*2.1. Effect of Alkaloids* **1**–**8** *on Noncyclic Electron Transport and H+-ATPase Activity*

Compounds **2** and **3** did not present an effect on noncyclic electron transport in preliminary tests. On the other hand, the other alkaloids inhibited noncyclic electron transport from H2O to methylviologen (MV) in chloroplasts isolated from *Spinacea oleracea* L. Arborinine (**1**) inhibited phosphorylating and uncoupled electron flow by 100% at 100 μM, which demonstrated that (**1**) behaves as a potent electron-transport inhibitor (Figure 2A). The basal electron flow was increased at low concentrations (around 15 μM), but electron flow at concentrations higher than 25 μM was decreased, inhibiting electron flow by 20% at 100 μM, which means that (**1**) binds to the CF1CF0-ATP*ase* complex, suggesting inhibitory activity on ATP synthesis. The results found, with regard to electron-transport reaction, a increment of the step as well as a decrease in the phosphorylating and uncoupled steps, indicating that (**1**) exhibited a dual effect by inhibiting both energy transfer and electron transport [17].

Compound **4** increased basal and phosphorylating electron transports by 80% and 40%, respectively, at the beginning of the illumination, and then decreased them, since the concentrations were higher than 80 μM (Figure 2B). As well as Compound (**1**), (**4**) decreased the uncoupled phase at concentrations close to 80 μM. Therefore, (**4**) did not demonstrate electron-transport inhibition, but rather acted as a decoupling agent. Graveoline (**5**) inhibited the basal, phosphorylating, and uncoupled electron transport by 40% at 300 μM, which suggested Hill reaction inhibitory behavior (Figure 2C).

Homolog mixture **6**–**8** inhibited energy transfer at 25 μM and showed slight inhibitory activity on electron-transport reactions at concentrations up to 100 μM (Figure 2D). Compounds **6**–**8** increased basal and phosphorylating electron transport by 230% and 140%, respectively. The uncoupled electron transport showed a small increase in concentrations below 100 μM. In this way, the mixture behaved mainly as an energy-transfer inhibitor and showed electron-transport inhibitory activity at higher concentrations.

**Figure 2.** Effect of the alkaloids isolated from *R. graveolens* on electron flow. Control-rate values for electron transport from basal, phosphorylating, and uncoupled conditions were 450, 620, and 1200 μequiv e<sup>−</sup> h−<sup>1</sup> mg−<sup>1</sup> chlorophyll (Chl)−1, respectively. Panel (**A**): Compound **1**; Panel (**B**): Compound **4**; Panel (**C**): Compound **5**; and Panel (**D**): Mixtures **6**–**8**.

When there is a significant increase on the basal electron-transport step, as observed for Compounds **1**, **4**, and **6**–**8**, this is an indication that the compounds are acting on the ATP–synthase complex [17]. Cyclic electron transport is happening normally, as can be observed in the basal reaction, due the behavior of the chloroplasts in the reaction medium. The percentage of the basal curve means that the effect is happening over the ATP–synthase complex once the basal reaction works harder to equilibrate this damage, thus increasing the speed of action.

Due to this, ATP*ase* analysis for Compounds **1**, **4**, and **6**–**8** was needed to confirm if they interfere on the CF1CF0-ATP*ase* complex, acting by direct inhibition of ATP synthesis. The experiments (Table 1) revealed that Compound **4** binds to CF1CF0-ATP*ase* complex exerting a direct inhibition of the H+ gradient dissipation and the Compounds **1** and **6**–**8** act as energy transfer inhibitors (H+-ATP*ase* inhibitor) [17].

The electron-transport increase on the basal reaction up to 100% indicates that the compounds acted on the ATP–synthase complex, blocking the energy transfer or acting as proton-transfer decoupling. This behavior was observed for Compounds **1** and **6**–**8** through the increase of the basal step for Compound **4** by the increment of the basal and phosphorylating reactions [18,19].

To confirm if Compounds **1** and **6**–**8** act as energy-transfer inhibitors, and if Compound **4** acts as a decoupling agent, we performed H+-ATP*ase* assays to verify their effect on the catalytic unit of the H+-ATP*ase* complex (CF0-CF1) [17]. Compounds **1** and **6**–**8** inhibited the energy transfer, as they decreased the inorganic phosphate (Pi) concentrations in the reaction medium by 25% at 100 μM and 300 μM, respectively. Corroborating the electron-transport data, both compounds are inhibitors of the CF0-CF1 enzymatic site of the ATPase complex (Table 1). In its turn, Compound **4** increased Pi concentration by 18% at 100 μM, which confirmed its proton-gradient uncoupling profile.


**Table 1.** Effect of Compounds **1**, **4**, and **6**–**8** on inorganic phospate (P*i*).

#### *2.2. Uncoupled PSI and PSII Electron-Flow Determination*

To localize the inhibition sites of the alkaloids on the thylakoid electron-transport chain, their effects on PSI and PSII (including partial reactions) were evaluated employing artificial donors and acceptors of electrons, as well as appropriate inhibitors [20]. Arborinine (**1**) inhibited uncoupled electron transport on PSII from water to DCBQ (from H2O to QB) and the partial reactions from water to sodium silicomolybdate (SiMo) (from H2O to QA) by 60% at 400 μM (Table 2). There were no significant results (<4%) for the reactions from DPC to 2,6-dichlorophenolindophenol (DCPIP) (from P680 to QB).

**Table 2.** Effects of arborinine (**1**) on photosynthetic electron transport on photosystem II (PSII). Note: DCPIP, 2,6-dichlorophenolindophenol.


*a* (μequiv e<sup>−</sup> h−<sup>1</sup> mg−<sup>1</sup> Chl−1), *b* (%), *c* (μM DCPIPred mg−<sup>1</sup> Chl−1).

The polarographic measures indicated that **1** inhibited the passage from H2O to QA, that is, on both sides of the electron transport on PSII. The first inhibition site (H2O to SiMo) occurs in the enzyme where water photo-oxidation happens, and the other at DPC (donates electron at P680) to DCPIPox (accepts electrons at QB site), located at the water-splitting enzyme complex (OEC) and between the range of electron flow from P680 to QA. These results indicated that **1** inhibited PSII at the span of electron transport from H2O to QA due the fact that SiMo accepts electrons exactly at the QA site. Table 2 shows that the span of electron transport from P680 to QB was not inhibited in all concentrations. Compound **1** inhibited the PSI uncoupled electron transport from reduced DCPIP to MV by 50% at 200 μM (Table 3). However, no changes were observed on inhibitory activity at higher concentrations.


**Table 3.** Effects of arborinine (**1**) on photosynthetic uncoupled electron transport at PSI.

*a* (μequiv e<sup>−</sup> h−<sup>1</sup> mg−<sup>1</sup> Chl−1) *b* (%).

#### *2.3. Chl a Fluorescence Measurements in Spinach Leaf Discs*

The Chl *a* fluorescence assay is a widely used tool to evaluate the photosynthetic apparatus in plants submitted to different stresses, as well as to provide detailed information about the structure and function of PSII [10,20]. For this experiment, all alkaloids were evaluated at 150 and 300 μM. Compounds **1**–**4** showed very low activity during the experiment, less than 20% compared to negative control (data not shown).

Compound **5** increased dV/dt0 and decreased PI*abs*, both parameters by 60% at 150 μM, which represents a stressful event occurring in the plant. The association of these parameters suggests that the natural redox process of photosynthesis was interfered with (Figure 3A). Parameters PSI0, PHI(E0), S*m*, ET0/CS0, and ET0/RC were reduced by 40%, which directly represents that electron transport on the redox process was interrupted, indicating damage to PSII. The decrease in S*m* demonstrates that not all absorbed energy was used, and then it was eliminated from the process. Energy dissipation was confirmed through the increase of the nonphotochemical "de-excitation" constant (K*n*) by 40% and the quantum yield (t = 0) of dissipation energy (PHI(D*o*)) by 20%. Thus, the energy contained in the system was released as heat or transferred to another molecule.

Compounds **6**–**8** were active at both concentrations during the leaf-disc fluorescence assay (Figure 3A,B). The PI*abs* parameter showed a decrease of 70% at 150 μM, indicating a nontraditional photosynthesis process. Parameters PSI0, PHI(E0), ET*0*/CS*<sup>0</sup>* and ET0/RC were reduced by 40, 40, 60 and 40%, respectively, at 150 μM. These decrements represent damage in electron transport on PSII, showing that the calculated quantum-yield values for the electron transport decreased in the process of the flux being inhibited. The reduction of ET0/CS0*,* ET0/RC, and RC/CS0 parameters by 30% indicated that the electron transport was being blocked, as well as a reduction in reaction centers participating in the process. Like Compound **5**, the increment was promoted by the mixture of analogs on the dV/dt0, Sm*K*, K*n*, and PHI(D0) parameters.

**Figure 3.** Radar plot of Compounds **5** and **6**–**8** effects on Chl *a* fluorescence parameters calculated from an *OJIP* transient curve. Panel (**A**) 150 μM, and Panel (**B**) 300 μM.

A *J* band (2 ms) was observed at the *OJIP* transient curve for Compound **5** (150 μM), which indicates inhibition at the quinone level, on the acceptor side of PSII (Figure 4A). An increase at *J* step can be understood as evidence for reduced-form QA accumulation (QA−) due electron-transport deceleration beyond QA [21]. Since the PSII electron flux was inhibited, the maximum PSII microelectrons field carries less QA−. This aspect corroborated the reduction of the PSI0 and PHI(E0) quantum parameters. The results of the fluorescence emission on spinach-leaf discs confirm the in vitro electron transport results, which revealed Compound **5** acting as a Hill reaction inhibitor.

The same *J* band was observed when the mixture of quinolone alkaloids was submitted to the assays (Figure 4B), which confirms that Compounds **6**–**8** also behave like 3-(3,4-dichlorophenyl)- 1,1-dimethylurea (DCMU), inhibiting the acceptor side of the PSII [10]. The Chl *a* experiment also showed the appearance of the *I* band near 30 ms (Figure 4C), which exclusively refers to the efficiency of the quinone pool. This event indicates whether plastoquinones are active or not during the QA reduction process. When the *I* band is found in negative values (on the graph), it suggests that the QA pool is functioning excellently, and an increase in the *J* band is also observed, that is, this indicates that the interaction site is the reaction center (P680). The transient bands show exactly this effect (Figure 4A,B). Phase *I* appears when a dynamic equilibrium is reached between the reduction of the plastoquinone pool by the electron flow from the PSII and its oxidation due to PSI activity [22].

**Figure 4.** Panel (**A**) Appearance of the J-band in the presence of Compound **5** at 150 μM. Panel (**B**) Appearance of the J-band in the presence of Compounds **6**–**8** at 150 and 300 μM. Panel (**C**) Appearance of the I-band in the presence of **6**–**8** at 150 and 300 μM. Panel (**D**) Radar plot of Compounds **5** and **6**–**8** effects on Chl *a* fluorescence parameters calculated from OJIP curve of sprayed *Lolium perenne* plants after 72 h.

#### *2.4. In Vivo Assays: Chl a Fluorescence Determination in Intact L. Perenne Leaves*

The in vivo Chl *a* fluorescence experiment represents a powerful tool to evaluate the performance of the photosynthesis system in living plants without causing any damage to them [23]. To evaluate compound activity, solutions at 150 and 300 μM were sprayed on the leaves of *L. perenne* plants. However, only Compound **5** and the mixture **6**–**8** were tested because they presented the best results on a semivivo assay. After 24, 48, and 72 h of treatment, Chl *a* fluorescence transients were measured and the OJIP parameters were calculated employing Biolyser HP software.

Data showed that **5** and **6**–**8** on plants after 24 and 48 h were insignificant, but a small variation on photosynthetic parameters was observed after 72 h at 150 μM (Figure 4D). In short, the in vivo results are less significant than the results observed on the semivivo assay. We justify this because there are many natural obstacles that the compounds have to transcend to reach their target (the chloroplast), for example, cell walls and membranes [23].

#### *2.5. Dry Biomass Determination*

Dry biomass results were obtained using *L. perenne* plants 15 days after compound application. Compound **5** and the mixture **6**–**8** were evaluated at 150 and 300 μM. The other compounds were not tested, as they did not present any activity in the previously assays. DCMU, a herbicide, was used as positive control (Table 4). Fortunately, the best results were observed for the lowest concentration, 150 μM. Treatments **5** and **6**–**8** decreased dry biomass by 20% and 23%, respectively, compared to negative control. These are significant results since they behaved like DCMU, which reduced 23% of the biomass of the target plant.


**Table 4.** Dry biomass assay for Compounds **5** and **6**–**8**. Note: DCMU, 3-(3,4-dichlorophenyl)-1,1-dimethylurea.

Based on in vitro, semivivo, and in vivo approaches, Compound **5** acts as a photosynthetic electron-transport inhibitor and as a plant-growth regulator. Mixture **6**–**8**, on the other hand, acts as an electron-transport and energy-transfer inhibitor, as well as plant-growth regulator. Our results showed that almost all alkaloids behaved as photosynthesis inhibitors once some of them acted as Hill reaction inhibitors. Through fluorescence measurement, we could observe the presence of transient bands *J* and *I* (obtained from *OJIP*-test). These steps suggest that compounds isolated from *R. graveolens* inhibited electron flow on the acceptor side of PSII, exactly like DCMU does. Therefore, the aim of our work was to present that natural products still could be employed on programs to lead to new scaffold models for herbicides in the future, since natural products remains an interesting alternative to replace the commercial herbicides.

#### **3. Materials and Methods**

#### *3.1. Alkaloid Isolation from Ruta Graveolens*

The ethanolic extract (203.6 g) from *Ruta graveolens* leaves was solubilized in CH3OH:H2O (1:3, *v*:*v*) and extracted by liquid–liquid partition with hexane and dichloromethane to obtain the partitioned extract fractions.

The dichloromethane fraction (14.6 g) from *Ruta graveolens* leaves was subjected to a chromatographic column using silica gel 60 (70–230 mesh), employing as a mobile phase increasing hexane, dichloromethane, acetone, and methanol concentrations to obtain 6 fractions (1–6). Fraction 5 (0.746 g) was subjected to a new chromatographic procedure using Sephadex LH-20 with isocratic elution formed by dichloromethane:methanol (1:1, *v*:*v*) to afford arborinine (**1**, 18.4 mg) and 1,4-dihydroxy-2,3-dimethoxy-*N*-methylacridone (**2**, 16.5 mg) [24].

1-hydroxy-3-methoxy-*N*-methylacridone (**3**, 17.9 mg) was obtained from the dichloromethane:hexane fraction of *Ruta graveolens* leaves using silica gel (70–230 mesh) and solvents of increasing polarity (hexane, dichloromethane, acetone, and methanol), followed by a second chromatographic purification over Sephadex LH-20 with isocratic elution dichloromethane: methanol (1:1, *v*:*v*) [25].

From the methanol fraction of *Ruta graveolens* leaves, the *N*-methyl-4-methoxy-2-quinolone (**4**, 19.3 mg) and graveoline (**5**, 14.1 mg) compounds were purified using a chromatographic column with silica gel as support, and hexane, dichloromethane, acetone, and methanol as the mobile phase [26].

The ethanolic extract (7.0 g) from the *Ruta graveolens* roots was solubilized with methanol:water (1:3 *v*:*v*), and subjected to liquid–liquid extraction with hexane to provide the respective fraction (1.53 g). The hexanic fraction was subjected to purification using silica gel (70–230 mesh). The mobile phase was composed of increasing portions of hexane, dichloromethane, acetone, and methanol to obtain 8 fractions (I–VIII). Fraction II (0.105 g) was subjected to new chromatographic purification by Sephadex LH-20 with isocratic elution dichloromethane:methanol (3:7, *v*:*v*) to obtain a homolog mixture of **6**, **7**, and **8**. The mixture was analyzed with GC-MS. The instrument was set to an initial temperature of 150 ◦C, and maintained at that temperature for 1 min. At the end of this period, the oven temperature was increased to 300 ◦C, at the rate of 10 ◦C/min, and maintained for 20 min. The chromatogram presented 3 peaks at retention times (tR) 11.5 min (**6**, *m*/*z* 313), 12.0 min (**7**, *m*/*z* 327), and 13.0 min (**8**, *m*/*z* 341). Based on the GC-MS experiment, a ratio of 8:1:1 (based on the peak areas) was estimated for **6**, **7**, and **8 [27]**.

**Compound 1.** 1H-NMR (200 MHz, CDCl3) δ: 3.81 (s, 3H, *N*-Me), 3.92 (s, 3H, 3-OMe), 4.00 (s, 3H, 2-OMe), 6.23 (s, 1H, H-4), 7.23 (ddd, *J* = 8.0, 6.8 and 0.7 Hz, 1H, H-7), 7,50 (dl, *J* = 8.0 Hz, 1H, H-5), 7.73 (ddd, *J* = 8.0, 6.8 and 1.4 Hz, 1H, H-6), 8.42 (dd, *J* = 8.0 and 1.4 Hz, 1H, H-8), 14.75 (s, 1H, OH). 13C-NMR (100 MHz, CDCl3) δ: 34.1 (*N*-Me), 56.0 (C3-OMe), 60.8 (C2-OMe), 86.8 (C-4), 105.8 (C-9a), 114.5 (C-5), 120.8 (C-8a), 121.5 (C-7), 126.2 (C-8), 130.2 (C-2), 134.6 (C-6), 140.5 (C-4a), 142.0 (C-5a), 156.2 (C-1), 159.3 (C-3), 180.8 (C-9).

**Compound 2.** 1H-NMR (200 MHz, CDCl3) δ: 3.96 (s, 3H, *N*-Me), 3.99 (s, 3H, 2-OMe), 4.03 (s, 3H, 3-OMe), 7.32 (ddd, *J* = 8.0, 7.5 and 1.6 Hz, 1H, H-7), 7.48 (dl, *J* = 8.7 Hz, 1H, H-5), 7.76 (ddd, *J* = 8.7, 7.5 and 1.6 Hz, 1H, H-6), 8.36 (dd, *J* = 8.7 and 1,6 Hz, 1H, H-8), 14.69 (*s*, 1H, 1-OH). 13C-NMR (50 MHz, CDCl3) δ: 44.0 (*N*-Me), 61.0 (C2-OMe), 61.5 (C3-OMe), 109.4 (C-9a), 116.6 (C-5), 121.3 (C-8a), 122.1 (C-8), 126.2 (C-7), 134.6 (C-6), 134.7 (C-2), 140.0 (C-3), 146.1 (C-5a), 151.5 (C-4), 155.8 (C-1), 157.0 (C-4a), 182.3 (C-9).

**Compound 3.** 1H-NMR (200 MHz, CDCl3) δ: 3.77 (s, 3H, *N*-Me), 3.90 (s, 3H, OMe), 6.30 (s, 2H, H-2 and H-4), 7.30 (ddd, *J* = 8.0, 7.2 and 1.6 Hz, 1H, H-7), 7.40 (dl, *J* = 8.0 Hz, 1H, H-5), 7.73 (ddd, *J* = 8.0, 7.2 and 1.6 Hz, 1H, H-6), 8.44 (dd, *J* = 8.0 and 1.6 Hz, 1H, H-8), 14.75 (s, 1H, OH), 13C-NMR (50 MHz, CDCl3) δ: 33.3 (*N*-Me), 55.6 (OMe), 90.1 (C-4), 94.1 (C-2), 105.0 (C-9a), 114.4 (C-5), 121.0 (C-8a), 121.4 (C-7), 126.7 (C-8), 134.1 (C-6), 142.0 (C-5a), 144.0 (C-4a), 166.0 (C-1), 166.1 (C-3), 180.0 (C-9).

**Compound 4.** 1H-NMR (200 MHz, CDCl3) δ: 3.64 (s, 3H, *N*-Me), 3.92 (s, 3H, OMe), 6.23 (s, 1H, H-3), 7.96 (dd, *J* = 8.0 and 1.5 Hz, 1H, H-5), 7.21 (ddd, *J* = 8.0, 7.1 and 1.5 Hz, 1H, H-6), 7.34 (dl, *J* = 8.0 Hz, 1H, H-8), 7.58 (ddd, *J* = 8.0, 7.1 and 1.5 Hz, 1H, H-7). 13C-NMR (50 MHz, CDCl3) δ: 28.8 (*N*-Me), 55.3 (OMe), 96.1 (C-3), 113.8 (C-8), 116.2 (C-4a), 121.4 (C-6), 123.1 (C-5), 131.0 (C-7), 139.4 (C-8a), 162.4 (C-4), 163.6 (C-2).

**Compound 5.** 1H-NMR (400 MHz, CDCl3) δ: 3.59 (s, 3H, *N*-Me), 5.99 (s, 2H, H-7 ), 6.26 (s 1H, H-3), 6.78 (dd, *J* = 1.6 and 0.4 Hz, 1H, H-2 ), 6.81(dd, *J* = 8.0 and 1.6 Hz, 1H, H-6 ), 6.84 (dd, *J* = 8.0 and 0.4 Hz, 1H, H5 ), 7.34 (ddd, *J* = 8.0, 6.8 and 1.6 Hz, 1H, H-6), 7.49 (dl, *J* = 8.4 Hz, 1H, H-8), 7.64 (ddd, *J* = 8.4, 6.8 and 1.6 Hz, 1H, H-7), 8.37(dd, *J* = 8.0 and 1.6 Hz, 1H, H-5).

**Compounds 6**–**8**. 1H-NMR (400 MHz, CDCl3) δ: 0.87 (t, *J* = 8.0 Hz, 3H, CH3-9 ), 1.26–2.37 (2H-2 to 2H-8 , overlapping with the signals of **7** and **8**), 3.05 (qt, *J* = 8.0 Hz, 2H, H-1 ), 4.12 (s, 3H, *N*-Me), 6.70 (s, 1H, H-3), 7.54 (tl, *J* = 8.0 Hz, 1H, H-8), 7.77 (tl, *J* = 8 Hz, 1H, H-6), 8.19 (tl, *J* = 8.0 Hz, 1H, H-5), 8.19 (tl, *J* = 8.0 Hz, 1H, H-7).
