High Order Chromatin Structure
Using Transmission Electron Microscopy and Digital Image Analysis
The height of the border on the insert can be used as a fiduciary
mark. The height = 430nm.
Transmission Electron Micrographs of Chromatin
Two light micrographs of a polytene chromosome(courtesy of J.S. Yoon)
High Order Structure
One puzzle about the chromosome is the how 30 nm fibers are arranged around the central core or axis. There is evidence that fibers are attached to scaffold and are radially arranged, as can be appreciated visually after denaturing the chromatin. Al
though there is little data to show how the fibers are condensed within these radial loops, a considerable amount of condensation must be required in order to fit the loop into the small radius of the chromosome.
The Solution
The Bowling Green group has broken the problem down into two separate questions. The first is whether there is orderly packing of the fiber. Jim Olesen, a former graduate student in the group, answered this question using a body of methods known as
stereology. The two-dimensional appearance of structures in sections is used to deduce features of the original three-dimensional organization. In this case, outlines of 30 nm fibers were traced from the digital image. Then their positions were pinpoin
ted and the distance between each pair of fibers, also called intercepts, was determined. Plotting a distribution function of the distances gave a sigmoidal inflection wherever there was a lamellar spacing. This theory is discussed in a forthcoming pap
er -- Olesen & Heckman (submitted).
The Technical Problems
Why has the problem of the higher order structure been so challenging? The group's recent work suggests that there were a number of technical problems. Deducing structure from a degraded image is one of these. The conventional protocols for embeddi
ng biological samples for electron microscopy involve extensive crosslinking, oxidation, and solvent extraction, especially of protein components. These changes, if drastic enough, could lead to a situation that is known in engineering parlance as "Garba
ge-in-garbage-out (GIGO)." In other words, the amount of information that can be extracted is limited by the quality of the specimen. Since the problem relates to scrambled information at the outset, image processing and analysis tools are mainly ineffe
ctive to address it.
The group solved the problem by embedding specimens directly, without prior fixation, in an innocuous organic polymer. They also got insight into another technical problem, in addition to GIGO. This was the fractal nature of the image of chromatin i
n thin sections. The image of the wall of the 30 nm fiber appeared to break down into smaller units, as the size of the array of pixels used in collecting the image was increased. The origin of this phenomenon appeared to be the contrast around the 10 n
m fiber, which was made visible as the magnification increased.
A description of methods and technical problems is found in: Heckman et al. (1995) Journal of Computer-Assisted Microscopy 7: 235-252.
E-mail heckman@bgnet.bgsu.edu to request a reprint.
Return to Carol Heckman's Top Page
Return to BGSU
Dept. of Biol. Scie. Top Page
Last updated 96-12-04