Bacterial cytoplasm as an effective cell compartment for producing functional VHH-based affinity reagents and Camelidae IgG-like recombinant antibodies

Abstract

Background: The isolation of recombinant antibody fragments from displayed libraries represents a powerful alternative to the generation of IgGs using hybridoma technology. The selected antibody fragments can then be easily engineered into (multi)-tagged constructs of variable mass and complexity as well as reconstituted into Camelidae IgG-like molecules when expressed fused to Fc domains. Nevertheless, all antibody constructs depend on an oxidizing environment for correct folding and consequently still belong to the proteins difficult to express in bacteria. In such organisms they are mostly produced at low yields in the periplasmic space.

Results: We demonstrate that fusion constructs of recombinant antibodies in combination with multiple tags can be produced at high yields and totally functional in the cytoplasm of bacteria expressing sulfhydryl oxidase. The method was applied to structurally demanding molecules such as VHHs fused to SNAP and Fc domains and was validated using the antibody-derived reagents in a variety of immune techniques (FACS, ELISA, WB, IP, SPR, and IF).

Conclusions: The collected data demonstrate the feasibility of a method that establishes a totally new approach for producing rapidly and inexpensively functional Camelidae IgG-like monoclonal antibodies and antibody-based reagents containing multiple disulfide bonds and suitable for both basic research and clinical applications.

Figure 1

Functional validation of the anti-HER2 SNAP-tagged VHH A10 expressed in the bacterial cytoplasm. Cytoplasmic-expressed A10-SNAP fusion protein was eluted at almost homogeneity (E) after a single metal affinity chromatographic (IMAC) step. Both a portion of the fusion protein and SNAP alone accumulated in the pellet (P). FACS analysis (filter band-pass: 564-606 nm) was performed using HER2 negative (MCF10A) and positive (SKBR3) cells and used to compare different A10 constructs and detection methods: i) myc + His-tagged antibodies expressed in bacterial periplasm in combination with anti-His and a secondary antibodies; ii) SNAP + His-tagged antibody expressed in bacterial cytoplasm in combination with anti-His plus secondary antibody; iii) SNAP + His-tagged antibody expressed in bacterial cytoplasm directly linked to the SNAP-Surface 549 (NEB) chromophore. The same cell lines were used in combination with A10-SNAP for cell ELISA and for immunostaining HER2 at the cell membrane of negative (MCF10A) and positive (SKBR3) cells.

Figure 2

Comparison between antibody constructs expressed in the bacterial cytoplasm and periplasm. a) The affinity of the VHH A10 for its antigen HER2 (purified ectodomain) was measured by Surface Plasmon Resonance. Three constructs were compared: A10 tagged myc + His and expressed in the bacterial periplasm, A10 tagged SNAP + His and A10 tagged SORT + His, both expressed in the bacterial cytoplasm. The graphs correspond to a single experiment representative of at least three repeats. The analyte concentrations were in the range 300-3.5 nM (myc-tag), 100-3.5 nM (SNAP-tag), and 30-0.35 nM (SORT-tag). b) The anti-HER2 antibodies A10 fused to GFP was expressed in the cytoplasm and directly assessed by FACS (filter band-pass: 515-545 nm) using HER2 negative (MCF10A) and positive (SKBR3) cells. c) The anti-HER2 antibodies A10 and C8 were expressed fused to the LPTEG peptide (SORT tag) in both the periplasm and the cytoplasm and the resulting samples were assessed by FACS (filter band-pass: 564-606 nm) using HER2 negative (MCF10A) and positive (SKBR3) cells, and the 6xHis tag for visualization.

Figure 3

Functional validation of Fc-VHH fusion antibodies. Fc-VHH constructs corresponding to the clones A10 and C8 were expressed in the bacterial cytoplasm, purified, and finally tested by cell ELISA and compared with the monoclonal anti-HER2 therapeutic monoclonal antibody trastuzumab by FACS (filter band-pass: 564-606 nm) using HER2 negative (MCF10A) and positive (SKBR3) cells. C8-Fc was used for HER2 detection by WB and to immunoprecipitate HER2 from SKBR3 cells.

Figure 4

Identification of the membrane receptor HER2 by immunofluorescence. The IF staining pattern of the VHH A10 fused to either myc (periplasmic expression) or Fc (cytoplasmic expression) was compared with that of trastuzumab in HER2 neg and HER2 pos cells.

Figure 5

Avidity effect of the bivalent Fc-VHH construct. The avidity of the bivalent A10-Fc was compared with that of the bivalent constructs A10-alkaline phosphatase (AP, periplasmic bacterial expression) and trastuzumab (hybridoma expression). The graphs correspond to a single experiment representative of at least three repeats. The analyte concentrations were in the range 1-0.01 nM (myc-tag), 300-3.5 nM (AP-tag), and 1-0.01 nM (SORT-tag).

Figure 6

Antibody accumulation in xenografted tumors. HER2 positive tumors were recovered from treated mice 3 and 48 hours after antibody injection. Tumors were cut in two pieces along their median axis and their infrared fluorescence intensity detected at 770 nm.