Graphene-based sensing of oxygen transport through pulmonary membranes

Abstract

Lipid-protein complexes are the basis of pulmonary surfactants covering the respiratory surface and mediating gas exchange in lungs. Cardiolipin is a mitochondrial lipid overexpressed in mammalian lungs infected by bacterial pneumonia. In addition, increased oxygen supply (hyperoxia) is a pathological factor also critical in bacterial pneumonia. In this paper we fabricate a micrometer-size graphene-based sensor to measure oxygen permeation through pulmonary membranes. Combining oxygen sensing, X-ray scattering, and Atomic Force Microscopy, we show that mammalian pulmonary membranes suffer a structural transformation induced by cardiolipin. We observe that cardiolipin promotes the formation of periodic protein-free inter-membrane contacts with rhombohedral symmetry. Membrane contacts, or stalks, promote a significant increase in oxygen gas permeation which may bear significance for alveoli gas exchange imbalance in pneumonia.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1. Sensing platform for oxygen transport through pulmonary membranes.

a Schematic illustration of lipid membranes deposited on the FET-based sensor which serves as a solid support for stacked lipid bilayers. b Bright field optical microscopy image of the oxygen gas sensor. The white arrow indicates the graphene channel between gold electrodes. Channel length and width are 50 μm. Inset: photograph of the sensor which has total 50 FETs on a 1 cm × 1 cm SiO2/Si substrate. c Fluorescence microscopy image of fluorescently labeled lipid membranes coated on the FET sensor. Graphene is marked with a white arrow. Inset: photograph of the sensor coated with lipid films.

Fig. 2. Electric characteristics of graphene FET sensors.

a Schematic image describing how the FET-based sensor detects the oxygen gas using graphene. Black arrows represent the direction of drain current (Id). b Transfer characteristics of graphene FET. Drain voltage (Vd) is 1 V and gate voltage (Vg) is swiped from −5 to 10 V. Blue and red curve indicate Id–Vg curves of the FET sensor without and with lipid films, respectively. c Sensing response calibration graph of the oxygen gas sensor as function of oxygen percent from 0 to 80% with and without lipids. Dots represent the real response data. Blue and green lines are linear fits of with and without lipids, respectively (correlation coefficient (R2) = 0.998).

Fig. 3. Film thickness of model systems and lung extracts.

AFM 3D height images and cross-sectional profiles of a healthy model systems composed of 3:1 DPPC:DOPG, scan size 4 μm × 4 μm, b diseased model system consisting of 3:1 DPPC:DOPG with addition of cardiolipin and Ca2+, scan size 8 μm × 8 μm, c healthy bovine lipid–protein extract surfactant (BLES), scan size 11 μm x 11 μm, and d diseased BLES with cardiolipin and Ca2+. Blue arrows display scanning directions of profiles, scan size 10 μm × 10 μm.

Fig. 4. GISAXS of model diseased pulmonary membranes.

a Synchrotron GISAXS of the diseased lung membrane model system. White arrows indicate stalk phase peaks which appear along the qz direction at qxy~±0.05 Å−1. b Schematic illustration of a stalk phase consisting of spatially ordered inter-bilayer contacts across a membrane with four stacked bilayers. Blue and green areas represent possible in-plane domains. c–f In-house GISAXS diffraction patterns of healthy model systems (c) and model systems with Ca2+ 4 mol% (d), with Ca2+ 4 mol% + cardiolipin (CL) 3 mol% (e) and with Ca2+ 4 mol% + CL 8 mol% (f). Stalk-phase peaks are indicated by white arrows. Oxygen gas saturation times (tsatO2) depicted in the GISAXS images corresponds to the time that FET sensors take to get saturated with a flow of 20% oxygen gas.

Fig. 5. 1D GISAXS and AFM of model systems and lung extracts.

1D GISAXS intensity versus reciprocal space q (Å−1) profiles for BLES samples and model membrane systems in healthy (a) and diseased (b) states. The middle panels are schematic representation of stalks and orientation in healthy (c) and diseased (d) model (blue) and BLES (magenta) membrane systems. The space between bilayers (light blue) represents water layers. AFM height and phase images of healthy (e) and diseased (f) model membrane systems are represented in the figure. Inset in the healthy model (e) is 250 nm × 250 nm while inset in the diseased model (f) highlighting the pores in the diseased model system is 500 nm × 500 nm, white arrows point to pore-like defects. Height and phase AFM images of lung extract (BLES) in healthy (g) and diseased (h) states are also displayed.

Fig. 6. Mechanism of enhanced oxygen gas transport through stalk phases.

Schematic illustrations of model systems and lung extracts in healthy and diseased states and how their different structures affect oxygen gas permeability. Enlarged illustrations (circles) describe detailed structures of each system, displaying lipids (e.g., DPPC, DOPG, Cardiolipin, and BLES), surfactant proteins (SP-B and SP-C) and calcium ions.