Automated Methods for the Analysis of Skeletal Muscle Fiber Size and Metabolic Type - International Review of Cell and Molecular Biology - page 303

Figure 7.11 Automated fiber-by-fiber quantification of COX activity. (A) An image rep-
resenting COX activity is shown. (B) The fiber mask from the image, produced from
CyteSeer ® is shown along with the unique identification number for each fiber. (C)
A portion of the image (from the rectangular region in (A)) is shown along with values
that correspond to the average COX intensity of each fiber.

represents increased COX activity was imported into CyteSeer ® ( Fig. 7.11 ).
Since there is no COX activity in the extracellular space between the fibers,
the white spaces in the image were adequate for the algorithm to identify the
borders of the fibers, analogous to laminin-labeled muscle sections visualized
utilizing fluorescence methods, turning the original image into the template
for automated analysis of the fiber boundaries ( Fig. 7.11 B). To provide an
image in which the labeling was proportional to the enzymatic activity,
the image was duplicated, and processed, by converting to gray scale and
inverting the intensities (processed image not shown). Projection of the fiber
outline mask onto the processed image yielded a fiber-by-fiber index of
COX activity ( Fig. 7.11 C).
A quantitative, fiber-by-fiber analysis of mitochondrial enzymatic activ-
ity is likely to represent an improvement over current methods, as human
skeletal muscle is typically a mixture of Type I and Type II fibers. Thus,
methods that average enzymatic activity across the entire sample (as is
achieved by preparing tissue homogenates) do not account for potential var-
iability of Type I/Type II fibers likely to be found in skeletal muscle biopsies
especially in pathological samples.
A strategy to identify enzymatic activity and to correlate the enzymatic
activities with fiber type would be to perform bright-field analysis of mito-
chondrial enzymatic function on skeletal muscle tissue sections and then to
label the same sections for fiber-type specific biomarkers, such as MyHC- b

for Type I fibers. To test whether clinically relevant samples could be
colabeled with both colorimetric and fluorescent labels, a series of human
muscle biopsies were tested in a variety of ways. First, each biopsy specimen,
which had been previously stained via an enzymatic reaction for NADH