Efficient expression of nuclear transgenes in the green alga Chlamydomonas: synthesis of an HIV antigen and development of a new selectable marker

Abstract

The unicellular green alga Chlamydomonas reinhardtii has become an invaluable model system in plant biology. There is also considerable interest in developing this microalga into an efficient production platform for biofuels, pharmaceuticals, green chemicals and industrial enzymes. However, the production of foreign proteins in the nucleocytosolic compartment of Chlamydomonas is greatly hampered by the inefficiency of transgene expression from the nuclear genome. We have recently addressed this limitation by isolating mutant algal strains that permit high-level transgene expression and by determining the contributions of GC content and codon usage to gene expression efficiency. Here we have applied these new tools and explored the potential of Chlamydomonas to produce a recombinant biopharmaceutical, the HIV antigen P24. We show that a codon-optimized P24 gene variant introduced into our algal expression strains give rise to recombinant protein accumulation levels of up to 0.25% of the total cellular protein. Moreover, in combination with an expression strain, a resynthesized nptII gene becomes a highly efficient selectable marker gene that facilitates the selection of transgenic algal clones at high frequency. By establishing simple principles of successful transgene expression, our data open up new possibilities for biotechnological research in Chlamydomonas.

Keywords: Antigen; Chlamydomonas reinhardtii; HIV; Molecular farming; Selectable marker gene; Transformation.

Fig. 1

HIV-1 P24 gene variants and physical map of the expression vector used for algal nuclear transformation. a Relative codon adaptation (RCA) of the different P24 gene variants compared to the nuclear genome of Chlamydomonas reinhardtii. Blue bars indicate the relative adaptation (in %) of each codon in the reading frames of the three synthetic gene variants. The x-axis indicates the codon numbers within the gene (cf. Table 1). Variant CrP24 contains the most frequently used synonymous codon for all amino acids. b GC content (in %) and its distribution over the reading frames of the three P24 gene variants. The values were determined in a sliding window of 40 bp. CrP24 is indicated in green, P24w in blue and the chloroplast-optimized variant CpP24 in red (see color code in panel a). The least adapted variant (CpP24) has the lowest GC content and the fully codon-optimized gene version (CrP24) has the highest GC content (cf. Table 1). c Physical map of the transformation vector used for expression of the three P24 gene variants in Chlamydomonas reinhardtii. All variants were cloned into the same vector and are driven by identical expression elements. The arrows indicate the binding sites of primer pairs used for PCR analysis of transformed algal strains. The different P24 coding regions were inserted into an expression cassette derived from the Chlamydomonas reinhardtii PSAD locus (Fischer and Rochaix 2001) using the restriction sites NdeI and EcoRI (P_PSAD: PSAD promoter; T_PSAD: PSAD terminator). The paromomycin resistance gene aphVIII serves as selectable marker and is driven by the fused promoters from the HSP70A gene (P_HSP70) and the RBCS2 gene (P_RBCS2) of C. reinhardtii

Fig. 2

Analysis of CrP24 mRNA accumulation in transformants of strains UVM11 and Elow47. Ten independent transformants harboring the complete CrP24 cassette (based on PCR assays; see “Materials and methods” section) were selected randomly for each strain. 10 µg of total RNA was used for northern blot analysis. Transformant number five of strain UVM11 was selected as a standard for relative quantitation, and a dilution series of total RNA (0.5, 1, 2.5, 5 µg) of this line was loaded in all blots as a positive control (PC). The untransformed strain was used as negative control (NC). Asterisks indicate the expected transcript size of 1.1 kb. The ethidium bromide-stained gel prior to blotting is shown below each blot and serves as loading control. The whole reading frame of CrP24 was used as hybridization probe. Marker band sizes are given in kb at the left. Additional transcripts of larger size may originate from multicopy insertions in tandem and/or from insertion into endogenous genes in the genome. a Northern blot analysis of UVM11 transformants. Transformed clone number 1 shows a slightly shorter CrP24 transcript, presumably because of a small deletion or truncation. b Northern blot analysis of Elow47 transformants

Fig. 3

Immunoblot analysis of P24 protein accumulation in strains UVM11 and Elow47 transformed with the P24 gene variant that was codon optimized for the Chlamydomonas nuclear genome. The same ten CrP24 transformants from each strain that had been analyzed by northern blots (Fig. 2) were tested for P24 accumulation. 40 µg of total protein from each transformant were separated by SDS-PAGE. The untransformed strain was used as negative control (NC). A dilution series of recombinant His-tagged P24 protein (rP24) was loaded for semiquantitative analysis of P24 accumulation levels. The small size difference between the recombinant protein and the protein expressed in algal cells is due to the His-tag. The upper part of the gel was stained with Coomassie and served as loading control (shown below each blot). a Immunoblot analysis of UVM11 transformants. Note that transformed clone number 1 that accumulates a truncated CrP24 transcript (Fig. 2a), is the only strain that does not accumulate the P24 protein. The maximum P24 accumulation level is approximately 0.25 % of total cellular protein (strain 6). b Immunoblot analysis of Elow47 transformants. Clone number 3 shows a larger-than-expected protein band (of approximately 36 kDa), consistent with accumulation of a larger mRNA (Fig. 2b). The larger protein may originate from in-frame fusion with an endogenous gene

Fig. 4

Southern blot analysis of CrP24 transformants of strains UVM11 and Elow47. a Physical map of the transformation vector integrated into the Chlamydomonas nuclear genome. The EcoRI and EcoRV restriction sites used for RFLP analysis are indicated. The location of the EcoRV site in the flanking chromosomal DNA is hypothetical (and variable depending on the integration site of the transgenes). The hybridization probe (‘Probe’) and the restriction fragment it detects (‘DNA fragment’) are also indicated. aphVIII: paromomycin resistance gene (selectable marker); P_HSP70: promoter from the HSP70A gene; P_RBCS2: promoter from the RBCS2 gene; P_PSAD: promoter from the PSAD gene; T_PSAD: terminator from the PSAD gene. b Southern blot analysis of ten randomly picked CrP24 transformants of expression strain UVM11 (left panel) and ten randomly picked transformants of control strain Elow47 (right panel). Samples of 10 µg total DNA were digested with the restriction enzymes EcoRI and EcoRV and separated by agarose gel electrophoresis. DNA samples extracted from untransformed strains were used as negative control (NC) and digested with the same enzymes. The hybridization probe was generated by labeling a DNA fragment covering the entire coding region of CrP24. Fragment sizes of the molecular weight marker (M) are given at the left in kb. Note that the majority of the transformants harbors a single copy of the CrP24 transgene

Fig. 5

Codon usage, GC content and physical map of the expression vector used for transformation of Chlamydomonas with nptII gene variants and with aphVIII. a Relative codon adaptation (RCA) of the two nptII gene variants compared to the nuclear genome of Chlamydomonas. Blue bars indicate the relative adaptation (in %) of each codon in the reading frames of the two gene variants. The x-axis indicates the codon numbers within the gene. CrnptII contains the most frequently used synonymous codon for all amino acids. b GC content (in  %) and its distribution over the reading frames of the two nptII variants. The values were determined in a sliding window of 40 bp. CrnptII is indicated in green and the original (bacterial) EcnptII in blue. The fully codon-optimized CrnptII has a higher GC content. c Physical map of the transformation vectors used for expression of the paromomycin resistance gene aphVIII (left panel) and the two nptII gene variants (right panel) in Chlamydomonas. The aphVIII gene is driven by the fused promoters from the HSP70A gene (P_HSP70) and the RBCS2 gene (P_RBCS2) of C. reinhardtii. The nptII coding regions were inserted into an expression cassette derived from the Chlamydomonas PSAD locus (Fischer and Rochaix 2001) using the restriction sites NdeI and EcoRI (P_PSAD: PSAD promoter; T_PSAD: PSAD terminator)

Fig. 6

Comparison of CrnptII and EcnptII transcript accumulation in Chlamydomonas strains UVM11 and Elow47 by northern blot analysis (a–d), and protein accumulation levels conferred by the two nptII variants as determined by immunoblot analysis (e–h). All analyzed strains were selected on kanamycin (50 µg mL⁻¹) and, therefore, are expected to express the nptII marker gene. For the northern blots (panels a–d), samples of 10 µg total RNA were electrophoretically separated in denaturing 1.2 % agarose gels. The gel blots were hybridized to a radiolabeled probe comprising a 1:1 mixture of the complete coding regions of both nptII variants. RNA samples isolated from the corresponding untransformed strains were used as negative control (NC). A dilution series (0.5, 1.0, 2.5 and 5.0 µg RNA) of an NptII-expressing algal clone (CrnptII transformant number 3 in strain UVM11) was loaded onto each gel as a positive control (PC) to facilitate comparison between blots. The band representing the 25S rRNA of the cytosolic 80S ribosome in the ethidium bromide-stained gel prior to blotting is shown below each blot as a loading control. Sizes of RNA marker bands are indicated in the right of each blot (in kb). The size of the nptII transcript is approximately 1.2 kb. For the immunoblots (panels e–h), samples of 40 µg total cellular protein were separated by SDS-PAGE and the same transformed strains were analyzed as in panel (a-d). A dilution series of recombinant NptII (rNptII) was loaded to facilitate semiquantitative analysis and comparison between blots. The size of the untagged protein expressed in Chlamydomonas is ~29 kDa, the slightly larger size of the rNptII is due to its His-tag. Protein samples of the untransformed strains were loaded as negative controls (NC). As a control for equal loading, the Coomassie-stained upper part of the gel is shown below each blot. a nptII mRNA accumulation in ten independent transgenic clones of expression strain UVM11 transformed with gene variant CrnptII. b nptII mRNA accumulation in ten independent clones of strain Elow47 transformed with gene variant CrnptII. c nptII mRNA accumulation in ten independent transgenic clones of expression strain UVM11 transformed with the EcnptII gene variant. d nptII mRNA accumulation in ten independent transgenic clones of strain Elow47 transformed with the EcnptII gene variant. e NptII protein accumulation in ten independent transgenic clones of expression strain UVM11 transformed with gene variant CrnptII. f NptII protein accumulation in ten independent clones of strain Elow47 transformed with gene variant CrnptII. g NptII protein accumulation in ten independent transgenic clones of expression strain UVM11 transformed with the EcnptII gene variant. Note that strains 4 and 5 show above-background expression of NptII, whereas in all other transformed clones, the signal is not stronger than that of the cross-reacting band of similar size in the NC. h NptII protein accumulation in ten independent transgenic clones of strain Elow47 transformed with the EcnptII gene variant

Fig. 7

Transformation efficiencies obtained with the two nptII gene variants in Chlamydomonas strains UVM11 and Elow47. a Analysis of transformation efficiencies obtained with CrnptII in expression strain UVM11 and the corresponding wild type-like strain Elow47 at different concentrations of kanamycin. The strains were transformed with the CrnptII cassette and selection was performed on media containing different concentrations of kanamycin (25, 50, 100 and 200 µg mL⁻¹). The number of kanamycin-resistant colonies was averaged from three independent transformation experiments. Error bars indicate the standard deviation. b Comparison of transformation efficiencies obtained with the two nptII variants in strains UVM11 and Elow47. Algal cells were transformed with either the CrnptII or the EcnptII cassette and selected on medium containing 50 µg mL⁻¹ kanamycin. The number of resistant colonies was averaged from three independent transformation experiments (scored 6 days after transformation). Error bars indicate the standard deviation

Fig. 8

Kanamycin resistance assays with CrnptII and EcnptII transformants of strains UVM11 and Elow47. The ten transformants per strain and gene variant that had been analyzed with respect to mRNA and protein accumulation levels were assayed for their antibiotic resistance by drop tests with three different cell concentrations (7 µl of cell suspensions with 10⁷, 10⁶ and 10⁵ cells mL⁻¹) on agar plates containing different concentrations of kanamycin (0, 25, 50, 75, 100 and 200 µg mL⁻¹). Untransformed UVM11 and Elow47 were used as negative controls. Note that all transformed clones shown here were initially selected on kanamycin (50 µg mL⁻¹) and, therefore, display some kanamycin resistance. However, on average, CrnptII transformants are more resistant to the antibiotic than EcnptII transformants, and UVM11 transformants tolerate higher kanamycin concentrations than Elow47 transformants

Fig. 9

Comparison of the detoxification efficiency of aphVIII and CrnptII for different antibiotics. Nine randomly selected transformants of UVM11 with either aphVIII or CrnptII (both initially selected on 25 µg mL⁻¹ kanamycin; Sizova et al. 2001) were assayed. The untransformed UVM11 strain was used as negative control (NC). Drop tests were performed using three different cell concentrations (7 µl of cell suspensions containing 10⁷, 10⁶ and 10⁵ cells mL⁻¹) on agar plates containing different concentrations of paromomycin (5, 10, 25 and 50 µg mL⁻¹), kanamycin (25, 50, 100 and 200 µg mL⁻¹) or G418 (2.5, 5, 10 and 25 µg mL⁻¹). Photographs were taken after 12 days. Note that, although initially selected on medium containing 25 µg mL⁻¹ kanamycin, some aphVIII transformants do not grow on kanamycin and paromomycin. This could be due to silencing of the (non-codon-optimized) aphVIII transgene during strain maintenance under non-selective conditions