Mots-Clés

analyse d'images, imagerie de cellules vivantes, quantification de la transcription

Description

In HIV patients, infection can be controlled by highly active anti-retroviral therapy (HAART). In spite of efficient suppression of viral replication, HAART does not lead to an HIV-1 cure, since integrated provirus can remain silent in a latent reservoir consisting mainly of resting CD4+ T memory cells. Upon HAART withdrawal the latent viruses become activated and provoke a viral rebound. To avoid latent virus re-activation, two major therapy approaches are currently under investigation: "block and lock and "shock and kill". The "block and lock" strategy aims at suppressing HIV-1 transcription to lock viruses in a deep latency state that would indefinitely prevent replication. In contrast, the "shock and kill" strategy is based at reactivating viral transcription helping to clear the infected cells by the immune system. So far, all the efforts of this approach did not allow to eliminate the latent reservoir due to incomplete HIV-1 reactivation, which was shown to be stochastic. This stochasticity is caused by random fluctuations of the HIV-1 promoter between ON and OFF states, leading to bursts of viral RNA production. Although these bursts were demonstrated by live cells imaging, their dynamics and mechanisms in T lymphocytes remain largely unknown and it is a subject of the studies of the team. We use live cell imaging to follow transcription of fluorescently labelled viral RNA with single molecule sensitivity. The HIV-1 transcriptional activity in movies is quantified and subjected to mathematical modelling to elucidate promoter limiting steps important for latency.  The goal of the CDD is to quantify numbers of RNA molecules in movies using existing python notebooks. The quantification includes segmentation of the cells in the movies, detection of single molecules and transcription sites and quantification of the RNA molecules numbers at the transcription sites based on their intensities. The quantification uses the principles of single molecule Fluorescent In Situ Hybridization (smFISH) quantification developed in python (see https://fish-quant.github.io/). Before quantification the movies will require preparation in ImageJ. We are looking for a candidate with a strong background in python and ImageJ, interested in image analysis and microscopy. The funding applications will be deposited to renew this CDD, the responses will be obtained by the end of 2024.