The mouse IAPE endogenous retrovirus can infect cells through any of the five GPI-anchored Ephrin A proteins

Abstract

The IAPE (Intracisternal A-type Particles elements with an Envelope) family of murine endogenous retroelements is present at more than 200 copies in the mouse genome. We had previously identified a single copy that proved to be fully functional, i.e. which can generate viral particles budding out of the cell and infectious on a series of cells, including human cells. We also showed that IAPE are the progenitors of the highly reiterated IAP elements. The latter are now strictly intracellular retrotransposons, due to the loss of the envelope gene and re-localisation of the associated particles in the course of evolution. In the present study we searched for the cellular receptor of the IAPE elements, by using a lentiviral human cDNA library and a pseudotype assay on transduced cells. We identified Ephrin A4, a GPI-anchored molecule involved in several developmental processes, as a receptor for the IAPE pseudotypes. We also found that the other 4 members of the Ephrin A family -but not those of the closely related Ephrin B family- were also able to mediate IAPE cell entry, thus significantly increasing the amount of possible cell types susceptible to IAPE infection. We show that these include mouse germline cells, as illustrated by immunohistochemistry experiments, consistent with IAPE genomic amplification by successive re-infection. We propose that the uncovered properties of the identified receptors played a role in the accumulation of IAPE elements in the mouse genome, and in the survival of a functional copy.

Figure 1. Screening strategy and identification of cDNA clones for the IAPE receptor.

A. Scheme of the screening strategy to identify the protein used as a receptor by the IAPE Env. More technical details are provided in the main text and in the Methods section. LV: lentiviral vector. B. Scheme representing the cellular localisation of the proteins encoded by the two candidate receptor genes. TMEM9 is an endosomal protein, whereas Ephrin A4 (encoded by the EFNA4 gene) is expressed at the cell membrane where it can interact with extracellular molecules, including the EphA proteins. C. Quantification of the mRNA level of the two genes identified as potential receptors for the IAPE elements. The amount of the corresponding mRNAs was measured by qRT-PCR performed on total RNA extracted from the parental Vero cells as a control, or the five clones found to be the most sensitive to infection by IAPE pseudotypes (re-numbered from 1 to 5 in this panel). Clones 1,4 and 5 contained an EFNA4 cDNA, and clones number 2 and 3 a TMEM9 cDNA. The mRNA levels were normalised using the RPLO gene as a reference, and the parental Vero cell line level was set at 1 for both genes.

Figure 2. Expression of EFNA4 is sufficient to render cells susceptible to infection by IAPE Env pseudotypes.

Following the screening strategy presented in Figure 1, EFNA4 was identified as a potential receptor for the IAPE Env. (A) To confirm this hypothesis, its cDNA (Hs_EFNA4 for Homo sapiens EFNA4) was re-cloned in a lentiviral vector (LV) and introduced into non-permissive Vero and WOP cells that were then subjected to infection with IAPE Env pseudotypes containing either the GFP or lacZ genes. (B) For the lacZ-containing pseudotypes, the cells were fixed and stained with X-Gal to reveal ß-galactosidase activity 3 days post infection. A photo of one representative field for each condition is presented. Note than in the case of the VSV-G pseudotypes, the supernatant was diluted 200 fold before its use for infection. (C) For the GFP-containing pseudotypes, the target cells were collected 3 days post infection and subjected to FACS analysis in order to quantify the proportion of GFP-positive cells, allowing precise calculation of the viral titres (see Methods for details). The results presented in B and C correspond to one representative experiment out of 3.

Figure 3. Ephrin A4 is a bona fide receptor for the IAPE Env.

(A) WOP cells transduced with the EFNA4 gene or a control gene were stained with the soluble His-tagged IAPE envelope SU subunit (or that of syncytin1 (syn1) as a control) followed by an Alexa488 anti-His antibody and subjected to FACS analysis. Only the cells transduced with EFNA4 bind the IAPE SU protein, and none of the cells were stained with the control syncytin1-SU. The expression level of Ephrin A proteins in the two populations was checked using the EphA2-Fc soluble protein (that can bind all Ephrin A proteins) and are shown in the small panels on the right. The data presented correspond to one representative experiment out of three. (B) 293T cells were transiently transfected with an expression vector for IAPE Env, or Ampho MLV Env as a control. At day 2 post transfection, cells were stained with a soluble Ephrin A4-Fc fusion protein, or a control Fc protein, followed by an Alexa 488 fluorescent anti-Fc antibody before being subjected to FACS analysis. Only the cells expressing the IAPE Env bind to the recombinant Ephrin A4 protein, The data presented correspond to one representative experiment out of three. Expression of the two envelope proteins in the transfected cells was checked by Western blot using specific antibodies, as shown on the right. (C) The soluble recombinant IAPE or syn1 His-tagged SU proteins were tested for interaction with Ephrin A4-Fc (or Fc-only and EphA2 controls) in a pull down assay as schematised on the left. Pellets were analysed by Western blot using an antibody directed against the His tag (upper part). The only interaction detected is between Ephrin A4-Fc and IAPE SU. We ensured that the 2 His-tagged SU proteins were produced in similar amount, as shown on the right (Before IP panel) and that all 3 Fc-tagged proteins were efficiently pulled down by the protein A-agarose beads and recovered in similar amount in the pellet, as shown on the lower panel (Western blot performed with an anti mouse IgG antibody).

Figure 4. Characterisation of several Ephrin A proteins as receptors for the IAPE Env protein.

(A) Comparison of the efficacy of human (Hs), mouse (Mm) and rat (Rn) Ephrin A4 proteins as receptors for IAPE Env pseudotypes. WOP cells were transduced with a lentiviral vector encoding one of the 3 versions of EFNA4, or a control gene. Three days later, they were challenged with GFP-marked lentiviral particles pseudotyped with the IAPE Env, the Friend ecotropic Env or no Env (none). Viral titres were measured by FACS three days post infection. The results (mean titre ± standard deviation) are from 5 independent experiments. (B) Test of the mouse Ephrin A and Ephrin B proteins for their activity as receptors for the IAPE Env pseudotypes. WOP cells were transduced with lentiviral vectors containing the different EFNA/B cDNA or a control gene, and tested for their ability to be infected with IAPE Env pseudotypes as described in (A). Results (mean ± standard deviation calculated from 3–4 independent experiments) are expressed as the percentage of the titre measured for each gene as compared to that obtained with the human EFNA4 (Hs_EFNA4, shown on the right). Asteriks indicate values significantly different (p<0.05) from that obtained with the untreated cells (unpaired Student test).

Figure 5. Expression of the Ephrin A proteins in mouse ovary and testis detected by immunohistochemistry.

Ephrin A proteins in these serial cryosections were labelled using a commercial soluble EphA7-Fc tagged protein (a soluble form of EphA7 that can bind all-Ephrin A proteins) (B and D), or a Fc-only control protein (A and C). A and B show high magnifications of a Graäfian follicle with the oocyte (filled arrowheads) surrounded by follicle cells (open arrowheads). C and D show cross-sections of seminiferous tubules, with spermatozoa in the central lumen. Filled arrowheads point to stained germinal cells located apart from the periphery of the tubule (inset: higher magnification of stained cells).

Figure 6. Classified retroviral receptors and dendrogram of the corresponding retroviral envelopes.

A. Scheme of the three types of proteins that are used as receptors by various retroviruses. Examples of each type are provided below, with the name of the receptors indicated on the left followed by the name of the virus(es) using them in brackets. B. Dendrogram of the envelopes listed in A. Those using receptors with multiple membrane domains are in yellow, single membrane domain in green and GPI anchor in purple.