Single‐Plasmid‐Based System for Efficient Noncanonical Amino Acid Mutagenesis in Cultured Mammalian Cells

We describe a new expression system for efficient non‐canonical amino acid mutagenesis in cultured mammalian cells by using the pyrrolysine tRNA synthetase/tRNACUAPyl pair. A significant improvement in the incorporation of non‐canonical amino acids into proteins was obtained by combining all the required genetic components into a single and compact vector that can be efficiently delivered to different mammalian cell lines by conventional transfection reagents.

nal suppressor tRNA by the evolvedt RNAs ynthetase, and translational elongation efficiency.T he latter may be affected by the identityo fb ases aroundt he in-frame stop-codonm utation and competition between the aminoacylated suppressor tRNA andrelease factors. [16][17][18][19] In at wo-plasmid-basedm ammalian expression system described by Gautier et al., [20] four copieso ft RNA CUA Pyl were simultaneously expressedi no rder to increase the intracellular levels of tRNA CUA Pyl (Figure 1, plasmids a and b). In another two-plasmid-based system recently described by Chin and coworkers, [17] significantly improved NCAA mutagenesis was achieved by using carefullys elected promoters and the co-expression of eight copieso ft RNA CUA Pyl together with mutated release factor 1. In both systems, expression of tRNA CUA Pyl was driven by at ype-3 pol III promoter,w hich enables efficient ex-  [20] Plasmid c is based on pBudCE4.1, with PylRS expression controlled by the CMV promoter andthe protein of interestisc ontrolledbyE F1a promoter. Plasmid d is basedo nt he same backboneand encodest wo PylT cassettes. In plasmid e,all components for NCAA mutagenesis in culturedm am-pression of foreignt RNA in mammalian cells. [14] Although the intracellular level of suppressor tRNA is al imiting factor in NCAA mutagenesis, increased levelso ft RNAc an have toxic effects and interferew ith downstream processes. In addition, the use of at wo-plasmid-based system can limit the integration of NCAA mutagenesis methods with other methodologies and assays;f or example, dual labeling of proteins,e xpressiono f more than one labeled protein, or assays (such as the luciferase assay) that require co-transfection with other plasmids.
Schultz and co-workers [21] demonstrated enhanced NCAA mutagenesis in mammalian cells by using as ingle-plasmidbased baculovirus expression system;t his plasmidw as subsequently used with chemical transfection methods. [22] Although significantly improving NCAA mutagenesis, baculovirus expression systemsa re less common in research laboratories. Importantly,c ombining all the required components on one plasmid resultedi nar elatively large construct (> 12 kb, not including the gene of interest). [21] The use of al arge plasmid in transient transfection methods can reduce transfection efficiency and expression level,a sw ell as limit the size of the protein to be studied.
Our goal, therefore, was to design am inimal expression vector for efficient NCAA mutagenesis by using the PylRS/ tRNA CUA Pyl pair.T of acilitatet he use of this vector,i tw ould have to be smalle nough to allow the expression of large proteins by commonc loning techniques and chemical transfection methods. The smalls ize of the plasmid might also enablee fficient co-transfection with additional plasmids, including cotransfection with as imilar vector for the expression of two different proteins.I na ddition, the designed plasmidm ust have a selectionm arker for stable transfection in mammalian cells.
We based our expression vector on the pBudCE4.1 backbone (Invitrogen) that enables the co-expressiono ft wo proteins from two independent promoters:h uman cytomegalovirus (CMV) immediate-early promoter,a nd the human elongation factor 1a-subunit (EF-1a)p romoter.T he selection marker, Zeocin,c an be used in both bacteria and mammalian cells, thereby eliminating the need for two selection markers. As a reporterg ene, we expressed the mCherry-TAG-eGFP-HA construct suggested by Gautiere tal. [20] mCherrye xpression serves as at ransfectionc ontrol, whereas eGFP expression is detected only if an NCAA is incorporated in response to an in-frame TAG codon. First, we cloned wild-type Methanosarcina mazei PylRS downstream of the CMV promoter and mCherry-TAG-eGFP-HA under the control of EF-1a promoter,t hereby creating plasmid pBud-PylS-mCherry-TAG-eGFP (Figure 1, plasmid c). Co-transfection of plasmids c and b led to as light improvement in amber suppression efficiency compared to co-transfection of plasmids a and b (Figure 2, lane 1v s. lane 5). Although plasmids a and c are based on different backbones, the improvement in amber suppression efficiency measuredw ith plasmid c can be attributed to the use of the stronger EF-1a promoter (comparedt oC MV) for the expression of mCherry-TAG-eGFP-HA (FigureS1i nt he Supporting Information).
In order to obtain high expression levels of tRNA CUA Pyl ,s everal copies of tRNA CUA Pyl are commonly expressed (usually four), with each copy under the control of an independent promot-er. [14,23,24] However, such tandema rrays are prone to homologous recombination. We therefore used two tRNAc assettes each comprising two different copieso ft RNA CUA Pyl (described by Chatterjee et al.). [21] In this cassette, aU 6p romoter and mutated (U25C) Methanosarcina mazei tRNA CUA Pyl are followed by an H1 promoter and Desulfitobacterium hafniense tRNA CUA Pyl ( Figure 1, plasmid d). [14] As light improvement in NCAA incorporation wasa chieved by using the new 4tRNA plasmid d compared to plasmid b (Figure 2, lane 1v s. lane 3). This can be attributed to the different promoters for tRNA CUA Pyl expression and the use of the U25C tRNA CUA Pyl mutant. [25] Am ore significant improvement in NCAA mutagenesis wasa chieved by using plasmids c and d rather than a and b (Figure 2, lane 1v s.  lane 7).
In order to create as ingle-plasmid-based system for NCAA mutagenesis, we combined plasmids c and d (both based on pBudCE4.1) to create pBud-PylS-mCherry-TAG-eGFP-4 tRNA (Figure 1, plasmid e). Using the single plasmid e,w ew ere able to achieve as ignificant improvement in NCAA mutagenesis ( Figure 2, lane 1v s. lane 9, Figure S2). Encoding the same components on two plasmids had ap ositive effect on NCAA mutagenesise fficiency (Figure 2, lane 1v s. lane 7);h owever, the combination of all components on as ingle plasmid had the most significant improvement (Figure 2, lane 7v s. 9). We tested all four possible orientationso ft he two tRNA cassettes and found that orientation had no effect ( Figure S3). We eventually selected opposing orientations for the tRNA cassettes as described in Figure 1( plasmid e). The size of the new vectori s 7.5 kb (excluding the protein of interest), which is significantly less than the 12 kb baculovirus expression vector that comprises currently the only single-plasmid-based mammalian expression system.
When using mCherry-TAG-eGFP-HA as ar eporter, fluorescent confocal microscopy ands ingle-cell fluorescencem easurement (Operetta, PerkinElmer)r evealed that the improved NCAA mutagenesis observed with the single plasmidcould be attributed not only to an overall increase in transfection efficiency and protein expression (Figures S4 and S5, mCherry expression), but also to as ignificant improvement in amber suppression efficiency ( Figures S4 and S5, eGFP expression). The overall yield of eGFP-150TAG expressed in HEK 293T cells from plasmid e was 0.57 mgp er 10 5 cells ( Figure S6).
New methodologies and tools for NCAA mutagenesis in cultured mammalian cells are usually validated in HEK 293 cells, where NCAA mutagenesis efficiency is known to be very high. To demonstrate the advantageo fo ur new expression system, we compared the incorporation level of 1 in four cell lines with known differences in NCAA mutagenesis efficiency (Figures 3a nd S2). The improvement in NCAA incorporation in HEK 293T,H CT116, and NIH3T3 cells was 12-to 20-fold. In HeLa cells, the improvement was even more dramatic, as incorporation of 1 by at wo-plasmid-basedm ammalian expression system wasbelow our detection limit.
In order to demonstrate the applicabilityo ft his expression vector to the study of biologically relevant systems, we incorporated Ne-acetyl-l-lysine (2)i nto one position (and even two positions) in the tumors uppressor protein p53,e xogenously expressed in ap 53-null cell line (Figure 4). [26] Acetylation of p53 at Lys120 (and to some extenta tL ys164)h as been shown to modulate the transcriptional activity of p53. [27][28][29] The mechanismsb yw hich p53 differentiates between different promoters as af unction of acetylation are unknown. In order to demonstrate the potentialo fo ur expression system in the study of p53 acetylation-dependent DNA-binding specificity, we expressed wild-type p53 as well as K120TAG, K164TAG, and K120,164TAG p53 mutantsi nt he HCT116-p53 À/À cell line ( Figure 4). With our expression system, single-acetylated p53 was expressed at the same level as endogenous p53 in response to DNA damage ( Figure S7). The expressionl evel of Lys120-acetylated p53 was cell-line dependent, and achieved up to 50 %o ft he exogenously expressed wild-type protein ( Figure S8). Moreover,w ea lso detectedd ouble-acetylated (K120Ac/K164Ac) p53, thus demonstrating the high level of NCAA mutagenesise fficiency obtained by the new single-plasmid-based system.
In summary,w eh ave designed ac ompact single-plasmidbased expression system for efficient NCAA mutagenesis in cultured mammalianc ells. The use of as ingle plasmid for ambers uppression in cultured mammalian cells enhances the efficiency of NCAA mutagenesis by improving ambers uppression as well as transfection efficiency (FiguresS4a nd S5). Moreover, our single-plasmid-based system will allow efficient co-transfection with additional plasmids, and the marker selection for mammalian cells can be used in stable transfections. Figure 3. EnhancedN CAA mutagenesis in cultured mammalian cells. The level of NCAA mutagenesis was estimated by using the mCherry-TAG-eGFP reporter.Protein was expressedint he presence or absenceo f1 from plasmids a and b or plasmid e in cell lines A) HEK 293T,B )HCT116, C) HeLa, and D) NIH3T3. Expressed mCherry-eGFP-HA was visualized by western blotting with antibodies against the C-terminal HA tag. Protein expression levels were quantifiedbyd ensitometry (ImageJ) [30] and normalizedt oa ctin (n = 3, mean AE SD). Figure 4. Expression of site-specifically acetylated p53 in HCT116-p53 (À/À) cells. Wild-type or lysine-acetylated p53 was transiently expressed in HCT116-p53 (À/À) cells by using our expressionsystem.C ellswereincubated in the presence( + +)o ra bsence( À)o f2 (5 mm). Expression levels of acetylated p53 are compared to the expression level of exogenously expressed wild-typep53 (lanes 1a nd 2, 50 %o ft otal protein) whichi si ndependent of 2.The improved expression system also enabled the expression of K120,K164 double-acetylated p53 (lanes 7, 8). DN'-p53 are known N'-truncated variants of p53.