Riboflavin transporter: evidence of a role as entry receptor for chimpanzee endogenous retrovirus

Abstract

Endogenous retroviruses (ERVs) are remnants of ancestral viral infections in germ cells that constitute a substantial proportion of the mammalian genome and are assumed to provide molecular fossil records of ancient infections. Analysis of these sequences may reveal the mechanisms of virus-host co-evolution, viral endogenization, and extinction. Chimpanzee endogenous retrovirus 1 (CERV1), a gamma retrovirus, is estimated to have circulated within primates for ~10 million years, although it is now apparently extinct. In this study, we aimed to gain an understanding of how the extinct CERV1 was transmitted and endogenized. On the basis of the identification of CERV1 fossils in the primate genome and using the expression-cloning method with the human cDNA library, we found that riboflavin transporter human SLC52A2 served as a receptor for CERV1 entry. The ectopic expression of human and chimpanzee SLC52A2 and its related SLC52A1 in heterogenic cells confers susceptibility to infection by CERV1 and porcine endogenous retrovirus (PERV). Virus interference experiments have shown that CERV1 inhibits infection by PERV and vice versa. This finding indicates that CERV1 and PERV belong to the same virus interference group. CERV1 shows infection in a wide range of human and primate cells. Notably, CERV1 infection is observed in human cell lines that express human SLC52A2 abundantly but hardly express human SLC52A1. Although CERV1 has been established to be present at high copy numbers in the great apes (Pan troglodytes, Pan paniscus, and Gorilla gorilla) and 15 Old World monkey species of the Cercopithecinae and Colobinae subfamilies, it is absent in humans and orangutans. CERV1 gene expression is observed in primates, including chimpanzees, suggesting that CERV1 has co-evolved with its hosts. Our results suggest that ERVs may have conferred resistance to viral infections in a convergent evolutionary manner. These findings are significant not only for advancing the field of paleovirology but also in terms of gaining an understanding of the potential risks of viral infection with respect to xenotransplantation, such as that from pigs to humans.

Keywords: CERV1; endogenous retrovirus; paleovirology; riboflavin transporter.

Figure 1

Infection of CERV1, PERV-A, and PERV-A/C. (A) Infection of LacZ-coding CERV1, PERV-A, and PERV-A/C Env-pseudotyped viruses in MDTF-huSLC52A2, MDTF-huSLC52A1, MDTF-chSLC52A2, MDTF-chSLC52A1, and MDTF cells. (B) Infection of LacZ-coding CERV1 10346 and CERV1-10346 L241P Env-pseudotyped viruses in MDTF-huSLC52A2, MDTF-huSLC52A1, MDTF-chSLC52A2, MDTF-chSLC52A1, and MDTF cells. (C) Alignment of the amino acid sequences of Envs of CERV1, CERV1-10346, and CERV1-10346 L241P. Envs are shown using single-letter amino acid codes. Leucine (L) and proline (P) at position 241 in CERV1 Env are indicated for comparison. Dots, conserved amino acid residues; SU, surface unit; TM, transmembrane domain. The data represent the averages of three independent experiments, with the standard deviations shown (A and B).

Figure 2

Interference group comparison of CERV1, PERV-A, and PERV-A/C. Infection of CERV1, PERV-A, and PERV-A/C Env-pseudotyped viruses in (A) PK15 cells, (B) HEK293T cells transiently expressing CERV1 Env (293/CERV1 cells) or empty plasmid (293/empty cells), and (C) HEK293T cells persistently infected with PERV-A/C (293T//PERV-A/C inf. cells) or 293T cells. Comparisons with HEK293T/empty or control cells were performed using Student’s t-test (^*P < .01) in (B, C). The y-axis indicates infectious units per millilitre of virus (A, C) or relative virus titre (%) (B). The infectious titres for 293T/empty cells (B) are as follows. CERV1: (1.38 ± 0.62) × 10⁵ IU/ml PERV-A: (8.64 ± 4.36) × 10³ IU/ml, PERV-A/C: (2.09 ± 0.12) × 10⁵ IU/ml, Ampho-MLV: (9.16 ± 0.52) × 10⁴ IU/ml. The data represent the averages of three independent experiments, with the standard deviations shown.

Figure 3

Comparison of the infectivity of CERV1, PERV-A, and PERV-A/C in human and primate cell lines. The infectivity of CERV1, PERV-A, and PERV-A/C Env-pseudotyped viruses was assessed in human and primate cell lines. The data represent the averages of three independent experiments, with the standard deviations shown.

Figure 4

Expression of SLC52A1 and SLC52A2 in human and non-human primate cell lines. Quantification of SLC52A2 (A) and SLC52A1 (B) transcripts by quantitative RT-PCR in cell lines. The x-axis indicates the analysed samples. The y-axis indicates the expression level normalized to the expression of GAPDH. In each experiment, the normalized expression in HeLa cells is set to 1 for comparison across cell lines. (C) Expression of SLC52A1 and SLC52A2 in cell line. The x-axis indicates the analysed samples. The y-axis indicates TPM.

Figure 5

Expression of SLC52A1 and SLC52A2 genes in human and non-human primate tissues. (A–D) Bar graphs showing the expression levels of the SLC52A1 and SLC52A2 genes. The x-axis represents the tissues investigated, and the y-axis shows the normalized expression levels of SLC52A1 and SLC52A2 in TPM. N.A. indicates that expression levels in the specified tissue were not investigated due to the unavailability of RNA-seq data.

Figure 6

Phylogenetic analysis using the CERV1 Env amino acid sequences. The phylogenetic tree for the CERV1 Env amino acid sequences was constructed based on the neighbour-joining method. Bootstrap values >70% are shown adjacent to the respective branches. The tree is drawn to scale, with branch lengths representing the number of substitutions per site. The CERV1 env type is categorized as Clade 1, Clade 2, Clade 3, and Clade 4. The asterisk indicates the CERV1 Env reference sequence used in this study (Soll et al. 2010).

Figure 7

Evolutionary history of primate species with the integration timings of CERV1. The arrowheads indicate integration timings of the ancestor with CERV1. The red colour indicates the species infected with CERV1.

Figure 8

Expression of CERV1 Env in primate tissues. Box plot showing the CERV1 Env gene expression levels, analysed using publicly available RNA-seq data from P. troglodytes, M. fascicularis, and M. mulatta. Box-and-whisker plots depict the median as a line in the centre of a box bounded by the first and third quartiles, with whiskers to 1.5 times the interquartile range and outlier values shown as individual dots.