Stable 293 T and CHO cell lines expressing cleaved, stable HIV-1 envelope glycoprotein trimers for structural and vaccine studies

Abstract

Background: Recombinant soluble, cleaved HIV-1 envelope glycoprotein SOSIP.664 gp140 trimers based on the subtype A BG505 sequence are being studied structurally and tested as immunogens in animals. For these trimers to become a vaccine candidate for human trials, they would need to be made in appropriate amounts at an acceptable quality. Accomplishing such tasks by transient transfection is likely to be challenging. The traditional way to express recombinant proteins in large amounts is via a permanent cell line, usually of mammalian origin. Making cell lines that produce BG505 SOSIP.664 trimers requires the co-expression of the Furin protease to ensure that the cleavage site between the gp120 and gp41 subunits is fully utilized.

Results: We designed a vector capable of expressing Env and Furin, and used it to create Stable 293 T and CHO Flp-In™ cell lines through site-specific recombination. Both lines produce high quality, cleaved trimers at yields of up to 12-15 mg per 1 × 109 cells. Trimer expression at such levels was maintained for up to 30 days (10 passages) after initial seeding and was consistently superior to what could be achieved by transient transfection. Electron microscopy studies confirm that the purified trimers have the same native-like appearance as those derived by transient transfection and used to generate high-resolution structures. They also have appropriate antigenic properties, including the presentation of the quaternary epitope for the broadly neutralizing antibody PGT145.

Conclusions: The BG505 SOSIP.664 trimer-expressing cell lines yield proteins of an appropriate quality for structural studies and animal immunogenicity experiments. The methodology is suitable for making similar lines under Good Manufacturing Practice conditions, to produce trimers for human clinical trials. Moreover, any env gene can be incorporated into this vector system, allowing the manufacture of SOSIP trimers from multiple genotypes, either by transient transfection or from stable cell lines.

Figure 1

Vector for constitutive secretion of BG505 SOSIP.664 gp140 in a Flp-In™ based expression system, and stable cell line selection. (A) Design of the pAM/C construct for expressing BG505 SOSIP.664 gp140. The plasmid map shows the site of the env and furin gene insertions, the promoters and the Poly A sequences. (B) Intracellular Env expression in transfected 293 T and CHO cells. The histograms represent parental cells (red) and stable cell clones (blue); the numbers (top right of each panel) are the mean fluorescence intensity (MFI) values after staining with FITC-2G12. (C) Secretion of BG505 SOSIP.664 gp140 trimers by Stable 293 T and CHO cell clones. The trimer concentrations in the culture supernatants were determined by ELISA using 2G12 and bio-PGT145. (D) Fluorescent microscopy of stable cell clones. Cells were grown in an 8-well chamber slide, treated with Brefeldin A, fixed, permeabilized and stained for Env (FITC-2G12; green) or nuclear DNA (DAPI; blue). The left panels show parental 293 T and CHO cells, the right, the stable cell clones.

Figure 2

Sustained expression of BG505 SOSIP.664 gp140 by stable cell lines. (A) Intracellular Env expression with continued passage of the 293 T 13# 3–5 and CHO B-D7 stable cell lines (blue curves). The fixed and permeabilized cells were stained with FITC-2G12 (20 μg/ml) after culture for 6 h in the presence of Brefeldin A. MFI values for the Env-expressing clones are recorded in the top right corner of each histogram. Parental cells served as negative controls (red curves show MFI values: ranges 9.12-12.9 for 293 T cells and 7.23-13.2 for CHO cells). (B) Production of BG505 SOSIP.664 gp140 trimers by the stable cell lines throughout the culture period, as determined by ELISA using 2G12 and bio-PGT145. (C) Viability of stable cell clones during passage. The cells were stained with trypan blue, with the percentage of viable cells (parental vs. stable) shown as the passage number increases.

Figure 3

Biochemical characterization of cell line-derived BG505 SOSIP.664 gp140 trimers and comparator proteins. (A) BN-PAGE analysis of 2G12-purified Env proteins produced by the BG505 SOSIP.664-expressing 293 T and CHO stable cell lines. The gels were stained with Coomassie blue. The molecular weights of marker (M) proteins (thyroglobulin, 669 kDa and ferritin, 440 kDa) are indicated. (B) A similar BN-PAGE analysis, but of SEC-purified trimers. (C) SDS-PAGE analysis of SEC-purified BG505 SOSIP.664 trimers, under non-reducing (- DTT) and reducing (+ DTT) conditions, followed by Coomassie blue staining. Cleavage of gp120 from gp41_ECTO is assessed by the conversion of the gp140 band to gp120 in the presence of DTT; the released gp41_ECTO subunit is not always stained strongly. (D) Western blot analysis of a reducing SDS-PAGE gel to assess the quality of SEC-purified BG505 SOSIP.664 trimers. The blots were probed with the anti-gp120 MAb, ARP 3119. No Env degradation products arising from V3-clipping or other proteolysis events are visible. (E) Reducing SDS-PAGE analysis of SEC-purified BG505 SOSIP.664 trimers and comparator Env proteins, followed by Coomassie blue staining. The comparator proteins were also SEC-purified to yield the 120-kDa fraction for MN gp120 and the “trimer” fraction (i.e., proteins containing three gp120 and three gp41_ECTO subunits) for uncleaved CZA97.012 gp140. Multiple degradation and/or aggregation products derived from MN gp120 and CZA97.012 gp140 are visible, but none from the BG505 SOSIP.664 trimers. (F) Western blot analysis of the same gel shown in (E). The detection antibodies were a pool (1:1:1) of polyclonal HIV-Igs derived from subtype A, B and C infections. The various aggregates or degradation products seen in (E) are Env-based.

Figure 4

Antigenicity of cell line-derived BG505 SOSIP.664 gp140 trimers. (A) SPR sensorgrams for the PGT145 bNAb and the F105 non-NAb. Each MAb was captured onto the chip by an immobilized anti-Fc Ab, and the binding of BG505 SOSIP.664 gp140 trimers (200 nM) from the various cell sources was recorded as response difference (RU) after background correction: PGT145 bound strongly and similarly to the trimers of different origin, but F105 binding was undetectable. Each curve represents one of two similar replicates. (B) Representative binding curves in the 2G12-capture ELISA for the same PGT145 bNAb and F105 non-NAb to BG505 SOSIP.664 gp140 trimers from the various cell sources. The plotted OD values have been background corrected (i.e., with no gp120 present). Similar data were obtained in three experiments of the same design.

Figure 5

Negative stain EM images and DSC traces for BG505 SOSIP.664 gp140 trimers produced by 293 T and CHO cells. (A) Reference-free 2D class averages are shown for trimers from transiently transfected or Stable 293 T and CHO cell lines, as indicated. (B) Thermal denaturation profiles of BG505 SODSIP.664 trimers from the 293 T (left panel) and CHO stable cell lines (right panel).