Exposure of various human cell lines to ultraviolet (UV) light leads to rapid appearance of polyubiquitinated CDC25A (PolyUb-CDC25A) (Mailand et al. 2000). Instability of CDC25A during interphase and upon exposure to ionizing radiation (IR) is dependent on the SCF-BTrCP (Skp1:Cullin:F-box) E3 ubiquitin ligase complex (Donzelli et al. 2002, Busino et al. 2003, Jin et al. 2003). CDC25A co-immunoprecipitates (co-IPs) with SCF subunits CUL1 and SKP1 (Donzelli et al. 2002). Association of CDC25A with CUL1 increases upon IR treatment and CDC25A associates only with the active, neddylated form of CUL1 (Jin et al. 2003). CDC25A does not co-IP with SKP2 (Busino et al. 2003) and ectopic expression of a dominant-negative SKP2 mutant does not lead to accumulation of CDC25A (Jin et al. 2003). Of the multiple screened F-box subunits of the SCF complex, only BTRC and FBXW11 were found to interact with CDC25A in vivo (Busino et al. 2003) and in vitro (Jin et al. 2003), and the association increased upon IR treatment (Jin et al. 2003).

Degradation of CDC25A S124A mutant in interphase cells is substantially delayed, implying that the phosphorylation of CDC25A at S124 is the prerequisite for SCF-BTrCP mediated CDC25A proteolysis (Donzelli et al. 2002). In the study by Jin et al. 2003, phosphorylation of CDC25A at S124 was found to not be required for CDC25A ubiquitination in vitro. A quadruple serine-to-alanine substitution mutant of CDC25A, lacking CHEK1 phosphorylation sites S124, S178, S279, and S293 was found to still be ubiquitinated by SCF-BTrCP, but shorter ubiquitin conjugates were produced compared to wild-type CDC25A (Jin et al. 2003). CHEK1-mediated phosphorylation of CDC25A at S76 however, was found to be required for SCF-BTrCP-mediated ubiquitination of CDC25A (Jin et a. 2003). CDC25A that co-IPs with BTRC and FBXW11 is hyperphosphorylated (Busino et al. 2003). The DSGXXXXS motif in CDC25A, which includes S82 and S88, corresponds to DSGXXS or DSGXXXS phosphodegron motifs frequently seen in SCF-BTrCP substrates, which interact with the beta-propeller of BTRC/FBXW11 (Busino et al. 2003, Jin et al. 2003). Phosphorylation of both serines in this motif is a known requirement for SCF-BTrCP substrate binding and degradation and, accordingly, CDC25A S28A;S88A double mutant does not interact with BTRC or FBXW11 and exhibits a prolonged half-life (Busino et al. 2003). Ubiquitination of CDC25A S82A mutant by SCF-BTrCP is impaired, while CDC25A S88A mutant is ubiquitinated with efficiency similar to that of the wild-type CDC25A (Jin et al. 2003). Neither S82 nor S88 are CHEK1 phosphorylation sites (Jin et al. 2003). D81 in the CDC25A phosphodegron is required for interaction with a conserved arginine residue in BTRC/FBXW11 (R475 in BTRC, corresponds to R412 in FBXW11) and SCF-BTrCP-mediated ubiquitination of CDC25A (Jin et al. 2003). CDC25A S79A mutant is not ubiquitinated by SCF-BTrCP, and phosphorylation of CDC25A at S76, S79, S82, and S88 is required for efficient binding to and ubiquitination by SCF-BTrCP (Jin et al. 2003). Donzelli et al. 2004 confirmed the requirement for S76 and S82 phosphorylation for CDC25A association with and ubiquitination by the SCF-BTrCP but reported that phosphorylation of CDC25A at S88 was dispensable. While CDC25A S76A mutant could still associate with SCF-BTrCP, it could not be ubiquitinated, leading to a model in which phosphorylated S82 is a docking site for SCF-BTrCP and phosphorylation of S76 is a priming step for S82 phosphorylation (Donzelli et al. 2004). Kinases that phosphorylate CDC25A on S82 and require a priming phosphorylation on S76, leading to ubiquitin-mediated degradation of CDC25A, are NEK11 (Melixetian et al. 2009) and Casein kinases I CK1alpha (CSNK1A1) and CK1epsilon (CSNK1E) (Honaker and Piwnica-Worms 2010, Piao et al. 2011). In the light of all this evidence, products of CDC25A phosphorylation by NEK11 and CK1, phosphorylated on S82, are annotated as substrates of the SCF-BTrCP complex.

Expression of a dominant-negative BTRC mutant leads to accumulation of phosphorylated CDC25A (Jin et al. 2003). Expression of a CUL1 mutant that interferes with degradation of SCF substrates leads to stabilization of CDC25A (Donzelli et al. 2002). Expression of a dominant-negative mutant of CUL1, which binds to SKP1 but not RBX1, results in accumulation of CDC25A and other SCF-BTrCP substrates both in the absence of DNA damage and upon IR treatment (Jin et al. 2003). Addition of a recombinant BTRC to HeLa cell extracts (Busino et al. 2003) or ectopic overexpression of BTRC in HEK293T cells (Jin et al. 2003) stimulates ubiquitination of the wild-type CDC25A, while ubiquitination of CDC25A S82A;S88A double mutant is impaired (Busino et al. 2003). Knockdown of BTRC and FBXW11 by siRNA leads to accumulation of CDC25A protein without affecting CDC25A mRNA level, and the accumulated CDC25A is hyperphosphorylated if cells are pre-treated with IR (Busino et al. 2003, Jin et al. 2003). BTRC- and FBXW11-depleted cells show radioresistant DNA synthesis phenotype that can be rescued by CDC25A depletion (Busino et al. 2003).

It was reported that during the normal cell cycle CDC25A is phosphorylated on S88 by the Cyclin A:CDK2 complex (CCNA:CDK2), constituting a negative-feedback loop that limits the duration of CDC25A activity (Ditano et al. 2021).

Xenopus CDC25A does not have a DSG but a DAG motif at a corresponding site, and both Xenopus DAG and human DSG motifs are embedded in the PEST domain-like sequence (Kanemori et al. 2005). The Xenopus DAG motif is required for Xenopus CHEK1-induced degradation of Xenopus CDC25A (xCDC25A) but PEST domain deletion mutants of xCDC25A can still bind to BTRC, although much less efficiently (Kanemori et al. 2005). An additional DDG motif, conserved between Xenopus (amino acid residues 226 and 231) and human (amino acid residues 215-220) CDC25A, which resembles a doubly phosphorylated DSG motif, with aspartate residues mimicking phosphorylated serines, contributes to binding of xCDC25A to BTRC and to ubiquitin-mediated degradation of xCDC25A (Kanemori et al. 2005).