. fatua was shown to have developed resistance to flucarbazone-sodium in China. The known mutation sites and ALS gene relative expression that confer resistance to ALS inhibitor herbicides were not found in resistant population. In the study, the absorption and metabolic rates of R and S populations are basically similar, with no significant difference in different time periods. CYP92A6 and the Aldo/keto reductase family were highly expressed in the R population after the application of flucarbazone-sodium. The results will serve as a data resource for further studies on the molecular mechanisms of A. fatua to flucarbazone-sodium. Further studies should be conducted to validate gene function to confirm herbicide function involved in metabolism.

Materials and methods

Plant materials

The seeds of suspected resistant (R) population were harvested in 2019 from wheat fields in Shangqiu city, Henan Province (115.05°E; 34.24°N). The seeds of a susceptible population (S) were collected in 2019 from wheat field in Zhoukou city, Henan Province (114.15°E; 33.35°N). The collected seeds were placed in plastic bags, then transferred to the laboratory, air-dried, threshed, placed in paper bags and stored at 4 °C for subsequent experiments.

The seeds were sowed directly in 8 cm diameter pots containing drenched loam soil. After emergence, the plants were thinned out to 6 seedings per pot and then cultured in artificial climate chamber at 30/25℃, light/dark, 16/8 h at the Institute of Plant Protection, Chinese Academy of Agricultural Sciences, with regular watering and fertilization.

Effect of malathion and NBD-Cl to flucarbazone-sodium resistance

At the 2–3 leaf stage, flucarbazone-sodium (70% water-dispersible granules, Arysta LifeScience) was applied to S population at doses of 0, 3.94, 7.88, 15.75, 31.5, 63 and 126 g ai ha⁻¹ and to R population at doses of 0, 15.75, 31.5, 63, 126, 252 and 504 g ai ha⁻¹ using a flat fan nozzle supplying 450 L ha⁻¹ water at a pressure of 220 kPa. The recommended field dose of flucarbazone-sodium is 31.5 g ai ha⁻¹. To verify the presence of metabolic resistance, seedlings were treated with cytochrome P450 inhibitors malathion (Hi-Yield Chemical, Bonham, TX) at 1000 g ai ha⁻¹ and the GST inhibitor 4-chloro-7-nitrobenzofurazan (NBD-Cl; Sigma-Aldrich) at 270 g ai ha⁻¹, either alone or in combination with flucarbazone-sodium. Malathion and NBD-Cl was applied to seedlings 1 h and 48 h before flucarbazone-sodium treatment. After 21 days of treatment (DAT), the aboveground shoots were collected and weighed. The experiment was repeated twice with three replications.

ALS gene amplification and sequencing

Approximately 100 mg young leaf tissues were harvested from each A. fatua plant at the 2- to 3-leaf stage for genomic DNA extraction by EasyPure Plant Genomic DNA Kit (TransGen Biotech, Beijing, China). The ALS gene sequence was amplified with three primer pairs previously reported [20], which included nine possible mutation positions. The PCR products from the ALS gene were sequenced by Biomed Biotech (Beijing, China) after gel purification. The resulting sequence was compared with the ALS gene sequence of A. fatua from GenBank (NCBI) (accession number: JN175309) using Vector NTI 12.5 (SigmaPlot Software Inc., Chicago, IL).

ALS gene relative expression assay

Forty cDNA samples from two A. fatua populations were extracted and used to the quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) at 0 (control), 1, 3, 5, and 7 days after treatment (DAT) following treatment with flucarbazone-sodium (31.5 g a.i. ha⁻¹). The difference in ALS gene expression between treatment groups and control groups after application may indicate the variation in ALS gene expression between A. fatua populations. Based on previous study [30, 31], the Actin gene was used as an internal control (reference gene) to normalize expression data. Because the ALS gene maybe exhibits copy number variation (CNV) in A. fatua, primers were designed based on the conserved sequences of different ALS copies. The amplification efficiency of the primers (F: 5ʹ-TTGATGACCGTGTGACTGGG-3ʹ; R: 5ʹ-CTGCTGTTCCAACTCCTCGT-3ʹ) used for ALS gene verification was 102%, with a product fragment of 221 bp. The fold-change in gene expression was calculated using the 2 ^−ΔΔCT method [32]. The experimental design included four biological replications and three technical replications for each treatment. The data were subjected to analysis of variance (ANOVA), using the general linear model procedure, in SPSS v19.0 (SPSSInc., Chicago, IL, USA).

RNA isolation, cDNA library construction, and sequencing

The confirmed R and S A. fatua populations were grown and herbicide-treated as previously described. The aerial parts from treated and control plants were collected 24 h after treatment and the samples (three biological replicates, two treatments, and two populations) were immediately cryopreserved in liquid nitrogen. The samples from A. fatua seedlings control groups were identified as RC, SC, while the treatment groups after 24 h of flucarbazone-sodium spraying were marked RT, ST. Each sample was subjected to the extraction of total RNA, which was then analyzed for quantity before being used to build cDNA libraries. The cDNA library construction and the Illumina sequencing were both performed by The Majorbio Corporation. The data were analyzed on the online tool of Majorbio Cloud Platform (https://cloud.majorbio.com/page/tools/).

Differential expression analysis and validation of candidate resistance genes

Eight metabolism-related genes were chosen for validation based on the transcriptome data; the criterion for gene selection was greater than two-fold overexpression of the R population in comparison to the S population, with a statistically significant difference (P ≤ 0.05).

According to transcriptomics predictions of CDS sequences, unique primers for each candidate gene have been designed for qRT-PCR, and a preliminary check on cDNA samples from several A. fatua populations was carried out to evaluate the specificity of each primer pair. Dilutions of 1, 1:3, 1:9, 1:27 and 1:81 were performed on cDNA to establish standard curves and assess primer amplification efficiency accordingly. Primer information of candidate genes was listed in Table 4.

Table 4.

Primers used for qRT-PCR

Gene ID	Function annotation	Forward premier (5’-3’)	Reverse premier (5’-3’)
Pepsico1_Contig31308.path4	ABCA2	GAACAAGCTGCAAGCAACGA	TTCTGCATTTGCAGCAACCC
Pepsico1_Contig37457.path1	ABCC13	AAGTGGGCTTTGTACTCGCA	TGCTGCAAGAAGTGGGAGAG
Pepsico1_Contig15890.path1	ABCF1	TGACGTGGGTTTCAGCTACC	GCTTTTGGCTCCTTCTTGCC
Pepsico1_Contig30663.path3	ABC transporter C family MRP4	CCTAGTCCTGGTGCTTAGCG	GCTGCATGAACATGGACGAC
Pepsico1_Contig9175.path1	CYP71E1	GCAGCTTCCTCACTGGTTCT	GCCAGGGGGCTGCTTATTAT
Pepsico1_Contig10080.path3	CYP92A6	CTCTTCATCCTTAGCGGCGT	AACCGGTCGAACATCTTGCT
Pepsico1_Contig7137.path2	Aldo/keto reductase family	CAAGGTCTCGTGAAGGCTGT	TAGTTCACCTGGTTGACGGC
Pepsico2_Contig4209.path1	ABCC1	TGTTCCTGCCCTTCTTCGTC	GTCATGACACCCTGAACCGT

qPCR was carried out in three technical replicates using PerfectStart® Green qPCR SuperMix in an ABI 7500 instrument (Applied Biosystems, USA) on 96-well plates PCR-96M2-HS-C® (Axygen, Corning, NY, USA) with a sealer Axygen UltraClear Sealing Film (Axygen, Corning, NY, USA). The reactions were performed in a final volume of 20 μL consisting of 1 μL of cDNA, 10 μL of 2xPerfectstart® Green qPCR SuperMix, 0.4 μL of each of the forward and reverse primers (10 μM) and 8.2 μL of Nuclease-free Water. The following program was used: 30 s for 94 °C, 40 cycles of 94 °C for 5 s, and 60 °C for 34 s. The relative expression was calculated on the mean Ct values using the ∆Ct method [32].

Statistical analysis

The data were integrated and analyzed with SigmaPlot v.12.0 (Systat Software, Inc., San Jose, USA). The herbicide dose that induced a 50% reduction in fresh weight (GR₅₀) was determined using a four-parameter nonlinear logistic regression model as follows [33]:

$$\begin{array}{c}\text{Y = C +}\frac{\text{D - C}}{\left[1+{\left(\frac{X}{GR50}\right)}^b\right]}\\ \\ \end{array}$$

where C is the lower response limit, D is the upper response limit, X is the herbicide application rate, and b is the GR₅₀ curve slope. The GR₅₀ of the resistant biotype was divided by the GR₅₀ of the susceptible biotype to calculate the resistance index (RI).

Authors’ contributions

This study was designed by ZFH and XJL. YS, SAH and YNL wrote the main manuscript text. YS and RLW analyzed the data. The collection of plant materials was conducted by SHW and HHJ. HLC and ZFH provided helpful suggestion in data analysis and manuscript revision. All authors reviewed the manuscript.

Funding

National Key Research and Development Program of China, grant number 2023YFD1400501.

Data availability

The datasets supporting the conclusions of this article are included within the article and its additional files. The raw Illumina sequence reads have been deposited in the NCBI Sequence Read Archive (SRA) database with accession number PRJNA1148113 for S and PRJNA1148077 for R.

Declarations

Ethics approval and consent to participate

This research did not involve any human subjects, human material, or human data. A. fatua in current research did not belong to the endangered or protected species.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.