Ann R. Kusnadi

Production of recombinant proteins in transgenic plants: Practical considerations

This review is based on our recent experience in producing the first commercial recombinant proteins in transgenic plants. We bring forward the issues that have to be considered in the process of selecting and developing a winning transgenic plant production system. From the production point of view, transcription, posttranscription, translation, and posttranslation are important events that can affect the quality and quantity of the final product. Understanding the rules of gene expression is required to develop sound strategies for optimization of recombinant protein production in plants. The level of recombinant protein accumulation is critical, but other factors such as crop selection, handling and processing of transgenic plant material, and downstream processing are equally important when considering commercial production. In some instances, the cost of downstream processing alone may determine the economic viability of a particular plant system. Some of the potential advantages of a plant production system such as the high levels of accumulation of recombinant proteins, glycosylation, compartmentalization within the cell, and natural storage stability in certain organs are incentives for aggressively pursuing recombinant protein production in plants.

Commercial production of avidin from transgenic maize: characterization of transformant, production, processing, extraction and purification

We have produced in transgenic maize seed the glycoprotein, avidin, which is native to avian, reptilian, and amphibian egg white. A transformant showing high-level expression of avidin was selected. Southern blot data revealed that four copies of the gene are present in this transformant. The foreign protein represents >2% of aqueous soluble extracted protein from populations of dry seed, a level higher than any heterologous protein previously reported for maize. In seed, greater than 55% of the extractable transgenic protein is present in the embryo, an organ representing only 12% of the dry weight of the seed. This indicates that the ubiquitin promoter which is generally considered to be constitutive, in this case may be showing a strong tissue preference in the seed. The mature protein is primarily localized to the intercellular spaces.An interesting trait of the transgenic plants expressing avidin is that the presence of the gene correlates with partial or total male sterility. Seed populations from transgenic plants were maintained by outcrossing and segregate 1:1 for the trait. In generations T2–T4, avidin expression remained high at 2.3% (230 mg/kg seed) of extractable protein from seed, though it varied from 1.5 to 3.0%. However, levels of expression did not appear to depend on pollen parent or growing location. Cracked and flaked kernels stored at −29°C or 10 °C for up to three months showed no significant loss of avidin activity. Commercial processing of harvested seed also generated no apparent loss of activity. The protein was purified to greater than 90% purity by affinity chromatography after extraction from ground mature maize seed. Physical characterization of purified maize-derived avidin demonstrated that the N-terminal amino acid sequence and biotin binding characteristics are identical to the native protein with near identical molecular weight and glycosylation. This study shows that producing avidin from maize is not only possible but has practical advantages over current methods.

Process and Economic Evaluation of the Extraction and Purification of Recombinant β-Glucuronidase from Transgenic Corn

A process model for the recovery and purification of recombinant β‐glucuronidase (rGUS) from transgenic corn was developed, and the process economics were estimated. The base‐case bioprocessing plant operates 7500 h/year processing 1.74 million (MM) kg of transgenic corn containing 0.015% (db) rGUS. The process consists of milling the corn into flour, extraction of protein by using 50 mM sodium phosphate buffer, and rGUS purification by ion exchange and hydrophobic interaction chromatography. About 137 kg of rGUS of 83% (db) purity can be produced annually. The production cost amounted to

Production and Purification of Two Recombinant Proteins from Transgenic Corn

43 000/kg of rGUS. The cost of milling, protein extraction, and rGUS purification accounted for 6, 40, and 48% of annual operating cost, respectively. The cost of transgenic corn was 31% of the raw material costs or 6% of the annual operating cost. About 78% of the cost of buffer and water were incurred in the protein extraction section, while 88% of other consumables were from the purification section. The sensitivity analysis indicated that rGUS can be produced profitably from corn even at the 0.015% (db) expression level, assuming a selling price of

Commercial production of β-glucuronidase (GUS): a model system for the production of proteins in plants

100 000/kg GUS. An increase in rGUS expression levels up to 0.08% significantly improves the process economics.

Molecular Farming of Industrial Proteins from Transgenic Maize

This study reports the production, purification, and characterization of recombinant Escherichia coli β‐glucuronidase (GUS) and chicken egg‐white avidin from transgenic corn seed. The avidin and gus genes were stably integrated in the genome and expressed over seven generations. The accumulation levels of avidin and GUS in corn kernel were 5.7% and 0.7% of extractable protein, respectively. Within the kernel, avidin and GUS accumulation was mainly localized to the germ, indicating possible tissue preference of the ubiquitin promoter. The storage‐stability studies demonstrated that processed transgenic seed containing GUS or avidin can be stored at 10 °C for at least 3 months and at 25 °C for up to 2 weeks without a significant loss of activity. The heat‐stability experiments indicated that GUS and avidin in the whole kernels were stable at 50 °C for up to 1 week. The buffer composition also had an affect on the aqueous extraction of avidin and GUS from ground kernels. Avidin was purified in one step by using 2‐iminobiotin agarose, whereas GUS was purified in four steps consisting of adsorption, ion‐exchange, hydrophobic interaction, and size‐exclusion chromatography. Biochemical properties of purified avidin and GUS were similar to those of the respective native proteins.

Processing of transgenic corn seed and its effect on the recovery of recombinant β‐glucuronidase

We have generated transgenic maize seed containing β-glucuronidase(GUS) for commercial production. While many other investigators have demonstrated the expression of GUS as a scoreable marker, this is one of the first cases where a detailed characterization of the transgenic plants and the protein were performed which are necessary to use this as a commercial source of GUS. The recombinant β-glucuronidase was expressed at levels up to 0.7% of water-soluble protein from populations of dry seed, representing one of the highest levels of heterologous proteins reported for maize. Southern blot analysis revealed that one copy of the gene was present in the transformant with the highest level of expression. In seeds, the majority of recombinant protein was present in the embryo, and subcellular localization indicated that the protein was dispersed throughout the cytoplasm. The purified recombinant β-glucuronidase (GUS) was compared to native β-glucuronidase using SDS-PAGE and western blot analysis. The molecular mass of both the recombinant and native enzymes was 68 000 Da. N-terminal amino acid sequence of the recombinant protein was similar to the sequence predicted from the cloned Escherichia coli gene except that the initial methionine was cleaved from the recombinant GUS. The recombinant and native GUS proteins had isoelectric points (pI) from 4.8 to 5.0. The purified proteins were stable for 30 min at 25, 37, and 50 ° C. Kinetic analysis of the recombinant and native GUS enzymes using 4-methylumbelliferyl glucuronide (MUG) as the substrate was performed. Scatchard analysis of these data demonstrated that the recombinant enzyme had a Km of 0.20 mM and a Vmax of 0.29 mM MUG per hour, and the native enzyme had a Km and Vmax of 0.21 mM and 0.22 mM/h respectively. Using D-saccharic acid 1,4-lactone, which is an inhibitor of β-glucuronidase, the Ki of the native and recombinant enzymes was determined to be 0.13 mM. Thus, these data demonstrate that recombinant GUS is functionally equivalent to native GUS. We have demonstrated the expression of high levels of GUS can be maintained in stable germlines and have used an efficient recovery system where the final protein product, GUS, has been successfully purified. We describe one of the first model systems for the commercial production of a foreign protein which relies on plants as the bioreactor.

Strategies for Recombinant Protein Recovery from Canolaby Precipitation

Recombinant egg white avidin and bacterial B-glucuronidase (GUS) from transgenic maize have been commercially produced. High levels of expression were obtained in seed by employing the ubiquitin promoter from maize. The recombinant proteins had activities that were indistinguishable from their native counterparts. We have illustrated that down-stream activities in the production of these recombinant proteins, such as stabilizing the germplasm and processing for purification, were accomplished without any major obstacles. Avidin (A8706) and GUS (G2035) are currently marketed by Sigma Chemical Co.

Recombinant aprotinin produced in transgenic corn seed: Extraction and purification studies

The tools of plant biotechnology that have been developed to improve agronomic traits are now being applied to generate recombinant protein products for the food, feed, and pharmaceutical industry. This study addresses several processing and protein recovery issues that are relevant to utilizing transgenic corn as a protein production system. The gus gene coding for beta-glucuronidase (rGUS) was stably integrated and expressed over four generations. The accumulation level of rGUS reached 0.4% of total extractable protein. Within the kernel, rGUS was preferentially accumulated in the germ even though a constitutive ubiquitin promoter was used to direct gus expression. Fourth-generation transgenic seed was used to investigate the effect of seed processing on the activity and the recovery of rGUS. Transgenic seed containing rGUS could be stored at an ambient temperature for up to two weeks and for at least three months at 10 degrees C without a significant loss of enzyme activity. rGUS exposed to dry heat was more stable in ground than in whole kernels. The enzyme stability was correlated with the moisture loss of the samples during the heating. Transgenic seed was dry-milled, fractionated, and hexane extracted to produce full-fat and defatted germ fractions. The results of the aqueous extraction of rGUS from ground kernels, full-fat germ, and defatted-germ samples revealed that approximately 10 times more rGUS per gram of solids could be extracted from the ground full-fat germ and defatted-germ than from the kernel samples. The extraction of corn oil from ground germ with hot hexane (60 degrees C) did not affect the extractable rGUS activity. rGUS was purified from ground kernels and full-fat germ extracts by ion exchange, hydrophobic interaction, and size exclusion chromatography. Similar purity and yield of rGUS were obtained from both extracts. Biochemical properties of rGUS purified from transgenic corn seed were similar to those of E. coli GUS.

Improved adsorption to starch of a .beta.-galactosidase fusion protein containing the starch-binding domain from Aspergillus glucoamylase

Transgenic plants may prove to be one of the most economical systems for the large‐scale production of proteins and peptides. Our goal is to develop an approach for protein recovery from canola that will be adaptable to a wide variety of recombinant proteins. For recombinant protein recovery, the two downstream processes considered were extraction of target protein and purification of recombinant protein from host proteins and other impurities. In these experiments, T4 lysozyme has been added to the canola extracts as an example protein to simulate recovery of recombinant proteins while evaluating the merits of several candidate precipitation strategies to understand selectivity behavior and how it might be affected by the presence of canola components. The ability of precipitating agents, such as acid and the polyelectrolytes Glass H and poly(acrylic acid) (PAA), was investigated. Approximately 70% of extracted canola proteins were precipitated by acid addition to pH 5, leaving about 90% of T4 lysozyme still in solution. Targeting T4 lysozyme for the precipitate phase by addition of oppositely charged polyelectrolyte was not successful in the presence of canola components. Addition of 2.5 times the PAA dosage required to precipitate pure T4 lysozyme to the spiked canola extract brought down only 40% of the T4 lysozyme. For the comparable result with Glass H, a 9 times higher dosage was required.