Everything may die, nothing may be regenerated

It's 1991, 63 years after Santiago Ramon y Cajal, a histologist, neuroanatomist, and Nobel prize winner, wrote in his 1928 book, Degeneration and Regeneration in the Nervous System: "In adult centers the nerve paths are something fixed, ended, immutable. Everything may die, nothing may be regenerated. It is for the science of the future to change, if possible, this harsh decree." At the Salk Institute in La Jolla, Calif., Rusty Gage finds himself in possession of some interesting data.

His team genetically modified cells to overexpress fibroblast growth factor (FGF), and when the cells were mixed with hippocampal cultures, Gage observed a massive proliferation of neurons in the hippocampus. "We were testing the fibroblasts for their survival effects on neurons. And we saw this massive proliferation, which hadn't been recognized before with FGF," Gage says. "We had two events: proliferation, plus the cells could differentiate into neurons in culture." The findings were intriguing, but Gage was hesitant to jump to conclusions.

Neurogenesis defined: A rat hippocampal section stained for mature neurons (top left panel). The top middle panel shows a magnified view of the dentate gyrus granule cell layer with proliferating and migrating cells (orange). The top right panel shows a newly generated neuron (blue). Example of human NPCs cultured as a neurosphere (bottom left panel) in contrast to rat NPCs cultutred as adherent monolayers (bottom right panel).

At the time, much of the neuroscience community was skeptical of neurogenesis in the adult brain, Gage says. In the 1960s and 1970s experiments with radioactive thymidine labeling hinted that neurogenesis happened in vivo, but the resolution was poor, the evidence was inconclusive, and few believed in the phenomenon. Fernando Nottebohm at the Rockefeller University showed more convincingly in the 1980s that neurogenesis occurred in songbirds during vocal learning, but "everyone thought this was particular to birds," says Rockefeller's Bruce McEwen. Ramon y Cajal had established a dogma that persisted, despite attempts to disprove it, and it made perfect sense. "For one," says Gage, "how could a neuron divide? Neurons are too complex. And the second issue is, if you add new neurons in the brain you would disrupt the existing structure."

But the logic didn't slow down Gage's group. Several fellows in Gage's lab wanted to push the research further and see if the proliferative effect of FGF could happen in vivo. By 1995 Gage's lab had developed a method to isolate neural stem cells, grow them in culture, and transplant them into the rat hippocampus where they would differentiate. Also, a new microscope became available to image neurogenesis: confocal microscopy. Gage's lab had a beta station for a confocal microscope, which could locate dividing cells labeled with bromodeoxyuridine (BrdU). "I was still trying to decide in my own mind how much of my time I wanted to spend on this," says Gage, "but people just got excited about it."

Neurogenesis research was gaining momentum. At Rockefeller McEwen and his postdoc, Elizabeth Gould, had showed that hormones could affect neurogenesis in the dentate gyrus of rat.¹ "I think it really was a turning point in the acceptance of the phenomenon as being real," Gould says. In the late 1990s Gould observed neurogenesis in the adult tree shrew as well,² but the question remained: Did it occur in humans? "For us to believe it's more than an epiphenomenon, we needed to see if it occurs in humans," Gage says.

Fortunately, the experiment was essentially already set up. Some cancer patients received BrdU injections to check for dividing cells in tumor biopsies. Gage sent requests for brains to every pathologist he knew. When the brains arrived they were brought to the microscope room and indeed, BrdU lit up the confocal with labeled cells in the hippocampus and cortex. But the problem was in determining whether these cells were neurons. Double-labeling them with neuronal or glial markers was impossible because the tissue was fixed in paraffin.

The clinicians in Gage's lab were determined to figure out whether neurogenesis occurs in humans. Peter Eriksson, who had taken a sabbatical at the Salk, returned to the Sahlgrenska University Hospital in G￶teborg, Sweden, and gained permission from cancer patients injected with BrdU to use their brains once they died. Five brains, properly fixed and double- or triple-labeled with neuronal markers, afforded the first observations of neurogenesis in humans.³ "It was really exciting. You could imagine looking at these sections in the microscope for the first time. We'd bring people into the lab to have a look: 'Do you believe this?' Even though we had seen these newly born neurons in other species, it was really something to see it in humans."