Interspecies relationships between nosocomial pathogens associated to preterm infants and lactic acid bacteria in dual-species biofilms

Abstract

The nasogastric enteral feeding tubes (NEFTs) used to feed preterm infants are commonly colonized by bacteria with the ability to form complex biofilms in their inner surfaces. Among them, staphylococci (mainly Staphylococcus epidermidis and Staphylococcus aureus) and some species belonging to the Family Enterobacteriaceae are of special concern since they can cause nosocomial infections in this population. NETF-associated biofilms can also include lactic acid bacteria (LAB), with the ability to compete with pathogenic species for nutrients and space. Ecological interactions among the main colonizers of these devices have not been explored yet; however, such approach could guide future strategies involving the pre-coating of the inner surfaces of NEFTs with well adapted LAB strains in order to reduce the rates of nosocomial infections in neonatal intensive care units (NICUs). In this context, this work implied the formation of dual-species biofilms involving one LAB strain (either Ligilactobacillus salivarius 20SNG2 or Limosilactobacillus reuteri 7SNG3) and one nosocomial strain (either Klebsiella pneumoniae 9SNG3, Serratia marcescens 10SNG3, Staphylococcus aureus 45SNG3 or Staphylococcus epidermidis 46SNG3). The six strains used in this study had been isolated from the inner surface of NEFTs. Changes in adhesion ability of the pathogens were characterized using a culturomic approach. Species interactions and structural changes of the resulting biofilms were analyzed using scanning electron microscopy (SEM) and confocal laser scanning microscopy (CLSM). No aggregation was observed in dual-species biofilms between any of the two LAB strains and either K. pneumoniae 9SNG3 or S. marcescens 10SNG3. In addition, biofilm thickness and volume were reduced, suggesting that both LAB strains can control the capacity to form biofilms of these enterobacteria. In contrast, a positive ecological relationship was observed in the combination L. reuteri 7SNG3-S. aureus 45SNG3. This relationship was accompanied by a stimulation of S. aureus matrix production when compared with its respective monospecies biofilm. The knowledge provided by this study may guide the selection of potentially probiotic strains that share the same niche with nosocomial pathogens, enabling the establishment of a healthier microbial community inside NEFTs.

Keywords: Serratia marcescens; Staphylococcus aureus; Staphylococcus epidermidis; biofilms; klebsiella pneumoniae; lactic acid bacteria; nasogastric enteral feeding tubes; preterm infants.

Figure 1

Antimicrobial activity (mm) of L. salivarius 20SNG2 (black bars) and L. reuteri 7SNG3 (grey bars) against the selected strains of S. marcescens, K. pneumoniae, S. aureus and S. epidermidis. Asterisks show statistical differences between the LAB effect on these strains: * (p < 0.05), and *** (p < 0.001).

Figure 2

Representative images of biofilms (48 h) of L. reuteri 7SNG3 obtained by SEM at 1,000× (A) and 3,000× (B), and by CLSM (E). Representative images of biofilms (48 h) of L. salivarius 20SNG2 biofilms obtained by SEM at 1,000× (C) and 3,000× (D), and by CLSM (F). For CLSM image acquisition, cells (in green) were stained with SYTO 13. CLSM scale bar: 50 µm.

Figure 3

Changes in viable attached cell population (log₁₀ cfu/cm²) in dual-species biofilms. Red and green bars mean log reduction or increase compared with monospecies biofilms of the respective pathogenic strains. Filled and lined bars represent 24 h and 48 h samples, respectively. Asterisks show statistical differences between monospecies pathogenic biofilms and dual-species biofilms with one of the LAB strains: * (p < 0.05) and *** (p < 0.001). † indicates statistical differences between 24 h and 48 h samples (n=4).

Figure 4

Percentage of biomass variation in dual-species biofilms. Red and green bars mean biomass reduction and biomass increasing compared with monospecies biofilms of the respective pathogenic strains. Filled and lined bars represent 24 h and 48 h samples, respectively. Asterisks show statistical differences between monospecies pathogenic biofilms and dual-species biofilms with both LAB: * (p < 0.05), ** (p < 0.005) and *** (p < 0.001). † indicate statistical differences between 24 h and 48 h samples (n=4).

Figure 5

Representative SEM images (3,000×) of monospecies biofilms (48 h) of K. pneumoniae 9SNG3 (A), S. marcescens 10SNG3 (D), S. aureus 45SNG3 (G) and S. epidermidis 46SNG3 (J). Representative SEM images of dual-species biofilms of K. pneumoniae 9SNG3 (B), S. marcescens 10SNG3 (E), S. aureus 45SNG3 (H) and S. epidermidis 46SNG3 (K) with L. reuteri 7SNG3. Representative SEM images of dual-species biofilms of K. pneumoniae 9SNG3 (C), S. marcescens 10SNG3 (F), S. aureus 45SNG3 (I) and S. epidermidis 46SNG3 (L) with L. salivarius 20SNG2.

Figure 6

Representative CLSM images of monospecies biofilms (48 h) of K. pneumoniae 9SNG3 (A), S. marcescens 10SNG3 (D), S. aureus 45SNG3 (G) and S. epidermidis 46SNG3 (J). Representative CLSM images of dual-species biofilms of K. pneumoniae 9SNG3 (B), S. marcescens 10SNG3 (E), S. aureus 45SNG3 (H) and S. epidermidis 46SNG3 (K) with L. reuteri 7SNG3. Representative CLSM images of dual-species biofilms of K. pneumoniae 9SNG3 (C), S. marcescens 10SNG3 (F), S. aureus 45SNG3 (I) and S. epidermidis 46SNG3 (L) with L. salivarius 20SNG2. Scale bar: 50 µm.