Determination of the Total Phenolics Content and Antioxidant Activity of Extracts from Parts of Plants from the Greek Island of Crete

Oxidative damages are responsible for many adverse health effects and food deterioration. The use of antioxidant substances is well renowned, and as such, much emphasis is placed on their use. Since synthetic antioxidants exhibit potential adverse effects, plant-derived antioxidants are a preferable solution. Despite the myriads of plants that exist and the fact that numerous studies have been carried out so far, there are many species that have not been examined so far. Many plants under research exist in Greece. Trying to fill this research gap, the total phenolics content and antioxidant activity of seventy methanolic extracts from parts of Greek plants were evaluated. The total phenolics content was measured by the Folin–Ciocalteau assay. Their antioxidant capacity was calculated by the 2,2-Diphenyl-1-picrylhydrazyl (DPPH) scavenging test, the Rancimat method based on conductometric measurements, and the thermoanalytical method DSC (Differential Scanning Calorimetry). The tested samples were obtained from several parts of fifty-seven Greek plant species belonging to twenty-three different families. Both a remarkably high phenolic content (with gallic acid equivalents varying between 311.6 and 735.5 mg/g of extract) and radical scavenging activity (IC50 values ranged from 7.2 to 39.0 μg/mL) were found in the extract of the aerial parts of Cistus species (C. creticus subsp. creticus, C. creticus subsp. eriocephalus, C. monspeliensis, C. parviflorus and C. salviifolius), Cytinus taxa (C. hypocistis subsp. hypocistis, C. hypocistis subsp. orientalis and C. ruber), and Sarcopoterium spinosum. Furthermore, the sample of Cytinus ruber showed the highest protection factor (PF = 1.276) regarding the Rancimat method, which was similar to that of butylated hydroxytoluene (BHT) (PF = 1.320). The results indicated that these plants are rich in antioxidant compounds, potentiating their use either as food additives to enhance the antioxidant properties of food products, or protect them from oxidation, or as sources for the preparation of food supplements with antioxidant properties.


Introduction
Lipid peroxidation is a major cause of deterioration during processing and storage, which leads to losses of quality and nutritional value and the development of unpleasant flavors. In addition, oxidative stress, in which reactive oxygen molecules such as superoxide, hydroxyl, and peroxyl radicals are generated, has been suggested to be the cause of aging and various diseases in humans [1]. To overcome the abovementioned problems, the addition of antioxidants is required, since it assists in the preservation of flavor and color and in food quality deterioration avoidance. The most frequent antioxidants used to In addition, several chemical, instrumental, and sensory techniques are commonly used to monitor the oxidation in foods, predict their shelf stability, and evaluate their effectiveness as antioxidants in different lipid systems. Recently, several accelerated oxidation tests have been applied to examine the oxidative stability of edible oils and the ability of antioxidants to prolong their life [16]. The specificity and sensitivity of each method do not lead to a complete examination of all phenolic compounds in the examined extract. A combination of several tests could provide a more reliable assessment of its antioxidant activity [17]. Most methods are based on oxygen absorption and the formation of volatile oxidation products, e.g., the Rancimat method. However, other techniques, such as the Differential Scanning Calorimetry (DSC) method, have also been used for the investigation of the effects of flavonoids on the thermal auto-oxidation of palm oil and other vegetable oils [18].
The present study aimed to investigate possible new sources of natural antioxidants, which would be involved in the protection against diseases involving reactive oxygen species (ROS) and also be useful in food conservation. To this end, seventy methanolic extracts were prepared from fifty-seven Greek plant species (some of them not examined so far, to the best of our knowledge) and examined using the above-mentioned assays to obtain a better overview of their antioxidant capacity. The plants were collected from Crete, which is a Greek Island with unique flora, including interesting species and endemic plants. We aimed to study, highlight, and valorize these plant extracts as potential food additives. It is worth mentioning that the selected plant taxa, common and endemic, are good representatives of the Cretan flora.
Regarding the Cytinus taxa, there are only a few previous reports that examine these plants [19]. However, some phenolics have been identified, including phenolic acids (such as 5-O-caffeoylquinic acid), flavonoids (including flavones, apigenin derivatives, myricetin), and hydrolysable tannins (mainly gallotannins) [20]. The latter are of great importance because they can exhibit not only high antioxidant activity but also other bioactivities, such as antibacterial, anti-inflammatory, etc. [20]. Regarding the Arum creticum and Arum idaeum species, they were found to have almost the same content in polyphenols with Arum dioscoridis [21]. Additionally, the results obtained herein are in accordance with previous studies, which showed that the methanolic extracts of the above-mentioned extract are rich in polyphenols, such as tannins from Cytinus taxa [22], flavonoids, and catechin derivatives from Cistus species [23][24][25].

DPPH Radical Scavenging Activity
The concentration of an antioxidant for decreasing the initial DPPH concentration by 50% (IC 50 ) is a parameter widely used to measure antioxidant activity [26]. Between two samples, the one with the lower IC 50 value exhibits the higher antioxidant activity. The scavenging activity of the plant extracts is shown in Table 1. It is noteworthy that all extracts that had a high phenolic content (>150 mg/g) showed a remarkable capacity to inhibit the DPPH radical (>80% at 200 µg/mL). The most effective DPPH radical scavengers (IC 50 <50 µg/mL) were the extracts of Cytinus taxa (C. hypocistis subsp. orientalis, C. ruber, and C. hypocistis subsp. hypocistis), Cistus monspeliensis, C. salviifolius, C. parviflorus, C. creticus subsp. creticus, C. creticus subsp. eriocephalus, Origanum microphyllum, Sarcopoterium spinosum, Cynoglossum columnae, and Daphne sericea.

Protection against Sunflower-Oil-Induced Oxidative Rancidity
The results represent a comparative study of the antioxidant activity of the sample extracts and known antioxidants (BHT and α-tocopherol) based on their protection factor. All sample extracts and antioxidants are presented at a concentration of 100 ppm. In most cases, a protection factor higher than 1 was recorded, as shown in Table 1. The sample of Cytinus ruber showed the highest protection factor (PF = 1.276) in the Rancimat method, which was similar to that of BHT (PF = 1.320). Additionally, the sample of Berberis cretica L. showed a significantly high protection factor (PF = 1.138), which was higher than that of α-tocopherol (PF = 1.090).

Differential Scanning Calorimetry (DSC)
The thermal-oxidative decomposition of the pure extracts was studied using DSC. In comparison to the Rancimat method, DSC is concluded to be useful as a method employing Plants 2023, 12, 1092 6 of 15 milder conditions and a shorter time, which can be applied for the evaluation of the oxidative stability of samples containing volatile antioxidants and other lipid systems containing water [27]. An exothermic peak is observed in the range of 200 to 365 • C, related to the auto-oxidation process of the samples. Using the curves, the onset temperature (T o ) at which the auto-oxidation process begins is determined [28]. Cytinus taxa (C. hypocistis subsp. hypocistis, C. hypocistis subsp. orientalis, and C. ruber) showed the highest oxidative stability in the DSC method. Owing to the results of the statistical analysis (vide infra), more emphasis was placed on the extracts from the Rafflesiaceae family. The effects of the thermal profile of pure extracts (family Rafflesiaceae) compared to α-tocopherol are shown in Figure 1. The onset temperature (T o ) of the Rafflesiaceae family curves ranged from 300 to 335 • C and was similar to that of α-tocopherol (313 • C). The thermal-oxidative decomposition of the pure extracts was studied using DSC. In comparison to the Rancimat method, DSC is concluded to be useful as a method employing milder conditions and a shorter time, which can be applied for the evaluation of the oxidative stability of samples containing volatile antioxidants and other lipid systems containing water [27]. An exothermic peak is observed in the range of 200 to 365 °C, related to the auto-oxidation process of the samples. Using the curves, the onset temperature (To) at which the auto-oxidation process begins is determined [28]. Cytinus taxa (C. hypocistis subsp. hypocistis, C. hypocistis subsp. orientalis, and C. ruber) showed the highest oxidative stability in the DSC method. Owing to the results of the statistical analysis (vide infra), more emphasis was placed on the extracts from the Rafflesiaceae family. The effects of the thermal profile of pure extracts (family Rafflesiaceae) compared to α-tocopherol are shown in Figure 1. The onset temperature (To) of the Rafflesiaceae family curves ranged from 300 to 335 °C and was similar to that of α-tocopherol (313 °C). Thermal profile of plant extracts (family Rafflesiaceae) compared to α-tocopherol, as determined by the differential scanning calorimetry.

Statistics
A statistical analysis of the data presented in Table 1 was carried out in order to draw more conclusions. For the statistical analysis, only the plant extracts that exhibited significant antioxidant activity (≥50% scavenging of DPPH free radicals) were used.
In order to reduce the complexity of the multivariate data and obtain a better view of the results, a principal component analysis (PCA) was performed. As observed in Figure  2, the two main components that could account for 86.3% of the variation were chosen (Eigenvalues > 1), and this was considered to be a statistically significant parameter (p <0.0001). PC1 demonstrated a positive association with TPC and antioxidant assays and a negative correlation with IC50, and it explained 65.9% of the variability. With a positive association between IC50, TPC, and PF and a negative correlation between To and the percentage of DPPH radicals reduced, PC2 can account for 20.4% of the variance in the data.
According to the PCA plot in Figure 2, TPC, To, and DPPH all have nearly identical loading directions; however, PF has a different loading direction and clearly differs from the other variables in terms of IC50. As can be seen, TPC is more strongly, positively associated (>0.7) with the To parameter and is less strongly correlated (>0.4) with PF. Figure 1. Thermal profile of plant extracts (family Rafflesiaceae) compared to α-tocopherol, as determined by the differential scanning calorimetry.

Statistics
A statistical analysis of the data presented in Table 1 was carried out in order to draw more conclusions. For the statistical analysis, only the plant extracts that exhibited significant antioxidant activity (≥50% scavenging of DPPH free radicals) were used.
In order to reduce the complexity of the multivariate data and obtain a better view of the results, a principal component analysis (PCA) was performed. As observed in Figure 2, the two main components that could account for 86.3% of the variation were chosen (Eigenvalues > 1), and this was considered to be a statistically significant parameter (p < 0.0001). PC1 demonstrated a positive association with TPC and antioxidant assays and a negative correlation with IC 50 , and it explained 65.9% of the variability. With a positive association between IC 50 , TPC, and PF and a negative correlation between T o and the percentage of DPPH radicals reduced, PC2 can account for 20.4% of the variance in the data. Additionally, the highest correlation (0.797) was found between To and the % scavenging, which was found to be statistically significant (p <0.0001). Furthermore, it is well known that the IC50 and % scavenging of DPPH radicals correlate negatively. A higher antioxidant activity is associated with lower IC50 concentrations. Thus, higher TPC concentrations are reflected in lower IC50 results. The dendrogram that was created with the identification of the plant extracts that were considered to be the most comparable was the objective of the hierarchical cluster analysis. Ward's method is the criterion applied in the hierarchical cluster analysis. Cytinus ruber, which offers a strong justification for its superiority compared to all other plant extracts, was clustered separately in Figure 3. Other members of the same family (Rafflesiaceae)-notably, Cytinus hypocistis-were likewise grouped separately, which may be viewed as strong support for its superiority to all other plant extracts. According to the PCA plot in Figure 2, TPC, T o , and DPPH all have nearly identical loading directions; however, PF has a different loading direction and clearly differs from the other variables in terms of IC 50 . As can be seen, TPC is more strongly, positively associated (>0.7) with the T o parameter and is less strongly correlated (>0.4) with PF. Additionally, the highest correlation (0.797) was found between T o and the % scavenging, which was found to be statistically significant (p < 0.0001). Furthermore, it is well known that the IC 50 and % scavenging of DPPH radicals correlate negatively. A higher antioxidant activity is associated with lower IC 50 concentrations. Thus, higher TPC concentrations are reflected in lower IC 50 results.
The dendrogram that was created with the identification of the plant extracts that were considered to be the most comparable was the objective of the hierarchical cluster analysis. Ward's method is the criterion applied in the hierarchical cluster analysis. Cytinus ruber, which offers a strong justification for its superiority compared to all other plant extracts, was  Figure 3. Other members of the same family (Rafflesiaceae)-notably, Cytinus hypocistis-were likewise grouped separately, which may be viewed as strong support for its superiority to all other plant extracts.  Figure 4 shows the fit curves for antioxidant assays by TPC. In each plot, the linear fit and various statistics were displayed (i.e., equation, summary-of-fit, ANOVA, and parameter estimates). The linear fits, however, exhibited a low R 2 . Thus, curve fitting was carried out so as to have a better fit. Following that, the transformation fit had a higher R 2 than the linear fit. Regarding the % DPPH scavenging in relation to the TPC, a reciprocal curve fit was found to be the most suitable, with an R 2 value of 0.63. This was also the case for TPC and To (R 2 = 0.68). A logarithmic plot curve was found to be the most suitable in explaining the relation between IC50 values and TPC (R 2 = 0.80). Otherwise, a linear positive correlation between the total phenolic content and antioxidant activity was reported in the study of Skotti et al. [9].  Figure 4 shows the fit curves for antioxidant assays by TPC. In each plot, the linear fit and various statistics were displayed (i.e., equation, summary-of-fit, ANOVA, and parameter estimates). The linear fits, however, exhibited a low R 2 . Thus, curve fitting was carried out so as to have a better fit. Following that, the transformation fit had a higher R 2 than the linear fit. Regarding the % DPPH scavenging in relation to the TPC, a reciprocal curve fit was found to be the most suitable, with an R 2 value of 0.63. This was also the case for TPC and T o (R 2 = 0.68). A logarithmic plot curve was found to be the most suitable in explaining the relation between IC 50 values and TPC (R 2 = 0.80). Otherwise, a linear positive correlation between the total phenolic content and antioxidant activity was reported in the study of Skotti et al. [9].

Plant Material
The plant species and the parts used herein are presented in Table 2. The freshly collected plant parts were sorted out, dried in a room with active ventilation at ambient temperature, packed in bags, and stored at room temperature. All plants were collected in Crete, Greece, after 2017 and were identified by Dr. E. Kalpoutzakis. The voucher specimens were kept in the herbarium of the Laboratory of Pharmacognosy and Natural Products Chemistry, Department of Pharmacy, University of Athens, Greece. The specimen numbers and the places of the collection are also listed in Table 2. The plant families, genera, and species names are according to Dimopoulos et al. [29], except for the members of the genus Cytinus L., which are named in accordance with the Flora Europaea [30].

Preparation of the Plant Extracts
The pulverized plant materials (50 g) were defatted by maceration for 48 h with dichloromethane and subsequently extracted by maceration for 48 h with 0.5 L of methanol (analytical grade). The extraction step was repeated two more times. The three methanolic extracts were combined. Next, the organic solvent was removed by vacuum distillation. All residues were then stored in a dry place protected from light.

Determination of Total Phenolics in the Extracts
The concentration of total phenolic compounds in the MeOH extracts was determined spectrometrically using the Folin-Ciocalteu method [31], using gallic acid as a standard to prepare a calibration curve. A total of 1 mL of plant extract (10 g/L) was mixed with 5 mL of Folin-Ciocalteu reagent and 4 mL (75 g/L) of sodium carbonate, and after 1 h, the absorption of the reaction mixture was measured at 765 nm against a methanol blank, using a Shimadzu UV-1700 UV/vis spectrophotometer (Tokyo, Japan). The results were expressed as milligrams of gallic acid equivalent (GAE) per gram of extract, based on the reference gallic acid calibration curve (at a linearity range of 1-10 µg/mL, with the equation y = 0.0834x + 0.0925 and R 2 = 0.9967) generated for this study. All determinations were performed in triplicate.

DPPH Radical Scavenging Assay
The radical scavenging activity of the plant extracts against stable DPPH was determined spectrometrically according to a previously reported procedure [32]. Briefly, 100 µL of the sample solution (200 mg/L), diluted in dimethylsulfoxide, was added to 1.9 mL of a 315 µM DPPH solution (in ethanol) and allowed to react for 30 min at 37 • C. A blank sample was prepared by adding 100 µL of dimethylsulfoxide in the DPPH solution. Then, the absorbance was measured at 515 nm, and the % scavenging was calculated using the following equation: where A 0 and A are the absorbances of the blank solution and the sample, respectively. The IC 50 values correspond to the amount of each sample required to scavenge 50% of the DPPH free radicals. They were calculated from regression lines, where the abscissa represents the sample concentration, and the ordinate is the average percent reduction of the DPPH radical. Each IC 50 value corresponds to an average of three separate tests. Plant extracts that achieved lower than 50% scavenging of DPPH radicals were not further examined.

Protection against the Oxidative Rancidity of Sunflower Oil
The method used was adapted from Lalas and Tsaknis [33]. Two and a half grams of sunflower oil and an antioxidant (plant extract, BHT, or α-tocopherol, in various concentrations) were accurately weighed into the reaction vessel of a Rancimat 679 (Metrohm LTD, Herisau, CH 9101, Switzerland). At the same time, in another vessel, pure sunflower oil (iodine value: 115 g I/100 g) (Elais S.A., Athens, Greece) was added (without antioxidants) to be considered as a control sample. A total of 1 mL of the appropriate solvent (methanol or dichloromethane) was added in order to dissolve the antioxidant and mixed well. The conditions were set at a temperature of 90 • C and an airflow of 15 L/h. The protection factor (PF) was calculated as follows: PF = (induction period with antioxidant)/(induction period without antioxidant). A protection factor greater than 1 indicates the inhibition of lipid oxidation. The higher the value, the better the antioxidant activity [33].

Differential Scanning Calorimetry (DSC)
The antioxidant action of extracts was estimated using the DSC method with a Perkin Elmer DSC-6 calorimeter (Perkin Elmer Corp., Norwalk, CT, USA). Oxidative stability was determined using the method of Tan and Che Man [34]. A total of 4 mg of the sample extracts (or α-tocopherol for comparison) was placed in DSC aluminum pans closed with lids perforated by a hole (internal diameter: 1 mm) in the center in order to allow the sample to be in contact with the oxygen stream. The purge gas foaming the reaction atmosphere was oxygen. The starting temperature of oxidation was determined as the onset temperature of the oxidation peak. The temperature program was: heat from 30 • C to 180 • C (at a rate of 100 • C/min), hold for 1 min at 180 • C, and, finally, heat from 180 • C to 390 • C (at a rate of 10 • C/min).

Statistics
Principal component analysis (PCA), hierarchical cluster analysis, and statistical analysis were all carried out using the JMP ® Pro 16 (SAS, Cary, NC, USA) software. Each plant extract was subjected to three separate analyses, with three replicates of each determination described above.

Conclusions
During the screening of fifty-seven plants in this work, Cytinus taxa (C. hypocistis subsp. hypocistis, C. hypocistis subsp. orientalis, and C. ruber), Cistus species (C. creticus subsp. creticus, C. creticus subsp. eriocephalus, C. monspeliensis, C. parviflorus, and C. salviifolius), and Sarcopoterium spinosum were found to be the most promising ones. All these extracts showed a high phenolic concentration and significant free radical scavenging activity. Since the reports for the TPC and antioxidant activity of most of the examined plant species are scanty and sparse, the results of this study can be used as a benchmark for future studies on the same plant species. Moreover, plant species that were overlooked or not thoroughly examined were highlighted as potential candidates so that they can be further studied and used for industrial purposes.