Indian Journal of Ophthalmology

: 1987  |  Volume : 35  |  Issue : 1  |  Page : 11--16

Developing human optic nerve in prenatal period changes in the numbers of retinal axons.

S Wadhwa, V Bijlani 
 A. I. I. M. S. New Delhi - 110 029, India

Correspondence Address:
S Wadhwa
A. I. I. M. S. New Delhi - 110 029


A quantitative estimation of the number of axons in the optic nerves of human foetuses ranging from approximately 8 to 26 weeks has been made. This study shows that from 8 weeks there is a steady increase in the number of axons with subsequent decrease by 26 weeks, during the period of lamination in the dorsal lateral geniculate nucleus (dLGN).

How to cite this article:
Wadhwa S, Bijlani V. Developing human optic nerve in prenatal period changes in the numbers of retinal axons. Indian J Ophthalmol 1987;35:11-16

How to cite this URL:
Wadhwa S, Bijlani V. Developing human optic nerve in prenatal period changes in the numbers of retinal axons. Indian J Ophthalmol [serial online] 1987 [cited 2023 Jun 10 ];35:11-16
Available from:

Full Text


Naturally occurring cell death is a virtually universal phenomenon in the vertebrate nervous system. Ganglion cells of the retina and their axons which constitute the optic nerve are among the structures which undergo regressive events. Developmental studies in the chick [1],rat [2],[3],[4],[5], cat [6] and monkey [7] have demonstrated an excess of axons in the optic nerve, early in development with subsequent elimination of the superfluous axons. A number of factors have been put forth which may underlie this loss. In the present study the number of axons have been estimated in the human fetal optic nerves at different stages of gestation in the prenatal life to provide comparative observations in man to enable extrapolation of experimental data obtained from investigations of various visual defects particularly primates.


Seven optic nerves obtained from human foetuses ranging in gestational age from 8 to 26 weeks were processed for electron microscopy. The foetuses were obtained from hysterotomies performed for medical termination of pregnancy in the first half of gestation and autopsies on foetuses from second half of gestation who died after premature delivery. The gestational ages were calculated on the basis of crown-rump length [8] and biparietal diameter. The time elapsed between fixation of tissue and death of foetus varied from thirty to forty five minutes. 2-3 mm segment of optic nerve was taken from the intra-orbital part of the nerve immediately behind the eye ball. Specimens were fixed in Karnovsky's fixative for 12-14 hours and after fixation in 1% osmium tetroxide for two hours were dehydrated and embedded in araldite. Semithin 1 nim sections were cut perpendicular to the long axis of the nerve. When the full face of the nerve was obtained, thin sections were picked up on copper grids, stained with uranyl acetate and lead citrate and examined under the Philips 300 electron-microscope. Cross sectional area of each optic nerve was estimated from light micrograph of an adjacent semithin section excluding the nerve sheath and central blood vessels, if present. One to four electron-micrographs were taken from similar positions in each grid square such that the entire cross-section of the nerve was uniformly sampled thus eliminating biased sampling. The micrographs were taken at a magnification of x 5600 to x 20,000 and printed at a final magnification of x 10, 164 to x 39,787. All axonal profiles except for glial processes, growth cones and degenerating axonal, profiles were counted. The total number of axons were estimated by multiplying the number of axons counted with the ratio of total cross-section area (Ax) measured in the 1 um section to total area sampled (A) which was measured from the electron-micrographs. The value so obtained was corrected for 10% compression along the axis of ultrathin sections [6].


Light microscopic observations: With increasing age the cross-sectional area of the nerves enlarges from 0.046mm 2 at 8 weeks to 1. 533mm 2 at 26 weeks [Table 1]. At 8 weeks the nerve is enveloped by a single layer of cells while the mesenchyme begins to condense concentrically around it. Glioblasts are uniformly inter spread amongst the axon with attempt to isolate the axon fascicles. No connective tissue infiltration is seen at this stage [Figure l]. At 12 weeks a similar appearance exists with axon. fascicles being more clearly apparent. At 14-15weeks the connective tissue envelope thickens and increasing number of astrocytic profiles are seen. Few glioblasts are also present.

The astroglia are seen to surround and completely delimit the axonal fascicles. At 24 weeks there is increase in the thickness of septae surrounding the centrally placed axon fascicles while at 26 weeks considerable connective tissue infiltration is seen all over [Figure 2].

Electron-microscopic observations; The axons at all ages are rounded profiles hut occasionally oval or elongated. The axoplasm is pale and contains microtubules, filaments and very occasionally mitochondrion. Vesicles and vacuoles are seen in some profiles. Growth cones are identifiable by their irregular profiles, some organelles and vesicles [Figure 3]. Some large axonal profiles with vacuolation and/or dense material are seen and were regarded as degenerating axons [Figure 4]. Astrocytic processes contain dense bundles of filaments and ribosomal. clusters [Figure 5]. However, certain profiles are difficult to identify, as to whether they are astrocytic processes or axon in which case they were not included in the axonal counts. No myelinated fibres were seen upto 26 weeks which was the last age period studied [Figure 6].

Quantitative observations: At 8 weeks the axonal count is estimated at 0.33 million which increases to 1.84 million at 12 weeks and 4.29 million at 14-15 weeks. At 16-17 weeks there is a further increase to 7.24 million. Thereafter progressively lower estimates were obtained being 4.41 million at 19 weeks, 3.38 million at 24 weeks and 2.21 million at 26 weeks. In the period between 16-17 and 19 weeks 2.81 million fibres are lost while between 19 and 26 weeks another 2.2 million axons are eliminated [Table 1] & [Figure 7].


The present study shows an abundant production of retinal. axons which reaches a maximal count around 16 - 17 weeks of gestation with a tendency to decline to twice the adult levels around 26 weeks of prenatal life by elimination of excess axons. Approximately 53% axons are eliminated between 16 - 17 and 26 weeks. The loss is rapid between .16 - 17 and 19 weeks when 24% axons are removed over a period of three weeks while the remaining loss occurs gradually during the next six weeks. Similar pattern of excessive production followed by a loss has been demonstrated to occur in the chick, rat, cat and monkey [1],[2],[3],[4],[5],[6],[7]. Provis et al (9) while studying human fetal optic nerves also demonstrated an excessive production of fibres till 16-17 weeks followed by a rapid decline till 20 weeks and a gradual fall till 29 weeks with stabilization of adult values at 32 weeks of gestation. While there is a similarity in the general pattern of a rise in numbers of optic nerve fibres till 16 - 17 weeks and subsequent sharp decline till 19 weeks followed by a gradual fall till 26 weeks in the present study and that of Provis et al [9] there is a difference in the absolute numbers of axons estimated at different gestational ages in the two studies. This difference may be attributed to the different methods used in estimating the total counts of fibres in the optic nerves. The sharp reduction in the number of axons between 16 - 17 and 19 weeks of gestation occurs during the period of beginning of segregation of retinal inputs into the dorsal lateral geniculate nucleus (dLGN) while a more steady loss occurs pari passu with the appearance of distinct lamination in the nucleus and thereafter [10],[11]. During the period of lamination an initially diffuse overlapping projection of retinal axons into the dLGN has been shown in cat [12] and monkey [13] to be segregated into discrete laminar bands receiving input from the two eyes separately. In addition other target areas of retinal fibres like superior colliculi are also seen to be reorganized which may be associated with reduction in the number of optic nerve fibres [14]. The axons may also include transient aberrant retinoretinal fibres [15],[16]. In fact .it is now evident that these three regressive events are due to selective death of relevant retinal ganglion cells [16],[17],[18 and reduction in the number of optic nerve fibres may be attributed to these regressive events.


A quantitative estimation of the number of axons in the optic -nerves of human fetuses ranging from approximately 8 to 26 weeks has been made on electron micrographs. At 8 weeks there is an estimated 0.33 million axons in the' optic nerve. At 12 weeks the number increases to 1.84 million. At 14 - 15 weeks approximately 4.29 million axon are present. By 16 - 17 weeks a peak count of 7.24 million is reached which declines to 3.38 million axons by 24 weeks. The number continues to reduce to 2.21 million by 26 weeks. Thus an excessive production of axons is followed by rapid elimination with subsequent steady loss during the period of lamination in the dorsal lateral geniculate nucleus (dLGN) whence there is reportedly a segregation of inputs from the two eyes into the separate layers of dLGN.

Various reasons for this loss which is also seen to occur in the rat, cat and monkey have been postulated., In man this loss seemingly occurs over a protracted period of time.[18]


1Raeger, G., Raeger, U., 1978, Exp. Brain rtes., 33: 65
2Lam, K., Sefton, A.J. and Bennett, R., 1982, Dev. Brain Res., 3: 487
3Perry, V.H., Henderson, Z. and Linder, R., 1983, J. Camp. Neurol., 219: 356.
4Sefton, A.J. and Lam, K;, 1984, Exp. Brain Res., 57, 107.
5Crespo, D., O'Leary, D.D.M. and Cowan, W.N., 1985, Dev. Brain Res., 19: 129
6Ng, A.Y.K. and Stone, J., 1982, Dev. Brain Res., 5:263
7Rakic, P. and Riley, K.P., 1983, Science, 219: 1441.
8Hamilton, W.J., Boyd, J.D. and P1ossman, J.W., 1962 Human Embryology. 2nd Edition. W.Heffer and Sons Ltd., Cambridge.
9Provis, J.11., Van Driel, D., Billson, F.A. and Russel, P., 1985, J. Comp. Neurol, 283: 92
10Hitchcock, P.F. and Hickey, T.L., 1980, J. Comp. Neurol., 194: 395
11Damayanti, N., Wadhwa, S. and Bijlani, V., 1983, Indian J. Pled. Res., 77:279.
12Shatz, C.J.1., 1982, J. Neurosci., 3: 482
13Rakic, P., 1976., Nature, 261:467
14Land, P.W. and Lund, R.D., 1979., Science, 205:698
15Bunt, S.M. and Lund, R.D.,1981, Brain Res., 211: 399
16Bunt, S.11., Lund, R.D. and Land, P.W., 1983 Dev. Brain. Res., 6: 149
17O'Leary, D.D.11., Fawcett, J.W. and Cowan, W.P1., 1983, Soc. Neurosci. Abst., 9: 856.
18O'Leary, D.D.M. Fawcett, J.W. and Cowan, W.N., .1984, Soc. Neurosci. Abst., 10: 464.