Biosynthesis and Genetic Incorporation of 3,4-Dihydroxy-L-Phenylalanine into Proteins in Escherichia coli

Abstract

While 20 canonical amino acids are used by most organisms for protein synthesis, the creation of cells that can use noncanonical amino acids (ncAAs) as additional protein building blocks holds great promise for preparing novel medicines and for studying complex questions in biological systems. However, only a small number of biosynthetic pathways for ncAAs have been reported to date, greatly restricting our ability to generate cells with ncAA building blocks. In this study, we report the creation of a completely autonomous bacterium that utilizes 3,4-dihydroxy-L-phenylalanine (DOPA) as its 21^st amino acid building block. Like canonical amino acids, DOPA can be biosynthesized without exogenous addition and can be genetically incorporated into proteins in a site-specific manner. Equally important, the protein production yield of DOPA-containing proteins from these autonomous cells is greater than that of cells exogenously fed with 9 mM DOPA. The unique catechol moiety of DOPA can be used as a versatile handle for site-specific protein functionalizations via either oxidative coupling or strain-promoted oxidation-controlled cyclooctyne-1,2-quinone (SPOCQ) cycloaddition reactions. We further demonstrate the use of these autonomous cells in preparing fluorophore-labeled anti-human epidermal growth factor 2 (HER2) antibodies for the detection of HER2 expression on cancer cells.

Graphical Abstract

Introduction

Genetic Code Expansion technology enables the site-specific incorporation of noncanonical amino acids (ncAAs) into proteins in living cells, a methodology that has proven to be a powerful tool for investigating and manipulating protein structure and function.[1–9] This technology involves the generation of bioorthogonal translational machinery consisting of (a) an evolved aminoacyl-tRNA synthetase (aaRS) and its corresponding tRNA, (b) a “blank” (normally noncoding) codon, and (c) a high concentration of intracellular ncAA. Innovative improvements in the efficiency of components involved in bioorthogonal translation and the expansion of codons available for ncAAs, now enable the incorporation of more than 300 ncAAs in prokaryotic and eukaryotic cells.[10–22] However, current methods for achieving high intracellular ncAA concentrations mainly rely on the exogenous addition of chemically synthesized ncAA to the culture medium. Most of these ncAAs are not commercially available, and chemical syntheses of these ncAAs often involve multistep organic preparations with time-consuming purification steps and low yield.[23–25] Furthermore, the membrane penetration ability of ncAAs with charged, highly hydrophobic, or hydrophilic structures can be poor, greatly limiting the efficiency of genetic code expansion technology.[24,26–28] Thus, developing cells with the autonomous ability to synthesize ncAAs by itself is essential for generating cells and organisms with improved genetic code expansion.[29–33]

3,4-dihydroxy-L-phenylalanine (DOPA), generated via post-translational modification, is a key amino acid component of many protein biomaterials, including mussel foot proteins, squid beak polymer, and others.[34–37] Taking advantage of the unique adhesion property of DOPA, mussels can resist a force of up to 400 N in the wave-swept environment of the ocean.[38] Similarly, the cross-linking property of DOPA increases the mechanical strength of squid beak material.[39] To incorporate this unique reactive catechol group at desired sites in proteins, several groups have reported the evolution of translational machinery for site-specific incorporation of DOPA into bacterial proteins using Genetic Code Expansion.[40–44] The utility of DOPA-containing proteins has been demonstrated with engineered redox proteins, fluorophore-labeled proteins, and crosslinked proteins.[40,41,43,45–51] However, all these methods rely on the exogenous addition of chemically synthesized DOPA to the culture medium. The efficiency of genetic DOPA incorporation is greatly limited by this exogenous feeding approach since high concentrations of DOPA in the cell culture are reported to be toxic.[40,52] To overcome this limitation, Lee and others reported the biosynthesis of DOPA from catechol, pyruvate, and ammonia in Escherichia coli.[41] As an attempt to bypass the requirement of exogenously adding these chemical precursors, Thyer et. al introduced an oxygenase-based DOPA biosynthetic pathway to bacterial cells to yield biosynthetic DOPA for the Genetic Corde Expansion.[43] However, a significant level of misincorporation of tyrosine was observed using this system.[43] Therefore, there is an unmet need for the development of autonomous cells that can produce homogenous DOPA-containing proteins.

In this study, we report the generation of a completely autonomous E. coli strain that can biosynthesize DOPA from tyrosine and genetically incorporate DOPA into proteins. We initially compared several reported aaRS/tRNA pairs for DOPA incorporation, one of which led to the production of pure DOPA-containing proteins without the misincorporation of tyrosine. With this efficient and bioorthogonal DOPA incorporation system, we found that mouse tyrosine hydroxylase together with tetrahydromonapterin (MH4) cofactor recycling enzymes was capable of biosynthesizing large quantities of DOPA for genetic incorporation into proteins. (Fig. 1) To our delight, the yield of DOPA-containing proteins from autonomous cells is greater than that from control cells fed exogenously with 9 mM DOPA. We used this autonomous system to produce DOPA-containing anti-HER2 single-chain fragment variable (ScFv). The utility of these fragments was demonstrated by site-specific functionalization using either oxidative coupling or strain-promoted oxidation-controlled cyclooctyne-1,2-quinone (SPOCQ) cycloaddition chemistry.

Fig. 1: Generation of a completely autonomous E.coli strain with DOPA as its 21st Amino Acid.

Mouse hydroxylase efficiently converts tyrosine to DOPA in the presence of MH4 recycling pathway and molecular oxygen. The biosynthesized DOPA is then site-specifically incorporated into the anti-HER2 antibody using an orthogonal chPheRS-2/tRNA pair. The resulting anti-HER2-DOPA antibody can be site-specifically functionalized by either oxidative coupling or SPOCQ cycloaddition reaction. DOPA: L-3,4-dihydroxyphenylalanine; MH4: tetrahydromonapterin; DHMR: dihydromonapterin reductase; PCD: pterin-4α-carbinolamine dehydratase; aaRS: aminoacyl-tRNA synthetase; HER2: human epidermal growth factor 2.

Results

DOPA was first genetically incorporated into proteins by the Schultz Group in 2003 using an engineered Methanococcus jannaschii TyrRS/tRNA^Tyr (DOPARS).[40] Although the DOPARS was obtained from multiple rounds of positive and negative selection, mass spectrometry analysis revealed significant Tyr incorporation in addition to that of DOPA. Besides DOPARS, a chimeric pair of pyrrolysyl-tRNA synthetase/tRNA pair and PheRS/tRNA from human mitochondria was also reported capable of genetically incorporating DOPA into proteins.[42] To investigate which reported aaRS/tRNA suppression plasmid pair was capable of consistent and efficient incorporation of DOPA, we first cloned DOPARS and chPheRS into separate pUltra vectors, in which aaRS and tRNA expression are driven by the trc and proK promotor, respectively. [53] DOPA incorporation efficiency was evaluated using a fluorescence assay with a superfolder green fluorescent protein (sfGFP) mutant carrying an amber codon at position 134, designated as pLei-sfGFP-D134TAG. Bacterial cells harboring pLei-sfGFP-D134TAG and either of the two pUltra plasmids were grown in M9G medium containing different concentrations of DOPA addition. Compared to expression without exogenous DOPA, improved green fluorescence was observed in the presence of DOPA for cells expressing either DOPARS or chPheRS (Fig. 2A). To investigate the specificity of these two synthetases for DOPA incorporation, sfGFP proteins were purified by Ni²⁺-NTA affinity chromatography and characterized using ESI-MS. Multiple peaks indicative of both DOPA and tyrosine incorporation were observed for sfGFP proteins purified from cells expressing DOPARS (Fig. 2B and 1–2). In contrast, proteins from cells expressing chPheRS exhibited only a single peak of 27960 Da in agreement with the expected mass of sfGFP-D134-DOPA. This establishes that chPheRS has superior specificity for DOPA incorporation (Fig. 2C and S1–2), and accordingly this construct was used for further study. To explore the effect of sfGFP protein expression level on DOPA incorporation efficiency, we screened two additional reporter plasmids encoding sfGFP with an in-frame amber codon driven by either a T5 promoter (pET22b-T5-sfGFP*) or a T7 promoter (pET28a-T7-sfGFP*).[54] Among all three reporter plasmids, pET22b-T5-sfGFP* exhibited the highest fold increase in fluorescence in the presence of 9 mM exogenous DOPA, compared to cells without DOPA feeding (Fig. S3). Thus, we used pET22b-T5-sfGFP* as the reporter plasmid for further studies.

Fig. 2: Construction of a completely autonomous E.coli with DOPA by introducing orthogonal translational machinery and DOPA biosynthetic pathway.

(A) Comparison of the efficiency of DOPARS/tRNA pair and chPheRS/tRNA pair to produce DOPA-containing sfGFP. (B) ESI-MS analysis and sfGFP purified from cells harboring DOPARS/tRNA pair. (C) ESI-MS analysis and sfGFP purified from cells harboring chPheRS/tRNA pair. (D) Comparison of the efficiency of different hydroxylases to produce DOPA-containing sfGFP. (E) Time-dependent expression of sfGFP151DOPA in the autonomous cells. (F) Optimization of sfGFP151DOPA production in the autonomous cells by screening different concentrations of Vitamin C. (G) ESI-MS analysis of sfGFP purified from control cells with external 9 mM DOPA addition. (H) ESI-MS analysis of sfGFP purified from the autonomous cells without external DOPA addition. (I) SDS-PAGE analysis of sfGFP and anti-HER2 ScFv antibody expressed in M9G in the presence (+) or absence (−) of 9 mM DOPA, or when inducing the DOPA biosynthetic pathway.

Aromatic compounds can undergo enzymatic hydroxylation in the presence of molecular oxygen and reduced cofactor.[55–58] To identify hydroxylases suitable for DOPA biosynthesis from tyrosine, we tested a variety of aromatic amino acid hydroxylases from different species, including Mus musculus (MmH), Xanthomonas campestris (XcH), Mesorhizobium ciceri (McH), Schistosoma mansoni (SmH), and Oryctolagus cuniculus (OcH), using the fluorescence assay with a superfolder green fluorescent protein. In addition to the hydroxylase being tested, we also introduced an artificial MH4 recycling pathway consisting of E. coli dihydromonapterin reductase (DHMR) and the Pseudomonas aeruginosa pterin-4α-carbinolamine dehydratase (PCD), because all hydroxylases are known to use MH4 as their cofactor.[59] We first subcloned the hydroxylase and MH4 recycling pathway into pET22b-T5-sfGFP*, which was designated as pET22b-T5-sfGFP*-XH-MH4R. Next, the suppression plasmid pUltra-chPheRS was transformed into bacterial cells harboring pET22b-T5-sfGFP*-XH-MH4R plasmids expressing the various hydroxylases. Protein expression in M9G medium was examined after 16 h, in parallel with expression by control cells containing pET22b-T5-sfGFP* and pUltra-chPheRS. As expected, control cells lacking hydroxylase only exhibited strong sfGFP fluorescence in the presence of 9 mM exogenous DOPA. Among all the hydroxylases tested, we found that cells expressing mouse hydroxylase (MmH) yielded the highest fluorescence signal. (Fig. 2D) To our delight, the sfGFP fluorescence from cells biosynthetically producing DOPA was even higher than that of control cells fed exogenously with 9 mM DOPA. This observation is consistent with the previous report that mouse tyrosine hydroxylase (MmH) could recognize tyrosine for DOPA biosynthesis while the tryptophan (SmH, OcH) and phenylalanie hydroxylase (XcH, OcH) prefer their native subtrates. [57–60]

To further test the influence of protein expression time and medium redox level on the production of DOPA-containing proteins in these autonomous cells, we quantified protein expression under different expression time and Vitamin C concentrations. Vitamin C was previously added to prevent DOPA oxidation.[61] We found 1 mg/mL Vitamin C and 16 hour expression time to be optimal. (Fig. 2E–F) To further confirm the yield and purity of sfGFP-DOPA produced from autonomous cells, sfGFP proteins were purified by Ni²⁺-NTA affinity chromatography and characterized with ESI-MS and SDS-PAGE. We observed a single peak of 27612 Da from sfGFP-DOPA purified from control cells with 9 mM DOPA addition, which corresponds to the mass of sfGFP-DOPA without the first methionine residue. (Fig. 2G and S1–2). The mass of sfGFP-DOPA isolated from autonomous cells was 27747 Da, which is in good agreement with the calculated mass of 27744 Da. (Fig. 2H and S1–2) Compared to a wildtype sfGFP protein yield (90 mg/L) and sfGFP-DOPA (2.5 mg/L) from control cells fed with 9 mM DOPA, 3.1 mg of the sfGFP protein were produced from 1 liter of autonomous cells. (Fig. 2I) Taken altogether, these data demonstrate that we have created a bacterium possessing both an efficient DOPA biosynthetic pathway and machinery for bioorthogonal genetic incorporation of DOPA.

Having an electron-rich aromatic ring, DOPA can be oxidized to L-dopaquinone, which has been demonstrated to be an important partner for both oxidative coupling and strain-promoted oxidation-controlled cyclooctyne-1,2-quinone (SPOCQ) cycloaddition reactions.[62,63] Using tyrosinase, the Francis group was able to activate either N- or C-terminal tyrosine residues to o-quinone and then label the resulting electrophile with a variety of nucleophiles, including amines, anilines, and thiols.[64] Proteins containing a site-specific DOPA, isolated from the autonomous cells described above, should provide a flexible handle for this type of oxidative coupling chemistry. (Fig. 3A) To test this hypothesis, a single-chain fragment variable (ScFv) of the trastuzumab antibody was used as a model protein. Trastuzumab targets the human epidermal growth factor 2 (HER2) receptor that is highly expressed on a variety of cancer cells, and has been approved by the FDA for treatment of cancer.[65] The ScFv fragment of Trastuzumab is of special interest because of its enhanced tissue and tumor penetration and its ease of preparation from bacterial cells.[66,67] To determine if DOPA-containing ScFv can be prepared using DOPA autonomous cells, the ScFv sequence with an amber stop codon at position 113 was substituted in place of the sfGFP gene sequence within pET22b-T5-sfGFP*-MmH-MH4R, yielding pET22b-T5-ScFv*-MmH-MH4R.[30] Cells transformed with both pET22b-T5-ScFv*-MmH-MH4R and pUltra-chPheRS were used for protein expression in M9G medium, in parallel with control cells harboring pET22b-T5-ScFv* and pUltra-chPhe. The resulting DOPA-containing ScFv* (anti-HER2-DOPA) species were purified by Ni²⁺-NTA affinity chromatography and characterized by ESI-MS and SDS-PAGE. The mass of anti-HER2-DOPA isolated from autonomous cells is 27577 Da, which matches the calculated mass of 27576 Da (Fig. 3B and S1–2). Comparing with wildtype protein yield (50 mg/L), the yields of anti-HER2-DOPA isolated from autonomous cells and controls cells fed with 9 mM DOPA were 0.8 mg/L and 0.7 mg/L, respectively. With anti-HER2-DOPA in hand, we reacted it for 1 hour with p-aminophenylalanine (pAF) in pH 6.5 PBS buffer in the presence of K₃Fe(CN)₆. ESI-MS analysis confirmed an efficient product formation with an observed mass of 27753 Da, which agrees with the expected mass of 27753 Da. (Fig. 3C and S4) The same reaction carried out with WT ScFv yielded no product, demonstrating the bioorthogonality of the reaction (Fig. S5). In addition to oxidative coupling chemistry, DOPA also exhibits great potential for strain-promoted chemistry (Fig. 3A). The Delft group reported that the cycloaddition between quinone and strained cyclooctyne has a fast reaction rate, which they termed it as strain-promoted oxidation-controlled cyclooctyne-1,2-quinone (SPOCQ) cycloaddition.[63] To test whether anti-HER2-DOPA could be used for the SPOCQ cycloaddition, we incubated anti-HER2-DOPA with bicyclononyne (BCN)-PEG3-Amine for 15 minutes in pH 8 PBS in the presence of K₃Fe(CN)₆. ESI-MS analysis of the reaction product revealed a mass of 27943 Da in perfect agreement with the expected mass of 27943 Da. (Fig. 3D and S4) Biosynthesized DOPA incorporated into proteins using autonomous cells provides a versatile handle for both oxidative coupling reaction and SPOCQ cycloadditions. After confirming the reactivity of anti-HER2-DOPA from the autonomous cells, we further used this protein for the preparation of fluorophore-labeled anti-HER2 antibody and examined its ability to detect HER2-positive cells. (Fig. 3E) Oxidative coupling chemistry was selected for this application due to the ease of synthesizing its coupling partner. The fluorophore, 4-dimethylaminophthalimide, was coupled to an aniline moiety via an amide bond formation to yield aniline-DMAP. The anti-HER2-DOPA was first reacted for 1 hour with aniline-DMAP in pH 6.5 PBS buffer in the presence of K₃Fe(CN)₆, followed by purification on a desalting column. As expected, we could observe both anti-HER2-DOPA and anti-HER2-DMAP on SDS-PAGE gel after coomassie statining, but a blue fluorescent band was only found for anti-HER2-DMAP on SDS-PAGE gel before coomassie statining. (Fig. 3F) To confirm the product formation, we also conducted ESI-MS analysis of the product. The observed mass was 27939 Da, which corresponds to the expected mass of 27938 Da (anti-HER2-DMAP). To verify the preservation of its biological activity after this chemical transformation, we incubated the anti-HER2-DMAP with HER2-positive SK-BR-3 cells and HER2-negative MDA-MB-468 cells for 1 hour before imaging. Confocal fluorescence imaging confirmed that the green fluorescence signal of DMAP was only observed with HER2-positive SK-BR-3 cells, and not with HER2-negative MDA-MB-468 cells. These results demonstrate that the functionalized anti-HER2-DOPA still retains its biological activity. (Fig. 3H)

Fig. 3: Site-specific modifications of anti-HER2-DOPA produced from the autonomous cells with two bioorthogonal reactions.

(A) Site-specific functionalizations of anti-HER2-DOPA using oxidative coupling and strain-promoted oxidation-controlled cyclooctyne-1,2-quinone (SPOCQ) cycloaddition reactions. (B-D) ESI-MS analysis of anti-HER2-DOPA before reaction (B) and after reaction with BCN-PEG3-amine (C), and pAF (D). (E) Preparation of fluorophore-labeled anti-HER2-DOPA using the oxidative coupling reaction. (F) ESI-MS analysis of anti-HER2-DMAP. (G) SDS-PAGE analysis of anti-HER2-DOPA and anti-HER2-DMAP, visualized by coomassie staining (left) and fluorescent imaging (right). (H) Binding of anti-HER2-DMAP in SK-BR-3 and MD-MBA-468 cells visualized by confocal microscopy. Cells were incubated with 100 nM anti-HER2-DMAP before imaging. Scale bar represents 10 μm.

In conclusion, we have generated a completely autonomous bacterial strain that utilizes DOPA as its 21^st amino acid for protein synthesis. This engineered strain can biosynthesize DOPA from tyrosine without exogenous addition of precursors, and genetically incorporate DOPA into proteins with high efficiency and fidelity. The production of DOPA-containing proteins from these autonomous cells is more efficient than that from cells exogenously fed with 9 mM DOPA. The unique catechol moiety of DOPA enables us to site-specifically modify DOPA-containing proteins produced from the autonomous cells using either oxidative coupling or SPOCQ cycloaddition reactions. The utility of this strain was further explored by preparing fluorophore-conjugated HER2-ScFv and using this probe to detect HER2 expression on breast cancer cells. The generation of cells with DOPA as a 21^st amino acid will provide powerful tools for the development of novel therapeutic protein conjugates, and will also promote the evolution of proteins with novel material properties.

Material and methods

Plasmids Construction

pUltra-DOPARS is a gift from Dr. Peter G. Schultz. To obtain pUltra-chPheRS, chPheRS-2 reported by Lin Group was amplified with Da607 and Da608 and assembled to NotI-digested pUtra-polyRS, which yields pUltra-chPheRS-MjYtRNA.[42] The tRNA of pUltra-chPheRS-MjYtRNA was changed to chPheT reported by Lin Group by gibson assembling Da615 and Da616-amplified pUltra-chPheRS-MjYtRNA and Da613 and Da614 amplified chPheT, which generates pUltra-chPheRS. To make pET22b-T5-sfGFP*-MmH-MH4R, MmH-MH4R cassette was amplified from pBbE1K-3 (addgene# 71722) with Da150 and 205 and assembled into pET22b-T5-sfGFP* vector amplified with Da156 and 204. The other pET22b-T5-sfGFP*-XH-MH4R containing hydroxylases from different species are from our previous report.[32] To generate pET22b-T5-ScFv*-MmH-MH4R, ScFv* was cloned from pBad-ScFv113* by Da218 and Da219 and assembled into the pET22b-T5-sfGFP*-MmH-MH4R vector amplified by Da216 and Da217. [30] Sequences of all DNA oligos were summarized in table S1 of supporting information. Sequences and NCBI reference numbers of all hydroxylases were included in table S2 of supporting information.

Expression and Purification of Proteins.

E. coli DH10B cells, co-transformed with pUltra-chPheRS/pUltra-DOPARS and pLei-sfGFP-D134TAG were grown in 2YT medium at 37°C. The protein expression was carried out in M9G. When the OD600 of the cell culture reached 0.6, protein expression was induced by the addition of IPTG to a final concentration of 1 mM. At the same time, indicated concentration of DOPA was added and 0.5 mg/mL Vitamin C was supplied to prevent oxidation of DOPA. After growth for an additional 16 h at 30 °C, cells were harvested by centrifugation at 4,750 × g for 10 minutes and used for protein purification and measurement of normalized fluorescence. Protein purification was carried out with Ni²⁺-NTA resin (Qiagen) following the manufacturer’s instructions. The purified proteins were then analyzed by ESI-MS.

E. coli BL21(DE3) cells, co-transformed with pUltra-chPheRS and pLei-sfGFP-D134TAG/pET22b-T5-sfGFP*/pET28a-T7-sfGFP* were grown in 2YT medium at 37°C. The protein expression was carried out in M9G. When the OD600 of the cell culture reached 0.6, protein expression was induced by the addition of IPTG to a final concentration of 1 mM. At the same time, indicated concentration of DOPA and 0.5 mg/mL Vitamin C were added. After growth for an additional 16 h at 30 °C, cells were harvested by centrifugation at 4,750 × g for 10 minutes and used for measurement of normalized fluorescence.

E. coli DH10B cells, co-transformed with pUltra-chPheRS and pET22b-T5-sfGFP* were grown in 2YT medium at 37°C as control cells. To identify the hydroxylase is most efficient for the biosynthesis of DOPA, pUltra-chPheRS was transformed with one of the pET22b-T5-sfGFP*-XH-MH4R (XH represents MmH, XcH, McH, SmH, or OcH). The protein expression was carried out in M9G. When the OD600 of the cell culture reached 0.6, protein expression was induced by the addition of IPTG to a final concentration of 1 mM and 0.5 mg/mL Vitamin C was added to prevent the oxidation of DOPA. At the same time, 9 mM DOPA were added to controls cells without the DOPA biosynthetic pathway. After growth for another 16 hours at 30 °C, cells were harvested by centrifugation at 4,750 × g for 10 minutes and used for protein purification and measurement of normalized fluorescence. Proteins purification was carried out with Ni²⁺-NTA resin (Qiagen) following the manufacturer’s instructions. The purified proteins were then analyzed by ESI-MS and SDS-PAGE. The optimization of DOPA biosynthesis was done following the same protocol except varying the amount of Vitamin C and the expression time.

E. coli BL21(DE3) cells, co-transformed with pUltra-chPheRS and pET22b-T5-ScFv* were grown in 2YT medium at 37°C as control cells. To produce anti-HER2-DOPA with biosynthesized DOPA, pUltra-chPheRS and pET22b-T5-ScFv*-MmH-MH4R were transformed into E. coli BL21(DE3) cells. Protein expression was carried out in M9G. When the OD600 of the cell culture reached 0.6, protein expression was induced by the addition of IPTG to a final concentration of 1 mM and 0.5 mg/mL Vitamin C was added to prevent the oxidation of DOPA. At the same time, 9 mM DOPA were added to controls cells without the DOPA biosynthetic pathway. Cells were grown for another 16 hours at 30 °C. Cells were harvested by centrifugation at 4,750 × g for 10 minutes and proteins were purified on Ni-NTA resin (Qiagen) following the manufacturer’s instructions. The purified proteins were then analyzed by SDS-PAGE and ESI-MS.

Expression and Fluorescence Measurement of sfGFP

After sfGFP expression with the method described above, 1 mL cells were harvested by centrifugation at 4,750 × g for 10 minutes and then washed with 1 mL PBS (pH 7.4) for 3 times. The resulting cell pellet was resuspended in 1 ml PBS (pH 7.4). Cell fluorescence was measured using excitation/emission wavelengths of 475/512 nm. The error bars in Fig. 2 and S1 represent for the standard deviations.

Bioorthogonal reactions of anti-HER2-DOPA

The anti-HER2-DOPA antibody (10 μM) dissolved in PBS (pH 6.5) was mixed with 500 μM pAF and 2 mM K₃Fe(CN)₆ at room temperature for 1 hour. The resulting product was analyzed by ESI-MS. The anti-HER2-DOPA antibody (5 μM) dissolved in PBS (pH 6.5) was mixed with 10 μM K₃Fe(CN)₆ at room temperature for 5 minutes, followed by the addition of 500 μM aniline-DMAP. After 1 hour reaction, the resulting product was desalted with Amicon 3K protein concentrator and analyzed by ESI-MS snd SDS-PAGE. The anti-HER2-DOPA (5 μM) dissolved in PBS (pH 8) was mixed with 10 μM K₃Fe(CN)₆ at room temperature for 5 minutes, followed by the addition of 1000 μM BCN-PEG3-amine. After reaction at room temperature for 15 minutes, the resulting product was analyzed by ESI-MS.

Fluorescence Imaging

Confocal fluorescent imaging of living cells was performed using a Zeiss LSM710 confocal microscopy. SK-BR-3 cells and MDA-MB-468 cells were cultured in complete medium (RPMI 1640 Medium or Dulbecco’s modified Eagle’s Medium, respectively, supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin) at 37°C in atmosphere containing 5% CO₂. SK-BR-3 cells and MDA-MB-468 cells were grown to about 80% confluency in 8-well confocal imaging plate. Cells were incubated with 100 nM anti-HER2-DMAP for 1 hour at 37 °C. Cells were then washed with PBS (pH 7.4) for three times. Cells were fixed with 4% PFA in PBS for 15 minutes at room temperature, followed by washing with PBS (pH 7.4) for 3 times. Nucleus staining was performed right before imaging by incubating cells with 10 μg/mL Hoechst for 10 min. After washing cells with PBS (pH 7.4) for 3 times, confocal imaging was carried out.

Mass Spectra Methods For Proteins

A single quadrupole mass spectrometer (Agilent: G7129A) coupled with 1260 infinity II Quaternary Pump (Agilent: G7111B) was used for all the protein samples with Pursuit 3 Diphenyl (100 × 2.0 mm) column. Water with 0.1% formic acid and ACN with 0.1% formic acid were the organic and aqueous mobile phases, respectively. Flow gradient was initially set at 5% ACN, 15% ACN at 0.1 min, 55% ACN at 4.5 min and then back to 10% ACN at 5 min. Spectra were deconvoluted using the Maximum Entropy deconvolution algorithm in the BioConfirm software.

Highlights:

1. A system for site-specific DOPA incorporation with high efficiency and fidelity was reported.

2. DOPA biosynthetic pathway was selected to produce DOPA-containing protein.

3. Incorporation efficiency of biosynthetic DOPA into proteins in cells was more than that with external DOPA feeding.

4. DOPA-containing proteins were used for protein conjugate using two bioorthogonal chemistries.

5. We created a bacterial strain with DOPA as its 21st amino acid building block.