Endogenous Retrovirus

An endogenous retrovirus (ERV) is a relic viral sequence embedded within the nuclear genome of a jawed vertebrate or other eukaryote, originating from an ancestral exogenous retrovirus that infected a host germline cell (an oocyte, spermatozoon, or early embryonic blastomere). During a canonical retroviral infection, the viral enzyme reverse transcriptase converts the single-stranded viral RNA genome into linear double-stranded DNA, which the viral integrase enzyme covalently splices into a host chromosome to form a provirus. Whereas proviral integration in somatic cells dies with the individual host, integration into the germline allows the provirus—provided it does not disrupt an essential host gene—to be transmitted vertically to offspring in accordance with classical Mendelian inheritance and potentially reach fixation across a species through genetic drift or positive selection.

In Homo sapiens, human endogenous retroviruses (HERVs) and their fragmented derivatives account for approximately 8.3% of the total nuclear genome (roughly 250 million base pairs distributed across ~450,000 individual loci)—more than four times the ~1.5% of the genome that encodes canonical host proteins. Over tens of millions of years of vertebrate evolution, the vast majority of ERV loci have accumulated nonsense mutations, frameshifts, internal deletions, and repressive epigenetic modifications that render them replication-incompetent. Nevertheless, ERVs serve as a major reservoir of evolutionary innovation: host genomes have repeatedly co-opted (exapted) retroviral envelope proteins to drive mammalian placenta formation and repurposed retroviral long terminal repeats (LTRs) as tissue-specific cis-regulatory enhancers.

Canonical Provirus Structure and Solo LTR Formation

A newly integrated, full-length ERV provirus spans approximately 7 to 11 kilobases (kb) and mirrors the genomic organization of exogenous retroviruses: 5'-LTR – gag – pro – pol – env – 3'-LTR. The three core protein-coding regions fulfill distinct structural and enzymatic roles:

• gag (Group-Specific Antigen): Encodes the structural polyprotein cleaved into the matrix (MA), capsid (CA), and nucleocapsid (NC) proteins that form the inner viral core.
• pro and pol (Protease and Polymerase): Encode the aspartyl protease (PR) required for polyprotein maturation, the RNA-dependent DNA polymerase and RNase H domains of reverse transcriptase (RT), and the integrase (IN) endonuclease.
• env (Envelope): Encodes the surface (SU) and transmembrane (TM) glycoproteins responsible for receptor binding and membrane fusion during viral entry.

Flanking the coding sequences are two identical Long Terminal Repeats (LTRs), each typically 300 to 1,200 bp in length and subdivided into U3 (unique 3'), R (repeat), and U5 (unique 5') regions. Because reverse transcription duplicates the U3 and U5 ends prior to integration, the 5' and 3' LTRs are 100% identical in sequence at the moment of insertion, flanked by a characteristic 4- to 6-bp host target site duplication (TSD). The U3 region harbors potent RNA Polymerase II core promoters, TATA boxes, polyadenylation signals, and binding motifs for host transcription factors.

Over evolutionary time, the two identical flanking LTRs of a full-length provirus frequently undergo intrachromosomal homologous recombination (ectopic unequal crossover). This recombination event excises the internal gag–pol–env coding sequence and leaves behind a single hybrid solo LTR. Approximately 85% to 90% of all HERV insertions in the human genome exist today as solo LTRs. Because nucleotide substitutions accumulate independently in the 5' and 3' LTRs of intact proviruses at the neutral host mutation rate (roughly 1.0 to 2.2 × 10⁻⁹ substitutions per site per year in primates), comparing the sequence divergence between paired 5' and 3' LTRs provides a molecular clock to date the original germline integration event.

Classification of Human Endogenous Retroviruses (HERVs)

Phylogenetic analysis of the highly conserved pol reverse transcriptase domain classifies vertebrate ERVs into three principal classes corresponding to exogenous retroviral genera, traditionally named using the single-letter amino acid code of the host transfer RNA (tRNA) that primes reverse transcription at the primer-binding site (PBS):

Among all human families, HERV-K (HML-2)—human mouse mammary tumor virus-like group 2—represents the most recently active lineage. Several dozen HERV-K (HML-2) loci are unique to Homo sapiens (absent in chimpanzees and bonobos), and at least 11 to 36 loci remain insertionally polymorphic within modern human populations, retain intact open reading frames for functional Gag, Pol, or Env proteins, and can assemble non-infectious viral-like particles in teratocarcinoma cells and early embryos.

Host Defense: KRAB-Zinc Finger Proteins and KAP1/TRIM28

Unchecked transcription or retrotransposition of ERVs threatens host genomic integrity through insertional mutagenesis, ectopic recombination, and aberrant activation of nearby proto-oncogenes. Consequently, tetrapod genomes have evolved an adaptive intracellular surveillance network centered on Krüppel-associated box domain zinc finger proteins (KRAB-ZFPs). Encoded by more than 350 genes in the human genome—many clustered on chromosome 19—KRAB-ZFPs engage in an evolutionary arms race with emerging retroelements by evolving C-terminal C₂H₂ zinc finger arrays that bind specific DNA sequence motifs within ERV LTRs or primer-binding sites.

Upon binding an ERV locus during early embryogenesis, the N-terminal KRAB domain recruits the scaffold corepressor KAP1 (TRIM28), which assembles a macromolecular silencing complex comprising the histone methyltransferase SETDB1 (ESET), Heterochromatin Protein 1 (HP1), the NuRD chromatin remodeling and deacetylase complex, and the de novo DNA methyltransferases DNMT3A and DNMT3B. SETDB1 catalyzes the trimethylation of histone H3 at lysine 9 (H3K9me3), while DNMT3A/3B methylate CpG dinucleotides across the LTR into 5-methylcytosine (DNA methylation), locking the provirus into transcriptionally silent constitutive heterochromatin that persists in somatic tissues.

Exaptation: Syncytins in Placentation and Cis-Regulatory Networks

Despite stringent epigenetic repression of most loci, natural selection has domesticated numerous ERV elements to perform indispensable host physiological functions—a process termed exaptation. The premier example of protein-coding exaptation is the independent capture of retroviral env genes to drive the morphogenesis of the mammalian placenta. In primates, the ERVW-1 gene on chromosome 7q21.2 (derived from a HERV-W insertion ~25–30 million years ago) encodes Syncytin-1, and the ERVFRD-1 gene on chromosome 6p24.2 (derived from a HERV-FRD insertion ~40 million years ago) encodes Syncytin-2. Retaining the ancestral fusogenic mechanics of the retroviral transmembrane glycoprotein, Syncytin-1 binds the sodium-dependent neutral amino acid transporter ASCT2 (SLC1A5) on cytotrophoblast cells, mediating cell–cell membrane fusion to generate the multinucleated syncytiotrophoblast layer at the maternal–fetal interface. Syncytin-2 additionally retains a conserved retroviral immunosuppressive domain (ISD) that helps protect the semi-allogeneic fetus from maternal T-cell rejection.

Remarkably, convergent capture of distinct retroviral env genes (syncytin-A and syncytin-B in muroid rodents, syncytin-Ory1 in lagomorphs, syncytin-Car1 in carnivores, and syncytin-Rum1 in ruminants) has occurred at least 10 independent times across eutherian and marsupial mammals, explaining the striking anatomical diversity of mammalian placentas. Beyond protein coding, thousands of solo LTRs have been exapted as cis-regulatory enhancers and alternative promoters: the salivary amylase gene (AMY1C) relies on an inserted HERV-E LTR for parotid-specific expression, and primate-specific MER41 LTRs harbouring STAT1/IRF1 binding motifs act as interferon-γ (IFNG)-inducible enhancers coordinating the innate immune transcriptional network.

Can human endogenous retroviruses produce infectious viruses today?

No naturally occurring replication-competent HERV capable of horizontal person-to-person transmission has been identified in modern humans. Even the youngest family, HERV-K (HML-2), contains disabling point mutations or deletions in at least one essential gene at every known fixed locus. However, researchers have reconstructed a functional consensus provirus (HERV-K_CON or Phoenix) in the laboratory by combining intact reading frames from multiple loci, proving that ancestral HERV-K elements were once fully infectious exogenous retroviruses.

How do scientists determine when an endogenous retrovirus integrated into the genome?

Researchers use two independent dating methods. First, comparative genomics maps orthologous ERV integration sites (sharing identical flanking host target site duplications) across primate species; an ERV present at the exact same chromosomal locus in humans, chimpanzees, and gorillas but absent in orangutans must have integrated between 9 and 14 million years ago. Second, because the 5' and 3' LTRs are identical at integration, measuring their accumulated nucleotide divergence and dividing by twice the host neutral mutation rate estimates the integration age.

What role do HERVs play in human diseases such as cancer and autoimmunity?

In healthy adult somatic cells, HERVs are silenced by DNA methylation and H3K9me3 heterochromatin. In many cancers (including melanoma, germ-cell tumors, and ovarian cancer) and autoimmune or neurodegenerative disorders (such as multiple sclerosis and amyotrophic lateral sclerosis), global epigenetic hypomethylation reactivates HERV transcription. This can create oncogenic chimeric transcripts via LTR promoter activation or trigger a "viral mimicry" innate immune response when double-stranded HERV RNA is sensed by cytoplasmic MDA5 and RIG-I receptors.

1. Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C., Baldwin, J., ... & International Human Genome Sequencing Consortium. (2001). "Initial sequencing and analysis of the human genome." Nature, 409(6822), 860–921. https://doi.org/10.1038/35057062
2. Mi, S., Lee, X., Li, X., Veldman, G. M., Finnerty, H., Racie, L., ... & McCoy, J. M. (2000). "Syncytin is a captive retroviral envelope protein involved in human placental morphogenesis." Nature, 403(6771), 785–789. https://doi.org/10.1038/35001608
3. Chuong, E. B., Elde, N. C., & Feschotte, C. (2016). "Regulatory evolution of innate immunity through co-option of endogenous retroviruses." Science, 351(6277), 1083–1087. https://doi.org/10.1126/science.aad5406
4. Rowe, H. M., Jakobsson, J., Mesnard, D., Rougemont, J., Reynard, S., Aktas, T., ... & Trono, D. (2010). "KAP1 controls endogenous retroviruses in embryonic stem cells." Nature, 463(7278), 237–240. https://doi.org/10.1038/nature08674