Extraction optimization and reactivity of 7α-acetoxy-6β-hydroxyroyleanone and ability of its derivatives to modulate PKC isoforms

Protein kinase C is a family of kinases that play important roles in carcinogenesis. Medicinal plants from Plectranthus spp. (Lamiaceae) are a well-known source of interesting abietanes, such as 7α-acetoxy-6β-hydroxyroyleanone (Roy). This study aimed to extract and isolate Roy from P. grandidentatus Gürke, comparing two extraction methods (CO2 supercritical and ultrasound-assisted acetonic extraction), and design new royleanone derivatives for PKC modulation focusing on breast cancer therapy. The concentration of Roy in the extracts was determined by HPLC–DAD. The supercritical extraction method yielded 3.6% w/w, with the presence of 42.7 μg mg−1 of Roy (yield of 0.13%), while ultrasound-assisted acetonic extraction yielded 2.3% w/w, with the presence of 55.2 μg mg−1 of Roy (yield of 0.15%). The reactivity of Roy was investigated aiming at synthetizing new ester derivatives through standard benzoylation and esterification reactions. The benzoylated (Roy-12-Bz) and acetylated (Roy-12-Ac) derivatives in the C12 position were consistently prepared with overall good yields (33–86%). These results indicate the 12-OH position as the most reactive for esterification, affording derivatives under mild conditions. The reported di-benzoylated (RoyBz) and di-acetylated (RoyAc) derivatives were also synthesized after increasing the temperature (50 °C), reaction time, and using an excess of reagents. The cytotoxic potential of Roy and its derivatives was assessed against breast cancer cell lines, with RoyBz emerging as the most promising compound. Derivatization at position C-12 did not offer advantages over di-esterification at positions C-12 and C-6 or over the parent compound Roy and the presence of aromatic groups favored cytotoxicity. Evaluation of royleanones as PKC-α, βI, δ, ε, and ζ activators revealed DeRoy’s efficacy across all isoforms, while RoyPr showed promising activation of PKC-δ but not PKC-ζ, highlighting the influence of slight structural changes on isoform selectivity. Molecular docking analysis emphasized the importance of microenvironmental factors in isoform specificity, underscoring the complexity of PKC modulation and the need for further exploration.

including apoptosis, survival, differentiation, proliferation, and migration 3 .PKC family includes 10 isoforms, encoded by nine functionally and structurally related genes, that are classified as classical (α, βI, βII and γ), novel (δ, ε, η, φ), and atypical (ζ and λ\ι) PKCs 3,4 .PKC isoforms are composed of a conserved N-terminal regulatory region (C1 and C2 domain), and a C-terminal catalytic region (ATP binding and phosphotransferase activity) 5 .PKCs are activated by endogenous calcium, diacylglycerol (DAG), phospholipids and/or phorbol esters (PS) 6 .PS bind to the C1 domain and represent a competitive ligand for the physiological substrate DAG 7 .Conversely, atypical PKCs are structurally and functionally distinct from other PKCs, with a single C1 domain unresponsive to DAG/phorbol and lacking a C2 domain 8 .Nonetheless, PKC-ζ can be activated by lipid components such as arachidonic acid and ceramide 9 .It is generally accepted that most PKC activators bind to the regulatory domain 3 .Thus, the C1 domain has become the most attractive target for the design of selective PKC activators 6,7,10 .On the other hand, PKC inhibitors may interact with PKC at different sites of the protein [11][12][13][14][15] .
Members of the PKC family were originally categorized as oncogenes; however, individual isoforms exhibit context-dependent activities and may have distinct or even opposite roles in the regulation of cellular processes involved in cancer development.This complexity poses challenges in designing selective PKC modulators 8,16,17 .Moreover, PKCs share high sequence homology and structural similarity among themselves and with other kinases 2,8,10 , leading to limited availability of selective modulators 5,6 .In this context, PKC-α has deserved increased attention in breast cancer research.Its role is intricate, as it can function as both an anti-proliferative agent and a growth stimulant depending on the context 18 .While PKC-α generally protects cancer cells 2 , studies have shown that its activation with phorbol 12-myristate 13-acetate (PMA) inhibits epidermal growth factorinduced cell spreading and chemotaxis in MDA-MB-231 cells 19 .Additionally, overexpression of PCK-α in MCF-7 breast cancer cells alters the expression of other PKC isoforms and decreases estrogen receptor (ER) levels 8,20 , potentially contributing to the switch from ER-positive to ER-negative state 20 .On the other hand, downregulation of PKC-α is associated with invasive ductal carcinoma 21 and is observed in epithelial cells of advanced breast tumors 22 .

Roy quantification by HPLC-DAD
The amount of Roy in acetonic and CO 2 supercritical extracts was quantified by HPLC-DAD.The HLPC method used had been previously optimized by Bernardes et al. 24 Extracts were analyzed at a concentration of 5 mg mL −1 .One aliquot of 20 µL of each extract was injected in triplicate and the amount of Roy was calculated with resource to one linear calibration curve.In order to detect Roy with good sensitivity, 270 nm was selected as the detecting wavelength for the analysis, according to the λ max of Roy UV spectra.Temperature was not controlled.After settling the analytical conditions, the HPLC method was evaluated for several validation characteristics such as specificity, linearity, limits of detection (LOD), and quantification (LOQ).Specificity was demonstrated by the peak purity of Roy.No interferences were observed at Roy's retention time, indicating that the peak corresponded to pure Roy.Linearity was determined through the analyses of four concentrations of the analyte, and the correlation between observed peak areas and respective concentration afforded the linear calibration graphic.Regression analyses showed good linearity, with R 2 of 0.99.The retention time (Rt), regression equation, correlation coefficient (R 2 ), and LOD and LOQ values, obtained for Roy are presented in Table 1.
The identification of Roy´s peaks in each chromatogram of the extract was achieved through the retention time and the UV spectra analysis.The values obtained were expressed as means ± SD (at least n = 3).In this work, a total of 3.1 ± 0.05 g of Roy was calculated in both extracts.The CO 2 supercritical extract revealed the presence of 42.7 ± 0.7 μg mg −1 (yield of 0.13%), while the ultrasound-assisted acetonic extract showed 55.2 ± 1.0 μg mg −1 of Roy (yield of 0.15%).These results indicated that acetonic extraction was slightly more effective than the CO 2 supercritical extraction.Further, it has been taken under consideration that CO 2 supercritical extraction requires higher costs due to the need for CO 2 , dedicated equipment and training, making it more time-consuming and economically unfavorable.Conversely, acetonic extraction offers simplicity, speed, and higher yield, making it a more attractive option.Although acetonic extraction requires a significant amount of solvent, acetone can be recovered after evaporation and reused for other applications.
The amount of Roy present in the CO 2 supercritical extract was slightly lower than that reported by Bernardes et al., (57 μg mg −1 ) 24 .However, our study showed a significantly higher yield of Roy in ultrasound acetonic extract (55.2 μg mg −1 ) compared to the previous report (8.04 μg mg −1 ) 24 .This significative difference can be attributed to variations in the ultrasound extraction parameters.While Bernardes et al. reported a single ultrasound-assisted acetonic extraction of 60 min, our study employed three cycles of 30 min each.Therefore, a higher amount of Roy was extracted from the plant when multiple cycles of limited duration were performed.This optimization of the acetonic extraction method resulted in a higher extraction efficiency, demonstrating the importance of process optimization for maximizing yield.

Roy reactivity
The preparation of derivatives from royleanones has been well described, with ester derivatives proving to be stable 27,[33][34][35][36][37] .Hence, we investigated the esterification of Roy using benzoyl chloride (Table 2) and acetic anhydride (Table 3) as reagents.To determine the optimal reaction conditions, we explore several parameters, including the quantities of reagents, base, catalyst, temperature, and work up conditions (Fig. 2).Under the reaction conditions reported in Table 2, benzoylation of Roy always afforded two different products, namely, Roy-12-Bz and RoyBz.Both compounds were obtained as yellow powder, which tends to crystallize in n-hexane and whose structures were confirmed by NMR spectroscopy and compared with the literature 34,38,39 .Curiously, it was not possible to obtain the mono-benzoylated product in position C-6 (Fig. 2) following this protocol, and to our knowledge, the preparation of benzoylated derivative at C-6 has not been achieved.
The number of equivalents needed, and the addition rate of the reagent were evaluated.The reagent was portion wise (or dropwise) added to the reaction.Dropwise addition of 15 eq of benzoyl chloride led to the formation of Roy-12-Bz in 1 h, with 69% of isolated yield.The same conditions were tested at 0 °C, but no significant changes were observed.Conversely, when the excess of benzoyl chloride was added at once, the reaction was very fast, and all the starting material Roy was consumed in 5 min, affording the same derivative in 36% isolated yield.Heating up to 50 °C and using pyridine as the solvent facilitated the formation of Roy-12-Bz in 30 min, further confirming the high reactivity of C-12-OH towards the benzoylation.
For the successful preparation of RoyBz in good yields (50-79%), the reaction mixture was left to react overnight with an excess of all reactants at 50 °C.Lower yields (28%) were obtained when DMAP was added in presence of an excess of the reactants and heated up to 50 °C for 2 h.Similar reactivity was observed for the reaction with acetic anhydride.Both 7α-acetoxy-6β-hydroxy-12-Oacetylroyleanone (Roy-12-Ac) and 7α,6β-diacetoxy-12-O-acetylroyleanone (RoyAc) were successfully prepared, but the synthesis of the 6-OH derivative was not achieved (Fig. 2).Both compounds were isolated as yellow powder, with a tendency to crystallize in Hex.The structure of RoyAc was confirmed by NMR analysis and compared with literature data 35 .Roy-12-Ac has been synthesized for the first time in this study, and full structural characterization has been reported herein.www.nature.com/scientificreports/At room temperature, the derivative Roy-12-Ac was obtained in 15 min using one equivalent or small excess of acetic anhydride with overall good yields (65-77%).Adding one equivalent of acetic anhydride at 0 °C resulted in the total consumption of Roy within 15 min (as observed by TLC), but subsequent degradation occurred during the isolation process, yielding a small amount of starting material alongside a lower yield of Roy-12-Ac (57%).However, the use of a small excess of acetic anhydride at 0 °C led to the highest isolation yield for Roy-12-Ac (87%).On the other hand, to obtain RoyAc the reaction mixture was allowed to react overnight with an excess of all reactants at 50 °C.Further purification of the resulting crude allowed the isolation of two products, namely, monofunctionalized Roy-12-Ac (50% yield) and desired RoyAc (48% yield).
Optimization of the work-up conditions was attempted based on previous hemi-synthetic studies 35 .The treatment of the reaction mixture with a 5% solution of HCl, intended to neutralize the pyridine in the reaction environment, led to increased decomposition and consequently worse Roy-12-Ac isolation yield (33%).Stopping the reaction with a 5% solution of NaOH to quench residual acetic anhydride caused full decomposition of the product and cleavage of the acetoxy group in position C-7, resulting in the tri-hydroxylated product 6β,7αdihydroxyroyleanone (DiRoy), isolated in 91% yield as yellow powder, which tended to crystallize in Hex.The structure of DiRoy was confirmed by NMR spectroscopy and compared with the literature 35,40 .Therefore, treatment of the crude mixture with even weak acidic or alkaline solutions led to the degradation of the desired product through ester hydrolysis.
To explore breast cancer biology and screening for new anti-tumoral agents, widely used cell lines MCF-7 and MDA-MB-231/MDA-MB-468 offer distinct characteristics.MCF-7 cells, being ER-positive, PR-positive, and HER2-negative, exhibit slower growth and less aggressiveness.In contrast, triple-negative MDA-MB-231/ MDA-MB-468 cells are more aggressive, invasive, and possess higher metastatic potential 43 .The cytotoxic effects of Roy 25 , DiRoy 25 , RoyBz 42 , and RoyAc 42 against MCF-7 cell line were previous evaluated using the MTT colorimetric method.Overall, the results obtained in this study through the SRB assay remain previous.Roy displayed significant anti-proliferative activity against all cell lines tested (IC 50 of 2.09 µM, 0.30 µM, 0.35 µM and 1.02 µM on MCF-7, MDA-MB-231, MDA-MB-468, and HFF-1, respectively).A compound with cytotoxic effects on normal cells can have significant side effects and safety concerns 44 , thus chemical modifications can be crucial role to improve selectivity towards cancer cells.All derivates were active against the tested cancer cell lines with the analogue RoyBz emerging as the most promising derivative with lower IC 50 values (IC 50 of 1.66 µM, 1.50 µM and 1.40 µM on MCF-7, MDA-MB-231 and MDA-MB-468, respectively) and selectivity against cancer cell lines (IC 50 of 3.58 µM on HFF-1 cell line).Additionally, RoyBz was the only derivative to surpass the cytotoxic activity of the parent compound Roy in the cell line MCF-7.These results suggest that derivatization on position C-12 does not confer any advantage compared to derivatization at both positions or when compared to the parent compound Roy.Additionally, it appears that aromatic groups favor the cytotoxicity against the MCF-7 cells (IC 50 of 1.66 µM for RoyBz), compared to aliphatic chains (IC 50 values of 4.45 µM for RoyAc, and 28.3 µM for RoyPr 42 ).On the other hand, DiRoy, a royleanone structure with three hydroxyl groups, and DeRoy, the structure with a double bound in the positions C-6 and C-7, exhibit a significant decrease in the cytotoxic effects, as IC 50 values above 20 μM suggest a weak cytotoxicity 45 , supporting the hypothesis that the acetoxy group at position C-7 may play a crucial role in cytotoxic activity against breast cancer cell lines.

Effect of compounds on individual PKC isoforms
PKC isoforms are key targets for cancer therapy due to their significant roles in cancer development and progression.The selective-PKC-δ activation observed for RoyBz is well established 5 .Nonetheless, it is important to understand the effect of Roy on PKC isoforms.Due to the specificity of the assay, it was decided to evaluate only few compounds: the natural ones Roy and DeRoy and two derivatives, one aromatic ester RoyBz and  27 as PKC-α, -βI, -δ, -ε, and -ζ activators using the yeast PKC assay.The results can be consulted in Table 5.
For the determination of EC 50 values were considered the concentration of compound that caused 50% of the maximal growth inhibition caused by the positive controls (phorbol 12-myristate 13-acetate, PMA, for cPKCs and nPKCs; arachidonic acid, ARA, for PKCζ), which was set as 100%.Data are mean ± SEM of four independent experiments.nt, not tested; ND, non-determinable (when the maximal response achieved was lower than 50% growth inhibition).
When analyzing the results (Table 5), RoyBz recorded a notable selective activity towards the PKC-δ isoform 5 .On the other hand, Roy activated all PKC isoforms except for PKC-δ.However, the obtained EC 50 (half maximal effective concentration) values for this compound were not significant when compared to the controls used (PMA and ARA).Curiously, DeRoy displayed promising EC 50 values on all the isoforms with increased potential comparing to PMA and ARA.Although not selective for a specific isoform, its efficacy to activate isoforms from the three PKC sub-families makes it a valuable positive control for PKC evaluation assays.The introduction of propanoyl groups (RoyPr) in Roy, led to an overall decrease of the EC 50 values.Interestingly, RoyPr seems to be also an activator of PKC-δ, but not of PKC-ζ.In a similar manner to PMA, RoyPr only activates the classical and novel PKC isoforms.These results suggest that slight changes in the structure affect the selectivity to each isoform.These preliminary results promote further studies, with new esters, to better understand the selectivity to PKC isoforms.
To get more insights on the origins of selectivity and activity, a molecular docking strategy was used based on our previous results using AutoDock 5,46 that provided an explanation for the activity of RoyBz in PKC-δ, by showing similar interaction patterns between this molecule and 13-acetylphorbol (PRB; positive control; c.f. Figure 1 of Bessa et al. 5 ).To focus on breast cancer, the research was centered on PKC-α.Taking as example the PKC-δ and PKC-α isoforms, the difference in the PRB binding region is due to M239G, W252Y, V255I, and K256H.Although only the first two are immediately next to PRB, the sidechains face away.However, these mutations can probably change the lipid/water microenvironment around this binding site.Therefore, for a closer comparison, instead of using the 2ELI NMR structure (available in Protein Data Bank 47 ), the PKC-δ 1PTR X-ray crystal structure was mutated and reverted to the PKC-α isoform.
The top-ranking poses of RoyBz in both α and δ isoforms are depicted in Fig. 3.As observed, although both poses have the same OAc group inserted in the groove, the RoyBz pose in the variant has a benzoyl group  protruding with the carbonyl pointing to the middle of the membrane.Although the free energy of binding for RoyBz in PKC-α is reduced from − 7.87 to − 6.38 kcal mol −1 , it is not enough to explain the inactivity with the α isoform and selectivity towards δ (PRB has an interaction of − 7.54 kcal mol −1 with δ).However, according to the Orientations of Proteins in Membranes (OPM) database information 48 , the PRB molecule sits inside the molecular membrane on the cytoplasmic inner leaflet.Taking this into consideration and the fact that RoyBz has the benzoyl group located in a region inappropriate for favorable interactions with the lipid tails, and considering that these facts are not accounted in the docking evaluation, this could explain the differences found between variants.This means that not only the predicted docking score needs to be considered, but also the capacity of the molecules to partition in the bilayer and the possible changes in the nature of the amino acids producing also changes in the microenvironment hydrophilic/lipophilic balance.Overall, the antitumoral effect of Roy, its analogues, and DeRoy was investigated in breast cancer cells, with a focus on evaluating their impact on PKC isoforms, which are pivotal targets in cancer therapy.The study of these royleanones revealed significant cytotoxic activity against breast cancer cell lines.Roy exhibited potent anti-proliferative effects across all tested cell lines.The importance of selective cytotoxicity against cancer cells to minimize off-target effects was underscored by the superior activity of RoyBz, which demonstrated lower IC 50 values and greater selectivity towards cancer cells compared to normal fibroblasts.Furthermore, the activation profile of PKC isoforms by royleanone derivatives revealed distinct effects.RoyBz selectivity activated PKC-δ, a critical target in colon cancer therapy due to its role in cancer progression 5 .Roy, DeRoy and RoyPr exhibited varying activation profiles across multiple PKC isoforms.The introduction of propanoyl groups in RoyPr affect its activity in the cell MCF-7 line and altered its isoform selectivity, suggesting that structural modifications can influence both cytotoxicity and PKC activation profiles.
These findings underscore the complex interplay between chemical structure, cytotoxic efficacy, and isoformspecific PKC activation, guiding future research towards optimizing royleanone derivatives for targeted cancers that are positively influenced by the modulation of PKC isoforms, particularly in breast cancer treatment.Further exploration of new derivatives and their effects on PKC isoforms is warranted to refine their therapeutic potential and elucidate their specific mechanisms of action in cancer cells.This exploration may lead to the discovery of promising PKC modulators with high potency and isoform-selectivity within this compound family.

Extraction and isolation
Acetone ultrasonic-assisted extraction was adapted from Bernardes C.E.S. et al. 24 The leaves and steams of P. grandidentatus were air dried.2.334 kg of the dried plant was grinded to powder and with acetone (15 × 3.4 L), in an ultrasound equipment.The ultrasonic bath (Sonorex Super RK 510 H; Bandelin, Berlin, Germany) operated at room temperature, for 30 min, for 3 times, at 35 Hz with maximum input power of 320 W. Filtration and evaporation of the solvent (under vacuum, 40 °C) yielded a residue of 52.74 g (2.3% w/w).
Supercritical fluid extraction was carried out in an apparatus, equipped with a 500 cm 3 sample cell, which had been previously described by Pereira et al. 50a sample of 130.11 g of powder plant was extracted with supercritical CO 2 for 4 h at 40 °C and 230 bar, using a fixed CO 2 flow rate of 0.3 kg h −1 .The supercritical fluid extract was recovered by washing the collection vessel and tubing the expansion line with acetone.The solvent was subsequently removed in a rotary evaporator and yielded a residue of 4.67 g (3.6% w/w).
The isolation process was adapted from Bernardes et al. 24 The crude extract was subjected to sequential liquid and dry flash chromatographic separations.The liquid flash chromatographic column used silica gel (Merck 9385) as stationary phase and CH 2 Cl 2 as eluent.The dry flash chromatography used silica gel (Merck 9385) as stationary phase and mixtures of Hex: AcOEt and AcOEt: MeOH as eluents.The fractions obtained were compared with one sample of Roy by TLC (Eluent Hex: AcOEt 90:10, 80:20, or 70:30).Roy was obtained from recrystallization from Hex.

Quantification of Roy by HPLC-DAD
The quantification of Roy in P. grandidentatus extracts was adapted from Matias et al. 2019 32 .The quantification was performed using HPLC-DAD and complementary spectroscopic methodologies.The quantification of the identified compound was carried out in a Dionex Ultimate 3000 UHPLC system with diode array detector (DAD; Thermo Fisher Scientific Inc. MA, USA), equipped with a Nucleodur 100-5 C18ec, (250 × 4.0 mm i.d., 5 µm) column, from Macherey-Nagel and Thermo Scientific™ Chromeleon™ 7.3 Chromatography Data System software (Thermo Fisher Scientific Inc. MA, USA).Each sample was analysed (after 20 µL injection) and a gradient elution mixture composed of solution A (methanol), solution B (acetonitrile), and solution C (0.3% trifluoroacetic acid in water) was used as follows: 0 min, 15% A, 5% B, and 80% C; 10 min, 70% A, 30% B, and 0% C; 25 min, 70% A, 30% B, and 0% C; 28 min, 15% A, 5% B, and 80% C; and 31 min, 15% A, 5% B, and 80% C. The flow rate was set at 1 mL min −1 .Compound identification was based on retention time.The time of analysis was 31 min, including the stabilization of the RP-18 column.For quantification and identification purposes, Roy was detected using chromatograms corresponding to 270 nm, and its content in the plant extracts was estimated from the peak areas based on a calibration curve obtained with an authentic standard of Roy.All analyses were performed in triplicate.LOD and LOQ were assessed based on Signal-to-Noise approach 51 . Vol.:(0123456789)

Synthesis general procedure
For the general procedure, Roy (around 20 μmol) was dissolved in of dichloromethane (DCM, 2 mL) or pyridine (0.5-1 mL) with stirring (400 rpm) in a 5 mL round bottom flask, at room temperature, under heating or in an ice bath.When DCM is used as solvent, excess of pyridine was added (2.5-18 eq).Then, benzoyl chloride (or acetic anhydride) (1-100 eq) was added to the reaction flask.Reactions were followed by TLC (eluent DCM: Acetone, 98:2) until total consumption of Roy, then concentrated under reduced pressure.Products purification was performed by preparative chromatography, using as the eluent a mixture of CHCl 3 :Acetone (99:1) to purify RoyBz and CH 2 Cl 2 :Acetone (99:1) for Roy-12-Bz, Roy-12-Ac and, RoyAc.The purity of the compounds was assessed by HPLC, according to the methodology described above.All compounds exhibited a purity exceeding 95%.

Molecular docking
Molecular docking experiments were conducted in a similar way as our previous successful predictions for PKCδ 5 , with AutoDock v4.2.6 46 .The 1PTR PKC-δ isoform structure was obtained from the PDB and the needed mutations (M239G, W252Y, V255I and K256H) performed in MOE 52 .The aminoacid protonation states were assigned using the Protonate 3D module within MOE and exported as PDB file.All tested molecules where built and energy minimized in MOE and their energy minimized using default parameters.The PBD files were converted to the respective PDBQT ligand or receptor files using python scripts available in MGLTools.Docking poses and interactions were visually inspected within MOE.

Figure 2 .
Figure 2. Reaction scheme and conditions tested for esterification reactions of Roy.

Figure 3 .
Figure 3. Top ranked docking poses for RoyBz in the α and δ isoforms (gray and yellow carbons, respectively).The blue line marks the position where the phospholipid carbon tails begin insertion in the membrane, according to OPM.

Table 1 .
Parameters used for the calibration curve.

Table 2 .
Reactional conditions tested for esterification with benzoyl chloride a .
aThe reactions were carried out with 10 mg of Roy.b Isolated Yields.n.i., not isolated.

Table 3 .
Reactional conditions tested for esterification with acetic anhydride a .
aThe reactions were carried out with 3 to 5 mg of Roy.b Isolated Yields.

Table 4 .
Cytotoxicity of the derivatives against human breast adenocarcinoma cell lines.Inhibitory concentration (IC 50 ), reported in µM, was based on dose-response curves.nt, not tested; Doxo, Doxorubicin.

Table 5 .
EC 50 values of compounds tested on individual PKC isoform, using the yeast PKC assay.