Functional Improvement of Human Cardiotrophin 1 Produced in Tobacco Chloroplasts by Co-expression with Plastid Thioredoxin m

Abstract

Human cardiotrophin 1 (CT1), a cytokine with excellent therapeutic potential, was previously expressed in tobacco chloroplasts. However, the growth conditions required to reach the highest expression levels resulted in an impairment of its bioactivity. In the present study, we have examined new strategies to modulate the expression of this recombinant protein in chloroplasts so as to enhance its production and bioactivity. In particular, we assessed the effect of both the fusion and co-expression of Trx m with CT1 on the production of a functional CT1 by using plastid transformation. Our data revealed that the Trx m fusion strategy was useful to increase the expression levels of CT1 inside the chloroplasts, although CT1 bioactivity was significantly impaired, and this was likely due to steric hindrance between both proteins. By contrast, the expression of functional CT1 was increased when co-expressed with Trx m, because we demonstrated that recombinant CT1 was functionally active during an in vitro signaling assay. While Trx m/CT1 co-expression did not increase the amount of CT1 in young leaves, our results revealed an increase in CT1 protein stability as the leaves aged in this genotype, which also improved the recombinant protein's overall production. This strategy might be useful to produce other functional biopharmaceuticals in chloroplasts.

Keywords: cardiotrophin-1, thioredoxin, plastid transformation, bioactivity, tobacco.

Figure 1

Integration of Trxm and ct1 genes by fusion or co-expression into the plastid genome and homoplasmy verification. (a) Map of the WT, Trxm-CT1 fusion and Trxm/CT1 co-expression plastid DNA. Transgenes were cloned into the intergenic region between the rrn16/trnV and 3′rps12 genes. The Trxm-CT1 fusion is driven by the psbA promoter and the rps16 terminator. In co-expression, CT1 is also driven by PpsbA but Trx m is driven by the rrnG10L promoter. Arrows within boxes show the direction of transcription. The selectable gene aadA is driven by the rrn promoter and psbA terminator. Probes for the Southern blot are shown over the corresponding sequences. The sizes of the predicted bands when DNA was digested with BglII are indicated. (b–d) Southern blot analysis for WT and two independent plants per construct is shown. The same blot was probed with FS (b), CT1 (c) and Trx m (d) probes. (e) Seed assays confirming homoplasmy of regenerated plants. WT seedlings bleach out on the spectinomycin-containing medium whereas seedlings from both transplastomic genotypes are resistant (green color). rrn16, trnV, 3′rps 12: original sequences of the chloroplast genome; aadA: aminoglycoside 3′-adenylyltransferase; Prrn: 16SrRNA promoter; PpsbA: psbA promoter and 5′-untranslated region; TpsbA: psbA terminator; PrrnG10L: rrn promoter and the gene 10 leader from phage T7; Trps16: rps16 terminator; Trxm: thioredoxin m; ct1: cardiotrophin-1; Trxm-CT1: fusion plants; Trxm/CT1: co-expression plants.

Figure 2

CT1 accumulation in fusion and co-expression tobacco plants. (a) Western blot analysis of CT1 and Trx m. Ten micrograms of total protein (TP) extracted from WT and transplastomic genotypes were loaded onto a 13% SDS-PAGE gel. Commercial human CT1 (20 ng) was used as standard. (b) Quantification of recombinant CT1 in transplastomic genotypes by densitometry of Western blots. A dilution series of commercial hCT1 was included. Protein loading, whose amounts (µg of total protein) are indicated over the blots, was adjusted to become comparable to the hCT1 standard. Biological replicate extracts from three different plants of each genotype are shown. (c) Representation of CT1 quantification in young tobacco leaves. Results are the mean ± SE of two measurements for three independent transgenic plants per construct and are shown as a percentage of TP. Different letters indicate significant differences (t-test, p ≤ 0.01). Free-CT1: free-CT1 expressing plants; Trxm-CT1: fusion plants; Trxm/CT1: co-expression plants; hCT1: commercial human cardiotrophin 1.

Figure 3

Levels of ct1 and Trxm transcripts in transgenic tobacco plants. (a) A Northern blot analysis was performed with RNA extracted from tobacco leaves of WT and transplastomic genotypes. Three plants for each construct were analyzed. Twenty micrograms of total RNA were electrophoresed, blotted and hybridized with CT1 (top panels) and Trx m (middle panels) specific probes. Ethidium bromide stained rRNA was used to assess loading (lower panels). (b) Expected transcription patterns for the different constructs integrated into the chloroplast genome. Horizontal discontinuous arrows above genes show monocistronic (x and y) and dicistronic (z) transcripts, and their expected sizes are indicated. Arrows within boxes show the direction of transcription. Free-CT1: free-CT1 expressing plants; Trxm-CT1: fusion plants; Trxm/CT1: co-expression plants.

Figure 4

CT1 protein accumulation in tobacco chloroplasts as a function of leaf age. (a) Comparison of CT1 accumulation in a developmental series of five alternating leaf samples from the top to the bottom of the plant. Representative Western blots are shown. For each genotype, 2 µg of total protein were loaded. (b) CT1 densitometric quantification of tobacco leaves in different developmental stages. Leaves are numbered from the top (youngest) to the bottom (oldest) of the plant: 1, 3, 5, 7, and 9. Results are the mean ± SE of two measurements for three independent transgenic plants per construct. For each leaf, different letters indicate significant differences (t-test, p ≤ 0.05). Free-CT1: free-CT1 expressing plants; Trxm-CT1: fusion plants; Trxm/CT1: co-expression plants.

Figure 5

Bioactivity of recombinant human CT1 produced in tobacco chloroplasts. (a) Representative Western blot for P-STAT-3 (Tyr705) and actin (loading control) in the HepG2 cell line. Cells were stimulated with 25 ng per well of commercial hCT1 as positive control (C+). Cells treated with WT crude extract and non-stimulated cells (C-) were used as negative controls. For all the transplastomic genotypes, cells were treated with plant extract equivalent to 25 ng per well of soluble recombinant CT1. Two independent plants per transplastomic genotype are shown. (b) Densitometric analysis of P-STAT-3 in stimulated HepG2 cell line extracts. The results are presented as the fold increase relative to the free-CT1 genotype (= 1) using the ratios between the densities of the P-STAT-3 bands and the corresponding actin bands. The data are presented as the mean ± SE of three individual experiments (6 plants/genotype). Different letters indicate significant differences (t-test, p ≤ 0.05). P-STAT-3: phosphorylated-signal transducer and activator of transcription-3; free-CT1: free-CT1 expressing plants; Trxm-CT1: fusion plants; Trxm/CT1: co-expression plants.

Figure 6

Disulfide oxidoreductase activity of Trx m protein from Trxm-CT1 and Trxm/CT1 genotypes. The dithiothreitol (DTT)-dependent insulin reduction assay of Trx m fused and co-expressed with CT1 was determined in a reaction mixture containing 4 μM of purified Trxm-CT1 fusion protein or 4 μM of purified Trx m/CT1 protein, supplemented with 0.5 mM DTT. Assays in the absence of Trx showing no activity were used as negative controls. Trx from E. coli was used at 1 and 3 μM as positive control.