FGF2 was previously reported to promote neurogenesis and improve survival in a study of a transgenic mouse model of Huntington's disease (see Jin et al., 2005). FGF2 is elevated in the AD brain (see Cummings et al., 1993), possibly as a reaction to neuron damage. How this relates to pathology is unclear, but some studies suggest that FGF2 inhibits the maturation of new neurons in vitro (see Chen et al., 2007). "One of the leading opinions is that cell proliferation is increased in Alzheimer's disease, but its maturation is impaired," Ikezu told ARF. "Since FGF2 is upregulated in the Alzheimer's disease brain, [some scientists] argue that it drives neuronal stem cell proliferation, which may be antagonistic to cell differentiation."

FGF2 occurs in four isoforms—a low-molecular-weight form found in the cytoplasm that is also secreted from the cell, and three high-molecular-weight varieties that reside primarily in the nucleus and are not secreted. The various forms appear to have different biological activities. Many previous studies of FGF2 were done by bathing cells with the low-molecular-weight form, said Ikezu. In this study, he and his team wanted to know how boosting the larger, intracellular forms would affect mouse models of AD.

First author Tomomi Kiyota and colleagues generated APP/PS1 double-transgenics by crossing Tg2576 mice carrying the Swedish APP mutation with mice harboring the M146V presenilin 1 (PS1) mutation. The team bilaterally injected adeno-associated virus carrying the FGF2 gene—the gene for green fluorescent protein as a control—into mouse hippocampi either at four months of age or at seven to eight months, before and after symptoms began, respectively. Three to four months after the treatment, they tested mice for spatial memory using a radial arm water maze. They followed up with neuropathological analysis one month later.

Mice treated with AAV-FGF2 pre-symptomatically did not develop memory deficits, unlike controls, while deficits were reversed in post-symptomatically treated mice. The results suggest that FGF2 expression "can protect from, and reverse, memory acquisition and recall impairments commonly seen in the APP+PS1 mice," the authors wrote. To figure out which underlying cell processes led to the benefits, the team examined brain tissue. Staining for doublecortin, a marker of newly born neurons, told the researchers that FGF2-treated transgenic mice had significantly more premature neurons in the subgranular zone of the dentate gyrus relative to AAV-GFP-injected controls. In that layer of the hippocampus, FGF2-injected APP+PS1 mice also took up more BrdU, a marker of cell proliferation. Older, symptomatic animals activated the transcription factor C-fos when injected with the FGF2 vector, suggesting that growth factor encouraged synaptic gene expression. The team then looked at long-term potentiation (LTP) in the J20 mouse, which has known LTP deficits. Pre-treating those mice with AAV-FGF2 preserved synaptic plasticity. Together, the data suggest that injected FGF2 restores cell proliferation, maturation, and LTP.

What about Aβ pathology? FGF2 promoted Aβ clearance and reduced its production. Mice treated pre-symptomatically produced less total Aβ in the hippocampus, while the number of compact plaques were fewer than in controls. Post-symptomatic treatment reduced plaques as well. Tests of cultured mouse primary microglia demonstrated that low-molecular-weight FGF2—the secreted kind—enhanced microglial uptake of Aβ in a dose-dependent manner. Further, expression of high-molecular-weight FGF2 in mouse primary cortical neurons reduced Aβ production.

The results indicate that in mice, FGF2 treatment reduces Aβ pathology and rescues neuronal function after symptoms have begun. "That is important because most patients are symptomatic," said Ikezu. Whether this particular strategy could ever work in AD patients is unknown, but other growth factors such as GDNF (see ARF related news story on Love et al., 2005) and NGF (see ARF related news story on Tuszynski et al., 2005) have shown some efficacy in clinical trials.

"I think the study adds evidence that stimulating neurogenesis might be a promising clinical strategy for Alzheimer's disease," said David Greenberg, Buck Institute, Novato, California. He does caution, however, that since the authors did not test the effect of knocking down neurogenesis in this study, they could not provide a definitive link between the cognitive improvement and neurogenesis.

In terms of future treatment for disease, a gene therapy may not be the best treatment option for people because it is so invasive, said Ikezu. But there are other ways of enhancing neurotrophic factor production and neurogenesis in the brain. For instance, Fahnestock, Cotman, and colleagues have found that enriched diet and environment enhance BDNF production in dogs (see Fahnestock et al., 2010). Exercise has also been reported to increase FGF2 production in the rodent hippocampus (see Gómez-Pinilla et al., 1997). And pharmaceuticals, such as antidepressants, have been shown to stimulate neurogenesis in animals and are already approved for use in people, said Greenberg, though the link between antidepressants and neurogenesis is controversial (for a review see Hanson et al., 2011). "The simplest thing to do would be to study drugs that are already approved and known to be safe in people and try them in some of these neurodegenerative disorders," suggested Greenberg.—Gwyneth Dickey Zakaib.

References:
Kiyota T, Ingraham KL, Jacobsen MT, Xiong H, Ikezu T. FGF2 gene transfer restores hippocampal functions in mouse models of Alzheimer's disease: Therapeutic implications for neurocognitive disorders. Proc Natl Acad Sci USA, 2011 Oct 31. Abstract