Epidermodysplasia verruciformis, or EV, is a rare genodermatosis defined by a lifelong susceptibility to infection by β-genus human papillomaviruses. Patients develop disseminated flat, wart-like lesions beginning in childhood, and a substantial fraction go on to develop cutaneous squamous cell carcinoma, particularly on sun-exposed skin. Genetic work over the past two decades has traced the condition to loss-of-function variants in the TMC6 and TMC8 genes and to disruption of the CIB1–EVER1–EVER2 complex, painting a picture in which keratinocytes—the principal cells of the epidermis—lack an intrinsic defense that normally keeps β-HPV in check. When persistent viral infection combines with ultraviolet light damage, malignant transformation becomes far more likely, making EV a natural experiment in virus-driven skin carcinogenesis.

The cGAS–STING pathway sits at the crossroads of viral detection and cancer biology. The enzyme cyclic GMP-AMP synthase, or cGAS, patrols the cytoplasm for double-stranded DNA that should not be there—whether from an invading virus or from the host genome itself, leaked through damaged nuclear membranes or packaged into micronuclei generated by chromosomal instability. Upon binding such DNA, cGAS produces a cyclic dinucleotide second messenger that activates STING on the endoplasmic reticulum, triggering TBK1–IRF3 signaling and the production of type I interferons. Because cytosolic DNA can arise both from viral infection and from genomic stress, the pathway links the two great themes of EV biology: chronic papillomavirus presence and the DNA damage that ultraviolet exposure inflicts on genetically susceptible skin.

To determine whether this pathway is expressed differently in EV-associated tumors, the team constructed tissue microarrays from 47 EV-associated cutaneous squamous cell carcinomas and 83 non-EV tumors from immunocompetent patients. The specimens came from the dermatopathology archive of the university’s medical school, and EV diagnoses rested on a characteristic clinical history of childhood-onset disseminated flat warts together with compatible histopathology. Thirteen EV patients contributed 46 of the linked tumors, meaning some patients supplied multiple cancers—a feature the investigators explicitly accounted for in their statistics using generalized estimating equations with patient identity as the clustering variable. Previous molecular work had confirmed β-HPV infection in nine of the thirteen patients. Two dermatopathologists independently reviewed all cases, and up to four tumor cores per sample were arrayed for staining.

The quantification itself was rigorously digital. Immunostained slides were scanned on an Aperio ScanScope and analyzed in QuPath, with image analysis performed blinded to whether a tumor came from an EV or non-EV patient and restricted to manually delineated tumor regions. Tumor cells were classified as negative, weak, moderate or strong using fixed optical-density thresholds applied uniformly across all slides, and an H-score—combining intensity and percentage of positive cells on a 0 to 300 scale—was calculated for each core, with the tumor-level value defined as the median across valid cores. STING was detected cytoplasmically, p53 nuclearly, and BCL2 cytoplasmically, all within identical staining batches for both groups to eliminate batch effects.

The headline result was unambiguous: STING expression was substantially higher in EV-associated tumors. The median H-score reached 192.4 in EV cancers compared with 149.5 in non-EV cancers, and after adjustment for histological grade the difference was estimated at 37.2 H-score units, a highly significant gap. Intriguingly, the size of that gap depended on how differentiated the tumors were. A formal interaction test found the EV-versus-non-EV difference varied by histological grade, with the largest estimated difference—72.2 units—appearing in well-differentiated grade 1 tumors, significantly exceeding the differences seen in grades 2 and 3. A substantial difference was also seen in carcinoma in situ. Whether this grade dependence reflects biology at early stages of malignant progression or changes as tumors dedifferentiate remains an open question.

p53 told a parallel story. The protein accumulated to much higher levels in EV tumors—median H-score 97.7 versus 36.3—and EV tumors were more than three and a half times as likely to fall in the highest tertile of p53 expression after cluster-aware ordinal regression. The association persisted after adjustment for grade and other covariates. BCL2, by contrast, was barely expressed in either group and showed no independent relationship with EV status, a result consistent with earlier immunohistochemical surveys of non-melanoma skin cancer. Notably, STING and p53 levels were positively correlated across the cohort, and the correlation survived adjustment for grade, EV status and patient clustering, hinting at convergent stress signaling within these tumors.

The authors are careful about what the staining can and cannot show. Total STING immunohistochemistry measures protein abundance, not pathway activation; recent experimental work has shown that the E6 protein of HPV-8, a β-papillomavirus classically associated with EV, reduces STING phosphorylation and blunts downstream interferon signaling in keratinocytes even though cGAS is recruited to micronuclei. Likewise, p53 staining cannot distinguish wild-type p53 stabilized by cellular stress from accumulation of mutant protein, and archival material was not available for TP53 sequencing. Experimental evidence cuts both ways: p53 activation can increase STING expression, while mutant p53 can disrupt STING–TBK1–IRF3 signaling. The São Paulo findings therefore define an altered protein-expression phenotype whose functional meaning—heightened antiviral alarm, senescence-associated stress, or both—awaits phospho-STING, phospho-TBK1, phospho-IRF3 and interferon-stimulated-gene readouts in matched tumor, precursor and non-neoplastic EV skin samples.

The study’s limitations are those of a retrospective, single-institution series: only thirteen unique EV patients, no individual cumulative ultraviolet exposure data, β-HPV typing available for only part of the cohort, and no molecular HPV assessment of the non-EV tumors. Yet the signal survived multivariable adjustment for age, sex, grade and anatomical region, as well as sensitivity analyses that aggregated to the patient level, restricted to tumors with multiple valid cores, and overlapped the age distributions of the two groups. As a next step, the researchers argue that comprehensive viral typing, viral-load measurement and TP53 sequencing should accompany functional dissection of the cGAS–STING axis, and that pharmacological modulation of the pathway should wait until functional differences are firmly established. If the elevated STING phenotype proves to reflect intact or even amplified DNA-sensing, EV-associated skin cancers could become an informative model for how the immune system perceives virus-driven malignancy—and, ultimately, a testing ground for therapies that deliberately turn the alarm up.

Subject of Research: Differential STING and p53 protein expression in epidermodysplasia verruciformis-associated versus non-EV cutaneous squamous cell carcinoma

Article Title: Differential STING and p53 expression in epidermodysplasia verruciformis-associated versus non-EV cutaneous squamous cell carcinomas

Article References: Fróes, L. A. R., de Oliveira, W. R. P., Stahlschmidt, P., da Cruz Silva, L. L., Pereira, N. V., & Sotto, M. N. (2026). Differential STING and p53 expression in epidermodysplasia verruciformis-associated versus non-EV cutaneous squamous cell carcinomas. Medical Oncology, 43(10), Article 286. https://doi.org/10.1007/s12032-026-03417-0

Image Credits: AI Generated

DOI: 10.1007/s12032-026-03417-0

Keywords: epidermodysplasia verruciformis, cutaneous squamous cell carcinoma, STING, cGAS, p53, BCL2, betapapillomavirus, immunohistochemistry, tissue microarray, innate immunity, skin cancer, HPV

The biological logic of STING makes such a map unusually valuable. Under normal conditions, STING resides in an inactive, self-inhibited conformation on the endoplasmic reticulum membrane, its ligand-binding domain held in a closed orientation. When cyclic dinucleotides, either bacterial second messengers or the mammalian cyclic GMP-AMP synthesized by the sensor cGAS after DNA leakage into the cytoplasm, bind to the ligand-binding pocket, STING rotates into an open state, exits the endoplasmic reticulum through the Golgi apparatus, and initiates a phosphorylation cascade through TBK1 and IRF3 that drives type I interferon and inflammatory cytokine production. The protein therefore operates as a conformational switch under tight negative control, and mutations that nudge the equilibrium in either direction produce distinctive disease phenotypes.

Gain-of-function variants of STING sit at one end of the clinical spectrum. Amino acid substitutions clustered in the dimerization interface and the lid region that covers the cyclic dinucleotide binding pocket destabilize the closed, inactive conformation, allowing spontaneous ligand-independent activation. Patients carrying such variants develop the autoinflammatory syndrome now classified as STING-associated vasculopathy with onset in infancy, or SAVI, characterized by systemic inflammation, interstitial lung disease, and vascular pathology that mirrors constitutive interferon signaling. Additional gain-of-function alleles have been linked to familial chilblain lupus and related interferonopathies, and mouse models carrying equivalent substitutions recapitulate the lethal inflammatory phenotype, confirming that the mutated protein itself, rather than an upstream sensing defect, drives pathology.

Loss-of-function variants occupy the opposite pole of the atlas. Biallelic inactivating mutations in the STING-encoding TMEM173 gene have been identified in patients presenting with a combined immunodeficiency marked by recurrent and severe viral infections, particularly respiratory viruses, alongside pulmonary disease. These alleles cluster in distinct structural neighborhoods: some disrupt ligand binding, others impair the conformational rearrangements needed for trafficking to the Golgi, and still others destabilize the protein so that steady-state abundance collapses. Intriguingly, a founder allele common in certain populations ablates STING function with apparently modest fitness cost, a reminder that evolutionary pressure from pathogens can shape the distribution of immune gene variants in ways that remain only partly understood.

Cancer adds a third interpretive layer to the mutation catalog. Tumors frequently acquire mutations that silence STING signaling, because intact STING activity, by recruiting and activating antigen-presenting cells within the tumor microenvironment, opposes immune evasion. Loss-of-function alterations of STING pathway components have been documented across a range of malignancies, including colorectal and gastric cancers, and correlate with diminished T-cell infiltration and poorer responses to immune checkpoint blockade. Conversely, pharmacological activation of STING with synthetic cyclic dinucleotide agonists has emerged as a major strategy in cancer immunotherapy, with numerous candidates advancing through preclinical and clinical evaluation. The atlas therefore serves a translational purpose: a clinician or drug developer can query whether a given tumor-associated substitution is predicted to ablate, enhance, or leave unaffected STING signaling, and can reason about how that functional assignment might inform immunotherapy selection.

What distinguishes a systematic atlas from a simple variant list is the integration of structural and biophysical annotation. High-resolution crystal and cryogenic electron microscopy structures of STING from human and multiple animal species have defined the ligand-binding pocket, the dimer interface, the hydrophobic lid, and the transmembrane segment, while molecular dynamics simulations have mapped the conformational transitions connecting inactive and active states. By projecting every cataloged missense mutation onto these structural ensembles, the resource makes mechanistic hypotheses explicit: substitutions can be classified by their proximity to the ligand pocket, their predicted effect on dimer stability, their likely influence on the trafficking sequence that shuttles STING from endoplasmic reticulum to Golgi, or their impact on the post-translational modification sites, including palmitoylation and phosphorylation, that modulate signaling intensity and duration.

The clinical translation dimension of the atlas extends beyond cancer. STING agonists are being explored as vaccine adjuvants and as treatments for chronic viral infections, where a transient burst of innate stimulation could strengthen adaptive immune responses. At the same time, excessive STING activation has been implicated in sterile inflammatory diseases, neurodegeneration, and aspects of aging biology, prompting the parallel development of STING inhibitors. An evidence-based map of how sequence variation alters STING function is directly relevant to both efforts, because it identifies which patient subgroups carry hypomorphic or hypermorphic alleles that could shift the therapeutic window. Precision dosing and patient stratification for STING-targeted drugs may ultimately depend on genotyping knowledge of exactly the kind this resource consolidates.

The methodological logic of atlas-building also deserves attention. Rather than relying on any single assay, the resource aggregates evidence from patient phenotypes, reporter assays measuring interferon promoter activation, protein localization studies, ligand-binding measurements, and animal models, assigning each variant a functional class supported by converging data. This convergence-based approach mitigates a chronic weakness of the field, in which individual studies using different cell lines and stimulation conditions have occasionally reported conflicting results for the same variant. By standardizing nomenclature, linking each entry to structural context, and flagging variants whose functional assignment rests on limited evidence, the atlas establishes a framework that future experimental work can extend systematically rather than idiosyncratically.

Several open questions frame the next phase of this effort. The functional consequences of many rare missense variants observed in large human sequencing cohorts remain untested, and the interplay between STING sequence variation and other innate immune genes, including cGAS itself and the downstream interferon receptor machinery, is only beginning to be explored. The pharmacological landscape is similarly incomplete: structural differences between human and rodent STING complicate the translation of agonists and inhibitors across species, and variants that alter drug binding pockets could produce patient-to-patient differences in treatment response. A living, continuously updated atlas offers a shared reference point for addressing these gaps, converting the accumulated knowledge of two decades of STING research into a practical instrument for genetic diagnosis, drug development, and, ultimately, the individualized treatment of the inflammatory, infectious, and malignant diseases in which this remarkable signaling protein sits at the center.

Subject of Research: Systematic functional and structural cataloging of STING protein mutations linked to autoinflammatory disease, immunodeficiency, and cancer immunotherapy

Article Title: Atlas-ing STING mutations to advance fundamental understanding and clinical translation

Article References: Zhang, B.-C., & Paludan, S. R. (2026). Atlas-ing STING mutations to advance fundamental understanding and clinical translation. Cell Research. https://doi.org/10.1038/s41422-026-01294-w

Image Credits: AI Generated

DOI: 10.1038/s41422-026-01294-w

Keywords: STING, innate immunity, TMEM173, SAVI, interferon signaling, cGAS, cyclic dinucleotides, loss-of-function variants, gain-of-function variants, cancer immunotherapy, STING agonists, autoinflammatory disease

At the center of the finding is the cGAS–STING pathway, a two-component innate immune circuit that has become one of the most intensively studied signaling axes in immunology. The enzyme cGAS, or cyclic GMP–AMP synthase, functions as a sensor of double-stranded DNA in the cytosol, a location where DNA should not ordinarily reside. Under normal conditions, genomic DNA is sequestered in the nucleus and mitochondrial DNA is enclosed within the double membranes of mitochondria. When DNA appears in the cytosol, whether from invading viruses, bacteria, or leaking from damaged host organelles, cGAS binds it and catalyzes the synthesis of a second messenger molecule called cyclic GMP–AMP, or cGAMP. This small cyclic nucleotide then binds STING, the stimulator of interferon genes, an adaptor protein anchored in the endoplasmic reticulum membrane. Activated STING initiates a phosphorylation cascade through TBK1 and IRF3 that drives the expression of type I interferons and a broad program of inflammatory cytokines. Seminal work by Ablasser, Sun, Chen and colleagues established the molecular identities and ligands of this pathway, and subsequent studies demonstrated that self-DNA, not just microbial DNA, can trigger it, linking the pathway to sterile inflammation and aging across multiple tissues.

Lei and colleagues now show that the aging oocyte becomes a site of precisely this kind of misfired DNA sensing. In aged oocytes, mitochondrial function deteriorates, and the integrity of the mitochondrial compartment, which normally keeps mitochondrial DNA physically separated from the cytosol, is compromised. The team found evidence that mitochondrial DNA leaks into the oocyte cytosol, where cGAS detects it and initiates production of cGAMP. The resulting STING activation establishes a state of chronic, low-grade inflammatory signaling within the oocyte itself. This is significant because oocytes are extraordinarily long-lived cells; the oocytes that a woman ovulates in her forties were formed during her own embryonic development and have endured decades of metabolic and oxidative stress. The accumulation of mitochondrial damage over that timespan provides a plausible mechanistic basis for why cytosolic DNA leakage emerges as a hallmark of the aged oocyte, and why the researchers could tie the inflammatory signature directly to reproductive decline rather than to systemic aging factors alone.

Perhaps the most remarkable aspect of the study is the discovery that the inflammatory signal does not remain confined to the oocyte. Using an elegant combination of genetic, pharmacological and cell biological approaches, the researchers demonstrated that cGAMP generated within the aged oocyte travels to the surrounding granulosa cells through connexin 37, or CX37, gap junctions. Gap junctions are arrays of intercellular channels that directly connect the cytoplasm of adjacent cells, allowing the passive diffusion of ions, metabolites and small signaling molecules. In the ovarian follicle, oocyte–granulosa cell gap junctional communication is well documented and essential: it allows granulosa cells to nourish the oocyte, deliver cAMP and other regulators that maintain meiotic arrest, and coordinate the developmental dialogue between the germ cell and its somatic support cells. CX37, in particular, is known to form the gap junctions that physically couple the oocyte to the surrounding cumulus cells, and knockout studies going back to work by Simon and colleagues in 1997 showed that its loss disrupts folliculogenesis and ovulation.

Lei and colleagues turned this communication conduit into an inflammatory superhighway. Because cGAMP is small enough, roughly 675 daltons, to pass through gap junction pores, the second messenger synthesized in the oocyte diffuses into the coupled granulosa cells, where it activates STING in those cells. The granulosa cells, which are the somatic workhorses of the follicle responsible for hormone production, oocyte maturation support and ovulation, respond to STING activation by mounting a sterile inflammatory program, upregulating interferon-stimulated genes and pro-inflammatory mediators. In effect, the aged oocyte recruits its own support network into a state of chronic inflammation. The researchers propose that this oocyte-to-granulosa signaling axis constitutes a cell-nonautonomous mechanism of tissue aging: a damaged germ cell actively spreads its stress phenotype to neighboring cells, degrading the function of the entire follicular unit. This concept resonates with broader observations that STING signaling contributes to age-related inflammation, or inflammaging, in tissues ranging from muscle to brain, but the demonstration of a gap junction-mediated propagation mechanism in the ovary is novel.

The consequences of this inflammatory relay for ovarian function are substantial. The authors present evidence that the cGAS–STING-driven inflammation in granulosa cells contributes to the functional deterioration of the aging ovary, including diminished follicular quality and impaired reproductive capacity. Experiments in which the pathway was genetically or pharmacologically interrupted, either by deleting cGAS or STING, blocking gap junction communication, or interfering with cGAMP synthesis, mitigated the inflammatory activation in granulosa cells and preserved markers of ovarian health. Prior work had already implicated the cGAS–STING axis in ovarian aging: studies from Navarro-Pando and colleagues showed that dampening this pathway alleviated age-related ovarian decline in mouse models, and other groups had connected mitochondrial stress and cytosolic DNA sensing to follicular dysfunction. The new study advances the field by identifying the oocyte as the initiating cell and by defining a specific intercellular transmission route, converting a correlation between STING activation and ovarian aging into a mechanistic circuit with defined anatomical wiring.

The identification of CX37 gap junctions as the conduit for cGAMP propagation is of particular translational interest because gap junctions are pharmacologically tractable. Drugs that modulate gap junctional communication exist, and the study suggests that transiently reducing oocyte–granulosa coupling in aged ovaries, or selectively blocking cGAMP transfer, might interrupt the inflammatory spread without eliminating the essential metabolic support that gap junctions provide in young follicles. Alternatively, interventions that stabilize mitochondrial membranes in aged oocytes, preventing the initial escape of mitochondrial DNA into the cytosol, would act upstream of the entire cascade. Compounds that improve mitochondrial quality control, reduce reactive oxygen species, or promote mitophagy could, in principle, decrease the burden of cytosolic mitochondrial DNA and thereby blunt cGAS activation at its source. The study thus offers multiple points of entry for future therapeutic development aimed at extending reproductive lifespan.

Beyond reproductive medicine, the findings contribute to a growing conceptual framework in geroscience: that age-related tissue dysfunction can propagate through second messengers transmitted between cells. cGAMP has previously been shown to traverse gap junctions in other contexts, transferring antiviral states between neighboring cells, a phenomenon sometimes described as a form of innate immune bystander signaling. The ovarian study extends this idea to a physiological aging process and identifies a specific connexin isoform responsible. It also adds to evidence that the oocyte is not a passive victim of the aging ovarian environment but an active participant that can shape the behavior of surrounding somatic cells. This reframing has implications for assisted reproduction, where the quality of the oocyte’s somatic environment is known to influence embryo development, and for the broader effort to understand how individual aged cells impose inflammatory phenotypes on otherwise healthier tissue neighbors.

Important questions remain. The extent to which the mechanisms defined in experimental models translate to human ovarian aging will require validation in human follicles, which are accessible only in limited quantities and at defined stages. The relative contribution of oocyte-derived cGAMP compared with other inflammatory triggers in the aging ovary, including cellular senescence in stromal compartments and systemic inflammatory factors, remains to be quantified. Whether chronic STING activation in granulosa cells causes irreversible loss of follicles or reversible functional impairment is another open issue, as is the question of whether manipulating gap junctional coupling early in life could have unintended consequences for follicular development. Nevertheless, by tracing an unbroken mechanistic line from mitochondrial DNA leakage in aged oocytes through cGAS activation, cGAMP synthesis, CX37-dependent intercellular transfer and STING-driven inflammation in granulosa cells, Lei and colleagues have provided one of the most complete mechanistic accounts of a mammalian tissue aging process to date.

The broader significance of this work lies in its demonstration that the ovary is not merely a passive target of systemic aging but an organ whose decline is orchestrated, at least in part, by an internally generated inflammatory program. The oocyte, the longest-lived cell in the body and the custodian of the species’ genetic continuity, emerges as both the origin and the propagator of the inflammatory signal that undermines its own follicular niche. If future studies confirm these mechanisms in human ovaries and identify safe ways to intervene, the slow fade of female fertility might one day be delayed not by replacing lost eggs but by quieting the inflammatory conversation that aged oocytes impose on their surroundings, extending the reproductive window and improving ovarian health in aging women.

Subject of Research: Leaky oocytes propagate cGAS–STING signaling

Article Title: Leaky oocytes propagate cGAS–STING signaling

Article References: Biswas, S., & Stout, M. B. (2026). Leaky oocytes propagate cGAS–STING signaling. Nature Aging. https://doi.org/10.1038/s43587-026-01153-8

Image Credits: AI Generated

DOI: 10.1038/s43587-026-01153-8

Keywords: Leaky, oocytes, propagate, cGAS, STING, signaling, scientific research