Function of a retroviral envelope protein in the placenta of a viviparous lizard

Retroviruses are enveloped RNA viruses able to integrate their genome as a DNA copy into the genome of the infected cell. For this, two viral enzymes are required: the reverse transcriptase producing a DNA copy from the RNA genome and the integrase. When a retrovirus infects an oocyte or a sperm cell, the integrated viral DNA copy, also called provirus, will finally be integrated in the genome of all cells of the growing organism. Furthermore, the provirus will be transmitted to the progeny of this organism like a normal cellular gene. These proviruses are called endogenous retroviruses in contrast to exogenous retroviruses such as HIV-1, which infects and integrates only in special target cells but not into the germline (reviewed in ref. 1). Although these integrated proviruses still produce infectious virus particles in some species, in most cases, after millions of years of evolution, the proviruses are not replication-competent anymore due to deletions or mutations. This was the reason why it was thought for a long time that endogenous retroviruses represent junk DNA. However, it was shown recently that envelope proteins of endogenous retroviruses play an important physiological role in placentogenesis (reviewed in refs. 2, 3). Retroviruses carry two envelope proteins on their surface, the surface and transmembrane envelope proteins, which are responsible for binding to the cellular receptor and fusion of the viral membrane with the cellular membrane. In the last few years, it was shown that the envelope proteins of endogenous retroviruses were utilized for the generation of syncytiotrophoblast structures in the placenta of numerous eutherian mammals, including humans and marsupials (Table 1). Reflecting their function, they are called syncytins. In PNAS, Cornelis et al. (4) now report the detection of such an “enslaved” retroviral envelope protein in the placenta of a viviparous lizard, the Mabuya lizard. These findings are of importance in demonstrating that lizards were also able to utilize envelope proteins of endogenous retroviruses and that evolution to placentas is inseparably associated with retrovirus-derived syncytins. Based on sequence analysis of RNA expressed in the placenta of the lizard, four different syncytin-like sequences were identified, Mab-env1 to Mab-env4. According to definition, a syncytin should be expressed in the placenta, should be fusogenic, and should be highly conserved (e.g., subject to purifying selection). Since Mab-env2, Mab-env3, and Mab-env4 are only weakly expressed in the placenta, only Mab-env1 deserves the definition syncytin-Mab1.

Table 1.

Syncytins in different vertebrate species

Subdivision, infraclass, or suborder	Order, suborder, family, or genus	Example	Name of the syncytin	Refs.
Eutherian mammals (placentalia)	Tenrecidae (tenrecs)	Lesser hedgehog tenrec	Syn-Ten1	(10)
	Lagomorpha (lagomorphs)	Rabbit	Syn-Ory1	(11)
	Rodentia (rodents)	Mouse	Syn-A, Syn-B	(12)
	Haplorrhini (haplorhines)	Humans	Syn-1, Syn-2	(13, 14)
	Ruminantia (ruminants	Cow	Syn-Rum1	(15)
	Carnivora (carnivorants)	Dog, cat	Syn-Car1	(16)
Marsupials	Didelphimorphia (American marsupials)	Opossum	Syn-Opo1	(17)
Lizards	Mabuya (skinks)	Mabuya sp.	Syn-Mab1	(4)

Although the majority of lizards are oviparous (egg-laying), around 20% of them, including the Mabuya lizards, reproduce via viviparity (live birth). Embryos are nourished via a placenta-like structure. Cornelis et al. (4) show expression of syncytin-Mab1 in this placenta, and that the protein is fusogenic (e.g., able to fuse cells and to generate syncytia). The receptor of the envelope protein, which is required for the fusogenic activity of the syncytin, was identified. It is the protein MPZL1, a signal-transducing transmembrane protein involved in cell migration. MPZL1 was strongly expressed in the placenta at the maternofetal interface and colocalizes with Mab1. This indicates that Mab1 and MPZL1 could interact and contribute to the maternal syncytium. Compared with the other known syncytins (Table 1), Mab-Env1 showed some interesting features. First, it was expressed not only in the placenta but also in many other tissues of the animal. Second, the noncoding polymerase gene corresponding to the envelope protein Mab-Env1 belongs to gammaretroviruses, but is not closely related to a group of gammaretroviruses that include murine leukemia virus, porcine endogenous retrovirus, Syn-Opo1, Syn-Rum1, Syn-1, Syn-2, and Syn-Car1. Third, Mab-Env1 has only two cysteines in a loop in the transmembrane protein (C-X₇-C) in contrast to typical gammaretroviral loops containing a third cysteine, which allows covalent binding with the surface protein (C-X₆-CC) (Fig. 1A). This suggests that it belongs to a potentially unknown retroviral family. Interestingly, Mab-Env3 and Mab-Env4 cluster with alpharetroviruses and present characteristics of avian gamma-type envelopes, confirming that these viruses are also either remotely related gammaretroviruses or belong to an unknown family. Since a functional Mab-Env1 was not found in lizards closely related to Mabuya, it was suggested that it entered the Mabuya genus 25 million years ago.

Fig. 1.

Structure of a retroviral envelope protein (A) and a comparison of the sequences of the immunosuppressive domains of different exogenous and endogenous retroviruses (B) are shown. The syncytins of Marubya lizards are framed in red. Note that syncytin-Mab1 has only two cysteines. BaEV, baboon endogenous retrovirus; BERV, bovine endogenous retrovirus; BERV-P, bovine endogenous retrovirus; C-CC, cysteine loop; CKS-17, consensus sequence of the gammaretroviruses; Env, envelope protein; ERV-3, endogenous retrovirus 3; FeLV, feline leukemia virus; FP, fusion peptide; GaLV, gibbon ape leukemia virus; HERV-FRD, human endogenous retrovirus FRD; HERV-K, human endogenous retrovirus K; HERV-W, human endogenous retrovirus W; HERV4-1, human endogenous retrovirus 4-1; ISU, immunosuppressive domain; KoRV, koala retrovirus; Mab-Env1–4, endogenous retroviruses of Marubya lizards; MMTV, mouse mammary tumor virus; MPMV, Mason Pfizer monkey virus; MSD, membrane-spanning domain; MuLV, murine leukemia virus; PERV, porcine endogenous retrovirus; SIVagm, SIV African green monkeys; SIV cpz, simian immunodeficiency virus chimpanzee; SIV sm, SIV sooty mangabey; SRV-1, simian retrovirus-1; SU, surface envelope protein; TM, transmembrane envelope protein. Identical amino acids and conservative exchanges (L = V = I) in all virus groups or in a single group are stained.

Like most retroviruses, all four transmembrane envelope proteins Mab-Env1–4 contain an immunosuppressive domain (Fig. 1), in addition to the fusion peptide and the membrane-spanning domain. The immunosuppressive domain is highly conserved among retroviruses. Synthetic peptides corresponding to this domain; recombinant and viral transmembrane envelope proteins; and noninfectious particles of several retroviruses, including HIV-1, have been shown to have immunosuppressive activity, modulating the cytokine release of normal human peripheral blood mononuclear cells and modulating the gene expression in these cells (reviewed in refs. 2, 5). Tumor cells not growing to tumors in immunocompetent mice grow to tumors when expressing the transmembrane protein of different retroviruses, indicating the immunosuppressive activity of these proteins (6–8). Single mutations in the immunosuppressive domain abrogated their immunosuppressive activity (7–9). Although the immunosuppressive activity of the Mabuya endogenous transmembrane protein was not analyzed, the sequence homology suggests such an activity (Fig. 1B). On the other hand, it cannot be excluded that the mutations in position 1, 2, and 3 in comparison to the gammaretrovirus consensus sequence CKS-17 abrogate this activity.