Antibiofilm Potential of Lavandula Preparations against Campylobacter jejuni

ABSTRACT

New approaches for the control of Campylobacter jejuni biofilms in the food industry are being studied intensively. Natural products are promising alternative antimicrobial substances to control biofilm production, with particular emphasis on plant extracts. Dried flowers of Lavandula angustifolia were used to produce essential oil (LEO), an ethanol extract (LEF), and an ethanol extract of Lavandula postdistillation waste material (LEW). The chemical compositions determined for these Lavandula preparations included seven major compounds that were selected for further testing. These were tested against C. jejuni for biofilm degradation and removal. Next-generation sequencing was used to study the molecular mechanisms underlying LEO actions against C. jejuni adhesion and motility. Analysis of LEO revealed 1,8-cineol, linalool, and linalyl acetate as the main components. For LEF and LEW, the main components were phenolic acid glycosides, with flavonoids rarely present. The MICs of the Lavandula preparations and pure compounds against C. jejuni ranged from 0.2 mg/ml to 1 mg/ml. LEO showed the strongest biofilm degradation. The reduction of C. jejuni adhesion was ≥1 log₁₀ CFU/ml, which satisfies European Food Safety Authority recommendations. Lavandula preparations reduced C. jejuni motility by almost 50%, which consequently can impact biofilm formation. These data are in line with the transcriptome analysis of C. jejuni, which indicated that LEO downregulated genes important for biofilm formation. LEW also showed good antibacterial and antibiofilm effects, particularly against adhesion and motility mechanisms. This defines an innovative approach using alternative strategies and novel targets to combat bacterial biofilm formation and, hence, the potential to develop new effective agents with biofilm-degrading activities.

IMPORTANCE The Lavandula preparations used in this study are found to be effective against C. jejuni, a common foodborne pathogen. They show antibiofilm properties at subinhibitory concentrations in terms of promoting biofilm degradation and inhibiting cell adhesion and motility, which are involved in the initial steps of biofilm formation. These results are confirmed by transcriptome analysis, which highlights the effect of Lavandula essential oil on C. jejuni biofilm properties. We show that the waste material from the hydrodistillation of Lavandula has particular antibiofilm effects, suggesting that it has potential for reuse for industrial purposes. This study highlights the need for efforts directed toward such innovative approaches and alternative strategies against biofilm formation and maintenance by developing new naturally derived agents with antibiofilm activities.

INTRODUCTION

The food industry is in a constant battle against economic losses caused by the formation and prevalence of bacterial biofilms on food manufacturing surfaces. Moreover, bacterial biofilms, as a persistent form of bacterial lifestyle in food processing environments, are a significant public health problem (1). One of the most prevalent foodborne pathogens and the major cause of human intestinal infections in developed countries is Campylobacter jejuni (2, 3). Human infections occur mainly by consumption of undercooked meat, especially poultry, or food that has been in contact with a contaminated surface (4).

Campylobacter jejuni can form biofilms on different abiotic surfaces used in the food industry (e.g., polystyrene, glass, and stainless steel), which is crucial for its environmental survival under various conditions, and thus remains a significant food-safety challenge in the food industry (5). A major problem is also the widespread antibiotic resistance of C. jejuni to a wide range of antibiotics as well as resistance to various biocides used for disinfection in the food industry (6–8).

Cost-effective chemical and physical methods are used for biofilm eradication in the food industry; however, these show only limited success (9). As biofilms contribute to bacterial pathogenicity and resistance against antibiotics and biocides, innovative strategies are needed to inhibit the bacterial properties involved (e.g., motility and adhesion) or to degrade or remove biofilms as a control against foodborne pathogens. Natural products have been shown to represent promising alternatives that can be effective antimicrobial agents against numerous bacterial targets that are important for biofilm formation and do not create selection pressure for the emergence of further antibiotic-resistant bacteria (10).

One such commonly used crop worldwide belongs to the Lamiaceae family and the genus Lavandula, some strains of which are Lavandula angustifolia Mill (true lavender, Lamiaceae), Lavandula latifolia Medik (spike lavender), and Lavandula × intermedia Emeric ex Loisel (lavandin) (11). These have been used since ancient times for medicinal purposes and more recently in industrial processes (12, 13).

The genus Lavandula grows around the Mediterranean and in southern Europe, northeastern Africa, the Middle East, southwest Asia, and southeast India (14). These aromatic plants have a rich essential oil content, for which sedative, carminative, antidepressive, antispasmodic, antimicrobial, and anti-inflammatory effects have been reported (15, 16). Along with one of their main components, linalool (17, 18), lavender and lavandin essential oils can have strong antibiofilm effects against the bacteria Staphylococcus aureus, Escherichia coli, and Campylobacter spp. (19–21). However, there have not been any studies that have investigated the potential molecular mechanisms of the antibiofilm actions of Lavandula essential oil, and so the present study investigates this.

In addition to alternative strategies using low doses of bioactive phytochemicals to control bacterial biofilms, this study has focused also on the use of by-products of essential oil distillation. More recently, the demand for Lavandula essential oils has grown hugely, which has resulted in greater cultivation and production of these plants. Approximately 200 tons of lavender oil and 1,200 tons of lavandin oil are produced annually, and there has been a concomitant increase in the by-products that are accumulated as waste materials and hydrolates or aromatic waters after essential oil production (22–24). These by-products and especially waste from the agro-food industry also represent a potentially important source of effective phytochemicals (10). For example, waste materials from spike lavender have antioxidant properties (13), and lavender hydrolates have antifungal and antibacterial properties that enhance the flavor and taste of foods (25). Reuse of these waste materials will partially solve one of the major problems—their disposal (26). In addition, the reuse of such a material is environmentally friendly, as the use of disinfectants, which are commonly used in the food industry, can lead to additional unnecessary chemicals in the environment (8). However, current studies have been focused only on planktonic cell suspensions, and there have not been any reports on antibiofilm effects of such Lavandula by-products. This is important, as biofilms are the most prevailing growth lifestyle in bacteria, and it is well known that in biofilms, the effectiveness of antimicrobials is significantly reduced (up to 1,000-fold) compared to that in the planktonic state (27).

In this study, Lavandula preparations and selected pure compounds were used to determine whether they have antibiofilm effects against C. jejuni NCTC 11168. Dried flowers of Lavandula angustifolia were used to produce an essential oil (LEO), an ethanol extract (LEF), and the ethanol extract of Lavandula postdistillation waste material (LEW). First, degradation of C. jejuni biofilms formed on an abiotic surface using these Lavandula preparations was investigated. Furthermore, strategies to target motility and adhesion of C. jejuni were introduced, with the aim to prevent or reduce biofilm formation in its early stages. Finally, an RNA sequencing (RNA-seq) approach was used to better understand the molecular mechanisms underlying the actions of LEO against C. jejuni biofilms and thus indicate innovative Campylobacter control strategies.

RESULTS

Dried flowers of Lavandula were used to produce LEO, the essential oil, LEF, an ethanol extract of flowers prior to distillation, and LEW, an ethanol extract of Lavandula postdistillation waste material. Each of these was initially analyzed for their chemical composition. Then, the main pure compounds were selected and included in the anti-Campylobacter analysis and MIC determinations, with the aim to compare the activities of the Lavandula preparations with those of the pure compounds. This analysis started with the innovative strategy of degradation and removal of these preformed biofilms of C. jejuni NCTC 11168. Second, the first phases of biofilm formation in which C. jejuni adhesion and motility are the predominant factors were targeted. Finally, a molecular level evaluation of the mode of action of LEO was performed.

Phytochemical analysis and identification of flavonoids.

The ethanolic extract of the Lavandula flowers, LEF, and the ethanolic extract of Lavandula postdistillation waste material, LEW, were analyzed for their phenolic compound contents using liquid chromatography-photo diode array-electrospray ionization mass spectrometry. LEF and LEW had phenolic acid glycosides as their major constituents. In addition, rosmarinic acid and the flavone apigenin-7-O-glucoside were detected. Flavonoids only had a minor role in the composition of LEF and LEW. Table 1 summarizes the peaks that were identified for both LEF and LEW. Figure S1 in the supplemental material illustrates a representative UV chromatogram. Compound 4 in Table 1 is most probably identical to the o-coumaric acid-2-O-glucoside previously described for L. angustifolia (28), as indicated by the UV spectral data; however, m-coumaric acid as an aglycon cannot be excluded at this point.

TABLE 1.

Identification of the main common flavonoids in the Lavandula ethanolic extract and in the ethanolic extract of waste material from hydrodistillation

No.	Compound identified	Full scan MS (m/z)	Fragment ions (MS2; m/z)a	UV maximum (nm)b
1	Coumaric acid hexoside I	325 [M−H]−	163 (100), 119 (25)	263, 290 sh
2	Caffeic acid hexoside	387 [M+HCOOH−H]−	341 (100), 207 (25)	302
3	Ferulic acid hexoside I	355 [M−H]−	193 (100), 149 (20)	302
4	Coumaric acid hexoside II	371 [M+HCOOH−H]−, 325 [M−H]−	325 (100)	277, 290 sh
5	Ferulic acid hexoside II	401 [M+HCOOH−H]−, 355 [M−H]−	355 (100)	295, 319
6	Apigenin-7-O-glucoside	431 [M−H]−	269 (100)	268, 334
7	Rosmarinic acid	359 [M−H]−	161 (100), 179 (30), 223 (10)	292 sh, 328

^aNumbers in parentheses are percent relative intensities.

^bsh, shoulder.

According to the gas chromatography-mass spectrometry analysis, the Lavandula essential oil, LEO, was particularly rich in 1,8-cineol, a cyclic monoterpene (36.1%), and the terpene alcohol linalool (15.8%). LEO was also rich in the furanoids cis-linalool oxide (7.9%) and trans-linalool oxide (5.9%). Two other main components were linalyl acetate (6.4%) and the monoterpenoid camphor (5.9%). Other components identified at >1% were α-terpineol, octen-3-yl-1-acetate, lavandulyl acetate and thymol, with the content of borneol close to 1% (Table 2).

TABLE 2.

Identification of the main components of the Lavandula essential oil

Retention time (min.)	Retention indexa	Compound identifiedb	Quantification (% of total)c
5.50	932	α-Pinene	0.27
5.90	946	Camphene	0.17
6.68	975	β-Pinene	0.39
7.79	1012	Hexyl acetate	0.20
8.18	1023	p-Cymene	0.32
8.41	1030	1,8-Cineol	36.11
9.88	1071	cis-Linalool oxide (furanoid)	7.90
10.47	1087	trans-Linalool oxide (furanoid)	5.87
10.95	1100	Linalool	15.79
11.43	1112	Octen-3-yl-1-acetate	1.14
12.65	1142	Camphor	5.93
13.12	1153	Nerol oxide	0.61
13.53	1163	Borneol	0.97
13.90	1172	cis-Linalool oxide (pyranoid)	0.66
14.03	1175	Terpinen-4-ol	0.56
14.60	1189	α-Terpineol	2.19
14.72	1192	Hexyl butanoate	0.44
16.63	1237	Cuminaldehyde plus hexyl-2-methyl butanoate	0.43
17.41	1256	Linalyl acetate	6.37
18.91	1291	Lavandulyl acetate	1.62
18.97	1293	Thymol	1.41
22.00	1365	Neryl acetate	0.32
22.80	1384	Geranyl acetate	0.61
30.57	1579	Caryophyllene oxide	0.97
34.39	1682	α-Bisabolol	0.21

^aLinear retention index relative to n-alkanes on HP-5-MS column.

^bCompounds identified by mass spectral libraries (63–65; laboratory database) and retention index (RI) comparison (63).

^cQuantification by normalization (area percent method) without considering calibration factors.

According to these analyses, 1,8-cineol, linalool, linalyl acetate, camphor, (−)-borneol, p-coumaric acid, and trans-ferulic acid were selected as the individual pure compounds to be included in further analysis.

Anti-Campylobacter activity of Lavandula preparations and pure compounds.

To evaluate the anti-Campylobacter activities of the Lavandula preparations and the seven chosen pure compounds as the main phytochemical compounds of LEO, LEF, and LEW, their MICs against C. jejuni NCTC 11168 were determined (Table 3). LEO and the pure compounds linalool, (−)-borneol, camphor, and 1,8-cineol had medium anti-Campylobacter activities, with MICs against C. jejuni NCTC 11168 of 0.25 mg/ml, with the last two compounds with MICs of 0.5 mg/ml. A stronger anti-Campylobacter effect was seen for linalyl acetate, with a MIC against C. jejuni NCTC 11168 of 0.2 mg/ml. Interestingly, LEW had the same anti-Campylobacter activities as LEF and also with p-coumaric acid and trans-ferulic acid, with MICs of 1 mg/ml.

TABLE 3.

MICs against C. jejuni NCTC 11168 for the Lavandula preparations (LEO, LEF, and LEW) and pure compounds

Treatment	MIC (mg/liter)
Lavandula essential oil	0.25 ± 0.06
Lavandula ethanol extract	1.00 ± 0.25
Lavandula waste material	1.00 ± 0.25
Linalool	0.25 ± 0.06
Linalyl acetate	0.20 ± 0.05
Camphor	0.50 ± 0.12
(−)-Borneol	0.25 ± 0.06
1,8-Cineol	0.50 ± 0.12
p-Coumaric acid	1.00 ± 0.25
trans-Ferulic acid	1.00 ± 0.25

Degradation of preformed biofilms by the Lavandula preparations.

To determine the effects of these Lavandula preparations on C. jejuni biofilms, a novel approach was developed based on culturing C. jejuni at the air-liquid interface for 48 h. After treatment with the Lavandula preparations, this allowed measurement of the relative area of the glass surface covered by the biofilm by using confocal laser scanning microscopy. The images of the untreated cultures showed the biofilm structure of the viable bacteria, which were mainly dispersed into small microcolonies (Fig. 1). The influence of different concentrations of LEO, LEF, and LEW on the biofilm structure and the cell viabilities were determined after 24 h of treatments (Fig. 2). Interestingly, when the biofilms were exposed to LEO, LEF, and LEW, there was significant degradation of the biofilms across all of these treatments. LEO was the most effective, as it effectively removed the bacterial biofilms. LEF and LEW had similar effects for C. jejuni biofilm removal (Fig. 2). After the treatments of the C. jejuni biofilms with LEF and LEW at 8× MIC, the relative surface cover with live cells was 0.016 ± 0.005 to 0.031 ± 0.002, respectively. Surprisingly, for LEO, the same was seen for concentrations as low as 1× MIC and 2× MIC, with a relative surface cover of 0.043 ± 0.005 and almost zero, respectively. Interestingly, even for LEO at 0.5× MIC, the relative cover of the glass surface was affected, with a relative surface cover of 0.38 ± 0.15. Additionally, the lowest concentration of LEW was more effective than LEF, with relative surface cover for LEW at 2× MIC of 0.51 ± 0.09 compared with LEF at 2× MIC of 0.88 ± 0.01.

FIG 1.

Representative fluorescence microscopy images of C. jejuni NCTC 11168 biofilms grown on glass surfaces treated with different Lavandula preparations. (A) Positive control incubated in MH broth with addition of Campylobacter growth supplement for 72 h in a microaerobic atmosphere at 42°C. (B to D) After 48 h of growth, the mature biofilms were exposed to the Lavandula preparations for 24 h: essential oil (LEO) (B), ethanol extract (LEF) (C), and ethanol extract of Lavandula postdistillation waste material (LEW) (D), all here at 2× MIC (see Table 3), under the same conditions as for panel A. Scale bars, 200 μm.

FIG 2.

Relative C. jejuni NCTC 11168 biofilm coverages of the glass surface following exposure to the Lavandula preparations, in comparison to the untreated C. jejuni control (dashed line). After 48 h of growth, the mature biofilms were exposed to the Lavandula preparations for 24 h: LEO (essential oil) at 0.5×, 1×, and 2× MIC; LEF (ethanol extract) and LEW (ethanol extract of Lavandula postdistillation waste material) at 2×, 4×, and 8× MIC. Data are relative values of biofilm cover of the glass surface presented as means ±standard deviations. *, P < 0.05 versus control.

Modulation of Campylobacter adhesion by Lavandula preparations and pure compounds.

Campylobacter jejuni adhesion to an abiotic surface was used as a strategy to target one of the first steps of biofilm formation. The effects of LEO, LEF, and LEW and the pure compounds on adhesion of C. jejuni NCTC 11168 to a polystyrene surface was measured as the numbers of adhered cells (CFU/ml) after 4 h, 8 h, and 24 h of treatment with subinhibitory concentrations (Fig. 3). Different times were selected to show these effects on adhesion to the abiotic surface and for the mechanisms involved, which are important for reversible (4 h and 8 h) and irreversible (24 h) attachment.

FIG 3.

Relative adhesion of C. jejuni NCTC 11168 to a polystyrene surface after 4 h (A), 8 h (B), and 24 h (C) of exposure to the Lavandula preparations and pure compounds (as indicated) at 0.25× MIC in comparison to that of untreated C. jejuni controls (dashed lines). Attached cells were suspended by sonication, and their concentrations were determined by plate counting. Data are relative values of adhered cells as means ±standard deviations. * P < 0.05 versus control. LEO, essential oil; LEF, ethanol extract; LEW, ethanol extract of Lavandula postdistillation waste material.

The relative adhesion of the C. jejuni cells was significantly lower after treatments with all of these Lavandula preparations, with the strongest effects seen after 4 h (Fig. 3). The most effective treatments here were with LEW at 4 h, when the relative adhesion was 0.002 ± 0.001 (Fig. 3A) and 0.04 ± 0.01 for LEF at 8 h (Fig. 3B). These data also indicated that LEO, LEF, and LEW all had comparable effects on C. jejuni adhesion to a polystyrene surface after 24 h of incubation, where the relative adhesion was 0.08 ± 0.07 for all Lavandula preparations (Fig. 3C).

Interestingly, the relative adhesion after the treatments with LEO, LEF, and LEW was comparable and in the same ranges as those for camphor, (−)-borneol and p-coumaric acid at 4 h and 24 h of treatment and for p-coumaric acid and trans-ferulic acid at 8 h of treatment. Also, the other pure compounds, linalool, linalyl acetate, and 1,8-cineol, were more effective after 4 h, where the relative adhesion was 0.21 ± 0.08 for linalool and 0.17 ± 0.05 for both linalyl acetate and 1,8-cineol (Fig. 3A).

Modulation of Campylobacter motility by Lavandula preparations and pure compounds.

As one of the first steps that are important for biofilm formation, cell motility was also targeted. The influence of LEO, LEF, and LEW and the pure compounds was measured on soft agar and compared with untreated cells. Their subinhibitory concentrations all significantly decreased C. jejuni NCTC 11168 motility (Fig. 4). LEO, LEF, and LEW had comparable effects on C. jejuni relative motility, at 0.47 ± 0.03, 0.45 ± 0.10, and 0.45 ± 0.09, respectively, in comparison with the motility of untreated C. jejuni colonies. The greatest antimotility effect of the pure compounds was seen for (−)-borneol, at 0.41 ± 0.03 relative to the motility of the untreated C. jejuni colonies.

FIG 4.

Relative motility of C. jejuni NCTC 11168 after exposure to the Lavandula preparations and pure compounds (as indicated) at 0.25× MIC (see Table 3) in comparison to that of the untreated C. jejuni control (dashed line). The diameters of swarming colonies were measured after 48 h of incubation on soft agar. Data are relative values as means ± standard deviations. *, P < 0.05 versus control. LEO, essential oil; LEF, ethanol extract; LEW, ethanol extract of Lavandula postdistillation waste material.

Effects of LEO on the transcriptome of C. jejuni NCTC 11168.

It is important to evaluate the mechanisms and the potential targets behind these reductions in biofilm-related contamination and for eradication of biofilm-related infections. According to the effects on C. jejuni biofilm degradation and on motility and adhesion in the early phase of biofilm formation, we selected LEO to extend this study to the molecular mechanisms behind these actions against C. jejuni NCTC 11168. For the biofilms, LEO at its MIC was already sufficient, while to effectively reduce C. jejuni motility and adhesion to polystyrene, LEO was effective at subinhibitory concentrations (i.e., 0.25× MIC). Thus, the mechanism of these LEO actions against C. jejuni were evaluated with the major focus on the genes modulated and the cell processes they are involved in.

Campylobacter jejuni NCTC 11168 in exponential growth phase was treated with LEO for 30 min and with a dimethyl sulfoxide (DMSO) control to eliminate any influence of the solvent used for LEO. The transcriptomic analysis was for the differentially expressed genes in comparison to expression in control cultures (P ≤ 0.05), with a view to predicting their generic functions and, of greater interest, their known functions. Strain-specific RNA-seq was analyzed using Ion Torrent technology. The reads were mapped to the annotated Campylobacter jejuni subsp. jejuni reference genome, with the data available at NCBI under BioProject accession number PRJNA57587.

(i) Differential gene expression for C. jejuni NCTC 11168 treated with LEO.

Detailed analysis showed that 326 genes were differentially expressed in C. jejuni NCTC 11168 when treated with LEO, with normalisation to the DMSO control (log₂ fold change [log₂FC] ≥ 1; P ≤ 0.05). Among these, 138 genes were upregulated and 188 genes were downregulated (Table S1).

The LEO treatment upregulated 42 genes that are involved in the synthesis of ribosomal and export proteins in C. jejuni (Table S1; Fig. S2). This led to global changes in gene expression of various pathways, including the tricarboxylic acid (TCA) cycle and those of pyruvate metabolism, nicotinate, and nicotinamide metabolism, folate biosynthesis, aromatic amino acid biosynthesis, and terpenoid and porphyrin biosynthesis (see Fig. S2). The genes in the main efflux system of C. jejuni involved in β-lactam resistance were also upregulated (i.e., cmeA, cmeB, and cmeC) (Fig. S2).

Of the 188 downregulated genes, 73 are involved in synthesis of transmembrane proteins and are important for the transport of different proteins and ions. Many genes involved in the iron uptake system were also downregulated, including the ABC transporters and TonB box (i.e., chuA, chuB, tonB3, tonB2, exbB1, exbB2, Cj0178, and ceuC) (Table S1; Fig. S2). Moreover, four genes involved in stress defense were downregulated (i.e., cstA, hrcA, dcuA, and ppk) (Table S1).

It is worth noting that 24 of the differentially expressed genes are involved in C. jejuni motility and biofilm formation (Table 4). Most of the genes that are involved in flagellar assembly of C. jejuni were downregulated, and 18 of them had a fold change lower than −1 (log₂FC) (Fig. 5).

TABLE 4.

Differentially expressed genes that are important for biofilm formation of C. jejuni NCTC 11168^a

Gene	Log2 fold change	Description
folK	−1.9	2-Amino-4-hydroxy-6-hydroxymethyldihydropteridine pyrophosphokinase, with genes important for flagellar synthesis
motA	−1.1	Flagellar motor proton channel
flaG	−1.0	Flagellar protein
Cj0719c	−1.2	Involved in PLPb homeostasis, with genes important for synthesis of flagella
pseF	−1.3	Pseudaminic acid cytidylyltransferase, catalyzes the final step in the biosynthesis of pseudaminic acid, a sialic-acid-like sugar that is used to modify flagellin
maf4	−1.4	Uncharacterized protein, with flaA gene, which enables synthesis of flagellin
flgG	−1.0	Flagellar basal-body rod protein
flgI	−1.1	Flagellar P-ring protein
Cj1467	−1.4	Uncharacterized protein, with genes important for synthesis of flagella
fliQ	−1.1	Flagellar biosynthetic protein
kpsM	−1.9	Capsule polysaccharide export system inner membrane protein
kpsS	−1.2	Polysaccharide modification protein
murB	−1.0	UDP-N-acetylenolpyruvoylglucosamine reductase, involved in cell wall formation
mrdB	−1.2	RodA protein homolog, involved in cell wall formation
bamD	−2.1	Outer membrane protein
waaC	−1.2	Lipopolysaccharide heptosyltransferase
lgt	−1.5	Prolipoprotein diacylglyceryl transferase
Cj0262	1.0	Methyl-accepting chemotaxis signal transduction protein
lpxA	1.3	Involved in biosynthesis of lipid A
peB3	1.9	Major antigenic peptide, with flaA and flaC genes, which enable synthesis of flagellin
peB2	1.0	Major antigenic peptide

^aDifferential expression was defined as a false discovery rate P value of ≤0.05.

^bPLP, pyridoxal 5′-phosphate.

FIG 5.

Pathway map of the C. jejuni NCTC 11168 flagellar assembly (Cje02040). This assembly showed upregulated (red) and downregulated (green) genes in C. jejuni after exposure to the essential oil LEO at 0.25× MIC for 30 min (see Table 3). Those genes are essential for locomotion and flagellum-dependent motility, which represent two features necessary for biofilm formation of C. jejuni. KEGG map of differentially expressed genes involved in flagellar assembly was constructed using Pathview (70) and the KEGG genome database (https://www.genome.jp/kegg/pathway.html).

(ii) Gene ontology enrichment analysis.

Gene ontology (GO) enrichment analysis of the significant differentially expressed genes (FC ≥ 1, P ≤ 0.05) in these LEO-treated samples compared with expression in controls showed the following two categories(Table 5): (i) biological processes with upregulation that included cellular nitrogen compound biosynthetic processes, tRNA aminoacylation, and nitrogen compound metabolic processes, and with downregulation for locomotion, cilium or flagellum-dependent cell motility, and biological regulation, and (ii) molecular functions with upregulation that included catalytic activity, nucleic acid and organic cyclic compound binding, and structural molecules, and with downregulation for transmembrane transport.

TABLE 5.

Gene ontology categories that are significantly enriched in terms of biological processes and molecular functions differentially regulated between LEO-treated samples and control samples

GO code	Term	No. of genes differentially expresseda	FDR P value	Regulation
Biological process
0040011	Locomotion	5/32	0.054026	Down
0071973, 0097588, 0001539, 0048870, 0006928	Cilium or flagellum-dependent cell motility	3/23	0.012148	Down
0065007	Biological regulation	19/83	0.062751	Down
0044271	Cellular nitrogen compound biosynthetic process	65/183	1.62E−10	Up
0006430	Transfer RNA aminoacylation	2/22	0.006264	Up
0006807	Nitrogen compound metabolic process	87/433	2.16E−08	Up
Molecular function
0046873	Metal ion transmembrane transporter activity	2/16	0.061105	Down
0015103	Organic anion transmembrane transporter activity	4/10	0.085316	Down
0019843	rRNA binding	27/39	1.51E−15	Up
0003735	Structural constituent of ribosome	36/52	1.00E−15	Up
0005198	Structural molecule activity	37/59	1.14E−14	Up
0097159, 1901363	Organic cyclic compound binding	93/436	0.006437	Up
0003676	Nucleic acid binding	54/175	4.16E−05	Up
0000049	tRNA binding	6/18	0.009744	Up
0005488	Binding	113/588	0.048065	Up
0140101	Catalytic activity, acting on a tRNA	2/43	0.048065	Up
0004812, 0016875	Aminoacyl-tRNA ligase activity	2/22	0.005088	Up
0140098	Catalytic activity, acting on RNA	9/64	0.069198	Up

^aAbsolute log₂ fold change of ≥1; false discovery rate [FDR] P value of ≤0.05 relative to total number of genes in enriched category.

DISCUSSION

In the present study, Lavandula preparations (i.e., LEO, LEF, and LEW) were used, with the aim to modulate the Campylobacter properties that are responsible for the initial phase of the multifactorial event of biofilm formation on abiotic polystyrene and glass surfaces. In doing so, we used a new approach of culturing and studying biofilms formed at the air-liquid interface. Also, we investigated LEO, LEF, and LEW as effective control strategies for biofilm degradation and removal under conditions that do not provide selection pressure for further development of antibiotic resistance of Campylobacter spp.

The occurrence of persistent and resistant C. jejuni cells in their biofilm form is associated worldwide with severe infections and gastrointestinal diseases in humans (4). The attachment of biofilms to different abiotic surfaces is responsible for major problems in the food industry, as the reversible binding turns into irreversible bonding and biofilm formation, with possible further dispersion and contamination of the food matrix (9, 29). Therefore, alternative strategies to either remove preformed biofilms or inhibit biofilm formation based on alternative antimicrobial agents of natural origin represent an innovative control approach, as we have described here.

For this purpose, dried flowers of Lavandula angustifolia were used to produce LEO, the essential oil, LEF, the ethanol extract of flowers prior to distillation, and LEW, the ethanol extract of Lavandula postdistillation waste material. Interestingly, LEF and LEW had similar chemical compositions. As expected, their main constituents were phenols, with the detection of rosmarinic acid, the flavone apigenin-7-O-glucoside, and a number of phenolic acid glycosides. Torras-Claveria et al. (24) reported a comparable composition for Lavandula × intermedia waste material, which in their case contained glucosides of coumaric, ferulic, and caffeic acids in addition to rosmarinic acid, chlorogenic acid, and a number of flavone glycosides. As the identification in the present study was derived from mass spectrometry and UV data, the sugar moieties were designated only as hexosides, although glucosidation is most likely to occur, according to other reports for Lavandula spp. (24, 28). The complete chemical profile was determined here for the essential oil, LEO, with 1,8-cineol, linalool, and linalyl acetate as the major compounds. This confirms the origin of LEO as the genus Lavandula (17), although the high levels of 1,8-cineol and linalool oxides indicate a high degree of oxidation for LEO, most likely due to extended storage after purchase of the plant material. On this basis, the chemical composition of LEO was only partially comparable with those in extensive previous studies in which linalool and linalyl acetate have been seen as the main components (12, 30–32). According to the chemical compositions of LEO, LEF, and LEW, the following seven main compounds were selected for further testing: 1,8-cineol, linalool, linalyl acetate, camphor, (−)-borneol, p-coumaric acid, and trans-ferulic acid. The antibacterial effects of these pure compounds were then compared with the antibacterial effects of the Lavandula preparations (LEO, LEF, and LEW) against C. jejuni.

Linalyl acetate had the strongest effect, while (−)-borneol, camphor, and 1,8-cineol had comparable antibacterial effects to that of LEO. Similarly, Blažeković et al. (12) showed that essential oils of L. × intermedia Budrovka and L. angustifolia, as well as linalool and linalyl acetate, have antibacterial effects against numerous Gram-positive and Gram-negative bacteria. For comparison, the Lavandula preparations and pure compounds used in the present study had stronger antibacterial effects against the Gram-negative C. jejuni. However, according to the higher levels of oxidized compounds in LEO than in the essential oils from L. × intermedia Budrovka and L. angustifolia tested by Blažeković et al. (12), a stronger antimicrobial effect might have been expected. At the same time, antibacterial effects of plant preparations can depend strongly on the type and storage of the plant material, the method of preparation of the plant material, the type of microorganisms used, the inoculum volumes, and the culture medium used, along with the pH, temperature, and incubation time (33, 34).

The moderate but comparable antibacterial effects of the ethanol extract LEF and the waste material LEW, as well as for p-coumaric acid and trans-ferulic acid, showed that LEW contains a diverse pool of bioactive compounds with antimicrobial properties. This is consistent with previous studies on postdistillation thyme waste, pinot noir grape skins and seeds, and postdistillation juniper fruit waste, where the advantages of agricultural waste material as promising antimicrobial agents with potential industrial applications were demonstrated (35, 36).

The Lavandula preparations in this study were effective for degradation and removal of the C. jejuni biofilms. LEO had the strongest effects on these mature C. jejuni biofilms. Similarly, Dănilă et al. (37) reported that lavender essential oil is an effective antibiofilm agent against Staphylococcus epidermidis clinical strains. It has also already been shown that various essential oils are good antibiofilm agents on various surfaces (38, 39). As showed here, LEO is rich in secondary metabolites, which are probably responsible for the degradation of the extracellular polymer matrix of the biofilms (40). Part of the C. jejuni polymers in the extracellular matrix can also form a capsule (41). Indeed, this study has shown that the two major proteins that are involved in the formation of the capsule in C. jejuni (i.e., KpsM and KpsS) were downregulated. High levels of numerous secondary metabolites might also modulate the synthesis and activity of enzymes to cause capsule degradation while disturbing the uptake of nutrients. These can also alter the pH gradient of the cell cytoplasm and thus cause cell death (42).

This was also confirmed by GO analysis of C. jejuni NCTC 11168, where LEO was seen to modify different biological and metabolic processes. In addition, it was shown that LEF and LEW have comparable effects to each other on biofilm removal, which might be due to their similar chemical compositions. Particularly good antibiofilm effects of this Lavandula agriculture waste material, LEW, emphasize its potential as an antimicrobial agent with possible industrial applications for precoating abiotic surfaces and for their disinfection.

The Lavandula preparations were successful for the degradation of these mature C. jejuni biofilms and were further used to determine whether they can be used to prevent biofilm formation. The first phase of biofilm formation is the attachment of the cells to the surface (29). Here, these Lavandula preparations reduced the adhesion of C. jejuni to polystyrene after 4, 8, and 24 h of incubation. These reductions in the adhesion were by ≥1 log₁₀ CFU/ml, which satisfies the recommendations proposed by the European Food Safety Authority (3). These results are also comparable to the effects of lavandin essential oil and ethanol extracts from Urtica dioica on the adhesion of C. jejuni (21).

To potentially focus on the components of these Lavandula preparations that are particularly important for this reduced C. jejuni adhesion, the seven major components as pure compounds were also tested, with various effects seen. For example, (−)-borneol reduced C. jejuni adhesion the most after 24 h of treatment, although this effect was weaker after 4 h and 8 h. On the other hand, trans-ferulic acid reduced C. jejuni adhesion after 4 h and 8 h of treatment, but this was lost after 24 h. These differing phenomena might be due to effects of the pure compounds on different targets at different times. It is interesting to note, however, that these main components from the essential oil and ethanol extracts did not have as good of effects against C. jejuni adhesion as the Lavandula preparations. Likewise, Duarte et al. (20) showed that coriander essential oil had greater effects against C. jejuni and Campylobacter coli adhesion than the pure compound linalool. This indicates that it is the combination of the individual components in these Lavandula preparations that is important for these antiadhesion effects.

Furthermore, the effects of the Lavandula preparations and pure compounds were monitored for C. jejuni motility, which is important for both colonization of surfaces and survival in the environment through biofilm formation (10, 43). All three of the Lavandula preparations and all pure compounds indeed reduced the motility of C. jejuni. The Lavandula preparations all had similar effects on C. jejuni motility. Linalyl acetate and (−)-borneol had similar effects as LEO, while LEO had stronger antimotility effects than linalool and camphor. This also suggests that the combination of the pure compounds in LEO is important for these antimotility effects. For LEF and LEW, these both had stronger antimotility effects than p-coumaric and trans-ferulic acids, which suggests that these two compounds only contribute to the higher antimotility effects of the combined components of LEF and LEW.

The RNA-seq data provided new insights into the influence of LEO on gene expression in C. jejuni. It can be noted that many ribosomal genes were upregulated, which could be the result of a global change in the transcriptome that caused transcriptional reprogramming. Transcriptional reprogramming might indicate that the bacteria are under intense stress (44). When bacteria are under stress, the stringent response is activated, which affects the expression of ribosomal genes that are necessary for stress resistance (45). Upregulation of the CmeABC efflux system confirms that C. jejuni was under severe stress while exposed to LEO. Nevertheless, it can be seen that this transcriptional reprogramming affected the gene expression of many pathways, including the TCA cycle, pyruvate metabolism, nicotinate and nicotinamide metabolism, folate biosynthesis, aromatic amino acid biosynthesis, and terpenoid and porphyrin biosynthesis (see Fig. S2 in the supplemental material). Many transmembrane proteins were downregulated, including proteins involved in the iron-uptake system. Askoura et al. (46) have shown that iron acquisition genes are downregulated during acid stress, and so it can be hypothesized that exposure of C. jejuni to LEO also led to stress responses in these bacteria. It is important to note that the iron-uptake system is involved in the colonization, biofilm formation, and pathogenicity of this pathogen (47, 48), and so downregulation of this pathway might lead to reduction of these processes. Gene cstA, which encodes a carbon starvation protein, was also downregulated, and this gene is known to be involved in C. jejuni motility and autoagglutination (49). Three other downregulated genes are also important for the stress response of C. jejuni: ppk, dcuA, and hrcA (50, 51). Downregulation of these genes might also contribute to biofilm reduction.

The analysis here of the transcriptome of C. jejuni NCTC 11168 following treatment with LEO showed that most of the genes involved in flagellar assembly and modification of flagella were downregulated (Fig. 5), which helps to explain these observations at the physiological level. Some of the downregulated genes (e.g., folK, Cj0719c, maf4, and Cj1467) are also part of a network of genes that are important for flagellar synthesis. It is known that C. jejuni requires functional flagella for initial attachment and further biofilm formation (52), and so it is necessary to modulate genes involved in flagellar assembly to reduce biofilm formation. Motility in Campylobacter is regulated by a chemotactic signaling system that allows the organisms to follow favorable chemical gradients in environments (53). It is interesting to note that Cj0262c was upregulated, which is the gene that encodes the transducer-like protein and is involved in chemotaxis of C. jejuni (54). Furthermore, its closely related Tlp1-3 transducer-like proteins that are also involved in the C. jejuni chemotaxis sensory system, motility, and colonization (55) were not differentially expressed. From the results obtained at the molecular and physiological level, it can be concluded that LEO did not influence chemotactic behavior and, consequently, chemotactic motility of C. jejuni.

Furthermore, LEO also reduced the expression of genes that are important for formation of proteins and lipopolysaccharides of the outer bacterial membrane (i.e., kpsM, kpsS, bamD, waaC, and lgt) and are involved in the initial attachment of these cells to abiotic and biotic surfaces (56). In contrast, upregulation was seen for four of the main genes that are involved in the synthesis of outer membrane proteins and are important for cell adhesion and biofilm formation (i.e., peb2, peb3, Omp50, and porA) (57). Overall, these data indicate that downregulation of the flagellar genes, together with lower expression of genes involved in the iron-uptake system, in stress defense, and thus, in reduced cell adhesion and motility are the most interesting targets to combat biofilm formation. These data confirm that bacteria are under severe stress while exposed to such plant materials. These insights into complete transcriptome analysis open up new aspects and possibilities for research of new targets that should help in the control of this pathogen.

The results of the present study show the potential use of this Lavandula ethanol extract of postdistillation waste material, LEW, against preformed mature biofilms as well as against the adhesion and motility mechanisms involved in biofilm formation. These results offer new solutions for the use of such plant waste materials that still contain large amounts of bioactive molecules. The reuse of LEW might thus provide economic benefits in different industrial fields (e.g., pharmaceuticals and food industry) and might also help solve one of the global problems here: waste disposal (58, 59).

Conclusion.

Campylobacter jejuni is shown to be sensitive to Lavandula preparations and selected pure compounds. Lavandula preparations are shown to be particularly effective in the fight against one of the world’s most common foodborne pathogens, C. jejuni. The Lavandula preparations have relatively potent antibiofilm properties. Moreover, both physiological and molecular approaches confirmed modulation of the first steps in biofilm formation (i.e., adhesion and motility). Also, it has been shown that the postdistillation waste material of Lavandula flowers has particular antibiofilm effects against C. jejuni, which suggests that such waste material can be reused for industrial purposes. Therefore, further efforts can now be directed toward such innovative approaches for alternative strategies and novel targets against bacterial biofilms to find and develop new and effective agents with antibiofilm activities.

MATERIALS AND METHODS

Chemicals.

Muller-Hinton (MH) agar was from bioMérieux (Marcy-l’Etoile, France), MH broth was from Oxoid (Hampshire, UK), and Karmali agar was from Biolife (Milan, Italy). Glycerol solution was from Kemika (Zagreb, Croatia), Campylobacter growth supplement and phosphate-buffered saline (PBS) were from Oxoid, kanamycin, dimethyl sulfoxide (DMSO), resazurin, menadione, propidium iodide (PI), and TRI reagent were from Sigma-Aldrich (Steinheim, Germany), SYTO9 was from Thermo Fisher (Waltham, USA), and ethanol was from Merck (Darmstadt, Germany). PureLink DNase and PureLink RNA minikits, Qubit RNA HS, Ion Total RNA-seq kits v2, RNA-seq barcode BC primer, and Ion PI Hi-Q sequencing 200 kits were from Thermo Fisher Scientific (Carlsbad, CA, USA), NEXTflex poly(A) magnetic beads were from Perkin Elmer (Waltham, MA, USA), and high-sensitivity DNA kits were from Agilent Technologies (Santa Clara, CA, USA). The pure compounds linalool and 1,8-cineol were from Symrise (Holzminden, Germany), linalyl acetate and p-coumaric acid were from Honeywell Fluka (Charlotte, NC, USA), camphor was from (ICN Biomedicals Inc., Aurora, OH, USA), and (−)-borneol and trans-ferulic acid were from Carl Roth (Karlsruhe, Germany). The total ion chromatograms from the gas chromatography-mass spectrometry analysis showed the following purities: linalool, 94.88%; 1,8-cineol, 99.85%; linalyl acetate, 95.64%; camphor, 94.77%; and (−)-borneol, 96.65%. The photodiode array (PDA) total scan chromatograms from the high-pressure liquid chromatography (HPLC)-UV showed that p-coumaric acid consisted of the trans (98.33%) and cis (1.53%) isomers, while ferulic acid also consisted of the trans (97.87%) and cis (1.85%) isomers.

Lavandula.

Dried flowers of Lavandula angustifolia were from Kottas Heilkräuter (Vienna, Austria; control number KLA91232), from which the LEO, LEF, and LEW were produced. LEO was prepared by hydrodistillation (60), with 2 h distillation of 30 g flowers in 1 liter water in a Neo-Clevenger-type distiller, and then stored at 4°C. LEF and LEW were prepared by 4-h to 6-h ethanol extraction (Soxhlet extraction) of 20 g flowers or Lavandula postdistillation waste material in 500 ml 96% ethanol. These were then concentrated in a rotary evaporator (Laborota 4000; Heidolph Instruments, Germany) at 40°C and 17.5 kPa pressure and stored at 4°C.

Phytochemical analysis of LEF and LEW.

Identification of the flavonoids in LEF and LEW, and determination of the purities of the p-coumaric and ferulic acid, were carried out using liquid chromatography-photo diode array-electrospray ionization mass spectrometry. The dry ethanol extracts from the flowers and the waste material (5 mg) were dissolved in 1 ml ethanol and centrifuged prior to analysis. This analysis was performed on an HPLC system (Ultimate RS 3000 Dionex) that included a pump, autosampler, column compartment, and photodiode array detector, which was coupled to a mass spectrometer (LTQ XL; Thermo Scientific, Waltham, USA). The column (Luna phenyl-hexyl, 5 μm, 250 by 2 mm; Phenomenex, Torrance, CA, USA) was used with a gradient elution of 0.1% formic acid in water (eluent A) and acetonitrile (eluent B) as follows: 0→20 min, 10%→70% B; 20→40 min, 70%→100% B; 40→45 min, 100% B; 45→45.5 min, 100%→10% B; and 45.5→52 min, 10% B. The flow rate was 0.250 ml/min, and the column was maintained at 35°C.

Mass spectra were recorded in negative ion mode for the m/z range from 50 to 2,000 atomic mass units (amu), with data-dependent fragmentation (normalized collision energy, 35%). Mass spectrometry conditions were as follows: capillary temperature, 350°C; source temperature, 300°C; sheath and auxiliary gas flow, 40 and 10 arbitrary units (machine settings), respectively; source voltage, 3.5 kV; and capillary voltage, −17 V. The compounds eluted were determined by their ultraviolent and mass spectra and in comparison with the literature (24, 61, 62).

Gas chromatography-mass spectrometry analysis of LEO.

Identification of the main compounds in LEO and determination of the purities of the linalool, 1,8-cineol, linalyl acetate, camphor, and (−)-borneol were carried out by gas chromatography-mass spectrometry. LEO (10 μl) was dissolved in 990 μl hexane and further analyzed by gas chromatography (7890 A; Agilent Technologies, Santa Clara, CA, USA) and mass spectrometry (5975 C VL MSD; Agilent Technologies, USA), operating at 70 eV, with an ion source temperature of 230°C and interface temperature of 280°C. A split injection (injection volume, 0.2 μl; split ratio, 50:1) at a 240°C injector temperature was used for LEO. A fused silica capillary column was used (5% phenyl, 95% methyl polysiloxane; HP-5MS, 30 m by 250 μm by 0.25 μm; Agilent J & W, USA). The temperature program was 1 min at 60°C and then raised to 220°C at 3°C/min. The carrier gas was helium 5.6 at a flow rate of 0.9 ml/min. Data acquisition was performed using Agilent GC/MSD ChemStation version E.02.02 for the mass scan range of 40 U to 400 U. The compounds were identified by their retention indices according to Adams (63) and by comparing their mass spectra with spectral data libraries (63–65) and the laboratory own database.

Bacterial strains and growth conditions.

Campylobacter jejuni NCTC 11168 (National Collection of Type Culture) was used in this study. The strain was stored at −80°C in 20% glycerol and 80% MH broth. Prior to the experiments, the strain was subcultivated on Karmali agar for 24 h at 42°C under microaerobic conditions (85% N₂, 5% O₂, 10% CO₂). The strain was further subcultured microaerobically in MH broth for 24 h at 42°C, and bacterial counts were determined by spectrophotometric measurements of absorbance at 600 nm. The inoculum was prepared in MH broth at 10⁵ CFU/ml for determination of the MICs as well as for the assays that targeted C. jejuni motility and adhesion. For counting, the C. jejuni NCTC 11168 strain was plated on MH agar under the appropriate conditions (as described above), and the colonies were counted and expressed as CFU/ml.

Antimicrobial susceptibility.

The MICs were determined by the broth microdilution method, as previously described (66). Stock solutions of LEO, LEF, and LEW Lavandula preparations and the pure compounds linalool, linalyl acetate, camphor, (−)-borneol, 1,8-cineol, p-coumaric acid, and trans-ferulic acid were prepared in DMSO at 40 mg/ml. The DMSO in the MH broth did not exceed 1%.

Targeting Campylobacter biofilm degradation.

To determine the effects of the Lavandula preparations for degradation and removal of C. jejuni biofilms, a novel approach based on culturing C. jejuni on an air-liquid interface was developed, with measurement of the relative glass surface coverage after treatment with the Lavandula preparations. Optimal C. jejuni growth conditions were established on the air-liquid interface (e.g., optimal O₂ concentration). Twelve-well microtiter plates (Sarstedt, Nümbrecht, Germany) were used, with the following added to each well: 3.0 ml MH broth, supplemented with 0.4% (vol/vol) Campylobacter growth supplement (which contained 0.125 g sodium pyruvate, 0.125 g sodium metabisulfite, and 0.125 g ferrous sulfate). The medium was inoculated with a 1% (vol/vol) C. jejuni NCTC 11168 overnight culture that contained approximately 1 × 10⁸ CFU/ml. Autoclaved microscopy coverslips (20 by 20 mm; Brand, Wertheim, Germany) were used as the model for the glass surface, which were inserted and tilted to the side of each microtiter well after inoculation of the medium. To prevent desiccation and to minimize the proportion of dead cells, the cultures were incubated without shaking in a microaerobic atmosphere for 72 h at 42°C, in a damp environment. The spent medium in each well was replaced with fresh medium every 24 h.

Visible biofilms were formed at the air-liquid interface on the glass coverslips after 48 h of incubation and were then treated for 24 h with LEO at 0.5× MIC, MIC, or 2× MIC, as well as with LEF and LEW ethanol extracts at 2× MIC, 4× MIC, and 8× MIC, according to the MICs determine as bacterial suspensions (see “Bacterial strains and growth conditions” and “Antimicrobial susceptibility”). Nontreated biofilms from C. jejuni NCTC 11168 were used as the negative controls. The biofilms were analyzed under confocal microscopy after 72 h of incubation.

After incubation, the microscopy coverslips were first washed in PBS to remove weakly adhered cells. This was performed by carefully and slowly submerging the microscopy coverslips in PBS twice in a row by holding them with forceps. The excess liquid was then removed by tapping the edge of the glass on a paper towel. The microscopy coverslips were then stained with 50 μl of a mixture of 20 mM PI and 5 mM SYTO 9 for 15 min (Live/Dead test) (57) in an aerobic atmosphere at room temperature in the dark. The mixture of the PI solution and SYTO 9 was prepared by adding 3 μl PI and 3 μl SYTO 9 to 1 ml physiological solution. PI cannot enter intact cells, while dead cells with a compromised cell membrane allow the passage of PI. When PI binds to DNA, it emits red fluorescence upon excitation. On the other hand, SYTO 9 stains DNA in cells green, regardless of their membrane integrity. Thus, dead cells appear under fluorescence microscopy as red and only those that are only green are alive (57).

Confocal microscopy (LSM 800 Axio Observer Z1 inverted microscope; Zeiss, Germany) was performed using the 20× (numerical aperture [NA], 0.4) lens objective. SYTO 9 fluorescence was excited with a 488-nm diode laser, and PI fluorescence was excited with a 561-nm diode laser, operated at 0.5% power. We always considered the upper 600 μm of the biofilm-covered glass surface, starting at the air-liquid interface in the direction toward the liquid. Experiments were performed in three or more biological replicates, and 10 images were acquired for each sample (5 by 2 mosaic images; total analyzed surface, 1470 μm by 607 μm). In controls, this contained >400,000 viable cells. The images for each slide were analyzed using Fiji software (version 1.52c) to obtain the biofilm surface cover. Individual channel images were converted to binary format by manually setting the threshold value that fully separated the biofilm from the background. Random noise was reduced for all of the images using the despeckle function. The biofilm surface coverage was obtained by using our ImageJ custom macro, where surface coverage of biofilm was calculated by first summing the areas covered by green cells (SYTO 9) and red cells (PI) and then subtracting the area of cells that were simultaneously green and red. The area covered by red cells was under 2% in all cases.

Targeting Campylobacter adhesion.

The adhesion of C. jejuni NCTC 11168 was analyzed under treatments with the Lavandula preparations and pure compounds at subinhibitory concentrations. Inocula were prepared as described above (see “Bacterial strains and growth conditions”) and treated with the LEO, LEF, and LEW Lavandula preparations and pure compounds at 0.25× MIC. They were then transferred (200 μl) to 96-well polystyrene microtiter plates (Nunc 266 120 polystyrene plates; Nunc, Denmark) and incubated microaerobically at 42°C for 4, 8, and 24 h. The supernatants with nonadherent cells were removed from each well and rinsed three times with PBS. Then, 200 μl PBS was added, with sonication for 10 min (28 kHz, 300 W; Iskra PIO, Šentjernej, Slovenia). The adhesion of cells was examined as CFU per milliliter, as previously described (35). The untreated culture acted as the negative control.

Targeting Campylobacter motility.

Antimotility assays were performed as previously described (21). Briefly, C. jejuni NCTC 11168 cultures were treated with LEO, LEF, and LEW Lavandula preparations and pure compounds at 0.25× MIC for 24 h in a microaerobic atmosphere at 42°C. After treatment, 1 μl of the treated cultures was placed in the middle of plated soft agar. The plates were incubated for 48 h in a microaerobic atmosphere at 42°C. After this incubation, the diameters of the swarming colonies were measured. The untreated culture acted as the negative control.

RNA sequencing.

(i) RNA isolation and quantification. The molecular mechanisms of LEO action in C. jejuni NCTC 11168 were analyzed for treatment with subinhibitory concentrations of the LEO Lavandula preparation. C. jejuni NCTC 11168 was grown microaerobically in 42 ml of MH broth for 16 h at 42°C to the middle exponential phase and then treated with LEO at 0.25× MIC for 30 min. The negative control was the culture treated with 1% DMSO under the same conditions, and the positive control was the nontreated culture. The cells were harvested by centrifugation (5,000 × g, 5 min, 4°C), and resuspended to a 1-ml cell suspension, which contained approximately 10⁹ CFU/ml. Cell lysis for isolation of total RNA was performed using RNA isolation (TRI) reagent (Sigma-Aldrich), DNase treatment using PureLink DNase kits (Thermo Fisher Scientific), and purification of isolated RNA using PureLink RNA minikits (Thermo Fisher Scientific), according to the manufacturers’ instructions. Quantification and qualification of the total RNA quality were determined using Qubit RNA HS assay kits (Thermo Fisher Scientific) and fluorimeter measurements (Qubit v4; Thermo Fisher Scientific). mRNA was enriched from the total RNA using magnetic beads (NEXTflex poly(A); Perkin Elmer).

(ii) Ion Torrent library preparation and sequencing.

For next-generation RNA-sequencing (RNA-seq), transcriptome libraries were constructed using Ion Total RNA-seq kit v2 (Thermo Fisher Scientific). Briefly, mRNA samples were enzymatically fragmented and purified using magnetic beads. Afterwards, ion adaptors were hybridized onto the fragmented mRNAs and ligated, and reverse transcription was performed. The prepared cDNA samples were purified using magnetic beads, and each cDNA sample was barcoded with an Ion Xpress RNA-Seq barcode BC primer (Thermo Fisher Scientific). The cDNA libraries were purified, and the concentration and size distribution of cDNA libraries were determined using a 2100 Bioanalyzer and high-sensitivity DNA kits (Agilent Technologies). The barcoded cDNA libraries prepared were diluted to the same molar concentrations, pooled in equal volumes, and amplified using an Ion OneTouch 2 system with the accompanying Ion PI Hi-Q OT2 200 kits. Sequencing was performed on an Ion Proton system, using Ion PI Hi-Q sequencing 200 kits (Thermo Fisher Scientific). The datasets generated and analyzed during the current study are available in the NCBI Sequence Read Archive (SRA) repository (https://www.ncbi.nlm.nih.gov/sra/) under BioProject accession PRJNA747749 and SRA accession numbers SRR15183216, SRR15183214, SRR15183213, and SRR15183212.

(iii) Data analysis.

The bioinformatics analysis was performed using a CLC Genomics Workbench (version 12.0.3) and a CLC Genomics server (version 11.0.2). Prior to the differential expression analysis, quality control of the sequencing reads and trimming of the adapter sequences were performed using the “trim reads” tool. Sequencing reads from each library were subjected to differential expression analysis using the RNA-Seq Analysis 2.21 tool (CLC Genomics server, 20.0.2). Campylobacter jejuni subsp. jejuni NCTC 11168 (ATCC 700819) complete genome sequence and genome annotations from the NCBI Nucleotide database (accession number NC_002163.1) were used as the reference genome sequences (67). To compare gene expression between samples treated with LEO and the controls, the “differential expression in two groups 1.1” tool was used. Genes with absolute log₂ fold change of ≥1 and a false-discovery rate (FDR) P value of ≤0.05 were considered differentially expressed. Differentially expressed genes were further analyzed via the STRING Consortium 2020, which provides functional enrichment analysis of protein-protein interaction networks in the STRING mapper tool (https://string-db.org/cgi/input.pl?sessionId=oO8HWWKYl5Fd&input_page_show_search=on). To construct a network of differentially expressed genes using Cytoscape, the NCBI protein identifiers of differentially expressed genes were searched in String database for the organism Campylobacter jejuni subsp. jejuni NCTC 11168 (ATCC 700819), with a cutoff of 0.8 confidence (score) and a maximum of 10 additional interactors. Afterwards, the network was clustered using the MCL Cluster algorithm implemented in the Cytoscape plugin clusterMaker (68) to determine clusters and functional interactions. For each cluster, the built-in functional enrichment available from the stringApp was used to obtain enriched terms.

Gene enrichment analysis was performed using GO_MWU (https://github.com/z0on/GO_MWU) (69), which uses Mann-Whitney U tests and the Benjamini-Hochberg (B-H) FDR corrections of P values to define which enriched GO categories are significantly represented by either upregulated or downregulated genes. GO categories with B-H FDR P values of <0.1 were considered significantly enriched by either upregulated or downregulated genes.

Statistical analysis.

All of the experiments were carried out in triplicates as three or more independent experiments. The data are expressed as means ±standard deviations, with analysis using Origin 2018 (OriginLab, Northampton, MA, USA). Statistical analysis was performed in IBM SPSS Statistics 23 (Statsoft Inc., Tulsa, OK, USA). To determine distribution of data, a Kolmogorov-Smirnov test of normality was performed and statistical significances were determined using Mann-Whitney tests for two independent means. Data were accepted as significant at a P value of <0.05.

Data availability.

The datasets generated and analyzed during the current study are available in the NCBI Sequence Read Archive (SRA) repository under the BioProject accession PRJNA747749 and SRA accession numbers SRR15183216, SRR15183214, SRR15183213, and SRR15183212.