How and why forgetting pays out:
see our recent publicaton
Can fruit flies learn values?
In classical conditioning experiments an aversive value gets associated with a specific odour by co-administrating an electric shock. The intensity of the electric schok is relfelcted int eh probabilty that the fruit flies avoide the conditioned odour. 
Current models based on a "Hebbian-type" of plasticity predict that flies would more reliably avoid the odor the more frequently the electric shock is applied. Yet, preliminary experiments indicate that fruit flies stop to accumulate the vaoidance probability while the electric shock is repeatedly applied. The learne aversive vlaue is adequantly encoding the intensity of the shock. We explroe the biolog of this value learning both mby modeleing and experimetns. 
Imaging synapses in the fly brain with super resolution microscopy
Formation of memories during the learning process implies modifications in synapses, due to the plasticity of the nervous system. Still unclear is whether analogous changes underlie the process of memory loss. We want to study structural changes, at the synaptic level, related to forgetting. As a model system we chose the olfactory memory system of the Drosophila brain. The fly brain has well defined regions involved in learning and associative memory, such as the neuropil of the mushroom bodies as well as the neurons connected to them; moreover it provides great advantages in manipulating the expression of genes in specific cells. These advantages, as well as the resolution required to observe changes within synapses, make single molecule localization microscopy the ideal technique for our study.
We are working on the implementation of two color direct Stochastic Optical Reconstruction Microscopy (dSTORM) and aim to image the pre- and postsynaptic proteins of the fsynapses within the Calyx.  We further plan to image not only the surface of the tissue but deeper into the tissue with 3D STORM.
Transcriptomic analysis of forgetting in the fly brain
The main center in the fly brain required for learning and formation of memory is the mushroom body. To identify the changes in gene expression during the aquisiton of memories and during forgetting of longterm memories we performed transcriptomic experiments. Targeted DamID (TaDa) is an efficient technique to perform a cell-type-specific (or genome-wide) binding profiling of a protein of interest without requiring cell isolation. In the SynaptiX project, we are interested in the study of transcriptional changes during the process of forgetting. Therefore, we focused on the DamID technique by studying the binding of the RNA polymerase II, which represents a marker for transcriptional profiling. The specific methylation of Adenines at GATC sites by the Dam-Pol2 construct will be representative of expressed genes in the Drosophila melanogaster genome. However, the analysis of TaDa-seq data presents two major challenges: the correct normalization of the samples and the reduction of the background noise. To overcome such issues, a control Dam-only (not containing the protein of interest, i.e. RNA Polymerase II) is always performed in addition to the Dam-Pol2 experiment. Finally, the log2 ratio of reads mapping respectively to Dam-Pol2 and Dam-only samples is represented. This representation allows the identification of expressed genes.
The central analysis of the TaDa-seq reads corresponds to the mapping to the genome and quantifying the mapped reads between the different samples. For the mapping process, we use the final release of the Drosophila melanogaster genome from Ensembl (release 82, September 2015) as reference. The TaDa-seq datasets are analyzed with the bioinformatics tool bowtie2 for genome indexing and reads mapping. The quantification and location of mapped reads is possible through several tools, however the classical RNA-sequencing tools do not take into account the specificity of the TaDa-seq analysis. Therefore, we use a recently released tool: Damidseq_pipeline that counts the reads mapping to a specific region of the genome while taking into account normalization of the reads as well as reduction of the background noise. Additional tools such as bedtools are also helpful in such analyzes to be able to study the mapping at a single base resolution. An overview of the TaDa-seq data analysis pipeline is represented in Figure 2
In TaDa-seq data analysis the main comparison is performed between the Dam-Pol2 and Dam-only samples. We focus on the repartition of sequencing reads in the two samples and take into account several time points to identify gene expression changes. The representation of the log2 ratio of mapping reads between the Dam-Pol2 and Dam-only samples allow us to distinguish genes with a significant binding of Pol2 in the flanking region. This binding of Pol2 is a strong evidence of the expression of this gene in the given sample. Based on the different time points in the learning protocol (paired and unpaired learning), we are able to compare the gene expression changes affected during the forgetting process. Although this technology will not inform us directly about the gene expression levels (compared to RNA-sequencing), the comparison based on several statistical tests (such as Student’s t-test and ANOVA) will give us confidence in the identification of any change in the expression of a gene of interest. Therefore, we will be able to identify increase or decrease of expression for the list of genes that might be involved in the process of forgetting. In addition, this experiment will allow us the identification of novel genes taking part of this process.
Based on the initial TaDa-seq samples, we developed a web-application based on the R programming language that allows us to analyze the reads distribution and visualize the GATC sites methylation peaks for a gene of interest. The CrebB gene is of particular interest in the SynaptiX project, as it is known to be involved in the learning process in Drosophila melanogaster while expressed in the mushroom body cell type. The expression of CrebB in the mushroom body cell type is visible in Figure 3.