Alternatively spliced brain-derived neurotrophic factor (BDNF) transcripts are targeted to distinct cellular compartments in neurons but the mechanisms underlying this sorting are unknown. Although only some BDNF isoforms are targeted to dendrites, we have found that the coding region common to all BDNF transcripts contains a constitutively active dendritic targeting signal and that this signal is suppressed in transcripts containing exons 1 or 4, which are restricted to the cell soma and proximal dendrites. This dendritic targeting signal is mediated by translin, an RNA-binding protein implicated in RNA trafficking, and is disrupted by the G196A mutation associated with memory deficits and psychiatric disorders. Molecular modeling and mutational studies indicate that the G196A mutation blocks dendritic targeting of BDNF mRNA by disrupting its interaction with translin. These findings implicate abnormal dendritic trafficking of BDNF mRNA in the pathophysiology of neuropsychiatric disorders linked to the G196A mutation.neuropsychiatric disorders ͉ neurotrophins S everal lines of evidence indicate that targeting of BDNF mRNA to dendrites plays a key role in mediating synaptic plasticity (1-4). However, the molecular mechanisms regulating this process and the differential subcellular localization of alternatively spliced BDNF transcripts, remain to be clarified.Multiple BDNF transcripts are generated by alternative splicing of one 5Ј exon with a shared 3Ј exon containing the entire BDNF coding region and either a short or long 3Ј UTR sequence (5, 6). In recent studies, we have demonstrated that BDNF transcripts differ in their subcellular localization (7). Exon 1 and 4 transcripts are localized in the cell soma, while exon 2 and 6 transcripts show a somato-dendritic localization. Thus, splice variants appear to encode spatial localization signals used to preferentially regulate BDNF expression in different subcellular domains (2, 3). A recent study has suggested that the long 3Ј UTR contains signals necessary for dendritic targeting of BDNF transcripts (4). However, it is unlikely that this mechanism can fully account for the differential dendritic targeting displayed by BDNF transcripts because more than one-third of exon 4 transcripts, which are retained in the soma, contain the long 3Ј UTR. Conversely, more than one-half of exon 6 transcripts, an isoform that displays targeting to dendrites, contain the short 3Ј UTR. To help define the mechanisms underlying differential localization of BDNF transcripts, we have tested the hypothesis that additional signals might be encoded by other BDNF mRNA regions.

Long-lasting forms of synaptic plasticity and memory require de novo protein synthesis. Yet, how learning triggers this process to form memory is unclear. Translin/trax is a candidate to drive this learning-induced memory mechanism by suppressing microRNA-mediated translational silencing at activated synapses. We find that mice lacking translin/trax display defects in synaptic tagging, which requires protein synthesis at activated synapses, and long-term memory. Hippocampal samples harvested from these mice following learning show increases in several disease-related microRNAs targeting the activin A receptor type 1C (ACVR1C), a component of the transforming growth factor-β receptor superfamily. Furthermore, the absence of translin/trax abolishes synaptic upregulation of ACVR1C protein after learning. Finally, synaptic tagging and long-term memory deficits in mice lacking translin/trax are mimicked by ACVR1C inhibition. Thus, we define a new memory mechanism by which learning reverses microRNA-mediated silencing of the novel plasticity protein ACVR1C via translin/trax.

SUMMARY Genetic defects in the microRNA (miRNA) generating enzyme, dicer, are increasingly linked to disease. Loss of miRNA in dicer deficiency is thought to be due to loss of miRNA-generating activity. Here, we demonstrate a previously unknown catabolic mechanism driving miRNA depletion in dicer deficiency. We developed a Dicer-antagonist assay revealing a pre-miRNA degrading enzyme that competes with pre-miRNA processing. We purified this pre-miRNA degrading activity using an unbiased chromatographic procedure and identified the ribonuclease complex Translin/Trax (TN/TX). In wild type dicer backgrounds, pre-miRNA processing was dominant. However, in dicer deficient contexts, TN/TX broadly suppressed miRNA. These findings indicate that miRNA depletion in dicer deficiency is due to the combined loss of miRNA-generating activity and catabolic function of TN/TX. Importantly, inhibition of TN/TX mitigated loss of both miRNA and tumor suppression with dicer haploinsufficiency. These studies reveal a potentially druggable target for restoring miRNA function in cancers and emerging dicer deficiencies.

Translin and Trax are components of an evolutionarily conserved RNA binding complex. Deletion of Translin in yeast, Drosophila and mouse produces a dramatic loss of Trax protein indicating that its stable expression is dependent on its association with Translin. Analysis of Translin KO mice has revealed multiple behavioral abnormalities and alterations in levels of transcripts encoding synaptic proteins. A confluence of localization, biochemical and RNA trafficking studies supports the view that this complex mediates dendritic trafficking of RNAs, a process thought to play a critical role in synaptic plasticity. However, further studies are needed to define its RNA cargoes, its precise role in this process, and how its binding activity and localization are regulated. Nevertheless, there is sufficient evidence to suggest that the Translin/Trax complex be included among the cadre of RNA binding complexes, such as Staufen and CPEB, that regulate dendritic trafficking of RNA in neurons.

Background: Antinuclear antibody pattern recognition is vital for autoimmune disease diagnosis but labor-intensive for manual interpretation. To develop an automated pattern recognition system, we established machine learning models based on the International Consensus on Antinuclear Antibody Patterns (ICAP) at a competent level, mixed patterns recognition, and evaluated their consistency with human reading. Methods: 51,694 human epithelial cells (HEp-2) cell images with patterns assigned by experienced medical technologists collected in a medical center were used to train six machine learning algorithms and were compared by their performance. Next, we choose the best performing model to test the consistency with five experienced readers and two beginners. Results: The mean F1 score in each classification of the best performing model was 0.86 evaluated by Testing Data 1. For the inter-observer agreement test on Testing Data 2, the average agreement was 0.849 (?) among five experienced readers, 0.844 between the best performing model and experienced readers, 0.528 between experienced readers and beginners. The results indicate that the proposed model outperformed beginners and achieved an excellent agreement with experienced readers. Conclusions: This study demonstrated that the developed model could reach an excellent agreement with experienced human readers using machine learning methods.