FeLIX is a restriction factor for mammalian retrovirus infection

Abstract

Endogenous retroviruses (ERVs) are remnants of ancestral viral infections. Feline leukemia virus (FeLV) is an exogenous and endogenous retrovirus in domestic cats. It is classified into several subgroups (A, B, C, D, E, and T) based on viral receptor interference properties or receptor usage. ERV-derived molecules benefit animals, conferring resistance to infectious diseases. However, the soluble protein encoded by the defective envelope (env) gene of endogenous FeLV (enFeLV) functions as a co-factor in FeLV subgroup T infections. Therefore, whether the gene emerged to facilitate viral infection is unclear. Based on the properties of ERV-derived molecules, we hypothesized that the defective env genes possess antiviral activity that would be advantageous to the host because FeLV subgroup B (FeLV-B), a recombinant virus derived from enFeLV env, is restricted to viral transmission among domestic cats. When soluble truncated Env proteins from enFeLV were tested for their inhibitory effects against enFeLV and FeLV-B, they inhibited viral infection. Notably, this antiviral machinery was extended to infection with the Gibbon ape leukemia virus, Koala retrovirus A, and Hervey pteropid gammaretrovirus. Although these viruses used feline phosphate transporter 1 (fePit1) and phosphate transporter 2 as receptors, the inhibitory mechanism involved competitive receptor binding in a fePit1-dependent manner. The shift in receptor usage might have occurred to avoid the inhibitory effect. Overall, these findings highlight the possible emergence of soluble truncated Env proteins from enFeLV as a restriction factor against retroviral infection and will help in developing host immunity and antiviral defense by controlling retroviral spread.IMPORTANCERetroviruses are unique in using reverse transcriptase to convert RNA genomes into DNA, infecting germ cells, and transmitting to offspring. Numerous ancient retroviral sequences are known as endogenous retroviruses (ERVs). The soluble Env protein derived from ERVs functions as a co-factor that assists in FeLV-T infection. However, herein, we show that the soluble Env protein exhibits antiviral activity and provides resistance to mammalian retrovirus infection through competitive receptor binding. In particular, this finding may explain why FeLV-B transmission is not observed among domestic cats. ERV-derived molecules can benefit animals in an evolutionary arms race, highlighting the double-edged-sword nature of ERVs.

Keywords: endogenous retrovirus; feline leukemia virus; restriction factor.

Fig 1

Endogenization of enFeLV in domestic cats. (A) Phylogenetic tree of enFeLV carrying the env gene based on LTRs. Intact Env, FeLIX, and Trunc-C4 are indicated by green, red, and blue circles, respectively. (B) Chromosome position of enFeLV proviral integration in domestic cat genome. The number of viral integrations is indicated in the top bar. (C) Integration time of enFeLV proviruses in domestic cat genome. (D) Structure of enFeLV Env classification (four groups). Group 1: intact, intact full-length Env has approximately 666 amino acids (aa). Group 2: like-Trunc-C4, defective Env like-Trunc-C4 (length > 300 aa), which has signal peptide (SP), variable region A (VRA), and variable region B (VRB) in Env surface unit (SU) but lacks transmembrane unit (TM) region. Group 3: like-FeLIX (length < 300 aa), which has only SP, VRA, and VRB in Env SU. Group 4: Env has a length between 140 and 273 aa and lacks VRA, VRB, or the TM region.

Fig 2

Infectivity and receptor usage of enFeLV Env-pseudotyped viruses. (A) Infectivity of enFeLV (clone1, clone2, and clone3) and FeLV-B (GA, MZ40-5B, KG20-5B, FO36-5B, and ON-T) Env-pseudotyped viruses. (B) Infectivity of Env mutants (E345G and N394K) from enFeLV clone1. (C) Western blot analysis of enFeLV Envs proteins in HEK293T cells using anti-FeLV gp70 and anti-FeLV p15e antibodies. Actin was used as a control. (D) Infection assay of enFeLV (clone1/WT, clone1 E345G, clone2, clone3, and AGTT) in Mus dunni tail fibroblast (MDTF)-fePit1, MDTF-fePit2, and MDTF (empty vector) as target cells for receptor usage. (E) Infection assay of FeLV-B (GA, MZ40-5B, KG20-5B, FO36-5B, and ON-T) and FeLV-B mutants (GA Q73R, FO36-5B K66D, and ON-T R73Q) in MDTF-fePit1, MDTF-fePit2, and MDTF (empty vector) as target cells for receptor usage. (F) Interference assay of FeLV-B/GA and FeLV-B/ON-T. AH927 cells pre-infected with either FeLV-B/GA or FeLV-B/ON-T were infected by the Env-pseudotyped viruses. FeLV-C/Sarma and MuLV 4070A were used as controls. Viral titers are indicated on the x-axis. The infectious units were determined by counting the number of log₁₀-galactosidase (LacZ)-positive cells per milliliter of virus indicated on the y-axis. Virus infection titers with standard deviations were averaged from three independent experiments. Mock represents the negative control.

Fig 3

Structural schematic and expression of truncated Envs. (A) Schematic representation of the structure of FeLIX-N249 (asparagine at position 249), FeLIX-D249 (aspartic acid at position 249), and Trunc-C4. The RBD of FeLV-B/GA with myc-tag is also shown. PRR, proline-rich region. The number of amino acids is indicated on the right side. (B) Detection of proviruses encoding FeLIX in the domestic cat (n = 22) and European wild cat (n = 9) genome. (C and D) FeLIX expression in feline tissues and cell lines. Quantification of feline FeLIX transcripts using quantitative RT-PCR in feline tissues and cell lines. The x-axis indicates the analyzed samples. The y-axis indicates the expression level normalized to the expression of peptidylprolyl isomerase A (PPIA). Normalized expression in PBMCs and 3201 cells is shown as 1 in feline tissues and cell lines, respectively. LN, lymph node. Expression of FeLIX-N249 and D-249 was determined via RT-PCR and sequencing in feline tissues (spleen, bone marrow, mandibular LN, and lung) and indicated feline cell lines. (F) Detection of FeLIX-N249 and D-249, Trunc-C4, and Fe-B-RBD in cell culture supernatants and (G) in cell lysates from HEK293T cells transfected with indicated plasmids. (H) Detection of FeLIX in the supernatant of indicated cells. Red asterisks indicate truncated Env proteins. FeLIX-N249 and D-249, as well as Trunc-C4, were analyzed via immunoprecipitation and western blotting using a goat anti-FeLV gp70 antibody (81-S-210-2, NCI), while Fe-B-RBD was assessed with an anti-Myc antibody.

Fig 4

Inhibitory effect of the truncated Env proteins derived from enFeLV for enFeLV and FeLV-B infection and restriction mechanism. (A) Inhibition assay using FeLIX-N249 and FeLIX-D249 for evaluating the infection of Env-pseudotyped viruses, enFeLV (clone1 E345G, clone2, clone3, and AGTT), (B) FeLV-B (GA, MZ40-5B, KG20-5B, FO36-5B, and ON-T), and FeLV-B mutants (GA Q73R, FO36-5B K66D, and ON-T R73Q) in AH927 cells. (C) Dose-dependent inhibition of FeLIX for Env-pseudotyped viral infection (FeLV-B/ON-T, FeLV-B/GA, and enFeLV-AGTT) in AH927 cells. (D) The replication-competent viruses assessed included FeLV-B/GA, FeLV-B/ON-T, and FeLV-A carrying the enFeLV-AGTT env gene, and FeLV-A/61E in AH927 cells. Restriction mechanism in (E and F) MDTF-fePit1 and (G) MDTF-fePit2 cells. FeLV-A/clone33, FeLV-C/Sarma, and FeLV-E/TG35-2 were also used in this assay. FeLIX-N249, FeLIX-D249, FeB-RBD, and the empty vector/mock were sourced from the supernatants of HEK293T cells transfected with their respective expression vectors. Each supernatant was added to the culture for 2 h. Subsequently, cells were infected with the Env-pseudotyped virus. The infectious units (IU) shown on the x-axis were determined by counting the number of log₁₀-galactosidase (LacZ)-positive cells per milliliter of virus indicated on the y-axis. Virus infection titers with standard deviations represent the means of three independent infection experiments. Comparisons were performed using the Student’s t-test (*P < 0.01).

Fig 5

Inhibitory effect of FeLIX from the supernatant of 3201 cells on enFeLV and FeLV-B infection. (A) Inhibition assays of Env-pseudotyped enFeLV and FeLV-B viruses were conducted using the culture supernatant from 3201 cells in AH927 cells as target cells. (B) The culture supernatants from the 3201 cells were treated with either goat anti-FeLV gp70 antibody or normal goat serum (control), after which the culture supernatant from which FeLIX was removed was used for the inhibition assays for enFeLV and FeLV-B. (C) The culture supernatant from the 3201 cells and the culture supernatant from which FeLIX was removed were used for the infection assay for FeLV-T. The infectious units were determined by counting the number of log₁₀-galactosidase (LacZ)-positive cells per milliliter of virus (x-axis). Virus infection titers with standard deviations represent the means of three independent experiments. Medium represents the negative control. Comparisons were performed using the Student’s t-test (*P < 0.01).

Fig 6

Inhibitory effect of FeLIX for non-feline mammalian retrovirus infection. (A) Infection of non-feline mammalian retroviruses (GaLV, KoRV, HPG, and MuLV 4070A) for feline, AH927, and CRFK cells; (B) for receptor usage in MDTF/fePit1, MDTF/fePit2, and MDTF/empty vector. Supernatants from HEK293T cells transfected with expression vectors encoding FeLIX-N249 and FeLIX-D249 were used for inhibition assays against non-feline mammalian retroviruses (GaLV, KoRV, HPG, and MuLV 4070A) in (C) AH927 cells, (D) MDTF-fePit1 cells, and (E) MDTF-fePit2 cells. (F) Dose-dependent inhibition of FeLIX for KoRV Env-pseudotyped viral infection in AH927 cells. (G) Inhibition assays of Env-pseudotyped virus of KoRV, GaLV, HPG, and 4070A ampho-MuLV were conducted using the culture supernatant from 3201 cells in AH927 cells. (H) The culture supernatants from the 3201 cells were treated with either goat anti-FeLV gp70 antibody or normal goat serum (control). Subsequently, the culture supernatant from which FeLIX was removed was used for the inhibition assays for KoRV, GaLV, and HPG infection. The infectious units were determined by counting the number of log₁₀-galactosidase (LacZ)-positive cells per milliliter of the virus. The viral titers are illustrated as the log number of infectious units per milliliter. Virus infection titers with standard deviations represent the means of three independent experiments. Mock represents the negative control. Comparisons were performed using the Student’s t-test (*P < 0.01).