Discovery of Novel Pentacyclic Triterpene Acid Amide Derivatives as Excellent Antimicrobial Agents Dependent on Generation of Reactive Oxygen Species

Developing new agricultural bactericides is a feasible strategy for stopping the increase in the resistance of plant pathogenic bacteria. Some pentacyclic triterpene acid derivatives were elaborately designed and synthesized. In particular, compound A22 exhibited the best antimicrobial activity against Xanthomonas oryzae pv. oryzae (Xoo) and Xanthomonas axonopodis pv. citri (Xac) with EC50 values of 3.34 and 3.30 mg L−1, respectively. The antimicrobial mechanism showed that the compound A22 induced excessive production and accumulation of reactive oxygen species (ROS) in Xoo cells, leading to a decrease in superoxide dismutase and catalase enzyme activities and an increase in malondialdehyde content. A22 also produced increases in Xoo cell membrane permeability and eventual cell death. In addition, in vivo experiments showed that A22 at 200 mg L−1 exhibited protective activity against rice bacterial blight (50.44%) and citrus canker disease (84.37%). Therefore, this study provides a paradigm for the agricultural application of pentacyclic triterpene acid.


Introduction
Plant diseases caused by the bacterial infestations of plant pathogens are the primary types of crop diseases [1,2]. The literature reports that yield losses due to plant diseases account for 10% to 15% of total yield each year, which causes enormous losses to agricultural production [3][4][5]. In particular, the Xanthomonas oryzae pv. oryzae (Xoo) is one of the main culprits of rice diseases and poses a severe threat to worldwide rice quality and yield [6,7]. Various control measures have been used to respond to pathogenic bacterial attacks. Of note, chemical pesticides are currently the most common and effective method for controlling plant bacterial diseases, providing an essential guarantee for global food production in recent decades [8,9]. However, the use of such traditional chemicals also brings some intractable problems, such as bacterial resistance [10], environmental pollution [11], ecological problems, and human safety concerns [12]. Therefore, an increasing need to develop agricultural bactericides that are both environmentally friendly and highly effective in controlling plant bacterial diseases is present.
Reactive oxygen species (ROS) are vital in bacterial resistance and antibiotic sterilization [13,14]. In recent years, antimicrobial drug employing ROS as a component of their mechanism of action are being developed in the pharmaceutical field and have been successfully applied for the prevention of surgical site infections [15,16], intra-vascular line care [17], and the eradication of multi-resistant organisms [18]. However, few studies on initiators that attack plant pathogens by inducing intracellular ROS have been published. Thus, discovering antimicrobial agents that cause oxidative stress mechanisms in plant bacteria is an appealing method for addressing such bacteria.
Interestingly, natural products have some excellent advantages, such as biodegradability, extensive sources, and a low susceptibility to drug resistance [19][20][21]. Thus, developing new pesticides based on natural products is a sustainable agricultural disease treatment strategy. Moreover, using natural products as active leads for structural derivation to optimize their molecular structures is gradually becoming essential for developing green, efficient, and novel pesticides. 18β-glycyrrhetinic acid (18β-GA) and ursolic acid (UA) belong to the pentacyclic triterpenes, which are active ingredients of herbal medicine with various biological properties, such as antioxidant [22], antibacterial [23], anticancer [24], and anti-inflammatory [25]. Our previous work found that the title compounds obtained by the structural derivatization of 18β-GA and UA could lead to a reduction in bacterial growth and even death by causing an excessive accumulation of ROS in Xoo and Xanthomonas axonopodis pv. citri (Xac) [26][27][28]. In addition, among agrochemicals, the synthesis and biological activity research of compounds containing amide skeletons have been a hot spot in agrochemical research because of their wide range of biological activities, such as antibacterial [29], insecticidal [30], and herbicidal activities [31]. Based on these properties, an active fragment splicing approach was used to introduce amide groups into the 18β-GA and UA backbones to obtain a series of title compounds and to investigate whether they could induce bacterial death by causing an inducement of the ROS pathway ( Figure 1).
lished. Thus, discovering antimicrobial agents that cause oxidative stress mechanisms in plant bacteria is an appealing method for addressing such bacteria.
Interestingly, natural products have some excellent advantages, such as biodegradability, extensive sources, and a low susceptibility to drug resistance [19][20][21]. Thus, developing new pesticides based on natural products is a sustainable agricultural disease treatment strategy. Moreover, using natural products as active leads for structural derivation to optimize their molecular structures is gradually becoming essential for developing green, efficient, and novel pesticides. 18β-glycyrrhetinic acid (18β-GA) and ursolic acid (UA) belong to the pentacyclic triterpenes, which are active ingredients of herbal medicine with various biological properties, such as antioxidant [22], antibacterial [23], anticancer [24], and anti-inflammatory [25]. Our previous work found that the title compounds obtained by the structural derivatization of 18β-GA and UA could lead to a reduction in bacterial growth and even death by causing an excessive accumulation of ROS in Xoo and Xanthomonas axonopodis pv. citri (Xac) [26][27][28]. In addition, among agrochemicals, the synthesis and biological activity research of compounds containing amide skeletons have been a hot spot in agrochemical research because of their wide range of biological activities, such as antibacterial [29], insecticidal [30], and herbicidal activities [31]. Based on these properties, an active fragment splicing approach was used to introduce amide groups into the 18β-GA and UA backbones to obtain a series of title compounds and to investigate whether they could induce bacterial death by causing an inducement of the ROS pathway ( Figure 1).
This study further explores the potential of 18β-GA with UA for agricultural applications. A series of title molecules containing amide groups were prepared, and their antibacterial activities were evaluated in vitro and in vivo. Finally, a series of biochemical experiments investigated their possible antibacterial mechanism.   This study further explores the potential of 18β-GA with UA for agricultural applications. A series of title molecules containing amide groups were prepared, and their antibacterial activities were evaluated in vitro and in vivo. Finally, a series of biochemical experiments investigated their possible antibacterial mechanism.

Synthesis of Title Compounds
The previously reported method synthesized a series of novel 18β-GA amide derivatives and UA amide derivatives. The target compounds were designed and synthesized as shown in Schemes 1 and 2. In brief, those compounds were synthesized by acid-amine condensation [36]. Nuclear magnetic resonance ( 1 H NMR, 13 C NMR) and high-resolution mass spectrometry (HRMS) were used to confirm the structures of the title compounds (See Figure S1-S114). (The synthesis method of the title compounds are described in the supporting information).

Synthesis of Title Compounds
The previously reported method synthesized a series of novel 18β-GA amide derivatives and UA amide derivatives. The target compounds were designed and synthesized as shown in Schemes 1 and 2. In brief, those compounds were synthesized by acidamine condensation [36]. Nuclear magnetic resonance ( 1 H NMR, 13 C NMR) and high-resolution mass spectrometry (HRMS) were used to confirm the structures of the title compounds (See Figure S1-S114). (The synthesis method of the title compounds are described in the supporting information).

Synthesis of Title Compounds
The previously reported method synthesized a series of novel 18β-GA amide derivatives and UA amide derivatives. The target compounds were designed and synthesized as shown in Schemes 1 and 2. In brief, those compounds were synthesized by acidamine condensation [36]. Nuclear magnetic resonance ( 1 H NMR, 13 C NMR) and high-resolution mass spectrometry (HRMS) were used to confirm the structures of the title compounds (See Figure S1-S114). (The synthesis method of the title compounds are described in the supporting information).

Results of Anti-Xoo and Anti-Xac of the Title Compounds In Vitro
The preliminary antibacterial activity test results (Table 1) showed that most studied compounds were less effective anti-Xoo and anti-Xac agents. However, compounds A 7 , A 19 , and A 20 led to the effective inhibition of the above two phytopathogenic bacteria at 100 and 50 mg L −1 , respectively. In addition, the literature studies have shown that due to the exposure of the piperazine and piperidine groups, NH favors antimicrobial activity, for example, as seen in the structure of xacin [37]. Therefore, based on the above antimicrobial activity test results and the literature research, compounds in the -Boc groups were considered to have improved antibacterial activities (Schemes 3 and 4). The results are shown in Table 2, and the antibacterial activities of the compounds obtained after -Boc had all increased and showed excellent activity. The effective concentrations of 50% maximum antibacterial activity against Xoo and Xac (EC 50 ) were further determined based on the title compounds with initial screening inhibition rates greater than 50%. The results (Table 3) showed that compound A 22 exhibited optimal antibacterial effects with EC 50 values of 3.34 and 3.30 mg L −1 , respectively, which were significantly better than the positive control bismerthiazol (BT, anti-Xoo: EC 50 = 28.34 mg L −1 , anti-Xac: EC 50 = 101.64 mg L −1 ) and thiodiazole copper (TC, anti-Xoo: EC 50 = 119.87 mg L −1 , anti-Xac: EC 50 = 86.54 mg L −1 ). This finding suggests that amines containing piperazine and piperidine rings can significantly promote the antibacterial activity of 18β-GA and UA after the removal of the -Boc group. Their corresponding structure-activity relationship (SAR) was analyzed for A-series title compounds: (1) the title compounds containing a piperidine ring showed a better inhibitory activity against Xoo and Xac than those containing piperazine rings; (2) the studied compounds esterified at the C-3 position hydroxyl group of 18β-GA showed a reduced inhibitory activity against Xoo and Xac, suggesting that the C-3 position hydroxyl group may be a crucially active group; and (3) the antibacterial activity of the target compounds with NH exposed on the piperidine and piperazine rings was superior to that of compounds without NH groups, such as compounds A 22 (anti-Xoo: EC 50

ROS Fluorescence Imaging
It has been reported that pentacyclic triterpenoid derivatives often inhibit the growth of cancer cells and are dependent on ROS activity [38,39]. However, their mechanism of action in the field of pesticides has rarely been reported. Therefore, we chose compound A 22 to investigate whether it could induce intracellular ROS accumulation and cause oxidative damage. We used 2 ,7 -dichlorofluorescein diacetate (DCFH-DA) fluorescence analysis to measure the level of ROS production. The results are shown in Figure 2, in which ROS production showed a significant increase in a concentration-dependent manner, and the fluorescence intensity of Xoo cells gradually increased with increasing concentrations of compound A 22 (12.5 and 25 mg L −1 ) compared to the control (0 mg L −1 ). This finding suggests that compound A 22 can interfere with the balance of the oxidation-reduction system of the Xoo cell and induce ROS overproduction, thus causing oxidative stress and cell damage.

CAT and SOD Enzyme Activity Assay Results
Catalase and superoxide dismutase (CAT and SOD, respectively) are the two major antioxidant enzymes in living organisms and are the most direct enzymes for scavenging ROS [40]. As shown in Figure 3, CAT and SOD enzyme activities showed a significant decrease after the action of compound A 22 (50 mg L −1 , 25 mg L −1 ) compared to the control (0 mg L −1 ). Therefore, this finding further suggests that compound A 22 interfered with the oxidoreductase activity and disturbed the redox state in Xoo cells, causing a slight amount of oxidative damage. ance of the oxidation-reduction system of the Xoo cell and induce ROS overproduction, thus causing oxidative stress and cell damage.

CAT and SOD Enzyme Activity Assay Results
Catalase and superoxide dismutase (CAT and SOD, respectively) are the two major antioxidant enzymes in living organisms and are the most direct enzymes for scavenging ROS [40]. As shown in Figure 3, CAT and SOD enzyme activities showed a significant decrease after the action of compound A22 (50 mg L −1 , 25 mg L −1 ) compared to the control (0 mg L −1 ). Therefore, this finding further suggests that compound A22 interfered with the oxidoreductase activity and disturbed the redox state in Xoo cells, causing a slight amount of oxidative damage.

Lipid Peroxidation
One of the hazards of ROS accumulation is its triggering or exacerbating cell membrane lipid peroxidation, the induction of which plays a crucial role in cell death [41]. Malondialdehyde (MDA) is a highly reactive compound produced during lipid peroxidation under conditions associated with oxidative stress and can reflect the extent of cellular peroxidative damage [42]. As shown in Figure 4, the MDA content of compound A22 was 2-3-fold higher at high drug concentrations (25 mg L −1 and 50 mg L −1 ) compared to the control (0 mg L −1 ), and MDA production corresponded to ROS levels, indicating that compound A22 can induce excessive ROS production, accelerate lipid peroxidation, and eventually lead to bacterial death.

Lipid Peroxidation
One of the hazards of ROS accumulation is its triggering or exacerbating cell membrane lipid peroxidation, the induction of which plays a crucial role in cell death [41]. Malondialdehyde (MDA) is a highly reactive compound produced during lipid peroxidation under conditions associated with oxidative stress and can reflect the extent of cellular peroxidative damage [42]. As shown in Figure 4, the MDA content of compound A 22 was 2-3-fold higher at high drug concentrations (25 mg L −1 and 50 mg L −1 ) compared to the control (0 mg L −1 ), and MDA production corresponded to ROS levels, indicating that compound A 22 can induce excessive ROS production, accelerate lipid peroxidation, and eventually lead to bacterial death.

Lipid Peroxidation
One of the hazards of ROS accumulation is its triggering or exacerbating cell membrane lipid peroxidation, the induction of which plays a crucial role in cell death [41]. Malondialdehyde (MDA) is a highly reactive compound produced during lipid peroxidation under conditions associated with oxidative stress and can reflect the extent of cellular peroxidative damage [42]. As shown in Figure 4, the MDA content of compound A22 was 2-3-fold higher at high drug concentrations (25 mg L −1 and 50 mg L −1 ) compared to the control (0 mg L −1 ), and MDA production corresponded to ROS levels, indicating that compound A22 can induce excessive ROS production, accelerate lipid peroxidation, and eventually lead to bacterial death.

H2O2 Plate Assay Results
It has been shown that oxidative stress can be triggered intracellularly by adding exogenous hydrogen peroxide (H2O2) [43,44]. Therefore, we investigated whether adding exogenous H2O2 would affect the bactericidal efficiency of compound A22. Results are shown in Figure 5 in which bacterial growth inhibition was not significant in the Petri dishes containing only H2O2 or compound A22. However, interestingly, in the Petri dishes containing a mixture of H2O2 and compound A22, we observed that the fungus cake shrank and became lighter in color. It is possible that the addition of exogenous H2O2 caused an acceleration of the burst of oxidative stress induced by compound A22 in Xoo and accelerated a more rapid bacterial death.

H 2 O 2 Plate Assay Results
It has been shown that oxidative stress can be triggered intracellularly by adding exogenous hydrogen peroxide (H 2 O 2 ) [43,44]. Therefore, we investigated whether adding exogenous H 2 O 2 would affect the bactericidal efficiency of compound A 22 . Results are shown in Figure 5 in which bacterial growth inhibition was not significant in the Petri dishes containing only H 2 O 2 or compound A 22 . However, interestingly, in the Petri dishes containing a mixture of H 2 O 2 and compound A 22 , we observed that the fungus cake shrank and became lighter in color. It is possible that the addition of exogenous H 2 O 2 caused an acceleration of the burst of oxidative stress induced by compound A 22 in Xoo and accelerated a more rapid bacterial death.

Cell Membrane Damage Detection Results
The generation of intracellular oxidative stress can modify phospholipids and proteins on the membrane through a process involving peroxidation and lead to changes in membrane permeability and the disruption of the phospholipid bilayer, which eventually result in cell death [45]. Among bacteria, the outer membrane of Gram-negative

Cell Membrane Damage Detection Results
The generation of intracellular oxidative stress can modify phospholipids and proteins on the membrane through a process involving peroxidation and lead to changes in membrane permeability and the disruption of the phospholipid bilayer, which eventually result in cell death [45]. Among bacteria, the outer membrane of Gram-negative bacteria consists of lipopolysaccharides, phospholipids, outer membrane proteins, and lipoproteins [46]. The outer membrane is the primary physical barrier for bacteria and is closely associated with bacterial pathogenicity and drug resistance [47]. Therefore, the effect of compound A 22 on the permeability of the Xoo outer membrane was determined using the hydrophobic probe, N-phenyl-1-naphthylamine (NPN). The results are shown in Figure 6A. The fluorescence intensity was significantly enhanced as the compound concentration increased, which indicated that compound A 22 effectively destroyed the outer membrane of the Xoo cells. Because the propidium iodide (PI) dye does not penetrate live bacterial cell membranes, it then binds to DNA. It emits fluorescence when cell membrane damage occurs, and the increased fluorescence is measured as an indicator of cell membrane permeability [48]. Therefore, to study the inhibitory effect of compound A 22 on the Xoo cell membrane, Xoo was incubated with different concentrations of compound A 22 and fluorescent nucleic acid dye PI. The results are shown in Figure 6B in which the fluorescence intensity gradually increased in a drug concentration-dependent manner, indicating that compound A 22 could disrupt the permeability of the Xoo cell membrane. The above results reveal that compound A 22 can cause damage to cell membranes and a loss of activity and eventually lead to bacterial death.

In Vivo Anti-Xoo Effect of Compound A22
The aim of our work was to demonstrate that the title compounds exert ex antimicrobial efficacy in plants; therefore, we validated them using pot experimen ble 4 and Figure 7 show that compound A22 exhibited excellent in vivo antibacte tivity against the rice bacterial leaf blight. Its therapeutic activity was 44.74%, protective activity was 50.44%; both superior to the control BT (curative activity: 3 protective activity: 40.87%) and TC (curative activity: 38.60, protective activity: 3 In particular, the control effect was found to significantly improve after the addi 0.1% (v/v) organosilicon (OSi) or orange peel essential oil (OPO) additives. The c activity was 59.65% (A22-OPO) and 60.53% (A22-OSi) against the rice bacterial leaf Furthermore, the toxicity results (See Figure S115) of title compound A22 were even non-toxic toward rice plants.

In Vivo Anti-Xoo Effect of Compound A 22
The aim of our work was to demonstrate that the title compounds exert excellent antimicrobial efficacy in plants; therefore, we validated them using pot experiments. Table 4 and Figure 7 show that compound A 22 exhibited excellent in vivo antibacterial activity against the rice bacterial leaf blight. Its therapeutic activity was 44.74%, and its protective activity was 50.44%; both superior to the control BT (curative activity: 39.47%, protective activity: 40.87%) and TC (curative activity: 38.60, protective activity: 37.39%). In particular, the control effect was found to significantly improve after the addition of 0.1% (v/v) organosilicon (OSi) or orange peel essential oil (OPO) additives. The curative activity was 59.65% (A 22 -OPO) and 60.53% (A 22 -OSi) against the rice bacterial leaf blight. Furthermore, the toxicity results (See Figure S115) of title compound A 22 were low or even non-toxic toward rice plants.

In Vivo Anti-Xoo Effect of Compound A22
The aim of our work was to demonstrate that the title compounds exert excellent antimicrobial efficacy in plants; therefore, we validated them using pot experiments. Table 4 and Figure 7 show that compound A22 exhibited excellent in vivo antibacterial activity against the rice bacterial leaf blight. Its therapeutic activity was 44.74%, and its protective activity was 50.44%; both superior to the control BT (curative activity: 39.47%, protective activity: 40.87%) and TC (curative activity: 38.60, protective activity: 37.39%). In particular, the control effect was found to significantly improve after the addition of 0.1% (v/v) organosilicon (OSi) or orange peel essential oil (OPO) additives. The curative activity was 59.65% (A22-OPO) and 60.53% (A22-OSi) against the rice bacterial leaf blight. Furthermore, the toxicity results (See Figure S115) of title compound A22 were low or even non-toxic toward rice plants.  The Tukey's HSD test was used to perform one-way ANOVA analysis between the different treatments; different uppercase letters indicate the values of control efficiency with a significant difference among different treatment groups at a level of p < 0.05.

In Vivo Anti-Xac Effect of Compound A 22
As shown in Table 5 and Figure 8, compound A 22 exhibited excellent in vivo antibacterial activity against Xac. Its curative activity was 58.86%, and its protective activity was 84.37%, which were superior to the control drug TC (curative activity: 51.51%, protective activity: 74.44%). In summary, the compound A 22 exhibited excellent antibacterial activity in vitro and a good control effect in vivo. The Tukey's HSD test was used to perform one-way ANOVA analysis betwee different treatments; different uppercase letters indicate the values of control effici with a significant difference among different treatment groups at a level of p < 0.05.

In Vivo Anti-Xac Effect of Compound A22
As shown in Table 5 and Figure 8, compound A22 exhibited excellent in vivo bacterial activity against Xac. Its curative activity was 58.86%, and its protective act was 84.37%, which were superior to the control drug TC (curative activity: 51.51%, tective activity: 74.44%). In summary, the compound A22 exhibited excellent antibac activity in vitro and a good control effect in vivo.

Instruments and Chemical Substances
The JEOL-ECX-500 spectrometer (500 MHz, JEOL Ltd., Tokyo, Japan) or a Bruker Biospin AG-400 spectrometer (400 MHz, Bruker Optics, Ettlingen, Germany) was used to measure the 1 H NMR and 13 C NMR with tetramethylsilane (TMS) as an internal standard and deuterated chloroform as the solvent. High-resolution mass spectrometry (HRMS) data were obtained on a Q-Exactive Orbitrap MS apparatus (Thermo Fisher Scientific, Waltham, MA, USA). Olympus BX53 microscope (Olympus, Tokyo, Japan) and a FluoroMax ® -4P fluorescence spectrophotometer (HORIBA Scientific, Paris, France) tested bacterial fluorescence images and fluorescence intensity, respectively. Bacteria enzymatic activity test and in vitro antibacterial test were performed using a Cytation™ 5 multi-mode readers (BioTek Instruments, Inc., Winooski, VT, USA). 18β-GA and UA were the starting material purchased from Energy-Chemical (Anhui Zesheng Technology Co., Ltd., Anhui, China).

General Protocols
The wild-type (WT) Xanthomonas oryzae pv. oryzae (Xoo) strain ZJ173 was kindly provided by Ming-Guo Zhou (Nanjing Agricultural University, Nanjing, China). The Xanthomonas axonopodis pv. citri (Xac) strain was preserved in our lab. In vitro and in vivo antimicrobial bioassays of the title molecules and measurement of enzyme activities were performed according to previously reported methods [49][50][51][52].

ROS Detection Assay
The ability of compound A 22 to affect redox homeostasis in Xoo cells was assessed by measurements of ROS levels. The ROS assay kit (S0033, Beyotime, Shanghai, China) was used to evaluate Xoo cellular ROS levels. First, plant pathogens were cultured to logarithmic growth (OD 595 = 0.2) and then treated with compound A 22 (0, 6.25, 12.5, and 25.0 mg L -1 ) for 12 h at 28 • C. Next, bacterial cells were harvested by centrifugation (6000 rpm, 3.5 min, and 4 • C), washed twice with sterile water, and resuspended. Finally, 100 µL of the bacterial solution containing 1 µL of S0033 staining solution was incubated for 20 min at 28 • C and further observed under an Olympus BX53 microscope.

Lipid Peroxidation Assay
Malondialdehyde (MDA) is living organisms' natural lipid oxidation product [53]. Some fatty acids are oxidized and gradually degraded into complex compounds, including MDA [42]. Therefore, MDA levels can be detected as an indicator of lipid peroxidation. Xoo cells were collected and washed with precooled phosphate-buffered saline (PBS) at a pH of 7.2. After that step, bacteria were resuspended in PBS and were subject to ultrasonication (SONICS ultrasonic crusher VCX150 (Newtown, CT, USA), power 150 W (30%), sonication 3 s, interval 10 s, repeated 60 times at 0 • C). The supernatant was centrifuged at 8000× g, 4 • C for 10 min, and the MDA content was measured according to the kit instructions (S0131S, Beyotime, Shanghai, China). Protein concentrations were determined using the Bradford method. The optical density (OD) value was measured using an enzyme marker.

Plate Assay
The hydrogen peroxide (H 2 O 2 ) plate assay was based on the method reported in the literature with some modifications [54,55]. Nutrient agar (NA) containing 15 g L −1 agar powder, H 2 O 2 (0.05 mM) and compound A 22 (12.5 mg L −1 ), NA with H 2 O 2 (0.05 mM) and compound A 22 (12.5 mg L −1 ), and a blank control were prepared. A 10 µL sample of bacterial solution was spread on the NA Petri dish, covered in a 28 • C incubator, incubated for 48 h, after which the size and growth of the bacterial cake were observed and photographs were taken. Three replicates were performed for each treatment, and the experiment was repeated three times.

Assay of Outer Membrane Damage
The effect of the title compound A 22 on the Xoo cell outer membrane was investigated according to the method reported in the literature with some modifications [56]. First, compound A 22 was added to bacteria at different concentrations (6.25, 12.5, 25.0, and 50.0 mg L −1 ) for incubation. Next, Xoo were centrifuged (6000 rpm, 3.5 min, 4 • C), washed, and resuspended in PBS to obtain a Xoo cell resuspension. N-phenyl-1-naphthylamine (NPN) dye (10 M, 50 µL) was added to the resulting resuspension solution and incubated at 28 • C for 30 min. After incubation, the bacterial suspension was removed and washed twice with PBS. Finally, the fluorescence intensity was measured using a fluorescence spectrophotometer (λ ex = 350 nm, λ em = 415 nm). Three parallel experiments were conducted for each group.

Study of Inner Membrane Permeabilization
The cell membrane permeability experiment was performed according to the previously reported method with slight modifications [57]. Xoo cells (OD 595 = 0.2) were cultured with different doses of A 22 (6.25, 12.5, 25.0, and 50.0 mg L −1 ) and then collected (6000 rpm, 3.5 min, 4 • C) and washed twice with PBS. After that step, propidium iodide (PI, 20 µM, 10 µL) was added to stain the bacteria, and fluorescence intensity was tested using a fluorescence spectrophotometer (λ ex = 535 nm, 5 nm slit width; λ em = 617, 5 nm slit width). Three parallel experiments were performed for each group.

Conclusions
This paper describes the synthesis of 18β-GA and UA derivatives, which were screened for their antibacterial activity against two phytopathogenic bacteria. Compound A 22 showed the best activity against Xoo and Xac with EC 50 values of 3.34 mg L −1 and 3.30 mg L −1 , respectively. Compound A 22 was then used to investigate the antimicrobial mechanism of action of Xoo. First, the assay results, such as increased ROS levels, decreased CAT and SOD enzyme activities, and increased MDA content, indicate that compound A 22 could disrupt the redox system in Xoo cells, trigger oxidative stress, and cause lipid peroxidation. Second, NPN and PI staining revealed that compound A 22 also triggered Xoo cell outer membrane damage and increased its permeability, ultimately leading to bacterial death. Finally, we demonstrated the excellent antibacterial activity of compound A 22 both in vitro and in vivo using pot tests. In summary, the novel 18β-GA and UA derivatives can be used as lead compounds for pesticide development by inducing an ROS burst.