Resistance Mechanisms to Glyphosate in Lamarckia aurea (L.) Moench Found in Southern Spain

Abstract

Glyphosate has been used for roadside weed control in southern Spain for over 40 years, and most populations of goldentop (Lamarckia aurea L.) Moench have putatively developed resistance to this active ingredient. The physiological and biochemical basis for glyphosate resistance in this weed has been investigated. Dose–response studies indicated that the resistant biotype (R) was almost 13 times more resistant to glyphosate compared to a known susceptible biotype (S). Studies of foliar glyphosate retention and ¹⁴C-glyphosate uptake/translocation showed no significant differences between both L. aurea biotypes. Basal 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) activity (µmol µg⁻¹TSP min⁻¹) showed similar values between R (0.82 ± 0.04) and S (0.75 ± 0.05) biotypes. On the other hand, the resistance factor (I₅₀R/I₅₀S) did not show a difference between the two biotypes. Therefore, it was concluded that target-site (TSR) resistance mechanisms are not involved in glyphosate resistance in this weed species. The metabolism of glyphosate to form the non-toxic metabolites aminomethylphosphonic acid (AMPA), glyoxylate, and sarcosine was greater and faster in the R compared to the S biotype; thus, glyphosate resistance is due to non-target-site resistance (NTSR) mechanisms. This paper is the first report of glyphosate resistance in L. aurea in the world.

1. Introduction

Glyphosate is a foliar, systemic, and broad-spectrum herbicide. The mode of action for this herbicide is the inhibition of the 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS, EC 2.5.1.19) enzyme, leading to the accumulation of shikimate [1]. Furthermore, tryptophan, phenylalanine, and tyrosine, which are aromatic amino acids, are not synthesized due to EPSPS inhibition, a process that will be lethal for plants exposed to glyphosate one week after treatment [2]. Resistance to glyphosate can be due to target-site resistance (TSR) and non-target-site resistance (NTSR) mechanisms [3,4]. Non-target-site resistance mechanisms include reduced absorption and translocation, sequestration in vacuoles, and metabolization of the herbicide to compounds that pose no toxicity towards the plant [5,6,7]. TSR mechanisms are due to the overexpression of the target protein or structural changes in the herbicide-binding site [4,8]. To date, 273 species (156 dicots and 117 monocots) have been described as resistant to herbicides, of which 60 cases belong to glyphosate [9]. Some recent cases of glyphosate resistance include barley grass (Hordeum glaucum Steud.) [10], goosegrass (Eleusine indica (L.) Gaertn.) [11], Asia Minor bluegrass (Polypogon fugax Nees ex Steud.) [12], and annual bluegrass (Poa Annua L.) [13]. These are some of the most recently confirmed cases in monocotyledonous weeds, and they illustrate how glyphosate resistance is spreading across a wide range of disturbed habitats, from roadside verge and fence lines (H. glaucum) to intensively managed cropping systems (E. indica, P. fugax, and P. annua), underlining the increasing risk of glyphosate resistance in grasses worldwide.

Goldentop (Lamarckia aurea L.) is unmistakable grass, because it has an open, but unilateral, inflorescence with many long, fragile, tremulous spikelets and a somewhat shiny golden color. It is a very common shrub measuring 5–20 cm, characterized by its flat-bladed leaves with a prominent midrib; an upper leaf sheath that is somewhat inflated; a unilateral panicle, which is golden yellow, with numerous branches holding fascicles of spikelets, composed of 3–5 sterile spikelets surrounding a fertile one; subequal glumes; and awned lemma. It reproduces by seed. It is found in cultivated fields, roadsides, disturbed areas, and calcicolous xerophytic meadows. It is distributed throughout the Mediterranean region.

For more than two decades, two methods of weed control along major land routes have been used: The first option is mechanical control in areas where heavy machinery can enter and/or manual control with small machinery. The second option is chemical control, currently the most widely used, with systemic herbicides, which have a rapid response and longer-lasting control than the first [14]. The latter carries the risk of selecting herbicide-resistant weeds if a rotation system of molecules with different modes of action (MOA) is not implemented [15]. In 2022, our laboratory carried out a survey on the A-4 highway from Córdoba to Montoro (southern Spain) in areas where we knew that glyphosate had been used for at least 10 or 15 consecutive years (two applications/year). In Montoro, we observed that the grass L. aurea escaped the action of glyphosate, while it had good control over other monocot- and dicotyledonous weeds. Therefore, the present paper was carried out to (a) determine the susceptibility level of two (R and S) L. aura biotypes to glyphosate in non-cultivated areas of southern Spain and (b) study the mechanisms of resistance endowing glyphosate resistance to this species.

2. Materials and Methods

2.1. Chemicals

Two forms of glyphosate were used in this study: glyphosate (¹⁴C-glyphosate (glycine-2-¹⁴C), 95% radiochemical purity, 273 Mbq mmol⁻¹ specific activity, Institute of Isotopes Co., Ltd., Budapest, Hungary) and commercially formulated glyphosate (Roundup Ultimate^® SL, 480 g ae L⁻¹ as isopropyl amine salt, Bayer CropScience, Pontevedra, Spain).

2.2. Plant Materials

Lamarckia aurea seeds suspected to be resistant to glyphosate (R) were collected on the sides of the highway (A-4) in the province of Córdoba Montoro (38°01′18″ N 4°22′58″ W), where the herbicide glyphosate had been the only control method for 10–15 years. Susceptible (S) biotype seeds were collected from a livestock lot without the use of herbicides at the University of Córdoba (Spain). The seeds were sown in containing peat. The trays were maintained at field capacity (FC) and topped by a layer of parafilm. The trays were then placed in a growth chamber for incubation. The growth chamber was set at a temperature of 28/18 °C (day/night) with a 12 h photoperiod, a light intensity of 350 μmol m⁻² s⁻¹, and a relative humidity of 60%. After the seedlings developed two or three true leaves, they were transplanted into plastic pots (10 cm × 10 cm × 6.3 cm) (one plant per pot) and filled with soil/peat (4:1 v/v). The pots were placed in a greenhouse with controlled conditions during the experiments (12 h photoperiod, a day/night temperature of 28/20 °C, and relative humidity of 60–70%, with constant water (field capacity)). Seeds were collected in October 2023, and all greenhouse and laboratory experiments described in this study were carried out between December 2023 and July 2024.

2.3. Dose–Response Curves

To study the susceptibility level of L. aurea biotypes, dose–response assays with glyphosate were carried out. At the BBCH 13-14 growth stage, the biotypes were exposed to glyphosate using a laboratory sprayer system (SBS-060 De Vries Manufacturing, Hollandale, MN, USA) equipped with 8002 flat-fan nozzles delivering 250 L ha⁻¹ at a 50 cm height above the plants. The doses of glyphosate which were applied (10 plants per dose) were as follows: 0, 25, 50, 100, 250, 500, 750, 1000, and 1500 g ae ha⁻¹. After 28 days, the plants were harvested to estimate the dose that causes a 50% reduction in growth (GR₅₀). Data is expressed as percentages of the control (untreated).

2.4. Glyphosate Retention

Five plants of R and S biotypes of L. aurea were assayed in a completely randomized design following the methods developed by Vazquez-Garcia et al. [16]. The plants were exposed to glyphosate at 360 g ae ha⁻¹ plus 100 mg L⁻¹ of Na-fluorescein at the BBCH 13-14 stage in the spraying system described in the dose–response assays. Afterwards, the methodology elucidated by Yanniccari et al. [17] was followed. The absorbance of fluorescein was measured by a spectrofluorometer (F-2500, Hitachi, Tokyo, Japan) at a 490/510 nm wavelength. The assay was carried out with two experimental runs, and the pooled results (µL of sprayed solution retained per g dry weight) were analyzed.

2.5. ¹⁴C-Glyphosate Uptake and Translocation

Five plants at three- to four-leaf stages of both R and S L. aurea biotypes were used for ¹⁴C-glyphosate absorption and translocation assays. The experiments were arranged in a completely randomized design.

Two mixes were used in this experiment. The first (mix A) was glyphosate (¹⁴C-glyphosate) plus commercially formulated glyphosate (Roundup Ultimate 48%), which was prepared with a final specific activity of 100,000 dpm μL⁻¹ and 360 g ae ha⁻¹. Then, the second leaf of each plant was exposed to one drop (1 μL) of mix A based the method described by Vázquez-García, et al. [18].

The second solution (mix B) was prepared with distilled water and acetone at a 1:1 (v/v) ratio for the purpose of washing the non-absorbed ¹⁴C-glyphosate, which was performed 24, 48, and 96 h after treatment (HAT). This procedure consisted of three replications for each sample (1 mL of mix B for each replication). The leaves of five plants treated from each biotype were washed at each evaluation time. The solution was then recovered in scintillation vials, and 2 mL of a scintillation cocktail was added and analyzed by a scintillation counter (LS 6500, Beckman Coulter, Brea, CA, USA). ¹⁴C-glyphosate uptake was calculated as a percentage of recovered radioactivity using Equation (1) [19]:

$$\begin{array}{l}\% of 14C-herbicide absorbed\\ =\left({\displaystyle \frac{14C in combusted tissue}{\left(14C in combusted tissue + 14C in shoot washes\right)}}\right)\end{array}$$

(1)

Prewashed plants that were utilized to determine absorption were extracted from the pots. Then, the roots were washed promptly so as to rinse off the substrate. The plants were separated into treated leaves, shoots, and roots. Thereafter, the tissues were placed in cellulose cones and dried in a stove for four days at 60 °C. The samples were then subjected to combustion using a sample oxidizer (Packard 307) for 180 s. The produced ¹⁴CO₂ was measured in Carbosob/Permafluor E+ (3/7 v/v) (Packard Instruments Co., Downers Grove, IL, USA). The scintillation counter apparatus was used to quantify the radioactivity of the plant tissues. Translocation was measured based on the method suggested by Vázquez-García et al. [20].

2.6. Glyphosate Metabolism

The 1st procedure in this assay was carried out under the conditions explained in the section related to the dose–response assays. Glyphosate at 360 g ae ha⁻¹ was sprayed over five plants of both R and S L. aurea biotypes. The plants that had not been treated with glyphosate were regarded as the control. Subsequently, the plants were cut from the soil surface and rinsed with distilled water to eliminate surplus glyphosate from leaf surfaces and substrate residues from roots. The plants were then subjected to liquid N₂ following drying by a paper towel for 180 s. Both treated and untreated samples were kept at a temperature of −80 °C until the beginning of metabolite analysis [21]. The extraction of glyoxylate, Amino methyl phosphonic acid (AMPA), and sarcosine were performed based on the method suggested by Rojano-Delgado et al. [22,23]. Metabolite standards were used to generate calibration curves (Sigma-Aldrich, St. Louis, MI, USA). Data is expressed as percentages of the sum of glyphosate plus metabolites recovered. Data was arranged in a completely randomized design with five replications (one replica–one plant).

2.7. EPSPS Activity

The activity level of EPSPS was measured in the four L. aurea biotypes. Five g of fresh foliar tissue (3 to four leaves) were collected, frozen in liquid nitrogen, and then grounded in a mortar to produce a fine powder [8]. The grounded samples were then kept at a temperature of −40 °C until analyses. The method suggested by Dayan et al. [24] and Bracamonte et al. [25] was used to extract the EPSPS enzyme; determine the total soluble protein (TSP, basal activity without glyphosate); and estimate the EPSPS inhibition rate by applying glyphosate at 0 (control), 0.25, 0.5, 1, 2.5, and 5 μM. The results are expressed as a percentage of the amount (μmol) of inorganic phosphate (Pi) released per μg of TSP min⁻¹ (μmol Pi μg⁻¹ TSP min⁻¹) relative to the control. The experiment was conducted in a completely randomized design (CRD) with five replications.

2.8. Statistical Analysis

A three-parameter log-logistic function, Equation (2) was fitted to the dose–response data using the R software (drc package) [26] to estimate the glyphosate rate required for a 50% reduction in plant growth or EPSPS activity inhibition for each L. aurea biotype (GR₅₀ and I₅₀, respectively).

$$\mathrm{y}={\displaystyle \frac{\mathrm{d}}{\left\{1+\exp \left[\mathrm{b}\left(\log \mathrm{x}-\log \mathrm{e}\right)\right]\right\}}}$$

(2)

in which y is the percentage of growth or enzyme inhibition, d is the upper limit, e is GR₅₀ or I₅₀, and b is the slope at e, respectively. A resistance factor (RF) was calculated for R and S biotypes (GR₅₀R/GR₅₀S or I₅₀R/I₅₀S).

For the rest of the assays, analysis of variance was carried out following the establishment of the pre-requisite conditions for the analysis (e.g., homogeneity of variances and normal distribution of error). Comparison of means was performed using the least significant difference (LSD) test in the SAS v.9.0 software.

3. Results

3.1. Dose–Response Curves

The glyphosate rate required to reduce the fresh weight (GR₅₀) by 50% of the S and R L. aurea biotypes was 40.69 ± 1.38 and 528.21 ± 9.88 g ae ha⁻¹, respectively, with an RF of 12.98 ± 0.50 for these parameters. Also, the dose required for a 90% reduction in the fresh weight (GR₉₀) was 970.34 ± 27.84 and 139.77 ± 8.63 g ae ha⁻¹ for R and S biotypes, respectively (Figure 1).

3.2. Spray Retention

The amount of glyphosate solution retained on the leaf did not differ between R and S L. aurea biotypes. The spray retention values for S and R biotypes were 747.8 ± 95.6 and 712.6 ± 98.8 µL of glyphosate retained per g dry weight, respectively.

3.3. EPSPS Activity

Basal EPSPS activity (µmol µg⁻¹TSP min⁻¹) showed similar values between L. aurea R (0.82 ± 0.04) and S (0.75 ± 0.05) biotypes. On the other hand, the I₅₀ for R (0.61 ± 0.03 µM) and S (0.58 ± 0.03 µM) did not show differences between the resistance factor of the two biotypes (RF = I₅₀R/I₅₀S = 1.05 ± 0.07) (Figure 2).

3.4. ¹⁴C-Glyphosate Uptake and Translocation

¹⁴C-glyphosate uptake and translocation did not show differences between S and R plants of L. aurea at all the time points studied, and only the simple effect of time was significant at p < 0.05. The maximum uptake of glyphosate was observed at 96 HAT. Twenty-four hours after treatment, 66.2% of the applied glyphosate had remained in the leaves, which was reduced to 35.7% at 96 HAT. The remaining glyphosate in the shoots and roots was 20.6 and 13.2% at 24 HAT, which went up by 9.3 and 21.1% at 96 HAT, respectively (Figure 3).

3.5. Glyphosate Metabolism

The metabolites aminomethylphosphonic acid (AMPA), glyoxylate, and sarcosine were produced, as shown in Figure 4. Glyphosate metabolism in the R biotype of L. aurea was very rapid and efficient; by 24 HAT, the plants had metabolized about 70% of the glyphosate into non-toxic metabolites (58% to AMPA, 11% to glyoxylate, and 0% to sarcosine). After 48 HAT, glyphosate metabolism in the resistant biotype reached about 82%, with AMPA, glyoxylate, and sarcosine shares of 50%, 21%, and 11%, respectively. The resistant biotype had almost completely eliminated the toxic product 96 HAT, while the S biotype only metabolized about 6%, 10%, and 20% of glyphosate (mostly into glyoxylate) at 24, 48, and 96 HAT, respectively (Figure 5).

4. Discussion

Compared with the S biotype, the glyphosate rate required for a 50% reduction in L. aurea growth was about 13 times higher in the R biotype. The RF value in dose–response studies may vary due to various factors, including the species and the mechanism responsible for herbicide resistance [27]. Also, the susceptible biotypes that were used in this study may have different degrees of susceptibility to herbicides, which can influence the resistance factor [28].

Glyphosate retention was statistically similar between the R and S biotypes of L. aurea. The herbicide amount that enters the plant during foliar application depends on the degree of spray retention, which underlines the significance of this attribute in the efficacy of herbicides [29]. Some of the factors that alter spray retention are as follows: (1) Orientation of leaves: erect and upright leaves retain less herbicide in comparison with horizontal leaves [30]; (2) Thickness of cuticles: the thicker the cuticle, the lower the herbicide retention [31]; and (3) Waxes: The high content of these structures can lead to decreased herbicide retention due to their hydrophobic nature [32]. In accordance with the results of the present study, spray retention was also not responsible for glyphosate resistance in horseweed (Conyza Canadensis L. Cronq) in the United States [33] and red brome (Bromus rubens L.) in southern Spain [34].

Similar values of the EPSPS gene activity were recorded in the R and S L. aurea biotypes. Altered EPSPS activity indicates target-site resistance to glyphosate, which includes single nucleotide polymorphism in the EPSPS gene, an increased gene copy number, and EPSPS gene overexpression [18]. The results of the present study indicate that TSR mechanisms are not involved in glyphosate resistance in L. aurea, and there is no need to conduct EPSPS gene sequencing, as it is clearly ruled out [19,35]. While TSR has been reported to contribute to glyphosate resistance in goosegrass from China [11], horseweed from Mexico [36], and plumeless thistle (Carduus acanthoides L.) from Argentina [37], no evidence of TSR has been observed in glyphosate-resistant Asia Minor bluegrass [12], prickly Russian thistle (Salsola tragus L.) [38], and Johnsongrass (Sorghum halepense (L.) Pers.) [18].

Resistant and susceptible biotypes of L. aurea showed similar glyphosate uptake and translocation. Uptake and translocation are the main NTSR mechanisms in glyphosate resistance [39]. Sequestration of glyphosate in vacuoles of the plant and thus failure of the herbicide to reach the chloroplast is reported to be responsible for reduced translocation of this herbicide [40]. It is suggested that ABC transporter proteins are involved during this process in glyphosate-resistant horseweed biotypes [3]. Glyphosate resistance due to translocation has also been reported in Johnsongrass [18] and Palmer amaranth (Amaranthus palmeri S. Watson) [19]. Conversely, the results indicate that neither uptake nor translocation is responsible for the resistance of the R biotype to glyphosate, indicating that changing herbicide formulations and adjuvant use might not help to overcome resistance [41].

A significant difference was observed between the R and S biotypes in terms of glyphosate metabolism. Glyphosate is metabolized into aminomethylphosphonic acid (AMPA) and glyoxylate mediated by a glyphosate oxidoreductase (GOX) and sarcosine via a C-P lyase [18]. Metabolic herbicide resistance poses a serious threat to the sustainable production of agricultural products due to its difficult-to-predict nature [42]. Weeds endowed with metabolic herbicide resistance have the potential to become resistant to various herbicides with different modes of action. Moreover, they may even resist herbicides that are yet to be introduced to the market [43]. Metabolic resistance to glyphosate has also been reported in Asia Minor bluegrass [12] and rigid ryegrass (Lolium rigidum Gaud.) [44]. Many weed species can become resistant to glyphosate in non-agricultural areas and enter cropping systems carrying resistance mechanisms to share with other weed populations [45].

Although alternative control options have not been experimentally evaluated in this work, research and guidelines for non-agricultural areas offer practical solutions for the management of glyphosate-resistant weeds. In some regions of the world, integrated management plans have been developed for these areas, such as rotating tank mixes of herbicides on bare soil and combining early-season mowing to avoid seed rain while promoting reseeding of native grasses [46,47]. From a scientific point of view, diversification of modes of action and cultural tactics are essential to delay the emergence of resistance, and always acting preventively is economically preferable to waiting until resistance has taken hold [28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48]. Therefore, herbicide rotation, residual pre-emergence treatments, selective mowing, and rapid revegetation should help to contain glyphosate-resistant L. aurea on Mediterranean roadsides.

5. Conclusions

This is the first study on herbicide resistance in L. aurea in the world. The resistance factor in the resistant biotype of L. aurea was 13 times greater than that of the susceptible biotype. No evidence of target-site resistance was observed in the studied plant. Also, non-target-site resistance mechanisms of herbicide retention, uptake, and translocation did not endow resistance to L. aurea. Glyphosate metabolism was the only mechanism responsible for glyphosate resistance in the studied populations. Also, some morphological differences were observed between the R and S biotypes of L. aurea during this study. This infers a possible fitness cost imposed by glyphosate resistance in this species, and further studies are required to investigate this issue. In addition, investigating alternative chemical and non-chemical methods—such as alleleopathy studies—to mitigate glyphosate resistance in L. aurea is suggested.