**Chromatographic, Chemometric and Antioxidant Assessment of the Equivalence of Granules and Herbal Materials of Angelicae Sinensis Radix**

**Valentina Razmovski-Naumovski 1,2,3, Xian Zhou 2, Ho Yee Wong 3, Antony Kam 3,4, Jarryd Pearson <sup>2</sup> and Kelvin Chan 2,3,5,\***


Received: 30 May 2020; Accepted: 19 June 2020; Published: 23 June 2020

**Abstract: Background:** Granules are a popular way of administrating herbal decoctions. However, there are no standardised quality control methods for granules, with few studies comparing the granules to traditional herbal decoctions. This study developed a multi-analytical platform to compare the quality of granule products to herb/decoction pieces of Angelicae Sinensis Radix (Danggui). **Methods:** A validated ultra-performance liquid chromatography coupled with photodiode array detector (UPLC-PDA) method quantitatively compared the aqueous extracts. Hierarchical agglomerative clustering analysis (HCA) and principal component analysis (PCA) clustered the samples according to three chemical compounds: ferulic acid, caffeic acid and Z-ligustilide. Ferric ion-reducing antioxidant power (FRAP) and 2,2-Diphenyl-1-picrylhydrazyl radical scavenging capacity (DPPH) assessed the antioxidant activity of the samples. **Results:** HCA and PCA allocated the samples into two main groups: granule products and herb/decoction pieces. Greater differentiation between the samples was obtained with three chemical markers compared to using one marker. The herb/decoction pieces group showed comparatively higher extraction yields and significantly higher DPPH and FRAP (*p* < 0.05), which was positively correlated to caffeic acid and ferulic acid, respectively. **Conclusions:** The results confirm the need for the quality assessment of granule products using more than one chemical marker for widespread practitioner and consumer use.

**Keywords:** Angelicae Sinensis Radix; antioxidant; Danggui; granules; herb; multivariate analysis; ultra-performance liquid chromatography

#### **1. Introduction**

Granule formulations have become the most popular delivery form for Chinese medicinal herbs and are used as an alternative to herb and decoction pieces in herbal prescriptions worldwide including China, Japan, USA and Europe [1]. For practitioners and consumers, granules are convenient in terms of easier administration (granules are added to water instead of boiling herbs in water which are then strained), transport (less bulky than herbs) and storage (protected from microbes and moisture). There is potential for better quality control of granules using good manufacturing practice (GMP) processes which would assure the reproducibility of products. This would promote clinical consistency as solvent ratios to herbs and boiling times of herbs/decoction pieces are not patient-dependent [2,3]. However, standardised quality control procedures for granules are limited. In recent times, ultra-performance liquid chromatography (UPLC) has analysed the granule formulations of popular herbs such as *Panax ginseng* (Araliaceae), *Salvia miltiorrhiza* (Lamiaceae), *Panax notoginseng* (Araliaceae) and other common composite formulae [2,4–7]. However, there are few comparative studies regarding the actual quality and efficacy of granules compared to the traditional herbal decoction, and the variations between granule formulations from different manufacturers [2–4]. This calls for a simple and rapid multi-method approach to guarantee the reliability and bioequivalence of herbal products to ensure their clinical efficacy [8].

In this study, the herb/decoction pieces and granule products of Angelicae Sinensis Radix, also known as Danggui in Chinese, are evaluated [9]. Danggui, the dried root of *Angelica* (A.) *sinensis* (Oliv.) Diels (Umbelliferae), is one of the most popular Chinese materia medica and is used in dietary supplements and cosmetics globally [10]. Originally listed as top grade in the Shennong's Classic of Herbology and nowadays described as 'female ginseng', Danggui is used in gynaecological disorders such as painful dysmenorrhea, postpartum weakness and treating menopause [11]. The herb is known to regulate blood circulation, have antioxidant activity, and is widely used in cardiovascular diseases such as atherosclerosis and hypertension [12]. Danggui is present in over 80 composite formulae of traditional Chinese medicine (TCM).

Despite the popularity of Danggui, there is no quality assessment of Danggui granules [13]. With the general consensus of using a multi-method approach in assessing the quality of herbal products, the present study evaluated the differences between the Danggui samples using chromatography, chemometrics and antioxidant activity. Three chemical markers (ferulic acid, caffeic acid and Z-ligustilide) were quantified using the developed UPLC method. Hierarchical agglomerative clustering analysis (HCA) and principal component analysis (PCA) grouped the samples according to the content of the three markers. The results were compared to using either ferulic acid or Z-ligustilide as the single chemical marker as specified by the Pharmacopoeia of the People's Republic of China (PPRC) [9] and World Health Organisation (WHO) guidelines, respectively [14]. Coupled with the statistical clustering analysis, correlating the chemical markers to antioxidant activity provided a comprehensive study of the differences between the products. Any variations between the products may imply possible pharmacological differences which need to be addressed in terms of correct dosages to patients.

#### **2. Materials and Methods**

#### *2.1. Plant Materials and Reagents*

Ten commercial Danggui granule products (coded as G1–G10) were produced by companies in mainland China, Hong Kong and Taiwan, and were either purchased from their distributors in Australia or directly from the manufacturers. Product names have been omitted as consent for disclosure was not sort. One herb (coded as R2) and four decoction piece samples (coded as R1, R3, R4, R5) were sourced from Min Xian, Gansu Province in China [9]. The region where the herbal material was sourced from is well known for Danggui and considered the best quality according to TCM. They were purchased from Australia, mainland China and Hong Kong. The samples were authenticated by Dr George Li from the Faculty of Pharmacy, The University of Sydney, Australia. The taxonomic identification was carried out macroscopically and microscopically according to the descriptions in the Pharmacopoeia of People's Republic of China (PPRC) [9]. Voucher specimens were deposited at NICM Health Research Institute, Western Sydney University, Australia. They were labelled as for granules: G(number)(company)(date)AS and raw materials: R(number)(date)AS.

The three reference chemical markers (caffeic acid, ferulic acid and Z-ligustilide) were purchased from Chengdu Biopurify Phytochemicals Ltd. (Sichuan, China) and were graded > 98% HPLC purity. The compounds were verified with liquid chromatography–mass spectrometry (LC–MS). Chloroform, formic acid and acetonitrile were obtained from Ajax Finechem (Taren Point, Australia). Methanol was purchased from Fisher Scientific (Loughborough, UK) and ethyl acetate was purchased from Biolab Ltd. (Scoresby, Australia). Water was obtained from a Milli-Q Reagent Water System (Millipore, Burlington, MA, USA). All the solvents mentioned were HPLC-grade. For the antioxidant assays, DPPH, Trolox, sodium acetate trihydrate, glacial acetic acid, TPTZ (2, 4, 6-tripyridyl-s-triazine), hydrochloric (HCl) acid and ferric chloride hexahydrate were purchased from Sigma-Aldrich Corp (St. Louis, MO, USA).

#### *2.2. Preparation of the Extracts and Standards*

In this study, it was anticipated that the granule manufacturing process involved the large-scale extraction of herbs with boiling water to reflect the traditional decoction, followed by spray-drying or fluidised bed drying and formulation with excipients [5]. Thus, to remove most of the water-soluble excipients, 1 g of the Danggui granule sample was suspended in methanol (10 mL) and sonicated for 30 min at 40 ◦C. The sonicated mixture was centrifuged at 4000 rpm for 10 min and the supernatant removed. The extraction was repeated two more times. The combined supernatants were concentrated by a rotary evaporator to dryness at 50 ◦C. Here, the residue obtained from the granules after the methanol extraction was assumed to be equivalent to the raw herb water extract without excipients.

The herb and decoction pieces of Danggui were ground by an electric blender and passed through a 500 μm aperture sieve. The powder (1 g) was refluxed with boiling water (30 mL) for 30 min and the extraction repeated two more times. The sample was then centrifuged at 4000 rpm for 10 min. The supernatant was transferred and evaporated to dryness at 50 ◦C. This was followed by the same extraction procedure as described for the granule samples to allow comparison of the samples as methanol extracts. The solutions were prepared by re-dissolving the dry extract residues with methanol followed by filtrating into the final testing samples through the filter syringes (0.2 μm).

The individual standard stock solutions of the chemical markers caffeic acid, ferulic acid and Z-ligustilide were prepared at the concentration of 2 mg/mL in methanol. To minimise the impact of the stability, the standards and samples were freshly prepared each day and protected from heat, moisture and light.

#### *2.3. Determination of Chemical Marker Content*

UPLC analyses were performed using a Waters Acquity ultra performance liquid chromatography (UPLC)® H series consisting of a H class quaternary solvent manager, an Acquity sample manager-FTN, an Acquity column oven and an Acquity Photodiode Array Detector (PDA) detector. The chromatographic separation was achieved using an Acquity UPLC BEH C18 column (50 mm × 2.1 mm, 1.7 μm) maintained at 40 ◦C [15].

The UPLC condition was based on our in-house HPLC method with modifications to the gradient condition [16–18]. The mobile phase consisted of 1% formic acid in water (A) and acetonitrile (B) (95:5, *v*/*v*), with a gradient elution as follows: 0–10 min, 5–12% B; 10–15 min, 12–20% B; 15–20 min, 20–100% B, 100% B for 5 min and reconditioning the column isocratically with 5% B for 4.5 min. The flow rate was 0.3 mL/min. The injection volume was 2 μL and the detection wavelength was set at 325 nm, which was similar to previous studies which monitored for ferulic acid and Z-ligustilide [15,17,19].

The UPLC method was validated in terms of linearity, repeatability and accuracy according to ICH guidelines [20]. Linearity testing was carried out by running six different concentrations of each chemical marker (caffeic acid (0.005–2 mg/mL), ferulic acid (0.005–2 mg/mL) and Z-ligustilide (0.01–0.3 mg/mL) in triplicate. Partial least square regression method was used to obtain the regression equations in the form of y = ax + b, where x is the concentration of the reference chemical marker and y was the peak area [21]. The limit of detection (LOD) and the limit of quantification (LOQ) were determined by standard deviation (SD) approach, where LOD = 3.33 × (SD of y-intercept/mean of slope) and LOQ = 10 × (SD of y-intercept/mean of slope). For repeatability, the intra-day precision was evaluated by running six concentrations of each marker three times within a day, whilst inter-day precision was examined on three separate consecutive days. To determine the accuracy of the method, a recovery assay was performed in triplicate by spiking two known concentrations (100 and 150 μg/mL) of the mixed standards (caffeic acid, ferulic acid) to one representative decoction piece and granule sample [22]. Percentage recovery (%) ± RSD was calculated by the equation: % = ((mean detected content − mean original content)/mean of spike content) × 100%.

#### *2.4. Antioxidant Activity Assays*

The 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay was performed as previously described [23]. The test samples were mixed with DPPH radical solution (0.24 mg/mL DPPH in methanol) and incubated for 30 min in the dark. The absorbance was determined at 515 nm. (±)-6-Hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid (Trolox) was used for the calibration curve. All values were expressed as milligrams Trolox equivalents (TE) per gram of dried weight (DW) (mg TE/g DW).

The ferric ion reducing antioxidant power (FRAP) assay was performed as previously described [24]. The FRAP working solution was prepared by mixing 10 volumes of 300 mM acetate buffer (pH 3.6), 1 volume of 10 mM 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ) in 40 mM HCl and 1 volume of 20 mM ferric chloride (FeCl3·6H2O). The test samples were mixed with pre-warmed FRAP reagent (37 ◦C) and incubated for 30 min at 37 ◦C. The absorbance was measured at 595 nm. The standard curve and the results of TE were obtained by the same approach as described above.

#### *2.5. Statistical Analyses*

The yields (reported as percentage of the dry weight of the herb) is the mean of three extractions. One of the extracts from the same sample was analysed three times by UPLC, with the final quantitative results from UPLC analyses expressed as the mean ± standard deviation (SD). Quantitative results were reported as milligrams per grams of the DW of the raw herb (mg/g) equivalent. Non-parametric test (SPSS 20.0 software, IBM, Chicago, IL, USA) was conducted to determine whether the content of each chemical marker analysed by UPLC was significantly different between individual samples.

The chosen markers were considered as variables in the following HCA and PCA statistical analysis. HCA grouped the individual Danggui samples into clusters based on the degree of the similarity of the variables. The HCA results were expressed as a dendrogram using Ward's linkage algorithm and squared Euclidean distances (SPSS 20.0 software, IBM, Chicago, IL, USA). The different linkage criteria applied in the dendrogram revealed the degree of similarity between each sample. The length of the linkage between each sample/group represents the degree of similarity. Thus, the shorter the linkage, the more similarity there is between each group. PCA was performed using XLSTAT 2019.1 by Addinsoft (New York, USA) which reduced the original variables into two major principal components (PCs). These two PCs maintained the greatest possible variance of the original variables (three chemical markers) [7]. The PCA results were represented in a biplot (score plot and loading plot), where the score plot showed the clusters and outliers of the samples, and the loading plot demonstrated the correlation of the PC to the original variables. In the biplot, a point represented each individual sample, and the distance allocated between samples revealed the degree of their similarity in terms of the content of the chemical markers.

For the antioxidant assays, the data were expressed as the mean ± standard deviation (SD) of three repeat measurements and was analysed using independent-samples t-test and non-parametric analysis by SPSS. For this study, *p* < 0.05 was considered as statistically significant. Pearson correlation coefficient (r) by SPSS (20.0 software, IBM, Chicago, IL, USA) evaluated the strength of the correlation of the chemical markers to the antioxidant activities.

#### **3. Results**

#### *3.1. Extraction Yields*

The mean yield of each sample is shown in Table 1. The yield results of the granule products were converted according to their concentrated ratio (as listed on the package) so that comparison to the original herbal material could be made. The yields of the herb/decoction pieces (33.2–44.8%) as a group were comparatively higher than that of the granules (2.7–12.9%).

**Table 1.** Average contents (mg/g, means ± SD, n = 3) of the chemical markers in the ten Danggui granule samples (G1–G10) and the five herb/decoction piece samples (R1–R5) analysed by UPLC-PDA.


G = granule; SD = standard deviation; ND = not detected. <sup>a</sup> Hospital-grade, use as directed by doctor, <sup>b</sup> Available in Australia; <sup>c</sup> Granule to raw herb ratio as specified by the manufacturer, thus 1 g granule is produced by 3 g herb etc.; <sup>d</sup> Average yield converted by granule ratio; \* Significantly different within the raw samples (*p* < 0.05); \*\* Significantly different within the granule samples (*p* < 0.05); \*\*\* Significantly different within the granule samples (*p* < 0.05).

#### *3.2. UPLC-PDA Quantification of the Chemical Markers*

The representative chromatograms of the mixed chemical markers, the granule (G1) and herb (R2) extracts are shown in Figure 1, with a total run time of 30 min. In this study, the three chemical markers, caffeic acid, ferulic acid and Z-ligustilide and their calibration curves produced good correlations between the peak area and concentration as shown in Table 2, with the correlation coefficients r2 > 0.997 for all analytes. The LODs and LOQs were in the range of 0.701–3.268 μg/mL and 2.106–9.813 μg/mL, respectively. The intra-day and inter-day RSD were 1.5–2.77% and 2.6–4.11%, respectively (Table 2). This suggests that the method had reasonable instrumental and method precision [22].

The addition of known amounts of the compounds to the samples is recommended for recovery testing of herbal compounds [22]. Granule sample 2 (G2) and decoction piece sample 3 (R3) were randomly chosen as representative Danggui samples from each group. The average recoveries (%) were for G2: 89.3 ± 1.1% (caffeic acid) and 99.7 ± 1.2% (ferulic acid); R3: 94.1 ± 2.1% (caffeic acid) and 99.6 ± 2.1% (ferulic acid).

The developed UPLC-PDA method simultaneously quantified the three marker compounds in the Danggui water extract herb and granule samples, and the results are shown in Table 1. In this study, the amount of caffeic acid, ferulic acid and Z-ligustilide in all the samples ranged from 0.004–0.041, 0.030–0.503 and 0.005–0.526 mg/g DW, respectively. In the granule samples, G5 and G10 had a relatively higher amount of Z-ligustilide (0.526 mg/g DW and 0.183 mg/g DW, respectively) compared to the rest of the samples.

Nonparametric independent-samples t-testing of the raw herb samples revealed that R2 had significantly higher ferulic acid (*p* < 0.05) and caffeic acid (*p* < 0.05), and there was no significant difference in Z-ligustilide content (*p* > 0.05).

Both caffeic acid and Z-ligustilide were not significantly different between the granule and decoction piece/raw herb groups (*p* > 0.05). However, the amount of ferulic acid was found to be significantly different (*p* < 0.05) between the two groups. In terms of ferulic acid and caffeic acid, G7 (higher content) and G10 were significantly different (*p* < 0.05). In terms of Z-ligustilide, G5 (higher content) (*p* < 0.05) and G10 (*p* < 0.05) were significantly different.

**Figure 1.** UPLC chromatograms detected under the developed mobile phase system at 325 nm. These chromatograms show: (**a**) three marker compounds: 1 = caffeic acid (0.1 mg/mL), 2 = ferulic acid (0.2 mg/mL), 3 = Z-ligustilide (0.1 mg/mL); (**b**) granule 1 (G1) sample (10 mg/mL); (**c**) raw herb 2 (R2) sample (10 mg/mL). AU = absorbance units.


**Table 2.** Calibration curves, detection limits and quantification limits (*n* = 6) of the three chemical markers in Danggui by UPLC-PDA.

Relative standard deviation RSD (%) = 100 × standard deviation (SD)/mean; y, peak area; x, the concentration of each reference chemical marker (mg/mL); R2, coefficient of determination; LOD, limit of detection (3.33 × (SD of y-intercept/mean of slope)); LOQ, limit of quantification (10 × (SD of y-intercept/mean of slope)).

#### *3.3. Multivariate Analysis Using HCA and PCA*

According to the dendrogram generated from HCA, the majority of the samples were divided into two main clusters. Specifically, R1, R3, R4, R5, G7 and G10 (relatively higher amounts of caffeic acid and ferulic acid) were classified into one cluster (Group 1), whereas G1–G4, G6, G8 and G9 were grouped into another cluster (Group 2) representing relatively lower amount of the marker acids. G5 (highest amount of Z-ligustilide) and R2 (highest amount of ferulic acid) were different to these two main groups (Figure 2).

**Figure 2.** Hierarchical agglomerative clustering analysis (HCA) dendrogram of Danggui samples using SPSS 20.0 software (Chicago, USA). Ward's method as amalgamation rule and the squared Euclidean distance as metric were employed to set up the clusters. The length of the linkage between each sample/group represents the degree of similarity. G: granule samples; R: raw herbs/decoction piece samples. Group 1: R1, R3, R5, G7, R4, G10; Group 2: G1, G2, G3, G4, G6, G8, G9. R2 and G5 are outliers.

PCA was also performed to determine the main chemical markers influencing the equivalence of Danggui raw materials and granules. Based on eigenvalues > 1, the first two principal components (PC), PC1 and PC2, were used to differentiate the samples according to the input data. From the result, the first two PCs could explain 53% and 47% of the variance of the three chemical markers, respectively. According to the loading matrices from the PCA biplot, the test samples were separated in PC1 by the differences in the chemical content of ferulic acid and caffeic acid, whilst PC2 was mainly due to the chemical content of Z-ligustilide. Similar to the results of hierarchical clustering, two major groups are set up in the PCA biplot (Figure 3). The decoction pieces and G7 (Group 1) were in close proximity and showed a higher content of caffeic acid and ferulic acid, with R2 demonstrating the highest amount of caffeic acid and ferulic acid. G5 was considered as an outlier of the samples due to its excessively high amount of Z-ligustilide. The PCA loading plot indicates that the Z-ligustilide content may have more

influence on the discrimination of G5 and G10. The rest of the granules (Group 2) were near each other and represented generally lower amounts of the three chemical markers.

**Figure 3.** Biplot from principal component analysis (PCA) of Danggui samples (PC1 vs. PC2) based on the three components Z-ligustilide, caffeic acid and ferulic acid using Unscrambler 10.3 from Camo AS software (Trondheim, Norway). In the biplot, a point represented each individual sample, and the distance allocated between samples, revealed the degree of their similarity in terms of the content of the chemical markers. PC: principal component. G: granule samples; R: raw herbs/decoction piece samples. Group 1: R1, R3, R5, G7, R4; Group 2: G1, G2, G3, G4, G6, G8, G9.

The HCA plot of the three markers was compared to using a single marker and the combination of the markers (Supplementary Material). Ferulic acid as the sole marker grouped G5 with the granules, showing no distinct difference (Figure S1). The HCA of the two markers (ferulic acid and Z-ligustilide) showed similar results to the original three marker HCA, with R2 showing more similarity to the decoction pieces (Figure S2). Using Z-ligustilide, all samples were grouped together, with G5 on its own (Figure S3). Using caffeic acid, there were three groups: G7 and G10 grouped with R2; G4 and G5 was with the rest of the herb/decoction pieces, with G5 showing some similarity with the granules (Figure S4).

#### *3.4. Antioxidant Activity*

In the DPPH and FRAP assays (Table 3), all the samples showed antioxidant activity, with R2 (highest amount of ferulic acid) showing significantly higher activity in both assays using independent-samples t-test (*p* = 0.017 and 0.002, respectively), whereas G5 (highest amount of z-ligustilide) was comparable (*p* = 0.421 and 0.483, respectively). A significant difference in antioxidant activity was shown between the herb/decoction piece samples as a group and the granules as a group (*p* < 0.05) in the FRAP assay. For the two major groups established by HCA and PCA, DPPH and FRAP antioxidant activities were compared by independent-samples t-test and found a significant difference (*p* = 0.027) between Group 1 and 2 in the FRAP assay.

Pearson correlation coefficient analysis investigated the correlation between antioxidant activity and the chemical markers of all the Danggui samples (Table 4). Positive and significant correlations were observed between the amount of ferulic acid and the antioxidant activities of the FRAP assay (0.791, *p* < 0.01). Caffeic acid showed significant correlation with the antioxidant activities of Danggui as measured by the DPPH assay (0.582, *p* < 0.05). In contrast, the amount of Z-ligustilide and samples' antioxidant activity was negatively and not significantly correlated to either DPPH and FRAP assay (−0.202 and −0.229, *p* > 0.05).


**Table 3.** Trolox equivalent (TE) of the granule and herb/decoction piece samples of dried weight (DW) using DPPH and FRAP assays, respectively.

<sup>a</sup> Values were the average of triplicate tests; SD = standard deviation.

**Table 4.** Pearson correlation between the three chemical markers and antioxidant activities of the samples in the DPPH and FRAP assays.


\* Correlation is significant at the 0.05 level (two-tailed). \*\* Correlation is significant at the 0.01 level (two-tailed).

#### **4. Discussion**

As demand grows for traditional Chinese medicines, so does the need for efficient ways of administrating herbal medicines. Thus, it is important to compare new formulations such as granules to the original herb. This is the first study that compares Danggui granules to the raw products. The yield (reported as percentage of the dry weight of the herb) is indicative of the herb dosage a patient is consuming. The yields were higher for the herb/decoction pieces and were lower than the 48% water-soluble extractives in the Hong Kong Chinese Materia Medica Standards (HKCMMS) [25]. In comparison, the yield of the granule products was lower. Granule size was nonuniform in the samples, and this will affect the extraction process above. Smaller particles may be extracted more efficiently or be missed as they make their way to the bottom of the bottle. To minimise this variability, each granule bottle was shaken before sampling to redistribute the particles [26,27].

The quality control of traditional Chinese medicines and their products is a challenge for industry due to the complexity of the formulations (using a holistic approach to treat disease), as well as high outlay costs for analytical instrumentation. Danggui has a complex chromatogram because of the number of individual constituents, the possible degradation and isomerisation of the organic acids and phthalides present [28,29]. Qualitative approaches such as thin layer chromatography (TLC) are highly recommended by pharmacopoeias and monographs to compare fingerprints of samples; however, it is does not usually allow for the quantification of compounds which will confirm their quality [25,30]. For this quality study on Danggui, the visual analysis of the TLC result failed to accurately determine the quantitative difference between the compounds of the samples as the LOD and concentrations of

the compounds were low and close to signal noise, and calibration curves could not be constructed (data not shown).

Other studies have used UPLC coupled with MS to investigate the chemical profile of *A. sinensis* [15,31]. This analysis would have expensive set up costs for examining herbal material. To separate the polar and non-polar constituents within a reasonably short running time, a UPLC condition was determined and optimised in this study. By adding 1% formic acid to water, the solvent system showed good separation of the constituents simultaneously, with a run time of 30 min (compared to 60 min for high-performance liquid chromatography methods) and good separation [32]. The optimised UPLC method and resulting chromatograms were able to quantify caffeic acid, ferulic acid and Z-ligustilide.

The disparity of the ferulic acid, caffeic acid and Z-ligustilide content between the granule samples indicates differences in the manufacturing processing of Danggui which may not mimic traditional water decoctions of TCM. It is interesting to note that R2 was a raw herb sample rather than a decoction piece (which has gone through a processing procedure such as smoke-drying). Its extraction in water favoured the polar compounds such as the organic acids. However, it has been reported that techniques such as steam distillation and other solvents such as ethanol may be used to enhance the extraction of the less polar components in the herb for granule production at the expense of the polar acids such as ferulic acid. Spray or vacuum drying may be used for heat sensitive compounds and for compounds in trace amounts [33]. For water insoluble compounds such as Z-ligustilide, one company mentions the use of carbon dioxide extraction [34]. Studies have established pharmaceutical approaches using methanol and hexane extraction to obtain a high content of Z-ligustilide as a lead compound for pharmacological studies [17,28,35]. Another issue could be adulteration, in which the extracts may be spiked with the known marker compound to reach the regulatory amount [14].

As differences in the chemical content of Danggui products could affect their efficacy, the identification and quantification of chemical markers is necessary to determine the quality of Danggui granules. In the study, the amount of ferulic acid in the samples (0.003–0.05%) was less than the 0.05% minimum requirement as stated in the PPRC for the quality assessment of Danggui [9]. In the PPRC, 70% ethanol is the nominated solvent, with no less than 45% ethanol-soluble extractives. This solvent will give a different chemical profile compared to water as a solvent which is used in home decoctions. However, no information is given in the PPRC regarding the standard amount of caffeic acid and Z-ligustilide for Danggui. The monograph for Radix Angelicae Sinensis states that a "sample contains not less than 0.6% (of Z-ligustilide) calculated with reference to the dried substance" [25]. In this study, Z-ligustilide in most of the samples were lower than the monograph and the standard range of 0.5–5% as reported by the WHO which is based on 100% methanol as the solvent [14,28]. In addition, the amount of Z-ligustilide detected in the extracts (0.005–0.526 mg/g DW) was lower due to the extraction in water (traditional decoction) than the content (1.26–37.7 mg/g from non-water solvents) found in previous studies [17,29,36].

HCA and PCA differentiated the Danggui samples based on the contents of caffeic acid, ferulic acid and Z-ligustilide on their own and in combination. In this study, a minimum of two chemical marker compounds (ferulic acid and Z-ligustilide) was required to differentiate Danggui products. This agrees with our previous findings where five rather than the nominated three chemical markers were required to differentiate raw and granule products of *Panax notoginseng* [7]. Thus, it is recommended that the WHO, the PPRC and other pharmacopoeias/monographs incorporate a minimum percentage value of at least two chemical standards, which will reflect traditional water extracts.

DPPH, along with FRAP, are commonly used to measure antioxidant activity and the methods with other herbal products have been widely published. Unlike biological cell assays, these assays have stable reaction responses and are cheap and quick for industry use. The contribution of caffeic acid and ferulic acid to antioxidant activity of Danggui was confirmed in a previous study [37]. Our findings indicated that phenolic acids such as ferulic acid are the key determinants influencing the antioxidant activities of Danggui as found in a previous study [38]. Thus, the chemical content of ferulic acid is an important chemical marker to ensure the correlation of the antioxidant activities to the different Danggui samples. One study revealed that Danggui extracts prepared with either water or 20% ethanol with an extraction time of 15 min yielded the best antioxidant activity [39]. As expected, Z-ligustilide did not correlate to antioxidant activity.

A limitation of the present study is that the Danggui was sourced from one region which is the region recommended for quality Danggui. As expected, the results showed that the decoction pieces were consistent in composition. Future studies could include comparing granules to raw decoctions in clinical trials to gauge clinical efficacy.

#### **5. Conclusions**

In the present study, UPLC coupled with multivariate analysis and antioxidant activity provided a rapid method for assessing differences in Danggui products. Comprehensive quality standardisation processes in pharmacopoeias and monograph publications are required to guide the regulation and standardisation of the production of commercial herbal granules. With the increased use of herbal medicinal granules around the world, this study will provide important information for standardisation committees, industry, practitioners and consumers on the quality control of herbs and its medicinal products. It is vital that patients are better informed about their health and treatment choices and are aware of what they are consuming. More importantly, practitioners will need to determine the correct dosages for their patients so that they do not undermine the efficacy of the herb and the patient's care. Thus, granule dosages would need to equate to the decocted raw product.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2305-6320/7/6/35/s1, Hierarchical agglomerative clustering analysis (HCA) dendrograms of Danggui samples using SPSS 20.0 software (IBM, Chicago, IL, USA). Dendrograms show different combinations of the markers. Ward's method as amalgamation rule and the squared Euclidean distance as metric were employed to set up the clusters. G: granule samples; R: raw herbs/decoction piece samples. Figure S1: Ferulic acid as the single marker, Figure S2: Ferulic acid and Z-ligustilide as the two markers, Figure S3: Z-ligustilide as the single marker, Figure S4: Caffeic acid as the single marker.

**Author Contributions:** Conceptualisation and supervision, K.C. and V.R.-N.; methodology, all authors; validation, H.Y.W., X.Z. and V.R.-N.; formal analysis, H.Y.W. and X.Z.; investigation, H.Y.W.; resources, V.R.-N., X.Z., J.P., A.K., K.C.; writing—original draft preparation, H.Y.W., V.R.-N. and K.C.; writing—review and editing, V.R.-N., X.Z. and K.C.; project administration, V.R.-N. and K.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** The project was supported by The Joint Chair in Traditional Chinese Medicine (JCTCM) Program, funded by the Office of Science and Research in NSW, the University of Sydney and Western Sydney University, Australia.

**Acknowledgments:** The authors would like to thank from the University of Sydney: Ka Ho Wong (analytical assistance), and from Western Sydney University: Leila Hejazi (LC/MS unit manager), Paul Fahey (statistical support), Samiuela Lee (general laboratory support) and Alan Bensoussan (Director of NICM) for his never-ending support and acquirer of the original program funding.

**Conflicts of Interest:** The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. As a medical research institute, NICM Health Research Institute receives research grants and donations from foundations, universities, government agencies, individuals and industry. Sponsors and donors also provide untied funding for work to advance the vision and mission of the Institute. The authors declare no competing financial interests.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **Uncharted Source of Medicinal Products: The Case of the** *Hedychium* **Genus**

### **Wilson R. Tavares 1, Maria do Carmo Barreto <sup>1</sup> and Ana M. L. Seca 1,2,\***


Received: 10 April 2020; Accepted: 27 April 2020; Published: 28 April 2020

**Abstract:** A current research topic of great interest is the study of the therapeutic properties of plants and of their bioactive secondary metabolites. Plants have been used to treat all types of health problems from allergies to cancer, in addition to their use in the perfumery industry and as food. *Hedychium* species are among those plants used in folk medicine in several countries and several works have been reported to verify if and how effectively these plants exert the effects reported in folk medicine, studying their essential oils, extracts and pure secondary metabolites. *Hedychium coronarium* and *Hedychium spicatum* are the most studied species. Interesting compounds have been identified like coronarin D, which possesses antibacterial, antifungal and antitumor activities, as well as isocoronarin D, linalool and villosin that exhibit better cytotoxicity towards tumor cell lines than the reference compounds used, with villosin not affecting the non-tumor cell line. Linalool and α-pinene are the most active compounds found in *Hedychium* essential oils, while β-pinene is identified as the most widespread compound, being reported in 12 different *Hedychium* species. Since only some *Hedychium* species have been investigated, this review hopes to shed some light on the uncharted territory that is the *Hedychium* genus.

**Keywords:** *Hedychium*; traditional medicine; coronarin D; villosin; anti-acetylcholinesterase; antidiabetic; anti-inflammatory; antimicrobial; antioxidant; antitumor

#### **1. Introduction**

Since the beginning of the history of mankind there was always a connection between plants and human health, as they were used as food and medicines [1]. The traditional herbal medicine outlined the foundations from which modern medicine developed and is still largely practiced around the world [2], particularly in Asian and developing countries [3,4]. This popular knowledge, also known as folk medicine, gives a good indication to scientists looking for sources of new compounds with pharmaceutical potential. Thus, medicinal plants and their derived natural compounds have become an increasing topic of investigation and interest [5,6].

According to "The Plant List" database [7], the genus *Hedychium* (Zingiberaceae family) comprises 93 species with accepted scientific plant names that, with the exception of *Hedychium peregrinum* N.E.Br. that is endemic to Madagascar [8], are native to wooded habitats in tropical and temperate Asia (i.e., China, Indian subcontinent and Southeast Asia) [8–10]. Members of this genus are well distributed worldwide, being easily found particularly throughout tropical Asia, Australia, Fiji, New Caledonia, New Guinea, New Hebrides, Samoa and the Solomon Islands [8,10,11], with some species being considered invasive in some places: e.g., *Hedychium coronarium* J. Koenig in Brazil [12] and *Hedychium gardnerianum* Sheppard ex Ker-Gawl. in Azores Archipelago [13] and Hawaii [14].

*Hedychium* species are medium-size rhizomatous perennial monocotyledonous plants that can be easily recognized by their characteristic striking foliage and terminal spikes that produce diversified numerous short-lived flamboyant flowers with several hues and fragrances varying depending on the species [15]. These features give them a high ornamental value, being cultivated worldwide mostly for this purpose and for its use in the perfumery industry, since, besides the aromatic flowers, *Hedychium* species rhizomes also originate strongly scented oils [16,17].

The use of *Hedychium* species in folk medicine is common in several countries since they are easily harvested directly from nature or obtained at local markets [18]. These plants are reported to possess analgesic, antimicrobial, antidiabetic, anti-inflammatory, antitumor, anti-allergic, anthelmintic and antioxidant properties [19–22]. In Table 1, it is summarized the different *Hedychium* species with reported traditional medicinal use in literature over different geographic areas.


**Table 1.** *Hedychium* species with reported traditional medicinal use.


In addition to the traditional medicinal uses stated in Table 1, *Hedychium* species are also included in the diet of some populations, like in Thailand where the flowers of *Hedychium forrestii* Diels can be boiled to become a beverage [45] or in India where the fruit of *H. spicatum* may be cooked and eaten with lentils in savory dishes [42]. Moreover, the rhizome of *H. coronarium* is also included in the diet of some populations of South East Asia, being consumed as a vegetable or as a food flavoring spice [46].

The traditional uses mentioned above show that several *Hedychium* species are used to treat a wide spectrum of diseases. These uses also show that *Hedychium* species should be considered as promising sources of new bioactive natural compounds and that is why these species have been the target of research by the scientific community. In recent years, several studies have been published on the phytochemical characterization of *Hedychium* species, as well as on the evaluation of the biological activities exhibited by their organic extracts, essential oils and pure compounds, with some of them showing very interesting results. Recently, literature reviews have been published focusing only on specific species, i.e., *H. coronarium* [20,47] and *H. spicatum* [21,48]. This work aims to update the available information that were not mentioned in the previous reviews, as well as involving all the other *Hedychium* species, their bioactivities and their bioactive isolated compounds. The research for this review was made combining the terms *Hedychium*, phytochemical and biological activities in the databases Web of Science, PubMed and Scopus and were considered only the published works involving *Hedychium* species whose binominal Latin name is an accepted name on the The Plant List database [7].

### **2. In Vitro and In Vivo Activities of** *Hedychium* **Extracts and Essential Oils**

Taking into account the traditional uses of *Hedychium* species, several works have been carried out to elucidate how effectively plants can exert the reported biological effects. The following is a compilation and discussion of the most current works on this subject, in which essential oils and extracts of *Hedychium* species are studied and their biological activities are ascertained.

#### *2.1. Anti-Acetylcholinesterase*

The inhibition of the enzyme acetylcholinesterase (AChE) is one of the pathways to countering the cholinergic deficit associated with cognitive dysfunction diseases like in Alzheimer's disease [49]. Arruda and colleagues [50] showed that the leaf essential oil of *H. gardnerianum* collected from four different locations could inhibit AChE action, mainly mixed inhibition, presenting IC50 values ranging from 1.03 ± 0.14 mg/mL to 1.37 ± 0.27 mg/mL, a value not statistically different from the value displayed by the AChE inhibitor standard compound α-pinene that presented an IC50 value of 1.43 ± 0.07 mg/mL. This work showed no statistically significant difference between the activity of samples taken in different geographical areas [50].

#### *2.2. Antidiabetic*

Deficiency in insulin secretion, insulin action or both, results in chronic hyperglycemia, the main characteristic of diabetes mellitus [51], the main treatment to this condition being the use of anti-diabetic drugs that can control glucose levels in the blood [52].

An in vivo study [53] was carried out to assess the effect of *H. coronarium* aqueous extract to lower blood glucose level in induced-type 2 diabetes mellitus (T2DM) animal models (streptozotocin (STZ)-induced T2DM Wistar rats and C57BKSdb/db mice, a mice model with a mutation that results in chronic hyperglycemia, pancreatic beta cell atrophy, low insulin level and obesity). After 28 days, the daily dose of *H. coronarium* aqueous extract (8.928 mg/kg for the STZ-induced T2DM rats and 17.71 mg/kg for the C57BKSdb/db mice) significantly increased glucose tolerance in both diabetic models, when compared with the group treated with distilled water (control group). In addition, the treatment also helped to maintain optimal β-cell structure, moderately increased insulin, improved the lipid profile and decreased aldosterone level in STZ-induced T2DM model.

In another in vivo assay [54], after 14 days of treatment, using an oral dose of 0.3 mL of essential oil from rhizomes of *H. spicatum*, was observed the reduction of blood glucose and urea levels in rats with diabetes induced by intraperitoneal injection of a solution of alloxan monohydrate (150 mg/kg). This result is similar to those obtained in the group of rats treated with the reference drug glibenclemide. Furthermore, it was noticed that the Islets of Langerhans regained their normal shape after the treatment period [54].

#### *2.3. Anti-Inflammatory*

Inflammation is a vital defense mechanism that works to ensure good health [55], but uncontrolled inflammation may lead to serious repercussions [56] and so it is important to continue research into products that can help in its control.

An in vivo study [57] with rats demonstrated the anti-inflammatory effect of a single oral dose (200 mg/kg) of aqueous and ethanolic extracts of *H. spicatum* rhizome against carrageenan-induced paw edema. Measurements of the edema volume were taken in a successive interval of 1 h, 2 h and 3 h and significant decrease in paw edema volume was detected since the beginning, with the aqueous extract reporting a 28.10% decrease in inflammation and the ethanolic extract a 25.62% decrease in inflammation. Although none of the extracts performed as well as the positive control compound indomethacin (41.32% decrease in inflammation), they both proved to present no acute toxicity in a concentration as high as 2000 mg/kg, with the rats never showing secondary toxic effects like coma, convulsion, salivation, increased motor activity or death. This dose of 2000 mg/kg was previously utilized in a similar work [58] where the ethanolic extract of *H. spicatum* reported a 55.54% of anti-inflammatory activity inhibition against carageenan-induced edema in rats.

#### *2.4. Antimicrobial*

A healthy human body is a symbiosis between human and microbial components [59]. However, sometimes that symbiotic balance can be disturbed, and human health can be impaired by pathogenic microorganisms (i.e., bacteria, fungi, parasites or viruses), the use of effective antimicrobial drugs being needed to restore health normality [60,61].

Noriega et al. [62] showed that, among five different plants, the essential oil of *H. coronarium* rhizome exhibited the most relevant antibacterial activity against *Listeria grayi* (MIC value = 0.45 mg/mL) and *Streptococcus mutans* (MIC value = 0.18 mg/mL) and even against the Gram-negative bacteria *Klebsiell oxytoca* (MIC value = 0.90 mg/mL). The authors point out the compounds 1,8-cineole and terpinen-4-ol as responsible for the reported activity [62]. In another work [63], *H. coronarium* leaves essential oil was also pointed out to have antibacterial activity against different bacterial strains, i.e., *Escherichia coli* (MIC value = 3.90 μL/mL), *Staphylococcus aureus* (MIC value = 7.81 μL/mL) and *Pseudomonas aeruginosa* (MIC value = 15.62 μL/mL). These two works are presented here also as examples of two constraints which are common in a variety of scientific papers. First, no work reports, as a comparative term, the activity exhibited by a standard antibacterial compound, determined under the same experimental conditions as the essential oil samples. Without these data it is very difficult to assess the true potential of the samples tested. Second, the MIC values are expressed in non-comparable units. Fortunately, one of the works [62] presents the density of the essential oil, making it possible to convert one of the sets of results [against *Listeria grayi* (MIC value = 0.50 μL/mL), *Streptococcus mutans* (MIC value = 0.20 μL/mL), *Klebsiell oxytoca* (MIC value = 1.0 μL /mL)], allowing to conclude that the essential oil from rhizome is more active as antibacterial agent than leaves essential oil. Regrettably, some papers do not present enough experimental data to allow a unit conversion. Additionally, Ray et al. [64] reported that the essential oil extracted from the rhizome of *H. coronarium* is an effective antifungal agent since it exhibited activity against *Candida albicans* (MIC = 3.12 μg/mL), *Aspergillus flavus* and *Fusarium oxysporum* (MIC value of 6.25 μg/mL for both species), these MIC values being much lower than those reported for antibacterial activity by Noriega et al. [62].

Another work [65] found that 20 μL of *Hedychium matthewii* S. Thomas, B. Mani & S. J. Britto rhizome essential oil could be as effective as 30 μg of the standard antibiotic amoxicillin, since it exerted nearly the same growth inhibition effect against several strains of Gram- positive and Gram-negative bacteria (*viz. Bacillus cereus*, *Staphylococcus aureus*, *Enterobacter aerogens*, *Salmonella paratyphi*, *Salmonella typhii*, *Escherichia coli*, *Vibrio parahaemolyticus*, *Proteus vulgaris*, *Klebsiella pneumoniae* and *Pseudomonas aeruginosa*). Furthermore, it could be pointed out that *Streptococcus haemolyticus* and *Vibrio cholerae* were more susceptible towards the essential oil (20 μL) than towards amoxicillin (30 μg).

The activity of *H. spicatum* flowers essential oil was evaluated against the Gram-negative bacteria *Borrelia burgdorferi* in stationary phase cycle and it was found out that a 0.1% (v/v) essential oil concentration could eradicate *B. burgdorferi* (100 μL) with no regrowth [66]. This is one of the few published works that evaluates the antibacterial activity in the stationary-phase of growth.

A different work [67] found that a combination treatment using essential oil of *H. spicatum* rhizomes and γ-radiation was effective against *Fusarium graminearum*, inhibiting both the fungal growth in maize grains and the production of the toxic mycotoxins deoxynivalenol and zearalenone in a dose-dependent way, with a complete inhibition at the concentration of essential oil 1.89 mg/g and 4.1 kGy of γ-radiation. Combinational treatment proved to be better than individual treatment, since complete inhibition of *F. graminearum* required the essential oil concentration of 3.15 mg/g or 6 kGy of γ-radiation.

It is not just the essential oils of *Hedychium* species that have been evaluated concerning antimicrobial activity. Arora and Mazumder [68] evaluated the activity of *H. spicatum* rhizomes methanolic extract and the antibiotic ciprofloxacin against different bacterial strains (*viz. Shigella boydii*, *Shigella soneii*, *Shigella flexneri, B. cereus, V. cholerae*, *E. coli, S. aureus, Ps. aeruginosa* and *K. pneumoniae*) at the concentrations of 200 to 1200 μg/mL. The results showed a similar inhibition effect for both antibiotic and extract, *B. subtilis* being the bacteria with greater susceptibility to the extract and antibiotic.

Another work [69] evaluated the anthelmintic activity of methanolic, ethanolic, hydromethanolic, hydroethanolic and aqueous rhizome extracts of *H. spicatum* against *Hemonchus contortus*, with the results showing that the methanolic extract were as effective as the positive control compound thiabendazole on time taken for paralysis and time taken for death (tested concentrations 20, 40 and 60 mg/mL).

#### *2.5. Antioxidant*

Oxygen metabolism is fundamental for human life but its reaction products, like reactive oxygen species (ROS), can increase oxidative stress, causing damage to cells and tissues [70] that, with time, leads to the development or aggravation of several chronic diseases [71]. Thus, therapeutic antioxidant agents are key to mitigate the oxidative stress impact in human health, with natural plant-derived products being the main investigation focus of search [72].

Noriega et al. [62] evaluated the antioxidant activity of the essential oil extracted from the rhizome of *H. coronarium*, reporting IC50 values of 9.04 ± 0.55 mg/mL and 2.87 ± 0.17 mg/mL for 1,1-Diphenyl-2-picrylhydrazyl (DPPH) and 2,2 -azino-bis(3-ethylbenzothiazoline-6-sulphonic acid (ABTS) assays, respectively. In a similar work, Ray and colleagues [64] also evaluated the antioxidant activity of the essential oil from *H. coronarium* rhizome, but from ten distinct regions of India, obtaining activity values higher than those indicated in the work of Noriega et al. [62] (IC50 values range from 0.57 to 2.19 mg/mL for the DPPH assay; and 0.12 to 0.67 mg/mL for the ABTS assay), but lower than the positive control 2,6-di-tert-butyl-4-methylphenol (BHT) (IC50 = 0.12 ± 0.01 mg/mL on the DPPH assay, and 0.08 ± 0.01 mg/mL on the ABTS assay). It could be pointed out that Ray et al. [64] also demonstrate, very clearly, that the geographical origin of the samples is a relevant variable for the level of activity displayed. The same conclusion can be drawn from the results obtained by Arruda et al. [50], where the DPPH antioxidant activity of *H. gardnerianum* leaf essential oil collected from four different locations ranged from EC50 = 8.46 ± 0.90 μg/mL to 31.14 ± 2.70 μg/mL (EC50 = 31.00 ± 0.19 μg/mL for BHT). In a more recent work, Ray et al., [73] studied the antioxidant activity of *Hedychium greenii* W. W. Smith. and *Hedychium gracile* Roxb. rhizomes essential oils by the same methodology (DPPH and ABTS assays), with *H. greenii* showing higher antioxidant activity (IC50 values of 16.73 ± 0.19 μg/mL for DPPH and 12.18 ± 0.16 μg/mL for ABTS assays) than *H. gracile* sample (IC50 values of 46.94 ± 0.6 μg/mL for DPPH and 31.13 ± 0.29 μg/mL for ABTS assays), and slightly higher than the positive control BHT (IC50 = 18.94 ± 0.3 μg/mL and IC50 = 14.21 ± 0.27 μg/mL for DPPH and ABTS assays, respectively). These results [73], when compared with those obtained in the works mentioned above [50,64], show that the level of antioxidant activity of essential oils exhibits variability between different *Hedychium* species (IC50 values range from 8.46 to 2190 μg/mL) higher than geographical variability (IC50 values range from 0.57 to 2.19 mg/mL for the DPPH assay).

Zhao et al. [74] compared essential oils and ethanolic extracts from rhizomes of different species from the Zingiberaceae family in terms of its antioxidant capacity by DPPH assay. The ethanol extracts of *H. coronarium* and *H. gardnerianum* proved to be the best antioxidant samples presenting IC50 values of 0.94 μg/mL and 1.59 μg/mL, respectively, even better than the reference compounds trolox (IC50 = 10.19 μg/mL) or ascorbic acid (IC50 = 8.37 μg/mL). Essential oils of these plants were also tested but unfortunately the authors presented the results as a graphic which does not allow the reading of numerical values of antioxidant activity.

Usha et al. [75] compared the hydromethanolic rhizome extract of different species also from Zingiberaceae family in terms of its antioxidant capacity and found out that *Hedychium* sp. reported the best results, with the lowest IC50 value on DPPH assay (36.4 μg/mL). This activity was correlated with its high phenol and flavonoid content. Unfortunately, the authors do not specify neither the *Hedychium* species that was used nor the IC50 value of the ascorbic acid used as positive control, which makes impossible to compare with other published works.

Another work [69] evaluated, through ABTS, DPPH and nitric oxide (NO) free radical scavenging assays, the antioxidant activity of methanolic, ethanolic, hydromethanolic, hydroethanolic and aqueous rhizome extracts of *H. spicatum*. The results showed the methanolic extract as the most antioxidant extract, presenting the lowest EC50 values for all the assays (EC50 ABTS value = 24.93 mg/mL, EC50 DPPH value = 8.31 mg/mL and EC50 NO value = 3.57 mg/mL). However, this extract is much less active than the positive control ascorbic acid (EC50 = 1.63 mg/mL to ABTS assay, EC50 = 0.049 mg/mL to DPPH assay and EC50 = 0.10 mg/mL to NO assay) and since the extract EC50 values are very high, it should be considered an inactive extract.

In an in vivo study, Choudhary and Singh [76] demonstrated the antioxidant potential of *H. spicatum* rhizome, since an improvement in the oxidative stress state of white leghorn cockerels (*Gallus gallus domesticus*) was observed after the rhizome powder was added to the animal diet, following chronic exposure to indoxacarb.

#### *2.6. Antitumor*

Cancer is a complex disease that is a major cause of death worldwide [77], with several treatments but no cure [78]. In the light of the aggressive and not always effective treatments in current medicine, the demand for safer and better anticancer compounds have turned the search to natural products as another therapeutic approach to cancer [79].

Ray and colleagues [80] demonstrated the antiproliferative time-dependent effect of *H. coronarium* rhizome ethanol extract against human cervical carcinoma HeLa cells, without affecting the viability of non-tumor human umbilical vein endothelial cells (HUVEC). After 24, 48 and 72 h of incubation, the observed IC50 values were 17.18 ± 0.46, 15.32 ± 0.68 and 12.57 ± 0.32 μg/mL, respectively. Although the positive control drug camphothecin presented a far greater inhibitory effect against HeLa cells (IC50 values of 0.82 to 0.98 μg/mL), it is also more toxic to the HUVEC cells (IC50 value for 24 h = 10.13 ± 0.62 μg/mL) than the *H. coronarium* ethanol extract (IC50 value for 24 h > 320 μg/mL), which means that the extract presents a higher selective cytotoxicity. In addition, the same study shed some light on the mechanism whereby the extract exerts its antitumor activity. It denotes the modulation of the expression of proapoptotic and antiapoptotic protein levels together with an increase of ROS generation and consequent oxidative stress induction in HeLa cells that led to an apoptosis-mediated G1 phase cell arrest as the main cause of HeLa cells migratory capacity inhibition.

In another study [81], the methanolic extract of *H. spicatum* rhizomes was described as possessing a dose-dependent cytotoxicity activity against human liver hepatocellular carcinoma cell line HepG2, testing concentrations in the range of 25 to 3000 μg/mL. The concentrations tested and the IC50 value (281.917 μg/mL) are very high, and the authors do not provide the cytotoxicity of a positive control nor do they evaluate the effects of such concentrations on non-tumor cells. The results obtained in the studies performed in these conditions, should be considered with many reservations as the effects observed using such high concentrations are non-specific. On the other hand, the researchers should take into account that 20 μg/mL is the limit established by the National Cancer Institute to consider an extract active enough to justify continuing its study [82], so the tested extract should be considered inactive against HepG2 cells line.

The in vitro cytotoxicity, by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, of *H. spicatum* rhizome chloroform extract was assessed against colorectal adenocarcinoma (Colo-205) cell line, human epidermoid carcinoma (A-431) cell line, human breast adenocarcinoma (MCF-7) cell line, human lung adenocarcinoma (A549) and Chinese hamster ovary (CHO) cell lines [83]. The results show that the extract presented cytotoxicity against all cell lines exhibiting IC50 values ranging from 37.45 ± 0.90 μg/mL to 63.21 ± 1.19 μg/mL, including against non-tumor cell line CHO (39.52 ± 0.06 μg/mL), indicating that the *H. spicatum* rhizome chloroform extract have small potential as a good anticancer drug since it affected in a similar way both tumor and non-tumor cell lines. Results like these shows how difficult it is to find an ideal anti-tumor drug that affect only the tumor cells, leaving the non-tumor cells undamaged. In addition, it would have been interesting if the authors had also tested a reference compound, since it would have enriched their work.

#### *2.7. Hepatoprotective*

The liver is a vital organ, capable of detoxifying the body from endogenous and/or exogenous substances detrimental to the organism, and which is responsible for the regulation of diverse functions and physiological processes, such as the metabolism of carbohydrates and fats and the secretion of bile [84]. Exposure to drugs and chemicals can cause liver injury which, taking into account all the functions inherent to the liver, is a major health problem [85]. Thus, compounds that can protect the liver, stimulate hepatic function or help to regenerate hepatic cells, while simultaneously being less toxic and more effective are of great interest, with natural sources being identified as good search option [86].

A study [87] indicated that *H. spicatum* possess hepatoprotective properties since its three rhizome extracts (methanolic, ethanolic and aqueous) exerted protection on HepG2 cells against paracetamol-induced toxicity. The IC50 values were 282, 356 and 515 μg/mL for the methanolic, ethanolic and aqueous extracts, respectively, which translates in a cytoprotection percentage of 16%, 13% and 9%, respectively. Compared to the 19% cytoprotection provided by the control substance silymarin (IC50 = 110 μg/mL), the hepatoprotective effect of the extracts is not huge but it is worth mentioning at least the methanolic extract.

A study which was also carried out to evaluate the potential hepatoprotective effect was the in vivo study [88]. where cockerels were fed for 16 weeks with rhizome powder of *H. spicatum*, while simultaneously receiving a dose of indoxacarb intended to cause chronic toxicity. The results of the liver analysis show that, when compared with the control group (indoxacarb administration without the added *H. spicatum* rhizome powder to the cockerels diet), *H. spicatum* rhizome ameliorated the damages caused in cockerels by indoxacarb in the duration of the experiment. Apparently, the treatment with *H. spicatum* modulated the expression levels of several different hepatic genes, such as those involved in metabolization of indoxacarb (cytochrome P450 1A1), in the immune system (interleukin 6 (IL-6)) and in antioxidant function (catalase (CAT), superoxide dismutase (SOD) and glutathione peroxidase (GPx)).

#### *2.8. Insecticide*

Control of mosquito population is crucial, particularly in developing countries, since they act as vectors of several pathogens and parasites responsible for various worrisome diseases, e.g., dengue, filariasis, malaria, West Nile or yellow fever [89,90]. In order to reduce or eliminate the human contact with the vector, a wide range of methods exists with insecticides being a top choice in case of mosquitoes [91]. However, with insecticide resistance being a problem in recent years [92], the search for better substances with insecticide potential is imperative.

Kalimuthu and colleagues [93] carried out an interesting work where *H. coronarium*-synthesized silver nanoparticles (AgNPs) were produced and their toxicity towards larvae and pupae of the dengue vector *Aedes aegypti* was assessed, as well as their synergy with *Mesocyclops formosanus* predation over *A. aegypti* larvae. The toxicity of aqueous *H. coronarium* rhizome extract was also assessed. The results indicate that both *H. coronarium* formulations tested, aqueous rhizome extract and AgNPs, were toxic against *A. aegypti* in a dose-dependent manner. Aqueous *H. coronarium* rhizome extract caused toxicity with LC50 values from 0.688% against larval instar I to 1.882% dose against pupae stage of *A. aegypti*, while AgNPs demonstrated its toxicity with LC50 values varying from 24.264 ppm for larval instar I till 348.68 ppm for pupae of *A. aegypti*. Once again, we are faced with a work whose authors express results in non-comparable units and do not provide the necessary data for their conversion, significantly reducing the impact of this work. Nevertheless, AgNPs were found to be stable over time in aquatic environment and since a positive synergy was reported with *M. formosanus* predation on young *A. aegypti* larvae, its combined use could lead to a higher efficacy in removing the larval population of dengue mosquitoes from aquatic areas.

In another work [94], *Hedychium larsenii* M. Dan and C. Sathish Kumar rhizomes essential oil was evaluated regarding its toxicity against larvae of mosquito vectors of diseases, namely *Anopheles* *stephensi* (malaria), *A. aegypti* (dengue) and *Culex quinquefasciatus* (St. Louis encephalitis). The results demonstrate that the essential oil exerted larvicidal activity over the different larvae with the LC50 values of 82.02, 88.60 and 96.40 μg/mL for *A. stephensi*, *A. aegypti* and *C. quinquefasciatus*, respectively. Again, the lack of a tested reference compound impairs any conclusion taken from these results.

### **3. Secondary Metabolites from** *Hedychium* **Species and Its Activities**

The diverse bioactivities observed on different *Hedychium* species/extracts are intrinsically linked to the compounds present in each one, so the need and interest in the phytochemical study of these extracts/species becomes clear. Several relevant works managed to isolate compounds from *Hedychium* extracts and carried out different assays to ascertain the bioactive potentials of those compounds. In Table 2 the compounds isolated from *Hedychium* extracts are gathered, as well as their bioactivities and the *Hedychium* species where they have already been identified. A figure with the chemical structures of the compounds (Figure 1) listed in this table is present after Table 2. It should be clarified that, for each compound in Table 2, only the highest activity value for each activity from each reference is presented, with some values converted from μg/mL to μM to facilitate comprehension and comparison of the different activities.


**Table 2.** Secondary metabolites isolated from *Hedychium* extracts with proven activities.


**Table 2.** *Cont.*

\* Only the highest activity value; \*\* Value after unit conversion from <sup>μ</sup>g/mL to <sup>μ</sup>M; # MG132-carbobenzoxy-Leu-Leu-leucinal positive control; † The authors do not indicate the extract prepared.

**Figure 1.** Chemical structure of the compounds referred on Table 2.

Taking the information of Table 2 into account, it is possible to identify that *H. coronarium* provided the highest number of isolated compounds and that the antitumor activity is the most reported bioactivity in the above-mentioned studies. On the other hand, the labdane-type diterpene is the most frequent family of compounds in the genus *Hedychium*, and some flavonoids and simple phenolic compounds are also identified.

Villosin (**26**) can be pointed out as the most promising antitumor compound, since it presented a highest and selective cytotoxicity against NCI-H187 cell line with an IC50 value of 0.40 μM, without toxicity against the non-tumor Vero cell line at 166.42 μM and presenting better results than the positive control compound ellipticine (i.e., IC50 value against NCI-H187 of 1.79 μM and IC50 value against Vero of 7.47 μM). Coronarin D (**7**) appears also as one interesting compound, since recent works report its antibacterial activity against *B. cereus* to be better than the positive control oxacillin.

In addition to these compounds, hedyforrestin B (**1**) and hedyforrestin C (**2**) should also be noted, since their antitumor activities against the NCI-H187 cell line are slightly lower (less than 1.7 times) than that shown by the reference compound ellipticine and with selectivity indices of 14.5 and 4.8, respectively.

On the other hand, compound isocoronarin D (**11**) should be highlighted since it exhibits activity against a broad spectrum of tumor cell lines (i.e., A549, human cervical carcinoma (HeLa), human hepatocellular carcinoma (HepG2), human acute promyelocytic leukemia (HL-60), human cholangiocarcinoma (HuCCA-1), human epidermoid carcinoma (KB), human breast adenocarcinoma (MDA-MB-231), human acute lymphoblastic leukemia T-lymphoblasts (MOLT-3), mouse lymphoma neoplasm (P388), human hepatocellular carcinoma (S102) and human hormone-dependent breast cancer (T-47D)), with IC50 values between 2.14 to 36.1 μM, better than etoposide or doxorubicin which are toxic only to some of these cell lines, and being more active against HepG2 (IC50 = 16.6 μM) than the reference compound etoposide (IC50 = 23.8 μM) [98].

Bearing in mind that all these compounds have hydroxyl groups and double bonds in their chemical structure, it is suggested that these compounds could be lead compounds, and researchers in the field of medicinal chemistry should use these labile functional groups to carry out structural modifications, in order to obtain more active derivatives and to determine the structure/activity relationships.

In addition, there are some works which require a critical analysis. Zhao and colleagues [96] isolated six labdanes from *H. longipetalum* rhizome that exhibited NO production inhibitory effects in lipopolysaccharides (LPS) and interferon gamma (IFN-γ)-induced murine macrophages RAW 264.7 cell line. The most active compound is yunnancoronarin A (**6**) (IC50 = 1.86 μM), but less active than the positive control carbobenzoxy-Leu-Leu-leucinal (MG132) (IC50 = 0.17 μM). Unfortunately, the authors do not mention the extraction and chromatographic procedures they carried out to isolate these compounds, which would have been a valuable information.

Another study that looks promising, but which actually shows very questionable results, is the one carried out by Kiem and colleagues [37]. They isolated compounds from rhizomes of *H. coronarium* methanol extract and investigated their anti-inflammatory potential through inhibition of pro-inflammatory cytokines production in LPS-stimulated bone marrow-derived dendritic cells (BMDC). The results are not acceptable and do not allow to infer conclusions since they are presented with associated standard errors greater than 20% (e.g., IC50 IL-6 inhibition value = 7.57 ± 2.02 μM) and in some cases close to 100% (e.g., IC50 IL-12p40 inhibition value = 0.19 ± 0.11 μM). This work [37] was only mentioned here to point out to all authors the need to present reliable data in their works, aiming always to show results with standard error less than 10%.

In other lines of work, several studies (e.g., Reddy et al. [83], Chimnoi et al. [98] and Endringer et al. [101]) assessed the antitumor potential of isolated compounds from *Hedychium* extracts without following the best guidelines for evaluating the cytotoxic potential of compounds. In fact, the authors did not test a reference compound in the same experimental conditions, and did not test the isolated compounds against a non-tumor cell line, which makes it difficult to draw conclusions. Regrettably, without these results, it is not possible to conclude about the efficacy and selectivity of the isolated compound compared to the drugs already available on the market.

In addition to compounds **28** and **29** (Figure 1), Carvalho and colleagues [104] also isolated the compounds 3-(2-hydroxyethoxy)xanthone (**30**) and oplopanone (**31**) from *H. gardenerianum* rhizome acetone extract, but the two compounds (Figure 2) do not present any reported activity and, therefore, were not included in Table 2. Since they belong to families of organic compounds well-known for their broad spectrum of activities (flavonoids and terpenes) [107], it would be worth investigating the biological activity of these compounds.

It is a fact that the availability of a specific compound in a plant can depend on several factors, like the geographic location where the plant developed [108] and/or the season when it was harvested [109]. Thus, different studies can present different percentages of the total content of the same compound which makes it difficult sometimes to make comparisons between the same plants. This fact is particularly relevant with regard to essential oils, where the majority of published studies refers to quantitative chemical analysis. These studies reveal a complex composition and a huge variability in the content of each compound, depending on geographic, seasonal and species factors, which is reflected in the variability of the biological activity level of the respective essential oils, already highlighted in point 2.

**Figure 2.** Chemical structure of the compounds **30** and **31**.

*Hedychium* species are not different, with several compounds being identified with distinct percentages on its essential oils. However, a deeper analysis of the published works allows to identify some compounds that, with some slight differences, appear repeatedly as the most abundant compounds in their essential oils. In Table 3 are gathered the five most abundant compounds identified in essential oils from *Hedychium* species as well as their activities and the species where they have already been identified. The respective structures are presented on Figure 3.


**Table 3.** The five most frequent and abundant chemical compounds identified in essential oils from *Hedychium* species.

**Table 3.** *Cont.*


**Figure 3.** Chemical structure of the compounds referred on Table 3.

As it is possible to see on Table 3, linalool (**35**) proved to have promising antitumor potential since it presented cytotoxicity against U937 cell line (i.e., IC50 = 2.59 μM), better than the positive control 5-FU against the same cell line (i.e., IC50 = 4.86 μM). It would have been interesting if the authors had tested the compounds cytotoxicity against a non-tumor cell line, but unfortunately that was not the case.

From the five most abundant and most frequent present compounds in essential oils of *Hedychium* species, β-pinene (**34**) is the most widespread compound among species being identified in 12 *Hedychium* species, mainly rhizomes but also in some cases from flower and leaf essential oil. The compounds α-pinene (**33**) and linalool (**35**), exhibit a broad range of bioactivities, being anti-acetylcholinesterase, anti-allergic, antidepressive, antidiabetic, anti-inflammatory, antimicrobial, antitumor, fumigant and neuroprotective agents.

The antimicrobial activity of α-pinene (**33**), β-pinene (**34**) reported by Leite et al. [115] presented in Table 3 should be noted, which appears as μL/mL and the authors do not provide the necessary data to convert it to μM or μg/mL. Thus, it is impossible to compare the exhibited activity with other published results and even to compare with the positive control used in this study.

Despite not being so abundant as the compounds referred in Table 3, the isolation of two compounds from *H. larsenii* rhizomes essential oil could be mentioned, i.e., *ar*-curcumene (**37**) and *epi*-β-bisabolol (**38**) (Figure 4), that presented insecticide properties against diseases mosquito vectors larvae *A. stephensi*, *A. aegypti* and *C. quinquefasciatus* [94]. The results show that the most affected vector was *A. stephensi* with compounds **37** and **38** presenting a LC50 values of 51.65 and 66.02 μM, respectively. Unfortunately, the lack of a tested reference compound is a handicap in this work.

**Figure 4.** Chemical structure of the compounds **37** and **38**.

Taking together, Tables 2 and 3 offer a summary view point of the works carried out in recent years that permitted the isolation of some compounds from *Hedychium* genus, being ascertained their bioactivities. This allows to easily identify where there is work already successfully developed and which paths have not yet been explored.

#### **4. Conclusions**

*Hedychium* genus is undoubtedly proven to be a valuable group of medicinal plants, being present in several folk medicines around the world where it is known to treat allergies, cancer, diabetes, inflammation, rheumatism and skin problems, as well as being also used as an analgesic, antimicrobial, anti-helminthic, antioxidant and insect repellent. In addition, some *Hedychium* species are part of human diet, being cooked as a vegetable, used as a spice or drunk as a beverage.

Several works explored *Hedychium* species in order to confirm if and how effectively these plants exert the reported biological effects on folk medicine, studying their essential oils, extracts and their isolated compounds. Taking into account the results of the literature in recent years, *Hedychium* species have been proven to possess interesting pharmaceutical activities, i.e., anti-acetylcholinesterase, antidiabetic, anti-inflammatory, antimicrobial, antioxidant, antitumor and hepatoprotective, as well as having potential to develop insecticides.

Phytochemical works have been carried out in *Hedychium*, mainly on *H. coronarium* and *H. spicatum*, but also on other less known species, leading to the isolation of interesting compounds that, in some cases, proved to be better than reference compounds. An example is coronarin D (**7**), possessing antifungal, antitumor and antibacterial properties, being more effective than the positive control oxacillin against *B. cereus* in antibacterial assays. Isocoronarin D (**11**), villosin (**26**) and linalool (**35**) can be pointed out as very promising antitumor compounds since they exhibited better cytotoxicity towards tumor cell lines than the reference compounds used, and in case of villosin (**26**) without toxicity on non-tumor cell line. Furthermore, the most bioactive compounds found in *Hedychium* essential oils can be highlighted as α-pinene (**33**) and linalool (**35**), since they are reported as presenting a wide spectrum of bioactivities. In addition, being identified in 12 different *Hedychium* species to this date, β-pinene (**34**) is the most widespread compound in *Hedychium* essential oils.

*Hedychium* species as proved to be a very rich genus that can still have a lot to offer to the scientific community. Moreover, the discovery in recent years of four new *Hedychium* species (i.e., *Hedychium chingmeianum* N.Odyuo and D.K.Roy [128], *Hedychium putaoense* Y.H.Tan and H.B.Ding [129], *Hedychium viridibracteatum* X.Hu [130] and *Hedychium ziroense* V.Gowda and Ashokan [131], may bring new compounds with pharmaceutical potential to the equation.

**Author Contributions:** M.d.C.B. and A.M.L.S. conceptualized and revised the paper; W.R.T. conducted the research and wrote the first draft. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by project MACBIOPEST (MAC2/1.1a/289), program Interreg MAC 2014–2020 co-financed by DRCT (Azores Regional Government), supporting W.R. Tavares's grant, as well as by FCT—Fundação para a Ciência e Tecnologia, the European Union, QREN, FEDER, and COMPETE, through funding the cE3c center (UIDB/00329/2020) and the LAQV-REQUIMTE (UIDB/50006/2020).

**Acknowledgments:** Thanks are due to the University of Azores and University of Aveiro.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**



#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review Genista tridentata* **L.: A Rich Source of Flavonoids with Anti-Inflammatory Activity**

#### **Diana C. G. A. Pinto \*, Mark A. M. Simões and Artur M. S. Silva**

LAQV-REQUIMTE & Department of Chemistry, University of Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal; mark.simoes@outlook.com (M.A.M.S.); artur.silva@ua.pt (A.M.S.S.)

**\*** Correspondence: diana@ua.pt

Received: 18 May 2020; Accepted: 29 May 2020; Published: 30 May 2020

**Abstract: Background:** *Genista tridentata* L. is an endemic species from the Iberian Peninsula used in Portuguese traditional medicine to treat inflammation-related diseases; this and other health-promoting effects are usually associated with the flavonoids produced by this species. In fact, anti-inflammatory properties were established for several of these flavonoid derivatives. **Methods:** A careful survey of the reported data, using mainly the Scopus database and *Genista tridentata* and *Pterospartum tridentatum* as keywords, was done. We have examined the papers involving the plant and those about the most relevant flavonoids anti-inflammatory activity. **Results:** The literature survey demonstrates that species are used to treat several health problems such as antihyperglycemia, hypertension, and inflammatory episodes. It was also possible to establish its richness in flavonoid derivatives, from which several are potential anti-inflammatory agents. **Conclusions:** From our described and discussed analysis, it can be concluded that *Genista tridentata* is an excellent source of bioactive flavonoids. Moreover, its traditional use to treat inflammation episodes may be due to its flavonoid content, from which genistein, biochanin A, rutin, and daidzein can be emphasized.

**Keywords:** *Genista tridentata*; *Pterospartum tridentatum*; isoflavones; flavonols; anti-inflammatory; genistein; biochanin A; rutin; daidzein

#### **1. Introduction**

Inflammation is a natural defense mechanism involved in the body's healing process, in which the body is protected from pathogens or abnormal cells [1]. However, if the inflammation is prolonged in time or serious, it can damage the healthy tissues and cause several diseases, such as cancer [2], Alzheimer's and Parkinson's diseases [3]. Therefore, the development of new anti-inflammatory drugs is still a demand, and plant secondary metabolites are considered a priority—in particular, those found in medicinal plants [4].

Among the plants used in Portuguese traditional medicine, *Genista tridentata* L. can be highlighted due to the important applications reported [5]; in fact, the plant, locally named carqueja, is in several regions called the "plant that heals everything" [5], and among its applications is the use to treat inflammatory diseases [6].

Flavonoids, a large family of natural compounds, are usually associated with anti-inflammatory activity [7], and most recently, we demonstrated that *G. tridentata* is rich in flavonoid derivatives [8], including some for which anti-inflammatory activities have been described. As examples, genistein, daidzein [9,10], and biochanin A [11,12] can be highlighted.

The most promising anti-inflammatory flavonoids that can be isolated from *G. tridentata* will be discussed in this review, emphasizing their mode of action and in vivo studies. Hopefully, this will help the scientific community to understand their involvement in inflammatory processes and consequently endorse the design for novel derivatives. Furthermore, the traditional medicine applications of

*G. tridentata* will also be addressed and discussed. To accomplish this survey, we used mainly the Scopus database (69 articles), but also Web of Science (61 articles) and PubMed, mostly for the anti-inflammatory activity. The keywords used were the accepted name (*Genista tridentata*), the most common synonym (*Pterospartum tridentatum*), and also the less common one (*Chamaespartium tridentatum*). Naturally, in the survey there were also the flavonoid names combined with anti-inflammatory activity. Relevance was given to the most recent biological evaluations and the in vivo studies and the clinical trials. In all cases, the papers involving both the plant and the most relevant flavonoids anti-inflammatory activities.

### **2.** *Genista tridentata***: Traditional Applications and Biological Activities**

*Genista tridentata* L. is a bush endemic to the Iberian Peninsula where it grows wildly. Unfortunately, its taxonomy is a little controversial, and consequently, the literature survey is more complicated. The most found scientific name is *Pterospartum tridentatum* (L.) Willk., which is considered by some taxonomists [13] as the correct name, but other authors used *Chamaespartium tridentatum* (L.) P.E. Gibbs [14]. However, according to the Plant List database [15], these are synonyms of *Genista tridentata* L. and there are eleven other synonyms and three infraspecific *taxa* [15]. However, in our survey, only the abovementioned synonyms were found—it seems that the other synonyms and infraspecific *taxa* are not used in articles involving chemical profile and/or anti-inflammatory evaluations. Although we used all names in the literature survey, herein, we will refer the species by the accepted name reported in the Plant Lista database [15].

*Genista tridentata* is an Angiosperm belonging to the Leguminosae family [15], which grows spontaneously under Mediterranean thermal conditions, where it is known as carqueja [16]. *G. tridentata* is a perennial shrub that can reach up to one meter in height, with stems of woody and rigid consistency. The roots are well-liked and quite long and sometimes intertwine in the roots of other companion species. The stems are woody, erect or prostrate with laterally winged branches, forming false leaves of dark green color, cut out and of coriaceous consistency. Thus the branches have a flattened shape with two or three wing-shaped expansions, with an articulated appearance, ending with two or three teeth. The leaves, persistent, alternating, unifoliolate and triangular, appear to be tridentate, by the leaflets being united to the stipulations. The flowers are of an intense yellow and are arranged in corymbiform inflorescences, in groups of 3 to 10, gathered in small and tight bouquets. They have an induction in the sepals that line them. The fruit is an oblong-linear pod 10 to 12 mm long [17].

Despite the abovementioned disagreement in the *G. tridentata* taxonomy, the vernacular designation, carqueja, is referred to in the ethnopharmacological surveys. Consequently, it is possible to mention here that *G. tridentata* is used in the Iberian Peninsula, particularly in Portugal, in traditional medicine, mainly to treat influenza, cold, cough, stomach troubles, and nervousness, and is also used as a tonic, hepatic protector, sedative, cicatrizant, and diuretic [6,14,18,19]. In these applications, the population mainly uses extracts of the plant flowers, leaves, or the aerial parts. Consequently, it is suggested that the plant presents several therapeutic properties, from which antispasmodic, antihypertensive, and anti-inflammatory properties can be emphasized [6,14].

The flowers are used in folk medicine for the treatment of various disorders, including those relating to the respiratory system, digestive tract, nervous system, urinary system and dermatology; it has also been indicated for diabetes control [16,20] and is sometimes used in mixtures with other plants for this purpose [20]. Some authors referred to the use of *P. tridentatum* for the treatment of colds, stomach pains, intestinal problems, kidney disease, liver and gallbladder problems and also for rheumatism [21]. It was also indicated for pneumonia, bronchitis and tracheitis, headaches, cough, for low blood pressure levels and high levels of cholesterol, diabetes and even in weight loss programs. This species is known for its diuretic, purgative, laxative, hypotensive, hypoglycemic effects, and for its digestive properties [14,22]. The infusion of dried flowers is considered an excellent emollient [21].

One vital point that should be herein mentioned is the obligation to have scientific validations of the claimed properties, an aspect that it is not at all strange to the scientific community [23]. In this regard, several evaluation studies involving *G. tridentata* extracts were reported and will be herein presented and discussed. Most of the studies were performed using the flowers or the aerial parts extracted with polar solvents and in vitro antioxidant evaluations (Table 1).


**Table 1.** Biological assays of *Genista tridentata* extracts.



AA, antioxidant activity; GI50, values correspond to the concentration that causes 50% inhibition of cell proliferation; IC50, values corresponded to the extract concentration that inhibits in 50% the oxidation and inflammatory process; MIC, minimum inhibitory concentration.

The authors achieved the extract antioxidant activity index or antioxidant potential through several assays, from which DPPH• (2,2-diphenyl-1-picrylhydrazyl radical) scavenging assay and β-carotene bleaching test are the most common. However, it is interesting to note that some authors used other less common tests, such as lipid peroxidation inhibition, through the decrease in TBARS (thiobarbituric acid reactive substances) [26,29–31], and, more recently, the oxidative hemolysis inhibition assay [31]. These diversifications in the assays are, in our opinion, very good because they can establish in more detail the *G. tridentata* health-promoting potential. Altogether, the reported results show that this species presents moderate to strong antioxidant activity, and apparently, the flower extracts and the water extracts are more active [25,27].

Another interesting feature in these reports is the fact that all authors obtained the total phenolic content and/or the total flavonoid content, and some established the polyphenolic profile or identified some of the phenolic compounds present [25,28–31]. In doing so, they associated the antioxidant activity to the polyphenolic content. On the other hand, some aspects of these reports are less enthusiastic, since the reported values are in different units; the positive controls used are different, making it impossible to perform comparisons.

Other evaluations, such as antifungal [32], antibacterial [31,33] agents, cytotoxicity activity in tumor and non-tumor cells [31], and even the immunostimulatory activity of the *G. tridentata* polysaccharides [34] were also performed. Additionally, Ferreira et al. also performed in vivo and in vitro toxicological assays and concluded that short-term use is safe [8,28].

The anti-inflammatory evaluation of the *G. tridentata* extracts and mainly those reports which were recently achieved [8,31,35] will be the focus of this review. In the most recent evaluations, the authors tested parts of the plant separately and established that the anti-inflammatory effects of plant extracts could occur through different mechanisms. Moreover, the roots, which are not used in traditional medicine, also presented strong anti-inflammatory activity [8]. Likewise, the antioxidant and anti-inflammatory activity is associated with the species richness in polyphenolic compounds, particularly flavonoids.

### **3. Structural Pattern of the Flavonoids Isolated from** *Genista tridentata*

Several authors demonstrated that *G. tridentata* produces several flavonoids; these metabolites are those that most contribute to the plant anti-inflammatory activity. Therefore, herein the flavonoids that were isolated from *G. tridentata* extracts or identified in will be discussed.

From the several established profiles, it is evident that the only classes of flavonoids detected were isoflavones **1**, flavones **2**, flavonols **3**, flavanones **4** and flavanonols **5** (Figure 1), and the major ones are isoflavones and flavonols (Tables 2 and 3).

**Figure 1.** Structure of the classes of flavonoid derivatives found in *G. tridentata*.

**Table 2.** Isoflavones and flavones produced by *G. tridentata*.


Glc = glucoside unit; GlcA = glucuronide unit; Ac = acetyl.

**Table 3.** Flavonols produced by *G. tridentata*.


Glc = glucoside unit; Rha = rhamnoside unit.

As far as we could find, the first report on the *G. tridentata* flavonoids allowed the isolation of four isoflavone derivatives **1a** to **1d**, and one flavonol **3a** (Tables 2 and 3) [20]. Four years later, the same research group found two other isoflavones **1e** and **1f**, and flavonols **3b** and **3c**, (Tables 2 and 3) [36]. The first flavone derivatives were just reported in 2012 and were luteolin derivatives **2a** and **2b**

(Table 2) [28]. Flavanonols were just uncovered, for the first time, in 2014 and are taxifolin derivatives, whereas flavanone derivatives were only reported in 2020 (Figure 2) [8,29].

**Figure 2.** Structure of some flavonoid derivatives found in *G. tridentata*.

Our literature survey showed that the compounds indicated in Table 2 were found by several authors, with the exception of isoflavones **1i** and **1j**, and all the flavone derivatives that were just reported once [8,28,33]. Furthermore, through the analysis of Table 2, it is possible to detect that most of the flavonoids present one or more hydroxy groups and almost all are linked to saccharide units. Usually, this substitution pattern is associated with the anti-inflammatory property of a flavonoid [37].

It is important to complement the information listed in Table 2 with the information that other isoflavone glycosides were described, namely biochanin A hexoside [8,29,31,33] and genistein hexoside [8,31], but the authors did not identify the hexose nor its position in the isoflavone ring. There are also references describing the presence of a methylbiochanin A or a methylprunetin [8,29–31] derivative. In all of these cases, although this is important information about the *G. tridentata* profile, it was not included in Table 2 because its structure is not fully established. This suggests that some investment in phytochemical studies involving *G. tridentata* extracts is still needed.

The substitution pattern of flavanonol derivatives includes several hydroxy groups and glucosides, as well as a disaccharide unit. The most referred derivatives were isoquercitrin **3a** and rutin **3c** (Table 3), and again we are in the presence of compounds having the required substitution pattern for being promising anti-inflammatory agents [37]. Additionally, quercetin hexoside derivatives were also found, but the authors were again unable to identify the hexose or its position [29–31].

Finally, we can find the identification and isolation of flavanonols and flavanones (Figure 2). Some authors have reported the presence of taxifolin [33] or its glucosides [8], whereas others just mention hexoside derivatives [29–31]. One fact is consistent—*G. tridentata* produces taxifolin derivatives. The last examples were recently reported and apart from being slightly different, they were isolated from the plant roots [8], which is also uncommon due to the fact that most of the works were performed using flowers or aerial parts. This highlights that some parts of the plant should still be studied.

#### **4. Flavonoids with Anti-Inflammatory Activity**

In the previous section, we showed the richness of at *G. tridentata* in flavonoids; additionally, the major class, that is isoflavones, is commonly associated with beneficial anti-inflammatory properties [10]. Yu et al. discussed, in their excellent review [10], the possible isoflavones anti-inflammatory mechanisms, of which herein we highlight the main points (Table 4). Still, we suggest that our readers consult the original review for details. According to the authors, isoflavones may be involved in the scavenging of reactive oxygen species and, in doing so, they prevent the production of peroxynitrite, species that can oxidize low-density lipoproteins. With this effect, isoflavones can prevent cell membrane damage. However, they can also act by inhibiting the production of pro-inflammatory cytokines and chemokine species such as *IL-1*β, *IL-6*, *IL-12* and *TNF-*α, or by inhibiting pro-inflammatory enzymes, such as cyclooxygenase, nitric oxide synthases, lipoxygenase and phospholipase A2, enzymes involved in the production of inflammatory mediators. Finally, there is also evidence that isoflavones can be involved in the regulation of *NF-*κ*B* factor signaling and, through that regulation, decrease the production of pro-inflammatory cytokines (Table 4) [10,38,39].

Spagnuolo et al. discussed the flavonoids neuroprotective potential, in particular flavonols, another family well represented in *G. tridentata* [40]. There is some evidence, at least in in vitro studies, that these flavonoids reduce neuroinflammation also by regulating important signaling pathways such as *NF-*κ*B* and MAPKs (Table 4) [40].


**Table 4.** Anti-inflammatory effects of the selected flavonoids.


Considering all these pieces of evidence and the fact that several flavonoids were found in *G. tridentata*, we selected some significative examples to discuss their anti-inflammatory potential, and Table 4 summarizes the effect and mechanism of action of the selected flavonoids.

#### *4.1. Biochanin A and Prunetin*

Biochanin A **1h** and prunetin **1d** are isomeric natural isoflavones (Figure 3) produced by *G. tridentata* not as the major components, but in small amounts, 4.8% (μg/g) for biochanin A **1h** and 4.1% (μg/g) for prunetin **1d** [29]. Some derivatives are also reported, and in particular, the methyl derivative that was not fully identified [29]; in fact, if there is no evidence of mass spectra fragments containing the characteristic A ring fragment [8] or the compound was isolated [20], it is possible to confuse these isomers. One fact is consistent—*G. tridentata* produced one or both.

**Figure 3.** Biochanin A **1h** and prunetin **1d** structures.

As far as we could find, prunetin **1d** was isolated for the first time in 1952 from *Pterocarpus angolensis* DC. [85] and biochanin A **1h** was isolated from *Cicer arietinum* L. in 1945 [86]. Although these isoflavones' natural occurrence seems to be similar, from the biological evaluation point of view, biochanin A **1h** has been extensively studied, and several health benefits were attributed to its consumption as well as its possible use to develop new drugs [87,88], and anti-inflammatory activity is among those biological properties.

In this century, several evaluations regarding the biochanin A **1h** anti-inflammatory activity have been performed (Table 4), and the first example is the study of Kalayciyan et al. [41], in which the compound potential to treat the Behçet's disease was established. The main anti-inflammatory effect of the compound is to decrease the secretion of interleukin-8 (*IL-8*), a potent leukocyte chemotactic factor known to induce inflammation [41]. More recently, it was also proved that biochanin A **1h** inhibits the *IL-8* expression in lipopolysaccharides (LPS)-stimulated human vascular endothelial cells in a dose-dependent manner [42], as well as in focal cerebral ischemia/reperfusion in rats [43]. The biochanin A **1h** effects on other interleukins levels, such as *IL-1*β, *IL-6*, *IL-10*, and *IL-18*, were evaluated in the last feew years, with *IL-1*β being the most studied one [44–54]. All these studies proved the inhibitory effect that biochanin A **1h** has on these inflammatory cytokines. However, the most important aspect is the fact that some of the studies were performed in vivo [43,46,49–52], which is a forward step to establish this compound pharmacological potential.

The inhibition of another important pro-inflammatory species, such as TNF-α, was also evaluated by several authors [42–49,52–56], as well as the inhibiting pro-inflammatory enzymes [49,55] and key phosphorylation steps [44,48,56,57]. All of these studies suggested that biochanin A **1h**'s anti-inflammatory effect occurs by suppressing the pathways NF-κB and MAPK [53,56–58], but is also associated with the up-regulation of PPAR expression [43,45,53,54]. Prunetin **1d**, a much less studied compound, also presents potent in vitro [56,59,60] and in vivo [59] anti-inflammatory activity, and apparently, its mechanism of action is also associated with the inhibition of the NF-κB pathway [59].

It should be highlighted that several of the studies mentioned above included the evaluation of cytotoxic effects, and all demonstrated that both isoflavones do not affect the viability of the cells, and in the subsequent tests the authors used noncytotoxic concentrations. From these studies, essential facts arose—prunetin **1d** should be subjected to more evaluations. Moreover, pharmacodynamic and pharmacokinetic parameters of both isoflavones should be evaluated in order to implement some clinical trials in the future.

#### *4.2. Daidzein*

Daidzein **1j** (Figure 4) is a natural isoflavone with a significant occurrence, mainly in fruits and nuts [89], which is the reason why humans are exposed to it and also to its health benefits [90]. In fact, several pharmacological properties are attributed to this isoflavone [91], including anti-inflammatory potential [10,91]. Although daidzein **1j** occurrence in *G. tridentata* is rare, only one report on its identification was reported (Table 2), we decided to include here the most recent works on its anti-inflammatory activity, since its occurrence seems to be exclusively in the plant roots [8]. This fact gives importance to that part of the plant, while importance usually is only given to the flowers and aerial parts, which are the ones used traditionally.

**Figure 4.** Daidzein **1j** structure.

The most recent studies involved in vivo studies with daidzein **1j**—the reasons why are herein highlighted. Due to its occurrence in common fruits [89], daidzein **1j** is present in mankind's diet, and it is a nontoxic compound [52]. These recent studies confirmed daidzein **1j**'s strong anti-inflammatory activity as well as settling on its mechanism of action (Table 4). Mainly, daidzein **1j** strongly affects various pathways, including NF-κB, p38MAPK, and TGF-β1. Regardless of this potential as an anti-inflammatory drug, as far as we could find, daidzein **1j** is not involved in clinical trials.

#### *4.3. Genistein*

Genistein **1e** (Figure 5), like daidzein **1j**, occurs naturally in everyday food, such as fruits and nuts [89], and as far as we could find, it is non-toxic for humans [92], which was also recently reinforced by Kumar et al. [93]. The pharmacological potential of genistein **1e** is well documented [94]; more recently, an overview regarding their mechanism of action in cancer models was published [95], and in some aspects, the anticancer and the anti-inflammatory activities are associated.

**Figure 5.** Genistein **1e** structure.

Regarding the anti-inflammatory activity studies, it should be emphasized that, recently, there are more in vivo studies, meaning that scientists are interested in giving this natural isoflavone new medicinal applications. From the reported results, we select a few (Table 4) that demonstrategenistein **1e**'s potential to become an anti-inflammatory drug.

It can be seen that like the isoflavones mentioned above, genistein **1e** targets the same pathways, with an emphasis on the upregulation of the PPARγ signaling pathway and downregulation of the NF-κB signaling pathway, as well as the decrease in several inflammatory mediators (Table 4). In light of the referred studies, genistein **1e** is a candidate to be used in the prevention or treatment of inflammation-related diseases. For example, it could be used to target microRNAs, which is considered a therapeutic target for liver disease. In fact, the results show that the anti-inflammatory activity of genistein **1e** downregulated microRNA expression of liver inflammation [70] but also pro-inflammatory cytokines species such as *IL-1*β and *TNF-*α [70,73]. Another interesting example is its ability to attenuate NF-κB inflammatory signaling in the brain with consequent inhibition of pro-inflammatory cytokines release, which gives genistein **1e** the possibility to become a new drug able to relieve chronic sleep deprivation's adverse effects [71]. Furthermore, there is some evidence supporting that genistein **1e** can, through its anti-inflammatory activity, prevent cardiovascular diseases [74]. Altogether, these findings suggest that genistein is a good candidate for future clinical trials.

#### *4.4. Rutin*

Rutin **3c** (Figure 6) is amongst the most found flavonoids in *G. tridentata* (Table 3), for which several biological and pharmacological properties have been established and reviewed through the years [96–100]. Some more specific activities, such as antidiabetic effects [101], reestablishment of the immune homeostasis [96,102], neuroprotective effects [98,103,104] and anticancer effects [98,99,105] were also addressed. Furthermore, some toxicological studies were also performed [98,106] as well as pharmacokinetic [98], bioavailability [99] and formulation development [100]. It should be emphasized that the mentioned properties prompted some clinical trials using rutin **3c** [107,108] and although the results are not remarkable, they at least confirm that it is safe to use rutin **3c**.

**Figure 6.** Rutin **3c** structure.

Obviously, rutin **3c**'s anti-inflammatory activity was also evaluated and several interesting results were reported (Table 4). It is known that in general, flavonoids decrease the production of pro-inflammatory interleukins, mainly IL-1β, IL-6, and IL-8, but also tumor necrosis factor α (TNF-α). There is evidence that rutin **3c** anti-inflammatory mechanism also involves the downregulation of these pro-inflammatory species [76–80]. The results show that rutin **3c** can also exert its anti-inflammatory activity through other mechanisms (Table 4), from which can be highlighted the inhibition of the HMGB1 signaling pathway through the downregulation of TLR4 and RAGE expressions [75] and also the inhibition of the MPO activity [80]. The last one is an important example because it provides evidence that rutin **3c** can be a possible therapeutic agent for autoimmune diseases [80].

Collectively, the results demonstrate that rutin **3c** attenuates inflammation through several mechanisms and is a nontoxic compound, so clinical trials more focused on its anti-inflammatory potential should be implemented. In this regard, Kalita and Das [109] studied the efficiency of a rutin **3c** formulation to be used in the treatment of inflammations through the long-term delivery via the skin. Their results, although preliminary, are sufficiently good to encourage future investigations.

#### *4.5. Taxifolin*

Our last example is taxifolin (Figure 2), which, as shown in the previous section, occurs in *G. tridentata,* mainly linked to sugar moieties. Nevertheless, we specify here some interesting studies due to the fact that in a living organism, it is possible to obtain the aglycone. The taxifolin anti-inflammatory potential has been known, at least, since 1971 [110] and recently Sunil and Xu published an interesting review on taxifolin's health benefits [111]. Some important aspects arose from this review: the first is the broad biological potential of taxifolin, mainly using in vitro evaluations, but also that the anti-inflammatory and toxicological evaluations are still scarce. The few examples (Table 4) suggest that its mechanism of action is similar to the one reported for the other flavonoids, that is also mainly targets the NF-κB and MAPK pathways. Although, the anti-inflammatory assessments are scarce, they suggest taxifolin's potential to be a drug candidate for the treatment of inflammations, suggesting that it should be further investigated.

#### **5. Conclusions**

This survey demonstrates beyond any doubt that *G. tridantata* is a source of bioactive metabolites, some of which present interesting anti-inflammatory activities which, in turn, contribute to the extracts' anti-inflammatory activity. Amongst our findings, the toxicological evaluations of both extracts and pure compounds are important and contribute to establishing *G. tridentata*'s medicinal value as well as the secondary metabolites' pharmacological value. However, in our opinion, some efforts on the plant taxonomy should be made to prevent confusion in the data reported. Moreover, we think that an extra effort on clinical trials, mainly concerning the pure compounds used as drugs, should be performed.

**Author Contributions:** D.C.G.A.P. and M.A.M.S. performed the literature survey; D.C.G.A.P. and A.M.S.S. conceived and wrote the paper. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** Thanks are due to the University of Aveiro and FCT/MCT for the financial support for the LAQV-REQUIMTE (UIDB/50006/2020) through national founds and, where applicable, co-financed by the FEDER, within the PT2020 Partnership Agreement.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**



#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **Genus** *Stachys***: A Review of Traditional Uses, Phytochemistry and Bioactivity**

#### **Ekaterina-Michaela Tomou, Christina Barda and Helen Skaltsa \***

Department of Pharmacognosy and Chemistry of Natural Products, Faculty of Pharmacy, School of Health Sciences, National & Kapodistrian University of Athens, Panepistimiopolis, Zografou, 15771 Athens, Greece; ktomou@pharm.uoa.gr (E.-M.T.); cbarda@pharm.uoa.gr (C.B.) **\*** Correspondence: skaltsa@pharm.uoa.gr; Tel.: +30-2107274593

Received: 11 August 2020; Accepted: 25 September 2020; Published: 29 September 2020

**Abstract: Background**: The genus *Stachys* L. (Lamiaceae) includes about 300 species as annual or perennial herbs or small shrubs, spread in temperate regions of Mediterranean, Asia, America and southern Africa. Several species of this genus are extensively used in various traditional medicines. They are consumed as herbal preparations for the treatment of stress, skin inflammations, gastrointestinal disorders, asthma and genital tumors. Previous studies have investigated the chemical constituents and the biological activities of these species. Thus, the present review compiles literature data on ethnomedicine, phytochemistry, pharmacological activities, clinical studies and the toxicity of genus *Stachys*. **Methods**: Comprehensive research of previously published literature was performed for studies on the traditional uses, bioactive compounds and pharmacological properties of the genus *Stachys*, using databases with different key search words. **Results**: This surey documented 60 *Stachys* species and 10 subspecies for their phytochemical profiles, including 254 chemical compounds and reported 19 species and 4 subspecies for their pharmacological properties. Furthermore, 25 species and 6 subspecies were found for their traditional uses. **Conclusions**: The present review highlights that *Stachys* spp. consist an important source of bioactive phytochemicals and exemplifies the uncharted territory of this genus for new research studies.

**Keywords:** *Stachys* L.; traditional uses; pharmacological activities; phytochemicals; bioactive compounds

#### **1. Introduction**

The genus *Stachys* L., a large member of the Lamiaceae family, comprises more than 300 species, dispersing in temperate and tropical regions of Mediterranean, Asia, America and southern Africa [1–3]. Up to now, the most established and comprehensive classification of the genus is introduced by Bhattacharjee (1980), categorizing into two subgenera *Betonica* L. and *Stachys* L. [2,3]. The subgenus *Stachys* includes 19 sections, while the subgenus *Betonica* comprises 2 sections [1]. However, the two subgenera present important botanical and phytochemical differences which differentiate them [1,4,5].

*Stachys* species grow as annual or perennial herbs or small shrubs with simple petiolate or sessile leaves. The number of verticillate ranges from four to many-flowered, usually forming a terminal spike-like inflorescence. Calyx tubes are tubular-campanulate, 5 or 10 veined, regular or weakly bilabiate with five subequal teeth. Corolla has a narrow tube, 2-lipped; upper lip flat or hooded and generally hairy, while the lower lip is 3-lopped and glabrous to hairy. The nutlets are oblong to ovoid, rounded at apex [6].

The genus name derived from the Greek word «stachys (=στα´χυς) », referring to the type of the inflorescence which is characterized as "spike of corn" and resembles to the inflorescences of the species of genus *Triticum* L. (Gramineae). In ancient times, the name "stachys" referred mainly to the species *Stachys germanica* L. whose inflorescence is like an ear and is covered with off-white trichome [7]. The Latin name of the genus is trifarium (=tomentose) [8].

Historically, Dioscorides mentioned the species *S. germanica* L. with the name "stachys" [9]. However, in late Byzantine era, 'Nikolaos Myrepsos' included some species of the genus *Stachys* (*S. germanica* L., *S. o*ffi*cinalis*(L.) Travis, *S. alopecuros*(L.) Benth.) in his medical manuscript "Dynameron". Precisely, *S. o*ffi*cinalis* and *S. alopecuros* were probably included in 11 recipes, under the names vetoniki, drosiovotanon, lauriole, kakambri, while *S. germanica* was added in 1 recipe referred as stachys [10].

Many species of the genus are extensively used in traditional medicine of several countries, having various names. For instance, the species *S. recta*, known as yellow woundwort, is called as "erba della paura" (="herb that keeps away fear") in Italy, attributing to the anxiolytic properties of its herbal tea, while *S. lavandulifolia* Vahl is called as "Chaaye Koohi" in Iran [11–13]. In addition, herbal preparations of *Stachys* spp. are widely consumed in folk medicine to treat a broad array of disorders and diseases, including stress, skin inflammations, stomach disorders and genital tumors [3,14,15]. Specially, the herbal teas of these plants, known as "mountain tea", are used for skin and stomach disorders [12,16]. The latter common name could lead to a misinterpretation since the herbal remedies of any *Sideritis* species are globally known with the same name.

In the international literature, *Stachys* species have been broadly studied through several phytochemical and pharmacological investigations, justifying their ethnopharmacological uses. Of special pharmacological interest are considered the anti-inflammatory, antioxidant, analgesic, renoprotective, anxiolytic and antidepressant activity [3,17–19]. The range of the therapeutic properties attributed to these species have been associated to their phytochemical content. Therefore, genus *Stachys* has received much attention for the screening of its bioactive secondary metabolites from different plant parts. In general, more than 200 compounds have been isolated from this genus, belonging to the following important chemical groups; terpenes (e.g., triterpenes, diterpenes, iridoids), polyphenols (e.g., flavone derivatives, phenylethanoid glycosides, lignans), phenolic acids and essential oils [3,5,14,20–22].

Consequently, plants of genus *Stachys* are considered a great source of phytochemicals with therapeutic and economic applications. Given the increasing demand for natural products, many *Stachys* species have been cultivated for uses in traditional medicine, in food market, in cosmetic industry and for ornamental reasons [21,22]. Despite the widely uses of the specific species and the large amount of research studies, there has been no recent comprehensive review including all the latest data of the specific genus and its contribution in medicine. Up to now, the available reviews are centered to the phytochemical profile and biological activities of *Stachys* spp. in correlation to chemotaxonomy approach [3,21–23]. Thus, this review summarizes the current state of knowledge on the traditional uses, phytochemistry, pharmacological activities, clinical studies and toxicity of the genus *Stachys* L.

#### **2. Materials and Methods**

A comprehensive search on previous studies was conducted on scientific databases such as PubMed, Scopus, Google scholar and Reaxys, including the years 1969–2020. The search terms "Stachys", "Stachys compounds", "Stachys phytochemicals", "Stachys pharmacological" and "Stachys traditional uses" were used for data collection. Searches were performed for other potential studies by manual screening references in the identified studies. In total, 161 publications describing the traditional uses, bioactive compounds, pharmacological properties and the toxicity of the genus *Stachys* were included, excluding articles focuses on taxonomy, botany and agronomy. The traditional medicinal uses of *Stachys* species were reported in Table 1, while the isolated specialized products were categorized by species in Tables 2–15, with the attempt of the discrimination between publications describing metabolites isolation (including NMR data) or identification/screening (by means of HPLC, LC-MS, etc.). The chemical structures of the bioactive compounds were showed in Tables 16–29. The reported biological activities of extracts/compounds of the last five years were mentioned by *Stachys* species in Table 30. The general characteristics of the analyzed studies in the current review are showed in Table 31. According to recent publications which support the division of the genus *Stachys* based on Bhattacharjee (1980), the classification in the present review is formed on this latter

study. The species name and their synonyms are quoted as reporting in databases "Plant list" or "Euro + Med" or "IPNI" [24–26].

### **3. Traditional Medicinal Uses of Genus** *Stachys*

Several *Stachys* spp. have been used in various ethnomedicines for thousands of years. A plethora studies mentioned their diverse traditional medicinal uses. In the current review, a detailed description of the available data of the traditional uses of *Stachys* spp. is shown in Table 1, reporting 25 species and 6 subspecies of this genus. A careful overview of the specific table reveals that the ethnomedicinal use of *Stachys* spp. is particularly in the area covering of Mediterannean to Iran. Most of the species are consumed as herbal teas for the treatment of infections, common cold, gastrointestinal disorders, inflammation, skin disorders/wounds, asthma and anxiety.

The species *S. a*ffi*nis*is widely used in Chinese traditional medicine for several uses such as common cold, heart disease, pain relief, antioxidant activity, ischemic brain injury, dementia and gastrointestinal related diseases [27–30]. Another species applied in Chinese folk medicine is *S. geobombycis*, known as DongChongXiaCao, which is recommended as tonic and interestingly, this species is also used in Europe and Japan [22].

In Iran, several species are applied as traditional therapeutic agents in various conditions, including *S. acerosa* [31], *S. fruticulosa* [32], *S. byzantina* (known in Farsi as "lamb's ear" or "lamb's tongue" or "sonbolehe noghrehi" or "zabanehe bare") [33–35], *S. inflata* (local names; poulk or "Ghol-e-Argavan") [31,36,37], *S. lavandulifolia* (known as "Chaaye Koohi") [12,13,31,38–44], *S. pilifera* [31,45], *S. schtschegleevii* [32,34,46], *S. sylvatica* [47] and *S. turcomanica* [34]. Of considerable interest, *S. sylvatica* (common name "hedge woundwort") is recommended for the treatment of women with polycystic ovary syndrome (PCOS) [47].

Furthermore, in Turkish folk medicine, the species *S. cretica* subsp. *anatolica*, *S. cretica* subsp. *mersinaea*, *S. iberica* subsp. *georgica*, *S. iberica* subsp. *stenostachya, S. kurdica*, *S. lavandulifolia* and *S. obliqua* are used mainly to treat colds, cough, stomach ache and as antipyretic agents, while *S. sylvatica* is applied in cardiac disorders [22,48–50].

In Italy, the infusions of the leaves of *S. annua* and *S. recta* are used to wash the face to reveal headache [51], whereas the aerial parts of the subspecies *S. annua* subsp. *annua*, known as "stregona annual" or "erba strega", are consumed as anti-catarrhal, febrifuge, tonic and vulnerary [52]. The decoction of the aerial parts of *S. recta* is also consumed as purative and for bad luck/spirit [53,54]. Interestingly, *S. annua* and *S. arvensis,* as well as the subspecies *S. recta* subsp. *recta* are applied against evil eye [11,51,52,55]. Moreover, in an area of central Italy, the species *S. o*ffi*cinalis* is used as oily extract to treat wounds and to dye wood yellow [29,54]. To be mentioned that *S. recta* is listed in the European Pharmacopeia, as well as *S. o*ffi*cnalis* is mentioned in Anthroposophic Pharmaceutical Codex (APC) [22]. However, Gören (2011) reported that some species (e.g., *S. annua, S. recta and S. sylvatica*) have been mentioned to be poisonous [22].

In North Greece, the infusion and decoction of *S. iva* are consumed against common cold and gastrointestinal disorders [56]. In addition, Fazio et al. (1994) reported different formulations of the Greek species *S. mucronata* applied in Greek tradition medicine. Precisely, the decoction of this species is consumed as an antirheumatic and antineuralgic agent, as well as the juice of fresh leaves is applied in wounds and ulcers. Moreover, the infusion of fresh leaves has antidiarrhoic effect, while the infusion of roots is purgative [57].

In addition to traditional medicinal uses, some species of genus *Stachys* are also consumed as edible plants, vegetables and food additives like the tubers of *S. a*ffi*nis* (known as Chinese artichoke/chorogi; China/Japan) in China and Japan [22,27], the aerial parts of *S. lavandulifolia* in Iran [31], or the *S. palustris* in Poland [22,58]. The latter species is also included in the diet in Sweden, Ukraine and Great Britain [22]. Moreover, the dried powder of *S. palustris* is used as an additive for bread in Europe, thus it is known as "mayday flour" [22].

*Medicines* **2020**, *7*, 63

The infusion of the aerial parts of *Stachys* sp. LAM is used as traditional remedy for colic, gases and swollen stomach in Peru [22,59]. It is noteworthy to mention that a few species have been used in veterinary such as *S. germanica* and *S. o*ffi*cinalis* [30,54].


*Stachys*specieswithreportedtraditionalmedicinal




#### *Medicines* **2020** , *7*, 63



#### *Medicines* **2020** , *7*, 63

#### **4. Chemical Composition**

Various non-volatile chemical constituents have been reported from different species of genus *Stachys*, categorizing into important chemical groups including fatty acids, alkaloids (e.g., stachydrine, turiaine), triterpenes, phytosterols, phytoecdysteroids, diterpenes, iridoids, flavonoids, phenylpropanoid glucosides, acetophenones, phenylethanoid glycosides, lignans, phenolic acids, megastigmanes and polysaccharides [3,20,21,23,67]. The present survey was focused on all the above groups, excluded fatty acids and alkaloids due to the limited available studies. This section summarizes the phytochemicals from the genus *Stachys* which are mainly responsible for its pharmacological benefits, presented in Tables 2–15. To be mentioned that large number of phytochemicals were mainly discovered from the aerial parts, leaves and a few were found in stems and roots.

**2.**Flavonesisolatedfrom*Stachys*



**Table 2.***Cont.*


**Table2.***Cont.*


**Table2.***Cont.*



**Table 2.***Cont.*




