
The PROTEIN PURIFICATION AND CHARACTERIZATION Team was headed by George Phillips, Jr., PhD, and was responsible for purifying all recombinant protein produced by the project. This team also confirmed that the expressed protein corresponded to the expected product.
An efficient and semi-automated pipeline system was established for high-throughput purification of E. coli-expressed proteins. The protocols have been optimized to work with unlabeled proteins, proteins labeled with Se-Met for X-ray crystallography, and proteins labeled with 15N or 13C;15N for NMR spectroscopy.
Key Steps of Protein Purification
Two-Step Purification of Recombinant Protein. Recombinant protein was purified using a two-step chromatography procedure, with an optional polishing step where ionic or size exclusion columns are used to improve the purity of the target protein. Once proteins were processed through the concentration stages, all samples were sent to the ESI-MS and MALDI-MS analysis for quality assurance before X-ray crystal screening or NMR HSQC data acquisition.
Key Features of the System. The key features of this system included an HPLC that used a binary gradient pump to purify six proteins sequentially by applying gradient elution from six independent Ni-IDA columns. One-step desalting and subtractive Ni-IDA chromatography removed His-tagged proteins from the target protein. Three such systems were used.
The construct used always yielded an N-terminal serine following TEV protease cleavage. Sesame (LIMS) was used for data capture and analysis. Details and statistics for each purification step, including % TEV cleavage, yields, and purity, were recorded.
Optimization of Purification Protocols. In order to optimize the pipeline, we developed protocols for automated protein production. We found that the presence of chaotrophic agents such as ethylene glycol and imidazole in the initial sonication buffer are critical to obtain high purity of fusion proteins. Optimum concentration of ethylene glycol and imidazole are 20% (w/v) and 35 mM, respectively. The optimized protocol allowed purification of native, SeMet-, 15N-, and 13C;15N-labeled proteins up to 150 mg of protein from 2L culture volume. The purity of target proteins was typically greater than 90%, suitable for structural determination by X-ray crystallography and NMR.
Experimental bioinformatic data were used to improve efficiency of current E. coli protein production pipeline. The most common reason for failures of target purification was poor TEV proteolysis, particularly when the percentage of cleavage was less than 70%. The results emphasized the requirement for screening methods that can assess the TEV proteolysis at small-scale cell culture stage. We tested expression vectors that contain different proteases cleavage sites and different linker sequences and developed high-throughput methods for screening for cleavage and solubility at the small-scale expression stage.
We also utilized a highly ordered data storage system, Lamp module in Sesame (LIMS), to monitor the quality control of purification processes and to store the biochemical properties of purified protein such as mobility and purity on SDS-polyacylamide gel, UV-visible spectra, MALDI-MS, and ESI-MS.
Expression Vectors and Large-Scale Cell Growth. We designed a (His)n-MBP fusion tag system (n = 6 or 8) to overcome the low solubility of recombinant eukaryotic proteins and to provide a generic Ni-IMAC purification strategy. The pVP13 and pVP16 expression vectors used for these studies were derived from pQE80 (Qiagen, Valencia, CA) to express an N-terminal fusion protein consisting of (His)n-MBP and a linker region containing the TEV protease site contiguous with the second residue of the target protein. Either E. coli Rosetta or B834 strains were used to produce unlabeled-, 15N-, and 15N/13C-, or SeMet-labeled proteins, respectively. Cells were inoculated in a 2-liter polyethylene terephthalate bottle which contained 500 ml of Terrific Broth or auto-induction medium and incubated in a shaker at 250 rpm, 25oC for 22-24 h. Cells were harvested by centrifugation at 5000 x g for 20 min.
Protein Purification Protocols. The overall philosophy of the protein purification pipeline was to automate the protocols as much as possible while preserving protein quality sufficient for structural studies. Protein purification processes were as follows:
Step 1: Cell lysis and preparation of the soluble fraction
Step 2: 1st IMAC capture of His-tagged fusion proteins
Step 3: Desalting of fusion proteins into TEV proteolysis buffer
Step 4: TEV proteolysis of fusion tags
Step 5: 2nd IMAC removal of tag and isolation of target proteins
Step 6: Desalting of targets
Step 7: Concentration of targets
Step 8: Drop-freezing of targets
Steps 2, 3, 5, and 6 were performed on the ÄKTA Purifier controlled by Unicorn 4.12 software. Detail description of purification processes was reported.
Confirm Identify and Integrity of Proteins. All purified proteins were subjected to MALDI-TOF and ESI mass spectrometry to confirm identity and integrity, determine oligomeric state, and investigate possible ligands. Incorporation levels of Se-Met and 15N and 13C isotopes are also determined by ESI-MS.
Use of the System for Inventory Control: All cell pastes grown in Large-Scale got a barcode that link3e to Sheherazade. The record contained the expression, solubility, and cleavage results for the protein candidate for purification, plus cell mass. An inventory report was developed in Sheherazade that returns all the testing results, vector information, and whether or not a particular sample is an outside request. This report is fed into an Access based database, which has been coded to sort the cell pastes, allowing for easy selection. Outside requests received the highest priority and, then, cell pastes are sorted based on suitability scores.