Genetic code expansion is applied for the introduction of an unnatural amino acid bearing a biorthogonal functional group on a carrier protein at a defined site. The biorthogonal function is further used for the site-selective coupling of a carbohydrate antigen to provide a homogeneous glycoconjugate vaccine.
Genetic code expansion is a powerful tool to introduce unnatural amino acids (UAAs) into proteins to modify their characteristics, to study or create new protein functions or to have access to protein conjugates. Stop codon suppression, in particular amber codon suppression, has emerged as the most popular method to genetically introduce UAAs at defined positions. This methodology is herein applied to the preparation of a carrier protein containing an UAA harboring a bioorthogonal functional group. This reactive handle can next be used to specifically and efficiently graft a synthetic oligosaccharide hapten to provide a homogeneous glycoconjugate vaccine. The protocol is limited to the synthesis of glycoconjugates in a 1:1 carbohydrate hapten/carrier protein ratio but amenable to numerous pairs of biorthogonal functional groups. Glycococonjugate vaccine homogeneity is an important criterion to ensure complete physico-chemical characterization, thereby, satisfying more and more demanding drug regulatory agency recommendations, a criterion which is unmet by classical conjugation strategies. Moreover, this protocol makes it possible to finely tune the structure of the actual conjugate vaccine, giving rise to tools to address structure-immunogenicity relationships.
Glycoconjugate vaccines are essential elements of the vaccine arsenal available for the prophylactic treatment of infectious diseases. They are safe, well-tolerated and efficient in a broad age group including young infants. They provide the optimal defense against infections caused by capsulated bacteria like meningococcus, pneumococcus or Haemophilus influenzae type b1. Glycococonjugate vaccines are made of purified bacterial polysaccharides that form the capsules of bacteria or synthetic oligosaccharides that mimic these surface-expressed polysaccharides2, which are covalently linked to a carrier protein. The presence of a carrier protein is essential to promote protective humoral immune responses directed against the antigenic determinant expressed by the carbohydrate antigens3. Apart from a careful selection and production of the carbohydrate antigen, the features known to exert an influence on the efficacy of a glycoconjugate vaccine are: the nature of the carrier protein, the conjugation chemistry (including the nature and the length of the linker if used), or the saccharide/protein ratio3. Obviously, the positions at which the saccharide is conjugated to the protein as well as the number of connectivity points are relevant for immunogenicity. To date, these two parameters have hardly been studied because the preparation of the glycoconjugates remains largely empirical. Their synthesis usually relies on the use of amine or carboxylic acid functions of, respectively, lysine or aspartic/glutamic acid side-chain residues present on the carrier protein sequence. This leads not to a single but to a heterogeneous mixture of glycoconjugates.
Playing on the reactivity, accessibility or distribution of the amino acid residues in the protein gives rise to more defined glycoconjugates that are more reliable to document the effect of saccharide/protein connectivity4. A step forward towards this goal can be achieved by applying protein glycan coupling technology, a recombinant process that allows the production of controlled glycoconjugate vaccines in cell factories5,6. However, the glycosylation exclusively takes place at an asparagine residue within D/EXNYS/T sequons (whereby X is any out of the 20 natural amino acids), not naturally present on the carrier proteins.
Site selective mutagenesis and in particular incorporation of cysteines to exploit their highly and selective reactivity appears as an alternative7,8. Production of carrier proteins incorporating UAAs in their sequence can offer even more flexibility for homogeneous glycoconjugate vaccine preparation. More than 100 UAAs have been developed and further incorporated into various proteins9,10. Many of them contain bioorthogonal functions usually used to carry out post translational modifications11 or to graft biophysical probes12 or drugs13 but which are ideal handles for further conjugation with carbohydrate antigens. Successful examples have been claimed by Biotech14 using cell-free protein synthesis15 but preparation of glycoconjugate vaccines according to this strategy still waits for becoming popularized.
Application of the in vivo strategy for the production of mutated carrier protein needs a modified translational machinery that includes a specific codon, a tRNA recognizing the codon and an aminoacyl-tRNA synthetase (aaRS) which specifically catalyzes the transfer of the UAA on the tRNA (Figure 1)16. The pyrrolysine amber stop codon suppression is one of the most widely used methods to incorporate UAA, in particular the propargyl-lysine (PrK)17. The latter can in turn react with azido-functionalized carbohydrate haptens to provide fully defined, homogeneous glycococonjugates. In the present manuscript we describe how to synthesize the propargyl-L-lysine, an UAA carrying an alkyne handle, how to incorporate it into a target protein during its translation in a bacteria and finally how to perform conjugation between the modified protein and a hapten carrying an azide function using click chemistry.
1. Synthesis of the UAA: propargyl-lysine (PrK)
2. Production of the recombinant protein modified by PrK
3. Removal of the histidine tag by TEV protease digestion
4. Assessment of the unnatural amino acid propargyl-lysine accessibility and functionality for click chemistry
NOTE: Conjugate the mPsaA with 6-hexachloro-fluorescein-azide using the protocol described by Presolski et al.20 for click chemistry.
5. Conjugation of mPsaA with an azido-functionalized carbohydrate antigen (Pn14TS-N3) by click chemistry
In this project, a homogeneous glycoconjugate vaccine was prepared using the amber stop codon suppression strategy to introduce an UAA at a defined site (Figure 1). Pneumoccocal surface adhesin A was selected as the carrier protein moiety. This protein is highly conserved and expressed by all strains of Streptococcus pneumoniae22. It is highly immunogenic and previously used as a carrier in pneumococcal vaccine formulations21,23. As a proof-of-concept, the UAA propargyl-lysine efficiently charged by the wild type pyrrolysyl-tRNA synthetase (PylRS)/tRNA pair of the archaea Methanosarcina mazei was investigated. The propargyl-lysine is commercially available but can be advantageously prepared from Boc-L-lysine in only two synthetic steps (Figure 2). An amber codon was generated at a desired position in a pET24d plasmid containing the mPsaA gene. This plasmid was co-transformed with a pEVOL plasmid (a kind gift from Edward Lemke (EMBL19) containing orthogonal tools necessary to incorporate the propargyl-lysine, into competent E. coli BL21(DE3) strain. Positive co-transformed clones were selected using 25 µg/mL kanamycin and 30 µg/mL chloramphenicol. The plasmid pEVOL contains originally not one but two copies of the gene coding for MmPylRS to incorporate the propargyl-lysine residue: the first copy is under the control of a constitutive promoter while the expression of the other one is inducible in the presence of arabinose. However, we have noticed no dramatic decrease of propargyl-lysine incorporation if the MmPylRS gene under the control of constitutive promoter is suppressed.
The propargyl-lysine was introduced at position 32 in replacement of a lysine near the N-terminus of the PsaA. Any residue with a surface-exposed side-chain can virtually be exchanged in view of carrying out further conjugation. The mutated protein was produced in its mature form (mPsaAK32PrK) with inclusion of a cleavable 6-histidine tag sequence at its C-terminus. The efficiency of the mPsaAK32PrK production was checked by SDS-PAGE and Western Blot analysis using an anti-Histidine tag antibody, when growth was performed in the presence or the absence of the UAA propargyl-lysine and in comparison with the production of the wild type mPsaA (Figure 3). Visualization revealed a protein band at an expected molecular weight (Lanes 4, Figure 3A & 3B). The presence of a full-length protein strongly indicates the successful incorporation of the PrK into mPsaA. The intensity is, however, lower than that observed for wild type mPsaA (Lanes 2, Figure 3A & 3B). Leakage (i.e., production of the full-length protein without incorporation of the UAA) and premature release of the protein by the Release Factor RF1 during translation are two main drawbacks frequently encountered during this process. On one hand, no band at the expected molecular weight is visualized in the absence of propargyl-lysine meaning that no leakage occurred (Lanes 3, Figure 3A & 3B) and indirectly confirmed that the band observed on Lanes 4 corresponds to mPsaAK32PrK. On the other hand, no band can be seen at low molecular weight that could correspond to the truncated form of mPsaA (Lane 4 on Figure 3A). The mPsaAK32PrK was then purified by affinity chromatography, with a typical yield of 8 mg/L (in comparison to 12-20 mg/L for the wild type protein) and the incorporation of the propargyl-lysine residue was finally confirmed by mass spectrometry (Figure 4). The histidine tag was removed upon proteolytic cleavage using TEV protease (Figure 4). The stability of the mPsaAK32PrK thus obtained was assessed by circular dichroism, which showed that the structure of the protein was not disturbed by the mutation of the Lysine 32 into a propargyl-lysine (data not shown).
Having the mPsaAK32PrK, the reactivity of the alkyne for click chemistry was assessed using an azido-functionalized fluorescein and further used to conjugate a synthetic oligosaccharide antigen β-2-azidoethyl d-Galp-(1→4)-β-d-Glcp-(1→6)-[β-d-Galp-(1→4)]-β-d-GlcpNAc (Pn14TS) (Figure 5). This tetrasaccharide is related to the S. pneumoniae type 14 capsular polysaccharide and has previously been conjugated to mPsaA using different conjugation chemistries8,21,24. Experiments here were done in comparison with wild type mPsaA as a control. The histidine tag was first removed upon proteolytic cleavage using the TEV protease. The digested mPsaAK32PrK and mPsaA WT were then conjugated to the fluoroprobe (Figure 6A) or Pn14TS (Figure 6B). The reaction was assessed by SDS-PAGE. The small increase in the molecular weight of the sample between lane 6 and 7 (Figure 6B) indicates a successful conjugation with the tetrasaccharide Pn14TS. Finally, the glycoconjugate was purified by gel filtration and its identity confirmed by mass spectrometry (Figure 6C). The conjugation by click chemistry being quantitative the majority of the mPsaAK32PrK was conjugated with the Pn14TS-N3 as illustrated by the mass spectrometry results (Figure 6C).
Figure 1: Incorporation of propargyl-lysine (PrK) into mPsaA during translation using an orthogonal pyrrolysyl-tRNA synthetase/tRNA pair and TAG codon reassignment25. During translation, endogenous synthetases catalyze the link between amino acids and corresponding tRNAs. Then, loaded tRNAs are used by the ribosomal machinery to generate the neo-synthesized polypeptide. According to the amber stop codon suppression strategy, an orthogonal aminoacyl-tRNA synthetase (aaRS) (herein a pyrrolysyl-tRNA synthetase from M. mazei), loads an UAA (herein PrK) on its cognate tRNA which designed anticodon can read the amber stop codon (TAG) on the mRNA. This specific recognition directs the incorporation of the UAA into the specific site on the target protein. Figure reproduced from Wang et al.25. Please click here to view a larger version of this figure.
Figure 2: Propargyl-lysine synthesis. (A) Steps of propargyl-lysine synthesis. Insert: Monitoring of the deprotection of Boc-l-Lys(prop-2-ynyloxycarbonyl)-OH intermediate: thin-layer chromatography on 0.25 mm silica gel plates with fluorescent indicator (GF254) and visualised by charring with vanillin in sulfuric acid/ethanol (1.5:95 v/v); eluent: CH2Cl2/MeOH (9:1), left lane: Boc-l-Lys(prop-2-ynyloxycarbonyl)-OH (Rf 0.90), right lane: crude propargyl-lysine, (Rf 0.38). 400 MHz 1H (B) and 13C NMR spectra (C) of propargyl-lysine recorded in D2O. Please click here to view a larger version of this figure.
Figure 3: Analysis of crude cell samples. (A) SDS-PAGE analysis and (B) Western blot analysis on crude cell samples. Lane 1: unstained protein marker; Lane 2: crude cell extract of wild type mPsaA; Lane 3: crude cell extract of mPsaAK32TAG grown in the absence of PrK; Lane 4: crude cell extraction of mPsaAK32TAG grown in the presence of PrK. Conditions: 12% acrylamide gel, running at 100 V, 2 h. SDS-PAGE stained by Coomassie blue; Western Blot revealed using anti-histidine tag antibody and secondary antibody coupled with AlexaFluor680. Please click here to view a larger version of this figure.
Figure 4: Histidine tag removal and mass spectrometry analysis. (A) SDS-PAGE analysis. Lane 1: unstained protein marker; Lane 2: crude cell extract; Lane 3: unbound fraction; Lane 4: wash fraction with 10 mM imidazole; Lane 5: wash fraction with 20 mM imidazole; Conditions: 12% acrylamide gel running at 100 V for 2 h, and stained by Coomassie blue; (B) MALDI-TOF-MS spectra of (top) mPsaA WT, theoretical MW 33 103 Da, found 33 106 Da and (bottom) mPsaA K32PrK, theoretical 33 184 Da, found 33 192 Da. The found masses are within the expected margin error. Please click here to view a larger version of this figure.
Figure 5: Schematic representation of the conjugation strategy by click chemistry. A single tetrasaccharide bearing an azide is specifically coupled to its complementary biorthogonal alkyne group on mPsaA K32PrK (mPsaA representation based on the 1PSZ PDB file, with a resolution of 2.0 Å26). Please click here to view a larger version of this figure.
Figure 6: Histidine-tag digestion and conjugation of mPsaA with (A) fluorescein-N3 and with (B) Pn14TS. (A) SDS-PAGE Lanes 1-3: WT mPsaA; Lane 4-6: mPsaAK32PrK; (B) Lane 1: unstained protein marker; Lane 2-4: WT mPsaA; Lane 5-7: mPsaAK32PrK. 2 µg protein sample/lane, 12% acrylamide, 100 V, 2 h; (C) MALDI-TOF-MS spectra of the Pn14TS-mPsaAK32PrK theoretical MW 34 091 Da, found 34 088 Da. Please click here to view a larger version of this figure.
Supplemental File 1. Please click here to view this file (Right click to download).
Site-directed mutagenesis is a straightforward strategy to incorporate specific amino acids at a defined position of a protein which remains barely used with the aim of preparing glycoconjugate vaccines7,8,14. Classical mutagenesis based on the 20 natural amino acids approach is highly efficient since no modification of the translation machinery is required. Cysteine mutations are usually targeted to further explore the unique thiol reactivity either directly or in two steps (e.g., after its modification into a dehydroalanine intermediate, a strategy called post-translational mutagenesis)27,28. Genetic code expansion is perhaps even more attractive and flexible since it allows the direct incorporation of a wide range of UAAs with diverse functionalities9,10. While several UAAs can be incorporated simultaneously within a protein29, the number of mutations is usually more limited. We herein applied the related amber stop codon strategy to introduce a single propargyl-lysine in a carrier protein. The incorporation can take place at any position provided the sidechain of the initial amino acid was surface-exposed, a criterion easily determined from X-ray crystallographic structures or in silico modeling. Moreover, it is not limited to propargyl-lysine but can be extended to any UAA functionalized with a biorthogonal function which will later serve as an anchor to graft the incoming carbohydrate antigen and for which an orthogonal aaRS/tRNA pair exists.
One of the drawbacks of the strategy is the possible production of truncated protein, resulting from the release of the peptidyl sidechain when reading the amber stop codon, as a side-product. Even if we did not observe any truncated form here (probably degraded by the bacteria because of it very small size), a histidine tag has been added at C-terminus of the protein to facilitate the purification of the expected full-length mutated protein from impurities and noticeably from the truncated protein (which by essence does not express the histidine tag sequence). This can become essential if the UAA incorporation is carried out near the protein C-terminus since purification cannot be attempted using alternative chromatography techniques like gel filtration.
For most applications removal of the histidine tag is not mandatory. However, it may be useful regarding the design of glycoconjugate vaccine as part of the immune system may be diverted against the tag sequence. For this proof-of-concept, we inserted an amino acid length sequence specifically cleaved by the TEV protease which leaves five extra amino acids on the carrier protein after digestion.
The conjugation step between the alkyne of the propargyl-lysine and a representative synthetic oligosaccharide Pn14TS related to a pneumococcal capsule and bearing a complementary azide was carried out according to a click chemistry protocol reported by Presolski et al.20 If necessary, completion of the reaction can easily be reached by increasing the reaction time or by modifying the ratio between the alkyne, azide and copper reactants and reagents. Copper salts are eliminated by treatment with excess EDTA followed by a short purification by steric exclusion chromatography.
The glycoconjugate obtained with the technique described in the present work can then be used to immunize mice. Having such fully-defined and easily modulated glycoconjugate in hands provides invaluable tools to evaluate the impact of the hapten/protein carrier connectivity on the immune response8. Since increasing the hapten/protein ratio is often correlated with enhanced anti-hapten humoral response when using short haptens30, one might be interested in testing conjugates with multiple haptens. The incorporation of multiple UAAs however needs some adjustments of the protocol as the incorporation of an UAA in the protein tends to decrease the yield of protein production due to the RF1 activity.
In definitive, this method is a powerful tool to gain access to homogeneous glycoconjugate vaccines facilitating their physico-chemical characterization and further carbohydrate antigen/carrier connectivity-immunogenicity relationship studies.
The authors have nothing to disclose.
E.C. gratefully acknowledges the financial support from La Région Pays de la Loire (Pari Scientifique Program "BioSynProt"), in particular a doctoral fellowship to T.V. We also acknowledge Dr Robert B. Quast (INRA UMR0792, CNRS UMR5504, LISBP, Toulouse, France) for his precious technical advices.
AIM (autoinductif medium) | Formedium | AIMLB0210 | Solid powder |
Boc-Lys-OH | Alfa-Aesar | H63859 | Solid powder |
BL21(DE3) | Merck Novagen | 69450 | E. coli str. B, F– ompT gal dcm lon hsdSB(rB–mB–) λ(DE3 [lacI lacUV5-T7p07 ind1 sam7 nin5]) [malB+]K-12(λS) |
Dialysis membrane | |||
DNAseI | |||
Filter 0.45 µm | |||
L-arabinose | |||
lysozyme | |||
Ni-NTA resin | Machery Nagel | Protino | Ni-NTA beads in suspension into 20% ethanol |
Pall centrifugal device | |||
pET24d-mPsaAK32TAG-ENLYFQ-HHHHHH | this study | same as pET24d-mPsaA-WT but with a K32TAG mutation in the mPsaA gene | |
pET24d-mPsaA-WT | this study | pET24d plasmide with the Wt mPsaA gene cloned between the BamHI and XhoI restriction sites with a TEV protease sequence followed by a His6 tag at the C-terminal end of mPsaA gene and carrying the Kanamycine resistance gene | |
pEVOL plasmid | gift fromEdward Lemke EMBL (ref 19) | plasmide with p15A origin, two copies of MmPylRS (one under GlnS promoter and one under pAra promoter), one copy of the tRNACUA under the ProK promoter, the chloramphenicol resistance gene | |
Propargyl chloroformate | Sigma-Aldrich | 460923 | Liquid |
Sonicator | Thermo Fisher | FB120-220 |