Article | Published:

Genetic code expansion in stable cell lines enables encoded chromatin modification

Nature Methods volume 13, pages 158164 (2016) | Download Citation


Genetically encoded unnatural amino acids provide powerful strategies for modulating the molecular functions of proteins in mammalian cells. However, this approach has not been coupled to genome-wide measurements, because efficient incorporation of unnatural amino acids is limited to transient expression settings that lead to very heterogeneous expression. We demonstrate that stable integration of the Methanosarcina mazei pyrrolysyl-tRNA synthetase (PylRS)/tRNAPylCUA pair (and its derivatives) into the mammalian genome enables efficient, homogeneous incorporation of unnatural amino acids into target proteins in diverse mammalian cells, and we reveal the distinct transcriptional responses of embryonic stem cells and mouse embryonic fibroblasts to amber codon suppression. Genetically encoding N-ɛ-acetyl-lysine in place of six lysine residues in histone H3 enables deposition of pre-acetylated histones into cellular chromatin, via a pathway that is orthogonal to enzymatic modification. After synthetically encoding lysine-acetylation at natural modification sites, we determined the consequences of acetylation at specific amino acids in histones for gene expression.


Genetic code expansion enables the site-specific incorporation of 'designer' amino acids into proteins produced in cells1. Numerous useful unnatural amino acids have been incorporated, facilitating new approaches for studying outstanding problems in biology2,3,4,5.

Genetic code expansion uses orthogonal aminoacyl-tRNA synthetase/tRNACUA pairs that direct the incorporation of unnatural amino acids in response to an amber stop codon introduced at a desired site in a gene of interest1. The PylRS (encoded by PylS)/tRNAPylCUA (encoded by PylT) pair from Methanosarcina species (commonly M. mazei, as used here, or M. barkeri) has emerged as a particularly versatile system for genetic code expansion1. This pair has been evolved to direct the incorporation of numerous structurally and functionally diverse designer amino acids, and it has been developed for the incorporation of unnatural amino acids in diverse hosts, including Escherichia coli, Saccharomyces cerevisiae, Caenorhabditis elegans, Drosophila melanogaster and mammalian cells1.

Current methods for incorporating unnatural amino acids in mammalian cells are based primarily on transient transfection and/or transient expression6,7,8,9, limiting the scope of most unnatural amino acid mutagenesis experiments to cell lines that can be efficiently transfected. Because transient expression (including viral transduction7) experiments lead to heterogeneous expression levels, current approaches severely limit the ability of investigators to couple precise perturbations—which may be effected by unnatural amino acid mutagenesis—to global genomic, epigenomic, transcriptomic, metabolomic or proteomic measurements. An ideal method for introducing unnatural amino acids into cells would express PylS, PylT and a gene of interest containing an amber codon from an integrated locus, facilitating uniform levels of unnatural amino acid incorporation for all cells in a clonal population. The multiple copies of PylT commonly required for efficient unnatural amino acid incorporation in mammalian cells8 are too large and too repetitive for retroviral packaging10 and cannot be straightforwardly installed in diverse cell types.

PiggyBac transgenesis enables the rapid and efficient integration of large and complex sequences into the genome of mammalian cells and many other hosts11. Here we demonstrate that PiggyBac transposon–mediated integration of an optimized PylS/PylT cassette containing multiple copies of PylT enables the generation of stable lines that show robust, efficient and uniform incorporation of unnatural amino acids.

Histone proteins, which package DNA in the eukaryotic nucleus, are densely post-translationally modified12, and understanding the in vivo functions of these modifications is of intense interest. N-ɛ-acetyl-lysine (AcK) and other post-translational modifications, as well as their nonremovable or selectively removable analogues (including lysine methylation13,14, trifluoro-acetyl-lysine15, crotonyl-lysine16,17, butyryl-lysine and propionyl-lysine16), have been genetically encoded into recombinant histones expressed in E. coli using an acetyl-lysine–tRNA synthetase/tRNAPylCUA pair18 (or other PylRS/tRNAPylCUA derivatives); this approach, along with other powerful strategies to synthetically install chromatin modifications19, allows scientists to address the mechanistic consequences of specific modifications on chromatin structure and function in vitro. In vitro experiments can provide a clear link between molecular cause and effect20,21,22,23 but are abstracted from the appreciable complexity of the cellular environment. In contrast, in vivo experiments typically provide a wealth of correlative information about changes in the chromatin state in a native context24, but it is commonly impossible to infer causation from these experiments.

Manipulation of cellular chromatin modifications would provide a route to understanding causation and mechanism. However, it is very challenging to selectively perturb a modification at a particular site on a histone. Manipulations of enzymes that modify histones have pleiotropic effects and may affect the modification of nonhistone proteins as much as that of histone proteins25,26, confounding efforts to assign causation to a particular modification state. One can recruit modified enzymes to genomic loci by synthetically tethering them to DNA-binding modules (such as catalytically inactive Cas9 variants)27. The locus-specific consequences observed in these experiments may result from post-translational modification of many sites on the nucleosome and other chromatin-bound proteins, and therefore they do not provide a causative link between a modification at a specific site in a protein and its consequences.

The direct genetic encoding of post-translationally modified amino acids (and their nonremovable analogues) into histones in mammalian cells would provide a way to study the specific consequences of defined modifications in histones, without the pleiotropic effects that may be mediated by modifying enzymes. Here we demonstrate that AcK can be genetically programmed into defined sites in histones and deposited into cellular chromatin. This approach reveals genes that can be regulated by the site-specific acetylation of a histone.


PiggyBac integration of the orthogonal PylRS/tRNACUA pair

We developed an approach for integration of a PylS and PylT expression cassette8 embedded in a PiggyBac targeting vector (Fig. 1a) into the mammalian genome. A resistance marker was expressed from an internal ribosome entry site in the PylS transcript to enable the selection of integrants that express PylS and PylT. By delivering the gene of interest containing the stop codon on an analogous vector, we doubled the number of PylT genes introduced.

Figure 1: Stable PiggyBac integration carrying the unnatural amino acid incorporation machinery.
Figure 1

(a) PiggyBac targeting vectors used in transient transfection and the generation of stable cell lines. Ins, insulators; IRES, internal ribosome entry site; prom, promoter; pA, polyadenylation signal; NeoR and PuroR represent resistance markers; pink triangles denote inverted terminal repeats. Amber codon between mCherry and EGFP is indicated as “TAG.” (b) Flow cytometry analysis of transiently transfected HEK293 cells and mouse ESCs (mESC) 48 h after transfection and the addition of 0.5 mM CpK. Cells were gated for live and single-cell populations. (c) Flow cytometry analysis of a HEK293 cell line (performed once) with the unnatural amino acid incorporation machinery integrated, cultured 48 h without (top) and with (bottom) the addition of 0.5 mM CpK. In b and c, numbers in plots indicate the percentage of cells in each quadrant.

Transient transfection experiments with the PiggyBac constructs confirmed the efficient, amino acid–dependent incorporation of N-ɛ-((tert-butoxy)carbonyl)-l-lysine (BocK) or N-ɛ-(((2-methylcycloprop-2-en-1-yl)methoxy)carbonyl)-l-lysine (CpK), both known PylRS substrates13,28, in response to the amber stop codon in mCherry-TAG-EGFP in HEK293 cells and mouse embryonic stem cells (ESCs) (Fig. 1b and Supplementary Fig. 1); in both cell lines, the majority of mCherry+ cells were also EGFP+.

However, flow cytometry revealed several drawbacks of transient transfection for unnatural amino acid incorporation. The low transfection efficiency of some cell lines, such as ESCs, means that many cells do not receive the unnatural amino acid incorporation machinery and/or gene of interest. For cells that are transfected, there is great variability in stop codon read-through and unnatural amino acid incorporation; this results from variability in both the expression level of the gene of interest and the expression level of the unnatural amino acid incorporation machinery (Fig. 1b and Supplementary Fig. 1), consistent with the copy number of PylT being a major determinant of unnatural amino acid incorporation efficiency8.

To address the limitations of transient transfection approaches, we created stable cell lines for unnatural amino acid incorporation (Supplementary Fig. 2). We cotransfected HEK293 cells with the two targeting vectors (Fig. 1a) and a plasmid carrying the PiggyBac transposase gene. After selection with puromycin (5 μg ml−1) and G418 (1,000 μg ml−1), we isolated single clones with strong mCherry fluorescence by flow cytometry. The clonal cell population responded uniformly to the addition of 0.5 mM CpK (Fig. 1c and Supplementary Fig. 2). Additional experiments demonstrated the generality of the approach across different cell lines (Supplementary Fig. 3).

Efficient unnatural amino acid incorporation in mouse ESCs

Integration of the unnatural amino acid incorporation machinery in mouse ESCs would provide new approaches for studying stem cell function and developmental biology29, and it might provide access to differentiated cell lines for unnatural amino acid incorporation.

We transfected an ESC line (E14) with 4×PylT/PylS, 4×PylT/sfGFP150TAG (Fig. 2a) and PiggyBac transposase and selected stable integrants with puromycin (5 μg ml−1) and G418 (1,000 μg ml−1) from 3 to 7 d after transfection. More than half of the cells in the resulting polyclonal pool expressed sfGFP when BocK or CpK was added, whereas we observed no sfGFP fluorescence in the absence of unnatural amino acid (Supplementary Fig. 4). We derived individual clones by flow cytometry that maintained robust unnatural amino acid–dependent expression, consistent with high-fidelity incorporation of the unnatural amino acid in response to the amber stop codon, and exhibited a defined and homogeneous level of sfGFP expression (Fig. 2b and Supplementary Fig. 4). We also generated E14 cell lines expressing 4×PylT/PylS and 4×PylT/mCherry-TAG-EGFP with the same protocol and observed near-quantitative, uniform and dose-dependent incorporation of BocK and CpK (Supplementary Fig. 5). These experiments demonstrate that the unnatural amino acid incorporation machinery can be stably integrated into ESCs and used to direct the site-specific incorporation of unnatural amino acids.

Figure 2: PiggyBac-mediated generation and differentiation of mouse ESC lines for unnatural amino acid mutagenesis.
Figure 2

(a) PiggyBac targeting vectors used to generate stable mouse ESC lines from an E14 line containing a bidirectional EF1 and U6 expression cassette, internal ribosome entry sites (IRES), resistance markers (NeoR, PuroR), insulators (Ins) and inverted terminal repeats (pink triangles). The top vector is 4×PylT/PylS, and the bottom vector is 4×PylT/sfGFP150TAG. sfGFP150TAG encodes superfolder GFP with an amber codon at codon 150. (b) Representative images of a clone grown in the presence or absence of 0.2 mM CpK for 48 h. Scale bar, 100 μm. (c) EB differentiation protocol used to produce beating cardiomyocyte aggregates in the presence (bottom) or absence (top) of 0.2 mM CpK. Scale bars, 100 μm. (d) Flow cytometry analysis of cells differentiated via the EB protocol in the absence of CpK and incubated with 0.5 mM CpK for 24 h. Multiple clones were derived, and two were characterized in detail.

To demonstrate the differentiation potential of ESCs containing the unnatural amino acid incorporation machinery, we performed a well-established embryoid body (EB) differentiation protocol, in the presence of CpK (0.2 mM), to produce cardiac myocytes from a cell line bearing 4×PylT/PylS and 4×PylT/sfGFP150TAG. After 7 d the EBs were smaller than those grown in the absence of CpK (Fig. 2c), but they nonetheless formed beating cardiomyocyte assemblies when replated (Supplementary Video 1). The differentiated tissue retained the propensity for unnatural amino acid incorporation, albeit at a lower efficiency, as judged by the appearance of sfGFP fluorescence upon the addition of CpK to cardiac myocytes that were differentiated in the absence of unnatural amino acid (Fig. 2d). Manipulating mouse ESCs may provide a facile in vitro strategy for deriving differentiated cell types that can site-specifically incorporate unnatural amino acids.

Cell-specific transcription responses to amber suppression

We leveraged the ability to generate stable cell lines to define the general and cell-type-specific effects of amber suppression on gene expression. In two mouse ESC (E14) and two mouse embryonic fibroblast (MEF) cell line clones in which 4×PylT/PylS and 4×PylT/sfGFP150TAG were integrated, we observed 1,302 and 36 common and significant (P < 0.005) changes in gene expression after the addition of 0.2 mM CpK (48 h), respectively (Fig. 3a,b and Supplementary Fig. 6). Only 11 genes identified in ESCs were also dysregulated in MEF cells (Supplementary Fig. 6). These observations suggest that, unlike in ESCs, which are known to be plastic30, amber suppression has a minimal effect on gene expression in MEFs.

Figure 3: RNA-seq analysis of E14 ESC lines incorporating unnatural amino acids (CpK or AcK).
Figure 3

(a) RNA-seq analysis of E14 cell line bearing 4×PylT/PylS and 4×PylT/sfGFP150TAG. The plot shows whole-transcriptome fragments per kilobase of exon per million fragments mapped (FPKM) in the presence versus absence of 0.2 mM CpK for 48 h. Significantly (P < 0.005) up- and downregulated genes from two biological replicates are colored red and blue, respectively; other genes are in black. Further analysis and biological replicates are shown in Supplementary Figure 6. (b) RNA-seq analysis of 3T3 MEF cell line bearing 4×PylT/PylS and 4×PylT/sfGFP150TAG. The plot shows whole-transcriptome FPKM values in the presence and absence of 0.2 mM CpK for 48 h. Up- and downregulated gene sets defined in ESCs (a) are colored red and blue, respectively. (c) RNA-seq analysis of E14 cell line bearing 4×PylT/AcKS-TAGDendra2. The plot shows whole-transcriptome FPKM values in the presence and absence of 10 mM AcK for 24 h. Significantly up- and downregulated genes from a are colored red and blue, respectively. (d) RNA-seq analysis of wild-type (WT) E14 cell line in the presence and absence of 10 mM AcK for 24 h. Significantly up- and downregulated genes from a are colored red and blue, respectively; other genes are in black. All experiments were performed in duplicate. Data in a and b were additionally acquired with two independent clones.

Strikingly, the 1,302 genes that were misregulated in the PylS/sfGFP150TAG/8×PylT E14 lines after the addition of CpK were similarly perturbed in the 4×PylT/AcKS cell line in the presence of AcK (Fig. 3a,c,d and Supplementary Fig. 6). This correlation suggests that the transcriptional changes we observed in ESCs were an effect of amber suppression and may have been independent of the specific unnatural amino acid incorporated at the amber stop codon.

Many misregulated genes were not stem cell specific, and the majority of stem cell–specific genes were unaffected (Supplementary Fig. 7), although known pluripotency factors, including Pou5f1, Nanog and Sox2 (ref. 31), were downregulated 1.5-fold to 2-fold (Supplementary Fig. 7). Upregulated genes fell into a variety of categories related to development (via Gene Ontology)32 (Supplementary Fig. 7), and there was a significant enrichment of primary metabolic enzymes among the downregulated genes, suggesting that amber suppression may have a repressive effect on primary metabolism in mouse ESCs, through an unknown mechanism (Supplementary Fig. 7). Notably, we did not observe upregulation of genes involved in protein-folding stress responses, which suggests that amber suppression does not cause a large number of proteins to be aberrantly elongated and misfolded, consistent with previous observations in mammalian cells33.

Encoded site-specific histone acetylation in cells

We developed a two-step strategy to generate a panel of cell lines with well-matched unnatural amino acid incorporation efficiency that incorporated an unnatural amino acid at distinct positions in a target protein or proteins (Supplementary Fig. 8). We used this strategy to generate a series of histone H3 variants in which acetyl-lysine was cotranslationally incorporated (Fig. 4a and Supplementary Fig. 8). In the first step, we cotransfected E14 cells with the 4×PylT/AcKS-TAGDendra2 cassette and PiggyBac transposase and selected with 10 μg ml−1 puromycin. We isolated a single 4×PylT/AcKS clone by flow cytometry, using a Dendra2 fluorescent reporter fused in-frame behind the AcKS ORF, separated by an amber codon, as a measure for efficient incorporation of acetyl-lysine at amber codons (Fig. 4a). Incorporation was dose dependent, and maximal at 10 mM AcK (Supplementary Fig. 9). In the second step, we cotransfected an E14 AcKS/4×PylT-TAGDendra2 cell line with the PiggyBac transposase plasmid and a 4×PylT/H3.2(XXTAG)-HA or 4μPylT/H3.3(XXTAG)-HA cassette, where XX indicates the position in the histone-expressing gene targeted for conversion of a lysine codon to TAG. We selected with 10 μg ml−1 puromycin and 2 mg ml−1 G418 and isolated polyclonal pools that were closely related to the parental cell line and to each other.

Figure 4: Site-specific incorporation of acetyl-lysine into histone H3 in ESCs.
Figure 4

(a) PiggyBac targeting vectors used to generate stable mouse ESC lines. The top vector is 4×PylT/AcKS-TAGDendra2, and the bottom vector is 4×PylT/H3.3(XXTAG)-HA. Variants of the bottom vector replace H3.3 with H3.2, vary the position of the amber codon (XX) or omit the amber codon to give wild-type H3.3 or H3.2. Pink triangles denote inverted terminal repeats; abbreviations are defined as in Figures 1 and 2. (b) Western blot of histone cell lines bearing 4×PylT/AcKS-TAGDendra2 combined with 4×PylT/H3.2-HA, 4×PylT//H3.2(XXTAG)-HA, 4×PylT/H3.3-HA or 4×PylT/H3.3(XXTAG)-HA (designated as wild-type (WT) or KXXTAG in the figure, respectively) with a C-terminal triple HA-tag and 4×PylT. Synthetic histone expression was detected in acid-extracted chromatin by western blotting against HA-tag; an endogenous histone H3 loading control is shown. Expression of histone H3.2 and H3.3 genes containing TAG codons is dependent on the addition of AcK (10 mM for 24 h). The experiment was performed in duplicate. (c) Amber-suppression efficiency in all cell lines from b, as measured by flow cytometry analysis of Dendra2 fluorescence, in the presence of 10 mM AcK. (d,e) Expression of HA-tagged histone H3 without (d) and with (e) amber codon in single cells as measured by fluorescent staining with anti-HA conjugated to Alexa Fluor 647 and flow cytometry analysis in single experiments. Data for the entire panel of H3 variants are shown in Supplementary Figure 10.

Western blotting against a C-terminal HA tag demonstrated AcK-dependent read-through of the amber codon in H3(XXTAG)-HA–containing cell lines (Fig. 4b). Although H3.3-HA was expressed continuously whereas expression of H3.3(XXTAG)-HA was induced by the addition of AcK for 24 h, the levels of H3.3 produced from H3.3(XXTAG)-HA reached 50–100% of the levels in the H3.3-HA control in most positions (Fig. 4b). All resulting cell lines showed a defined and comparable level of acetyl-lysine incorporation at an amber stop codon (Fig. 4c), and levels of histone expression were comparable between H3.2 and H3.3 pairs with the TAG codon at the same position (Fig. 4d,e and Supplementary Fig. 10). However, there was some variability in the level of acetyl-lysine incorporation at distinct sites in histone H3 (Fig. 4b), consistent with context effects in amber-suppression efficiency34, and we observed minimal incorporation for H3.3(K9TAG)-HA.

Genetically directed acetylation in chromatin

We focused on detecting H3.3 acetylation in chromatin, because, unlike H3.2, H3.3 undergoes dynamic, replication-independent deposition35. H3.3-HA from each expressed H3.3(XXTAG)-HA variant was deposited into chromatin, as judged by anti-H3.3 and anti-HA immunoblotting (Supplementary Fig. 11). The amount of transgenic H3.3 produced in the 24-h window after the addition of AcK to cells was a small fraction of the total H3.3 (Supplementary Fig. 11); this is observed in many histone transgenesis experiments and is a consequence of the large number of endogenous histone genes and tight control of cellular histone protein levels36. We directly detected acetylation on Lys37 (K37ac) and K56ac on histone H3.3-HA in chromatin-extracted histones (Supplementary Fig. 11), as well as in immunoprecipitated soluble and nucleosomal H3.3-HA (Supplementary Fig. 11). We also detected K27ac and K64ac on HA-tagged histones immunoprecipitated from nucleosomal histones (Supplementary Fig. 11). Notably, we detected K37ac, K56ac and K64ac from transgenic H3.3-HA only when acetylation was genetically encoded at these sites, showing that genetically encoded acetylation produces higher levels of acetylation than endogenous enzymes do at these sites. In contrast, we detected K27ac on all transgenic H3.3-HA, consistent with the efficient acetylation of K27 by enzymes in mouse ESCs. Transient expression experiments in HEK293T cells confirmed that K27ac can be specifically incorporated into chromatin and detected using anti-H3K27ac (Supplementary Fig. 11). We note that less K27ac was detected with anti-K27ac when the histones were pre-acetylated at K56, suggesting possible negative cross-talk of acetylations on histone H3. We did not detect K9ac or K23ac on the tagged histone species, because the antibodies used did not provide the necessary sensitivity (despite their apparent ability to detect the abundant K9 or K23 acetylation, respectively, on the endogenous histones) and/or because the encoded acetylation at those sites was rapidly lost through deacetylation. In the future, it would be interesting to investigate genetically encoding nonhydrolyzable analogues of post-translational modifications.

Consistent with the role of K56ac in histone deposition in human cells37, the majority of H3.3 K56ac-HA was in the chromatin-bound fraction (Supplementary Fig. 11). The K56ac signal resulting from H3.3(K56TAG)-HA expression was easily detectable even in the presence of the bulk endogenous histones, suggesting that the encoded acetylation was retained in H3.3K56ac-HA and that endogenous levels of K56ac are extremely low38,39.

Encoded H3 acetylation regulates transcription

To investigate changes in gene expression resulting from genetically encoded acetylation in H3.3, we carried out mRNA-seq on all H3.3 cell lines. We identified 16 candidate genes in H3.3(XXTAG)-HA cell lines (Fig. 5a) that showed (1) significant (P < 0.005) upregulation relative to the 4×PylT/AcKS-TAGDendra2 parental cell line (which provided a control for the effects of amber suppression with AcK) and (2) significant (P < 0.005) and more than twofold upregulation relative to the 4×PylT/AcKS-TAGDendra2, 4×PylT/H3.3-HA cell line (which provided a control for amber suppression and the effects of expressing the nonacetylated histone in the H3.3(XXTAG)-HA cell lines). As H3.3 with encoded acetylation provides only a fraction of H3 in the cell, the genes identified may be a subset of those regulated by H3 acetylation. We compared the expression levels in all H3.3(XXTAG)-HA cell lines and in the H3.3-HA control cell line to those in the parental cell line (Fig. 5a; Supplementary Fig. 12 shows absolute expression counts and a wider range of controls). Notably, a subset of genes identified, including Tex13, Xist, Gm14327 and Vmn1r65, were particularly strongly induced in H3.3 (K37TAG)–HA and H3.3 (K56TAG)–HA cell lines, for which we have confirmed site-specific acetylation on H3.3 in chromatin. We thus focused on the gene XIST (X-inactive specific transcript), which gives rise to the noncoding XIST RNA required for X inactivation during development40. We confirmed by reverse-transcription quantitative PCR (RT-qPCR) that XIST expression in the H3.3 K27TAG, K37TAG and K56TAG cell lines was higher than in the control H3.3 cell line in the presence of 10 mM AcK (Fig. 5b and Supplementary Fig. 12). Furthermore, we did not observe XIST induction in H3.2 K56TAG cell lines in the presence of AcK (Fig. 5b). XIST induction increased with H3.3 K56Ac expression levels (Supplementary Fig. 12). The proteins expressed from H3.3-HA, H3.3 (K37TAG)-HA and H3.3 (K56TAG)-HA were specifically enriched at the transcription start site and coding region of the XIST gene, as judged by anti-HA chromatin immunoprecipitation (ChIP) (Fig. 5c,d and Supplementary Fig. 13).

Figure 5: Genetically encoded, site-specific histone acetylation in chromatin regulates gene expression.
Figure 5

(a) RNA-seq analysis of E14 ESC lines bearing 4×PylT/AcKS-TAGDendra2 and 4×PylT/H3.3(XXTAG)-HA. Heat map of genes that were significantly (P < 0.005) upregulated in at least one H3.3(XXTAG) cell line with respect to both the parental 4×PylT/AcKS-TAGDendra2 line and the matched control 4×PylT/AcKS-TAGDendra2, 4×PylT/H3.3-HA line. Biological duplicates were performed for each condition. (b) RT-qPCR analysis of noncoding XIST RNA expression in H3.3(56TAG) cell lines, with and without the addition of 10 mM AcK (24 h) to the medium, and in a matched H3.2(56TAG) control cell line with AcK added, showing that XIST upregulation is dependent on K56 acetylation of histone H3.3. Error bars indicate the 95% confidence interval of three technical replicates. (c) Locations of qPCR primers used in d. Up, universal primer site; TSS, transcription start site; Ex, exon; In, intron. (d) ChIP assay validating variant-specific incorporation of H3.3K56TAG-HA3 in the presence of 10 mM AcK for 24 h at the Xist locus at primer sites tiling the Xist locus. ChIP assay was performed using α-HA magnetic beads. Error bars indicate the 95% confidence interval of three technical replicates. MajSat, major satellite.


The ability to rapidly generate stable cell lines with defined amber-suppression efficiency will be broadly enabling for unnatural amino acid mutagenesis. We anticipate that our approach for creating cell populations that have well-matched amber-suppression efficiency and express different mutants of proteins of interest will prove particularly useful. A key application of stable cell lines for unnatural amino acid mutagenesis will be in the coupling of precise molecular, spatial and temporal perturbations effected by site-specific unnatural amino acid incorporation with measurements of cell-, organism- or ecosystem-wide consequences. Such approaches will require careful control for the effects of amber suppression itself, and we have demonstrated that these effects are cell-type specific and require direct measurement in control experiments.

The approach we have reported for genetically installing post-translational modifications (and, by extension, their nonhydrolyzable analogues) in chromatin and a recently described approach that uses protein trans-splicing to assemble a ubiquitinated histone in chromatin within isolated cell nuclei36 are orthogonal to enzymatic modification. Orthogonal routes to chromatin modification allow the consequences of post-translational modifications at defined sites on a histone to be defined, without the pleiotropic effects of modifying enzymes at other sites on histones or on other proteins. In the future, orthogonal routes to chromatin modification may be extended to facilitate the site-specific and orthogonal modification of a histone deposited at defined genomic loci. Given the potential of histone modifications to control epigenetic phenomena, we anticipate that orthogonal approaches to chromatin modification may provide insight into whether specific modifications cause epigenetic change. Orthogonal approaches to chromatin modification may form a basis for 'synthetic epigenetics' in which precise (natural or unnatural) modification states are programmed into cells to control cell and organism fate.



The following primary antibodies were used for western blotting: anti-GFP (sc-9996; Santa Cruz Biotechnologies), anti-mCherry (ABE3523; SourceBiosciences), anti-Flag M2 (A8592; Sigma), anti-actin (4967; Cell Signaling Technology), anti-H3 (ab1791; Abcam), anti-H3K56ac (07-677; Millipore), anti-H3K37ac (61587; ActiveMotif), anti-H3K64ac (ActiveMotif 39545, a gift from Robert Schneider, Institut Génétique Biologie Moléculaire Cellulaire) and anti-H3K27ac (ab4729; Abcam).

ESC culture.

E14 ESCs41 were acquired from Babraham Institute (Cambridge, UK) and were not further authenticated. ESCs were cultured under standard conditions (10% CO2, 90% humidity, 37 °C) in knockout DMEM (10829-018; Life Technologies), 2 mM Glutamax (Gibco), 0.1 mM non-essential amino acids (Gibco), 15% ESC-grade fetal bovine serum (FBS) (Gibco), 0.1 mM β-mercaptoethanol, and leukemia inhibitory factor (ESG1107; Millipore). ESCs were routinely tested for mycoplasma.

Transient transfection.

Cells were transfected with TransIT-293 (HEK293) or TransIT-2020 (all other cell types) according to the manufacturer's protocol. Amino acid was added at transfection, and cells were incubated for 24–48 h.

Generation of stable cell lines.

Cells were transfected in a six-well plate with TransIT-293 (HEK293) or TransIT-2020 (all other cell types) with 2.5 μg of plasmid DNA, PiggyBac targeting plasmid(s) and Super PiggyBac Transposase Plasmid (SBI) at a ratio of 5:1. After 48 h, cells were split 1:6 into a six-well plate each, and selection antibiotic was added. Optimal antibiotic concentrations (5–10 μg mL−1 puromycin, 0.5–2 mg mL−1 G418) were experimentally determined. Cells were grown for at least 7 d under selection for polyclonal pools or sorted by flow cytometry into wells of single cells after 3–5 d of selection for clonal cell lines. Experiments were performed at early passages (<10 for polyclonal pools and <20 for clonal cell lines). Antibiotic resistance was periodically confirmed in long-term cultures.

Western blotting of nucleocytoplasmic extracts.

Cells were seeded in a 24-well plate (2 × 105 cells per well) and incubated with or without amino acid at the concentrations and times indicated in figures. Cells were washed with PBS and lysed by the addition of 100 μL of ice-cold RIPA buffer (Sigma) supplemented with Protease Inhibitor Cocktail (Roche). After 10 min of incubation on ice, lysates were cleared by centrifugation (10 min at 24,000 r.p.m. and 4 °C), 20 μL of 6× SDS sample buffer were added, and samples were incubated at 95 °C for 10 min. 10 μL were loaded per lane on a NuPAGE gel. Proteins were transferred using an iBlot onto a nitrocellulose membrane, blocked with 5% milk powder and a mixture of Tris-buffered saline and Tween-20 (TBST), and incubated with primary followed by secondary antibodies (goat anti-rabbit IgG (H+L)–horseradish peroxidase (HRP) conjugate, Bio-Rad, 1721019; goat anti-mouse IgG (H + L)-HRP conjugate, Bio-Rad, 1706516). Primary antibody was typically incubated overnight at 4 °C in 5% milk and TBST, whereas secondary antibody was incubated in TBST for 1 h at room temperature. HRP signal was visualized using Illuminata Forte (Millipore) substrate and a ChemiDoc digital imager (Bio-Rad). For acetylation-specific antibodies, milk was replaced with purified BSA (A9418; Sigma). Acetylation-specific antibodies were used at 1:500 to 1:5,000 dilutions with consistent results.

Flow cytometry analysis of live and fixed cells.

Cells were trypsinized and triturated to dissociate aggregates and taken up in full medium before centrifugation. Cells were centrifuged at 500g for 5 min, washed in PBS and 5% FBS, and resuspended in PBS for analysis. For fixation, cell suspension was treated with 1% paraformaldehyde for 10 min at room temperature, quenched with full medium, centrifuged (500g for 5 min) and resuspended in PBS for analysis. For immunofluorescent staining, cells were permeabilized in PBS with 5% FBS and 0.01% Triton X-100 and incubated with fluorescent dye–conjugated antibody for 1 h, after which they were washed three times with PBS, centrifuged (500g for 5 min) and resuspended in PBS for analysis.

Western blotting of chromatin extracts.

Cells were grown in a 24-well plate with or without amino acid at the concentrations and times indicated in figures. Cells were washed with PBS and lysed by the addition of 500 μL of ice-cold lysis buffer (1% Triton X-100 in PBS) supplemented with Protease Inhibitor Cocktail (Roche). After 10 min of incubation on ice, nuclei were pelleted by centrifugation (10 min at 10,000 r.p.m. and 4 °C) and resuspended well in 100 μL of 0.4 N H2SO4. The suspension was rocked at room temperature for 1 h, and insoluble material was removed by centrifugation (10 min at 24,000 r.p.m. and 4 °C). Supernatant was neutralized with 6× SDS sample buffer and 1 M Tris base. SDS-PAGE and western blotting were performed.

Analysis of nucleosomal histones.

Cells were trypsinized from two T125 tissue culture plates (2 × 107 cells) and washed twice with DMEM with 10% FBS and PBS. For isolating nuclei, cells were resuspended in ice-cold hypotonic buffer (10 mM HEPES, pH 7.5, 20 mM NaCl, 0.01% Triton X-100). Nuclei were pelleted, washed with hypotonic lysis buffer and pelleted again, and resuspended in digestion buffer (10 mM HEPES, pH 7.5, 20 mM NaCl, 0.01% Triton X-100). MNase was added, and nuclei were incubated at 37 °C for 10 min. Nuclei were lysed in RIPA buffer supplemented with 10 mM EDTA, and insoluble material was removed by centrifugation. Supernatant was used for immunoprecipitation of tagged histone H3 using anti-HA magnetic beads (Pierce; 88836). Digestion to mononucleosomes was confirmed by purification of a DNA sample and agarose gel electrophoresis.


RNA-seq was carried out from two biological replicates for each clone and condition, in two independent clones for ESCs and MEF cells. An RNEasy Kit (Qiagen) was used to isolate total RNA from one six-well plate per replicate. RNA was quality controlled using a Bioanalyzer (RNA integrity number (RIN) > 9). Library generation and sequencing were carried out using the Illumina Truseq Kit at Bejing Genomics Institute (BGI) Hong Kong. We used TopHat2 and/or Bowtie2 to align reads to the mouse genome (build mm10). Reads per kilobase of transcript per million mapped reads and differential expression values were calculated using CuffLinks and/or CuffDiff. Data analysis was performed using R and cummeRbund42. cummeRbund was used to confirm sample quality and comparable dispersion across samples. Gene ontology analysis was performed using tools at (ref. 32). Statistically significant changes (P < 0.005) were extracted from CuffDiff2 and were based on a Poisson model distribution corrected for both count uncertainty and count overdispersion43.


Total RNA was reverse transcribed using the Superscript III kit (Life Technologies) and random d(T)20. qPCR was carried out using Power SYBR Green master mix (Life Technologies) and IDT PrimeTime primers (Xist 3–4, Mm.PT.58.10791082; Xist 6–7, Mm.PT.58.43610994; Actb, Mm.PT.58.33540333).


ChIP was performed essentially as described by the Zhao lab44,45. ESCs were trypsinized, resuspended in full medium, pelleted and washed twice with ice-cold PBS before being flash-frozen. Frozen cells were resuspended in MNase digestion buffer, and the protocol was carried out as described using anti-HA magnetic beads (Pierce; 88836). ChIP-qPCR primer sequences are presented in Supplementary Table 1.

EB differentiation.

EB differentiation was performed according to a well-established protocol46. In short, ESCs were cultured on gelatin in the presence of LIF and then trypsinized and aggregated in hanging drops in ES medium without LIF for 2 d. EB aggregates were then cultured in suspension for 5 d, plated in a 24-well pate coated with 0.1% gelatin and maintained in ES medium without LIF.

Plasmid sequences.

pPiggyBac 4×U6-PylT(U25C)/EF1-MmPylS-IRES-Puro:


pPiggyBac 4×U6-PylT(U25C)/EF1-sfGFP(150TAG)-IRES-Neo:


Accession codes.

RNA-seq data are available at the NCBI Gene Expression Omnibus under accession number GSE73823.


Primary accessions

Gene Expression Omnibus


  1. 1.

    Expanding and reprogramming the genetic code of cells and animals. Annu. Rev. Biochem. 83, 379–408 (2014).

  2. 2.

    & Designer proteins: applications of genetic code expansion in cell biology. Nat. Rev. Mol. Cell Biol. 13, 168–182 (2012).

  3. 3.

    , , , & Selective, rapid and optically switchable regulation of protein function in live mammalian cells. Nat. Chem. 7, 554–561 (2015).

  4. 4.

    & Optical control of protein function through unnatural amino acid mutagenesis and other optogenetic approaches. ACS Chem. Biol. 9, 1398–1407 (2014).

  5. 5.

    & Cellular incorporation of unnatural amino acids and bioorthogonal labeling of proteins. Chem. Rev. 114, 4764–4806 (2014).

  6. 6.

    , & Light-activated kinases enable temporal dissection of signaling networks in living cells. J. Am. Chem. Soc. 133, 2124–2127 (2011).

  7. 7.

    , , , & Efficient viral delivery system for unnatural amino acid mutagenesis in mammalian cells. Proc. Natl. Acad. Sci. USA 110, 11803–11808 (2013).

  8. 8.

    , , & Efficient multisite unnatural amino acid incorporation in mammalian cells via optimized pyrrolysyl tRNA synthetase/tRNA expression and engineered eRF1. J. Am. Chem. Soc. 136, 15577–15583 (2014).

  9. 9.

    et al. Genetic incorporation of multiple unnatural amino acids into proteins in mammalian cells. Angew. Chem. Int. Edn. Engl. 52, 14080–14083 (2013).

  10. 10.

    et al. Genetically encoding unnatural amino acids in neural stem cells and optically reporting voltage-sensitive domain changes in differentiated neurons. Stem Cells 29, 1231–1240 (2011).

  11. 11.

    et al. Efficient transposition of the piggyBac (PB) transposon in mammalian cells and mice. Cell 122, 473–483 (2005).

  12. 12.

    , , , & SnapShot: histone modifications. Cell 159, 458–458 (2014).

  13. 13.

    , , , & Genetically encoding Nɛ-methyl-L-lysine in recombinant histones. J. Am. Chem. Soc. 131, 14194–14195 (2009).

  14. 14.

    , , & A genetically encoded ɛ-N-methyl lysine in mammalian cells. ChemBioChem 11, 1066–1068 (2010).

  15. 15.

    , , & Expanding the genetic code of yeast for incorporation of diverse unnatural amino acids via a pyrrolysyl-tRNA synthetase/tRNA pair. J. Am. Chem. Soc. 132, 14819–14824 (2010).

  16. 16.

    , & Synthesis of epsilon-N-propionyl-, epsilon-N-butyryl-, and epsilon-N-crotonyl-lysine containing histone H3 using the pyrrolysine system. Chem. Commun. (Camb.) 49, 379–381 (2013).

  17. 17.

    , , , & Site-specific incorporation of epsilon-N-crotonyllysine into histones. Angew. Chem. Int. Edn. Engl. 51, 7246–7249 (2012).

  18. 18.

    , & Genetically encoding Nɛ-acetyllysine in recombinant proteins. Nat. Chem. Biol. 4, 232–234 (2008).

  19. 19.

    & Histones: at the crossroads of peptide and protein chemistry. Chem. Rev. 115, 2296–2349 (2015).

  20. 20.

    et al. Regulation of transcription through acetylation of H3K122 on the lateral surface of the histone octamer. Cell 152, 859–872 (2013).

  21. 21.

    et al. A method for genetically installing site-specific acetylation in recombinant histones defines the effects of H3 K56 acetylation. Mol. Cell 36, 153–163 (2009).

  22. 22.

    et al. Acetylation of histone H3 at lysine 64 regulates nucleosome dynamics and facilitates transcription. eLife 3, e01632 (2014).

  23. 23.

    et al. Accelerated chromatin biochemistry using DNA-barcoded nucleosome libraries. Nat. Methods 11, 834–840 (2014).

  24. 24.

    & Histone modification: cause or cog? Trends Genet. 27, 389–396 (2011).

  25. 25.

    et al. Lysine acetylation targets protein complexes and co-regulates major cellular functions. Science 325, 834–840 (2009).

  26. 26.

    et al. Regulation of cellular metabolism by protein lysine acetylation. Science 327, 1000–1004 (2010).

  27. 27.

    Dynamics and memory of heterochromatin in living cells. Cell 149, 1447–1460 (2012).

  28. 28.

    et al. Proteome labeling and protein identification in specific tissues and at specific developmental stages in an animal. Nat. Biotechnol. 32, 465–472 (2014).

  29. 29.

    Embryonic stem cell differentiation: emergence of a new era in biology and medicine. Genes Dev. 19, 1129–1155 (2005).

  30. 30.

    et al. Microfluidic single-cell whole-transcriptome sequencing. Proc. Natl. Acad. Sci. USA 111, 7048–7053 (2014).

  31. 31.

    & The transcriptional and signalling networks of pluripotency. Nat. Cell Biol. 13, 490–496 (2011).

  32. 32.

    et al. Gene ontology: tool for the unification of biology. Nat. Genet. 25, 25–29 (2000).

  33. 33.

    et al. Genetic code expansion enables live-cell and super-resolution imaging of site-specifically labeled cellular proteins. J. Am. Chem. Soc. 137, 4602–4605 (2015).

  34. 34.

    , & Translational readthrough potential of natural termination codons in eucaryotes—the impact of RNA sequence. RNA Biol. 12, 950–958 (2015).

  35. 35.

    , & New functions for an old variant: no substitute for histone H3.3. Curr. Opin. Genet. Dev. 20, 110–117 (2010).

  36. 36.

    , , & Chemical tagging and customizing of cellular chromatin states using ultrafast trans-splicing inteins. Nat. Chem. 7, 394–402 (2015).

  37. 37.

    , , & CBP/p300-mediated acetylation of histone H3 on lysine 56. Nature 459, 113–117 (2009).

  38. 38.

    et al. Histone deacetylase inhibitors globally enhance h3/h4 tail acetylation without affecting h3 lysine 56 acetylation. Sci. Rep. 2, 220 (2012).

  39. 39.

    , , & Acetylated histone H3K56 interacts with Oct4 to promote mouse embryonic stem cell pluripotency. Proc. Natl. Acad. Sci. USA 110, 11493–11498 (2013).

  40. 40.

    , , & Silencing of the mammalian X chromosome. Annu. Rev. Genomics Hum. Genet. 6, 69–92 (2005).

  41. 41.

    , , , & HPRT-deficient (Lesch-Nyhan) mouse embryos derived from germline colonization by cultured cells. Nature 326, 292–295 (1987).

  42. 42.

    et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat. Protoc. 7, 562–578 (2012).

  43. 43.

    et al. Differential analysis of gene regulation at transcript resolution with RNA-seq. Nat. Biotechnol. 31, 46–53 (2013).

  44. 44.

    et al. Native chromatin preparation and Illumina/Solexa library construction. Cold Spring Harb. Protoc. 2009 Pdb.prot5237 (2009).

  45. 45.

    et al. Global analysis of the insulator binding protein CTCF in chromatin barrier regions reveals demarcation of active and repressive domains. Genome Res. 19, 24–32 (2009).

  46. 46.

    & In vitro differentiation of mouse embryonic stem (mES) cells using the hanging drop method. J. Vis. Exp. 17, e825 (2008).

Download references


This work was supported by the UK Medical Research Council (U105181009 and UPA0241008 to J.W.C.). S.J.E. was supported by a European Molecular Biology Organization Long-Term Fellowship (ALTF 1232-2011) and the Herchel Smith Fund. O.S.W. was funded through the PhD program in Chemical Biology and Molecular Medicine at the University of Cambridge. We are grateful to T. Elliott (Medical Research Council Laboratory of Molecular Biology (MRC-LMB), Cambridge, UK) for providing CpK, to R. Schneider (Institut Génétique Biologie Moléculaire Cellulaire, Illkirch, France) for providing anti-H3K64ac (ActiveMotif 39545), to the MRC-LMB FACS facility (M. Daly, F. Zhang and V. Romashova) for support, and to J. Sale for helpful comments on the manuscript.

Author information


  1. Medical Research Council Laboratory of Molecular Biology, Cambridge, UK.

    • Simon J Elsässer
    • , Russell J Ernst
    • , Olivia S Walker
    •  & Jason W Chin
  2. Department of Chemistry, Cambridge University, Cambridge, UK.

    • Simon J Elsässer
    • , Olivia S Walker
    •  & Jason W Chin
  3. Department of Medical Biochemistry and Biophysics, Karolinska Institutet, Stockholm, Sweden.

    • Simon J Elsässer


  1. Search for Simon J Elsässer in:

  2. Search for Russell J Ernst in:

  3. Search for Olivia S Walker in:

  4. Search for Jason W Chin in:


S.J.E. and J.W.C. conceived the experimental strategy, analyzed the data and wrote the paper. S.J.E. performed most experiments and analyzed and interpreted the RNA-seq data. O.S.W. and R.J.E. performed qPCR experiments and analysis.

Competing interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to Jason W Chin.

Integrated supplementary information

Supplementary information

PDF files

  1. 1.

    Supplementary Text and Figures

    Supplementary Figures 1–13 and Supplementary Table 1


  1. 1.

    Beating cardiomyocyte aggregates form in the presence of CpK

    Beating cardiomyocyte aggregates differentiated from mouse ESC via embryoid body formation in the presence of 0.2 mM CpK.

About this article

Publication history





Further reading