Biochemical and structural characterization of separins, proteinases required for the separtion of sister chromatids

In eukaryotic cells, the DNA replication takes place during the S-phase of the cell cycle. However, separation of the newly replicated chromosomes does not occur until the metaphase to anaphase transition in mitosis. Up to this point, the sister chomratids are held together by a complex of proteins known as cohesin. Chromosome separation is then achieved by activation of a protein called separin which carries out specific proteolysis of a subunit of the cohesin complex. This phenomenon appears to be conserved amongst all eukaryotes. The separin protein has been recently shown to be a cysteine proteinase, related by sequence similarity over about 120 amino acids to the caspases, gingipains and legumains, themselves all members of the CD clan of cysteine proteinases. At present, nothing is known about the enzymatic mechanism, substrate specificity or three- dimensional structure of the separin enzymes. It is the goal of this project to carry out such a characterisation. The following approach will be taken. Firstly, it is essential to establish a recombinant expression system for separins, as it is not feasible to isolate sufficient material from eukaryotic cells undergoing mitosis. Thus, cDNAs encoding the proteolytic domains of separins from S.cerevisiae, S. pombe, D. melanogaster and H. Sapiens will be introduced into standard prokaryotic expression vectors, so that the proteins can be produced at high levels and subsequently purified, either by the use of affinity labels or by standard methods of protein chemistry. Separin proteins from four different organisms will be expressed for two reasons. Firstly, the behaviour of a protein in an expression system cannot be predicted. Only experimentation can determine which proteins can be stably expressed at high levels in a recombinant system; the use of separins from different species will enable the most suitable to be chosen. Once sufficient amounts of pure separin have been prepared, assays will be developed for enzymatic activity using fluorogenic substrates. These will be hexapeptides comprising the six amino acids preceding a cleavage site of separin on the processed cohesin (the sites are known for S. cerevisiae and S. pombe separins and can be predicted by sequence similarity for the other two organisms). Optimal conditions and enzymatic parameters can be determined with the assay. As the separation of the sister chromatids is an essential event for cell division and hence cell growth, it represents a target for inhibition and thus for the development of antibiotics. However, as the process is also found in mammalian cells, it is important that inhibitors can be found which are specific for the fungal enzyme and do not affect the mammalian ones.To design such inhibitors, a thorough knowledge of the substrate specificities of the separins is essential. This is a major goal of this project. The approaches will be the use of fluorgenic peptide libraries on the one hand and a genetic screen, either in a bacterial or in a yeast cell, on the other. Finally, this knowledge and its implications for inhibitor design will be complemented and extended by the determination of the three-dimensional crystal structure of the proteolytic domain of a separin.

The goal of this project was to characterise both biochemically and structurally the protein separin. This protein is responsible for the cleavage of cohesin and thus separation of the chromosomes during mitosis and meiosis. To reach this goal, sufficient amounts of the purified protein must be available. Initially, we applied different bacterial expression systems to produce the protein, using separin cDNAs from organisms as varied as baker`s yeast, the fruit fly and man. Despite the use of different cDNAs and several different expression systems, we were unable to generate sufficient soluble separin protein. The hydrophobicity of the separin proteins appears to cause them to precipitate, even at low concentrations. Experiments to renature the proteins from inclusion bodies also failed. Using bioinformatics, we then looked for a related prokaryotic protein as it seemed possible that such a protein might be more soluble in E.coli. The separins are members of the group of proteins containing a domain referred to as the caspase-haemoglobinase fold. Investigation of published bioinformatics data on this group of proteins revealed that proteins of certain cyanobacteria such as Anabaena variabilis and Nostoc species possess two proteins which show similarity to the separins in regions around their putative active site. These proteins are named HetF and HetF-like. The two proteins are involved, as their name suggest, in heterocyst formation. These organelles are required for nitrogen fixation and are produced in an environment lacking nitrogen in a combined form. The heterocysts are produced in order to maintain a low oxygen environment, thus protecting the nitrogenase from oxidation. Interestingly, the function of these proteins in heterocyst formation is not known, although they appear in some way to regulate the activity of the protein HetR, itself a proteinase. We expressed the HetF-like and HetF proteins of A. variabilis and Nostoc sp. in E.coli. Successful expression in E.coli and purification in milligram amounts of the HetF-like protein of A. variabilis was achieved. Screening for protein crystals was attempted in two laboratories in Vienna and Barcelona, but was not successful. The problem appeared to be the low solubility of the protein at the high concentrations required for screening. Protease digestion of the HetF-like protein showed that 90% of the molecule was resistant to digestion, indicating correct folding of the molecule. We assume that the proteinase sensitive regions were not structured and led to the precipitation. Future experiments will concentrate on screening the proteinase resistant domain of HetF-like for crystals to enable the structure of a protein of this interesting group of proteins to be determined.