Quantitative analysis of the operation and control of oxidative protein folding in the yeast endoplasmic reticulum

All living cells possess the ability to secrete proteins from their interior to their exterior environment. This ability serves a variety of purposes, including the transformation of nutrients into a form where they can be taken up by the cell, communication with other cells, and the formation of scaffold structures on which cells can grow. The ability of cells to secrete proteins is also exploited by the bioprocessing industry, which re-programs cells to make and secrete protein-based frontline drugs against debilitating diseases like cancer, multiple sclerosis and arthritis.
Part of the secretion processes in higher (eukaryotic) cells is to ensure that secreted proteins adopt a structure in which they have optimal activity. Proteins are polymeric strings of amino acids, and their folded structure depends on interactions between individual amino acids within them. A specific type of interaction that is critical to the activity of many proteins occurs when two cysteine amino acids form bonds between the sulphur atoms they contain: this is known as a disulphide bond. Disulphide bond formation occurs as an integral part of the secretion process, and involves a cascade of specific enzymes. These enzymes remove an electron from the interacting cysteines, allowing them to form a bond between them that determines the affected protein's shape. The electron is then passed between different enzymes and ultimately onto an oxygen atom, which reacts with water to form hydrogen peroxide. Since the latter is toxic if present in large amounts, it has to be removed in a further series of reactions. The entirety of these reactions is called the oxidative protein folding (OPF) pathway and is the focus of this project.
There are fundamental differences between the OPFs of different types of higher cells. We will explore differences between the OPFs of two specific cell types (simple yeast cells and complex human cells) to improve our understanding of the molecular machinery involved in oxidative folding. Such knowledge will also improve our ability to manipulate the pathway by genetic engineering in order to generate better producing cells for the bioprocessing industry.
Yeast cells only secrete relatively small amounts of proteins, and their OPF machinery therefore evolved to operate on a minimal enzyme set. In contrast, many human cells are prolific secretors, due to their need to communicate extensively with other cells in the body, to produce enzymes for the digestion of food, or to produce molecules of the immune system. Human cells therefore have a much more complex OPF, with different forms of the OPF enzymes that are only act on specific types of target proteins. Interestingly, human cells are also able to use the toxic hydrogen peroxide to drive the OPF reactions, whereas yeast cannot do this.
We will use a three-pronged strategy to exploit these differences: 1), we will isolate the enzymes of the OPF machinery from yeast and human cells and will study their detailed properties in test tubes; 2) we will use the information from these experiments to generate a computational model that can predict properties of the OPF pathways inside cells; and 3) we will use predictions made with the computational model to change properties of the yeast OPF enzymes, and to mix them with human enzymes, in living yeast cells. Overall, this strategy will enable us to better understand how the OPF machinery functions, and in the longer term will enable us to engineer yeast cells that are better suited for use in bioprocessing applications.

The oxidative protein folding (OPF) pathway in the ER generates disulphide (SS)-bonded native proteins, yet we know surprisingly little about quantitative aspects of the pathway, how it is regulated or how it links to endogenous reducing processes. By combining experimental in vivo, in vitro and modelling work on the yeast and human OPF pathways, we will provide important new insights into the operation and control of SS-bond formation in eukaryotic cells. The focus will be on processes that convert reduced proteins into native, export-competent SS-bonded proteins. Our study will be aided by development of a computational model, which will initially focus on the simpler yeast pathway but subsequently extended to include heterologous components.
Yeast strains expressing homologous and recombinant human proteins will be used to determine i) the levels of secretion of the target proteins, ii) the abundance and redox status of Ero1p and Pdi1p, iii) the intracellular levels of reduced, native and intermediate forms of the target proteins, and iv) the extent of ER stress responses. In parallel, we will analyse the pathway in vitro using purified components (e.g. Ero1p and Pdi1p) with reduced ribonuclease as substrate. We will extend the assay by inclusion of recently-discovered components which use H2O2 to reoxidise reduced PDI and thus derive kinetic parameters and identify stoichiometries of the various catalysts that optimise production of active folded RNAase.
The computational model will be used to make quantitative predictions which will suggest in vivo manipulations of yeast. The yeast OPF pathway will engineered by varying the levels of components, by introducing novel capabilities for using H2O2 generated in the core pathway, and by adding the capabilities of various human PDI isoforms to facilitate OPF in both the in vitro and in vivo systems with the long term aim of engineering yeast strains optimally designed to produce high-value biopharmaceuticals.

This project will have a significant impact on UK industry, the UK economy and in the longer term the wider population through medical benefits.

Impact on Industry.
Within the biopharmaceutical industry, the need for the cost-effective and robust production of pharmaceutical grade recombinant proteins is now a major priority. Improving the fold, solubility and purity of high value rPs remains a major challenge for this sector, in particular post-translational modifications such as glycosylation and disulphide bond formation that are prevalent in many of the high value target proteins such as recombinant antibodies. Our study will deliver new knowledge on the pathway that ensures the correct formation of disulphide bonds i.e. the oxidative protein folding pathway. This knowledge will be of benefit to end users who seek to rationally engineer improved eukaryotic host cells that are optimised in this pathway. A number of recent reviews (e.g. Idiris et al. 2010, Appl Microb Biotech 86: 403) have highlighted disulphide bond formation as the most promising target for improving key parameters that will increase the efficacy and accuracy of protein secretion.

Impact on the Economy.
Global sales of recombinant therapeutic proteins including monoclonal antibodies (mAbs) were estimated to be $160 billion in 2013 (BCC Research 2008). The bioprocessing industry is thus a major economic factor world-wide and the UK is a major player in the sector, with nearly all international companies having a UK base. In addition, numerous UK-based SMEs are highly active in this sector. Even a minor improvement in host cell capability will translate into a measureable impact on this sector. The systems level information we obtain with yeast will provide a platform on which step changes in production platforms can be made - both yeast and mammalian cells - that will have long-lasting economic benefits.

Impact through Medical Benefits.
Biopharmaceuticals are state-of-the-art treatments for many high-impact diseases including multiple sclerosis, cancer, and inflammatory diseases. These diseases have adverse impact beyond the affected individuals themselves including carers and dependent family members. Improvements in the production of recombinant proteins thus have high potential for positive impact felt in wider society. In the longer term, more efficient production of higher quality drugs will benefit the Health Service provision and patient.