Role of FBXW5-loss in Centrosome Abnormalities and Cell Physiology

Abstract

Centrosome duplication is a tightly regulated process during cell division and should, like DNA replication, occur only once per cell cycle. The equal distribution of centrosomes to the cell poles, with simultaneous migration of chromosomes, ensures genomic stability during mitosis. In most tumors, the centrosome cycle is disrupted, although it remains unclear whether centrosome abnormalities are the cause or consequence of the tumor formation process. In previous work Puklowski et al. demonstrated that the ubiquitin E3 ligase SCFFBXW5 plays a central role in the initiation of the centrosome cycle. Upon reduction of cellular SAS6 (spindle assembly abnormal protein 6 homolog) by RNAi-induced destabilization of FBXW5, centrosome amplifications and the formation of multinucleated cells were observed [1]. In the present work, such destabilization of FBXW5, so far performed only in human cancer cells, was applied to untransformed murine fibroblasts and liver cells. It was shown that in vitro the loss of FBXW5 - in addition to the increased occurrence of supernumerary centrosomes - also leads to reduced cell division and cell migration. Further studies in synchronized FBXW5-suppressed cells revealed that the growth and migration disturbances are accompanied by the increased appearance of multipolar spindles. Presumably, this phenotype leads to a delay in the transition from mitosis to G1 phase and thus to a later onset of the next cell cycle. Although genomic instability and tumor development have been postulated to also occur with the formation of multipolar spindles [2], no transformation was observed in vitro in FBXW5-deficient cells. Downregulation of FBXW5 was also attempted in vivo. Hydrodynamic tail vein injection of transposon-based shRNAs, or CRISPR/Cas9 RNAs, was aimed to specifically reduce FBXW5 in mouse liver. However, no tumor formation or other pathophysiological consequences were observed - neither in wild-type nor in p53-, p21-, and p19arf-deficient mice. Presumably, the depletion of FBXW5 alone is not a sufficient trigger of tumorigenesis. Yet, the loss of FBXW5 could not be demonstrated beyond doubt in vivo. To uncover potential regulatory mechanisms that might suppress transformation of FBXW5-deficient cells, a genome-wide CRISPR/Cas9 library screen was performed. Here, different regulators, such as replication protein A2 (RPA2) or cell stress-associated genes like ang5 or map4k2 were identified.