USING GUT-ON-A-CHIP AND 4D IMAGING TO UNDERSTAND PATHOGEN INVASION

Abstract

Real time mechanobiology study challenge

In the intricate dance between pathogens and the human body, mechanical cues wield significant influence, defining gene expression, adhesion dynamics, and even morphogenesis. From the density of extracellular matrix, shear forces of fluid flow in urinary tract or blood vessels tortuosity, the field of mechanobiology intertwines with fundamental biological processes. However, capturing these dynamics in real-time, especially within complex microenvironments like the gut, has remained a challenge because confocal microscopes are too slow to image peristaltic motion [1].

Gut-on-chip models

The development of Organ-on-a-Chip (OoC) technology, where miniature ecosystems replicate physiological conditions with remarkable fidelity, has opened new opportunities to study mechanobiological events, particularly for pathologies that are exclusive to humans and cannot be replicated in animal models. Through the integration of diverse cell types and porous membranes facilitating homeostatic conditions, OoC platforms mimic the microarchitecture of organs while introducing essential elements like flow and peristalsis. Yet, the quest for comprehensive insights faces hurdles, particularly in achieving real-time 3D imaging and studying transient events.

The study involves the invasion mechanisms of two pathogens:

The amoeba Entamoeba histolytica (responsible for amebiasis)

Amebiasis results from the presence of the amoeba Entamoeba histolytica, a parasite that exclusively targets humans. While many infections show no symptoms, once the parasite breaches the intestinal lining, it can trigger painful diarrhea accompanied by blood and ulcer formation. In its severe manifestations, amebiasis can escalate to the formation of abscesses in vital organs such as the liver, lungs, and brain.

The bacteria Shigella flexneri (responsible for shigellosis)

Shigellosis stems from infection with Shigella bacteria, which are distinct variants of Escherichia coli. These bacteria harbor a virulence plasmid, allowing them to infiltrate only human intestinal epithelial cells and subsequently the mucosal layer. This invasion triggers significant inflammation accompanied by extensive tissue damage.

Aim of the study

The aim of this work was to pioneer 4D live imaging of OoC under cyclic deformation, mirroring the peristaltic motions of the intestine. Beyond visualization, this approach unveils the rheology of tissues, offering a new perspective into the spatiotemporal distribution of mechanical stress. With a focus on the enteric barrier, the study presents the interplay between mechanical cues and pathogenic assaults, highlighting the virulence mechanisms of E. histolytica and S. flexneri.

Various parameters were observed: death of host cells, tissue connectivity, pathogens tracking, penetration and colonization, and stress correlation.

Experiment Setup

Materials:

Gut-on-chip (Emulate)
T2i spinning-disk confocal microscope (Nikon)
ORCA-Flash 4.0 digital CMSO camera (Hamamatsu)
OB1 MK3+ flow controller(Elveflow)
MSF flow sensor (Elveflow)
Elveflow Software Interface (Elveflow)

4D live imaging optimization

First of all, acquiring videos under peristaltic conditions presents several challenges, that were addressed in this study:

Alignment: Sequences often fall out of phase with the cyclic peristaltic motion. To address this, alignment into a 4D stack is necessary, accounting for the time lag between cycles.
Tissue curvature: The curvature of the tissue can lead to images being out of focus, needing a 2D projection.
Movement correction: Out-of-plane movement can occur, necessitating motion correction techniques to ensure accurate imaging.

Gut-on-chip infection

The gut-on-chip developed by Emulate consists of two chambers (as presented in their previous work [2]). The upper chamber is perfused with culture medium and hosts cultured Caco2 cells, while the lower chamber is solely dedicated to perfusing culture medium. A transverse membrane enables mechanical stretching to simulate peristaltic movement. The bacteria and amoeboid infection were consequently obtained by injecting infected cultures to the system.

Microfluidic peristaltic stress setup

The microfluidic setup was monitored with the ESI software from Elveflow and monitored with OB1 pressure controller on a 3 channels setup (see figure 4):

Channel 1: Medium perfusion to the top channel of the intestine chip, containing Caco2 cells (epithelium from human colon). Associated with MSF flow sensor (30 µl/hour).

Channel 2: Medium perfusion to the bottom channel of the intestine chip. Associated with MSF flow sensor (30 µl/hour).

Channel 3: Lateral vacuum stretch to reproduce peristalsis of the colon (magnitude 10% on 0.15 Hz).

Key Findings

Notes for experiment control parameters: WP: with peristalsis-like stretch; WOP: without peristalsis-like stretch; Shigella-mxiD: avirulent strain; TSAR Shigella strain expressed GFP upon T3SS activation (marker of virulence gene induction allowing penetration into host cells); Cysteine protease inhibitor (E64) prevents from the cell membrane degradation activity of amoeba.

Development of a tissue rheology model

In order to compare local stress and local virulence, a tissue rheology study model was established. However, the chip encasement prohibits the use of rheological probes, and the PDMS scaffold alters mechanical behavior compared to the pressure pump command. Consequently, a robust mathematical model was developed based on video analysis to address these challenges.

Dynamical invasion of S. flexneri

For the study of S. flexneri infection dynamics, the researchers maintained a monitoring period of 1 to 2 hours post-infection, comparing results with an avirulent strain (Shigella-mxiD). They observed two pivotal parameters: the extent of bacterial colonies and the individual duration of penetration.

Peristalsis boost bacterial colonization, starts earlier and expends more rapidly

Indeed, the number of bacteria significantly increases with peristaltic movement (Fig. 5A). Also, during the first 2 hours of infection, TSAR activation passes through two distinct phases demonstrating an earlier and more extensive induction of virulence with peristalsis (Fig. 5B).

Local mechanical stress accelerates cell to cell spreading

The effects of local mechanical stress on S. flexneri were investigated by comparing the speed of TSAR activation among individual bacteria. In order to achieve this, stress maps of the tissue were computed utilizing the blue membrane probe Pro12A and compared to individual bacteria TSAR activation. As illustrated in Fig. 6, it was observed that bacteria exhibited significantly earlier activation in regions of elevated epithelial stress.

Dynamical invasion of E. histolytica

For the infection study of E. histolytica, a rigorous 7-hour surveillance period through 30-minute or 1-hour intervals was employed, contrasting with a control featuring a cysteine protease inhibitor (E64) that prevents cell degradation. Three crucial parameters were observed: cell death, tissue connectivity, and pathogen penetration.

Peristalsis boosts amoeboid tissue degradation and tissue invasion

With fluorescent microscopy analysis, it has been observed a 10% increase of death rate with peristalsis (WP) in 3 hours after amoeboid infection compared to static tissue (WOP, Fig. 7A). To address the tissue impairment, the cells connectivity was also observed and presented a failure point around 3 to 4 hours after the challenge (Fig. 7B).

The process of tissue degradation was also observed in fluorescent microscopy (Fig. 8). Briefly, E. histolytica (red) degraded the actin (purple) on the brush border, phagocyted dead cells (yellow), and cleaved E-cadherin (green) junctions during the process of infection. Peristaltic motion actively promotes this mechanism, highlighting the intimate connection between infection mechanisms and tissue mechanics, as elucidated in this study.

Cysteine proteases secreted by E. histolytica are essential for efficiently degrading and invading the human colonic tissue. In experimental chip conditions involving cysteine proteases inhibitor (E64), no more degradation of the tissue and parasite invasion was observed, WP and WOP (Fig. 9). This underscores the essential role of cysteine proteases activity in efficiently breaking down and invading tissue, validating that the OoC is an accurate model to study the initial steps of amoebiasis.

Local mechanical stress enhances amoeboid penetration

Ultimately, it was determined that peristalsis reduces the migration of amoeba on the epithelium compared to conditions without peristalsis, thereby facilitating tissue penetration. By comparing the speed of individual amoeba penetration to local stress levels using the same correlation method with local stress, it was revealed that parasites were statistically more successful in areas experiencing higher stress induced by peristaltic movement.

Conclusion

Peristalsis plays a role in renewing gut tissue by shedding epithelial cells and microbiota. Inhibiting gastrointestinal motility usually promotes bacterial overgrowth, indicating that peristaltic motion is fundamentally crucial for minimizing the risk of infection. Nevertheless, the findings demonstrate that peristaltic motion serves as a determinant factor for the invasion of both S. flexneri and E. histolytica, despite their significant differences in size, life cycle, and infection mechanisms. These pathogens have evolved to exploit colonic environmental cues, suggesting a reevaluation of how we investigate host-pathogen interactions within their target organ niche. An overview of the influence of peristaltic motion and local mechanical stress with the two pathogens infection is represented in Fig. 10.

Authors Information

Nathalie Sauvonnet is the head of the tissue hemostasis group at Institut Pasteur. The group focused on intracellular trafficking of eukaryotic cells and how it regulates cell and tissue organization. They are experts in cutting-edge technology using high resolution microscopy, single molecule tracking, robust image analysis, statistical classification, crispR-cas9 edition to organ-on-chip.

Aleix Boquet-Pujadas has an academic background in mathematics and physics from Universitat Autònoma de Barcelona. In 2019, he defended his PhD in applied mathematics from Sorbonne Université in Paris. His thesis demonstrated variational approaches in inverse problems for imaging-based characterization of cellular dynamics under the supervision of Jean-Christophe Olivo-Marin.

Elisabeth Labruyère, a senior researcher at the Institut Pasteur, has extended her expertise in the mechanobiology of intestinal invasion by the pathogenic amoeba Entamoeba histolytica. She focused on the interplay between amoebic and environmental factors, both biological and mechanical, influencing its migration through colonic tissue.

References

[1] A. Boquet-Pujadas et al., “4D live imaging and computational modeling of a functional gut-on-a-chip evaluate how peristalsis facilitates enteric pathogen invasion,” 2022. [Online]. Available: https://www.science.org

[2] A. Grassart et al., “Bioengineered Human Organ-on-Chip Reveals Intestinal Microenvironment and Mechanical Forces Impacting Shigella Infection,” Cell Host Microbe, vol. 26, no. 3, pp. 435-444.e4, Sep. 2019, doi: 10.1016/j.chom.2019.08.007.