A novel on-a-chip system with a 3D-bioinspired gut mucus suitable to investigate bacterial endotoxins dynamics

Abstract

The possible pathogenic impact of pro-inflammatory molecules produced by the gut microbiota is one of the hypotheses considered at the basis of the biomolecular dialogue governing the microbiota-gut-brain axis. Among these molecules, lipopolysaccharides (LPS) produced by Gram-negative gut microbiota strains may have a potential key role due to their toxic effects in both the gut and the brain. In this work, we engineered a new dynamic fluidic system, the MINERVA device (MI-device), with the potential to advance the current knowledge of the biological mechanisms regulating the microbiota-gut molecular crosstalk. The MI-device supported the growth of bacteria that are part of the intestinal microbiota under dynamic conditions within a 3D moving mucus model, with features comparable to the physiological conditions (storage modulus of 80 ± 19 Pa, network mesh size of 41 ± 3 nm), without affecting their viability ($\sim$ 10⁹ bacteria/mL). The integration of a fluidically optimized and user-friendly design with a bioinspired microenvironment enabled the sterile extraction and quantification of the LPS produced within the mucus by bacteria (from 423 ± 34 ng/mL to 1785 ± 91 ng/mL). Compatibility with commercially available Transwell-like inserts allows the user to precisely control the transport phenomena that occur between the two chambers by selecting the pore density of the insert membrane without changing the design of the system. The MI-device is able to provide the flow of sterile medium enriched with LPS directly produced by bacteria, opening up the possibility of studying the effects of bacteria-derived molecules on cells in depth, as well as the assessment and characterization of their effects in a physiological or pathological scenario.

Keywords: Dynamic culture; Gut-brain-axis; Lipopolysaccharide; Microbiota; Microenvironment; Microfluidic devices.

Graphical abstract

Fig. 1

Schematic representation of the MI-device rationale (bottom, centre): it has been designed to fill the gap between the classical gut-on-a-chip approach (left, representative picture of a gut-on-chip device) and the in vitro modelling of the intestinal mucus (right, representative image of a hydrogel). The dynamicity and high level of control over the mass-transport phenomena that are typical of fluidic devices are synergically coupled with the bio-similarity in terms of viscoelastic properties and microstructure of the optimized intestinal mucus model.

Fig. 2

A) Stability of the intestinal mucus model in a time period of 21 days. B) Viscoelastic properties of the model expressed in terms of storage (G′) and loss (G″) moduli in the frequency spectra of 0.1–20 Hz. C) Generalized Maxwell model results fitting (red colour) the experimental data (blue colour). (Data obtained by three independent experiments with n = 5).

Fig. 3

A) Release profile of the LPS-FITC from the I-Bac3Gel to the medium in time. B) Release rate of the LPS-FITC and the Weibull model fitting the experimental data. (Data obtained by three independent experiments with n = 5).

Fig. 4

Viability of ATCC and CCUG E. coli (A-B and C-D respectively) after being cultured in the I-Bac3Gel (3D) or suspension (pK) for different time periods (24, 48 and 72 h) in Transwell-like insert with transparent (T) or translucent (L) PET membranes to separate the hydrogels from the medium. (Data obtained by five independent experiments with n = 3).

Fig. 5

Concentration of LPS quantified in the medium added to the insert with transparent (T) or translucent (L) membrane in contact with ATCC (A and B) or CCUG (C and D) E. coli cultured in suspension (pK) or within the intestinal mucus model (3D). (Data obtained by five independent experiments with n = 3).

Fig. 6

Design of MI-device for 3D dynamic culture of bacteria. (A) Exploded and combined vision of the three components (apical and basal components and Transwell-like insert) composing the MI-device. B) Assembled MI-device with red and blue colours representing the fluid dynamics of the apical and basal hemi-chambers, where the flows of culture medium (red, apical hemi-chamber) and of I-Bac3Gels (green, basal hemi-chamber) are separated by the membrane of the Transwell-like insert (in blue). C) Real pictures of the Transwell-like inserts and of the MI-device (bottom and top visions) showing the pre-filling inflow of the I-Bac3gels. D) Focus on the functional area of the MI-device where the exchange of molecules occurs. The I-Bac3Gel with the embedded bacteria flows in the basal hemi-chamber. Thanks to the porous membrane, the bacteria-derived molecules, including LPS, can diffuse into the apical hemi-chamber (for technical and dimensional details, see Refs. [[80], [81], [82]]) and can be collected for further use.

Fig. 7

3D dynamic culture of ATCC (A) and CCUG (B) cultured in the MI-device mounted with transparent (T) or translucent (L) commercial inserts. The dashed lines represent the upper and lower ranges of CFUs counted for 3D static culture for ATCC (left) and for CCUG (right). (Data obtained by five independent experiments with n = 2).

Fig. 8

Concentration of LPS quantified in the apical hemi-chamber of the MI-device mounted with transparent and translucent (T and L, respectively) inserts for the 3D dynamic culture of ATCC (A and B) or CCUG (C and D) E. coli. The concentration of LPS of the 3D static cultures (3D(L) and 3D(T)) were reported for comparison. (Data obtained by five independent experiments with n = 2).

Fig. 9

Effect on Caco-2 cells of the LPS-containing medium obtained by 3D dynamic culture of ATCC E. coli in the MI-device after 72 h. Integrity and morphology of the epithelial cell monolayer were measured respectively by means of TEER quantification (A) and light microscopy (B) (representative picture, scale bar of 100 μm). (Data obtained by two independent experiments, each assessed in triplicate).