Probing and modeling cell-cell communication in 2D biomimetic tissues

TL;DR

The study addresses how gap-junction-like transport operates in tissues by building biomimetic 2D hexagonal networks of Droplet Interface Bilayers (DIBs) decorated with α-Hemolysin nanopores. Calcein diffusion is quantified and interpreted with a Continuous Time Random Walk (CTRW) model on hexagonal lattices, revealing that the inverse diffusion time scales as $\lambda \sim c_m^{1.6}$, indicating nonlinear pore-formation kinetics. A hex-lattice CTRW framework reproduces the observed spatiotemporal diffusion profiles, while deviations at high pore load point to lipid composition effects on pore adsorption and transport. These findings establish a controllable, quantitative platform for probing intercellular transport mechanisms and offer insights for tissue-scale modeling of cell–cell communication.

Abstract

In tissues, cells in direct physical contact with each other can exchange ions or molecules via protein clusters called gap junctions that form channels across the membranes of adjacent cells. Here, we use a simplified biomimetic approach, coupled with theoretical modeling, to unravel the physical mechanisms controlling such transport. Tissues are mimicked with 2D hexagonal networks of monodisperse aqueous droplets connected by lipid membranes called Droplet Interface Bilayers (DIBs), decorated with $α$-Hemolysin ($α$HL) transmembrane proteins forming nanopores through heptamerization in the membrane. The diffusion of calcein across 2D DIB networks is thoroughly studied using epifluorescence microscopy at various $α$HL concentrations. The results are successfully confronted with a Continuous Time Random Walk model in hexagonal networks, with an average waiting time increasing nonlinearly with the concentration of pore monomers.

Figures (5)

• Figure 1: (A) Sketch of the droplet printer setup. The printing head, mounted on a vibration exciter is positioned above an oil+lipid pool placed on the XY motorized translation stage. A syringe pump imposes a flow of the aqueous phase through a capillary tubing, attached to the printing head. Inset: principle of the printing technique -- images reproduced from valet2018quasistatic showing the detachement of an aqueous droplet as it crosses an oil/air interface (the white scale bar is 400 $\mu$m long). (B) Droplets printed using the setup imaged from below. (C) Principle of the experiment: the Calcein molecules contained in a source droplet diffuse in the network of DIBs connected with $\alpha$Hemolysin nanopores. (D) The DIB network after compacting the droplets observed in bright field. The source (magenta symbol), first (red), second (blue) and third (green) neighbors positions have been overplotted on the image. The thin blue lines represent the Delaunay tessellation used to determine the connectivity of the droplets.
• Figure 2: Histogram of the droplet diameters. The solid line is a gaussian fit, yielding $\bar{d}$ = 157 $\pm$ 10 $\mu$m. Inset: histogram of the coordination number.
• Figure 3: A-C - Fluorescence images of a DIB network, with $c_m$=125$\mu$g/mL at different times ($t$=0; 6 and 16 hours, respectively) showing the diffusion of calcein in the network. D) The corresponding normalized intensity (presence probability) as a function of time, for the source (upper pannel, black diamonds), the first/second/third neighbors (lower pannel, red disks/blue squares, green triangles). The error bars are standard error of the mean for curves sharing the same rank of neighborhood. The solid lines are fits with Eq. \ref{['HexaFit']}. On the upper pannel, a control experiment without any nanopores has been overplotted, showing that the source intensity (magenta crosses) remain constant over time.
• Figure 4: Selection of normalized probabilities $P(t)/P_{max}$ curves for first (A) second (B) and (C) third neighbors. Different colors denote different $\alpha$HL monomer concentrations $c_m$ (see legend on panel C). Solid lines are the best fits using Eq(\ref{['HexaFit']}). Insets: log-log plot of the presence probability $P(t)$. Solid lines are polynomial fits at short times, using respectively (A) $P\sim \lambda (t-t_0)$ , (B) $P\sim (\lambda (t-t_0))^2$ and (C) $P\sim (\lambda (t-t_0))^3$.
• Figure 5: (A) Dependence of the characteristic rate $\lambda$ with the $\alpha$HL monomer concentration. Different symbols are different fit methods to compute $\lambda$. Error bars are standard error on the mean with typically 80 fits per concentration. The line is the best power-law fit of the data $\lambda\sim c_m^n$, with $=1.6\pm 0.2$. Inset: same plot, but differentiating the rank of neighborhood, between first (red), second (blue) and third (green) neighbors. Here, the error bars are standard deviation of data.