Enhancing carotenoid production in Rhodotorula mucilaginosa KC8 by combining mutation and metabolic engineering

Abstract

Rhodotorula mucilaginosa has been considered as a potential industrial yeast due to its unicellular and fast-growing characteristics, and its ability to produce carotenoids, including torularhodin. However, its low total carotenoid production limits its commercial application. In this study, mutation breeding and metabolic engineering were employed to enhance carotenoid production in the R. mucilaginosa strain KC8. After chemical–physical mutagenesis, R. mucilaginosa K4 with a 67% greater concentration of carotenoids (14.47 ± 0.06 mg L⁻¹) than R. mucilaginosa KC8 (8.67 ± 0.07 mg L⁻¹) was obtained. To further enhance carotenoid production, gene HMG1 encoding the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase was introduced from another yeast, Saccharomyces cerevisiae, and overexpressed in R. mucilaginosa K4. The carotenoid production of HMG1-gene-overexpression transformant G1 reached 16.98 mg L⁻¹. To relieve the feedback inhibition of ergosterol, and to down-regulate ergosterol synthesis, ketoconazole, an ergosterol synthesis inhibitor, was added at a concentration of 28 mg L⁻¹. The carotenoid production of the transformant G1 reached 19.14 ± 0.09 mg L⁻¹, which was 121% higher than in R. mucilaginosa KC8. This suggests that a combination of chemical–physical mutagenesis, overexpression of the HMG1 gene, and adding ketoconazole is an effective strategy to improve carotenoid production.

Introduction

Carotenoids are a class of pigments synthesized as hydrocarbons or their oxygenated derivatives by plants and microorganisms. Carotenoids have received considerable attention because of their important biological functions. Carotenoids are used as antioxidants to reduce cellular or tissue damage (Rao and Agarwal 1999; Rao and Agarwal 2000), and as coloring agents for food products (Kong et al. 2010). Although commercial carotenoids are obtained mainly by chemical synthesis or extraction from tomatoes, the low cost and high efficiency of microbial synthesis has gained more and more attention. Several microbes, including bacteria, filamentous fungi, algae, and yeasts produce a broad range of carotenoids, including torularhodin, γ- and β-carotene, lycopene, torulene and astaxanthin (An et al. 2001; Frengova et al. 2004; Aksu and Eren 2005; Jeon et al. 2006; Kuzina and Cerda-Olmedo 2006; Raja et al. 2007).

A few carotenoid-producing microorganisms have been used for commercial purposes, such as Blakeslea trispora (used for β-carotene and lycopene production) and Xanthophyllomyces dendrorhous (used for astaxanthin production). Although some other carotenoid-producing microorganisms, such as yeasts, contain lower concentrations of carotenoids compared to the industrialized microorganisms, their ability to produce specific and valuable orange-red carotenoids like torularhodin and torulene have aroused attention (Frengova and Beshkova 2009). Current studies suggest that torularhodin and torulene can be applicable to prevent certain types of cancer [such as prostate tumors (Du et al. 2015)] and enhance the immune system as the precursors of vitamin A and hormones (Breierová et al. 2008). However, due to the undefined effects on human health, they have not been approved for use in foods or as cosmetic additives individually. More research is required on these pigments to promote their development and application. Libkind and Van Broock (2006) reported a yeast strain, R. mucilaginosa CRUB 006, containing 83.4% of torularhodin, 10.8% of β-carotene and 5.7% of torulene. Yeasts are more convenient than filamentous fungi or algae for large-scale production in fermenters due to their unicellular nature and high growth rate. R. mucilaginosa is considered to be a valuable carotenoid source and has potential commercial value. However, its low carotenoid production limits its commercial application. A potential solution to this could be to modify the strain through mutagenesis or metabolic engineering methods to enhance carotenoid production.

Carotenoids are derived from successive addition of an active isoprene unit, isopentenyl diphosphate (IPP), followed by enzymatic cyclization, oxidation, reduction, isomerization, and/or hydroxylation (McGarvey and Croteau 1995). Two IPP biosynthetic pathways have been reported in the literature: the mevalonate (MVA) pathway (Homann et al. 1996) and the 2-C-methyl-d-erythritol 4-phosphate (MEP) pathway (Rohmer et al. 1993). In contrast to some microorganisms, yeast uses only the MVA pathway to convert acetyl-CoA to IPP, which serves as a precursor for carotenoid biosynthesis (Boucher and Doolittle 2000). The biosynthesis of IPP in the MVA pathway begins with the conversion of acetyl-CoA to MVA through acetoacetyl-CoA and β-hydroxy-β-methylglutaryl coenzyme A (HMG-CoA) in an irreversible reaction catalyzed by HMG-CoA reductase (Disch and Rohmer 1998; Withers and Keasling 2007). HMG-CoA reductase, which is encoded by the HMG1 gene, is considered as a key regulatory point in the isoprenoid pathway (Ruiz-Albert et al. 2002). According to the available literature, overexpression of HMG1 gene (encoding HMG-CoA reductase) has successfully increased carotenoid accumulation. For example, a twofold increase in lycopene content was achieved in an engineered yeast, Candida utilis, through overexpression of HMG-CoA reductase (Shimada et al. 1998). Wang and Keasling (2002) expressed the HMG1 gene from S. cerevisiae in a filamentous fungus, Neurospora crassa, under the control of the strong, constitutive glyceraldehyde-3-phosphate dehydrogenase (GPD) promoter, resulting in 6- and 1.5-fold increases in lycopene and neurosporaxanthin production, respectively, relative to the wild-type strain.

Generally, genes are transferred from a wild strain to an engineered organism with a clear genetic background to produce improved or novel substances. At present, the metabolic engineering of carotenoids has focused mainly on non-carotenogenic microbes such as Escherichia coli, S. cerevisiae, or C. utilis, including the introduction of exogenous carotenoid genes from carotenogenic microbes or plants and other metabolic regulation methods. Very little published information is available on the metabolic engineering of wild-type carotenogenic microbes, such as B. trispora, R. mucilaginosa and Dunaliella salina. However, metabolic engineering of wild-type strains for secondary metabolite production can take advantage of the native pathways, which could be helpful for maintaining the genetic stability of the secondary metabolites.

In our previous study, we isolated a yeast strain, R. mucilaginosa KC8, which can produce carotenoids: mainly torularhodin and β-carotene. In this study, we used both mutagenesis and metabolic engineering methods to improve carotenoid production in R. mucilaginosa KC8. In addition, the effect of the ergosterol synthesis inhibitor, ketoconazole, on carotenoid production of the transformant was evaluated. As far as we are aware, this is the first reported use of metabolic engineering of a wild yeast strain of R. mucilaginosa to improve carotenoid production.

Materials and methods

Microorganisms and culture conditions

R. mucilaginosa KC8 was maintained at the Key Laboratory for Microorganisms and Functional Molecules. The strain was grown on yeast extract peptone dextrose (YPD) solid medium containing 20 g L⁻¹ glucose, 20 g L⁻¹ yeast extract, and 10 g L⁻¹ peptone at 28 °C, and subcultured thereafter every 30 days. E. coli DH5α was used for routine DNA manipulation.

Mutagenesis

R. mucilaginosa KC8 was cultured in YPD liquid medium until it reached the late log phase of the growth. The cells were collected and washed three times by centrifugation. The washed cells were resuspended and adjusted to 10⁸ cells mL⁻¹. The cell suspension was treated with atmospheric and room temperature plasma (ARTP—a mutation breeding system developed by Si Qing Yuan Biotechnology, Beijing, China) (Zhang et al. 2014). The operating parameters of the ARTP were as follows: helium gas flow rate QHe = 10.0 standard liters per minute (slpm), RF power input 120 W, and the treatment time 120 s (the maximum mutation rate was obtained under these conditions). The treated cells were diluted and then spread on YPD plates. Single colonies with a more intensive pink color than the wild type strain were selected for shake-flask fermentation. The mutant with maximum carotenoid production was chosen for NaNO₂ mutagenesis in the presence of 0.025 mol L⁻¹ NaNO₂ for 15 min. The mutant obtained from NaNO₂ mutagenesis was treated under UV (15 W, 20 cm) for 80 s, and then by UV-NaNO₂ composite mutagenesis. Finally, the mutant K4, with the highest carotenoid production, was selected for the subsequent study.

DNA manipulation and cloning procedure

The 5.8 s rDNA1 of R. mucilaginosa K4 was amplified by PCR using chromosomal DNA as template with the primers A1-F (5′-GCCCTCGAGCTGCAGAACCAATGCATTGGTCCGTAGG TGAACCTGCGG-3′, XhoI restriction site underlined) and A1-R (5′-CCGAGATCTG AAGATCTTC CTCCGCTTATTGATATGC-3′, BglII restriction site underlined). The 5.8 s rDNA2 of R. mucilaginosa K4 was amplified by PCR using chromosomal DNA as template with the primers A2-F (5′-GCGGGATCCCTGCAGAACCAATGCATTGGTCCGTAGGTGAACC TGCGG-3′, BamHI restriction site underlined) and A2-R (5′-GGCTCTAGAGAAGATCTT CCTCCGCTTATTGATATGC-3′, XbaI restriction site underlined). The oligonucleotides HMG-F (5′-GGACTAGTATGGACCAATTGGTGAAAACTGAAG-3′, SpeI restriction site underlined) and HMG-R (5′-CCGCTCGAGTTAGGATTTAATGCAGGTGACG-3′, XhoI restriction site underlined) were used for amplifying the HMG1 gene using S. cerevisiae genomic DNA as a template. All of these amplified PCR products were verified by sequencing.

Construction of expression vector

Plasmid pPIC-2rDNA-G-H-C, used for overexpression of the HMG1 gene, was constructed. To assure strong expression in R. mucilaginosa, a GPD promoter and CYC terminator were placed upstream and downstream of the HMG1 gene. The PCR products of the HMG1 gene were digested by SpeI and XhoI and ligated between the SpeI and XhoI sites of the plasmid P416 GPD, which contains the GPD promoter and CYC terminator. The fragment containing the HMG1 gene, GPD promoter, and CYC terminator was named G-H-C. And the plasmid containing fragment G-H-C was named P416 GPD-G-H-C. G-H-C was amplified by PCR using cloning vector pMD18-G-H-C as the template with the primers G-H-C-F (5′-GCGGCGGCCGCCAGT TCGAGTTTATCATTATCAAT-3′, NotI restriction site underlined) and G-H-C-R (5′-CCGGCGGCCGCGCGGCCGCGCAAATTAAAGCCAACGAG CG-3′, NotI restriction site underlined). The fragment 5.8 s rDNA1 and plasmid pPICZαA were digested by restriction endonucleases XhoI and BglII. The linear fragments were treated by T4 DNA polymerase to construct a new plasmid, named pPIC-rDNA1. Similarly, the fragment 5.8 s rDNA2 was digested and ligated into pPIC-rDNA1 and named pPIC-2rDNA. The fragment G-H-C was cut down by restriction endonuclease NotI, and then subcloned into plasmid pPIC-2rDNA, which was predigested with the same restriction sites. The resulting recombinant plasmid was named pPIC-2rDNA-G-H-C. The construction procedure of the expression vector is summarized in Fig. 1.

Fig. 1

Construction procedure of the expression vector

Transformation

Transformation of R. mucilaginosa K4 was carried out by electroporation as described by Kondo et al. (1995). According to the available literature (Takahashi et al. 2014; Mannazzu et al. 2015) and many trials, a few modifications of this method were made to improve the efficiency of transformation. Briefly, S. cerevisiae cells were grown in 200 mL YPD medium to the log phase [optical density at 600 nm (OD₆₀₀), 2.5; 10⁸ cells mL⁻¹]. The cells were harvested and washed twice with ice-cold sterilized water and once with ice-cold 1 M sorbitol. The cells were resuspended in 2.5 mL ice-cold 1 M sorbitol (10¹⁰ cells mL⁻¹). Plasmid pPIC-2rDNA-G-H-C was linearized by restriction enzyme BamHI digestion. The yeast cell suspension (50 μL) was mixed with 10 μL plasmid DNA solution and transferred to a 0.2-cm cuvette for electroporation. In the standard protocol, an electric pulse of 660 V was delivered, and internal resistance was set at 800 or 1000 V, with a capacitance of 25 mF (Bio-Rad Gene Pulser Xcell™, Hercules, CA). After the electric pulse was delivered, the transformed cells were spread on YPD plates containing 150 μg mL⁻¹ zeocin and incubated at 28 °C. Transformants with the phenotype of interest were transferred to slants and plates for further analysis. Integration of the expression cassettes into the genome was confirmed by using PCR with the corresponding primers described above.

Fermentation conditions

Cell suspensions (2 mL) were transferred to 150-mL conical flasks containing 30 mL YPD liquid seed medium. The flasks were cultivated in a rotary shaker at an agitation rate of 180 rpm at 28 °C for 36 h. Fermentation was carried out with 8% (v/v) seed medium inoculum in 250-mL conical flasks containing 50 mL fermentation medium. The fermentation medium (Liu et al. 2016) contained the following (g L⁻¹): glucose 31.8, yeast extract 1, (NH₄)₂SO₄ 2, KH₂PO₄ 1, MgSO₄·7H₂O 0.5, CaCl₂ 0.1, and NaCl 0.3, and the pH was adjusted to 7.0. All media were sterilized at 121 °C for 20 min. Cultures were maintained for 96 h, and the cells were then harvested to determine the dry cell weight and carotenoid concentrations.

Analytical techniques

At appropriate time intervals, a 10-mL sample was collected from the culture broth, and then centrifuged at 5000 g for 20 min. The sediment was washed with distilled water and re-centrifuged (three times). Dry biomass weight was determined after drying at 105 °C overnight. The extraction and quantification of total carotenoid concentration was performed as described by Irazusta et al. (2013). Total carotenoid quantification was performed spectrophotometrically at 490 nm by using the extinction coefficient of carotenoids (ε_{1% 1 cm} = 2680, according to Simpson et al. 1964).

Results

Mutation rate and carotenoid production by mutants

The highest mutation rates and carotenoid production following treatment by different mutagenesis methods is shown in Table 1. The mutation rates following treatment by atmospheric and room temperature plasma (ARTP) were obviously higher than those by other traditional mutagenesis methods. After several rounds of mutation-screening, R. mucilaginosa K4 showed a maximum carotenoid production (14.47 ± 0.06 mg L⁻¹), which was 67% higher than the parent strain (8.67 ± 0.07 mg L⁻¹).

Table 1 Maximum carotenoid production and mutation rates treated by different mutation methods. ARTP Atmospheric and room temperature plasma, UV ultra violet. Values are means of triplicate ± standard deviation (SD)

Clones and transformation

Genes encoding HMG-CoA reductase (HMG1) have been cloned and sequenced from various microorganisms (Homann et al. 1996). A cDNA encoding the HMG-CoA reductase of the HMG1 gene was cloned from S. cerevisiae genomic DNA. Sequence analysis of the amplified fragments indicated they were identical to the known HMG1 gene in S. cerevisiae. The amplified fragments of 5.8 s rDNA from R. mucilaginosa K4 have also proved correct by agarose gel electrophoresis and DNA sequencing. To verify the expression vector, G-H-C was amplified by PCR using the expression vector pPIC-2rDNA-G-H-C as a template, with primers G-H-C-F and G-C-H-R. The PCR products were recovered for sequencing. Sequence analysis of the PCR products showed that the amplified fragments consisted of the GPD promoter, HMG1, and the CYC terminator, and were identical with their homologous genes. After construction of the expression vector, R. mucilaginosa K4 was transformed with plasmid pPIC-2rDNA-G-H-C, and then plated onto YPD medium supplemented with zeocin. Colonies with red-pink color were selected for further characterization. The presence of expression cassettes was confirmed through PCR with the primers HMG-F and HMG-R. The PCR products of the transformants appeared as a clear 1.6-kb band in the electrophoresis gel. Sequence analysis of the purified PCR products indicated that the 1596-bp amplified fragments were identical to the HMG1 gene of S. cerevisiae. Therefore, the HMG1 gene of S. cerevisiae had been successfully introduced into R. mucilaginosa K4.

Effects of overexpression of HMG1 on biomass and carotenoid concentration

More than 20 transformants with redder colony color were selected on zeocin-supplied YPD plates. After secondary screening with flask-shaking fermentation, three strains showed elevated carotenoid production, and were designated as G1, G2, and G3, respectively. The biomass and carotenoid production of the transformants and the wild strain after 96-h fermentation are shown in Table 2. The carotenoid concentrations of the transformants were markedly higher than that of R. mucilaginosa K4, whereas they had negligible difference in biomass. Among these transformants, R. mucilaginosa G1 showed a maximum carotenoid concentration (1.13 ± 0.01 mg g⁻¹ dry biomass) which was 17.9% higher than the parent strain R. mucilaginosa K4 (0.97 ± 0.01 mg g⁻¹ dry biomass). Table 3 shows that the carotenoid-producing ability of R. mucilaginosa G1 remained relatively stable after eight generations.

Table 2 Biomass and carotenoid production of transformants and strain K4. Values are means of triplicate ± SD

Table 3 Carotenoid production of transformant G1 through generations. Values are means of triplicate ± SD

Effects of ketoconazole on biomass and carotenoid concentration

Blocking sterol biosynthesis can drive more precursors into the carotenoid biosynthesis pathway compared to that of applying a single over-expression. Therefore, we studied the effects of the ergosterol synthesis inhibitor, ketoconazole, on biomass and carotenoid concentration of the transformant R. mucilaginosa G1 strain. As shown in Fig. 2, a certain amount of ketoconazole significantly stimulated carotenoid biosynthesis. The highest carotenoid volumetric production (19.14 ± 0.09 mg L⁻¹) was achieved after supplying 28 mg L⁻¹ ketoconazole, which was 12.5% higher than that without the addition of ketoconazole.

Fig. 2

Effects of ketoconazole on biomass and carotenoid production of the transformant R. mucilaginosa G1. Three replicates of each sample. Error bars SD of mean values. ** P < 0.01 (statistically significant difference between the value and control)

Discussion

For a wild yeast strain, mutagenesis is an effective method to improve secondary metabolite production. Among various mutagenesis methods, ARTP is a novel plasma generating system driven by a radio-frequency (RF) power supply with water-cooled, bare-metallic electrodes. ARTP has many advantages, such as low cost, low plasma temperature, flexible operation, and security, and thus shows promising application in biotechnology. Recently, ARTP has been used extensively as an effective mutation tool for various microorganisms (Lu et al. 2011; Qiang et al. 2014; Zhang et al. 2014). Its mutation effects were also obvious in our yeast strain, R. mucilaginosa K4. However, combined with other chemical and physical mutagenesis methods, the mutation rate was increased further.

Just as with strains that have been treated by different kinds of mutagenesis, its hard to obtain ideal effects through physical and chemical mutation breeding to increase production. Metabolic engineering could solve this problem by orientational modification. The delivery of exogenous genes into yeast cells is a crucial technique in the genetic manipulation procedure. Electroporation is simple and convenient, and can be adapted to a wide variety of yeast cells. Initially, the transformation frequency was low. Through constant experimentation, the voltage of the electric pulse and the amount of the plasmid were identified as the key factors for successful electroporation. HMG-CoA reductase is the rate-controlling enzyme of the mevalonate pathway, the metabolic pathway that produces carotenoids, sterol, and other isoprenoids (Goldstein and Brown 1990; Gardner and Hampton 1999). The introduction and overexpression of the HMG1 gene enhanced the amount of HMG-CoA reductase, and carotenoid production consequently increased. The successful amplification of the HMG1 gene using R. mucilaginosa G1 genomic DNA as a template indicated that the HMG1 gene had been integrated into genomic DNA. The stable inheritance of the carotenoid-producing ability of R. mucilaginosa G1 also supported this hypothesis.

The increase of common precursors resulting from over-expression of the HMG1 gene had also been previously found to be able to increase the flux of the squalene and ergosterol biosynthetic pathway (Shimada et al. 1998). Meanwhile, in the sterol biosynthetic pathway, HMG-CoA synthase is regulated by the feedback inhibition of end products (Goldstein and Brown 1990). Excessive ergosterol might reduce HMG-CoA synthase activity, resulting in the limited increment of overall flux. Ketoconazole, an inhibitor of ergosterol biosynthesis, has been applied successfully by Tang et al. (2008) to increasing carotenoid production in B. trispora. To relieve the feedback inhibition, and to down-regulate the sterol synthesis branch, ketoconazole was added in the present study. Although the increment was less than that from mutation methods, the regulation strategies turned out to be effective. The low increment may still be ascribed to the insufficiency of common precursors. If the flux of the common biosynthetic pathway is further enhanced, the effects of ketoconazole on carotenoid production would be more significant. Certainly, the economy and safety of the ketoconazole should be considered before it is used in commercial carotenoid products. It can be concluded that the increase of carotenoid content in R. mucilaginosa G1 is due to more precursors from over-expression of the HMG1 gene fluxing into the carotenoid synthetic pathway in the presence of an ergosterol synthesis inhibitor.

The present study suggests that the combination of overexpressing the HMG-CoA reductase gene and adding ergosterol synthesis inhibitors is an effective strategy for a wild yeast strain to further improve carotenoid production after mutagenesis. However, this is a preliminary study and the increase of carotenoid production was still very limited. At its present carotenoid production, R. mucilaginosa G1 is not ready for commercial application; however, its carotenoid-producing ability was much higher than other R. mucilaginosa strains, such as an isolate from the Brazilian ecosystem with a maximum carotenoid production of 745 μg L⁻¹ (Maldonade et al. 2012), another R. mucilaginosa isolate from soil with a carotenoid production of 355 μg g⁻¹ (Aksu and Eren 2005), and an R. mucilaginosa mutant A23 with a production of 734.58 μg/g dry biomass after UV irradiation (Issa et al. 2016). Therefore, further efforts should be made to increase transformation rates, increase the number of screening transformants and investigate the expression conditions of exogenous genes. In addition, these strategies should be investigated with the industrial strain, B. trispora, where metabolic pathways and gene information are more complicated.