Characterisation of N-terminal fragments of Retinoblastoma Binding Protein 6 for structural analysis
Retinoblastoma Binding Protein 6 (RBBP6) is a 200 kDa RING finger-containing protein that plays a role in 3'-end poly-adenylation of mRNA transcripts as well as acting as an E3 ubiquitin ligase against a number of proteins involved in tumourigenesis, including p53. Since the human protein is too large and poorly structured for heterologous expression in bacteria, it would be advantageous to identify smaller fragments suitable for expression in bacteria. Many E3 ubiquitin ligases form homo-dimers and dimerisation is important for their activity; structural studies of the isolated RING finger of RBBP6 showed that it forms a weak homo-dimer. This poses the question of whether the complete RBBP6 protein forms homo-dimers in vivo, and, if so, whether a fragment of RBBP6 containing the RING finger could be identified which would be suitable for structural as well as functional studies. Such a construct would allow detailed investigation of the homo-dimeric state of the fragment, the relationship between dimerisation and ubiquitination activity, and the role of domains such as the DWNN domain and zinc finger in ubiquitination. A fragment consisting of the first 335 residues of RBBP6, dubbed R3 because it contained the first three domains of the protein, was expressed, along with three variants expressing mutations known to disrupt the dimerisation of the isolated RING finger. Size exclusion chromatography showed that R3 forms a strong homo-dimer that was not disrupted by the mutations, suggesting that additional parts of R3 outside of the isolated RING finger form part of the interface. To identify whether this included the DWNN domain or the zinc finger, a shorter fragment dubbed R2, excluding the N-terminal DWNN domain, was cloned and expressed. This was also found to form a strong homo-dimer, suggesting that the DWNN domain may not form an essential part of the dimer interface. Availability of the RING finger samples and monomerising mutations allowed investigation of whether the RING finger from RBBP6 was able to auto-ubiquitinate itself. Using a fully in vitro ubiquitination assay supplemented with intact proteasomes purified from human cell lysates, we found that wild type RING auto-ubiquitinates itself very efficiently, catalysing its own destruction in the proteasome. This provides an answer to the question of why RBBP6 is so difficult to detect in mammalian cells. Surprisingly, monomeric mutant RING fingers were also able to auto-ubiquitinate and catalyse their own destruction, although perhaps not as efficiently as wild type. This result would appear to rule out the hypothesis that dimerisation of RBBP6 is required for ubiquitination activity. Finally, samples of the RING finger from human MDM2 were expressed in bacteria and used to investigate whether the RING fingers of RBBP6 and MDM2 interact directly with each other. If so, this may provide a mechanism whereby RBBP6 and MDM2 cooperate in ubiquitination of p53. The results of a GST pull down assay using GST-MDM2-RING as ''bait'' and RBBP6-RING as ''prey'' provides evidence that such an interaction between the RING does exist. This work lays the foundation for future structural studies of the RING-RING hetero-dimer using protein Nuclear Magnetic Resonance Spectroscopy.