The numbers of RGCs in untreated animals and animals that had received an intraorbital crush of the optic nerve were determined in flat-mounted retinas that were stained with antibodies to Brn-3a, a reliable marker for RGCs.
58 Quantitative analysis of retinas from untreated animals (
n = 6) revealed the presence of 3993.8 ± 54.4 (mean ± SEM) RGCs/mm
2. In animals that received an intraorbital crush of the optic nerve and intravitreal transplantation of control-NS cells, RGC numbers decreased to 189.8 ± 13.5 RGCs/mm
2, 97.0 ± 12.8 RGCs/mm
2, 50.2 ± 4.2 RGCs/mm
2, and 62.7 ± 5.0 RGCs/mm
2 at 1, 2, 3, and 4 months after the lesion, respectively (
n = 6 for each postlesion interval;
Figs. 4E–H,
Fig. 5). In animals with intravitreally grafted CNTF-NS cells, in comparison, we detected 529.0 ± 20.9 RGCs/mm
2 1 month after the lesion, 429.8 ± 30.7 RGCs/mm
2 2 months after the lesion, 320.5 ± 11.5 RGCs/mm
2 3 months after the lesion, and 302.7 ± 4.3 RGCs/mm
2 4 months after the lesion (
n = 6 for each postlesion interval;
Figs. 4A–D,
Fig. 5). CNTF-treated retinas thus contained 2.8-, 4.4-, 6.4-, and 4.8-fold more surviving RGCs than control retinas at the 1-, 2-, 3-, and 4-month postlesion interval, respectively. This difference between RGC numbers in CNTF-treated and control retinas was statistically significant at all postlesion time points analyzed (
P < 0.001 according to the Student's
t-test;
Fig. 5). Anterograde axonal tracing experiments performed in a fraction of animals from the different experimental groups at the different postlesion intervals confirmed complete transections of RGC axons in all lesioned nerves analyzed.