Un nouveau mécanisme d'expulsion des centrosomes est nécessaire au maintien de l’activité métabolique dans les cellules polyploïdes

Centrosomes act as the main microtubule (MTs) organizing centers (MTOC) of animal cells. Each centrosome is composed of two centrioles that recruit and assemble the pericentriolar material (PCM) responsible for MT nucleation. Although centrosome number is tightly regulated in the cell, many exceptions can be found. This is the case of centrosome amplification in multiciliated cells or centrosome elimination in programmed polyploid Drosophila salivary gland (SGs) cells.Polyploidy is defined by the presence of more than two sets of all chromosomes and is achieved through different pathways. Endoreplication is the most common physiological way to generate polyploid cells by alternating G and S phase without entering mitosis. Interestingly, many endoreplicating cells lose their centrosomes in a not well understood mechanism.With the aim of characterizing the mechanism of centrosome elimination, I used Drosophila SGs as a model system. These are highly secretory tissues composed of secretory cells surrounding a tubular lumen, in which they secrete a glue essentiel for pupariation. These secretory cells undergo several rounds of endoreplication generating cells with up to 2048C with the chromosomes attached one to another, polyten chromosome.I first analyzed centrosome behavior during embryonic SG development, since centrosomes disappear at those developmental stages. Using immunofluorescence and live imaging approaches, I identified a model for centrosome elimination, that is initiated when centrosomes localize at the apical membrane at early embryonic stage. Then centrosomes are release into the lumen at later stages of embryogenesis. While analyzing the membrane behavior, I could detect that the centrosomes localized at the apical membrane were engulfed in an actin vesicle enriched with active myosin II (MyoII). This suggested a possible role of MyoII in the elimination process, which was confirmed using a hypomorphic mutation (MyoIImut) Surprisingly, an increase in centrosome numbers in MyoIImut SGs was also detected compared to WT SGs, suggesting that if centrosomes were not eliminated from the cell, they are capable of duplicating throughout development.I then confirmed that MyoII activity is essential for centrosome elimination as only wild type MyoII, not inactive MyoII, could rescue the MyoIImut phenotypes in L3 SGs. I also used another strategy to maintain centrosomes in the cell without affecting the whole cytoskeleton, by overexpressing Sak/Plk4 (Plk4OE) known to induce centrosome amplification. Interestingly, centrosomes were present in Plk4OE SG cells, and they were only eliminated with the constitutive active MyoII, suggesting that centrosome elimination is a time restricted process dependent on MyoII activity. Centrosomes in the lumen were polyubiquitinated for proteasome degradation until they disappeared in prepupae stages. Using different mutations, I could identify the involvement of autophagy and autolysosome exocytosis in centrosome release in the lumen.I next investigated the purpose of centrosome elimination in SG cells. Maintenance of centrosomes in SG cells had no effect on cell size or ploidy and on MT nucleation. Surprisingly however, I could identify defects in mitochondria structure and organization. This was reinforced with seahorse experiments showing reduced respiration and ATP production, uncovering a novel role for centrosome elimination in ensuring proper mitochondria organization to maintain metabolic and respiratory activity in polyploid SG cells.In conclusion my PhD work revealed a novel mechanism involved in centrosome elimination with key functions in safeguarding the mitochondria cytoskeleton ensuring high metabolic energy production.