Pick up any article on neuronal development in adulthood, and there is a good chance you will read that the birth of new neurons has been observed in the hippocampal region of the brain in every mammalian species examined, including humans. This idea underlies the view — widespread among neuroscientists — that analysis of such neurogenesis in animals can benefit our understanding of learning, emotional disorders and neurodegenerative disease in humans. But in a paper in Nature, Sorrells et al.1 report that, unlike in other mammals, the last new neurons in the human hippocampus are generated in childhood. These findings are certain to stir up controversy.
Today, it is common knowledge that the brain can change according to needs and demands. But it was not always so. In the 1960s, biologist Joseph Altman reported that new neurons are generated in the adult brain, specifically in a hippocampal subregion called the dentate gyrus, which is now known to be crucial for memory2. Further research languished, however, owing to scepticism about the brain’s capacity for such dramatic plasticity. It wasn’t until the 1990s, with the development of improved techniques for visualizing brain cells, that acceptance of adult neurogenesis became widespread3.
Although the scope and function of neurogenesis remain debatable, there has been a general consensus that the hippocampus is one region in which adult neurogenesis exists in humans as it does in animals. This is based on several studies. For example, one study in patients given a synthetic nucleoside molecule called bromodeoxyuridine (BrdU) showed that it had been incorporated into the DNA of dividing cells in the dentate gyrus4. Another found that protein markers of neurogenesis in animals were present in post-mortem human brain tissue5, and a third used radiocarbon dating to identify hippocampal-neuron turnover6. However, methodological challenges make human studies difficult to interpret, and more are required to make definitive conclusions.
Sorrells et al. set out to address this need using classic immunohistochemical techniques in which specific antibodies are bound to proteins of interest, revealing their locations in tissue. The authors used this strategy to count neural precursor cells, proliferating cells and immature neurons in samples from 59 human subjects, spanning fetal development through to old age (Fig. 1). They found streams of all three cell types migrating from an embryonic ‘germinal zone’ to the developing dentate gyrus at 14 weeks of gestation. By 22 weeks, migration was reduced, and immature neurons were largely restricted to the dentate gyrus. And there were many fewer immature neurons at one year of life than at earlier stages. The oldest sample containing immature neurons was taken from a 13-year-old individual. These findings are in stark contrast to the prevailing view that human hippocampal neurogenesis extends throughout adult life.
Is it possible to reconcile the findings with previous human data? Although direct comparisons are difficult, Sorrells et al. offer some explanations. For example, they find that DCX and PSA-NCAM, two proteins that reliably mark immature neurons in animals, can label mature neurons and non-neuronal glial cells in humans. Indeed, the authors show that these two markers unambiguously identify immature neurons only if both are expressed in a cell. Similarly, the group demonstrated that it is possible to obtain BrdU-like immunohistochemical labelling in tissue that did not actually contain BrdU. Nonspecific labelling could therefore have led to false-positive results in previous studies.
The researchers’ careful approach also speaks to the challenges of performing neurogenesis work in humans. Animal studies have shown that PSA-NCAM is modified by previous experiences7 and that DCX degrades if tissue is not rapidly preserved8. An apparent loss of neurogenesis could therefore reflect changes in marker expression, especially if stringent criteria are used to define new neurons. Given that there are debates about hippocampal precursor-cell identity even in rodents9, it is also possible that we simply do not know what to look for in humans.
Sorrells et al. minimized these issues in several ways. First, they observed neurogenesis in the hippocampus of infants and children, which served as a positive control. Second, they used a variety of adult samples to minimize the possibility that problems with tissue health or preservation could confound their results. Third, they used diverse markers of neurogenesis to gain multiple lines of evidence. Nonetheless, further investigation will be needed to see whether Sorrells and colleagues’ conclusions will stand the test of time.
How do the authors’ findings fit with the animal literature? With a bit of conceptual recalibration, they might fit quite well. Rodents are born with relatively immature nervous systems, so adult rodent neurogenesis could be a decent model of neurogenesis in children or adolescents. Given that depression, schizophrenia and Alzheimer’s disease are rooted in early hippocampal defects, even neurons generated in childhood could have a key role in the aetiology of disease in humans. In addition, primate data10 suggest that new neurons in humans could go through an extended period of maturation (years or even decades) relative to what occurs in rodents, during which time they might have enhanced plasticity and important functional properties. Thus, whereas the continual addition of new neurons might provide plasticity in adult rodents, the prolonged development of neurons could provide a similar plasticity in adult humans.
At the other end of the developmental spectrum, even in rodents, neurogenesis is very low by middle age2. Thus, Sorrells and colleagues’ human data again are not wholly inconsistent with the animal literature. If the focus of rodent studies were shifted to identifying the mechanisms by which neurogenesis diminishes over time, and to how neurogenesis can be enhanced to offset pathology caused by age and disease, we just might be able to translate the authors’ sobering findings into discoveries that improve human health.
Nature 555, 315-316 (2018)
Sign up for the daily Nature Briefing email newsletter
Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.