Chemical Composition of Cynara Cardunculus L. var. altilis Heads: The Impact of Harvesting Time

Cardoon is a multi-purpose crop with several industrial applications, while the heads (capitula) are edible and commonly used in various dishes of the Mediterranean diet. Several reports in the literature study the chemical composition of the various plants parts (leaves, flower stalks, bracts, seeds) aiming to industrial applications of crop bio-waste, whereas for the heads, most of the studies are limited to the chemical composition and bioactive properties at the edible stage. In the present study, cardoon heads were collected at six different maturation stages and their chemical composition was evaluated in order to determine the effect of harvesting stage and examine the potential of alternative uses in the food and nutraceutical industries. Lipidic fraction and the content in fatty acids, tocopherols, organic acids, and free sugars were determined. Lipidic content decreases with the maturation process, while 22 fatty acids were detected in total, with palmitic, oleic, and linoleic acids being those with the highest abundance depending on harvesting time. In particular, immature heads have a higher abundance in saturated fatty acids (SFA), whereas the samples of mature heads were the richest in monounsaturated fatty acids (MUFA). The α-tocopherol was the only isoform detected being present in higher amounts in sample Car B (619 µg/100 g dw). Oxalic, quinic, malic, citric and fumaric acids were the detected organic acids, and the higher content was observed in sample Car E (15.7 g/100 g dw). The detected sugars were fructose, glucose, sucrose, trehalose and raffinose, while the highest content (7.4 g/100 g dw) was recorded in sample Car C. In conclusion, the maturation stage of cardoon heads influences their chemical composition and harvesting time could be a useful means to increase the quality and the added value of the final product by introducing this material in the food and nutraceutical industries.


Introduction
Cynara cardunculus L., or commonly known as cardoon, belongs to the Asteraceae family which is one of the largest families of the plant kingdom with more than 2000 species. Cynara cardunculus comprises three botanical varieties, all native to the Mediterranean basin, the wild cardoon (var. sylvestris), the domesticated cardoon (var. altilis DC), and the globe artichoke (var. scolymus) [1,2]. This crop has been gaining attention due to the high biological and industrial potential that it has

Plant Material
Head (capitulum) samples of Cynara cardunculus var. altilis DC cv. Bianco Avorio (Fratelli Ingegnoli Spa, Milano, Italy) were collected from the experimental farm of the University of Thessaly in Velestino, in central Greece (22.756 E, 39.396 N), during the growing period of 2017-2018 ( Figure 1). Heads were collected from 15 individual plants (n = 15) at the beginning of the flower's development, in full maturity, and in seed ripening stages. The climate conditions during the growing period and the procedure used for the collection and treatment of plant material were described by Mandim et al. [19]. Briefly, Car A corresponds to principal growth stage (PGS) 5 (harvest: 26 April 2018), Car B corresponds to PGS 5/6 (harvest: 10 May 2018), Car C corresponds to PGS 6 (harvest: 24 May 2018), Car D corresponds to PGS 6/7 (harvest: 04 July 2018), and finally the samples Car E and Car F correspond to PGS 7 and PGS8, respectively (harvest: 09 August 2018 and 29 August 2018, respectively) [19,26]. The plant material and the growing condition have been described in previous reports by our team. Briefly, bract samples were collected from 8-year-old plants sexually propagated from seeds in 2010. Soil parameters were the following, as previously described by the authors [27]: loam texture (48% Sand; 29% Silt; 23% Clay); Organic matter: 1.3%; pH: 7.9; EC: 1.4 dS/m; NO −3 : 9.49 mg/kg; P: 74.53 mg/kg; K exch : 0.98 cmol c /kg; Ca exch : 13.96 cmol c /kg; Mg: 4.32 cmol c /kg. Prior to crop establishment a base dressing was applied by using 50 kg/ha N, 90 kg/ha P 2 O 5 and 40 kg/ha K 2 O. After crop establishment, nitrogen fertilizers were applied with side dressing at each growing period and before plant regrowth (100 kg/ha N). Plant density was 40,000/ha with distances of approximately 0.6 m between rows and 0.4 m within rows. Irrigation was applied monthly during the first growing period (staring on April and until July) with water cannons, whereas in the following years irrigation was applied only twice in each growing period (on April and May) due to the extensive root system that plants form after the second year of establishment. Weed control was applied with hoeing after plant regrowth at each growing period, since at later growth stages plant is very competitive against weeds. No pesticides and fungicides were applied. Climate conditions during the experimental period (shoot emergence until senescence) are presented in Figure 2.

Fatty Acids
For the analysis of fatty acids composition, the lipidic fraction was extracted with petroleum ether through a Soxhlet extraction system at 120 • C. After a transesterification process, the fatty acids content was analyzed by Gas-liquid Chromatography (GC), coupled to a Flame Ionization Detector (FID) at 260 • C and according to the analytical conditions previously described [27]. The identification and quantification of fatty acids were performed by comparing the relative retention times of FAME peaks from samples with standards (reference standard mixture 47,885-U), using Clarity DataApex 4.0 software (Prague, Czech Republic). The results were expressed as relative percentages and in mg per 100 g dw of each detected fatty acid.

Tocopherols
Tocopherols composition was analyzed using a high-performance liquid chromatography system (HPLC, Knauer, Smartline system 1000, Berlin, Germany) coupled to a fluorescence detector (FP-2020; Jasco, Easton, USA) programmed for excitation at 290 nm and emission at 330 nm, according to the procedure previously described [28]. The qualitative and quantitative analysis were performed using the Clarity 2.4 software (DataApex, Prague, Czech Republic) and was achieved through comparison of the chromatographic data (retention times and spectra) with commercial standards, using the internal standard method. Tocopherols standards (α-, β-, γ-, and δ-isophorms,) were used for compounds identification and quantification by the internal standard method. Results were expressed in mg per 100 g of dry weight (dw) and were processed using the Clarity 2.4 software (DataApex, Prague, Czech Republic).

Organic Acids
Organic acids were determined following the procedure previously described [29]. The samples were analyzed by Ultrafast Liquid Chromatography (UPLC, Shimadzu 20A series, Kyoto, Japan) coupled to a Diode Array Detector (UFLC-PDA, Shimadzu Corporation, Kyoto, Japan). Data were analyzed using the LabSolutions Multi LC-PDA software (Shimadzu Corporation, Kyoto, Japan). The identification was accomplished through the comparison of the retention times and the spectra obtained to the commercial standards (oxalic, quinic, malic, ascorbic, citric and fumaric acids), and their respective calibration curves were used to determine the quantity based on the area of the peaks. Results were expressed in g per 100 g of dw.

Free Sugars
Free sugars content was analyzed by High Performance Liquid Chromatography (HPLC, Knauer, Smartline system 1000) coupled to a refractive index detector (RI detector, Knauer Smartilne 2300, Knauer, Berlin, Germany), according to the analytical conditions previously described [30]. Data was analyzed using the Clarity 2.4 software (DataApex) and the identification was performed through the comparison with standards (D (-)-fructose, D (+)-sucrose, D (+)-glucose, D (+)-trehalose and D (+)-raffinose pentahydrate (Sigma-Aldrich, St. Louis, MO, USA). The quantification was performed using melezitose (Matreya, PA, USA) as internal standard (IS). The results were processed through Clarity 2.4 software (DataApex, Prague, Czech Republic) through the comparison of retention times, UV-Vis and mass spectra of the sample compounds with those obtained from the available standards and the literature information available and presented in g per 100 g of dry weight (dw).

Statistical Analysis
All the performed experiments were executed in triplicate. Results were presented as mean value ± standard deviation. Means and standard deviations were calculated using Microsoft Excel. Differences among samples were analyzed using SPSS Statistics software (IBM SPSS Statistics for Mac OS, Version 26.0. Armonk, NY: IBM Corp.). The results were subject to an analysis of variance (ANOVA), while the Tukey's HSD test (α = 0.05) was used to assess the significant differences between the samples. For the comparison between two samples, a two-tailed paired Student's t-test was applied to assess the statistical differences (α = 0.05).
Moreover, a Principal Component Analysis (PCA) was performed in order to examine the contribution of each variable to the total diversity and classify the studied maturation stages according to their chemical composition and nutritional value by using the statistical software Statgraphics 5.1.plus (Statpoint Technologies, Inc., VA, USA).

Lipid Fraction and Fatty Acids Composition
In Table 1 are presented the lipid and tocopherols content of cardoon heads collected at different maturation stages. The highest total lipidic fraction and a-tocopherol content was observed at early maturity stages (Car A and Car B for total lipidic fraction and α-tocopherol content, respectively) and immature heads (Car A) had 10.9 times higher lipidic content than the sample of late maturity (Car F). In contrast, the progress of maturation process resulted in a gradual decrease of total lipid fraction, whereas α-tocopherol content showed fluctuating trends with the lowest content being observed in Car C samples (mid-maturity stages). In particular, the highest amount was detected in the sample Car B (PSG 5/6) (619 µg/100 g dw), while sample Car C (PSG 6) presented the lowest abundance (25 µg/100 g dw). Table 1. Lipidic fraction and tocopherols of Cynara cardunculus heads at various stages of maturity (mean ± SD; n = 3).

Organic Acids and Free Sugars
In Table 3 are presented the results regarding the content in organic acids of Cynara cardunculus heads, in relation to maturity stage. The main detected organic acids were oxalic, quinic and malic acid, followed by citric and fumaric acids which were detected in lower amounts. Moreover, a great variation in individual organic acids content was observed at different maturity stages. In particular, at early stages (samples Car A and B) malic acid was the most abundant organic acid (1.45 and 2.31 g/100 g dw, respectively), while at late stages (sample F) oxalic acid was the richest organic acid (12.1 g/100 g dw) followed by quinic acid (3.3 g/100 g dw).  The free sugars composition of cardoon heads in relation to maturity stage is presented in Table 4. Sucrose, glucose and raffinose were the main detected sugars, followed by fructose and trehalose, while a great variation in sugar composition was observed in response to maturity stage. In particular, sample Car C had the highest content in total sugars (7.4 g/100 g dw), with raffinose and fructose (1.8 and 1.64 mg/100 g dw, respectively) being present in higher amounts; the same sample was also the only one containing all the detected sugars. Samples Car D and Car E revealed the lowest content of total sugars (1.03 mg/100 g dw). Moreover, in the most advanced maturation stages (samples Car D-F) trehalose was the free sugar being recorded in higher amounts (0.34-0.96 mg/100 g dw), whereas raffinose was not detected. In the remaining samples car B and C, the presence of raffinose and sucrose stands out, except for sample Car B where sucrose was not detected.

Principal Component Analysis (PCA)
Principal component analysis (PCA) is used to reduce multivariate data complexity as a method of identifying patterns and expressing data in ways that highlight similarities and differences, and further identify groups of samples according to their maturation stage [31,32]. The first four principal components (PCs) were associated with Eigen values higher than 1 and explained 97.2% of the cumulative variance, with PC1 accounting for 43.8%, PC2 for 22.9%, PC3 for 19.4% and finally PC4 for 11.1%. PC1 was positively correlated to lipidic fraction, raffinose and total sugars, malic, citric and fumaric acid, linoleic and α-linolenic acid and PUFA content, whereas it was negatively correlated to oxalic and quinic acid, total organic acids and MUFA content. PC2 was positively correlated to sucrose, quinic acid, palmitic and oleic acid and SFA content, whereas it was negatively correlated to trehalose, linoleic and α-linolenic acid and PUFA. Similarly, PC3 was positively correlated to α-tocopherol, tregalose and oxalic acid content, whereas it was negatively correlated to fructose, sucrose, raffinose, total sugars, malic and citric acid and oleic acid. Finally, PC4 was positively correlated to fructose, sucrose, total organic acids, oleic acid and MUFA content, whereas it was negatively correlated to palmitic and stearic acid and SFA content. These results indicating a correct application of the PCA allowing differentiation between the tested maturity stages, as shown in the corresponding scatterplot ( Figure 4). Moreover, the plot suggests that the differences in the chemical composition of the tested samples are correlated with the maturation stage. The early (sampes Car A and B) and mid and late stages (samples Car D and F) are closely positioned, whereas samples Car C and Car E are clearly distinct due to the very low α-tocopherol and very high organic acids content, respectively, compared to the rest of the tested maturity stages. The loading plot ( Figure 5) of the first two components revealed groups of positively correlated variables, namely the upper right quadrant comprising fructose, sucrose, raffinose, total sugars, lipidic fraction, SFA, palmitic acid, malic acid and fumaric acid; the lower right quadrant comprising tocopherols, glucose, citric acid, PUFA and and linoleic acid; the upper left quadrant comprising quinic acid, oxalic acid, total organic acids and oleic acid; the lower left quadant comprising trehalose and α-linolenic acid

Discussion
It could be suggested that the maturation stage of cardoon heads affects the fatty acids content which do not present a similar qualitative and quantitative profile over the maturation process. For example, pentadecanoic acid (C15:1) was detected only in senescent head samples (Car E and F), whereas docosadienoic acid (C22:2) was only detected in immature heads (samples Car A and B). The fatty acid profile of cardoon has already been studied by our team who studied cardoon seeds collected at full maturity and when heads were dry and senesced and the influence of harvesting time was suggested [33]. Moreover, harvesting time is also associated with variable climate conditions as shown in Figure 2, which according to the literature may also affect the compositional profile of globe artichoke [20]. Therefore, the dry conditions during the period of March-May in our study could be implicated in the observed differences among the tested maturation stages. Analogous fatty acids composition in cardoon heads has also been previously described by Petropoulos et al. [34], who studied different cardoon genotypes and verified that palmitic and linoleic acids were present in higher abundance in all the studied heads. Palmitic and linoleic acids were also suggested as the main fatty acid in globe artichoke heads by Dosi et al. [35] although they reported significantly higher amounts of linoleic than palmitic acid (55.20 and 34.80 mg/100 g fw, respectively) compared to our study. The studies of the lipidic fraction in cardoon tissues mostly refer to its seeds [36][37][38], as a result of the great interest for its industrial potential for the biodiesel production, although fatty acids composition of stalks, capitula and leaves has been also reported [39]. The high content in fatty acids present in cardoon heads, particularly the essential linoleic and oleic acids, is an added value to this multi-purpose crop that could be used for the production of these acids in industrial scale [40]. To the best of our knowledge, this is the first report that evaluated the influence of the physiological stages of cardoon heads collected during all the flowering stage.
Regarding fatty acids classification, our results showed that the saturated fatty acids (SFA) were the most abundant class of fatty acids in heads of early maturity (Car A-C), whereas monounsaturated fatty acids (MUFA) were the predominant ones in the remaining samples (samples D-F). Furthermore, the maturation stage of the samples reveals a strong influence both on the fatty acids profile, as well as on the proportion of PUFA/SFA and n-6/n-3, parameters which are associated with the nutritional value and the functional properties of food products [33,41]. Sample Car B was the only sample with a PUFA/SFA value higher than 0.45 and an n-6/n-3 ratio value lower than 4.0, both characteristics being associated with good nutritional properties. This evidence agrees with the popular medicine, since in the Mediterranean cousin cardoon heads are used when immature [19,42,43]. With the results obtained in this study, we corroborate the consumption of immature cardoon heads, preferably harvested in the beginning of May, based on the conditions of the growing location.
The tocopherols content detected in cardoon heads is presented in Table 1. The α-tocopherol was the only isoform identified and was detected in all the studied samples. Tocopherols are antioxidant compounds with high capacity to undergo oxidation reactions, therefore fluctuations of environmental conditions that may induce plant stress throughout the growth cycle of the heads could be a justification for the observed fluctuations in tocopherols content [23]. As shown in Figure 2, variable climatic conditions prevailed during and prior the harvesting period in our study, especially a dry period during March-May which could be associated with stressful conditions that resulted in an increase of tocopherols content. In a previous study, we studied the bioactive properties of cardoon heads also analyzed in the present work and despite the antioxidant capacity associated with tocopherols, the anticipation that the sample Car B with the highest tocopherols content would show the greater antioxidant activity was not verified. This fact suggests that other classes of compounds, such as phenolic compounds, could be also related to the demonstrated antioxidant potential [19], while Kukić et al. [44] suggested that β-sitosterol possessed a strong antioxidant capacity in extracts obtained from cardoon bracts. Despite that, the highest content of α-tocopherol in this sample could be associated with the highest content in PUFAs (see Table 2) highlighting the protective effects of tocopherols against lipid peroxidation [45,46]. Similarly to tocopherols composition of heads, other plant tissues of cardoon do not present a wide variety of tocopherols isoforms and the α-tocopherol was the only isoform detected in cardoon seeds [23,47], contrary to the cardoon seed oil which was rich in αand δ-tocopherols [47][48][49][50] and leaf blades that contained all tocopherol isoforms [51].
The organic acids profile (Table 3) was also different between samples with a great variation during the maturation process of the detected organic acids: oxalic, quinic, malic, citric, and fumaric. The sample collected at the principal growth stage (PGS) 7 (Car E) revealed the highest content in organic acids (15.7 mg/100 g dw), and sample Car F (PGS 8) the lowest abundance (0.89 mg/100 g dw). Immature to mid maturity heads (samples Car A-C) had malic acid as the most abundant organic acid (1.45-2.31 mg/100 g dw). In sample D, the organic acid present in higher quantities was citric acid (0.84 g/100 g dw), while in advanced maturation stages, oxalic and quinic acids were the organic acids present in higher amounts. Moreover, considering the concomitant decrease in total sugars content at late maturity stages, the increase of organic acids especially in sample Car E could be explained by the increased requirements in osmolytes as a mechanism of cardoon plants to overcome stress conditions induced by the high mean and max temperatures and the concomitant low precipitation (see Figure 2) [52,53]. Compared to other reports, cardoon heads revealed greater variety and lower abundance of organic acids than seeds [23,47], while organic acids profile was similar to aerial parts [54]. Finally, the high quinic acid in sample Car E could be valorized for its confirmed high antioxidant potential [55,56].
The qualitative and quantitative information regarding the free sugars composition of cardoon heads are presented in Table 4. The samples analyzed revealed significant differences in free sugars content throughout the maturation process. Therefore, in early stages (sample car A) sucrose and raffinose were the main detected sugars, while sugar composition was altered in the following stage (sample car B) and raffinose and glucose were the most abundant sugars. Similarly, in sample C sucrose, fructose and raffinose were detected in the highest amounts, whereas in the following stages (samples D-F) total sugars content was significantly reduced consisting mainly of trehalose. Considering that the tested samples were collected under environmental conditions [19] could justify the observed differences in sugar composition between the various maturity stages. According to Petropoulos et al. [34] who studied the influence of the genotype on different parameters of cardoon heads, significant differences between the genotypes tested were also suggested. Previous studies identified sucrose [31,34,57] and glucose [58,59] as the most abundant sugars. These differences with the present work could be justified by the effect of factors such as the geographic location and growing conditions [58] or the stage of maturity of the analyzed samples [60], especially at early maturity stages where heads are edible and sweet in taste. In the same context and considering that artichoke is a rich source of inulin [61,62], free sugars content at late maturity stages (samples car D-E) exhibits a significant reduction probably due to the inulin formation and the accumulation of storage carbohydrates [63,64]. This argument could also explain the high content of trehalose at late maturity stages where the amounts of other free sugars is reduced, while the evolving environmental conditions (increasing temperatures, water shortage) could pose plants under stress and increase the needs of osmolytes for stress tolerance [52,65]. Finally, the decrease of total free sugars at late maturity stages should be linked with the lignification process and the hard texture of cardoon heads, since lignin biosynthesis involves the binding of non-structural carbohydrates for the formation of lignin carbohydrate complexes (LCC) [65].

Conclusions
Cardoon heads are widely consumed in several dishes of the Mediterranean countries, such as salad and soups, due to their richness in health-promoting compounds and to their well-recognized nutritional value. However, the stage when heads are edible is relatively short and is highly affected by environmental conditions during harvesting that may result in heads of hard texture due to high fibers content which make them inedible. Therefore, a considerable number of heads are considered as waste with potential suggested uses the energy and biomass production. With the present study we aimed to analyze cardoon heads chemical composition and evaluate the potential of alternative uses that will increase the added value of the crop. Our results allowed a complete characterization of the chemical composition of cardoon heads and the evaluation of the effect that the maturity stage has on the lipid, fatty acids, tocopherols, organic acids, and free sugars content. To the best of our knowledge, this is the first report that studied the influence of the maturity stage on the composition in lipidic content, fatty acids, tocopherols, organic acids, and free sugars content. In particular, immature heads (samples Car A and B) exhibited the highest content in lipidic fraction and α-tocopherol, respectively. Moreover, samples of mid-(sample Car C) and late maturity (sample Car E) had the highest total sugars and total organic acids content, respectively. With the present results we verified that the maturation stage of cardoon heads had a significant influence on the chemical composition. The obtained information should be implemented by different application areas, e.g., nutraceuticals and food supplements, in order to obtain the maximum potential of cardoon heads utilization, as well as to improve the production and commercialization techniques.