The anti-biofilm activity of lemongrass (Cymbopogon flexuosus) and grapefruit (Citrus paradisi) essential oils against five strains of Staphylococcus aureus

Authors


Correspondence

Carol A. Phillips, School of Health, The University of Northampton, Boughton Green Road, Northampton NN2 7AL, UK. E-mail: carol.phillips@northampton.ac.uk

Abstract

Aims

To determine the sensitivity of five strains of Staphylococcus aureus to five essential oils (EOs) and to investigate the anti-biofilm activity of lemongrass and grapefruit EOs.

Methods and Results

Antimicrobial susceptibility screening was carried out using the disk diffusion method. All of the strains tested were susceptible to lemongrass, grapefruit, bergamot and lime EOs with zones of inhibition varying from 2·85 to 8·60 cm although they were resistant to lemon EO. Lemongrass EO inhibited biofilm formation at 0·125% (v/v) as measured by colorimetric assay and at 0·25% (v/v) no metabolic activity was observed as determined by 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT) reduction. Grapefruit EO did not show any anti-biofilm activity. Following exposure to lemongrass EO extensive disruption to Staph. aureus biofilms was shown under scanning electron microscopy.

Conclusions

In comparison to the other EOs tested, lemongrass exhibited the most effective antimicrobial and anti-biofilm activity.

Significance and Impact of the Study

The effect of lemongrass EO highlights its potential against antibiotic resistant Staph. aureus in the healthcare environment.

Introduction

Staphylococcus aureus is an important human pathogen whose ability to acquire resistance mechanisms and other pathogenic determinants has added to its emergence in both acute and community healthcare settings (Said-Salim et al. 2005; Zetola et al. 2005). Although the majority of the Staph. aureus infections described in community settings has been associated with methicillin susceptible Staph. aureus (MSSA) strains (Said-Salim et al. 2005) rather than methicillin resistant Staph. aureus (MRSA) strains, infections caused by the latter are linked with higher subsequent treatment costs than those caused by MSSA strains (Gordon and Lowy 2008). Over 50% of healthcare associated MRSA strains are known to be resistant to both β-lactam and non-β-lactam antibiotics while the community acquired (CA) Staph. aureus strains are susceptible to non-β-lactam antibiotics (Fey et al. 2003; Weber 2005; King et al. 2006). The emergence of CA infections especially CA MRSA strains is of particular importance as they possess molecular characteristics absent in hospital MRSA strains highlighted in a review by Chambers and Deleo (2009). These characteristics include; a staphylococcal chromosomal cassette mec (SCCmec) type IV allele and genes which encode the virulence factor Panton Valentine-Leukocidin (PVL) (King et al. 2006). PVL is a cytotoxin which causes leucocyte destruction and tissue necrosis (Weber 2005; Loughrey et al. 2007), and both MSSA and MRSA strains from community settings can carry the PVL gene (Rasigade et al. 2010).

Biofilms have been defined by Donlan and Costerton (2002), as immobile communities of organisms attached to a substratum or to each other, embedded in a matrix of extracellular polymeric substances and showing an altered phenotype. Biofilm formation has been identified in Pseudomonas sp. (Klausen et al. 2003; Ghafoor et al. 2011), Candida sp. (Ramage et al. 2002; Silva et al. 2011), Enterococcus sp. (Seno et al. 2005; Mohamed and Huang 2007) and Staph. aureus (Donlan and Costerton 2002; Knobloch et al. 2002; Yarwood et al. 2004). The ability to form biofilms is also known to provide the organisms with protection against antibiotics (Yarwood et al. 2004).

Interest in natural antimicrobials has grown in recent years and the most important and well researched of these are plant products which have many medicinal and antimicrobial properties (Bourne et al. 1999; Cowan 1999). Essential oils (EOs) extracted from plants have been shown to possess antimicrobial activity in in vitro assays against a range of bacteria including known antibiotic resistant strains (Fisher and Phillips 2006; Warnke et al. 2009). EOs have been used as topical antimicrobials (Barker and Altman 2010; Dai et al. 2010), as dental and oral treatments (Jeon et al. 2011; Palombo 2011), and for burns and wound healing (Edwards-Jones et al. 2004; Thakur et al. 2011). Recently, the use of EOs in vapour phase has also been shown to be anti-bacterial and anti-fungal as reviewed by Laird and Phillips (2012).

In recent years, reports of studies on the anti-biofilm activity of EOs have been increasing. For example, cinnamon EO against Candida sp. (Pires et al. 2011) a citrus EO in vapour phase against Enterococcus faecium (Laird et al. 2012) and lemongrass EO against biofilm formation in Listeria monocytogenes (De Oliveira et al. 2010). The resistance of biofilm-associated organisms is estimated at 50–500 times more than planktonic cells (Jabra-Rizk et al. 2006). The effect of different EOs on biofilms has been investigated with studies showing sufficient eradication of biofilms of Pseudomonas aeruginosa (Kavanaugh and Ribbeck 2012) and Staph. aureus and Staphylococcus epidermidis (Nuryastuti et al. 2009; Nostro et al. 2007) at similar concentrations as corresponding planktonic cells. In other in vitro studies, there was increased activity in biofilms exposed to EOs when compared to their effect on cells in suspension (Al-Shuneigat et al. 2005; Karpanen et al. 2008).

The aim of this present study was to determine the anti-staphyloccocal activity of a range of EOs against five strains which included those of CA origin using in vitro screening assays and to further investigate the anti-biofilm effect of the EOs found to be the most effective after initial screening.

Materials and methods

EOs and components

The EOs used in this study were lemongrass (Cymbopogon flexuosus), grapefruit (Citrus paradisi), lime (Citrus aurantifolia), bergamot (Citrus bergamia) and lemon (Citrus limon) obtained from Belmay Plc., Northampton, UK. Two known EO components limonene (Sigma-Aldrich, Dorset, UK) and citral (95%, natural; SAFC Supply Solutions, St Louis, MO, USA) were also investigated.

Micro-organisms

The test micro-organisms used in this study included: three MSSA strains (a hospital MSSA isolate, MSSA NCTC 13297 and a CA PVL positive MSSA) and two MRSA strains (one CA MRSA MW2, and a PVL positive CA MRSA strain). MSSA NCTC 13297 was obtained from Health Protection Agency, London, UK, while the other strains were provided by Professor Mark Fielder at Kingston University, UK. Stock cultures were stored on beads at −80°C. Working cultures were maintained on Brain Heart Infusion (BHI) agar, sub-culturing weekly for a maximum of 3 weeks, to maintain viability and colony characteristics. For inoculum preparation, single colonies were picked from a BHI agar plate into BHI broth (CM225; Oxoid Ltd, Basingstoke, UK) and incubated overnight at 37°C. Enumeration was on BHI agar (CM0375; Oxoid).

Disc diffusion

The screening method was adapted from Prabuseenivasan et al. (2006) and the British Society for Antimicrobial Chemotherapy (BSAC) guidelines Version 10 (Andrews and Howe 2011). Briefly, 100 μl of each of the EOs was deposited onto sterile 2 cm diameter filter paper discs placed on the surface of BHI plates previously spread with 100 μl of a 108 CFU ml−1 overnight culture. The plates were left to dry for 15 min in a sterile environment, inverted and incubated at 37°C for 24 h. The diameters of zones of inhibition (ZOI) were measured using Vernier calipers. The controls were bacterial cultures without EO exposure.

Minimum inhibitory and minimum bactericidal concentrations

The method used was adapted from Hammer et al. (1998). An aliquot (20 μl) of a 108 CFU ml−1 overnight culture was added to wells of a sterile 96-well micro-titre plate. Each EO was diluted in BHI broth containing 0·5% (v/v) Tween 20 and added to wells to give final EO concentrations of 0·03, 0·06, 0·12, 0·5, 1, 2 and 4% (v/v). The positive control wells contained BHI broth and cells without EOs while the negative control wells contained BHI only. Optical density (OD) was measured at 595 nm using a microplate reader (Bio-Rad 680XR, Hertfordshire, UK) and again after incubation for 24 h at 37°C. The Minimum Inhibitory Concentration (MIC) was determined as the lowest EO concentration at which the OD at 24 h of the inoculum remained the same or reduced compared with the initial reading.

For Minimum Bactericidal Concentration (MBC) determination, 10 μl was taken from each well after incubation and spot inoculated (Hammer et al. 1998) onto BHI agar and incubated for 24 h at 37°C. The concentration at which no growth was observed on subculture was determined as the MBC.

Minimum biofilm inhibitory concentration (MBIC)

Inhibition of biofilm formation was assessed using a method adapted from Nostro et al. (2007). An aliquot (100 μl) from an overnight culture diluted in BHI broth supplemented with 1% (w/v) glucose to 108 CFU ml−1 was dispensed into each test well of a 96 well plate. In all, 100 μl of the EO concentrations 0·06–4% (v/v) for lemongrass EO and 1–4% (v/v) for grapefruit EO were added into the wells. The negative control was BHI broth only whereas the positive control contained cell cultures alone with no added EO. Following 24 h incubation at 37°C, the contents of the wells were decanted and each well gently rinsed twice with 300 μl of sterile phosphate buffered saline (PBS) (pH: 7·3 ± 0·3). The plates were air dried for 30 min, stained with 0·1% (w/v) crystal violet for 30 min at room temperature (Wijman et al. 2007), washed three times with PBS (200 μl per well) and dried. The crystal violet was then solubilized using 10% (v/v) glacial acetic acid and the OD measured at 595 nm using a Microplate reader (Bio-Rad 680XR). The MBIC was determined as the EO concentration at which the OD ≤ negative control (Pettit et al. 2005; Sandoe et al. 2006). Each experiment was performed in quadruplicate and performed on four separate occasions.

Minimum biofilm eradication concentration (MBEC)

The method used was similar to that described by Kwiecinski et al. (2009). After biofilm formation for 48 h, the medium was discarded and the wells gently rinsed twice with PBS. A total of 200 μl of the EOs (lemongrass or grapefruit) were serially diluted and added into the wells ranging from 0·06 to 4% (v/v) for lemongrass EO and 1 to 4% (v/v) for grapefruit EO. The plates were then incubated for 24 h at 37°C after which the wells were washed with PBS and stained using the Crystal Violet (CV) staining method as described previously. The positive control was biofilm without EO. The concentration at which already established biofilms were removed from the bottom of the treated wells was determined as the MBEC (Muli and Struthers 1998; Ceri et al. 1999).

Biofilm metabolism assay – XTT reduction

This method is based on reduction of tetrazolium salt XTT [2, 3-bis (2-methyloxy-4-nitro-5-sulphophenyl)-2H-tetrazoluim-5-carboxanilalide] and was performed to determine the metabolic activity of the biofilm formed using methods described by Cerca et al. (2005) and Laird et al. (2012). Stock solutions of XTT in PBS (5 mg ml−1) and menadione (1 mmol l−1) were prepared. At the start of each experiment, a fresh solution of XTT/menadione was prepared at a ratio of 12·5/1 v/v. Biofilms were formed for 48 h in the wells of 96 well plates and 200 μl of the XTT/menadione mix was added into each test and control well, incubated in the dark at 37°C for 24 h and the OD measured at 450 nm.

Biofilm viability assay (CFU ml−1)

Biofilm viability was measured using a method adapted from Pettit et al. (2005). Following 24 h exposure of biofilms to lemongrass or grapefruit EO, a sterile scraper was used to dislodge each biofilm into the micro-titre wells, and 100 μl of the well contents removed and spread onto BHI agar. Plates were incubated for 24 h at 37°C before enumeration.

Gas chromatography mass spectrometry

GC/MS analysis of the EOs was performed using the TurboMass (Perkin-Elmer, Buckinghamshire, UK) instrument with column 1 stationary phase Rtx 1 column (60 m × 0·25 mm i.d.; film thickness: 0·25 μm; Restek). The oven temperature programme was: initial temperature of 50°C; increasing by3°C min−1 to 265°C and held for 13 min. Helium was used as the carrier gas with a 1 μl injector volume, an injector temperature of 285°C and a split ratio 30 : 1. The MS was performed with an EI+ source and operated in scan mode, from 35 to 350 m/z at a detector temperature of 300°C. The compounds were identified by comparing retention times and mass spectra with those of standards or their retention indices with published data and their mass spectra with the National Institute of Standards and Technology (NIST) library.

Scanning electron microscopy (SEM)

Following preliminary investigations of biofilm formation, PVL CA MSSA was selected for SEM observations due to its higher biofilm OD values compared with the other strains (data not shown). In all, 2 cm diameter sterile stainless steel discs (Goodfellows Cambridge Ltd, Huntingdon, UK) were immersed in six well plates (Nunclon Surface, Roskilde, Denmark) containing 5 ml of BHI broth supplemented with 1% (w/v) glucose for 48 h. A total of 100 μl of a 108 CFU ml−1 overnight culture was then added and the plates incubated for 48 h in a shaking incubator. After incubation the discs were removed and gently rinsed with sterile PBS to remove loosely attached cells and re-suspended in 0·125, 0·5 and 1% (v/v) lemongrass or 4% (v/v) grapefruit EO. After exposure to the EOs, the discs were washed three times with PBS and fixed with 2·5% (v/v) glutaraldehyde in PBS solution for 2 h at 4°C, washed twice with PBS and dehydrated for 10 min using a graded ethanol series; 30, 50, 70, 90, 100% (v/v). The samples were then dried prior to coating with gold and observed using a Hitachi S-3000 Scanning Electron Microscope (Hitachi High-Technologies Europe, Maidenhead, UK).

Statistical analysis

Statistical analysis was conducted using spss version 17.0 (IBM, Armonk, NY, USA). Significance levels was set at P = 0·05. After assumptions of normality and variances of homogeneity were checked, one way analysis of variance (anova) was performed.

Results

Screening

No ZOIs were observed with either lemon EO or limonene (Table 1) whereas lemongrass EO and citral, the major component in lemongrass EO (Table 2), both completely cleared the plate of bacterial growth with a ZOI of >8·60 cm. Grapefruit, lime and bergamot EOs produced ZOIs ranging from 2·85 to 4·63 cm (Table 1). Consequentially, lemongrass and grapefruit EOs were selected for determination of MICs and MBCs and for anti-biofilm activity, as these were the most effective at >8·60 and 3·48 cm respectively (Table 1). Although ZOI produced by the latter was only marginally more effective than lime EO (ZOI = 3·47 cm), it was selected for further studies as it has been shown to have anti-bacterial activity and potential in other antimicrobial applications (Williams et al. 2007; Uysal et al. 2011).

Table 1. Zones of inhibition (cm) ± SE measured after exposure to essential oils and components against Staphylococcus aureus strains
 LemongrassGrapefruitLimeBergamotCitralLemonLimonene
  1. CA, community acquired; MRSA, methicillin resistant Staph. aureus; MSSA, methicillin susceptible Staph. aureus; PVL, Panton Valentine-Leukocidin.

Hospital MSSA8·6 (±0)3·96 (±0·17)4·17 (±0·18)3·36 (±0·07)8·6 (±0)0 (±0)0 (±0)
PVL CA-MSSA8·6 (±0)3·23 (±0·06)2·90 (±0·17)2·89 (±0·07)8·6 (±0)0 (±0)0 (±0)
MSSA NCTC 132978·6 (±0)3·35 (±0·07)3·27 (±0·05)3·12 (±0·11)8·6 (±0)0 (±0)0 (±0)
CA-MRSA (MW2)8·6 (±0)3·43 (±0·04)3·57 (±0·12)2·85 (±0·09)8·6 (±0)0 (±0)0 (±0)
PVL CA-MRSA8·6 (±0)3·42 (±0·11)3·58 (±0·08)2·85 (±0·09)8·6 (±0)0 (±0)0 (±0)
Mean (cm)8·603·483·473·018·600·000·00
Table 2. GC/MS analysis of the essential oils showing major components
Components Cymbopogon flexuosus Citrus aurantifolia Citrus paradisi Citrus bergamia
  1. *Neral is the Z-isomer of Citral (also known as Citral B) and geranial is the E-Isomer of Citral (Citral A).

Alpha Pinene0·202·300·501·30
Methyl heptenone2·00
Myrcene1·002·321·00
Limonene0·4047·3093·5038·50
Linalool1·500·1511·40
Citronellal1·00
Neral*33·002·000·20
Geranial*47·002·900·40
Geranyl acetate1·50
Beta carypohyllene4·001·00
Beta pinene22·700·407·20
Alpha terpinene0·25
Para cymene0·250·50
Gamma terpinene7·506·00
Trans alpha bergamotene1·10
Beta bisabolene2·70
Decanal0·26
Nootkatone0·10
Alpha thujene0·30
Sabinene1·10
Terpinolene0·30
Linalyl acetate27·90

The MIC for lemongrass EO at 0·06% (v/v) for all strains tested was lower than that for grapefruit EO at 0·5% (v/v) for MSSA NCTC 13297 and 1% (v/v) for the other strains. Similarly, the MBC for lemongrass EO was the same for all the strains tested (0·125% v/v), while for grapefruit the MBC was 2% (v/v) for the hospital MSSA strain and MSSA NCTC 13297, and 4% (v/v) for the three community strains; PVL CA MSSA, CA-MRSA (MW2) and PVL CA MRSA (Table 3).

Table 3. Minimum inhibitory concentration, minimum bactericidal concentration, minimum biofilm inhibitory concentration (MBEC) and minimum biofilm eradication concentration (MBEC) (% v/v) for lemongrass essential oil (EO) and grapefruit EO against Staphylococcus aureus strains
 Lemongrass EOGrapefruit EO
MIC (%)MBC (%)MBIC (%)MBEC (%)MIC (%)MBC (%)MBIC (%)MBEC (%)
  1. CA, community acquired; MRSA, methicillin resistant Staph. aureus; MSSA, methicillin susceptible Staph. aureus; PVL, Panton Valentine-Leukocidin.

Hospital MSSA0·060·1250·06>412>4>4
PVL CA-MSSA0·060·1250·125>414>4>4
MSSA NCTC 132970·060·1250·125>40·52>4>4
MRSA MW20·060·1250·125>414>4>4
PVL CA-MRSA0·060·1250·125>414>4>4

Lemongrass EO prevented biofilm formation at 0·06% (v/v) for the hospital MSSA strain and 0·125% (v/v) for the other strains tested (Table 3) which, for four of the five strains were the same concentration as the MBC. However, lemongrass EO did not remove already formed biofilms (MBEC) at any of the concentrations tested, i.e., 0·06–4% (v/v). Grapefruit EO did not either prevent biofilm formation or remove already formed biofilms at 1–4% (v/v) (Table 3).

Inhibition of metabolic activity occurred in the presence of lemongrass EO after 24 h for all five Staph. aureus strains at 0·125 and 0·06% (v/v) with no significant difference in the reduction brought about by these two concentrations (Fig. 1a). At 0·25% (v/v), no metabolic activity was observed (results not shown). Grapefruit EO did not reduce the metabolic activity as measured by the XTT assay after 24 h incubation. When the effect of grapefruit EO was compared for four strains (not including PVL CA MSSA), there was no statistical significant difference as determined by the one-way Anova for 2% (P = 0·390) and 1% (P = 0·259) EO, although at 4% there was a statistical significant difference (P = 0·039) with PVL CA MRSA having a higher metabolic activity than the other three strains.

Figure 1.

Changes in metabolic activity following 24 h exposure of biofilms of Staphylococcus aureus strains to (a) lemongrass essential oil (EO) (□ 0·125%, image 0·06%, ■, control) and (b) grapefruit EO (image 1%, image 2%, □ 4%, ■ control) as determined by the XTT assay (control = biofilms not exposed to EO; N = 4 for each treatment and for each strain).

There was a significant difference between the metabolic activity of PVL CA MSSA and the other strains at all the concentrations of grapefruit EO tested (at 4%, P = 0·036; at 2%, P = 0·038; at 1%, P = 0·012) with PVL CA MRSA having a metabolic activity approximately 2·7 times that of the control compared to the other four strains (Fig. 1b).

Following lemongrass EO treatment for 24 h, biofilms showed total loss of viability at concentrations of 0·125 – 4% (v/v) dependent on strain (results not shown), although at 0·06 (v/v) some viable cells were recovered (Fig. 2a) there was no significant difference between the treated and control viable counts (P = 0·57) and with no significant differences between the results for the five strains (P = 0·49). Grapefruit EO treated biofilms showed no reduction in viability at any of the concentrations tested (Fig. 2b).

Figure 2.

Effects of (a) lemongrass essential oil (EO) (□ 0·125%, image 0·06%, ■ control) and (b) grapefruit EO (image 1%, image 2%, □ 4%, ■ control) on the relative biofilm viability of Staphylococcus aureus strains following 24 h exposure as determined by the CFU ml−1 assay (control = biofilms not exposed to EO; N = 4 for each treatment and for each strain).

When biofilms were treated with 2 and 1% grapefruit EO, there was no statistically significant difference between the viable counts obtained for all strains (P value at 2% = 0·70; P value at 1% = 0·72). At 4%, there was a statistical significant difference between the strains (P = 0·049) which can be attributed to the larger variation observed for the hospital MSSA strain although there was no statistical differences between the other four strains (P = 0·45).

Following biofilm quantification, the PVL CA MSSA strain consistently showed increased biofilm formation compared to the other strains tested, and therefore was chosen for SEM. After 24 h exposure to lemongrass EO, the control (Fig. 3a) showed intact biofilm structure, and at 0·125% (v/v) (Fig. 3b) it was observed that the integrity of the biofilm structure was disrupted. At 0·5% (v/v) lemongrass EO, there was evident damage on the biofilm structure (Fig. 3c) and at 1% (v/v) of lemongrass EO treatment, no biofilms were observed on the discs although biofilm debris remained (Fig. 3d). When PVL CA MSSA was treated with 4% (v/v) grapefruit EO, no effect on biofilm formation and integrity was observed in comparison to the control (results not shown).

Figure 3.

Scanning electron micrographs of Panton Valentine-Leukocidin community acquired methicillin susceptible Staphylococcus aureus following treatment with lemongrass essential oil at (a) 0% (control), (b) 0·125% (c) 0·5% and (d) 1% (v/v) after 24 h exposure (magnification ×5000, Scale 10 μm). Arrows indicate biofilm formation (a, b), biofilm disruption (c) and biofilm debris (d).

Discussion

Lemongrass EO at low concentrations between 0·03 and 0·06% (v/v) was effective at inhibiting the growth of all five Staph. aureus strains, and at 0·125% (v/v) the effect of lemongrass EO was bactericidal. The results presented here are consistent with those of a previous study (Barbosa et al. 2009) in which it was demonstrated that lemongrass EO inhibited the growth of Gram positive bacteria, including Staph. aureus at a concentration of 0·05% (v/v). In this present study the MIC for grapefruit EO was higher than that for lemongrass EO for all the strains, i.e., between 0·5 and 2% (v/v), while bactericidal activity was observed between 2 and 4% (v/v) EO. To date, there are very few studies that have investigated the antimicrobial activity of grapefruit EO although it has been shown to possess both anti-fungal and anti-bacterial activity (Viuda-Martos et al. 2008; Uysal et al. 2011).

When the effects of the components were compared to the overall effect of the EO, contrasting results were observed. First, citral, the major component in lemongrass EO also showed a ZOI of 8·6 cm (i.e., no growth on plate) at screening and similar MIC profiles to lemongrass EO (data not shown) suggesting that citral may be responsible for the majority of the anti-bacterial activity. This high activity by citral has been previously reported (Hayes and Markovic 2002; Da Silva et al. 2008; Aiemsaard et al. 2011). Limonene is the major component in the grapefruit EO used in this study, at approximately 94%, but it did not show any antimicrobial effect as demonstrated by the screening results which has also been observed in previous studies by Fisher and Phillips (2006) and Inouye et al. (2001). In comparison to this lack of activity by limonene, grapefruit EO produces inhibition zones in the Staph. aureus strains between 3·23 and 3·96 cm which suggests that other components of the grapefruit EO are involved in the anti-bacterial activity observed in this study. Onawunmi et al. (1984) observed that myrcene showed no anti-bacterial activity per se, but enhanced activity of other EO components when in combination which suggests that the presence of other components in small amounts could enhance the EO antimicrobial activity. Although the individual components of EOs are important, they act in a synergistic manner so that the EO exhibits a greater anti-bacterial activity than the sum of that brought about by its components (Gill et al. 2002).

The results of this present study demonstrate that lemongrass EO possesses anti-biofilm activity at low concentrations between 0·06 and 0·125% (v/v) which has been reported previously (Aiemsaard et al. 2011). As biofilm formation is a survival mechanism but also contributes to virulence and persistence (Vuong et al. 2004; Soto et al. 2006), it has been suggested that preventing biofilm attachment is a way of dealing with the problem of biofilms in the food industry (Sinde and Carballo 2000). Therefore considering the results presented here, there may be a possible potential for lemongrass EO use in food processing environments. The effect on the organoleptic properties of the foodstuff at the anti-biofilm concentrations would need to be determined, although lemongrass per se is generally recognized as safe and is used as a food ingredient world-wide.

To our knowledge, this is the first time the anti-biofilm activity of grapefruit EO has been reported. The results described here demonstrate that although grapefruit EO is bactericidal at 2–4% (v/v) against the different Staph. aureus strains tested, it has limited or no activity against biofilm formation. This suggests that biofilm formation could offer protection against EOs, or at least against grapefruit EO. Previous studies have shown however that when grapefruit EO was combined with other EOs against MRSA (not in biofilms), there was synergistic activity and improved antimicrobial potential (Edwards-Jones et al. 2004) hence, combining grapefruit EO with other EOs or antimicrobial compounds might also enhance its activity against biofilms. The synergistic action of EOs against surface adhered cells has previously been demonstrated by the results of a study by De Oliveira et al. (2010) who reported a 100% log reduction of a mature L. monocytogenes biofilm after 60 min contact time with a combination of lemongrass and citronella EOs.

Both lemongrass and grapefruit EOs were unable to eradicate already established biofilms (Table 2). The inability of antimicrobial compounds to remove biofilm deposits has been observed previously (Lin et al. 2011). As biofilms develop, the pioneer cells undergo irreversible attachment leading up to maturation (Mittelman 1998) and at this point, removal of biofilms is said to be difficult and would require mechanical force or chemical disruption (De Oliveira et al. 2010). In addition, Pitts et al. (2003), after an investigation of reductions in Ps. aeruginosa and Staph. epidermidis biofilms using a range of chemical agents such as hydrogen peroxide and 1 mol l−1 sodium chloride, suggest that such reductions are micro-organism and antimicrobial agent specific further highlighting the difficulty with regard to biofilm removal. For example, 1 mol l−1 sodium chloride significantly reduces Ps. aeruginosa biofilms but not those of Staph. epidermidis, while hydrogen peroxide was more effective against the latter than it was against the former (Pitts et al. 2003). It has been suggested (Kelly et al. 2012) that biofilm prevention is preferable to disruption and removal due to factors such as the nature of biofilms, adherence of staphylococci to surfaces and treatment problems associated with biofilms.

Biofilms tolerate high amounts of antibiotic between 10–1000 fold when compared to planktonic cells (Yarwood et al. 2004; Resch et al. 2006; Kelly et al. 2012), and in this study, lemongrass at twice the MIC and same concentration as the MBC prevented biofilm formation highlighting antimicrobial activity as well as its potential as an anti-biofilm agent. Exposure of the biofilms to grapefruit EO showed no reduction in metabolic activity in four of the five Staph. aureus strains and an increased metabolic activity in the PVL positive MSSA strain. The reasons for such increase in metabolic activity are unknown. However, a study by Kwiecinski et al. (2009) using the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] colorimetric assay reported an increase in metabolism within Staph. aureus biofilms when treated with tea tree oil (TTO) concentrations lower than the MBEC which, it was suggested, could be a result of a stress response. Whether this is the case for the increase in metabolic activity in PVL positive MSSA strain when treated with grapefruit EO remains to be determined.

Lemongrass EO at the MBC and MBIC concentrations disrupts PVL CA MSSA biofilms and at 0·5% (v/v), and at 1% (v/v), total destruction of the biofilm was observed. Kwiecinski et al. (2009) has suggested that TTO treatment of Staph. aureus biofilms causes damage to the extra cellular matrix and damage to the biofilm structure was observed on treatment with 1% (v/v) lemongrass EO suggesting a similar mode of action. With the grapefruit EO, SEM analysis did not show any disruption of the biofilm structure which further confirms the lack of anti-biofilm activity and indicates the importance of biofilm formation as a protective mechanism against EOs.

This study is one of few that have investigated the anti-biofilm properties of EOs especially the effect of lemongrass EOs against Staph. aureus biofilms. Where previously (Bearden et al. 2008), investigated commercial formulations containing EOs against CA MRSA, this is first study that has demonstrated the anti-biofilm activity of lemongrass EO against biofilms of CA MSSA and MRSA including PVL positive strains. Further studies would be required to evaluate any potentially toxic effect of lemongrass EO; however, the antimicrobial and anti-biofilm properties provide another option for future antimicrobial therapeutic interventions in both clinical and industrial applications.

Acknowledgements

E.C.A. acknowledges the financial support provided from NHS Northamptonshire and The University of Northampton, UK through the Centre of Health and Wellbeing Research, for a Postgraduate Studentship. The authors also thank Belmay Plc., Northampton, UK for supplying the oils and for the GC/MS analysis and Professor Mark Fielder, Kingston University UK for supply of the clinical isolates.

Ancillary