Journal of the American Oil Chemists' Society

, Volume 90, Issue 2, pp 167–182

Single-Cell Oils as a Source of Omega-3 Fatty Acids: An Overview of Recent Advances


    • Fermentation and Metabolic Engineering GroupOcean Nutrition Canada Limited
  • Mercia C. Valentine
    • Fermentation and Metabolic Engineering GroupOcean Nutrition Canada Limited
Review Article

DOI: 10.1007/s11746-012-2154-3

Cite this article as:
Armenta, R.E. & Valentine, M.C. J Am Oil Chem Soc (2013) 90: 167. doi:10.1007/s11746-012-2154-3


Omega-3 fatty acids, namely docosahexaenoic acid and eicosapentaenoic acid, have been linked to several beneficial health effects (i.e. mitigation effects of hypertension, stroke, diabetes, osteoporosis, depression, schizophrenia, asthma, macular degeneration, rheumatoid arthritis, etc.). The main source of omega-3 fatty acids is fish oil; lately however, fish oil market prices have increased significantly. This has prompted a significant amount of research on the use of single-cell oils as a source of omega-3 fatty acids. Some of the microbes reported to produce edible oil that contains omega-3 fatty acids are from the genus Schizochytrium, Thraustochytrium and Ulkenia. An advantage of a single cell oil is that it usually contains a significant amount of natural antioxidants (i.e. carotenoids and tocopherols), which can protect omega-3 fatty acids from oxidation, hence making this oil less prone to oxidation than oils derived from plants and marine animals. Production yields of single cell oils and of omega-3 fatty acids vary with the microbe used, with the fermentative growing conditions, and extractive procedures employed to recover the oil. This paper presents an overview of recent advances, reported within the last 10 years, in the production of single cell oils rich in omega-3 fatty acids.


Omega-3 fatty acidsSingle cell oilsDocosahexaenoic acidEicosapentaenoic acidFermentationThraustochytridsOmega-6 fatty acids


Polyunsaturated fatty acids (PUFA), and specifically the omega-3 fatty acids (FA) docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) (Fig. 1) have been linked to several health benefits [1]. Just during 2011, several studies have suggested there are positive health effects from EPA and DHA consumption, such as lowering the risk of vision loss due to eye macular degeneration [2], reduction of blood vessel stiffness [3], and relief of anxiety and inflammation [4]. DHA consumption has been associated with reducing risk of colds in babies from moms that took this dietary supplement [5]. All the latter, plus numerous studies that suggest there are health effects for the heart and brain, and many other benefits for treating several illnesses and diseases including asthma, rheumatoid arthritis, schizophrenia, depression, multiple sclerosis, migraine headaches, etc [6]. Even low DHA levels have been associated to suicide risk [7].
Fig. 1

Chemical structures of the omega-3 FA’ EPA (20:5n-3) (a) and DHA (22:6n-3) (b)

Microorganisms are significant producers of oil [8]. Table 1 includes phototrophic and heterotrophic microorganisms and their reported oil percentages. Some of these microbial oils may contain significant amounts of omega-3 and omega-6 FA such as EPA/DHA and arachidonic acid (ARA), respectively. Most of oil producing microorganisms studied within the past decade have been eukaryotes; and have been found worldwide, both along coastlines and in the open ocean. Using microorganisms to produce omega-3 FA is still a relatively new field, and research on this area has been growing significantly within the last few years. Studied eukaryotes include Pavlova lutheri [9], which is a form of algae, Mortierella alpina (a type of fungus) [10], and Crypthecodinium cohnii [11] from the super family Alveolata. Heterokonts are also eukaryotes that have some oil producing species. This group is divided into subgroups: the PX clade [12], the Oomycetes [13], the Bacillariophyta group [14, 15], and the Labyrinthulida [1618]. There are also oleaginous bacterial species, such as Moritella dasanensis [19] and Shewanella sp. [20]. Table 2 shows different classes of microorganisms that produce a large range of biomass and FA.
Table 1

Oil content of several microorganisms (includes phototrophic and heterotrophic organisms)


Oil content (% dry weight)

Botryococcus braunii


Chlorella sp.


Crypthecodinium cohnii


Cylindrotheca sp.


Dunaliella primolecta


Isochrysis sp.


Monallanthus salina


Nannochloropsis sp.


Neochloris oleoabundans


Nitzschia sp.


Phaeodactylum tricornutum


Schizochytrium sp. (thraustochytrids)


Tetraselmis suecica


Adapted from Singh et al. [8]

Table 2

Biomass and FA produced with various microorganisms


Biomass (g/L)

Total oil (g/L)

Prominent PUFA (g/L)


Mortierella alpina DSA-12



ARA (18.8)


Pythium irregulare ATCC 10951



EPA (0.89)


Phaeodactylum tricornutum UTEX 640



EPA (0.5)


Aurantiochytrium strain BL10



DHA (16.8)


Crypthecodinium cohnii ATCC 30556



DHA (0.14)


Thraustochytrium sp. ONC T18



DHA (4.6)


PUFA Polyunsaturated fatty acids, ARA arachidonic acid (C20:4n-3), EPA eicosapentaenoic acid (C20:5n-3), DHA docosahexaenoic acid (C22:6n-3), N/Av not available

Single cell oils (SCO) are microbial oils considered a promising oil alternative to those from fish and land based plant sources. The term single cell oils was first introduced by Professor Ratledge in 1976 to define oils produced by single celled microorganisms such as yeasts and molds [21]. He indicated that although bacteria are also single cell microorganisms, they do not tend to accumulate edible oils. Recombinant bacteria such as Escherichia coli are not likely suitable to produce edible oils, however, this is being attempted to produce lipid biofuel [22]. Thus, SCO have attracted attention for their commercial development as they can be a significant source of specialty lipids such as PUFA. Also, SCO have been suggested to be a promising source of biofuels. Two examples of current commercial SCO rich in omega-3 FA are oil with DHA from a Schizochytrium [23], and EPA from the genetically modified yeast Yarrowialipolytica [24].

Thraustochytrids, which includes the genera Schizochytrium and Thraustochytrium, are among the most promising microorganisms for producing omega-3 FA, with reported oil contents of >50 % (dry basis) and more than 30 % DHA within the total oil produced [25]. Reported isolates of thraustochytrids would produce both EPA and DHA; however, only one (often DHA) is produced in significant amounts. In addition to omega-3 FA, these types of microorganisms have been reported to produce carotenoids and co-enzymes [26] in an oil composed of >99 % triacylglycerides [27], with a low cholesterol content (<1.5 mg/g) [28]. Thraustochytrids are shown in Fig. 2.
Fig. 2

Oil producing thraustochytrid cells; intracellular oil is stored in compartments/vesicles (a); scanning electron microscope (SEM) image (b). Scale bar 5 μm

Thraustochytrids are often mistakenly called microalgae when discussing their potential biotechnological applications. Although they are closely related to brown algae, thraustochytrids are not algae, and there is no literature that classifies them as such. They are non-photosynthetic marine protists within the kingdom Chromista, phylum Labyrinthulomycota. These are heterotrophic and absorptive microorganisms that produce heterokont zoospores, and have been generally associated with, and isolated from, decomposing algal and plant materials and from sediments and offshore water samples. Thus, their nutrition in the natural environment is primarily saprotrophic, and are considered weak parasites of algae [29, 30].

Thraustochytrids are relatively unknown microorganisms, and new environmental isolates have been reported around the world. That is why taxonomic classification of thraustochytrids is not settled. Based on morphology, chemotaxonomic features and 18S rRNA gene phylogeny, a taxonomic rearrangement was recently proposed for the genus Schizochytrium sensu lato, with three different genera proposed: Schizochytrium sensu stricto, Aurantiochytrium, and Oblongichytrium gen. nov. [31]. There is a need of using a combination of phylogeny, morphology and profile of produced products (i.e. specific FA and carotenoids) in order to carry out a more accurate classification of these microorganisms.

This paper presents an overview of recent advances made in three areas of optimization for producing SCO rich in omega-3 FA: upstream and downstream processing, and metabolic engineering.

Upstream Processing

A number of variables have been considered to optimize growth and fatty acid accumulation by microorganisms. Factors include carbon source, nitrogen source, salinity, dissolved oxygen, pH, length of growth, temperature, and culture mode, which includes a two-stage fermentation process.

Carbon Source

One of the first things to determine when growing a microorganism is its optimal carbon source. Preferred carbon form and concentration can vary widely between species. The most common tested carbon source is glucose, as the metabolic biochemical pathway that converts this sugar into FA is relatively efficient and is found in most microorganisms. The maximum theoretical lipid yield (Yoil/glucose) for consumed glucose is 31 %, and 30 % for glycerol (Yoil/glycerol) [27, 32]. This is paramount to consider when assessing microbial technologies that may suggest sugar to lipid yields above the maximum theoretical limit. Indeed, in the scientific literature, it is very rare to see results of glucose to lipid yields above 20 % [33]. Glucose has been used at different starting concentrations within the media culture. The lowest concentration of glucose found was 1 g/L for growing bacteria Shewanella [20]. Generally, used glucose levels have been in the range of 5–40 g/L. However, the heterokont Monodus subterraneus grew best with a glucose concentration of 60 g/L [12]. In one study using Crypthecodinium cohnii, the most biomass, 3.13 g/L was produced at a starting glucose concentration of 20 g/L [34]. For the same microorganism, the highest DHA accumulation, 53.4 % of total fat, was at 5 g/L of glucose. The preferred carbon source of recently discovered microorganisms remains to be determined. For example, Moritella dasanensis has the potential to grow on a variety of carbon sources including trehalose, d-arabinose, and glycerol [19].

Using glycerol as a carbon source for producing SCO has become more common in the past 5 years because large amounts of raw glycerol are available as a by-product from the biofuel industry, for which there is an insufficient market [13]. The DHA-producing microalgae Schizochytrium limacinum can consume pre-treated glycerol as a carbon source, and actually thrives better with glycerol compared to glucose [35]. Nevertheless, soap and methanol remnants in glycerol by-product inhibited cell growth of this thraustochytrid [36]. Figure 3 depicts how both glucose and glycerol are metabolized for fatty acid production (triacylglycerol oil).
Fig. 3

Microbial metabolism of glucose and glycerol to produce triacylglyceride oil

The EPA-producing fungus Pythium irregulare is also capable of using crude glycerol [13]. Two of the large impurities found in crude glycerol are methanol and soap, and differing explanations have been provided about how these impurities affect the growth and fatty acid production by their respective microorganisms. It was also found that both soap and methanol inhibited biomass production and fatty acid accumulation of Pythium irregulare [13]. On the contrary, Chi et al. [35] found that if soap concentration was <5 %, cell growth for Schizochytrium limacinum was improved, also noting that if methanol remained under 20 g/L there was little effect on growth. In any case, glycerol appears to be a promising potential carbon source for several omega-3 FA-producing microorganisms. This, besides the benefit of reducing a waste (glycerol by-product), could reduce the cost of producing omega-3 FA.

In a similar attempt of reusing waste products from other industries, Crypthecodinium cohnii CCMP 316 was grown on carob pulp syrup [37]. Carob plant’s pulp has a high sugar content, which, after extraction, is turned into syrup that can be used as a carbon source in fermentations. Using a diluted supply of carob pulp syrup supplemented with sea salt and yeast extract yielded 42 g/L of biomass with 1.9 g/L of DHA [37]. In another example, Thraustochytriidae sp. AS4-A1 (Ulkenia sp.) was grown on waste products from a potato chip plant and a brewery [38]. The carbon source was a mixture of starch, glucose, maltose and ethanol. The crude by-product allowed 8.1 g/L of dry biomass and 0.5 g/L of DHA to be produced. However, supplementing the waste by-product with a mixture of yeast extract and MSG (monosodium glutamate) as well as B vitamins produced better biomass and DHA, 14.9 g/L (dry) and 2.6 g/L, respectively [38].

Other waste carbon sources such as inulin [39] and thin-stillage [40] have been used to produce oil with yeast and fungus, respectively. Inulin hydrolysates, from Jerusalem artichoke, with fructose and glucose as fermentable sugars, were used to produce microbial oil by yeast Rhodotorulamucilaginosa, producing up to 19.5 g/L cell biomass with 52 % oil content on a dry basis [39]. Thin-stillage is a major by-product from dry-milling corn-ethanol plants; this substrate produced 23 g/L of cell mass (Pythium irregulare) in a 9-day process, with an EPA yield of 243 mg/L and productivity of 27 mg/L/d [40].

In some cases, oil is added to the fermentation media where it can be directly converted into the desired FA. Thus, by adding oil, biomass and FA, the yield could be improved [13]. Some examples of oils added to the fermentation media are flaxseed oil and soybean oil. While vegetable oils are added to complement the carbon source, in one case, soybean oil, or soybean lecithin, was used as the only carbon source for Labyrinthula sp. strain L72 [41]. Soybean oil at a concentration of 5 g/L produced 0.32 g/L of DHA while soybean lecithin, at the same concentration, only produced 0.17 g/L. In comparison, glucose as a carbon source at 5 g/L only allowed 0.01 g/L of DHA to be produced. Some microorganisms can consume nutrients under less than ideal conditions. The ectoplasmic network of Labyrinthula species can digest bacteria [42]. In this case, bacteria Psychrobacter phenylpyruvicus was dispersed in the agar for Labyrinthula sp. S3-2 to consume, and when Labyrinthula sp. S3-2 was incubated with just soybean oil and base agar media, 5.1 % of the total oil was DHA. Nevertheless, DHA production improved (8.7 % of total oil) when incubated with P. phenylpyruvicus and soybean oil in base agar media.

Carbon is generally the most expensive fermentation ingredient when producing omega-3 FA, thus a current challenge regarding bioprocess optimization is finding and enhancing strains able to efficiently metabolize cheap and widely available sugars. A significant monosaccharide candidate is the five-carbon sugar (5-C) xylose. Xylose, besides glucose, is the most abundant sugar monomer in hemicellulose. Typically, and in comparison to hexoses such as glucose, microbial metabolism of xylose is inefficient; thus, this sugar is accumulated in large amounts in industrial waste carbon feedstocks (i.e., agricultural and forestry wastes). This makes xylose the most abundant and underused carbon source from waste streams. Xylose has been tested for growing thraustochytrid Aurantiochytrium sp. [43], yeast Cryptococcus curvatus [44], and fungus Mortierella isabellina [33].

Moreover, fast growing seaweed has been used as a novel carbon feedstock in heterotrophic processes to produce ethanol. Seaweed, when compared to hemicellulose, is easier to convert into fermentable sugars, resulting in less or no fermentation inhibitory compounds that result from hydrolysis of lignocellulosic materials. When hydrolyzing red seaweed, for example, glucose and galactose are mostly produced. However, galactose microbial metabolism is usually not efficient, and efforts have been made to improve its microbial conversion to ethanol [45]. Although early efforts at using seaweed fermentable sugars have been able to produce ethanol, this carbon source could also be used to produce microbial oil. This research area poses a significant opportunity, especially when aiming at production of microbial oils rich in bioactives (i.e. omega-3 FA).

Nitrogen Source

Nitrogen is another important factor in culture growth. Nitrogen is used for biomass accumulation (first phase of growth) by assisting in amino acid synthesis among other processes. Once the nitrogen supply runs out, cells begin to accumulate FA using the remaining carbon provided (second phase of growth). There are many nitrogen sources available for microbial growth. In most cases, nitrogen optimization has not been targeted. However, in one study, a combination of monosodium glutamate and yeast extract was found to be ideal for growing a thraustochytrid [25].

Yeast extract and peptone are common nitrogen sources, which are often used together. For example, to grow a thraustochytrid-like microorganism (strain 12B), a nitrogen concentration of 1 % peptone and 1 % yeast extract have been commonly reported [16]. Crypthecodinium cohnii ATCC 30556 used a combination of tryptone and yeast extract [11]. To grow Nitzschia laevis, small concentrations of both nitrate, ammonium and urea have been used [46, 47]. Sometimes nitrogen can be fed into the system through acid/base control (i.e. HNO3/NH4OH). Mortierella alpina DSA-12, had an initial concentration of 10 g/L of corn steep powder as well as 14 % (v/v) NH4OH for base control [48].

In microbial production of omega-3 FA, nitrogen is the second most expensive ingredient after sugar (carbon source). Its cost increases significantly when nitrogen comes from protein sources such as yeast extract, soy peptone and glutamate. Thus, an alternative for cost reduction is to use inorganic nitrogen such as NH4OH and (NH4)2SO4. However, replacing protein with mineral nitrogen would likely result in the need of increasing addition of trace elements and vitamins; this because protein nitrogen sources are complex nutrients that already contain some of these extra ingredients.

Also, rendered animal protein [49] and municipal wastewater [44] have recently been used as a nutrient source to produce microbial oil. These type of wastes can be a potential abundant and economical source of not only nitrogen, but also phosphorous in the case of wastewater. Research to define feasibility and optimization of waste use in fermentation processes to produce high value bioactives could be worthwhile. However, it would be expected that regulatory regulations will play a significant role at defining if, and how, such nitrogen feedstocks could be used to produce microbial bioactives aimed at human and animal markets.

Culture Mode

A two-stage bioprocess for cell growth and lipid production is common for oleaginous microorganisms [50], such as the yeast Lipomyces starkeyi [51]. Cell proliferation occurs at the first phase because all media nutrients are in excess with constant low oil content (approximately 10 % of cell dry weight). The second phase for oil production starts when the microbe is stressed due to nitrogen exhaustion. Here the microorganism continues to assimilate the carbon source (sugar); however, due to nitrogen limitation (used for DHA/RNA and protein synthesis) most of cell functions focus on oil accumulation for an energy reserve [52].

Microbial production of omega-3 FA is a relatively new emerging research field, and the large majority of peer reviewed literature refers to growing microorganisms in flasks and small reactors at the laboratory bench scale. Culture methods include batch mode, fed-batch and continuous culture modes [53, 54]. In a batch mode, all media and nutrients are supplied right from the start. Fed-batch mode is similar to batch mode in that some nutrients are supplied at the beginning, but at some point, additional media or nutrients are added in. Continuous cultures are those that have fresh media being pumped in while old media and some cells are pumped out continuously. Bioprocess scale also plays a role, in the case of a Thraustochytrium sp. grown at different levels (agar plates, flasks and automated bioreactors), it was determined that variety of FA decreased from agar plates to bioreactors [25].

A research opportunity for producing microbial omega-3 FA could be the use of semi-continuous, instead of fed-batch, fermentation culture mode. Nevertheless, this requires significant research in order to determine, for each candidate microorganism, the time for ideal partial harvesting based on maximum oil productivity, the amount of cell biomass to be left behind as seed to continue the process and proceed at replenishing fermentation media, among other operating factors.

Dissolved Oxygen

Earlier biochemical practices recommended high Dissolved Oxygen (DO) levels in the fermentation media in order to boost formation of double bonds in FA, such as PUFA DHA [55]. Nevertheless, a significant increase in unsaturated FA, under limited fermentation oxygen conditions, was observed when producing lipids with Candida [21].

The relationship between DO effects and desaturation of FA requires further investigation [50]. During initial cell growth of a fermentation process, a high oxygen consumption rate is expected [50]. Within this active cell replication stage, high DO levels are difficult to maintain because microorganisms use oxygen faster than it can be replaced when growing in a medium with an abundant organic matter (i.e. sugar), thus reducing oxygen levels in the media environment [56].

When producing microbial oil, limited oxygen (or low DO levels) is not a fermentation operating condition only applicable to a particular group of heterotrophic microorganisms. Low DO occurs when producing FA with a variety of microorganisms such as yeast, fungi and marine protists. DO levels in fermentation media have been linked to significant production of microbial oil with yeast or mold in a two-phase fermentation batch process [52]. This has been observed in thraustochytrids [50], and yeast (Candida and Rhodotorula) [57], where cell oxygen uptake was higher during the active cell replication fermentation stage, thus with a low DO saturation percentage in the fermentation media. Subsequently, cell oxygen demand dropped dramatically when nitrogen became limited in the medium, a nutrient limitation condition suggested to trigger oil accumulation within the fermentation second phase. This is expected because cell replication is significantly minimized when nitrogen is limited, thus the microorganism prioritizes energy accumulation in the form of oil. Additionally, this is on the basis that fat, a chemically reduced compound, does not require oxygen for its synthesis from malonyl-coenzyme A [57].

Oxygen transfer is a complex measurement as it involves a gas to liquid phase change. It is affected by several factors such as chemical composition of fermentation media, temperature, pressure and surface area of air/oxygen bubbles, volume of gas introduced per unit reactor volume per unit time, speed of agitation, type of sparger, oxygen sensor, or a combination of the above factors [56]. Therefore, the DO percentage in the fermentation medium is not an absolute measurement, and it is difficult to measure both at high and low percentages. For example, two fermentations (in exactly same fermentation vessels with identical DO sensors) where the only difference is either the media composition or temperature, is likely to show different DO readings at any same given sampling time.

DO effects on lipid desaturation of SCO is highly debatable and still remains to be investigated further [50]. For this, it is important to consider that DO is probably not the only factor contributing to the lipid profile. Factors such as temperature, carbon to nitrogen ratio, phosphorous, among others, have been shown to also influence lipid content and profile [46].


Generally, the pH level of fermentation cultures is not nearly as controlled as are the carbon and nitrogen sources. pH is important as the intracellular pH level is maintained primarily by proton pumps, such as K+ proton pumps and Na+ proton anti-port pumps [34]. Use of these protons is energy dependent; when cells are exposed to pH environments that are not optimal, they are forced to use energy to maintain the pH. Because this energy is diverted to pH maintenance, cell growth is reduced.

It is common to report the initial pH of fermentations; however, change in the pH at the end of fermentation experiments is rarely indicated. Scott et al. [27] reported a change from 7.4 to 6.6 at the end of a 6-day fermentation with thraustochytrids. When producing SCO with Cunninghamella echinulata and Mortierella isabellina, the pH of the medium did not change in relation to the carbon source used (glucose, glycerol and xylose) after 8 days of fermentation [58].

Another study [34] looked at the effect of pH when growing Crypthecodinium cohnii ATCC 30556, observing that this microorganism was able to grow at pH levels between 4 and 10. At the extreme ends of that range, biomass production was very poor, producing <0.5 g/L, while DHA at pH 4 and 10 were 44.8 and 45.6 %, respectively, relative to total fat. Optimum pH level was found to be 7.2 and produced a dry cell weight concentration of 2.73 g/L with DHA at 56.8 % of the total oil.

Typically, pH is studied with the goal of defining optimal cell growth and oil productivity. However, an opportunity exists of studying pH as a way to reduce sterilization needs in fermentation processes. Some oil-producing microbes are capable of growing at low pH, which could be a way to control contamination and, thus, reducing process energy inputs significantly. Nevertheless, especially for microbial omega-3 FA processes, regulatory authorities would need to address matters regarding production of these compounds under non-sterile environments. For low pH fermentations, new environmental isolates may have an advantage over recombinant microorganisms, as the latter ones often acquire a high sensitivity to drastic pH changes. Such research may help at increasing the understanding of a relatively new research area, microbial consortia systems, where controlled (i.e. through pH control) presence of other microbes could provide insights of beneficial relationships among microbes. Recently, a symbiotic relationship was discovered between photosynthetic algae and bacteria, in which bacteria provides vitamin B12 to algae in exchange of photosynthate or carbon substrate [59, 60]. Such understanding could help in developing bioprocesses where culture sterilization needs could be either reduced or removed, thus saving on energy consumption, which could greatly improve the economics of algae oil producing technologies.

Temperature and Growth Period

Temperature and the growth period greatly affect biomass accumulation and fatty acid yield. Overall, a higher temperature increases cell growth, while cooler temperatures promote accumulation of FA [20]. In a study that looked into the effect of different temperatures on Crypthecodinium cohnii ATCC 30556, it was determined that at 30 °C the highest biomass produced was 2.42 g/L, with 23.61 % DHA relative to total oil. At 15 °C, the maximum DHA was produced, 57.64 % of total oil [11]. While the majority of references described keeping the same temperature throughout microbial growth, several used a dual temperature system where cultures were grown at a specific temperature for biomass growth, and then a significantly cooler temperature for fatty acid accumulation. For example, in growing Mortierella alpina, the culture was grown for 2 days at 28 °C, followed by 2 days at 20 °C [61].

When growing P. irregulare on dry-milling derived thin stillage, a high EPA productivity was achieved by switching the fermentation temperature throughout a 9-day fermentation process: 30 °C for initial 4 days, 25 °C during following 3 days, and 15 °C during last 2 days. This bioprocess yielded 23 g/L of cell biomass and >240 mg/L of EPA [40].

A more extreme example is that of Shewanella sp. KMG327, where biomass accumulation phase occurred between 16 and 20 °C, while fatty acid accumulation occurred at 4 °C [20]. In a similar study with Crypthecodinium cohnii, a dual temperature system was also used [11]. Cells were grown at 25 °C for 48 h, and then exposed to a cooler environment at 15 °C for 24 h, increasing DHA production to 57.81 % of total oil. This DHA increase at lower temperatures was attributed to the microorganism trying to maintain proper membrane lipid fluidity by the addition of more PUFA. Besides temperature, the length of the fermentation can be affected by different factors including microorganism doubling time and how the culture is being fed and maintained, and also the desired outcome for the experiment (i.e. oil productivity targets).

Although lower temperatures may increase omega-3 FA production, temperatures below 20 °C could also diminish cell and oil productivities. Thus, a proper strategy could include first using a higher temperature (≥25 °C) to maximize cell growth during the fermentation first stage (exponential phase), and then switch to a lower temperature (<25 °C) to increase biosynthesis of omega-3 FA (lipid accumulation phase).


Many oil-producing microorganisms are capable of growing at a variety of salinity levels. Salt, in particular Na+ ions are required for growth by many microorganisms as it facilitates entrance of K+ ions, which are also required for growth [62]. Microbial growth has been reported at different salt levels with different organisms requiring different amounts. For example, Thraustochytrid striatum can grow with a salinity level between 2 and 50 g/L [25]. Salt can be obtained from several sources; natural sea water is often used at salt concentrations from 50 to 100 % of sea water [11, 13, 41, 63]. Other times, sea water is artificially created by the addition of various minerals such as magnesium, iron, potassium, and various others [64]. Other salts can be used as in the case of Moritella dasanensis, which can grow in NaCl [19], or Monodus subterraneus, which uses sodium acetate [12]. NaCl, CaCl2, pH and temperature were found to affect EPA productivity by Nitzschialaevis. Specifically, 14 and 0.1 g/L of NaCl and CaCl2, respectively, yielded an optimal EPA productivity of 28 mg/L/day [65].

For thraustochytrid growth, Na+ is required due to its role in cell osmotic regulation, and not because of the direct Na+ involvement in cell metabolism. However, there is evidence that suggests that thraustochytrid cells grow in the absence of Na+; provided that cell turgor is adjusted by other means such as compatible solutes [66]. Turgor is the pressure that pushes the cytoplasm against the cell wall (semi-permeable membrane), caused by the osmotic pressure differences between the outside and inside of the cell. Na+, Cl and K+ affect osmotic pressure, and it is still not clear their relative contribution in thraustochytrids’ turgor pressure [66].

As both Na+ and Cl could greatly affect microbes’ cell osmotic adjustment mechanism, this opens the possibility of using fermentation media with high salinity to control microbial growth in a non-sterile fermentation via osmotic regulation. This could be feasible if the oil-producing strain has a significant robustness against high-salinity conditions. Nevertheless, fermenter corrosion caused by Cl contained in salt (NaCl) should be considered. For this, other sources of Na+ could be investigated.


There has been significant discussion regarding effects of light on growth of omega-3 FA-producing microorganisms. Phototrophic microorganisms are those that use photosynthesis to get chemical energy from light energy [67]. Monodus subterraneus UTEX 151 produces 32.6 % EPA (relative to total oil) and has an optimal growth under 10 klux of light [12]. Pavlova lutheri (strain SMBA60) was grown at its optimal light condition of 11.7 klux to obtain a significant biomass and lipid growth [9].

While examining differences in Thraustochytrid growth in light and darkness, it was determined that in most cases these organisms did not require light to grow, but Thraustochytrium aureum was one species that grew better in light [63]. Other organisms can grow under light conditions that roughly mimic light from the natural environment. Labyrinthulid sp. L95-2 grew under a 12 h light fluorescent light cycle followed by 12 h of darkness [68]. Finally, there are several cases having conflicting views about the benefits of light. Nitzschia laevis UTEX 2047 is a microorganism that can thrive under photoautotrophic, heterotrophic, and a combination of both those conditions. In one study, a light intensity of 60 μE/m2 s was determined for optimal growth [15]. However, the following year, it was determined that darkness was optimal [46]. Optimization of media components had a great effect on those studies. Although there are many theories on how light can affect growth, each microorganism is different in the way it responds to light. Thus the question of how light will affect growth must be investigated for each species individually.

Light also affects the biosynthesis of minor components such as carotenoids. In a study with a thraustochytrid, blue light increased carotenoid (astaxanthin) biosynthesis [69]. Further studies could be aimed at defining specific light conditions to reach carotenoid concentrations that maximize the chemical stability protection of a microbial oil rich in omega-3 FA.

Downstream Processing

After growth and harvesting of microbial biomass, oil must be efficiently extracted. Downstream processing for oil extraction is the least investigated area of bio-processing. Nonetheless, there is no universal method that will yield the best oil recovery for all microorganisms. In some microbial cells, cell walls are particularly thick and do an excellent job of maintaining oil within the cytoplasm, thus a cell disruption method must be employed to aid the recovery process. A few of these techniques include sonication, bead milling, and the use of a French press. Shen et al. [70] experimented with several different extraction techniques on two different microorganisms, Scenedesmusdimorphus, and Chlorellaprotothecoides. For S. dimorphus the best method was milling, obtaining 25.3 % of lipids. For C. protothecoides the best extraction method involved the use of a bead-beater, obtaining 18.8 % of lipid recovery. Results are summarized in Table 3. Milling was performed by passing the material through a mill three times. The French press was operated at 1,500 psig, with five passes. Sonication was done with a 100-watt ultrasonic processor, with 2 passes at 2 min each. The bead-beater was used with 1 mm glass beads, with two passes at 2 min each. Soxhlet extraction was done on cells with no cell disruption. This study proved that same extraction method may not work for different microbial biomasses.
Table 3

Effect of same cell disruption methods on algae lipid yields (%) from two microbial biomasses: Scenedesmusdimorphus and Chlorellaprotothecoides


Wet milling

French press



Soxhlet extraction

S. dimorphus






C. protothecoides






Adapted from Shen et al. [70]

Other techniques do not involve a separate cell disruption method, but instead combine extraction and fatty acid esterification in one step with a mixture of solvents. The basis behind solvent based extractions is the concept that solvents will extract compounds of similar polarity. Specifically, a solvent, or a solvent mixture, used must be able to enter the biomass efficiently, interact with the lipids and form a solution that later can be easily disassembled to obtain the lipid fraction [71]. An extraction method that exemplifies this concept is the Bligh and Dyer method [72]. This procedure uses two solvents: chloroform and methanol. In this case, lipids which have a polar nature will associate with polar solvents like methanol, and other cellular constituents will associate with chloroform. Water is added to the mixture after the solvents have extracted the lipids, which then causes separation into two layers, an organic and an aqueous layer. The corresponding lipid layer is removed for further processing [72].

The majority of extraction techniques use mostly the same solvents, but some may be in different proportions or added at different times. The original Bligh and Dyer method uses chloroform, methanol and water in a ratio of 2:2:1.8. Zhu et al. [73] also used this ratio on wet biomass to obtain an oil yield of 27.6 % from Mortierella alpina. Other combinations of solvents include methanol, acetyl chloride and hexane [74], or hexane coupled with ultrasonics [75]. When comparing several oil extraction methods for extracting FA from a thraustochytrid biomass, a miniaturized version of the Bligh and Dyer method, and ultrasonic energy to breakdown cells, offered the highest extraction yields of FA [28]. Table 4 shows oil extraction results from several microorganisms.
Table 4

Oil extraction from various microorganisms


Cell amount (g)

Extraction method

Solvent(s) used

Lipid yielda (%)


Mortierella alpina

Bligh and Dyer

Chloroform, methanol



Monodus subterraneus and Phaeodactylum tricornutum

70 and 500, respectively

Simultaneous fatty acid extraction and transesterification

Methanol, acetyl chloride and hexane



Scenedesmus sp.


Bligh and Dyer

Chloroform, methanol and saline solution



Scenedesmus sp.


Ultrasonic extraction with sonication

Chloroform, methanol



Scenedesmus sp.


Ultrasonic extraction with sonication




Thraustochytrium sp.


Miniaturized Bligh and Dyer

Chloroform, methanol and buffer solution



N/Av Not available

a% of dry cell biomass

Another extraction method that has been employed is the use of supercritical fluids. A super critical fluid is a fluid that has been brought above its defined critical temperature and pressure [76]. Above these parameters, a solvent has excellent properties that allow them to increase their extraction capabilities. For example, the ability to dissolve materials like a liquid and the ability to diffuse through materials like a gas [76]. There are different solvents that can be used, and the most widely used one is supercritical CO2. Dimethylether is another solvent used under supercritical conditions. As shown in Table 5, supercritical fluid extraction (SFE) has the ability to extract a whole range of lipids depending on the type of microorganism and operating extraction conditions [7780]. Table 5 provides details on how much oil was extracted from various microorganisms using supercritical fluids. In these examples, all cell biomass used was dry as SFE does not work efficiently with samples that contain high moisture.
Table 5

Comparison of supercritical fluid extraction of lipids from various microbial biomasses


Amount (g)



Lipid yield (%)


Mortierella alpina



T: 316°K

P: 4 MPa



Spirulina platensis



100 g of glass beads

T: 313°K

P: 70 MPa



Spirulina platensis



T: 328°K

P: 70 MPa



Crypthecodinium cohnii



T: 323°K

P: 30 MPa



Liagora boergesenii



T: 328°K

P: 3.44 MPa



T Temperature, P pressure, N/Av not available

To date, oil extraction with organic solvents (i.e. hexane) coupled with mechanical disruption remains the most efficient and economical technology to extract oil from microorganisms on an industrial scale. Other extraction technologies that minimize or avoid organic solvents in the extraction process are showing promising results at the laboratory and pilot levels (i.e. pulse electric field, microwaves, ultrasonics, enzymes, etc.). However, these remain to be fully demonstrated and adapted to the industrial scale, as well as to determine their economics [81].

Most current research and development efforts to produce microbial oil, including microbial oil for biofuels, are focused mostly in strain development and optimization of upstream processing. Downstream processing has been the least investigated area, even when it may add more than 40 % of the cost to the finished microbial oil product due to high energy consumption required for this process. Hence, there is a need for developing cost-effective extraction technologies that can be scalable beyond the pilot plant level. Specifically, there is a need for technologies to both conduct oil extraction directly from wet cell biomass (not dry) and to avoid use, or at least use reduced amounts, of organic solvents, which would ease compliance with regulatory requirements for producing SCO intended for human consumption. This is a challenging task when the aim is to extract an SCO rich in omega-3 FA through a process that requires to combine high extraction yields with an effective protection against FA chemical oxidation.

Metabolic Engineering

Genetically based strain improvement is the direct enhancement of product formation or cellular properties through modification of specific biochemical reactions using recombinant DNA techniques [82]. This produces a genetically modified microorganism. There are multiple methods to go about genetic modification, including production of new products by new pathways, removing or decreasing activity of an enzyme, and amplifying genes to increase existing products [83].

Sakuradani et al. [84] did work with the 1S-4 strain of Mortierella alpina. It was made Δ-12 desaturase deficient, which increased Δ-5 and Δ-6 desaturase activities. The resulting mutant was called the JT-180 strain. Improvement of this mutant was moderate in comparison to the parent; the mutant was able to produce more ARA in comparison to the parent, 2.0 versus 1.2 g/L. Another mutant was created by the addition of a new gene, EL2, in the 1S-4 strain, determining that EL2 controls elongation of both C18 n-3 and n-6 to the C20 counter-parts. The mutant containing this mutation produced 3.6 g/L of ARA, while parent strain only had 1.9 g/L. The EL1 gene was over-expressed in a different mutant, and improvement of this strain included an 18 % increase in ARA production [84].

Another type of genetic manipulation occurs when cells are exposed to a mutagen, screened for potential enhancements, and cultivated as potentially enhanced species. This is known as classic mutagenesis with the purpose of strain improvement. Chemicals such as hydroxylamine, ethyl methane sulfonate and nitrous acid have been used to alter bases, thus mutate single-stranded DNA [85]. Two widely used mutagens are UV radiation and N-methyl-N′-nitro-N-nitrosoguanidine (MNNG). UV radiation has many effects on DNA including cyclobutane pyrimidine dimers and CC to TT mutations [86], while MNNG reacts with DNA to form alkylated bases which can mismatch and cause point mutations [85]. Lian et al. [17] applied both UV and MNNG treatments on the microorganism Schizochytrium sp. M209059. After finding optimum levels of exposure, a mutant (HX-308 M) was created. This presented a 34.84 % increase in oil production. DHA yield also almost doubled from 0.11 g of DHA per g of biomass to 0.20 g/g. Therefore, after optimizing UV radiation exposure procedure and discovering proper culturing conditions unique to each microorganism, strains with more desired properties may arise.

In another example with UV radiation, Meireles et al. [9] performed an experiment on microorganism Pavlova lutheri, which can produce EPA and DHA. UV radiation with a 254-nm lamp was used at a distance of 20 cm above the culture for 15 or 22 min. After initial radiation with UV light, there were a few mutants that displayed higher EPA and DHA levels. One of the best performing mutants was named II#2. It produced 4.38 mg of DHA per gram of dry biomass and 9.5 mg/g of EPA, in comparison to 3.51 mg/g of DHA and 5.94 mg/g of EPA for the wild type. Under optimal culturing conditions, II#2 produced an overall yield of 10.6 and 23.1 mg/g, while the wild type only produced 8 and 17.4 mg/g, of DHA and EPA, respectively.

Nannochloropsis oculata has also been subjected to random mutagenesis with the chemical N-methyl-N-nitrosourea, via selection pressure, using a known chemical inhibitor of acetyl-CoA carboxylase activity (quizalofop), to produce EPA mutants [87]. Mutants QUIZ1 and QUIZ2 produced 28.3 and 25.2 mg/g of EPA (relative to dry cells), respectively. This represented 18.4 and 5.4 % EPA increase for QUIZ1 and QUIZ 2, respectively, compared to the parent strain (23.9 mg/g EPA). Several fatty acid metabolic pathways are involved in the biosynthesis of PUFA (Fig. 4). This adds significant complexity when regulation for a specific fatty acid (i.e. EPA and DHA) is targeted. The challenge increases when a relatively unknown non-oxygen dependant pathway, such as the polyketide synthase (PKS) pathway, is involved in DHA biosynthesis. PUFA biosynthesis in thraustochytrids seems to be via the PKS pathway [88]. The PKS pathway has been demonstrated in a Schizochytrium sp. (a thraustochytrid) [89]. Nevertheless, this pathway may not be typical for all thraustochytrids and it is possible that other thraustochytrids may only have the standard fatty acid synthesis pathway.
Fig. 4

Biochemical pathways involved in microbial fatty acid biosynthesis

This particular PUFA biosynthesis was first reported as the PKS pathway in 2001, at the genetic level in both prokaryotic (Shewanella sp.) and eukaryotic (Schizochytrium sp.) cells; suggesting that in cold water ecosystems, this pathway could be a source of PUFA for fish and mammals [89]. Two significant precedents for this work on this fatty acid pathway’s potential involvement in PUFA biosynthesis were reported in an isolate of a marine bacterium closely related to Shewanella putrefaciens [90], and in psychrophilic marine bacteria [91].

Another strain improvement technique is genome shuffling. This is a method that speeds up evolution by allowing multiple parental DNA to recombine, and leads to progeny with characteristics from the parents involved [92]. Resulting organisms may not be considered genetically modified since manipulation of specific genes is not involved [92]. In classical strain improvement techniques, an organism is exposed to a mutagen of some sort, and then a screening for the mutant that best meets the initially set out criteria is conducted. Genome shuffling uses protoplast fusion which can be isolated from cells by the use of enzymes to gently remove the cell wall or puncture it. For this, two different strains are treated with an enzyme solution to obtain the protoplasts. Next, protoplasts are mixed and treated with poly(ethylene glycols) (PEG), washed, plated, then examined for mutants [93].

Zhao et al. [94] wanted to get a new fungal strain that could use a wide variety of carbon sources as well as producing more ARA. Genome shuffling between Diasporangium sp. and Aspergillus niger was used. Ability to use many carbon sources and general robustness of A. niger were some of the inherent properties used to select this microorganism. After doing 3 rounds of genome shuffling, a suitable hybrid named F1 was found [94]. F1 was a true hybrid because it had important characteristics from both of initial parental strains. F1 was able to grow on all 8 carbon sources tested: carboxymethyl-cellulose (CMC), maltose, soluble starch, lignin, d-(+)-cellobiose, corn straw, d-(+)-xylose, and d-glucose.

Aspergillus niger did the same, but Diasporangium sp. grew poorly on CMC, lignin, corn straw and xylose [94]. The hybrid also had the ability to produce ARA, which is a characteristic of Diasporangium sp. but not of A. niger. In terms of ARA, the F1 hybrid was able to produce more than the parent strain. In one example, 14.22 % of ARA was produced by the hybrid. As a comparison, only 4.95 % was produced by the parent strain. Under optimal fermentation conditions, F1 hybrid produced 13.45 g/L of biomass, which offered a 32.12 % improvement over the parent strain. The ARA yield was also greater from the hybrid with 0.81 versus 0.41 g/L from the parent strain [94]. Thus, genome shuffling may be a promising technique to enhance oleaginous species.

Other ways of using metabolic engineering include providing an organism with an ability that was not previously present. This was through insertion of DNA that originated in a different species; these are transgenic microorganisms. Yu et al. [95] engineered a cyanobacterium, Synechococcus, to produce EPA as a novel product. This was accomplished by the sub-cloning of essential open reading frames from a plasmid isolated from Shewanella sp. SCRC-2738. The new plasmid, pJRDEPA-S, was then transferred into the cyanobacterium through conjugal gene transfer [95]. Under different temperature and light intensities, these transgenic cyanobacteria produced various amounts of EPA. At a temperature of 23 °C and a light intensity of 40 lux, 1.79 mg/L of EPA were produced, while under a higher light intensity (1,000–1,500 lux), 2.24 mg/L of EPA were produced. Changing other aspects of the culture conditions also plays a significant role. When the temperature was lowered to 18 °C, under a light intensity of 800–1,000 lux with some agitation at 100 rpm, 3.86 mg/L of EPA was produced.

Saccharomyces cerevisiae has also been modified genetically to produce EPA [96]. This through heterologous recombination of five fatty acid desaturases and an elongase; Δ-5 desaturases came from Paramecium tetraurelia (a protozoan) and from the microalgae Ostreococcus tauri and Ostreococcus lucimarinus. Approximately 0.5 % EPA was produced, relative to the total FA, in an oil that does not have any EPA when it comes from the parent yeast strain.

Recently, genetic modification approaches have also been used to facilitate downstream processing, and initial research have shown promise for the future of oil extraction from microorganisms. This may involve the creation of genetically modified microorganisms that secrete FA out of cells as they are being produced, thus easing the oil extraction process. Michinaka et al. [97] created a strain of yeast that not only produces free FA, but secretes them into the extracellular fraction. The original parent strain was Saccharomyces cerevisiae, which had a reduction in the activity of acyl-CoA oxidase. These cells were exposed to ethyl methane sulfonate (EMS), a mutagen that causes base changes from G to A, which results in G-C to A-T base pair mutations. The resulting strain, B-1, was then again exposed to EMS, and the mutant YTS51 was found through the ole1 overlay assay, which was able to secrete FA into the medium at a concentration 17 times higher than the parent stain B-1 [97].

Although bacteria are also single cell microorganisms, they do not tend to accumulate edible oils [22]. Furthermore, due to regulatory matters, it may be challenging to use bacteria, especially recombinant, such as E. coli to produce edible oils. However, a genetically modified E. coli was reported for producing FA [98]. The latter through expression of plant-derived genes that encode thioesterases that prefer varied chain length fatty acyl-carrier proteins (acyl-ACP’s). The resulted recombinant bacteria had the ability to excrete FA through the membrane. A similar concept has also been demonstrated for Saccharomyces cerevisiae [99]. Oil yields for both types of modified microorganisms were low. However, these early results on cellular FA excretion could be the base for future improvements that would ease downstream processing of SCO. Consequently, this may reduce processing costs due to removal of the energy required to mechanically disrupt microbial cell walls. This, in addition to potential impacts in SCO production for high value lipids, would benefit production of lipid biofuels, where highly significant cost reductions are targeted. A major challenge for making FA cellular excretion efficient would be to minimize or remove consumption of released free FA by either the producing microbe or other microbes such as bacteria present in mixotrophic cultures.

Another important aspect within genetic improvement of strains is the need for efficient high-throughput screening methods. This in order to quickly identify significantly enhanced traits within thousand of cells that have been changed either through genetic engineering or classic mutagenesis. This has been attempted for EPA-producing marine bacteria from the genus Shewanella using reduction of the dye 2,3,5-triphenyltetrazolium chloride (TTC). Its color change, from colorless to bright red, reflects in colonies that contain significant amounts of EPA [100]. Also, a GC–MS sub-microscale in situ (SMIS) assay for quick FA analysis in small cell samples (250 μg in 100 samples per day) has the potential for an efficient and highly sensitive high-throughput screening method [101]. This could speed up the initial identification of cell lines with a significant ability of omega-3 FA production.

Chlorella is an oil-producing alga, and has a long history of being produced commercially to make dietary supplements, as this alga can be a significant source of nutrients such as carotenoids and vitamin B12 [102]. Chlorella protothecoides has even been used to make lipid-rich flour for food applications [103]. In terms of its biology, Chlorella is a relatively well known, and to date is perhaps the most viable microalga for metabolic engineering to produce oils with customized fatty acid profiles. Chlorella can grow in both autotrophic and heterotrophic conditions to produce lipids, and it has been grown heterotrophically to maximize lipid productivity; this by using an organic carbon source (i.e. sugar and glycerol) instead of CO2 fixing [104]. Regardless of any genetic modification to which Chlorella and other microbes may be subjected to, maximum theoretical sugar to lipid yields would not be overcome (see previous section on Carbon Source).

Significant research breakthroughs for microbial oil production will come from advances in metabolic engineering as the body of biological knowledge, including information on full genetic sequences, increases. In addition to E. coli and Saccharomyces cerevisiae, public full genomes for other oil-producing microbes are limited, with the exception of Chlorella [105]. Meanwhile, full genome sequencing for other oil-rich microbes such as Botryococcusbraunii is underway [106, 107]. To date, no full genome has been published for any thraustochytrids [108]. Also, specific genetic improvements, particularly for omega-3 FA production, would need to be weighed against public perception and meeting of regulatory requirements for genetically modified microorganisms.


There is a need to optimize bio-processing to reduce the costs of producing SCO rich in omega-3 FA, and producing other potential products from single cells. Advances will likely come from a combination of improvements in the areas of upstream (fermentation) and downstream (oil extraction) processing, and in metabolic engineering (Fig. 5). The current demand to produce microbial DHA and EPA as an alternative to fish oil sources has stimulated significant advances in optimization within the last 10 years. Research on SCO as a source of omega-3 FA and other potential bio-products is relatively new, and many aspects regarding their production optimization, which could result in cost savings on the industrial scale, require further investigation.
Fig. 5

Bioprocess optimization to produce SCO rich in omega-3 FA

Within upstream processing (fermentation) of heterotrophic microorganisms there is a wide scope for improvements, such as in optimizing growth and oil production of newly discovered microorganisms. Other key areas include the use of low-cost carbon and nitrogen sources, clarification of the relationship between DO and FA unsaturation, fermentation culture mode, temperature, and length of fermentation. Of course, any processing changes need to respect regulatory requirements for commercial production of edible microbial oils.

Any further breakthroughs in optimization will require an increased understanding of these microorganisms’ biology, which is currently relatively unknown, especially for thraustochytrid-like microorganisms and those that are recent environmental isolates. Their lipid biochemistry and molecular biology will play a key role in elucidating the relationship between their genetic material and key fatty acid conversions, particularly those that relate to potential enhancements for omega-3 fatty acid production. Novel molecular biology approaches may even have a significant impact, in the long term, at improving the oil extraction component of bio-processing. However, any benefits of genetic engineering improvement would need to be weighed against the risks of negative consumer perception of genetically modified organisms.

Last but not least, there is a clear need to investigate how to reduce the costs of downstream processing technologies for oil extraction. It will be a challenging task to develop a technology that not only eliminates or reduces effects on the chemical stability of omega-3 FA, but that is also economically and ecologically sustainable. Traditional extraction methods using organic solvents are often criticized for potential harmful health and environmental impacts. Thus, technological improvements must focus on both cost-effectiveness (high extraction yields at low cost) at the industrial scale and on reducing the environmental footprint of current traditional extraction processes.

Copyright information

© AOCS 2012