Submitted by for the award of the degree of

171 Figure 4.10 (a) Schematic representation of a (4–8–1) neural network (having three neurons in the input layer, eight neurons in the hidden layer, and one in the output layer). b) The prediction performance of ANN models for the hIFN-γ production. 221 Figure 5.3 (a) Schematic representation of a (4–8–1) neural network (having three neurons in the input layer, eight neurons in the hidden layer, and one in the output layer). b) The prediction performance of ANN models for the hIFN-γ production.

Introduction

Biopharmaceuticals
Interferon gamma (IFN-γ)
Development in production of recombinant hIFN-γ
Purification of recombinant hIFN-γ
Pichia pastoris: As a versatile host expression system
Bioprocess development for High cell density cultivation of Pichia pastoris
Metabolic engineering and flux balance analysis
Objectives of present study
Approach
Thesis organization
Reference

Reports of heterologous production of recombinant hIFN-γ (rhIFN-γ) in several systems are available that include adenovirus (Xu et al., 1997), E. Çelik et al., 2010, investigated the regulatory effect of methanol feeding in the presence and absence of sorbitol on metabolic flux distributions of P.

Review of Literature

Historical perspective of human interferon gamma

Later, Gray and Goeddel, 1982 deciphered the structure of the hIFN-γ gene, and further cloning of the gene encoding the protein was carried out during the 1980s, leading to a more detailed analysis of the protein (Devos et al ., 1982).

Genomics & proteomics of hIFN-γ

Four helices (A, B, C, and D) from one subunit and two from the other (E' and F') interact to form one of two distinct, symmetrical domains of the protein. In contrast, many other studies found that truncation in the C-terminal domains and more than 9 amino acids decreased the biological activity of the recombinant protein.

JAK-STAT signaling pathway of interferon signal transduction

It is also heat sensitive and undergoes irreversible thermal denaturation in solution at a temperature range of 40±5⁰C (Younes and Amsden, 2002). Mikulecký et al., 2016, revealed the crystal structure of IFN-γR2 and reported the importance of N-glycosylation for the stability of IFN-γR2 and also revealed the role of IFN-γR2 in the receptor specificity of hIFN-γ.

Mechanism of IFN-γ

The other route of antitumor action of IFN-γ involves indirect stimulation of the immune system, which triggers a series of immunological responses. Both humoral and cellular effects are produced and may depend on the dose of IFN-γ administered and its timing in relation to the immune process (Yasui et al., 1986).

Biological significance of hIFN-γ

The role of IFN-γ has already been well established in the treatment of chronic granulomatous disease (Marciano et al., 2004; Further IFN-γ produced by a company such as IMMUNEX® has shown a positive effect during clinical trials in atopic dermatitis (Panahi) . et al., 2012).

Current status of hIFN-γ

Recently, an Iranian company called Exir Pharma Ltd started manufacturing interferon-gamma under the trade name γ-Immunex (http://www.drugbank.ca/drugs/DB00033). The Indian GMO Research Information System (IGMORIS) recommended clinical trials for Human Interferon gamma-1b in 2009.

Table 2. 1. List of organizations producing & marketing Interferon-Gamma as a therapeutic

Bioprocessing of human interferon gamma (hIFN-γ)

Expression of hIFN-γ in E.coli
Expression of recombinant hIFN-γ in other protein production systems

To overcome this barrier, other expression systems are used in addition to E.coli (Leister et al., 2013). Eukaryotic systems such as Xenopus laevis oocytes have also been reported to be used in heterologous expression of hIFN-γ (Taya et al., 1982).

Purification of hIFN-γ

Purification of insoluble proteins requires complex extraction and expensive denaturation and refolding procedures. Techniques include immunoaffinity chromatography with monoclonal antibodies (Honda et al., 1987), size exclusion chromatography (Reddy et al., 2007; Vandenbroeck et al., 1993) and ion exchange chromatography (Haelewyn and De Ley, 1995; Petrov). et al., 2010).

Reverse micellar extraction as advanced method for protein purification

Diffusion of biomolecules within micelles is governed by electrostatic, steric, and hydrophobic interactions between proteins and opposing micelles (Hong et al., 2000; Umesh Hebbar et al., 2008a). The selectivity of the RME system can be affected by various parameters such as pH, type of salt and its concentration, type of solvent, type of surfactant and its concentration, water content and volume ratio of the aqueous phase to the organic phase (LIU et al , 2008).

Recombinant protein production

However, isolating and purifying these proteins from natural sources is tedious and results in low concentration, which significantly increases downstream processing costs and the risk of infectious contamination during extraction (Porro et al., 2005). It was observed that more than half of protein-based biopharmaceuticals are produced in microbial systems (~30% in Escherichia coli and ~20% in Saccharomyces cerevisiae), while the remainder is mainly produced by mammalian cells ( Martínez et al., 2012 ).

Pichia pastoris: As a platform for recombinant protein production

The enzymes involved in the methanol utilization pathway are alcohol oxidase, catalase and dihydroxacetone synthase, which are found in the microbodies, peroxisomes. These compounds are further metabolized in the cytosol to finally recycle xylulose-5-phosphate in a cyclic pathway.

Expression Design

A third of the glyceraldehyde-3-phosphate produced becomes available for central metabolism and biomass production. Factors such as promoters, selectable markers, phenotypes of the strains, controlling proteolysis, post-translational modifications, gene dosage, are the main factors that determine the mRNA level, protein folding, translocation and translation of the recombinant proteins.

Promoters

Inducible promoters
Constitutive promoters

Recent developments have been made regarding the increase of the Aox2 gene using the truncated version of the AOX2 promoter. However, it may have the disadvantage of cytotoxic effects of the constitutive production of the foreign proteins in Pichia pastoris and is therefore not so widespread.

Figure 2.10 Generic representation of AOX based promoter for Pichia pastoris

Expression strains

Auxotrophic strains
Protease deficient strains
Other efficient strains

Proteolysis of secreted heterologous protein can occur during their vesicular transport by resident proteases (Werten et al., 1999) or cell wall-associated proteases (Kang et al., 2000), various strategies have been applied to overcome this problem, including protease deficient strains. Lower growth rates and transformation efficiencies are observed in protease-deficient strains and are therefore highly unstable (Cereghino and Cregg, 2000).

Genomic Integration of Vectors in the P. pastoris genome

Insertion
Gene replacement
Multiple gene insertions

They can be confirmed by Southern blot analysis or direct expression of a foreign assay (Cereghino and Cregg, 2000). Multicopy gene insertions can be performed by constructing vectors that comprise the entire expression cassette (Brierley et al., 1990).

Selectable markers

Signal processing and glycosylation in Pichia pastoris

Processing of the secretion signals
Glycosylation in Pichia pastoris

The absence of sialic acid also resulted in rapid removal of the protein from the mammal's bloodstream. It has been reported that expression of trans-sialidase under the control of the AOX1 promoter in the absence of co-expressed mannosidase does not contain the terminal 1,6-linked mannose residues normally found in P.

Optimization of recombinant protein production in Pichia pastoris

Fusion FC protein has also been reported to be successfully expressed in glyco-engineered Pichia pastoris strain (Jacobs et al., 2008; Laukens et al., 2015). This problem can be solved by increasing the GC content in gene synthesis (Prabhu et al., 2016b).

Protein folding

Strategies for these proteolytic processes include selection of protease-deficient host strains (Gleeson et al., 1998; Goodrick et al., 2001), modification of protein structure to resist the protease (Gustavsson et al., 2001), addition of protease inhibitors in bioreactor ( Holmquist et al., 1997), changing the pH of the culture medium (Jahic et al., 2003; Kang et al., 2000), supplementing the medium with casamino acids and peptone (Clare et al., 1991) ) and adding protease inhibitors (Shi et al., 2007). The expression system has also been optimized for culture level factors such as pH, temperature, dissolved oxygen (DO) content, methanol monitoring, etc. (Cos et al., 2006; Looser et al., 2015; Potvin et al., 2012).

Proteolysis

It has been observed that when proteasomes such as PrA are overexpressed, part of the enzyme can be secreted. Most of the proteases are stress dependent such as starvation, change of carbon sources, heat and pH changes, or toxic chemicals.

Metabolic flux analysis of recombinant protein secretion by Pichia pastoris

Systems biology provides detailed information about the genome that aids in the manipulation of targeted genes in the cellular network, characterizes the functional behavior of the biological system from a holistic perspective, and identifies novel biological entities that contribute to the improved production of chemicals and materials ( Lee et al., 2005). Later, Çelik et al., 2010 reported a central metabolic model of Pichia pastoris, consisting of 141 reactions and 102 metabolites.

Optimization of the production of gurmarin, a sweet taste suppressing protein secreted by the methylotrophic yeast Pichia pastoris. Enhancing secretory production of heterologous proteins in Pichia pastoris by targeting protein folding.

Development of cellular engineering strategy for improved production

Background

Overexpression of heterologous proteins sometimes impairs the expression mechanism of the host system and results in metabolic burden (Mattanovich et al., 2004). Later, the unfolded protein in ER was subjected to endoplasmic reticulum-associated degradation (ERAD), which ultimately contributed to low protein expression levels (Zhang et al., 2006).

Materials and Methods

Strains, Vectors, Kits and medium
Cloning of hIFN-γ in Pichia pastoris
PCR reaction for hIFN-γ full length amplification using Q5 DNA polymerase
EcoRI –XbaI digestion of hIFN-γ full length amplified gel eluted PCR product
EcoRI –XbaI digestion of pPICZαA vector
Ligation reaction of pPICZ-α plasmid with full length hIFN-γ double digested
Transformation of ligation reaction into the E.coli, Top10 cells
Linearization of pPICZαA-hIFN-γ construct by using enzyme SacI
Transformation of pPICZαA-hIFN-γ plasmid into Pichia pastoris
Determination of Mut phenotype
Cloning of molecular chaperons
Cloning of codon optimized hIFN-γ gene
Expression studies
Optimization of process parameters using one factor at a time experiments
Qualitative protein analysis by SDS-PAGE
Western blot analysis
Production of hIFN-γ on bench top fermenter
Model parameter estimation
Enzyme linked immunosorbent assay (ELISA) of IFN-γ
Dry cell weight (DCW)
Statistical analysis

In this study, pKANB-PDI pKANB is based on a vector carrying a copy of the PDI gene: Genr V tem. Columns were prepared with 1 ml of 50% Ni-NTA resin (Invitrogen) and equilibrated in buffer (25 mM sodium phosphate, 0.5 mM NaCl, pH 8.0). Samples were then loaded onto Ni-NTA columns and washed with 5 column volumes of equilibration buffer and protein was eluted with elution buffer (25 mM phosphate buffer, pH 8.0, containing 150 mM imidazole and 0.5 mM NaCl).

Table 3.1 Oligonucleotides used in this study

Results and discussion

Cloning of hIFN-γ in Pichia pastoris
Effect of single chaperons on hIFN-γ in Pichia pastoris
Synergetic effect of chaperons on the production of hIFN-γ
Effect of chaperons on growth characteristics of Pichia pastoris
Kinetic studies on Bench top reactor
Synthesis of codon-optimized gene and expression in P. pastoris
Effect of Temperature on rhIFN-γ expression
Effect of pH on rhIFN-γ expression
Effect of Methanol concentration on rhIFN- γ expression
Effect of Agitation on rhIFN-γ expression
Effect of Inoculum concentration on rhIFN-γ concentration
Purification of rhIFN-γ and western blotting analysis

In our previous study, we deduced that involvement of chaperones affects the growth characteristics of recombinant Pichia pastoris (Prabhu et al. 2016). This result means that the chaperone relieves the inhibition of yeast growth induced by overexpression of a foreign protein (Zhang et al., 2006).

Figure 3.1 (a) PCR amplification of 432bp hIFN-γ gene fragment M- 1kb ladder, Lane 1- amplified cDNA of 432bp hIFN- γ gene (b): Clone confirmation by release check analysis using EcoRI and XbaI enzyme Lane 1,2,3- positive clones showing

Conclusion

The results showed that the production of rIFN-γ was related to growth and 1.98 mg/L of protein was produced. Further codon optimization of the gene encoding hIFN-γ resulted in a significant increase in protein production reaching 1.8 mg/L, this yield being due to more efficient translational efficiency and more stable mRNA.

Expression of heterologous proteins in Pichia pastoris: a useful experimental tool in protein engineering and production. Enhanced secretion of heterologous proteins in Pichia pastoris after overexpression of Saccharomyces cerevisiae chaperone proteins.

Optimization of nutrient concentration and development of substrate

Background

The major problem associated with this medium is the formation of precipitates at higher ionic strength (Cos et al., 2006). Furthermore, Jahic et al., 2002 showed that the low maintenance energy of Pichia pastoris in (glycerol/methanol) medium allows it to grow at high cell density.

Materials and Methods

Strain and media
Primary screening of carbon and nitrogen sources and their effect on hIFN-γ
Effect of aeration on Protein expression
Effect of casamino acid on Protein expression
Optimization of screened components by Box-Behnken Design
Artificial neural network linked with genetic algorithm (ANN-GA) as a
Genetic Algorithm (GA)
Validation of the optimized conditions
Stirred tank bioreactor cultivation
Kinetic modeling
Kinetic Experiment
Substrate inhibition Growth Models for recombinant P. pastoris
Model for validation by statistical comparisons
Dual substrate models
Non-linear regression analysis
Analytical Methods
Estimation of gluconate and methanol

The test data were used to assess the predictive ability of the generated model (Song et al., 2004; Yasin et al., 2014). In this study, tansig was used as a transfer function between input and hidden layer, and the output produced by the hidden layer became input to the output layer, where purelin transferred functions and produced output the same as hidden layer (Khayet and Cojocaru, 2012). The equation for Purelin and tansig respectively is given below:.

Table 4.1. Coded values of independent variables for Plackett- burman screening

Results and discussion

Effect of carbon source on hIFN-γ production
Effect of Nitrogen source on hIFN- γ production
Effect of Aeration and casamino acid on hIFN-γ Production
Screening of essential medium components
Optimization of screened variables for maximization of hIFN-γ
Hybrid ANN-GA modelling

Validation of Box behnkhen and ANN-GA

Unstructured Model Prediction in batch reactor
Effect of initial substrate concentration on cell growth, specific growth rate
Modeling of growth kinetics of individual substrate on recombinant
Sensitivity Analysis of Kinetic Parameters and model acceptability analysis
Dual substrate growth kinetics for Recombinant Pichia pastoris

Similarly, significant effect was observed when glutamic acid was used as nitrogen source where approximately 7 mg/L hIFN-γ production was achieved. In this present investigation, we achieved a maximum of 30 mg/L hIFN-γ production from Pichia pastoris.