Sequence Determinants in Gammaretroviral Env Cytoplasmic Tails Dictate Virus-Specific Pseudotyping Compatibility

Abstract

Viruses can incorporate foreign glycoproteins to form infectious particles through a process known as pseudotyping. However, not all glycoproteins are compatible with all viruses. Despite the fact that viral pseudotyping is widely used, what makes a virus/glycoprotein pair compatible is poorly understood. To study this, we chose to analyze a gammaretroviral glycoprotein (Env) whose compatibility with different viruses could be modulated through small changes in its cytoplasmic tail (CT). One form of this glycoprotein is compatible with murine leukemia virus (MLV) particles but incompatible with human immunodeficiency virus type 1 (HIV-1) particles, while the second is compatible with HIV-1 particles but not with MLV particles. To decipher the factors affecting virus-specific Env incompatibility, we characterized Env incorporation, maturation, cell-to-cell fusogenicity, and virus-to-cell fusogenicity of each Env. The HIV-1 particle incompatibility correlated with less efficient cleavage of the R peptide by HIV-1 protease. However, the MLV particle incompatibility was more nuanced. MLV incompatibility appeared to be caused by lack of incorporation into particles, yet incorporation could be restored by further truncating the CT or by using a chimeric MLV Gag protein containing the HIV-1 MA without fully restoring infectivity. The MLV particle incompatibility appeared to be caused in part by fusogenic repression in MLV particles through an unknown mechanism. This study demonstrates that the Env CT can dictate functionality of Env within particles in a virus-specific manner.IMPORTANCE Viruses utilize viral glycoproteins to efficiently enter target cells during infection. How viruses acquire viral glycoproteins has been studied to understand the pathogenesis of viruses and develop safer and more efficient viral vectors for gene therapies. The CTs of viral glycoproteins have been shown to regulate various stages of glycoprotein biogenesis, but a gap still remains in understanding the molecular mechanism of glycoprotein acquisition and functionality regarding the CT. Here, we studied the mechanism of how specific mutations in the CT of a gammaretroviral envelope glycoprotein distinctly affect infectivity of two different viruses. Different mutations caused failure of glycoproteins to function in a virus-specific manner due to distinct fusion defects, suggesting that there are virus-specific characteristics affecting glycoprotein functionality.

Keywords: Env incorporation; HIV-1; MLV; R peptide; assembly; cytoplasmic tail; fusogenicity; glycoprotein; retrovirus.

FIG 1

Mutations in the Env cytoplasmic tail cause virus-specific incompatibility. (A) Schematic of Env cytoplasmic tail variants. Numbering follows from the N terminus of MLV Env TM. (B) Relative percent infectivity with different Env variants normalized to that of MLV Env. Averages and standard deviations from three independent experiments are shown. (C) Relative percent infectivity with various mutated Env proteins normalized to that of MLV Env. Numbering follows from the N terminus of MLV Env TM. Averages and standard deviations from three independent experiments are shown. **, P < 0.01; *, P < 0.05 (by unpaired and two-tailed Student's t test for each Env).

FIG 2

GaLV incompatibility with HIV-1 particles is modulated by residues 171N, 178K, and 179I. (A) Amino acid sequences of MLV and GaLV Env cytoplasmic tails. Chimeric mutations are indicated in boldface. (B) Relative percent infectivity with different chimeric Env proteins. (C) Comparison of amino acid sequences between MLV and GaLV in the region of residues 170 to 186. Variable amino acids are indicated in gray font. (D) Relative percent infectivity of Env point mutants. (E) Relative percent infectivity of MLV Env combination mutants. MLV Env proteins with K171N, Q178K, and A179I mutations (MLV NKI) and the V170I additional mutation (MLV INKI) are indicated. Infectivity is normalized to that of MLV Env. Averages and standard deviations from three independent experiments are shown. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns, not significant (by unpaired and two-tailed Student's t test for each Env).

FIG 3

Env incompatibility is expression level independent. Relative percent infectivity with increasing amounts of Env (20 ng, 80 ng, 320 ng, and 1,280 ng), normalized to that with 20 ng of MLV Env is shown. Averages and standard deviations from three independent experiments are shown.

FIG 4

I170P Env is incorporation defective in MLV particles, but MLV Env NKI is maturation defective in HIV particles. (A) Western blotting was performed with purified virus on SU (HA), p15E TM, and CA. Relative percent infectivity normalized to that of MLV/GaLV Δ8 is shown from the same viral medium used for Western blotting. Shown is a representative example of three independent experiments. (B) Virus capture assay with HIV-1 and MLV iGluc vectors and GFP-tagged Envs. Viruses in the collected supernatant were captured with GFP-antibody coated on ELISA plates, washed, and infected with 293FT cells with cocaptured VSV-G, and Gaussia luciferase activity was measured. Env incorporation was represented by signal from the infected cells with captured virus. Each sample was normalized to level of the positive control (straight infection without wash). Averages and standard deviations from three independent experiments are shown.

FIG 5

I170P Env incompatibility is partially alleviated by truncation of the Env CT. (A) Schematic diagram with amino acid sequences with Env truncations. (B) Relative percent infectivity normalized to that of MLV/GaLV Δ8. The average and standard deviation from three independent experiments for each are shown. (C) Western blot of purified HIV-1 and MLV particles with the indicated Env shows Env incorporation and capsid release. Relative percent infectivity of the medium used in Western blotting, normalized to that of MLV/GaLV Δ8, is shown. Shown is a representative example of two independent experiments.

FIG 6

The I170P Env incompatibility is partially alleviated by replacing MLV MA. (A) Schematic diagram of chimeric MLV Gag constructs used. (B) Western blot of purified HIV-1 and MLV particles with the indicated Env shows Env incorporation and capsid release. Averages and standard deviations from six independent experiments including infectivity from medium used for Western blotting are shown. The Western blot image is a representative example of two independent experiments. (C) Relative percent infectivity normalized to that of HIV-1 with MLV/GaLV Δ16. Averages and standard deviations from three independent experiments are shown. (D) Virus capture assay with MLV, hMA-MLV, and Src-MLV Gag with iGluc vectors and GFP-tagged Envs. Env incorporation was represented by signal from the infected cells with captured virus. Each sample was normalized to the level of positive control (straight infection without wash). Averages and standard deviations from three independent experiments are shown.

FIG 7

I170P Env is fusogenic out of virions but not within MLV particles. (A) Cell-to-cell fusogenic activity of indicated Env proteins relative to that of MLV Δ26 Env. (B) Virus-to-cell fusogenic activity of the indicated Env proteins relative to that of wild-type MLV Env. Env incorporation in MLV Gag-Cre virus was determined by Western blotting. A representative example of two Western blots is shown. Fusion assays represent the averages and standard deviations from at least three independent experiments. **, P < 0.01; ***, P < 0.001; ns, not significant (by unpaired and two-tailed Student's t test).

FIG 8

Hydrophobicity is important at the 170th residue of TM. (A) Schematic diagram of Env CT with mutations. Below is the comparison of the CTs from different gammaretroviruses. (B) Relative percent infectivity with mutant Env normalized to that of MLV/GaLV Δ8. (C) Relative percent infectivity of HIV-1, MLV, and RSV normalized to that of MLV/GaLV Δ20 with HIV-1. The average and standard deviation from three independent experiments for each are shown. FeLV, feline leukemia virus; RD114, a feline endogenous retrovirus; XMRV, xenotropic murine leukemia virus-related virus.

FIG 9

The 170th residue in MLV/GaLV Env CT is critical for MLV infectivity. Representative Env incorporation with the indicated Env from three independent experiments is shown. Average relative infectivity normalized to that of MLV/GaLV Δ26 (170I*) and standard deviations from three independent experiments are shown. ****, P < 0.0001; ns, not significant (by unpaired and two-tailed Student's t test).