Theoretical Framework of Endogenous Viral Biosynthesis

Application of cellular stress to cells bearing integrated EVE induces a dose-dependent increase in transcription with a measurable activation threshold.

IFN-gamma treatment increases HERV expression in human cells via ISGF3 transcription factors.

Under optimal induction conditions (I > I_crit), cells bearing a complete EVE can produce virions.

HERV-K102 activation by IFN-gamma creates a positive feedback loop amplifying the type I IFN response.

Deregulated HERV-K102 expression contributes to autoimmune disease pathogenesis.

HERV-K102 is significantly upregulated in response to interferon signaling

HERV-K102 at locus 1q22 is upregulated by IFN; HERV-K102 transcripts dominate HML-2 RNAs during pro-inflammatory macrophage polarization

IFN factors induce transcription factor binding to the LTR region upstream of HERV-K102

IFN-alpha induces STAT1 and IRF1 binding to LTR12F upstream of HERV-K102, triggering its transcription

HERV-K102 activation forms a positive feedback loop amplifying the type I interferon response

Type I interferons increase HERV-K102 expression in myeloid cells (monocytes, neutrophils) in SLE and RA patients; implication in an IFN amplification loop

Genomic integration of HERV-K102 at locus 1q22. The retroviral cassette LTR12F-gag-pol-env-LTR3 is inserted into chromosome 1 and can be transcribed into messenger RNA under the effect of inducing factors.

HERV-K102 induction cascade: IFN-alpha activates the JAK/STAT pathway, inducing STAT1 and IRF1 binding on LTR12F, triggering HERV-K102 transcription. Transcripts amplify the IFN response via a positive feedback loop.

BEVI model dynamics: temporal evolution of concentrations E(t) (integrated EVE), T(t) (viral transcripts), P(t) (viral proteins) and V(t) (assembled particles). Differential equations (1)-(5) predict a cascade activation with a saturation regime for V(t). Illustrative model based on the paper equations.

Model predictions (2018) compared to independent experimental confirmations (2023, 2025). Corroboration was not anticipated by the study authors; results are independently convergent.

HERV-K102 upregulated by IFN signaling (H1/H2)

Transcription factors bind upstream LTR (H2)

HERV-K102 at 1q22 dominates HML-2 RNAs during pro-inflammatory macrophage polarization

IFN-alpha induces STAT1/IRF1 binding on LTR12F upstream of HERV-K102

Type I IFNs increase HERV-K102 in myeloid cells of SLE and RA patients, forming an IFN amplification loop

Illustrative BEVI simulator: HERV-K102 induction cascade

Below threshold (I < I_crit) non-productive biosynthesis

Above threshold (I > I_crit) productive activation

ILLUSTRATIVE MODEL based on the publication's differential equations. Not a live biological simulator. Values do not represent real biological data.

Molecular virology · Endogenous biology · Mathematical framework

Theoretical Framework of Induced Endogenous Viral Biosynthesis

A mathematical and experimentally falsifiable model in molecular virology DESVAUX G.J.Y., December 2018

The discovery of endogenous viral elements (EVE) represents one of the most significant advances in modern virology. These virus-derived sequences have colonized eukaryotic genomes over hundreds of millions of years of co-evolutionary history. Human endogenous retroviruses (HERVs) represent approximately 8.3% of the human genome more than the totality of coding exons. Yet their biological functions and the conditions of their reactivation remained largely theoretical at the time of this publication (2018).

This work proposes a quantitative mathematical framework the BEVI model to describe when and how inducing factors (particularly interferon) can lift the epigenetic repression of these sequences, leading to their transcription and potentially to the assembly of viral particles. The explicit formulation of falsifiable hypotheses (H1-H5) is at the core of the scientific approach adopted.

2. Endogenous Viral Elements and Human Retroviruses

Definition 1: Endogenous Viral Element (EVE)

An endogenous viral element is a nucleotide sequence integrated into an organism's genome, derived from an ancestral virus that infected germ cells of a common ancestor.

HERVs represent ~8.3% of the human genome.

Human endogenous retroviruses (HERVs) are the most studied subset of EVEs. Unlike exogenous retroviruses, they are hereditarily transmitted in the germline. The vast majority are epigenetically silenced by DNA methylation and histone modifications but they retain the capacity to be re-expressed under specific stimuli.

The conserved retroviral architecture is: LTR-gag-pol-env-LTR. For HERV-K102, this cassette is localized at chromosomal locus 1q22, with an LTR12F region at the 5' position that is particularly sensitive to regulation by IFN-inducible transcription factors.

3. The BEVI Model and Mathematical Formalization

The BEVI model is based on a three-phase conceptual architecture: Phase 1 - Latency (EVE integrated, epigenetically silent); Phase 2 - Induction (IFN, stress, demethylation lift repression); Phase 3 - Biosynthesis (transcription, translation, virion assembly). This progression is governed by coupled differential equations.

Intensity of the inducer factor (0-1)

Equations (1)-(5) of the BEVI model

Quadratic assembly kinetics with saturation at Vmax

Equation (3) the induction sigmoid function f(I) is the central element of the model. It implies the existence of an activation threshold I_crit below which biosynthesis is non-productive, and above which it becomes productive (Activation Theorem, equation 6 of the publication). This prediction is directly experimentally falsifiable: one can measure the response of HERV-K102 at different IFN-alpha concentrations and verify the sigmoid shape.

4. Induction Cascade and HERV-K102 Feedback Loop

The IFN -> STAT1/IRF1 -> LTR12F -> HERV-K102 molecular cascade

The BEVI model predicts that interferon-alpha (IFN-alpha) is the most efficient inducing factor for HERV-K102. The proposed mechanistic pathway: IFN-alpha activates the JAK/STAT pathway (inducer factor I in equation 3), producing active STAT1 and IRF1 complexes capable of binding to the LTR12F region at the 5' position of HERV-K102. This binding lifts transcriptional repression, allowing HERV-K102 mRNA production.

The most important consequence of this activation is the formation of a positive feedback loop: HERV-K102 transcripts themselves contribute to amplifying the type I IFN response. This type of positive feedback mechanism is generally associated with strongly amplified and potentially pathological immune responses. The mathematical model allows precise calculation of the runaway condition for this feedback.

In accordance with Popper's criterion, each hypothesis of the BEVI model is formulated in a way that allows experimental refutation. The publication also details concrete experimental protocols for testing each hypothesis (THP-1, HEK293T, A549 cultures; IFN-gamma treatment at increasing doses; RNA-seq, RT-qPCR, Western Blot, electron microscopy).

Oxidative stress: ROS inducing DNA damage via ATM/ATR, I increase of +0.3 to +0.7

Immune stress: interferons via JAK/STAT pathway, I increase of +0.2 to +0.5

Metabolic stress: hypoxia via HIF-1alpha, I increase of +0.2 to +0.4

It is important to clarify that the publications by Russ et al. (2023) and Takahashi et al. (2025) were carried out independently and without reference to the BEVI model. The convergence between the model predictions (formulated in 2018) and the experimental results obtained subsequently constitutes an indirect but significant corroboration of the plausibility of the proposed theoretical framework.

Russ et al. 2023: HERV-K102 and the IFN response

Russ E. et al. (2023, Microbiology Spectrum, 11(2), e04438-22) showed that the HERV-K102 provirus at locus 1q22 is significantly upregulated in response to interferon signaling. They specifically documented that IFN-alpha induces STAT1 and IRF1 binding to LTR12F upstream of HERV-K102, and that HERV-K102 transcripts constitute the majority of HML-2 RNAs during pro-inflammatory macrophage polarization.

Takahashi et al. 2025: positive IFN feedback loop

Takahashi Y. et al. (2025, BMC Genomics, 26(1)) showed that type I interferons increase HERV-K102 expression in myeloid cells (monocytes and neutrophils) from patients with systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA). These results are consistent with hypothesis H4 of the BEVI model: HERV-K102 activation forms a positive feedback loop amplifying the type I IFN response, and this loop could play a role in the pathogenesis of autoimmune diseases.

If the BEVI model is more broadly confirmed, it opens therapeutic perspectives. Associated pathologies include: multiple sclerosis (HERV-W), systemic lupus erythematosus (HERV-K102), rheumatoid arthritis (HERV-K), and certain cancers (ERV-9, HERV-K). Envisaged therapeutic approaches include reverse transcriptase inhibitors, demethylating agents, targeted immunomodulators, and anti-HERV antibodies.

Simplification of signaling pathways: cellular stress pathways are interconnected in complex ways and the model does not capture all cross-talks.

EVE heterogeneity: there are thousands of EVE copies in the human genome with different characteristics; the model focuses on HERV-K102 as a case study.

Species specificity: the model is primarily based on human data and may not generalize directly to other organisms.

This theoretical framework is published under an open licence on Zenodo (DOI 10.5281/zenodo.19006097). It is a theoretical article with explicitly falsifiable hypotheses. The comparison between model predictions and independent experimental data constitutes the appropriate scientific validation for this type of work.