The Cellular Phenotype of Roberts Syndrome Fibroblasts as Revealed by Ectopic Expression of ESCO2

et al. (2009) The Cellular Phenotype of Roberts Syndrome Fibroblasts as Revealed by Ectopic Expression of ESCO2. PLoS ONE 4(9): e6936. doi:10.1371/journal.pone.0006936 The Cellular Phenotype of Roberts Syndrome Fibroblasts as Revealed by Ectopic Expression of ESCO2 Petra van der Lelij 0 Barbara C. Godthelp 0 Wouter van Zon 0 Djoke van Gosliga 0 Anneke B. Oostra 0 Ju rgen Steltenpool 0 Jan de Groot 0 Rik J. Scheper 0 Rob M. Wolthuis 0 Quinten Waisfisz 0 Firouz Darroudi 0 Hans Joenje 0 Johan P. de Winter 0 Peter E. Warburton, Mount Sinai School of Medicine, United States of America 0 1 Department of Clinical Genetics, VU University Medical Center , Amsterdam , The Netherlands , 2 Department of Toxicogenetics, Leiden University Medical Center , Leiden , The Netherlands , 3 Division of Molecular Biology, Netherlands Cancer Institute , Amsterdam , The Netherlands , 4 Department of Pathology, VU University Medical Center , Amsterdam , The Netherlands Cohesion between sister chromatids is essential for faithful chromosome segregation. In budding yeast, the acetyltransferase Eco1/Ctf7 establishes cohesion during DNA replication in S phase and in response to DNA double strand breaks in G2/M phase. In humans two Eco1 orthologs exist: ESCO1 and ESCO2. Both proteins are required for proper sister chromatid cohesion, but their exact function is unclear at present. Since ESCO2 has been identified as the gene defective in the rare autosomal recessive cohesinopathy Roberts syndrome (RBS), cells from RBS patients can be used to elucidate the role of ESCO2. We investigated for the first time RBS cells in comparison to isogenic controls that stably express V5- or GFP-tagged ESCO2. We show that the sister chromatid cohesion defect in the transfected cell lines is rescued and suggest that ESCO2 is regulated by proteasomal degradation in a cell cycle-dependent manner. In comparison to the corrected cells RBS cells were hypersensitive to the DNA-damaging agents mitomycin C, camptothecin and etoposide, while no particular sensitivity to UV, ionizing radiation, hydroxyurea or aphidicolin was found. The cohesion defect of RBS cells and their hypersensitivity to DNA-damaging agents were not corrected by a patient-derived ESCO2 acetyltransferase mutant (W539G), indicating that the acetyltransferase activity of ESCO2 is essential for its function. In contrast to a previous study on cells from patients with Cornelia de Lange syndrome, another cohesinopathy, RBS cells failed to exhibit excessive chromosome aberrations after irradiation in G2 phase of the cell cycle. Our results point at an S phase-specific role for ESCO2 in the maintenance of genome stability. - . These authors contributed equally to this work. Introduction Roberts syndrome (RBS) is a rare autosomal recessive disease characterized by growth retardation and congenital abnormalities. RBS patients typically have limb malformations involving symmetric reduction in the number of digits, and the length or presence of bones in the arms and legs, but the severity of these abnormalities is quite variable, even within families [1]. Survival is generally poor, as most cases of RBS end in spontaneous abortion, still-birth, or neonatal death. Cells from RBS patients show specific cytogenetic characteristics, mainly consisting of metaphase chromosomes displaying repulsion at heterochromatin regions or centromere splitting leading to a railroad-track appearance of chromosomes. RBS is caused by mutations in ESCO2 [2]. ESCO2 is one of the two human orthologs of the Saccharomyces cerevisiae protein Eco1/Ctf7, a putative acetyltransferase required for the establishment of sister chromatid cohesion during S phase [3,4]. In addition, Eco1 is important to maintain sister chromatid cohesion after the introduction of double strand breaks in G2/M phase of the cell cycle [57], suggesting that the establishment of cohesion is also essential for postreplicative repair of double strand breaks. EBV-immortalized lymphoblastoid cell lines from Roberts syndrome patients have previously been claimed to be hypersensitive to the growth-inhibiting effect of mitomycin C (MMC) and gamma irradiation [8,9]. However, these studies were not entirely conclusive since isogenic control cell lines were not available or not included. Gordillo et al showed that a lymphoblastoid cell line from a patient homozygous for the missense mutation W539G in the acetyltransferase domain of ESCO2 was as sensitive to MMC as lymphoblasts from RBS patients lacking ESCO2 mRNA and protein due to nonsense or frameshift mutations [8], indicating that the ESCO2 acetyltransferase domain is important for its function. This missense mutation reduced the acetyltransferase activity of ESCO2 in vitro, which suggests that it is actually pathogenic. Nevertheless, to explore the function of the human ESCO2 protein, there is a need for functionally corrected cell lines from Roberts syndrome patients. In this study we investigate ESCO2-deficient immortalized skin fibroblasts from a Roberts syndrome patient in comparison to isogenic ESCO2 complemented cell lines to document the cellular phenotype of ESCO2-deficient cells. Cells lacking functional ESCO2 appeared to be characterized by a chromatid cohesion defect and by hypersensitivity to the DNA-damaging agents mitomycin C, camptothecin, and etoposide. Immortal fibroblasts from a Roberts syndrome patient are functionally complemented by epitope-tagged ESCO2 To be able to study the role of ESCO2 in sister chromatid cohesion and DNA damage response, SV40 immortalized, ESCO2-deficient fibroblasts from a Roberts syndrome patient (VU1199-F SV40) were stably transfected with cDNA constructs encoding either V5- or GFP-tagged ESCO2 protein. As a negative control, a patient-derived mutation in the acetyltransferase domain of GFP-ESCO2 was generated (GFP-ESCO2 (W539G)). Upon neomycin selection, several clones were obtained, which expressed the V5-ESCO2 or GFP-ESCO2 proteins at levels much higher than endogenous ESCO2 in wild type fibroblasts (Figure 1A). As expected from their molecular weights GFPESCO2 ran more slowly in the SDS-PAGE gel than V5-ESCO2; Western blotting with a GFP-specific antibody was used to demonstrate that the GFP signal was exclusively derived from the GFP-ESCO2 fusion protein, since no uncoupled GFP molecules were detected (Figure 1A). Both wild type and mutant ESCO2 proteins localized in the nucleus, where ESCO2 is supposed to perform its function, but the expression levels seemed to vary between cells (Figure 1B). Since overexpression and/or the V5 or GFP tags may interfere with the activity of ectopically expressed proteins, we tested functional complementation by cytogenetic analysis. Metaphase spreads revealed a strong reduction in the number of railroad chromosomes in cell lines stably transfected with wild type ESCO2 (Figure 1C), demonstrating that the V5- and GFP-tagged ESCO2 proteins are indeed functional, in spite of their re (...truncated)