The research team, led by Zhilan Zhang and Ping Zhou of Xiangya Hospital of Central South University together with colleagues at Central South University Xiangya School of Medicine Affiliated Haikou Hospital, began by examining human hepatocellular carcinoma samples. They found that several immune-related molecules were prominently expressed in tumor tissue: LAG-3, an inhibitory receptor expressed on exhausted T cells; CXCL10, a chemokine secreted in response to inflammatory signals; CXCR3, the receptor on T cells that binds CXCL10; and CD8, the defining marker of cytotoxic T lymphocytes. This molecular signature hinted that the CXCL10/CXCR3 axis, a well-known trafficking pathway that guides activated T cells into inflamed tissue, was active in the liver tumor microenvironment, and that LAG-3-mediated suppression might be restraining the very T cells the pathway was drawing in.

To test the functional significance of these observations, the investigators turned to hepatocellular carcinoma-bearing mice. When the animals received microwave ablation, a technique that uses electromagnetic energy to generate lethal heat within tumor tissue, the researchers observed a striking change in the tumor-infiltrating lymphocyte population: LAG-3 expression rose on subsets of the infiltrating T cells. In other words, ablation did not merely shrink the tumor; it reshaped the immune landscape, drawing lymphocytes into the remaining tumor while simultaneously increasing the prevalence of the inhibitory checkpoint that can render those lymphocytes dysfunctional. This observation provided a mechanistic explanation for why local ablation alone often fails to prevent recurrence and why pairing it with checkpoint blockade could be advantageous.

LAG-3, or lymphocyte-activation gene 3, has attracted intense interest in oncology because it regulates T cell exhaustion through pathways that are distinct from those of the better-known checkpoint molecules PD-1 and CTLA-4. By binding its ligands and transmitting inhibitory signals, LAG-3 dampens the proliferative and cytotoxic capacity of T cells. Blocking LAG-3 with antibodies releases this brake, and several anti-LAG-3 agents are already in clinical development for solid tumors and hematologic malignancies. The new study asked a specific and clinically important question: does LAG-3 blockade complement microwave ablation in hepatocellular carcinoma, and if so, through what molecular circuitry?

The answer, according to the mouse experiments, is a clear yes. Compared with either microwave ablation or anti-LAG-3 therapy alone, the combined treatment produced a synergistic anti-tumor effect. Mice receiving both interventions survived significantly longer and showed markedly inhibited tumor growth. Beyond the gross measures of tumor burden, the combination also remodeled the tumor immune microenvironment in ways that favored immune attack: tumor-infiltrating lymphocytes increased in number, serum levels of CXCL10 rose, the population of CXCR3-positive CD8+ T cells expanded, and the cytotoxic activity of CD8+ T cells was enhanced. The convergence of increased chemokine production with greater numbers of chemokine-receptor-bearing killer cells suggested that the CXCL10/CXCR3 axis was the engine driving the therapeutic synergy.

To confirm that the chemokine axis was genuinely required rather than merely correlated, the researchers performed two decisive loss-of-function experiments. First, they blocked CXCL10 in mice receiving the combined therapy. Neutralizing the chemokine weakened CD8+ T cell function, demonstrating that the chemokine signal is necessary for the enhanced cytotoxic response. Second, they depleted or blocked CD8+ T cells themselves under the combined regimen. This maneuver promoted tumor growth and impaired the anti-tumor benefit, establishing CD8+ T cells as the essential cellular mediators of the combination effect. Together, the two experiments trace the causal chain: microwave ablation and LAG-3 blockade act together to elevate CXCL10, CXCL10 engages CXCR3 on CD8+ T cells to recruit and activate them, and activated CD8+ T cells execute the tumor killing.

The mechanistic picture is biologically plausible and fits with established immunology. Thermal injury from ablation is known to release tumor antigens and danger signals that provoke local inflammation, and inflammatory cytokines such as interferon-gamma induce CXCL10 production in stromal and immune cells. At the same time, the influx of newly activated T cells into this inflammatory environment creates a larger pool of cells vulnerable to LAG-3-mediated inhibition, which likely explains why LAG-3 expression climbed on tumor-infiltrating lymphocytes after ablation in the study. Removing that constraint with an anti-LAG-3 antibody allows the newly recruited CD8+ T cells to proliferate, produce cytotoxic molecules such as granzyme B, and sustain their attack on residual tumor cells, including microscopic deposits that ablation cannot physically reach.

Hepatocellular carcinoma remains one of the most lethal and rapidly increasing cancers worldwide, frequently diagnosed at an advanced stage when curative resection or transplantation is no longer feasible. Immune checkpoint inhibitors have transformed the treatment landscape in recent years, but only a fraction of patients respond durably, and resistance remains a central clinical problem. Locoregional therapies such as microwave ablation are standard of care for early-stage disease, yet recurrence is common. A regimen that combines the antigen-releasing and inflammation-generating effects of ablation with checkpoint blockade that preserves T cell function could address both limitations simultaneously, converting an otherwise localized treatment into an in situ cancer vaccine while ensuring the recruited immune cells retain full killing capacity.

The authors emphasize that the therapeutic potential of combining microwave ablation with LAG-3 blockade had already been demonstrated in various cancers, but its specific efficacy and molecular mechanisms in hepatocellular carcinoma had remained unclear. By identifying the CXCL10/CXCR3 pathway as the mechanistic bridge, the study fills that gap and provides biomarkers that could be used to monitor or stratify patients. Serum CXCL10 levels, the frequency of CXCR3-positive CD8+ T cells, and LAG-3 expression on tumor-infiltrating lymphocytes are all measurable in clinical settings and could serve as pharmacodynamic indicators of whether the combination is engaging its intended immune circuitry in human trials.

The study was supported by the Natural Science Foundation of Hainan Province, and all animal procedures were approved by the Animal Care and Ethical Standards Committee of Central South University Xiangya School of Medicine Affiliated Haikou Hospital. The authors declared no competing financial interests. As with any preclinical finding, important caveats apply before the results can inform patient care. Mouse models of hepatocellular carcinoma do not fully recapitulate the immunosuppressed, cirrhotic, hepatitis- or metabolically driven liver environment in which human tumors arise, and the dosing, timing, and sequencing of ablation relative to checkpoint blockade will require careful optimization in clinical studies. Nevertheless, the identification of a defined chemokine-dependent mechanism gives the field a concrete target around which to design combination trials, and it reinforces a growing consensus in immuno-oncology: the most effective treatments will be those that simultaneously generate the raw materials of an immune response and remove the brakes that prevent that response from succeeding.

Historically, activation of CD8⁺ T cells necessitates antigen processing and presentation by major histocompatibility complex class I (MHC-I) molecules on antigen-presenting cells (APCs). Conventional dendritic cells type 1 (cDC1s) have been recognized as pivotal players in cross-presentation — a process whereby exogenous antigens are presented on MHC-I molecules — which is critical for cytotoxic T cell priming in viral infections, tumor immunity, and with certain vaccine modalities such as protein- and cDNA-based vaccines. Despite this, the precise role of cDC1 cells and the associated cross-presentation machinery in the context of mRNA–LNP vaccines had not been firmly established, prompting a detailed investigation in this latest study led by Jo, Li, Thakur, and colleagues.

The researchers provide compelling evidence that, contrary to prior assumptions, effective CD8⁺ T cell priming following mRNA–LNP vaccination does not solely depend on cDC1 cells or the canonical WDFY4-dependent cross-presentation pathway. Utilizing genetically engineered mouse models deficient in cDC1 cells and components essential for classical cross-presentation, the team demonstrated that CD8⁺ T cell responses were maintained. This indicates a redundancy in dendritic cell subsets capable of instigating cytotoxic T cell immunity, thus broadening the understanding of APC roles in response to mRNA vaccines.

One of the pivotal findings is that both cDC1 and cDC2 dendritic cell subsets can independently prime CD8⁺ T cells, suggesting a level of functional plasticity that can compensate for the absence of one subset. Crucially, though these individually primed CD8⁺ T cells exhibited distinct phenotypic characteristics, both subsets were capable of mediating potent anti-tumor immunity and the formation of immunological memory. This finding has profound implications for vaccine design, highlighting the resilience and adaptability of cellular immune responses elicited by mRNA–LNP platforms.

Delving further into the mechanisms underlying these observations, the study uncovers the significant role of a process known as “cross-dressing,” wherein cDCs acquire peptide–MHC-I complexes directly from non-hematopoietic cells. This alternative pathway of antigen presentation substantially contributes to the priming of CD8⁺ T cells during mRNA vaccination. Notably, the effectiveness of cross-dressing relies on type I interferon signaling, a critical component of the innate immune response that enhances the ability of dendritic cells to stimulate T cell responses.

This discovery sheds light on why mRNA–LNP vaccines can potently activate CD8⁺ T cells against antigens that may not be directly encoded by the vaccine itself, a phenomenon that could not be easily explained by classical antigen presentation pathways alone. The induction of cross-dressing by mRNA vaccines potentially broadens the spectrum of antigen targets, implying that these vaccines might harness unconventional but highly efficient immune activation routes.

Importantly, the study’s insights challenge and expand the current paradigms of immune activation by nucleic acid vaccines. By demonstrating that mRNA–LNP vaccines bypass strict reliance on cDC1 and cross-presentation, the research opens avenues for optimizing vaccine formulations to exploit multiple dendritic cell subsets and innate immune pathways, potentially enhancing the breadth, potency, and durability of CD8⁺ T cell responses.

The broader implications extend to cancer immunotherapy, where robust and durable cytotoxic T cell responses are critical for tumor clearance. The ability of mRNA vaccines to stimulate CD8⁺ T cells through unconventional dendritic cell activation pathways may translate into improved strategies for cancer vaccine development. Moreover, understanding the role of cross-dressing could inform approaches to circumvent immune evasion mechanisms employed by tumors or persistent viral infections.

From a mechanistic perspective, the study also underscores the intricate interplay between innate signaling pathways, such as type I interferon, and antigen presentation processes. Type I interferons appear to orchestrate the acquisition of peptide–MHC-I complexes by dendritic cells, reinforcing the notion that successful vaccine-induced immunity depends on finely-tuned coordination between innate and adaptive immune components.

The findings encourage revisiting the design of adjuvants and delivery systems within mRNA vaccines to harness or amplify these unconventional pathways. Tailoring vaccine constructs to promote enhanced cross-dressing and engagement of both cDC1 and cDC2 subsets could yield more potent and broadly effective vaccines, not only against infectious diseases but also in immuno-oncology.

In summary, the research led by Jo et al. reveals an unexpected flexibility in dendritic cell-mediated CD8⁺ T cell priming by mRNA–LNP vaccines, highlighting cross-dressing as a substantial contributor to their immunogenic profile. This revelation enriches the conceptual framework of vaccine immunology and provides a platform for innovation in next-generation vaccine strategies focused on eliciting robust cellular immunity.

As the field progresses, these mechanistic insights furnish a foundation for developing mRNA vaccines capable of eliciting comprehensive immune protection through multiple complementary antigen presentation pathways. Such advances hold promise for addressing emerging infectious diseases and improving therapeutic vaccine design for cancer and chronic infections globally.

Subject of Research:
Unconventional pathways of CD8⁺ T cell priming induced by mRNA vaccines involving dendritic cell cross-dressing and type I interferon-dependent mechanisms.

Article Title:
mRNA vaccines engage unconventional pathways in CD8⁺ T cell priming.

Article References:
Jo, S., Li, L., Thakur, C. et al. mRNA vaccines engage unconventional pathways in CD8⁺ T cell priming. Nature (2026). https://doi.org/10.1038/s41586-026-10353-6

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41586-026-10353-6

Vaccines traditionally function by inducing strong antibody responses that can neutralize pathogens upon initial exposure. However, high mutation rates in viruses often enable them to evade these antibody-mediated defenses over time, rendering vaccines less effective. This limitation has underscored the importance of eliciting robust T cell responses, especially those mediated by CD8+ cytotoxic T lymphocytes, which identify and destroy infected cells and can recognize viral mutations more flexibly. Enhancing these responses remains a central challenge in vaccinology, one that the latest research from the University of Pennsylvania seeks to address through mRNA vaccine platforms augmented with IL-12.

IL-12 is a key immunoregulatory cytokine involved in the differentiation and activation of T cells. It promotes the development of T helper 1 (Th1) cells and stimulates the production of interferon-gamma (IFN-γ), thereby enhancing cellular immunity against intracellular pathogens and malignancies. While IL-12 is naturally secreted during infections, delivering it exogenously as part of a vaccine adjuvant requires sophisticated technology to ensure localized, controlled expression without systemic toxicity. The advent of lipid nanoparticle (LNP)-encapsulated mRNA vaccines offers an ideal vehicle to safely deliver IL-12, harnessing the body’s own cells to produce the cytokine with precision.

The study published in Science Immunology demonstrates the transformative potential of IL-12 mRNA-LNP adjuvants in preclinical mouse models. By co-administering IL-12 encoding mRNA alongside mRNA vaccines targeting SARS-CoV-2 and influenza, researchers observed a pronounced amplification of antigen-specific CD8+ T cell responses. These cytotoxic T cells exhibited enhanced functional profiles, including increased proliferation, cytokine production, and cytolytic activity, translating to superior protection against viral challenge. Moreover, the IL-12 adjuvant improved immunity against non-viral threats, such as melanoma tumors and Listeria monocytogenes bacterial infections, illustrating broad applicability.

This work addresses a historical bottleneck in vaccine science—the difficulty in eliciting strong and durable CD8+ T cell responses. Traditional adjuvants have had limited success in this domain, often focusing more on antibody generation. The flexibility of mRNA technology allows for the co-delivery of immunomodulatory messages like IL-12, enabling finely tuned immune modulation. According to senior author Christopher A. Hunter of Penn Vet, the synergy between mRNA vaccine platforms and IL-12 adjuvants points toward a future where vaccines are not only more effective but also require fewer doses, potentially reducing side effects and improving compliance.

The implications of IL-12 mRNA vaccines extend well beyond infectious diseases. Cancer immunotherapy stands to benefit from this innovation, as mounting a vigorous T cell-mediated attack against tumors is essential for successful treatment. IL-12’s capacity to invigorate cytotoxic lymphocyte responses may address the immunosuppressive tumor microenvironment, boosting the efficacy of existing or novel tumor vaccines and immunotherapies. Susan M. Domchek, director of the Abramson Cancer Center’s Basser Cancer Interception Institute, emphasizes the clinical promise of this technology, expressing optimism about its rapid translation into treatments for patients at high risk of developing cancer.

Central to these discoveries is the collaborative ecosystem within the University of Pennsylvania, bringing together experts in cytokine biology, vaccine research, and nanoparticle engineering. The fusion of Anthony T. Phan’s focus on CD8+ T cells, Drew Weissman’s pioneering work in mRNA vaccine development—recognized globally through his 2023 Nobel Prize—and Mohamad-Gabriel Alameh’s expertise in nanoparticle design has culminated in this groundbreaking study. This multidisciplinary approach underscores how academic environments catalyze novel biomedical solutions.

Further exploration of cytokine mRNAs as vaccine adjuvants is underway, as the team investigates additional immune modulators that can be similarly encoded and delivered. The potential to tailor vaccine-induced immunity through rational design of mRNA adjuvants represents a paradigm shift, moving beyond conventional empiricism to mechanistic precision in immunization strategies. For instance, ongoing collaborations aim to determine if IL-12 can enhance HIV vaccine candidates and adapt this technology to veterinary infectious diseases such as avian influenza, broadening the impact across human and animal health.

From a practical standpoint, IL-12 inclusion in mRNA vaccine regimens could reduce the number of necessary booster shots and vaccine dosages. By intensifying cellular immunity, vaccines become more potent with fewer administrations, which can decrease cost, logistical burdens, and patient discomfort. This advance has profound public health implications, especially for resource-limited settings and populations hesitant about frequent injections or associated side effects.

Funded by the National Institutes of Health’s Adjuvant Discovery Program, the Basser Cancer Interception Institute, and other entities, this research exemplifies the importance of sustained investment in fundamental immunology and vaccine science. Continued support accelerates translation from bench to bedside, fostering innovations that have the potential to reshape preventive medicine and immunotherapy at large. The research team’s comprehensive publication outlines both mechanistic insights and translational benefits, positioning IL-12 mRNA-LNPs as next-generation vaccine adjuvants.

In addition to enhancing immune protection against viruses that cause respiratory illnesses, this approach holds promise for combating evolving pathogens and malignancies that have traditionally evaded durable immune control. IL-12’s role as a molecular “booster” of T cell immunity could provide a crucial backup when neutralizing antibodies wane or fail. As global health challenges continue to evolve, such refined immunomodulation strategies may become indispensable tools in the fight against infectious and non-infectious diseases.

Overall, the incorporation of IL-12 into mRNA vaccine platforms represents a sophisticated and emerging frontier that combines immunology, molecular biology, and nanotechnology. It offers a compelling example of how understanding cytokine biology can be leveraged through innovative platforms to produce vaccines that not only prevent disease but also potentially transform therapeutic approaches against cancer. The coming years will reveal how this approach performs in clinical trials and its ultimate impact on public health worldwide.

Subject of Research: Animals

Article Title: An Il12 mRNA-LNP adjuvant enhances mRNA vaccine–induced CD8 T cell responses

News Publication Date: 6-Jun-2025

Web References:
http://dx.doi.org/10.1126/sciimmunol.ads1328

Keywords: mRNA vaccines, vaccine research, T cell responses, cytokines, SARS CoV 2, COVID 19 vaccines

Dendritic cells are crucial orchestrators of the immune system’s capacity to detect and eliminate malignancies. These highly specialized antigen-presenting cells bridge innate and adaptive immunity by priming CD8⁺ T cells, the immune system’s frontline cytotoxic effector cells. However, tumours often co-opt suppressive molecular pathways to impair DC function, thereby blunting effective anti-tumour T cell responses. STAT3, a transcription factor frequently activated in the tumour microenvironment, has long been implicated in immune evasion through its suppressive influence on dendritic cell activity.

Zhou and colleagues developed an innovative therapeutic approach using SD-36, a highly selective molecule engineered to degrade STAT3 protein within DCs, thus removing this suppressive "brake" on the immune system. Their experiments began with testing SD-36 efficacy in immune-deficient mouse models bearing MC38 colon carcinoma tumours. Notably, low-dose SD-36 treatment failed to inhibit tumour growth in NSG mice, which lack both innate and adaptive immunity, and similarly showed no effect in Rag1-deficient mice that have innate but defective adaptive immunity. These findings underscored the essential role of an intact adaptive immune system for SD-36’s antitumour activity.

To further pinpoint the immune effectors involved, the authors selectively depleted CD8⁺ T cells in wild-type mice bearing tumours and found that SD-36’s anti-tumour efficacy was completely abolished. This pivotal observation confirmed that the therapeutic benefit relied fundamentally on CD8⁺ T cell function. Flow cytometric analyses demonstrated that SD-36 significantly increased the proportion of tumour-infiltrating CD8⁺ T cells expressing critical cytotoxic molecules such as TNF, IFNγ, and granzyme B, across multiple tumour models. This effect delineates a scenario in which STAT3 degradation in DCs indirectly mobilizes a robust cytotoxic T cell response, effectively curtailing tumour progression.

Delving into the mechanistic underpinnings, the study elegantly revealed the role of classical dendritic cells type 1 (cDC1s) in mediating SD-36’s therapeutic effects. In Batf3-deficient mice, which lack cDC1s, SD-36 failed to suppress tumour growth, indicating that these DC subsets are indispensable for the drug’s efficacy. Further dissection using STAT3 knockout mice substantiated that the presence of functional STAT3 in DCs was necessary for SD-36 to exert its anti-tumour activity, as genetic ablation of STAT3 negated the compound’s benefits.

Intriguingly, the authors uncovered a dynamic reprogramming of transcription factor signaling within cDC1s following SD-36 treatment. Phosphorylation levels of STAT3 were markedly diminished, while STAT5 phosphorylation was enhanced—signifying a molecular switch within DCs. This shift was accompanied by upregulation of maturation and co-stimulatory molecules such as MHC class I, MHC class II, and CD80, hallmark indicators of enhanced antigen-presenting capacity and DC activation. The data suggest that STAT3 acts as a negative regulator that suppresses STAT5-driven DC maturation, and that its targeted degradation effectively “releases the brakes” on DC function.

Profound mechanistic clarity was gained through adoptive transfer experiments where wild-type or STAT3-deficient cDC1s were introduced into Batf3-deficient mice. SD-36 restored anti-tumour immunity only when STAT3-competent DCs were present, cementing the conclusion that DC-intrinsic STAT3 degradation is crucial for therapeutic efficacy. Additional genetic validation with STAT5b knockout DCs illustrated that SD-36’s beneficial effects absolutely required functional STAT5 signaling, reinforcing the concept of a STAT3/STAT5 balance that dictates DC phenotype and immune outcomes.

The therapeutic potential of STAT3 degradation was further amplified by combining SD-36 with immune checkpoint blockade (ICB) targeting PD-L1. In highly immunogenic MC38 tumours, anti-PD-L1 therapy alone suppressed tumour growth, whereas in poorly immunogenic B16F10 melanomas, PD-L1 blockade was ineffective. Remarkably, SD-36 monotherapy slowed tumour progression in both models, and the combination with anti-PD-L1 yielded profoundly synergistic inhibition of tumour growth. Complementary ex vivo studies with human ovarian cancer-derived DCs and T cells showed enhanced polyfunctional T cell priming following dual SD-36 and PD-L1 blockade treatment, highlighting translational prospects.

Building upon this foundation, the authors introduced a second-generation STAT3 degrader, SD-2301. Unlike SD-36, which recruits the cereblon–cullin 4A E3 ligase complex for protein degradation, SD-2301 employs a high-affinity VHL ligand to engage the VHL–cullin 2 complex, resulting in significantly improved potency. In vivo experiments revealed that SD-2301 achieved superior STAT3 degradation in DCs and demonstrated greater efficacy in controlling tumour progression at substantially lower doses compared to SD-36.

Functionally, SD-2301 mirrored SD-36 in enhancing effector CD8⁺ T cell populations, with increased expression of IFNγ and granzyme B, alongside upregulation of DC maturation markers within the tumour microenvironment. Crucially, SD-2301 exerted no deleterious effects on tumour vascularization or animal body weight, supporting its safety profile. As with SD-36, SD-2301 synergized robustly with PD-L1 checkpoint blockade to further restrain tumour growth, reinforcing the concept of STAT3 degradation as a powerful adjunct to existing immunotherapies.

Pharmacokinetic profiling of SD-2301 showed favorable attributes including slow clearance and high plasma exposure, critical parameters for clinical translation. High selectivity for STAT3 versus other STAT family members was confirmed in human peripheral blood mononuclear cells, suggesting a minimized risk of off-target effects. These findings collectively underscore the therapeutic promise of targeted STAT3 degradation strategies to reprogram DCs and invigorate anti-tumour immunity.

This seminal work by Zhou et al. thus elegantly illuminates a novel immunotherapeutic paradigm wherein simultaneous inhibition of immunosuppressive STAT3 and activation of stimulatory STAT5 in dendritic cells unleashes potent cytotoxic T cell responses against cancer. The development of highly selective degrader molecules like SD-36 and SD-2301 provides powerful chemical tools to manipulate this axis, opening new avenues for combination therapies with immune checkpoint blockade. The clarity of mechanism, robust preclinical efficacy, and promising translational relevance make this approach a compelling candidate for advancing cancer immunotherapy.

As the field moves forward, targeting transcription factor balance within antigen-presenting cells may become a cornerstone strategy for overcoming tumour immune evasion. By specifically enhancing DC function without globally suppressing STAT3 in all cells, such approaches could minimize collateral immunosuppression and toxicities. The dual effects of STAT3 degradation—releasing immune suppression while promoting DC maturation and T cell priming—may be especially advantageous in “cold” tumours resistant to conventional immunotherapies. Moreover, the synergistic potential with PD-L1 blockade suggests that future clinical regimens could be designed to maximize durable remissions.

In summary, the findings reveal a sophisticated interplay between STAT3 and STAT5 pathways controlling dendritic cell programming that can be harnessed to boost tumour immunity. Utilizing PROTAC technology to selectively degrade STAT3 in DCs represents a leap forward in immuno-oncology, demonstrating that finely tuned modulation of transcriptional networks in immune cells can powerfully reshape anti-cancer immune responses. This study paves the way for a new class of combinatorial therapies that strategically liberate dendritic cells from tumour-induced checkpoints, ultimately empowering the adaptive immune system to achieve sustained tumour control.

Subject of Research: The role of STAT3 and STAT5 transcription factor balance in dendritic cells governing anti-tumour immunity and the therapeutic potential of STAT3 degradation.

Article Title: STAT5 and STAT3 balance shapes dendritic cell function and tumour immunity

Article References:
Zhou, J., Tison, K., Zhou, H. et al. STAT5 and STAT3 balance shapes dendritic cell function and tumour immunity.
Nature (2025). https://doi.org/10.1038/s41586-025-09000-3

Image Credits: AI Generated