**1. Introduction**

Neglected and underutilized endogenous plants are an appealing option to increase food variety [1]. The Mediterranean basin has a plenty of wild and semi-domesticated plants that have been exploited by rural communities for centuries and even during food crises, thus earning the name of "famine food" [2]. A declining consumption of noncultivated edible plants is occurring due to social factors and the diffusion of the so-called "Western-style diet" [3,4]. Nonetheless, NUS are receiving increased scientific attention due to nutritional benefits, richness of bioactive components, and suitable micronutrient contents [3,5]. Numerous non-cultivated species are considered weeds in intensive crop systems and therefore, are largely ignored by researchers, farmers, and consumers [2]. These non-commodity plants offer the opportunity to enrich urban-style diets at an affordable cost [2,5]. For instance, several wild and semi-domesticated plants are consumed raw in salads, such as purslane and borage [6]. These halophytes are commonly richer in bioactive compounds than typical salad crops and can provide nutritious food while ensuring a more diversified food basket and a sustainable diet [2].

Purslane (*Portulaca oleracea* L.) is generally seen as a common species among summer crops [7,8] but it was already considered a medicinal and food source by the Ancient Egyptian civilization [9]. In Southern Italy, this herbaceous plant is typically harvested in the wild, although some family farms leave an area for this species to grow, with the harvest typically sold in small markets serving a specific community or area. Purslane is potentially suitable for hydroponics due to the ease of harvest (i.e., stems and leaves are edible), mechanical properties (i.e., the succulent leaves and stems are suitable for motorized harvest), and amenability to the cut-and-come again strategy (i.e., ability to regrow and to produce roots). In floating systems, the cultivation cycle can last from a minimum of 13 days in a summer cycle [10] to approximately 3 months from sowing to the final harvest [7,8]. Purslane is consumed in the Mediterranean basin fresh, cooked or as a dried vegetable [8]. In Southern Italy, it is employed mainly as fresh salad, often with wild rocket. It is rich in molecules with antioxidant potentials such as alfa-tocopherol, beta-carotene, and ascorbic acid [7–9,11]. This species has a high content of proteins, carbohydrates, and minerals (iron, phosphorus, magnesium, calcium, and potassium) [1,7]. On the other hand, mature leaves of purslane frequently contain a high amount of oxalates (between spinach and tea) [12] and nitrates (similar to spinach and celery) [13], which make this species more suitable for occasional consumption. Nonetheless, it has been pointed out that these traits have sufficient intraspecific variability for the selection of improved lines for ready-to-eat products [14].

Borage (*Borrago officinalis* L.) is an annual herb [15–17], probably native to Syria [15], naturalized in the Mediterranean basin and common in Asia Minor, Europe, North Africa, and America [15,18]. In several countries, this herb is cultivated in open-field conditions mainly to extract oil from seeds. Especially in Europe, *B. officinalis* is also grown for culinary and medicinal uses and often harvested in the wild [17,18]. Nonetheless, a recent study indicated the suitability of borage to produce ready-to-eat fresh cut leaves [19]. To our knowledge, information on the cultivation of borage in soilless systems is limited to the study of the effect of salinity on yield and seed characteristics [20]. Borage smells as cucumber, while leaves, stems, and sometimes flowers, are eaten (cooked or raw) in soups and salads, as well as in vegetable and meat dishes [17]. The same authors mentioned that in Northern Spain, borage leaves, petioles, and stems are eaten fresh or moderately fried in salads, whereas in Italy, borage flowers and leaves are eaten in omelets, stews, soups, condiments or pickled and in oil. Borage is rich in fatty acids and is consumed under the belief that it is a treatment for various diseases such as diabetes, arthritis, multiple sclerosis, and eczema [16]. In addition, borage is characterized by tannins, saponins, flavonoids (kaempferol, quercetin, and isorhamnetin), and phenolic acids (p-coumaric, vanilic, chlorogenic, rosmarinic, and caffeic) [17].

In recent years, the production of microgreens to complement that of mature plants has become a trendy market opportunity for novel foods, due to their rich phytochemical content and sensory value [21]. These young leafy greens enhance the human diet by representing not only a different source of bioactive compounds but also by combining vivid colors and tastes. Currently, several horticultural species have been evaluated and exploited to produce microgreens [22].

There is a wide consensus that we need to reverse the abandonment of NUS by changing their reputation (e.g., old-fashioned and associated with the rural lifestyle) and microgreens offer interesting nutritional and social benefits. Producing and selling NUS microgreens to populations who already are accustomed to specific plant species is a practical way to actively encourage people to improve the nutritional value of their diet by harnessing traditional biodiversity. Moreover, microgreens also represent a way to promote NUS in an urban cultural context. The exploitation of the NUS qualities as microgreens is strongly limited by insufficient awareness of their nutritional value. The current study aimed to characterize two underutilized species, purslane and borage, grown as microgreens, with the goal of promoting their value as a novel and sustainable food complement. Specifically, we assessed yield, macro- and micronutrients using an inductively coupled plasma mass spectrometer (ICP-OES), carotenoids by a high-performance liquid chromatographic method with diode-array detection (HPLC-DAD), and polyphenols by ultra-high-performance liquid chromatography coupled to quadrupole orbitrap high-resolution mass spectrometry (UHPLC-Q-Orbitrap HRMS).

### **2. Materials and Methods**

### *2.1. Plant Material and Growth Condition*

Seeds of borage (*Borrago officinalis* L.), also known as starflower, and of purslane (*Portulaca oleracea* L.), also known as duckweed, were obtained from "Pagano Costantino & F.lli S.R.L" (Scafati, Italy) and "Nehme Establishment for Trade & Agriculture" (Batroun, Lebanon), respectively. Sowing density was 40,000 seeds m<sup>−</sup><sup>2</sup> for borage and 80,000 seeds m<sup>−</sup><sup>2</sup> for purslane. Plants were sown and grew in a climatic chamber (KBP-6395F, Termaks, Bergen, Norway) in 204 cm<sup>2</sup> plastic trays filled with 650 cm<sup>3</sup> of peat-based substrate (Special Mixture, Floragard, Vertriebs-GmbH, Oldenburg, Germany). The macronutrient supply of the substrate (electric conductivity (EC): 282 μS cm<sup>−</sup>1; pH: 5.48) is described elsewhere [23]. Fertigation was applied daily using a modified (quarter-strength) Hoagland's nutrient solution (NS) prepared with osmotic water (EC: 100 ± 25 μS cm<sup>−</sup>1). The NS had an EC of 500 ± 50 μS cm<sup>−</sup>1, a pH of 6 ± 0.2, and the following mineral composition: 2.0 mM NO3 −-N, 0.25 mM S, 0.20 mM P, 0.62 mM K, 0.75 mM Ca, 0.17 mM Mg, 0.25 mM NH4 +-N, 20 μM Fe, 9 μM Mg, 0.3 μM Cu, 1.6 μM Zn, 20 μM B, and 0.3 μM Mo. Light was provided by light-emitting diode (LED) panels (K5 XL 750, Kind LED Grow Light, Santa Rosa CA, USA) with a 12 h photoperiod. The photosynthetic photon flux density was 300 ± 15 μmol m<sup>−</sup><sup>2</sup> s<sup>−</sup><sup>1</sup> measured at the tray level. The temperature and relative humidity were set at 24 ◦C day and 18 ◦C night ( ± 2 ◦C) and 65% day and 75% night ( ±5%), respectively. Each species was replicated three times and randomly placed on the shelf of the climate chamber. Trays were relocated daily across the shelf to avoid time-invariant position effects among experimental units.

### *2.2. Sampling and Morphometric Measurements*

At the emergence of the first true leaf, 23 days after sowing (DAS), borage and purslane microgreens were harvested with scissors by cutting at the substrate level (Supplementary Figure S1). Fresh weight (fw) was measured, and yield expressed in kg fw m<sup>−</sup>2. Each replicate was divided into homogeneous sub-samples for the destructive analyses. A pool was immediately frozen and stored at −80 ◦C for the determination of total ascorbic acid and chlorophylls. Another pool was weighed, placed in a forced-air oven (65 ◦C) until constant weight, for dry weight (dw) assessment and the subsequent analysis of the mineral composition. An additional pool was first snap-frozen in liquid nitrogen and then cold-lyophilized (Christ, Alpha 1–4, Osterode, Germany) for the quantification of phenolic compounds, carotenoids (lutein and β-carotene), and antioxidant activities.

### *2.3. Antioxidant Activity Measurements*

The free radical scavenging activity was quantified with a 2,2-diphenyl-1-picrylhydrazyl (DPPH)-based method using a previously described procedure with few modifications [24]. The DPPH radicals were formed by dissolving 4 mg in 10 mL of methanol. Samples were diluted with the same solvent to obtain a DPPH radical working solution (DRWS) with an absorbance of 0.90 ( ±0.02) at 517 nm. A mixture of 1 mL of DRWS and 200 μL of sample was incubated for 10 min at room temperature, and absorbance was spectrophotometrically read at 517 nm. The activity was expressed as TEAC (mmol Trolox equivalents kg−<sup>1</sup> dw of sample).

The ferric reducing antioxidant activity was measured using a FRAP assay [25] with few modifications. Briefly, the FRAP working solution was prepared by mixing 1.25 mL of 10 mM 2,4,6- tripyridyl-striazine (TPTZ) in HCl (40 mM), 1.25 mL of FeCl3 (20 mmol) in water, and 12.5 mL of 0.3 M sodium acetate buffer (pH 3.6). The reaction of the FRAP

solution (2.850 mL) and samples (150 μL) was incubated at room temperature for 4 min and then, absorbance was read at 593 nm. The results were expressed as TEAC (mmol Trolox equivalents kg−<sup>1</sup> dw of sample).

The ABTS-scavenging activity was evaluated according to the previously published procedures with minor modifications [26]. The 2,2-azinobis-(3-ethylbenzothiazoline-6-sulfonate) (ABTS•+) radicals were generated by mixing 5 mL of 7 mM ABTS aqueous stock solution with 88 μL of 2.45 mM aqueous potassium persulfate, diluted with ethanol to a working solution with an absorbance of 0.700 ± 0.002 at 734 nm. Subsequently, 100 μL of sample and 1 mL of the above resulting solution were mixed and incubated for 3 min at room temperature. Absorbance was then read at 734 nm. The results were expressed as TEAC (mmol Trolox equivalents kg−<sup>1</sup> dw of sample).

#### *2.4. Quantification of Chlorophylls, Catotenoids, and Total Ascorbic Acid*

Chlorophylls were spectrophotometrically quantified according to a previously published procedure [27]. Briefly, samples were weighed, and pigments extracted in 90% acetone. Aliquots were read at 662 and 645 nm using a Hach DR 4000 spectrophotometer (Hach Co., Loveland, CO, USA). Total chlorophyll was estimated as the sum of chlorophyll a and b and expressed in mg kg−<sup>1</sup> fw. Total ascorbic acid was determined as previously described [28] and expressed as mg ascorbate equivalents 100 g<sup>−</sup><sup>1</sup> fw.

Carotenoids (β-carotene and lutein) were quantified by HPLC-DAD essentially as reported [29]. The apparatus comprised a 1200 Series quaternary pump and a 1260 Diode Array Detector Separation (Agilent Technologies, Santa Clara, CA, USA), equipped with Gemini C18 (Phenomenex, Torrance, CA, USA) reverse phase columns (250 × 4.6 mm, 5 μm). Calibration curves were built with using β-carotene and lutein commercial standards (Sigma-Aldrich, Milan, Italy) in the 5 to 100 μg mL−<sup>1</sup> range. Results were expressed in μg g<sup>−</sup><sup>1</sup> dw.

### *2.5. Analysis of Macro- and Micro-Minerals by ICP-OES*

Minerals (P, K, Ca, Mg, Na, Mn, Fe, Zn, Cu, Se, B, Cr, Mo, Ni, Al, Ba, Cd, and Pb) were quantified by inductively coupled plasma-optical emission spectrometry (Spectroblue, Spectro Ametek, Berwyn, PA, USA) [30]. Briefly, 1 g of dried plant tissue was processed by microwave-assisted digestion (MLS-1200, Microwave Laboratory Systems, Milestone, Shelton, CT, USA) in 10 mL of a 3:1 (*v/v*) solution of nitric acid and fuming hydrochloric acid. The slurry was brought to a final volume of 50 mL with ultra-pure water (Merck Millipore, Darmstadt, Germany). For non-alkaline elements (Fe, Mn, Mo, Se, and Zn), the calibration curve was built in the 1.0 to 100 μg L−<sup>1</sup> interval and the quantity of the minerals expressed in μg g<sup>−</sup><sup>1</sup> dw. For alkaline elements (P, K, Ca, Mg, and Na), the calibration curve was built in the 100 μg L−<sup>1</sup> to 10 mg L−<sup>1</sup> range and the quantity of the minerals expressed in mg g<sup>−</sup><sup>1</sup> dw. For the determination of the accuracy, we used standard reference material (BCR CRM 142R-Commission of the European Communities, 1994). The recovery range was in the 86% to 98% interval.

### *2.6. Analysis of Polyphenols by UHPLC-Q-Orbitrap HRMS*

Lyophilized plant tissue (100 mg) was extracted using 5 mL of a methanol/water (60:40, *v/v*) solution by sonication for half an hour. The mixture was centrifuged (4000 rpm, 15 min), the supernatant filtered through Whatman paper, and then aliquots (10 μL) were used for anthocyanins and polyphenols quantification with a UHPLC system (Dionex UltiMate 3000, Thermo Fisher Scientific, Waltham, MA, USA) coupled to a Q-Exactive Orbitrap mass spectrometer (UHPLC, Thermo Fisher Scientific, Waltham, MA, USA) essentially as described [29,31]. Chromatographic separation was carried out in a Luna Omega PS 1.6 μm column (50 × 2.1 mm, Phenomenex, Torrance, CA, USA) and identification was performed with a Q-Exactive Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) operating in fast negative/positive ion switching mode. Two scan events (full scan MS and all ion fragmentation, AIF) were

set for the compounds. Data processing was carried out with the Xcalibur software 3 (Xcalibur, Thermo Fisher Scientific, Waltham, MA, USA). Polyphenols were expressed in μg g<sup>−</sup><sup>1</sup> dw. The individual phenolic compounds were identified and quantified by comparison with the available standards as described [29,31].

### *2.7. Statistical Analysis*

Morphometric measurements were independently carried out on the three replications. All instrumental determinations for each of the three biological replications were performed in two technical replicates. For all variables, the equality of the means between the two species was evaluated with an unpaired two-tailed Student's *t*-test.
