A Recombinant Collagen–mRNA Platform for Controllable Protein Synthesis

We have developed a collagen–mRNA platform for controllable protein production that is intended to be less prone to the problems associated with commonly used mRNA therapy as well as with collagen skin-healing procedures. A collagen mimic was constructed according to a recombinant method and was used as scaffold for translating mRNA chains into proteins. Cysteines were genetically inserted into the collagen chain at positions allowing efficient ribosome translation activity while minimizing mRNA misfolding and degradation. Enhanced green fluorescence protein (eGFP) mRNA bound to collagen was successfully translated by cell-free Escherichia coli ribosomes. This system enabled an accurate control of specific protein synthesis by monitoring expression time and level. Luciferase–mRNA was also translated on collagen scaffold by eukaryotic cell extracts. Thus we have demonstrated the feasibility of controllable protein synthesis on collagen scaffolds by ribosomal machinery.

. Trimeric structure of collagen. Blue, purple and green indicate three α 1 chains, respectively. (A) Glycine-Proline-Proline. (B) Glycine-Proline-Cysteine. The sulfhydryl groups of cysteines are shown as yellow sticks. The figure is made by PyMol [1] using PDB 1K6F and modeling of C instead of P in the original structure.

I. Cloning of MBP-collagen
Since collagen triple helix is formed in a C-to N-terminal direction and the correct trimerization of the C-terminus is crucial for collagen assembly, the use of C-terminal tags is avoided. We derived the collagen scaffold from a part of human collagen type I, α1 chain (NCBI accession number: NP_000079) with additional 10 Gly-Pro-Pro repeats at both ends for increased stability. MBP domain was fused to the N-terminus to improve the solubility and yield of collagen in E. coli. The his 6 -tag enables purification of MBP-collagen. A bacteriophage T4 fibritin foldon domain at the C terminus serves as a nucleation site to facilitate the correct folding of collagen triple helix. [2] The following collagen sequence was designed with a His tag (brown) and TEV protease cleavage site (red) at the N terminus. Natural collagen molecule contains no cysteine in the GXY repeats region. To provide the functional sulfhydryl group for further reaction with amine modified mRNA, we inserted two Glycine-Proline-Cysteine (GPC) triplets into our recombinant collagen (underlined).
Collagen mimic was optimized for E. coli codon usage and synthesized with a 5' BamHI and NdeI and 3' SacI followed by a stop codon. It was cloned into pUC57 by GeneWiz (South Plainfield, NJ).
Two separate restriction digests were performed: 5 μ g of collagen-pUC57 plasmid (donor) or 1μ g of pET-25b(+) plasmid (recipient) was digested with SacI and NdeI. The reaction mixture contained 5μ l of 20 U/μ l of each enzyme, 100 μg/ml BSA and 1× NEB buffer 4. The reaction was carried out at room temperature for 2.5-3 h. The Collagen-pUC57 plasmid digest was run on a 1% agarose gel and the bands of collagen gene (651 bp) from the double-cut product were cut out of the gel and purified with MEGA quick-spin TM total fragment DNA purification kit (JH Science, USA). 1 μ l of 10 U/μ l CIP (NEB, USA) was added to 50 μ l of pET-25b(+) plasmid digest and incubated for 30 min at 37 °C. The reaction mix was treated with 1 μ l CIP again for 30 min at 37 °C, then purified with MEGA quick-spin TM total fragment DNA purification kit. The ligation of collagen gene into the cut pET-25b(+) plasmid was performed with the ligation mix (Takara, Japan). The reaction mixture for the ligation included approximately 16 ng of cut collagen and 9 ng of pET-25b(+) plasmid digest. The reaction was carried out at room temperature for 30 min, then directly transformed into highly-efficient competent E. coli DH5a cells prepared as previously described. [3,4] Collagen gene was subcloned into a pET-MBP-TevH or pET-GB1-TevH plasmid by transfer-PCR. [5] Primers were ordered from Sigma (Rehovot, Israel). Plasmid purification was U/μ l DpnI (NEB, USA) was added to 10 μ l of TPCR product followed by incubation for 1-2 h at 37 °C . The DNA was then directly transformed into competent E. coli DH5a cells.
Colony PCR screening was performed using T7 and PetRev primers (Table 1) 301 AAC GGC AAG CTG ATT GCT TAC CCG ATC GCT GTT GAA GCG TTA TCG CTG ATT TAT AAC AAA 421 AAA GCG AAA GGT AAG AGC GCG CTG ATG TTC AAC CTG CAA GAA CCG TAC TTC ACC TGG CCG

II. Recombinant collagen expression and purification
The constructed plasmids were transformed into chemically competent E. coli BL21(DE3). Cells were grown overnight at 37 ˚C in LB media supplemented with 30 μg/ml kanamycin. Cultures were diluted 1:100 in fresh LB media containing appropriate selective antibiotics as above and grown for additional 2.5 h until the level of OD 600nm was 0.6-0.8. Protein expression was induced by

III. Characterization of MBP-collagen
Since the elution buffer for MBP-collagen (TBS: 50 mM Tris-Cl, 150 mM NaCl, 1mM DTT) has significant effect on the CD spectra at low wavelength (below 210 nm), TBS buffer was exchanged to water using Amicon Ultra-15 device (30K, Millipore) before CD measurement. Circular Dichroism Spectroscopy (CD) spectra were recorded using an Applied Photophysics-Chirascan spectrometer. Collagen sample was buffer exchanged into pure water before CD measurements. Data were collected in 2 nm increments with a 3 s averaging time, 1 nm bandpass, and 0.1 cm pathlength.
The sample was heated from 20°C to 90°C at 5°C increments (heating rate 1°C /min). The ellipticity (mdeg) was monitored with a 5 s averaging time, 1 nm bandpass, and 0.1 cm path-length.

IV. PCR and in vitro transcription
A pair of primers was used for PCR amplification of the eGFP gene under regulation of T7 promoter using pIVEX-GFP plasmid [6] as a template (GFP-F and GFP-R, Table 1). The reverse S7 primer is located upstream to the transcription terminator resulting in run-off PCR fragment lacking a terminator. In this construct the translation stop codon is found 108 bp upstream to the 3' end of mRNA. This distance is sufficient for accommodating a functional ribosome. PCR conditions for eGFP amplification were as follows: A single denaturation step (95 ˚C, South Korea) and used for the subsequent transcription reaction. For luciferase gene amplification, similar process was performed except that PT7CFE1-Chis-luciferase plasmid was used as PCR template, while Luc-F and Luc-R were used as forward and reverse primers, respectively ( Table 1).
The PT7CFE1-Chis-luciferase plasmid was constructed by cloning the luciferase gene into the multiple cloning sites of the plasmid pT7CFE1-Chis (Pierce).
The PCR fragment (43 ng/μl) was in vitro transcribed at 37° C for 2.5 hours using T7 RNA

VI. In vitro translation of eGFP and luciferase on collagen scaffolds
eGFP synthesis on collagen scaffolds was measured using the E. coli cell-free translation assay. S12 cell free extract was prepared from Escherichia coli strain BL21(DE3) according to previous report. [