Akihiro Fujikawa, Kei Nakayama, Teiichi Furuichi, Manabu Abe, Shuji Shigenobu, Nobuyuki Shiina, Masaharu Noda, Akira Futatsugi, Rie Ohashi, Yo Shinoda, Katsuhiko Mikoshiba, Maya Yamazaki, and Kenji Sakimura
Local regulation of synaptic efficacy is thought to be important for proper networking of neurons and memory formation. Dysregulation of global translation influences long-term memory in mice, but the relevance of the regulation specific for local translation by RNA granules remains elusive. Here, we demonstrate roles of RNG105/caprin1 in long-term memory formation. RNG105 deletion in mice impaired synaptic strength and structural plasticity in hippocampal neurons. Furthermore, RNG105-deficient mice displayed unprecedentedly severe defects in long-term memory formation in spatial and contextual learning tasks. Genome-wide profiling of mRNA distribution in the hippocampus revealed an underlying mechanism: RNG105 deficiency impaired the asymmetric somato-dendritic localization of mRNAs. Particularly, RNG105 deficiency reduced the dendritic localization of mRNAs encoding regulators of AMPAR surface expression, which was consistent with attenuated homeostatic AMPAR scaling in dendrites and reduced synaptic strength. Thus, RNG105 has an essential role, as a key regulator of dendritic mRNA localization, in long-term memory formation., eLife digest Messages pass from one nerve cell to the next across gaps called synapses. The first neuron releases chemical signals from the end of its long, thin nerve fiber. The second receives the message at receptors on branching structures known as dendrites. Each connection has a corresponding bump called a dendritic spine. As animals learn, these can grow larger, strengthening the connection. This is the basis of how memories form. To strengthen a synapse, the cell must transport the materials to the dendritic spine. The cell makes copies of the genetic instructions to strengthen the synapse in the form of messenger RNA (often shortened to mRNA). But, this happens in the body of the cell, a long way from the dendrites themselves. The mRNA travels from the cell body to the dendrites in collections of molecules referred to as ‘RNA granules’. One of the key components of the RNA granule system is a protein called RNG105/caprin1. Now, Nakayama, Ohashi et al. have engineered mice to delete the gene for RNG105/caprin1, revealing its effect on memory. Mice lacking RNG105/caprin1 struggled to make long-term memories. Unlike their normal counterparts, these mutant mice did not become accustomed to new environments or objects. They also found it more challenging to learn the position of a hidden platform in a water-based maze. Lastly, over time, the mutant mice forgot to be fearful of a dark chamber where they had received a small electric shock. Memories form in a part of the brain called the hippocampus and the dendritic spines in this region were smaller in mice lacking RNG105/caprin1. Furthermore, when the nerve cells from this part of the brain were grown in Petri dishes, they did not respond normally to stimulation. The dendritic spines of normal cells increased in size, but those on the cells lacking RNG105/caprin1 got smaller compared to normal cells. A closer look revealed that the distribution of mRNA in brain cells from mice lacking RNG105/caprin1 differed from that of normal mice. Some pieces of genetic information failed to make it from the cell body to the dendrites. This included mRNA involved in making regulators of a component of dendritic spines called the AMPA receptor. The AMPA receptor detects the chemical messenger, glutamate, and is crucial for memory formation. These findings further our understanding of long-term memory and open the way for future research into human disease. Mutations in RNA granule components, including RNG105/caprin1, have links to conditions such as amyotrophic lateral sclerosis (ALS) and autism spectrum disorder (ASD). Further investigation could reveal new targets for drug treatment.