Allelopathic Potential and Cytotoxic, Genotoxic, and Antigenotoxic Effects of Tecoma stans Flowers (Bignoniaceae)

Abstract

Weed control is mainly carried out using synthetic herbicides, which represent 62.6% of the total pesticides sold. However, some plants produce allelochemicals that inhibit the growth of other plants, and these substances can be isolated and used as natural herbicides. This study aimed to evaluate the allelopathic, cytotoxic, genotoxic, and antigenotoxic potential of the ethanol extract (EE), hexane (HEX), dichloromethane (DCM), ethyl acetate (EA) and hydroethanol (HE) fractions obtained from Tecoma stans flowers. Nuclear magnetic resonance (NMR) was used to characterize the compounds present in the samples. The allelopathic activity was tested using Allium cepa and Lactuca sativa seeds, and the cytotoxicity, genotoxicity and antigenotoxicity were evaluated using A. cepa seeds. The saturated and unsaturated fatty acids ω-3 and ω-6, terpenes, flavonoids, and phenolic acids with coumaroyl or glycosyl derivatives were characterized in the samples. The HEX and DCM fractions significantly inhibited germination and root growth, effects associated with fatty acids and phenolic compounds. The EA fraction exhibits genotoxic potential at higher concentrations tested. The extract and fractions reduced the genotoxicity induced by glyphosate and atrazine, reversing chromosomal abnormalities. These results demonstrate the possible use of the extract and fractions as natural sources of allelochemicals, but safe dosage validation is required.

1. Introduction

Since 2008, Brazil has been the world’s largest pesticide consumer. Herbicides used predominantly for the control of invasive plant species account for 62.6% of total pesticide sales [1,2]. The widespread use of these substances produces severe environmental consequences, including the contamination of soil, groundwater, and surface water and disruptions to ecosystems and food chains. Moreover, several pesticides have been linked to serious health risks, affecting agricultural workers and consumers, which ingest pesticide residues through food [3,4]. The continued use of the same herbicide or products with the same mechanism of action favors the emergence of resistant weeds, which has become a major challenge for farmers, scientists, and agribusiness [5].

Glyphosate, an herbicide commonly used in the tomatoes, potatoes, lettuce, apples, strawberries, grapes, bananas, and oranges cultures [6], has been widely used since its introduction in the 1970s, and its use has increased approximately one hundredfold compared to its initial period. Glyphosate is a non-selective herbicide that acts on the shikimic acid pathway. It inhibits the action of the enzyme 5-enolpyruvylshikimate-3-phosphate synthase, preventing its binding to phosphoenolpyruvate and, consequently, preventing or strongly reducing the synthesis of the essential amino acids’ tryptophan, tyrosine, and phenylalanine [7]. Van Bruggen et al. [8] reported a correlation between the increase in glyphosate use and several human diseases, including kidney damage, cancer, and neurological and emotional disorders, such as autism, depression, and Alzheimer’s. Atrazine is another herbicide widely used in Brazil, primarily for weed control in sugarcane and corn cultures. It is a triazine derivative that acts as an herbicide by interfering with photosystem II, binding to the D1 subunit of plastoquinone, culminating in carbon fixation reduction [9]. However, according to the European Union and the U.S. Environmental Protection Agency, atrazine has significant risks to human health, with neurological and reproductive effects due prolonged ingestion [10]. It is considered a probable carcinogen and has been associated with an increase in cases of bladder and lung cancer, and multiple myeloma in rural workers, due to the ingestion of contaminated water [11,12]. Despite these risks, this herbicide was among the best-selling in Brazil between 2009 and 2019 [13], reinforcing the need for its replacement with less toxic alternatives.

The use of synthetic herbicides causes several implications for human health, including DNA damage, cellular alterations associated with various types of cancer in the digestive, urinary, and respiratory systems, and, in severe cases, death resulting from herbicide poisoning [3,4,5,6,7,8,9,10,11,12,13,14]. However, some plants produce allelochemicals, which are substances that inhibit the growth of other plant species. These compounds can be isolated and utilized as natural herbicides, offering a sustainable strategy for weed management in agricultural production systems [12,15]. The allelopathic mechanisms of plants are associated with secondary compounds, such as alkaloids, coumarins, essential oils, flavonoids, terpenoids, steroids, quinones, tannins, and phenols [15,16], which are released by volatilization, leaching, decomposition, or root exudation [16]. In this interaction, one plant acts as a donor of allelochemicals and the other as a recipient, being affected by exposure to these substances [17]. Allelochemicals present structural and biological diversity, exhibit bioactivity with lower toxicity, and action mechanisms distinct from synthetic herbicides, making them promising natural compounds for the development of new products [15,18,19].

Natural products have unknown adverse effects, and therefore, it is important to assess the toxicity of plant extracts and their fractions to ensure their safe use in crops and for human and animal health [20]. Genotoxicity assays are essential to identify DNA damage and to determine the safety of plant extracts. Allium cepa (monocotyledonous) and Lactuca sativa (eudicotyledonous) are widely used in these tests. A. cepa is the most used due to its sensitivity to genotoxic agents, with large and visible chromosomes, which facilitate the detection of chromosomal aberrations; in addition to high cell division rates, allowing these effects to be observed in a short period of time [21,22]. A. cepa test allows to assess the cytotoxicity by the changes in root growth rate, cell viability, mitotic index, and genotoxicity by the formation or increase in chromosomal aberrations in root tip cells [21]. This assay has also identified compounds with antigenotoxic, antimutagenic and/or anticarcinogenic potential [23]. Antigenotoxicity refers to the action of natural or synthetic agents that prevent or reverse genetic damage, acting through different mechanisms, such as inhibition of genotoxic enzymes, direct DNA repair, or antioxidant activity against reactive oxygen species [24,25]. Phenolic compounds have antioxidant properties and can act alone or in conjunction with other substances. Studies have shown that galangin prevents DNA damage caused by UVB radiation and that catechins protect against oxidative cell damage [26,27].

Tecoma stans, popularly known as “ipê-mirim” or “yellow-sin”, belongs to the Bignoniaceae family and its phytochemicals compounds (monoterpene alkaloids, phytosterols, polyphenols, fatty acids, flavonoids, terpenes) have presented a broad biological potential, such as allelopathic, antidiabetic, antioxidant, antimicrobial, anti-inflammatory, antinociceptive, and larvicide effects. In folk medicine, T. stans leaves and flowers are used to treat diabetes, digestive disorders, fever, and infections [28,29]. Leaves are the most studied, in terms of phytochemical characterization and biological activities, including some reports of the allelopathic potential of extracts obtained from T. stans [29,30,31,32], and toxicity on breast cancer (MCF-7) and mouse embryo fibroblast (MEF) cell lines [29]. However, scientific knowledge about the biological effects of T. stans flowers is still limited, and no studies were found about its allelopathic potential. So, this study aimed to assess the allelopathic activity and the cytotoxic, genotoxic, and antigenotoxic effects of T. stans flowers. The results may contribute to the discovery of new natural compounds with phytotoxic activity, which in the future could promote safer crops for human health and the sustainable use of biodiversity.

2. Materials and Methods

2.1. Chemicals

Ethanol was purchased from Isofar (Duque de Caxias, RJ, Brazil); hexane, dichloromethane, ethyl acetate, hydrochloric acid (HCl), acetic acid, sodium hydroxide (NaOH), and glacial acetic acid were acquired from CRQ—Cromato Produtos Químicos (São Paulo, SP, Brazil). Methanol-d₄, 3-(trimethylsilyl) propionic-2,2,3,3-d₄ acid sodium salt (TSP-d₄) and 2-(N-morpholino)ethanesulfonic acid were purchased from Sigma-Aldrich (Darmstadt, Germany). Glyphosate was obtained from Citromax (São Lourenço do Oeste, SC, Brazil), atrazine from Nortox (Arapongas, PR, Brazil), and orcein from Himedia (Kelton, PA, USA).

2.2. Collection and Preparation of Plant Material

T. stans flowers were collected in Divinópolis, central-western region of Minas Gerais State, Brazil (20°10′44″ S, 44°55′6″ W), in April 2018, according to proof of registration for collection of botanical material carried out with SISBIO n° 30006, on 20 March 2018, Registration at IBAMA: 5282691. Fertile samples were collected, and the vouchers were identified by Andreia Fonseca Silva at the PAMG Herbarium, at the Agricultural Research Company of Minas Gerais (EPAMIG) as Tecoma stans (L.) Juss. ex Kunth (PAMG 58284). This work has authorization to access the genetic heritage with Access Registration No. A6704FE, dated 10 August 2019, granted by SISGEN/CGEN/MMA, by the Brazilian Biodiversity Law (13.123/2015).

Plant material was collected from three shrubs and healthy flowers were selected. The fresh flowers (500 g) were dried in an oven (Edutec FDC-1000 Serials, Sapucaia do Sul, RS, Brazil) at 40 °C for 7 days to stabilize the biomass (150 g). Plant material was crushed in a knife mill (Marconi MA 048, Piracicaba, SP, Brazil). Then, 150 g of the material was placed in 1350 mL of 70% ethanol (1:9), and turbo-extraction (Ultra-Turrax, Marconi MA-102 Plus, Piracicaba, SP, Brazil) was carried out at 3000 rpm. After extraction, the filtrate obtained was taken to a rotary evaporator (IKA HB10 digital, Staufen, Germany) at 40 °C with a rotation of 70 rpm to evaporate the solvent.

Subsequently, 40 g of the ethanol extract (EE) solubilized in 100 mL of 70 °GL ethanol was partitioned with solvents in increasing order of polarity, using a separation funnel. The partition started by adding 50 mL of hexane to obtain the hexane (HEX) fraction, and the process was repeated two more times, totaling 150 mL of hexane. Then, the extraction was repeated with the solvents dichloromethane (DCM) and ethyl acetate (EA), and finally, the hydroethanol (HE) residue was collected.

2.3. Nuclear Magnetic Resonance (NMR) of the Ethanol Extract and Its Fractions

Samples (15 mg) were dissolved in 800 μL of methanol-d₄ containing 0.01% (w/v) TSP-d₄ (3-(trimethylsilyl) propionic-2,2,3,3-d₄ acid sodium salt), vortexed (Vortex mixer, Global Trade Technology XH-C, Jaboticabal, SP, Brazil) for 1 min, and placed in an ultrasound bath (Sanders SoniClean, Santa Rita do Sapucaí, MG, Brazil) for 20 min. After centrifugation at 17,000× g for 15 min, the supernatants (700 μL) were transferred to 5 mm diameter NMR tubes. The ¹H NMR experiments were performed in a Bruker Avance Neo 600 MHz (Fällanden, Switzerland) and the spectra were acquired at 300 K with a spectral window (SW) of 16 ppm, a number of digitized points (TD) of 64 K, with HDO (deuterated water) signal pre-saturation, acquisition number (NS) of 128, acquisition (AQ), and waiting times before each acquisition (d1) of 3.2 s and 5.0 s, respectively. All spectra were obtained using the zgcppr pulse sequence and processed using a 0.3 Hz line broadening before the Fourier transform. The phases and baselines were automatically corrected using the TopSpin 4.4.0 program. Finally, the spectra were calibrated using the TSP-d₄ signal at 0.00 ppm.

The ¹H-¹H COSY (Correlation Spectroscopy) homonuclear correlations were recorded with a standard 90°-90° pulse sequence and field gradient in the z-axis direction. These experiments were acquired with an AQ(F2) of 192 ms, an NS of 4, and a d1 of 2.0 s, under a SW of 10 ppm in both dimensions, with TD(F1) and TD(F2) set to 1024 and 2048, respectively. The Fourier transform was applied, and the transformed data were symmetrized. Two-dimensional J-Resolved NMR spectra were acquired using jresgpprqf pulse program, 16 scans per 128 increments in F1 and 2K for F2 using spectral widths of 13,157.895 Hz in F2 and 120 Hz in F1. A 2.0 s relaxation delay was employed. The J-Resolved spectra were symmetrized and tilted, and then calibrated.

2.4. Analysis of Seed Germination

The tests were carried out with seeds of eudicotyledonous Lactuca sativa (lettuce) var. aurelia (Topseed^®, Lot: 073792, São Paulo, SP, Brazil) and monocotyledonous Allium cepa (onion) cv. pear-shaped bay (Topseed^®, Lot: 073432, São Paulo, SP, Brazil). No sanitizing process was performed on the seeds prior to germination. Germination and growth were conducted in 2-(N-morpholino) ethanesulfonic acid (MES) buffer solution with a pH adjusted to 6.0–6.2 with a NaOH solution. The extract and fractions of T. stans were tested at concentrations of 0, 250, 500, 750, and 1000 µg/mL solubilized in MES buffer solution in quadruplicate, according to Amado et al. [33]. Thus, 25 seeds (n = 2900) were placed on Petri dishes disposable and sterile (90 × 15 mm) containing filter papers and 4 mL of buffer solution containing T. stans samples or 4 mL of the control (negative control: MES; and positive controls: atrazine and glyphosate for A. cepa and atrazine for L. sativa).

The Petri dishes were closed, sealed with film paper, and incubated in a B.O.D. Incubator Germination Chamber (Cienlab, Campinas, SP, Brasil) in the dark at 25 °C for 12 days for A. cepa seeds and 7 days for L. sativa seeds [34]. The number of germinated seeds was counted daily. Based on these data, the germination speed index (GSI) was calculated [35]. In the “first count” parameter, seeds germinated up to the 4th day for L. sativa and up to the 6th day for A. cepa were counted, and seed vigor was observed. The number of seeds germinated was used to determine the “germination potential” [36]. Analyses were performed in quadruplicate, and pure MES solution was evaluated as a negative control.

The first count (seed vigor) was calculated using the following equation:

$$\mathrm{F}\mathrm{i}\mathrm{r}\mathrm{s}\mathrm{t} \mathrm{c}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}={\displaystyle \frac{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} {\mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s}}^{\mathrm{’}} \mathrm{g}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d} \mathrm{u}\mathrm{p} \mathrm{t}\mathrm{o} \mathrm{t}\mathrm{h}\mathrm{e} n \mathrm{d}\mathrm{a}\mathrm{y}}{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s} }}\times 100$$

(1)

where n = 4th for L. sativa, and n = 6th for A. cepa.

The germination potential was calculated using the following equation:

$$\mathrm{G}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{p}\mathrm{o}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}={\displaystyle \frac{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s} \mathrm{g}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}}{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s}}}\times 100$$

(2)

2.5. Seedling Growth Analysis

Once the total germination time for L. sativa and A. cepa was completed, the plates were removed from the germination chamber and cooled to −20 °C for 24 h to interrupt the growth process. Then, the plates were opened, and the length of each root and epicotyl was measured, in millimeters, using a ruler. Subsequently, the effects on the growth of seedlings were calculated using the following formula:

$$\% \mathrm{of} \mathrm{growth} = \left[\right(\mathrm{Ma} - \mathrm{Mc})/\mathrm{Mc}] \times 100$$

(3)

where Ma is the average root or epicotyl length of the seedlings and Mc is the average root or epicotyl length of the seedlings for the negative control (MES). Thus, positive values represent stimulation of the studied parameter, and negative values represent inhibition.

2.6. Cytogenotoxicity Assay

For the cytogenotoxicity test, 30 A. cepa seeds were germinated on Petri dish containing filter paper and 4 mL of distilled water, in a total of 35 Petri dishes. No sanitizing process was performed on the seeds prior to germination. After 12 days of germination in B.O.D. in the dark at 25 °C, the seedlings were removed from the plates containing water, and 10 specimens (n = 870) were transferred to Petri dishes containing 4 mL of the extract, fractions, and controls at concentrations of 250, 500, 750, and 1000 μg/mL [33]. MES buffer solution was used as a negative control, and glyphosate (GLI) and atrazine (ATZ) were used as positive controls. Thirty A. cepa seeds were used in each Petri dish in the germination assay to obtain an adequate number of viable individuals, to allow a consistent evaluation of germination parameters. After seed germination, 10 seedlings were transferred to another Petri dish so that the roots would reach sufficient length to obtain meristematic regions suitable for analysis (typically around 1–2 cm), maintaining the statistical robustness of the results.

After 24 h in the presence of the extract, fractions, or controls, the roots were collected, fixed in 1 mL of Carnoy (3:1, ethanol: glacial acetic acid) in 2-mL microtubes and stored in a freezer (Consul CRM37EBANA, Joinville, SC, Brasil) at −20 °C until slide preparation.

2.7. Antigenotoxicity Assay

The antigenotoxic assay was adapted from methodology proposed by Sousa et al. [37], and a post-treatment exposure protocol was chosen. Initially, 30 seeds of A. cepa were placed to germinate on Petri dish containing filter paper and 4 mL of distilled water (totaling 30 dishes with 30 seeds each). No sanitizing process was performed on the seeds prior to germination. After 12 days of germination in B.O.D. at 25 °C, the seedlings were removed, and 10 specimens were transferred to Petri dishes (66 dishes containing 10 seeds each), containing 4 mL of glyphosate solution (500 μg/mL—based on high AI and low NI observed in the cytogenotoxicity assay). After 24 h in B.O.D., the seedlings were washed with distilled water and treated for 24 h with the extract and fractions at concentrations of 250, 500, 750, and 1000 μg/mL, in triplicate, according to Amado et al. [33]. The roots were collected, fixed in 1 mL of Carnoy (3:1, ethanol:glacial acetic acid) in 2-mL microtubes and stored in a freezer at −20 °C until slide preparation. The antigenotoxic assay was repeated with atrazine at a concentration of 500 μg/mL using the same methodology described above (n = 1260).

2.8. Preparation of Glass Slides

The microtubes were removed from the freezer, and using tweezers, at least 3 root tips were selected and washed with distilled water for 5 min [38]. They were then hydrolyzed with 1 N HCl for 15 min in a water bath (Nova Instruments 1236, Piracicaba, SP, Brazil) at 40 °C. The roots were washed with ice-cold distilled water for 5 min. Cuts were made to collect only the apical meristem of the roots, which was transferred to the slide containing 1 drop of 2% orcein previously diluted in 45% acetic acid (5:3 v:v).

The roots were chopped with surgical blades and crushed. The slides were analyzed under a Primo Star Zeiss trinocular light optical microscope (Oberkochen, Germany) at 40× magnification coupled to a Canon PowerShot A650 IS camera, 12.1 Megapixels, 6× optical zoom (Hong Kong, China). Cell counting was performed using AxionVision software (AxioVs40V 4.8.0.0 version). At least 2100 cells (3 × 700 cells) were analyzed using the scanning technique for each of the 5 concentrations tested: 0, 250, 500, 750, and 1000 μg/mL (n = 60,900 to citogenotoxicity assay and n = 90,300 to antigenotoxiciy assay).

In the antigenotoxicity test using 500 μg/mL glyphosate, the protocol for preparing slides differed due to the sensitivity of the roots to the treatment. Thus, the root tips were washed with distilled water for 5 min, hydrolyzed with 1 N HCl for 3 min in a water bath at 30 °C, and washed again with ice-cold distilled water for 5 min. Subsequently, the previously described staining and crushing techniques were performed.

The parameters analyzed during cell counting were the mitotic (MI), chromosomal aberration (AI), and necrosis indices (NI). The types of chromosomal aberrations evaluated for each phase of the cell division cycle are described in

Table 1. Chromosomal aberrations considered for each mitotic phase found in A. cepa.

Mitotic Phase	Types of Aberrations
Interphase	Presence of heterochromatins, micronuclei, nuclear buds and formation of more than 3 nucleoli
Prophase	Presence of micronuclei, wandering chromosomes or fragments and aneuploidy
Metaphase	C-metaphase, sticky chromosomes, micronuclei, wandering chromosomes, aneuploidies and polyploidies
Anaphase	Anaphase bridges, micronuclei, wandering chromosomes and lagging chromosomes
Telophase	Micronuclei, wandering chromosomes, aneuploidies and heterochromatins

Table organized by the author based on adaptations [39,40,41].

.

The mitotic index was calculated using the following equation [42]:

$$\mathrm{MI} = (\mathrm{m}/\mathrm{T}) \times 100$$

(4)

where m is the number of cells in mitosis, and T is the total number of cells.

The aberration index was obtained using the following equation [42]:

$$\mathrm{AI} = (\mathrm{a}/\mathrm{T}) \times 100$$

(5)

where a is the number of cells with aberrations, and T is the total number of cells.

The necrosis index was calculated using the following equation:

$$\mathrm{NI} = (\mathrm{n}/\mathrm{T}) \times 100$$

(6)

where n is the number of necrotic cells, and T is the total number of cells.

The antigenotoxic effect was obtained using the following equation [43]:

$$\mathrm{Low} \mathrm{AI} (\%) = \left[\right({\mathrm{AI}}_{\mathrm{pc}} - {\mathrm{AI}}_{\mathrm{s}})/({\mathrm{AI}}_{\mathrm{pc}} - {\mathrm{AI}}_{\mathrm{nc}}\left)\right] \times 100$$

(7)

where AI is the aberration index, pc is the positive control (glyphosate or atrazine), s is the extract or fractions of T. stans flowers, and nc is the negative control (MES).

2.9. Statistical Analysis

All experimental data were expressed as mean ± standard error. Before statistical analysis, the Shapiro–Wilk test and Kolmogorov–Smirnov test were applied to evaluate the normality of the data, and the Brown–Forsythe test or Bartlett’s test was applied to determine the homogeneity of variance. The data for each group were analyzed through one-way ANOVA, and post hoc Dunnett’s test was used to compare significant differences between experimental groups with controls. Statistical analysis was performed using the GRAPHPAD PRISM 7.0 software (San Diego, CA, USA) with p < 0.05 significance levels.

3. Results and Discussion

3.1. Analysis of ¹H NMR Spectra of T. stans Samples

A metabolic profile from a natural product extract can be obtained using different techniques. All of them have advantages and disadvantages in their performance [44]. In this study, NMR was chosen because it is fast, efficient, highly reproducible, does not need pre-derivatization before analysis, is a universal qualitative and quantitative technique, is not selective and does not depend on the compound’s chemical characteristics, such as polarity and acidity (pKa) [45]. Little information has been reported about the flower extract and fractions composition of T. stans. Therefore, an undirected analysis of the samples was carried out in this study, with the aim of providing a fingerprint of the extract and its fractions, revealing structural information that would allow us to identify some analytes in the mixtures and that would add information about the main alterations in the different fractions obtained. This data can help predict the classes of natural products potentially involved in the cytotoxic, genotoxic, and antigenotoxic effects of the extract and fractions of T. stans flowers.

Figure 1 shows the ¹H NMR spectrum of T. stans ethanol extract. It is possible to observe by characteristic chemical shifts that this extract contains many compounds belonging to different natural product classes, such as alkaloids, phenylpropanoids, polyketides, sugars, and terpenoids, with the sugar region signals predominating [5.13 (d, J = 3.68 Hz) to 3.12 (m) ppm] [46].

Figure 2 shows the expansion of the spectrum 1 in this aromatic region, and Figure 3 shows the comparison among the ¹H NMR spectra of the EE and its fractions. We can analyze the spectrum by region. Starting from the aromatic region, it may be considered that flavonoids, phenolic acids, and other aromatic compounds are present in this extract (Figure 2). The liquid chromatography coupled to diode array detector and mass spectrometer (LC-DAD-MS) analysis of the EE and its fractions revealed the presence of caffeoyl and p-coumaroyl derivatives [47,48]. The intense signals at δ 7.58 (d, J = 15.95 Hz), 7.25 (m), 6.79 (m), 6.67 (m), and 6.26 (d, J = 15.86 Hz) agree with coumaroyl derivatives, and they could be observed in the extract and its fractions. The solvent partition enhances the aromatic compounds in the EA fraction, followed by the DCM fraction.

The second region of the spectrum considered is between 6.00 and 3.00 ppm, where most signals are likely to belong to carbohydrates. NMR spectra indicated that EE and its EA and HE fractions have dominant signals in this region, whereas the DCM fraction contains lower concentrations (Figure 3).

In this region, it is possible to observe signals characteristic of anomeric protons resonances at δ_H 5.37 (d, J = 3.67)/δ_C 92.29; δ_H 5.16 (d, J = 2.00)/δ_C 101.57; δ_H 5.09 (d, J = 3.67)/δ_C 92.54; δ_H 4.46 (d, J = 7.78)/δ_C 96.93 assigned to sucrose, rhamnose, α-glucose, β-glucose [49]; whose J¹ connections were also observed at HSQC and COSY spectra (Figures S1 and S2).

The third region of the ¹H NMR spectra is between 3 and 0.5 ppm, characteristic of organic acids, amino acids, terpenes, and fatty acid signals. In this region, the fractions exhibit a notable difference in metabolic profile, indicating that solvent partitioning served as the primary separation method (Figure 4).

Considering the difference in metabolic profile observed by the ¹H NMR spectra of the fractions, an exploratory analysis of the spectra of these fractions was conducted. The ¹H NMR spectrum of the HEX fraction shows a more straightforward metabolic profile. The significant signals are consistent with saturated (2.25, 1.55, 1.27–1.31, and 0.88 ppm) and unsaturated fatty acids ω-3 and ω-6 (2.81, 0.96, 2.77, 0.88–0.91, 5.32–5.33 ppm) [50]. These signal attributions were confirmed by 2D NMR analysis (Figure S3).

Although the DCM fraction still displayed some signals of fatty acids, other signals were observed and assigned to phenylethanol glycosides, monoterpenoid alkaloids, and flavonoids. Signals, such as 2.54, 2.18, 0.93, and 0.89 ppm, are characteristic of methyl groups of monoterpene backbones present in monoterpenoid alkaloids and iridoids [51], and they could be observed in the DCM and EA NMR spectra. Additionally, the DCM fraction likely presents a great diversity of terpenes or monoterpenoid alkaloids due to the higher intensity of the signals in this region. Hexosides of caffeoyl, coumaroyl, and fenylethanol derivatives were putatively identified by mass spectrometry [47,48]. Signals between 7.86 and 6.26 ppm, aligned with the signals from 4.86 to 3.5 ppm, are in agreement with the presence of these compounds [52].

The ethyl acetate fraction signals in the upfield region of the spectrum also displayed different profiles. It presented a singular doublet at 1.07 ppm (J = 6.11 Hz) coupled with a carbon at 18.19 ppm, one singlet at 2.54 ppm coupled with a carbon at 28.61 ppm, and multiplets of different intensities among 3.77–2.78, 1.85–1.58, and 1.45–1.33 ppm.

Analysis of 2D spectra allowed the assignment of the doublet at 1.07 ppm (J = 6.11 Hz) to a methyl group of a deoxy sugar such as rhamnose or fucose. This signal is coupled with a multiplet at 3.54 ppm, which in turn is coupled to another multiplet at 3.89 ppm (Figure 5), characteristic of the deoxysugar rhamnose present in verbascoside, as characterized by MS in the DCM and EA fractions [47,48].

The COSY spectrum showed the coupling of multiplets at 3.72 and 2.77 ppm (Figure 5), which are correlated in the HMBC spectrum with the signals at 70.79, 115.76, 119.83, and 129.99 ppm (Figure 6), corresponding to carbons from the sugar and aromatic ring moieties of phenylethyl glycosides.

The HMBC correlation between δ_H 4.89 with δ_C 74.86, 80.19, and 166.84, together with correlations shown by proton δ_H 7.58 with δ_C 113.73, 121.84, and 166.84 (Figure 6), reinforced the presence of coumaroyl or caffeoyl glycosides, such as verbascoside, isoacteoside, or forsythoside A, which are isomers that were putatively characterized by their mass data from LC-DAD-MS analyses of the T. stans flower extract and its fractions [47,48]. Verbascoside, isoverbascoside, and other phenylethyl glycosides have been described previously in leaves, flowers, and fruits of T. stans [47,48,52,53]. The chemical shifts related to the compounds putatively detected in T. stans flower extract and fractions are listed in the Table S1.

3.2. Germination Response of Seeds to Extract and Fractions of T. stans

Table 2. Analysis of seed germination observed for L. sativa and A. cepa treated with the EE and HEX, DCM, EA, and HE fractions and controls.

L. sativa					A. cepa
	μg/mL	1st Count (%)	Germination Rate (%)	GSI	μg/mL	1st Count (%)	Germination Rate (%)	GSI
MES	0	78.0 ± 5.8	93.0 ± 2.2	11.2 ± 0.5	0	88.0 ± 3.7	100.0 ± 0.0	8.9 ± 0.4
EE	250	63.0 ± 3.5 a	72.0 ± 4.6 a	9.5 ± 0.7	250	74.0 ± 4.0 a	89.0 ± 1.2 a	5.7 ± 0.4 a
	500	67.0 ± 3.9	72.0 ± 1.9 a	8.6 ± 0.4 a	500	69.0 ± 6.4 ac	95.0 ± 2.2 c	5.8 ± 0.2 ac
	750	71.0 ± 3.5	78.0 ± 1.3 a	8.6 ± 0.3 a	750	81.0 ± 2.9 c	92.0 ± 4.2 c	6.0 ± 0.2 ac
	1000	71.0 ± 2.9	72.0 ± 3.3 a	8.5 ± 0.4 a	1000	72.0 ± 4.2 ac	95.0 ± 1.2 bc	5.8 ± 0.3 abc
HEX	250	67.0 ± 2.2	76.0 ± 2.7 a	10.6 ± 0.6	250	50.0 ± 5.8 abc	81.0 ± 3.9 ab	3.8 ± 0.3 abc
	500	66.0 ± 4.0	67.0 ± 3.9 a	8.5 ± 0.5 a	500	33.0 ± 7.4 ab	85.0 ± 3.9 ac	3.1 ± 0.2 ab
	750	67.0 ± 3.5	74.0 ± 1.3 a	8.1 ± 0.2 a	750	25.0 ± 2.2 ab	80.0 ± 4.2 ac	2.9 ± 0.1 abc
	1000	70.0 ± 4.4	77.0 ± 3.9 a	8.4 ± 0.3 a	1000	21.0 ± 6.1 abc	85.0 ± 2.9 ac	2.8 ± 0.2 abc
DCM	250	63.0 ± 2.2 a	76.0 ± 3.3 a	9.9 ± 0.3	250	66.0 ± 10.0 ab	81.0 ± 7.1 ab	5.6 ± 0.7 a
	500	63.0 ± 2.9 a	72.0 ± 4.2 a	7.6 ± 0.4 a	500	43.0 ± 8.9 ab	69.0 ± 3.9 abc	3.6 ± 0.3 ab
	750	42.0 ± 3.0 ab	58.0 ± 5.5 ab	4.6 ± 0.4 ab	750	74.0 ± 4.8 ac	88.0 ± 3.8 ac	5.4 ± 0.4 abc
	1000	52.0 ± 6.0 a	65.0 ± 6.0 a	4.7 ± 0.4 ab	1000	78.0 ± 4.4 c	92.0 ± 1.9 c	6.1 ± 0.4 ac
EA	250	68.0 ± 2.8	78.0 ± 1.3 a	9.7 ± 0.8	250	79.0 ± 7.1	97.0 ± 3.5 c	6.6 ± 0.3 a
	500	67.0 ± 3.5	76.0 ± 1.9 a	11.0 ± 0.5	500	74.0 ± 6.1 c	93.0 ± 1.2 c	5.8 ± 0.4 ac
	750	70.0 ± 1.3	74.0 ± 3.0 a	10.4 ± 0.2	750	77.0 ± 3.5 c	95.0 ± 2.9 c	6.4 ± 0.1 ac
	1000	71.0 ± 2.9	73.0 ± 2.9 a	9.9 ± 0.5	1000	64.0 ± 4.2 ac	90.0 ± 4.0 c	5.5 ± 0.3 abc
HE	250	70.0 ± 4.0	73.0 ± 5.1 a	11.12 ± 1.2	250	78.0 ± 3.0	92.0 ± 3.8	6.3 ± 0.3 a
	500	68.0 ± 5.0	71.0 ± 5.1 a	9.1 ± 0.8 a	500	66.0 ± 6.7 abc	91.0 ± 3.5 c	5.9 ± 0.4 ac
	750	63.0 ± 3.9 a	74.0 ± 6.9 a	11.3 ± 0.7 b	750	76.0 ± 9.4 c	92.0 ± 3.3 c	5.7 ± 0.5 abc
	1000	71.0 ± 1.2	81.0 ± 1.2 a	11.9 ± 0.7	1000	73.0 ± 8.3 c	90.0 ± 1.3 ac	5.6 ± 0.4 abc
ATZ	250	71.0 ± 1.1	77.0 ± 2.2 a	9.5 ± 0.5	250	77.0 ± 5.1	96.0 ± 1.9	6.8 ± 0.7 a
	500	67.0 ± 2.2	73.0 ± 3.9 a	9.2 ± 0.5 a	500	75.0 ± 4.4	88.0 ± 1.9 a	6.6 ± 0.3 a
	750	66.0 ± 2.3	74.0 ± 4.0 a	8.9 ± 0.6 a	750	72.0 ± 3.3 a	90.0 ± 1.3 a	6.9 ± 0.2 a
	1000	64.0 ± 4.2 a	75.0 ± 3.9 a	10.2 ± 0.8	1000	79.0 ± 2.9	82.0 ± 4.0 a	7.1 ± 0.4 a
GLI	250	-	-	-	250	78.0 ± 5.5	82.0 ± 5.5 a	5.9 ± 0.7 a
	500	-	-	-	500	44.0 ± 6.3 a	53.0 ± 6.4 a	2.9 ± 0.4 a
	750	-	-	-	750	13.3 ± 1.3 a	20.0 ± 3.5 a	1.0 ± 0.2 a
	1000	-	-	-	1000	3.2 ± 2.5 a	9.0 ± 3.1 a	0.4 ± 0.2 a

MES: 2-(N-morpholino)ethanesulfonic acid, EE: ethanol extract, HEX: hexane fraction, DCM: dichloromethane fraction, EA: ethyl acetate fraction, HE: hydroethanol fraction, ATZ: atrazine, GLI: glyphosate. Results are means ± standard error (n = 4). ^a Statistically significant difference in relation to the MES (p < 0.05). ^b Statistically significant difference in relation to atrazine, considering samples at the same concentrations (p < 0.05). ^c Statistically significant difference in relation to glyphosate, considering samples at the same concentrations (p < 0.05).

presents the first count, germination rate, and GSI for L. sativa and A. cepa seeds treated with the T. stans extract and fractions, the negative control (MES), and positive controls (atrazine and glyphosate).

The effects caused by the T. stans extract and fractions were independent of the concentration, as they exhibited statistically significant differences in germination compared to MES at lower concentrations. The inverse behavior of the dose-dependent effect on concentration may be related to the difficulty in absorbing allelochemicals present in high concentrations. Limitations in the absorption and translocation of allelochemicals at high concentrations can occur. At these concentrations, allelopathic compounds may undergo molecular aggregation, precipitation, or even induce changes in the permeability of cell membranes, reducing their effective entry into target plant tissues. Furthermore, high concentrations can trigger defense mechanisms, such as the activation of antioxidant systems and metabolizing enzymes, decreasing the phytotoxic effect [54]. However, at lower concentrations, the absorption of allelochemicals would have better efficiency and, therefore, a greater inhibitory effect [55]. At lower concentrations, allelochemicals tend to exhibit greater bioavailability, favoring their absorption by the roots and young tissues of recipient plants. In this sense, the compounds can interfere more efficiently in essential physiological processes, such as cell division, enzymatic activity, protein synthesis, and hormonal balance, resulting in a greater inhibitory effect on germination and plant growth. This phenomenon characterizes a hormesis-type response, widely described in allelopathy studies, in which low doses exert more effective effects than high doses [56].

The HEX and DCM fractions promoted greater inhibition of seed germination in both species tested, with statistically significant differences compared to MES for L. sativa seeds, and compared to MES, and ATZ and/or GLI for A. cepa seeds. Fatty acids and phenolic compounds, respectively, were characterized in these fractions by LC-DAD-MS, as previously described [47,48], which was corroborated by the NMR analyzes reported in this work. Phenolic compounds are the chemical substances most associated with the allelopathic effects of vegetables [57,58]. These compounds, as well as fatty acids are used for integrated pest management in organic agriculture due to their toxic effects on the growth and development of plant species [59]. In addition, flavonoids inhibit plant germination and growth [60].

Regarding GSI, the extract and HEX, DCM (500, 750, and 1000 μg/mL), HE (500 μg/mL) fractions and the positive control (ATZ) (500 and 750 μg/mL) exhibited lower values than the negative control (MES), with statistically significant differences in relation to L. sativa seeds. For A. cepa seeds, the T. stans samples and the positive controls (ATZ and GLI) decreased the GSI compared to MES. The HEX fraction (250 μg/mL) showed a lower value compared to GLI at the same concentration, and at all concentrations compared to ATZ (p < 0.05). The EE and EA (1000 μg/mL), DCM (500 and 750 μg/mL) and HE (750 and 1000 μg/mL) showed low GSI values compared to ATZ at the same concentrations. The positive control GLI was also tested in trials with L. sativa seeds; however, it caused necrosis and contamination in all treatments. The high toxicity of this herbicide has already been mentioned in the literature, which corroborates our results [61,62].

3.3. Seedling Growth Analysis

The epicotyls and roots of the seedlings were measured to determine whether the tested treatment inhibited or induced seedling growth. The results for L. sativa are presented in Figure 7. All treatments affected the growth of the root and epicotyl of L. sativa. In addition, the greatest effect of the extracts and fractions occurred on the root system, which showed a greater reduction in its development compared to the epicotyl. Compared to ATZ, the EE (250 μg/mL), HEX (750 μg/mL), DCM (250 and 750 μg/mL), EA (250 μg/mL) and HE (750 μg/mL) inhibited root growth more markedly (p < 0.05), while the EE, HEX, DCM and EA (500 μg/mL), and EE, DCM and EA (1000 μg/mL) showed similar inhibition. For epicotyl growth, the extract and fractions promoted less inhibition than ATZ at the tested concentrations (p < 0.05), except for DCM (750 and 1000 μg/mL). The DCM fraction stood out at concentrations of 750 and 1000 μg/mL, with inhibition of 61.6% and 57.3% on the epicotyls, respectively, compared to the negative control (p < 0.05), with the result being similar to the ATZ at these concentrations.

Atrazine acts as an inhibitor of photosystem II (PSII), blocking the flow of electrons in the photochemical phase of photosynthesis. This blockage occurs through the binding of the molecule to the QB site of the D1 protein, preventing the transfer of electrons from plastoquinone and interrupting the electron transport chain. As a consequence, there is a reduction in the production of ATP and NADPH, compromising energy-dependent reactions and resulting in decreased carbon assimilation and the synthesis of carbohydrates, sugars, and other metabolites essential for plant growth. This metabolic imbalance leads to the formation of reactive oxygen species, damage to cell membranes, and ultimately, plant death. Atrazine exhibits predominantly xylem mobility and is mainly absorbed by the roots, which is why it is preferentially applied to the soil and is widely used as a pre-emergent herbicide [63].

Phenolic compounds, such as caffeoyl and coumaroyl derivatives were characterized in EE and in the DCM and EA fractions by LC-DAD-MS [47,48] and NMR. According to the literature, these compounds can stimulate the activity of oxidative enzymes, which are involved in the synthesis of defense compounds and in the lignification of the cell wall. The unregulated activation of these enzymes can lead to the accumulation of reactive oxygen species (ROS) and oxidative stress in the root tissue, resulting in damage to membranes and organelles [64]. Furthermore, one of the most consistent mechanisms reported for phenolic compounds in allelopathy is their ability to alter cell membrane permeability. This mechanism can result in increased permeability to electrolytes; lipid peroxidation and disruption of the lipid bilayer, compromising root cell homeostasis and reducing their elongation and differentiation capacity [65,66,67]. Phenolic compounds interfere with hormonal activity, membrane permeability, photosynthesis, respiration, and organic compound synthesis [68]. In this sense, the phenolic derivatives observed in T. stans samples contributed to the reduction in root length with these treatments for L. sativa seeds.

For A. cepa seeds (Figure 8), the best results were observed in treatments with the HEX and DCM fractions of T. stans, which inhibited root growth by more than 50% at a concentration of 500, 750, and 1000 μg/mL, compared to the negative control (p < 0.05), and showed an effect similar to that of the ATZ control at all tested concentrations. Regarding the percentage of inhibition of the epicotyl, the HEX fraction stood out, as it affected the growth of A. cepa at all concentrations in relation to the negative control and with values higher than the inhibition potential of the positive control ATZ (p < 0.05). The DCM fraction also showed a greater inhibitory effect than ATZ at concentrations of 500 and 1000 μg/mL (p < 0.05). The EE and fractions exhibited less inhibition of root and epicotyl growth compared to GLI (p < 0.05).

It was also possible to observe atrophy and deficiency of root differentiation in the A. cepa seedlings treated with the HEX fraction. Changes in the root growth rate, shape, and color are macroscopic parameters that indicate cytotoxic damage [69]. Analysis of the growth of A. cepa seedlings treated with the T. stans flower extract and fractions showed changes in the growth and shape of the roots compared to the negative control. Glyphosate changed the morphology of A. cepa roots during the germination process, mainly causing necrosis.

The mechanism of action of glyphosate is widely described in the literature. This herbicide interferes with plant growth and cell division by acting as a functional analog of auxin, one of the main phytohormones regulating plant growth. Glyphosate interacts with the same action sites as endogenous auxin, promoting an unregulated increase in auxin signaling in plant tissues. As a consequence, several metabolic processes dependent on this hormone are altered, including cell elongation, tissue differentiation, and cell cycle regulation. This physiological imbalance results in growth inhibition, chlorosis of meristematic tissues, and subsequently necrosis, culminating in a slow and progressive death of the plants [63].

Phenolic compounds and alkaloids present in the DCM fraction showed allelopathic potential. Alkaloids play a significant role in the allelopathic activity of vegetables, inhibiting plant growth via several mechanisms, including interference with DNA, enzyme activity, protein biosynthesis, and membrane integrity in developing plants [70]. However, one of the current challenges in allelopathy is the determination of the specific mode of action of allelochemicals, as they have a diverse chemical nature and multiple target sites in higher plants [71].

Studies evaluating the allelopathic activity of T. stans flowers were not found in the literature. Cipriani et al. [30] evaluated the allelopathic effect of the aqueous extract obtained from T. stans leaves on the germination and growth of L. sativa seeds, obtaining an inhibition percentage of 98% (the concentration used was not specified) in relation to the group control. In this study, carbocyclic iridoids and alkaloids derived from the acetic acid pathway, including iridoid alkaloids and benzoic acid derivatives, appear to be the main candidates for phytotoxic substances present in T. stans extracts [30]. In fact, phytotoxic (allelopathic) effects have already been described for glycosylated and potentially water-soluble iridoid derivatives [72,73]. Our results corroborate those reported in the literature, as glycosylated iridoids were annotated in the extract and fractions [47,48].

The bioherbicidal capacity of methyl palmitate has been demonstrated, among other methyl esters isolated from Lantana camara flowers on some weed species and suggested that these compounds may have inhibitory effects on the generation of ATP and electron transport in chloroplasts and mitochondria, similar to other phytotoxins [74]. In our study, saturated and unsaturated fatty acids ω-3 and ω-6 were characterized in the HEX fraction, which may be related to the allelopathic potential found for this fraction.

Some research has already highlighted the allelopathic potential of species from the Bignoniaceae family; for example, the hexane fraction of Jacaranda micrantha flowers inhibited the growth of the epicotyl and radicle of L. sativa [75]. The chloroform fraction from the stems of Tynanthus micranthus inhibited the germination of L. sativa at a concentration of 0.8 mg [76]. The extract obtained by cold infusion of the leaves, stem bark, and roots of Handroanthus serratifolius reduced the length of Zea mays seedlings [77].

3.4. Genotoxic Potential of T. stans

The MI, AI, and NI were analyzed. The results of genotoxicity tests with A. cepa meristematic cells against extracts from T. stans flowers are represented in

Table 3. Results of genotoxicity tests observed in A. cepa treated with the EE and HEX, DCM, EA, and HE fractions and positive controls (ATZ and GLI).

Sample	Treatments (μg/mL)	MI	AI	NI
MES	0	4.52 ± 0.66	5.61 ± 1.42	0
EE	250	2.95 ± 0.90 b	4.52 ± 0.26 b	0.09 ± 0.09
	500	2.66 ± 0.37	5.71 ± 2.85 bc	0.14 ± 0.14
	750	2.47 ± 0.09	8.90 ± 1.71 c	0.42 ± 0.21
	1000	1.66 ± 0.26 abc	12.98 ± 1.82 c	0.47 ± 0.12 c
HEX	250	1.23 ± 0.17 abc	4.71 ± 1.24 b	0
	500	1.57 ± 0.21 ab	8.19 ± 2.02 bc	0
	750	1.76 ± 0.91 a	9.76 ± 1.06 c	0
	1000	1.33 ± 0.12 abc	11.61 ± 1.24 c	0 bc
DCM	250	2.19 ± 0.20 abc	5.57 ± 0.42 b	0
	500	1.70 ± 0.50 ab	7.25 ± 0.94 bc	0
	750	1.38 ± 0.34 a	8.71 ± 1.13 c	0
	1000	1.28 ± 0.21 abc	10.95 ± 1.83 c	0.09 ± 0.09 c
EA	250	3.14 ± 0.50 b	8.80 ± 1.53	0.19 ± 0.12
	500	3.04 ± 0.28	13.00 ± 2.42 a	0.38 ± 0.12
	750	3.71 ± 0.71	16.00 ± 1.00 ac	1.71 ± 0.82
	1000	1.80 ± 0.88 bc	17.00 ± 1.14 ac	0.38 ± 0.17 c
HE	250	4.10 ± 0.41	5.10 ± 0.84 b	0
	500	2.71 ± 0.42	5.50 ± 0.82 bc	0.66 ± 0.00
	750	3.19 ± 0.40	10.23 ± 3.26 c	0.28 ± 0.28
	1000	2.00 ± 0.65 abc	10.38 ± 0.84 c	0.28 ± 0.28 c
ATZ	250	5.57 ± 0.54	13.42 ± 2.21 a	2.14 ± 1.09
	500	5.42 ± 1.24	16.71 ± 1.91 a	0.23 ± 0.09
	750	3.76 ± 0.25	10.66 ± 2.14	1.95 ± 1.14
	1000	7.85 ± 0.86	19.23 ± 2.69 a	3.19 ± 1.27
GLI	250	5.13 ± 0.51	7.41 ± 0.70	1.47 ± 1.26
	500	2.80 ± 1.95	18.66 ± 1.76 a	0.85 ± 0.49
	750	1.42 ± 0.73 a	29.19 ± 4.14 a	0.95 ± 0.12
	1000	6.76 ± 3.56	48.95 ± 6.88 a	27.09 ± 3.04 a

MI: Mitotic Index, AI: Aberration Index, NI: Necrosis Index. MES: 2-(N-morpholino)ethanesulfonic acid, EE: ethanol extract, HEX: hexane fraction, DCM: dichloromethane fraction, EA: ethyl acetate fraction, HE: hydroethanol fraction, ATZ: atrazine, GLI: glyphosate. Results are mean ± standard error (n = 4). ^a Statistically significant difference in relation to the MES (p < 0.05). ^b Statistically significant difference in relation to atrazine, considering samples at the same concentrations (p < 0.05). ^c Statistically significant difference in relation to glyphosate, considering samples at the same concentrations (p < 0.05).

.

T. stans samples showed a decrease in MI, especially for the HEX and DCM fractions, with a statistically significant difference compared to the negative control (p < 0.05). Studies have shown that a decrease in MI causes root growth inhibition [78,79], and this was observed in the allelopathy assays carried out in this study with the extract and fractions of T. stans. Plant extracts that selectively and moderately reduce MI can be explored for the development of natural herbicides [80,81].

As a defense mechanism for the immune response, higher plants release phytohormones and secondary metabolites that play a fundamental role in activating the immune response and preventing division in cellular areas affected by physical or chemical action [82]. In this sense, the decrease in MI may have been triggered by a defense mechanism of A. cepa to phytochemical compounds from T. stans extracts and fractions. Furthermore, phenolic compounds may also be related to the decrease in the mitotic index found in A. cepa [83,84].

Previous studies have shown a decrease in MI after treatments with extracts of plant species analyzed using the A. cepa test [85,86]. The genotoxic and antiproliferative effects of the aqueous extract of the bark of Handroanthus chrysotrichus (Bignoniaceae), popularly known as ipê-amarelo have shown a reduction in MI at all tested concentrations (5, 10, and 15 g/L) compared to the negative control, and there was no genotoxic action of the extracts on the A. cepa cell cycle [87].

The potential of plant extracts that inhibit mitotic growth could be explored for use in cancer cells as a possible treatment strategy. The cytotoxic potential of T. stans extracts against human cancer cell lines has been evaluated, and the extract obtained from the stems of T. stans (var. stans) showed a cytotoxic concentration (CC₅₀) between 0.0156 and 0.5533 μg/mL on tumor cell lines, and the leaf extract showed a CC₅₀ of 39.89–200.0 μg/mL [88]. The extract obtained from the stem of T. stans (var. angustata) was also cytotoxic, presenting a CC₅₀ of 0.0841–80.25 μg/mL, and the leaf extract showed a moderate CC₅₀ between 24.22 and 200.0 μg/mL. In addition, other studies have reported the anticancer activity of T. stans [89,90,91].

The positive controls, ATZ and GLI, increased the MI at a concentration of 1000 μg/mL but were not statistically significant compared to the negative control. However, an MI higher than the negative control is the result of increased cell division, which may be harmful due to disordered cell proliferation and tumor tissue formation [92]. The mitotic activity count, among other characteristics, such as NI, is a parameter used as a marker of tumor proliferation [93]. Studies have shown that glyphosate can induce cancer cell growth [94].

Regarding AI, some treatments with T. stans samples presented chromosomal anomalies; however, only the EA fraction at concentrations of 500, 750, and 1000 μg/mL showed a statistically significant difference compared to the negative control. These results suggest that the EA fraction exhibits genotoxic potential at higher concentrations tested. The genotoxicity observed for the EA fraction at high concentrations is likely related to its enrichment in semipolar secondary metabolites, particularly phenolic compounds and flavonoids, which at high doses can induce oxidative stress. The excessive production of reactive oxygen species (ROS) can result in DNA damage, including strand breaks, chromosomal aberrations, and mitotic spindle disruption, simultaneously overloading cellular detoxification and DNA repair mechanisms, thus increasing chromosomal instability [27]. In contrast, the T. stans extract and other fractions were not considered genotoxic.

In the positive controls, ATZ and GLI (Figure 9), a greater number of micronuclei, anaphase bridges, sticky chromosomes (Figure 9G), aneuploidy (Figure 9K), and polyploidy, among other aberrations, as well as necrosis (Figure 9L) were observed.

A greater amount of necrosis was observed in treatments with the EA fraction at a concentration of 750 μg/mL, and ATZ at concentrations of 250, 750, and 1000 μg/mL (Table 3); however, this necrosis rates were not statistically significant compared to the negative control (MES). The positive control (GLI) caused high necrosis rate at a concentration of 1000 μg/mL (Figure 9L), with a statistical difference compared to the negative control. The increase in NI may indicate toxicity of the tested substances, suggesting significant genotoxic potential. This index is an important complementary measure to the other parameters evaluated in genotoxic assays with A. cepa, providing a comprehensive view of the effects of the tested substances on plant cells [95,96].

In addition, the presence of micronuclei was observed in the EA fraction at a concentration of 1000 μg/mL (Figure 9H) and in the ATZ and GLI controls at a concentration of 500 μg/mL. The appearance of micronuclei during cell division is related to the mutagenic or genotoxic effects of biological agents. Furthermore, anaphase bridges were observed in treatments with the EA fraction at a concentration of 1000 μg/mL (Figure 9F) and in positive controls. These anomalies are formed by the breakage and fusion of chromosomes and chromatids and by changes in the activation of replication enzymes [97]. Some phytochemical compounds, such as flavonols, polyphenols, alkaloids, and tannins, have been cited for causing chromosomal damage in high concentrations [43]. In this sense, it is important to determine the safe dosage for the correct use of natural products to avoid adverse effects and ensure the effectiveness of these products.

No further studies were found that evaluated the genotoxic activity in vitro of T. stans.

3.5. Antigenotoxic Potential of T. stans

To determine the antigenotoxic potential of T. stans samples, A. cepa seedlings were treated with glyphosate and atrazine at a concentration of 500 μg/mL for 24 h, subsequently treated with the extract and fractions at concentrations of 250, 500, 750, and 1000 μg/mL, and left at B.O.D. for another 24 h. The results of the antigenotoxic activity of T. stans are shown in

Table 4. Results of the antigenotoxicity assay observed in A. cepa treated with the EE and HEX, DCM, EA, and HE fractions previously treated with atrazine and glyphosate.

Atrazine					Glyphosate
Samples	μg/mL	MI	AI	Low AI (%)	Samples	μg/mL	MI	AI	Low AI (%)
MES	0	4.38 ± 1.45	4.19 ± 1.04	-	MES	0	4.09 ± 0.60	4.28 ± 1.59	-
ATZ	500	4.23 ± 0.88	13.90 ± 1.81 a	-	GLI	500	1.38 ± 0.47 a	17.71 ± 1.08 a	-
EE	250	1.38 ± 0.33 a	6.95 ± 0.59 b	71.56	EE	250	2.14 ± 0.37	9.04 ± 1.40 c	64.53
	500	1.14 ± 0.32 ab	9.09 ± 1.25 a	49.50		500	1.14 ± 0.43 a	5.33 ± 1.61 c	92.19
	750	1.07 ± 0.27 ab	10.77 ± 1.33 a	32.17		750	3.14 ± 1.62	6.42 ± 1.15 c	84.04
	1000	0.95 ± 0.45 ab	11.19 ± 1.21 a	27.94		1000	2.42 ± 0.41	7.18 ± 1.97 c	78.40
HEX	250	1.52 ± 0.41	4.57 ± 0.64 b	96.07	HEX	250	1.52 ± 0.26 a	7.00 ± 2.00 c	79.78
	500	1.19 ± 0.25 ab	7.38 ± 1.85 b	67.15		500	2.33 ± 0.45	2.23 ± 0.66 c	115.24
	750	1.09 ± 0.20 ab	9.23 ± 0.78 a	48.03		750	1.42 ± 0.51 a	4.90 ± 0.67 c	95.39
	1000	0.85 ± 0.50 ab	9.29 ± 0.49 a	47.42		1000	1.19 ± 0.41 a	4.47 ± 0.38 c	98.58
DCM	250	1.09 ± 0.20 ab	3.71 ± 0.08 b	104.90	DCM	250	1.33 ± 0.49 a	8.38 ± 2.71 c	69.50
	500	1.04 ± 0.37 ab	6.14 ± 0.59 b	79.90		500	0.76 ± 0.42 a	6.61 ± 1.99 c	82.62
	750	1.23 ± 0.04 ab	9.80 ± 0.53 a	42.15		750	0.52 ± 0.12 a	7.28 ± 0.43 c	77.65
	1000	1.04 ± 0.19 ab	10.29 ± 0.92 a	37.11		1000	0.38 ± 0.12 a	6.42 ± 2.45 c	84.04
EA	250	3.00 ± 0.45	7.95 ± 2.12 b	61.27	EA	250	3.14 ± 0.45	8.14 ± 0.41 c	71.27
	500	3.28 ± 1.14	10.71 ± 1.23 a	32.84		500	2.61 ± 0.54	7.33 ± 1.83 c	77.30
	750	2.04 ± 0.49	11.38 ± 0.26 a	25.98		750	4.14 ± 2.26 c	12.80 ± 4.13 a	36.52
	1000	1.61 ± 0.17	12.47± 1.69 a	14.70		1000	3.09 ± 0.48	13.90 ± 2.99 a	28.36
HE	250	1.90 ± 0.76	6.85 ± 1.39 b	72.54	HE	250	1.80 ± 0.58	9.90 ± 2.18 c	58.15
	500	2.14 ± 0.57	8.57 ± 1.93 b	54.90		500	2.47 ± 0.59	4.23 ± 0.73 c	100.35
	750	1.95 ± 0.66	9.76 ± 0.83 a	42.64		750	0.71 ± 0.43 a	4.19 ± 0.71 c	100.70
	1000	1.57 ± 0.45	10.52 ± 0.93 a	34.80		1000	1.52 ± 0.80	3.76 ± 1.34 c	103.90

MI: Mitotic Index, AI: Aberration Index; Low AI (%): Antigenotoxic effect. MES: 2-(N-morpholino)ethanesulfonic acid, ATZ: atrazine, GLI: glyphosate EE: ethanol extract, HEX: hexane fraction, DCM: dichloromethane fraction, EA: ethyl acetate fraction, HE: hydroethanol fraction. Results are mean ± standard error (n = 4). ^a Statistically significant difference in relation to the MES (p < 0.05). ^b Statistically significant difference in relation to atrazine at concentration of 500 μg/mL (p < 0.05). ^c Statistically significant difference in relation to glyphosate at concentration of 500 μg/mL (p < 0.05).

.

A. cepa seeds treated with the extract and fractions of T. stans (pre-treated with ATZ) showed a reduction in MI compared with MES and ATZ, with a statistically significant difference (p < 0.05) in relation to the two controls for EE and HEX (500, 750, and 1000 μg/mL), and DCM at all concentrations tested. For seeds pre-treated with glyphosate, there was also a reduction in MI, with a statistically significant difference (p < 0.05) for EE (500 μg/mL), HEX (250, 750, and 1000 μg/mL), DCM (all concentrations) and HE (750) compared to MES.

When determining the antimutagenic potential of a sample, values lower than 25% inhibition of mutagenic activity indicate a weak or non-antimutagenic effect. Values between 25 and 40% reduction in AI indicate a moderate effect. However, values greater than 40% indicate a strong antimutagenicity of the analyzed compound [98]. In this sense, the T. stans extract and fractions reduced the genotoxicity caused by glyphosate and atrazine.

When analyzing AI, there was a statistically significant decrease in all treatments with T. stans extract and fractions compared to the aberrations caused by GLI at a concentration of 500 μg/mL. The strongest effect antigenotoxic was observed in cells pre-treated with glyphosate, since the extract and fractions promoted a reduction in the AI greater than 40%, except EA fraction at concentrations 750 and 1000 μg/mL. However, in treatments with ATZ (500 μg/mL), the lower the concentration, the greater the decrease in AI. Thus, the HEX and DCM fractions stood out at a concentration of 250 μg/mL, which reduced the rate of aberrations caused by ATZ by more than 95%.

The most common chromosomal aberration recorded in glyphosate and atrazine was C-metaphase. According to the literature, C-metaphase is a reversible aberration, but chromosomal breaks, anaphase bridges, and sticky chromosomes are aberrations that are difficult to correct [99]. Plant extracts may have antigenotoxic potential due to the presence of bioactive compounds, which protect DNA against genotoxic damage. Antioxidant compounds can minimize DNA damage, as they are responsible for the primary defense against oxidative damage and maintain the integrity of genetic material, contributing to the prevention of genotoxic events [100].

Studies have demonstrated the antioxidant activity of T. stans related to the presence of phenolic compounds in its extracts [47,48,101]. Phenolic compounds neutralize free radicals and reactive oxygen species, which are generated during exposure to a toxic agent [102]. In addition, some secondary metabolites activate DNA repair pathways, facilitating the correction of damage caused by a toxic agent, and/or inhibit enzymes involved in the activation processes of genotoxic compounds [103].

Previous studies have indicated that some flavonoids have antigenotoxic and antimutagenic activities. In these studies, flavonoids apigenin, quercetin, and luteolin were noted in the T. stans extracts and fractions [47,48,104,105]. It has been demonstrated that polyphenolics, such as luteolin, quercetin, and rosmarinic acid, have protective effects against oxidative damage to DNA in PC12 cells, a neuronal cell model [106]. In this sense, the antigenotoxic activity of T. stans may be related to the phenolic compounds characterized in its extracts and fractions.

According to some studies, substances with antigenotoxic properties have also shown anticancer properties [99], indicating that these compounds can help prevent the formation of tumors by protecting cells’ DNA against damage that could lead to cancer development [107]. By identifying antigenotoxic substances from plant products, it is possible to use them in the development of chemopreventive medicines to protect human DNA against damage. This could have significant implications for the treatment and prevention of numerous human pathologies, including cardiovascular diseases, premature aging, chronic diseases, inflammatory conditions, metabolic syndromes, and neurodegenerative disorders [108].

Reports in the literature have corroborated with the activities assessed in this work, which can be correlated with terpenes, alkaloids, phenolic compounds, and fatty acids characterized in the extract and fractions of T. stans. Radha and May [109] discussed the role of terpenoids as natural compounds with allelopathic activity, emphasizing that these compounds, especially volatile monoterpenes and sesquiterpenes can affect neighboring plants through exudates or the emission of volatile substances, influencing the germination, growth, and development of competing species. A study with extracts and fractions of Zanthoxylum rhoifolium demonstrated the induction of genotoxicity, with evidence of chromosomal alterations in in vitro/in planta assays. Alkaloids with a benzophenanthridine nucleus appear to be involved in this effect, possibly through interaction with Topoisomerase II, contributing to DNA damage [110].

Phenolic compounds are discussed as one of the main classes of plant secondary metabolites capable of protecting biological systems against genetic damage induced by physical, chemical, or biological agents. Flavonoids, phenolic acids, tannins, and other polyphenols act as antigenotoxic agents, reducing mutations, chromosomal aberrations, and micronucleus formation [111]. Freire [112] addresses fatty acids and their derivatives as central components of plant ecological signaling, highlighting that, in addition to structural and energetic functions, lipids act as signaling molecules capable of modulating defensive responses and chemical interactions between plants. The study emphasizes that these compounds actively participate in allelopathy, competition, and indirect communication processes, integrating responses to biotic and abiotic stress.

4. Conclusions

The ethanol extract and fractions (HEX, DCM, EA, and HE) of T. stans flowers exhibited promising allelopathic activity on seeds of monocotyledonous and eudicotyledonous species, but the EA fraction exhibited genotoxic potential at higher concentrations tested (500, 750, and 1000 μg/mL). T. stans flower extract and fractions also demonstrated antigenotoxic potential by reducing glyphosate- and atrazine-induced damage. Antigenotoxic compounds from flower extracts can be used as allelochemicals in agriculture to reduce the harmful effects of herbicides by promoting more sustainable agricultural practices. This study is the first to evaluate the allelopathic, genotoxic, and antigenotoxic effects of T. stans flowers.