Postdoctoral researcher Hong Wang flits in and out of a dark, curtained-off area in Ryohei Yasuda’s lab at Duke University. Wang is fiddling with a slice of mouse brain that floats in an electrolyte broth similar to the fluid that naturally bathes the brain. Once she has the setup just right, she will use one of Yasuda’s specially built microscopes to look at enlargement of a single neuron’s spine, a tiny bud that receives signals from neighboring cells.
Yasuda, an HHMI early career scientist, is a biophysicist by training. For his Ph.D. dissertation, he developed ways to image single molecules tethered to glass slides. To better understand the molecular intricacies of learning, memory, and behavior, he has shifted his interests to imaging molecules in single synapses in brain slices. Hundreds of molecules are needed to transform a discrete event into a long-lasting memory, and in his latest study, his team reports exactly when and where in the neuron two important molecules are active.
“The synapse is small but not simple,” Yasuda says. “It’s got a lot of complicated machinery, and it’s the perfect place to use my biophysical background.”
When one neuron fires to another, the synapse between the cells changes in strength—a process called synaptic plasticity. Changes lasting minutes or longer are thought to encode long-term memories. In particular, the receiving neuron goes through many changes as receptors and other proteins shuttle to its spines. Scientists have had trouble figuring out how each of the many molecular players behaves in relation to the others, in both space and time. Yasuda’s been particularly interested in finding the molecules responsible for turning short-term events into longer-term cell changes.
Studying chemical signaling in living spines has been difficult because each spine contains only a few copies of each protein, and some signals happen fast. In 2003, as a postdoctoral researcher at Cold Spring Harbor Laboratory working with Karel Svoboda (now a group leader at Janelia Farm), Yasuda developed two-photon fluorescence lifetime imaging microscopy (2pFLIM) , a method that allows scientists to quantify protein–protein interactions in living cells at single-synapse resolution. Yasuda and Svoboda were among the first to apply the technique to neuroscience questions. Since arriving at Duke in 2005, Yasuda has built four microscopes and has designed sensors for 2pFLIM that change fluorescence lifetime when a particular molecule is active.
Until recently, scientists speculated that Ca2+/calmodulin-dependent kinase II (CaMKII)—an enzyme that’s abundant in neurons and is well known to be important in synaptic plasticity—was responsible for the change. Yasuda tested this long-standing hypothesis. His team measured CaMKII’s activity for 30 minutes in response to stimulation and, to their surprise, found that it turned on for only about 1 minute. That study was published in Nature in 2009. CaMKII is important for integrating the calcium signals that rush into a neuron when it’s stimulated, but the enzyme’s short activity time indicates that other molecules must come into play later in the synaptic strengthening process. Yasuda wanted to know what those molecules were.
He decided to look at one aspect of synaptic strengthening: the filamentous scaffolds that enlarge the spine, increasing its contact area with the sending neuron. Focusing on RhoA and Cdc42, two molecules involved in lengthening scaffold filaments, Yasuda’s group showed that both turn on later and for longer periods compared with CaMKII. “There are probably more complicated processes downstream, but [RhoA and Cdc42] look like mediators of short signal to long signal,” Yasuda says. The group published its results April 7, 2011, in Nature.
Although RhoA and Cdc42 turn on and off at similar times, their spatial patterns are different, the group found. RhoA activation diffused beyond the stimulated spine, extending a few millimeters along the receiving neuron, whereas Cdc42 activation was restricted to the stimulated spine. The importance of these different spatial patterns is unknown, but Yasuda suspects that a spreading signal instructs the cell to release more resources to the spine. Conversely, signaling that’s restricted to the spine may help compartmentalize memory storage there and thus maximize overall memory capacity, he adds.
Yasuda’s team is pursuing several follow-up studies. Because many diseases are linked to a failure in synaptic plasticity, he plans to measure molecular activity within individual spines in mouse models of mental retardation and Alzheimer’s disease.
On this particular day in the lab, Wang gets the images she needs. She is looking at brain slices from a mouse lacking the Cdc42 protein, comparing the structure and plasticity of the spines in these mice with those in healthy controls. Yasuda expects the Cdc42 mutant to have lower levels of synaptic plasticity. With another research group at Duke, Yasuda’s team is also taking the mutant mice through a battery of learning and memory tests—Yasuda’s first foray into animal behavior. “It is definitely exciting to see if Cdc42 is important for actual learning and memory,” he says.