Pichia pastoris versus Saccharomyces cerevisiae: a case study on the recombinant production of human granulocyte-macrophage colony-stimulating factor
- Anh-Minh Tran†1,
- Thanh-Thao Nguyen†1,
- Cong-Thuan Nguyen1,
- Xuan-Mai Huynh-Thi1,
- Cao-Tri Nguyen1,
- Minh-Thuong Trinh1,
- Linh-Thuoc Tran1,
- Stephanie P. Cartwright2,
- Roslyn M. Bill2 and
- Hieu Tran-Van1Email authorView ORCID ID profile
© The Author(s) 2017
Received: 30 September 2016
Accepted: 27 March 2017
Published: 4 April 2017
Recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) is a glycoprotein that has been approved by the FDA for the treatment of neutropenia and leukemia in combination with chemotherapies. Recombinant hGM-CSF is produced industrially using the baker’s yeast, Saccharomyces cerevisiae, by large-scale fermentation. The methylotrophic yeast, Pichia pastoris, has emerged as an alternative host cell system due to its shorter and less immunogenic glycosylation pattern together with higher cell density growth and higher secreted protein yield than S. cerevisiae. In this study, we compared the pipeline from gene to recombinant protein in these two yeasts.
Codon optimization in silico for both yeast species showed no difference in frequent codon usage. However, rhGM-CSF expressed from S. cerevisiae BY4742 showed a significant discrepancy in molecular weight from those of P. pastoris X33. Analysis showed purified rhGM-CSF species with molecular weights ranging from 30 to more than 60 kDa. Fed-batch fermentation over 72 h showed that rhGM-CSF was more highly secreted from P. pastoris than S. cerevisiae (285 and 64 mg total secreted protein/L, respectively). Ion exchange chromatography gave higher purity and recovery than hydrophobic interaction chromatography. Purified rhGM-CSF from P. pastoris was 327 times more potent than rhGM-CSF from S. cerevisiae in terms of proliferative stimulating capacity on the hGM-CSF-dependent cell line, TF-1.
Our data support a view that the methylotrophic yeast P. pastoris is an effective recombinant host for heterologous rhGM-CSF production.
KeywordsS. cerevisiae P. pastoris rhGM-CSF Fermentation Bioactivity TF-1 cell
Human granulocyte-macrophage colony-stimulating factor (hGM-CSF), a glycosylated cytokine, plays a vital role in proliferation and differentiation of granulocytes and macrophages from bone marrow progenitor cells . hGM-CSF also enhances cytotoxicity of macrophages and monocytes towards tumor cells , as well as tumor presentation of dendritic cells [1, 3]. Due to its stimulatory capacity on hematopoietic stem cells, recombinant hGM-CSF (rhGM-CSF) is recommended for therapeutic use in combination with chemo- or radio-therapy for cancer or transplantation patients .
hGM-CSF comprises 127 amino acids. The presence of 2 N- and 4 O-glycosylation sites leads to different glycoforms with molecular weights ranging from 14 to 60 kDa . The glycosylation of the protein affects its pharmacokinetics, in vivo half-life, immunogenicity and cytotoxicity .
rhGM-CSF is produced from bacteria, baker yeast or mammalian hosts . rhGM-CSF produced in Escherichia coli is biologically active but unstable in human plasma. It induces an immune reaction because N-formyl methionine (fMet) is its first amino acid. rhGM-CSF produced in mammalian hosts has a similar glycosylation pattern to the native human protein, but production rates are slow [6, 7]. The product from baker’s yeast, Saccharomyces cerevisiae, is glycosylated and approved by the FDA for the treatment of neutropenia and leukemia in combination with chemo- or radio-therapy for cancer or transplantation patients. The drawbacks of using S. cerevisiae are hyper-glycosylated products and low cell density growth.
The methylotrophic yeast Pichia pastoris has emerged as an alternative host due to its shorter and less immunogenic glycans, higher density cell growth and higher secreted protein yields than S. cerevisiae. In this study, we show that P. pastoris secretes a higher yield of more active recombinant rhGM-CSF than S. cerevisiae.
In silico codon optimization
hGM-CSF coding sequences were from Gene Bank (Accession M11220). In silico codon optimization was done using DNA 2.0 software with P. pastoris and S. cerevisiae codon tables and optimized parameters . The optimization evaluation was done via http://www.gcua.schoedl.de with low (<20%) frequency codon usage displayed as hatched bars and very low (<10%) frequency codon usage as white bars.
Expression of rhGM-CSF in P. pastoris and S. cerevisiae
Recombinant P. pastoris and S. cerevisiae rhGM-CSF clones were expressed as described previously [9, 10]. The expression of the rhGM-CSF was analyzed by SDS-PAGE and immuno blot probed with an antibody against hGM-CSF (LifeSpan BioScience). Briefly, recombinant S. cerevisiae strain BY4742/hgm-csf was cultured in selective CSM-ura medium (6.7 g/L yeast nitrogen base with ammonium sulphate, 0.1 M sodium phosphate, supplemented with amino acids lacking uracil ) supplemented with 2% glucose until the OD600 of the culture was 0.5–1. The cells were then harvested and transferred into CSMG-ura medium (CSM supplemented with 20 g/L galactose) with shaking at 230 rpm at 30 °C. The pH 6.0 of each medium was stabilized by 0.1 M sodium phosphate buffer. The sample supernatants were collected after 72 h of induction and stored for further analysis.
Recombinant P. pastoris X33::hgm-csf was grown in 10 mL BMGY medium (2% peptone, 1% yeast extract, 0.34% yeast nitrogen base, 1% ammonium sulfate, 1% glycerol, and 0.4 mg/L biotin, buffered with 1/10 volume of pH 6.0 potassium phosphate buffer) at 30°C with constant shaking at 250 rpm until the culture reached an OD600 of 2–4. The cells were harvested and re-suspended in 10 mL BMMY (0.5% methanol is substituted for 1% glycerol in BMGY) with the same growth conditions. For induction, methanol was added every 24 h to a final concentration of 0.5%. After 72 h, culture supernatants were collected by centrifugation for further analysis.
Fermentation of recombinant P. pastoris and S. cerevisiae
Recombinant yeast S. cerevisiae strain BY4742/hgm-csf was batch-cultured in selective CSM(-ura) medium, pH 5.0 supplemented with 2% glucose until the culture reached an OD600 of 0.5 to 1, and then transferred into 3.18 L YP medium to a final OD600 of 0.5 in a 5 L LiFlus-GX fermenting vessel system (Biotron). Others growth parameters were temperature at 30 °C, aeration rate 1.5 vvm (through a filter), and agitation speed 720 rpm. Every 24 h, 40 g/L galactose was added at a rate of 6.7 mL/h. Samples were harvested at every 4 h and analyzed. Growth of the yeast cultures was monitored by optical density measurements at 600 nm. Protein secretion in the fermented medium was analyzed by SDS-PAGE and the protein concentration was determined using Bradford’s method . The supernatant was stored at −20 °C for later purification.
For P. pastoris expression , a flask containing 150 mL BMGY containing 0.4 µg/mL biotin was inoculated from a frozen glycerol stock. The inoculum seed flask was grown at 30 °C, 250 rpm, and 24 h until OD600 = 2–6. A sterilized fermenter containing 2.5 L Fermentation Basal Salts medium supplemented with 4% glycerol and 11 mL PTM1 trace salts was prepared, the pH adjusted to 5.0 with ammonium hydroxide, the temperature set to 30 °C, agitation at 750 rpm and aeration to 5.0 vvm air. The culture in the inoculum seed flask was completely transferred into the fermenter in which the batch culture was grown for 24 h, until the glycerol was completely consumed. Glycerol feeding was initiated for about 4 h at a feed rate to 18.15 mL/h 50% w/v glycerol containing 1.2% v/v PTM1 trace salts. Methanol feeding was then initiated for 72 h at a feed rate of 100% methanol containing 1.2% v/v PTM1 trace salts as follows: 9 mL/h for 3 h; 18 mL/h for 3 h and then 27 mL/h. The supernatant was analyzed by SDS-PAGE every 4 h and the protein concentration was determined using Bradford’s method . The supernatant was stored at −20 °C for later purification.
Ion exchange chromatography and hydrophobic interaction chromatography purification of recombinant rhGM-CSF
rhGM-CSF was purified using hydrophobic interaction chromatography. A HiPrep Phenyl Fast Flow column was equilibrated in buffer A [(NH4)2SO4 2 M, CH3COONa 20 mM, pH 5.0]. Approximately 20 mL of filtered supernatant was loaded onto the column at 1 mL/min. The column was then washed with 10 column volumes of buffer B. The bound protein was subsequently eluted with buffer B (Na2HPO4 10.14 mM; KH2PO4 1.76 mM; NaCl 136.89 mM; KCl 2.68 mM, pH 7.4) in a step-wise manner, ranging from 5, 10, 20, 40, 60 and 100% buffer B. The target fractions were identified by SDS-PAGE. The protein was eluted maximally at 10 to 20% buffer B.
rhGM-CSF was alternatively purified using anion exchange chromatography at pH 7.5. A HiTrap Q Sepharose Fast Flow column was equilibrated in 50 mM Tris-HCl. Approximately 20 mL of filtered and desalted supernatant was loaded onto the column at 1 mL/min. The column was then washed with 10 column volumes of 50 mM Tris-HCl. The bound protein was subsequently eluted in a step-wise manner with the NaCl concentration optimized over a range of 0–1 M NaCl in 50 mM Tris-HCl. The target fractions were identified by SDS-PAGE. The protein was eluted maximally at 0.1–0.2 M NaCl.
Bioactivity of purified rhGM-CSF
In silico optimization removes low frequency codon usage in both yeasts
Differences in the expression profile of rhGM-CSF in the two yeast species
Recombinant hGM-CSF is highly secreted from P. pastoris
Ion exchange chromatography is more efficient in recovery of rhGM-CSF than hydrophobic interaction chromatography
Purification yield of rhGM-CSF from P. pastoris and S. cerevisiae using IEC with % elution buffer or HIC
GM-CSF from P. pastoris (IEC) 10%
GM-CSF from P. pastoris (IEC) 20%
GM-CSF from P. pastoris (HIC)
GM-CSF from S. cerevisiae (IEC) 10%
GM-CSF from S. cerevisiae (IEC) 20%
GM-CSF from S. cerevisiae (HIC)
Purification yield (%)
Recombinant hGM-CSF produced from P. pastoris shows high bioactivity on a GM-CSF-dependent cell line
Bioactivity of purified rhGM-CSF from P. pastoris and S. cerevisiae using IEC or HIC
Response curve range (OD unit)
GM-CSF from P. pastoris (IEC)
11.8 × 106
GM-CSF from P. pastoris (HIC)
5.0 × 106
GM-CSF from P. pastoris (crude)
0.53 × 106
GM-CSF from S. cerevisiae (IEC)
0.36 × 105
GM-CSF from S. cerevisiae (crude)
0.27 × 105
Four publications describe rhGM-CSF expression in three different yeast systems, including the well-known yeasts S. cerevisiae, P. pastoris and Yarrowia lipolytica [14–17], but none conducts a head-to-head comparison. In the present study, we used an in silico-optimized synthetic hGM-CSF gene in two yeast expression vectors. The expression patterns were similar to those previously published [14, 15, 17].
Using fed-batch fermentation, we obtained yields of rhGM-CSF of 9.5 and 180.3 mg/L from S. cerevisiae and P. pastoris respectively. The latter is higher than has been previously described , possibly because we used single-copy rather than high-copy integration. All published protocols on the purification of rhGM-CSF combine two different purification steps to get above 95% purity [14–17]. In our study, we used a one-step procedure to obtain greater than 95% purity.
The activity of purified rhGM-CSF from S. cerevisiae on TF-1 cells was 0.36 × 105 LU/mg, which is lower than that of the commercially available drug Leukine (rhGM-CSF). The discrepancy could be due to the glycosylation moiety added to rhGM-CSF  which in turn could have an effect on its bioactivity. The drug Leukine contains a non-glycosylated rhGM-CSF in contrast to our purified, glycosylated rhGM-CSF. Purified rhGM-CSF from P. pastoris had an activity of 1.2x107 LU/mg. This activity is almost equal to the standard sample announced by WHO  and is two times higher than that of Leukine. Compared to previous data , our purified rhGM-CSF is 15 times less active. When compared with our E. coli-derived rhGM-CSF (non-glycosylated), our data show that rhGM-CSF is 1.5 times lower than that of E. coli origin (unpublished data). With the same purity, rhGM–CSF from S. cerevisiae was 327 times less active than that from P. pastoris.
In this study, we have been able to compare the pipeline from gene to recombinant protein between S. cerevisiae and P. pastoris. Our results suggest that rhGM-CSF expressed in P. pastoris is more active than the same protein from S. cerevisiae. To our knowledge, the data presented here are the first head-to-head comparison of S. cerevisiae and P. pastoris as host cells for rhGM–CSF production.
recombinant human granulocyte-macrophage colony-stimulating factor
Food and Drug Administration
human erythroleukemic cell line
sodium dodecyl sulfate-polyacrylamide gel electrophoresis
effective dose 50
ion exchange chromatography
hydrophobic interaction chromatography
World Health Organization
AMT and TTN initiated the study, prepared the recombinant clones and participated in the data analysis and interpretation. TCN and MTT carried out the fermentations. XMH performed the expression analysis. AMT and CTN performed the purification. TTN and HTV conducted bioassay evaluation. TLT participated in the experimental design. SC and RB participated in the experimental design and helped to draft the manuscript. HTV participated in the study design, coordinated the data analysis and interpretation, and drafted the manuscript. All authors contributed to the final version of the manuscript. All authors read and approved the final manuscript.
We thank Prof. Toshio Kitamura, Institute of Medical Science, University of Tokyo for providing the TF-1 cell line.
The authors declare that they no competing interests.
Availability of supporting data and materials
Underlying data is contained in Additional files 1 and 2. This data has also been deposited at Figshare: https://figshare.com/articles/Data_of_Fig_5/4762642 and https://figshare.com/articles/Data_of_Fig_7/4762636.
Research funding was from the Science and Technology Research Program through the Department of Science and Technology Hochiminh City (Grant Number: 77/2013/HĐ-SKHCN). The British Council supported a visit of HTV to Aston University. The funding bodies have no role in the design, collection, analysis, interpretation and reporting of data or in writing the manuscript.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Metcalf D. The granulocyte-macrophage colony-stimulating factors. Science. 1985;229(4708):16–22.View ArticlePubMedGoogle Scholar
- Grabstein KH, Urdal DL, Tushinski RJ, Mochizuki DY, Price VL, Cantrell MA, Gillis S, Conlon PJ. Induction of macrophage tumoricidal activity by granulocyte-macrophage colony-stimulating factor. Science. 1986;232(4749):506–8.View ArticlePubMedGoogle Scholar
- Ghosh PK, Bhardwaj D, Karnik R. Human granulocyte-macrophage colony-stimulating factor: the protein and its current & emerging applications. Indian J Biotechnol. 2007;6(4):435–41.Google Scholar
- Nicola AN. GM-CSF. In: Oppenheim JJ, Feldmann M, Durum SK, editors. Cytokine reference: a compendium of cytokines and other mediators of host defense, vol. 1. 1st ed. New York: Academic Press; 2001. p. 900–5.Google Scholar
- Dorr RT. Clinical properties of yeast-derived versus Escherichia coli-derived granulocyte-macrophage colony-stimulating factor. Clin Ther. 1993;15(1):19–29.PubMedGoogle Scholar
- Armitage JO. Emerging applications of recombinant human granulocyte-macrophage colony-stimulating factor. Blood. 1998;92(12):4491–508.PubMedGoogle Scholar
- Bollati-Fogolin M, Forno G, Nimtz M, Conradt HS, Etcheverrigaray M, Kratje R. Temperature reduction in cultures of hGM-CSF-expressing CHO cells: effect on productivity and product quality. Biotechnol Prog. 2005;21(1):17–21.View ArticlePubMedGoogle Scholar
- Villalobos A, Ness JE, Gustafsson C, Minshull J, Govindarajan S. Gene Designer: a synthetic biology tool for constructing artificial DNA segments. BMC Bioinform. 2006;7:285–92.View ArticleGoogle Scholar
- Lien PTK, Truong DT, Hieu TV. Cloning and expression of human granulocyte-macrophage colony stimulating factor (hGM-CSF) in Pichia pastoris. J Sci Technol Dev. 2013;16(3):13–21.Google Scholar
- Minh TA, Thao NT, Cartwright SP, Bill RM, Hieu TV. Cloning, expression and isolation of recombinant protein hGM-CSF in Saccharomyces cerevisiae. J Biotechnol. 2015;13(2A):643–9.Google Scholar
- Darby RAJ, Cartwright SP, Dilworth MV, Bill RM. Which yeast species shall I choose? Saccharomyces cerevisiae versus Pichia pastoris (review). In: Bill RM, editor. Recombinant protein production in yeast—methods and protocols. Berlin: Springer Science; 2012. p. 14.Google Scholar
- Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976;72:248–54.View ArticlePubMedGoogle Scholar
- Invitrogen. Pichia fermentation process guidelines. Carlsbad: Invitrogen Corporation; 2002.Google Scholar
- Miyajima A, Otsu K, Schreurs J, Bond MW, Abrams JS, Arai K. Expression of murine and human granulocyte-macrophage colony-stimulating factors in S. cerevisiae: mutagenesis of the potential glycosylation sites. EMBO J. 1986;5(6):1193–7.PubMedPubMed CentralGoogle Scholar
- Bhatacharya P, Pandey G, Mukherjee KJ. Production and purification of recombinant human granulocyte-macrophage colony stimulating factor (GM-CSF) from high cell density cultures of Pichia pastoris. Bioprocess Biosyst Eng. 2007;30(5):305–12.View ArticlePubMedGoogle Scholar
- Gasmi N, Lassoued R, Ayed A, Tréton B, Chevret D, Nicaud JM, Kallel H. Production and characterization of human granulocyte–macrophage colony-stimulating factor (hGM-CSF) expressed in the oleaginous yeast Yarrowia lipolytica. Appl Microbiol Biotechnol. 2012;96(1):89–101.View ArticlePubMedGoogle Scholar
- Srinivasa Babu K, Pulicherla KK, Antony A, Meenakshisundaram S. Cloning and expression of recombinant human GMCSF from Pichia pastoris GS115-A progressive strategy for economic production. Am J Ther. 2014;21(6):462–9.View ArticlePubMedGoogle Scholar
- Mire-Sluis AR, Das RG, Thorpe R. The international standard for granulocyte-macrophage colony stimulating factor (GM-CSF) evaluation in an international collaborative study. J Immunol Methods. 1995;179(1):127–35.View ArticlePubMedGoogle Scholar