The prospects for recombinant proteins from transgenic animals

Abstract

The use of GM animals to produce human recombinant therapeutic proteins has progressed only slowly. But recent successes and new technologies could speed up development.

Mammalian cell cultures have been the gold standard for producing human recombinant proteins as therapeutics almost since the dawn of genetic engineering 30 years ago. But this technology suffers from several drawbacks that prevent cost‐efficient scaling up of production levels and requires expensive cell cultures that generate only small yields of the desired proteins. Other protein expression systems based on bacteria or yeast can be more easily scaled up and have been successfully used to produce human recombinant proteins, but their more primitive translational machinery creates other challenges 1.

The key point is that recombinant human proteins require appropriate post‐translational modifications that alter their functional properties and ensure correct folding into the final tertiary form. Such post‐translational modifications (PTMs) include reversible addition of chemical groups such as phosphate, carbohydrates in glycosylation or polypeptides in ubiquitylation. If the PTMs are substantially different from those that would have occurred in a human cell, the protein may end up being degraded by the ubiquitination pathway, or fold incorrectly, or could even have negative side effects.

Nonmammalian production systems

Bacteria, in particular E. coli, were first used to produce recombinant protein because they enable rapid and simple expression and because E. coli's genetics, genome sequence and physiology are well known. However, the cell wall of E. coli contains toxic pyrogens and the expressed proteins lack some common PTMs. Yeasts, being eukaryotes, have more extensive PTM machinery that can carry out most common translations, including acetylation, amidation, hydroxylation, methylation, N‐linked glycosylation, O‐linked glycosylation, phosphorylation, pyrrolidone carboxylic acid, sulphation and ubiquitylation, but glycosylation remains a challenge when producing recombinant proteins in yeasts. Since 60% of recombinant proteins approved for therapeutic use are glycoproteins, the adequate glycosylation pattern is critical when using yeast as host 2.

Given that 60% of recombinant proteins approved for therapeutic use are glycoproteins, the adequate glycosylation pattern is critical…

Both insects and plants have also been used for producing human recombinant proteins with some success. Insect systems became attractive after a seminal paper in 1983 reporting that baculoviruses produced massive amounts of two proteins, polyhedrin and p10, through two strong promoters and that the corresponding genes could be dispensed for virus propagation in insect cells 3. Since, baculovirus insect systems have been used to produce a number of recombinant proteins, including an influenza vaccine for human use 4.

More recently, plant‐based systems have emerged as another alternative because of their ability to cheaply produce large amounts of protein. Recombinant proteins produced by plants include phytase from Aspergillus niger, which releases phosphates from phytates, a form of phosphorus in animal feeds. Plant systems have a number of potential advantages, including better expression than microbes and lower complexity than mammalian systems. There is also minimal risk of contamination with pathogenic microbes or viruses. There are though a number of constraints and challenges, the main one being again incorrect post‐translational modifications of the recombinant protein. Plant proteins lack the terminal galactose and sialic acid residues commonly found in animals, but have α‐(1,3) fucose and β‐(1,2) xylose that animals lack. As a result, glycoproteins from plants can lead to immune reactions and alter pharmacokinetic properties.

First achievements

Indeed, it was the desire to find expression systems capable of correct post‐translational modifications of recombinant proteins that led scientists to use mammalian cell culture in the first place. At about the same time, researchers started to work on transgenic animals but their progress was slower because of technological challenges, more complicated approval mechanisms, as well as negative public acceptance given the kick back against genetically modified organisms.

But the potential for sustained production of protein under controlled conditions and harvesting it through milk, urine or eggs in the case of birds, encouraged further development. The first such recombinant protein approved for human use was antithrombin, which the European Medicines Agency (EMA) cleared in 2006 as a drug for preventing blood clotting in patients with hereditary antithrombin deficiency. (https://www.ema.europa.eu/en/medicines/human/EPAR/atryn). It was subsequently approved by the Food and Drug Administration in 2009.

… it was the desire to find expression systems capable of correct post‐translational modifications of recombinant proteins that led scientists to use mammalian cell culture in the first place.

The drug is made from the milk of goats that were genetically modified by micro‐injection of human antithrombin genes into the cell nucleus of their embryos; it is sold under the brand name ATryn by rEVO Biologics (formerly GTC Biotherapeutics) based in Massachusetts, USA. This first example demonstrated the scale potential, since a single genetically modified goat can produce as much antithrombin in a year as 90,000 blood donations. It also highlighted other important factors concerning the choice of animals: goats were selected because they reproduce more rapidly than cattle and produce more milk than rabbits or mice. They were therefore in the sweet spot between yield per animal and rate of reproduction.

The next breakthrough followed in 2011 when the EMA approved a recombinant C1‐esterase inhibitor from the milk of transgenic rabbits for the treatment of hereditary angiooedema. Sufferers from this condition have a deficiency in this inhibitor protein, which normally modulates inflammatory responses. This results in spontaneous and inappropriate inflammation that can become life‐threatening when it affects the respiratory system.

Chickens and eggs

This second approval of a therapeutic product produced by transgenic animals spawned a number of biotech companies, including PPL Therapeutics in the UK, Hematech and Genzyme in the United States, Nexia Biotechnologies in Canada and Pharming in the Netherlands. It also galvanized research at academic institutions, such as the University of Edinburgh in the UK whose Roslin Innovation Centre recently published results promoting chicken eggs as a production system for human proteins 5. They have so far produced two proteins with therapeutic potential: IFNalpha2a, which has antiviral and anticancer effects, and macrophage‐CSF, which can help stimulate repair of damaged tissues.

The Edinburgh group chose to work on chickens because they offer several advantages over mammalian systems, despite the greater genetic distance from humans. These relate primarily to cost, scale and purity of the proteins produced, according to Helen Sang, Head of Division of Developmental Biology at The Roslin Institute. “Egg white is a much simpler material from which to purify proteins as it is made up of a relatively small number of proteins and has no fat, unlike milk,” she explained. “The generation time of chickens is about 6 months compared to around 15 months for cows. Then a single male can be bred with a very large number of hens to produce many transgenic offspring over a short period. These two attributes mean that a flock of transgenic hens laying eggs with recombinant protein can be built up very quickly. Conversely, if a protein is very active and only a small amount is required it can be produced by rearing a small number of hens, so it is a flexible system.” To give an indication for one of the recombinant proteins, three eggs would yield a single human dose of the resulting drug so that one chicken would produce 100 a year. This could then scale up to the production of a popular therapeutic, given say a battery of 100,000 chickens, which equates to about 10 average broiler farms in the United States.

However as Lissa Herron, Head of the Avian Biopharming Business Unit at Roslin Technologies, pointed out, it will take years before any drugs derived from this pipeline will be approved for human use. “As with any drug, we need to meet the regulatory requirements, which would probably take 5 plus years to get to the point where our manufacturing is up to standard, and if we're developing a novel drug, an additional 5–10 years for clinical trials,” she said. There is though some prospect of drugs being approved for animal use sooner than that. “There are absolutely stringent regulatory hoops for animal health products, but there are some faster routes to market that aren't available in human health,” Herron commented. “There is substantial interest in protein therapeutics from the companion animal market, for things like cancer and autoimmune disorders, while for agriculture, we really need to have a low price and easy drug delivery. We hope the chicken system can deliver that.”

Egg white is a much simpler material from which to purify proteins as it is made up of a relatively small number of proteins and has no fat, unlike milk.

Although these human therapeutics from eggs may be a decade or more away, one of the few other recombinant drugs from transgenic animals to gain marketing authorization does come from chicken eggs. Both the FDA and EMA in 2015 approved Sebelipase alfa, a recombinant form of the enzyme lysosomal acid lipase (LAL) under the tradename Kanuma for treatment of people with lysosomal acid lipase deficiency (LAL‐D). It is an orphan drug that would not be profitable without government assistance, as the incidence of LAL‐D varies between 1 in 40,000 and 1 in 300,000 (https://ghr.nlm.nih.gov/condition/lysosomal-acid-lipase-deficiency#statistics).

LAL‐D is caused by deleterious mutations in the LIPA gene—which codes for the lysosomal acid lipase—and results in an inability to break down certain fats and a toxic build‐up of fatty substances in cells and tissues. The most severe form is early‐onset Wolman disease whereby lipids accumulate throughout the body, mostly in the liver, during the first weeks of life. Symptoms include an enlarged liver and spleen, low weight gain, vomiting, diarrhoea and poor absorption of nutrients from food. The less severe cholesteryl ester storage disease starts later in life with symptoms again associated with malabsorption, such as diarrhoea and vomiting, still with risk of cirrhosis developing. Sebelipase alfa was shown to alleviate these symptoms in a phase‐3 trial 6. The FDA approval in fact came in two parts. The Centre for Drug Evaluation and Research (CDER) approved the human therapeutic application of the medication, while the Centre for Veterinary Medicine (CVM) approved the recombinant DNA construct in genetically engineered chicken.

Technological advances

While human therapeutics from transgenic animals is still a small market, gene‐editing technology holds great promise to accelerate development and approvals, even if it will not necessarily lead to a plethora of new therapeutics given the need for clinical trials to gain approval. Yehuda Ben‐Shahar, who studies the genetics and evolution of behaviour in Drosophila at Washington University in St. Louis, noted that techniques such as CRISPR are not game changers by themselves, but they help to reduce the cost and increase the speed at which research can be done. “Homologous recombination has been used for a long time,” he said. “All CRISPR does is help the efficiency of the process by enabling nicks to be made in DNA in a targeted place.”

In addition, various other techniques, especially insertion of artificial noncoding introns within and adjacent to genes, can help to boost production of the target protein 7. “When making transgenic animals producing recombinant proteins we have a choice between cDNA (complementary DNA), the full‐length gene and a mini gene to use in a genetic construct,” said Mikhail Shepelev from the Institute of Gene Biology, Russian Academy of Sciences, Laboratory of Molecular Oncogenetics in Moscow. Such a “mini” gene has had the long introns of the original gene replaced with short sequences of around 100 single nucleotide units. The theory is that retention of these smaller introns elevates production of the target protein in various ways, while the resulting mini gene is still a lot more compact than the full‐length version. This exploits the phenomenon known as “intron‐mediated enhancement,” where transcription rates can be increased, protein export accelerated or efficiency of mRNA translation boosted 8.

…various other techniques, especially insertion of artificial noncoding introns within and adjacent to genes, can help to boost production of the target protein

However, this approach does have a cost. “For every new gene or protein you will have to design the minigene, analyse sequences, find appropriate endonuclease restriction sites and generate multiple plasmids when constructing the minigene,” Shepelev explained. “This is time and effort consuming.” His team and others have therefore been developing artificial introns that can be standardized and then inserted in a genetic construct after removal of all the original introns, to boost production of the protein. “We can simply put several artificial introns into a genetic construct in between endogenous exons thus increasing the expression level of the transgene compared to cDNA without introns,” Shepelev said.

There is also further potential for synthetic biology to design larger constructs or whole metabolic pathways for precision manufacturing of a recombinant protein in a target mammalian. “This is not happening yet, because getting large pieces of synthetic DNA into animal cells is a challenge, but it is one we are thinking about,” Roslin's Sang said. As more advanced gene‐editing techniques become available, along with growing experience from synthetic biology and a better understanding of how to manipulate and control post‐translational modification, the scientific and economic potential of mammalian animal production systems improves. It also helps to create a healthy competition between different production systems, especially yeast, insect, plant and mammalian ones.

EMBO Reports (2019) 20: e48757