Long-term memory (LTM) formation requires new protein synthesis and new gene expression. Based on our work in Aplysia, we hypothesized that the rRNA genes, stimulation-dependent targets of the enzyme Poly(ADP-ribose) polymerase-1 (PARP-1), are primary effectors of the activity-dependent changes in synaptic function that maintain synaptic plasticity and memory. Using electrophysiology, immunohistochemistry, pharmacology and molecular biology techniques, we show here, for the first time, that the maintenance of forskolin-induced late-phase long-term potentiation (L-LTP) in mouse hippocampal slices requires nucleolar integrity and the expression of new rRNAs. The activity-dependent upregulation of rRNA, as well as L-LTP expression, are poly(ADP-ribosyl)ation (PAR) dependent and accompanied by an increase in nuclear PARP-1 and Poly(ADP) ribose molecules (pADPr) after forskolin stimulation. The upregulation of PARP-1 and pADPr is regulated by Protein kinase A (PKA) and extracellular signal-regulated kinase (ERK)—two kinases strongly associated with long-term plasticity and learning and memory. Selective inhibition of RNA Polymerase I (Pol I), responsible for the synthesis of precursor rRNA, results in the segmentation of nucleoli, the exclusion of PARP-1 from functional nucleolar compartments and disrupted L-LTP maintenance. Taken as a whole, these results suggest that new rRNAs (28S, 18S, and 5.8S ribosomal components)—hence, new ribosomes and nucleoli integrity—are required for the maintenance of long-term synaptic plasticity. This provides a mechanistic link between stimulation-dependent gene expression and the new protein synthesis known to be required for memory consolidation.

The transition from short-term to long-term forms of synaptic plasticity requires protein synthesis and new gene expression. Most efforts to understand experience-induced changes in neuronal gene expression have focused on the transcription products of RNA polymerase II—primarily mRNAs and the proteins they encode. We recently showed that nucleolar integrity and activity-dependent ribosomal RNA (rRNA) synthesis are essential for the maintenance of hippocampal long-term potentiation (LTP). Consequently, the synaptic plasticity and memory hypothesis predicts that nucleolar integrity and activity dependent rRNA synthesis would be required for Long-term memory (LTM). We tested this prediction using the hippocampus-dependent, Active Place Avoidance (APA) spatial memory task and found that training induces de novo rRNA synthesis in mouse dorsal hippocampus. This learning-induced increase in nucleolar activity and rRNA synthesis persists at least 24 h after training. In addition, intra-hippocampal injection of the Pol I specific inhibitor, CX-5461 prior to training, revealed that de novo rRNA synthesis is required for 24 h memory, but not for learning. Using qPCR to assess activity-dependent changes in gene expression, we found that of seven known rRNA expression variants (v-rRNAs), only one, v-rRNA IV, is significantly upregulated right after training. These data indicate that learning induced v-rRNAs are crucial for LTM, and constitute the first evidence that differential rRNA gene expression plays a role in memory.

Theparaneoplastic Ma antigen (PNMA)genes are associated with cancer-induced paraneoplastic syndromes that present with neurological symptoms and autoantibody production. How PNMA proteins trigger a severe autoimmune disease is unclear.PNMAgenes are predominately expressed in the central nervous system with little known functions but are ectopically expressed in some tumors. Here, we show that PNMA2 is derived from a Ty3 retrotransposon that encodes a protein which forms virus-like capsids released from cells as non-enveloped particles. Recombinant PNMA2 capsids injected into mice induce a robust autoimmune reaction with significant generation of autoantibodies that preferentially bind externalspikePNMA2 capsid epitopes, while capsid-assembly-defective PNMA2 protein is not immunogenic. PNMA2 autoantibodies present in cerebrospinal fluid of patients with anti-Ma2 paraneoplastic neurologic disease show similar preferential binding to PNMA2spikecapsid epitopes. These observations suggest that PNMA2 capsids released from tumors trigger an autoimmune response that underlies Ma2 paraneoplastic neurological syndrome.