Reducing conditions are the key for efficient production of active ribonuclease inhibitor in Escherichia coli

Abstract

Background: The eukaryotic RNase ribonuclease/angiogenin inhibitors (RI) are a protein group distinguished by a unique structure - they are composed of hydrophobic leucine-rich repeat motifs (LRR) and contain a high amount of reduced cysteine residues. The members of this group are difficult to produce in E. coli and other recombinant hosts due to their high aggregation tendency.

Results: In this work dithiothreitol (DTT) was successfully applied for improving the yield of correctly folded ribonuclease/angiogenin inhibitor in E. coli K12 periplasmic and cytoplasmic compartments. The feasibility of the in vivo folding concepts for cytoplasmic and periplasmic production were demonstrated at batch and fed-batch cultivation modes in shake flasks and at the bioreactor scale.Firstly, the best secretion conditions of RI in the periplasmic space were evaluated by using a high throughput multifactorial screening approach of a vector library, directly with the Enbase fed-batch production mode in 96-well plates. Secondly, the effect of the redox environment was evaluated in isogenic dsbA+ and dsbA- strains at the various cultivation conditions with reducing agents in the cultivation medium. Despite the fusion to the signal peptide, highest activities were found in the cytoplasmic fraction. Thus by removing the signal peptide the positive effect of the reducing agent DTT was clearly proven also for the cytoplasmic compartment. Finally, optimal periplasmic and cytoplasmic RI fed-batch production processes involving externally added DTT were developed in shake flasks and scaled up to the bioreactor scale.

Conclusions: DTT highly improved both, periplasmic and cytoplasmic accumulation and activity of RI at low synthesis rate, i.e. in constructs harbouring weak recombinant synthesis rate stipulating genetic elements together with cultivation at low temperature. In a stirred bioreactor environment RI folding was strongly improved by repeated pulse addition of DTT at low aeration conditions.

Figure 1

Growth with EnBase in MWPs of a library of 36 constructs with periplasmic RI production and luciferase activity of the different clones. A: OD₆₀₀'s are shown for all cultures of the vector library. Label sizes represent promoter strength [from weakest (pC, smallest labels) to strongest (pCTU, largest labels)]. Label shapes and colours represent ribosome binding strengths: T7 (red squares), lac (black triangle), var (blue triangle). Sizes and colours of the specific growth rate labels (circles) also indicate promoter and ribosome binding site strengths. B: The periplasmic aggregation signal measured as luminescence in samples of the RI expression library 5 h after induction. The data represent three independent cultivation experiments of the whole library.

Figure 2

SDS-PAGE images of total (T), soluble (S), or insoluble (IN) protein fractions of samples from shake flask cultures of E. coli RV308 dsbA⁺ with the following periplasmic constructs: pCUvar-malE-RI (A, F), pCUvar-ompA-RI (B, G), pCUvar-phoA-RI (C, H) and pCUlac-pelB-RI (D, I) 4 h after induction. Gels represent protein fractions after RI production without DTT (A-E) or with 12 mM DTT (F-J). Gels E and J represent insoluble protein fractions of the constructs without DTT (E) or with 12 mM DTT (J). Numbered lanes: 1 - total protein fraction 10 min before induction, 2 and 3 - soluble and total protein fractions 4 h after induction. Numbered lanes in gels E and J represent the insoluble fractions with the plasmids pCUvar-malE-RI (1A, 1F), pCUvar-ompA-RI (2B, 2G), pCUvar-phoA-RI (3C, 3H) and pCUlac-pelB-RI (4D, 4I). Protein size marker: PageRuler™ Protein Ladder Plus (Fermentas). For growth conditions see Materials and Methods. The amounts of applied protein are normalised.

Figure 3

SDS-PAGE images of total cell extracts (T), soluble (S), or insoluble (IN) protein fractions from periplasmic production constructs of E. coli RV308 dsbA^-. For vectors and conditions see Fig. 2.

Figure 4

RI protein activities in total soluble and periplasmic fractions after 4 h of batch RI production in different E. coli RV308 dsbA⁺ (A) and dsbA^- (B) constructs at 22°C with 0, 2, 6, 12 or 18 mM of DTT. Culture conditions: glucose-MSM medium, induction with 0.2 mM IPTG, preinduction temperature 37°C, shift to the respective temperature at the time of induction.

Figure 5

RI protein activities in total soluble fractions (A) and SDS-PAGE images of normalised total (T), soluble (S) and insoluble (IN) cell extracts (B) after 4 h of batch cytoplasmic RI production in shake flask batch cultures in E. coli RV308 pCUlac-His₆-RI at 22°C with 0, 2, 6, 12 or 18 mM of DTT. Lanes in (B): 1 - total protein fraction 10 min before induction, lanes 2 to 4 - soluble, total and insoluble protein fractions 4 h after induction. Gels represent protein fractions after 4 h of RI production without DTT at 37°C (gel 1) or 22°C (gel 2), or with 12 mM DTT at 37°C (gel 3) or 22°C (gel 4).

Figure 6

RI protein activities in total soluble and periplasmic fractions [in kU (gCWW)^-1] of E. coli RV308 pCUlac-pelB-RI and E. coli RV308 pCUlac-His₆-RI after EnBase cultivation in shake flasks with RI induction for 4 h at 22°C without or with 12 mM DTT (A), and SDS-PAGE images of total (T), soluble (S) and insoluble (IN) protein extracts (normalised to equal amounts protein) (B). Left gel: E. coli RV308 dsbA⁺ pCUlac-pelB-RI (periplasmic expression construct); right gel: E. coli RV308 pCUlac-His₆-RI (cytoplasmic production construct). Lanes: 1 - total protein fraction 10 min before induction, 2 to 4 - soluble, total, and insoluble protein fractions 4 h after induction with 12 mM of DTT. Protein size marker: PageRuler™ Protein Ladder Plus (Fermentas). Data originate from three experiments.

Figure 7

Growth curves of E. coli RV308 pCUlac-pelB-RI (periplasmic expression construct) and E. coli RV308 pCUlac-His₆-RI (cytoplasmic production construct) without (control) and with RI production in a batch and a fed-batch process with exponential glucose feeding in a 10 L bioreactor. For conditions see Material and Methods. Black symbols: control culture without induction; red symbols: batch bioreactor cultivation, blue symbols: fed-batch bioreactor cultivation.

Figure 8

The amount of reduced DTT during batch shake flask and fed-batch bioreactor processes with microarerobic production mode, was measured by using Measure-iTTM Kit (Invitrogen). The samples for reduced DTT evaluation were taken every synthesis hour starting with DTT addition moment. The data is derived from three assays.

Figure 9

RI activities in the total soluble protein fraction (white bars) and in the periplasmic fraction (grey bars) of E. coli of RV308 pCUlac-pelB-RI and RV308 pCUlac-His₆-RI from samples of batch and fed-batch bioreactor cultivations without and with addition of DTT. Data derive from three activity assays.