Date Published: April 4, 2017
Publisher: Public Library of Science
Author(s): Rubén Cárdenes, Chong Zhang, Oxana Klementieva, Stephan Werner, Peter Guttmann, Christoph Pratsch, Josep Cladera, Bart H. Bijnens, Thomas Abraham.
Structural analysis of biological membranes is important for understanding cell and sub-cellular organelle function as well as their interaction with the surrounding environment. Imaging of whole cells in three dimension at high spatial resolution remains a significant challenge, particularly for thick cells. Cryo-transmission soft X-ray microscopy (cryo-TXM) has recently gained popularity to image, in 3D, intact thick cells (∼10μm) with details of sub-cellular architecture and organization in near-native state. This paper reports a new tool to segment and quantify structural changes of biological membranes in 3D from cryo-TXM images by tracking an initial 2D contour along the third axis of the microscope, through a multi-scale ridge detection followed by an active contours-based model, with a subsequent refinement along the other two axes. A quantitative metric that assesses the grayscale profiles perpendicular to the membrane surfaces is introduced and shown to be linearly related to the membrane thickness. Our methodology has been validated on synthetic phantoms using realistic microscope properties and structure dimensions, as well as on real cryo-TXM data. Results demonstrate the validity of our algorithms for cryo-TXM data analysis.
Biological membranes provide specialized permeability barriers for cells and cell organelles, in which the interplay of lipids and membrane proteins facilitates basic processes of respiration, photosynthesis, protein and solute transport, signal transduction, and motility . The cell membranes of almost all living organisms and many viruses are made of two layers of lipid molecules (a bilayer) containing various types of protein molecules, which are typically 5–10nm in thickness. Nuclei and other sub-cellular structures are also surrounded by one or more lipid bilayers. The composition and content of membrane can be modified in different kinds of cells, including the healthy as well as the pathological ones. The modifications are extensive enough to alter membrane fluidity and affect a number of cellular functions . Such dynamic nature of membranes are also reflected as their structural changes, such as thickening or reduction in membrane thickness, or lipid bilayer width [3–6].
In order to evaluate the performance of our methodology of segmentation and subsequent membrane quantification, the above described experiments have been carried out. The first set of experiments uses the phantom made from simplified geometries similar to biological structures with realistic properties of a cryo-TXM system, so as to validate our segmentation algorithm. The second set of experiments uses the phantom mimicking biological membranes with variations, in order to validate our membrane quantification metric. And the last set of experiments uses the measured data obtained from a synchrotron soft X-ray microscope, to investigate the feasibility of applying our method to real biological problems.
The proposed 3D segmentation algorithm introduces the compatible ridge selection concept to better deal with noisy membrane structures. It has been validated on a synthetic image phantom using realistic microscope properties and structure dimensions, including the simulation of the missing wedge effects of cryo-TXM. Our segmentation approach constitutes minimal user effort with very satisfactory results, and we consider it only part of the contribution of the paper, that could eventually be changed by other method. Once properly segmented, a membrane quantification procedure is proposed through measuring features on membrane grayscale profiles perpendicular to the 3D membrane surfaces. These measures have been investigated using a synthetic membrane phantom that incorporates the PSF of the microscope. The newly introduced area metric has shown to be linearly related to the membrane thickness and the absorption coefficient of the membrane constituent material. This enables us to extract information about structural changes in membrane thickness or in the concentration of membrane components. Our segmentation and quantification algorithms require very little user interaction, thus reducing greatly the amount of time spent in manual analysis typically found in studies with data from soft X-ray transmission microscopes. The presented phantom validation, as well as real data experiments, have demonstrated the feasibility towards exploring the capability of cryo-TXM to detect changes in biological membranes at near native-state. Although we have shown theoretically that it is possible to obtain valid measures with a zone plate with a spatial resolution of 40nm, a better quantification with a zone plate with higher spatial resolution could be expected. The real data sets we have investigated are plasma and nuclear membranes of SH-SY5Y human neuroblastoma cells treated with Amyloid β-peptide to investigate its interaction with these membranes. From the results obtained with our method, we have validated our algorithms and showed that they work. Although results are preliminary, this pilot study has initiated a potential interesting biological question to be further investigated on larger numbers of datasets. We also expect that the methodology is not limited to the application addressed here, but could be applied to other type of cells and organelles to analyze their properties.