It's news that would make Ben Barres extremely happy. Last week in Nature, an article describes a role for astrocytes in memory. Specifically, the researchers found that long term potentiation depends on the release of D-serine from astrocytes.
The role for astrocytes in long-term potentiation (and well, everything) remains somewhat controversial. Neurons can synthesize D-serine themselves, leading to debate over whether astrocytes were really necessary, despite results showing that release of D-serine from glia allows LTP in cultures. In authors come down in favor of a role for astrocytes, showing that "induction of NMDAR-dependent LTP at excitatory hippocampal synapses depends on the availability of NMDARs provided by the Ca2+-dependent release of D-serine from a local astrocyte." The authors go on to suggest that neighboring astrocytes will have distinctive effects on their local synapses, but will also likely have an extended influence, potentially allowing LTP of synapses contacted by the activated astrocyte, extending LTP into a ranged alteration of potentially hundreds of additional neurons/synaptic contacts.
And for those of you not acquainted with Dr. Ben Barres, or why the news that astrocytes are important would make him happy, I refer you to his Stanford academic profile, where you can read about "The Mystery and Magic of Glia".