to BioTechniques free email alert service to receive content updates.
New Site of Long-term Memory in Fruit Flies

Jeffrey M. Perkel

Using a photoconvertible fluorescent protein and cold-sensitive protein synthesis inhibitor, researchers have discovered a new site for memory formation in fruit flies.

Bookmark and Share

More than a century ago, Pavlov trained his dogs to salivate when a bell rang. That conditioned reflex isn’t limited to dogs, or even vertebrates. Fruit flies do it, too, and researchers have now identified a new neuroanatomical structure, called the “dorsal-anterior-lateral” (DAL) neurons, that is critical to the process.

Ann-Shyn Chiang of the National Tsing Hua University, Taiwan, led the research team that first conditioned fruit flies to respond to certain odors as if they were electrically shocked. Because protein expression is a key step in memory consolidation, the group then identified the neurons in which protein synthesis was required to retain or consolidate those memories.

Anterior montage view of a Drosophila brain within its head capsule, showing part of the olfactory associative long-term memory retrieval circuits. Source: Science/AAAS

As Chiang and colleagues report in the Feb. 10 issue of Science,the site of that protein expression was unexpected (1). A brain region called the “mushroom body” (MB) has long been associated with memory formation. Yet disruption of protein synthesis in the MB after training did not impact memory formation in this study. Blocking protein synthesis in the two adjacent DAL neurons did.

“The DAL neurons can now be added to the neural system underlying long-term olfactory memory in Drosophila,” said Ronald Davis, neuroscience chair at the Scripps Research Institute Florida. “This is the major conceptual contribution.”

“It certainly goes against what a lot of people thought in the field,” added Glenn Turner, an assistant professor at Cold Spring Harbor Laboratory (CSHL). “Most people thought it [memory consolidation] happened in the mushroom body. This paper suggests it doesn’t.”

The conventional wisdom in neurobiology, explains Josh Dubnau of CSHL, who authored a perspective accompanying the article (2), is that memory consolidation occurs as neurons strengthen synaptic connections via a change in gene expression inside the cells receiving the stimuli in the first place.

“But in this paper, the synapse is signaling not to the nucleus of its own cell, but to cells in a different region, and that means there must be communication between those two regions and that gene expression changes in one region affect synaptic connections in another,” he said.

The authors thus posit the existence of a neuronal circuit joining MB and DAL neurons — a systems-level connectivity that has not previously been demonstrated in invertebrate brains.

To draw that conclusion, Chiang and his team developed two genetic tools, which they placed under the control of different regulatory elements to express in specific neural subpopulations in the fly. The first is a photoconvertible fluorescent protein called KAEDE. Normally, KAEDE fluoresces green; when irradiated with UV light, it switches red. By periodically illuminating the flies with UV light, the team monitored new protein expression in specific cellular subsets, as the newly synthesized protein would fluoresce a different color.

The second tool was a cold-sensitive protein synthesis inhibitor called RICINCS, which they likewise could express in specific neural subsets to block protein synthesis where — and when — they wanted. Changing the temperature from 30°C to 18°C rapidly deactivated the inhibitor, and vice versa.

Although Davis believes the study is a major contribution, he doubts that protein expression is required only in DAL neurons and not at all in the MB. One alternative explanation, he suggested, is that the two cell types (DAL and MB neurons) could be “differentially sensitive” to RICINCS protein, leading to an apparent lack of synthesis in MB cells.

“It is easy to account for the results by redundancy in function,” he explained. There are approximately 4000 mushroom body neurons but only two DAL neurons implicated in long-term memory. “You can probably lose the function of many mushroom body neurons before you see an effect, but you will see an effect immediately when only two comprise one of the nodes of the neural system.”

Be that as it may, however, there’s no denying that DAL neurons have some role to play in long-term memory formation. The challenge now is to identify precisely what that role is, and the molecular actors that make it happen.


  1. Chen, C.-C., et al. 2012. Visualizing long-term memory formation in two neurons of the Drosophila brain. Science 335:678–85.
  2. Dubnau, J.. 2012. Ode to the mushroom bodies. Science, 335:664–5.