The Plant Journal (2015) 82, 717–729

doi: 10.1111/tpj.12844

TECHNICAL ADVANCE

Improved and versatile viral 2A platforms for dependable and inducible high-level expression of dicistronic nuclear genes in Chlamydomonas reinhardtii Thomas M. Plucinak1, Kempton M. Horken1, Wenzhi Jiang1, Jessica Fostvedt1,†, Sanh Tan Nguyen1,‡ and Donald P. Weeks1,* 1 Department of Biochemistry, University of Nebraska, Lincoln, NE 68588-0664, USA Received 24 November 2014; revised 20 March 2015; accepted 25 March 2015; published online 6 April 2015. *For correspondence (e-mail [email protected]). † Present address: AccliPhot Marie Curie Program, Martin Luther Universita€t, Halle, Germany. ‡ Present address: Department of Molecular Genetics, Ohio State University, Columbus, OH, 43210 USA.

SUMMARY A significantly improved viral 2A peptide system for dependable high-level expression of dicistronic genes in Chlamydomonas reinhardtii has been developed. Data are presented demonstrating that use of an especially proficient ‘extended FMDV 2A’ coding region allows production of two independent protein products from a dicistronic gene with almost complete efficiency. Importantly, results are also presented that demonstrate the utility of this 2A system for efficient high-level expression of foreign genes in C. reinhardtii, which has not previously been reliably achievable in this algal model system. To expand the versatility of the 2A expression system, a number of commonly used selectable marker proteins were assessed for their compatibility with the extended FMDV 2A peptide. Additional experiments demonstrate the feasibility and utility of 2A-containing dicistronic systems that rely on a strong conditional promoter for transcriptional control and a low-expression marker gene for selection. This strategy allows easy and efficient delivery of genes of interest whose expression levels require regulation either to mitigate potential toxicity or allow differential expression under controlled experimental conditions. Finally, as an additional practical demonstration of the utility of the extended FMDV 2A system, confocal fluorescence microscopy is used to demonstrate that native and foreign proteins of interest bearing post-translational remnants of the extended FMDV 2A peptide localize correctly to various cellular compartments, including a striking demonstration of the almost exclusive localization of the Rubisco small subunit protein to the pyrenoid of the C. reinhardtii chloroplast in cells maintained under ambient CO2 concentrations. Keywords: 2A, dicistronic gene, gene expression, protein production, Chlamydomonas reinhardtii, toxic protein, FMDV 2A.

INTRODUCTION The unicellular green alga Chlamydomonas reinhardtii (Chlamydomonas hereafter) has long served as a model organism for studies of photosynthesis (Moroney et al., 1989), cellular motility (Pedersen et al., 2008), and, more recently, potential biofuel synthesis (Radakovits et al., 2010; Veyel et al., 2014). Among its useful features are its publicly available nuclear genome sequence and the ability to transform all three of its genomes (nuclear, mitochondrial and chloroplast) with foreign DNA (Boynton et al., 1988; Grossman, 2005; Merchant et al., 2007). Although © 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd

transformation of the nuclear genome is possible with moderate efficiency (Kindle, 1990; Shimogawara et al., 1998), there are a number of difficulties associated with the integration and expression of foreign DNA. First, transgenes are integrated randomly during transformation. This leads to a high degree of expression variability among transformants due to position effects (Kindle, 1990). Second, transformation efficiency is quite low compared to other organisms such as Saccharomyces cerevisiae and Escherichia coli (Kindle, 1990; Shimogawara et al., 1998). 717

718 Thomas M. Plucinak et al. Using electroporation under optimal conditions, only 1 in approximately 25 000 cells are transformed in a stable manner (Shimogawara et al., 1998). Third, transgene constructs are often fragmented or rearranged prior to or during integration, including separation of the selectable mark gene from the gene of interest (GOI). This often necessitates extensive screening to identify desired transformants for further study (Shimogawara et al., 1998; Dent et al., 2005). Fourth, transgenes are commonly subject to silencing (Wu-Scharf et al., 2000; Jeong et al., 2002), resulting in unreliable long-term expression of GOIs. One method that has been utilized to address some of these difficulties is to fuse the coding regions of a GOI to a selectable marker gene, such that a single promoter is used to drive expression of the fusion protein (Fuhrmann et al., 1999). This method increases the chances that primary transformants express the desired GOI, allows more reliable gene expression over time, and reduces the amount of screening required to identify suitable transformants. However, there is a substantial risk that fusion of a GOI to a marker gene will hinder the function of the GOI, the marker or both. Viral 2A peptides represent another mechanism by which two or more genes may be translated from a single dicistronic mRNA in eukaryotic organisms. The 2A peptide system is utilized by several viral species whose hosts include a broad variety of eukaryotes (Donnelly et al., 2001a,b; Doronina et al., 2008). Most minimal 2A peptide domains are approximately 19 amino acids long, and are used to separate functional protein domains in viral polyproteins (Donnelly et al., 2001a). 2A peptides catalyze an event termed a ‘ribosomal skip’ during translation that results in production of independent peptides from one mRNA (Doronina et al., 2008). Following this event, the N–terminal protein partner retains at its C–terminus all of the 2A amino acids except the C–terminal residue, which is always a proline. This proline remains covalently attached to the N–terminus of the downstream protein (Donnelly et al., 2001a,b). 2A peptides have been shown to function in all eukaryotes tested to date (de Felipe et al., 2006). However, they do not function in prokaryotic organisms (Donnelly et al., 1997), because eukaryotic-specific translation termination cofactors appear to be required to facilitate the 2A mechanism (Donnelly et al., 2001b; Doronina et al., 2008). 2A peptides from various sources catalyze the ribosomal skip process with various efficiencies (Donnelly et al., 2001a). The standard 2A peptides (19 amino acids) used in most studies catalyze this reaction with approximately 90% efficiency (Donnelly et al., 2001b), whereas longer 2A peptides (> 27 amino acids) exhibit 96–99% efficiency (Donnelly et al., 2001b). Use of longer 2A peptides results in production of near-stoichiometric levels of proteins separated by 2A peptides (Donnelly et al., 2001a; de Felipe et al., 2006). In addition, because the ‘skip’ mechanism occurs during translation, proteins separated by a 2A

peptide are free to localize independently. However, successful targeting of all 2A-separated proteins to specific subcellular locations is not always achieved (de Felipe et al., 2006). With regard to Chlamydomonas, utilization of the 2A system offers several advantages for simultaneous expression of multiple GOIs. First, the 2A system allows fusion of a GOI to a selectable marker that may drastically reduce screening time and make gene expression more reliable. Second, depending on the GOI and marker combination, significant levels of GOI over-expression may be obtained. Third, marker and GOI proteins have the potential to localize independently. We and others (Rasala et al., 2012, 2013) have provided preliminary reports of the utility of the 2A system for achieving expression of dicistronic messenger RNAs in Chlamydomonas. Here we provide a much expanded characterization of the 2A system in Chlamydomonas. This includes the introduction and verification of use of an extended 2A peptide (39 amino acids) from foot and mouth disease virus (FMDV) as a highly efficient system for simultaneous production of two proteins from a dicistronic gene, quantitative comparison of this extended 2A system with a standard system utilizing a 19 amino acid 2A peptide, comparison of the efficiency at which the 2A system and traditional ‘adjacent selection’ systems (see below) achieve expression of a GOI, as well as examination of the compatibility of the 2A system with a number of selectable marker genes. In addition, we demonstrate the utility of inducible promoters in conjunction with low-expression selectable marker genes to allow initial high-efficiency transformation under low constitutive expression conditions, followed by subsequent inducible over-expression of GOIs from dicistronic 2A constructs. RESULTS AND DISCUSSION Chlamydomonas processes 2A peptides efficiently, and an extended form of the FMDV 2A peptide is processed in an almost quantitative manner To determine whether various 2A peptides are properly processed in Chlamydomonas, and, if so, how efficiently, a series of marker genes was constructed comprising the bleomycin resistance gene (Bler) and the fluorescent mCherry protein separated by a 2A peptide coding region (Figures 1a and 2). 2A peptides are highly varied in their amino acid sequences as well as the efficiency by which they are processed in various eukaryotic organisms (Donnelly et al., 2001a). The most commonly used 2A peptides are usually approximately 19 amino acids long. However, it has been shown that the inclusion of a longer native N– terminal region substantially improves 2A performance (Donnelly et al., 2001a). The 2A peptide used in previous Chlamydomonas studies (Rasala et al., 2012, 2013) was based on the minimal FMDV 2A peptide, which has been

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

Improved FMDV 2A system for Chlamydomonas 719

(a)

(b)

Figure 1. Strategy for verification of 2A function in Chlamydomonas. (a) The dicistronic 2A-containing construct used to assess 2A function in vivo. If the mRNA transcribed from the Bler–2A–mCherry dicistronic gene is translated into a polypeptide that is cleaved when it reaches the 2A region, separate Bler and mCherry proteins will be produced, with the Bler protein localizing solely in the nucleus and the mCherry protein localizing to both the cytoplasm and the nucleus. Lack of separation at the 2A site during translation will produce a Bler– 2A–mCherry chimera that localizes exclusively in the nucleus. (b) Amino acid sequences of the 2A peptides used in this study: TaV 2A, derived from the Thosea asigna virus (TaV); extended FMDV 2A, derived from the foot and mouth disease virus (FMDV); *FMDV 2A, extended FMDV 2A made non-functional by replacement of two amino acids (NP) that are essential for 2A activity with non-canonical amino acids (TA).

Figure 2. Gene constructs used in the present study. Genes in RbcSP vectors are driven by the Chlamydomonas RbcS2 gene promoter and terminate with the RbcS2 gene termination region. Genes in PsaDP vectors are driven by the Chlamydomonas PsaD gene promoter and terminate with the PsaD gene termination region. 2A, extended FMDV 2A coding sequence; asterisk, non-functional (mutant) extended FMDV 2A coding region; Bler, bleomycin resistance gene exons divided by one or two introns; mCherry, mCherry fluorescence protein gene; C, coding region for the CIA5 C–terminal peptide epitope; 4, coding region for four repeats of the flexible peptide EAAAR; FKB12, coding region for the Chlamydomonas FKB12 gene; RbcS2 Coding, coding region of the RbcS2 gene; RbcS2 Term, RbcS2 gene termination region; AphVIII, paromomycin resistance gene driven by the Chlamydomonas PsaD gene promoter.

shown to be as much as 45-50% less efficient than the extended FMDV 2A system in allowing production of fulllength protein from the gene downstream of the 2A coding sequence (i.e. there is markedly less ‘pseudo-termination’ during the ‘ribosomal skip’ event with the extended FMDV 2A system) (Donnelly et al., 2001a). Thus, an extended version of the FMDV 2A peptide (Figure 1b) was

synthesized and evaluated. Moreover, we reasoned that when designing complex 2A-based polycistronic genes such as those successfully produced recently by Rasala et al. (2014), it may be beneficial to use multiple dissimilar 2A peptides to minimize the chances of recombination occurring between their respective DNA coding regions during or after transformation (Thomas and Rothstein,

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

720 Thomas M. Plucinak et al. 1989; Ma and Mitra, 2002; Szymczak et al., 2004). Therefore (as described below), another commonly used 2A peptide derived from the Thosea asigna virus (TaV 2A) (Figure 1b) (de Felipe et al., 2006), which has also been reported to promote efficient ‘ribosomal skip’ activity (Donnelly et al., 2001a), was also chosen for evaluation and comparison with the extended FMDV 2A. The Bler and mCherry marker genes were chosen to assess the efficacy of the 2A mechanism in Chlamydomonas for a number of reasons. First, they both result in easily detectable phenotypes. The Bler protein allows selection of transformants that are resistant to the antibiotic zeocin (Lumbreras et al., 1998), while the mCherry protein is easily detectable via fluorescence microscopy. In addition, imaging of fluorescent proteins that emit maximally in the red range (Shaner et al., 2004) leads to less background from chlorophyll autofluorescence in photosynthetic organisms compared to the more typically employed GFP. Second, both of these proteins have been shown to tolerate various fusion partners (Fuhrmann et al., 1999; Shaner et al., 2004). Third, their native localization patterns in eukaryotic cells are different. The Bler protein and Bler/GFP chimeras have been shown to localize predominantly to the nucleus of Chlamydomonas (Fuhrmann et al., 1999), while RFP-like proteins such as mCherry are found scattered throughout the cytoplasm and nucleus in Chlamydomonas (Knobbe et al., 2015) and in animal cells (Tsien, 1998) when they lack a localization signal. The Bler protein is required at relatively high levels because it must form 1:1 complexes with the antibiotics bleomycin or zeocin in order to neutralize their ability to create DNA double-strand breaks (Gatignol et al., 1988). Accordingly, fusion of a marker protein such as mCherry to Bler ensures that its expression level is adequate for detection. If these two markers are sepa-

Ble-mCherry

Ble-2A-mCherry

Figure 3. Fluorescence imaging of Chlamydomonas cells. WT cells (left panels) were transformed with either a Bler–mCherry gene construct (center panels) or a Bler–TaV 2A–mCherry gene construct (right panels). Chlorophyll autofluorescence is shown in green and mCherry fluorescence is shown in red.

Merge

mCherry

Chlorophyll

Wild-Type

rated by a properly functioning 2A peptide, mCherry is detected in the cytoplasm and nucleus. Conversely, if a particular 2A peptide is not processed properly, the vast majority of the mCherry signal is detected in the nucleus (Figure 1a). Indeed, when the Bler and mCherry proteins were fused without a 2A peptide (Figure 2, construct pGY3), the overwhelming majority of the mCherry signal was detected in the nucleus, a structure that, in Chlamydomonas, is confined to the anterior end of the cell and contained within a cytoplasmic cup formed by the surrounding chloroplast (Figure 3). A similar localization was previously observed with a GFP/Bler chimeric protein (Fuhrmann et al., 1999). When the TaV and extended FMDV 2A peptides were inserted between the Bler and mCherry proteins (Figure 2, constructs pKJ9 and pKV1, respectively), mCherry signal was present in abundance in the cytoplasm, indicating that the 2A peptides were being processed correctly and that the two marker proteins were localizing independently as intended (Figure 3). The mCherry protein remaining in the nucleus of these transformants may have either been unprocessed full-length Bler–2A–mCherry protein or free mCherry protein that diffused through the nuclear pores from the cytoplasm due to its relatively small size (approximately 27 kDa). To confirm these observations at the protein level and assess the processing efficiency of the extended FMDV 2A peptide compared to the minimal TaV 2A peptide in Chlamydomonas, an epitope tag derived from the C–terminus of the Chlamydomonas transcription activator CIA5 (Wang et al., 2005) was added to the C–terminus of the mCherry protein to allow its detection during protein blot analysis (Figure 2, construct pNK1). A highly sensitive and specific polyclonal antibody has been raised against this epitope (Wang et al., 2005). An additional advantage of using this

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

Improved FMDV 2A system for Chlamydomonas 721 Figure 4. Protein blot analysis of the efficiency of FMDV 2A processing in Chlamydomonas cells. Expected sizes of indicated bands are as follows: CIA5, approximately 100 kDa; full-length unprocessed Bler–FMDV 2A–mCherry–EpCIA5, 48.6 kDa, mCherry–EpCIA5, 30.4 kDa. (a) Proteins from transformed cells expressing an mCherry Bler–FMDV 2A–mCherry–EpCIA5 gene construct detected on a protein blot using polyclonal antibodies raised against a synthetic CIA5 epitope (EpCIA5). (b) Proteins from transformed cells expressing an mCherry Bler–FMDV 2A*–mCherry–EpCIA5 gene construct producing a non-functional FMDV 2A peptide.

(a)

WT WT (Ble-mCherry) WT (Ble-TaV 2A-mCherry) WT (Ble-FMDV 2A-mCherry-CIA5 Ep) 1

2

3

4

5

6

7

8

9

10

11

CIA5 Full Length mCherry mCherry: % processed 73

(b)

99

98 96

84 85

98 95 83

99

88

WT WT (Ble-mCherry) WT (Ble-FMDV-mCherry) WT (Ble-*FMDV 2A-mCherry) 1

2

3

4

5

6

7

8

9

CIA5 Full Length mCherry

epitope is that the CIA5 protein is constitutively expressed under standard growth conditions, and may serve as an internal loading control for protein blot experiments (Xiang et al., 2001; Wang et al., 2005). Utilization of this epitope to tag the mCherry protein enabled detection of processed and unprocessed mCherry protein species. Accordingly, protein blot analysis of proteins extracted from a number of pKV1 transformants (previously confirmed by fluorescence microscopy to be expressing mCherry) revealed distinct and prominent bands corresponding in size to free tagged mCherry in ten of 11 cases, and far weaker bands corresponding in size to unprocessed full length Bler–2A– mCherry–EpCIA5 (Figure 4a). Although precise quantification of the 2A peptide processing efficiency in this experiment was obscured because several lanes contained an overabundance of mature mCherry protein, estimates from more lightly loaded lanes indicate processing efficiencies of the extended FMDV 2A peptide approaching 100% (Figure 4a). In comparison, the cleavage efficiency of the TaV 2A peptide in a strain expressing the pKJ9 construct (Bler–TaV 2A–mCherry–EpCIA5) (Figure 4a, third lane from the left, and Figure S1) was markedly poorer than that

observed in lanes containing equivalent loading of products from the extended FMDV 2A-containing gene (Figure 4a, construct pKV1, lanes 1, 3, 4 and 8). No prominent bands other than the native wild-type (WT) CIA5 protein were detected in non-transformed cells or cells expressing a Bler–mCherry fusion protein (Figure 4a, first and second lanes, respectively). Western blot analysis of cells expressing Bler–FMDV 2A*–mCherry–EpCIA5 (Figure 2, construct pNK1), in which essential residues in the 2A active domain were mutated (FMDV 2A*; Figure 1b), revealed that all of the detectable mCherry protein corresponded in size to the full-length unprocessed protein (Figure 4b). These experiments demonstrate that, while both the extended FMDV 2A peptide and the TaV 2A peptide are processed efficiently in Chlamydomonas, processing of the extended FMDV 2A is almost quantitative. Therefore, for expression of 2A-containing dicistronic genes in Chlamydomonas, the extended FMDV 2A peptide should be favored in most circumstances. Because the bleomycin binding protein is needed in a 1:1 ratio with bleomycin (zeocin, phleomycin and related antibiotics) to provide protection from the DNA cleavage

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

722 Thomas M. Plucinak et al. activity of the antibiotic (Gatignol et al., 1988), high-level expression of the binding protein is required to confer bleomycin resistance. Such high-level expression was shown to be essential for successful Chlamydomonas genetic transformation during initial development of the bleomycin resistance gene as a selectable marker for this organism (Lumbreras et al., 1998). As a result, the experiments described above in which the bleomycin resistance gene was used as the selectable marker gene provide immediate evidence that constructs utilizing the 2A system support dependable high-level production of any protein whose gene is included with the bleomycin resistance gene in a 2A-containing dicistronic gene. The 2A expression platform allows expression of GOIs at dependably high levels and significantly out-performs adjacent selection with respect to GOI expression frequency Although we have shown that a marker gene may be expressed at easily detectable levels with the Bler–2A system, it is perhaps more significant to be able achieve dependable high-level expression of native GOIs in Chlamydomonas. To determine whether the Bler marker (which is required at high levels for 1:1 sequestration of zeocin) is capable of accomplishing this goal when used in conjunction with the extended FMDV 2A, the Bler– FMDV 2A gene and its mutant counterpart possessing an inactive 2A peptide (FMDV 2A*) were coupled with the coding region of the Chlamydomonas FKB12 gene to allow expression of the CrFKB12 protein in vivo (Figure 2, constructs pLZ1 and pMA4). The small FKB12 protein (approximately 12 kDa) is a cis-trans prolyl isomerase that, among other things, assists with protein folding (Crespo and Hall, 2002; Crespo et al., 2005). When the antibiotic rapamycin is present, it forms a complex with FKB12 that then binds and inhibits the target of rapamycin (TOR) protein. The TOR protein is a highly conserved eukaryotic protein that plays an essential role in coordinating cell growth and development in response to nutrient availability (Crespo and Hall, 2002; Crespo et al., 2005). Therefore, rapamycin treatment severely inhibits growth in most eukaryotic organisms. In Chlamydomonas, the FKB12 protein is non-essential, and fkb12 mutants are resistant to the effects of rapamycin (Crespo et al., 2005). Complementation of fkb12 mutants therefore cannot be directly selected by restoration of the ability of transformed cells to grow. Rather, complementation is detected by the restoration of sensitivity to rapamycin (Crespo et al., 2005), which is determined by replica plating of transformants on plates containing medium with and without rapamycin. FKB12 is considered to be a highabundance protein (Nigam et al., 1993; Crespo et al., 2005). Therefore, this protein offers the opportunity to examine the utility of the 2A expression platform in achiev-

ing dependable and efficient expression of a specific native GOI at physiologically relevant levels, and to demonstrate the markedly higher efficiency of the 2A system in achieving co-expression of two independent proteins from a single dicistronic gene compared to transformation of Chlamydomonas cells with a construct bearing two adjacent genes. The pLZ1 construct, containing a functional 2A coding sequence between the bleomycin resistance gene and the FKB12 gene, and the pMA4 construct, containing a nonfunctional 2A coding sequence (Figure 2), were used to transform an fkb12 mutant derived from Chlamydomonas strain CC3491. Transformants were selected on zeocin, and then randomly screened via replica plating for fkb12 complementation (i.e. restoration of rapamycin sensitivity). Screening was performed using 10 lM rapamycin, a concentration that is sufficiently high to allow clear-cut distinction between complemented and non-complemented cells (i.e. a level that allows moderate growth rates of fkb12 mutants, but causes almost complete growth inhibition of fkb12 mutants complemented with the FKB12 gene). A large proportion of randomly selected zeocin-resistant transformants screened displayed WT sensitivity to rapamycin (Figure S2A). This indicated that the 2A peptide system was functioning efficiently, and that the FKB12 protein was being expressed at functional levels and localizing correctly within the cell. When the mutant version of the 2A DNA sequence was used to bridge the Bler and FKB12 genes (pMA4), restoration of rapamycin sensitivity was not observed (Figure S2A), probably due to disruption of the FKB12–rapamycin–TOR complex due to steric hindrance and/or improper FKB12 localization. To verify these results, this same experiment was repeated using a different fkb12 mutant generated using transient expression of the CRISPR/Cas9 gene editing technique in Chlamydomonas strain CC503 (Jiang et al., 2014) to target the FKB12 gene. This fkb12 mutant contained a 21 bp deletion and a 4 bp insertion (sequence shown in Figure S2B), resulting in a frameshift within the coding region of the FKB12 gene. The results of this second experiment were almost identical to those of the first experiment, and again demonstrated high efficiency of the extended FMDV 2A system in supporting co-transformation and co-expression of two genes of interest from a single 2A-containing dicistronic construct. Currently, the most commonly employed method for expressing a GOI in Chlamydomonas is to position an independent selectable marker gene adjacent to a GOI expression cassette on a single plasmid that is subsequently used for transformation (Lumbreras et al., 1998; Shimogawara et al., 1998; Fischer and Rochaix, 2001). This method is referred to as adjacent selection. As previously mentioned, adjacent selection is unreliable and inefficient in achieving GOI expression, often requiring labor-intensive and tedious screening procedures. To directly com-

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

Improved FMDV 2A system for Chlamydomonas 723 pare the rates of co-expression of a GOI and a selectable marker gene linked through a 2A peptide coding region with the rates obtained by adjacent selection, an additional plasmid was constructed containing independent FKB12 and selectable marker gene expression cassettes. The selectable marker gene included in this plasmid was the commonly used AphVIII gene that provides resistance to the antibiotic paromomycin (Sizova et al., 2001) (Figure 2, construct pME2). Randomly chosen Parr transformants generated in the fkb12 mutant were subsequently assayed for fkb12 complementation. The complementation rate obtained using this adjacent selection construct (pME2) was 10%, while that achieved with the 2A expression platform (construct pLZ1) was 93% (Figure 5). The 2A expression platform facilitates subcellular localization studies in Chlamydomonas One of the most useful and widely implemented applications of fluorescent proteins is as localization markers for novel or unstudied proteins of interest. However, such studies are difficult to perform in Chlamydomonas due to the poor control of transgene expression discussed above. By coupling a highly expressed selectable marker such as Bler to the fluorescent protein-tagged protein of interest through a 2A peptide bridge, unreliable expression of the protein of interest resulting from position effects, gene silencing and transgene fragmentation may be mitigated or eliminated. Moreover, because localization of the 2A fusion partner is in most cases completely independent of localization of the selectable marker protein (de Felipe et al., 2006), a single highly expressed marker gene such as Ble may be used to perform a diverse range of protein localization studies. To demonstrate the utility of extended FMDV 2A-based expression vectors in allowing precise localization of proteins of interest in Chlamydomonas, coding regions for the Bler-2A-FKB12 versus adjacent selecon 100

Complementaon frequency (%)

90

93%

80 70

Rubisco small subunit protein (RBCS2), the CrCIA5 epitope (Figure 2) and mCherry were joined and expressed as a fusion with the FMDV 2A/Bler coding regions (Figure 2, construct pQN4). Previously, it was shown using immunogold labeling techniques that RBCS2 localizes predominantly (90%) to the chloroplast pyrenoid under ambient CO2 conditions, with a decrease to approximately 50–60% when cells are shifted to high CO2 concentrations (Borkhsenious et al., 1998; Mitchell et al., 2014). After transformation of WT cells with plasmid pQN4 and selection on zeocin-containing agar plates, randomly selected colonies were grown in liquid culture under both low (ambient) CO2 and high CO2 conditions and screened for mCherry fluorescence. Of 14 colonies examined, 12 (86%) abundantly expressed mCherry. More importantly, in all 12 of these samples, mCherry fluorescence was detectable only in the pyrenoid (Figure 6a) under low CO2 conditions. However, when cells expressing mCherry-tagged Rubisco were placed under high CO2, a distinct decrease in florescent signal was observed relative to the signal detected in cells maintained in low CO2, and a portion of the mCherrytagged Rubisco was clearly dispersed into the stroma (Figure 6a, bottom row versus middle row; enlarged view provided in Figure S3). The exclusive localization of the mCherry fluorescence signal to the pyrenoid under low CO2 conditions (Figure 6a and Figure S3) contrasts with the results shown in Figure 3 demonstrating that the Bler selectable marker protein localized almost exclusively to the nucleus while ‘free’ mCherry protein was distributed throughout the nucleus and the cytoplasm. Free mCherry and Bler–mCherry protein species were undetectable in the chloroplast in the earlier experiments (Figure 3). To further confirm expression and correct processing of the pQN4 translation products, protein blot analysis was performed using whole-cell extracts prepared from four independent mCherry-positive transformants. In all four cases, the vast majority of the protein product from the pQN4 construct had been efficiently separated at the 2A site to yield the expected RBCS2 chimera (approximately 50 kDa; Figure 6b). Also as expected, only the native CIA5 protein was detected in non-transformed control cells (Figure 6b). Evaluation of the functionality and efficacy of various selection marker/2A peptide combinations

60 50 40 30 20 10

10%

0 Ble-2A (pLZ1)

Adjacent (pME2)

Figure 5. Comparison of complementation frequencies obtained with the 2A system (using construct pLZ1) and adjacent selection (using pME2). For each of three independent repetitions, at least 96 zeocin- or paromomycin-resistant colonies were randomly screened for FKB12 complementation via replica plating onto rapamycin-containing TAP plates.

As shown above, use of the Bler selectable marker gene in combination with the 2A system allows dependable highlevel co-expression of various GOIs. However, there are numerous circumstances under which use of other or additional selectable marker genes in combination with the 2A/ GOI system is likely to be advantageous (e.g. when high constitutive levels of expression of a GOI cannot be tolerated). Furthermore, situations may arise in which it is desirable to transform a single cell with multiple dicistronic 2A constructs, perhaps each with its own strength and type

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

724 Thomas M. Plucinak et al.

(a)

Chlorophyll

mCherry

Merge

WT

QN4 Low CO2

Figure 6. Analysis of WT cells expressing mCherrytagged RbcS2 expressed from plasmid pQN4. (a) Fluorescence imaging of non-transformed WT cells grown under low (ambient) CO2 conditions (top row) and WT cells transformed with the plasmid pQN4 grown under low CO2 conditions (middle row) or high CO2 conditions (bottom row). Chlorophyll autofluorescence is shown in green and mCherry fluorescence is shown in red. (b) Protein blot analysis of whole-cell extracts of four individual pQN4 transformants in which CIA5 and RbcS2–mCherry–EpCIA5 are detected using an antibody to the CIA5 epitope. The three protein bands detected in the protein blot are CIA5 (approximately 100 kDa), the unprocessed, full-length RbcS2–2A–mCherry–EpCIA5 (approximately 70 kDa), and mCherry–EpCIA5 (successfully processed from RbcS2–2A–mCherry–EpCIA5; approximately 55 kDa).

QN4 High CO2

(b)

Processing eff. (%)

72

85

91

of promoter (e.g. conditional/non-conditional). For these reasons, a number of selectable markers commonly used in Chlamydomonas research were evaluated for compatibility with the extended FMDV 2A peptide and to determine their expression level under standard selection conditions when coupled to the FMDV 2A peptide. In addition, because there are benefits and drawbacks to placing a GOI at either the C- or N–terminal end of a 2A peptide, both of these orientations were tested for many of the selectable marker genes evaluated.

92 To evaluate these selectable marker/2A combinations in a simple manner, the Gaussia luciferase (Gluc) gene (Ruecker et al., 2008; Shao and Bock, 2008) was used as the fusion partner in all cases. The GLUC protein metabolizes the substrate coelenterazine to generate visible light that may be easily quantified using a luminometer. Gluc–2A– selectable marker and selectable marker–2A–Gluc expression vectors were constructed (Figure S4A) using the following selectable marker genes: Bler (zeocin resistance), AphVIII (paromomycin resistance) (Sizova et al., 2001),

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

Improved FMDV 2A system for Chlamydomonas 725 AphVII (hygromycin resistance) (Berthold et al., 2002), AadA (spectinomycin resistance) (Meslet-Cladiere and Vallon, 2011), ARG7 (arginino succinate lyase, for restoration of arginine auxotrophy to arg7 mutants) (Debuchy et al., 1989) and RBCS2 (Rubisco small subunit 2, for restoration of photoautotrophic growth to rbcs null mutants) (Khrebtukova and Spreitzer, 1996). As the results summarized in Figure S4B demonstrate, the expression levels and compatibilities of these selectable marker/2A combinations vary considerably. The Bler protein functions well and promotes robust Gluc expression in either orientation (Gluc– 2A–Bler or Bler–2A–Gluc). In contrast, the AphVIII, AphVII and AadA markers all behave similarly, in that functional dicistronic gene constructs were not obtained with these markers in position 1. When the markers are present in position 2, selectable marker gene function is obtained, but Gluc expression is almost undetectable. When the ARG7 cDNA occupies position 2, there is approximately tenfold greater Gluc expression compared to the expression levels obtained with the AphVII, AphVIII and AadA selectable markers. Dicistronic genes containing ARG7 in position 1 were not evaluated. Although RBCS2 is not a commonly used selection marker, it was chosen for evaluation because it is one of the most highly expressed proteins in photosynthetic organisms, and rbcs null mutants (e.g. T60–3) exist that may be complemented with RBCS2 genomic DNA (Khrebtukova and Spreitzer, 1996). However, a dicistronic gene comprising Gluc in position 1 and RBCS2 in position 2 did not complement the T60–3 rbcs mutant. Interestingly, when the Gluc marker was excised from this plasmid, leaving only an FMDV 2A–RBCS2 fusion, the resulting gene was able to complement T60–3 cells as efficiently as the complete WT RBCS2 gene (Figure S4B). Although fusion of GLUC to FMDV 2A–RBCS2 prevented proper RBCS2 functionality in this particular transgenic context, future experiments are required to determine whether it is possible to express other GOIs using the FMDV 2A–RBCS2 vector. In summary, six selectable markers available for use in Chlamydomonas were assessed for compatibility with the extended FMDV 2A peptide. Among the four commonly used antibiotic resistance markers tested (Bler, AphVII, AphVIII and AadA), all were active when placed downstream of the FMDV 2A peptide, but only the Bler protein performed well when placed upstream of the 2A peptide (Figure S4). Interestingly, the AphVII, AphVIII and AadA proteins appear not to tolerate fusions to their C–termini (Figure S4). The active sites of the homologous AphVII and AphVIII proteins overlap with their C–terminal domains (Sizova et al., 2001; Berthold et al., 2002), which probably renders this region sensitive to modification. When designing a 2A expression cassette, it is important to consider the benefits and drawbacks of placing a GOI on either the 50 or 30 side of the 2A peptide coding region. In

the case of Chlamydomonas, which has a propensity for cleaving and rearranging newly introduced DNA molecules prior to chromosomal integration, placement of the GOI upstream of the 2A and selectable maker gene sequences has the advantage of helping to ensure that transformants express the GOI. On the negative side, the protein produced from a GOI placed upstream of a 2A peptide coding sequence will always possess the residual 2A peptide (minus the 2A C–terminal proline) at its C–terminus. Conversely, proteins encoded by GOIs placed downstream of a 2A peptide coding region always retain the 2A-derived proline at their N–termini (Donnelly et al., 2001b). The degree to which a protein tolerates these relatively minor alterations must ultimately be determined empirically. However, it appears that instances in which the C–terminal 2A polypeptide or the N–terminal 2A proline interfere with the function of the protein of interest are limited (de Felipe et al., 2006). Utilization of the heat shock protein 70A gene promoter allows robust inducible expression of a 2A dicistron when coupled with a low-expression marker gene Numerous circumstances exist in which constitutive expression of a particular protein may be inhibitory, toxic or otherwise undesired. For example, certain proteins such as custom zinc finger nucleases (Szczepek et al., 2007) or CRISPR/Cas9 (Jiang et al., 2014) may cause toxicity if constitutively expressed, but may be tolerated if their expression is only transiently induced. Because several commonly available selectable markers used for transformation of Chlamydomonas (Figure S4) rely on enzymatic inactivation of antibiotics, and therefore are effective even when expressed at very low levels, we reasoned that coupling such selectable marker genes to a weak constitutive, but strongly inducible, promoter may allow us to initially select for stable transformation under non-inducing conditions where expression of potentially toxic GOIs is very low. We further reasoned that, if stable transformation is achieved under such conditions, subsequent high-level transient expression of coupled GOIs may also be observed under inducing conditions. With this strategy in mind, we designed and evaluated an inducible Chlamydomonas 2A-containing vector (Figure 7a, pPY1) that coupled expression of the potent selectable marker gene AphVIII (whose product is required in small amounts) to the Gluc gene (whose gene product, luciferase, may be readily measured), using the heat-inducible Chlamydomonas heat shock protein 70A (HSP70A) gene promoter (Schroda et al., 2000) to control transcription. To evaluate the expression strategy outlined above, we subjected WT cells to electroporation in the presence of pPY1 and selected transformants at 23°C on solid paromomycin-containing TAP medium. From hundreds of transformants, 12 colonies were randomly selected and grown

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

726 Thomas M. Plucinak et al.

(a) HSP70A Pro

Gluc

2A

AphV

(Par r)

PsaD Term

pPY1 (b)

1 400 000

131

25°C 1 200 000

40°C

Raw RLU

1 000 000

36

800 000

45

600 000

27

18

400 000

promoter, for example, may be induced more strongly than the HSP70A promoter (Ohresser et al., 1997; Schroda et al., 2000), the HSP70A promoter was used in our experiments due to the ease and rapidity with which its activity may be controlled. And finally, because the utility of using low-expression marker proteins to promote low-level constitutive expression of proteins of interest with the 2A system is limited, employing inducible promoters such as these to control expression of 2A-containing dicistronic genes (or polycistronic genes; Rasala et al., 2014) significantly increases their usefulness, and augments the overall utility and versatility of the 2A expression system. CONCLUSIONS

200 000

1 0 Control

1

4

5

6

7

Samples assayed Figure 7. Transcriptional control of a 2A dicistronic gene by the heat shock protein 70A gene promoter. (a) Diagram of the expression cassette contained in plasmid PY1 to assay heat shock inducibility. The extended FMDV 2A peptide was used in this construct. (b) Heat-responsive expression of the Gaussia luciferase (Gluc) protein. Gluc activity from sample cultures was measured at 25°C both prior to initiation of heat shock and 15 min after completion of a 45 min heat shock at 40°C. The cell densities of all samples were equalized based on cell density readings taken at OD750 prior to Gluc activity measurements. The numbers above the bars represent heat-dependent Gluc signal induction factors relative to samples of the same cells that were not heat shocked. Values are means  standard deviation for three biological replicates, expressed as raw RLU (relative luminescence units). Control refers to non-transformed Chlamydomonas CC3491 cells; numbered samples correspond to cultures derived from five independent CC3491 pPY1 transformants.

at 23°C to mid-log phase in liquid TAP medium containing paromomycin. After shifting the cultures to 40°C for 45 min, nine of the 12 cultures screened expressed the GLUC protein in a heat-responsive manner. The degree to which exposure to elevated temperature increased GLUC expression was examined more thoroughly using four of these transformants. This experiment demonstrated induction levels ranging from 18- to 131-fold above non-induced levels (Figure 7b). As expected, control non-transformed cells produced only background levels of luminescence under both heat shock and non-heat shock conditions. These experiments provide proof of concept that a Chlamydomonas inducible promoter may be successfully used to initially enable selection to occur under very low level GOI expression conditions, and subsequently allow highlevel expression of 2A-containing dicistronic transgenes upon exposure to inducing conditions. Although we chose the HSP70A promoter for these initial experiments, a number of other Chlamydomonas conditional promoters have also been characterized, including those that control the CYC6 (Ferrante et al., 2008), NIT1 (Ohresser et al., 1997) and CAH1 (Kucho et al., 1999) genes. Although the NIT1

The present study provides a clear demonstration of the power of 2A expression systems in allowing efficient generation of transgenic lines with dicistronic genes as well as reliable, high-level and controllable expression of recombinant genes from the nuclear genome of Chlamydomonas. The versatility afforded by the large array of selectable marker configurations and the expression control strategies described here provides a user-friendly molecular toolbox that significantly simplifies and facilitates durable multi-gene engineering in Chlamydomonas. Utilization of this toolbox has the potential to expedite a broad range of algal genetic engineering and basic research endeavors. EXPERIMENTAL PROCEDURES Chemicals and reagents All standard chemicals and reagents were purchased from SigmaAldrich (http://www.sigmaaldrich.com). Restriction enzymes were purchased from New England Biolabs (https://www.neb.com).

Plasmid construction The pSP124 vector (Lumbreras et al., 1998), consisting of the Bler gene containing two copies of the RbcS2 first intron under the control of the RBCS2 promoter, was used for assembly of all RbcS promoter-based constructs, and was obtained from Saul Purton (University College London, Institute of Structural and Molecular Biology). To allow construction of Bler–GOI fusions, an in-frame NdeI site (50 -CATATGGTCCTGCTCCTCGGCCAC-30 ) as well as an EcoRI site (50 -GAATTCCCGACGTCGACCCACTC-30 ) (restriction sites are underlined) were added directly after the last codon of the Bler gene. A plasmid containing an mCherry cDNA was obtained from Edgar Cahoon (Department of Biochemistry, University of Nebraska-Lincoln, NE). The mCherry cDNA was modified by PCR to add an ApaI restriction site (underlined) to the 50 terminus (50 AACCCGGGCCCCATGGTGAGCAAGG-30 ) and EcoRI and NheI sites (underlined) to the 30 terminus (50 -GATGAATTCTTAGTAGGTA CCGTTGCTAGCCTTGTACAGCTCGTCCATG-30 ). TaV and FMDV 2A cDNAs codon-optimized for expression in Chlamydomonas were synthesized via megaprimer PCR and included 50 NheI and NdeI sites and 30 ApaI sites that were used for ligation into the NdeI and ApaI sites at the 30 end of the Bler gene and the 50 end of the mCherry cDNA. To create pGenD-based Bler–2A expression vectors containing PsaD promoter and termination regions, the original pGenD plasmid (Fischer and Rochaix, 2001) was digested

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

Improved FMDV 2A system for Chlamydomonas 727 using ApaI, blunt-ended using T4 DNA polymerase, and re-ligated to ensure that the ApaI site at the end of the 2A genes was unique in this construct. The Bler gene (containing the second RbcS2 intron but not the first) was amplified by PCR from the pSP124 vector using oligonucleotide primers that added in-frame 50 NdeI and 30 NheI restriction sites. This PCR product, together with the FMDV 2A cDNA and the various fusion genes, was then ligated into the PsaD vector using the unique NdeI, NheI, ApaI and EcoRI restriction sites (Figure 2). Plasmid pME2 was constructed by placing the CrFKB12 cDNA into the NdeI and EcoRI sites in plasmid pJR38 (Neupert et al., 2009), which contains a PsaDP cassette for expressing cDNAs, adjacent to an AphVIII expression cassette. The plasmids constructed for the experiments shown in Figure S2 were all based on the construct pLZ1 (Figure 2). The various marker genes and the Gluc gene were amplified via PCR using oligonucleotide primers inserting either 50 NdeI/30 NheI or 50 ApaI/30 EcoRI restriction sites to construction the various gene combinations examined. To construct plasmid pPY1 (Figure 7), the forward PCR primer 50 -TCTAGAAGGCTTGACATGATTGGTGC-3′ (XbaI site underlined) and the reverse primer 50 -CATATGTACTGACTCTTAAGCGAGTTGAGTGG-3′ (NdeI site underlined) were used to amplify the Chlamydomonas HSP70A gene promoter, including the full 50 UTR, from WT genomic DNA. The NdeI and XbaI sites were used to replace the PsaD promoter in the Gluc– 2A–AphVIII plasmid constructed for the experiments depicted in Figure 7, placing the HSP70A promoter start codon in-frame with the Gluc cDNA.

Chlamydomonas growth conditions and transformation protocol The Chlamydomonas strain CC3491, which lacks cell walls, was used for all experiments. Standard electroporation transformation conditions (Shimogawara et al., 1998) were used. In brief, cells were grown in TAP medium (Gorman and Levine, 1965) under continuous illumination (approximately 80 lM/m2/sec) and rotation at 140 rpm, to a density of approximately 3 9 106 cell/ml. Cells were pelleted by centrifugation at 2,500 g for 5 min at 25°C and brought to a concentration of 4 9 108 cells/ml in TAP medium + 60 mM sucrose. For each individual transformation, 1 lg of circular plasmid DNA was used to transform 1 9 108 cells. Cells were allowed to recover in TAP medium + 60 mM sucrose with shaking at 140 rpm at room temperature under constant illumination for 20–24 h before plating onto solidified TAP medium containing appropriate levels of antibiotics. Plates were maintained at 25°C under constant illumination. The antibiotic resistance genes and antibiotic concentrations used were: AphVIII/ 20 lg/ml paromomycin, Bler/10 lg/ml zeocin, AadA/150 lg/ml spectinomycin, and Aph7/10 lg/ml hygromycin). The arginine-requiring mutant strain CC1618 was transformed with vectors containing a WT ARG7 gene followed by selection for complementation of arginine auxotrophy on Petri plates containing TAP medium with 1% agar, but lacking arginine, under constant illumination at room temperature. The Chlamydomonas mutant strain T60–3 (Khrebtukova and Spreitzer, 1996) (obtained from Robert Spreitzer, Biochemistry Department, University of Nebraska-Lincoln), which lacks functional RBCS1 and RBCS2 genes, was transformed with RBCS2 constructs, followed by selection for photoautotrophic growth at room temperature, under constant illumination in 5% CO2 on TAP minus acetate (TP) plates lacking a carbon source.

FKB12 complementation assay A rapamycin-resistant strain of CC3491 was generated and used for complementation experiments using a WT Chlamydomonas FKB12 gene. Following transformation and selection on the appro-

priate antibiotic, a small sample of cells from 7–10-day-old colonies were transferred into 100 ll TAP medium in 96-well plates, and allowed to acclimatize for 12–24 h in the light at room temperature. Samples were thoroughly mixed by pipetting, and then replica-plated in 5 ll volumes onto plates containing solidified TAP medium or TAP medium + 10 lg/ml rapamycin. CC3491 and the rapamycin-resistant CC3491 mutant were used as controls. The plates were wrapped in plastic wrap and kept under continuous light at room temperature. Colonies on plates were photographed 3-5 days after spotting.

Protein blotting Samples of Chlamydomonas colonies to be analyzed were picked into 2 ml TAP medium in test tubes and shaken at 285 rpm, 25°C, under continuous illumination until the density reached approximately 3–6 9 106 cells/ml. A volume of cells containing 3 9 105 were centrifuged for 5 min at 2500 g at 25°C, and re-suspended in a one-tenth the original volume of water. The re-suspended cells were then diluted with 1 volume of 2 x SDS–PAGE buffer (1 M Tris/HCl pH 6.8, 4% SDS, 2% glycerol, 20 mg bromophenol blue, 14 ll/ml 2–mercaptoethanol) and boiled for 5 min. Aliquots (10 ll) of each sample were loaded into wells of a Tris/glycine SDS–PAGE gel (http://www.bio-rad.com/) and fractionated by electrophoresis. For Western blotting, the membrane was blocked using TBST (Tris-buffered saline containing 0.05% Tween–20) in 5% not-fat dry milk. Primary CIA5 antibody (Ab153, Xiang et al., 2001) was added at a 1:8000 dilution to TBST in 5% non-fat dry milk. Horseradish peroxidase-conjugated donkey anti-rabbit secondary antibody (GE Healthcare, http://www.gelifesciences.com) was added at a dilution of 1:2500. The membrane was then treated with SuperSignal West Pico chemiluminescent substrate (Thermo Scientific, http:// www.thermoscientific.com/), followed by image capture using an Odyssey FC imaging system (LI-COR Biosciences; http://www.licor.com/).

Confocal fluorescence microscopy Images of live Chlamydomonas cells placed on a microscope slide and covered with a cover slip were captured using a Nikon (http:// www.nikonmetrology.com) A1 confocal imaging system mounted on a Nikon Eclipse 90i microscope with a 1009 objective. mCherry and chloroplast fluorescence signals were acquired sequentially with 561.5 nm excitation and 570–620 nm emission and 641 nm excitation and 662-737 nm emission, respectively, and pseudo-colored red and green for visualization.

Heat shock treatment Prior to heat shock treatment, samples were grown under continuous light to mid-log phase (approximately 3-6 x 106 cells/ml) in TAP medium and brought to equal cell densities based on readings at OD750. A volume of 2 ml of each culture were shaken in a test tube at 275 rpm, under constant illumination of approximately 80 lmol/m2/sec, at either 25°C or 40°C for 45 min. Immediately after completion of the heat shock treatment, cell samples were transferred to room temperature and allowed to equilibrate/ recover for 15 min with shaking and illumination conditions identical to those above, prior to assessing gene expression levels and the efficiency of 2A-mediated protein cleavage by protein blot analysis.

Luciferase assay Prior to luciferase assays, 7–9-day-old colonies of transformed Chlamydomonas cells, grown on TAP plates under antibiotic

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

728 Thomas M. Plucinak et al. selection and constant illumination at 25°C, were individually collected and suspended in 99 ll TAP liquid medium. Then 1 ll of 1 mM coelenterazine was mixed into each sample immediately prior to measurement of luminescence using a model 2030–000 luminometer (Turner Biosystems Instruments; http://www.promega.com). An integration period of 5 sec was used for each measurement. To assess the luciferase activity from Chlamydomonas cultures, 99 ll of culture (containing approximately 6 9 106 cells/ml) was mixed with 1 ll of 1 mM coelenterazine, immediately followed by measurement of luminescence as described above. Addition of substrate directly to non-transgenic WT Chlamydomonas cultures at a density of approximately 6 9 106 cells/ml produced a background reading of ≤ 10 000 relative luminescence units.

ACKNOWLEDGMENTS This work was supported by the US National Science Foundation (MCB-0952533 and EPSCoR-1004094) and the US Department of Energy (DOE DE-EE0001052 and DOE CAB-COMM DOE DEEE0003373).

SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article. Figure S1. Protein blot analysis of the efficiency of TaV 2A processing in Chlamydomonas cells. Figure S2. FKB12 complementation assay. Figure S3. Enlargement of portions of Figure 6 (a) to better visualize mCherry-tagged Rubisco in the chloroplast of cells transformed with construct pQN4. Figure S4. Evaluation of various selectable markers for compatibility with the extended FMDV 2A peptide.

REFERENCES Berthold, P., Schmitt, R. and Mages, W. (2002) An engineered Streptomyces hygroscopicus aph7″ gene mediates dominant resistance against hygromycin B in Chlamydomonas reinhardtii. Protist, 153, 401– 412. Borkhsenious, O.N., Mason, C.B. and Moroney, J.V. (1998) The intracellular localization of ribulose-1,5-bisphosphate carboxylase/oxygenase in Chlamydomonas reinhardtii. Plant Physiol., 116, 1585–1591. Boynton, J.E., Gillham, N.W., Harris, E.H. et al. (1988) Chloroplast transformation in Chlamydomonas with high velocity microprojectiles. Science, 240, 1534–1538. Crespo, J.L. and Hall, M.N. (2002) Elucidating TOR signaling and rapamycin action: lessons from Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev., 66, 579–591. Crespo, J.L., Diaz-Troya, S. and Florencio, F.J. (2005) Inhibition of target of rapamycin signaling by rapamycin in the unicellular green alga Chlamydomonas reinhardtii. Plant Physiol., 139, 1736–1749. Debuchy, R., Purton, S. and Rochaix, J.D. (1989) The argininosuccinate lyase gene of Chlamydomonas reinhardtii: an important tool for nuclear transformation and for correlating the genetic and molecular maps of the ARG7 locus. EMBO J., 8, 2803–2809. Dent, R.M., Haglund, C.M., Chin, B.L., Kobayashi, M.C. and Niyogi, K.K. (2005) Functional genomics of eukaryotic photosynthesis using insertional mutagenesis of Chlamydomonas reinhardtii. Plant Physiol., 137, 545–556. Donnelly, M.L., Gani, D., Flint, M., Monaghan, S. and Ryan, M.D. (1997) The cleavage activities of aphthovirus and cardiovirus 2A proteins. J. Gen. Virol., 78, 13–21. Donnelly, M.L., Hughes, L.E., Luke, G., Mendoza, H., ten Dam, E., Gani, D. and Ryan, M.D. (2001a) The ‘cleavage’ activities of foot-and-mouth disease virus 2A site-directed mutants and naturally occurring ‘2A-like’ sequences. J. Gen. Virol., 82, 1027–1041.

Donnelly, M.L., Luke, G., Mehrotra, A., Li, X., Hughes, L.E., Gani, D. and Ryan, M.D. (2001b) Analysis of the aphthovirus 2A/2B polyprotein ‘cleavage’ mechanism indicates not a proteolytic reaction, but a novel translational effect: a putative ribosomal ‘skip’. J. Gen. Virol., 82, 1013–1025. Doronina, V.A., Wu, C., de Felipe, P., Sachs, M.S., Ryan, M.D. and Brown, J.D. (2008) Site-specific release of nascent chains from ribosomes at a sense codon. Mol. Cell. Biol., 28, 4227–4239. de Felipe, P., Luke, G.A., Hughes, L.E., Gani, D., Halpin, C. and Ryan, M.D. (2006) Eunum pluribus: multiple proteins from a self-processing polyprotein. Trends Biotechnol., 24, 68–75. Ferrante, P., Catalanotti, C., Bonente, G. and Giuliano, G. (2008) An optimized, chemically regulated gene expression system for Chlamydomonas. PLoS ONE, 3, e3200. Fischer, N. and Rochaix, J.D. (2001) The flanking regions of PsaD drive efficient gene expression in the nucleus of the green alga Chlamydomonas reinhardtii. Mol. Genet. Genomics, 265, 888–894. Fuhrmann, M., Oertel, W. and Hegemann, P. (1999) A synthetic gene coding for the green fluorescent protein (GFP) is a versatile reporter in Chlamydomonas reinhardtii. Plant J. 19, 353–361. Gatignol, A., Durand, H. and Tiraby, G. (1988) Bleomycin resistance conferred by a drug-binding protein. FEBS Lett., 230, 171–175. Gorman, D.S. and Levine, R.P. (1965) Cytochrome f and plastocyanin: their sequence in the photosynthetic electron transport chain of Chlamydomonas reinhardii. Proc. Natl Acad. Sci. USA, 54, 1665–1669. Grossman, A.R. (2005) Paths toward algal genomics. Plant Physiol., 137, 410–427. Jeong, B.R., Wu-Scharf, D., Zhang, C. and Cerutti, H. (2002) Suppressors of transcriptional transgenic silencing in Chlamydomonas are sensitive to DNA-damaging agents and reactivate transposable elements. Proc. Natl Acad. Sci. USA, 99, 1076–1081. Jiang, W., Yang, B. and Weeks, D.P. (2014) Efficient CRISPR/Cas9-mediated gene editing in Arabidopsis thaliana and inheritance of modified genes in the T2 and T3 generations. PLoS ONE, 9, e99225. Khrebtukova, I. and Spreitzer, R.J. (1996) Elimination of the Chlamydomonas gene family that encodes the small subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase. Proc. Natl Acad. Sci. USA, 93, 13689– 13693. Kindle, K.L. (1990) High-frequency nuclear transformation of Chlamydomonas reinhardtii. Proc. Natl Acad. Sci. USA, 87, 1228–1232. Knobbe, A.R., Horken, K.M., Plucinak, T.M., Hakim, E.B., Cerutti, H. and Weeks, D.P. (2015) SUMOylation by a stress-specific SUMO E2 conjugase is essential for survival of Chlamydomonas reinhardtii under stress conditions. Plant Physiol., 167, 753–765. Kucho, K., Ohyama, K. and Fukuzawa, H. (1999) CO2-responsive transcriptional regulation of CAH1 encoding carbonic anhydrase is mediated by enhancer and silencer regions in Chlamydomonas reinhardtii. Plant Physiol., 121, 1329–1338. Lumbreras, V., Stevens, D.R. and Purton, S. (1998) Efficient foreign gene expression in Chlamydomonas reinhardtii mediated by an endogenous intron. Plant J. 14, 441–447. Ma, C. and Mitra, A. (2002) Intrinsic direct repeats generate consistent posttranscriptional gene silencing in tobacco. Plant J. 31, 37–49. Merchant, S.S., Prochnik, S.E., Vallon, O. et al. (2007) The Chlamydomonas genome reveals the evolution of key animal and plant functions. Science, 318, 245–250. Meslet-Cladiere, L. and Vallon, O. (2011) Novel shuttle markers for nuclear transformation of the green alga Chlamydomonas reinhardtii. Eukaryot. Cell, 10, 1670–1678. Mitchell, M.C., Meyer, M.T. and Griffiths, H. (2014) Dynamics of carbon concentrating mechanism induction and protein re-localisation during the dark to light transition in synchronised Chlamydomonas. Plant Physiol., 166, 1073–1082. Moroney, J.V., Husic, H.D., Tolbert, N.E., Kitayama, M., Manuel, L.J. and Togasaki, R.K. (1989) Isolation and characterization of a mutant of Chlamydomonas reinhardtii deficient in the CO2 concentrating mechanism. Plant Physiol., 89, 897–903. Neupert, J., Karcher, D. and Bock, R. (2009) Generation of Chlamydomonas strains that efficiently express nuclear transgenes. Plant J. 57, 1140–1150. Nigam, S.K., Jin, Y.J., Jin, M.J., Bush, K.T., Bierer, B.E. and Burakoff, S.J. (1993) Localization of the FK506-binding protein, FKBP 13, to the lumen of the endoplasmic reticulum. Biochem. J. 294(Pt 2), 511–515.

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729

Improved FMDV 2A system for Chlamydomonas 729 Ohresser, M., Matagne, R.F. and Loppes, R. (1997) Expression of the arylsulphatase reporter gene under the control of the nit1 promoter in Chlamydomonas reinhardtii. Curr. Genet., 31, 264–271. Pedersen, L.B., Veland, I.R., Schroder, J.M. and Christensen, S.T. (2008) Assembly of primary cilia. Dev. Dyn., 237, 1993–2006. Radakovits, R., Jinkerson, R.E., Darzins, A. and Posewitz, M.C. (2010) Genetic engineering of algae for enhanced biofuel production. Eukaryot. Cell, 9, 486–501. Rasala, B.A., Lee, P.A., Shen, Z., Briggs, S.P., Mendez, M. and Mayfield, S.P. (2012) Robust expression and secretion of xylanase1 in Chlamydomonas reinhardtii by fusion to a selection gene and processing with the FMDV 2A peptide. PLoS ONE, 7, e43349. Rasala, B.A., Barrera, D.J., Ng, J., Plucinak, T.M., Rosenberg, J.N., Weeks, D.P., Oyler, G.A., Peterson, T.C., Haerizadeh, F. and Mayfield, S.P. (2013) Expanding the spectral palette of fluorescent proteins for the green microalga Chlamydomonas reinhardtii. Plant J. 74, 545–556. Rasala, B.A., Chao, S.S., Pier, M., Barrera, D.J. and Mayfield, S.P. (2014) Enhanced genetic tools for engineering multigene traits into green algae. PLoS ONE, 9, e94028. Ruecker, O., Zillner, K., Groebner-Ferreira, R. and Heitzer, M. (2008) Gaussia-luciferase as a sensitive reporter gene for monitoring promoter activity in the nucleus of the green alga Chlamydomonas reinhardtii. Mol. Genet. Genomics, 280, 153–162. Schroda, M., Blocker, D. and Beck, C.F. (2000) The HSP70A promoter as a tool for the improved expression of transgenes in Chlamydomonas. Plant J. 21, 121–131. Shaner, N.C., Campbell, R.E., Steinbach, P.A., Giepmans, B.N., Palmer, A.E. and Tsien, R.Y. (2004) Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol., 22, 1567–1572. Shao, N. and Bock, R. (2008) A codon-optimized luciferase from Gaussia princeps facilitates the in vivo monitoring of gene expression in the model alga Chlamydomonas reinhardtii. Curr. Genet., 53, 381–388.

Shimogawara, K., Fujiwara, S., Grossman, A. and Usuda, H. (1998) Highefficiency transformation of Chlamydomonas reinhardtii by electroporation. Genetics, 148, 1821–1828. Sizova, I., Fuhrmann, M. and Hegemann, P. (2001) A Streptomyces rimosus aphVIII gene coding for a new type phosphotransferase provides stable antibiotic resistance to Chlamydomonas reinhardtii. Gene, 277, 221–229. Szczepek, M., Brondani, V., Buchel, J., Serrano, L., Segal, D.J. and Cathomen, T. (2007) Structure-based redesign of the dimerization interface reduces the toxicity of zinc-finger nucleases. Nat. Biotechnol., 25, 786– 793. Szymczak, A.L., Workman, C.J., Wang, Y., Vignali, K.M., Dilioglou, S., Vanin, E.F. and Vignali, D.A. (2004) Correction of multi-gene deficiency in vivo using a single ‘self-cleaving’ 2A peptide-based retroviral vector. Nat. Biotechnol., 22, 589–594. Thomas, B.J. and Rothstein, R. (1989) The genetic control of direct-repeat recombination in Saccharomyces: the effect of rad52 and rad1 on mitotic recombination at GAL10, a transcriptionally regulated gene. Genetics, 123, 725–738. Tsien, R.Y. (1998) The green fluorescent protein. Annu. Rev. Biochem., 67, 509–544. Veyel, D., Erban, A., Fehrle, I., Kopka, J. and Schroda, M. (2014) Rationales and approaches for studying metabolism in eukaryotic microalgae. Metabolites, 4, 184–217. Wang, Y., Sun, Z., Horken, K.M., Im, C.-S., Xiang, Y., Grossman, A.R. and Weeks, D.P. (2005) Analyses of CIA5, the master regulator of the carbonconcentrating mechanism in Chlamydomonas reinhardtii, and its control of gene expression. Can. J. Bot., 83, 765–779. Wu-Scharf, D., Jeong, B., Zhang, C. and Cerutti, H. (2000) Transgene and transposon silencing in Chlamydomonas reinhardtii by a DEAH-box RNA helicase. Science, 290, 1159–1162. Xiang, Y., Zhang, J. and Weeks, D.P. (2001) The Cia5 gene controls formation of the carbon concentrating mechanism in Chlamydomonas reinhardtii. Proc. Natl Acad. Sci. USA, 98, 5341–5346.

© 2015 The Authors The Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2015), 82, 717–729