Esco2 and cohesin regulate CRL4 ubiquitin ligase ddb1 expression and thalidomide teratogenicity

Abstract

Cornelia de Lange syndrome (CdLS) and Roberts syndrome (RBS) are severe developmental maladies that arise from mutation of cohesin (including SMC3, CdLS) and ESCO2 (RBS). Though ESCO2 activates cohesin, CdLS and RBS etiologies are currently considered non-synonymous and for which pharmacological treatments are unavailable. Here, we identify a unifying mechanism that integrates these genetic maladies to pharmacologically-induced teratogenicity via thalidomide. Our results reveal that Esco2 and cohesin co-regulate the transcription of a component of CRL4 ubiquitin ligase through which thalidomide exerts teratogenic effects. These findings are the first to link RBS and CdLS to thalidomide teratogenicity and offer new insights into treatments.

Keywords: CRL4 ubiquitin ligase; Roberts syndrome (RBS); birth defects; cohesinopathies; cornelia de lange syndrome (Cdls); thalidomide.

Figure 1.

Esco2 KD and Smc3 KD phenotypes include reduced body and eye size, and an increase in abnormal otolith development. (a) Representative images of control embryos (SC MO injected), Esco2 KD (esco2-ATG MO injected) and Smc3 KD (smc3-ATG MO injected) embryos. For all experiments 24–40 replicates were analyzed and at least 3 independent trials were performed. (b) Quantification of body size from MO injected embryos were compared to un-injected WT embryos to obtain percent similarity. Graph reveals significant reductions of body length in Esco2 KD and Smc3 KD compared to control embryos (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (c) Quantification of eye size from MO injected embryos were compared to un-injected WT embryos to obtain percent similarity. Graph reveals significant reductions of eye size in Esco2 KD and Smc3 KD compared to control embryos (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (d) Graph shows percent of normal, fused, small, or absent otolith phenotypes with MO treatments. Data reveals 95% of control embryo otoliths exhibit normal phenotype, while Esco2 KD and Smc3 KD embryos exhibited 27% and 0% normal otolith phenotypes, respectively. An increase in abnormal otolith phenotypes was observed with KD treatments with predominantly fused phenotypes in Esco2 KDs and absent phenotypes in Smc3 KDs. Scale bar: 100 μm.

Figure 2.

Phenotypes of thalidomide treated embryos overlap with Esco2 KD and Smc3 KD embryos. (a) Representative images of control embryos (WT treated with DMSO) and thalidomide (thal.) treatments (WT treated with 200 μM, 400 μM, and 800 μM concentrations of thalidomide). For all experiments 16–28 replicates were analyzed and at least 3 independent trials were performed. (b) Quantification of body size after drug treatment were compared to un-treated WT embryos to obtain percent similarity. Bar graph reveals a significant reduction of body length with thal. treatments compared to DMSO treated controls (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (c) Quantification of eye size after drug treatment were compared to un-treated WT embryos to obtain percent similarity. Bar graph reveals a significant reduction of eye size with all thal. treatments compared to DMSO treated controls (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (d) Graph shows percent of normal, fused, small, or absent otolith phenotypes with drug treatments. Data reveals 100% of control embryo otoliths exhibit normal phenotype, while 200 μM, 400 μM, and 800 μM thal. treatments had 31%, 36% and 63% normal otolith phenotypes, respectively. An increase in absent otolith phenotypes was observed with increasing concentrations of thalidomide. Scale bar: 100 μm.

Figure 3.

Exogenous ddb1 overexpression rescues Smc3 KD phenotypes. (a) Representative images of control embryos (WT injected with ddb1 mRNA), Smc3 KD (smc3-ATG MO injected) and Smc3 KD + ddb1 mRNA (smc3-ATG MO co-injected with ddb1 mRNA) embryos. For all experiments 26–40 replicates were analyzed and at least 3 independent trials were performed. (b) Quantification of body size from injected embryos were compared to un-injected WT embryos to obtain percent similarity. Bar graph reveals a significant rescue of body length in Smc3 KD + ddb1 mRNA compared to Smc3 KD alone (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (c) Quantification of eye size from injected embryos were compared to un-injected WT embryos to obtain percent similarity. Bar graph reveals a significant rescue of eye size in Smc3 KD + ddb1 mRNA compared to Smc3 KD alone (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (d) Graph shows percent of normal, fused, small, or absent otolith phenotypes with MO treatments. Data reveals 0% of Smc3 KD embryos exhibited normal otoliths, while 70% of Smc3 KD + ddb1 mRNA embryo otoliths were rescued to normal levels. A decrease in absent otolith phenotypes was observed with ddb1 mRNA co-injections compared to KD alone. Scale bar: 100 μm.

Figure 4.

Exogenous ddb1 overexpression exacerbates Esco2 KD phenotypes. (a) Representative images of control embryos (WT injected with ddb1 mRNA), Esco2 KD (esco2-ATG MO injected) and Esco2 KD + ddb1 mRNA (esco2-ATG MO co-injected with ddb1 mRNA) embryos. For all experiments 29–38 replicates were analyzed and at least 3 independent trials were performed. (b) Quantification of body size from injected embryos were compared to un-injected WT embryos to obtain percent similarity. Bar graph reveals a significant reduction of body length in Esco2 KD + ddb1 mRNA compared to Esco2 KD alone (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (c) Quantification of eye size from injected embryos were compared to un-injected WT embryos to obtain percent similarity. Bar graph reveals a significant reduction of eye size in Esco2 KD + ddb1 mRNA compared to Esco2 KD alone (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (d) Graph shows percent of normal, fused, small, or absent otolith phenotypes with MO treatments. Data reveals 27% of Esco2 KD embryos exhibited normal otoliths, while only 7% of Esco2 KD + ddb1 mRNA embryo otoliths were normal with an absent phenotype largely observed. Scale bar: 100 μm.

Figure 5.

Ddb1 KD phenotypes overlap cohesinopathies and thalidomide teratogenicity phenotypes. (a) Representative images of control embryos (WT injected with ddb1 mRNA), Ddb1 KD (ddb1-SB MO injected), Ddb1 KD + low ddb1 mRNA (ddb1-SB MO co-injected with 25ng/μl ddb1 mRNA) and Ddb1 KD + high ddb1 mRNA (ddb1-SB MO co-injected with 100ng/μl ddb1 mRNA) embryos. For all experiments 24–60 replicates were analyzed and at least 3 independent trials were performed. (b) Quantification of body size from injected embryos were compared to un-injected WT embryos to obtain percent similarity. Bar graph reveals a significant reduction of body length in Ddb1 KD that is rescued by ddb1 mRNA (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (c) Quantification of eye size from injected embryos were compared to un-injected WT embryos to obtain percent similarity. Bar graph reveals a significant reduction of eye size in Ddb1 KD that is partially rescued with ddb1 mRNA (error bars represent s.e.m., one-way ANOVA with Turkey’s multiple comparison, *P < 0.05). (d) Graph shows percent of normal, fused, small, or absent otolith phenotypes with MO treatments. Data reveals 21% of Ddb1 KD embryos exhibited normal otoliths, while 85% of Ddb1 KD + high ddb1 mRNA embryo otoliths were normal. Scale bars: 100 μm.