The membrane proteins signal peptide peptidase, signal peptide peptidase like and presenilin are intramembrane aspartyl proteases located in the endoplasmic reticulum, plasma membrane and organelle. These membrane proteins are able to catalyze a hydrolytic reaction in a hydrophobic space. The downstream consequences of these reactions impact a variety of cellular functions such as cytokine production, inflammatory responses, embryogenesis, and immune system regulation. Additionally, the aspartyl proteases such as signal peptide peptidase and presenilin, a part of the γ-secretase complex, hydrolyze peptides leading to pathogen maturation and Alzheimer’s disease, respectively. Electron crystallography offers the unique aspect of studying membrane proteins in a near native state. Determining the structures of Haloarcula morismortui and Methanoculleus marisnigri JR1 signal peptide peptidases by electron crystallography may provide insight into how a hydrolysis reaction occurs in a hydrophobic environment and how the protein determines which transmembrane signal peptides to cleave. Additionally, structure determination may help answer questions regarding why human presenilin, part of the γ-secretase complex, incorrectly processes amyloid precursor protein into amyloid-beta peptides leading to Alzheimer’s disease. Such structural data may not only shed light on how amyloid precursor protein is processed but how other proteins are processed by signal peptide peptidase leading to immune responses, cell signaling, and pathogen maturation. In addition, structure-function data may have an impact on pharmaceutical drug designs that targets signal peptide peptidase, signal peptide peptidase like, and/or presenilin. To determine the structure of aspartyl proteases, two archaeal signal peptide peptidases were used for two-dimensional crystallization trials to be able to study their structure by electron crystallography. Haloarcula morismortui and Methanoculleus marisnigri JR1 signal peptide peptidases, both human signal peptide peptidase homologues, were recombinantly over-expressed and purified. During dialysis trials, various lipid-to-protein ratios, sodium chloride concentrations, temperatures, detergents and a variety of other variables were tested. Methanoculleus marisnigri JR1 signal peptide peptidase showed the most promising results in terms of crystallinity. Optimizing dialysis conditions, specifically narrowing the lipid to protein ratio, resulted in two-dimensional crystals. Ordered arrays measuring up to 200 nm x 200 nm were observed. These ordered arrays have been shown to be reproducible amongst multiple batches of purified Methanoculleus marisnigri JR1 signal peptide peptidase. Preliminary projection maps of negatively stained ordered arrays show unit cell dimensions of a = 178 Å, b = 160 Å, γ = 92.0 Å and a = 175 Å, b = 167 Å, γ = 92.0 Å. The monomer measurements are approximately 70 Å by 80 Å. This is the first time a signal peptide peptidase homologue has been crystallized by two-dimensional crystallization.

No description available.

Crystallization is often a major bottleneck to macromolecular structure determination. This is particularly true for membrane proteins, which have hydrophobic surfaces that cannot readily form crystal contacts. Of the roughly 109,000 protein structures in the PDB, only about 539 represent unique membrane proteins, despite immense interest in membrane proteins from both a biological and therapeutic standpoint. Membrane protein crystallization has been facilitated by the development of new detergents, lipidic cubic phase methods, soluble protein chimeras, and non-covalent protein complexes. The design process of protein fusion constructs and non-covalent antibody fragments specific for each target membrane protein, however, is costly and time-consuming. An improved, more general method of membrane protein co-crystallization is needed. This dissertation details the development of two approaches for cost-effective non-covalent crystallization chaperones: (1) Engineered hypercrystallizable Fab antibody fragment with high affinity for EYMPME (EE epitope), which form complexes with EE-tagged soluble and membrane proteins. (2) Engineered monomeric streptavidin (mSA2) for complexation with biotinylated membrane proteins. Both methods are generalizable through insertion of a short epitope into a surface-exposed loop of a membrane protein by site directed mutagenesis. Crystallization trials of representative chaperone-membrane protein complexes and possible difficulties with the approach are discussed.

Signal peptide peptidase (SPP) and SPP-like (SPPL) aspartyl intramembrane proteases are known to contribute to sequential processing of type II-oriented membrane proteins referred to as regulated intramembrane proteolysis. The ER-resident family members SPP and SPPL2c were shown to also cleave tail-anchored proteins, including selected SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) proteins facilitating membrane fusion events. Here, we analysed whether the related SPPL2a and SPPL2b proteases, which localise to the endocytic or late secretory pathway, are also able to process SNARE proteins. Therefore, we screened 18 SNARE proteins for cleavage by SPPL2a and SPPL2b based on cellular co-expression assays, of which the proteins VAMP1, VAMP2, VAMP3 and VAMP4 were processed by SPPL2a/b demonstrating the capability of these two proteases to proteolyse tail-anchored proteins. Cleavage of the four SNARE proteins was scrutinised at the endogenous level upon SPPL2a/b inhibition in different cell lines as well as by analysing VAMP1-4 levels in tissues and primary cells of SPPL2a/b double-deficient (dKO) mice. Loss of SPPL2a/b activity resulted in an accumulation of VAMP1-4 in a cell type- and tissue-dependent manner, identifying these proteins as SPPL2a/b substrates validated in vivo. Therefore, we propose that SPPL2a/b control cellular levels of VAMP1-4 by initiating the degradation of these proteins, which might impact cellular trafficking.