Skip to main content
Download PDF

Transient co-expression with three O-glycosylation enzymes allows production of GalNAc-O-glycosylated Granulocyte-Colony Stimulating Factor in N. benthamiana

Plant Methods volume 14, Article number: 98 (2018) Cite this article

Abstract

Background

Expression of economically relevant proteins in alternative expression platforms, especially plant expression platforms, has gained significant interest in recent years. A special interest in working with plants as bioreactors for the production of pharmaceutical proteins is related to low production costs, product safety and quality. Among the different properties that plants can also offer for the production of recombinant proteins, protein glycosylation is crucial since it may have an impact on pharmaceutical functionality and/or stability.

Results

The pharmaceutical glycoprotein human Granulocyte-Colony Stimulating Factor was transiently expressed in Nicotiana benthamiana plants and subjected to mammalian-specific mucin-type O-glycosylation by co-expressing the pharmaceutical protein together with the glycosylation machinery responsible for such post-translational modification.

Conclusions

The pharmaceutical glycoprotein human Granulocyte-Colony Stimulating Factor can be expressed in N. benthamiana plants via agroinfiltration with its native mammalian-specific mucin-type O-glycosylation.

Background

Plants have emerged as alternative expression systems for the production of pharmaceutical proteins [1,2,3,4]. While there are several advantages to using plants as expression systems, such as low cost of production and maintenance, fast scalability, biological safety, and proper protein folding and assembly [1,2,3, 5], there are also some limitations concerning bioactivity or quality of the produced proteins [6]. Although plants are known to perform post-translational modifications, just like mammalian cells do, differences in glycosylation patterns between plant and animal cells might represent a major drawback that can impact the biochemical properties of the plant-derived recombinant protein [5,6,7,8,9,10].

With regard to the differences in glycosylation patterns between humans and plants, plant N-glycosylation differs mainly in the attachment of α(1,3)-fucose and β(1,2)-xylose to the core N-glycan, while N-glycans in human and mammalian cells present α(1,6)-fucose and no xylose is attached. Further maturation of the plant core N-glycan typically results in a biantennary structure, occasionally with terminal Lewis A epitopes. Terminal Lewis A epitopes are rarely observed in human proteins, but are widely distributed in plant glycoproteins. Maturation of human N-glycans will present a multi-antennary structure with two or more terminal branches, further extended with galactose and sialic (neuraminic) acid [4, 11]. Major efforts to humanize N-glycosylation patterns in plants have focused on retaining the recombinant protein in the ER to avoid further modification of the core N-glycan with α(1,3)-fucose, β(1,2)-xylose and Lewis A, abolishing expression of glycosyltransferases responsible for the attachment of α(1,3)-fucose and β(1,2)-xylose, and further humanization has included the expression of human β(1,4)-galactosyltransferase and heterologous enzymes required to generate sialylation, which are present in human cells but absent in plants [4, 11,12,13,14].

In the case of O-glycosylation, patterns differ significantly between plants and human cells, and specifically from the human mucin-type O-glycosylation. In plants, protein O-glycosylation is characterized by the presence of O-glycans attached to hydroxyprolines (Hyp) and Ser residues. Typically, Hyp residues in plants are decorated with arabinogalactan polysaccharides or arabinose oligosaccharides, and Ser residues with galactose [6, 11, 15, 16]. While human mucin-type O-glycosylation is characterized by the attachment of N-acetylgalactosamine (GalNAc) to Ser or Thr residues, which can be further elongated with other sugars to form specific and highly complex glycans. Although mucin-type O-glycosylation is widely present in human proteins, plants lack the enzymatic machinery to perform this specific glycosylation [11, 15, 16].

Recently, a few, but significant, advances in engineering O-glycosylation in plants to produce mucin-type glycans were reported. Among the recent achievements, Daskalova et al. [6] successfully expressed an O-glycosylated human Mucin 1 peptide derivative, which was detected exclusively as a glycoform in Nicotiana benthamiana. This was achieved by co-expressing the Yersinia enterocolitica UDP-GlcNAc 4-epimerase, in charge of converting UDP-GlcNAc to UDP-GalNAc in the cytoplasm, and a Caenorhabditis elegans UDP-GlcNAc/UDP-GalNAc transporter, responsible for the transport of the sugar donor to the Golgi lumen, together with the human GalNAc transferase 2.

In 2012, Yang et al. [10] were also able to generate mucin-type O-glycosylation in N. benthamiana, using a similar approach. They transiently expressed a Pseudomonas aeruginosa UDP-GlcNAc 4-epimerase, together with the human GalNAc-T 2 and 4, and a human 3.5 tandem repeat of Mucin1. The derivative mucin peptide was GalNAc-O-glycosylated with up to three and five GalNAc residues, when expressed with GalNAc-T2 and GalNAc-T 2 together with GalNAc-T 4, respectively. The mucin-type O-glycosylation was also demonstrated by mass spectrometry on a tandem repeat of MUC16 and on interferon α2b. These authors [17] were also able to establish mucin-type O-glycosylation in Arabidopsis thaliana and Nicotiana tabacum BY-2 cells, and encountered a high degree of proline hydroxylation and Hydroxyproline linked arabinosides, plant-specific O-glycosylation, on the substrate model protein.

The human Granulocyte-Colony Stimulating Factor (G-CSF) is a cytokine that stimulates the production, proliferation, differentiation and activation of neutrophil stem cells, raising the levels of neutrophils in the blood stream, and thus protecting the organism against bacterial, fungal and viral infections [18,19,20,21,22]. This cytokine is an important glycoprotein that is used as treatment in patients with neutropenia to reduce opportunistic infections, especially in cancer patients undergoing chemotherapy or radiotherapy [20, 23, 24]. Due to this clinical relevance, this protein is one of the most widely sold pharmaceuticals, thus making its production for clinical use very important [24,25,26,27]. G-CSF is mucin-type O-glycosylated at a single Thr residue. It has been successfully expressed in CHO cells, Escherichia coli, yeast and plants [28,29,30,31,32]. Nevertheless, only the mammalian platform variant is reported to be mucin-type O-glycosylated, while the yeast-derived variant is decorated with mannose residues. Several studies have shown the impact of glycosylation on protein stability [33,34,35,36,37], pointing out the relevance of production and correct glycosylation of this pharmaceutical.

In this project, our objective is to transiently express the human Granulocyte-Colony Stimulating Factor as a model pharmaceutical protein in N. benthamiana plants, via Agrobacterium infiltration, together with the genes required for the synthesis of N-acetylgalactosamine (GalNAc)-O-glycosylation, allowing the production of a plant-derived human-specific O-glycosylated G-CSF.

Experimental procedures

Construction of binary vectors

The coding sequence corresponding to the human G-CSF variant 2 (NCBI: NM_172219) was chemically synthesized (GenScript, Piscataway, NJ, USA) in three different fragments to build the variants used in this study. The three fragments consisted of the nucleotide sequences corresponding to the secretory signal peptide (pUC57 Nat-SP), the mature G-CSF fragment with a mutated glycosylation site that encodes an Ala, instead of the native Thr, and an extra 5′-end Met (pUC57 G-CSF A), and a 5′-end enterokinase coding sequence fused to the mature G-CSF fragment with the native glycosylation site (pUC57 EkG-CSF). The Sec-G-CSF coding sequence was built by digesting the pUC57 Nat-SP with NcoI and Bsu361 and using the digested fragment to replace the 5′-end EkG-CSF fragment of the pUC57 EkG-CSF, which was released with the same restriction enzymes, producing the pUC57 Sec-G-CSF. The Cyt-G-CSF coding sequence was built by releasing the 3′-end of the pUC57 G-CSF A with SapI cutting upstream the non-glycosylation site and downstream the stop codon. The same restriction enzyme was used to digest the pUC57 EkG-CSF, releasing the same fragment but containing the native glycosylation site (Thr), and the released fragment was cloned into the SapI opened pUC57 G-CSF A, generating the pUC57 Cyt-G-CSF with the glycosylation site. The Zera-G-CSF coding sequence was built by digesting the pUC57 EkG-CSF with NcoI and SacI, releasing the whole fragment, and ligating it into a pUC18 ZeraF1V vector available in the lab [38], by releasing the F1V fragment with the same enzymes and thus generating the pUC18 Zera-G-CSF. The different G-CSF coding sequences generated were then PCR amplified to introduce the flanking BbsI A (5′-end Sec-G-CSF: Mt114F—GGGGGGGAAGACATCATGGCTGGACCTGCCACTC), (5′-end Cyt-G-CSF: Mt23F—GGGGGGGAAGACATCATGACACCCTTAGGACC), (5′-end Zera-G-CSF: Mt20F—GGGGGGGAAGACGTCATGAGGGTGTTGCTCGTTGC) and B (3′-end of all G-CSF variants: M113R—CCCCCCGAAGACAGAGCTCGGGTTGAGCAAGGTGAC) sites, and removal of the stop codon to allow in frame fusion with the appropriate tag. The BbsI A site contains the recognition sequence for the restriction enzyme, followed by the specific overhang CATG, and the BbsI B site contains the specific overhang AGCT, allowing future GoldenGate cloning and disruption of the stop codon for further in-frame cloning with the respective tag. The resulting amplified product was cloned in pUC57 vector via EcoRV, and the mutated sites were verified by DNA sequencing and restriction digestions. The resulting pUC57 G-CSF variant without Stop codon (pUC57 Sec-G-CSF BbsI AB w/o Stop, pUC57 Cyt-G-CSF BbsI AB w/o Stop, pUC57 Zera-G-CSF BbsI AB w/o Stop) was then used to clone the coding sequence into the pENTR4 BbsI AB vector (a GoldenGate compatible pENTR4 vector generated in the lab), via GoldenGate, using the restriction enzyme BbsI and DNA ligase in a single cutting-ligation reaction. The resulting pENTR4 G-CSF variant without Stop codon (pENTR4 Sec-G-CSF w/o Stop, pENTR4 Cyt-G-CSF w/o Stop, pENTR4 Zera-G-CSF w/o Stop) was then used to clone the corresponding coding sequence into the binary vector via Gateway LR reaction, allowing the in-frame cloning of the G-CSF variant with the desired C-terminal tag, eYFP (pGWB 641) or c-Myc (pGWB 617) [39]. The binary vectors were used to transform Agrobacterium tumefaciens strain AGL-1.

Other constructs used in this work were the suppressor of post-transcriptional gene silencing p19 from Cymbidium ringspot tombusvirus [40] to increase accumulation levels; SecGFP (as protein secretion control), SecGFP:KDEL (for ER localization signal) previously published [41], and STtmd:mRFP (as Golgi marker for co-localization) [42]. Agrobacterium strain AGL-1 was used for pGWB 642, and Agrobacterium strain EHA105 was used for p19, GFP, and STtmd:mRFP constructs. The binary vectors used to generate O-glycosylation were as follows: pH7WG2 GNE, containing the Y. enterocolitica gne gene coding for UDPGlcNAc-4 epimerase; pH7WG2 GT, containing the C. elegans UDPGlcNAc/UDPGalNAc Transporter gene; and pH7WG2 GNT2, containing the human GalNAc Transferase 2 coding sequence, previously published [6]. All were electroporated into the Agrobacterium strain AGL-1.

Agrobacterium infiltration

Agrobacterium tumefaciens strains were cultured to an optical density at 600 nm (OD600) of 0.5–0.8. The cells were then collected by centrifugation at 1000 g for 30 min. The pellets were resuspended in Agro-infiltration solution (3.2 g/L Gamborg´s B5 medium and vitamins, 20 g/L sucrose, 10 mM MES pH 5.6, 200 μM 4′-Hydroxy-3′-5′-dimethoxyacetophenone) to a final OD600 of 1.0, followed by incubation at room temperature for 1 h, with gentle agitation. The suspension was then used for needle-less infiltration of the abaxial leaf epidermis through the stomata of N. benthamiana plants [43].

Confocal analysis

Protein subcellular visualization was determined by imaging the abaxial epidermal cells of leaf samples, with an Olympus LSM FV1200. Different lasers allowed for imaging of the different fluorescent tag fusion proteins. For GFP imaging, the tag was excited with a 488 Argon laser and detected at 500–545 nm. For eYFP imaging, the tag was excited at 515 and detected at 530–545 nm. For mRFP imaging, the tag was excited at 559 nm with a He/Ne laser and detected at 570–545 nm. The Imaris software (version 7.6.1, Bitplane Scientific Software, Bitplane, Zurich, Switzerland) was used to generate 3D images from z-stack confocal images. Line-sequential scanning mode was used for co-localization imaging.

Preparation of total protein extracts

Four leaf discs (approximate fresh weight of 30 mg) from at least three biological replicates per sample were collected with a 7 mm diameter cork borer, put into a 2 ml tube with three 2.3 mm zirconia/silica beads (Bio Spec Products Inc, Cat. No. 11079125z) and frozen in liquid N2. Collected leaf discs were pulverized in a Mixer Mill (Retsch, Haan, Germany) for 1 min at 30 Hz, in previously frozen homogenizer blocks. Pulverized tissue was spun down for 1 min at 1000 g and 300 μL Protein Extraction Buffer (PBST0.1%, 2% PVPP, 1 mM EDTA pH 8.0, 1 mM PMSF, 1 μg/mL Leupeptin, 100 mM Sodium L-ascorbate) or Reducing Extraction Buffer (50 mM Tris, pH 8.0, 1% SDS, 20 mM DTT) were added to the sample. Samples were vigorously vortexed 3 times for 5 s and centrifuged at 20,000g for 15 min at 4 °C. Cleared supernatant was transferred to a new tube and TSP (Total Soluble Protein) was quantified using the Bio-Rad Bradford [44] protein assay reagent (Bio-Rad, Cat. No. 5000006).

SDS-PAGE western blotting and lectin blot

Serological assays were performed by resolving the protein sample in a sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to a PVDF or Nitrocellulose membrane. Recombinant proteins were detected with a 1:5000 dilution of the primary mouse anti-c-Myc monoclonal antibody (GenScript, Cat. No. A00864) or 1:5000 dilution of the primary rabbit anti-GCSF polyclonal antibody (GeneTex, Cat. No. GTX31157), and 1:3000 dilution of the goat anti-mouse IgG HRP-conjugated secondary antibody (Bio-Rad, Cat. No. 170-6516) or 1:3000 dilution of the goat anti-rabbit IgG-conjugated secondary antibody (Bio-Rad, Cat. No. 1706515). In the case of lectin blots, a 1:1000 VVA-HRP conjugate (EY Laboratories, Cat. No. H-4601-1) was used. Blotted membranes were visualized with the enhanced chemiluminescence (ECL) detection kit (GE Healthcare, Cat. No. RPN2232), following the manufacturer recommendations, and imaged with the DNR Bio-Imaging System MicroChemi (RANCOM A/S, Birkerød, Denmark). Band analysis was performed using the TotalLab TL 100 software (Nonlinear Dynamics, Durham, NC).

Protein purification

Protein extract was transferred to a spin column, and anti-c-Myc beads were added to the column, following the manufacturer instructions (MBL International, Cat. No. 3305). The sample was incubated at 4 °C for 1 h, with gentle agitation end-over-end. The column was briefly centrifuged for 10 s, and the flow through was recovered. Three washes were performed by adding 200 μL of the washing buffer provided with the kit, and briefly centrifuging. Finally, 20 μL of elution peptide was added to the column, the sample was incubated for 5 min at 4 °C, and the purified protein was recovered by centrifuging the column for 10 s. Elution was repeated twice.

Analysis of O-glycans by mass spectometry

SDS-PAGE excised bands were Trypsin/Chymotrypsin (6.25 mg/L each) digested [45] at 30 °C overnight. The peptide digests were analyzed using an Easy-nLC 1000 nano-flow system with a 100 µm × 2 cm Acclaim C18 PepMap™ trap column and a 75 µm × 15 cm Acclaim C18 PepMap™ analytical (Thermo Scientific, MA, USA) coupled to a Q-Exactive™ Quadrupole Orbitrap mass spectrometer (Thermo Scientific, MA, USA). The flow rate was 300 nL min−1 and 10 µL of the protein digest was injected. 97% mobile phase A (LC/MS Optima water, 0.1% formic acid) was decreased to 90% over 3 min. Peptides were eluted with a linear gradient from 10 to 35% mobile phase B (LC/MS Optima acetonitrile 0.1% formic acid) over 21 min followed by 35–90% over 3 min and maintained for 8 min. The nanospray voltage was set at 2.1 kV, capillary temperature 275 °C, and S-lens RF level 55. The Q-Exactive was operated in top 10 data-dependent acquisition mode with a full scan mass range of 400–2000 m/z at 70,000 resolution, automatic gain control (AGC) of 1e6 and maximum injection time (IT) of 250 ms. The MS/MS scans were acquired at 17,500 resolution, AGC of 2e5, maximum IT of 50 ms, intensity threshold of 8e4, normalized collision energy of 27 and isolation window of 1.2 m/z. Unassigned, singly and > 4 charged peptides were not selected for MS/MS and a 20 s dynamic exclusion was used. The Thermo .raw files were converted to mascot generic format using Proteowizard v2 [46] and the MS/MS scans were searched against the target/reverse human G-CSF amino acid sequence and the N. benthamiana proteome (Sol Genomics Network, accessed Jan 10th, 2015) using X! tandem [47] search algorithm operated from the SearchGUI v.2.35 [48] interface and processed in PeptideShaker v1.3.6 [49]. A 3 ppm precursor ion mass error and a 0.02 Da product ion error were used along with carbamidomethylation as a constant modification and oxidation of methionine, Hex(1)NAc(1) of Thr, HexNAc of Thr as variable modifications. A 1% FDR rate was used at the protein, peptide and peptide spectrum match level.

Statistical analysis

The statistical analyses were performed with the Minitab 18 software (Minitab Inc., PA, USA). A one-way analysis of variance (one-way ANOVA) was performed followed by Tukey test to find significance differences between means (statistical difference was defined as p ≤ 0.05).

Results

  1. 1.

    Transient expression of G-CSF in N. benthamiana leaves.

To explore the feasibility of using N. benthamiana transient expression for the production of a mammalian glycoprotein, like G-CSF, several factors need to be explored, such as the accumulation levels of the recombinant protein, proper protein folding and native post-translational modifications. To address these issues, the native secretory human G-CSF (Sec-G-CSF) coding sequence, including its native secretory signal sequence, was used to build the corresponding expression cassettes, in order to target the model protein to the secretory pathway (Fig. 1). Two other G-CSF variants were also built, which would target the recombinant protein to the cytosol and as ER-derived protein bodies. These variants were used as controls to determine protein subcellular localization, as well as to evaluate the impact of protein accumulation. The cytoplasmic variant (Cyt-G-CSF) corresponds to the mature amino acid sequence with an N-terminal methionine, allowing the accumulation of the protein in the cytosol. The protein body variant (Zera-G-CSF) was achieved by fusing the N-terminus of the mature G-CSF coding sequence (no signal peptide) to the γ-Zein ER-accumulating domain: Zera® peptide [38]. The different constructs were fused to an eYFP tag in order to track expression and proper subcellular localization of the recombinant protein. All of the G-CSF variant constructs-containing Agrobacterium strains were co-infiltrated with an Agrobacterium strain containing the p19 construct, a post-transcriptional gene silencing suppressor from Cymbidium ringspot tombusvirus (CymRSV) [40].

Fig. 1
figure 1

Schematic representation of constructs generated in this work and used for Agrobacterium-mediated transient expression in N. benthamiana leaves. Sec-G-CSF:eYFP, secretory G-CSF variant fused to C-terminal eYFP; Cyt-G-CSF:eYFP, cytoplasmic G-CSF variant fused to C-terminal eYFP; Zera-G-CSF:eYFP, G-CSF fused to N-terminal Zera peptide and C-terminal eYFP; Sec-G-CSF:cMyc, secretory G-CSF fused to C-terminal c-Myc4x. 35S-P, Cauliflower Mosaic Virus 35S promoter; nos-T, nopaline synthase transcription terminator; G-CSF, human Granulocyte-Colony Stimulating Factor coding sequence; Nat-SP, native secretory signal peptide from human G-CSF; Zera, γ-Zein ER-accumulating domain; Ek, enterokinase cleavage site; eYFP, enhanced yellow fluorescent protein for confocal and serological detection; c-Myc4x, four repeats of the c-Myc tag for detection and purification; attB1 and attB2, linker regions derived from recombination sites generated after Gateway cloning. Schematic representation not drawn to scale

Full size image

Expression of G-CSF:eYFP was first detected by confocal microscopy (Fig. 2a–l). Infiltrated N. benthamiana leaves were monitored from 2 to 8 dpi (days post-infiltration). Sec-G-CSF:eYFP was detected in the ER (endoplasmic reticulum), as a multiple dot-like pattern indicating a potential localization in the Golgi, as well as in the apoplast (Fig. 2a–c), suggesting localization in the secretory pathway. The cytoplasmic variant, Cyt-G-CSF:eYFP, was detected in the cytosol, and no signal was detected in the apoplast (Fig. 2d–f). Cytoplasmic eYFP (CyteYFP) was used as control for cytoplasmic accumulation pattern, also allowing the visualization of an apoplast with no presence of recombinant protein (Fig. 2g–i). Finally, the Zera-G-CSF:eYFP was found forming protein bodies due to the N-terminal fusion of the Zera® peptide (Fig. 2j–l).

Fig. 2
figure 2

Transient expression of G-CSF:eYFP variants in N. benthamiana plants. Expression of Sec-G-CSF:eYFP (a–c), Cyt-G-CSF:eYFP (d–f), CyteYFP (g–i), Zera-G-CSF:eYFP (j–l). White arrows indicate apoplast localization (c) or absence of signal in the apoplast (e, f, h, i). White arrow indicating Golgi localization (b). Inset white boxes (e, f) depict microscopic zoom in shown in f and i, respectively. Samples collected at 4 dpi. Size bar: 5 μm. m Western Blot detection of Sec-G-CSF:eYFP (Secretory variant, 49 kDa), Cyt-G-CSF:eYFP (Cytoplasmic and mature protein variant, 49 kDa), Zera-G-CSF:eYFP (Zera fused variant, 61 kDa), eYFP (Cytoplasmic eYFP, 29 kDa). All samples were treated with PBS + Tween (PBST) extraction buffer and reducing extraction buffer (SDS-DTT). n Band quantification of SDS-DTT treated samples. 20 µg TSP of PBS samples were loaded on the gel. 9 µL (equalized volume to the smallest PBS sample amount) of SDS-DTT samples was loaded on the gel. Black arrows denote monomeric G-CSF:eYFP variants. eGEHK: protein standard. Proteins were detected with GFP antibody. Samples collected at 6 dpi. Columns denoted with a different letter are significantly different (p ≤ 0.05) using one-way ANOVA and followed by Tukey test. Error bars are standard deviation of the means

Full size image

Identity of the transiently expressed G-CSF:eYFP recombinant proteins was further corroborated by western blot (Fig. 2m, n). Sec-G-CSF:eYFP was detected as a monomeric form with an apparent molecular weight of 49 kDa, which corresponds to the mature G-CSF:eYFP protein (after signal peptide cleavage). The same molecular weight was observed with Cyt-G-CSF:eYFP, confirming signal peptide processing of the Sec-G-CSF:eYFP. Furthermore, multimers were observed with Cyt-G-CSF:eYFP. In the case of Zera-G-CSF:eYFP a weak band was detected when protein extraction was performed with PBST (PBS + Tween), but the Zera® fusion variant was properly extracted when the plant tissue sample was treated in denaturing and reducing conditions, using an SDS-DTT containing extraction buffer. Multimers were also observed with Zera-G-CSF:eYFP. eYFP cleavage was observed with all the different variants, and the cleaved eYFP band was corroborated using the CyteYFP expressing samples. It has to be noted that both Sec-G-CSF and Cyt-G-CSF were successfully extracted with PBST extraction buffer (Additional file 1: Fig. S3), but Zera-G-CSF could only be properly extracted with a reducing extraction buffer (Fig. 2; Additional file 2: Fig. S2).

G-CSF:eYFP accumulation was then assessed by densitometry analysis (Figs. 2n, 3a–d), with samples collected at 4, 6 and 8 dpi for each variant. We found that Sec-G-CSF reached an accumulation level of 17 mg/Kg F.W. (Fresh Weight), followed by Cyt-G-CSF:eYFP with 40 mg/Kg F.W., while Zera-G-CSF:eYFP reached 117 mg/Kg F.W., at 6 dpi (Figs. 2n, 3; Additional file 1: Fig. S3). We also observed that accumulation was stable at 4 and 6 dpi for Sec-G-CSF and Cyt-G-CSF, beginning to drop at 8 dpi, but not significantly (Fig. 3). There is no significant impact concerning efficiency of the extraction buffer of Sec-G-CSF (PBST or SDS-DTT) (Additional file 1: Fig. S3), but higher levels of accumulation were observed with Cyt-G-CSF when samples were treated with PBST, in comparison to SDS-DTT (Additional file 1: Fig. S3). Zera-G-CSF was the only variant that showed a significant increase in accumulation levels on the time-course, in comparison to the other two variants, reaching up to 135 mg/Kg F.W. at 8 dpi (Fig. 3), but as previously mentioned Zera-G-CSF could only be properly extracted under denaturing and reducing conditions using an extraction buffer containing SDS-DTT (Figs. 2, 3; Additional file 2: Fig. S2). This result was expected since previous studies indicate the usage of denaturing and reducing conditions for extraction of recombinant proteins fused to Zera [38, 50, 51]. The previous observations highlight the vulnerability of G-CSF to proteolytic degradation, and especially in the case of the secretory G-CSF.

Fig. 3
figure 3

Time-course accumulation of transiently expressed G-CSF:eYFP variants in N. benthamiana plants. Western blot detection of Sec-G-CSF:eYFP (Secretory variant, 49 kDa) (a), Cyt-G-CSF:eYFP (Cytoplasmic and mature protein variant, 49 kDa) (b), Zera-G-CSF:eYFP (Zera fused variant, 61 kDa) (c). d Band quantification of Western blot detected proteins. Samples collected at 4, 6 and 8 dpi. Four leaf discs from three different plants were collected for each sample. All samples were treated with reducing extraction buffer (SDS-DTT). Equal amounts were used for all samples. Black arrows denote monomeric G-CSF:eYFP variants. eGEHK: protein standard. Proteins were detected with GFP antibody. Columns denoted with a different letter are significantly different (p ≤ 0.05) using one-way ANOVA and followed by Tukey test. Error bars are standard deviation of the means

Full size image
  1. 2.

    Sec-G-CSF localization in the secretory pathway.

To determine the feasibility of producing a mucin-type O-glycosylated G-CSF recombinant protein, we then corroborated that the native secretory signal peptide used in this work was targeting G-CSF to the secretory pathway. For this purpose, we determined the co-localization of Sec-G-CSF with Rat sialyltransferase, a Golgi resident enzyme, fused with monomeric red fluorescent protein (STtmd:mRFP) as a Golgi marker, which would suggest trafficking of secretory G-CSF along the secretory network (Fig. 4). Sec-G-CSF:eYFP was observed to localize in the ER as previously observed (Fig. 2) but also localized in the Golgi together with STtmd:mRFP (Fig. 4a–c). Although only a few Golgi bodies were captured where Sec-G-CSF would co-localize with the Golgi marker, a previously tested secretory GFP (SecGFP) construct from the lab [52] was used as a positive control for co-localization in the Golgi, and a similar pattern was observed, being SecGFP also localized in the ER (Fig. 4e–g). Thus, both Sec-G-CSF:eYFP and SecGFP were detected throughout the secretory pathway, in the ER, Golgi and in the apoplast (Fig. 4a–h). As a negative control for secretion, an ER-retained GFP (GFP:KDEL) construct was used, which would be directed to the secretory pathway but retrieved to the ER. The GFP:KDEL negative control was not found to co-localize with STtmd:mRFP, and it was not observed in the apoplast (Fig. 4i–k). The previous observations indicate that the native G-CSF secretory signal peptide indeed targets G-CSF to the secretory pathway. The same constructs were used to monitor apoplast signal, and only Sec-G-CSF:eYFP and SecGFP were observed in the apoplast, while GFP:KDEL was not, corroborating the previous observations (Fig. 4d, h, l). Although Sec-G-CSF was not exclusively observed in the Golgi, where O-glycosylation is traditionally referred to begin (Additional file 3: Fig. S1), the fact that the recombinant protein was observed in the apoplast suggested secretion of the protein to certain extent, thus trafficking from the ER to the apoplast. Moreover, some studies [53,54,55,56] have also suggested that the attachment of GalNAc to substrate proteins might also occur in regions of the ER (Additional file 3: Fig. S1), therefore we could consider to subject Sec-G-CSF to the O-glycosylation machinery.

Fig. 4
figure 4

a, b Golgi and apoplast localization of Sec-G-CSF in N. benthamiana plants. Sec-G-CSF:eYFP (A, C, D), SecGFP (E, G, H), GFP:KDEL (I, K, L). Sec-G-CSF, SecGFP and GFP:KDEL were co-infiltrated with STM:mRFP (B, C, F, G, J, K). White arrows depict Golgi localization (A, B, E, F) and colocalization (C, G) with STM:mRFP. White arrowheads (images on the right) depict localization in apoplast (D, H), or apoplast without presence of recombinant protein (I). Agrobacterium AGL-1 was carrying Sec-G-CSF. SecGFP, GFP:KDEL, STM:mRFP and p19 were delivered by EHA105. 4 dpi. Size bar: 5 μm. a shows the colors as originally displayed, while b shows a color palette adjusted for those who have difficulty viewing red and green colors

Full size image
  1. 3.

    O-Glycosylation of Sec-G-CSF.

Once we confirmed that Sec-G-CSF was targeted for secretion, the Sec-G-CSF coding sequence was co-expressed with the O-glycosylation machinery (glyco-machinery) comprised by the Y. enterocolitica UDP-GlcNAc/GalNAc 4-epimerase, the C. elegans UDP-GlcNAc/GalNAc transporter, and the human GalNAc-Transferase 2, to allow the mammalian-specific GalNAc-O-glycosylation of G-CSF. In this case, the eYFP tag was replaced by a C-terminal c-Myc4x tag (Sec-G-CSF:cMyc) for detection and purification purposes. Expression of Sec-G-CSF:cMyc was determined by western blot with a molecular weight of 28 kDa (Fig. 5a). Sec-G-CSF:cMyc reached accumulation levels of 49 mg/Kg F.W. and 31 mg/Kg F.W., at 4 and 6 dpi, respectively (Fig. 5b).

Fig. 5
figure 5

Western and Lectin Blot of O-glycosylated Sec-G-CSF. a Western blot detection of Sec-G-CSF:cMyc (Secretory variant, 28 kDa). Samples collected at 4 and 6 dpi. Four leaf discs from three different plants were collected for each sample. 20 µg TSP of PBS samples were loaded on the gel. Black arrow denotes monomeric Sec-G-CSF:cMyc. G-CSF: protein standard. Proteins were detected with G-CSF antibody. b Band quantification of Western blot detected Sec-G-CSF:cMyc. c c-Myc Detection: 0.25 µL eluent (Sec-G-CSF without glyco-machinery: -) after c-Myc purification; and 2.5 µL eluent (10x more than negative control) (Sec-G-CSF with glyco-machinery: +) after c-Myc purification. Detection using the c-Myc antibody. GalNAc Detection (Lectin Blot):: Equalized amounts of c-Myc purified Sec-G-CSF (without and with glyco-machinery: -, +). GalNAc detection using VVA-HRP. GelCode Blue: GelCode Blue Staining of c-Myc purified Sec-G-CSF (without and with glyco-machinery: -, +). Sec-G-CSF:cMyc co-expressed with p19 and glyco-machinery (∓) (O-Glycosylation machinery: UDP-GlcNAc/GalNAc epimerase, UDP-GlcNAc/GalNAc transporter, and GalNAc-glycosyltransferase) (OD600 ratios; Sec-G-CSF:cMyc: 0.5; p19: 0.1; GNT2: 0.2; epimerase/transporter: 0.1/strain). Columns denoted with a different letter are significantly different (p ≤ 0.05) using one-way ANOVA and followed by Tukey test. Error bars are standard deviation of the means

Full size image

Once expression of Sec-G-CSF:cMyc was assessed, the secretory variant was then co-expressed with the O-glycosylation machinery. Sec-G-CSF:cMyc co-expressed with and without the O-glycosylation machinery was purified by c-Myc monoclonal antibody affinity chromatography, and analyzed via SDS-PAGE, c-Myc detection by western blot and GalNAc-O-glycosylation detection by lectin blotting (Fig. 5c). Sec-G-CSF:cMyc was detected both when co-expressed with or without the O-glycosylation machinery by SDS-PAGE and western blot using a c-Myc specific antibody (Fig. 5c). For detection of the O-glycosylated residue, the lectin Vicia villosa agglutinin conjugated with HRP (VVA-HRP), which is specific for GalNAc-O-T/S moieties, was used as probe (Fig. 5c). VVA-HRP reacted only with the Sec-G-CSF:cMyc co-expressed with the glycosylation machinery, but not with the Sec-G-CSF:cMyc expressed alone, indicating proper GalNAc-O-glycosylation of Sec-G-CSF:cMyc. To further corroborate the specific attachment of the GalNAc to the Thr-163 residue (Additional file 4: Fig. S4), Sec-G-CSF:cMyc co-expressed with and without the O-glycosylation machinery was purified and in-gel Trypsin/Chymotrypsin digested. A Trypsin/Chymotrypsin digestion system was used to generate a smaller peptide around the predicted O-glycosylation site that would be more amenable to MS/MS analysis. The resulting peptides were identified by nano LC–MS/MS and GalNAc attachment to the specific Thr-163 site was confirmed by the identification of an N-acetylhexosamine (HexNAc) derived from the peptide where the Thr-163 is localized, as indicated by the characteristic HexNAc product ion at m/z 204 (Additional file 4: Fig. S4; Additional file 5: Table S1). Nevertheless, Sec-G-CSF expressed alone could not be detected during the MS/MS analysis; moreover, analyses targeted to detect plant-specific post-translational modifications, such as proline hydroxylation and further plant-specific O-glycosylation, which are likely to occur during secretion of recombinant proteins, were not carried out in the present study. Thus, future studies will have to be carried out in order to fully characterize the recombinant protein.

Discussion

Accumulation has been a major focal point for the production of recombinant proteins in plants [4, 38, 57, 58]. Although it may be assumed that post-translational modifications, such as glycosylation which might have an impact on protein stability or activity [6, 11], would be produced by plants and other eukaryotic expression platforms, most studies concerning expression of glycoproteins do not address this issue [9, 11, 16, 32, 59,60,61,62,63]. In this study, we address the feasibility of introducing the mammalian-specific mucin-type O-glycosylation to the pharmaceutical glycoprotein human G-CSF, in N. benthamiana plants via transient expression.

G-CSF is commercially produced in E. coli and CHO cells, well-established expression platforms. The bacterial system is capable of producing the recombinant protein in its mature form, with an extra N-terminal Met, and the protein is not glycosylated [35, 64]. The mammalian system is capable of decorating the recombinant protein with Gal-GalNAc-O-glycosylation with up to two sialic acids, at the single Thr residue reported to be mucin-type O-glycosylated in humans [35, 65].

Studies addressing the glycosylation differences between these two commercially available G-CSF’s have determined that lack of glycosylation has an impact on the protein’s susceptibility to degradation, decreasing the biological activity of the cytokine, and thus an increased amount of the non-glycosylated form is required to obtain a similar biological effect as that obtained with the glycosylated form [33, 34, 37]. Other studies have also suggested that the lack of glycosylation might promote protein aggregation, which would hamper the biological activity of the protein [30, 66,67,68].

Due to the pharmaceutical relevance of this glycoprotein, several alternative production platforms have been considered. Among these platforms, yeasts have been explored due to the secretory system they possess, which is similar to that of higher order eukaryotes [29, 69]. Recombinant hG-CSF has been successfully expressed in Saccharomyces cerevisiae and Pichia pastoris [29, 31, 66], but recent analysis reported that the recombinant P. pastoris-produced protein was not mucin-type O-glycosylated. Instead, the O-glycosylation Thr site used in mammalian cells was mannose O-glycosylated [70]. In the case of the S. cerevisiae-derived recombinant protein, glycosylation was apparently absent [66]. Furthermore, the yeast-derived recombinant protein was found mainly as multimers, which were not biologically active and had to be denatured and re-natured to recover proper protein folding and regain biological activity. The issue with multimerization is believed to result from lack of proper protein folding, as it has also been observed when G-CSF was expressed using the filamentous fungus Aspergillus niger, where glycosylation could not be determined [67].

Recently, G-CSF production has been achieved in tobacco and rice suspension cells [59, 62] and tobacco plants [60], targeting the protein to the secretory pathway; ER-retained in BY-2 cells [61]; in lettuce chloroplasts [32]; and transiently in N. benthamiana [63]. These studies do not report multimerization as it was observed in the case of the yeast and A. niger reports, and the protein was proved to be biologically active. However, glycosylation was not addressed in any of the previous studies.

In the present study, we have successfully expressed recombinant GalNAc-O-glycosylated G-CSF in N. benthamiana plants. Mammalian-specific O-glycosylation of the plant-derived G-CSF was achieved by targeting the recombinant protein to the secretory pathway, using its native secretory signal peptide and co-expressing it with a GalNAc-O-glycosylation machinery [6]. The recombinant secretory G-CSF (Sec-G-CSF) was confirmed to be GalNAc-O-glycosylated via lectin blot. Further mass spectrometry analysis corroborated the proper GalNAc-O-glycosylation at the single Thr residue reported to be mucin-type O-glycosylated in the native protein. Although plant-specific glycosylation is known to occur on proteins targeted to the secretory pathway, more specifically O-glycosylation at hydroxylated prolines, analyses in the present study were not addressed to characterize such post-translational modifications. Thus, further analytical studies will be required to fully characterize the glycosylation profile of the recombinant protein, as well as to determine glycosylation ratios. Nevertheless, corroboration of proper GalNAc-O-glycosylation of the Sec-G-CSF variant highlights the feasibility to further elongate the GalNAc-O-glycosylation moiety and imitate native mucin-type O-glycosylation. This objective would require the co-expression of additional glycosyltransferases and other proteins, as it has been demonstrated by recent studies where the GalNAc-O-glycosylated sites of recombinant proteins transiently expressed in N. benthamiana plants, were further elongated with Galactose and sialic acid [14,15,16, 71, 72].

In relation to multimerization, Sec-G-CSF did not seem to multimerize, in contrast to the other two variants targeted to the cytoplasm or expressed as ER-derived protein bodies (Cyt-G-CSF and Zera-G-CSF, respectively) (Figs. 2, 3; Additional file 3: Fig. S3). As reported previously by other studies where recombinant G-CSF was expressed in yeast and A. niger [29, 31, 66, 67], multimerization had a negative impact on biological activity and it has been suggested that it is caused by the lack of proper protein folding, by escaping the ER processing [67]. The fact that Sec-G-CSF was glycosylated does not only support the evidence that the native secretory signal peptide did indeed target the recombinant protein to the secretory pathway, since the GalNAc-Transferase is known to locate and exert its activity in the Golgi apparatus or in subregions of the ER (Additional file 3: Fig. S1) [15, 49,50,51,52], but localization of G-CSF in the secretory pathway also suggests proper protein processing by the ER, which together with further glycosylation, might decrease the nature of oligomerization of G-CSF, as suggested by other studies [30, 66,67,68]. Lack of multimerization and proper glycosylation are positively related to biological activity in this case, indicating the potential of the expression platform used in this project for the production of this glycoprotein. Nevertheless, proper studies related to the impact of glycosylation and ER processing to multimerization and biological activity will have to be carried out in future work.

It must also be noticed that the accumulation levels obtained with the secretory G-CSF were the lowest in comparison to the other two variants (Cyt-G-CSF and Zera-G-CSF). Co-expression with the glycosylation machinery decreased the accumulation levels of Sec-G-CSF:cMyc even more (approximately 10–30 × decrease was observed during the experiments, data not shown) (Fig. 5). Such decrease was expected since the expression of several recombinant proteins and their respective markers would cause an exhaustive usage of the translation machinery, therefore it would be advisable to generate a single multiple expression cassette containing only the required transgenes, assuring at the same time that all genes get expressed in the same cells [16]. Likewise, future subcellular relocalization of the recombinant G-CSF together with glycosyltransferases, as well as developing stable transgenic lines with the corresponding glycosylation machinery, should also be considered as an attempt to increase accumulation levels, while allowing the production of mucin-type O-glycosylated G-CSF in plants [4].

Conclusions

In conclusion, we have successfully expressed GalNAc-O-glycosylated G-CSF in N. benthamiana plants via agroinfiltration, offering an alternative system for the production of this pharmaceutical protein with its native mammalian-specific post-translational modification. Further studies would be required to explore elongation of the glycosylation moiety and approaches to increasing accumulation of this recombinant glycoprotein.

References

  1. Daniell H, Singh ND, Mason H, Streatfield SJ. Plant-made vaccine antigens and biopharmaceuticals. Trends Plant Sci. 2009;14:669–79.

    Article  CAS  Google Scholar 

  2. Sharma AK, Sharma MK. Plants as bioreactors: recent developments and emerging opportunities. Biotechnol Adv. 2009;27:811–32.

    Article  CAS  Google Scholar 

  3. Sack M, Hofbauer A, Fischer R, Stoger E. The increasing value of plant-made proteins. Curr Opin Biotechnol. 2015;32:163–70.

    Article  CAS  Google Scholar 

  4. Dicker M, Schoberer J, Vavra U, Strasser R. Subcellular targeting of proteins involved in modification of plant N- and O-glycosylation. In: Castilho A, editor. Methods mol. biol. New York: Humana Press; 2015. p. 249–67.

    Google Scholar 

  5. Faye L, Boulaflous A, Benchabane M, Gomord V, Michaud D. Protein modifications in the plant secretory pathway: current status and practical implications in molecular pharming. Vaccine. 2005;23:1770–8.

    Article  CAS  Google Scholar 

  6. Daskalova SM, Radder JE, Cichacz ZA, Olsen SH, Tsaprailis G, Mason H, et al. Engineering of N. benthamiana L. plants for production of N-acetylgalactosamine-glycosylated proteins–towards development of a plant-based platform for production of protein therapeutics with mucin type O-glycosylation. BMC Biotechnol. 2010;10:62.

    Article  Google Scholar 

  7. Bardor M, Faveeuw C, Fitchette AC, Gilbert D, Galas L, Trottein F, et al. Immunoreactivity in mammals of two typical plant glyco-epitopes, core α(1,3)-fucose and core xylose. Glycobiology. 2003;13:427–34.

    Article  CAS  Google Scholar 

  8. Gomord V, Fitchette AC, Menu-Bouaouiche L, Saint-Jore-Dupas C, Plasson C, Michaud D, et al. Plant-specific glycosylation patterns in the context of therapeutic protein production. Plant Biotechnol J. 2010;8:564–87.

    Article  CAS  Google Scholar 

  9. Jin C, Altmann F, Strasser R, Mach L, Schähs M, Kunert R, et al. A plant-derived human monoclonal antibody induces an anti-carbohydrate immune response in rabbits. Glycobiology. 2008;18:235–41.

    Article  CAS  Google Scholar 

  10. Yang Z, Drew DP, Jørgensen B, Mandel U, Bach SS, Levery SB, et al. Engineering mammalian mucin-type O-glycosylation in plants. J Biol Chem. 2012;287:11911–23.

    Article  CAS  Google Scholar 

  11. Vukušić K, Šikić S, Balen B. Recombinant therapeutic proteins produced in plants: towards engineering of human-type O- and N-glycosylation. Period Biol. 2016;118:75–90.

    Article  Google Scholar 

  12. Khan AH, Bayat H, Rajabibazl M, Sabri S, Rahimpour A. Humanizing glycosylation pathways in eukaryotic expression systems. World J Microbiol Biotechnol. 2017;33:1–12.

    Article  CAS  Google Scholar 

  13. Castilho A, Neumann L, Gattinger P, Strasser R, Vorauer-Uhl K, Sterovsky T, et al. Generation of biologically active multi-sialylated recombinant human EPOFc in plants. PLoS ONE. 2013;8:1–13.

    Article  Google Scholar 

  14. Dicker M, Tschofen M, Maresch D, Konig J, Juarez P, Orzaez D, et al. Transient glyco-engineering to produce recombinant IgA1 with defined N- and O-glycans in plants. Front Plant Sci. 2016;7:18.

    Article  Google Scholar 

  15. Strasser R. Challenges in O-glycan engineering of plants. Front Plant Sci. 2012;3:1–6.

    Article  Google Scholar 

  16. Strasser R. Engineering of human-type O-glycosylation in Nicotiana benthamiana plants. Bioengineered. 2013;4:191–6.

    Article  Google Scholar 

  17. Yang Z, Bennett EP, Jørgensen B, Drew DP, Arigi E, Mandel U, et al. Toward stable genetic engineering of human O-glycosylation in plants. Plant Physiol. 2012;160:450–63.

    Article  CAS  Google Scholar 

  18. Houston AC, Stevens LA, Cour V. Pharmacokinetics of glycosylated recombinant human granulocyte colony-stimulating factor (lenograstim) in healthy male volunteers. Br J Clin Pharmacol. 1999;47:279–84.

    Article  CAS  Google Scholar 

  19. Reeves G. Overview of use of G-CSF and GM-CSF in the treatment of acute radiation injury. Health Phys. 2014;106:699–703.

    Article  CAS  Google Scholar 

  20. Chavez-Tapia NC, Mendiola-Pastrana I, Ornelas-Arroyo VJ, Noreña-Herrera C, Vidaña-Perez D, Delgado-Sanchez G, et al. Granulocyte-colony stimulating factor for acute-on-chronic liver failure: systematic review and meta-analysis. Ann Hepatol. 2015;14(5):631–41.

    CAS  PubMed  Google Scholar 

  21. Hermesh T, Moran TM, Jain D, López CB. Granulocyte colony-stimulating factor protects mice during respiratory virus infections. PLoS ONE. 2012;7:2–9.

    Article  Google Scholar 

  22. Barreda DR, Hanington PC, Belosevic M. Regulation of myeloid development and function by colony stimulating factors. Dev Comp Immunol. 2004;28:509–54.

    Article  CAS  Google Scholar 

  23. Bönig H, Silbermann S, Weller S, Kirschke R, Körholz D, Janssen G, et al. Glycosylated vs non-glycosylated granulocyte colony-stimulating factor (G-CSF)—results of a prospective randomised monocentre study. Bone Marrow Transplant. 2001;28:259–64.

    Article  Google Scholar 

  24. Dale DC. Colony-stimulating factors for the management of neutropenia in cancer patients. Drugs. 2002;62:1–15.

    Article  CAS  Google Scholar 

  25. Duncan N, Hewetson M, Atra A, Dick G, Pinkerton R. An economic evaluation of the use of granulocyte colony-stimulating factor after bone marrow transplantation in children. Pharmacoeconomics. 1997;11:169–74.

    Article  CAS  Google Scholar 

  26. Barnes G, Pathak A, Schwartzberg L. Pharmacoeconomics of granulocyte colony-stimulating factor: a critical review. Adv Ther. 2014;31:683–95.

    Article  CAS  Google Scholar 

  27. Whyte S, Cooper KL, Stevenson MD, Madan J, Akehurst R. Cost-effectiveness of granulocyte colony-stimulating factor prophylaxis for febrile neutropenia in breast cancer in the United Kingdom. Value Health. 2011;14:465–74.

    Article  Google Scholar 

  28. Mire-Sluis AR, Das RG, Thorpe R. The international standard for granulocyte colony stimulating factor (G-CSF). Evaluation in an international collaborative study. Participants of the collaborative study. J Immunol Methods. 1995;179:117–26.

    Article  CAS  Google Scholar 

  29. Suk Yang D, Soon Bae C, Lee J. Production of recombinant human granulocyte-colony-stimulating factor in high cell density yeast cultures. Biotechnol Lett. 1997;19:655–9.

    Article  Google Scholar 

  30. Bae CS, Yang DS, Lee J, Park YH. Improved process for production of recombinant yeast-derived monomeric human G-CSF. Appl Microbiol Biotechnol. 1999;52:338–44.

    Article  CAS  Google Scholar 

  31. Lasnik MA, Porekar VG, Stalc A. Human granulocyte colony stimulating factor (hG-CSF) expressed by methylotrophic yeast pichia pastoris. Pflugers Arch. 2001;442:R184–6.

    Article  CAS  Google Scholar 

  32. Sharifi Tabar M, Akbar Habashi A, Memari HR. Human granulocyte colony-stimulating factor (hg-csf) expression in plastids of Lactuca sativa. Iran Biomed J. 2013;17:158–64.

    PubMed  PubMed Central  Google Scholar 

  33. Carter CRD, Keeble JR, Thorpe R. Human serum inactivates non-glycosylated but not glycosylated granulocyte colony stimulating factor by a protease dependent mechanism: significance of carbohydrates on the glycosylated molecule. Biologicals. 2004;32:37–47.

    Article  CAS  Google Scholar 

  34. Carter CRD, Whitmore KM, Thorpe R. The significance of carbohydrates on G-CSF: differential sensitivity of G-CSFs to human neutrophil elastase degradation. J Leukoc Biol. 2004;75:515–22.

    Article  CAS  Google Scholar 

  35. Höglund M. Glycosylated and non-glycosylated recombinant human granulocyte colony-stimulating factor (rhG-CSF)—what is the difference? Med Oncol. 1998;15:229–33.

    Article  Google Scholar 

  36. Mattii L, Azzarà A, Fazzi R, Carulli G, Chimenti M, Cecconi N, et al. Glycosylated or non-glycosylated G-CSF differently influence human granulocyte functions through RhoA. Leuk Res. 2005;29:1285–92.

    Article  CAS  Google Scholar 

  37. Martin-Christin F. Clinical report Granulocyte colony stimulating factors: how different are they? How to make a decision? Anticancer Drugs. 2001;12:185–91.

    Article  CAS  Google Scholar 

  38. Alvarez ML, Topal E, Martin F, Cardineau GA. Higher accumulation of F1-V fusion recombinant protein in plants after induction of protein body formation. Plant Mol Biol. 2010;72:75–89.

    Article  CAS  Google Scholar 

  39. Nakamura S, Mano S, Tanaka Y, Ohnishi M, Nakamori C, Araki M, et al. Gateway binary vectors with the bialaphos resistance gene, bar, as a selection marker for plant transformation. Biosci Biotechnol Biochem. 2010;74:1315–9.

    Article  CAS  Google Scholar 

  40. Silhavy D, Molnár A, Lucioli A, Szittya G, Hornyik C, Tavazza M, et al. A viral protein suppresses RNA silencing and binds silencing-generated, 21- to 25-nucleotide double-stranded RNAs. EMBO J. 2002;21:3070–80.

    Article  CAS  Google Scholar 

  41. Conley AJ, Joensuu JJ, Menassa R, Brandle JE. Induction of protein body formation in plant leaves by elastin-like polypeptide fusions. BMC Biol. 2009;7:48.

    Article  Google Scholar 

  42. Latijnhouwers M, Hawes C, Carvalho C, Oparka K, Gillingham AK, Boevink P. An Arabidopsis GRIP domain protein locates to the trans-Golgi and binds the small GTPase ARL1. Plant J. 2005;44:459–70.

    Article  CAS  Google Scholar 

  43. Kapila J, De Rycke R, Van Montagu M, Angenon G. An Agrobacterium-mediated transient gene expression system for intact leaves. Plant Sci. 2005;122:101–8.

    Article  Google Scholar 

  44. 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.

    Article  CAS  Google Scholar 

  45. Finnie C. Plant proteomics. Plant Proteomics. 2007;1072:1–253.

    Google Scholar 

  46. Kessner D, Chambers M, Burke R, Agus D, Mallick P. Proteowizard: open source software for rapid proteomics tools development. Bioinformatics. 2008;24:2534–6.

    Article  CAS  Google Scholar 

  47. Craig R, Beavis RC. TANDEM: matching proteins with tandem mass spectra. Bioinformatics. 2004;20:1466–7.

    Article  CAS  Google Scholar 

  48. Vaudel M, Barsnes H, Berven FS, Sickmann A, Martens L. SearchGUI: an open-source graphical user interface for simultaneous OMSSA and X!Tandem searches. Proteomics. 2011;11:996–9.

    Article  CAS  Google Scholar 

  49. Vaudel M, Burkhart JM, Zahedi RP, Oveland E, Berven FS, Sickmann A, et al. PeptideShaker enables reanalysis of MS-derived proteomics data sets. Nat Biotechnol. 2015;33:22–4.

    Article  CAS  Google Scholar 

  50. Torrent M, Llompart B, Lasserre-Ramassamy S, Llop-Tous I, Bastida M, Marzabal P, et al. Eukaryotic protein production in designed storage organelles. BMC Biol. 2009;7:5.

    Article  Google Scholar 

  51. Joseph M, Ludevid MD, Torrent M, Rofidal V, Tauzin M, Rossignol M. Proteomic characterisation of endoplasmic reticulum-derived protein bodies in tobacco leaves. BMC Plant Biol. 2012;12:36.

    Article  CAS  Google Scholar 

  52. Saberianfar R, Joensuu JJ, Conley AJ, Menassa R. Protein body formation in leaves of Nicotiana benthamiana: a concentration-dependent mechanism influenced by the presence of fusion tags. Plant Biotechnol J. 2015;13:927–37.

    Article  CAS  Google Scholar 

  53. Stanley P. Golgi glycosylation. Cold Spring Harb Perspect Biol. 2011;3:1–13.

    Article  Google Scholar 

  54. Gill DJ, Chia J, Senewiratne J, Bard F. Regulation of O-glycosylation through Golgi-to-ER relocation of initiation enzymes. J Cell Biol. 2010;189:843–58.

    Article  CAS  Google Scholar 

  55. Van den Steen P, Rudd PM, Dwek RA, Opdenakker G. Concepts and principles of O-linked glycosylation. Crit Rev Biochem Mol Biol. 1998;33:151–208.

    Article  Google Scholar 

  56. Bennett EP, Mandel U, Clausen H, Gerken TA, Fritz TA, Tabak LA. Control of mucin-type O-glycosylation: a classification of the polypeptide GalNAc-transferase gene family. Glycobiology. 2012;22:736–56.

    Article  CAS  Google Scholar 

  57. Ahmad A, Pereira EO, Conley AJ, Richman AS, Menassa R. Green biofactories: recombinant protein production in plants. Recent Pat Biotechnol. 2010;4:242–59.

    Article  CAS  Google Scholar 

  58. Alvarez ML, Pinyerd HL, Topal E, Cardineau GA. P19-dependent and P19-independent reversion of F1-V gene silencing in tomato. Plant Mol Biol. 2008;68:61–79.

    Article  CAS  Google Scholar 

  59. Hong S-Y, Kwon T-H, Lee J-H, Jang Y-S, Yang M-S. Production of biologically active hG-CSF by transgenic plant cell suspension culture. Enzyme Microb Technol. 2002;30:763–7.

    Article  CAS  Google Scholar 

  60. Tabar MS, Solouki M, Tohidfar M, Sadeghizadeh M. Expression of human granulocyte-colony stimulating factor (hG-CSF) gene in tobacco (Nicotiana tabacum). Aust J Crop Sci. 2012;6:135–40.

    CAS  Google Scholar 

  61. Nair NR, Chidambareswaren M, Manjula S. Enhanced heterologous expression of biologically active human granulocyte colony stimulating factor in transgenic tobacco BY-2 cells by localization to endoplasmic reticulum. Mol Biotechnol. 2014;56:849–62.

    Article  CAS  Google Scholar 

  62. Hong S-Y, Kwon T-H, Jang Y-S, Kim S-H, Yang M-S. Production of bioactive human granulocyte-colony stimulating factor in transgenic rice cell suspension cultures. Protein Expr Purif. 2006;47:68–73.

    Article  CAS  Google Scholar 

  63. Zvereva AS, Petrovskaya LE, Rodina AV, Frolova OY, Ivanov PA, Shingarova LN, et al. Production of biologically active human myelocytokines in plants. Biochem Biokhimii͡a. 2009;74:1187–94.

    Article  CAS  Google Scholar 

  64. Perez-Perez J, Martinez-Caja C, Barbero J, Gutierez J. DnaK/DnaJ supplementation improves the periplasmic production of human granulocyte-colony-stimulating factor in Escherichia coli. Biochem Biophys Res Commun. 1995;210(2):524–52.

    Article  CAS  Google Scholar 

  65. Oh-eda M, Hasegawa M, Hattori K, Kuboniwa H, Kojima T, Orita T, et al. O-linked sugar chain of human granulocyte-colony stimulating factor protects it against polymerization and denaturation allowing it to retain its biological activity. J Biol Chem. 1990;265:11432–5.

    CAS  PubMed  Google Scholar 

  66. Bae CS, Yang DS, Chang KR, Seong BL, Lee J. Enhanced secretion of human granulocyte colony stimulating factor directed by a novel hybrid fusion peptide from recombinant Saccharomyces cerevisiae at high cell concentration. Biotechnol Bioeng. 1998;57:600–9.

    Article  CAS  Google Scholar 

  67. Kraševec N, Milunović T, Lasnik MA, Lukančič I, Komel R, Porekar VG. Human granulocyte colony stimulating factor (G-CSF) produced in the filamentous fungus aspergillus niger. Acta Chim Slov. 2014;61:709–17.

    PubMed  Google Scholar 

  68. Raso SW, Abel J, Barnes JM, Maloney KM, Pipes G, Treuheit MJ, et al. Aggregation of granulocyte-colony stimulating factor in vitro involves a conformationally altered monomeric state. Protein Sci. 2005;14:2246–57.

    Article  CAS  Google Scholar 

  69. Böer E, Steinborn G, Kunze G, Gellissen G. Yeast expression platforms. Appl Microbiol Biotechnol. 2007;77:513–23.

    Article  Google Scholar 

  70. Gong B, Burnina I, Lynaugh H, Li H. O-linked glycosylation analysis of recombinant human granulocyte colony-stimulating factor produced in glycoengineered Pichia pastoris by liquid chromatography and mass spectrometry. J Chromatogr B Anal Technol Biomed Life Sci. 2014;945–946:135–40.

    Article  Google Scholar 

  71. Dicker M, Maresch D, Strasser R. Glyco-engineering for the production of recombinant IgA1 with distinct mucin-type O-glycans in plants. Bioengineered. 2016;5979:484–9.

    Article  Google Scholar 

  72. Castilho A, Neumann L, Daskalova S, Mason HS, Steinkellner H, Altmann F, et al. Engineering of sialylated mucin-type O-glycosylation in plants. J Biol Chem. 2012;287:36518–26.

    Article  CAS  Google Scholar 

Download references

Authors’ contributions

IAR-A, RM and GAC conceived, designed the experiments. IAR-A performed all the experiments. JBR designed, performed and interpreted MS/MS analysis. IAR-A, JBR, SGL, RM and GAC helped to write, improve and review the latest version of this manuscript. All authors read and approved the final manuscript.

Acknowledgements

Authors would like to thank Centro de Biotecnologia-FEMSA and Centro de Agrobiotecnologia at Tecnologico de Monterrey, Mexico, in particular the Research Group of Bioprocessing and Synthetic Biology (Dr. Marco Antonio Rito-Palomares); and the Center for Protein Identification and Development (CIDPRO) (Dr. Sergio Román Othón Serna-Saldívar) for infrastructure and economic support. Special thanks are also given to CONACyT for Grant (265404/227702-351384) to Israel A. Ramírez-Alanis. We also would like to especially thank Hong Zhu and Angelo Kaldis for technical assistance, Reza Saberianfar for providing confocal microscopy expert assistance, and the AAFC foreign research participant program. We also thank Sasha Daskalova for providing the glycosylation related vectors. We thank Dr. Tsuyoshi Nakagawa (Shimane University) for providing Gateway binary vectors that contain the bar gene, which was identified by Meiji Seika Kaisha, Ltd.

Competing interests

The authors declare that they have no competing interests.

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

Consent for publication

Not applicable.

Ethics approval and consent to participate

Not applicable.

Funding

Not applicable.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Author information

Authors and Affiliations

  1. School of Engineering and Sciences, Tecnologico de Monterrey, Campus Monterrey, Av. Eugenio Garza Sada 2501 Sur, C.P. 64849, Monterrey, NL, Mexico

    Israel A. Ramírez-Alanis, Silverio García-Lara & Guy A. Cardineau

  2. Arizona State University, Phoenix, AZ, 85004-4467, USA

    Guy A. Cardineau

  3. Agriculture and Agri-Food Canada, London, ON, Canada

    Justin B. Renaud & Rima Menassa

  4. Department of Biology, University of Western Ontario, London, ON, Canada

    Rima Menassa

Authors
  1. Israel A. Ramírez-Alanis
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Justin B. Renaud
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Silverio García-Lara
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Rima Menassa
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Guy A. Cardineau
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Guy A. Cardineau.

Additional files

Additional file 1: Figure S3.

Impact of extraction buffer on G-CSF. A Western Blot detection of Sec-G-CSF:eYFP and Cyt-G-CSF:eYFP extracted with PBST0.1 % (gels on the left) or under reducing and denaturing extraction conditions (SDS-DTT) (gels on the right). B Band quantification of Western blot detected proteins. Samples collected at 4, 6 and 8 dpi. Four leaf discs from different leaves were collected from each biological sample. 20 μL TSP of PBST0.1 % treated sample or equivalent volume of SDS-DTT treated sample were loaded on the gel. Black arrows denote monomeric G-CSF:eYFP variant. p19: negative control. eGEHK: protein standard. Proteins were detected with GFP antibody. Band quantification of Western blot detected proteins (Graph). Columns denoted with a different letter are significantly different (p ≤ 0.05) using one-way ANOVA and followed by Tukey test. Error bars are standard deviation of the means

Additional file 2: Figure S2.

Secretory and Cytoplasmic G-CSF can be extracted with a PBS-based extraction buffer, but Zera-G-CSF can only be extracted under denaturing and reducing conditions. Western Blot detection of G-CSF:eYFP variants transiently expressed in N. benthamiana plants. Samples collected at 4 dpi. Upper panel: Plant 1; Lower panel: Plant 2. Sec-G-CSF:eYFP (Secretory variant, 49 kDa), Cyt-G-CSF:eYFP (Cytoplasmic and mature protein variant, 49 kDa), Zera-G-CSF:eYFP (Zera fused variant, 61 kDa), eYFP (Cytoplasmic eYFP, 29 kDa), p19 negative control. Four leaf discs from two different plants were collected for each sample. 50 µg TSP of PBST samples or equivalent volume of SDS-DTT samples were loaded on the gel. Black arrows denote Zera-G-CSF:eYFP. Gray arrows denote Sec-GCSF-:eYFP and Cyt-G-CSF:eYFP. Red asterisks denotes a faint band corresponding to Zera-G-CSF:eYFP extracted with PBST extraction buffer and its corresponding proper extraction with SDS-DTT extraction buffer. Proteins were detected with anti GFP

Additional file 3: Figure S1.

Compartmentalization of GalNAc attachment to target proteins in the secretory pathway. GalNAc-Transferases are traditionally referred to be localized in the Golgi Apparatus (Cis-Golgi), but recent studies suggest also ER localization, thus proposing attachment of GalNAc in subregions of ER and proximal Golgi compartment. GalNAc-T, GalNAc transferase. ER, Endoplasmic Reticulum. ERGIC, intermediate ER-Golgi Compartment. Grey box, GalNAc

Additional file 4: Figure S4.

MS/MS identification of glycosylated Sec-G-CSF:cMyc-derived peptide. A Schematic illustration of released peptide after Trypsin/Chymotrypsin in gel digestion of c-Myc purified Sec-G-CSF:cMyc expressed alone or co-expressed with the O-glycosylation machinery. Native glycosylation site (Thr-163) is denoted by underlining. B Extracted ion chromatograms of predicted QQMEELGMAPALQPTQGAMPAFASAF peptide derived from Sec-G-CSF:cMyc expressed alone (upper panel), not being detected; or co-expressed with the O-glycosylation machinery (lower panel), where it was detected. C MS/MS of c-Myc purified and Trypsin/Chymotrypsin in gel digested Sec-G-CSF:cMyc-derived peptide (QQMEELGMAPALQPTQGAMPAFASAF) co-expressed with the O-glycosylation machinery, showing identified b- and y-ions. Modified Thr-163 is marked in red. Complete list of detected product ions is shown in Supplementary Table S1

Additional file 5: Table S1.

y- and b- product ions detected in GalNAc-O-glycosylated QQMEELGMAPALQPTQGAMPAFASAF derived peptide and associated mass errors

Rights and permissions

Open Access This 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.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ramírez-Alanis, I.A., Renaud, J.B., García-Lara, S. et al. Transient co-expression with three O-glycosylation enzymes allows production of GalNAc-O-glycosylated Granulocyte-Colony Stimulating Factor in N. benthamiana. Plant Methods 14, 98 (2018). https://doi.org/10.1186/s13007-018-0363-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13007-018-0363-y

Keywords

Plant Methods

ISSN: 1746-4811