Structural and physicochemical characterization and purity assessment of myrsinoic acids A and B, active compounds isolated from Rapanea ferruginea barks

2.1 Chemicals and reagents

Analytical grade hexane and ethyl acetate were purchased from Vetec (Rio de Janeiro, Brazil). Deuterated chloroform (CDCl₃) was purchased from Merck (White House Station, NJ, USA).

For HPLC analysis, all solvents used were HPLC grade (Tedia, Fairfield, Ohio, USA). The water was purified using a Milli-Q system (Millipore, Massachusetts, USA). All solutions were filtered through 0.45 μm membrane (Millipore, Massachusetts, USA).

2.2 Isolation and purification of myrsinoic acids A and B

Isolation procedures were performed according to Hess et al. (2010) and Bella-Cruz et al. (2013) as briefly described below: a mixture of hydroethanolic extracts of stem bark, branches and leaves of R. ferruginea (43 g) was subjected to silica gel column chromatography with hexane/ethyl acetate to obtain myrsinoic acids A and B in their pure forms. The fractions eluted with 20% and 40% of ethyl-acetate in hexane gave MAA (2.3 g), and MAB (4.2 g), respectively. After solvent evaporation, MAA presented as a yellow oil, and MAB as a crystalline white powder.

2.3 Structural characterization

2.4 HPLC analysis

Chromatographic analysis was carried out on a HPLC Shimadzu® LC 20-AT system, consisting of a quaternary pump and a Shimadzu SPD-M20A photodiode array detector, SIL-20A HT auto-sampler, with a Shimadzu CTO-10AS VT column oven equilibrated at 35 °C, and software LC Solution was used. The chromatographic column used was Phenomenex Kinetex® C₁₈ column (150 × 4.6 mm) with core-shell particles of 2.6 μm. The separation was achieved on a gradient method. The method was chosen according to the best separation of myrsinoic acids A and B at a concentration of 50 μg mL⁻¹ in methanol solution. The best mobile phase was methanol–acetonitrile–water (acidified to pH 2.5 with phosphoric acid) at a flow rate of 0.9 mL min⁻¹. The gradient elution was programmed as follows: initial isocratic condition during 2 min (25:5:70), 2–5 min (25:30:45), 5–10 min (25:60:15), 10–12 min (20:70:10), 12–15 min (15:75:10) and 15–20 min (15:84:1) then returns to the initial condition until 30 min of analysis. The analysis was monitored at 260 nm for MAA and at 270 nm for MAB. The purity of MAA and MAB was calculated through the presence of other peaks in the chromatograms of the compounds, expressed in relative area.

The specificity of the method was established according to the ICH guidelines (2005) using the PDA detector for the peak purity evaluation, and the samples were analyzed under forced degradation conditions (acidic, alkaline, oxidative and photodegradation), as described below.

2.5 Stability data

In order to develop a gradient stability-indicating LC–UV method, a forced degradation study of MAA and MAB was carried out. Both compounds were initially dissolved in methanol (50% of the volume) and the volume was made up with the degraded solutions as follows: hydrolytic stress studies were performed (at 25 °C) in acid (1 M HCl, 24 h), basic (1 M NaOH, 24 h) and oxidative (30% H₂O₂, 6 h) conditions, at final concentration of 100 μg mL⁻¹ solutions of MAA and MAB. The MAA oil and MAB powder were submitted under conditions of visible light (2.4 mi lux h⁻¹) and UVA (400 W h m⁻²) irradiation. After exposure, sample solutions of 100 μg mL⁻¹ in methanol were analyzed by the HPLC method. The stressed samples were assayed by comparison with non-degraded reference standards.

To confirm the temporal stability of MAA and MAB in methanol and acetonitrile, 2.5 mg of each sample was weighted and dissolved in 25 mL of each solvent. The solutions were stored in the following three conditions: light shielding at room temperature, light shielding at 15 °C and light shielding at 2 °C. The sampling periods were 0, 4, 8, 24, 48 h and 14 days. Each sample was evaluated by HPLC. Statistical analysis was carried out by one-way ANOVA using experimental design tools of Statistica® 6.0 software followed by Tukey test. The level of significance used was p < 0.05.

2.6 Physicochemical characterization

3.1 Structural characterization

In order to confirm the chemical structure of MAA and MAB, it was carried out the spectroscopic and spectrometric analysis of NMR (¹H and ¹³C), FTIR and UV. In addition, it was determined, for the first time, the fragmentation patterns by mass spectrometry and the Raman spectroscopy.

3.1.2

3.1.2 Fourier transform infrared spectroscopy (FTIR)

The MAA ATR-FTIR spectrum, presented in Fig. 2A, showed absorption peaks at 2918 cm⁻¹, corresponding to CH_sp3 stretch. A broad overlapping OH band in the range of 3400–2600 cm⁻¹ was observed, which together with the peak at 1678 cm⁻¹ (C⚌O), characterizes the carboxylic acid. At 1600 cm⁻¹ and 1201 cm⁻¹, the peaks observed correspond to the C⚌C and C—O stretch. The DRS-FTIR spectrum for MAB, shown in Fig. 2B, displayed a stretch peak at 3367 cm⁻¹ with O—H corresponding to alcohol. A prominent absorption band between 3000 and 2600 cm⁻¹ and intense peak at 1701 cm⁻¹ correspond to carboxylic acid. The peak at 1605 cm⁻¹ corresponds to the C⚌C stretch, confirmed by the intense bending peak at 960 cm⁻¹. The 1282 and 1180 cm⁻¹ bands correspond to the C—O stretch, concerning aryl-alkyl-ether in the molecule.

Close

Figure 2

FTIR spectra of myrsinoic acid A (A) and myrsinoic acid B (B).

Export to PPT

3.1.3

3.1.3 Raman spectroscopy

Raman spectroscopy is an analytical technique that is widely applied to the chemical and physicochemical characterization of solid compounds as it is a non-destructive method and needs only minimal sample preparation. Furthermore, in contrast to IR spectroscopy, it is not disturbed by the presence of water. Raman spectra are a highly specific method, serving as the fingerprints of the molecules (Láng et al., 2013). Raman spectroscopy was performed complementary to DRIFT, with the purpose of elucidating and confirming the compound structure. The MAB Raman spectrum is presented in Fig. 3. It showed the main characteristic bands at 1700–1500 cm⁻¹, relating to the aromatic ring and COOH vibrations, respectively, in 1400 cm⁻¹ the vibration relative to CH₃CH₂, and the double bond signal is demonstrated at 1300 cm⁻¹, which is in agreement with the previous data (Vandenabeele et al., 2000).

Close

Figure 3

Raman spectrum of MAB.

Export to PPT

3.1.4

3.1.4 Ultraviolet spectroscopy

The MAA UV spectrum (Fig. 4) presented two maximum absorption peaks, at approximately 205 and 258 nm. MAB showed maximum absorption at 218 and 269 nm (Fig. 4). Specific absorption ( ${\text{E}}_{1\text{cm}}^{1\%}$ ) was 287.2 and 229.0 for MAA and MAB, respectively.

Close

Figure 4

(A) Chromatographic profiles of MAA at 260 nm (R_T of 16.03 min), and of MAB at 270 nm (R_T of 15.47 min) and its respective UV spectra; Chromatographic profiles of MAA at 260 nm (B) and MAB at 270 nm (C) after degradation studies: (1) non-degraded samples at 100 μg mL⁻¹; (2) samples after stress with 1 M HCl 24 h; (3) samples after stress with 1 M NaOH 24 h; (4) samples after stress with 30% H₂O₂ 6 h; (5) samples after stress with visible light (2.4 mi lux h⁻¹); (6) and samples after stress with UVA (400 W h m⁻²).

Export to PPT

3.1.5

3.1.5 Mass spectrometry with direct sample inlet (DI-MS)

The insertion of MAA into the mass spectrometer resulted in a spectrum (Fig. 5A) with a molecular ion peak at m/z 342 and a base peak at m/z 123, which was generated by cleavage between C1′ and C2′ bond (C₉H₅). The MAB mass spectrum (Fig. 5B) showed the molecular ion peak at m/z 358 and the base peak at m/z 69 corresponding to the fragment (C₅H₉), generated by the cleavage of the C1′′ and C7 bond. For this compound, an intense peak was observed at m/z 109, which is suggestive of the C1′ and C2 cleavage bond, with loss of water.

Close

Figure 5

DI-MS (EI mode) mass spectrum of MAA (A) and MAB (B) and suggested major fragmentation mechanisms.

Export to PPT

3.2 HPLC analysis

The HPLC method was developed considering the parameters of resolution factor (R) between peaks, runtime and the capacity of being stability-indicative. The use of an acid mobile phase was required. MAA and MAB had retention times (R_T) of 16.03 ± 0.01 and 15.47 ± 0.01 min, respectively (Fig. 4). Separation was considered satisfactory (R = 4.3). The calculated purity of MAA was 76.20% and MAB of 98.90%. It was not observed supplementary peaks in another wavelength (200–400 nm), for both compounds. The HPLC method showed good specificity, as the peak purity through PDA was >99.99% for both compounds, and there was not apparent co-elution of degradation products with MAA and MAB peaks after the stress tests described below.

3.3 Stability data

After dissolved in the degraded solutions, both compounds remained visually soluble. Significant degradation was observed after acid hydrolysis (1 M HCl, 24 h) of MAA (72.22%) and MAB (43.44%). Chromatographic profile of degraded MAA showed the presence of at least six different impurities with R_T between 12.4 and 15.3 min, as observed in Fig. 4A-2. The MAB chromatogram after acid hydrolysis presented seven impurity peaks between 11.2 and 17.1 min (Fig. 4B-2).

Degradation was not observed when MAA and MAB were exposed to alkaline hydrolysis condition (Fig. 4A-3 and B-3, respectively), probably due to the formation of a stable salt.

MAA submitted to oxidative hydrolysis (30% H₂O₂, 6 h), showed significant degradation (36.60%), and supplementary peaks were observed at about 12.63 and 12.88 min (Fig. 4A-4). MAB exposed to the same conditions exhibited degradation of 6.80%, and supplementary peaks were detected at 14.06 and 14.33 min (Fig. 4B-4).

After visible light exposure (2.4 mi lux h⁻¹) of the MAA sample, the yellowness of the oil decreased. Its chromatographic analysis showed a significant degradation of 40.74% and a different peak with R_T about 14.0 min (Fig. 4A-5). When MAB was submitted at this stress condition, a color change was observed in the white powder, which turned yellow. The HPLC profile, presented in Fig. 4B-5, showed a MAB degradation of 29.93% and elution of the degradation products at between R_T 12.8 and 14.6 min. When MAA was exposed to UVA irradiation (400 W h m⁻²), the oil presented a dark yellow coloration and degradation of 45.01% was observed. As observed in Fig. 4A-6, the same impurities were observed, which appeared after visible light exposure, at R_T of 14.08 min. MAB presented color change to dark yellow after UV exposure. This compound also had significant degradation (21.97%), and the same degradation products were observed at between R_T 12.8 and 14.6 min (Fig. 4B-6), for the MAB after visible light exposure.

The temporal stability test indicates that the MAA and MAB methanol solutions were stable when stored at light shielding at room temperature until 14 days. MAB acetonitrile solutions showed stability at the room temperature, after 14 days. However significant degradation of MAA acetonitrile solution was observed after 14 days, and no change in the chromatographic profile of MAA, probably due to the formation of nonchromophore degradation products.

Both MAA and MAB in methanol or acetonitrile solutions were stable until 14 days, when stored at light shielding at 2 °C or 15 °C. The statistical results showed no significant difference among samples. The RSD (%) for assay of MAA and MAB during solution stability was within <2%.

3.4 Physicochemical characterization

Using HPLC, the partition coefficients (log P) of each compound were determined by plotting the log P values against (log k) of the known compounds (Fig. 6) (OECD, 1989), where t₀ was 3.31 min.

Close

Figure 6

Correlation between log P and log k was determined by the isocratic method for reference compounds: benzyl alcohol, phenol, benzoic acid, benzene, toluene and naphthalene.

Export to PPT

The following calibration curve was obtained: $\text{log}{P}_{\mathit{ow}}=2.1311·\text{log}k+2\text{,}244$ ( $r^2=0.8773\text{,}n=6$ ). MAA and MAB exhibited log k values of +0.52 and +0.48, in that order, and the log P values were 3.30 and 3.22, respectively.

These data are extremely useful in the drugs biopharmaceutical classification. The literature reports that compounds with log P of between 2 and 3 show optimum skin permeability coefficient values (Hadgraft et al., 2000). Log P values found for MAA and MAB were similar to those of non-steroidal anti-inflammatory agents (NSAIDs) such as diclofenac, ibuprofen, indomethacin and naproxen, which showed significant skin permeability after topical application (Hadgraft et al., 2000; Singh and Roberts, 1994).

Lipophilicity is one of the most important descriptors to be identified for CNS penetration. Blood-brain barrier (BBB) penetration is considered optimal when the log P values are in the range of 1.5–2.7, with a mean value of 2.1 (Pajouhesh and Lenz, 2005). However, it is seen that CNS drugs differed considerably in their lipophilicity, not following this rule. For example, some antidepressants present a mean log P of 3.10 (Ghose et al., 1999), as well as some drugs used in the treatment of Alzheimer’s disease, such as donepezil (log P value of 3.08–4.11) (Sozio et al., 2012) and Memantine, which has a log P value of 3.28 (Sonkusare et al., 2005). Thus, the log P of MAA can support the effects on memory observed in a previous study conducted by our research group (Costa et al., 2003). The absorption, excretion, and physiological barrier penetration may be related to the log P and pK_a values of drug, highlighting the importance of predicting both of these parameters in drugs.

TGAs of MAA revealed no mass loss from the sample over temperature range of 20–150 °C (Fig. 7A). Upon heating, MAA displayed a gradual loss of mass between 160 °C and 400 °C, maintained until 450 °C. The results evidence the low humidity of 1.42 ± 0.75% for this compound. The DSC thermogram for MAA showed an exothermic event at 272.1 ± 9.09 °C (Fig. 7A), characteristic of the process of thermal decomposition that occurs concomitant to heat release. As myrsinoic acid A is an oil, it was not possible observe its melting peak; thus, purity could not be ascertained using DSC.

Close

Figure 7

DSC and TG thermograms of MAA (A) and MAB (B).

Export to PPT

The TGA plot of MAB shows that thermal degradation of the sample starts at 250 °C, with a progressive loss of mass until 350 °C, maintained up to 450 °C (Fig. 7B). The MAB humidity was 0.75 ± 0.01%. The DSC thermogram showed a sharp endothermic peak with maximum at 123.6 ± 0.1 °C, due to the compound melting (Fig. 7B). The exotherm event at 298.6 ± 10.3 °C corresponded to significant loss of mass and thermal degradation. The purity of MAB was determined as 100.74 ± 0.18 mol%, by the Van’t Hoff equation.

The active pharmaceutical ingredient (API) can exist in different crystalline forms. Formulators should take advantage of this, where appropriate, when some unfavorable technological and biopharmaceutical properties of APIs need to be improved. The literature contains many examples of the suitability of using one crystalline form as opposed to another (Martino et al., 2007). In general, amorphous structure presents greater solubility than crystalline form, which is a desirable characteristic for a new pharmaceutical compound. As an example, low levels of amorphous character induced in griseofulvin by milling, increased its solubility of twofold or more (Elamin et al., 1994). The X-ray diffraction patterns related to myrsinoic acid B are shown in Fig. 8A. According to the diffractograms, it is possible to observe a semi-crystalline behavior for the sample, with an evident reflection at 5.2°, at low 2θ angles. These results, taken together with the DSC data, indicate that the phenolic compound MAB presents a semi-crystalline structure, what means that the sample has two phases, crystalline and amorphous.

Close

Figure 8

X-ray diffraction spectra (A) and SEM photomicrograph (enlargement of 1000×) (B) of MAB powder.

Export to PPT

The morphological analysis of drugs is an easy way to monitor and detect changes related to variations in crystal synthesis process or even in pharmaceutical processing (Maximiano et al., 2010). The photomicrograph of Fig. 8B shows the morphology of MAB crystallized by a solution of hexane–ethyl acetate (8:2) evaporated at room temperature with magnification of 1000×. It indicates a semi-crystalline compound, with the presence of some crystals with an orthorhombic appearance and particles without defined shape, which is characteristic of an amorphous compound, and is in agreement with the previous results (DSC and XRPD).

Finally, the solubility analysis of markers revealed that MAA was very slightly soluble in methanol and acetonitrile, while MAB was slightly soluble in both solventes. Both markers were practically insoluble in water.