Codon Optimization

Each organism has codons, which it uses more or less frequently. Because of this, even if different transcribed RNA sequences result in the same translated amino-acid sequence, expression of the protein can vary based on the codons that are used. Therefore, codon optimization is needed for optimal gene expression. In our case, we needed to reverse engineer the proteins that allow for crocin synthesis in Crocus sativus into DNA sequences, which use the optimal codons for our unconventional yeast Rhodotorula toruloides.

The most frequently used codons of R. toruloides result in a DNA sequence that is high in G-C content [1]. The G-C content of the R. toruloides genome is 62% [2]. Companies which provide DNA synthesis, however, can not synthesise large DNA fragments with a high G-C content. Although these companies, such as IDT and Twist Bioscience, offer online sequence optimization, they don’t offer codon optimization for our unconventional yeast R. toruloides. Therefore, we used DNA Chisel [3], a python library that allows for the use of custom codon usage tables.

With no added constraints for the codon optimisation task, the optimised sequence had a high GC content and highly repetitive regions - traits that don’t allow for custom DNA synthesis. Different constraints, such as maximum GC content and maximum amount of repetitive sequences, were added. This resulted in a balancing act where optimal codon usage for our yeast was balanced with the criteria needed to synthesise the DNA. Although the sequences



DNA construct creation

To achieve crocin synthesis in R. toruloides 3 genes from Crocus sativus need to be added. Additionally, an antibiotic resistance gene is needed to select successful transformants.

Level 0 parts

7 new parts for use in R. toruloides were created. These parts are:

  • 1. Codon optimized zeaxanthin 7,8(7',8')-cleavage dioxygenase from Crocus sativus.
  • 2. Codon optimized crocetin glucosyltransferase 2 from Crocus sativus.
  • 3.Terminator from R. toruloides (NP11) mRNA-heat shock 70kDa protein gene
  • 4. Codon optimized kanamycin resistance gene from Escherichia coli
  • 5. Promoter from R. toruloides (NP11) glucose-6-phosphate isomerase (PGI) gene
  • 6. Codon optimized beta-carotene hydroxylase from Crocus sativus.
  • 7. Codon optimized hygromycin resistance gene from Escherichia coli
  • To create these parts, restriction sites used in BioBrick and RFC1000 assembly were removed. RFC1000 prefix and suffix for each part were added to the ends. Flanking the RFC1000 prefix and suffix, each part also has a BioBrick prefix and suffix, which allows for easy PCR amplification for all parts using the same primers and also allows the use of the conventional Biobrick assembly method.

    Level 1 Part Assembly

    We used the iGEM RFC1000 standard for the assembly of our DNA constructs. This method uses the type IIs restriction enzyme BsaI to digest and T4 ligase to ligate the level 0 parts. Each level 1 part includes a promoter, transcribed sequence and a terminator. We used the same promoter PPGI for all level 1 parts. The PPGI promoter is the strongest constitutive promoter reported for R. toruloides NP11 [1].

    Before the Golden Gate Assembly (GGA) was performed each level 0 part was amplified by PCR using Taq polymerase [2]. After the PCR amplification, the required parts for GGA (promoter, transcribed sequence and terminator for each transcriptional unit (TU) ) were mixed together (7 ul each) and purified using Thermo Scientific GeneJET PCR purification kit. The GGA was performed with the purified mix for each transcriptional unit. For each transcriptional unit 10 ul the purified mix, 2 ul 10x T4 ligase buffer, 0.5ul of BsaI, 0.5 ul of T4 ligase and 7ul of water were added in a PCR tube. The GGA mix was cycled between 5 min 37 C and 5 min 16 C for 10 cycles, followed by 5 min enzyme inactivation at 60 C.

    Level 2 Assembly

    Before level 2 assembly each transcriptional unit was amplified by PCR with primers that contained overhangs with SapI restriction recognition sites and fusion sites according to the iGEM RFC1000 standard. For each transcriptional unit 10ul of the GGA product was used in a 50ul PCR reaction [2]. The PCR reaction was performed according to [3] protocol.

    After PCR amplification gel electrophoresis was performed and the DNA fragments near 3kb were purified. Level 2 assembly was performed by mixing the purified transcriptional units 1-4 (4 ul each), adding 2 ul 10x T4 ligase buffer, 0.5 ul SapI and 0.5 ul T5 ligase. The GGA mix was cycled between 5 min 37°C and 5 min 16°C for 10 cycles, followed by 5 min enzyme inactivation at 60°C.

         [1] Wang, Y., Lin, X., Zhang, S., Sun, W., Ma, S., and Zhao, Z. K. (2016) 
         Cloning and evaluation of different constitutive promoters in the oleaginous yeast Rhodosporidium toruloides.
          Yeast, 33: 99– 106. doi: 10.1002/yea.3145.
         [2] https://international.neb.com/protocols/0001/01/01/taq-dna-polymerase-with-standard-taq-buffer-m0273

    Transformation, colony PCR and qPCR

    R. toruloides does not have a described 2-micron plasmid ORI sequence. One of the described methods for transformation uses Agrobacterium tumefaciens, which has been proven to be capable of infecting fungi [1]. The Agrobacterium tumefaciens-mediated transformation (ATMT) has been successfully used to incorporate foreign DNA into the genome of R. toruloides [2]. Since ATMT is a time consuming method that presents multiple difficulties we chose a protocol described by Nora et al. to incorporate the linear DNA construct into R. toruloides [3]. This method relies on the non-homologous end joining mechanism where the foreign DNA is inserted in random locations of the yeast genome.

    We tested if the DNA has successfully been integrated by adapting a previously described yeast colony PCR protocol [4].

      [1] Hooykaas PJJ, van Heusden GPH, Niu X, Reza Roushan M, Soltani J, Zhang X, van der Zaal BJ. Agrobacterium-Mediated
      Transformation of Yeast and Fungi. Curr Top Microbiol Immunol. 2018;418:349-374. doi: 10.1007/82_2018_90. PMID: 29770864.
      [2] Park YK, Nicaud JM, Ledesma-Amaro R. The Engineering Potential of Rhodosporidium toruloides as a Workhorse for
      Biotechnological Applications. Trends Biotechnol. 2018 Mar;36(3):304-317. doi: 10.1016/j.tibtech.2017.10.013. Epub 2017 Nov 10.
      PMID: 29132754. 
      [3] Nora, L.C., Wehrs, M., Kim, J. et al. A toolset of constitutive promoters for metabolic engineering of Rhodosporidium
      toruloides. Microb Cell Fact 18, 117 (2019). https://doi.org/10.1186/s12934-019-1167-0
      [4] Lõoke M, Kristjuhan K, Kristjuhan A. Extraction of genomic DNA from yeasts for PCR-based applications. Biotechniques. 2011
      May;50(5):325-8. doi: 10.2144/000113672. PMID: 21548894; PMCID: PMC3182553.

    Uracil auxotrophic strain creation

    Using auxotrophy-complementing genes as selection markers eliminates the need for antibiotic based positive clone identification. A strain unable to synthesise a compound necessary for its growth is auxotrophic and can survive only if it can be taken from the environment. The auxotrophy-complementing genes therefore can be used to create a prototroph that is able to survive in minimal media without the previously necessary amino acid, nucleotide or other organic compound. Widespread antibiotic pollution promoting the resistance to them is the main reason auxotrophy genetic markers should be preferred.

    R. toruloides has been transformed with resistance genes for hygromycin, bleomycin and zeocin which are some of the effective antibiotics for the wild type yeast [1]. In our opinion auxotrophic selection methods should be the first to be implemented for novel yeasts. The creation of auxotrophic strains would allow for future genetic engineering that forgoes environmentally harmful and often exotic and expensive antibiotics. Therefore we set to create a uracil auxotrophic R. toruloides strain and design a complementing gene to be used as a selection marker for our and future iGEM and research projects.

    URA3 gene encodes orotidine 5-phosphate decarboxylase which is necessary for the synthesis of uracil - a nucleotide essential for cell survival. The Ura3p also converts 5-Fluoroorotic acid to 5-fluorouracil which is a suicide inhibitor that causes cell death. Knowing that URA3 strains are killed off by FOA5 and ura3 strains are unable to grow without supplemented uracil both positive and negative selection experiments can be designed.

    FOA5 toxicity to wild type strains can be used to screen for useful mutants that are uracil auxotrophic, since those would not be affected by FOA5 presence. We are using a protocol developed by [2] to isolate a uracil auxotrophic strain of R. toruloides which could be later confirmed by sequencing.

         [1] Park YK, Nicaud JM, Ledesma-Amaro R. The Engineering Potential of Rhodosporidium toruloides as a Workhorse for
         Biotechnological Applications. Trends Biotechnol. 2018 Mar;36(3):304-317. doi: 10.1016/j.tibtech.2017.10.013. 
         Epub 2017 Nov10. PMID: 29132754. 
         [2] Wang F, Yue L, Wang L, Madzak C, Li J, Wang X, Chi Z. Genetic modification of the marine-derived yeast Yarrowia
         lipolytica with high-protein content using a GPI-anchor-fusion expression system. Biotechnol Prog.
          2009 Sep-Oct;25(5):1297-303. doi: 10.1002/btpr.235. PMID: 19743190. 

    Medium optimization for higher beta-carotene yields

    Amounts of R. toruloides carotenoid production can be highly variable. There are many factors that affect carotenoid yields, such as composition of growth medium, temperature and lighting conditions. Furthermore, it is mentioned in literature that additives such as asparagine can boost carotenoid production. In order to achieve higher yields of our final product (crocin), it was important to make as much beta-carotene as possible since carotene is the starting point of our biosynthesis cycle.

    For our experiments we prepared four different media. Medium 1: defined medium with salts (10 g/L glucose; 0.52 g/L KH2PO4; 0.52 g/L MgSO4; 4 g/L NH4NO3; 10 g/L asparagine); medium 2: SD minimal medium (20 g/L glucose; 6.7 g/L yeast nitrogen base); medium 3: SD minimal medium with added arginine (10 g/L); Medium 4: YPD medium with salts (40 g/L glucose, 7 g/L KH2PO4, 2 g/L (NH4)2SO4, 2 g/L Na2SO4, 1.5 g/L MgSO4·7H2O, 1.5 g/L yeast extract). All media were prepared with deionized water and sterilised by autoclaving at 121°C. Tetracycline (10 mg/L) was added to sterilised and cooled solutions. For each experiment, 50 mL of medium was transferred to sterile 250 mL flask and inoculum was added from an SD night culture for the final OD to be around 0.2.

    We chose four different growth conditions for each medium: 30°C with light; 30°C without light; 20°C with light; 20°C without light. Prepared cultures were put in incubators and grown for 120 hours while shaking at 160 rpm. Optical density at 600 nm was measured after 120 hours. The final OD for most of the culture reached 20. Cells were harvested with centrifugation (7000g, 2 min) and washed with distilled water twice. The carotenoids were extracted from wet cell biomass and quantified via UV-Vis Spectroscopy.