Introduction

Microalgae are photosynthetic microorganisms comprising a diverse group of eukaryotic and cyanobacteria members present in a wide range of environments, being highly abundant in oceans, lakes and rivers, as well as extreme environments including desert and polar regions (Lee 2008). Microalgae also have a huge ecological role, accounting for approximately 50% of the world’s organic carbon fixation (Field et al. 1998; León and Fernández 2007). Their huge diversity is only compared with their biotechnological potential. Current research is focused on exploiting diverse bioproducts made by these microorganisms, including polyunsaturated fatty acids (PUFAs, like DHA and EPA), proteins, vitamins, polysaccharides, and carotenoids with high biotechnological impact, mainly in food products (Pulz and Gross 2004; Sousa et al. 2008), biodiesel production (Dickinson et al. 2017), bioremediation (Muñoz and Guieysse 2006), and cosmetics (Enzing et al. 2012). Nowadays, the microalgae market produces nearly 9000 t of dry matter per year, with a global market size of nearly US$ 6.5 billion (Mobin and Alam 2017).

Among high-value microalgal compounds, the red pigment astaxanthin has the highest demand in the market due to its antioxidant properties. It belongs to the xanthophyll family and can be found in some microalgae, bacteria, yeasts, and marine animals (Lorenz and Cysewski 2000; Ambati et al. 2014). The unicellular freshwater microalgae Haematococcus pluvialis Fotow 1844 (recently proposed to be a synonym of Haematococcus lacustris (Girod-Chantrans) Rostafinski 1875 by Nakada and Ota 2016) is the richest source of astaxanthin, a powerful antioxidant with nutraceutical and pharmacological applications. Also, astaxanthin is key for pigmentation in farmed salmon, trout, and poultry (Sommer et al. 1992; Gutiérrez et al. 2012). In humans, astaxanthin-based dietary supplements have demonstrated reduction of several disorders, such as neurodegenerative diseases (Guerin et al. 2003), cancer (Chew et al. 1999; Tanaka et al. 2012), ocular diseases (Cort et al. 2010), skin diseases (Tominaga et al. 2012), and on the other hand, immune response enhancement (Park et al. 2010). Also, it has been reported that it acts as an anti-inflammatory agent (Speranza et al. 2012) and an inhibitor of LDL cholesterol (Iwamoto et al. 2000; Choi et al. 2011). Despite that the pigment is mostly used for aquaculture, the market of astaxanthin-based products for human consumption is constantly growing, being estimated at US$ 35–60 million in 2008 (Frost 2008). From 2011 to 2016, the global production of natural astaxanthin increased from 43,279 to 73,717 kg, with a rate of 11.24%. This shows an expected US$1.5 billion market for natural astaxanthin by 2020, mainly through the market of capsules, ingredients for cosmetics, drinks, and functional food (Nguyen 2013).

The market price of natural astaxanthin is mainly determined by its concentration and purity. Regarding H. pluvialis dry biomass, prices may range from US$2500 to 7000 kg−1, depending on its astaxanthin concentration (Bauer and Minceva 2019; Harith et al. 2020). Meanwhile, the price of astaxanthin oleoresin (usually obtained by supercritical CO2 extraction) may range from US$8000 to $15,000 kg−1, being usually commercialized with astaxanthin concentrations of 5, 7, 10, or 15% (personal communication from Astax Chile SPA). Global production is mainly concentrated in Europe, North America, China, and other Asian countries.

Despite the importance of the existing market for natural astaxanthin, over 95% of the available pigment is produced by chemical synthesis. Even though several studies have reported that the optimization of different culture parameters could reduce production costs, making the production of natural astaxanthin competitive, or even cheaper, than the synthetic product is still not possible (Li et al. 2011; Koller et al. 2014). The astaxanthin molecule has two asymmetric carbon atoms in positions 3 and 3′. Consequently, there are different optical isomers, or enantiomers: (3S,3′S); (3R,3′R) and (3R,3′S). Synthetic astaxanthin comprises a mixture of the three isomeric forms, while natural astaxanthin has optically pure (3S, 3′S)-chirality (Renstrøm et al. 1981; Grung et al. 1992), which provides stronger pigmentation in aquaculture (Østerlie et al. 1999) and also a much higher oxygen radical absorbance capacity (ORAC) when compared with synthetic astaxanthin (Naguib 1998) and the rest of the carotenoids (Chew and Park 2004; Palozza et al. 2009; Capelli et al. 2013). Moreover, synthetic astaxanthin is not approved for human consumption (Capelli and Cysewski 2012; Shah et al. 2016), nor associated with any other natural stabilizing compounds (e.g., β-carotene, lutein, and PUFAs) as found together in natural astaxanthin (Holtin et al. 2009; Schmidt et al. 2011). Striking differences between natural and artificial astaxanthin have already been reported by Capelli et al. (2013), showing that in terms of anti-oxidant properties, the synthetic compound is 20 times less effective than its natural counterpart. It also raises serious concerns about safety for human consumption, since artificial astaxanthin is considered a different chemical compound.

Given the above-mentioned background and its reported properties, consumers have already become aware of the benefits of natural astaxanthin. Therefore, optimizing the commercial production of natural astaxanthin is critical in order to supply its increasing global demand (Lorenz and Cysewski 2000; Liu et al. 2014; Shah et al. 2016).

Haematococcus pluvialis as a source of astaxanthin

Compared with other microorganisms that naturally produce astaxanthin, such as the red yeast Xanthophyllomyces dendrorhous and the microalga Chromochloris zofingiensis, H. pluvialis has the capacity to produce and accumulate significantly higher amounts of the pigment (Rise et al. 1994; Boussiba et al. 1999; Borowitzka 1999; Liu et al. 2014; Curtain 2000). Hence, H. pluvialis is considered the most promising microalgae for astaxanthin production, accumulating up to 4% of cell dry weight (DW), the highest content registered on nature (Boussiba 2000). However, these concentrations have only been reported under laboratory-controlled conditions, or by some commercial producers that report up to 6% DW (Table 1). Haematococcus pluvialis astaxanthin is currently offered as a dietary supplement, mainly in the USA and Japan (Sarada et al. 2002; Liu 2010), but it is also available in many other countries. Since this microalga has been generally recognized as safe (according to GRAS status), its commercialization should be guaranteed worldwide.

Table 1 Some industrial producers of astaxanthin derived from Haematococcus. Mean culture yields on % DW were provided when the information was available from the industries website. NS, not specified
Full size table

Haematococcus pluvialis is a common, single cell chlorophyte, distributed worldwide in temperate regions. It undergoes important morphological changes from green active cells to red immobile cells (Sun et al. 2015). When green vegetative cells are submitted to stressful environmental factors, such as high light intensity, high temperatures, external pH changes, high salt concentration, and nutritional stress, the induction of large quantities of astaxanthin occurs, changing the cellular morphology to aplanospores, a red non-motile encysted phase (Zhang et al. 2009; Shah et al. 2016). Despite its exceptional characteristics for astaxanthin production, H. pluvialis growth rates are relatively slow, compared with other microalgae. Also, it is sensitive to polluted environments and contamination by other microorganisms (Olaizola 2000).

This review is focused on the diverse biotechnological approaches made to enhance the production of astaxanthin from these microorganisms and supply its increasing global demand. For additional information about H. pluvialis’s biology and culture, an excellent general review focusing on several aspects of astaxanthin production, including current challenges for commercial scale production and global market perspectives, can be found in Shah et al. (2016).

Astaxanthin biosynthesis

Astaxanthin biosynthesis in H. pluvialis is a complex process, highly up-regulated under stress conditions, which coincides with the accumulation of triacylglycerols (TAGs). Both compounds are deposited in cytosolic lipid bodies during the “red” or encysted aplanospore stage of H. pluvialis culture.

Astaxanthin is a C40 tetraterpene, belonging to the carotenoid family. It is biosynthesized by head-to-tail condensation of two isoprene isomers: isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAPP). IPP can be originated from two dissimilar pathways: the mevalonate pathway (MVA), located in the cytosol, and the non-mevalonate pathway (DOXP), located in the chloroplast (Lichtenthaler et al. 1997; Lichtenthaler 1999; Eisenreich et al. 2001). Haematococcus pluvialis lacks three key enzymes of the MVA pathway, mevalonate kinase (MVK), phosphomevalonate kinase (PMK), and mevalonate pyrophosphate decarboxylase (MVD) (Gwak et al. 2014), but instead, it has the full set of enzymes for the DOXP pathway located inside the chloroplast. This is required for the conversion of photosynthesis-derived products into IPP.

In the following step, IPP undergoes isomerization to DMAPP. It has been long assumed that this conversion is catalyzed exclusively by isopentenyl pyrophosphate isomerase (IPI; Lichtenthaler 1999; Sun et al. 1998). However, recent transcriptomic studies suggest that no ipi genes of H. pluvialis (IPI1 and IPI2) are up-regulated during cellular accumulation of astaxanthin (Gwak et al. 2014). Different reports have suggested that another enzyme of similar activity, 4-hydroxy-3-methylbut-2-enyl diphosphate reductase (HDR), would be most likely responsible for catalyzing the interconversion between IPP and DMAPP (Hoeffler et al. 2002; Rohdich et al. 2002; Gwak et al. 2014). Further study is required to assess the contribution of these three enzymes for astaxanthin accumulation.

Elongation of the isoprenoid chain is initiated with a molecule of DMAPP and subsequent linear additions of three IPP molecules, catalyzed by geranylgeranyl pyrophosphate synthase (GGPS) (Young and Britton 1993; Cunningham and Gantt 1998). The final step of this process ends in a C20 compound, geranylgeranyl pyrophosphate (GGPP), a shared precursor among other isoprenoids. Then, phytoene synthase (PSY) catalyzes the head-to-tail condensation of two GGPP molecules to form the first tetraterpene carotenoid, phytoene (Cunningham and Gantt 1998). The PSY enzyme has been seen to be up-regulated in H. pluvialis cells stressed with high-light, undergoing transformation from green zoospores to red aplanospore stage (Steinbrenner and Linden 2001; Vidhyavathi et al. 2008; Gwak et al. 2014). Two classes of PSY have been found in green algae like Ostreococcus and Micromonas, while some other green algae like Chlamydomonas reinhardtii and Chlorella vulgaris only possess one class of PSY (Tran et al. 2009). Different PSY genes have been cloned and characterized from microalgae, including H. pluvialis (Steinbrenner and Linden 2001) and Chromochloris zofingiensis (Cordero et al. 2011). The Haematococcus PSY has a N-terminal extension similar to a chloroplast targeting sequence, indicating that PSY targets the chloroplast of H. pluvialis (Steinbrenner and Linden 2001).

Subsequently, four successive desaturation reactions of phytoene are catalyzed by phytoene desaturase (PDS) and a ζ-carotene desaturase (ZDS). In the first two reactions, PDS converts phytoene to phytofluene and ζ-carotene. ZDS catalyzes two further reactions converting ζ-carotene to neurosporene and lycopene (Cunningham and Gantt 1998). Plastid terminal oxidases PTOX1 and PTOX2 act as co-factors for electron transfer between the C40 carotenoid intermediates, plastoquinone, and the final electron acceptor, oxygen (Li et al. 2010; Nawrocki et al. 2015). The enzyme PTOX1 is co-regulated with the astaxanthin synthesis in H. pluvialis (Wang et al. 2009; Nawrocki et al. 2015). The reactions catalyzed by PDS are among the rate-limiting steps of the pathway (Chamovitz et al. 1993), and mutations in this gene may confer a higher astaxanthin synthesis capacity (Li et al. 2010). PDS activity is inhibited by the herbicide norflurazon (NF) that competes with the cofactor plastoquinone (PQ) (Breitenbach et al. 2001).

Both terminal ends of the lycopene molecule undergo cyclization by lycopene cyclases (LCY-e and LCY-b), yielding α-carotene (lutein precursor) and β-carotene (astaxanthin precursor). Cyclization is a branching point in the carotenoid biosynthesis in most organisms (Shah et al. 2016). In H. pluvialis, most carbon flux is directed to β-carotene synthesis (Borowitzka 1999), and high level of LCY-b transcripts has been observed under stress conditions (Lorenz and Cysewski 2000; Gwak et al. 2014). The final two-oxygenation steps, catalyzed by β-carotene ketolase (BKT) and β-carotene hydroxylase (CrtRb), are rate-limiting steps for astaxanthin synthesis (Vidhyavathi et al. 2008; Linden 1999; Steinbrenner and Linden 2001). Even though the reactions can proceed in any order, BKT shows a higher substrate specificity towards β-carotene when compared with zeaxanthin, favoring initial addition of a keto group before the enantio-selective hydroxylation from canthaxanthin to astaxanthin, catalyzed by CrtR-b (Lotan and Hirschberg 1995). Enantio-selectivity of astaxanthin synthesis is one of the main advantages of H. pluvialis astaxanthin over its synthetic counterpart, representing a crucial part for the nutraceutical market. Since astaxanthin has two identical chiral centers at positions 3 and 3′, it can exist in four different configurations, yielding three different isomers: (3R,3′S); (3R,3′R); (3S,3′S), depending on the spatial orientation of the hydroxyl groups in the chiral carbon. In the course of the chemical synthesis, these isomers are present in a ratio of 2:1:1, respectively, yielding only 25% of the naturally occurring (3S,3′S) isoform. Haematococcus pluvialis synthesizes the (3S,3′S) stereoisomer of astaxanthin (Grung et al. 1992), resulting in a much sought-after product in the nutraceutical market. Astaxanthin biosynthesis pathway is presented in Fig. 1.

Fig. 1
figure 1

Astaxanthin biosynthetic pathways on H. pluvialis. IPI, isopentenyl pyrophosphate isomerase; HDR, 4-hydroxy-3-methylbut-2-enyl diphosphate reductase; GGPS, geranylgeranyl pyrophosphate synthase; PSY, phytoene synthase; PDS, phytoene desaturase; ZDS, ζ-carotene desaturase; LCY-b, lycopene β-cyclase; LCY-e, lycopene ε-cyclase; BKT, β-carotene ketolase; CrtR-b, β-carotene 3,3′-hydroxylase

Full size image

Culture systems and harvesting process

Commercial production of astaxanthin from H. pluvialis is currently focused on a two-stage strategy. The first stage consists in the production of vegetative biomass under favorable conditions, achieving rapid growth at high cell densities. Subsequently, a second stage is carried out where astaxanthin biosynthesis is promoted by applying stressful media changes like salt addition, temperature rises, increases in incident light and CO2 addition (Christian et al. 2018). Nitrate starvation is also an effective method to increase the accumulation of astaxanthin. Recent studies have focused on intracellular nitrate starvation by using nitrate reductase (NR) inhibitors like Na2WO4, instead of changing the C/N ratio to create a nitrogen deficient culture media (Li et al. 2019). Furthermore, the addition of oxidizing substances can slightly induce oxidative stress and increase cellular carotenoid content (Guedes et al. 2011a, 2011b). These changes lead to the formation of H. pluvialis cysts, astaxanthin-rich cells that are easily harvested by passive settling and centrifugation. Finally, cell biomass is mechanically disrupted to release astaxanthin which is additionally dried. The final product can be encapsulated or subjected to astaxanthin extraction for pure nutraceutical formulations (Lorenz and Cysewski 2000; Guerin et al. 2003). Alternatively, Haematococcus astaxanthin can be produced indoors under mixotrophical conditions adding organic carbon sources like acetate or glycerol for astaxanthin induction (He et al. 2018; Wen et al. 2020; Zhang et al. 2020), obtaining higher biomass concentration, and therefore higher astaxanthin yields. Nevertheless, production costs and contamination risks when applying these strategies are higher. Recently, new approaches like the addition of melatonin, specific phytohormones, or even electrical treatments to H. pluvialis culture have demonstrated to enhance astaxanthin accumulation; however, none of them have been validated at industrial scale (Ding et al. 2018, 2019; Kim et al. 2018). Currently, the two most applied technologies for H. pluvialis astaxanthin production are open pond systems and closed photo-bioreactors (PBRs) (Gong and Bassi 2016). Cultivation in open systems requires less initial investment and maintenance, as set-up is cheaper. However, uncontrolled environmental conditions usually lead to low production and culture contamination. Closed systems have greater potential for optimization, higher mass transfer, higher production, and less contamination since they count with controlled operational conditions. All these advantages are associated to higher costs and complex set-ups (Singh and Sharma 2012; Borowitzka 2013; Acién et al. 2017; Moreno-Garcia et al. 2017). Currently, most PBR technologies are available for large-scale production by companies established in Europe and all over the world (Acién et al. 2017). Accordingly, the current industrial production relies only on open or closed systems as shown in Table 1.

Zhang et al. (2014) developed an alternative cultivation approach using an immobilized film method. However, the cultivation was not as efficient as the suspended algal cultures in PBRs due to poor light penetration (Zhang et al. 2016). Recently, Kiperstok et al. (2017) reported the possibility of continuous co-production of biomass and astaxanthin in a one-phase process at high light intensities up to 1015 μmol photons m−2 s−2. This method is still under research and development, and currently is not used at industrial scale.

Biotechnological alternatives for astaxanthin production improvement

Since the major natural source of astaxanthin is H. pluvialis, a variety of strains have been isolated around the globe. The maximum yield reached in photobioreactors at industrial scale production is 5–6% DW and a maximum of 2% DW in 200,000-L open pond raceway systems, as informed by some companies (Table 1). Nevertheless, there is wide variation among different natural strains, with some of them reported as hyperproducers. For example, Chekanov et al. (2014) described an artic strain able to accumulate up to 3–5.5% astaxanthin under prolonged stress conditions. Later, Wang et al. (2019) described a newly discovered strain (JNU35), which hyperaccumulated up to 5.6% DW astaxanthin after complex culture optimization. Notwithstanding the existence of natural strains, the need for higher yields has motivated research aimed to enhance H. pluvialis natural capacity to produce astaxanthin.

Classical Mutagenesis

This is the simplest method regularly used to obtain different varieties of microorganisms, enhancing particular characteristics. This technique induces random changes on the genetic material by applying chemical mutagen agents or UV radiation. Additionally, plasma exposure has been successfully used on this microalga. Classic mutagenesis can be broadly divided into two main steps: (i) exposure of cells to mutagenic agents and (ii) isolation of mutant cells with increased astaxanthin contents.

Methods for obtaining astaxanthin-overproducer mutants

In most studies, variables such as the time of exposure and concentration of different mutagen agents are adjusted in order to obtain high mortality rates, in the range of 85–95%, originating potentially mutant individuals that are selected later using several approaches.

Among chemical mutagens, most of H. pluvialis available literature describes the application of ethyl methanesulfonate (EMS), either in isolation or in combination with UV radiation exposure in order to obtain mutants with enhanced astaxanthin production (Tjahjono et al. 1994; Sun et al. 1998; Wang et al. 2016). The effectiveness of this compound to generate overproducers varies largely. The first report, described by Tjahjono et al. (1994), resulted only in a modest increase of 6% with respect to the untreated strain. Since then, several overproducers have reached more than twice the astaxanthin yield of their original strains (data summarized on Table 2).

Table 2 Astaxanthin yields reported after the modification of Haematococcus pluvialis using different mutagenesis strategies. Yields (in mg/g DW) of the base strain and the best mutant strain(s) are reported when available
Full size table

The mutagens N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) (Sandesh Kamath et al. 2008; Hu et al. 2008) or diethyl sulfate (DES) (Wang et al. 2016) are also employed with different dosages and exposure periods. UV radiation is also very effective at inducing changes in DNA, even when there is an intrinsic capacity of the microalgae to limit this damage (Shah et al. 2016). Revertant mutants can be easily avoided by including a period of dark incubation in order to evade the photoreactivation driven by light (Essen and Klar 2006). This method has been successfully applied by several authors (Chumpolkulwong et al. 1997; Sandesh Kamath et al. 2008) in order to obtain astaxanthin-enhanced strains. The combination of UV and chemical treatment has also been explored, with better astaxanthin yields obtained by subjecting the previously UV mutated cells to EMS treatment, reaching 2.5% DW (Chen et al. 2003), while more recent research have reported 4.7% DW (Wang et al. 2016). A third alternative consists in the exposure of cells to low-temperature plasma by atmospheric pressure dielectric barrier discharge plasma (DBD), a technique that produces energetic electrons and ions, reactive species, and UV radiation, used to induce gene mutations. Using this method, 3.3% DW astaxanthin yield has been achieved, with nearly 40% higher growth rates under high light irradiation (Liu et al. 2016). Later, the same group demonstrated that this mutant has higher astaxanthin accumulation due to higher photosynthetic activity, increased non-photosynthetic pigments protecting photooxidative stress, and better efficiency of CO2 utilization (Chen et al. 2020). The last alternative to induce genetic mutations is to exploit the high energy and penetration capabilities of γ rays. Using this technique, 2.4 times more astaxanthin has been demonstrated when applying γ-rays to an H. pluvialis strain, which also showed higher growth rates under high CO2 stress when compared with the unmodified strain (Cheng et al. 2016).

Methods for screening overproducing strains

After exposure to mutagenic agents, successful isolation of mutants relies on their capacity to survive and grow in the presence of compounds at inhibitory concentrations, either by color changes of the mutants, or capacity to survive under different light conditions. It is expected that surviving strains will have an increased capacity to synthesize carotenoids and, consequently, produce more astaxanthin.

Herbicide-based mutant selection

On many cases, diphenylamine (DPA) has been used as an indicator for mutant selection. DPA is an herbicide that interferes with the carotenogenic pathway by inhibiting β-carotene oxygenase and β-carotene hydroxylase, blocking cell capacity to produce astaxanthin under stress conditions (Fig. 2). Colonies capable of synthesizing astaxanthin will stay as red colonies. They represent candidate pigment overproducers or DPA tolerant strains, as opposed to wild-type cells, which will remain green under these conditions (Tripathi et al. 2001; Chen et al. 2003; Gómez et al. 2013; Wang et al. 2016).

Fig. 2
figure 2

Culture aspect of WT and a chemically induced mutant MT derived from the same strain a under favorable growth conditions and b under stress conditions, where a higher astaxanthin level is reflected by its darker red color. c WT strain reached ca 2% astaxanthin DW, while the mutant reached almost double concentration, as can be evidenced by the methanol extracts absorption spectra. Reprinted from enhanced protection against oxidative stress in an astaxanthin-overproduction Haematococcus mutant (Chlorophyceae), Zhengyu Hu, Yantao Li et al. European Journal of Phycology 43(4), Nov 1, 2008. © 2008 British Phycological Society, reprinted by permission of Informa UK Limited, trading as Taylor & Francis Group, www.tandfonline.com on behalf of British Phycological Society

Full size image

The concentration required for the selection varies depending on each experiment and the original strain used. Tripathi et al. (2001) used 15 μM DPA and a 1-stage mutagenesis scheme, yielding a strain with nearly 1% astaxanthin (DW), 2.46 times higher than the wild-type strain. In comparison, Wang et al. (2016) used 71 μM DPA and a three-stage mutagenesis breeding, isolating a colony with 4.7% astaxanthin (DW), 1.7 times higher than the base strain, and almost the highest astaxanthin concentration reported for any H. pluvialis mutant (Table 2).

Even when Tripathi et al. (2001) obtained a strain that nearly triplicated the yields of the base strain, its productivity was much lower than the strains obtained by Wang et al. (2016), emphasizing the requirement of a previous selection step of high yield base strains with commercial utility. Furthermore, the use of DPA does not always guarantee the selection of overproducers as reported by some groups, who relied on other selection agents to obtain overproducers (Chen et al. 2003; Sandesh Kamath et al. 2008). Differences on the capacity to select enhanced cells could be attributed to different susceptibilities of unrelated strains. Nevertheless, Chen et al. (2003) and Wang et al. (2016) used the same strain, 712, but reported different results. Chen does not report any remarkable strain selected using this herbicide, while the latter was able to obtain high yield overproducers by using DPA at higher concentrations and a multi-stage selection system. Therefore, culture conditions and selection schemes are very important for successful overproducer isolation.

Another herbicide, glufosinate, has strong effects on plants, increasing ammonia levels and halting photosynthesis. It has also been reported to inhibit cell growth and induce astaxanthin accumulation on H. pluvialis cells (Aflalo et al. 1999). It has been used as a replacement of DPA for the isolation of overproducing strains, reaching a maximum of 3.9% astaxanthin DW (Sandesh Kamath et al. 2008). However, not every herbicide is suitable to successfully select mutants, as demonstrated in compounds like norflurazon and fluridone, which were tested on early reports and are no longer used (Tjahjono et al. 1994; Chen et al. 2003).

Non-herbicide mutant selection

The selection under different herbicides is not the only option to isolate overproducing strains. For instance, nicotine, an inhibitor of lycopene cyclase that avoids astaxanthin formation has also been used by Tripathi et al. (2001) and Chen et al. (2003) to select resistant strains with enhanced astaxanthin production. Using nicotine, Tjahjono et al. (1994) were able to obtain resistant, but not overproducer strains. Apparently, the difference in nicotine concentrations (100 μM vs 200 μM) could be responsible for the negative results (Chen et al. 2003). Compactin, an inhibitor of isoprenoid biosynthesis and consequently carotenoids, was used to isolate strains that doubled the astaxanthin content of original strains (Chumpolkulwong et al. 1997). The same results were replicated later, obtaining a strain that doubles the yield of the base strain (Tripathi et al. 2001). Carotenoid biosynthesis target sites for these compounds are presented in Fig. 1. Other researchers completely discarded the use of inhibitors by using a high stress condition of blue and white light to select strains that accumulate twice as much carotenoids than the base strain. These selected strains have a brown color on the green background of non-over carotenoid producers (Sun et al. 1998). Following a similar visual selection, mutagenized cells cultivated in liquid media, followed by light stress with sodium acetate and ferrous sulfate additions, resulted in strains that nearly doubled the astaxanthin production, as demonstrated by light microscopy and reverse phase high-performance liquid chromatography (Hu et al. 2008) (Fig. 2). Differential growth of partially deficient photosynthetic mutants under heterotrophic and autotrophic conditions has also been exploited to select highly photosensitive mutants with increased astaxanthin accumulation (Hong et al. 2012) (Fig. 3). The selection of cells that present altered growth rates after mutagenesis could be achieved by direct cell counting and optical density determination on microplates, as an alternative screening method to isolate overproducer cells (Liu et al. 2016). Even when the later authors did use DPA and nicotine to determine effects on photosynthesis, the inhibitors were not used for the mutant screening. In Hong et al. 2018, Hong et al. reported a new technique for colorimetric selection of H. pluvialis mutants using azide, a chemical that acted as an excellent astaxanthin inducer, followed by an oil-based astaxanthin quantification, offering a high-throughput screening strategy without complex pretreatment processes.

Fig. 3
figure 3

General scheme for the isolation of highly photo-inducible, partially photosynthesis-deficient and photosensitive H. pluvialis mutants, which presented a lower photosystem II efficiency but higher astaxanthin productivity than WT cells. EMS-treated cells were grown sequentially in heterotrophic, autotrophic, mixotrophic, autotrophic, and inductive autotrophic conditions. During culture partially photosynthesis-deficient (PP), partially photosynthesis-deficient and photosensitive (PP–PS), and highly photo-inducible PP–PS mutants were isolated. Reprinted from Process Biochemistry, 47(12), Min-Eui Hong, Seung Phill Choi, Youn-Il Park, Young-Kee Kim, Won Seok Chang, Byung Woo Kim, Sang Jun Sim, Astaxanthin production by a highly photosensitive Haematococcus mutant, pp. 1972–1979, © 2012, with permission from Elsevier

Full size image

Genetic engineering

Some microalgae offer a great potential for genetic engineering in terms of production of recombinant proteins as well as high value compounds like carotenoids or PUFAs. Despite the breakthroughs done with the microalga Chlamydomonas reinhardtii as the model organism for doing genetic engineering, few efforts have been done in general with other microalgae. During the last decade, H. pluvialis has been a target for genetic engineering either by nuclear or chloroplast transformation (Steinbrenner and Sandmann 2006; Gutiérrez et al. 2012). Recently, Qiulan et al. (2018) published the first draft genome of H. pluvialis, a valuable genetic resource for understanding the genetic basis of astaxanthin production.

Nuclear transformation

Transformation experiments were mostly adopted from a biolistic approach. Teng et al. (2002) reported that transient expression of the lacZ gene could be observed in motile cells but not in non-motile cells. Later, the activity of the endogenous β-carotene ketolase (BKT) promoter was studied by transiently expressing the lacZ gene on H. pluvialis (Meng et al. 2005, Fig. 4). An alternative transformation method has been used through Agrobacterium (Kathiresan and Sarada 2009; Kathiresan et al. 2009) but was not repeated successfully by others researchers (Fig. 5).

Fig. 4
figure 4

Early proof of transient transformation on H. pluvialis, green-blue cells are stained by β-galactosidase activity derived by lacZ exogenous gene. a lacZ expression under control of H. pluvialis bkt promoter region. b lacZ expression under control of SV40 promoter-enhancer. c Negative control showing no β-galactosidase activity. Figure from Meng, Chun-Xiao; Teng, Chang-Ying, Cloning and Characterization of β-Carotene Ketolase Gene Promoter in Haematococcus pluvialis, Acta Biochimica et Biophysica Sinica, 2005, 37(4), 270–275, by permission of Oxford University Press

Full size image
Fig. 5
figure 5

Illustration showing the H. pluvialis transformation mediated by Agrobacterium tumefaciens, a common natural vector widely used for plant transformation. The co-cultivation step brings both cell types in close proximity, where previously transformed Agrobacterium transfers the genetic construct to the microalgae. Further selection of transformed strains is performed on antibiotic containing media, with resistance provided by the genetic construct previously transferred. TAP, Tris acetate phosphate media. Reprinted by permission from Springer Nature, Journal of Applied Phycology 21, Towards genetic improvement of commercially important microalga Haematococcus pluvialis for biotech applications, Shanmugam Kathiresan et al., © 2009

Full size image

Nowadays, an advanced shuttle-vector has been developed for efficient nuclear transformation based on an endogenous dominant marker that allows the production of safe transgenic microalgae strains without foreign DNA sequences (Sharon-Gojman et al. 2015). All previous research proved the possibility to genetically enhance H. pluvialis; however, they were not focused on astaxanthin enhancement. The first effort to improve astaxanthin accumulation by nuclear transformation was carried out with a mutated phytoene desaturase gene (pds) that confers resistance to norflurazon (Steinbrenner and Sandmann 2006), showing 26% higher astaxanthin levels after 48 h of high light exposure as compared with the wild type. Additionally, cloning an extra copy of the bkt gene directly into the nucleus allowed a four-fold increase in astaxanthin content under stress conditions (Kathiresan et al. 2015). Recently, the modification of H. pluvialis trophic capabilities has been achieved by nuclear transformation with the hexose uptake protein (HUP1) gene of the green microalga Parachlorella kessleri. The transformed H. pluvialis strains proliferate under very low illumination on nutrient media supplemented with glucose, and produce astaxanthin in total darkness under special conditions, or with subsequent exposure to increasing illumination levels, providing a new way to supply nutrients for astaxanthin development on low-light conditions (Waissman-Levy et al. 2019).

Although most genes and enzymes for algae-specific carotenogenic pathways are still unknown, a number of them from H. pluvialis have been cloned and sequenced (Takaichi 2011). A summary of different transformation efforts and their respective efficiencies is presented in Table 3.

Table 3 Summary of different genetic transformation approaches reported to date for H. pluvialis. Transformation efficiency is provided when available
Full size table

Chloroplast transformation

Since most of the carotenogenic pathway occurs in the chloroplast of green algae, the optimization of carotenoid production can be undertaken by metabolic engineering, either by the expression of genes directly inserted in the chloroplast or by nuclear expression followed by gene product translocation to the chloroplast. Expression of genes from the chloroplast genome guarantees the expression in the right place and is undoubtedly the best strategy in terms of protein accumulation since it allows control of exogenous DNA insertion. On the other hand, nuclear expression is associated with silencing and positional effects that could lead to lower protein expression (Pérez-España et al. 2016; Henríquez et al. 2016; Siddiqui et al. 2019). By using next generation sequencing, Bauman et al. (2018) revealed an extremely large 1.35-megabase chloroplast genome for H. pluvialis, reported as the largest assembled chloroplast genome of any plant or alga known to date. Following the successful chloroplast transformation of H. pluvialis driven by the endogenous rbcL promoter (Gutiérrez et al. 2012, Fig. 6), the metabolic engineering study presented by Galarza et al. (2018) overexpressed the endogenous pds gene in order to achieve a higher astaxanthin accumulation. After transformation, astaxanthin accumulation per dry weight was enhanced up to 67%.

Fig. 6
figure 6

Schematic presentation of an integration/expression vector. Transgene integration occurs by homologous recombination between flanking regions of chloroplast genome and vector. GOI, gene of interest. Crossed lines represent homologous recombination

Full size image

Polyploidy

Artificial induction of polyploidy is an interesting alternative for microalgae improvement, which may hold an enormous potential for increasing metabolite productivity. However, only a few studies have been carried out in this field. Ploidy levels refer to the number of chromosome sets present in a cell or individual. Considering “n” as the basic number, two sets are present on a diploid state (2n = 2x). The polyploidy state is the heritable condition, where more than two chromosome sets are present on a cell (Comai 2005). This is a very common condition in higher plants, with 30 to 80% of modern angiosperm plants originating from a polyploid ancestor or showing some degree of polyploidy (Meyers and Levin 2006). In polyploids, the increased number of chromosomes cause changes on physiology, phenotype, and fertility (Levin 1983), with enlarged cell size and bigger nuclei, chloroplast, and mitochondria (Stebbins 1970; Levin 1983). Additionally, the increase on cell size is reflected on the presence of larger tissues and organs, a phenomenon known as the Gigas effect (Morgan et al. 2003).

Artificial polyploidy induction of microalgae

Most of the artificially polyploidized microalgae reports are directed to genetic and cytological studies. In Chlamydomonas eugametos and C. reinhardtii, colchicine was used to study inheritance mechanisms, producing diploids and higher ploidy cells from haploid strains (Wetherell and Krauss 1956). In Chlorella sp., colchicine-derived polyploids showed an increased number of nuclei and endoplasmic reticulum (Wanka 1968). Polyploidy was also present in the cyanobacterium Anacystis nidulans (Griese et al. 2011). In this species, individual polyploids (4n) showed a higher growth rate, and consequently a potential increase in productivity, when compared with the original strains (Vonshak and Richmond 1981). More recently, the anti-mitotic agent demecolcine, closely related to colchicine, was used to induce diploidy in C. reinhardtii. This resulted in bigger cells and enhanced abiotic stress resistance while being more productive than their haploid counterpart under these conditions, demonstrating an enhancement of the polyploid strains (Kwak et al. 2017). Concerning carotenogenesis in H. pluvialis, increasing the copy number of genes is expected to also increase the astaxanthin content (Lee and Zhang 1999). Latest research indicate that Haematococcus presents a cell cycle that includes multiple DNA replication stages, being effectively polynuclear and polyploid in its latter aplanospore stage, but recovering its individual diploid stage at the end of multiple cytokinesis that produces flagellate zoospores (Reinecke et al. 2018).

Nonetheless, the only available reference mentioning polyploidization for increasing productivity in this species reported an increased growth rate, but a lower astaxanthin production per cell. The study was only focused on protein expression profiles, not giving further details about the methodology used to induce polyploidy (Kim 2006). However, it represents an early attempt at using polyploidy for Haematococcus improvement. In a more recent attempt to enhance astaxanthin content, our group isolated one polyploidized strain of H. pluvialis that had bigger aplanospores and nearly 50% more astaxanthin accumulated in DW than its original strain under laboratory conditions and semi-industrial set-ups (Fig. 7, unpublished data). This provided further evidence of the utility of this approach.

Fig. 7
figure 7

Light micrographs of H. pluvialis base strain (left) and polyploid strain (right) derived from the same base strain subjected to colchicine treatment, showing an increase in the cell size and ploidy levels representative of the gigas effect from polyploids (right). Bar represents 25 μm. Both images are taken from late aplanospore stage and show astaxanthin accumulation. Images from unpublished results from the authors

Full size image

Artificial polyploidy induction with protoplast fusion

Alternative methods to generate polyploidy without using antimitotic agents are available. The cell wall of flagellated Haematococcus can be digested by proteinase K treatment for protoplast preparation and it is possible to regenerate the cell wall from these cells (Tjahjono et al. 1994). Under appropriate conditions, these protoplasts can be fused, resulting in strains that present cells with twice as much volume and with three times higher antioxidant yields than the parental lines, clearly showing the Gigas effect, along with an enhancement of astaxanthin production (Tjahjono et al. 1994). Furthermore, H. pluvialis protoplasts can be fused with the chrysophyte Ochromonas danica, resulting in individuals with a mixture of metabolic capacities, as evidenced by their fatty acid profiles. They combine lipids only present in Haematococcus (C16:0 and C18:3n-6) or O. danica (C16:2 and C24:0). This combination also takes place in the interspecies fusants that show both fatty acid profiles, and combined growth abilities on a specific media, revealing genetic recombination between the two microalgal phyla. These heterofusants showed the presence of two unfused nuclei, lacking the capacity to produce astaxanthin and presenting an always-green color (Abomohra et al. 2016, Fig. 8). In this case, a heterokaryotic cell was obtained instead of a true allopolyploid, since there are two different nuclei on each cell. A graphic comparison of these methods of modification is shown in Fig. 9. The advantages and drawbacks of biotechnological approaches for astaxanthin yield improvement in H. pluvialis are presented on Table 4.

Fig. 8
figure 8

Transmission electron micrography of a fusant obtained between Haematococcus pluvialis and Ochromonas danica, showing the two unfused nuclei present on each cell indicated by arrows. Figure reproduced from Protoplast fusion and genetic recombination between Ochromonas danica (Chrysophyta) and Haematococcus pluvialis (Chlorophyta), Abd El-Fatah Abomohra, Mostafa El-Sheekh, et al., Phycologia 55(1), 2016, reprinted by permission of the publisher (Taylor & Francis Ltd., http://www.tandfonline.com)

Full size image
Fig. 9
figure 9

Comparison of the three methods for genetic improvement on Haematococcus pluvialis. a Modification inducer, including UV radiation or chemicals usually employed for random mutagenesis, DNA vectors with gene construct for genetic transformation, and chemical polyploidization agent, respectively. b DNA change resulting on each case, random mutation, DNA integration, or whole DNA duplication, respectively. c Final effects of modification, mutant cells with distinctive morphology or growth and pigment accumulation, modified organisms containing exogenous or rearranged DNA, and bigger cells with higher pigment production, respectively

Full size image
Table 4 A summary of advantages and disadvantages when using different biotechnological approaches for Haematococcus improvement
Full size table

Alternatives to Haematococcus pluvialis for astaxanthin production

Even though H. pluvialis is the most efficient producer of natural astaxanthin, its culture presents several problems, such as slow growth rate and low biomass when compared with other microalgae (Lorenz and Cysewski 2000; Olaizola 2000). Chromochloris zofingiensis can be used as an alternative producer since it has faster growth rate. Moreover, it can grow under heterotrophic conditions using glucose as energy source (Sun et al. 2008; Liu et al. 2014). Nevertheless, C. zofingiensis shows very limited astaxanthin productivity when compared with H. pluvialis. For example, the less productive strain of H. pluvialis, CCAP34/8, produces 1.1% astaxanthin DW (77 mg L−1), while the best reported C. zofingiensis strain (CCAP211/14), produces 0.37% astaxanthin DW and 25 mg L−1, both under photoautotrophic batch culture conditions (Del Campo et al. 2004; García-Malea López et al. 2006). Additionally, astaxanthin from C. zofingiensis consists in a mixture with 50% of other keto-carotenoids, not allowed for human consumption in USA, Europe and Australia (Liu et al. 2014). Furthermore, it is not recognized as GRAS (Yang et al. 2016), hindering its use for human consumption. A second alternative consists in the heterologous expression of the carotenogenic pathway in other hosts, since there is wide knowledge of the genes involved in astaxanthin biosynthesis. However, a highly engineered strain of Escherichia coli, BW-ASTA, only yielded 0.14% DW astaxanthin and required IPTG, an expensive inducer, hindering the commercial mass-scale production in this heterologous system (Lemuth et al. 2011).

Conclusion and perspectives

Even when H. pluvialis is the best natural source of astaxanthin known to date, its productivity still requires improvement. This challenge has been addressed from different biotechnological approaches. Classical mutagenesis has well-established protocols and is easy to perform but presents some drawbacks, since it is time-consuming, and its results are inherently unpredictable. Due to its random nature, there is no control over random changes at a genetic level, probably presenting multiple mutations that may have impact on growth and reproduction stages, including non-flagellated or unstable cells (Gómez et al. 2013), revertant mutants from DES or UV-treated cells (Wang et al. 2016), and UV and EMS mutants loss under continuous selection (Chumpolkulwong et al. 1997). The aforementioned data indicates that even though random mutagenesis is simple to perform, the resulting strains must be carefully tested to ensure they maintain the improved properties. Additionally, it is important to use a strain with the highest astaxanthin accumulation capability as a base for mutagenesis, since several studies that show increments that doubled the initial strain productivity are still behind the results of non-improved highly productive strains. As it is mentioned in the reported work of Tripathi et al. (2001), a 225% increase on astaxanthin yield was reached, but the original H. pluvialis strain only produced 0.43% on DW, well behind the 6% yields reported by some industries for commercial use. Additionally, the reported increase was achieved in mixotrophic conditions, further increasing the costs under a future industrial set-up. In absolute terms, several strains obtained by mutagenesis, presented in Table 2, displayed yields very close or higher than 4% DW, but since there is no information available in literature, these strains are not currently used for industrial production. However, it is still possible that some mutant strain is used under industrial secret, which we are not aware of.

Genetic engineering of carotenoid metabolic pathways has been employed in several species to increase yields, but presents specific challenges, such as detailed knowledge of carotenoid biosynthesis pathways, availability of nuclear and chloroplast genomes, and enzyme cellular localization knowledge. Even when H. pluvialis has been transformed by different techniques, there are only a few examples of astaxanthin enhancement in this microalga. In the first case, the use of a modified version of the pds gene for nuclear transformation, conferring resistance to a bleaching agent, increased the astaxanthin content of transformed strains (Steinbrenner and Sandmann 2006). Additionally, the overexpression of the bkt gene by nuclear transformation enhanced pigment production, as demonstrated by Kathiresan et al. (2015), and astaxanthin yield, by the insertion of the endogenous nuclear pds gene into the chloroplast (Galarza et al. 2018). On the other hand, Waissmann-Levy et al. (2019) stably integrated the gene HUP1 from P. kessleri into H. pluvialis, allowing hexose uptake and the capacity to survive with minimal or no light at all when growing in glucose-supplemented media. As the preceding examples demonstrate, both nuclear and chloroplast genomes of H. pluvialis have been genetically manipulated by either biolistic approaches or Agrobacterium mediated transformation, with variable transformation efficiencies (Table 3). Recently, the release of both chloroplast and nuclear genomes for H. pluvialis opened new opportunities for research and genetic engineering of this microalga. Only the chloroplast genome of this microalga is 1.35 Mb in size, being the largest assembled chloroplast to date (Bauman et al. 2018). Also, their nuclear genome is huge, reported to be 669 Mb and containing 18.545 genes in the first draft of Luo et al. (2019), one of the biggest sizes in currently sequenced green algae. The presence of six copies of the bkt gene in H. pluvialis nuclear genome, while only one copy is present in close relative species such as Volvox carteri or C. reinhardtii, might contribute to the high productivity of astaxanthin in this species (Luo et al. 2019). The three types of bkt genes in H. pluvialis, bkt1, bkt2, and bkt3, are isoform copies. According to the latest genome assembly of this species by Luo et al. (2019), bkt experienced multiple gene duplications during genome evolution, with one group containing bkt1, while the other group contains bkt2 and bkt3. The bkt gene family expansion may contribute to the high astaxanthin yield of H. pluvialis and, along with transcriptomic data, reinforce the usefulness of these resources for Haematococcus enhancement. In addition, other researchers have developed specific expression vectors for either nuclear or chloroplast transformation (Gutiérrez et al. 2012; Sharon-Gojman et al. 2015) that may be useful to enhance astaxanthin accumulation. Finally, the Crispr/Cas9 system could also be adapted for gene editing in H. pluvialis to achieve further improvement of carotenoids (Nymark et al. 2016). Despite transformation is an excellent tool for H. pluvialis improvement, several restrictions still exist for the commercialization of natural products derived from genetically engineered organisms (GMO) in markets aimed to human consumption, since some consumers may have concerns about the safety of GMO-derived products. For this reason, additional regulations may be required for the introduction of GMO microalgae into global markets (Beacham et al. 2017). These restrictions need to be addressed before attempting a commercial use for genetically engineered strains.

Artificial polyploidization techniques have been used to enhance agricultural traits for several decades in superior plants, resulting in important gains due to the Gigas effect, leading to bigger cells and higher amounts of secondary metabolites. Despite these benefits, artificial polyploidization of microalgae has only a few appearances in literature. H. pluvialis is regarded as diploid in its zoospore stage, but it presents an unusual cell cycle with octaploid and even higher ploidy levels before reaching cytokinesis (Reinecke et al. 2018), offering an excellent starting material for a stable fixing of these high ploidy levels by means of artificial polyploidy induction. But the performance of Haematococcus cells with stable high ploidy levels are unknown and can be low, since it has been described that cells possessing higher DNA loads have a slower growth (possibly due to the requirement of longer times on the DNA synthesis and mitosis stages of the cell cycle; Doyle and Coate 2019). In the case of H. pluvialis treated with colchicine, the evidence points out the utility of this approach, with cells presenting higher growth rates, but lower astaxanthin accumulation and differential protein expression profile that could explain this discrepancy (Kim 2006). In the same context, our group has successfully obtained polyploidized strains that showed bigger cells and more astaxanthin at laboratory and industrial conditions (deposited on BEA repository and protected by Chilean and international patents, unpublished results). Protoplast fusion techniques also appeared to be useful for H. pluvialis astaxanthin enhancement, resulting in a strain with three-fold higher carotenoid accumulation under Fe2+ stress conditions, and 30% higher production of this antioxidant than the original wild type cells, notably surpassing the yields from the mutant strains used to make the fusants (Tjahjono et al. 1994). Additionally, the inter-species fusants could hold great potential, not only to obtain better astaxanthin yields or enhanced growth characteristics, but also to “assemble” novel metabolic pathways in Haematococcus fusants, mixing the biochemical capabilities of different microalgae to obtain novel strains. This was demonstrated by Abomohra et al. (2016), who attempted to produce new strains that combine fatty acids, with potential pharmaceutical use, with astaxanthin production. Even when this research failed to obtain this objective, it demonstrated the combination of different fatty acid profiles into one new organism. Therefore, it may be possible to use this strategy to mix different metabolic pathways of important commercial value for biorefineries, especially where the production of several high value chemicals is desirable to reduce costs and maximize the benefits of the cultures.

Most of the literature cited shows researchers who evaluate the performance of the modified strains under laboratory conditions, with culture volumes typically under 500 mL, very uniform illumination and temperature, and active aeration by shaking or air bubbling. These uniform growth conditions are seldom encountered in industrial set-ups, with wide variation in light intensities, reaching cells that are near the surface in open-ponds, or near artificial light sources in photobioreactors. The same is true for temperature variations that can affect the growth of Haematococcus cells. Additionally, the culture media used in laboratory-scale experiments tends to be very different from the ones used industrially. Usually, the purity of reagents varies for industrial conditions and will be of lower quality for cost reasons, with several impurities that could hamper the cell growth and hinder the yields obtained at a laboratory scale.

As a final remark, the continued enhancement of H. pluvialis astaxanthin aims for competitive industrial production. It may be achieved by more than one of the biotechnological alternatives described in this review, starting with the best naturally occurring strains, to generate overproducing cells that can grow better in current industrial production systems, not requiring expensive chemical supplies or modification of the existing facilities. Furthermore, the ability to produce additional high-value compounds in addition to astaxanthin, by the integration of novel metabolic pathways by genetic engineering, will allow the integration into a biorefinery set-up, enhancing even more the economic value of novel H. pluvialis strains.