Bergamot was harvested in July 2014 from native plants in Descanso (Santa Catarina State, Brazil). After harvesting, the bergamot (Delicious Citrus Bergamia) was washed with abundant water and then peeled. Peels were cut in 1 cm² pieces and then stored at -18 °C in polypropylene bags. Neither in the peeling step nor in peel cutting, losses of bergamot peel oil were visually observed. Propane and carbon dioxide (CO₂) with minimum purity of 99.5% in the liquid phase was purchased from White Martins S.A. Staphylococcus aureus strains (ATCC 25923) and Escherichia coli strains (ATCC 25922) were used for the antimicrobial activity tests. Brain Heart Infusion Broth (BHI), 2,3,5-Triphenyl-tetrazolium chloride (TTC), and dimethylsulfoxide (DMSO, 99%) was purchased from Merck, Sigma-Aldrich and Vetec, respectively.

Before extraction, the previously cut bergamot peels were crushed in a slicer and sieved to obtain particles of mean size ≤ 2 mm. Typically, around 30 g (±0.05 g) of raw material was used in all extraction experiments. The equipment consisted of four components: a solvent gas cylinder, a high-pressure pump, an extraction vessel, a micrometric valve connected to collection flask and a temperature and pressure control systems. A detailed discussion about experimental apparatus, methodology and extraction conditions can be found in the works of Capeletto et al. (2016) [21Capeletto C, Conterato G, Scapinello J, et al. Chemical composition, antioxidant and antimicrobial activity of guavirova (Campomanesia xanthocarpa Berg) seed extracts obtained by supercritical CO2 and compressed n-butane. J Supercrit Fluids 2016; 110: 32-8.
[http://dx.doi.org/10.1016/j.supflu.2015.12.009] ] and Scapinello et al. (2015) [22Novello Z, Scapinello J, j Dal Magro, et al. Extraction, chemical characterization and antioxidant activity of andiroba seeds oil obtained from pressurized n-butane. Ind Crops Prod 2015; 76: 697-701.
[http://dx.doi.org/10.1016/j.indcrop.2015.07.075] ].

After preliminary tests, the extraction conditions for all assays were set as solvent flow rate of 3 mL min^-1 and 2 h of extraction time for experimental runs. In fact, preliminary tests showed that extraction times greater than 2 h led to negligible gains of the extract obtained and therefore above this time resulted in practice in solvent losses with additional costs and unfruitful results. Also, preliminary tests and experience from previous works of our research group show that good extraction conditions may be 55 °C and 350 bar for SCCO₂ (density 0.881kg/m³) and 55 °C and 40 bar (density 0.441kg/m³) for propane. The purpose for these extraction conditions was to increase the possibility of extraction a larger number of chemical compounds by raising vapor pressure of pure components while improving solvation trough adequate solvent extracting densities. For all extraction conditions using SCCO₂ and pressurized propane, bergamot peel oils obtained were collected in amber bottles and then stored under refrigeration (5 ºC ± 2 ºC). The overall extraction yield of bergamot peel oil was obtained by the relation between the total mass oil extracted by the mass of the sample. Based on triplicate experiments carried out for all the experimental conditions for both compressed solvents, the overall average standard deviation of the yields noticed was about 0.2 wt%.

The chemical composition of bergamot peel oil was determined using a Shimadzu gas chromatograph model Varian 3800 equipped with a split-injection port, flame-ionization detector and a HP-Innowax column (25 m x 0.25 mm x 0.5 μm). The chromatograph operating conditions were: split = 1:150; column flux = 0.92 mL.min^-1, detector temperature: 250 °C, injector temperature: 250 °C, oven temperature: 60 °C (8 min), 180 °C (4 °C min^-1), 180 °C - 230 °C (20 °C min^-1), 230 °C (20 min) using helium as carrier gas: split ratio: 50:1, 34 kPa, and injected volume of 1.0 µL. The composition was obtained from electronic integration measurements using flame ionization detection. The GC-MS analysis was performed on a GC-MSD system model HP 5973-6890, operating in EI mode at 70 eV, equipped with a cross-linked column capillary (HP-5, 30 m x 0.25 mm). Helium was used as carrier gas (56 kPa, 1 ml.min^-1). The identification of compounds in the bergamot peel oil was based on the retention index, determined as a function of a homologous series of C₇-C₃₀ n-alkanes, under identical experimental conditions, to the NBS mass Library (Massada, 1976) described by Felton et al. (2009) [24Adams RP. Identification of essential oil components by gas chromatography/mass spectrometry No 4^th ed. 2007.]. The relative amounts of the individual components were calculated based on the GC peak area (FID response).

The antimicrobial analyses to determine Minimum Inhibitory Concentration (MIC) were done for the bergamot peel oil extracted with pressurized propane and supercritical CO₂ (Table 1). The assays were carried out in triplicate on systems sterile microplates containing 96 cavities shaped bottom “U” containing 100 μL of Brain Heart Infusion Broth (BHI), following the adapted method described by Hentz and Santin (2007) ] [25Hentz SM, Santin NC. Avaliação da atividade antimicrobiana do óleo essencial de alecrim (Rosmarinus officinalis l.) contra Salmonella sp. Evidência Interdisciplinar 2007; 7: 293-100.]. A 200 μL volume of the sample of bergamot peel oil extracted at the concentration from 27 mg.mL^-1 (diluted in 10% DMSO) was filtered (0.45 μm Millipore filter) and inoculated in BHI medium. Also, the blank test was done with 100 µL of BHI in 10% DMSO. The bacterial inoculums of Staphylococcus aureus and Escherichia coli with a concentration of 0.5 in the McFarland scale (10⁸ UFC mL^-1) were diluted in 0.9% in a sterile saline solution and a volume of 5 µL (10⁴ UFC mL^-1) was deposited in each cavity of the microplate utilizing a micropipette. In each line of the microplates containing different concentrations of the bergamot peel oil extract, adjusted with DMSO 10%. The 96-well microplates were incubated in a bacteriological oven at 35 °C for 18 hours. Afterward, 20 µL de TTC 0.5% was added in each cavity of the microplate and again incubated for another three hours. The MIC was defined as the lowest concentration, enough to inhibit the microbial growth (Mann and Markham, 1998).

Table 1
Chemical composition of bergamot peel oil obtained from supercritical CO₂ (55 °C and 350 bar) and compressed propane extraction (55 °C and 40 bar).

The obtained extraction yields, defined as the weight percentage of the oil extracted with respect to the initial charge of the raw material in the extractor, were 0.48 wt% and 0.37 wt% for SCCO₂ and pressurized propane, respectively. Table 2 shows the bergamot peel oil composition obtained by SCCO₂ and pressurized propane extraction. It can be observed that more compounds were identified from the SCCO₂ (99.8%) extract compared to propane (~ 87%), which may be justified to the fact that in the experimental conditions used the propane has greater power of solvation and in this way extracted less volatile compounds, with higher molecular weight and consequently greater structural complexity that were not identifiable by GC-MS [26Illés V, Daood HG, Perneczki S, Szokonya L, Then M. Extraction of coriander seed oil by CO2 and propane at super- and subcritical conditions. J Supercrit Fluids 2000; 17: 177-86.
[http://dx.doi.org/10.1016/S0896-8446(99)00049-2] ].

Table 2
Minimum inhibitory concentration (MIC) of bergamot peel oil extracts.

Comparison of chemical profile obtained from the application of extraction solvents shows that the major constituent in both bergamot peel oils was limonene, an important monoterpene, precursor of many pharmaceuticals and chemicals [27Keasling JD. Manufacturing molecules through metabolic engineering. Science 2010; 330(6009): 1355-8.
[http://dx.doi.org/10.1126/science.1193990] [PMID: 21127247] ], with 45% observed for SCCO₂ and 21.5% for propane. These results are in accordance with the studies reported by Sawamura et al. (2006) [28Sawamura M, Onishi Y, Ikemoto J, Tu NTM, Phi NTL. Characteristic odour components of bergamot (Citrus bergamia Risso) essential oil. Flavour Fragrance J 2006; 21: 609-15.
[http://dx.doi.org/10.1002/ffj.1604] ], in which the authors reported limonene as the major component in bergamot essential oil.

Other important components identified in bergamot peel oil were linalyl acetate, linalool and geraniol with 23.6%, 11.3%, and 0.25% of the total components extracted with CO₂ and 19.7%, 15.1%, and 2.03% for propane, respectively. As reported by Elisabetsky et al. (1995) [29Elisabetsky E, Marschner J, Souza DO. Effects of Linalool on glutamatergic system in the rat cerebral cortex. Neurochem Res 1995; 20(4): 461-5.
[http://dx.doi.org/10.1007/BF00973103] [PMID: 7651584] ], compounds such as linalyl acetate and linalool may be associated with anxiolytic activity. According to Camargo and Vasconcelos (2014) [30Camargo SB, Vasconcelos DFSA. Atividades biológicas de Linalol: Conceitos atuais e possibilidades futuras deste monoterpeno. Rev Ciênc Méd Biol 2014; 13: 3-381.
[http://dx.doi.org/10.9771/CMBIO.V13I3.12949] ], linalool is a secondary metabolite, one of the most important substances in the pharmaceutical industry, used by popular medicine as anti-inflammatory, analgesic, hypotensive, and antimicrobial properties [10Sakurada T, Mizoguchi H, Kuwahata H, et al. Intraplantar injection of bergamot essential oil induces peripheral antinociception mediated by opioid mechanism. Pharmacol Biochem Behav 2011; 97(3): 436-43.
[http://dx.doi.org/10.1016/j.pbb.2010.09.020] [PMID: 20932858] , 31Julião LS, Tavares ES, Lage CLS, Leitão SG. Cromatografia em camada fina de extratos de tr{ê}s quimiotipos de Lippia alba (Mill) N.E.Br. (erva-cidreira). Rev Bras Farmacogn 2003; 13: 36-8.
[http://dx.doi.org/10.1590/S0102-695X2003000300014] , 32Costa AM, Buglione CC, Bezerra FL, Martins PCC, Barracco MA. Immune assessment of farm-reared Penaeus vannamei shrimp naturally infected by IMNV in NE Brazil. Aquaculture 2009; 291: 141-6.
[http://dx.doi.org/10.1016/j.aquaculture.2009.03.013] ].

Results reported in this research work agree with those published by Dugo et al. (2000) [33Dugo P, Mondello L, Dugo L, Stancanelli R, Dugo G. LC-MS for the identification of oxygen heterocyclic compounds in citrus essential oils. J Pharm Biomed Anal 2000; 24(1): 147-54.
[http://dx.doi.org/10.1016/S0731-7085(00)00400-3] [PMID: 11108548] ] and Russo et al. (2013) [34Russo R, Ciociaro A, Berliocchi L, et al. Implication of limonene and linalyl acetate in cytotoxicity induced by bergamot essential oil in human neuroblastoma cells. Fitoterapia 2013; 89: 48-57.
[http://dx.doi.org/10.1016/j.fitote.2013.05.014] [PMID: 23707744] ], who reported that bergamot peel oil has a non-volatile fraction (4-7%), characterized by coumarins and furocoumarins, and a volatile fraction (93-96%) with hydrocarbons as monoterpenes and sesquiterpenes such as limonene, γ-terpinene, α-β-pinene, β-myrcene, sabinene, β-bisabolene, and oxygenated derivatives, such as linalool, linalyl acetate, neral, geraniol, neryl and geranyl acetate.

From the MIC results, it was noticed that the bergamot peel oils were efficient as bacterial growth-inhibitor. However, the antibacterial effect was higher against Gram-positive bacteria (S. aureus) with a determined MIC of 31.2 μg.mL^-1. On the other hand, Gram-negative bacteria (E. coli) was more resistant, requiring higher concentrations of bergamot peel oil, 500 μg.mL^-1 of oil obtained with SCCO₂ and 125 μg.mL^-1 of with pressurized propane. The resistance of Gram-negative bacteria such as E. coli may be associated with a complex cell wall compared to Gram-positive bacteria, thereby increasing their resistance to the antibacterial agents. Ashok Kumar et al. (2011) [35Ashok kumar K, Narayani M, Subanthini A, Jayakumar M. Antimicrobial activity and phytochemical analysis of citrus fruit peels -Utilization of fruit waste. Int J Eng Sci 2011; 3(6): 5414-21.], using five different solvent extracts of Citrus lemon and Citrus simensis, analyzed the antimicrobial activity and reported a MIC of 12.5 mg.mL^-1 for E. coli and for 25 mg.mL^-1 for S. aureus, these results is similar to the strong antimicrobial activity of the bergamot oil obtained in this work.

The difference in the minimum inhibitory concentration for E. coli for the two bergamot peel oils obtained with CO₂ (500 μg.mL^-1) and propane (125 μg.mL^-1) may be related to differences regarding chemical composition of oils obtained, and might be associated with the compounds γ-terpinene (3.25% for CO₂ and 6.08% propane), geraniol (0.25% CO₂ and 2.03% propane), β-caryophyllene (0.84% CO₂ and 1.96% propane), neryl acetate (0.47% CO₂ and propane 2.61%), trans- α -bamamotene (0.08% CO₂ and 1.17% propane), β-bisabolene (1.27% CO₂ and 3.04% propane) and linalool (11.35% CO₂ and 15.16% propane).

According to Sokovié et al. (2010) [36Soković M, Glamočlija J, Marin PD, Brkić D, van Griensven LJ. Antibacterial effects of the essential oils of commonly consumed medicinal herbs using an in vitro model. Molecules 2010; 15(11): 7532-46.
[http://dx.doi.org/10.3390/molecules15117532] [PMID: 21030907] ], linalool has higher antimicrobial effects against Gram-positive bacteria compared to Gram-negative ones, with an effect on protein denaturation or dehydration of the cells. Limonene, a cyclic hydrocarbon, can accumulate in the biological membrane and change its structure and function [37Sikkema J, de Bont JAM, Poolman B. Interactions of cyclic hydrocarbons with biological membranes. J Biol Chem 1994; 269(11): 8022-8.
[PMID: 8132524] ]. In addition, limonene is reported as an antifungal agent [38Chee HY, Kim H, Lee MH. In vitro Antifungal Activity of Limonene against Trichophyton rubrum. Mycobiology 2009; 37(3): 243-6.
[http://dx.doi.org/10.4489/MYCO.2009.37.3.243] [PMID: 23983542] ], bacteriostatic [39Jaroenkit P, Matan N, Nisoa M. In vitro and in vivo activity of citronella oil for the control of spoilage bacteria of semi dried round scad (Decapterus maruadsi). Int J Med Aromatic Plants 2011; 1: 234-9., 40Van Vuuren SF, Viljoen AM. Antimicrobial activity of limonene enantiomers and 1,8-cineole alone and in combination. Flavour Fragrance J 2007; 22: 540-4.
[http://dx.doi.org/10.1002/ffj.1843] ], and anti-bactericidal plasma membrane of the bacteria and causes the loss of the bacterial membrane integrity [41Zukerman I. Effect of oxidized d-limonene on micro-organisms. Nature 1951; 168(4273): 517-8.
[http://dx.doi.org/10.1038/168517a0] [PMID: 14875145] ]. A study of E. coli inactivation by terpenes and terpenoids (such as carvacrol or citral) demonstrated the sublethal lesions in the cytoplasmic membrane [42Ait-Ouazzou A, Cherrat L, Espina L, Lorán S, Rota C, Pagán R. The antimicrobial activity of hydrophobic essential oil constituents acting alone or in combined processes of food preservation. Innov Food Sci Emerg Technol 2011; 12: 320-9.
[http://dx.doi.org/10.1016/j.ifset.2011.04.004] ] suggesting the bactericidal membrane rupture as the inactivation mechanism [43Espina L, Gelaw TK, de Lamo-Castellví S, Pagán R, García-Gonzalo D. Mechanism of bacterial inactivation by (+)-limonene and its potential use in food preservation combined processes. PLoS One 2013; 8(2): e56769.
[http://dx.doi.org/10.1371/journal.pone.0056769] [PMID: 23424676] ].

According to Prado et al. (2013) [44Prado ACP, Manion BA, Seetharaman K, Deschamps FC, Barrera Arellano D, Block JM. Relationship between antioxidant properties and chemical composition of the oil and the shell of pecan nuts. Ind Crops Prod 2013; 45: 64-73. [Caryaillinoinensis (Wangenh) C. Koch].
[http://dx.doi.org/10.1016/j.indcrop.2012.11.042] ], Wang et al. (2008) [45Wang YS, He HP, Yang JH, Di YT, Hao XJ. New monoterpenoid coumarins from Clausena anisum-olens. Molecules 2008; 13(4): 931-7.
[http://dx.doi.org/10.3390/molecules13040931] [PMID: 18463594] ], Sartoratto et al. (2004) [46Sartoratto A, Machado ALM, Delarmelina C, Figueira GM, Duarte MCT, Rehder VLG. Composition and antimicrobial activity of essential oils from aromatic plants used in Brazil. Braz J Microbiol 2004; 35: 275-80.
[http://dx.doi.org/10.1590/S1517-83822004000300001] ] and Michielin et al. (2009) [47Michielin EMZ, Salvador AA, Riehl CAS, Smânia A Jr, Smânia EFA, Ferreira SRS. Chemical composition and antibacterial activity of Cordia verbenacea extracts obtained by different methods. Bioresour Technol 2009; 100(24): 6615-23.
[http://dx.doi.org/10.1016/j.biortech.2009.07.061] [PMID: 19683436] ], it is possible to classify the antimicrobial agents based on MIC values. Michielin et al. (2009) [47Michielin EMZ, Salvador AA, Riehl CAS, Smânia A Jr, Smânia EFA, Ferreira SRS. Chemical composition and antibacterial activity of Cordia verbenacea extracts obtained by different methods. Bioresour Technol 2009; 100(24): 6615-23.
[http://dx.doi.org/10.1016/j.biortech.2009.07.061] [PMID: 19683436] ] classified the extracts as strong inhibitors with MIC up to 500 μg.mL^-1, moderate inhibitor for MIC from 600 to 1500 μg.mL^-1, and low inhibitor for MIC over 1600 μg.mL^-1, allowing positioning the bergamot peel oils obtained in this work as a strong inhibitors agent against Gram-positive and Gram-negative bacteria.

It should be mentioned that many compounds present in the essential oil from vegetable matrices, e.g. limonene, becomes effective in combination with heat in the inactivation of some bacteria such as E. coli and Listeria monocytogenes [42Ait-Ouazzou A, Cherrat L, Espina L, Lorán S, Rota C, Pagán R. The antimicrobial activity of hydrophobic essential oil constituents acting alone or in combined processes of food preservation. Innov Food Sci Emerg Technol 2011; 12: 320-9.
[http://dx.doi.org/10.1016/j.ifset.2011.04.004] ] due to a synergistic effect on pathogens inactivation, making them able for use in food preservation, keeping the organoleptic properties of the fresh food products [43Espina L, Gelaw TK, de Lamo-Castellví S, Pagán R, García-Gonzalo D. Mechanism of bacterial inactivation by (+)-limonene and its potential use in food preservation combined processes. PLoS One 2013; 8(2): e56769.
[http://dx.doi.org/10.1371/journal.pone.0056769] [PMID: 23424676] ].