Cost-effective production of tag-less recombinant protein in Nicotiana benthamiana

Abstract

Plants have recently received a great deal of attention as a means of producing recombinant proteins. Despite this, a limited number of recombinant proteins are currently on the market and, if plants are to be more widely used, a cost-effective and efficient purification method is urgently needed. Although affinity tags are convenient tools for protein purification, the presence of a tag on the recombinant protein is undesirable for many applications. A cost-effective method of purification using an affinity tag and the removal of the tag after purification has been developed. The family 3 cellulose-binding domain (CBM3), which binds to microcrystalline cellulose, served as the affinity tag and the small ubiquitin-related modifier (SUMO) and SUMO-specific protease were used to remove it. This method, together with size-exclusion chromatography, enabled purification of human interleukin-6 (hIL6) with a yield of 18.49 mg/kg fresh weight from leaf extracts of Nicotiana benthamiana following Agrobacterium-mediated transient expression. Plant-produced hIL6 (P-hIL6) contained less than 0.2 EU/μg (0.02 ng/mL) endotoxin. P-hIL6 activated the Janus kinase-signal transducer and activator of transcriptional pathways in human LNCaP cells, and induced expression of IL-21 in activated mouse CD4⁺ T cells. This approach is thus a powerful method for producing recombinant proteins in plants.

Keywords: Nicotiana benthamiana; bdSENP1; bdSUMO; cellulose-binding domain; human interleukin-6; plant-based expression system; proteolytic tag removal.

Figure 1

Agrobacterium‐mediated transient expression enables high levels of MCS‐hIL6 production in Nicotiana benthamiana. (a) Design of a chimeric hybrid construct for high level protein expression, purification, and affinity tag removal. The plant expression binary vector p1300‐C3bdSU‐hIL6 contains a chimeric hybrid construct consisting of various domains, including a plant codon‐optimized M domain, a cellulose‐binding domain (CBM3), a small ubiquitin‐related modifier (bdSUMO) derived from Brachypodium distachyon, and human interleukin‐6 (hIL6) flexible linkers (L). In addition, the BiP leader sequence and an ER retention signal HDEL were fused to the 5’ and 3’ ends of the chimeric construct, respectively. The chimeric construct was placed under the double enhancer‐containing CaMV 35S promoter (d35S) and a strong translational enhancer sequence (5′ UTR) followed by the HSP terminator (HSP‐t) from Arabidopsis thaliana. The N‐linked glycosylation site (Asp‐44) of hIL6 is indicated in gray. (b) Western blot analysis of MCS‐hIL6. Nicotiana benthamiana leaf tissues were harvested at the indicated time points after infiltration of p1300‐C3bdSU‐hIL6 with (+) or without (−) the p38 silencing suppressor. Total leaf extracts (30 μg) were separated using 12.5% SDS‐PAGE and analyzed by western blotting with anti‐IL6 antibody. (c) Coomassie brilliant blue (CBB) staining. A membrane identical to that used for western blot analysis was stained with CBB. The large subunit of the rubisco complex (RbcL) was used as a loading control. M: molecular weight standards; WT: wild‐type N. benthamiana leaf tissue extracts. The arrow indicates the position of MCS‐hIL6 fusion protein bands (60–65 kDa).

Figure 2

CBM3 in MCS‐hIL6 irreversibly binds to microcrystalline cellulose (MCC) beads. (a, b) Coomassie brilliant blue (CBB) staining and western blot analysis. Total plant leaf extracts (TP) were incubated with MCC beads at 4 °C for 1 h. Proteins in the supernatants were collected as the unbound (UB) fraction. MCC beads were washed four times and each wash‐off fraction (W1 to W4) was collected separately. After the fourth wash, the MCC beads were treated with 80% glycerol and the supernatant was collected (E fraction). Finally, the MCC beads were boiled in SDS loading buffer and the supernatant was collected (B fraction). These fractions were analyzed using 12.5% SDS‐PAGE followed by CBB staining (a) or western blot analysis with anti‐IL6 antibody (b). M: pre‐stained molecular weight standard; WT: wild‐type N. benthamiana leaf tissue extracts. Arrows indicate the position of MCS‐hIL6 (60–65 kDa).

Figure 3

MCS‐hIL6 is cleaved by His:bdSENP1 both in crude leaf extracts and when immobilized on microcrystalline cellulose (MCC). (a) bdSENP1‐mediated cleavage of bdSUMO in MCS‐hIL6. Total leaf extracts were treated with (+) or without (−) His:bdSENP1 and analyzed by western blotting with anti‐IL6 antibody. The large subunit of the rubisco complex (RbcL) stained with CBB was used as a loading control. (b, c) bdSENP1‐mediated cleavage of bdSUMO in MCS‐hIL6 immobilized on MCC beads. Total protein extracts were incubated with MCC beads. After binding, the MCC beads were washed twice and treated with (+) or without (−) His:bdSENP1. Proteins in the supernatant and MCC bead fractions were collected separately and analyzed by western blotting with anti‐IL6 antibody (b) or anti‐CBM3 antibody (c). M: molecular weight standards; WT: wild‐type total leaf extracts; TP: total leaf extracts; UB: unbound fraction; W1 and W2: first and second wash‐off fractions, respectively; S: supernatant after His:bdSENP1 treatment; NS: supernatant without His:bdSENP1 treatment; B: proteins released from MCC beads by boiling.

Figure 4

Production of hIL6 from N. benthamiana leaf tissue via MCC‐based affinity purification, SUMO‐based removal of the affinity tag and size‐exclusion chromatography. (a) Schematic illustrating purification of CBM3‐bdSUMO‐tagged protein and removal of the affinity tag. (b, c) Production of hIL6. Fresh leaf tissues (40 g) were homogenized in extraction buffer and total protein extracts were prepared. The total protein extracts were incubated with MCC beads at a 10: 1 ratio. After incubation at 4 °C, the supernatant was collected (unbound fraction, UB) and the MCC beads were washed four times in washing buffer (W). The MCC beads were treated with 10 μg His:bdSENP1 at 4 °C for 6 h. The supernatant was collected (On‐bound cleavage) and His:bdSENP1 was removed by passing the supernatant through a Ni²⁺‐NTA affinity column (His:bdSENP1 removed). The flow‐through fraction containing hIL6 was collected (cellulose purified fraction) and applied to the size‐exclusion column. Fractions were collected from the size‐exclusion column. The fractions obtained at each step were analyzed using SDS‐PAGE followed by CBB staining (b) or western blot analysis with anti‐IL6 antibody (c). M: pre‐stained molecular weight standards; WT: wild‐type total leaf extracts; T: total leaf extracts of MCS‐hIL6; UB: unbound fraction; W: wash‐off solution.

Figure 5

Plant‐produced hIL6 has a high mannose‐type N‐glycosylation. Purified hIL6 was treated with (+) or without (−) endo‐H and analyzed by western blotting with anti‐IL6 antibody.

Figure 6

Plant‐produced hIL6 is active in the STAT3 signalling pathway in animal cells. (a, b) Phosphorylation of STAT‐3 (p‐STAT3) and quantification of p‐STAT3 levels. Total protein extracts from LNCaP cells, treated with plant‐produced hIL6 (P‐hIL6) or commercial E. coli‐produced hIL6 (E‐hIL6) for 30 min (a) or for the indicated period of time (b), were analyzed by western blotting with anti‐p‐STAT3 (phosphorylated STAT3), anti‐STAT3, or anti‐β‐actin (ACTB) antibodies. β‐actin was used as a loading control. The intensity of p‐STAT3 bands was measured using ImageJ software to quantify the amount of p‐STAT3. Three independent experiments were performed. Error bars, SEM (n = 3). Statistical analysis was performed using the Student's t‐test between P‐hIL6 and E‐hIL6 at the indicated time points; ns: no significant difference. (c) IL6‐induced IL‐21 expression. FACS‐sorted CD25⁻ total CD4⁺ T cells from mouse splenocytes were cultured in the presence of different quantities of P‐hIL6 or E‐hIL6. IL‐21 production was measured by ELISA after 3 days of treatment. The data are the means ± SEM (n = 3). Data were analyzed using the Student's t‐test; *: P < 0.04; ns: no significant difference.