Bioactive Potential: A Pharmacognostic Definition through the Screening of Four Hypericum Species from the Canary Islands

In this work, we propose a general methodology to assess the bioactive potential (BP) of extracts in the quest of vegetable-based drugs. To exemplify the method, we studied the anticancer potential (AP) of four endemic species of genus Hypericum (Hypericum canariense L, Hypericum glandulosum Aiton, Hypericum grandifolium Choisy and Hypericum reflexum L.f) from the Canary Islands. Microextracts were obtained from the aerial parts of these species and were tested against six human tumor cell lines, A549 (non-small-cell lung), HBL-100 (breast), HeLa (cervix), SW1573 (non-small-cell lung), T-47D (breast) and WiDr (colon). The methanol–water microextracts were evaluated further for cell migration, autophagy and cell death. The most promising bioactive polar microextracts were analyzed by UHPLC–DAD–MS. The extraction yield, the bioactivity evaluation and the chemical profiling by LC–MS suggested that H. grandifolium was the species with the highest AP. Label-free live-cell imaging studies on HeLa cells exposed to the methanol–water microextract of H. grandifolium enabled observing cell death and several apoptotic hallmarks. Overall, this study allows us to select Hypericum grandifolium Choisy as a source of new chemical entities with a potential interest for cancer treatment.


Introduction
The chemical bioprospecting of new therapeutic agents, pesticides, nutraceuticals, and other products with biotechnological applications from plants is one of research lines with the greatest impact worldwide [1][2][3]. The chemical bioprospecting protocol includes (a) the selection and extraction of appropriate plants, (b) the evaluation of the bioactive potential of the extracts to select the species with the highest bioactive potential, and (c) the bioguided isolation of compounds from the best candidates [4]. Thus, the rationalization of each stage of the chemical bioprospecting contributes to accelerate the discovery of compounds with high added value [5,6].
Bioactive potential (BP) is a very often employed term in the search of bioactive compounds from living organisms (bacteria, fungi or plants) but its definition is not clear yet in the literature. BP can be subdivided into agrochemical or pharmacological potential depending on the field of the research, i.e., agrochemistry or pharmacology, respectively [3,7]. In drug discovery, the bioactive potential of extract from plants as a source of high-value compounds has mainly focused the bioactivity evaluation and

Results and Discussion
Our evaluation protocol of the AP of plants consists of three sequential steps. Every step provides relevant information to help in the assessment of the AP. In the next Sections, we will present our implementation of the methodology and we will discuss the results and the criteria we followed to prioritize the Hypericum species from the Canary Islands as a source of bioactive compounds.

Selection of Plants and Microextraction
H. canariense, H. glandulosum, H. grandifolium and H. reflexum were selected considering chemosystematic and ethnopharmacological approaches [7]. The extraction process of these species started with the maceration of their aerial part at a small scale (microextraction). For this purpose, to perform the extraction in a sequential maceration, we selected as solvents water, methanol and dichloromethane. The procedure we followed is an adapted methodology previously described by Macías in the search of phytotoxic compounds [43]. This methodology was designed regarding the vast range of specialized metabolites in the Hypericum genus (range of polarities and bioactivities) [22,44] and as a way to solve the difficulty to obtain them using a single extraction [45]. Therefore, each plant dried material (approximately 10 g) was sequentially macerated under the same conditions from polar to non polar solvents considering a sample ratio of 20:1 (v/w). Firstly, a methanol/water (1:1) mixture (MW), then methanol (MM), and finally a dichloromethane/methanol (1:1) mixture (DM) were used, as depicted in Figure 1.  [7]. The extraction process of these species started with the maceration of their aerial part at a small scale (microextraction). For this purpose, to perform the extraction in a sequential maceration, we selected as solvents water, methanol and dichloromethane. The procedure we followed is an adapted methodology previously described by Macías in the search of phytotoxic compounds [43]. This methodology was designed regarding the vast range of specialized metabolites in the Hypericum genus (range of polarities and bioactivities) [22,44] and as a way to solve the difficulty to obtain them using a single extraction [45]. Therefore, each plant dried material (approximately 10 g) was sequentially macerated under the same conditions from polar to non polar solvents considering a sample ratio of 20:1 (v/w). Firstly, a methanol/water (1:1) mixture (MW), then methanol (MM), and finally a dichloromethane/methanol (1:1) mixture (DM) were used, as depicted in Figure 1. After the extraction process, three microextracts from each plant were obtained (MW, MM and DM). Thus, a total of 12 microextracts were prepared. For each microextract, the dry mass was weighed and the extraction yield, expressed in percentage, was calculated. The results are shown in Table 1. After the extraction process, three microextracts from each plant were obtained (MW, MM and DM). Thus, a total of 12 microextracts were prepared. For each microextract, the dry mass was weighed and the extraction yield, expressed in percentage, was calculated. The results are shown in Table 1. The extraction yield is a pharmacognostic parameter used for considering any vegetal as a source of leads or drugs [14]. It also enables the calculation of the total activity (TA) (Equation (1), ), a useful tool for the comparison of crude extracts during screening of plants as a source of bioactive compounds [11]. In our study, the MW and MM microextracts from the aerial parts of the four Hypericum species gave the highest yields. In contrast, the DM microextracts provided the lowest yields, except for H. glandulosum. In general, these data suggested that Hypericum species from the Canary Islands were rich in medium and polar compounds. These results also match with a recently published study on Hypericum species from the UK [14]. These 12 microextracts were submitted to the biological evaluation at two levels.

Biological Evaluation
The scientific literature holds several examples of non-mammalian and mammalian cell-based assays that have been used for the identification and characterization of bioactive natural products [46]. In the therapeutic area of cancer, screenings based on mammalian cell cultures have enabled the rapid study of a multitude of natural product extracts. In drug discovery programs of plant extracts (such as the NCI's), the first biological assessment is an antiproliferative test against tumor cell lines (a general or primary bioassay) [47,48].

Antiproliferative Activity
Following the guidelines of the NCI, the antiproliferative activity of the 12 microextracts (MW, MM and DM) obtained from the aerial parts of H. canariense (CMW, CMM and CDM), H. glandulosum (GMW, GMM and GDM), H. grandifolium (LMW, LMM and LDM), and H. reflexum (RMW, RMM and RDM) were evaluated against six human tumor cell lines: A549, HBL-100, HeLa, SW1573, T-47D and WiDr. The maximum test concentration was set at 250 µg/mL, according to NCI guidelines [49,50]. The results, expressed as 50% growth inhibition (GI 50 ), are given in Figure 2 (Table S1). According to NCI calculations, GI 50 values represent approximate values. Taken as a whole, the GI 50 values enabled an arrangement of the plants in the following decreasing order of activity: H. grandifolium > H. canariense > H. glandulosum > H. reflexum. Our results are in agreement with a previous study where the antiproliferative activity of essential oils and polar extracts (methanol-acetone 1:1) from the aerial parts of H. canariense, H. grandifolium and H. reflexum were evaluated against the human tumor cell lines A375 (malignant melanoma), MDA-MB-231 (breast adenocarcinoma) and HCT116 (colon carcinoma) [42]. In that study, the polar extract from H. grandifolium also exhibited a significant antiproliferative activity and resulted the most active one. Noteworthy, our work represents the first study reporting antiproliferative activity of H. glandulosum extracts.
When we are considering the solvent extraction, the sequence obtained (in decreasing activity) for our samples was DM > MM > MW. In the particular case of H. grandifolium, the DM and MM microextracts (i.e., GDM and GMM, respectively) were the most active ones against all the tumor cell lines tested. They displayed the lowest GI 50 values (in the range 2.7-13 µg/mL) among the twelve assayed microextracts ( Figure 2). However, these GI 50 values denote a highly cytotoxic profile (GI 50 ≤ 20 µg/mL) according to the NCI's guidelines [50]. Noteworthy, the MW microextracts induced moderate cytotoxicity (GI 50 Molecules 2022, 27, 6101 5 of 18 21-200 µg/mL). In our experience, highly cytotoxic samples induce large cell killing and biological effects are more difficult to be observed. In contrast, samples with moderate cytotoxicity allow to measure relevant phenotypic changes in treated cells. According to this criterion and our interest in the pharmacognostic study of polar extracts, the hydroalcoholic microextracts were the prioritized for further testing. When we are considering the solvent extraction, the sequence obtained (in decr activity) for our samples was DM > MM > MW. In the particular case of H. grandi the DM and MM microextracts (i.e., GDM and GMM, respectively) were the most ones against all the tumor cell lines tested. They displayed the lowest GI50 values range 2.7-13 µg/mL) among the twelve assayed microextracts ( Figure 2). However GI50 values denote a highly cytotoxic profile (GI50 ≤ 20 µg/mL) according to the guidelines [50]. Noteworthy, the MW microextracts induced moderate cytotoxicit 21-200 µg/mL). In our experience, highly cytotoxic samples induce large cell killin biological effects are more difficult to be observed. In contrast, samples with mo cytotoxicity allow to measure relevant phenotypic changes in treated cells. Accord this criterion and our interest in the pharmacognostic study of polar extrac hydroalcoholic microextracts were the prioritized for further testing.

Total Activity
The TA (Equation (1)) is a significant pharmacognostic parameter f quantification of the antimicrobial activity of plant extracts and it is very useful rational screening of diverse plants. This tool has been used as a part of the evalua the antibacterial activity of extracts from south African species [12]. TA, expres (mL/g), was defined as "the largest volume to which the bioactive compounds in one g plant material can be diluted and still inhibit the growth of the tested organisms Analogously, we have envisioned the calculation of the TA when consideri antiproliferative activity of plant extracts against human solid tumor cells. Fro pharmacological point of view, the growth inhibition (GI50 in µg/mL) indicat potency of extracts, while TA represents the efficacy of the extracts [11].

= 1 /
In this work, we calculated the TA of the Hypericum species to compa antiproliferative efficacy of their microextracts. To the best of our knowledge, this first time that the TA is reported to assess the antiproliferative activity duri screening of the of plant species.
Since our protocol involves the sequential extraction using three solvent s (Figure 1), we considered appropriate to compute the TA for each individual extra sum all the TAs corresponding to a virtual total extract from each plant and each ce With this calculation, each TA value represents a theoretical activity of each theoric

Total Activity
The TA (Equation (1)) is a significant pharmacognostic parameter for the quantification of the antimicrobial activity of plant extracts and it is very useful in the rational screening of diverse plants. This tool has been used as a part of the evaluation of the antibacterial activity of extracts from south African species [12]. TA, expressed in (mL/g), was defined as "the largest volume to which the bioactive compounds in one gram of plant material can be diluted and still inhibit the growth of the tested organisms" [11]. Analogously, we have envisioned the calculation of the TA when considering the antiproliferative activity of plant extracts against human solid tumor cells. From the pharmacological point of view, the growth inhibition (GI 50 in µg/mL) indicates the potency of extracts, while TA represents the efficacy of the extracts [11].
In this work, we calculated the TA of the Hypericum species to compare the antiproliferative efficacy of their microextracts. To the best of our knowledge, this is the first time that the TA is reported to assess the antiproliferative activity during the screening of the of plant species.
Since our protocol involves the sequential extraction using three solvent systems (Figure 1), we considered appropriate to compute the TA for each individual extract and sum all the TAs corresponding to a virtual total extract from each plant and each cell line. With this calculation, each TA value represents a theoretical activity of each theorical total extract against each cell line that complies with the definition provided by Eloff [11]. Therefore, Figure 3 (Table S2)  Therefore, Figure 3 (  Based on the GI50 data ( Figure 2), the TA values ( Figure 3 and Table S2) and interest in polar extracts, we selected for further studies the MW microextracts fro canariense (CMW), H. glandulosum (LMW) and H. grandifolium (GMW). The si microextract of H. reflexum (i.e., RMW) were also tested for comparison purposes. A point, many different phenotypic experiments can be considered to depict the ov mode of action of the microextracts and give information about possible cell targets. T experiments are generally termed as secondary bioassays. The literature reports di secondary bioassays for the potential anticancer activity of Hypericum sp. extracts [19 Herein, we selected as secondary tests cell migration and vacuole formation, two rele hallmarks for cancer treatment, which have not been reported for the (four) Hypericu (included in this work). Cell migration experiments give an idea of the antimeta potential of the tested samples [51], while vacuole formation is an indicator of autop [52] and drug resistance mechanisms [53], among others.

Cell Migration Disturbances in A549 Cells
Cell migration is the ability of a cancer cell to undergo movement and inv allowing it to relocate within the tissues [51]. This process allows neoplastic cells to lymphatic and blood vessels for dissemination into the circulation metastasize in di organs. The non-small-cell lung cancer (NSCLC) cell line A549 is considered as a h metastatic cancer cell line [54] and represents a good in vitro model to asses antimigratory properties using the wound healing assay [55].
To study the effects on A549 cell migration of the MW microextracts of the Hypericum species, we selected two doses, a high (100 µg/mL) and a low dose (50 µg/ and two time points, 6 and 24 h ( Figure 4). The use of this early time point en avoiding, at minimum, the effect of cell proliferation in the assay, limiting the respon the antimigratory activity of the microextract. With the exception of CMW Based on the GI 50 data (Figure 2), the TA values ( Figure 3 and Table S2) and our interest in polar extracts, we selected for further studies the MW microextracts from H. canariense (CMW), H. glandulosum (LMW) and H. grandifolium (GMW). The sibling microextract of H. reflexum (i.e., RMW) were also tested for comparison purposes. At this point, many different phenotypic experiments can be considered to depict the overall mode of action of the microextracts and give information about possible cell targets. These experiments are generally termed as secondary bioassays. The literature reports diverse secondary bioassays for the potential anticancer activity of Hypericum sp. extracts [19][20][21]. Herein, we selected as secondary tests cell migration and vacuole formation, two relevant hallmarks for cancer treatment, which have not been reported for the (four) Hypericum sp. (included in this work). Cell migration experiments give an idea of the antimetastatic potential of the tested samples [51], while vacuole formation is an indicator of autophagy [52] and drug resistance mechanisms [53], among others.

Cell Migration Disturbances in A549 Cells
Cell migration is the ability of a cancer cell to undergo movement and invasion allowing it to relocate within the tissues [51]. This process allows neoplastic cells to enter lymphatic and blood vessels for dissemination into the circulation metastasize in distant organs. The non-small-cell lung cancer (NSCLC) cell line A549 is considered as a highly metastatic cancer cell line [54] and represents a good in vitro model to assess the antimigratory properties using the wound healing assay [55].
To study the effects on A549 cell migration of the MW microextracts of the four Hypericum species, we selected two doses, a high (100 µg/mL) and a low dose (50 µg/mL), and two time points, 6 and 24 h (Figure 4). The use of this early time point enables avoiding, at minimum, the effect of cell proliferation in the assay, limiting the response to the antimigratory activity of the microextract. With the exception of CMW all microextracts, were able to reduce the closure of the wound when compared to control wells after 6 and 24 h. GMW was able to inhibit cell migration after six hours of treatment in a concentration dependent manner, being the most active of all microextracts assayed. Interestingly, CMW, which showed antiproliferative activity, failed to reduce the migration of A549 cells. Meanwhile, both LMW and RMW did not show correlation between dose and inhibition of cell mobility. microextracts, were able to reduce the closure of the wound when compared to cont wells after 6 and 24 h. GMW was able to inhibit cell migration after six hours of treatme in a concentration dependent manner, being the most active of all microextracts assaye Interestingly, CMW, which showed antiproliferative activity, failed to reduce t migration of A549 cells. Meanwhile, both LMW and RMW did not show correlati between dose and inhibition of cell mobility.  . Wound closure assay results on A549 cells when using 100 µg/mL (a) and 50 µg/mL (b MW microextracts. Wound closure percentage with respect to time = 0 is given on the right t corner of each image.

Formation of Acidic Vacuoles in HeLa Cells
Considering the antiproliferative profile of the Hypericum extracts, CMW, GM LMW and RMW were selected to perform acridine orange (AO) staining in order evaluate the formation of acidic vacuoles in the cells after the treatment. AO is fluorescent dye that binds to acidic compartments. Differences in fluorescence intens could represent changes in cell response to the plant microextract, of particular interest the formation of acidic vacuoles [56]. As a positive control we used 10 µM of tamoxif (TAM), an established anticancer drug whose ability to enhance cell vacuolization h been reported previously [57]. As depicted in Figure 5a, GMW showed the highest fo change in RFU over control, similar to the effect produced by TAM, whilst the rest of t microextracts did not produce a representative fold change when treated with 100 µg/m

Formation of Acidic Vacuoles in HeLa Cells
Considering the antiproliferative profile of the Hypericum extracts, CMW, GMW, LMW and RMW were selected to perform acridine orange (AO) staining in order to evaluate the formation of acidic vacuoles in the cells after the treatment. AO is a fluorescent dye that binds to acidic compartments. Differences in fluorescence intensity could represent changes in cell response to the plant microextract, of particular interest to the formation of acidic vacuoles [56]. As a positive control we used 10 µM of tamoxifen (TAM), an established anticancer drug whose ability to enhance cell vacuolization has been reported previously [57]. As depicted in Figure 5a, GMW showed the highest fold change in RFU over control, similar to the effect produced by TAM, whilst the rest of the microextracts did not produce a representative fold change when treated with 100 µg/mL for 24 h. Taking this into account and considering the previous result, we proceeded to evaluate GMW microextract effects on HeLa cells under fluorescence microscopy. Images taken agreed with the data obtained by spectrofluorimetric assays, with this microextract enhancing the accumulation of the dye inside clearly visible vacuoles (Figure 5b).The accumulation of AO inside acidic vacuoles produces a metachromatic shift from green to red fluorescence [58]. for 24 h. Taking this into account and considering the previous result, we proceeded to evaluate GMW microextract effects on HeLa cells under fluorescence microscopy. Images taken agreed with the data obtained by spectrofluorimetric assays, with this microextract enhancing the accumulation of the dye inside clearly visible vacuoles (Figure 5b).The accumulation of AO inside acidic vacuoles produces a metachromatic shift from green to red fluorescence [58]. Overall, the results might anticipate a different chemical profile of these microextracts.

Early Chemical Profiling of MW Microextracts
The final step for the evaluation of the anticancer potential was to analyze the chemical composition of the selected bioactive MW microextracts from the aerial parts of H. canariense, H. glandulosum and H. grandifolium (i.e., CMW, LMW and GMW, respectively). In an effort to estimate their chemical qualitative composition in terms of specialized metabolites, they were analyzed by UHPLC-DAD-MS (Table 2). Methanol and acetone-methanol extracts from H. canariense and H. grandifolium were analyzed by LC-MS and reported previously [33,42]. This work represents the first preliminary chemical study of H. glandulosum extract. Overall, the results might anticipate a different chemical profile of these microextracts.

Early Chemical Profiling of MW Microextracts
The final step for the evaluation of the anticancer potential was to analyze the chemical composition of the selected bioactive MW microextracts from the aerial parts of H. canariense, H. glandulosum and H. grandifolium (i.e., CMW, LMW and GMW, respectively). In an effort to estimate their chemical qualitative composition in terms of specialized metabolites, they were analyzed by UHPLC-DAD-MS (Table 2). Methanol and acetonemethanol extracts from H. canariense and H. grandifolium were analyzed by LC-MS and reported previously [33,42]. This work represents the first preliminary chemical study of H. glandulosum extract. Table 2 shows the chemical profiles of the three Hypericum MW microextracts at 280 nm. These chromatograms also show the UHPLC fingerprint of these species and the substantial differences in their chemical composition, both in terms of the number of peaks and in the proportion of common peaks ( Figure S1). We speculate that this differential profile is responsible for the differences observed in the biological activity of the microextracts (Figures 2, 4 and 5). According to these preliminary results, a total of forty one main compounds in terms of specialized metabolites were detected in the methanol-water microextracts CMW, GMW and LMW, respectively. Twenty four were known compounds tentatively identified based on their retention times, spectral properties (UV-Vis and MS/MS patterns) and from the comparison of literature. These known compounds belonged to the family of phenolic secondary metabolites and they have previously been described or isolated from Hypericum extracts [19,42,60,[62][63][64]. Among these known compounds 6, 9, 14, 15, 19, 20, 28 and 33 were ubiquitous in all MW microextracts. A careful analysis of the UV trace at 280 nm of CMW, GMW and LMW microextracts enabled the detection of at least the presence of eighteen, nineteen and twenty seven specialized metabolites, respectively. Additionally, these MW microextracts also contained a total of fifteen unidentified compounds whose spectral data suggested their novelty as natural products from Hypericum species since they were not found in the literature.
In spite of this preliminary analysis by UHPLC-DAD-MS of MW microextracts from H. canariense, H. glandulosum and H. grandifolium, several minor detected compounds cannot be identified due to many of them being co-eluted with major compounds. In this framework, further bioguided isolation studies might be required to clarify their identity.

Bioactive and Anticancer Potential
In this work, we discussed the microextraction with three different solvents, the evaluation of bioactivity of the obtained microextracts, and the chemical analysis of the most polar microextracts from the aerial parts of four Hypericum species from Canary Islands. These experiments might address the selection of the species with the highest bioactive potential. Therefore, we propose an empirical equation for the calculation of the bioactive potential (BP, Equation (2)) based on the previous works by Eloff and Cos [11,15], our experience on the screening of plants [65,66] and the results described in previous sections. Basically, BP and TA have similar definitions, but BP refers to the efficacy of the plant material in several related bioassays and the possibility to find new compounds as the responsible of the initial observed biological activity.
(BP) = (N + 1) × Total Activity × Chemical novelty (2) Chemical novelty = total number of detected compounds number of known compounds (3) (BP) = (N + 1) × Yield extraction × 1 Bioactivity × Chemical novelty (4) In summary, the calculation of BP involves three experimental components (Equation (4)): the extraction process (yield extraction), the bioactivity evaluation (bioactivity in primary bioassays), and the chemical analysis by LC-MS (chemical novelty, Equation (3)); and a fourth factor defined as N + 1, where N means how many times a sample becomes the most active one in a bioassay. This fourth factor (N + 1) implies that BP never could be zero since an extract might be bioactive in any sort of biological test (intrinsic bioactivity). Therefore, after applying Equation (4), the resulting value of BP enables the sorting and prioritizing a list of plants from the highest to the lowest BP during a screening. agrochemical or pharmacological studies) related to the search of high-value products from plants.
To exemplify the calculation of BP, considering our special interest in the anticancer activity in the polar extracts from of the aerial parts of Hypericum species, we applied and modified BP formulae to calculate the anticancer potential (AP) of methanol-water microextracts (MW) from H. canariense (CMW), H. grandifolium (GMW) and H. glandulosum (LMW) (Equation (5)). From the biological tests and regarding GI50 values, the MW microextract from H. reflexum (RMW) was not subjected to the chemical analysis by LC-MS, and consequently it was considered the sample with the less AP.
Anticancer Potential (AP) = (N + 1) × Total Activity × C total C known (5) The anticancer potential values from the MW were obtained based on one primary and two secondary bioassays related to anticancer activity ( Figure 6). From the first biological test we conclude that none of the MW microextracts were the most active one. Finally, GMW was selected as the most active one in two of the secondary bioassays: cell migration disturbances in A549 cells and formation of acidic vacuoles in HeLa cells. Table 3 shows the N and chemical novelty values considered for the calculation of the AP of MW microextracts from H. grandifolium, H. glandulosum and H. canariense. The TA for these MW microextracts were computed in Section 2.2.2 (Table S2). Overall, the AP values permit to sort the microextracts: GMW > LMW> CMW (H. grandifolium > H. glandulosum > H. canariense). These results also allow select H. grandifolium Choisy as the species with the highest AP as a source of polar potentially bioactive compounds.

MW Microextracts of H. Grandifolium Induce Cell Death in HeLa Cells
Based on this, we selected the MW microextract of H. grandifolium (GMW) to run a more detailed study on HeLa cells using label-free continuous live-cell imaging. Live-cell imaging opens a new window in the study of cell response to potential bioactive compounds. The ability to evaluate the behavior of a population at single cell level facilitates the study of phenotypic changes between subgroups of individuals. Moreover, live imaging enables checking the effects at continuous time points, changing the concept of dose-response relationship at fixed times.
HeLa cells were exposed to GMW and the effects were monitored for 15 h. For comparison purposes, we used TAM. Figure 7 depicts how GMW is able to induce death to HeLa cells earlier and in extended populations than TAM (Figure 7a). In consonance with AO staining results ( Figure 5), cells are highly vacuolated (Figure 7b) in the GMWtreated group. This feature could be induced not only by the pharmacological activity of the extract but as a cell response to an external agent, trying to isolate the exogenous substances. Since live imaging has enabled us to observe cell death progressively, we can differentiate several apoptotic hallmarks. Treatment with GMW induced nuclear condensation and cell shrinkage after 10 h of incubation in a broader way than TAM. For

MW Microextracts of H. Grandifolium Induce Cell Death in HeLa Cells
Based on this, we selected the MW microextract of H. grandifolium (GMW) to run a more detailed study on HeLa cells using label-free continuous live-cell imaging. Live-cell imaging opens a new window in the study of cell response to potential bioactive compounds. The ability to evaluate the behavior of a population at single cell level facilitates the study of phenotypic changes between subgroups of individuals. Moreover, live imaging enables checking the effects at continuous time points, changing the concept of dose-response relationship at fixed times.
HeLa cells were exposed to GMW and the effects were monitored for 15 h. For comparison purposes, we used TAM. Figure 7 depicts how GMW is able to induce death to HeLa cells earlier and in extended populations than TAM (Figure 7a). In consonance with AO staining results ( Figure 5), cells are highly vacuolated (Figure 7b) in the GMW-treated group. This feature could be induced not only by the pharmacological activity of the extract but as a cell response to an external agent, trying to isolate the exogenous substances. Since live imaging has enabled us to observe cell death progressively, we can differentiate several apoptotic hallmarks. Treatment with GMW induced nuclear condensation and cell shrinkage after 10 h of incubation in a broader way than TAM. For the same time period, cell death was not observed in the control group (Figure 7a, Supplementary material V1). The observed effects were analyzed and quantified with STEVE software (Figures 7c and S2-S4). Cell death is represented as a drastic decrease in the percentage of confluence in the GMWtreated samples. The formation of apoptotic bodies by cell shrinkage is depicted in the diminished cell area. In agreement with this last result, dry mass density is increased in the GMW-treated cells when compared to the control, reflecting the collapsing of cells caused by the extract.

General Experimental Procedures
Acetonitrile, methanol, and formic acid for UHPLC were purchased from Merck (Darmstadt, Germany). Water for UHPLC was purified with a Millipore Simplicity System (Bedford, MA, USA). Solvents for the extraction of plant material were of analytical grade.

Plant Material
The aerial parts of H. canariense, H. grandifolium, and H. reflexum (6 to 8 kg of fresh material) were collected in La Palma and Tenerife (Canary Islands, Spain), during the period between June and July 2019. 7 kg of the aerial parts of H. glandulosum, were collected in La Palma and Tenerife (Canary Islands, Spain), in June 2021. All the specimens were collected and identified by Dr. Pedro Luis Pérez de Paz, University of La Laguna. A voucher specimen of species was deposited at TFC Herbarium, University of La Laguna (Tenerife, Spain) ( Table 4). Plant material was dried in airy oven at 38 • C temperature for one week. Finally, plant material was stored at room temperature in the darkness.

Extraction of Plant Material
Ground aerial parts (10 g) from H. canariense, H. glandulosum, H. grandifolium, and H. reflexum, were sequentially extracted with the same volume (200 mL) of methanol-water (1:1, v/v, MW), methanol (MM), and dichloromethane-methanol (1:1, v/v, DM) for 24 h at room temperature. The resulting extracts were filtered and concentrated under reduced pressure to evaporate the less volatile solvent. All extracts were resuspended in water and lyophilized affording MW, MM and DM extracts, respectively. Thereafter, the microextracts were weighed and stored at 4 • C until further processing.

UHPLC-DAD-MS3 Analysis
UHPLC-DAD-MS analyses of raw and aqueous extracts were conducted using a previously established method [67] with some modifications. Thus, 10 mg of polar microextracts (CMW, GMW and LMW) were dissolved in 1 mL of methanol-water (1:1, v/v) and subjected to UHPLC analyses. The separation was carried out on a Kinetex XB-C 18 (150 mm × 2.1 mm × 1.7 µm, Phenomenex, Torrance, CA, USA). The mobile phases were 0.1% formic acid in water (A), and 0.1% formic acid in acetonitrile (B), and elution was conducted with the following gradient: 0 min-3% B, 60 min-26% B, 120 min-95% B. The flow rates were 0.3 mL/min. The UV-Vis spectra of the detected compounds were recorded over the 190-600 nm range. The chromatograms of these microextracts) were recorded at 280 nm. Mass spectra were recorded in the positive and negative ion modes. Compounds were characterized based on the maxima observed in their UV-Vis spectra and on their MS spectra. Additionally, the Reaxys database was searched for compounds previously detected and identified in Hypericum species.

Cell, Culture and Plating
The human solid tumor cell lines A549 (lung), HBL-100 (breast), HeLa (cervix), SW1573 (non-small-cell lung), T-47D (breast), and WiDr (colon) were used in this study. These cell lines were a kind gift from Prof. Godefridus J. Peters (VU Medical Center, Amsterdam, The Netherlands). Cells were maintained in 25 cm 2 culture flasks in RPMI 1640 supplemented with 5% FBS and 2 mM L-glutamine in a 37 • C, 5% CO 2 , 95% humidified air incubator. For all tests, exponentially growing cells were trypsinogen and resuspended in an antibioticcontaining medium (100 units penicillin G and 0.1 mg of streptomycin per mL). Single-cell suspensions were counted with Moxi Z. After counting, dilutions were made to give the appropriate cell densities for inoculation onto the appropriate (microtiter) plates.

Antiproliferative Tests
The tests were performed in 96-well plates using the SRB assay [68] with the following specifications. Cell seeding density was 2500 cells/well for A549, HBL-100, HeLa, and SW1573, and 5000 cells/well for T-47D and WiDr. Stock samples were dissolved in DMSO at an initial concentration of 100 mg/mL. Test samples were prepared by decimal serial dilutions of the stock sample to give a final test concentration of 250, 25 and 2.5 µg/mL. The incubation time was 48 h. The optical density of each well was measured at 530 (primary) and 620 nm (secondary) using a microplate reader (Power Wave XS, BioTek, Winooski, VT, USA). The antiproliferative activity, expressed as 50% growth inhibition (GI 50 ) values, was calculated according to the NCI formulas [49]. GI 50 represent approximate values. Following NCI guidelines, the concentrations giving PG values above and below PG = 50 were used to make interpolations on the concentration axis.

Cell Migration Assay
To study cell migration, we used the wound healing (scratch) assay [70]. Single-cell suspensions of A549 cells were seeded onto a 24 well plate at a density of 50,000 cells/well. Cells were incubated until they reached >90% confluence. Afterwards, a mark was drawn on the outside bottom of each well. This enabled finding the reference point when taking pictures. For each well, a scratch on the cell culture was made perpendicularly to the mark using a sterile p200 tip. Then, the medium was replaced with fresh medium with 2.5% FBS. This allows us to diminished the interference caused by cell proliferation during the time of assay, evaluating only the cell migration. Pictures were taken with a brightfield microscope (Axiovert 40 CFL, Zeiss, Germany) at one magnification (5X) using the software ZEN 2012 (blue edition v1.1.0.0) (accessed on 10 August 2022) at different time intervals (0, 6 and 24 h since the formation of the scratch). For the quantification of cell migration, TScratch, an image software based in MATLAB, was used adapting threshold of images to measure the area of the wound made [71]. Results are represented as percentage of the closed area as defined by Equation (6).

Acridine Orange Staining Assay
To evaluate the induction of acidic vacuoles by the plant extracts the acridine orange staining assay was used. Briefly, single-cell suspensions of HeLa were seeded onto a 6 well plate at a density of 100,000 cells/well. After 24 h extracts diluted in DMSO were added and incubated for an additional period of 24 h. Then, the medium was removed, cells were detached with trypsin, resuspended with PBS (200 µL), transferred to 1.5 mL tubes and centrifuge 1200 rpm × 5 min (Centrifuge 5418 R, Eppendorf). Each group of cell suspensions were incubated with 1 mL of acridine orange solution (20 µM in PBS) for 30 min at room temperature under dark conditions. After centrifuge at 1200 rpm for 5 min, cells were washed twice with PBS to remove the excess of staining, centrifuged one last time and resuspended in PBS. Then, 100 µL of each cell suspension were added onto a black-walled, clear-bottom, flat bottom 96 well plate and measured fluorescence intensity at 502/528 excitation/emission wavelengths using a microplate reader (Varioskan LUX VLBL0TD0, Thermo Fisher Scientific, Waltham, MA, USA). The number of cells was normalized with crystal violet staining and suspending the dye with 1% SDS in PBS, and finally measuring the optical density at 595 nm using a microplate reader as aforementioned.
For fluorescence microscopy images, cells were seeded onto a coverslip into a p6 well at the aforementioned seeding density. Treatments were made exactly the same as previously described and then staining was performed directly on the wells after removal of the medium. Two washes with PBS were made and coverslips were mounted fresh over glass slides to observe under fluorescence microscopy using SP5 Leica apparatus with LAS-AF software (Leica Microsystems) equipped with I3 (450-490) and N2.1 (515-560) excitation filters for green and red channels, respectively. Both channels were merged into individual images using ImageJ.

Live-Cell Imaging
A CX-A imaging platform microscope (Nanolive SA, Lausanne, Switzerland) was used to measure refractive indices, creating a holotomographic 3D image of the cells. HeLa cells were cultured at a density of 50,000 cells/well onto a IBIDI µ-Dish, 35 mm high (Martinsried, IBIDI, Germany) and treated with or without 100 µg/mL of extract right before the acquisition of the images. Data obtained were transferred to FIJI (NIH, USA) for image analysis. STEVE software (Nanolive SA) was used for the analysis of the refractive indices and obtention of the Confluency, Mean Cell Area, and Mean Average Dry Mass Density.

Conclusions
In summary, we have proposed a general methodology and empirical equation to assess the BP during the screening of extracts from plants. We applied the said protocol to four Hypericum species from the Canary Islands (H. canariense L., H. glandulosum Aiton, H. grandifolium Choisy and H. reflexum L.f.), specifically to evaluate the anticancer potential (AP) of their methanol-water microextracts. The assessment of AP involved the evaluation of the yield of extraction, the assessment of the antiproliferative activity of the microextracts and their dereplication by LC-MS. After the application of this pharmacognostic protocol, it was possible to assess the AP of the methanol-water microextracts and to select H. grandifolium as the best candidate to carry out the bioguided isolation of polar natural products with high added value.
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/molecules27186101/s1, Figure S1: LC-PDA-MSn chromatograms for Hypericum MW microextracts; Figure S2: Confluency obtained with STEVE software based on refractive indexes resulting from CX-A observation over time; Figure S3: Mean Cell Area obtained with STEVE software based on refractive indexes resulting from CX-A observation over time; Figure S4: Average dry Mass Density obtained with STEVE software based on refractive indexes resulting from CX-A observation over time; Table S1: Antiproliferative activity of Hypericum microextracts; Table S2: Total activity of Hypericum microextracts; Video S1: Live-cell imaging of MW microextract of H. grandifolium.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.