Development of measles virus-based shielded oncolytic vectors: suitability of other paramyxovirus glycoproteins

Abstract

Antibody-mediated neutralization may interfere with the efficacy of measles virus (MV) oncolysis. To circumvent vector neutralization, we sought to exchange the envelope glycoproteins, hemagglutinin (H) and fusion (F), with those from the non-crossreactive Tupaia paramyxovirus (TPMV). To sustain efficient particle assembly, we generated hybrid glycoproteins with the MV cytoplasmic tails and the TPMV ectodomains. Hybrid F proteins that partially retained fusion function, and hybrid H proteins that retained fusion support activity, were generated. However, when used in combination, the hybrid proteins did not support membrane fusion. An alternative strategy was developed based on a hybrid F protein and a truncated H protein that supported cell-cell fusion. A hybrid virus expressing these two proteins was rescued, and was able to spread by cell fusion; however, it was only capable of producing minimal amounts of particles. Lack of specific interactions between the matrix and the H protein, in combination with suboptimal F-protein processing and inefficient glycoprotein transport in the rescue cells, accounted for inefficient particle production. Ultimately, this interferes with applications for oncolytic virotherapy. Alternative strategies for the generation of shielded MV are discussed.

Figure 1

Functional characterization of TPMV F proteins with truncations of the cytoplasmic tail. (A) Examples of visual assessment of syncytium formation. Cells were co-transfected with a standard or mutated F-expression plasmids, the standard H-expression plasmid, and a GFP-expression plasmid. The fusion score was assessed 24-hours after transfection. The F-plasmids used for the four examples are: FΔ32 (fusion score: +++), FΔ38 (fusion score: ++), FΔ34 (fusion score: +) and FΔ40 (fusion score: -). (B) Sequence of the F-protein predicted cytoplasmic tail (CT) and part of the transmembrane (TM) segment; a vertical line separates the two regions. Deletion mutants were named by their extent, such as a deletion of 10 amino acids was named FΔ10. The fusion efficiency of each construct was tested after co-expression with the standard H_His-tag protein. The ectodomain of this protein is extended with a six-histidine tag, allowing for fusion-support function in Vero-αHis cells, expressing a pseudo-receptor recognizing the His-tag ^, . Fusion efficiently levels are denoted on the right using the scale illustrated in panel A.

Figure 2

Biochemical and functional characterization of F protein mutants. (A) Protein expression analysis. Cell lysates from transfected Vero-αHis cells used to assess fusion function were digested without or with EndoH (−/+), fractionated on 10% SDS-PAGE, and immunoblotted with a TPMV F specific antibody (F_ecto, upper panel). F₀: precursor protein, F₁: larger portion of the active F molecule (upper panel), actin: cellular actin (lower panel). (B) Schematic representation of the FΔ32 deletion mutant and the hybrid glycoprotein FΔ32/mvcyt (MV amino acids are bold typeface and underlined). Fusion scores are indicated with the same convention as in Figure 1.

Figure 3

Functional and biochemical characterization of H protein cytoplasmic tail truncations. (A) Sequence of the predicted 94 residue cytoplasmic tail (CT) and part of the transmembrane (TM) region; the two regions are separated by a vertical line. H protein deletions were named for the extent of their truncation, such that an H protein with a CT truncation of 10 amino acids was named HΔ10. The fusion efficiency of each construct when co-expressed with the full-length F protein in Vero-αHis cells is indicated on the right by the same convention outlined in figure 1. (B) H-protein expression analysis. Cell lysates were collected and fractioned on 7.5% SDS-PAGE and probed with the TPMV H specific antibody H_ecto (upper panel) or for cellular actin (lower panel). (C) Fusion of Vero-αHis cells after co-transfection of plasmids expressing FΔ32/mvcyt and HΔ90. A co-transfected plasmid expressing GFP allows for visualization of transfected cells. Cells were observed 24 hours after the beginning of transfection.

Figure 4

Time course of cell-associated and cell-free virus production. Growth analysis of (A) TPMV in TBF cells, (B) MVvac2(GFP)N and MVvac2(FΔ32/mvcyt-HΔ90)GFP in Vero-αHis cells. All viruses were inoculated at an MOI of ~0.03. Titers were determined by collection of cellular (circles) and supernatant (squares) fractions every 12 hours over a 60 hour time period for TPMV and MV (filled symbols), and every 24 hours over a 120-hour time period for the hybrid virus (open symbols). Vertical axis: titer by TCID₅₀/ml, horizontal axis: time post-infection.

Figure 5

FACS analysis of surface expression of MV and TPMV glycoproteins in CHO-K1 and Vero-αHis cells. Cells were mock-transfected (negative) or transfected with the plasmids encoding MV-H617, TPMV-H_His-tag or TPMV-H 90. After 24 hours, cells were stained with anti-Penta-His AlexaFluor 647 conjugated antibody.

Figure 6

Wild-type and truncated TPMV H protein transport kinetics. CHO-K1 cells were transfected with wild-type F and (A) H_His-tag or (B) H 90 or an empty vector (mock) for 24 hours before being starved for 30 minutes and pulsed with 100μCi/ml ³⁵S labeled methionine and cysteine for 1 hour and then chased for up to 4 hours. After the indicated chase period, cell lysates were collected and H was immunoprecipitated and digested without or with EndoH (−/+ lanes, respectively) followed by fractioning on 7.5% SDS-PAGE. (C) Kinetics of complex oligosaccharide acquisition. Vertical axis: percent EndoH resistance of three experiments for wild type H (diamonds) and HΔ90 (squares). Horizontal axis: time of chase (hours).

Figure 7

Processing kinetics of TPMV F and FΔ32/mvcyt hybrid proteins. (A and B) SDS-PAGE analysis. CHO-K1 cells were transfected with TPMV H_His-tag and either wild-type TPMV F (A), FΔ32/mvcyt (B), or an empty vector (mock in A and B), for 24 hours before being starved for 30 minutes and pulsed with 100μCi/ml ³⁵S labeled methionine and cysteine for 1 hour, then chased for up to 6 hours. Cellular lysates were collected at the indicated times and F protein was immunoprecipitated and digested without or with EndoH (−/+ lanes, respectively) followed by fractioning on 10% SDS-PAGE. The positions of F₀ and F₁ are indicated on the right. R: EndoH resistant band. S: EndoH sensitive band. (C) Kinetics of F-protein processing. The quantity of F₁ protein was measured in three independent experiments. Wild-type F: diamonds, FΔ32/mvcyt: squares. Vertical axis: percentage of F-protein processed. Horizontal axis: time of chase (hours).