1. Introduction

Vegetables are an important source of chemical diversity, useful in both agriculture and medicine (^{Kolewe et al., 2008}). Part of this diversity is due to substances called secondary metabolites, which belong to three main groups: terpenes, phenolic compounds and nitrogenous compounds (^{Zhong, 2011}; ^{Ortiz de Elgueta-Culebras et al., 2017}). These are produced as defence or protection mechanisms when adapting to the environment, and when facing stressful conditions, but are not essential for vital functions (^{Dixon and Paiva, 1995}; ^{Bourgaud et al., 2001}). ^{Medeiros et al. (2009}) point out that substances such as the alkaloids in plant extracts can activate secondary metabolism, as they themselves constitute an abiotic stress and therefore induce the production of polyphenols (^{Koyama et al., 2012}). The latter is reported as the concept of elicitation that is generated by an elicitor (^{Ramirez-Estrada et al., 2016}). This is relevant as polyphenols compose a group of natural antioxidants that have the ability to neutralise free radicals and play an important role in modulating enzymatic detoxification, resulting in beneficial effects for human health (^{Costa et al., 2009}; ^{Andarwulan et al., 2012}; ^{García-Tirado et al., 2012}; ^{Martini et al., 2017}; ^{Tighe-Neira et al., 2017}). In this sense, botanic extracts may be used to bio-stimulate secondary metabolism in other vegetables, as has been observed with extracts of marine algae as a stress-causing agent (^{Araujo et al., 2012}; Tighe et al., 2014). ^{Batish et al. (2008}) found similar results using caffeine in doses of 2000 µM on Phaseolus aureus Roxb. Similarly, U. europaeus, originating in Western Europe and considered an invasive weed (^{Matthei et al., 1995}; ^{Kaufman and Kaufman, 2013}), is a species of interest, as phytochemical studies have shown it contains secondary metabolites such as flavonoids, tannins, polyphenols and quinolizidine alkaloids including anagrine, cytosine, N-methylcytosine, and lupanine (^{Hornoy et al., 2012}). On the other hand, red hot chilli pepper (Capsicum annuum) and lettuce (Lactuca sativa) are important horticultural species with great nutraceutical properties (^{Cheng et al., 2014}; ^{Korkutata and Kavas, 2015}; ^{Sayin and Arslan, 2015}). These species are widely cultivated in greenhouse conditions, leading to constant production throughout the year (^{Batlang, 2008}; ^{Ioslovich, 2009}). Therefore, based on the principle mentioned by ^{Medeiros et al. (2009}) and ^{Koyama et al. (2012}), we hypothesized that U. europaeus extracts could influence the secondary metabolism of species of agronomic interest, increasing their total polyphenol content. In this work, we have selected Capsicum annuum L. and Lactuca sativa L as models to evaluate the effect of applying different doses of aqueous and methanolic extracts from Ulex europaeus L. on seedlings of the models species grown in laboratory and greenhouse conditions.

The assays were conducted in the Plant Biotechnology Laboratory of the School of Agronomy, and in the Chemistry of Natural Products Laboratory of the School of Environmental Sciences, both at Universidad Católica de Temuco. Greenhouses were facilitated by Agroindustry Daube Ltda., Camino Botrolhue km 5.5, Temuco, La Araucanía Region, Chile. Furthermore, U. europaeus was obtained from the Pillanlelbún Experimental Centre, School of Agronomy, Universidad Católica de Temuco, from which C. annuum (Cacho de Cabra) seeds, and L. sativa were also collected. After collection, U. europaeus was dried at room temperature, separated into above-ground (shoot) and below-ground (root) fractions, and ground to make the extracts. The extracts were obtained according to ^{Alghazeer et al. (2012}) with some modifications. Nonetheless, for specific, aqueous extracts we used the methodology described by ^{Benayad et al. (2013}). Thus, the extract was obtained by infusing 5 g of material in 100 mL of distilled water at 90°C. On the other hand, the methanolic extracts were obtained by incubating 10 g of plant material in 40 mL of methanol, changing solvent every 48 h by a period of six days (^{Karthishwaran et al., 2010}). All extracts were vacuum filtered with filter paper N°2 and then clarified by centrifugation at 5000 rpm for 10 min. The resulting fractions were then standardised in their concentration (total dissolved solids) and stored in chilled, sterilised plastic bottles until use.

The L. sativa and C. annuum seeds were disinfected by constant agitation in a Captan solution (0.5 g in 100 mL of distilled water) for 30 min. Subsequently, they were washed with distilled water and dried. C. annuum seeds were immersed in distilled water for 15 h before being used. For laboratory assays, seeds of both species were placed in glass bottles (215 mL) with three circles of N°1 filter paper inside. Fifty seeds of each species were deposited per bottle in a laminar flow bench (Arquimed, AHC - 5A1), and 1 mL solution applied per treatment (see Table 1). The bottles were sealed with plastic film and maintained at 21±2ºC, 100 µmol m^-2s^-1 and a photoperiod of 16/8 h light/darkness in a growth chamber (Inema). For greenhouse assays, a 250 m² greenhouse with a polyethylene cover was used. In addition, on a desk 1.20 m high, the germination and development of the previously-disinfected seeds was carried out, using containers of expanded polystyrene (1.25 x 10^-4 m³) and a mixture of soil, compost and sand in a ratio of 1:1:1, previously sterilised three times at 121ºC for 30 min (HL-340 vertical autoclave). Three mL of each treatment (see Table 1) was applied to each seedling with a manual 500 mL sprayer, every seven days on four occasions, once the first true leaf had appeared.

Five doses of each extract type (aqueous and methanolic) for shoot and root fraction of U. europaeus were used as treatments as described ^{Tighe et al. (2016}). Also, was considered distilled water as an absolute control, and caffeine p.a. as a positive control Table 1.

Table 1: Treatments used in laboratory and greenhouse assays. 

¹Ar = root aqueous, Mr = root methanolic, As = shoot aqueous, Ms = shoot methanolic, Ca = caffeine, C = control

A phytochemical analysis of the extracts was undertaken according to published methods (^{Pascual et al., 2002}; ^{Karthishwaran et al., 2010}; ^{Upadhyay et al., 2010}). Briefly, alkaloids were determined using Dragendorff reagent in drop tests on dye plates, and by thin-layer chromatography on silica gel plates visualized under UV-365 nm lights. Flavonoids were determined by reaction with metallic Mg in an acid medium. The tannins were identified with FeCl₃ solution at 10%, and confirmed with gelatine and gelatine-salt solution. The presence of saponins was determined by the foam test and confirmed with vanillin-sulphuric acid reagent.

The total polyphenol content was performed according to the Folin-Ciocalteau method (^{Georgé et al., 2005}). The reagents used were: acetone:water (7:3 v/v), sodium carbonate solution (75 g L^-¹), Folin solution (diluted in 10 parts of water) and gallic acid (5 - 25 mg L^-¹). A gallic acid curve was prepared as standard by determining the absorbance value in a spectrophotometer (Optizen) at 760 nm. The results were expressed in mg Equivalent of Gallic Acid (EGA) per 100 g of dry material.

The experimental units were placed in a block design completely at random, for both laboratory and greenhouse assays. For the separation of means, an ANOVA test, a Tukey multiple comparison test, or a Student t-test, all with 5% significance was performed as appropriate, using the statistical package JMP 5.0^®.

Alkaloids were not present in the aqueous extracts, whereas all metabolites, especially tannins and saponins were detected in the methanolic extracts (Table 2).

Table 2: Phytochemical analysis of Ulex europaeus L. extracts. 

(-): Negative result, (+): Positive result, (++): Positive result with strong evidence. Ar = root aqueous, As = shoot aqueous, Mr = root methanolic, Ms = shoot methanolic

The low abundance of alkaloids in the extracts may be due to the methods of extraction of the aqueous and methanolic fractions, as in the last step both involved a near neutral pH. As alkaloids are composed of basic nitrates, they are found in plants in the form of salts of organic acids, whose solubility is far greater at lower pHs. In U. europaeus, the quinolizidine alkaloids N-methylcytisine (12.2 mg kg^-1 DM), cytisine (83 mg kg^-1 DM) and anagrine (32 mg kg^-1 DM) have been reported; these play several roles in the plant, including allelopathic interrelations (^{Máximo and Lourenco, 2000}; ^{Máximo et al., 2006}; ^{Hornoy et al., 2012}). The genus Ulex is rich in flavonoids, particularly isoflavones and pterocarpanes (^{Máximo et al., 2002}). Tannins are polar phenolic derivatives and are therefore soluble in water. Saponins, which are a group of oily glycosides, are soluble in methanol and water, yet were not detected in the aqueous root extracts, probably due to the fact that they are found in low concentrations.

Differences were observed in the polyphenol content of leaf for both species cultivated in laboratory and greenhouse conditions (Figure 1). In C. annuum, values are significantly greater (P<0.05) in greenhouse conditions after five days of application of different U. europaeus extracts. On the other hand, the polyphenol contents observed in L. sativa are greater in the laboratory (P<0.05). The response variable between extracts was observed only in the laboratory and for both species. In this context, the polyphenol content of C. annuum is lower in C (P<0.05 than in the other treatments, except Ca. However, in L. sativa, the content in Ca is lower than in all other extracts, while for Ar it is higher (P<0.05). No other significant differences are present between the rest of extracts, including the control.

Figure 1: Polyphenol concentration obtained five days after application of methanolic and aqueous extracts of U. europaeus obtained from different organs. A) C. annuum leaf, B) L. sativa leaf. Different uppercase letters between bars of different environmental conditions and equal extracts, represent significant differences according to the Student t-test (P<0.05). Different lowercase letters between bars of equal environmental conditions and different extracts, represent significant differences according to the Tukey test (P <0.05) 

The exposure of plants to controlled (laboratory) and less controlled (greenhouse) conditions with the same treatments typically triggers greater total polyphenol levels in the greenhouse, given fluctuations in temperature, humidity, and other factors of stress (^{Koyama et al., 2012}). However, this behaviour is only observed in C. annuum. In this sense, ^{Divi et al. (2010}) in a study with Arabidopsis, assert that tolerance to thermal stress is related to the presence of brassinosteroids which, together with other phytohormones, generate a particular response in different genotypes.

In this part of the analysis, the origin (root or shoot) and nature (aqueous or methanolic) of the U. europaeus extracts did not constitute a differentiating factor in either of the two growth conditions. However, upon analysing the levels by condition (laboratory or greenhouse), they do indeed differ and behave as expected according to previous studies (Figures 2 and 3).

Figure 2: Polyphenol content of C. annuum after application of methanolic and aqueous extracts of U. europaeus obtained from different organs, under laboratory and greenhouse conditions. r² = linear regression analysis 

In laboratory-grown C. annuum (Figure 2), the total polyphenol concentration increases concomitant with the dose, except at the highest concentration, where plants suffer a decrease in polyphenols. Regression to extract values are highly significant (P<0.001), but with greater representativeness in Ar and Ca.

Other than detecting higher polyphenol content in the greenhouse, with respect to those found in the laboratory, it is not possible to identify a clear trend for the variable in these less controlled conditions; indeed, the regressions are not significant (P>0.05). In both cases, the observed C values correspond to the lowest in every condition, suggesting an overall positive effect of extracts and doses on the C. annuum.

The total polyphenol concentration in C. annuum can be explained, in part, by the absence or reduction of other abiotic stress factors in controlled laboratory conditions, except in the higher doses. Such doses may cause toxicity and be generally detrimental to the physiological state of the species, as corroborated partly by ^{Batish et al. (2008}) and ^{Medeiros et al. (2009}). Additionally, the stimulation of secondary metabolism by low doses of stress-causing agents also affects primary metabolism, as explained by ^{Cedergreen (2008}) in a study on biomass production. ^{Mitchell and Chassy (2009}) point out that stress may be less attractive in production, but the nutritional value per gram of product may be superior by increasing the content of certain metabolites that improve the plant's nutritional quality.

In L. sativa (Figure 3), the total polyphenol content in laboratory conditions presents a trend similar to that found in C. annuum, due to the direct relationship with increased dose, except at the highest dose. Despite the significance of the linear regression (P<0.05), the representativeness is low, except in Ca. However, in the greenhouse, no clear trend was seen in relation to the dosage (similar to C. annuum), with no significant regressions (P>0.05).

Figure 3: Polyphenol content of L. sativa after application of methanolic and aqueous extracts of U. europaeus obtained from different organs, under laboratory and greenhouse conditions. r² = linear regression analysis 

For L. sativa, the differences in the response for each dose and extract, in laboratory and in greenhouse, are striking. This phenomenon could be caused by the differences in the average temperature in the greenhouse which could be an additional stress factor for the species, thus acting as a co-variable. The temperature average was 25°C for the day and 16ºC for the night in test period, with a 9 ºC of thermal amplitude. At respect, ^{Ancillotti et al. (2015}) point out that heat shock in Nicotiana langsdorffii Weinmann alter the secondary metabolism tending to raise the polyphenol content. In this sense, ^{Helyes et al. (2015}) found that polyphenol content rise in Solanum melongena under heat stress. Similar observations by ^{Chua et al., (2015}) in Premna serratifolia L. confirm that polyphenol content in harvest early is lower than later in the same day; the authors affirm that the temperature had a significant effect. On this matter, ^{Rao and Ravishankar (2002}) and ^{Divi et al. (2010}) point out that temperature is indeed a factor which, under certain conditions, could increase secondary metabolites, although the response is heavily dependent on the species. Also, it is necessary to differentiate environment temperature than soil temperature, in the last case ^{Xue et al. (2018}) found more biomass in Alpine meadow higher temperatures, of what can be inferred the absence of stress; that is, greater primary than secondary metabolism and its metabolites. The above is not applicable to Ca in the laboratory, given that a decreasing response is observed as the dose increases. In this case, Ca presents the expected trend when a plant is subjected to high concentrations of phytotoxic agents (that go beyond stress), with a slight stimulation in low doses and a reduction in the polyphenol concentration in higher doses, acting as an elicitor (^{Rao and Ravishankar, 2002}).

In summary, doses of different Ulex europaeus L. extracts, as well as environmental conditions, affect total polyphenol content when applied to both Capsicum annuum L. and Lactuca sativa L. In general, at increasing doses of extracts, the content of these secondary metabolites also rises. This effect is clear for both species in controlled conditions, yet more variable in the presence of other stress factors - such as temperature - in less controlled (greenhouse) conditions. These results are the first known report of a relationship between the dose of the extracts of U. europaeus and the induction of polyphenolic compounds. From this work, future research could focus on the characterization and identification of the specific phytochemical compounds that are beneficial for human health, thus opening up the possibility of cultivating species with high antioxidant contents using natural products.