Chemical Composition of Essential Oils of Bulbs and Aerial Parts of Two Cultivars of Allium sativum and Their Antibiofilm Activity against Food and Nosocomial Pathogens

This work aimed to evaluate the chemical composition of the essential oils (EOs) of two cultivars of Allium sativum and their antibiofilm activity against the food pathogens Acinetobacter baumannii, Escherichia coli, Listeria monocytogenes, and Staphylococcus aureus. The crystal violet assay ascertained the susceptibility of the bacterial biofilms, while the MTT assay let to evaluations of the metabolic changes occurring in the bacterial cells within biofilms. Their chemical composition indicated some sulfuric compounds (i.e., allicin, diallyl disulfide, and allyl propyl disulfide), and decene as some of the main components of the EOs. The aerial parts and bulbs’ EOs from the two cultivars showed chemical differences, which seemed to affect the antibiofilm activity. The EOs from aerial parts of ‘Bianco del Veneto’ inhibited the biofilm formation of L. monocytogenes and E. coli (60.55% and 40.33%, respectively). In comparison, the ‘Staravec’ EO inhibited the cellular metabolism of E. coli (62.44%) and S. aureus (51.52%) sessile cells. These results indicate their possible use as preserving agents in the food industry and suggest their potential exploitation in the development of new formulations to avoid or limit nosocomial infections.


Introduction
Garlic (Allium sativum L.) is one of the most important species used since ancient times as a traditional medicine. Garlic is an annual herbaceous aromatic spice, a member of the Amaryllidaceae family, native to the countries of Central Asia; it was probably one of the first plants to be cultivated and used as food [1,2]. The binomial Allium sativum is a name with both Celtic and Latin origins. From the Celtic side, the term 'allium' means burning or pungent, while the word 'sativum' comes from the Latin side, which means planted or cultivated. On the other hand, the term garlic has Anglo-Saxon origins, with 'gar-leac' indicating its flowering stem [3]. There are numerous varieties of garlic, also classified according to the bulb's color and characterized by different compositions and biological activities [4,5]. Garlic is widely used as a flavoring agent. It possesses various therapeutic properties, and is used in the treatment of earache, whooping cough, colds, and stomach disorders, and in the prevention of cardiovascular diseases [6]. These activities have been confirmed by several clinical and epidemiological trials [7]. Garlic properties derive mainly from organosulfides (allicin, diallyl disulfide, and diallyl trisulfide) and flavonoids on the species, and within the species, on the cultivar. It, in turn, also contributes to influencing the biological properties, such as the antibacterial and anti-biofilm properties, that a specific plant and its derivatives (extracts, EOs) can then exert. For these reasons, we studied the chemical of the EOs of aerial parts and bulbs of two cultivars of A. sativum, the Italian 'Bianco del Veneto', and the Albanian 'Staravec', which have never been studied before, and their biofilm inhibitory capacity against A. baumannii, E. coli, L. monocytogenes, and S. aureus. In addition, we analyzed their effect on the bacterial cell metabolism within biofilm, which concurs to increase their virulence.
The presence of numerous sulfur compounds agrees with the available literature [32][33][34][35][36][37]. The amounts of allicin appeared very variable between the samples; in fact, it is present in minimal quantities in the aerial parts of the 'Staravec'. The other samples represent the main component, with percentages ranging between 17.5% and 62.2%. The presence of sulfur compounds related to diallylsulfide, the main constituents in the EO of the aerial parts of 'Staravec', agrees with the available literature. The garlic EO studied by Casella and coworkers [8] is poor in allicin but rich in diallyl disulfide, diallyl trisulfide, and diallyl tetrasulfide. Furthermore, these sulfur compounds influence the biological properties of the samples, including their antibacterial activity, as reported by Shang and coworkers [38].

Antibiofilm Activity Biofilm Inhibitory Capability of the EOs
The capacity of the EOs to inhibit bacterial biofilm formation and the metabolism of the bacterial cells within the biofilm was assessed through crystal violet and MTT tests, respectively. The tests were performed utilizing 10 µL/mL and 20 µL/mL of EOs, two concentrations abundantly lower than the minimal inhibitory concentration, calculated through the resazurin test ( Table 2). The data indicated that the bulbs of 'Bianco del Veneto' exhibited inhibitory activity on biofilm formation (Table 3), with inhibition percentages ranging from 18.59% (10 µL/mL vs. A. baumannii) up to 63% (64.29% with 20 µL/mL vs. L. monocytogenes and 63.18% vs. S. aureus). The 'Bianco del Veneto' EO aerial parts were effective against all four strains, mainly when used at 20 µL/mL, showing an efficacy, especially against E. coli and L. monocytogenes, against which its inhibitory action reached percentages of 60.55% and 50.52%, respectively. Concurrently, the 'Bianco del Veneto' bulbs' EO was always active, with an inhibitory action ranging between 48.90% (against E.coli) up to 64.29% (against L. monocytogenes), when used at 20 µL/mL. The EOs from the aerial parts of 'Staravec' showed less inhibitory activity, resulting in effectiveness against A. baumannii (45.61% inhibition), E. coli (27.56% inhibition), and S. aureus (26.31% inhibition), but it was utterly ineffective vs. L. monocytogenes. The EOs of the bulbs, on the other hand, proved to be extremely ineffective in inhibiting the formation of biofilm by all four microbial strains. Table 2. Minimal inhibitory concentration (µL/mL) of the EOs from the A. sativum cultivars necessary to inhibit the growth of the pathogenic bacterial strains Listeria monocytogenes, Escherichia coli, Acinetobacter baumannii, Staphylococcus aureus.
The experiments were performed in triplicate and reported as the mean (±SD). b : p < 0.01; compared with the tetracycline used as the control (ANOVA followed by Dunnett's multiple comparison test).
The correlation analysis, considering the chemical composition of the EOs and their inhibitory behavior exhibited against the bacterial biofilm formation, highlighted a different influence of some among the more abundant compounds. Thus, only diallyl sulfide, present in the EO of the aerial parts of the 'Bianco del Veneto', seemed to have some inhibitory activity vs. A. baumannii (Corr-Coeff = 0.142). Considering the 'Bianco del Veneto' bulbs' EO, where the resulting concentration of diallyl sulfide was equal to 27.9%, such actions arrived up to a concentration of 57.34% using 20 µL/mL of the same oil. The diallyl sulfide, therefore, seemed to counteract the negative effect on the inhibitory biofilm activity exhibited by the decene (Corr-Coeff = −0.19), allyl propyl disulfide (Corr-Coeff = −0.01), and allicin (Corr-Coeff = −0.377). As evidence of the negative effect exerted by allicin, we could highlight how the EO obtained from the 'Staravec' bulb's EO-containing 62.2% of allicin-was unable to exert any inhibitory biofilm effect on all four bacteria used in the tests. The inhibitory biofilm action exerted by the 'Bianco del Veneto'aerial parts' EOs against E. coli seemed to depend on the decene (Corr-Coeff = 0.488) and allyl propyl disulfide (Corr-Coeff = 0.42). In the case of the 'Bianco del Veneto' bulbs' EO, the good inhibitory biofilm action against E.coli seemed to be influenced not so much by the decene (absent in this EO) but rather by the diallyl sulfide. As we have seen previously, the EOs obtained from the 'Staravec' cultivar were much less effective in exhibiting inhibitory biofilm action. The correlation analysis indicated that allyl propyl disulfide affected the inhibitory biofilm formation shown by the aerial parts' EO (whose activity did not go beyond 27.56% with 20 µL/mL). Moreover, as we previously indicated, the EO of the 'Staravec' bulb, where allicin (Corr-Coeff = −0.383) is present at 62.2%, was practically ineffective vs. E. coli. The correlation coefficients were all positive in the case of L. monocytogenes. However, as shown in Table 3, the EOs extracted from the 'Staravec' proved utterly ineffective. Diallyl sulfide (Corr-Coeff = 0.493) was present at a negligible percentage (0.5%); the ineffectiveness of the inhibitory biofilm action exhibited by the 'Staravec' bulbs' EO seemed, in this case, to be influenced by the extremely high percentage of allicin (Corr-Coeff = 0.075). The tests were performed using 10 µL/mL and 20 µL/mL. All tests were performed in triplicate. Results are expressed as percentages (mean ± SD) and calculated assuming the control (untreated bacteria, for which we assumed an inhibitory value = zero). a : p < 0.1; compared with the control (ANOVA followed by Dunnett's multiple comparison test).
From the analysis of the correlation, the inhibitory biofilm action (up to 63.18%, see Table 3) exhibited by the EO of the bulb of 'Bianco del Veneto' against S.aureus, seemed to be influenced not so much by allicin, although it was abundant (Corr-Coeff = 0.018), but from the diallyl sulfide (Corr-Coeff = 0.532). Conversely, the positive influence of diallyl sulfide present in the aerial parts' EO of such a cultivar seemed to be somewhat overwhelmed by the presence of allicin, but above all by the presence of allyl propyl disulfide (Corr-Coeff = −0.305), which appeared to weakly inhibit the S. aureus's biofilm formation.
The available literature demonstrated that diallyl sulfides exhibited promising antimicrobial activity [37,[39][40][41][42]. Our results confirmed that, although allicin contributes to the antibacterial properties of hydroalcoholic extracts of garlic [43,44], it could exhibit a breaking effect on the biofilm inhibitory action of the A. sativum EOs.
Through the test with MTT, we observed that the action of EOs on the formation of bacterial biofilms did not always coincide with a similar effect on the metabolism of bacterial cells (Table 4). The tests were performed using 10 µL/mL and 20 µL/mL. All tests were performed in triplicate. Results are expressed as percentages (mean ± SD) and calculated assuming the control (untreated bacteria, for which we assumed an inhibitory value = zero). a : p < 0.1; compared with the control (ANOVA followed by Dunnett's multiple comparison test).
The action of the EOs on the metabolism of L. monocytogenes appeared opposite to what we saw in the test with crystal violet: in fact, the 'Bianco del Veneto' EOs, very effective in limiting the adhesion process of the biofilm by this bacterium, were utterly ineffective in affecting its metabolism. On the other hand, the 'Staravec' EOs confirmed their almost entirely ineffectiveness (apart from a 10.38% metabolic inhibition in the presence of 20 µL/mL of the EO obtained from the aerial parts). In this case, the correlation analysis highlighted a robust negative effect linked to the presence of allicin (Corr-Coeff = −0.748) and diallyl sulfide (Corr-Coeff = −0.724). The weak albeit inhibitory activity (10.38%) observed in the presence of 20 µL/ mL of the EO obtained from the aerial parts of 'Staravec' could be due to the presence of allyl propyl disulfide (Corr-Coeff = 0.284). The EOs' components might be able to act against its biofilm formation not so much by working on L. monocytogenes metabolism but with other mechanisms, as has been reported [45].
Against E. coli, an inhibitory effect on biofilm formation (50.52% with the EOs from the aerial parts of 'Bianco del Veneto') did not correspond to an identical inhibitory force on the metabolism of these bacterial cells, which was utterly absent. On the contrary, 'Starevec' EOs proved to be very efficient at affecting its metabolism, so much so that we observed percentages of metabolic inhibition never lower than 62.64% and 65.71% with 20 µL/mL of the bulb and aerial parts EOs, respectively.
The EOs, when used at the highest concentration, always exerted an inhibitory force on the bacterial metabolism of A. baumannii, being that the 'Staravec' EOs are more effective than the 'Bianco del Veneto' ones, especially in regards to the EO recovered from the bulbs (59.79% and 24.10%, respectively, when we tested 20 µL/mL). Correlation analysis allowed us to observe that allyl propyl disulfide (Corr-Coeff = 0.244) and allyl propyl disulfide (Corr-Coeff = 0.16) seemed to mitigate, in a certain sense, the negative effect of allicin (Corr-Coeff = −0.178) on the inhibitory metabolic capacity exerted by the 'Bianco del Veneto' aerial parts' EOs vs. A. baumannii. The lack of decene (Corr-Coeff = 0.553) in the EO of the 'Bianco del Veneto' bulb made it relatively weak in inhibiting the A.baumanni metabolism. Therefore, we could hypothesize that the greater inhibitory metabolic capacity exerted by the EO of the aerial parts of 'Staravec' vs. A. baumannii could be due to the almost total absence of allicin. On the other hand, the bulb EO of 'Staravec' showed a greater propensity to inhibit the metabolism of A. baumannii, probably due to allyl propyl disulfide and decene, expressing a positive correlation coefficient, although both are present at low concentrations.
The bacterial metabolism of S. aureus was, in turn, the most sensitive to the action of all the EOs tested. 20 µL/mL of the 'Bianco del Veneto' bulbs' EOs inhibited the adhesion process and bacterial biofilm formation (63.18% inhibition, Table 3). In addition, it exerted a potent inhibitory force on its cellular metabolism (61.44% inhibition).
Correlation analysis evidenced a strong influence exhibited by allicin (Corr-coeff = 0.449), which, on the other hand, was also the most abundant compound of these two EOs. The inhibitory effect of allicin on the S. aureus metabolism was weakened by the presence of allyl propyl disulfide (Corr-coeff = −0875) and decene (Corr-Coeff = −0.928). The different behavior exhibited by the EOs on the biofilm adhesion process and the metabolism of bacterial cells confirms previous studies on other Allium species and cultivars [28,46]. Casella and coworkers [8] demonstrated that the antibacterial activity of the EOs from different Allium species' fresh bulbs was mainly due to diallyl sulfides. This was partially ascertained by Polito et al. [28]. Taking  On the other hand, allicin positively influenced the capacity exhibited by those EOs to inhibit the metabolism of A. baumannii within the biofilm (Corr-coeff = 0.850). Thus, performing the correlation analysis based on the data reported in the work of Polito et al. [28], we might propose an almost generally negative influence of allicin and a positive effect exhibited generally by other sulfide compounds, such as diallyl sulfides and allyl propyl sulfides. In addition, other compounds, such as decene, are similar in their capacity to inhibit the formation of biofilms or the cell metabolism within the biofilm.
Yin and Cheng [39] reported some properties of diallyl sulfides, demonstrating their capacity to offer antioxidant and antimicrobial protection against pathogens, including E. coli, L. monocytogenes, and S. aureus. These compounds also acted on some bacterial enzymes involved in bacterial metabolism [47]. The EOs obtained from the two cultivars 'Bianco del Veneto' and 'Staravec', could inhibit biofilm formation and, in some cases, such as against E. coli and L. monocytogenes, were capable of affecting their metabolism. However, Fratianni et al. [48], analyzing the action of different types of honey on some pathogenic bacteria, observed that the inhibitory effect on biofilm formation is not always correlated with an analog effect on the bacterial metabolism within the biofilm. Similarly, the action of different seed oils against A. baumannii, E. coli, L. monocytogenes, P. aeruginosa, and S. aureus was sometimes different, depending on the microorganism [49].

Plant Material
Plants of the two cultivars of A. sativum, 'Bianco del Veneto' and 'Staravec', were collected in May-June 2020. The cultivars were grown in an experimental field at Pontecagnano (Salerno province, Southern Italy) on a fine-textured soil previously ploughed and fertilized. Cloves of both cultivars were planted on 15 November 2019, spaced 10 cm apart from each other and 10 cm deep in the soil, in rows spaced 50 cm apart, to obtain a density of 20 plants per square meter. According to a randomized block design with 3 reps, the 2 cultivars were arranged in 5 m 2 plots (2.0 m × 2.5 m). Furthermore, the standard agronomic practices of local garlic growers were followed. At harvest time, samples of 10 plants, randomly taken from each plot, were analyzed for morphological traits: the skin color of the bulbs and cloves, bulbs' mean weight, equatorial bulb diameter, number of cloves per bulb, cloves' mean weight. As shown in Table 5, the bulb and clove traits of garlic were significantly different between the cultivars. In particular, the mean weight and equatorial diameter of the 'Bianco del Veneto' bulbs were significantly higher than the 'Staravec' ones. In addition, the mean clove weight of 'Bianco del Veneto' was higher, whilst the number of cloves per bulb was significantly similar between the tested garlic cultivars.

Extraction of the EOs
Plant samples were cleaned from soil and other material residues and dried for about one week at room temperature. The plant material was divided into aerial parts and bulbs, separated and classified, and extracted with methanol at room temperature. This extraction was repeated three times, renewing the solvent. The extracts were then filtered using a paper filter and freed of excess methanol (Rotavapor R-100, BUCHI Italia s.r.l, 20007 Cornaredo, Italy). Subsequently, with the minimum amount of methanol, samples were placed in a flask half-filled with water and subjected to steam distillation, as reported in the European Pharmacopoeia [50]. The EOs obtained were solubilized in n-hexane, dried in a nitrogen atmosphere, and stored in amber vials in the refrigerator at 4 • C.

Composition of the EOs
The composition of the EOs was achieved by GC and GC-MS. GC analysis was performed using a Sigma 115 gas chromatograph (Perkin-Elmer Italia, Milano, Italy) equipped with a flame ionization detector (FID) and a non-polar HP-5 MS fused silica capillary column (30 m × 0.25 mm; 0.25 µm film thickness). Injector and detector temperatures were 250 • C and 290 • C, respectively. The analysis was carried out as follows: 5 min isothermally at 40 • C; the temperature was subsequently increased by 2 • C/min until 270 • C; and finally, the sample was kept in isotherm for 20 min. The analysis was also performed on an HP Innowax column (50 m × 0.20 nm; 0.25 µm film thickness). In both cases, He was used the carrier gas at a flow of 1.0 mL/min. GC-MS analysis was performed using a 6850 Ser II Apparatus (Agilent, Cernusco sul Naviglio, Italy) equipped with a DB-5 fused silica capillary column (30 m × 0.25 mm; 0.25 µm film thickness) and linked to an Agilent Mass Selective Detector (MSD 5973). Ionization voltage 70 V; ion multiplier energy 2000 V. The mass spectra were scanned in the range of 40-500 amu, with 5 scans per second. The chromatographic conditions were as above; the transfer line temperature was 295 • C. Most of the components were identified by comparison of their Kovats indices (Ki) with those of the literature [51][52][53] and by comparison of the mass spectra to those of pure compounds available in our laboratory or to those present in the NIST 02 and Wiley 257 mass libraries. The Kovats indices were determined in relation to a homologous series of n-alkanes (C10-C35), under the same operating conditions. For some compounds, the identification was confirmed by co-injection with standard compounds.

Antibacterial Activity Microorganisms and Culture Conditions
The Gram-positive, Staphylococcus aureus subsp.

Minimal Inhibitory Concentration (MIC)
The resazurin microtiter plate assay was used to determine the lowest concentration of the EOs capable of preventing visible growth of the microorganisms [54]. Multiwell plates were organized in triplicate; then, they were incubated at 37 • C and 35 • C, depending on the strain, for 24 h. The smallest concentration at which a colour change happened (from dark purple to colourless) fixed the MIC value.

Biofilm Inhibitory Action of the EOs
The EOs' capability to inhibit bacterial biofilm formation was assessed in flat-bottomed 96-well microtiter plates [46], using two concentration of EOs lower than those detected through MIC, which is applied generally to identify the concentration of a substance/molecule capable of inhibiting the growth of the planktonic cells. A total of 10 µL of the overnight bacterial cultures, standardized to 0.5 McFarland (1.5 × 10 7 cells/mL) with fresh culture broth, were put in each well. Then, 10 and 20 µL/mL of each EO and Luria-Bertani broth were added, to reach a final volume of 250 µL/well. Microplates were closed with parafilm, to prevent evaporation; they were incubated for 48 h at 37 • C (except for A. baumannii, incubated at 35 • C). Once removed the planktonic cells, wells were washed with sterile PBS, and we added 200 µL of methanol to fix the adhered cells. After 15 min, the methanol was removed, and the microplates were left to dry. The staining of the adhered cells was obtained by adding 200 µL of 2% w/v crystal violet solution, which was removed after 20 min. The wells were gently washed with sterile PBS and left to dry under the flow laminar cap. The addition of 200 µL of glacial acetic acid 20% w/v allowed for the release of the bound dye. The absorbance was measured at λ = 540 nm (Cary, Varian, Milano, Italy). The percent of adhesion was calculated with respect to the control; an inhibition of 0% was considered for cells without treatment. The tests were carried out in triplicate, and the average results were taken for reproducibility.

Inhibition of Cells' Metabolic Activity within the Biofilm
The EOs capacity to inhibit the metabolic activity of the bacterial cells was assessed through the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) method [55]. The method is different from MIC, as it is applied to the cells within the biofilm after the discarding of the non-adhered cells; furthermore, it was used considering concentrations of the EOs lower than those identified by MIC. Ten µL of the overnight bacterial cultures, standardized to 0.5 McFarland (1.5 × 10 7 cells/mL) with fresh culture broth, were put in each well. Then, 10 and 20 µL/mL of each EO and Luria-Bertani broth were added to reach a final volume of 250 µL/well. After 48 h of incubation, planktonic cells were removed, and 150 µL of PBS and 30 µL of 0.3% of MTT (Sigma, Milano, Italy) were added. Depending on the strain, microplates were kept at 37 • C and 35 • C. After 2 h, we removed the MTT solution, and two washing steps were performed with 200 µL of sterile physiological solution. Then, 200 µL of dimethyl sulfoxide (Sigma, Milano, Italy) was added to allow for the dissolution of the formazan crystals by the adhered cells, which was measured after 2 h, at λ = 570 nm (Cary, Varian, Milano, Italy).

Statistical Analysis
All assays were carried out in triplicate. Data from each experiment were expressed as the mean ± SD, and statistically analyzed by two-way ANOVA followed by Dunnet's multiple comparisons test, at the significance level of p < 0.05, using GraphPad Prism 6.0. The analysis also correlated the values of the antibiofilm activity of the EOs with their composition, using the free software environment for statistical computing and graphics, R (https://www.r-project.org/ last accessed on 18 May 2022).

Conclusions
This research demonstrated a chemical difference between the EOs of two cultivars of A. sativum. The chemical compositions confirmed the presence of sulfur compounds, particularly allicin, as reported in the literature. However, the percent amounts of allicin were variable between the two cultivars and between the aerial parts and bulbs of the same cultivar; indeed, allicin, rarely affected the inhibitory biofilm capabilities exhibited by EOs. Both the EOs from the aerial parts and the bulbs of the cv. 'Bianco del Veneto', unlike the cv. 'Staravec', were able to inhibit the biofilm formation of L. monocytogenes. On the other hand, the EOs from aerial parts bulbs of 'Staravec' acted against the cellular metabolism of E. coli.
Identifying substances with a broad spectrum of antibacterial activity against Grampositive and Gram-negative bacteria, and their virulence, is always helpful from an application point of view, both for food safety and for different areas such as hospitals.
The inhibitory action exhibited not only by the bulbs but also by the aerial parts of the two cultivars of A. sativum represents an aspect of particular interest. These results could allow for the more rational use and exploitation of all the A. sativum plants. Most people continue to harvest bulbs exclusively and produce a large amount of Allium plant waste, which could have important biological properties. This might give several advantages to the food industry, as the extracts/EOs of the aerial parts could be helpful in the formulation of natural preserving agents to prolong the shelf life of perishable products. Furthermore, other studies on these substances could give the industry new weapons to fight pathogenic bacteria that are becoming more easily resistant to conventional antibiotic therapies. Finally, as we have tried to highlight, the results show once again that the biological properties of a plant are the result of the synergistic or antagonistic actions of the singular substances that compose it. Biodiversity is also like this, which could provide a full advantage to the circular economy, which today represents the most crucial pillar from an environmental, economic, and social point of view.