**4. Discussion**

In the present study we demonstrated that cinnamon, marjoram, and thyme EOs and main components tested had good antibacterial and anti-biofilm forming effect on the investigated bacteria associated with food spoilage and outbreaks. Trans-cinnamaldehyde and thymol were the best inhibitors with MIC values below 1 mg/mL. As expected from the composition pattern of EO cinnamon, the trans-cinnamaldehyde and the EO exhibited similar effect against the bacterial biofilms studied. Besides this, thymol and trans-cinnamaldehyde are phenolics having the best antimicrobial activity among the EO compounds. For comparison, the monoterpene terpiene-4-ol did not exceed the effect of its parent oil marjoram.

Monoculture biofilms were significantly inhibited by the EOs and components in MIC/2 concentration which suggest that growth-reducing effect is not solely responsible for biofilm inhibition. Sub-lethal damage of the cell wall can negatively influence bacterial attachment to surfaces which is the first step in biofilm formation [29].

Dual-species biofilms responded to EOs differently. Contrary to our findings, Almeida et al. [37] reported that, in dual-species (*E. coli*/*L. monocytogenes*) biofilms, cell densities were not altered. In our case, *L. monocytogenes* outgrew *E. coli* in the control samples and most of the EOs eliminated this bacterium from the biofilm. In line with this, the study of Giaouris et al. [9] concluded that *Listeria* formed a stronger biofilm in mixed population, moreover, presence of other bacteria increased its growth. Co-culturing of *L. monocytogenes* and *S. aureus* resulted a strong biofilm, which is in agreement with the results of Millezi et al. [38]; only high concentrations of EOs (e.g., cinnamon EO: 1–2 mg/mL, marjoram EO, and terpinene: 4–8 mg/mL) and components inhibited their formation. *P. putida* proved to be more resistant to the oils and compounds than the *Listeria* and this bacterium was present in the population even at high agent concentrations. This is in accordance with the results of Giaouris et al. [9] who showed that co-culturing with *L. monocytogenes* within a dual-species biofilm increased the community resistance of *P. putida*.

In most cases, anti-biofilm formation effect showed no concentration dependence above a certain concentration, but the results of cell number determination were not congruent with this. In most cases, dual-species biofilms were inhibited at lower concentration than the MIC of the individual bacteria, but cell death occurred mainly at the higher MIC value or above. This discrepancy pointed to the fact that absorbance and cell enumeration data can differ significantly. Biofilms, inhibited to a high degree based on absorbance data, can still contain enough living cells to cause hygiene problems. Scanning electron microscopic images demonstrated the structural changes caused by EOs or components during biofilm development. Similarly to the altered biofilm structure visualized by confocal laser scanning electron microscope in a recent study [39], our SEM pictures also showed that the three-dimensional structure of matured biofilms disappeared after EO treatment and most of the treated cells have been burst. Studying the mechanism of EOs in more detail, Zhang et al. [40] detected leakage of electrolytes due to disruption of cell permeability after EO treatment which eventually lead to cell death. Along with our findings these results also support the fact that EOs induce severe membrane damage [25,40].

In conclusion, the EOs and EO components examined in this study could represent alternatives for elimination of *E. coli*, *L. monocytogenes*, *P. putida,* and *S. aureus* single and *L. monocytogenes*-*E. coli*/*S. aureus*/*P. putida* dual biofilms in vitro with different efficiency. The use of EOs as antimicrobial agents in real food systems is often limited due to their strong odor and taste. Therefore, our future investigations aim at novel approaches, such as encapsulation of EOs, that could potentially reduce the organoleptic impact and increase the antimicrobial activity [41].

**Author Contributions:** Conceptualization, E.B.K, J.K., and C.V.; methodology and data analysis, E.B.K., A.V., M.T., T.P., G.H., and V.L.B.; writing—original draft preparation, E.B.K.; writing—review and editing, J.K., C.V., and M.T.

**Funding:** This work was supported by the Hungarian Government and the European Union within the framework of the Széchenyi 2020 Programme through grants GINOP-2.3.2-15-2016-00012, GINOP-2.3.3-15-2016-00006 and EFOP-3.6.1-16-2016-00008. The research of M.T. was supported by grants through the János Bolyai Research Scholarship of the Hungarian Academy of Sciences and the UNKP-18-4 New National Excellence Program of the Ministry of Human Capacities. The research of G.H. was supported by the NKFI K 128217 Research Scholarship.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
