Study on Three Sarcocapnos Species as Potential Sources of Bioactive Compounds: Relation between Phenolic Content and Bioactivity by Multivariate Analysis

In this work, we report the phenolic composition and bioactivity of the aerial parts of three species of Sarcocapnos (S. enneaphylla, S. pulcherrima, and S. saetabensis) to study their potential as sources of bioactive compounds to revalorize them and contribute to the conservation of these plant species. Samples were collected in different locations in the province of Jaén (southeast of Spain), and qualitative and quantitative analyses of phenolic compounds were performed by high-performance liquid chromatography with diode array and mass spectrometry detection. S. enneaphylla presented the highest concentration of phenolic compounds (58 mg/g DE). The most abundant compound in S. enneaphylla and S. saetabensis was rutin (35 mg/g DE and 11.7 mg/g DE, respectively), whereas isorhamnetin-O-rutinoside was dominant in S. pulcherrima (11.5 mg/g DE). Several assays were performed to evaluate the potential bioactivity of the three species of Sarcocapnos. These assays included antioxidant and radical scavenging (ABTS and DPPH), reducing power (CUPRAC and FRAP), phosphomolybdenum and metal chelating, and enzyme inhibitory activity (acetylcholinesterase, amylase, butyrylcholinesterase, glucosidase, and tyrosinase). In general, all methanolic extracts presented the highest phenolic and flavonoid contents, as well as the highest radical scavenging, antioxidant, and enzyme inhibitory properties. This relationship between phenolics and bioactivity was confirmed by multivariate analysis.


Introduction
Plants are valuable sources of bioactive compounds of outstanding interest for the pharmaceutical industry. Due to the side effects of synthetic drugs, the interest in natural medicinal plants has increased in the last decade. Phenolic compounds are among the most important phytochemicals of plants due to their important health benefits [1]. e use of natural products is not only exclusive for human health but also for applications in livestock and aquaculture production systems [2]. It is thus important to characterize the composition and bioactivity of phytochemicals in lesserknown plants.
Sarcocapnos DC. (subfam. Fumarioideae, fam. Papaveraceae) is a plant genus endemic to the Western Mediterranean subregion (southwest Europe and northwest Africa) which has been scarcely investigated so far. Papaveraceae plants have been widely used due to their medicinal properties. Among them, several Sarcocapnos species have been consumed as infusions for the treatment of several illnesses. Sarcocapnos genus is composed of ten taxa: seven species and three subspecies [3,4]. Its main diversification center lies in the Iberian Peninsula, from which eight out of the ten accepted taxa are endemic [5]. Except for the northwest African endemism S. crassifolia (Desf.) DC., the rest of the six species are present in Spain. Some of them show a narrow distribution area circumscribed to the Baetic mountains of Southern Iberian Peninsula (Spain), mainly distributed in the area of eastern Andalusia [6].
All Sarcocapnos species are present in overhanging cliffs, mainly over carbonated substrata. Most of them are endangered species [7], usually threatened by their own biological and demographic features, natural processes (droughts and rock collapse), global climate change, and at a lesser extent, by human activities such as quarrying or climbing. Only a few of them are currently protected by law at a national or regional level. erefore, studies concerning their chemical composition and bioactivity are important to revalorize and to protect these plant species.
e genus Sarcocapnos has proved to be a rich source of isoquinoline alkaloids, most of them with a cularine skeleton [8][9][10]. However, to our best knowledge, the phenolic composition of Sarcocapnos species has not been reported to date. Also, there are no studies regarding the bioactivity of these plant species. We have therefore studied the phenolic composition and bioactivity of three species (all of them suffruticose chasmophytes) that coexist in the province of Jaén (southeast of Spain): S. enneaphylla, S. pulcherrima, and S. saetabensis.
S. enneaphylla (L.) DC. has the widest distribution area of the genus, from Southern France to Northern Africa, with a great number of populations throughout Spain [11]. Its flowers and leaves show the smallest dimensions of the genus. S. pulcherrima Morales and Romero García is an endemic species to southeastern Spain, with isolated populations in Eastern Andalusia [12]. And finally, S. saetabensis Mateo and Figuerola species seems to have a hybrid origin from S. enneaphylla and S. pulcherrima [13]. It is an endemic species to Eastern Spain from Catalonia to eastern Andalusia. e alkaloid compositions of the three Sarcocapnos species have been previously reported [8,9,14].
Considering the scarce data concerning the mentioned species, the first goal of this work is to detail the composition of the phenolic content of aerial parts of these Sarcocapnos species as well as their antioxidant activity and enzyme inhibitory properties against cholinesterase, amylase, glucosidase, and tyrosinase. e final aim is to develop novel pharmaceutical and food products based on the extracts of these plants (or specific isolated bioactive compounds), therefore valorizing these species as a source of bioactive compounds and hence contributing to their conservation.

Materials and Methods
All chemicals and reagents are given in File S1 of Supplementary Materials.

Plant Material and Sample Preparation. Sarcocapnos
leaves were collected by hand from different locations (two locations per plant species) in the province of Jaén (Southeast of Spain) as shown in Table 1. Plant material of each location is a pool of the aerial parts collected from approximately five different plants during the same day. Figure 1 shows the photographs of each Sarcocapnos species in their collection site. e taxonomical classification was confirmed by botanist Dr. Carlos Salazar-Mendías, and voucher numbers are also given in Table 1.
Extractions were carried out in two different media: methanol (MeOH; HPLC grade; Sigma-Aldrich) and water (Milli-Q Waters purification system; Millipore; Milford, MA, USA). Due to the small amount of plant material (most Sarcocapnos species are endangered), we selected MeOH and water, the most common solvents for the extraction of phenolic compounds. For both extractions, leaves were lyophilized (ModulyoD/23, ermo Savant; Waltham, MA, USA) and crushed with a grinder.
Methanol extractions were performed as follows: 2.5 g of dry material was extracted with 50 mL MeOH in an ultrasonic liquid processor (Qsonica Sonicators; Newton, CT, USA) with a power of 55 W and a frequency of 20 kHz, for 10 min (using 50% power) at room temperature. Extractions were done in triplicate. After sonication, solutions were filtered through Whatman No.1 filters. e solvent was evaporated under reduced pressure in a Hei-Vap Precision rotary evaporator (Heidolf; Schwabach, Germany) at 40°C. Dried extracts (DEs) were stored at -20°C until analysis.
On the other hand, the extractions with water were carried out in the following way: 2.5 g of dry material was extracted with 150 mL H 2 O at 100°C in a hot plate (C-MAG HS7, IKA; Staufen, Germany) for 30 min. Extractions were done in triplicate. After that, solutions were filtered in a vacuum pump (Vacuubrand; Wertheim, Germany) through Whatman No. 1 filters. Finally, the solvent was evaporated under reduced pressure, and the dried extracts were stored at -20°C until analysis.

HPLC Analysis of the Phenolic Compounds.
e extracts were analyzed with high-performance liquid chromatography with diode-array and mass spectrometry detection (HPLC-DAD-MS n ), operating in both negative and positive ion modes. Instrumentation and detailed conditions are given in File S2  Individual sock solutions of caffeic acid, neochlorogenic acid, coumaric acid, ferulic acid, hydroxytyrosol, sinapic acid, quercetin, kaempferol, and rutin were prepared in MeOH. We prepared calibration curves for caffeic acid, neochlorogenic acid, coumaric acid, ferulic acid, hydroxytyrosol, sinapic acid, quercetin, and rutin at concentrations 0.5-100 μg mL −1 in MeOH. Chromatograms were recorded at 280 nm for hydroxytyrosol; 320 nm for caffeic acid, neochlorogenic acid, coumaric acid, ferulic acid, and sinapic acid; and 350 nm for quercetin and rutin. Peak area (at the corresponding wavelength) was plotted vs analyte concentration to construct the calibration graphs. Repeatability (n � 9) and intermediate precision (n � 9, 3 consecutive days) were lower than 3 and 8%, respectively. e robustness of the method was assessed by measuring signals at ±2 nm of the optimum wavelength and by modifying the percentage of mobile phases (2% variation with respect to optimum conditions), observing variations lower than 5% in all cases.

Assays for Total Phenolic and Flavonoid Contents.
Total phenolic content (TPC) and total flavonoid content (TFC) were measured by spectrophotometric assays [15]. e obtained results were reported as standard equivalents of gallic acid and rutin for phenolics and flavonoids, respectively. Details for the protocols are provided in File S3 of Supplementary Materials.

Determination of Antioxidant and Enzyme Inhibitory
Effects. Regarding the antioxidant activity of Sarcocapnos extracts, different spectrophotometric assays were performed: ferrous ion chelating, phosphomolybdenum, FRAP, ABTS, CUPRAC, and DPPH [15]. Results are given as standard compounds equivalents of Trolox and EDTA. Details are given in File S4 of Supplementary Materials. e in vitro enzyme inhibitory effects of Sarcocapnos extracts were evaluated on five enzymes: α-amylase, α-glucosidase, acetyl-and butyryl-cholinesterases, and tyrosinase.
e enzyme inhibitory actions were assessed as kojic acid equivalents (KAE) for tyrosinase, galantamine equivalents (GALAE) for acetylcholinesterase (AChE) and butyrylcholinesterase (BChE), and acarbose equivalents (ACAE) for α-amylase and α-glucosidase. Assays were performed as previously reported [15,16], and the details for the protocols are provided in File S4 of Supplementary Materials.

Statistical Analysis.
e different Sarcocapnos extracts were analyzed in triplicate (there were 3 independent extracts for each plant species; Section 2.1), and the values were given as mean ± SD. Firstly, the data were submitted to a descriptive analysis (ANOVA one-way) followed by Tukey's test, using XLSTAT v.2018 (Addinsoft Inc) statistical software. A significance level of 5% was set for all analyses. Unsupervised principal component analysis (PCA) was performed in order to investigate similarities/differences between different samples and to identify the factors responsible for the distinguishing between that samples. Prior to principal component analysis, data were prepreprocessed (autoscaling). Afterwards, Hierarchical Cluster Analysis (HCA) was applied on the results of PCA to bring out the different clusters. To this end, the Euclidean similarity measure and complete linkage were chosen. Finally, Pearson's correlation was carried out to assess the relationship between total bioactive compounds and biological activities. All statistics were done under R v.3.6.1 software.

HPLC-MS Analysis.
Mass spectrometry was used for compounds characterization. A typical chromatogram is shown in Figure S1 of Supplementary Materials. We identified or tentatively characterized 37 compounds, of which more than 30% corresponded to flavonoids and more than 40% were phenolic acids. e characterization of the compounds was performed by using information available in scientific literature and the use of analytical standards. Compounds were numbered according to their order of elution ( ey were also characterized by the neutral losses previously mentioned, as well as the loss of acetylhexoside (204 Da) in compound 38. In all of them, the aglycone isorhamnetin was observed at m/z 315 (typical fragment at m/z 300; comparison with an analytical standard).
Compounds 31 and 33 were putatively characterized as kaempferol-O-rutinoside isomers due to the loss of rutinoside (308 Da) and the presence of the aglycone kaempferol at m/z 285 (the aglycone was identified as kaempferol due to the absence of MS 3 fragment ions at m/z 243 and 241, which would be indicative of luteolin; analytical standards of both kaempferol and luteolin were analyzed).
Compound 8 with [M − H] − ion at m/z 353 presented fragment ions at m/z 191 and 179, which corresponded to neochlorogenic acid (identified by comparison with an analytical standard).
Compounds 9 and 10 were characterized as 3-p-coumaroylquinic acid isomers based on the [M − H] − ions at m/z 337, MS 2 base peak at m/z 163, and comparison of their fragmentation pattern with bibliographic data [17].
Compound 14 was characterized as ferulic acid glucuronide; it exhibited deprotonated molecular ion at m/z 369 and suffered the neutral loss of 176 Da (glucuronide) to yield fragment ions at m/z 193 and 134, typical of ferulic acid. Compound 28 corresponded to ferulic acid (identified by comparison with an analytical standard).
Compound 19, [M − H] − at m/z 295, had major fragments at m/z 179 and 135, which indicated the presence of caffeic acid; it was characterized as caffeic acid cinnamyl ester [18].
Compound 23 presented [M − H + HCOOH] − at m/z 453 and was tentatively identified as coumaric acid-Ohexoside derivative (formate adduct) due to the neutral loss of hexoside (325 ⟶ 163) and the presence of coumaric acid   Journal of Analytical Methods in Chemistry (fragmentation 163 ⟶ 119). Compound 26 was identified as coumaric acid by comparison with an analytical standard.
Compound 29 was tentatively characterized as a coumaric acid derivative. Compound 32 displayed deprotonated molecular ion at m/z 683, with fragments at m/z 223 and m/z 205, typical of the sinapic acid fragmentation, so this compound was tentatively identified as sinapic acid derivative.  is compound was named as hydroxytyrosol-O-hexoside [20].
Compound 18 (m/z 342) was characterized as magnoflorine using the positive ion mode. is alkaloid presented the base peak at m/z 297, and other fragment ions at m/z 282 and 265 [22].

Quantification of Phenolic Compounds.
Twenty-seven compounds were quantified in the analyzed extracts of S. enneaphylla, S. pulcherrima, and S. saetabensis by HPLC-DAD using analytical standards of the corresponding chemical families.
In the three Sarcocapnos species under study, methanolic extracts always presented the highest concentration of phytochemicals, mainly due to the higher solubility of flavonoids in methanol than water. In addition, it implies that methanol is more efficient in cells walls degradation of Sarcocapnos species, probably due to their nonpolar character. Quantification data of all the extracts of Sarcocapnos are given in Tables 3-5.
Total individual phenolic content (TIPC) was defined as the sum of all the compounds quantified individually by HPLC-DAD. TIPC data are shown in Tables 3-5 as well as in Figure S2 of Supplementary Materials. It can be observed that the TIPC values were in the following order: S. enneaphylla > S. pulcherrima > S. saetabensis. is pattern is in agreement with flavonoids concentrations in the three Sarcocapnos species.

Total Phenolic and Flavonoid Contents.
e total phenolic and flavonoid contents of the Sarcocapnos extracts are illustrated in Figure 2 (also in Table S1 of Supplementary Materials). e highest total phenolic content was determined in S. enneaphylla (61.2 mg GAE/g), followed by S. saetabensis (50.6 mg GAE/g) and S. pulcherrima (43.8 mg GAE/g). Regarding the total flavonoid content, the trend was similar to the one for total phenolic content. is finding is consistent with the HPLC-DAD quantification results, which indicated the highest levels of phenolics and flavonoids (particularly rutin) in S. enneaphylla methanol extracts. As can be seen in Figure S3 of Supplementary Materials, the levels of total bioactive compounds exhibited significant differences for each Sarcocapnos species in terms of location. is fact was also mentioned by several authors who reported that the amounts of bioactive compounds could vary due to climatic of geographical conditions [25][26][27]. Besides, methanol has proved to be the best solvent for the extraction of phenolics and flavonoids from Sarcocapnos species. In earlier studies [28,29], the effects of solvents on total bioactive compounds were investigated, and methanol was also found to be the most effective for most of them, in agreement with our findings. Based on our information, there are no previous studies concerning the total amount of bioactive compounds in the genus Sarcocapnos. In this sense, these results will be a scientific starting point on this genus.

Antioxidant Ability.
e antioxidant abilities of the Sarcocapnos species are shown in Table 6 and are expressed in terms of standard equivalents (trolox (TE) and EDTA (EDTAE)). We used antioxidant assays involving different mechanisms (metal chelating, radical quenching, and reduction ability). Similar to the levels of total bioactive compounds, antioxidant activities changed in terms of location in each species. Except for metal chelating ability, the tested methanol extracts exhibited stronger antioxidant abilities when compared with water    extracts. e best radical scavenging ability in DPPH and ABTS assays was obtained for the methanol extract of S. enneaphylla (89 mg TE/g and 136 mg TE/g), followed by S. saetabensis (75.3 mg TE/g and 122 mg TE/g) and S. pulcherrima (66.8 mg TE/g and 105.3 mg TE/g). CUPRAC and FRAP assays include the transformation of Cu 2+ to Cu + and Fe 3+ to Fe 2+ , respectively. ese transformations could be important to prevent Fenton or Haber-Weiss reactions and thus could stop the production of hydroxyl radicals. From this point of view, the tested Sarcocapnos extracts exhibited significant reduction ability. e highest abilities were recorded in S. enneaphylla and S. saetabansis methanol extracts, and they were statistically similar (p > 0.05).
e results for phosphomolybdenum assay, which is based on the transformation of Mo (VI) to Mo (V), were similar to CUPRAC and FRAP, and again the highest ability was noted in S. enneaphylla with the value of 1.97 mmol TE/g. e results for radical scavenging and reducing power assays were in good agreement with the levels of total bioactive compounds. Our observations coincide with the previous studies, which found a good analogy between total bioactive compounds and antioxidant properties [30,31]. Concerning metal chelating ability, the results are inconsistent with the other assays. Metal chelating abilities of the water extracts were found to outperform methanol extracts. is rather contradictory finding may be due to the presence of nonphenolic chelators in the water extracts [32,33].

Enzyme Inhibition Properties.
Several enzymes are targets to manage several diseases, and thus, the symptoms could be decreased by inhibiting these enzymes. Due to the side effects of synthetic compounds, researchers tend to focus on natural sources for the development of novel inhibitors with lower side effects [34][35][36][37]. Hence, we investigated the enzyme inhibitory effects of Sarcocapnos species against cholinesterases (AChE and BChE), tyrosinase, amylase, and glucosidase. e results are shown in Table 7. Except for glucosidase, all enzymes were inhibited by the tested extracts. Regarding AChE inhibition abilities, the values ranged from 2.7 mg GALAE/g to 5.2 mg GALAE/g, while the values for BChE ranged from 10 mg GALAE/g to 22.2 mg GALAE/g. ese inhibition effects may be linked to the chemical profiles of the extracts. For example, the extracts contained some flavonoids such as rutin, isorhamnetin, or kaempferol, which have been previously reported as cholinesterase inhibitors [38][39][40]. Tyrosinase is a main enzyme in the melanogenesis pathway, and its inhibition is an important therapeutic way to control hyperpigmentation problems. As depicted in Table 7, the methanol extracts exhibited stronger tyrosinase inhibitor effects than water extracts. In the same way, we observed the highest inhibitor effects for methanol extracts against amylase and glucosidase. e tested water extracts were not active on glucosidase.
ese observations are consistent with the levels of total bioactive compounds. A similar linear correlation between total bioactive components and these inhibitions was reported by other authors [41][42][43][44]. In addition, some flavonoids (rutin, isorhamnetin, etc.) have been reported as significant inhibitors against these enzymes [45][46][47][48]. Taken together, the tested Sarcocapnos species can be appraised as sources of natural enzyme inhibitors to combat global health problems.

Multivariate
Analysis. An unsupervised exploratory multivariate analysis PCA was performed on biological activities datasets of Sarcocapnos species extracts derived from two solvents with different geographical origins. e main focus of doing PCA was to comprehensively screen the differentially expressed biological activities between Sarcocapnos species samples. In fact, two independent factors were involved as follows: geographical origins and solvent used for extraction. us, through PCA, we sought to determine which factor was responsible to distinguish between the biological activities of the samples, previously uncovered by the univariate analysis.
e contribution of biological activities on both retained principal component is shown in Figure 3(b); as can be seen, eight biological activities represented the most associated variables with PC1 while only AChE was most closely linked with PC2. By observing the score plots reported in Figures 3(c) and 3(d), it could be noticed that the extracts can be classified in terms of different factors. Concerning Figure 3(c), extracts were separated into two groups along PC1, based on the solvent used for extraction. erefore, the biological activities of studied species were affected by the varied polarity of both solvents used. Several reports highlighted the influence of the type of solvent on the biological activities of a large number of species [31,49,50].
is disparity reflects the availability of different bioactive compounds, involved in these biological activities, in each extraction solvent. As a result, selecting the appropriate solvent leads to a high recovery yield of molecules of biological interest.
Afterwards, by analyzing the score plots under another angle, the difference between the extracts of Sarcocapnos species can be also attributed to a geographical factor (Figure 3(d)). Nonetheless, this geographical factor took effect within methanolic extracts of species. Two subclusters were observed; indeed, samples obtained from specimens of Sarcocapnos species from Segura de la Sierra, Vilches, and Fuensanta de Martos, respectively, were distinguished from those harvested in Alcaudete, Jódar, and between Jaén and Otíñar. On the other hand, as regards to the extracts derived from water, a difference was observed only between the samples of S. enneaphylla. In more detail, specimen from Vilches was detached from that of Alcaudete along PC1.
Heatmap and HCA were produced on the result of PCA taking into account only the first two principal components of PCA, for visualizing differential bioactivities in the   different samples (Figure 4). Two major clusters and their respective two subclusters were obtained (Figure 4(a)). As can be seen, methanol extracts of all studied Sarcocapnos species were found to be more active against the vast majority of the evaluated biological activities, due to the higher solubility of the bioactive compounds in methanol. e highest values of total phenolic and total flavonoid contents (spectrophotometric analyses), as well as total individual contents (HPLC-MS), were obtained with extracts derived from methanol for all species. Hence, TPC, TFC, and total individual flavonoids content correlated very well with all biological activities except AChE and MCA (Figure 3(e)    positively linked with TFC, TPC, and total individual flavonoid content, respectively, while MCA assay was positively bound to total individual phenolic acid content (R � 0.93) at p < 0.05. Additionally, the inhibition of tyrosinase, amylase, glucosidase, and BChE was highest with the increasing content of TFC (R � 0.95; 0.94; 094; and 0.79), TPC (R � 0.9; 0.9; 0.89; 0.93) and total individual flavonoids content (R � 0.74; 0.73; 0.8; 0.64) respectively. According to the literature, phenolic compounds (including the flavonoids), ubiquitous in plants, are of great interest due to their considerable beneficial for human health, since they prove to be able to cut down the incidence of several chronic ailments such as diabetes and Alzheimer and skin diseases through their antioxidant properties. Instead, their antioxidant activities are strongly linked with their structure, especially the position and number of the hydroxyl groups, the conjugated double bonds, and the nature of substitutions on the aromatic rings. Afterwards, by analyzing the score plots under another angle, the difference between the extracts of Sarcocapnos species can be also attributed to a geographical factor (Figure 3(d)). Nonetheless, this geographical factor took effect within methanolic extracts of species. Two subclusters were observed; indeed, samples obtained from specimens of Sarcocapnos species from Segura de la Sierra, Vilches, and Fuensanta de Martos, respectively, were distinguished from those harvested in Alcaudete, Jódar, and between Jaén and Otíñar. On the other hand, as regards the extracts derived from water, a difference was observed only between the samples of S. enneaphylla. In more detail, specimen from Vilches was detached from that of Alcaudete along PC1.
Regarding Figure 4(b), the difference across subclusters III and IV as well as between S. enneaphylla specimens was found tangibly with AChE inhibition assay. As it happens, the highest activity was obtained within cluster III and S. enneaphylla specimen from Vilches (belonging the cluster I), respectively.
With regard to our results, the biological activities of studied species varied among the geographical origins. is observation is closely linked to the presence/absence of some key bioactive compounds under the influence of the variation of climatic conditions as well as the soil nutrient composition. is is in agreement with the report of Rocchetti [51], who underlined the impact of different geographical origins on phenolic composition of chardonnay wines.

Conclusion
In this work, we have presented the first report on the phenolic composition of three Sarcocapnos species as well as the relation with the bioactive properties. For the three species, methanol was the most suitable solvent, and the highest phenolic content as well as the highest bioactivities were found in extracts from S. enneaphylla, although the other species were also active in all the assays. Flavonoids were the most abundant compounds, followed by phenolic acids. Among flavonoids, rutin, kaempferol, and isorhamnetin glycosides were the main compounds. ese compounds have been previously reported to present significant bioactivity and, thus, may be the main contributors to the observed effects. Multivariate analysis confirmed the relationship between phenolics concentration and bioactivity. It also indicated a clear difference within the same species depending on the geographical origin. e results here presented may open new ways to valorize Sarcocapnos species as sources of natural compounds for the food (nutraceuticals or food supplements) or pharmaceutical industry and represent a starting point for additional research, being the final goal the industrial applications of Sarcocapnos species.

Data Availability
e data used to support the findings of this study are included within the article.

Conflicts of Interest
e authors declare that they have no conflicts to interest.