Comparison of estimates of neuronal number obtained using the isotropic fractionator method and unbiased stereology in day old chicks (Gallus domesticus)

Abstract

BackgroundThe relative size and neuronal density of brain regions are important metrics in both comparative and experimental studies in neuroscience. Consequently, it is imperative to have accurate, reliable and reproducible methods of quantifying cell number. New methodThe isotropic fractionator (IF) method estimates the number of neurons and non-neurons in the central nervous system by homogenizing tissue into discrete nuclei and determining the proportion of neurons from non-neurons using immunohistochemistry (Herculano- Herculano-Houzel and Lent, 2005). Comparison with existing methodOne of the advantages of IF is that it is considerably faster than stereology. However, as the method is relatively new, concerns about its accuracy remain, particularly whether homogenization results in underestimation of cell number. In this study, we compared estimates of neuronal number in the telencephalon and ‘rest of brain’ (i.e. the diencephalon and brainstem excluding the optic lobes) of day old chicks using the IF method and stereology. ResultsIn the telencephalon, there was a significant difference in estimates of neuronal number between the 2 methods, but not estimates of neuronal density (neurons/mg of tissue). Whereas in the ‘rest of brain’, there was a significant difference in estimates of neuronal density, but not neuronal number. In all cases, stereological estimates were lower than those obtained using the IF method. ConclusionDespite the statistically significant differences, there was considerable overlap (all estimates were within 16% of one another) between estimates obtained using the two methods suggesting that the two methods provide comparable estimates of neuronal number in birds.