Nerve fibre morphometry with transmission electron microscopy: Application of the nucleator probe in ImageJ

Stereology and semiautomated binary image histomorphometry are two common methods used for morphometry of nerve fibres. Nucleator probe can be used for the estimation of morphometric parameters like diameter, perimeter, area and volume of a structure that is approximately either a circle or a sphere. In this study, we estimated these parameters with the help of ImageJ software on calibrated transmission electron micrographs. We procured samples of the cochlear nerve (CN) during winter months, within 6-12 hours of death, to reduce post-mortem autolytic changes. The temporal bones containing the CN were fixed by immersion in chilled paraformaldehyde. After dissecting out from the petrous part of the temporal bone, the CN were osmicated and processed for embedding in resin. From the resin blocks, silver coloured (70 nm) ultrathin sections were cut and picked on 300-mesh copper grids, stained with uranyl acetate and lead citrate and viewed under Tecnai G2-20 transmission electron microscope. The transmission electron micrographs had scale bars embedded into them by the software at the time of imaging, and the morphometric parameters of randomly selected nerve fibres were measured using the ImageJ software. The ImageJ software could become a low-cost and dependable tool for nerve fibre morphometry.• Nucleator probe is used for the estimation of morphometric parameters like diameter, perimeter, area or volume• Morphometric parameters were estimated by the ImageJ software on calibrated transmission electron micrographs• The ImageJ software could become a low-cost and dependable tool for nerve fibre morphometry


a b s t r a c t
Stereology and semiautomated binary image histomorphometry are two common methods used for morphometry of nerve fibres. Nucleator probe can be used for the estimation of morphometric parameters like diameter, perimeter, area and volume of a structure that is approximately either a circle or a sphere. In this study, we estimated these parameters with the help of ImageJ software on calibrated transmission electron micrographs. We procured samples of the cochlear nerve (CN) during winter months, within 6-12 hours of death, to reduce post-mortem autolytic changes. The temporal bones containing the CN were fixed by immersion in chilled paraformaldehyde. After dissecting out from the petrous part of the temporal bone, the CN were osmicated and processed for embedding in resin. From the resin blocks, silver coloured (70 nm) ultrathin sections were cut and picked on 300-mesh copper grids, stained with uranyl acetate and lead citrate and viewed under Tecnai G 2 -20 transmission electron microscope. The transmission electron micrographs had scale bars embedded into them by the software at the time of imaging, and the morphometric parameters of randomly selected nerve fibres were measured using the ImageJ software. The ImageJ software could become a low-cost and dependable tool for nerve fibre morphometry.
• Nucleator probe is used for the estimation of morphometric parameters like diameter, perimeter, area or volume • Morphometric parameters were estimated by the ImageJ software on calibrated transmission electron micrographs • The ImageJ software could become a low-cost and dependable tool for nerve fibre morphometry Specifications

Method details Background
Age-related alterations in the peripheral nerves can be studied by estimations of the number of nerve fibres, their diameters (total and of axons alone) and the thickness of their myelin sheath. These measurements require reliable and efficient methods. The most reliable method used for the evaluation of nerves is unbiased design-based stereology [1] and semiautomated binary image histomorphometry [3 , 9] . Hunter et al. [4] found that stereological estimations with light microscopy resulted in higher than expected values for nerve fibre and axon diameter than what is seen with histomorphometry and manual counting with the ImageJ software. They proposed that histomorphometry has potential advantages over stereology when stereology is limited to light microscopy alone.
The probe that is used to make these estimations is a local probe called the nucleator [2 , 5] . Unlike other unbiased probes, the nucleator assumes that the cross-section of particle of interest is roughly circular and employs the summation of the measured radii from an apparent centre and then utilizes the formulae related to the circle to estimate the circumference, area and volume (2 r, r 2 , 4/3 r 3 ), respectively [2] . The truth is that the profiles of nerve fibres are rarely circular; they are more often irregular than circular ( Fig. 5 in [4] ). Probably, the overestimations found by Hunter et al. [4] on light microscopic images may be minimised by increasing the number of isotropic rays employed in the nucleator probe, though it will make estimations cumbersome for the user [4] . Another limitation that may have affected the results is related to the poor resolution of the light microscopic images. The finer details of the nerve fibres, their boundaries and myelin sheath are best observed by transmission electron microscopy [8] . Therefore, in this study, we have measured the dimensions of nerve fibres on transmission electron micrographs using isotropic rays by adapting the 'straight line' tool available in the ImageJ software on the basis of the principle of the nucleator probe [2] . The freely available ImageJ software could become an inexpensive and reliable tool for morphometry at centres that do not have the resources to purchase expensive software for image analysis. A bone saw was used to remove the bony skull cap of the cadavers during autopsy. The dura mater covering the superolateral surface of the cerebral hemisphere was removed. The temporal lobe of cerebral hemisphere was retracted slightly backwards to visualize the brainstem. The cranial II to VI nerves were cut with a sharp scalpel blade and the VII and VIII nerves were visible entering the internal acoustic meatus. The VII and VIII nerves were cut at that level and the remaining cranial nerves, blood vessels were cut to detach the brainstem from the base of the skull. The medulla oblongata was cut by a transverse incision at the level of upper border of C1 vertebra in the posterior triangle of the neck. The brain was removed gently from the cranial cavity. The apical  portion of the petrous part of the temporal bone, along with the vestibulocochlear and facial nerves in the internal acoustic meatus, was carefully chiselled out after removing the brain from the cranial fossae. The bones with the nerves were fixed by immersion in chilled 4% paraformaldehyde made in 0.1 M phosphate buffer (PB, pH 7.4).

Dissection of the CN
The petrous part of the temporal bone ( Fig. 1 ) was fixed in a bone holder and the internal acoustic meatus (IAM) was de-roofed with a diamond drill (EC300/55,958, Medtronic Xomed, Inc. Florida, USA) in order to expose the CN within the IAM. The CN was cut in a plane that was transverse to its long axis, as close as possible to its exit from the cochlea, with a sharp razor blade. Segments (about 2 mm length) of the isolated pieces of the nerve were transferred into chilled fixative composed of 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M PB for 48 h. The nerve segments were osmicated, dehydrated and processed for making resin blocks. In the resin blocks, the nerve segments were placed in such a manner that the cochlear end of the CN remained as the cutting surface ( Fig. 2 ) .
Steps of tissue processing 1. Primary fixation-The 2 mm long segment of the CN was immediately immersed in chilled fixative (    8. Embedding-with pure resin in embedding mould at RT and relative humidity less than 50%. 9. Polymerisation -molds were kept at 50 °C for 1 day and then at 60 °C until solidification of resin was completed. To check completion of polymerisation, mould spaces filled with only resin without any tissue in it was used (blank blocks) ( Fig. 3 ).   ( Fig. 5 ) 6. Selected the magnifying glass (digital zoom) from the tool bar in the ImageJ window ( Fig. 6 A) and left-clicked on the image to magnify the scale bar ( Fig. 6 B) 7. Selected the 'straight line' tool from the tool bar in the ImageJ window and drew a straight line that would measure the length of the scale bar already present in electron micrographs (here it is 10 μm) ( Fig. 6 B). 8. Opened 'Analyze' menu and selected 'Set Scale' ( Fig. 7 ). A dialogue box is opened in which the distance of pixels is based on the length of the line drawn over the scale bar present in the micrograph. Filled in the known distance (written on the scale bar e.g. 10) and unit of length (written on the scale bar e.g. μm). Then clicked OK ( Fig. 8 ) 9. Opened 'Analyze' menu and then clicked on 'Measure' or pressed on ctrl + M ( Fig. 9 A) to see the results ( Fig. 9 B), where length measured is the actual length of the scale bar. 10. Drew four isotropic rays from the apparent centre of the transverse section of an individual nerve fibre forming two perpendicular test lines and measured distance from the centre of the nerve fibre to the inner and outer boundaries of the myelin sheath. These boundaries provided us the axon radius and nerve fibre radius, respectively ( Figs. 10-13 ). 11. Opened File and saved the results ( Fig. 14 ).        12. Calculated the area of the inner (occupied by axoplasm) and outer (axoplasm and myelin sheath) boundaries using the formula: a = r 2 , where 'r' is the mean of the measured radii of the line drawn from the centre to the inner or outer boundary of the transverse section [2] . 13. Calculated the diameters as 2r and myelin thickness as outer radius minus the inner one [7] . 14. Calculated the G-ratio by dividing the axon diameter by the diameter of the whole nerve fibre.

Note
Please go through Fig. 15 A and B representing a myelinated axon and screen capture of the digitised image of ultrathin section in one square of the copper mesh grid (left), the list of parameters in use on ImageJ software during image analysis morphometry (right), respectively from our published research article based on the above method [5] .

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
Data will be made available on request.

Funding
This work was supported by the Department of Anatomy, All India Institute of Medical Sciences, New Delhi .