Neurons and synapses follow a “use it or lose it” paradigm—regular nerve firing promotes survival as well as maintenance of the dendrite tree. A paper in last week’s Neuron links neural activity and dendritic branching to memory, via nuclear calcium and vascular endothelial growth factor D (VEGF-D). Without this signaling, dendrites cannot maintain their full complexity, limiting learning and memory. The work reveals a role for VEGF-D above and beyond support for blood vessels, and suggests that therapeutics targeting the factor, such as those used in cancer treatment, could have unwanted side effects on cognition. First author Daniela Mauceri and senior author Hilmar Bading of the University of Heidelberg, Germany, led the study.
The evidence for VEGF-D’s involvement in arborization and memory is “very compelling,” wrote Albert La Spada of the University of California in San Diego in an e-mail to ARF.
When a nerve receives and transmits an action potential, it causes calcium to enter the nucleus, where the anion partners with calmodulin to activate gene transcription (reviewed in Greer and Greenberg, 2008). Bading’s team previously found that neuron survival depends on this nuclear calcium signaling (Lau and Bading, 2009; Papadia et al., 2005). In the current work, the scientists turned their attention to the role of calcium signaling in dendrite morphology.
When Mauceri expressed a calmodulin-blocking peptide, M13, in primary hippocampal neuron cultures, they observed shorter, less complex dendrite arbors with fewer, skimpier spines than in cells with normal calcium activity. The scientists went back to a transcriptome analysis they had already performed (Zhang et al., 2009) to identify VEGF-D as a candidate mediator of the arborization phenotype. VEGF-D turns on in response to both nuclear calcium and calcium/calmodulin-dependent protein kinase IV (CaMKIV), which is involved in dendrite and spine morphology (Chow et al., 2005).
To confirm the vascular factor’s role in arborization, Mauceri added recombinant VEGF-D protein to the cultured neurons. It rescued the dendrite phenotype caused by M13 treatment. She also used RNA interference to block VEGF-D production in neurons with normal calcium activity. The dendrites lost complexity just as if calcium were blocked. Unlike M13, neither VEGF-D nor its interfering RNA affected dendritic spines, suggesting that some other pathway involving calcium controls their density and morphology.
Next, the researchers took their work in vivo. Mauceri injected the hippocampi of adult mice with an adeno-associated virus carrying the anti-VEGF-D RNA, selectively knocking out the gene. The neural morphology mirrored that seen in vitro: short dendrites with less branching than normal.
One might think that, once dendrites set up their branches, they stay the same, Bading said. But the evidence says otherwise. “It looks like you really need something to maintain this arborization,” he said.
That maintenance is key to learning and memory, the researchers discovered. They tested mice that lacked hippocampal VEGF-D, and mice injected with a scrambled RNA interference sequence, in both the Morris water maze and a contextual fear-conditioning task. The VEGF knockdowns had trouble recalling the location of a hidden platform in the former, and were less likely to remember which chamber led to an unpleasant foot shock in the latter. “We saw a very dramatic phenotype,” Bading said. “Probably the circuitry is not connecting in the right way.”
Loss of VEGF-D activity could also be involved in neurodegeneration, Bading speculated. “I would be very surprised if there are not pathological conditions where VEGF-D expression falls…;which could then have impact on brain function,” he said. There is evidence of dendritic atrophy in aging and Alzheimer’s (see review by Dickstein et al., 2007). Other members of the VEGF family are involved in neurogenesis, neuroprotection, and learning and memory (Cao et al., 2004; Poesen et al., 2008; Le Bras et al., 2006; Licht et al., 2010). Certain VEGF-A alleles increase risk of amyotrophic lateral sclerosis due to reduced VEGF-A protein levels (Lambrechts et al., 2003; Lambrechts et al., 2008). However, Bading and colleagues did not find any evidence that VEGF members, other than VEGF-D, are involved in dendritic arborization.
While they are protective in nerves, VEGF proteins enhance angiogenesis, which can support tumor growth (Achen and Stacker, 2008; Stacker et al., 2001). Researchers are pursuing VEGF-D blockers as cancer treatments (Heath and Bicknell, 2009). Unfortunately, preventing VEGF-D activity in the nervous system could impair cognition. “If VEGF-D blockers get into the brain, it could be bad news,” Bading said.
Bading noted that there is an intriguing similarity between the protein’s neural and vascular roles: In both places, it promotes branching. The scientists have not yet worked out the full mechanism between VEGF-D signaling and dendritic arborization; VEGF-D appears to bind the type 3 VEGF receptor and activate the MAP kinase p38, but it is not clear if VEGF-D-mediated pathways toward angiogenesis and dendrite arborization are the same. La Spada suggested that a more detailed understanding of these mechanisms might help scientists design drugs that only target one or the other.
Mauceri D, Freitag HE, Oliveira AM, Bengtson CP, Bading H. Nuclear calcium-VEGFD signaling controls maintenance of dendrite arborization necessary for memory formation. Neuron. 2011 Jul 14;71(1):117-30. Abstract