Age-related cognitive decline affects up to one-quarter of adults over the age of sixty1. A healthy diet and regular exercise can help to prevent this decline, but as yet there are no treatments to reverse it2. Progress in understanding how the brain changes during development and ageing has sparked tantalizing ideas for harnessing youthful factors to slow age-associated cognitive changes — or even to rejuvenate the ageing brain. Writing in Nature, Iram et al.3 bolster this line of thinking by tapping into the cerebrospinal fluid (CSF), which bathes the brain tissue and contains several protein growth factors necessary for normal brain development.
The authors infused CSF from young adult mice (10 weeks old) into the brains of aged mice (18 months old) over 7 days. This treatment improved the memory recall of the old animals in a fear-conditioning task, in which they learnt to associate a small electric shock with a tone and flashing light. Iram and colleagues then used RNA sequencing to determine how CSF treatment altered gene expression in the hippocampus — a key memory centre in the brain that is often the focus of studies of age-associated cognitive decline.
Cells in the central nervous system called oligodendrocytes produce myelin, a fat- and protein-rich material that insulates neuronal fibres called axons. Myelination of axonal projections throughout the brain ensures that strong signal connections are maintained between neurons. Iram and colleagues found that genes that are typically expressed in oligodendrocytes were highly upregulated in old mice treated with CSF from young mice, compared with the animals’ aged counterparts treated with artificial CSF.
Previous work has demonstrated that successful fear-conditioning in mice requires oligodendrocyte proliferation and myelin formation, and that disruption of this process impairs memory4,5. The authors therefore examined whether treatment with the young CSF affected the proliferation and maturation of oligodendrocyte precursor cells (OPCs). Indeed, they found that young CSF more than doubled the percentage of OPCs actively proliferating in the hippocampus of old animals. This cellular change was followed three weeks later by an increase in myelin production. The findings strongly suggest that young CSF improves the cognitive abilities of aged mice by modulating oligodendrocytes.
The authors took a deeper dive into the pathways activated by young CSF using an established line of rat OPCs grown in cell culture. Gene transcription involves the formation of chains of various nucleoside molecules to make RNA, so Iram et al. added a labelled nucleoside to the culture medium in which the OPCs were grown — this enabled the authors to isolate and sequence newly made RNA transcripts, which had incorporated the labelled nucleoside. The greatest increase in gene expression in response to young CSF treatment was in serum response factor (Srf), which encodes a transcription factor that initiates cell proliferation and differentiation. Six hours after young human CSF administration to the OPCs, Srf expression had returned to baseline levels, but downstream targets related to the cell cycle and proliferation were upregulated. The authors confirmed that these SRF signalling pathways were also activated in old mice after young CSF administration.
CSF contains a rich cocktail of signalling molecules and growth factors, many of which could induce the SRF signalling pathways seen in the OPCs. Iram et al. identified candidate factors capable of inducing Srf expression in published protein databases from large-scale studies of CSF. Fibroblast growth factor 17 (FGF17) emerged as a compelling candidate. The authors showed that the protein is robustly expressed in mouse neurons, exhibits decreased expression in aged mice, and induces OPC proliferation in cultured rat OPCs.
FGF17 infusion into old CSF partially recapitulated the effects of the young CSF, both in vitro in OPCs and in vivo, improving the memory recall of aged mice (Fig. 1). Finally, the authors demonstrated that the blockade of FGF17 in cultured OPCs treated with young CSF was sufficient to inhibit OPC proliferation, and that treatment of young mice with FGF17 blockers impaired cognition. The researchers’ experiments strongly suggest that FGF17 is a CSF-borne factor crucial for cognition, and demonstrate that its effects are probably mediated by oligodendrocytes and myelination in the hippocampus.
Iram and colleagues’ finding adds FGF17 to a growing list of factors that affect neuronal development and cognition and that are known to change with ageing6,7. There has been particular interest in CSF and its signalling factors during brain development, when neuronal progenitors depend on these signals to proliferate and build the cerebral cortex8,9. The CSF also has roles in adult mice, for instance in influencing neuron production in the brain’s subventricular zone10. Previous work has often found that beneficial CSF factors originate in the choroid plexus — a sheet of tissue located in each ventricle of the brain that secretes CSF and forms a major barrier between the brain and the rest of the body. Unexpectedly, Iram and colleagues found that FGF17 in the CSF isn’t sourced by the choroid plexus, but, instead, by youthful neurons themselves, providing evidence that neuron-based signals are delivered by the CSF — a provocative hypothesis about protein and fluid distribution throughout the brain. How FGF17 is distributed in the CSF and delivered to target cells in the hippocampus presents a new direction of research.
Iram and colleagues have broken ground in the field of brain health and ageing by discovering that young CSF contains a factor that aids memory recall in older mice through oligodendrocyte maturation and myelination in the hippocampus. Not only does the study imply that FGF17 has potential as a therapeutic target, but it also suggests that routes of drug administration that allow therapeutics to directly access the CSF could be beneficial in treating dementia. Any such treatments will be hugely helpful in supporting our ageing population.