Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Historical Perspective on Recombinant Protein Production

The initial host for producing recombinant proteins was E. coli because the technology was first developed using this bacterium. As the technology progressed, other organisms were used as alternative hosts, each with a potential advantage to produce certain types of proteins. The production of proteins is now a 50 billion dollar industry with most of the production being carried out in E. coli, yeast, or Chinese hamster ovary (CHO) cells (Andersen and Krummen 2002). Other systems continue to be developed to fill some of the unmet needs, including insect cultures and transgenic animals, because no one system appears to have solved all of the needs for every protein application.

Plants can also be used as a host, but this approach has greatly lagged behind the progress seen in other systems. This is in part because, as a higher organism, recombinant plant technology was more complicated than that required for microorganisms. Furthermore, the initial focus in plants was on crop improvement, and only a few groups experimented with plants as a host to produce recombinant proteins for industrial and pharmaceutical purposes.

Despite the fact that the earliest work of making industrial and pharmaceuticals was done in plants, the concept of using plants to make recombinant proteins for this purpose was met with much skepticism from the general scientific community. A wide range of technical questions need to be addressed such as: Could plants express microbial or animal proteins? Could plants perform posttranslational modifications on animal proteins? Could they form dimers, trimers, tetramers, etc. as in their native host? Would the proteins be stable if left on the crop until harvest? And could the proteins be readily purified from the complex plant matrix? Most of the recombinant protein production systems used today rely on microbial or cell culture systems where the culture medium and environment are highly controlled and can be manipulated at a moment’s notice. Even though the origin of pharmaceutical and industrial products had its roots in plants, making these proteins in plants was clearly a paradigm shift from using dedicated growing vessels.

These technical questions were addressed by the scientific community with a host of excellent studies in a variety of systems providing insight to this general topic (Hood and Howard 2002). While there is still a lot to learn about plant-produced proteins, the technical question of can it be done has been answered, and today there is little doubt that plants can express a wide range of proteins.

The second question is, do any of these plant systems have the prerequisites for making commercially competitive proteins? Specifically, when is it practical to produce proteins in plants for commercial applications? Going beyond the basic theory, many groups have touted an array of potential advantages for a variety of different systems to produce recombinant proteins. As there are thousands of different plant systems that can be used, the question in front of us now is which, if any, of these are useful in a commercial sense? Some general guidelines have been proposed as to when certain plant systems may be preferred that may help at a very basic level (Howard and Hood 2005), but the technology continues to develop, and each protein has its own unique set of challenges. Nevertheless, we can now definitively answer the second question as to “can it be done?” since the first plant-produced recombinant proteins (β-glucuronidase and avidin) were commercialized 15 years ago (Hood et al. 1997; Witcher et al. 1998). This success addressed a host of practical questions as to functional equivalency, storage, and purification of the proteins. This commercialization event also highlighted some key advantages including using an animal-free source, reduced cost of raw material, and reduced cost of processing. Today there are more than 30 proteins (Table 14.1) that have been commercialized. These have been for industrial and health-related applications, and the first plant-produced recombinant therapeutic protein was announced last year (http://www.protalix.com/procellex-platform/overview-procellex-platform.asp).

Table 14.1 Commercialized plant-produced recombinant proteins
Full size table

Despite the commercialization of these products, this is an extremely modest amount of activity compared to the industry at large. The success in plants represents what is better labeled a niche market rather than mainstream. So now the question is, what is holding up plants to becoming one of the mainstream systems for protein production? The theoretical advantages touted in many review articles are convincing, and the basic science and commercialization questions have been addressed, yet the industry has moved very slowly. What follows is an opinion of why the industry has moved slowly, what needs to happen for plants to play a major role, and which applications are most likely to be affected and when.

2 Why Has It Taken So Long to Develop the Plant Platform?

If plants are a better platform for expressing proteins, then why has it taken so long for this to occur while other systems dominate the industry? The first reason is that basic technology funding for plants is only a fraction of the funding that is available to microbial and human health research. Therefore, it took longer to develop the basic scientific principles needed for recombinant technology in plants. This is only a partial answer of course and was exacerbated by several other factors. As plants are higher eukaryotes, they are much more complicated than microorganisms, which included a basic and practical understanding of aspects of protein accumulation as it differs from microorganisms such as regulation of expression in different tissues and intracellular locations and developmental stage of the plant. Therefore, more funding was needed to discern even the basic biology. Progress was further complicated by the fact that plants represent a kingdom with thousands of viable options for production, each with its own peculiarities. Even if we focus on cultivated plants, we still have hundreds of choices. Finally, the vast majority of the funding that went to plants for recombinant DNA technology was focused on crop improvement. While crop improvement was needed and some of the technology did cross over, only a few small groups with limited budgets initially tried to make the case for this work, and they did not have the backing of government institutions or private industry similar to that offered to non-plant protein production systems.

Funding to create plant production technology was not the only factor slowing technology development. Fear of the unknown and resistance to change usually play a role in the acceptance of any new technology, and this is no exception. Major corporations had established research and production platforms for microbial production. No enthusiasm existed to fund work that would replace billions of dollars of investment and dedicated production facilities. The fear of upsetting the status quo could be felt by anyone trying to suggest that another way to accomplish the goal of protein production may be available.

Finally, fear in the general population has also contributed to the delay in acceptance. The debate over genetic engineering (GE) has spilled over from the use of recombinant DNA in foods to the use of plants to produce pharmaceutical and industrial proteins in plants as well. The fact that bacteria and yeast are food organisms used extensively as host for genetically engineered proteins does not seem to penetrate into the public’s debate. The fact that many of the proteins that are proposed to be made in plants are already in the food chain also does not seem to matter. In general, safety is not the focus of these debates at all but rather emotional issues. Regulatory guidelines have been instituted for decades in order to produce pharmaceutical and industrial proteins from native or recombinant hosts. This also includes keeping these segregated from the food supply no matter if this is bacteria, yeast, eggs, or plants. In addition, new sets of guidelines (enforced by the USDA in the USA) have been introduced to specifically address the production of industrial and pharmaceutical recombinant proteins in plants. While fear and bias can eventually be replaced with logic and a risk-benefit analysis, the heightened sensitivity of genetically engineered crops has delayed reason and at times put safety concerns in the background.

3 What Is Needed to Use Plants as a Major Technology for Protein Production?

Several factors need to be addressed for plant production systems to become a mainstream technology in the production of recombinant proteins. These factors include overcoming both technical and nontechnical barriers. On the nontechnical front, there must be a willingness to fund the work to create the products, and the public must accept the products when produced. While this seems obvious and in many cases not contentious, the fact that these are GE products has raised concerns. While the GE debate is most active as it relates to food safety and there are no documented scientific reports of anyone getting sick from a GE food, this has not been sufficient to dissuade the public concerns. Although there is still a strong opposition based on perceived safety issues, GE food has become part of the mainstream.

Like most any new technology, theoretical risk (fear) usually wins over theoretical benefits (promises). When the promises are transformed into real benefits, then the risks are more keenly evaluated. This is true for the GE food debate and likely to be true with any opposition that may be encountered to plant-produced proteins. One advantage for plant-produced proteins that was not as clear for food applications is that the public has a clear choice to use the product or not and can evaluate their risk and their benefit directly. Therefore, it is hoped, but not guaranteed, that plant-produced proteins will be accepted as they have from the other hosts of industrial and pharmaceutical products.

The trigger for accepting plant-produced proteins has begun by providing an animal-free source of proteins as discussed below. In this case, the perceived risk of not using plants to produce the proteins is greater than the perceived risks of using them. In other words, the fear of GE plants apparently is not as bad as the real risk of having a viral or prion contamination in a therapeutic product. This formula of establishing direct benefits to the consumer will be the key to acceptance of this technology as it relates to consumers. The use of plant-produced proteins to make animal-free products illustrates an important point, the technology will not gain acceptance if it is at par with existing systems. The new technology must have some compelling advantage to overcome the status quo.

One compelling reason is if plants can provide a dramatic cost advantage. The cost of most proteins is inversely proportional to the concentration in the host tissue. In the past, protein levels produced in plants were not high enough to translate to a dramatic cost advantage, but this began to change in recent years. Levels of recombinant proteins in some grains and tobacco leaves for selected proteins are well within the realm of other systems as it relates to cost. Still, this only works today when coupled with other advantages that can offset the cost of changing systems. To gain long-term acceptance, several technology improvements can help decrease the overall cost and are discussed below.

4 Technology Improvements Needed for Industrial Applications

4.1 Accumulation of Recombinant Proteins in Host Tissue

This is the most crucial factor in reducing cost. High expression results not only in a lower cost of raw material but a lower cost in transportation, storage, processing, and purification. In microbial systems, more than 70 % of the energy inputs can be converted to the final product. In plants when the recombinant protein reaches at least 1 % of the weight of the tissue, it is considered very good expression but obviously is not nearly 70 % of the inputs. Clearly there is still much more room for improvement in plants to reach higher levels of accumulation for any protein. Some general rules for the best ways to express specific types of proteins in plants are needed.

4.2 Product Integrity

Proteins made in plants are usually very similar if not identical to those made in other systems. There are examples, however, where plants do not possess the same type of machinery as animals and modifications can be made. One example of this is when plants do not have the required enzymes for posttranslation modifications. The case most often cited is that of the differences in glycosylation between plants and animals. While both plants and animals share the basic backbone in protein glycosylation patterns, there are subtle differences in linkages and sugar composition. Theoretical debates over the potential for allergenicity have been waged over these differences. However, one can also make a theoretical argument that as we eat plant proteins routinely, this difference does not necessarily lead to allergenicity. Testing a number of proteins over the years will demonstrate how important this difference in glycosylation is in practice, but the cases studied to date have not shown any indication that these differences lead to allergenicity.

Another example of altered glycosylation patterns is the addition of sialic acid on some select animal proteins. While there is no evidence for the presence or absence of this sugar to alter activity or lead to allergenicity, sialic acid can prolong the retention time of proteins in the blood (Morell et al. 1971). The differences in glycosylation have led to modifications of the plant’s machinery to make glycosylation similar to that present in animals (Jez et al. 2013). These changes are not always essential, but having them included in future protein design eliminates concerns and would be a valuable contribution in the long term.

Glycosylation is not the only case when plants may require other posttranslational enzymes. As an example, in addition to the gene-encoding collagen, a second gene was introduced that performed the required hydroxylation to generate the preferred form of the protein (Xu et al. 2011). This example along with the glycosylation enzymes may represent an opportunity in the future where certain plants may be engineered to possess specific posttranslational enzymes that can be used for pharmaceutical purposes.

4.3 Tissue Processing/Purification

The cost of purification usually represents greater than 80 % of the overall cost of a typical pharmaceutical protein. Much of the protein purification technology can be adapted from that used in other non-plant systems, but there are some unique aspects that need to be addressed for plant systems. In the case of the grains, many options for established processing technologies are available that can be adapted. This is in contrast to vegetative tissue such as tobacco leaves where much of the technology had to be developed for this specific purpose. It will be important to understand what can be improved upon in downstream engineering or in modification of the starting genetic material to eliminate or reduce toxic compounds or agents that may interfere in purification.

Finally, the possibility of by-products or coproducts of plant tissue can help reduce the overall costs. If only one part of the host tissue expresses the protein of interest, the rest of the plant may be available for other products. One example of this is illustrated when using corn germ to produce recombinant proteins leaving the endosperm for other uses such as the production of ethanol (Howard and Hood 2007). While this is not a critical factor for higher priced pharmaceuticals, coproducts can become critical for some industrial proteins. To be able to practice this in full will require the infrastructure for an integrated production facility and the regulatory approvals to allow uses of the other parts of the plant.

4.4 Development Time

The time required to transform plants is measured in months as opposed to hours or days as in the case of most microorganisms. It is not uncommon for many plants to have a year's delay before the first prototype product can be evaluated. While in the long run this is not a major deterrent, there is no guarantee that the protein was made correctly, and waiting a year to make the second version is a major drawback. The time it takes to evaluate these new constructs needs to be significantly reduced. In the case of the transient expression system in tobacco leaves, the development time has been reduced to weeks (Fischer et al. 1999; Scholthof et al. 1996). While still much longer than that of microbes, it is practical to work around many of the limitations to make this competitive with overall developmental times in microbes. This situation however is not the case for stable transformation systems, especially in the grains. Significant resources will eventually be required to address this problem.

4.5 Product and Crop Containment

Regulatory guidelines are in place to assure product integrity and containment when expressing proteins in any host. Regulation has been used successfully for bacteria, yeast, eggs, and a variety of other microorganisms. While plants represent a new host, the same regulatory guidelines for production, storage, transportation, and containment apply. There are some unique aspects to plants such as growing the crop outdoors rather than in a contained building, which are significantly different. For this reason, USDA has also imposed guidelines for containment to ensure no inadvertent escape into the environment, or mixing into the food or feed chain.

In addition to containment of the organisms, several new genetic strains of microorganisms have been developed that have characteristics that benefit the accumulation of recombinant proteins such as enabling higher levels of accumulation. The concept of having genetic strains that are better suited for containment and expression in microorganisms has not crossed over to plants. This is one area that would greatly benefit the commercial application of the plant system and at the same time provide additional safeguards for containment. The proposal is that specific plant germplasm would be developed that is optimized for plant-produced pharmaceutical and industrial proteins. Some of the key attributes that may be included are shown in Table 14.2. It is important to recognize that none of the items listed are essential. Conditions for containment are met now mostly by isolation requirements, and protein accumulation is adequate for many products. Nevertheless, the changes below represent the future enablement of genetic safeguards for containment that would be much less cumbersome and costly than the isolation methods used today. The increase in protein expression would allow more proteins to be commercialized at a reduced cost, benefitting the industry and the public. It is recognized that it will require significant resources over a long period of time to incorporate many of these features, but in the long term this will reduce the overall cost of production and give the public greater confidence.

Table 14.2 Key attributes of future germplasm designed for plant-produced pharmaceutical and industrial proteins
Full size table

5 What Applications Are Most Likely to Use This Technology and When?

Every protein production platform has a list of advantages as to why it is better than other platforms using competitive technologies. From a commercial perspective, all of the arguments can all be broken down to cost. In a standard definition of the unit cost of production, this entails the generation of raw material, storage, transportation processing, and purification. Other factors also can also contribute to the overall cost as well, such as the cost of research needed to make the product, the cost of meeting regulatory requirements, the product formulations, the capital cost of building new production facilities, the cost of changing over to a new system, and even the cost of public education if needed to gain acceptance of new products. What this means is that in order to launch a new platform technology, it is not enough to compete on the traditional unit cost of production alone. Additional advantages must overcome the hidden cost of these barriers particularly when replacement of existing infrastructure is required. The acceptance of this new technology relies on this broader definition of cost that is not always well defined in economic terms but clearly present in making production decisions. Therefore, when we discuss the applications below, it is always in the context of this broader overall cost.

5.1 Fine Chemicals

The first commercial plant-produced products could be considered to be in the fine chemicals market, consisting of specialized proteins used for research, diagnostics, or other specialized uses. In general, the proteins are used to make other products. The advantage here is that if the proteins are functionally equivalent, they can be substituted directly for the existing product without making any other infrastructure changes and require little regulatory oversight. Easy introduction into the market place can occur and does not involve the general public.

While the first plant-produced products were functionally equivalent to native proteins and the unit cost of production was lower than the native protein, they did not offer any other clear advantage. A convincing reason as to why someone should stop using what they were accustomed to and switch to a new product was not obvious. This type of product with a relatively low volume, small market value, and large expense of research and development (R&D) cannot be justified to make new versions of the products. This was definitely true 15 years ago and, although the technology has improved dramatically, is still the case today. Even though the production of these proteins may have a clear demonstrated lower unit cost, the cost of developing the products must come down dramatically for this to be a significant contributor in the market. When the infrastructure for plant-produced proteins becomes mature, replacement products may be possible. This application remains a niche market, however, and will not bring the technology into the mainstream.

5.2 Animal-Free Source of Proteins for Health Care (Nontherapeutics)

The first significant commercial success in using plant-produced recombinant proteins is just now being realized in the production of animal-free proteins. The technology has found a home due to the regulatory impetus in Europe and in the USA to require pharmaceutical manufacturing to use animal-free systems whenever feasible. Many of these products recently introduced are used in the manufacturing of pharmaceuticals with animal cell culture systems that account for a large portion of the current manufacturing of therapeutics. While cell culture facilities are developed in a very clean environment, they often rely on animal proteins such as growth factors and processing enzymes to produce the final therapeutic product. These required proteins are not easily manufactured in microbes because of posttranslational requirements. Obtaining them from their native source is not always practical as is the case for some human proteins or desirable as in the case of some animal proteins that bring the threat of pathogen contamination. In general, these plant-made versions offer a good alternative, even though they are priced at a premium over the native proteins. This is a clear instance when unit cost of production is not the driving factor.

The trend toward more protein therapeutics and guidelines to use animal-free sources would suggest that this market will continue to increase in the near future. As many of these proteins require posttranslational modifications, microbial production is not suitable, providing the opening for plant-produced products. This trend, however, is not limited to the use in cell culture systems as illustrated by the case for trypsin (Chap. 4). In cases where microbial production is preferred, there still may be a requirement for posttranslational processing using plant-produced proteins.

5.3 Therapeutics

Therapeutics represents the largest market by far for recombinant protein products. This is also the hardest market to enter due to the expense and time-encumbered regulatory studies. It is difficult to justify the added expenses of repeating regulatory studies and replacing expensive manufacturing facilities for the same protein product that is simply a little less expensive to produce. The hurdle that must be overcome is that the product must have a large market to justify the expense and the cost reduction must be dramatic.

Another option is to produce new products rather than the same product in a different system. In this case making the new product in the plant system at the start of the regulatory process will eliminate the hurdle of duplicating regulatory studies. The catch is that it is difficult to have a company try an untested production system for their blockbuster product. This catch 22, however, may be broken at least in part with the introduction of taliglucerase by Protalix/Pfizer in the treatment of Gaucher’s disease. This represents the first plant-produced recombinant therapeutic to reach the market. While not a blockbuster drug, taliglucerase does represent a case where a regulatory path can be outlined in detail and should help provide confidence to large companies to test other plant-produced products.

A slow acceptance of this technology will continue for therapeutics until a blockbuster product is made. At that time, the plant platform will be considered mainstream and at par with other platforms where production decisions will be made as to which system can produce the product at the lowest unit cost. In addition, with mainstream acceptance, the production of generics should be more attractive assuming the unit cost is lower in the plant system. When this occurs, the market for an animal-free source of proteins used in for the manufacturing of therapeutics in cell culture production systems may start to decline albeit very slowly.

5.4 Vaccines

One area that has received much attention is the use of plants to produce vaccines. While no products are currently on the market, several clinical trials have been conducted, and many of the practical problems appear to have been resolved as products move ahead in the pipeline toward commercialization. Two very different approaches utilize two different types of platforms.

First, in the case of pathogens that mutate quickly, such as many of the flu viruses, having a eukaryotic production system that can be employed rapidly to keep pace with the mutating virus is a clear advantage. The transient expression of vaccines in tobacco leaves can meet this need, and it seems likely that the first products are underway as they move through the required regulatory studies. A few examples are illustrated in this book, but more are likely to emerge as soon as the first of these is commercialized.

The second application is the use of plants to make oral vaccines. While conceptually this sounds like a clear advantage, difficulties in expressing the antigens at high concentrations have limited progress. However, progress has been made on this front, and this technical barrier now seems to be resolved. The acceptance of this type of product is more complicated than that of parenteral vaccines. The oral route of administration requires more studies to assure that this method will provide adequate protection since the final product is not the same as that made from other sources. In addition, many vaccines are co-administered in one injection; therefore, replacing one of the injected vaccines does not eliminate the need for the injection. While these problems can be resolved in the long term, in the short term they are barriers. The approach to using a plant-made vaccine as a booster may be the best way for this new type of product to be introduced and a way to monitor the effects of the immune system in a large population with little risk.

An oral vaccine made from plants also illustrates an important point in commercialization. The new product has benefits well beyond the simple economic cost model for making a protein at a lower cost. The plant platform impacts the way the end user receives the product, and it disrupts current dogma. Not only can the product itself be produced less expensively but also the downstream needs of administration are also greatly reduced adding enormous value. While this example illustrates the greatest potential benefit, it also has the greatest barriers to overcome.

5.5 Food and Feed Additives

The public debate over GE foods has clearly disincentivized any work in this area. Ironically, however, from a safety perspective, the choice of using an animal-free source of proteins for processing food (particularly if it is made in a food crop with generally recognized as safe (GRAS) status) is much less risky than using the same protein made from animals that may carry contaminating pathogens. Once emotional issues have receded, logic will hopefully win in the longer term. The idea of making functional foods is not new, and having these plant-produced proteins replace animal proteins should be a lower risk alternative that the public can embrace once they get past the emotional reactions to GE foods. Once this occurs, food processing is likely to become a major application for plant-produced proteins.

In addition to reducing risk, a much greater incentive to use this approach is the potential to add proteins to food at a cost that would otherwise be prohibitive. The example of brazzein (Chap. 13) illustrates the case where the protein can add value without placing a large burden on cost since purification of the protein is not necessary. Another example is the use of feed or food supplements to aid in digestion. Protease and cellulase from native sources are already used in this manner, and these proteins have been expressed in plants as well (Chaps. 4 and 12). What is missing is the will to test these plant-produced enzymes and demonstrate that these will work in the specific applications. These proteins have the potential to turn the public’s attention away from the fear of GE food and put it on the function and safety of food products. In the future, labeling of GE food may even be a selling point to highlight that unlike other forms of adding new proteins to the crop that do not involve genetic engineering, these products have been safety tested and contain functional benefits that the consumer can appreciate.

5.6 Industrial Proteins

One advantage that has been touted consistently for plant-produced proteins is that extremely large volumes can be made at a very low cost. One example that illustrates this unmet need is the requirement to produce enzymes for the conversion to bioethanol from cellulose. Chapter 12 reviews this in more detail, and it is hard to imagine how other protein production systems can compete with the volume and cost parameters required for this type of project. The first technical problem to achieve very low cost is to have very high expression levels. The next problem is to obtain activity of the enzymes in crude extracts because purification of the enzymes would be cost prohibitive. In examples where only one enzyme is needed, this example sets a good precedent for volume and cost. This specific case will also require the action of other enzymes for bioconversion of cellulose that must be made in plants or microbes to have a fully functional application. The plant system has this potential, but it also illustrates the contrast with the fungal secreted enzymes. In plants there is a systematic and concrete approach of engineering one enzyme at a time which is in contrast to using fungal isolates that may secrete 20 enzymes and have been selected for the ability to deconstruct cellulose. Each production system has its advantages, and choice of a system depends ultimately on its final application.

6 Conclusions

The commercialization of plant-produced proteins has progressed slowly over the past 15 years since the first introduction of a commercial product demonstrating feasibility. Many factors have contributed to this slow progress, but in brief, the technology was not robust and predictable in the early stages to compete strictly on a cost basis with other existing platforms, and there was little motivation to fund technology improvements to a system that was considered a threat to existing platforms. In the last several years, however, the advantages of plant production systems beyond the unit costs are enabling the acceptance of the technology. The clear front-runner is the move into an animal-free source of proteins for cell cultures. This may soon be followed by an animal-free source of therapeutics, a rapid system for the production of parenteral vaccines, orally delivered vaccines, and industrial enzymes that can only be produced on the scale that a plant system can provide. The advantages of plant-produced proteins beyond the unit cost are the key to the initial commercialization. In the longer term as the technology becomes more engrained into the industry, this approach can be used for a variety of other proteins where plants can compete on unit cost as well.