Effects of phytogenic substances on growth and biofilm formation of Escherichia coli and Salmonella field isolates

The phytogenics garlic oil (Allium sativum), cinnamaldehyde, carvacrol, thymol, and thyme oil (Thymus vulgaris) were provided by Delacon Biotechnik GmbH, Engerwitzdorf, Austria. The major metabolites were identified and quantified by gas chromatography-mass spectrometry (GC-MS) analysis (Table 1). Rapeseed oil was used as lipophilic control substance for CV-biofilm assay (VFI GmbH, Wels, Austria).

E. coli field isolates were provided by the veterinarian pathology department of AGES GmbH Linz, Austria, and serotyped by the national reference laboratory for Escherichia coli of AGES GmbH Graz, Austria. Due to their high-level growth in pure culture in various organs and their mucoid or hemolytic properties, the field isolates O54:H21, O88:H8, O149:H10 (all three isolated from pig), and O8:H2 (isolated from cattle) were classified as pathogenic. E. coli O174:H2 was isolated from cattle and carries verotoxin genes (vtx1 and vtx2) as well as the enterohemolysin gene (E-hly). Orough:Hrough was isolated from piglet and is positive for verotoxin gene 2 (vtx2). E. coli ONT:H10 is a field isolate from small intestine of calf and possesses the heat-stable enterotoxin A gene (estA). E. coli O157:H7 was provided by the national reference laboratory for Escherichia coli of AGES GmbH Graz and was tested positive for both verotoxin genes (vtx1 and vtx2), the enterohemolysin gene (E-hly), and the intimin gene (eae). E. coli reference strain ATCC 8739 was purchased from Microbiologics Inc., Minnesota, USA (WDCM number 00012).

Salmonella enterica subsp. enterica field isolates were sero-typed by the national reference laboratory for Salmonella of AGES GmbH Graz. Salmonella Mbandaka was isolated from a process control sample of an Austrian feed manufacturer. This serotype was also detected in laying hen feed, flocks, and eggs as well as human fecal samples and traced back to an oil mill in Italy (personal communication). Salmonella Senftenberg was isolated from a soybean meal sample imported from Italy (RASFF notification No. 2013.1738). Salmonella Typhimurium (multi-resistant field isolate from human stool sample), Salmonella Enteritidis (field isolate from human stool samples), and Salmonella Infantis (multi-resistant field isolate from mature poultry) were provided by the national reference laboratory for Salmonella of AGES GmbH Graz. Salmonella Typhimurium reference strain ATCC 14028 was purchased from Microbiologics Inc. (WDCM number 00031).

Susceptibility testing of all E. coli and Salmonella serotypes was performed with broth micro-dilution method according to DIN EN ISO 20776-1:2019 (ISO, 2019). Stock solutions (200000 ppm) of garlic oil, cinnamaldehyde, carvacrol, thymol, and thyme oil were prepared in absolute ethanol (VWR International GmbH, Vienna, Austria) and serially diluted (1:1 dilutions) in Mueller-Hinton broth (Oxoid GmbH, Wesel, Germany) to obtain concentrations of 20000 ppm to 10 ppm. Each substance concentration was pipetted horizontally into Nunc^® polystyrene 96-well microtiter plates (Thermo Scientific, Waltham, Massachusetts, USA) in four technical replicates (50 μl per well).

The test organisms were incubated in 10 ml Mueller-Hinton broth for 16–20 h at 37°C. The cultures were then diluted with fresh Mueller-Hinton broth (approx. 1:500 dilutions) to achieve a final bacterial inoculum of 5 × 10⁵ to 2 × 10⁶ CFU ml⁻¹. 50 μl of the inoculum were added to each well to obtain a total test volume of 100 μl and final substance concentrations of 10000 ppm to 5 ppm. In addition, each microtiter plate contained four replicates with Mueller-Hinton broth alone as negative growth control (NC) and four replicates with bacterial inoculum without test substance as positive growth control (PC). The starting bacterial concentration was verified with colony count method. The microtiter plates as well as the colony count plates were incubated for 16–20 h at 37°C. The MIC was defined as the lowest concentration of the test substance that completely inhibits visible growth (no obvious cell pellet or turbidity).

The biofilm biomass was assessed with the microtiter plate assay by measuring all the attached biomass via crystal violet staining (defined as “CV-biofilm”). Within this definition, the field isolates E. coli O88:H8, O54:H21, and O174:H2 as well as Salmonella Mbandaka, Senftenberg, Infantis, and Typhimurium showed CV-biofilm formation and were exposed to garlic oil, cinnamaldehyde, carvacrol, thymol, thyme oil, and rapeseed oil as described in section “AST (Antimicrobial susceptibility testing)”. However, nutrient broth instead of Mueller-Hinton broth was used for serial dilutions and to adjust the bacterial inoculum to 2 x 10⁶ to 5 x 106 CFU ml⁻¹, which was verified with colony count method. After incubation for 16–20 h at 37°C, the amount of bacterial growth of the PC, at the fist sub-MIC, and at 5 ppm of each test substance was determined by OD600 measurement with the NanoDrop One photometer instrument (NanoDrop Micro-UV/Vis-Spectral photometer, Thermo Scientific, Waltham, Massachusetts, USA). The microtiter plates were then rinsed two times with water to remove the unattached cells and media components and dried for about 30 min at room temperature. For crystal violet staining, 125 μl of 0.1% (w/v) crystal violet solution (Alfa Aesar by Thermo Fisher GmbH, Kandel, Germany) in water was added to each well and incubated at room temperature for 10–15 min. Subsequently, the plates were rinsed 3–4 times with water and completely dried at room temperature. To quantify the adherent bio-mass, 125 μl 30% acetic acid solution (Merck KGaA, Darmstadt, Germany) in water (v/v) were added to the wells to solubilize the adherent crystal violet. The resulting solution was transferred to a new microtiter plate. Absorbance was measured at 550 nm with the PHOmo micro-plate reader (Anthos Mikrosysteme GmbH, Friesoythe, Germany). The assay was carried out on three different days with four technical replicates per day for each combination of test strain and substance concentration.

All the statistical analyses were performed with SAS 9.4 (SAS Institute, Inc., Cary, NC, USA). Parameters were analyzed for outliers and normal distribution with the UNIVARIATE procedure. To compare the differences among concentrations within different substances and strains, data were subjected to an analysis of variance (ANOVA) using PROC GLIMMIX according to the model: Yijk=μ+αi+βj+αβij+dayk+εijki=1,…,a;j=1,…,b;k=1…,n \matrix{{{Y_{ijk}} = \mu + {\alpha _i} + {\beta _j} + \alpha {\beta _{ij}} + {day}_k + {\varepsilon _{ijk}}} \hfill \cr {i = 1, \ldots,a;j = 1, \ldots,b;k = 1 \ldots,n} \hfill} where y_ijk is the observation in concentration i and strain j on day k; μ the overall mean; α_i the fixed effect of concentration i; β_j the fixed effect of strain j; αβ_ij the interaction effect of concentration i and strain j; day k the random effect of day k, and ε_ijk the random error with mean 0 and variance σ₂. In addition, a is the number of concentrations; b the number of strains; and n the number of days.

The MIC of each test substance was defined as the lowest concentrations that completely inhibits visible growth of all the E. coli or Salmonella serotypes. The highest potential for antimicrobial effects in E. coli field isolates and reference strain was observed with thymol at a value of 150 ppm, followed by carvacrol and cinnamaldehyde at 300 ppm and thyme oil at 600 ppm. MIC values of Salmonella field isolates and reference strain exposed to the tested phytogenics were comparable to those of E. coli, except for the MIC of carvacrol, which was in line with the value of thymol in E. coli. Garlic oil showed no inhibition of bacterial growth within the concentration range tested (up to 10000 ppm).

The MIC values for E. coli and Salmonella field isolates examined in the CV-assay were mostly in line with the MIC values from AST, with cinnamaldehyde, thymol, and carvacrol inhibiting growth at 300 ppm, thyme oil at 600 ppm and garlic oil showing no inhibition within the concentration range tested. The CV-biofilms for E. coli and Salmonella were calculated separately and are illustrated in Figure 1.

Figure 1

No interaction of the applied concentration and E. coli strains has been observed for any of the five phytogenics (P > 0.050). Cinnamaldehyde proved to be the most effective substance against E. coli CV-biofilm formation and showed a reduction of CV absorbance compared to the PC control even at a concentration as low as 5 ppm (−25.5%, P = 0.012). Carvacrol, thymol and thyme oil reduced E. coli CV-biofilm starting at a sub-MIC of 40 ppm (−47.8%, P = 0.001; −33.1%, P = 0.025; and −32.4%, P = 0.002, respectively). Garlic oil reduced E. coli CV-biofilm at a concentration of 10000 ppm by −80.9% (P < 0.001) compared to the PC.

MIC values determined for Salmonella with the CV-assay were identical to those of E. coli. The effects of phytogenics on Salmonella CV-biofilm formation were comparable, although an interaction of applied concentration and strain was found for garlic oil (P = 0.027) and thymol (P = 0.001), but not for cinnamaldehyde, carvacrol, and thyme oil (P > 0.050). Cinnamaldehyde was the most effective substance, showing reductions of Salmonella CV-biofilm at 40 ppm (−23.2%, P = 0.043). Carvacrol, thymol, and thyme oil reduced Salmonella CV-biofilms at 80 ppm by −65.6% (P < 0.001), −56.9% (P < 0.001) and −42.7% (P < 0.001), respectively. Garlic oil did not significantly affect CV-biofilm of Salmonella within the concentration range tested.

The count of viable cells was determined at the first sub-MIC and at 5 ppm substance concentration as well as for the PC. Generally, cell growth of both E. coli and Salmonella is influenced at the fist sub-MIC, with Salmonella being subject to higher fluctuations.

Rapeseed oil, which was used as lipophilic control substance, had no significant effect on CV-biofilm of E. coli and Salmonella.