Plant Cell Reports

, Volume 24, Issue 11, pp 629–641

Microalgae as bioreactors

Authors

  • Tara L. Walker
    • Cluster for Molecular Biotechnology, Science Research Centre and CRC for DiagnosticsQueensland University of Technology
  • Saul Purton
    • Department of BiologyUniversity College London
  • Douglas K. Becker
    • Cluster for Molecular Biotechnology, Science Research Centre and CRC for DiagnosticsQueensland University of Technology
    • Cluster for Molecular Biotechnology, Science Research Centre and CRC for DiagnosticsQueensland University of Technology
Review

DOI: 10.1007/s00299-005-0004-6

Cite this article as:
Walker, T.L., Purton, S., Becker, D.K. et al. Plant Cell Rep (2005) 24: 629. doi:10.1007/s00299-005-0004-6

Abstract

Microalgae already serve as a major natural source of valuable macromolecules including carotenoids, long-chain polyunsaturated fatty acids and phycocolloids. As photoautotrophs, their simple growth requirements make these primitive plants potentially attractive bioreactor systems for the production of high-value heterologous proteins. The difficulty of producing stable transformants has meant that the field of transgenic microalgae is still in its infancy. Nonetheless, several species can now be routinely transformed and algal biotechnology companies have begun to explore the possibilities of synthesizing recombinant therapeutic proteins in microalgae and the engineering of metabolic pathways to produce increased levels of desirable compounds. In this review, we compare the current commercially viable bioreactor systems, outline recent progress in microalgal biotechnology and transformation, and discuss the potential of microalgae as bioreactors for the production of heterologous proteins.

Introduction

As a consequence of increasing global demand for recombinant proteins in a variety of industrial, diagnostic and therapeutic applications, bioreactor systems are becoming more important for the production of large quantities of proteins, particularly in cases where traditional sources are limited due to cost and/or availability. While recent advances in recombinant DNA technology have seen the development of a number of novel bioreactor systems based on a diverse variety of living cells and organisms, there is no standard system for the efficient production of proteins on a commercial scale. Choice of bioreactor system for the production of a particular protein depends on development and production costs, protein integrity, purity and the level of expression in that system. The ever-growing demand for human recombinant proteins, as well as the increasing need for industrial and diagnostic proteins has been a major driving force in the development of new, safe and large-scale, platforms for protein production.

The most well-established bioreactor systems of bacterial and yeast fermentation and mammalian cell culture are currently used for the production of the majority of commercially available heterologous proteins (Lin Cereghino and Cregg 1999; Swartz 2001; Schutt et al. 1997). While bacterial and yeast fermentation has been used for centuries as a source of natural products, fermentation-based bioreactors have often been the expression system of choice for the production of small, simple recombinant proteins. Rapid growth and production rates combined with the immense physiological knowledge and availability of advanced genetic tools make bacteria and yeasts extremely powerful and versatile expression systems. Fermentation systems are easily developed and the production costs are low once initial outlay for fermentation and purification equipment is covered. There are, however, limitations with the established bioreactor systems. Bacterial fermentation is limited in application as bacteria are unable to perform post-transcriptional and post-translational modifications essential for the production of functional eukaryotic proteins. Such modifications include intron-splicing, glycosylation and multimeric protein assembly (Lin Cereghino and Cregg 1999; Swartz 2001). An additional complication is that high intracellular levels of heterologous proteins tend to result in the formation of protein aggregates as insoluble inclusion bodies. Furthermore, the presence of bacterial endotoxins and proteases makes purification difficult and can lead to adverse effects in humans. Although yeasts have the ability to post-translationally modify proteins, the pattern of glycosylation does not mirror that of higher organisms and usually involves hyperglycosylation (Fischer et al. 1999).

Bioreactors based on mammalian cell lines are currently the system of choice for the production of proteins for therapeutic use and 21 different diagnostic and therapeutic proteins have been approved in the USA (Kelley 2001; Chu and Robinson 2001). Mammalian bioreactors, whether cell culture-based or transgenic animals, have a number of disadvantages. Mammalian cell lines are difficult, and thus expensive, to culture in large volumes due to instability, poor oxygen supply, waste product accumulation and sensitivity to shear forces associated with agitation (Janne et al. 1998; Giddings et al. 2000; Chu and Robinson 2001). Low expression of heterologous proteins and the need for expensive media add further to costs. Product purity is paramount in proteins destined for therapeutic use and high-level purification of heterologous proteins in low quantities from supernatant is expensive. The use of transgenic animals has gained attention in recent years as a potential system for the high-level production of complex proteins that can be isolated from the milk of lactating animals (Echelard 1996). The establishment of transgenic animals as bioreactors is extremely expensive at US$500,000 per animal, labour-intensive and time-consuming during the scale-up to production size herds (Wall et al. 1992; Larrick and Thomas 2001). The major disadvantage of mammalian bioreactor systems is the issue of contamination of the isolated protein with potentially harmful disease-causing agents or oncogenic sequences (Scott et al. 1999).

Transgenic plants offer a number of advantages over microbial- or animal-based expression systems. Plants can be grown on an agricultural scale to produce high amounts of safe, functional recombinant proteins. Production can be scaled up to meet demand and require relatively low input compared to other systems (Mason and Arntzen 1995). For plant-based bioreactors to be commercially viable, expression levels of 5% of total soluble protein are usually required. Expression levels have been reported in transgenic plants ranging from 0.001% (Mason et al. 1992) to 46.1% (De Cosa et al. 2001) of total soluble protein. Plants have been shown to be capable of producing functional vaccines (Mason et al. 1992; Streatfield et al. 2003), antibodies (Hiatt et al. 1989; Ramirez et al. 2002), enzymes (Mor et al. 2001), hormones (Ganz et al. 1996), and a variety of other therapeutic proteins (Borisjuk et al. 1999). Importantly, plants are not hosts to major human pathogens that can contaminate bacterial and mammalian systems (Herbers and Sonnewald 1999). There is however at least one pathogen, Pseudomonas aeruginosa, that can cause disease in both plants and animals (Rhame et al. 1997), and some toxin-producing bacteria and fungi occur in plants. To date, plant viruses are not known to infect humans or animals, nor are potentially pathogenic human and animal viruses capable of replicating in plant cells (Miele 1997). There are, however, potential concerns that must be addressed. These include: containment of genetically-modified plants in the environment; allergic reactions to plant protein glycans and other plant antigens; product contamination by mycotoxins, pesticides, herbicides and endogenous plant secondary metabolites; and regulatory uncertainty for proteins requiring approval for human drug use.

Pharmaceutical proteins often require complex post-translational processing and folding for bioactivity. Plants have similar pathways of protein synthesis, secretion, folding and post-translational modification to animal cells (Cabanes-Macheteau et al. 1999). The majority of human therapeutic proteins produced in plants show significantly similar structural, biochemical and functional properties to proteins in humans and animal cell cultures (Sijmons et al. 1990; Cramer et al. 1999), although differences in protein processing, particularly glycosylation, do exist between plants and animals (Cabanes-Macheteau et al. 1999). Alterations to the glycosylation pattern may not specifically alter the activity of a protein, however other properties such as folding, stability, solubility, susceptibility to proteases, blood clearance rate and antigenicity, can be affected (Jefferis and Lund 1997).

Plant N-linked glycans do not contain the terminal sialic acid residues or mannose-6-phosphate found in mammalian glycoproteins. Sialic acid is present as the terminal sugar on many human serum glycoproteins and appears to function in their serum longevity and rates of clearance (Grinnell et al. 1991). The incorporation of this charged sugar residue into protein glycans has not been demonstrated in plants. Some carbohydrate moieties are unique to plants and may present an antigenic challenge to the immune system when administered regularly, resulting in sensitisation or immunisation (Ma and Hiatt 1996). The xylosyl and fucosyl residues on plant N-linked complex glycans have been demonstrated to be the key epitopes responsible for the allergenicity of plant glycoproteins in humans (Garcia-Casado et al. 1996). Glycosylation differences may only be a problem for therapeutic glycoproteins administered to patients by injection (Ma et al. 1998). This is consistent with our repeated exposure to plant oligosaccharides in foods, and the lack of novelty of plant glycans to the mucosal system. Metabolic engineering of the plant N-glycosylation pathway is probably the most effective way to negate the structural differences that exist between plant and mammalian N-glycans. Strategies currently being investigated include: retention of the protein in the endoplasmic reticulum (Gomord et al. 1997), isolation of plant mutants that are affected in the N-glycan biosynthesis pathway (von Schaewen et al. 1993; Rayon et al. 1999), knocking out specific Golgi enzymes by antisense technology (Wenderoth and von Schaewen 2000), and complementation with human glycotransferases (Wee et al. 1998).

Eukaryotic microalgae may provide an attractive alternative to the bioreactor systems currently in use. The term “microalgae” is used to describe a diverse photosynthetic group of prokaryotic (cyanobacteria) and eukaryotic organisms. Although the blue-green algae Spirulina serves as a commercial source of nutritional supplements marketed by Cyanotech, the focus of this paper is the eukaryotic green algae such as Chlamydomonas reinhardtii and Dunaliella salina, and the diatom Phaeodactylum tricornutum which have been used for many years as model systems for basic biological and ecological studies. Consequently, over the last 10 years, there has been significant progress in the development of genetic engineering technology for several green algal and diatom species. Algal research groups and biotechnology companies are now beginning to apply this technology to modify key metabolic pathways or synthesize high value recombinant proteins. Recently, there have been a number of reports on the production of antibodies, insecticidal proteins and vaccines in microalgae (e.g. Borovsky 2003; Geng et al. 2003; Mayfield et al. 2003; Sun et al. 2003; Banicki 2004).

Microalgal biotechnology

Potentially, microalgae provide all the benefits of plants coupled with high productivities associated with large-scale microbial production approaches. Microalgae have the same basic photosynthetic mechanism as higher plants. As they are of simple structure, being unicellular, filamentous or colonial, energy is directed into photosynthesis, growth and reproduction rather than maintaining differentiated structures. Consequently, the levels of protein in microalgae can be between 30–50% of the dry weight biomass (Benemann and Oswald 1996; Brown 2002). Furthermore, given that microalgae are microscopic in size and grown in liquid culture, nutrients can be maintained at or near optimal conditions potentially providing the benefits of high levels of controlled, continuous productivity similar to microbial fermentation.

While the use of microalgae by human populations is centuries old, efforts to mass culture microalgae began shortly after World War II in the USA, Germany and Japan as a potential source of food in a world experiencing a population explosion. Since then, mass culturing of microalgal species have been variously explored in the treatment of wastewater and control of water pollution, for atmosphere regeneration in biospheres (ie, spacecraft), as renewable fuels for transportation (biodiesel), as a source of high-value natural health products (nutriceuticals) and lately in the mitigation of greenhouse gases and the production of hydrogen as a fuel source (Benemann and Oswald 1996).

From a commercial perspective microalgae have been used as a natural source of high-value compounds for use in health and medical applications since the 1960s (Table 2).
Table 1

Comparison of the features of recombinant protein production for the various bioreactor systems currently available (modified from Fischer et al. 1999)

Features

Bacteria

Yeast

Mammalian cell culture

Transgenic animals

Transgenic plants

Transgenic microalgae

Production time

Short

Medium

Long

Long

Long

Short

Production cost

Medium

Medium

High

High

Low

Very low

Scale up cost

Higha

Higha

Higha

Higha

Low

Very low

Cost/storage

Cheap/-20°c

Cheap/-20°c

Expensive LN

Expensive LN

Cheap RT

Cheap RT

Production scale

Limited

Limited

Limited

Limited

Worldwide

Worldwide

Propagation

Easy

Easy

Hard

Feasible

Easy

Very easy

Distribution

Feasible

Feasible

Difficult

Difficult

Easy

Easy

Delivery vehicle

No

No

No

Yes

Possible

Possible

Gene size

Unknown

Unknown

Limited

Limited

Not limited

Not limited

Glycosylation

Absent

Incorrect

‘Correct’

‘Correct’

‘Correct’

‘Correct’

Multimeric protein assembly

No

No

No

Yes

Yes

Yes

Protein yield

Medium

High

Medium-high

High

High

b

Riskc

Yes

Unknown

Yes

Yes

Unknown

Unknown

Safety

Low

b

Medium

High

High

High

Ethical concerns

Low

Medium

Medium

High

Medium

Medium

aIn terms of large scale fermenters and associated equipment

bUnknown

cIn relation to contamination of human therapeutic proteins with viral sequences, oncogenes or endotoxins RT—Room temperature; LN—Liquid nitrogen

Table 2

Companies currently selling microalgal-based products or developing microalgae as bioreactors for protein production

Company

Microalgae

Products

Cyanotech (www.cyanotech.com)

Spirulina pacifica

Spirulina extracts as nutritional supplements, immunological diagnostics, aquaculture feed/pigments and food colouring

Martek Biosciences Corporation (www.martekbio.com)

Crypthecodinium cohnii

Nutritional fatty acids

Mera Pharmaceuticals (www.aquasearch.com)

Haematococcus pluvialis

Natural astaxanthin as a nutraceutical

Earthrise Nutritionals (www.earthrise.com)

Spirulina sp

Nutritional supplement to inhibit replication and infectivity of viruses including HIV, CMV, HSV and influenza A

PharmaMar (www.pharmamar.com)

Various

Anticancer drugs derived from marine microorganisms

Nikken Sohonsha Corporation (www.chlostanin.co.jp)

Chlorella sp. Dunaliella sp

Dietary supplements: polysaccharide N,β-1.3 glucan (Chlorella) and β-carotene (Dunaliella)

Nature Beta Technologies

Dunaliella bardowil

β-carotene powder

Cognis (www.cognis.com)

Dunaliella salina

Mixed carotenoids

Subitec GmbH (www.subitec.com)

Undisclosed

Polyunsaturated fatty acids

Solazyme (www.solazyme.com)

Undisclosed

No products yet brought to market

Phycotransgenics (www.phycotransgenics.com)

Chlamydomonas reinhardtii

Transgenic microalgae for animal health/feed, bioremediation, environmental monitoring and biopesticides

Algal Biotechnology (www.algalbiotechnology.com)

Chlamydomonas reinhardtii

Developing recombinant protein technology

In recent years, there has been a surge of interest in microalgal metabolites as a source of novel types of structures including potential drugs not found in the higher plants (Shimizu 1996; Cohen 1999). Species of Dunaliella, Haematococcus, Chlorella and Euglena produce carotenoids such as β-carotene, astaxanthin and canthaxanthin. These are used as pigments in food products and cosmetics, as vitamin A supplements and health food products, and as feed additives for poultry, livestock, fish and crustaceans (Borowitzka 1988; Lorenz and Cysewski 2000). For example, over 80% of the world's supply of natural β-carotene comes from the halophilic green alga Dunaliella salina harvested from large saline ponds of over several hundred hectares in Australia (Curtain 2000). Natural β-carotene contains both trans and cis isomers and sells for twice the price of the synthetic molecule which contains only the trans isomer. Carotenoids derived from natural sources are, however, beginning to command a premium price in a society concerned with the “organic” credentials of foodstuffs. For example, farmed salmon fed synthetic astaxanthin derived from petrochemical by-products have been implicated in an allergic reaction in humans. In contrast, clinical trials have shown natural astaxanthin, marketed as BioAstin (Cyanotech), acts as an anti-inflammatory in instances of arthritis, muscle pain and carpal tunnel syndrome. With the increasing cost of crude oil and significant advances in mass culture and harvesting techniques for microalgae leading to cost reductions, the natural products are beginning to compete favourably with the synthetic compounds. Microalgae could potentially be a future commercial source of a number of other vitamins including vitamin C, E and B12 (Borowitzka 1988).

Numerous species from various algal groups including diatoms and dinoflagellates are a rich source of long-chain polyunsaturated fatty acids (LCPUFA), especially the omega-3 polyunsaturated fatty acids, eicosapentaenoic, docosahexaenoic and arachidonic acid that are of value as nutritional supplements in humans and animals. Several companies currently market algal-derived LCPUFA preparations as healthcare products or as a source of LCPUFA in aquaculture feeds (Apt and Behrens 1999). There is also increasing commercial interest in the wide range of secondary metabolites present in algae that may have medicinal applications. A large number of pharmacologically active compounds have been reported in microalgae, including anticancer, antimicrobial, antiviral, and various neurological activities (Schwartz et al. 1990; Cannell 1993; Codd 1995; Moore 1996; Borowitzka 1999; Ördög et al. 2004) although no pharmacologically novel compounds have yet achieved clinical use. It should be noted, however, that while many compounds may hold promising potential in vitro, results from clinical studies may disprove the hype. For example, such trials have shown some microalgal-derived antibiotics are either inactive or toxic in vivo (Borowitzka 1999).

The ease of culture and harvesting of products from microalgal culture has a number of advantages from an industry perspective that have, in turn, led to an increased commercial interest in the biotechnology of this group of organisms. Most microalgae are photoautotrophs, requiring only sun, water and basic nutrients for maximal growth. The lack of a requirement for an exogenous carbon source for energy potentially makes the large-scale culture of microalgae in open ponds comparatively cheap (Apt and Behrens 1999). Microalgae such as Dunaliella and Chlorella grow in saline to hyper-saline waters and thus their large-scale culture does not compete with conventional agriculture for the limited resources of arable land and fresh water. Cultures can achieve high cell densities under conditions of high light and aeration, and be grown to volumes of several megalitres. High biomass yields can also be achieved in fermenter-based systems or photobioreactors with vigorous agitation and light. Some microalgae can also be grown in dark fermenters as heterotrophs, requiring a supply of sugars for energy and as a carbon source, where the high levels of biomass seen in open or closed photobioreactor systems can be achieved in lower volumes. Microalgae such as Dunaliella, Haematococcus and Chlorella do not produce toxins and are classified as food sources falling into the GRAS (Generally Regarded As Safe) category. Consequently, many high-value compounds produced by microalgae are administered as a powder of freeze-dried algae with no extraction undertaken. Since microalgae are simple in structure concerns about processing potential bioproducts out of complex tissues are obviated and the possibility exists through genetic engineering that recombinant proteins can be secreted directly into the culture medium in closed photobioreactors or targeted for sequestering into the periplasmic space.

The vision of harvesting high-value products from mass cultures of microalgae cheaply will, however, be difficult to realise. The few commercial species that are cultured in large ponds are extremophiles growing in a highly selective environment (high salt and/or high temperature) which precludes the growth of most other algae or bacteria. Species that grow in low salt and moderate temperature, such as Chlorella, are often overrun by other species in open pond culture systems and require large starter inoculums grown in closed photobioreactor systems. Developing techniques to select and maintain strains of commercial interest recalcitrant to growth in open pond systems is of high priority for the industry. Even in open ponds, however, the supply of CO2, O2 and sunlight beyond the first few centimetres depth becomes limiting and this restricts cell growth (Apt and Behrens 1999). Surface area, not depth, is the critical parameter contributing to cell concentration and thus productivity. The need for turbulent flow to remove O2 build up, maintain access to sunlight and nutrient exchange adds to the complexity and cost of mass culture and may not be suitable for fragile species. Nonetheless, the potential of high levels of biomass production, equivalent to the highest found in land-based plants (Benemann and Oswald 1996), maintain microalgae mass culture as an attractive option. In an industry in its infancy, there is considerable scope for improvement and innovation.

The potential for large-scale culture in conditions unsuitable for conventional crops makes microalgae a desirable target for both increased production of natural compounds by metabolic engineering and for exploitation as biological factories for the synthesis of novel high-value compounds. The establishment of routine methods for genetic manipulation of algae is still embryonic and to-date no algal recombinant products have been brought to market. Nevertheless, recent progress in the development of efficient genetic transformation procedures for Chlamydomonas reinhardtii and several other model species (Stevens and Purton 1997; Falciatore and Bowler 2002) indicate that molecular engineering of commercially important microalgae may soon be realised.

Microalgal transformation

The development of methods for algal transformation has advanced significantly in the last 10 years. Much of the progress has been due to the pioneering work on the transformation of the green alga Chlamydomonas reinhardtii (Lumbreras and Purton 1998; Fuhrmann 2002). The development of the molecular tools and the methodology for introducing novel genes into either the nuclear or chloroplast genome of C. reinhardtii has acted as a catalyst for similar developments in other algal species. Although routine transformation is currently achievable for only a handful of species including C. reinhardtii, Volvox carteri, several species of Chlorella and the diatom Phaeodactylum tricornutum; there have been initial reports of transformation success in a wide range of species as outlined in Table 3.
Table 3

Species of microalgae which have been transformation and the method used

Microalgae

Transformation method

Genome

Reference

Green microalgae

C. reinhardtii

Microprojectile bombardment

Nuclear

Debuchy et al. 1989

  

Chloroplast

Boynton et al. 1988

  

Mitochondrial

Randolph-Anderson et al. 1993

 

Electroporation

Nuclear

Brown et al. 1991

 

Glass-bead

Nuclear

Kindle 1990

 

Silicon carbide whiskers

Nuclear

Dunahay 1993

 

Agrobacterium tumifaciens

Nuclear

Kumar et al. 2004

Dunaliella salina

Electroporation

Nuclear

Geng et al. 2003

Chlorella ellipsoida

Polyethylene glycol

Nuclear

Jarvis and Brown 1991

Chlorella sorokiniana

Microprojectile bombardment

Nuclear

Dawson et al. 1997

Chlorella vulgaris

Electroporation

Nuclear

Chow and Tung 1999

Chlorella kessleri

Microprojectile bombardment

Nuclear

El Sheekh 1999

Haematococcus pluvialis

Microprojectile bombardment

Nuclear

Teng et al. 2002

Volvox carteri

Microprojectile bombardment

Nuclear

Schiedlmeier et al. 1994

Diatoms

Cyclotella cryptica

Microprojectile bombardment

Nuclear

Dunahay et al. 1995

Navicula saprophila

Microprojectile bombardment

Nuclear

Dunahay et al. 1995

Phaeodactylum tricornutum

Microprojectile bombardment

Nuclear

Apt et al. 1996

Cylindrotheca fusiformis

Microprojectile bombardment

Nuclear

Fischer et al. 1999

Dinoflagellates

Amphidinium spp

Silicon carbide whiskers

Nuclear

ten Lohuis and Miller 1998

Symbiodinium microadriaticum

Silicon carbide whiskers

Nuclear

ten Lohuis and Miller 1998

Red Algae

Cyanidoschyzon merolae

Electroporation

Nucleus

Minoda et al. 2004

Porphyidium spp.

Microprojectile bombardment

Chloroplast

Lapidot et al. 2002

Euglenoids

Euglena gracilis

Microprojectile bombardment

Chloroplast

Doetsch et al. 2001

The delivery of DNA into the nuclear genome of C. reinhardtii can be achieved using various techniques. The simplest method is to agitate a cell suspension in the presence of the DNA and glass beads. The abrasive action of the beads creates transient holes in the plasma membrane of the cells, and is most efficient if the proteinacious cell wall is removed by enzymatic treatment or cell wall deficient mutants are used (Kindle 1990). Other delivery methods are bombardment with DNA-coated microprojectiles (biolistics) (Debuchy et al. 1989), electroporation (Brown et al. 1991; Shimogawara et al. 1998), agitation with silicon-carbide whiskers (Dunahay 1993) and most recently Agrobacterium-mediated transfer (Kumar et al. 2004). For transformation of the chloroplast or mitochondrial genomes of C. reinhardtii, microprojectile bombardment is the method of choice (Boynton et al. 1988; Randolph-Anderson et al. 1993). For most other microalgal species, microprojectile bombardment has been used. These include Volvox carteri (Schiedlmeier et al. 1994), various Chlorella species (Jarvis and Brown 1991; Dawson et al. 1997; Kang et al. 2000), and the diatoms Phaeodactylum tricornutum, Cyclotella cryptica and Navicula saprophila (Dunahay et al. 1995; Apt et al. 1996; Zaslavskaia et al. 2001).

Selectable markers

Early selectable markers for Chlamydomonas nuclear transformation were cloned Chlamydomonas genes that were able to rescue auxotrophic mutants with mutations in the corresponding endogenous gene. These markers include NIA1 (formerly NIT1), which encodes nitrate reductase and rescues nit1 mutants to growth on nitrate as the sole nitrogen source (Fernandez et al. 1989; Kindle et al. 1989) and arg7, which encodes argininosuccinate lyase and rescues arg7 mutants to arginine-independent growth (Debuchy et al. 1989). A mutated version (CRY1) of the RPS14 gene encoding ribosomal protein S14, was the first dominant selectable marker allowing direct transformation of wild-type Chlamydomonas and conferring resistance to the translation inhibitors emetine and cryptopleurine (Nelson et al. 1994). Recently, transformation of C. reinhardtii with a mutated acetolactate synthase (als) or protoporphyrinogen oxidase (PPX1) gene has allowed recovery of colonies resistant to the herbicides sulfometuron methyl and S-23142 respectively (Kovar et al. 2002; Randolph-Anderson et al. 1998). Dominant markers based on bacterial antibiotic-resistance genes have also been developed for Chlamydomonas. These include the Sh.ble gene that confers resistance to the bleomycin family of antibiotics (Stevens et al. 1996), the aadA gene, that confers resistance to spectinomycin (Cerutti et al. 1997b), and the aphVIII gene conferring resistance to paromomycin (Sizova et al. 2001). For chloroplast transformation, two bacterial genes have been developed as dominant markers: aadA and the kanamycin-resistance marker aphA6 (Goldschmidt-Clermont 1991; Bateman and Purton 2000). The successful development of foreign genes as markers for C. reinhardtii has been due in part to the use of genes that show a similar codon bias to C. reinhardtii genes: namely a GC-bias for nuclear genes and an AT-bias for chloroplast genes (Stevens et al. 1996; Bateman and Purton 2000). As an alternative strategy, synthetic ‘codon-optimised’ versions of the foreign gene have been created for expression in the nucleus (Fuhrmann et al. 1999) or the chloroplast (Franklin et al. 2002).

The choice of selectable markers has proved more restrictive for other algal species. The use of recessive markers is not possible for algae in which the vegetative cell is diploid (e.g. diatoms), unlike the haploid C. reinhardtii. Furthermore, natural resistance of many microalgae to most antibiotics and herbicides commonly used for selection (Apt et al. 1996) has limited the choice of dominant markers. Finally, the high salt conditions required for the growth of many marine algae often reduces the activities of antibiotics (Allnutt et al. 2000). One marker that has proved particularly suitable for most algae is the ble gene encoding phleomycin resistance (Stevens and Purton 1997). Several properties of the phleomycin family may serve to make these antibiotics the preferred selection agent for transformation of algal species. Antibiotics of this family act in a non-specific manner by inducing damage to DNA within the cell and therefore have broad range toxicity. The natural resistance to many antibiotics and herbicides used in the transformation of higher plants is, thus, not witnessed with these antibiotics. Furthermore, the phleomycin family of antibiotics retain toxicity in high salt media. The small size of ble and a codon bias, which parallels that of Chlamydomonas, and other GC-rich algal species, add to the attractiveness of this selection system

Promoters

Also important is the choice of promoter to drive heterologous gene expression. Although the cauliflower mosaic virus 35S (CaMV 35S) promoter drives strong and constitutive expression in most dicotyledonous and some monocotyledonous plants (Benfey et al. 1990), it has not been shown to be a useful promoter in most algal species. Rather, in algal transformation, the most effective promoters have been derived from highly expressed algal genes. A widely used promoter for Chlamydomonas transformation is derived from the 5′ untranslated region of the C. reinhardtii ribulose bisphosphate carboxylase/oxygenase small subunit (RbcS2) (Stevens et al. 1996). It was also shown that transformation frequency was significantly increased when Chlamydomonas introns (particularly the first intron of RbcS2) were introduced into the coding region of the ble selectable marker gene (Lumbreras et al. 1998). This intron appears to contain a transcriptional enhancer element as it can act in an orientation-independent manner and is effective when placed either upstream or downstream of the promoter. Synthetic promoters have also been developed by fusing the promoter from the Chlamydomonas Hsp70A (heat shock protein 70A) gene to other Chlamydomonas promoters (Schroda et al. 2000). The Hsp70A promoter serves as a transcriptional enhancer of promoters RbcS2, β2-tubulin and Hsp70B leading to high-level expression under inducing conditions (Schroda et al. 2000). Successful transformation of the diatom Phaeodactylum tricornutum has also relied on the use of highly expressed endogenous promoters. In this case, the promoters from the fcp genes encoding the fucoxanthin chlorophyll-binding proteins have been used to drive expression of various dominant markers and reporter genes (Zaslavskaia et al. 2000).

Reporter genes

Several different reporter genes have been developed for both nuclear and chloroplast gene expression studies in C. reinhardtii. For nuclear expression, a popular reporter gene is the endogenous ARS gene, which encodes the periplasmic enzyme arylsulfatase. Endogenous ARS activity is normally undetectable since is expressed only under conditions of sulphur starvation. The arylsulphatase activity resulting from expression of promoter-ARS constructs in transgenic Chlamydomonas can be detected in colonies by spraying plates with the chromogenic substrate 5-bromo-4-chloro-3-indolyl sulphate (X-SO4) or assayed in solution with p-nitrophenyl sulphate or α-naphthylsulphate (Davies et al. 1992). More recently, two codon-optimised genes encoding green fluorescent protein (GFP) and luciferase have been developed for nuclear gene expression and protein localisation studies in Chlamydomonas and related green algae such as Volvox carteri (Fuhrmann et al. 1999, 2004). GFP-fusions in particular have proven to be a useful tool for the in vivo study of dynamic processes such as flagellar and centriole assembly and cell cycle events in Chlamydomonas (e.g. Ruiz-Binder et al. 2002). Similar fluorescent reporters based on GFP, luciferase and aequorin have been developed for gene expression and protein localisation studies in P. tricornutum (Falciatore and Bowler 2002; Zaslavskaia et al. 2000).

A modified gfp gene for Chlamydomonas chloroplast transformation has also been synthesised by optimising its codon usage to reflect that of the major chloroplast-encoded proteins (Franklin et al. 2002). Chloroplasts transformed with this modified gene under the control of the C. reinhardtii rbcL 5′- and 3′-UTRs were shown to accumulate ∼80-fold more GFP than those transformed with the native gene. The GFP accumulated to ∼0.5% of the total soluble protein (TSP), ∼50-fold higher than reports of uidA (GUS) expression under the control of the same promoter (Ishikura et al. 1999). In contrast to the high levels of GFP targeted to the nucleus (Fuhrmann et al. 1999), fluorescence in the chloroplast was not easily visualised using fluorescence microscopy (Franklin et al. 2002). This is most likely due to the chlorophyll and other pigments within the chloroplast absorbing much of the incident light targeted to GFP, or some of the light being emitted by the GFP being reabsorbed by the chloroplast pigments. GFP has been visualised in the chloroplasts of higher plants (Reed et al. 2001), but only when the levels of GFP accumulated to over 5% of the TSP, approximately 10-fold higher than levels presently obtained in C. reinhardtii. Recently successful GFP visualisation in the Chlamydomonas chloroplast has been reported using a gfp gene that had been modified for tobacco chloroplast expression (Komine et al. 2002). In addition to GFP, a second fluorescent reporter has now been developed based on the bacterial two-subunit luciferase. Mayfield and Schultz (2004) created a codon-optimised gene (luxCt) encoding a single polypeptide in which the two subunits are fused, and showed that luxCt is a sensitive reporter of expression levels in the chloroplast of living cells.

Novel transformation approaches need to be explored in order to achieve high-level expression of heterologous proteins in microalgae. For example, DNA viruses have been a useful source of promoters in higher plants to drive expression of transformation cassettes and this may apply to algae. The Phycodnaviridae, a family of large icosahedral double-stranded DNA viruses that infect eukaryotic algae, serve as an example. A promoter derived from the Paramecium bursaria chlorella virus (PBCV-1) has been shown to function extremely well both in higher plants and Chlorella (Mitra and Higgins 1994; Kang et al. 2000). Undoubtedly, a better understanding of the molecular biology of microalgae will lead to further advances in transgene expression. Towards this goal, the genome sequencing of several microalgal species including C. reinhardtii (Grossman et al. 2003), the rhodophyte Cyanidoschyzon merolae (Matsuzaki et al. 2004) and the diatom Thalassiosira pseudonana (Armbrust et al. 2004) will lead to important insights into the identity and expression of algal genes and the function of their products.

Current developments in microalgal bioreactor systems

The concept of “Molecular Farming” was first realised by Charles Arntzen and colleagues in 1992, with the successful production of a human protein in a transgenic plant (Mason et al. 1992). There are now hundreds of examples of the expression of recombinant proteins in plants many of which are now in human trials or at the commercial production stage (Mason and Arntzen 1995; Cramer et al. 1996; Fischer and Emans 2000). While research in this area has focused on the use of field crops such as tobacco, potato and banana, microalgae would make an economical alternative to current technologies. An increased demand for simple small-scale, high quality production (for example in the case of research proteins) may also favour sterile, fermentation approaches rather than broad acre crops. Indeed, algal biotechnology companies focused on producing recombinant proteins in fermentation systems using Chlamydomonas have begun to operate (Franklin and Mayfield 2004), although some of these are clearly still in the research and development phase (Table 2). In these instances, the high return on select high-value protein products offsets the costs on plant infrastructure, media requirements, harvesting and processing. The increase in the number of patents surrounding algal transformation and expression of heterologous proteins augurs well for the future development of an industry.

Routine nuclear transformation of microalgal species is central to future development of a biotechnology industry using microalgal bioreactor systems, however only three algal species (C. reinhardtii, V. carteri and P. tricornutum) can be “routinely” transformed. Certainly there is a huge potential market for natural products harvested from this group of organisms and metabolic engineering provides further opportunities through increased yield. Natural products such as the carotenoids β-carotene from Dunaliella and astaxanthin from Haematococcus attract premium prices and total markets exceed US$500 million. In the past, traditional mutagenesis and cultivation methods have been employed to increase yields, for example astaxanthin yields in one strain have been improved to 1.5–3% yield by dry weight. Metabolic engineering may provide a mechanism for further increasing yield as well as producing modified carotenoids with novel properties. Similarly, metabolic engineering in diatoms and dinoflagellates could provide improved yields of fatty acids and also provides opportunities for the production of modified oils and fatty acids for both the health food market and as a source of biofuels (Apt and Behrens 1999).

Light is the rate-limiting factor for cell growth in fermentation and open pond systems as it can only penetrate the first few centimetres into dense cultures. For example, outdoor cultivation systems realise a conversion of 3–4% of total solar energy to biomass, well below the theoretical expected value of 10%; a value which is also seen in laboratory conditions under low light (Benemann and Oswald 1996). The reason for the lower productivity is that the photosynthetic system of microalgae captures more photons than can be used preventing the flow of photons to cells further removed from the light source. The poor productivity of Dunaliella mass cultures has been attributed to this light limitation effect (Benemann and Oswald 1996). To obviate this effect mutagenesis approaches were utilised with the view of reducing the cellular content of light harvesting pigments and resulted in significant increases of photosynthetic rates and productivity in dense cultures of Chlamydomonas (Nakajima and Ueda 2000; Polle et al. 2000) and Dunaliella (Melis et al. 1999) in comparison to cultures of wildtype cells. Conversion of photoautotrophic microalgae to heterotrophy also provides an alternative solution to the cost of supplying light in closed photobioreactor systems. The problem of light saturation has been solved for diatoms with the trophic conversion of P. tricornutum, an obligate photoautotrophic organism, to heterotrophic growth (Zaslavskaia et al. 2001). The authors predict this will result in a 10–50-fold increase in dry biomass accumulation when compared to currently used light-dependent culture systems, therefore reducing production costs by an order of magnitude (Zaslavskaia et al. 2001). However, this system can only be viable on a large scale if the exogenous carbon source is extremely cheap.

Chloroplast genetic engineering is becoming a favoured method for protein production in higher plants as high levels of transgene expression can be achieved. For human somatotropin, an expression level of 7% total soluble protein (TSP) in chloroplasts was reported by Staub et al. (2000), which is about 300-fold higher than a similar gene expressed using nuclear transformation. De Cosa et al. (2001) were able to engineer tobacco chloroplasts to produce high levels of Btcry2Aa2 protein. The yield of protein was 46.1% TSP of mature leaves. This is the highest level of foreign protein expression reported so far in transgenic plants. More recently, the tetanus toxin was shown to accumulate to between 10–15% of total soluble protein in transgenic tobacco chloroplasts and the recombinant protein was able to induce protective levels of antibodies in mice (Tregoning et al. 2003). Although the nucleus has been the traditional target for transgene expression, chloroplast transformation has several advantages over nuclear transformation, including higher levels of expression than in the nucleus, the ability to introduce groups of genes as operons, and precise and targeted integration of recombinant DNA via homologous recombination to give predictable and uniform gene expression.

Two examples of heterologous protein expression in the C. reinhardtii chloroplast have recently been reported and demonstrate the feasibility of the approach in microalgae. Researchers at the Scripps Research Institute (La Jolla, California) engineered the chloroplast of C. reinhardtii to express a large single-chain antibody directed against the herpes simplex virus glycoprotein D in the chloroplast (Mayfield et al. 2003). The antibody was synthesised using codons optimised to reflect abundantly translated C. reinhardtii chloroplast mRNAs and its transcription was under the control of either the rbcL or atpA promoter and 5′ UTR and the rbcL 3′ UTR. After co-bombardment with a plasmid containing a modified 16S ribosomal RNA gene, transformants were selected on spectinomycin. The fully active recombinant antibody accumulated to ∼1% TSP in the transgenic chloroplasts, however levels would need to be increased signficantly for commercial viability. This was the first report on the use of a microalgae as a bioreactor. It has been estimated that costs for monoclonal antibody production in mammalian cell culture are approximately US$150 per gram, whereas production in plant systems costs approximately US$0.05 per gram (Mayfield et al. 2003). Algal systems are expected to further reduce the production costs more than 10-fold to approximately $US0.002 per litre due to cheaper media costs and the ability to grow large quantities in open pond systems (Mayfield et al. 2003). Research is also focusing on the expression of antibodies in the nuclear genome of C. reinhardtii. The second report involves expression of a foot and mouth disease virus VP1 and a cholera toxin B subunit fusion protein in the C. reinhardtii chloroplast (Sun et al. 2003). The fusion protein was expressed at approximately 3% TSP and was shown to bind to the intestinal membrane GM1-ganglioside receptor indicating its potential as a mucosal vaccine source.

Chlorella also shows promise as a protein expression system; one novel application being investigated involves the feeding of transgenic Chlorella expressing trypsin-modulating oostatic factor to mosquito larvae as a potential larvicide for mosquito control (Borovsky 2003). Other groups are examining Chlorella as a bioreactor for the production of growth hormones. Transient expression of human growth hormone (hGH) into the extracellular medium of C. vulgaris and C. sorokiniana has been reported (Hawkins and Nakamura 1999). Problems were encountered however, with instability of the foreign DNA and hGH production could not be detected after 1–2 months. No stable transformants could be established. More recently, the flounder growth hormone has been produced in C. ellipsoidea cells under the control of the CaMV 35S promoter. In this case, the recombinant protein was detected after seven passages in media devoid of the selective agent phleomycin, and when fed to flounder promoted a 25% increase in their growth after 30 days (Kim et al. 2002). The gene encoding the mature rabbit neutrophil peptide-1 (NP-1) has also been stably introduced into C. ellipsoidea cells. Highly efficient expression of biologically active NP-1 by the transgenic Chlorella cells was demonstrated using in vitro anti-microbial tests (Chen et al. 2001).

More recently a report describing the use of Dunaliella salina has been published (Geng et al. 2003). Here they report expression of the hepatitis B surface antigen in D. salina under the control of the maize ubiquitin-Ω promoter, a promoter that has previously been used for the expression of foreign genes in Chlorella ellipsoidea (Chen et al. 2001). While the rate of transient transformation was relatively low (between 6.6×10−5 and 1.9×10−3 cells as a frequency of total cells), at least 14 positive lines were confirmed by PCR and Southern blot analysis although stable integration of the transgene was not determined. The presence of the expected 25 kDa hepatitis B protein was detected in four lines by Western blot.

In addition to the recognised uses discussed above, other uses of microalgae continued to be explored. The possibility of using transgenic microalgae for bioremediation or as a biosensor of environmental toxins such as heavy metals has been explored (Siripornadulsil et al. 2002). Heavy metals such as cadmium are used in a wide variety of industrial processes, including plastic manufacturing, electroplating, and battery production. One mechanism by which plants and algae detoxify cadmium is the production of proline. Transgenic C reinhardtii cells expressing the mothbean Δ1-pyrroline-5-carboxylate synthase (P5CS) gene have 80% higher free-Pro levels than wild-type cells and show 80% increase in cadmium tolerance and a marked increase in binding of the metal.

A high lipid content, routinely 20% and potentially as high as 60% dry weight, resulted in an interest in developing biofuels from microalgae. After the oil embargo of 1973, the U.S. Department of Energy began funding projects to develop renewable transportation fuels from microalgae. While early projects concentrated on hydrogen and methane production, later funding under the Aquatic Species Program (ASP) was directed towards mass culture of microalgae grown in open ponds utilising CO2 from coal-fired power plants. The economic feasibility of such schemes required productivity levels at the theoretical maximum, open pond systems based on non-arable land, high oil prices and credits for greenhouse gas abatement (Benemann and Oswald 1996). While the economic feasibility of many of the schemes remained to be demonstrated when the ASP was terminated in 1996, research initiated under the DOE funding continues today. For instance, Melis and colleagues at the University of California at Berkeley have further developed the use of Chlamydomonas for hydrogen fuel production (Melis et al. 2000; Zhang et al. 2002). Under strictly anaerobic conditions Chlamydomonas uses light energy to reduce protons (H+) to hydrogen gas (H2), however, the production of H2 is normally suppressed by the presence of O2 which is also generated in the water-splitting reaction. Under sulphur-deficient conditions the rate of O2 production drops below that of respiratory O2 consumption, cultures become anaerobic and H2 gas is generated. The inexorable increase in oil prices remains an underlying economic factor in the exploration of alternate fuel sources and may drive the revisitation of many of the ideas explored between the 1970s and 1990s as part of the ASP. Indeed as recently as March 2005, the European Union announced a new research network termed SOLAR-H aimed at discovering new methods of producing hydrogen including by photobiological production using genetically engineered microalgae. Continued innovation could eventually permit the generation of hydrogen gas as a clean, renewable and economically viable fuel using the most plentiful natural resources of light and H2O.

Political as well as economic forces may well drive the future development of microalgal bioreactor systems. From concerns about global climate change and with breakthrough international developments such as the Kyoto Protocol, greenhouse gas mitigation has become a major focus of most countries including the U.S. Carbon-fixing rates of microalgae are an order of magnitude higher than land plants due to fast growth rates. The microalgal sequestration of CO2 generated by coal-fired power plants was a major focus of the DOE's ASP and research focussed on the efficiency and economics of fixation of effluent CO2 in open pond systems. In 2001, an international network on the biofixation of CO2 and greenhouse gas abatement with microalgae was established with many of the research goals built on the prior ASP outcomes. The network, which involves DOE's National Energy Technology Laboratory, Enitecnolgie and the International Energy Agency, has a 5-year plan to demonstrate the feasibility of microalgae-based technologies for greenhouse gas abatement and to achieve practical applications within a decade (Benemann 2003). To offset the high costs of using microalgae in open pond systems to sequester CO2, maximising productivity is a major issue of concern. Suggestions such as using the same ponds for wastewater treatment or using microalgae as animal feed and fertilizer provide some level of additional return albeit at a low level. Biofuels provide a higher level of additional return; however the highest return is achieved when targeting high-value products such as bioplastics for commercial use or nutritional supplements for human health (Benemann 2003). The genetic manipulation of rates of CO2 fixation, photosynthetic productivity along with levels of metabolite production provide avenues for further improvement in the economics of greenhouse gas mitigation using microalgae and may drive the development of bioreactors towards multipurpose systems using transgenic strains.

The potential of escape of transgenic organisms is, however, a concern when using microalgae in open pond systems. In land plants, the potential for gene flow via pollen to surrounding crops is cited as a major threat to crops marketed as GMO-free and to natural populations of the species. Although microalgae do not form pollen, they are able to spread through water and air systems; often aided by waterbirds and other vertebrate species. This raises concerns similar to transgenic land plants regarding introgression of genes into natural populations and raises the possibility that transgenic microalgae would be prohibited from open pond systems. Any large-scale cultivation of transgenic microalgae will be required by law, for the foreseeable future, to be fully enclosed and this may restrict their usage to small-scale production processes either in photobioreactors or in mixotrophic systems using light and a supplemented carbon substrate (Lee 2001).

Conclusions

Although microalgae have been used for decades as a commercial source of high-value nutritional supplements, this group of microscopic plants have become the subject of increasing interest from the perspective of the production of heterologous proteins for commercial and research purposes (He 2003; Franklin and Mayfield 2004). The ease and low cost of large-scale culture make this group of photosynthetic eukaryotes particularly attractive as bioreactors and allow diversification of existing industries concentrating on harvesting natural products. Microalgae share many attributes with higher plants, including patterns of glycosylation, and have a low risk of contamination by viruses or prions that infect animals. Unlike higher plants, however, microalgae have relatively fast growth rates, can achieve high cell densities under conditions of high light and aeration and be grown in volumes of megalitres. In systems where genetic transformation is routine, primary transformants can be generated in as little as two weeks. Because microalgae are simple in structure, protein purification is a simpler process and costs are therefore lower than higher plants where the protein has be isolated from complex tissues. In the case of green algae, these organisms fall into the GRAS (generally regarded as safe) category, meaning that administration of a protein can be by ingestion of dried algal powder. Cost of production is an over-riding consideration in developing bioreactor systems and estimates of costs of production of recombinant antibodies in microalgal bioreactor systems is US$0.002 per gram. This is much lower than estimates of recombinant protein production in mammalian cell culture of US$150 per gram and compares very favourably with agricultural scale production using transgenic corn at US$0.05 per gram (Mayfield et al. 2003). The potential of using microalgae, in particular Chlamydomonas, as bioreactors has already resulted in the formation of several biotech companies.

There is still much basic science to do before microalgae can be fully exploited. Only three species, C. reinhardtii, V. carteri and P. tricornutum, can claim to be routinely transformable and this list needs to be expanded to include the many industrially important species. The recent reports of nuclear transformation of Dunaliella and of Haematococcus augurs well for the industry that extracts high-value compounds from these species (see Table 3). One outcome of the current algal genomics projects should be acceleration in the development of transformation methodology for microalgae and their subsequent exploitation. The coming years should prove to be an exciting time with important insights into algal biology, the creation of improved strains for specialised commercial applications as well as their potential use as bioreactors for the production of commercially valuable recombinant proteins.

Copyright information

© Springer-Verlag 2005