Effects of Tween 80 on Growth and Biofilm Formation in Laboratory Media

Abstract

Tween 80 is a widely used non-ionic emulsifier that is added to cosmetics, pharmaceuticals, and foods. Because of its widespread use we need to understand how it affects bacteria on our skin, in our gut, and in food products. The aim of this study is to investigate how Tween 80 affects the growth and antimicrobial susceptibility of Staphylococcus aureus, Listeria monocytogenes, and Pseudomonas fluorescens, which are common causes of spoilage and foodborne illnesses. Addition of 0.1% Tween 80 to laboratory growth media increased the growth rate of planktonic S. aureus batch cultures, and it also increased the total biomass when S. aureus was grown as biofilms. In contrast, Tween 80 had no effect on batch cultures of L. monocytogenes, it slowed the growth rate of P. fluorescens, and it led to formation of less biofilm by both L. monocytogenes and P. fluorescens. Furthermore, Tween 80 lowered the antibacterial efficacy of two hydrophobic antimicrobials: rifampicin and the essential oil isoeugenol. Our findings underline the importance of documenting indirect effects of emulsifiers when studying the efficacy of hydrophobic antimicrobials that are dispersed in solution by emulsification, or when antimicrobials are applied in food matrixes that include emulsifiers. Furthermore, the species-specific effects on microbial growth suggests that Tween 80 in cosmetics and food products could affect the composition of skin and gut microbiota, and the effect of emulsifiers on the human microbiome should therefore be explored to uncover potential health effects.

Keywords: Staphylococcus aureus; Tween 80; biofilm; emulsifier; isoeugenol; polysorbate 80.

FIGURE 1

Growth curves of (A) Staphylococcus aureus, (B) Pseudomonas fluorescens, or (C) Listeria monocytogenes grown in either Tryptic Soy Broth (TSB) or TSB with 0.1 (v/v) % Tween 80. Error bars show standard deviation of the mean (n = 9). Growth rates and lag times are included. Statistical significance compared to control is marked by ^∗(p < 0.05).

FIGURE 2

Crystal violet staining of (A) S. aureus, (B) L. monocytogenes, (C) P. fluorescens after growth of biofilms for 0, 24, or 48 h in TSB followed by treatment with 0.1, 0.5, or 1.0% (v/v) Tween 80 for 24, 48, or 72 h. Error bars show standard deviation of the mean. All treatments were statistically significant (p < 0.05) from the control, except S. aureus 0.1% Tween (24 h→24 h); L. monocytogenes 0.1% and 0.5% Tween (48 h→24 h); and P. fluorescens 0.5% and 1% (48 h→24 h).

FIGURE 3

Development of biomass over time measured by crystal violet staining (OD₅₈₅ measurement) for S. aureus (A,B), L. monocytogenes (C,D), and P. fluorescens (E,F). Biofilms were grown in TSB for 24 h (A,C,E) or 48 h (B,D,F) before replacement of the supernatant with TSB containing 0.1% Tween 80 or 0.1% Tween 80 + 0.1% Oleic acid. Error bars show standard deviation of the mean (n = 18). Biofilms treated with Tween 80 or Tween 80 + Oleic acid were compared to control biofilms. Statistical significance (p < 0.05) is marked by ^∗ for Tween 80-treated biofilms and + for Tween 80 and Oleic acid-treated samples.

FIGURE 4

Confocal laser scanning microscopy images of three replicate S. aureus biofilms grown for 48 h in TSB media and transferred to either fresh media (A–C) or 0.1% Tween 80 (D–F) followed by growth for 24 h. SYTO 60 (red) stains living bacteria while TOTO-1 (green) stains dead bacteria and extracellular DNA. Scale bar = 20 μm.

FIGURE 5

Occupied area (in percentage) per layer in z-stacks from confocal laser scanning microscopy images. (A) Dead bacteria and eDNA, (B) Living bacteria. Error bars show standard deviation of the mean (n = 10). Statistical significance (p < 0.05) is marked by ^∗. Each layer corresponds to 0.53 μm.

FIGURE 6

Permeabilization of a model membrane after 1 h treatment with Tween 80, measured by the leakage of calcein from vesicles suspended in TSB. Leakage is measured as fluorescence intensity, as vesicles become increasingly fluorescent as the calcein concentration within decreases. Values are normalized to a control sample in which 100% permeabilization is achieved by addition of 0.5% (v/v) Triton X-100. Error bars show the standard deviation of the mean (n = 3).