An HPLC Method to Determine Phenolic Compounds of Plant Extracts: Application to Byrsonima crassifolia and Senna alata Leaves

Background: The Amazonian Region has a variety of medicinal plants with bioactive compounds, whose characterization could present the potential for sustainable development. Objectives: A method for separating, identifying, and quantifying a mixture of nine phenolic compounds (gallic acid, 3-hydroxybenzoic acid, p-coumaric acid, catechin, myricetin, rutin, quercetin, kaempferol


INTRODUCTION
Traditional communities in the Amazon region use a significant number of plant species for food and medical purposes.Nonetheless, many of these plants' chemical compositions, bioactive contents, biological properties, and medicinal efficacy are unknown. [1]n this context, Paula Filho (2018) [2] realized a phytochemical characterization of medicinal plants used by the population living on Cajari riverbanks in Amapá state.Among the studied plant species, the author reported the use of Acmella oleracea L., Byrsonima crassifolia (L.) Kunth, Cuminum cyminum L., Eryngium foetidum L., Inga edulis Mart., Pterocarpus rohrii Vahl, Senna alata (L.) Roxb and Talinum paniculatum (Jacq.),leaves.[5] Silva et al. (2007) [6] evaluated the phenolic compounds content in extracts obtained from 15 plant species from the Brazilian Amazon.These authors prepared extracts rich in phenolic compounds, highlighting the plants' potential as sources of bioactive agents.Bysornima crassifolia (L.) Kunth is a plant from Malpighiaceae family commonly known as muruci.The leaves, bark, and fruits of murici are reported to present diuretic and anti-inflammatory properties, according to traditional use in folk medicine in the Amazonian region. [6]The fruits are consumed in natura or used to produce sweets, ice cream, and juice. [2]Some studies indicated the content of phenolic compounds and the antioxidant and antimicrobial activities of extracts obtained from this species. [6,7]enna alata is a plant from the Fabaceae family used in traditional herbal medicine for different skin problems.This plant species, found in Ghana and Brazil, is reported to present antimicrobial, antifungal, and antioxidant activities, besides nutritional and other values. [8,9]The medical effect of Senna alata is attributed to the presence of active Cite this article: Cavalcante MA, Oliveira JS, Barreto MSS, Pinheiro LP, Cantuária PC, Borges WL, Silva GA, Souza TM.An HPLC Method to Determine Phenolic Compounds of Plant Extracts: Application to Byrsonima crassifolia and Senna alata Leaves.Pharmacog Res.2022;14(4):395-404.
components, such as anthraquinone (aloe emodin) and flavonoids (kaempferol 3-gentiobioside, and kaempferol). [10]uantification of phenolic compounds is commonly performed by spectrophotometric and chromatographic methods.Spectrophotometry is used to estimate the total phenolic content, whereas the individual classes of these substances can be identified by high-performance liquid chromatography (HPLC). [11]Folin-Ciocalteu is an example spectrophotometric method widely used for phenolic compound quantification.14] In contrast, chromatography consists of techniques for separating and quantifying the analytes present in plant matrices.The retention time allows the separation of different constituents, analytes, and interferents.17] Some factors, such as the type of column, detectors, mobile phase, and properties of the tested substances, influence the analysis of phenolic compounds by HPLC. [11]For example, reverse-phase liquid chromatography coupled with an ultraviolet detector or a diode array is widely used for routine quantitation analyzes of phenolic compounds. [18,19]hus, several studies are commonly carried out to develop chromatographic methods that enable the separation and identification of phenolic compounds in plant species, considering the type and number of phenolic compounds found in various matrices.This study developed and validated a simple, sensitive, and selective method for separating and identifying phenolic compounds using HPLC.This method was also applied to analyze aqueous and hydroethanolic extracts obtained from the leaves of the following Amazonian plants: Byrsonima crassifolia (L.) Kunth and Senna alata (L.) Roxb.

Plant material
Byrsonima crassifolia (L.) Kunth and Senna alata (L.) Roxb usage was registered in the National System for the Management of Genetic Heritage and Associated Traditional Knowledge, No. AE536D5.The Fresh leaves were obtained from the Brazilian Agricultural Research Corporation (EMBRAPA, Amapá, Brazil).The vegetal species' excicatas were identified by a taxonomist and deposited in the Amapaense Herbarium (HAMAB) under the codes 019453-HAMAB and 019458-HAMAB, respectively.The geographic coordinates of collection sites were the following: Byrsonima crassifolia (N 0°23'16.73732";W 0.387983), and Senna alata (N 0°0'57.77517';W 0.0160049").Fresh leaves were selected, washed, and dried in an air circulation oven at 60°C for 4 hr.The dried samples were ground to obtain a particle size of approximately 5 mm and then were stored at 4°C until analysis.

Preparation of the extract
Extracts of the plant species were prepared by suspension of 1 g of the dry plant powder in 6 mL of extraction solvent, water or ethanol:water mixture (1:1 v:v), at 30°C and 200 rpm for 1 hr, followed by filtration using Whatman paper No. 4. The residue was then extracted by adding 6 mL of the extraction solvent under the same condition.After this process, extracts were rotary evaporated under reduced pressure (LUCA-EV 01, Lucadema).22]

Preparation of sample solutions
For the Senna alata aqueous and hydroethanolic extracts, approximately 0.01 g of the dried sample was resuspended in 10 mL of water and ethanol:water (1:1 v:v) solution, respectively.For Byrsonima crassifolia aqueous and hydroethanolic extracts, 0.05 g of the dried sample was resuspended in 10 mL of water and ethanol:water (1:1 v:v) solution, respectively.The samples were stored at 4°C until analysis.

Preparation of standard solutions
The standard stock solution (1 g L -1 ) of ellagic acid, 3-hydroxybenzoic acid, p-coumaric acid, myricetin, quercetin, and kaempferol were prepared by dissolution in methanol.Rutin, catechin, and cyanidin standard stock solutions (1 g L -1 ) were produced by dissolution in ethanol whereas gallic acid was dissolved in acidified water (1% formic acid in ultrapure water).All solutions were stored at 4°C until analysis.Working solutions of the standards were prepared by dilution of stock solutions according to desired concentrations.

Ranges of the calibration curve
Aliquots of the standard stock solution were transferred to volumetric flasks and diluted to get concentration in the linear range described in Table 1.

HPLC instrumentation and chromatographic conditions
A reversed-phase C 18 HPLC column was used to separate phenolic compounds chosen among phenolic classes to draw a composition profile of the studied matrices.HPLC analysis was performed using a liquid chromatographer (Shimadzu, Prominence LC2030C) equipped with a binary pump, an online vacuum degasser, a diode array detector (DAD), an autosampler, a thermostatted column compartment, which accommodates 216 samples at a time and features a direct access rack system.System control and data analyses were performed using Labsolution software (Shimadzu).Before choosing the chromatographic conditions, several trials were carried out using different types of solvents and mobile phase concentrations to check the retention time (RT), peak shape, and tailing factor (peak symmetry) of the analyte.Separation was carried out in a Shim-pack VP-ODS C 18 column (5 µm, 150 x 4.6 mm).The mobile phase consists of a gradient elution using the proportions of solvent A (1% formic acid in water) to solvent B (acetonitrile) as follows: initial 5% B; 0 -7 min, 5% B; 7 -18 min, 5 -30% B; 18 -35 min, 30 -60% B; 35 -40 min, 60 -95% B with a flow-rate of 1 mL min -1 and the injection volume of 0.02 mL of samples and standards.The total running and post-running times were 40 and 5 min, respectively.The column temperature was maintained at 40°C throughout the analysis and the spectra were acquired in the 210 -800 nm range.Chromatograms were plotted at 254 and 277 nm.Using these chromatographic conditions, it was possible to confirm the retention time of phenolic compounds by injection of the corresponding standard separately.Identification of phenolic compounds in the extracts was based on the comparison of spectra and retention times of the samples against the external standards.The quantification was made by external standardization using analytical curves and limits of detection (LOD) and quantification (LOQ).

Validation of the Method
The chromatographic method was validated considering the guidelines recommended by the International Conference on Harmonisation (ICH). [23]Parameters such as specificity, linearity, repeatability, precision, LOD, LOQ, and robustness were evaluated to determine the method's performance.Since the test of extraction-method efficiency was out of the purpose of this work, the recovery was not calculated.

Specificity
The method specificity is the ability to distinguish between investigated analytes and other components, such as impurities, degradants, or excipients.The specificity was demonstrated by running a blank procedure, standard and sample.The chromatographic parameters such as column efficiency and peak symmetry were determined for the standards, according to the ICH guidelines Linearity The curve's linearity was determined by different known concentrations (Table 1) of the mix standards solution.For the linearity study, the standard solutions were injected five times (n = 5), and the peak area was measured.The calibration curve for each phenolic compound was constructed by plotting peak areas against concentration.Linear regression equations and correlation coefficient were determined.

Limit of detection and limit of quantification
The LOD is the lowest amount of analyte in a sample that can be detected but not necessarily quantitated under the stated experimental conditions whereas the LOQ is the lowest amount of analyte in a sample, which can be quantitatively determined with accuracy. [24]The LOD and LOQ were determined from a signal-to-noise ratio of 3 and 10, respectively, for the injection of standard solutions at low concentrations.In addition, the regions before and after chromatographic peaks were considered for calculating background noise.

Precision
Method precision was determined by repeatability, inter and intra-day precision.The inter and intra-day precision were determined using a minimum of three and six replicates for three different concentration levels, respectively: 10, 30, and 45 µg mL -1 for gallic acid, catechin, p-coumaric acid, cyanidin, and kaempferol, and 15, 30, 45 µg mL -1 for 3-hydroxybenzoic acid, rutin, myricetin, and quercetin.The inter-day precision variations were studied by different analysts.Peak area was measured and the results for precision were expressed as the percentage relative standard deviation (%RSD).

Robustness
The robustness of the HPLC method was evaluated considering the optimized parameters and slight and viable variations.The evaluated parameters were mobile phase composition (8 mL, 10 mL, and 12 mL of formic acid) and wavelength (250 nm, 254 nm, and 258 nm, except for catechin).Retention times and peaks area were checked for ruggedness by injecting the mixture of standards at 30 µg mL -1 .

Statistical Analysis
All the measurements were made in triplicates, and the results were expressed as mean values and standard deviation.Linear adjustment and determination of coefficients were realized in Excell software.Experimental data were subjected to analysis of variance (ANOVA) and Tukey's test at 5% significance using the Minitab® 19 software.

Chromatographic Method
For the method development and validation, a mixture of 10 phenolic compounds standards (4 phenolic acids, and 6 flavonoids) was injected into the HPLC system.For the mobile phase selection, different acetonitrile initial concentrations (ranging from 10% to 5%) were evaluated with a 50 min total running time.After optimizing the initial concentration (5%, v/v), the gradient time was decreased to obtain the best separation condition.It comprised a 40-min separation using gradient elution with linear steps, where the acetonitrile concentration varied from 5% to 95% (see details in the methodology section).Under these conditions occurred complete co-elution between ellagic acid and other compounds.To solve this problem, ellagic acid was eluted separately in the same method.Even though, this standard was not separated due to its solubilization in the mobile phase and interaction with the column (Figure 2).Therefore, it was not possible to use the ellagic acid standard in this method.Chromatograms of the nine external standards mixture (gallic acid, catechin, 3-hydroxybenzoic acid, p-coumaric acid, rutin, myricetin, cyanidin, quercetin, and kaempferol), were recorded at 254 nm and 277 nm, are presented in Figure 3.The investigated compounds showed suitable responses and were successfully separated at 254 nm, except for the catechin which presented a better definition at 277 nm.As the article's objective was to develop an optimized chromatographic method,

Method Validation
For each of the nine phenolic compounds, calibration curves based on the area under chromatographic peaks were plotted at five different concentrations to check the method's linearity (Figure 4).The calibration plots for the nine phenolic compound standards showed suitable linear fitting, as indicated by regression parameters in Table 2.The curves for all the investigated compounds exhibited correlation coefficients greater than 0.990.
The peak area was checked for repeatability and reproducibility by injecting the mixture of standards at concentrations of 10, 30, and 45 µg mL -1 for gallic acid, catechin, p-coumaric acid, cyanidin, and kaempferol and 15, 30, and 45 µg mL -1 for 3-hydroxybenzoic, myricetin, rutin, and quercetin using six replicates on consecutive days.Table 3   shows that in repeatability and reproducibility at three different concentration levels of phenolic compounds standard solution.Precision was evaluated based on the %RSD value.
The method was evaluated for changes of ±2 ml and ±4 nm in mobile phase composition and wavelength, respectively.The analytes' retention times were significantly affected by variations in mobile phase composition, as indicated by the one-way analysis of variance at a 5% level (Table 4).A slight reduction in retention times was detected for all compounds increasing the formic acid content (Table 5).In addition, Tukey tests showed a significant difference among means retention time at different mobile phase compositions for each external standard.ANOVA results also showed that retention times were not affected by the wavelength increase (Table 4).
For most of the external standards, the peaks' area changed significantly with mobile phase composition and wavelength (Table 6).However, the increase in formic acid content did not considerably affect cyanidin's peak area at a 5% level (p= 0.099).For quercetin (p= 0.338) and kaempferol (p= 0.518), no significant change in the peaks' area was also verified with the wavelength variation.As indicated by the mean values (Table 7), Tukey tests did not show any pattern of increase or decrease of peaks' area with the changes in mobile phase composition and wavelength.
The developed and validated method allowed the quantification of the nine distinct phenolic compounds in extracts of plant species.The mobile phase composition, column temperature, and wavelength parameters selected for the method were 1% formic acid aqueous solution, 40°C, and 254 and 277 nm, respectively.

Analysis of amazonian medicinal plants extracts
Identification of phenolic compounds was performed by comparing the extracts' retention times and spectra with those of the corresponding external standards.Quantification was based on external standards calibration.The chromatograms for the aqueous and hydroethanolic extracts of Senna alata and Byrsonima crassifolia are presented in Figure 5.The number and sort of phenolic compounds identified in the samples varied among the studied extracts.None of the extracts presented the peak associated with p-coumaric acid, but other peaks associated with unknown compounds were also detected.
The mean phenolic compounds contents in aqueous and hydroethanolic extracts of Byrsonima crassifolia, and Senna alata, expressed as mg g -1 of extract ± standard deviation, are presented in Table 8.Byrsonima crassifolia aqueous extract presented four phenolic compounds, whereas Senna alata hydroethanolic one presented six.Rutin was detected in the four evaluated extracts and its contents were significantly lower for Byrsonima crassifolia (2.51 and 5.97 mg g -1 ) when compared to Senna alata (12.17 and 11.34 mg g -1 ).Hydroethanolic extract of Senna alata  also presented kaempferol, quercetin, myricetin, 3-hydroxibenzoic acid, and cyanidin, but the two last substances were most abundant.Catechin (49.02 mg g -1 ) was the most abundant phenolic content in both Byrsonima crassifolia extracts (Table 3).

DISCUSSION
Validation of an analytical method consists of an evaluation of the procedure's suitability for the intended purpose.Data from method validation are required to ensure the quality, reliability, and consistency of analytical results. [25]A reversed-phase HPLC method was developed and validated to determine different phenolic compounds in aqueous and hydroethanolic extracts of Byrsonima crassifolia and Senna alata leaves.We used a mobile phase composed of acetonitrile gradients and 0.1% formic acid in water (v/v) which is typical in separating complex samples. [26,27]The developed method can be considered rapid as it allowed the separation of compounds at 254 nm in only 35 min.The optimized method was validated based on ICH guidelines. [23]ccording to ICH, linearity is associated with the ability to obtain test results that are proportional to the concentration of analyte in the   sample within a specified range.The LOD and LOQ varied between 6.2807 -14.8851 µg mL -1 and 6.8002 -16.0071 µg mL -1 , respectively (Table 2).Considering the used equipment and column, attained LOD and LOQ indicated that this method enabled an adequate detection and quantification of the phenolic compounds in the specified concentrations.
The precision of the analytical method demonstrates agreement between a series of measurements obtained by multiple sampling of the same homogeneous sample under the prescribed conditions. [23]Repeatability tests determined the precision of the experiments through solutions prepared and measured by the same analyst.Measurements made by distinct analysts, considering systematic errors, were also conducted to calculate the reproducibility (inter-day precision).Analyzing the results of the repeatability and reproducibility revealed that the relative standard deviation (RSD) of the peak area was less than 5% and 6%, respectively.The highest values of reproducibility were recorded for cyanidin and quercetin solution with a concentration of 45 µg mL -1 .%RSD values below 10% are acceptable by the AOAC manual for the Peer-Verified Methods Program, [28] so the method was found to be precise and reproducible.
The phase reverse HPLC method was succsesfully applied for analysis of the phenolic compounds in plant extracts, enabling the quantification of eight phenolic compounds present in the aqueous and hydroethanolic extracts of Byrsonima crassifolia and Senna alata matrix samples.
(2009) [31] identified naringin and apigenin as the major phenolic components in ethyl acetate fraction.Our results indicated that Senna alata leaves can also contain 3-Hydroxybenzoic acid, rutin, myricetin, cyanidin, quercetin, and kaempferol.The total phenolic content and extract composition depend on the extraction method and solvent's type and polarity. [32]Such aspects might explain the differences between obtained results and literature data.Furthermore, growing conditions and differences in plant genotypes also affect plant phenolic content and composition. [33]

CONCLUSION
A simple and precise HPLC method, using the usual 5 µm columns, was developed and validated to determine nine phenolic compounds found in plant extracts (gallic acid, catechin, 3-hydroxybenzoic acid, p-coumaric acid, rutin, myricetin, cyanidin, quercetin, kaempferol).It showed efficiency, precision, accuracy, and sensitivity, allowing timesaving during analysis.Therefore, this method is recommended for efficient assays in routine work.The developed method was successfully applied to analyze aqueous and hydroethanolic extracts of Byrsonima crassifolia and Senna alata leaves, respectively.Catechin was the main phenolic compound identified for the Byrsonima crassifolia extracts whereas 3-hydroxibenzoic acid and cyanidin were the most abundant substances in aqueous and hydroethanolic extracts of Senna alata, respectively.Further HPLC analyzes of extracts from these plants using other solvents will still be evaluated to investigate other phenolic compounds.

Figure 1 :
Figure 1: Extraction process employed for the studied plant species.

Table 3 : Repeatability and reproducibility of external standards.
*RSD: relative standard deviation.

Table 4 : ANOVA results for the effect of mobile phase composition and wavelength of retention times of standards compounds.
DF: Degree of freedom; *Par: parameter affecting retention times (mobile phase composition, column temperature, or wavelength).

Table 5 : Retention times of external standards ± standard deviation.
Means classified by the same letter for the same variable (mobile phase composition, column temperature or wavelength) and external standard did not differ according to the Tukey test at 5% significance.

Table 6 : ANOVA results for the effect of mobile phase composition and wavelength of peaks' area of standards compounds.
DF: Degree of freedom; *Par: parameter affecting peaks' area (mobile phase composition, column temperature, or wavelength).

Table 7 : Peaks area of external standards ± standard deviation.
Except for catechin: 273nm, 277nm e 281nm; Means classified by the same letter for the same variable (mobile phase composition, column temperature, or wavelength) and external standard did not differ according to the Tukey test at 5% significance. *