Retroviral host range extension is coupled with Env-activating mutations resulting in receptor-independent entry

Abstract

The extent of virus transmission among individuals and species is generally determined by the presence of specific membrane-embedded virus receptors required for virus entry. Interaction of the viral envelope glycoprotein (Env) with a specific cellular receptor is the first and crucial step in determining host specificity. Using a well-established retroviral model-avian Rous sarcoma virus (RSV)-we analyzed changes in an RSV variant that had repeatedly been able to infect rodents. By envelope gene (env) sequencing, we identified eight mutations that do not match the already described mutations influencing the host range. Two of these mutations-one at the beginning (D32G) of the surface Env subunit (SU) and the other at the end of the fusion peptide region (L378S)-were found to be of critical importance, ensuring transmission to rodent, human, and chicken cells lacking the appropriate receptor. Furthermore, we carried out assays to examine the virus entry mechanism and concluded that these two mutations cause conformational changes in the Env variant and that these changes lead to an activated, or primed, state of Env (normally induced after Env interaction with the receptor). In summary, our results indicate that retroviral host range extension is caused by spontaneous Env activation, which circumvents the need for original cell receptor. This activation is, in turn, caused by mutations in various env regions.

Keywords: Rous sarcoma virus; envelope glycoprotein; receptor-independent entry; retrovirus; virus entry.

Fig. 1.

Virus transmission to mammalian cells and envelope glycoprotein alteration. (A) Schema of rat and hamster infection with original PR-RSV-C virus. PR-RSV-C virus was injected into a chicken, the resulting tumor was minced and injected into a rat, which developed a tumor caused by XC-RSV. The virus rescued from XC cells was again multiplied in a chicken and injected into a Syrian hamster, where H20 tumor harboring H20-RSV was obtained (for details see Materials and Methods). (B) Diagram of the envelope glycoprotein domain structure with positions of described mutations responsible for mammalian tropism (black arrows) (14, 16) and found amino acid mutations (red arrows). Host-range regions (hr1/2) and variable regions (vr1/2/3) in the surface subunit (SU) are depicted in gray boxes. The fusion peptide (FP), heptad repeats (HR1/2), and membrane spanning domain (MSD) in the transmembrane subunit (TM) are shown in white boxes. (C) Comparison of gp85 amino acid sequences from original parental virus PR-RSV-C (sequence from GenBank V01197.1) (17) and viruses rescued from RSV-transformed rat (XC) or hamster (H20) cells. h and vr regions in SU are depicted in gray boxes. FP, HR1, HR2, and MSD in TM are shown in white boxes. Their position is depicted according to refs. , .

Fig. 2.

Viruses equipped with EnvH20 are able to infect hamster cells. (A) Amount of newly made env viral DNA in infected cells (NIL or NIL-Tvc) was quantified in different time points postinfection. The results were normalized to chGAPDH expression and are presented relative to the sample of NIL-Tvc infected PR-RSV-C on day 1. (B) Schematic diagram of chimeric reporter vector RCAS-EnvH20-GFP prepared from RCAS-EnvC-GFP and H20-RSV virus. Restriction sites and position of primers used for the env gene substitution are depicted. (C) Infection of NIL and NIL-Tvc cells with RCAS-GFP viruses containing EnvC or EnvH20 was scored by flow cytometry 3 d later. Titers in GFP-transducing units were determined as described in SI Materials and Methods. Error bars show the SD of two independent experiments in parallel. The limit of detection (signal from uninfected NIL cells) is marked with a dashed line. Significant differences are marked by asterisks (***P < 0.001). (D) Relative infectivity of NIL cells infected with RCAS-EnvC-GFP and RCAS-EnvH20-GFP. Infectivity is expressed as the percentage of the viral titer on NIL-Tvc cells. (E) Examples of FACS diagrams showing the percentages of GFP⁺ cells after infection of NIL cells with RCAS-EnvC-GFP and RCAS-EnvH20-GFP viruses.

Fig. 3.

Mutations in the first part of SU (D23G) and in the fusion peptide (L378S) are responsible for the virus-extended host range. (A) Schematic diagram showing restriction sites (dashed lines) that were used to divide EnvH20 into four parts, which were tested separately as chimeras with EnvC. Positions of the found amino acid mutations are shown with red arrows. (B and D) Infection of hamster cell lines NIL and NIL-Tvc (B), human cell lines HEK293, RPE1-hTERT, and chicken L15 cell line (Tvc⁻) (D) with RCAS-GFP viruses harboring EnvC-H20 chimeras and RCAS-EnvC-GFP single mutants L378S, G464S, and G464S was scored by flow cytometry 2 (HEK293) or 3 (NIL, RPE1-hTERT, L15, NIL-Tvc) days later. Titers were determined as described in SI Materials and Methods. The limit of detection (signal from uninfected cells) is marked with a dashed line. (C) Relative infectivity of hamster NIL cells with different RCAS-GFP viruses is expressed as the percentage of the viral titer on NIL-Tvc cells. Error bars show the SD of two independent experiments in parallel. Significant differences are marked by asterisks (*P = 0.05–0.01, **P = 0.01–0.001, ***P < 0.001). NS, not significant.

Fig. S1.

Virus harboring EnvH20 has a lower titer than the virus with original EnvC despite the same RT activity and VLP formation. (A) Infection of NIL-Tvc and DF-1 cells with RCAS-EnvC-GFP and RCAS-EnvH20-GFP viruses was scored by flow cytometry 2 (DF-1 cell) or 3 (NIL-Tvc) days later. Titers were determined as described in Materials and Methods. The results are presented as titers relative to RCAS-EnvC-GFP. (B) RT activity of RCAS-EnvC and RCAS-EnvH20-GFP was measured using the PERT assay described in SI Materials and Methods. The results of the PERT assay are presented relative to RCAS-EnvC-GFP. (C) VLP production of DF-1 cells infected with RCAS-EnvC-GFP or RCAS-EnvH20-GFP was estimated by Western blot analysis. Expression of Gag product p27 in the medium was determined with anti-p27 antibody.

Fig. 4.

EnvH20 has similar features as a receptor-primed Env. (A) The virus with EnvH20 is thermosensitive. Virus stocks of similar titers were incubated at 44 °C for the indicated time. The titers were determined on NIL-Tvc 3 d after infection using flow cytometry. The amount of proviral DNA in infected DF-1 cells was measured 1 d after infection using RT-PCR. The results are presented as the percentage of the original titer (full lines) or the amount of DNA (dashed lines) remaining after heating. (B) The virus with EnvH20 is inactivated by low pH. Purified viruses of similar titers were incubated with low (pH 5) or neutral (pH 7.5) treatment at 37 °C for 30 min before infection of DF-1 cells. The titers were determined by flow cytometry 2 d later. (C) Formation of TM oligomers triggered by increasing temperature. The virus was incubated for 20 min at the indicated temperature. Samples were lysed in Laemmli loading buffer and analyzed by SDS/PAGE without boiling. TM was detected by immunoblotting with an antibody against its C-terminal part. Only the 70-kDa isoform of TM is shown. (D) Formation of TM oligomers at low pH. The virus was incubated for 30 min at the indicated pH at room temperature. Samples were neutralized, lysed in Laemmli loading buffer, and analyzed by SDS/PAGE without boiling. TM was detected by immunoblotting with an antibody against its C-terminal part. (E) The virus with EnvH20 binds liposomes. Viruses of similar titers were incubated with or without liposomes at 37 °C and then the mixture was separated by sucrose gradient centrifugation. Sample fractions were collected and analyzed by Western immunoblotting using anti-p27 antibody. T, top; B, bottom. (F) The virus containing EnvH20 is inhibited with PMB. Purified viruses of similar titers were incubated at 37 °C for 30 min with increasing concentration of PMB. Viruses were diluted in medium and spinoculated on DF-1 cells. The percentage of GFP⁺ cells was measured by flow cytometry 2 d after infection. Titers below the limit of detection are marked with “<”. Error bars show the SD of the two independent experiments in parallel. Significant differences are marked by asterisks (*P = 0.05–0.01, **P = 0.01–0.001, ***P < 0.001). NS, not significant.

Fig. 5.

The level of Env activation correlates with the efficiency of receptor-independent entry. (A) Virus inactivation by low pH. Purified viruses harboring EnvC-H20 chimeras were incubated with low (pH 5) or neutral (pH 7.5) treatment at 37 °C for 30 min before infection of DF-1 cells. The titers were determined by flow cytometry 2 d later. (C) PMB inhibition of virus infection. Purified viruses harboring EnvC-H20 chimeras were incubated at 37 °C 30 min with increasing concentration of PMB. The titers were determined on DF-1 cells by flow cytometry 2 d after infection. Titers below the limit of detection are marked with “<”. Error bars show the SD of two independent experiments in parallel. (B and D) Correlation between relative NIL infectivity and virus inactivation at low pH (B) or PMB inhibition of virus infection (expressed as residual infectivity after 1 mM PMB) (D).

Fig. 6.

Mutation in TM facilitates entry into mammalian cells independent of SU. (A) Relative infectivity of human RPE1-hTERT cells with RCAS-EnvC-GFP, RCASBP-B-GFP, and RCAS-GFP viruses bearing chimeric Envs (SU-C or SU-B with TM from EnvH20) is expressed as the percentage of the viral titer on DF-1 cells. The percentage of GFP⁺ cells was scored by flow cytometry 2 (DF-1) or 3 (RPE1-hTERT) days after infection. (B) PMB inhibition of virus infection. Purified viruses were incubated at 37 °C for 30 min with increasing concentration of PMB. The titers were determined on DF-1 cells by flow cytometry 2 d after infection. Titers below the limit of detection are marked with “<”. Error bars show the SD of two independent experiments in parallel. Significant differences are marked by asterisks (**P = 0.01–0.001, ***P < 0.001). NS, not significant.

Fig. 7.

Proposed model of receptor-independent entry. (Left) Viral Env is normally activated by binding with the receptor. Activated Env takes a conformation enabling further changes that occur at low pH in endosomes and lead to viral and cell membrane fusion. (Middle) Mutations D32G and L378S change EnvC conformation and shift Env close to the active state, which normally follows receptor priming. These viruses are therefore able to be spontaneously activated with efficiency, depending on the mutation type. After low pH exposure, activated Env facilitates viral and cell membrane fusion. (Right) Consequences of Env activation can be observed in different steps. The level of Env activation positively correlates with increased formation of reactive thiolate and decreased stability at low pH. After activation, the conformation is changed: Heptad repeats are exposed and the fusion peptide is inserted into the liposome membrane. After low pH treatment, formation of the six-helix bundle can be determined by TM oligomerization assay.