HIV Drug Ritonavir Found to Degrade Key Cytomegalovirus Protein and Block Viral Replication

Human cytomegalovirus (HCMV) is one of the most widespread pathogens in the human population and a formidable threat to people whose immune systems are compromised. In newborns, transplant recipients, and patients living with HIV, the virus can trigger retinitis, interstitial pneumonia, and severe neurological impairment. Yet the antiviral arsenal against HCMV remains narrow. Drugs such as ganciclovir, which inhibits the viral DNA polymerase, and letermovir, which targets the terminase complex, act only during the mid-to-late stages of viral replication, and their long-term use is constrained by hematologic toxicity, the emergence of drug resistance, and poor activity against latent viral reservoirs. A new study published in PLOS Computational Biology now reports that an already approved drug, the HIV protease inhibitor ritonavir, can attack HCMV at a much earlier point in its life cycle by eliminating a master regulator of viral gene expression.

The target in question is the immediate-early protein 1, or IE1, a product of the viral major immediate-early gene that has long been considered undruggable. Through alternative splicing, the immediate-early transcript yields two major isoforms, IE1 and IE2, which together orchestrate the onset of lytic replication. IE1 coordinates the cascade activation of early and late viral genes, neutralizes host translational repressors such as PKR to favor viral protein synthesis, interferes with STAT1/STAT2 signaling to suppress type I interferon responses, and pushes infected cells from the G1 into the S phase of the cell cycle to create a replication-friendly environment. Because IE1 lacks a well-defined catalytic pocket or conventional active site, structure-based small-molecule design has historically been difficult to apply, leaving the protein without direct pharmacological inhibitors despite abundant evidence that disabling it halts viral replication.

A research team led by Wanfeng Liu, Yu Shi, Sunzhong Mao, Xiangyang Xue, and Shiyu Feng at Wenzhou Medical University and collaborators approached this problem through computational drug repurposing. Using the resolved crystal structure of the IE1 central core domain (PDB ID 6TGZ) as the receptor, they docked 890 compounds from an FDA-approved drug library with Discovery Studio 2019 and ranked hits by LibDock score. Six candidates exceeded the score threshold of 150, among them ritonavir, which topped the list with a score of 174.651. A second round of docking using CB-Dock2 for blind surface scanning and AutoDock-GPU for site-specific calculations identified five candidate pockets on the IE1 homodimer, with pocket 1 at the interface of the A and B subunits selected for analysis. Ritonavir again ranked first, showing the most favorable predicted binding energy of −16.16 kcal/mol.

To test whether the predicted interaction was physically plausible, the team ran 100-nanosecond molecular dynamics simulations of both unbound IE1 and the IE1–ritonavir complex in explicit solvent using GROMACS 2024 on the WeMol platform. Analyses of backbone root-mean-square deviation, radius of gyration, residue-level fluctuations, hydrogen-bond counts, solvent-accessible surface area, secondary-structure content, and principal components indicated that the complex remained stable without major perturbation of the global protein fold. MM-PBSA calculations on the equilibrated 50–100 nanosecond trajectory segment suggested energetically favorable binding driven largely by van der Waals contacts, with residue-level decomposition flagging GLU615, LEU523, PHE642, and PRO694 as potential hotspot residues at the predicted interface.

The decisive evidence came from surface plasmon resonance. Using a Biacore 1K system with recombinant IE1 immobilized on a CM5 sensor chip, the researchers injected serial dilutions of ritonavir ranging from 12.5 to 6400 nanomolar and observed concentration-dependent, reversible binding. Global fitting of the sensorgrams to a 1:1 Langmuir model yielded an equilibrium dissociation constant of 3.18 micromolar, while steady-state analysis gave a comparable value of 4.14 micromolar. The authors caution that docking scores were used only for relative ranking within identical simulation systems and cannot reproduce absolute affinities, but the combination of computational prediction and direct biophysical measurement provides the first indication that ritonavir binds IE1.

Binding, however, was only the beginning. In human foreskin fibroblast cells infected with the laboratory-adapted HCMV strain AD169, ritonavir reduced IE1 protein levels in a time- and dose-dependent manner, with a detectable decrease from day 2 post-infection. The same reduction appeared in HeLa cells engineered to express IE1 alone, showing that the effect does not depend on the full viral replication context, and immunofluorescence microscopy confirmed diminished nuclear IE1 signal in infected cells. Crucially, the drug appeared selective: the abundance of host proteins such as GAPDH, β-actin, α-tubulin, and HSP90 was unchanged, global cellular ubiquitination was not increased, and ectopically expressed viral proteins UL96, UL55, UL86, UL50, and UL83 were not appreciably reduced. Reverse-transcription quantitative PCR showed that IE1 mRNA levels were unaffected across five days of treatment, pointing to a post-transcriptional mechanism.

That mechanism turned out to be targeted protein degradation. In a cycloheximide chase assay that blocks new protein synthesis, ritonavir markedly shortened the half-life of IE1, producing roughly 40 percent greater degradation within nine hours compared with controls. Rescue experiments then discriminated between the two major intracellular degradation routes: the proteasome inhibitor MG132 partially reversed the ritonavir-induced loss of IE1, whereas the autophagy inhibitor chloroquine had no effect. Co-immunoprecipitation experiments in cells stably expressing Flag-tagged IE1 showed that ritonavir treatment substantially increased ubiquitin signals attached to IE1, indicating that the drug enhances IE1 ubiquitination and channels the protein into the ubiquitin–proteasome pathway. Although ritonavir is neither a PROTAC nor a molecular-glue degrader, the authors note that its mode of action is conceptually related to targeted protein degradation, an approach increasingly explored as an antiviral strategy.

The functional consequences for the virus were substantial. At concentrations that did not significantly affect host cell viability or apoptosis, ritonavir reduced extracellular viral DNA loads in the supernatant of infected cells from 31.52 to 11.75 copies per milliliter, an inhibition rate of approximately 62.7 percent, and TCID50 assays confirmed a dose-dependent drop in infectious particle production. A recombinant HCMV expressing GFP provided independent confirmation: supernatants from ritonavir-treated cultures produced roughly 2.2-fold lower GFP signals upon reinfection of fresh fibroblasts. Western blotting revealed that prolonged treatment suppressed not only IE1 but also the downstream viral proteins UL44, UL83, and UL84, consistent with collapse of the immediate-early regulatory cascade. Quantifying dose-response relationships, the team measured a half-maximal cytotoxic concentration of 57.36 micromolar and a half-maximal effective concentration of 6.10 micromolar for ritonavir, corresponding to a selectivity index of 9.40, a moderate but meaningful in vitro safety margin.

Perhaps the most clinically suggestive finding concerns combination therapy. When ritonavir was paired with ganciclovir, the two drugs acted synergistically: 5 micromolar ritonavir combined with 2 micromolar ganciclovir achieved 90.83 percent inhibition of viral replication, well above the 73.43 percent predicted by the Bliss independence model, and a 5-by-5 checkerboard analysis yielded an overall Bliss synergy score of 10.443 with a most synergistic area score of 12.10. The combination also suppressed IE1 more strongly than either drug alone. By contrast, ritonavir combined with letermovir showed no meaningful synergy, with observed effects falling below Bliss-predicted values and an overall synergy score of only 1.475. The authors attribute the ganciclovir synergy to the drugs’ distinct mechanisms, with ritonavir dismantling early viral regulation while ganciclovir blocks DNA synthesis, whereas the basis for the absence of synergy with letermovir remains unclear.

The study has limitations that the authors acknowledge openly. Their screening pipeline relied on conventional LibDock and AutoDock docking rather than newer AI-driven frameworks, without retrospective benchmarking using decoy-set enrichment or ROC-AUC evaluation. Mechanistically, MG132 only partially rescued IE1 levels, the responsible E3 ubiquitin ligase remains unidentified, and binding-site mutagenesis and IE1 rescue experiments are still needed to establish a definitive causal link between IE1 degradation and antiviral activity. Most experiments used laboratory cell lines and the AD169 strain rather than clinical isolates, drug-resistant strains, or animal models, and whether the effective concentrations are achievable in patients remains to be determined. Even so, the work validates IE1 as a druggable target, reveals a previously unknown direct antiviral activity of a widely used HIV protease inhibitor, and suggests a combination strategy that could matter particularly for patients co-infected with HIV and HCMV. Because the IE1 core domain is highly conserved among humans, rhesus macaques, and rats, the approach may also open the door to cross-species, broad-spectrum antiviral development built on pharmacological destabilization of essential viral regulators.

Subject of Research: Repurposing ritonavir to induce proteasomal degradation of the HCMV immediate-early protein IE1 as an antiviral strategy

Article Title: Repurposing ritonavir to induce proteasomal degradation of IE1 for inhibition of human cytomegalovirus replication

Article References: Liu, W., Shi, Y., Guo, X., Lou, C., Xie, Z., Shen, Y., Zhang, L., Mao, S., Xue, X., & Feng, S. (2026). Repurposing ritonavir to induce proteasomal degradation of IE1 for inhibition of human cytomegalovirus replication. PLOS Computational Biology, 22(10), e1014867. https://doi.org/10.1371/journal.pcbi.1014867

Image Credits: AI Generated

DOI: 10.1371/journal.pcbi.1014867

Keywords: human cytomegalovirus, ritonavir, drug repurposing, IE1 protein, ubiquitin-proteasome pathway, molecular docking, surface plasmon resonance, ganciclovir, antiviral therapy, targeted protein degradation, synergy, molecular dynamics simulation