Planta (2009) 229:873–883 DOI 10.1007/s00425-008-0879-x

O R I G I N A L A R T I CL E

Strategies to facilitate transgene expression in Chlamydomonas reinhardtii Alke Eichler-Stahlberg · Wolfram Weisheit · Ovidiu Ruecker · Markus Heitzer

Received: 12 October 2008 / Accepted: 17 December 2008 / Published online: 7 January 2009 © Springer-Verlag 2009

Abstract The unicellular green alga Chlamydomonas reinhardtii has been identiWed as a promising organism for the production of recombinant proteins. While during the last years important improvements have been developed for the production of proteins within the chloroplast, the expression levels of transgenes from the nuclear genome were too low to be of biotechnological importance. In this study, we integrated endogenous intronic sequences into the expression cassette to enhance the expression of transgenes in the nucleus. The insertion of one or more copies of intron sequences from the Chlamydomonas RBCS2 gene resulted in increased expression levels of a Renilla-luciferase gene used as a reporter. Although any of the three RBCS2 introns alone had a positive eVect on expression, their integration in their physiological number and order created an over-proportional stimulating eVect observed in all transformants. The secretion of the luciferase protein into the medium was achieved by using the export sequence of the Chlamydomonas ARS2 gene in a cell wall deWcient strain and Renilla-luciferase could be successfully concentrated with the help of attached C-terminal protein tags.

Electronic supplementary material The online version of this article (doi:10.1007/s00425-008-0879-x) contains supplementary material, which is available to authorized users. A. Eichler-Stahlberg · W. Weisheit · O. Ruecker · M. Heitzer Center of Excellence for Fluorescent Bioanalysis, University of Regensburg, Josef-Engert-Str. 9, 93053 Regensburg, Germany Present Address: M. Heitzer (&) Geneart AG, Josef-Engert-Str. 11, 93053 Regensburg, Germany e-mail: [email protected]

Similarly, a codon adapted gene variant for human erythropoietin (crEpo) was expressed as a protein of commercial relevance. Extracellular erythropoietin produced in Chlamydomonas showed a molecular mass of 33 kDa probably resulting from post-translational modiWcations. Both, the increased expression levels of transgenes by integration of introns and the isolation of recombinant proteins from the culture medium are important steps towards an extended biotechnological use of this alga. Keywords Chlamydomonas reinhardtii · Erythropoietin · Gene expression · Intron · luciferase · Secretion

Introduction Being a model system for photosynthesis and Xagellar function the eukaryotic unicellular green alga Chlamydomonas reinhardtii has recently also gained interest as an organism of biotechnological relevance (Leon-Banares et al. 2004). Simple salt based media, fast vegetative growth and high cell densities are the basis for a cost eVective cultivation and are—along with well characterized genetics—a major argument for a future industrial use of this alga (Franklin and MayWeld 2004). Beside the extraction of natural components (e.g. carotenoids) and the production of hydrogen using (sun)light as energy source (Melis 2007), the expression of recombinant proteins has been identiWed as one possible biotechnological goal (Fuhrmann 2004). In theory, transgenes can be expressed from any of the three algal genomes (nucleus, chloroplast, mitochondria) (Rochaix 2002), but only protein production within the single large chloroplast of Chlamydomonas has been brought to commercial relevance yet (Franklin and MayWeld 2005).

123

874

The molecular biology of the chloroplast allows a directed manipulation of its genome, since foreign DNA can be integrated by homologous recombination (Heifetz 2000). Moreover, mechanisms for transcriptional or post-transcriptional gene silencing are completely absent, which results in a reliable and constant level of recombinant protein over the time. However, comprising a prokaryotic like gene expression system, enzymes for the post-translational modiWcation of proteins such as formation of disulWde bonds or glycosylation are not found inside the chloroplast. Nevertheless it has been shown, that a large single chain antibody against herpes simplex virus glycoprotein D was expressed and assembled correctly to form fully functional dimers (MayWeld et al. 2003). Since then expression eYciencies were further optimized and the chloroplast of Chlamydomonas was reported to be a considerable alternative for the industrial production of antibodies and other recombinant proteins (MayWeld and Franklin 2005; MayWeld et al. 2007). Since most extracellular proteins need a correct posttranslational processing for proper folding or enzymatic activity, these proteins will not always be produced functionally inside plastids. In this case, transgene expression from the nuclear genome of C. reinhardtii oVers the possibility to direct recombinant proteins to the organelles of the classical secretory pathway containing all enzymes responsible for post-translational modiWcations (Griesbeck et al. 2006). A subsequent secretion into the culture medium additionally allows the puriWcation of the recombinant protein product from the culture supernatant rather than from a whole cell extract, which simpliWes downstream processing. Alternatively, proteins can be targeted and attached to the cell membrane, where antigenic epitopes were shown to induce an immune response when added as an edible vaccine to Wsh food (Sayre et al. 2001). Although important improvements have been developed, e.g. codon adaptation of transgenes and construction of suitable expression vectors (Fuhrmann et al. 1999; Heitzer and Zschoernig 2007), gene expression from the nucleus of Chlamydomonas is still too ineVective for industrial applications and needs to be optimized. It has been reported, that inclusion of endogenous intron sequences within the transgene led to increased transformation rates as a result of elevated amounts of selection marker protein for Chlamydomonas reinhardtii and its close multi-cellular relative Volvox carteri. When testing the gene encoding nitrate reductase (NITA) as a metabolic marker for Volvox, the insertion of NITA intron 1 or introns 9 and 10 within the cDNA resulted in a ten times higher number of transformants compared to cDNA without any introns (Gruber et al. 1996). Similarly, a mutant variant of the Chlamydomonas gene for acetolactate synthase for the selection against sulfometuron methyl yielded only resistant transfor-

123

Planta (2009) 229:873–883

mants if all physiological introns where present within the coding region (Kovar et al. 2002). Since transformation eYciencies of the bacterial derived selection marker genes ble and aph7⬘⬘ could also be increased two to tenfold upon the artiWcial insertion of one copy of intron 1 from the Chlamydomonas RBCS2 gene (Rubisco small subunit 2), endogenous intron sequences seem to exert a generally positive eVect on nuclear gene expression (Lumbreras et al. 1998; Berthold et al. 2002). Lumbreras et al. showed further, that the expression enhancing eVect of RBCS2 intron 1 was dependent on its position within the ble coding region and that it could be increased by a second copy of this intron. In this study, a detailed approach was used to analyse the inXuence of diVerent intron sequences and their position upon the expression of the reporter gene Renilla-luciferase (Fuhrmann et al. 2004) from the nuclear genome of Chlamydomonas reinhardtii. To estimate the suitability of this alga as a system for extracellular protein production, recombinant gene products were targeted to the culture medium and isolated using protein tags. The human gene for erythropoietin (Jacobs et al. 1985)—a small hormone exhibiting two disulWde bonds and speciWc N- and O-glycosylation at four positions in its mature form (Cheetham et al. 1998)—was chosen as an example for a protein of pharmaceutical interest.

Materials and methods Construction of plasmids With plasmid pRbcRL(Hsp196) (Fuhrmann et al. 2004) as a template, the crluc gene for Renilla-luciferase was ampliWed with oligonucleotides Rluc(Pst)fw (AAACTGCAGG CCAGCAAGGTGTACGACCCCGA) and T3 (ATTAAC CCTCACTAAAGGGA) by recombinant PCR. The product was digested with PstI (underlined) and BamHI and inserted back into pRbcRL(Hsp196)/XhoI/BamHI together with a DNA fragment encoding the signal peptide of the Chlamydomonas arylsulfatase gene ARS2, which was also ampliWed by PCR with oligonucleotides Ars1 (AAACTCG AGATGGGTGCCCTCGCGGTGTTC) and Ars2 (AAA CTGCAGGTCGGCCGCATGCGCAACCGA) from plasmid pJD55 (Davies et al. 1992) and cut with XhoI and PstI, to create plasmid pxx20. Then, the ARS2-crluc gene was excised with XhoI and BamHI and inserted into pHsp70A/ RbscS2-Chlamy. The resulting plasmid was fused to plasmid pUC-Arg7-lox-B by Cre/lox-mediated site-speciWc recombination to yield the tandem expression plasmid pAES12 (Heitzer and Zschoernig 2007). In the following, the sequences of intron 1, 2 or 3 from the RBCS2 gene of C. reinhardtii (Goldschmidt-Clermont

Planta (2009) 229:873–883

875

Fig. 1 Schematic drawings of luciferase constructs and average expression levels. The export sequence for secretion of the ARS2 gene (A) was attached to a codon adapted gene variant encoding the luciferase of Renilla reniformis (crluc). DiVerent combinations of intron 1, 2 and 3 of the Rubisco small subunit gene of Chlamydomonas (RBCS2) were inserted into an NruI and a SnaBI restriction site within the coding region. The expression of all genes was mediated by the constitutive chimeric HSP70A/RBCS2 promoter and the 3⬘ untranslated region of the RBCS2 gene. In plasmids pAES15 and pAES16 a C-terminal protein-tag (hexa-histidine-tag, strepII-tag) was added for protein

puriWcation. The average luciferase activity calculated from all luminescent transformants for each plasmid is given on the right. For comparison, the activity for pAES12 was set to 100% (right). A 63 bp of ARS2 encoding the signal peptide, ARS2 arylsulfatase 2 (GenBank AF333184), HSP70A 70 kDa heat shock protein (M76725), In1, In2, In3 intron 1, 2, 3 of RBCS2, RBCS2 Rubisco small subunit 2 (X04472), UTR untranslated region of RBCS2. The positions of open reading frames (start, stop), important restrictions sites (NruI, SnaBI) and protein tags (hexa-histidine-tag, strepII-tag) are marked

and Rahire 1986) were inserted into a NruI and/or a SnaBI restriction site within the luciferase coding region (S1 and Fig. 1). Therefore, the intron sequences were ampliWed from C. reinhardtii genomic DNA via PCR with a proofreading DNA-polymerase (VentR; New England Biolabs, Beverly, MA, USA) and a speciWc pair of primers (S1). After phosphorylation, the PCR products were directly inserted into plasmid pxx20 digested with NruI or SnaBI. The resulting expression cassettes were—as above for pAES12—transferred to plasmid pHsp70A/RbscS2Chlamy and fused to pUC-Arg7-lox-B to give plasmids pAES13-14 and pAES20-23, respectively. To create pAES15 and pAES16 a double stranded oligonucleotide linker encoding a hexa-histidine- (Janknecht et al. 1991) or strepII-tag (Voss and Skerra 1997) was inserted into a maintained SnaBI restriction site directly downstream of intron 3 (S1) before the transfer to pHsp70A/RbscS2Chlamy and the subsequent plasmid-fusion.

mydomonas reinhardtii resulting in the artiWcial gene crEpo (GenBank EU940697). The leader peptide of Chlamydomonas arylsulfatase gene ARS2 and a C-terminal hexahistidine-tag were added and the complete ARS2-crEpo-his6 gene was synthesized from overlapping HPLC-puriWed oligonucleotides (Fuhrmann et al. 2005). To allow the insertion of RBCS2 intron 2 the synthetic gene was artiWcially divided into two exons (crEpo1 and crEpo2) by recombinant PCR using the Expand-High Fidelity-PCR-system (Roche, Mannheim, Germany) with oligonucleotides Ep1fw1 (AAACTCGAGATGGGTGCCCTCGCGGTGTTC) together with Ep1rev7 (TTTGGATCCGGTCTCTTCACCGCC TCGCTCAGCAGGGC) and Ep2fw1 (AAAGAAGAC AAGCAGGTGCTCCGCGGCCAAGCCC) together with Ep2rev8 (TTTGGATCCTTAATGGTGGTGATGGTGG TG). Products were ligated into vector pGEM-T (Promega, Madison, WI, USA) to yield pGEM-Epo1 and pGEM-Epo2. For subsequent cloning, the above mentioned oligonucleotides carried recognition sites for XhoI (Ep1fw1, underlined), BsaI and BamHI (Ep1rev7), BbsI (Ep2fw1) and BamHI (Ep2rev8). BsaI and BbsI cut outside their recognition sequence and create four base overhangs. In this case, these overhangs were designed to be complementary to the Wrst (Ep1rev7, bold) or last (Ep2fw1) bases of RBCS2 intron 2

Synthesis of crEpo The amino acid sequence of human erythropoietin (GenBank P01588) without the leader peptide (amino acids 1–27) was translated to the nuclear codon usage of Chla-

123

876

to allow gapless insertion of the intron sequence between the two exons. Plasmid pHsp70A/RbscS2-Chlamy already contains a copy of RBCS2 intron1 within the 5⬘ untranslated region of the promoter upstream of the start ATG. To insert RBCS2 intron 3, the plasmid was digested with BamHI and Wlled with Klenow fragment. Intron 3 was ampliWed from pAES14 with oligonucleotides rbcS2I3-5⬘ (CGTAAGTCT GGCGAGAGCCCG) and rbcS2I3-3⬘ (TCTGCGGGCGCA CGGGAAATG) applying VentR-DNA-Polymerase (New England Biolabs, Beverly, MA, USA) and ligated directly into the reWlled BamHI restriction site after phosphorylation. In the resulting pxx343, the intron 3 sequence lies downstream of the stop codon followed by a restored BamHI-site (a 3⬘-BamHI-site was not restored). RBCS2 intron 2 was also obtained by recombinant PCR from plasmid pAES14 using the Expand-High FidelityPCR-system (Roche) with oligonucleotides BbsI-rbcS2I2fw (AAAGAAGACAAGTGAGCTTGCGGGGTTGCGAGC) and BsaI-rbcS2I2rev (AAAGGATCCGGTCTCACTGC AAGCAAGGGGATGAAGGG) and ligated into vector pGEM-T to obtain pGEM-In2. Recombinant restriction sites for BbsI, BamHI and BsaI were introduced for cloning (underlined), overhangs created upon digestion with BbsI and BsaI are bold. A fragment containing the complete crEpo1 was excised from pGEM-Epo1 with XhoI and BamHI and inserted into pxx343 cut with the same enzymes. The resulting plasmid pEpo3 was incubated with BamHI and BsaI to allow the insertion of intron 2 from pGEM-In2 double digested with BamHI and BbsI via the complementary overhangs to yield pEpo4. Similarly, the second part of crEpo, excised from pGEM-Epo2 with BamHI and BbsI, was inserted into pEpo4 cut with BamHI and BsaI. Finally, the fusion of the resulting plasmid pEpo5 to pUCArg7-lox-B yielded a large tandem expression vector (pEpo6) for crEpo and the selection marker ARG7 (Heitzer and Zschoernig 2007). Growth of algae and transformation The cell wall defective and arginine auxotrophic Chlamydomonas reinhardtii strain cw15arg¡ was used for all transformations according to the glass bead method (Kindle 1990). Algae were cultivated at 25°C under constant illumination in Tris–acetate–phosphate (TAP) medium with or without the addition of 100 mg/l arginine (Harris 1989). For transformation, 7.5 £ 107 cells were agitated in the presence of 1.5
g plasmid linearized with EcoRV and incubated on TAP-agar plates without arginine. After 8 days on selective medium, arginine prototrophic transformants were transferred to transparent 96-well-plates containing 200
l TAP per well or 24-well-plates with 2 ml

123

Planta (2009) 229:873–883

TAP per well. For larger scale cultures, 50 ml or 250 ml TAP-medium was grown as above but with gentle shaking. Measurement of luminescence To detect Renilla-luciferase expression in transformants, 50
l algal cultures (logarithmic growth phase, OD800nm = 0.9–1.1) diluted with 150
l TAP-medium were tested in 96-well-plates on a PolarStar Optima microplate luminometer (BMG Labtech, Jena, Germany). As substrate, coelenterazine (0.1 mM in 10% ethanol; PJK, Kleinblittersdorf, Germany) was automatically injected to a Wnal concentration of 5
M and light emission was detected for 25 s at the highest ampliWcation. To monitor the total luciferase activity of a culture (logarithmic growth phase, OD800nm = 0.9–1.1), 50
l TAP-medium containing 0.5% Triton X-100 was added to a 50
l culture aliquot and incubated for 2 min at room temperature for cell lysis. After centrifugation (2 min, 1,000£g) the complete supernatant (green, 100
l) was diluted with 100
l TAP and used for activity determination. Similarly, cells pelleted from 50
l algal cultures were mixed with 100
l TAP-medium containing 0.25% Triton X-100 to detected luciferase within or attached to the cells. As a control, plasmid pRbcRL(Hsp196) was transformed to obtain Chlamydomonas strains expressing Renilla-luciferase within the cytosol (Fuhrmann et al. 2004). Antibody production and immunodetection Plasmid pCrLuc is a pET16b-derivate (Novagen, Madison, WI, USA) containing the entire crluc gene for Renillaluciferase inserted into the XhoI and BamHI site adding an N-terminal deca-histidine-tag to the luciferase polypeptide chain (Fuhrmann et al. 2004). After transfer into Escherichia coli strain BL21(DE3) (Invitrogen, Carlsbad, CA, USA), luciferase expression from pCrLuc was induced by supplementing the culture with 1 mM isopropyl-
-D-1-thiogalactopyranoside (IPTG) and the total soluble protein was subjected to aYnity chromatography on Ni–NTA–agarose according to the manufacturer (Qiagen, Hilden, Germany). The bound luciferase protein was eluted by raising the imidazole concentration to 250 mM. For the production of polyclonal antibodies in rabbits, 2 mg puriWed protein dialysed against 50 mM Na–phosphate buVer pH 8.0 with 100 mM NaCl was transferred to Davids Biotechnologie (Regensburg, Germany). For immunodetection, protein extracts were separated on 10 or 12% SDS polyacrylamide gels and transferred to nitrocellulose membranes (Amersham Bioscience, Uppsala, Sweden) using the semi-dry blotting method. The above mentioned anti-luciferase-antibody was used in a 1:1,000 dilution for detection of Renilla-luciferase after blocking of

Planta (2009) 229:873–883

the membrane with 7% milk powder in PBS containing 0.5% Tween-20. As a secondary antibody a diluted (1:2,000) antirabbit IgG-antibody-alkaline phosphatase conjugate (SigmaAldrich, St Louis, MO, USA) was employed and detection was completed by a chromogenic reaction using nitroblue-tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP). Equally, detection of erythropoietin was performed, except that membranes were blocked with 1% BSA in PBS with 0.5% Tween-20 and a commercial polyclonal antibody raised against recombinant human erythropoietin (R&D Systems, Minneapolis, MN, USA) was used. AYnity chromatography of tagged luciferase and erythropoietin Luciferase protein was isolated from the medium of 500 mlcultures raised from C. reinhardtii cells transformed with pAES15 (hexa-histidine-tag) or pAES16 (strepII-tag) harvested at an OD800 nm between 1.0 and 1.5. After centrifugation (5 min, 25°C, 1,000£g) supernatants were lyophilised, resuspended in 50 ml distilled water and separated by aYnity chromatography on Ni–NTA–agarose (Qiagen, Hilden, Germany) or Strep–Tactin–Sepharose (IBA GmbH, Göttingen, Germany) at 4°C according to the instructions of the manufacturers. In detail, Ni–NTA–agarose columns were washed with 5 mM imidazole and tagged protein was eluted with 250 mM imidazole in 2 ml-fractions; strepII-tagged luciferase was detached from the Strep–Tactin resin using 2.5 mM D-desthiobiotin in 6.5 ml-fractions. For immunodetection, protein from 200
l of each chromatography fraction was precipitated with methanol/chloroform (Wessel and Flugge 1984) and resuspended directly in 20
l SDS polyacrylamide sample buVer. Ten or Wfty microliter of each fraction was analysed in luciferase assays in parallel. One litre culture medium (4 £ 250 ml culture) of a C. reinhardtii strain transformed with pEpo6 was used for the puriWcation of erythropoietin provided with a C-terminal hexa-histidine-tag (OD800nm = 1.0–1.6). As for luciferase, the supernatant was lyophilised, resuspended in 100 ml distilled water and applied to a Ni–NTA–agarose column. In this case, the aYnity resin was washed with 60 mM imidazole and eluted with 1 M imidazole. Chromatography fractions were dialysed against distilled water at 4°C, lyophilised again and Wnally resuspended in 100
l SDS polyacrylamide sample buVer for subsequent immunodetection.

Results Extracellular expression of Renilla-luciferase The leader peptide of the Chlamydomonas reinhardtii ARS2 gene encoding the extracellular enzyme arylsulfatase

877

(de Hostos et al. 1989) was chosen to target transgenic proteins to the culture medium. The sequence of the N-terminal 21 amino acid export signal was placed directly upstream of crluc (Fuhrmann et al. 2004), a Renilla-luciferase gene which was codon optimized for the expression from the Chlamydomonas nuclear genome (Fig. 1, pAES12). As strain cw15arg¡ possesses only a rudimentary cell wall (de Hostos et al. 1988), it was used in all expression experiments to allow an eYcient export of proteins out of the cells into the medium. Luminescence measurements of ten randomly chosen luciferase expressing transformants revealed, that approximately 65% of the total luciferase activity was found in the culture supernatant compared to only 10% for intracellular targeted luciferase without a signal peptide (Fig. 2a). Analysis of luciferase protein by immunoblotting conWrmed the protein export (Fig. 2b). InXuence of intron sequences on luciferase expression When using a longer fragment (287 bp) of the ARS2 gene encoding the export signal peptide and containing the Wrst genomic intron, slightly enhanced levels of luciferase activity could be detected in positive transformants compared to the above mentioned approach (not shown). As a similar positive stimulation was reported for the integration of introns on the expression of two selection markers in Chlamydomonas (Lumbreras et al. 1998; Berthold et al. 2002), a set of expression plasmids was constructed to analyse the eVect in detail (Fig. 1). Pre-existing NruI and SnaBI sites within the synthetic crluc gene were selected for the integration of intron sequences. The three intron sequences of the Chlamydomonas gene for the ribulose-1,5-bisphosphate carboxylase/ oxygenase small subunit RBCS2 were chosen (GoldschmidtClermont and Rahire 1986), as the expression cassette of plasmid pAES12 already comprised promoter, intron 1 and 3⬘ untranslated region (UTR) of the RBCS2 gene. Figure 1 displays number and positions of the diVerent introns within the luciferase expression cassette of the six plasmids constructed (pAES13 to 14, pAES20 to 23). For co-transformation of cw15arg¡ cells, each expression cassette was arranged on a tandem expression vector together with the selection marker ARG7 (Heitzer and Zschoernig 2007). For all constructs, 48 arginine auxotrophic transformants were analysed for luciferase activity and the average luciferase level was determined (Fig. 1). The comparison with algae transformed with plasmid pAES12 showed, that the integration of one additional intron sequence generally led to an increase in the average luciferase activity (Fig. 1, pAES21 to 23). The sequence of intron 3 was the most eVective resulting in an almost twofold improvement, whereas intron 1 and 2 had a less pronounced inXuence.

123

878

Planta (2009) 229:873–883

Fig. 2 Extracellular expression of luciferase. a After transformation with pAES12, the luciferase activity of ten luminescent transformants was measured in pelleted cells (black) and supernatant (grey, equal to activity in culture medium) to determine the luciferase export rate. Strains transformed with plasmid pRbcRL(Hsp196) enabling intracellular luciferase expression were used as a control (pRbc). For comparison, the total activity of each construct was set to 100%. b Analysis of luciferase produced in Chlamydomonas from plasmid pAES12 (extracellular expression) and pRbcRL(Hsp196) (pRbc, intracellular

expression). Two millilitre algal cultures (logarithmic growth phase, 5 £ 106 cells/ml) were separated into cell- and medium fraction by centrifugation. After precipitation of the total protein of the medium fraction with methanol/chloroform and resuspension in sample buVer, corresponding volumes were separated on a 10% SDS-gel and incubated with a polyclonal anti-luciferase antibody after electroblotting. The untransformed strain cw15arg¡ (C) was treated equally as a control

Fig. 3 Expression of luciferase from genes containing multiple introns. a Similarly to Fig. 2a, the export rate was determined for strains generated by transformation with plasmids pEAS13, pAES14 and pAES20–pAES23. For each construct, the luciferase activity of ten independent luciferase expressing transformants was analyzed in the supernatant (grey) and the cell pellet (black). The total activity of each construct was set to 100%. b Western analysis of Chlamydomonas

strains transformed with diVerent expression plasmids (pAES 13, pAES14 and pAES20–pAES23) using a polyclonal anti-luciferase antibody. For comparison, transformants with a similar luciferase activity level were chosen (not applicable for pAES13). The integration of one or two intron sequences into random sites of the coding region of crluc had no eVect on protein size

The insertion of a third intron furthermore enhanced crluc expression (pAES13, 14 and 20) by approximately the same extent, with the exception of plasmid pAES14, where a fourfold increase was observed (449%). In pAES14 all three introns of the RBCS2 gene were present in single copy and positioned in their physiological order, which seemed to form a synergistic eVect. The correct splicing of all inserted introns was checked by Western analysis, where only secreted luciferase protein of the correct size was detectable (Fig. 3b). As expected, the extent of luciferase export was also unaVected by the additional intron sequences and ranged from 60 to 75% of the total luciferase activity (Fig. 3a).

For a detailed analysis, all luciferase expressing transformants were classiWed into diVerent groups of luciferase activity according to their individual expression level. As integration of foreign DNA into the nuclear genome of C. reinhardtii occurs via non-homologous recombination, levels of transgene expression usually cover a broad range depending on the chromosomal position of transgene insertion (Schroda et al. 2002). Figure 4 shows that the expression of luciferase formed a similar distribution for each genetic construct transformed. The insertion of one or two additional introns resulted in a shifting of this distribution to higher activity values (Fig. 4b, d), reXecting the diVerent eYciencies of the three sequences (Fig. 4a, c).

123

Planta (2009) 229:873–883

879

Fig. 4 Distribution of expression activities. According to their individual luciferase activity all luminescent transformants of each plasmid were divided into diVerent groups of certain activity levels (e.g. <10,000 units, 10,000–25,000 units, etc.), the total number of transformants for each plasmid was set to 100%. As integration of transgenes occurred by non-homologous recombination events in

Chlamydomonas, expression levels of transformants vary compliant with the transcriptional status of the integration locus. Diagrams a–d illustrate the eVect of intron integration among the diVerent constructs. Note that uneven activity intervals were only chosen for an adequate presentation

AYnity chromatography of extracellular luciferase containing protein tags

in a Wvefold enrichment (Fig. 5a, c). A comparable result was achieved using the strepII-tag and a Strep–Tactin column (not shown).

Two plasmids, pAES15 and pAES16, were produced for the isolation of secreted luciferase by providing the coding sequence with short protein tags at the C-terminus. Plasmid pAES15 encoded an additional hexa-histidine-tag (Janknecht et al. 1991) downstream of the third intron, pAES16 a so called strepII-tag (Voss and Skerra 1997) in the same position (Fig. 1). For each construct, the supernatant of a 500 ml-culture of a transformant exposing a high luciferase level was subjected to aYnity chromatography on the corresponding resin. In both cases an enrichment of luciferase protein was observed (Fig. 5): histidine-tagged luciferase eYciently bound to Ni–NTA–agarose and elution resulted

Extracellular expression of human erythropoietin The human glycoprotein erythropoietin (Epo) was chosen to test the production and export of a protein of biotechnological signiWcance in Chlamydomonas. Therefore, the coding sequence of the Epo gene was adapted according to the nuclear bias of Chlamydomonas reinhardtii and the resulting gene crEpo (GenBank EU940697) was synthesised from overlapping oligonucleotides. The human export peptide (amino acids 1–27) was replaced by the arylsulfatase leader peptide and a C-terminal hexa-histidine-tag was

123

880

Planta (2009) 229:873–883

Discussion Secretion of luciferase

Fig. 5 AYnity chromatography of protein tagged luciferase from culture supernatants. The culture medium of Chlamydomonas cells transformed with pAES15 (hexa-histidine-tag) was lyophilized, resuspended in 1/10 volume distilled water and applied to Ni–NTA–agarose (pAES15). After a washing step bound protein was eluted with 250 mM imidazole (pAES15, 2 ml-fractions). a Luciferase activity of the diVerent fractions of the Ni–NTA chromatography. Ten microliter of each fraction were monitored for luminescence and normalized to the activity of the resuspended lyophilisate (L, set to 100%). b Western analysis for the fractions of the Ni–NTA chromatography shown in (a). Two-hundred microliters of each fraction was precipitated, resuspended in sample buVer and separated on a 10% SDS-gel. For detection, the anti-luciferase antibody was used. L resuspended lyophilisate, D Xow through, W washing fraction, E1–E4 elution fraction 1–4

added for puriWcation. The coding sequence was interrupted artiWcially by the insertion of RBCS2 intron 2 in a position similarly to the consensus splicing sequence of C. reinhardtii (Fig. 6a). Intron 3 was placed directly after the stop TAA of crEpo and as for the expression of luciferase the strong HSP70A/RBCS2 tandem promoter containing intron1 of the RBCS2 gene was used for transcription. After co-transformation of cw15arg¡ cells together with the ARG7 selection marker, transformants growing without the addition of arginine were tested for the presence of the erythropoietin expression cassette by genomic PCR (not shown) and Western analysis: out of 28 transformants containing the crEpo gene, 24 displayed a single signal around 33 kDa after detection with an antibody raised against recombinant human erythropoietin (Fig. 6b) and proved the correct production of the protein. As expected, the erythropoietin protein was successfully secreted into the culture medium (Fig. 6b). Due to its C-terminal histidine-tag, the crEpo gene product could partly be isolated from the supernatant by aYnity chromatography (Fig. 6c).

123

To eYciently use Chlamydomonas as an organism for the production of recombinant proteins, the export of the product protein into the culture medium would be favourable. The secretory pathway provides all possibilities of posttranslational modiWcation of the eukaryotic cell and Wnally facilitates the isolation of the protein from the algal medium. Sayre et al. (2001) employed the signal peptide of an extracellular arylsulfatase of Chlamydomonas (ARS2) to target antigens to the algal periplasm. We used the same amino acid sequence to mediate the export of Renilla-luciferase to the exterior of cell wall defect algae. In these cells the signal sequence successfully directed approximately 65% of the total luciferase activity (and approx. 60% of erythropoietin judged from the Western blot signals) to the culture supernatant, the rest probably remained within the cell organelles or stayed attached to the remnants of the cell wall. In comparison, only 10% of activity was found in the culture supernatant upon omission of this sequence. de Hostos et al. (1988) reported, that upon induction 90% of the arylsulfatase activity is found in the medium. Integration of introns Although in most cases introns are not encoding a protein sequence, it was proposed that an important part of the genetic information essential for an eukaryotic cell is maintained within introns (Mattick 1994). Introns take part in genome wide processes like exon shuZing (Liu and Grigoriev 2004) or aVect the expression of distinct genes by alternative splicing (Lareau et al. 2004) or by regulation of transcription (Palmiter et al. 1991; Liu et al. 1995). As for other organisms including plants (Chapman et al. 1991; Koziel et al. 1996), an enhanced expression of transgenes containing intron sequences has been observed for Chlamydomonas reinhardtii (Lumbreras et al. 1998; Berthold et al. 2002). To test if this positive eVect can generally be used for transgene expression in Chlamydomonas, the crluc gene encoding the luciferase of Renilla reniformis was employed as a model gene to analyse and quantify the inXuence of introns within the expression cassette. As the artiWcial promoter driving the expression of Renilla-luciferase already contained one copy of intron 1 of the RBCS2 gene of Chlamydomonas, the two other introns of this gene were utilized for this study. The insertion of one or two additional copies of RBCS2 intron 1, 2 or 3 into two randomly chosen positions of crluc resulted in an increased level of average luciferase activity in all cases. In a previous study it was shown, that the position of an intron within the transgene is crucial

Planta (2009) 229:873–883

881

Fig. 6 Expression of erythropoietin in Chlamydomonas. The human gene for erythropoietin was artiWcially divided into two exons and adapted to the nuclear codon usage of Chlamydomonas reinhardtii. a Schematic illustration of the expression cassette in plasmid pEpo6. The two exons of crEpo were interrupted by intron 2 (In2) of the RBCS2 gene. Intron 1 (In1) and 3 (In3) were positioned within the untranslated regions. The ARS2 signal peptide (A) and a C-terminal hexa-histidine-tag were added for extracellular targeting and puriWcation. A 63 bp of ARS2 encoding the signal peptide, ARS2 arylsulfatase 2 (GenBank AF333184), HSP70A 70 kDa heat shock protein (M76725), In1, In2, In3 intron 1, 2, 3 of RBCS2, RBCS2 Rubisco small subunit 2 (X04472), UTR untranslated region of RBCS2. The positions of open reading frame (start, stop) and histidine-tag are marked. b Localisation of erythropoietin expressed from plasmid pEpo6. An

aliquot of a culture (total) was divided into medium and cell pellet by centrifugation. After protein precipitation of the medium fraction corresponding volumes of medium (med.) and cells (cells) were applied on a 12% SDS-gel and analysed with a commercial anti-erythropoietin antibody. A culture of untransformed strain cw15arg¡ was treated equally as a control. c AYnity chromatography of recombinant erythropoietin from the culture medium of a strain transformed with pEpo6 using Ni–NTA–agarose. The diVerent elution fractions were dialysed against distilled water, lyophilized and resuspended in sample buVer. After gel electrophoresis (12% SDS-gel) and electroblotting the erythropoietin protein was detected using an anti-erythropoietin antibody. L resuspended lyophilisate, D Xow through, W washing fraction, E1–E4 elution fraction 1–4

for the extend of its eVect: the inXuence of RBCS2 intron 1 was strongly dependent on the promoter variant used and maximal near the 5⬘ end of the ble coding region (Lumbreras et al. 1998). When inserted 743 bp downstream of the start codon of crluc, RBCS2 intron 3 caused the greatest increase in expression, although it was shown that intron 1 is harbouring a transcriptional enhancer sequence (Lumbreras et al. 1998). As the sequence of intron 3 is highly conserved in the two RBCS genes of Chlamydomonas and displays a direct sequence repeat, it was argued to “be involved in the expression or regulation of the genes” before (Goldschmidt-Clermont and Rahire 1986). However, no further tests to elucidate the precise activity of intron 3, e.g. integration in reverse orientation or outside the coding region, were performed. An analysis of all luminescent transformants generated, revealed, that the intron eVects are not solely reXected by the average expression levels, but are also visible in diVerent groups of expression eYcacy resulting from diVerent transcriptional conditions at the random genomic integration sites. Therefore, the positive eVects raising the expression level are not restricted to certain transformants, but generally active, independent from the chromosomal position of the inserted expression cassette. The incorporation of a third intron sequence upstream of the stop codon had only an additive eVect, except for the

combination in pAES14 containing introns (1 + 2 + 3): when a single copy of each intron was present in their physiological order and in their approximate physiological position a more than fourfold increase in expression was detected. In concert with promoter and 3⬘ untranslated region of the RBCS2 gene present in all constructs, this arrangement mimics the composition of a wildtype Chlamydomonas gene to the maximum, creating a favourable pattern for transgene expression. Although it has been shown, that cDNA can be successfully expressed from the nuclear genome of Chlamydomonas, our results demonstrate a positive eVect of intronic sequences within the expression cassette (Fischer and Rochaix 2001). The mechanism how the introns increase transgene expression needs to be further investigated, but it might act on the transcriptional or post-transcriptional level. An activation on the transcriptional level would be reXected by increased amounts of de novo synthesized RNA as a result of a direct stimulation of the initiation of RNA-polymerase II—as reported before for RBCS2 intron1 (Lumbreras et al. 1998)—or as an indirect eVect by the prevention of gene silencing. In Chlamydomonas it was shown, that foreign transgenes are eYciently recognized and silenced (Schroda 2006). Alternatively, a positive inXuence on the post-transcriptional level may include an enhanced RNA maturation initiated by the additional splicing event (Bentley 2005).

123

882

Although the diVerent introns were artiWcially introduced into the crluc genes, the correct splicing of all intron sequences could be conWrmed by the aYnity chromatography of the correct protein product via protein tags and analysis by activity assays. In Western analysis using polyclonal antibodies only one distinct signal was detected for each expression cassette, excluding early termination of translation or read through over additional sequences caused by unspliced introns. Extracellular expression of erythropoietin For crluc, both integration sites were chosen coincidently without considering the physiological splicing sites of the Chlamydomonas RBCS2 gene. For the crEpo gene, a motif close to the wildtype sequence of intron 2 was used (CG/GTGAG…GCAG/GT). The third intron within the expression cassette of crEpo was integrated directly downstream of the stop codon, and thus its correct splicing could not be tested with the methods used. However, the positioning after the stop facilitated the construction of the expression vector, which could serve as a general platform for the expression of transgenes of a similar size: as intron 1 and 3 were integrated within the 5⬘ and 3⬘ untranslated region, this part of the expression cassette remains constant and only the sequence of intron 2 has to be placed into the coding region of a desired transgene. For transgenes with longer ORFs (>1 kb), the incorporation of multiple introns is conceivable. The analysis of the exported erythropoietin protein by immunoblotting showed an increased molecular weight: the calculated signal for an unmodiWed protein is 19 kDa, instead a single band was visible at 33 kDa, which is close to the size of the physiological human protein [34 kDa, Jelkmann 1992] and of recombinant erythropoietin expressed in tobacco cells [30 kDa, Matsumoto et al. 1995] and plants [32.5 kDa, (Cheon et al. 2004)]. For the functionality of many extracellular proteins including erythropoietin, a correct glycosylation is essential. Although for Chlamydomonas the O-glycosylation at hydroxyproline residues of cell-wall proteins is well characterized, only little is known about O-glycosylation at serine/threonine or N-glycosylation (Bollig et al. 2007). Since there are substantial diVerences in the O- and N-glycosylation patterns of recombinant proteins in plant and animal cells (SaintJore-Dupas et al. 2007), a detailed analysis of the secreted erythropoietin, e.g. by mass spectroscopy is indispensable. Accordingly, the proper folding and the correct formation of the two disulWde bonds essential for the biological function of erythropoietin have to be examined in the next steps. The attachment of protein tags (hexa-histidine and strepII) to Renilla-luciferase and erythropoietin for isolation proved to be a simple and fast way to accumulate recombinant

123

Planta (2009) 229:873–883

proteins from the culture medium of the algae and provides a method to concentrate protein to analyse its post-transcriptional modiWcation and physiological activity. However, in this study even expression yields from transgenes containing multiple introns were too low to collect enough protein for further analysis starting from a 1 l culture (with approx. 100
g/l recombinant erythropoietin in the supernatant or 0.03% of the dry weight, only roughly estimated using the detection limit of the commercial anti-erythropoietin antibody). Matsumoto et al. (1995) expressed recombinant human erythropoietin in tobacco cells and reported a yield of 0.0026% of the total extractable protein. Erythropoietin expression in leafs of whole tobacco and Arabidopsis plants was more pronounced, but caused retarded vegetative growth and male sterility (Cheon et al. 2004). No such changes in morphology or growth rate were found for Chlamydomonas strains expressing erythropoietin (not shown). The results of this study emphasized, that Chlamydomonas reinhardtii should be considered as an alternative organism for the expression of extracellular recombinant proteins in the future. Although expression levels have to be enhanced further to create a platform of commercial interest and to fully proWt from the low costs and the high cell densities already established for the cultivation of this alga (Walter et al. 2003), the new data obtained from the analysis of puriWed recombinant proteins will reveal their post-translational modiWcations and will be equally crucial for future biotechnological applications of Chlamydomonas. Acknowledgments We wish to thank Regina Groebner-Fererra for perfect technical assistance and Amparo Hausherr for the puriWcation of luciferase from E. coli. This work has been Wnancially supported in parts by the Bavarian Ministry of Economic AVairs, Infrastructure, Transport and Technology, the Bavarian Ministry of Environment, Public Health and Consumer Protection and the German Federal Ministry of Education and Research including an ExistSeed grant and a BioChance grant together with the Entelechon GmbH.

References Bentley DL (2005) Rules of engagement: co-transcriptional recruitment of pre-mRNA processing factors. Curr Opin Cell Biol 17:251–256 Berthold P, Schmitt R, Mages W (2002) An engineered Streptomyces hygroscopicus aph 7⬘⬘ gene mediates dominant resistance against hygromycin B in Chlamydomonas reinhardtii. Protist 153:401–412 Bollig K, Lamshoft M, Schweimer K, Marner FJ, Budzikiewicz H, WaVenschmidt S (2007) Structural analysis of linear hydroxyproline-bound O-glycans of Chlamydomonas reinhardtii—conservation of the inner core in Chlamydomonas and land plants. Carbohydr Res 342:2557–2566 Chapman BS, Thayer RM, Vincent KA, Haigwood NL (1991) EVect of intron A from human cytomegalovirus (Towne) immediate-early gene on heterologous expression in mammalian cells. Nucleic Acids Res 19:3979–3986

Planta (2009) 229:873–883 Cheetham JC, Smith DM, Aoki KH, Stevenson JL, HoeVel TJ, Syed RS, Egrie J, Harvey TS (1998) NMR structure of human erythropoietin and a comparison with its receptor bound conformation. Nat Struct Biol 5:861–866 Cheon BY, Kim HJ, Oh KH, Bahn SC, Ahn JH, Choi JW, Ok SH, Bae JM, Shin JS (2004) Overexpression of human erythropoietin (EPO) aVects plant morphologies: retarded vegetative growth in tobacco and male sterility in tobacco and Arabidopsis. Transgenic Res 13:541–549 Davies JP, Weeks DP, Grossman AR (1992) Expression of the arylsulfatase gene from the beta 2-tubulin promoter in Chlamydomonas reinhardtii. Nucleic Acids Res 20:2959–2965 de Hostos EL, Togasaki RK, Grossman A (1988) PuriWcation and biosynthesis of a derepressible periplasmic arylsulfatase from Chlamydomonas reinhardtii. J Cell Biol 106:29–37 de Hostos EL, Schilling J, Grossman AR (1989) Structure and expression of the gene encoding the periplasmic arylsulfatase of Chlamydomonas reinhardtii. Mol Gen Genet 218:229–239 Fischer N, Rochaix JD (2001) The Xanking regions of PsaD drive eYcient gene expression in the nucleus of the green alga Chlamydomonas reinhardtii. Mol Genet Genomics 265:888–894 Franklin SE, MayWeld SP (2004) Prospects for molecular farming in the green alga Chlamydomonas. Curr Opin Plant Biol 7:159–165 Franklin SE, MayWeld SP (2005) Recent developments in the production of human therapeutic proteins in eukaryotic algae. Expert Opin Biol Ther 5:225–235 Fuhrmann M (2004) Production of antigens in Chlamydomonas reinhardtii: green microalgae as a novel source of recombinant proteins. Methods Mol Med 94:191–195 Fuhrmann M, Oertel W, Hegemann P (1999) A synthetic gene coding for the green Xuorescent protein (GFP) is a versatile reporter in Chlamydomonas reinhardtii. Plant J 19:353–361 Fuhrmann M, Hausherr A, Ferbitz L, Schodl T, Heitzer M, Hegemann P (2004) Monitoring dynamic expression of nuclear genes in Chlamydomonas reinhardtii by using a synthetic luciferase reporter gene. Plant Mol Biol 55:869–881 Fuhrmann M, Oertel W, Berthold P, Hegemann P (2005) Removal of mismatched bases from synthetic genes by enzymatic mismatch cleavage. Nucleic Acids Res 33:e58 Goldschmidt-Clermont M, Rahire M (1986) Sequence, evolution and diVerential expression of the two genes encoding variant small subunits of ribulose bisphosphate carboxylase/oxygenase in Chlamydomonas reinhardtii. J Mol Biol 191:421–432 Griesbeck C, Kobl I, Heitzer M (2006) Chlamydomonas reinhardtii: a protein expression system for pharmaceutical and biotechnological proteins. Mol Biotechnol 34:213–223 Gruber H, Kirzinger SH, Schmitt R (1996) Expression of the Volvox gene encoding nitrate reductase: mutation-dependent activation of cryptic splice sites and intron-enhanced gene expression from a cDNA. Plant Mol Biol 31:1–12 Harris EH (1989) The Chlamydomonas sourcebook: a comprehensive guide to biology and laboratory use. Academic Press, San Diego Heifetz PB (2000) Genetic engineering of the chloroplast. Biochimie 82:655–666 Heitzer M, Zschoernig B (2007) Construction of modular tandem expression vectors for the green alga Chlamydomonas reinhardtii using the Cre/lox-system. Biotechniques 43:324, 326, 328 (passim) Jacobs K, Shoemaker C, Rudersdorf R, Neill SD, Kaufman RJ, Mufson A, Seehra J, Jones SS, Hewick R, Fritsch EF et al (1985) Isolation and characterization of genomic and cDNA clones of human erythropoietin. Nature 313:806–810 Janknecht R, de MartynoV G, Lou J, Hipskind RA, Nordheim A, Stunnenberg HG (1991) Rapid and eYcient puriWcation of native histidine-tagged protein expressed by recombinant vaccinia virus. Proc Natl Acad Sci USA 88:8972–8976

883 Jelkmann W (1992) Erythropoietin: structure, control of production, and function. Physiol Rev 72:449–489 Kindle KL (1990) High-frequency nuclear transformation of Chlamydomonas reinhardtii. Proc Natl Acad Sci USA 87:1228–1232 Kovar JL, Zhang J, Funke RP, Weeks DP (2002) Molecular analysis of the acetolactate synthase gene of Chlamydomonas reinhardtii and development of a genetically engineered gene as a dominant selectable marker for genetic transformation. Plant J 29:109–117 Koziel MG, Carozzi NB, Desai N (1996) Optimizing expression of transgenes with an emphasis on post-transcriptional events. Plant Mol Biol 32:393–405 Lareau LF, Green RE, Bhatnagar RS, Brenner SE (2004) The evolving roles of alternative splicing. Curr Opin Struct Biol 14:273–282 Leon-Banares R, Gonzalez-Ballester D, Galvan A, Fernandez E (2004) Transgenic microalgae as green cell-factories. Trends Biotechnol 22:45–52 Liu M, Grigoriev A (2004) Protein domains correlate strongly with exons in multiple eukaryotic genomes—evidence of exon shuZing? Trends Genet 20:399–403 Liu K, Sandgren EP, Palmiter RD, Stein A (1995) Rat growth hormone gene introns stimulate nucleosome alignment in vitro and in transgenic mice. Proc Natl Acad Sci USA 92:7724–7728 Lumbreras V, Stevens DR, Purton S (1998) EYcient foreign gene expression in Chlamydomonas reinhardtii mediated by an endogenous intron. Plant J 14:441–447 Matsumoto S, Ikura K, Ueda M, Sasaki R (1995) Characterization of a human glycoprotein (erythropoietin) produced in cultured tobacco cells. Plant Mol Biol 27:1163–1172 Mattick JS (1994) Introns: evolution and function. Curr Opin Genet Dev 4:823–831 MayWeld SP, Franklin SE (2005) Expression of human antibodies in eukaryotic micro-algae. Vaccine 23:1828–1832 MayWeld SP, Franklin SE, Lerner RA (2003) Expression and assembly of a fully active antibody in algae. Proc Natl Acad Sci USA 100:438–442 MayWeld SP, Manuell AL, Chen S, Wu J, Tran M, Siefker D, Muto M, Marin-Navarro J (2007) Chlamydomonas reinhardtii chloroplasts as protein factories. Curr Opin Biotechnol 18:126–133 Melis A (2007) Photosynthetic H2 metabolism in Chlamydomonas reinhardtii (unicellular green algae). Planta 226:1075–1086 Palmiter RD, Sandgren EP, Avarbock MR, Allen DD, Brinster RL (1991) Heterologous introns can enhance expression of transgenes in mice. Proc Natl Acad Sci USA 88:478–482 Rochaix JD (2002) The three genomes of Chlamydomonas. Photosynth Res 73:285–293 Saint-Jore-Dupas C, Faye L, Gomord V (2007) From planta to pharma with glycosylation in the toolbox. Trends Biotechnol 25:317–323 Sayre RT, Wagner RE, Siripornadulsil S, Farias C (2001) Transgenic algae for delivering antigens to an animal. US patent number 7,410,637 Schroda M (2006) RNA silencing in Chlamydomonas: mechanisms and tools. Curr Genet 49:69–84 Schroda M, Beck CF, Vallon O (2002) Sequence elements within an HSP70 promoter counteract transcriptional transgene silencing in Chlamydomonas. Plant J 31:445–455 Voss S, Skerra A (1997) Mutagenesis of a Xexible loop in streptavidin leads to higher aYnity for the Strep-tag II peptide and improved performance in recombinant protein puriWcation. Protein Eng 10:975–982 Walter C, Steinau T, Gerbsch N, Buchholz R (2003) Monoseptic cultivation of phototrophic microorganisms–development and scale-up of a photobioreactor system with thermal sterilization. Biomol Eng 20:261–271 Wessel D, Flugge UI (1984) A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal Biochem 138:141–143

123