Why Cortical Neurons Cannot Divide, and Why Do They Usually Die in the Attempt?
Neurons are the prime example of excitable cells able to process and transmit information through electrical and chemical signals. A current estimate suggests that the human brain contains approximately 8.6 × 109 neurons, among which 19% belong to the cerebral cortex, which makes up approximately 82% of the brain mass (Herculano‐Houzel, 2009). It is estimated that the average cortical neuron connects through synapses with at least 10,000 other neurons, thus allowing the cortex to function as the center for memory, learning, and reasoning (Gilbert, 2010). The structural complexity of the mammalian cerebral cortex arises through a process of cellular migration from germinal centers at the ventricular and subventricular zones of the neural tube, where the neuronal precursor cells (neuroblasts) reside. Neuroblasts divide to generate neurons that migrate into the cortical plate. For the migrating neurons, this event represents the end of their proliferating potential, so they become postmitotic (unable to divide) through the remainder of their life span (Rakic, 1974), which, in the case of human cortical neurons, could mean several decades. This postmitotic condition is a property of mature, terminally differentiated cells, and, in the case of cortical neurons, this condition is remarkably stable, instantiated by the fact that there are currently no reports of spontaneous cancer arising from human cortical neurons or from carcinogen‐treated cortical neurons (Rakic, 2006). For example, classical studies with rats have shown that intracerebral exposure to powerful chemical carcinogens results in brain tumors in about 35% of treated animals, but such tumors are either gliomas or sarcomas and are never neuronal in kind (Zimmerman and Arnold, 1941). Additional studies have shown that chemical carcinogens may cross the blood–brain barrier and then be metabolized inside neurons into mutagenic species; however, the result of this is not tumorigenesis but toxic neuronal death (Dutta et al., 2010). Currently, only one isolated article has suggested the tumorous transformation of mature neurons in transgenic mice by ad hoc experimental manipulations. However, the resulting tumors are always gliomas, with no remaining evidence of the neuronal phenotype (Friedmann‐Morvinski et al., 2012). This article revives the old claim that neuronal dedifferentiation into progenitor cells is a possible road for tumorigenesis in the brain, but there is scope for doubting whether the labeling and, thus, the identification of the artificially transformed neurons were truly unambiguous (Rakic, 2002). Indeed, the stability of the postmitotic state in mammalian cortical neurons is such that current expert opinion suggests that brain tumors may arise only from the stem or precursor cells naturally present in the postnatal brain but never from the terminally differentiated cortical neurons (Dirks, 2008; Demir et al., 2009; Ma et al., 2009).
It has been suggested that during evolution there has been a trade‐off between increasing brain complexity and the proliferating capacity of neurons (Kempermann et al., 2004; Rakic, 2006). This suggests that in mammals the preservation of acquired information and network architecture in the standing population of cortical neurons is more valuable to the organism than the potential introduction of new, “inexperienced” neurons to the cortex (Rakic, 1985; Kempermann et al., 2004). Thus, one may wonder whether the remarkable stability of the postmitotic state in the cortical neurons of mammals has been the subject of natural selection and whether it depends on the action of specific genes and their products.
Indeed, all biological states or processes controlled by genes and gene products can be inhibited, reverted, or bypassed by spontaneous or induced mutations in the genes involved.