A Blind Spot in Confocal Reflection Microscopy: The Dependence of Fiber Brightness on Fiber Orientation in Imaging Biopolymer Networks
Fourth entry by Kelly Miller, AP225 Fall 2011
A Blind Spot in Confocal Reflection Microscopy: The Dependence of Fiber Brightness on Fiber Orientation in Imaging Biopolymer Networks. L.M. Jawerth, S. Munster, D.A. Vader, B. Fabry, and D.A. Weitz. (2010). Biophysical Journal, 98, L01-L03.
Two different confocal imaging techniques were compared in the context of imaging networks of biopolymer fibers. The two techniques compared were: confocal reflection microscopy (CRM) and confocal fluorescence microscopy (CFM). Fluorescently labeled type I collagen networks were imaged using both techniques and the images were compared. The CRM system is not able to detect fibers above more of a 50 degree angle (from the imaging plane). For this imaging system the brightness decreases for more vertically oriented fibers and therefore, the 3D network structure appears in the image, to be aligned with the imaging plane. The other system discussed in this paper, CFM, exhibits little variation of fiber brightness with the angle of the fiber and, as a result, an isotropic collagen network is imaged. Overall, CFM detects approximately twice as many fibers as are visible with CRM, and therefore, yield a more complete structural picture for 3D fiber networks.
The point of this paper is to provide a simple model to predict the detected fiber brightness as a function of the fiber orientation in CRM.
Collagen is a ubiquitous protein in mammals that constitutes a primary component of connective tissue in the interstitial space between cells. It appears under the microscope as a branched network of fibers, each of which can be resolved with confocal microscopy. To understand how cell-matrix interaction depends on the local environment of the cell - it is important to image the exact 3-D fiber environment of the cell. The most commonly used technique for imaging collagen networks is confocal reflection microscopy which uses back-scattered light to form an image. This method has been used to successfully obtain information on the collagen network such as morphology of collagen networks, mesh size, location and the orientation of the fibers.
However, sometimes the imaging of these fibers occurs in an anisotropic manner - producing images of the fibers aligned primarily with the imaging plane. This might occur from the intrinsic properties of the sample or the imaging method itself. To try to figure out what is producing the anomalous effects, it is essential to use an alternative imaging modality to examine the collagen structure.
The alternative technique that was used for this purpose was the confocal fluorescence microscope (CFM) which uses laser light to excite fluorophores in the imaging sample and forms an image from the emitted light.
Data was simultaneously collected using CRM and CFM on fluorescently labelled fibers. The brightness and orientation of individual fibers were analyzed. It was found that fiber brightness decreases in CRM with increasing fiber angle - which means that fibers above an angle of 50 degrees from the imaging plane will be entirely undetected. Consequently, the collagen structure appears aligned with the imaging plane.
CFM, on the other hand, detects fibers with similar brightness, independent of their orientation - therefore twice as many fibers are exposed and an isotropic network is displayed.
The authors then provide a description of an experiment that they performed to predict the detected fiber brightness as a function of fiber orientation in the CRM. To figure out if the apparent anisotropy seen in CRM is an imaging artifact and not an intrinsic sample property, they rotated the same by 90 degrees. For both types of imaging technique the dat from the rotated case closely match those of the original sample before the reorientation. It was concluded that the apparent anisotropy in the CRM data does not rotate and therefore the apparent anisotropy must arise from a bias in the CRM imaging technique.
This paper, aside from its relevance to the subject of soft matter, was incredibly interesting to me because I have just learned all about confocal microscopy, from AP 217. One of the things we addressed in 217 was isotropic resolution which seems to be the issue in this paper. I also found it very interesting because I didn't realize that collagen was a polymer.
The authors suggest future directions to improving the poor collagen image quality that comes out of CRM and for figuring out exactly why this happens. One thought was to coat the collagen fibers with gold particles before imaging. It has been suggested that labeling the collagen fibers in this way may enhance the reflective properties of the sample, making the CRM a more effective imaging system. Further investigations (from when this paper was written) are needed to confirm whether this is the case.