Dwarf Pomegranate (Punica granatum L. var. nana): Source of 5-HMF and Bioactive Compounds with Applications in the Protection of Woody Crops

While the properties of edible pomegranate varieties have been widely explored, there is little information on ornamental types. In this study, possible alternatives for the valorization of dwarf pomegranate fruits have been explored. The characterization of their hydromethanolic extract by gas chromatography−mass spectrometry evidenced the presence of high contents of 5-hydroxymethylfurfural (a carbon-neutral feedstock for the production of fuels and other chemicals) and β- and γ-sitosterol stereoisomers. The microbicidal activity of the crude extract, both alone and in a conjugate complex with chitosan oligomers (COS), was investigated against three plant pathogenic microorganisms that cause significant losses in woody crops: Erwinia amylovora, E. vitivora, and Diplodia seriata. In in vitro assays, a strong synergistic behavior was found after conjugation of the bioactive constituents of the fruit extract with COS, resulting in minimum inhibitory concentration (MIC) values of 750 and 375 μg·mL−1 against E. amylovora and E. vitivora, respectively, and an EC90 value of 993 μg·mL−1 against D. seriata. Hence, extracts from the non-edible fruits of this Punicaceae may hold promise as a source of high value-added phytochemicals or as environmentally friendly agrochemicals.


Introduction
The pomegranate tree (Punica granatum L.), of the Lythraceae family, is a shrub native to the Middle East and the Mediterranean but is widely cultivated in warm-temperate regions around the world. The fruit from the edible varieties is the well-known pomegranate. A high genetic diversity of morphological and biochemical traits characterizes this species, with more than 500 cultivars identified worldwide [1].
The dwarf pomegranate (usually referred to as P. granatum var. nana, but also as Punica granatum 'Nana', Punica granatum nana, P. granatum var. nana Pers., and Punica nana L.) is a natural variant of the species. Even though it has sometimes been treated as the third species of Punica [2], in the work by Currò et al. [3] on microsatellite loci for pomegranate, it was evidenced that it did not show unique allele patterns.
This ornamental variety differs from the edible variety in its much smaller size, its lustrous, lanceolate leaves (ca. 2.5 cm long), and its small (ca. 5 cm) fruits. However, the

Elementary Analysis
The percentages of C, H, and N of P. granatum var. nana fruits (in dry matter wt%) are summarized in Table 1. The moisture content was 5.67%, and structural water (recorded by thermal analysis) was 10.79%.
The higher heating value (HHV) derived from the elemental analysis data was 17.1 kJ·g −1 . Table 1. Elemental analysis of P. granatum L. var. nana fruits (after drying and grinding). Values are presented as the mean of four replicates, followed by minimum and maximum values in parentheses.

Thermal Analysis
Thermogravimetry (TG), derivative thermogravimetry (DTG) and differential scanning calorimetry (DSC) curves for P. granatum var. nana fruits are shown in Figure S1. The exothermal effects that occur at 325 and 445 • C are due to holocellulose and lignin decomposition. The ash content at 600 • C was 4.4%.

Vibrational Characterization
The main infrared absorption bands in the spectrum of P. granatum L. var. nana fruits (dried and ground), together with their assignments, are shown in Table 2. The most prominent ones appeared at 1730, 1444, 1150, and 1018 cm −1 . The peak at 1730 cm −1 is attributed to unconjugated carbonyl stretching vibration in hemicelluloses (C−O stretching vibration in carboxyl, carbonyl, and acetyl groups); the peak at 1444 cm −1 to asymmetric C−H deformations; the peak at 1150 cm −1 to C−O−C vibrations in cellulose; and the peak at 1018 cm −1 to pectins or carotenes.

Antimicrobial Activity
The activity of the hydromethanolic extract of P. granatum var. nana fruits and its three main constituents was tested alone and after conjugation with chitosan oligomers (COS). The results against the two quarantine bacterial plant pathogens are summarized in Table 4. It can be seen that the activity of the extract was comparable to that of COS alone, β-sitosterol, and 5-HMF, reaching total inhibition of the two Erwinia spp. at 1500 μg·mL −1 (whereas DDMP led to full inhibition at half that dose). However, after conjugation with COS, a marked potentiation of the antibacterial action of the extracts was found, with minimum inhibitory concentration (MIC) values of 375 and 250 μg·mL −1 against E.

Antimicrobial Activity
The activity of the hydromethanolic extract of P. granatum var. nana fruits and its three main constituents was tested alone and after conjugation with chitosan oligomers (COS). The results against the two quarantine bacterial plant pathogens are summarized in Table 4. It can be seen that the activity of the extract was comparable to that of COS alone, β-sitosterol, and 5-HMF, reaching total inhibition of the two Erwinia spp. at 1500 µg·mL −1 (whereas DDMP led to full inhibition at half that dose). However, after conjugation with COS, a marked potentiation of the antibacterial action of the extracts was found, with minimum inhibitory concentration (MIC) values of 375 and 250 µg·mL −1 against E. amylovora and E. vitivora, respectively, similar to those recorded for COS−DDMP and better than those found for COS−β-sitosterol and COS−5-HMF.
As for the antifungal behavior, tested against D. seriata ( Figure 2 and Figure S4), the activity of the pure extracts was lower than that of chitosan, 5-HMF, and DDMP (for which total inhibition was achieved at 1500, 1000, and 750 µg·mL −1 , respectively) and much lower than that of β-sitosterol (for which mycelial growth was completely inhibited at 250 µg·mL −1 ). The activity again increased after conjugation with COS, resulting in total inhibition at 1000 µg·mL −1 . This inhibitory concentration was lower than that recorded for COS alone, which points to a synergistic behavior (see Tables 5 and 6 in which the effective concentrations and synergistic factors are summarized, respectively). However, the best results corresponded to the COS−β-sitosterol conjugate complex, for which total inhibition was achieved at 187.5 µg·mL −1 , followed by COS−DDMP and COS−5-HMF.

Thermal Analysis
The ash content at 600 • C for fruits of the nana variety (4.4%) was intermediate between those reported for seeds and the peel of other pomegranate varieties/cultivars (with ash contents as low as 1.5% for the seeds [26,27] and 6.8% for the peel [28]).

Vibrational Characterization
The bands present in the freeze-dried hydromethanolic extract were consistent with the presence of the main constituents identified by GC−MS: the band at 2849 cm −1 can be assigned to 5-hydroxymethylfurfural (C−H vibration of the aldehyde group), that at 1048 cm −1 (C−O strain) to β-sitosterol, and those at 1360, 1311, 1246, 1184, 1065 and 702 cm −1 to pyrogallol (with a 10 cm −1 shift) [31].

Phytoconstituents Identified by GC−MS
Taking into consideration that the hydromethanolic extraction mixture also solubilizes polar compounds (non-volatile) that cannot be detected by GC−MS without previous derivatization of the extract, a word of caution concerning the results seems necessary. It should be clarified that, in the work presented herein, such prior derivatization was not conducted because it has a number of drawbacks: it makes procedural preparation steps longer and costlier (which would have a negative impact on the economic viability of the crop protection treatments), the data acquisition process becomes more complex and longer because derivatization can sometimes lead to impurities, notwithstanding the uncertainty of conversion of compounds into derivatives and the use of toxic reagents [32]. On the other hand, the injection of non-volatile compounds may result in eventual damage to the GC capillary column.
Concerning the high concentration of 5-HMF in the extract (37% peak area), it was initially ascribed to a failure in the extraction operating conditions (unexpected rise of the temperature), but the bibliographic reports of a 21% (peak area) content in an ethanolic extract by Kumar and Vijayalakshmi [33], a 32.1% (peak area) content by Hamad et al. [11], and a 39.7% content by Mohamad and Khalil [35] encouraged us to repeat the extraction and determination, verifying that the result was consistent. This result is also in agreement with a study by Fischer et al. [36], who explored the thermal impact on the anthocyanin and phenolic co-pigments of pomegranate juices and on the formation of 5-HMF and other degradation products and concluded that upon forced heating at 90 • C for five hours, the 5-HMF contents increased only slightly. The finding of such a high 5-HMF content is important because it is a chemical used as a flavoring agent in the food industry and industrially as a carbon-neutral feedstock for the production of fuels and other chemicals [37,38]. Further, it has antibacterial properties [39]: for instance, Kaur et al. [40] assayed the antibacterial activity of 5-HMF, and its derivatives from pomegranate fruits were tested for their antimicrobial potential against Klebsiella sp., finding MICs in the 40-160 µg·mL −1 range.

Comparison of the Microbicidal Activity of the Extract
Reports on the antimicrobial activity of P. granatum var. nana are scarce in the literature: the antibacterial activity of dwarf pomegranate peel extracts (in water, aqueous ethanol, and aqueous ethanol−methanol) was tested against Salmonella enterica by Wafa et al. [19], with moderate results, and the hydromethanolic extract of P. granatum var. nana leaves was tested against S. rolfsii, R. solani, and F. oxysporum f.sp. lycopersici by Emam et al. [18]. To the best of the authors' knowledge, no studies on the activity of fruit extracts have been published to date.
Conversely, the microbicidal activity of edible pomegranate varieties has received more attention and has been recently addressed in review articles by Singh et al. [6], Pirzadeh et al. [7], and Chen et al. [8]. Since no reports are available against the plant pathogens studied here, results against other Erwinia spp., Xanthomonas spp. (since E. vitivora is also known as Xanthomonas ampelina), and Botryosphaeriaceae are presented for comparison purposes.
Hassan et al. [45] investigated the antibacterial activity of ethanolic extracts of the peel of 32 Egyptian pomegranate cultivars against E. carotovora and X. campestris, finding diameter zone of inhibition (DZI) values in the 10 to 37.5 mm range and in the 12.8 to 36.6 mm range, respectively, depending on the P. granatum cultivar. Mhaisgawali et al. [46] assessed the antibacterial activity of pomegranate bark against E. chrysanthemi and X. malvacearum, finding efficacies in terms of DZI values similar to those of streptomycin. Truchado et al. [47] reported that pomegranate extract at 20 µg·mL −1 caused significant inhibition of E. carotovora quorum-sensing signals, and Vlachou et al. [48] found that pomegranate extracts showed good control of E. carotovora pv. atroseptica and Xanthomonas spp. at a concentration of 1 mg/disc. Hussein et al. [49] tested five plant extracts in dimethyl sulfoxide, including P. granatum peel, against E. amylovora, concluding that the inhibition was greater than that of black cumin and thyme. However, none of these studies reported MIC values. The only study in which such information was available was the one by Khaleel et al. [50], who tested the efficacy of the ethyl acetate extract of P. granatum peel against E. carotovorum and X. gardneri, finding MIC values of 3.125 and 6.25 mg·mL −1 , respectively. In comparison, the hydromethanolic extracts of P. granatum var. nana fruits tested here were 2 to 4 times more effective (MIC = 1.5 mg·mL −1 ). Nonetheless, in view of the MIC values presented in Table S1, their effectiveness is still much lower than that of conventional antibiotics (for which MIC values less than 2 µg·mL −1 were obtained against E. amylovora and E. vitivora). However, it should be clarified that presently there are no antibiotics authorized as plant protection products in the EU. Some EU member states authorize their emergency use to control outbreaks, but the volumes used are negligible, and their application is strictly controlled [51].
Regarding the antifungal activity against Botryosphaeriaceae fungi, Matos et al. [52] observed no inhibition of Lasiodiplodia theobromae for pomegranate oil at concentrations of up to 0.6%, but Lorenzetti et al. [53] reported a significant inhibition (87%) of Diplodia macrospora at a concentration of 7.4% (i.e., at 74,000 µg·mL −1 ) of pomegranate peel extract. In comparison, the non-conjugated dwarf pomegranate fruit extract, for which an EC 90 of 4600 µg·mL −1 was estimated, should be much more effective (by a factor of 16).

On the Synergistic Behavior after Conjugation with Chitosan Oligomers
Although they do not refer to conjugated complexes, but to compounds, there are previous reports in the literature of improved antimicrobial activity by adding pomegranate extracts to chitosan coatings: for example, incorporation of pomegranate peel extracts in water and methanol into chitosan resulted in a significant reduction of the fungal growth and activity of Penicillium spp. [54,55], finding that chitosan enhanced the action of the plant extracts to inhibit pathogenic fungi. The incorporation of peel extracts into chitosan coatings also inhibited microbial growth, prolonged the shelf life, and maintained the sensory scores of pepper [56].
An explanation of the mechanism of action behind the enhanced antifungal effect in comparison with the use of individual natural agents has not been reported. According to Tayel et al. [55], the synergism can be tentatively explained by the various fungicidal components of each agent applied and by the fact that fungal pathogens are not readily resistant to multiple fungitoxicants. An alternative explanation would be based on enhanced solubility and bioavailability as a result of increased binding to specific negatively charged binding receptors on bacterial and fungal membranes.

Plant Material
The 10-year-old specimens of P. granatum var. nana under study were purchased at one of the flower market stalls in the Plaza Bib-Rambla (also named 'Plaza de las Flores') in Granada (Spain). After one year in a pot, they were planted in Rales (Asturias, Spain). A voucher specimen, identified and authenticated by Dr. P. Pablo Ferrer-Gallego, was deposited at the VAL (Herbarium of the Botanical Garden of the University of Valencia, Valencia, Spain), code VAL244427.
In 2020, 15 individuals were harvested to obtain fruit composite samples for analysis. The samples were shade dried and pulverized to a fine powder in a mechanical grinder. The moisture content was measured by weight loss after heating at 105 • C until a constant weight was reached.

Preparation of the Fruis Extract
The P. granatum var. nana fruit sample was mixed (1:20, w/v) with a methanol/water solution (1:1 v/v). The solution was heated in a water bath at 50 • C for 30 min and sonicated for 5 min (2.5 min−1 min stop−2.5 min) using a 1000 W, 20 kHz probe-type ultrasonicator (model UIP1000hdT; Hielscher Ultrasonics, Teltow, Germany). After centrifugation at 9000 rpm for 15 min, the supernatant was filtered using Whatman No. 1 paper. Aliquots of the final solution were freeze-dried for infrared vibrational analysis.

Plant Biomass and Fruit Extract Physicochemical Characterization
Elemental analysis (CHNS) of the fruits (once dried and ground) was carried out with a LECO CHNS-932 apparatus (St. Joseph, MI, USA). Calorific values were calculated from elemental analysis data by using the following equation: HHV = (0.341 × %C) + (1.322 × %H) − 0.12(%O + %N), where HHV is the higher heating value, i.e., the calorific value of the dry material expressed in kJ·g −1 , and %C, %H, %O, and %N represent the mass fractions (in wt% of dry material) [59].
Thermogravimetry and differential scanning calorimetry analyses were performed using a simultaneous TG-DSC2 (Mettler Toledo; Columbus, OH, USA) in N 2 :O 2 atmosphere (4:1) with a heating ramp of 20 • C·min −1 . According to the usual pyrolysis temperature conditions in oxygen bomb calorimeters, the ash content was estimated from the residue obtained upon heating up to 600 • C [60].
The infrared vibrational spectra were recorded with a Nicolet iS50 Fourier-transform infrared spectrometer (Thermo Scientific; Waltham, MA, USA) equipped with an attenuated total reflection (ATR) system. Spectra were collected with a resolution of 1 cm −1 in the 400-4000 cm −1 range, taking the interferograms that resulted from the co-addition of 64 scans. The spectra were processed using the advanced ATR correction algorithm [61] available in OMNIC TM software.
The hydroalcoholic fruit extract was studied by gas GC−MS at the STI facilities at the University of Alicante (Alicante, Spain), using a 7890A gas chromatograph coupled to a 5975C quadrupole mass spectrometer (both from Agilent Technologies; Santa Clara, CA, USA). The chromatographic conditions were: 3 injections/vial, 1 µL injection volume; 280 • C injector temperature, in splitless mode; 60 • C initial oven temperature, kept for 2 min, followed by a heating ramp of 10 • C·min −1 up to a final temperature of 300 • C, kept for 15 min. The chromatographic column used for the separation of the compounds was an HP-5MS UI (Agilent Technologies), with length = 30 m, diameter = 0.25 mm, and film thickness = 0.25 µm. Concerning the mass spectrometer conditions, the temperature of the electron impact source of the mass spectrometer was 230 • C, and that of the quadrupole was 150 • C; with a 70 eV ionization energy. For equipment calibration, test mixture 2 for apolar capillary columns according to Grob (Supelco 86501) and PFTBA tuning standards were used. The identification of the extract constituents was based on a comparison of their mass spectra and retention times with those of the authentic compounds and by computer matching with the National Institute of Standards and Techniques (NIST11) database and the monograph by Adams [62].

In Vitro Antimicrobial Activity Assessment
CLSI standard M07-11 [63] was followed for the determination of the antibacterial activity. An isolated colony of E. amylovora was first incubated at 30 • C for 18 h in TSB liquid medium. Starting from a 10 8 colony forming units (CFU)·mL −1 concentration, serial dilutions were conducted to obtain a final inoculum concentration of~10 4 CFU·mL −1 . The bacterial suspensions were then delivered to the surface of TSA plates, to which the products under investigation had previously been added at concentrations ranging from 62.5 to 1500 µg·mL −1 . The incubation of the plates was carried out at 30 • C for 24 h. The same procedure was followed for E. vitivora, albeit at a temperature of 26 • C. Readings were taken after 24 h. The MIC values were determined as the lowest concentrations at which no bacterial growth was observed. The experiments were run in triplicate, with each replicate consisting of three plates per treatment and concentration combination.
To compare the effectiveness of the natural compounds with that of standard antibiotics, ETEST ® gradient MIC strips (bioMérieux Clinical Diagnostics, Marcy-l'Étoile, France) were chosen. Well-isolated colonies from agar plates were suspended in 0.85% NaCl suspension medium to obtain an inoculum turbidity of 0.5 McFarland; a sterile swab was soaked in the inoculum and, after removing excess fluid, it was used to carefully streak the entire surface of the Mueller-Hinton agar plates three times (rotating the plate 60 • each time to evenly distribute the inoculum), allowing a drying time of 20 min; ETEST ® strips (with amikacin, benzylpenicillin, gentamicin, and tetracycline) were then positioned on the plates, using one strip per plate; plates were incubated overnight in an inverted position (lid down), and the MICs were read where the edge of the inhibition ellipse intersected the side of the strip.
As for the antifungal activity, it was assessed following EUCAST standard antifungal susceptibility testing procedures [64]. The agar dilution method was used in which aliquots of stock solutions were incorporated onto the PDA medium to obtain concentrations in the 62.5−1500 µg·mL −1 interval. Five-millimeter mycelial plugs from the margin of 1-week-old PDA cultures of D. seriata were transferred to plates that incorporated the aforementioned concentrations of each product (with three plates per treatment and concentration combination and two replicates). Incubation was conducted at 25 • C in the dark for seven days. PDA medium without any amendment was used as the control. Mycelial growth inhibition was calculated as ((d c − d t )/d c ) × 100, where d c and d t are the average diameters of the fungal colony of the control and of the treated fungal colony, respectively. The 50% and 90% effective concentrations (EC 50 and EC 90 , respectively) were calculated through PROBIT analysis in IBM SPSS Statistics v.25 (IBM; Armonk, NY, USA). Wadley's method [65] was chosen to determine the level of interaction.

Statistical Analysis
After checking that homogeneity and homoscedasticity requirements were met (using Shapiro-Wilk and Levene tests, respectively), the statistical analysis of the mycelial growth inhibition results for D. seriata was carried out by analysis of variance (ANOVA), followed by post hoc comparison of means through Tukey's test (p < 0.05), using IBM SPSS Statistics v.25.

Conclusions
Although the ash content of the inedible fruits of P. granatum var. nana impedes their use as biofuel, the finding by GC−MS in the hydroalcoholic extract of significant amounts (37%) of 5-hydroxymethylfurfural (key for the production of fuels and other chemicals), validates their usefulness as a promising energy feedstock. Moreover, due to their content of β-sitosterol (and other phytochemicals such as 2,3-dihydro-3,5-dihydroxy-6-methyl-(4H)-pyran-4-one), fruits of P. granatum var. nana can also be used as a source of agrochemicals for the control of fire blight and 'maladie d'Oléron', although conjugation with COS is required to promote the efficacy of the extracts, resulting in MIC values as low as 375 and 250 µg·mL −1 against E. amylovora and E. vitivora, respectively. The antifungal behavior of the conjugated complexes, tested against D. seriata, was moderate (EC 90 = 993 µg·mL −1 ), but suggests that they may have a broader spectrum of antibiotic activity.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

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 their relevance to an ongoing Ph.D. thesis.