Volatiles and Antifungal-Antibacterial-Antiviral Activity of South African Salvia spp. Essential Oils Cultivated in Uniform Conditions

Spontaneous emissions of S. dentata Aiton and S. scabra Thunb., as well as the essential oil (EO) composition of the cited species, together with S. aurea L., were investigated. The chemical profile of the first two species is reported here for the first time. Moreover, in vitro tests were performed to evaluate the antifungal activity of these EOs on Trichophyton mentagrophytes, Microsporum canis, Aspergillus flavus, Aspergillus niger, and Fusarium solani. Secondly, the EO antibacterial activity against Escherichia coli, Staphylococcus aureus, and Staphylococcus pseudointermedius was examined, and their antiviral efficacy against the H1N1 influenza virus was assessed. Leaf volatile organic compounds (VOCs), as well as the EOs obtained from the arial part of Salvia scabra, were characterized by a high percentage of sesquiterpene hydrocarbons (97.8% and 76.6%, respectively), mostly represented by an equal amount of germacrene D (32.8% and 32.7%, respectively). Both leaf and flower spontaneous emissions of S. dentata, as well as the EO composition, showed a prevalence of monoterpenes divided into a more or less equal amount of hydrocarbon and oxygenated compounds. Interestingly, its EO had a non-negligible percentage of oxygenated sesquiterpenes (29.5%). S. aurea EO, on the contrary, was rich in sesquiterpenes, both hydrocarbons and oxygenated compounds (41.5% and 33.5%, respectively). S. dentata EO showed good efficacy (Minimal Inhibitory Concentration (MIC): 0.5%) against M. canis. The tested EOs were not active against E. coli and S. aureus, whereas a low inhibition of S. dentata EO was observed on S. pseudointermedius (MIC = 10%). Once again, S. dentata EO showed a very good H1N1 inhibition; contrariwise, S. aurea EO was completely inactive against this virus. The low quantity of S. scabra EO made it impossible to test its biological activity. S. dentata EO exhibited interesting new perspectives for medicinal and industrial uses.


Introduction
Throughout the world, thousands of people are affected by dermatophyte infections which constitute the most common skin diseases. These infections, especially caused by fungal pathogens belonging to Trichophyton and Microsporum species [1], are lead to the fourth highest incidence of disease when compared to hundreds of different illnesses and injuries globally [2]. On the other hand, Staphylococcus aureus is also considered 2. Results and Discussion 2.1. Aroma Profile and EO Analyses 2.1. 1
There are still few studies concerning the flavor profile of South African sage species. In fact, according to the best of our knowledge, no work has reported the spontaneous emissions of S. dentata and S. scabra. The articles present in the literature, so far, focused on other species. Ascrizzi [26], studying the VOC composition of S. aurea and S. aurita L.f., evidenced the prevalence of monoterpene hydrocarbons (93.4%) and sesquiterpene hydrocarbons (44.4%), respectively. These results disagreed with our findings in S. dentata, in which oxygenated monoterpenes prevailed (57.9% in the flowers and 47.5% in the leaves), although a considerable percentage of monoterpene hydrocarbons was observed (greater than 40.0% both in the leaves and in the flowers). On the contrary, the volatile organic compounds of S. scabra leaves (97.8% sesquiterpene hydrocarbons) had a similar trend to S. aurita [26]. Myrcene and β-caryophyllene were the most abundant compounds in S. aurea (89.5%) and S. aurita (17.8%), respectively. Myrcene was present in a reduced percentage only in the leaves of S. dentata, while β-caryophyllene was more abundant in S. scabra, even though the prevalent compound was germacrene D.
Comparing the current results of the S. aurea EO with those reported in the literature, some substantial differences can be noted. First of all, the main compounds changed depending on the use of fresh or dry material. In fact, fresh leaves of S. aurea (native from South Africa but grown in Italy), analyzed by Serrato-Valenti [28], pointed out presented (34.7%), δ-3-carene (16.5%), and camphene (8.3%) as the predominant constituents. The oil extracted herein from dry material showed β-caryophyllene (12.5%) and epi-α-cadinol (10.2%) as the most abundant components, although a considerable percentage of δ-3carene (7.8%) was also noted. Moreover, both camphor and camphene were present, but in extremely low percentages (0.2% and 0.1%, respectively). The dried aerial parts of native South African S. aurea EO was characterized by monoterpene hydrocarbons (35.6%), with myrcene as the most abundant constituent (11.5%) [29]. These were in total disagreement with our findings, and the myrcene value did not exceed 1.0%.
The same species collected in the Western Cape region of South Africa in the vegetative stage was the subject of a more recent work [31]. The authors registered a yield of EO very low in comparison with that found herein (0.25% vs. 1.01%) but in agreement with the amount found by Kamatou [29]. They also observed that β-eudesmol (12.3%), α-eudesmol (12.4%), terpinene-4-ol (10.1%), and T-cadinol (7.6%) were the majority. This was in contrast with the results of this study, where the EO was eudesmol-free and terpinene-4-ol was of a very less percentage (0.2%).
Kamatou [29] examined the EO composition of S. africana-caerulea L. and S. lanceolata Lam. and found oxygenated sesquiterpenes as the predominant constituents (58.7% and 47.9%, respectively), mostly represented by spatulenol (29.1% and 18.3%, respectively) and caryophyllene oxide (14.0-15.0%). This chemical class, although present in all the examined species in consistent percentages (33.5% in S. aurea, 29.5% in S. dentata, and 15.8% in S. scabra), did not predominate in any of them. The highest percentage of caryophyllene oxide was found in S. aurea (3.6%) and S. scabra (3.4%), while its level in S. dentata did not exceed 0.2%. A small percentage of spatulenol, on the other hand, was present only in S. scabra.

Antimicrobial and Antiviral Activities
The examined EOs showed variable antimycotic degree toward the tested fungi ( Table 3). The microdilution test showed that only dermatophytes were sensitive to the sage Eos, particularly S. dentata. The low quantity of S. scabra EO made it impossible to test its antimicrobial and antiviral activity. Noteworthily, S. dentata EO demonstrated a good antifungal action on M. canis (MIC = 0.5%). The tested oils were not active against E. coli and S. aureus, whereas a low inhibition efficacy of Salvia dentata EO was observed on S. pseudointermedius (MIC = 10%). Regarding the antiviral activity, only S. dentata EO showed very good H1N1 inhibition. On the contrary, S. aurea was completely inactive against this virus (Table 4). Table 3. Results of microdilution testing of S. aurea and S. dentata EOs on selected fungal species.

EOs
Microsporum canis  As far as we know, no report is present in the literature on the biological activity of S. dentata EO. Only S. aurea EO was the subject of a few studies for its biological activity. The first one dates back to 1998 when Bisio and collaborators investigated its antimicrobial activity [33]. The authors tested this oil on 13 microorganisms. They found a nonsignificant effect on Gram-positive bacteria, especially S. aureus. These results were confirmed in this work. Later on, Russo tested the oil firstly on human melanoma cells (M14, A2058 and A375) [34] and then on prostate cancer [35]. In both works, the author affirmed the inhibition of growth and an apoptotic effect on all the tested cells.
Kamatou and collaborators [36] investigated other South African species and noted a poor antimicrobial activity of the three sage EOs. Later, van Vuuren [30] demonstrated the highest activity of S. africana-caerulea EO against the Brevibacillus agri foot-odor causing bacterium. This oil was characterized by viridiflorol (36.7%) and limonene (25.7%).
Viridiflorol was also found as the major constituent in Algerian S. algeriens Desf. and Iranian S. sclareopsis flower (71.1%) and leaf EOs (23.47%). Antifungal activity against Alternia solani and Fusarium oxysporum was noted using the flower of Algerian sage even though the best effectiveness was observed using the leaf EO (rich in benzaldehyde, eugenol, and phenylethyl) [37]. The S. sclareopsis leaf EO, however, evidenced a high antioxidant activity [38]. These results confirmed those found here, where the S. dentata EO, i.e., that with the highest amount of viridiflorol, was completely inactive on the tested Fusarium spp.
The investigation on Tunisian sage (S. officinalis L.), where camphor was one of main compounds (25.14%) as in our S. dentata EO, reported an interesting activity of this oil against S. aureus [39]. Their results disagreed with ours because S. dentata EO showed only slight activity. The observed activity of S. dentata EO could be due to the presence of both viridiflorol and camphor. Da Silva [40], in fact, demonstrated that viridiflorol, even though in a low amount, was more effective against S. aureus, while Gilabert [41] noted an inhibition of about 40% of the growth of human pathogenic bacteria using the same compound (viridiflorol) at 50 µg/mL. Trevizan [42] also showed the in vitro efficacy of the cited component but on Mycobacterium tuberculosis (MIC = 190.0 µg/mL), and they compared the in vivo anti-inflammatory activity to dexamethasone. This compound was also noted for its potent acetylcholinesterase inhibition [43]. A moderate antioxidant activity of Ferula vesceritensis, where viridiflorol was one of the main compounds (13.4%), was reported by Benchabne et al. [44].
Observing the results of antiviral activity once again, only the S. dentata EO evidenced very good action on H1N1 virus (Table 5). This oil presented a fair percentage of β-pinene, (3.2%), a compound known to have good anti-HSV activity, as stated by Orhan [45] and Astani [46], whereas, α-pinene reduced the infectivity of HSV-1 by >96% [47]. The cited constituent was present at a highest percentage in the active sage oil (10.2%). Falang carried out an in silico investigation of the plant constituents of some Nigerian medicinal species [48]. The authors showed that the phytochemical constituents of the selected plants had better binding affinity to several Covid-19 viral target proteins, testing the S. officinalis EO compounds borneol, camphor, and pinene. Thus, this plant was selected among nine others to proceed with in vitro studies. These constituents were almost exclusive to S. dentata essential oil and could be responsible for the antiviral action of this oil. Kamatou [36] assessed the antimalarial activity of three South African sage species and pointed out that the best antimalarial and anti-inflammatory activity was shown by S. runcinate EO, where β-caryophyllene (10.5%) was one of main compounds, as in S. aurea. The S. aurea EO used herein was ineffective on the tested virus. This could be explained, on one hand, by the fact that both malaria and influenza virus evidenced different sensitivity to this oil and, on the other hand, by the fact that the antimalaria activity may have been due to a synergic effect of other compounds [49,50].
β-Caryophyllene, a compound shared by different Salvia species, also suppressed HSV multiplication by more than 90% [51]. An investigation on Mosla dianthera EO, which showed a comparable amount of β-caryophyllene, confirmed its safe and effective therapeutic ability for the treatment of influenza and subsequent viral pneumonia [52]. Recently, Dunkic [53] demonstrated that, in addition to β-caryophyllene content, germacrene D might play an important role in the antiphytoviral activity. These results were in complete disagreement with those found in this work, where, although S. aurea contained large amounts of β-caryophyllene, it was completely inactive on the tested virus. Others compounds were found to inhibit virus replication with a dosage below the cytotoxic level, such as terpinen-4-ol, terpinolene, and α-terpineol [54]. All of these compounds were present in the analyzed sage samples, except for α-terpineol, which was present only in S. dentata EO, in a very low amount; however, this could explain the activity of this latter EO on H1N1 virus.

Origin and Cultivation Method of the Plant Material
The native South African Salvia spp. (Table S1)

Volatile Organ Compound and Essential Oil Analyses
The analysis of volatile organic compounds was performed on fresh plant using the solid-phase microextraction (SPME) method [55]. A sample of the fresh aerial parts of each Salvia sp. was placed separately in a glass jar, and then sealed with aluminum foil for 30 min (equilibration time) at room temperature (22 ± 1 • C). By the end, the fiber (PDMS, 100 µm) (St. Louis, MO, USA), previously preconditioned according to the manufacturer's instructions, was exposed to the headspace for 15 min. Once sampling was finished, the fiber was withdrawn into the needle and transferred to the injection port of the GC-MS instrument, where thermal desorbing and component analysis took place. It was not possible to analyze the spontaneous emissions of S. aurea because the fresh plant was not available. For the essential oil extraction, the dried aerial parts of each plant were hydrodistilled using the Clevenger apparatus according to the European Pharmacopoeia (EDQM, 2017). The obtained essential oil was kept at a temperature of 4 • C and away from light sources until analysis. A diluted oil, in n-hexane by HPLC (at 5%), was injected into GC-MS.

GC-MS Analyses
Gas chromatography-mass spectrometry (GC-MS) was used to determine VOC and EO components. The gas chromatograph used was an Agilent 7890B (Agilent Technologies Inc., Santa Clara, CA, USA). The mobile phase was represented by helium (He). The capillary column was an Agilent HP5-MS (Agilent Technologies Inc., Santa Clara, CA, USA), of 30 m length and 0.25 mm diameter. The stationary phase was linked to the internal surface of the column via covalent bonds and was stabilized by transverse bonds. The syringe was inserted into the gas chromatograph through the injector, allowing the adsorbent fiber to come out. The splitless method was used for injection; the injected sample was vaporized and transported to the carrier gas column. The temperature at the injector level was 220 • C. The separation column was contained in a thermostatic chamber, in which the starting temperature was 60 • C, increasing by 3 • C per minute up to 240 • C. The detector coupled to the gas chromatograph was an Agilent 5977B single-quadrupole mass spectrometer (Agilent Technologies Inc., Santa Clara, CA, USA), operating in full scan mode (1 scan/s), in the range 30-300 m/z.
The compounds were identified by comparing their retention times with those of pure reference samples and comparing their linear retention indices (LRIs), determined relative to a series of n-alkanes.

Evaluation of Antifungal Activity
The EOs were tested on dermatophytes (M. canis and T. mentagrophytes), isolated from animal hair samples and potentially mycotoxin-producing molds (A. niger, A. flavus, and F. solani) isolated from environmental sources. The molds were maintained on Potato Dextrose Agar at −20 • C. The antimycotic activity of EOs was investigated using a microdilution test as recommended by Clinical and Laboratory Standards Institute M38A 2 [60] (CLSI, 2008) and following the protocol described by Ebani [63], which evaluated the growth of the fungus in culture media added with scalar concentrations (v/v) at 5%, 4%, 3%, 2%, 1%, 0.25%, 0.2%, and 0.1%. All assays were performed in triplicate. Positive controls were achieved using itraconazole for dermatophytes and amphotericin B; the negative control was culture medium alone.

Antibacterial Activity
The tests were executed using three canine clinical isolates, specifically, one E. coli strain and one S. aureus strain previously isolated from dogs with urinary tract infections, and one S. pseudointermedius strain isolated from a dog with otitis. The antibacterial activity of essential oils was tested using both the diffusion agar method (Kirby-Bauer) and the broth microdilution test.

Agar Disc Diffusion Method (Kirby-Bauer Technique)
The agar disc diffusion method was executed following the procedures described by Clinical and Laboratory Standards Institute [64] and with some modifications as previously described. Briefly, 9 cm diameter petri dishes containing Muller-Hinton medium were sown, through the use of a swab, with the bacterial strain in order to obtain uniform bacterial growth. Sterile cellulose 6 mm discs soaked in a solution (10% v/v in dimethyl sulfoxide (DMSO)) of each EO were added. The in vitro sensitivity of all bacterial strains to chloramphenicol was assayed using the same method, and the results were interpreted as indicated by the National Committee for Clinical Laboratory Standards [65]. The plates were then incubated for 24 h at 37 • C. The diameters of inhibition zones (IZs) were measured in millimeters, and the tests were performed in triplicate.

Minimum Inhibitory Concentration
The MIC value (minimum inhibitory concentration) was determined by the broth microdilution method following the protocol previously reported [63]. In brief, an EO stock solution was prepared by adding 40 µL of each oil to 360 µL of BHI (Brain Heart Infusion) broth. The test involved the preparation of a series of decreasing scalar dilutions (halved) of the antimicrobial agent, to which the same amount of BHI was added. In each well of a 96-well sterile microplate, 95 µL of BHI was added to 95 µL of the stock solution (10% solution). Then, 5 µL of each bacterial suspension was inoculated into each well. The test was performed in a total volume of 100 µL with final EO concentrations of 10% to 0.5%. The same assay was performed simultaneously for microorganism growth control (tested agents and media) and sterility control (tested oil and media). All tests were performed in triplicate, with chloramphenicol (Oxoid Ltd., Basingstoke, UK) as a positive control.
The microplates were incubated at 37 • C for 24 h. The MIC value was defined as the lowest concentration, expressed as mg/mL, of each EO for which microorganisms showed no visible growth.

Antiviral Activity
Madin-Darby canine kidney (MDCK) cells, propagated in modified Eagle's medium (MEM; SIGMA, Milano, Italy) supplemented with 10% fetal bovine serum (FBS; SIGMA) and 1% penicillin/streptomycin (SIGMA), were used for the inhibitory viral plaque reduction assay (PRA). Briefly, six-well plates were seeded with 2.5 × 10 5 cells in 3 mL of growth medium and kept overnight in incubators at 37 • C with 5% CO 2 . On the day of infection, after removal of the growth medium, cell monolayers at 80-90% confluence were infected with 100 mL of influenza virus H1N1 (human pandemic variant A/Firenze/05/2017 H1N1) with a multiplicity of infection (MOI) of 0.01 in the presence or absence (MEM with DMSO alone) of different concentration (from 0.1% to 0.0001%) of each EO diluted in DMSO in a final volume of 0.3 mL and incubated for 1 h at 37 • C with 5% CO 2 . Then, after a washing step with PBS 1×, the overlay medium composed of 0.5% Sea Plaque Agarose (Lonza, Basel, Switzerland) diluted in propagation medium supplemented with L-1-tosylamido-2phenylethyl-chloromethyl-ketone-treated trypsin (2 mg/mL; Sigma, St. Louis, MO, USA) was added to each well. After 4 days of incubation at 37 • C, the monolayers were fixed with methanol (Carlo Erba Chemicals, Milan, Italy) and stained with 0.1% crystal violet (Carlo Erba Chemicals), and the viral titers were calculated on the basis of counting plaqueforming units (PFU). The percentage of PRA was calculated by dividing the average PFU of EO-treated samples by the average of untreated samples (viral positive control in the presence of DMSO alone): PRA = 100 − (PFU obtained with EOs at indicated dilution/PFU obtained with DMSO alone) × 100. All experiments were repeated at least twice.

Conclusions
The volatile emissions of S. dentata and S. scabra, as well as the EO composition of the cited species, together with S. aurea (South Africa sages), were investigated. The chemical profiles of the first two species were reported here for the first time. The EOs obtained in good amounts were tested for antifungal, antibacterial, and antiviral activity. It is worthy to note that the effect of S. dentata essential oil was startling and showed a fair to good activity on the tested pathogens. These plants were introduced in Italy for ornamental purposes by CREA-OF (Sanremo); however, according to these encouraging biological activities, the EO of S. dentata deserves deep investigation and could exhibit interesting new perspectives for medicinal and industrial uses. The S. sclarea EO presented compounds previously known for their good antiviral activity; therefore, the use of another extraction method would be interesting to increase its EO yield in order to verify this potential.