First author Wen-Biao Gan and colleagues used transgenic mice that express yellow fluorescent protein (YFP) in neuronal axons to investigate the dynamics of synapses. The authors chose to focus on neurons in the submandibular ganglion in the neck because these cells are easy to image in vivo. Furthermore, as the mice used do not express YFP in the postganglionic neurons, it is easier to decipher the often complex network of synaptic connections.

Using confocal microscopy, Gan and colleagues were able to image the same cells over a period of months and even years. They found that in young animals (40-45 days old), just over 70 percent of terminals were stable over a two-week period, but this number rose to over 80 percent for mice that were 32 months old. Over a one-month interval, the ratio fell slightly, as may be expected, to just less than 70 percent for young animals, but again, older mice (12 months) showed more synaptic stability (about 78 percent of terminal being stable).

This work seems to complement Gan's previous data showing that dendritic spines turn over quite slowly in the visual cortex. (In contrast, Karel Svoboda and colleagues at Cold Spring Harbor, have shown that spines can turn over quite rapidly, but their observations were made in the barrel cortex. See ARF related news story.) Gan suggests that stability of the axonal synapses "helps assure that functionally appropriate connectivity will not undergo refinement without good reason."

Also, in last week's Nature Genetics, Charles Cohen-Salmon and colleagues at the French Centre National de la Recherche Scientifique in Orleans and Marseille report that mitochondrial DNA (mtDNA) can influence learning and development in the brain, an impact that increases as animals age.

First author Pierre Roubertoux and coworkers made this connection after comparing two strains of congenic mice that harbor different mitochondrial genomes. Starting with NZB/BINJ mice (N) and CBA/H mice (H), Roubertoux crossed female N mice with male H mice to generate offspring with N mtDNA (mtDNA is inherited only from the mother). He then back-crossed the offspring with fully H males to completely remove N nuclear DNA (nDNA) from the animals, leaving them with H nDNA and N mtDNA. He carried out the same crosses to generate mice with N nDNA and H mtDNA.

Roubertoux then tested the animals in a variety of learning tasks. In the Morris water maze, where animals swim until they find a hidden platform, fully H animals learned fastest, N mice with H mtDNA were slightly slower, but both strains with N mtDNA learned much more slowly. For example, fully H animals at three months old took 45 seconds to find the platform initially, but after three days’ training slashed this time to only 20 seconds. H mice with N mtDNA, however, also took 45 seconds at first, but only managed to improve this to 32 seconds after three days’ training. Results were highly significant.

Roubertoux found that as the animals aged, the influence of the mtDNA became more pronounced. When he tested three-month-old animals for exploratory behavior on a platform drilled with holes, there was little difference between fully N mice and N mice with H mtDNA. By 12 months, however, N mice were exploring about 45 holes every 10 minutes, whereas their counterparts with H mtDNA were getting to over 50.

The results suggest that H mtDNA improves learning and exploratory behavior. To explain the differences, the authors compared brain anatomy of the animals, finding that N mtDNA seems to lead to greater brain weight-H mice with N mtDNA had almost twice the brain weight of fully H mice. Why this should be is unclear, but the authors did find polymorphic differences between the N and H mtDNA, most notably in complexes I and IV of the respiratory chain.

Being the site for oxidative phosphorylation, mitochondria generate most of the cells’ reactive oxygen species, and they have been implicated in a variety of neurodegenerative diseases (see ARF related news story and ARF news story). The authors conclude that "mtDNA, which, in most cases, if of unknown origin, may have unpredictable effects." It must be remembered, however, that these mice are congenic with respect to their nuclear DNA, so it remains to be seen whether polymorphisms in the relatively small mitochondrial genome can have as large an impact as those in the nuclear DNA.-Tom Fagan.

References:
Gan W-B, Kwon E, Feng G, Sanes JR, Lichtman JW. Synaptic dynamism measured over minutes to months: age-dependent decline in an autonomic ganglion. Nat Neurosci. 2003 Sep;6(9):956-960. Abstract

Roubertoux PL, Sluyter F, Carlier M, Marcet B, Maarouf-Veray F, Cherif C, Marican C, Arrechi P, Godin F, Jamon M, Verrier B, Cohen-Salmon C. Mitochondrial DNA modifies cognition in interaction with the nuclear genome and age in mice. Nat Genet. 2003 Sep;35(1):65-9. Epub 2003 Aug 17. Abstract