Options for determination of 2-D distribution of collagen fibrils in transmission electron micrographs-Application to the mammalian corneal stroma

Averaged spacing and 2-d organization of collagen fibrils in the posterior cornea of the rabbit eye assessed by transmission electron microscopy

PURPOSE

To assess how reproducible collagen fibril spacing might be in the corneal stroma as viewed by transmission electron microscopy by calculating averaged values for the 2-D organization.

METHODS

One cornea from 8 albino rabbits (2 kg) was fixed in situ to preserve natural shape. Thin sections were stained with 2% phosphotungstic acid (PTA) and images taken of fibrils from the central-posterior stroma. After projection at 250,000 X magnification, an overlay was prepared of the fibrils. Using a 500 × 500 nm region of interest (ROI), the distances to all fibril centers were measured to 2 nm resolution.

RESULTS

The sets of fibrils had average diameters between 32.4 and 36.1 nm (group mean ± SD of 34.4 ± 4.2 nm). The mean fibril density was 396 ± 21 per square micrometer, with a fibril area fraction of 38.7 ± 3.9%. The mean distance to the literal nearest neighbor fibril center was 43.2 ± 4.5 nm. A radial distribution analysis showed a distinct nearest neighbors peak at 51 nm. This nearest neighbors peak had an average amplitude of 2.236 ± 0.315, with a broader secondary peak being evident in all data sets centered at 93 nm with an average amplitude of 1.166 ± 0.093 (or 53.3 ± 7.1% of the nearest neighbors peak).

CONCLUSIONS

Overall, these results show that a predictable 2-D organization can be demonstrated for collagen fibrils in rabbit corneas when consideration is given to sample selection and preparation and the image analysis strategy.

Effect of processing methods for transmission electron microscopy on corneal collagen fibrils diameter and spacing

INTRODUCTION

The corneal tissue was processed in fixatives and embedded in resin for transmission electron microscopy to observe the ultrastructure of the collagen fibrils (CFs). The effect of these processing methods on the CF diameter and the interfibrillar spacing was studied.

METHODS

Four normal human corneal buttons were used for this study. A part of each cornea was fixed in 2.5% glutaraldehyde containing cuprolinic blue in sodium acetate buffer and embedded in spurr's resin (SpurrCB). A second part of each cornea was fixed in 2.5% glutaraldehyde + osmium tetroxide and embedded spurr's resin (SpurrOsm). The third part of each cornea was fixed in paraformaldehyde (4%) and embedded in LR White at 4°C (LRWhite). Ultrathin sections were stained with uranyl acetate and lead citrate.

RESULTS

In the tissue, fixed in SpurrCB, the diameter was 38.4 ± 5.9 nm and spacing between CF was 52.5 ± 5.3 nm. In the tissue fixed in SpurrOsm, the diameter was 28.37 ± 5.84 nm and spacing between CF was 45 ± 4.57 nm. In the tissue fixed in LR White, the CF diameter was 24 ± 2.3 nm and spacing between CF was 39.0 ± 4.2 nm. The diameters and interfibrillar spacing of the tissue processed by SpurrCB, SpurrOsm, and LRWhite were significantly different (P < 0.001) from one another.

CONCLUSION

Our study shows that there is a variation in the CF diameter and spacing depending on the method of fixation and embedding resins used. This needs to be considered when comparative studies using different methods are done.

Assessment of collagen fibril spacing in relation to selected region of interest (ROI) on electron micrographs--application to the mammalian corneal stroma

AIMS

To evaluate measurements of collagen fibril spacing using different shaped regions of interest (ROI) on transmission electron micrograph (TEM) images of rabbit corneal stroma.

METHODS

Following glutaraldehyde fixation and phosphotungstic acid staining, TEM images of collagen fibrils in cross section were projected at a final magnification close to 250,000 × to obtain overlays. Interfibril distances (IFDs; center-to-center spacing) were measured within different ROIs of the same nominal area (0.25 μm(2) ) but different shape (with the length to width, L:W, ratio from 1:1 to 6:1). The IFD distribution was analyzed, and the 2D organization assessed using a radial distribution analysis.

RESULTS

The fibrils had an average diameter of 35.3 ± 3.8 (SD) nm, packing density of 393 ± 4 fibrils / μm(2) and a fibril volume fraction of 0.39 ± 0.02. IFDs ranged from 29 to 1400 nm depending on the shape of the ROI, with average values ranging from 263 to 443 nm. By artificially selecting IFD data only to a radial distance of 250 nm, the average IFDs were just 145-157 nm. The radial distributions, to 250 nm, all showed a nearest neighbors first peak which shifted slightly from predominantly at 45-54 nm with more rectangular ROIs. The radial distribution profiles could be shown to be statistically different if the ROI L:W ratio was 2:1 or greater.

CONCLUSION

Selection of an ROI for assessment of packing density and interfibril distances should be standardized for comparative assessments of TEMs of collagen fibrils.

Observations on the ultrastructure of equatorial scleral collagen fibrils in sheep eyes

OBJECTIVE

To investigate collagen fibrils of the equatorial sclera in relation to the age-related changes in eye size in sheep.

ANIMALS STUDIED

Lambs and outbred ewes.

PROCEDURES

Sheep eyes (three lamb and three from adult outbred ewes), presumed disease-free, were processed for transmission electron microscopy (TEM) immediately postmortem. Tissue blocks from the equatorial region were sectioned across fibril bundles orientated along the equator. Micrographs including at least 500 fibrils were projected at 22,000× magnification for measures of fibril diameters (FDs).

RESULTS

Lamb eyes were smaller than those of adult ewes but equatorial scleral thickness was only marginally less at 0.232±0.013 vs. 0.254±0.012 mm (P value not significant). Scleral tissue was composed of compacted bundles of collagen fibers that tended to be rounder in outer compared to being flatter in inner regions. In typical (normal) appearing regions, FDs were distinctly larger (68-410 nm) in outer sclera compared to inner sclera (63-281 nm). Outer sclera FDs were bimodal averaging 192±58 nm, compared to unimodal distributions at inner locations averaging 156±48 nm (P<0.001). Some atypical regions, especially at outer-mid sclera locations, were also noted where the FD distribution was bimodal but also included numerous microfibrils (<50 nm diameter), with similar appearances being found for both lamb and adult ewe eyes.

CONCLUSIONS

The equatorial sclera is a mixture of rounder versus flatter collagen fiber bundles, the former being more likely to be made up of a mixture of both smaller and larger fibrils, as compared to slightly smaller fibrils.