SUMMARYRecent studies indicated that sortilin-related receptor 1 (SORL1) is a risk gene for late-onset Alzheimer's disease (AD), although its role in the aetiology and/or progression of this disorder is not fully understood. Here, we report the finding of a non-coding (nc) RNA (hereafter referred to as 51A) that maps in antisense configuration to intron 1 of the SORL1 gene. 51A expression drives a splicing shift of SORL1 from the synthesis of the canonical long protein variant A to an alternatively spliced protein form. This process, resulting in a decreased synthesis of SORL1 variant A, is associated with impaired processing of amyloid precursor protein (APP), leading to increased Aβ formation. Interestingly, we found that 51A is expressed in human brains, being frequently upregulated in cerebral cortices from individuals with Alzheimer's disease. Altogether, these findings document a novel ncRNA-dependent regulatory pathway that might have relevant implications in neurodegeneration.

The most frequently used technique to study the expression profile of genes involved in common neurological disorders is quantitative real-time RT-PCR, which allows the indirect detection of very low amounts of selected mRNAs in tissue samples. Expression analysis by RT-qPCR requires an appropriate normalization to the expression level of genes characterized by a stable, constitutive transcription. However, the identification of a gene transcribed at a very stable level is difficult if not impossible, since significant fluctuations of the level of mRNA synthesis often accompanies changes of cell behavior. The aim of this study is to identify the most stable genes in postmortem human brain samples of patients affected by Alzheimer’s disease (AD) suitable as reference genes. The experiments analyzed 12 commonly used reference genes in brain samples from eight individuals with AD and seven controls. After a careful analysis of the results calculated by geNorm and NormFinder algorithms, we found that CYC1 and EIF4A2 are the best reference genes. We remark on the importance of the determination of the best reference genes for each sample to be analyzed and suggest a practical combination of reference genes to be used in the analysis of human postmortem samples.

BackgroundMetformin is a widely used oral hypoglycemizing agent recently proposed as potential anti-cancer drug. In this study we report the antiproliferative effect of metformin treatment in a high risk neuroblastoma cell model, focusing on possible effects associated to different levels of differentiation and/or tumor initiating potential.MethodsAntiproliferative and cytotoxic effects of metformin were tested in human SKNBE2 and SH-SY5Y neuroblastoma cell lines and in SKNBE2 cells in which differentiation is induced by retinoic acid treatment or stable overexpression of NDM29 non-coding RNA, both conditions characterized by a neuron-like differentiated phenotype.ResultsWe found that metformin significantly inhibits the proliferation of NB cells, an effect that correlates with the inhibition of Akt, while AMPK activity resulted unchanged. Notably, metformin effects were modulated in a different ways by differentiating stimuli, being abolished after retinoic acid treatment but potentiated by overexpression of NDM29.ConclusionThese data suggest the efficacy of metformin as neuroblastoma anticancer agent, and support the requirement of further studies on the possible role of the differentiation status on the antiproliferative effects of this drug.

Recent advances in life sciences suggest that human and rodent cell responses to stimuli might differ significantly. In this context, the results achieved in neurotoxicology and biomedical research practices using neural networks obtained from mouse or rat primary culture of neurons would benefit of the parallel evaluation of the same parameters using fully differentiated neurons with a human genetic background, thus emphasizing the current need of neuronal cells with human origin. In this work, we developed a human functionally active neural network derived by human neuroblastoma cancer cells genetically engineered to overexpress NDM29, a non-coding RNA whose increased synthesis causes the differentiation toward a neuronal phenotype. These cells are here analyzed accurately showing functional and morphological traits of neurons such as the expression of neuron-specific proteins and the possibility to generate the expected neuronal current traces and action potentials. Their morphometrical analysis is carried out by quantitative phase microscopy showing soma and axon sizes compatible with those of functional neurons. The ability of these cells to connect autonomously forming physical junctions recapitulates that of hippocampal neurons, as resulting by connect-ability test. Lastly, these cells self-organize in neural networks able to produce spontaneous firing, in which spikes can be clustered in bursts. Altogether, these results show that the neural network obtained by NDM29-dependent differentiation of neuroblastoma cells is a suitable tool for biomedical research practices.