Estrogen receptor alpha, a nuclear hormone receptor, plays a pivotal role in the development and progression of breast cancer. Its aberrant activation drives tumor growth, making ERα a prime target for therapeutic intervention. Current treatments often involve selective estrogen receptor modulators or degraders; however, resistance mechanisms frequently emerge, limiting their long-term efficacy. This scenario has propelled scientists to seek alternative strategies that can modulate ERα stability and function more effectively. Bufalin, a steroid compound derived from traditional Chinese medicine, emerged as an intriguing candidate through a sophisticated AI-driven discovery pipeline.

The use of artificial intelligence in drug discovery represents a transformative shift in biomedical sciences. Traditional experimental methods are labor-intensive and time-consuming, often involving trial-and-error screening of vast chemical libraries. In contrast, AI algorithms can rapidly analyze complex biological and chemical datasets, identifying promising molecules with desired biological activities. In this study, the research team deployed advanced machine learning models designed to predict molecular glues — small molecules that facilitate protein-protein interactions leading to targeted protein degradation. By leveraging extensive databases of molecular structures and interaction profiles, AI identified Bufalin as a potential mediator capable of inducing ERα degradation.

Molecular glues have garnered significant attention as an innovative class of therapeutic agents. Unlike classical inhibitors that block active sites, molecular glues facilitate new interactions between target proteins and the cellular degradation machinery, effectively tagging the protein for destruction. This mechanism allows for highly selective modulation of protein levels within the cell. Bufalin’s identification as a molecular glue is particularly noteworthy because it opens new directions in modulating challenging targets like nuclear receptors, which have traditionally been difficult to drug due to their complex regulation and conformational dynamics.

The researchers employed a multi-layered validation approach to confirm Bufalin’s activity. Initial computational predictions were followed by biophysical and biochemical assays that demonstrated Bufalin’s ability to bridge ERα with E3 ubiquitin ligases, the enzymes responsible for tagging proteins for proteasomal degradation. Structural analyses, including cryo-electron microscopy and mass spectrometry, elucidated the tri-molecular complex formed by Bufalin, ERα, and the ligase, revealing the molecular basis of the induced proximity effect. These findings confirm that Bufalin does not merely inhibit ERα but promotes its active ubiquitination and subsequent degradation.

Beyond the mechanistic insights, cell-based experiments unveiled the functional consequences of Bufalin-induced ERα degradation. Cancer cell lines reliant on ERα signaling exhibited marked decreases in proliferation upon Bufalin treatment. Moreover, transcriptional profiling revealed downstream attenuation of estrogen-responsive genes, corroborating the effective dismantling of ERα-mediated signaling pathways. Importantly, comparative studies indicated that Bufalin’s mode of action differed fundamentally from existing selective estrogen receptor degraders (SERDs), potentially circumventing common resistance pathways.

One of the remarkable aspects of this research is its demonstration of AI’s role in unearthing bioactive natural products with previously unrecognized mechanisms. Bufalin had been studied mainly for its cardiotonic and anti-inflammatory properties; however, its capacity as a molecular glue expands its therapeutic relevance substantially. This finding exemplifies how AI can bridge traditional knowledge with modern molecular pharmacology, offering a new lens through which to explore natural compound libraries for drug discovery.

The study also highlights the importance of integrative approaches combining computational predictions with experimental validations. While AI can prioritize candidates rapidly, empirical evidence remains critical to decipher complex biological interactions and to understand pharmacodynamics and toxicity profiles. The researchers’ comprehensive methodology, encompassing in silico modeling, biochemical assays, and cellular analyses, set a rigorous standard for future work in this rapidly evolving field.

Bufalin’s potential therapeutic application extends into breast cancer treatment paradigms where hormone receptor status is a critical determinant. Since ERα-positive breast cancers constitute a majority of breast cancer diagnoses worldwide, the introduction of a molecular glue degrader offers a desperately needed option, especially for patients who develop resistance to endocrine therapies. Future clinical investigation will be necessary to evaluate Bufalin’s safety, efficacy, and pharmacological characteristics in vivo, but the preclinical results are undeniably promising.

This research also paves the way for the discovery of other molecular glue degraders targeting a broad spectrum of disease-relevant proteins. By refining and expanding AI models, the identification process can be diversified and accelerated, potentially transforming how pharmaceutical companies approach ‘undruggable’ targets. The modular nature of molecular glues allows for tailored interventions designed for selective degradation, reducing off-target effects and improving patient outcomes.

The discovery of Bufalin as an ERα molecular glue degrader exemplifies how blending AI with molecular biology can overcome longstanding drug development hurdles. This paradigm shift in drug design has far-reaching implications beyond oncology, potentially influencing treatments for neurodegenerative diseases, immune disorders, and viral infections, where aberrant protein regulation plays a pathogenic role. By targeting protein stability rather than merely function, clinicians may gain access to a new class of interventions with greater specificity and durability.

Furthermore, the study emphasizes the significance of multidisciplinary collaboration. Chemists, biologists, data scientists, and clinicians joined forces to translate AI-generated hypotheses into tangible experimental evidence. Such collaborative ecosystems are essential for harnessing the full power of AI-enhanced drug discovery, ensuring that computational advances are grounded in biological reality and clinical relevance.

In addition to its scientific merit, this breakthrough carries profound implications for drug affordability and accessibility. Artificial intelligence enables more cost-effective exploration of chemical space, potentially shortening timelines and reducing expenses associated with bringing novel therapeutics to market. This could democratize access to cutting-edge treatments, particularly for diseases with high unmet medical needs like hormone receptor-positive breast cancer.

Looking forward, the integration of AI-driven methods with emerging technologies such as single-cell proteomics, CRISPR screens, and high-throughput structural biology could further revolutionize our understanding of protein interactions and degradation pathways. Bufalin’s identification as a molecular glue may represent just the tip of an iceberg, with many more druggable mechanisms awaiting discovery through sophisticated computational and experimental synergies.

In conclusion, harnessing artificial intelligence to uncover Bufalin as a molecular glue degrader of estrogen receptor alpha represents a landmark achievement in contemporary biomedical research. This study not only sheds light on a novel mechanism to combat hormone-driven cancers but also showcases the transformative power of AI-guided drug discovery. As the pharmaceutical landscape evolves, the fusion of computational ingenuity with biological insight promises to unlock new frontiers in disease treatment, heralding an era of more precise, effective, and personalized medicine.

Subject of Research: Identification of Bufalin as a molecular glue degrader targeting estrogen receptor alpha using artificial intelligence.

Article Title: Harnessing artificial intelligence to identify Bufalin as a molecular glue degrader of estrogen receptor alpha

Article References:
Jiang, S., Liu, K., Jiang, T. et al. Harnessing artificial intelligence to identify Bufalin as a molecular glue degrader of estrogen receptor alpha. Nat Commun 16, 7854 (2025). https://doi.org/10.1038/s41467-025-62288-7

Image Credits: AI Generated

Traditional drug development typically begins with identifying a protein implicated in disease pathology, such as a mutated receptor on cancer cells, followed by exhaustive screening of molecular libraries to find compounds capable of binding effectively to that target site. This approach is not only costly and slow but also plagued by high attrition rates, with only a fraction of candidates advancing through costly clinical trials. The newly developed BInD model circumvents these limitations by directly designing drug molecules informed solely by the three-dimensional structure of the target protein, without reliance on any pre-existing molecular data or known binders. This capability signals a paradigm shift toward streamlined and more precise therapeutic design.

At the core of BInD’s innovation is its “simultaneous design” mechanism, which integrates molecular generation and binding evaluation into a unified process. Unlike prior AI drug design systems that separately generated candidate molecules and then assessed their binding propensity—often resulting in inefficiencies and suboptimal candidates—BInD models the complex interplay of non-covalent interactions between the prospective drug molecule and the protein’s binding pocket during the molecule’s construction. This approach ensures that every atom, covalent bond, and intermolecular interaction is instantiated in concert to optimally complement the target’s structural features, greatly enhancing the likelihood of producing molecules with high binding affinity and desirable stability.

The AI model’s architecture leverages a diffusion-based generative framework. Such diffusion models begin with random noise and progressively refine structures by simulating a stochastic denoising process, enabling the generation of highly realistic molecular geometries. This methodology is akin to the recent breakthroughs exemplified by AlphaFold 3, the Nobel Prize-winning tool renowned for accurately predicting protein folding and protein-ligand complexes in silico. However, while AlphaFold 3 outputs spatial atom coordinates primarily for prediction, BInD introduces chemically grounded constraints during molecule generation. These include empirical bond lengths, angular relationships, and atom-protein proximity data derived from chemical principles, greatly improving the chemical plausibility and synthetic feasibility of the designed molecules.

A unique distinction of BInD lies in its capacity for multi-objective optimization during the design phase. Drug discovery is a multifaceted challenge, requiring candidates not only to bind strongly to their target but also to exhibit favorable drug-like properties, such as bioavailability, metabolic stability, and minimized toxicity. Prior AI systems frequently optimized a limited subset of these parameters, often at the expense of others, leading to candidates unsuitable for clinical development. BInD’s architecture balances these diverse objectives simultaneously, generating molecules that harmonize binding affinity with pharmacokinetic and physicochemical properties, potentially accelerating the pipeline from initial design to viable therapeutic candidates.

To further enhance its design capabilities, the team incorporated a knowledge-based guidance system grounded in established chemical laws, which steers the diffusion process toward chemically sound configurations. This innovation ensures that the model respects fundamental molecular constraints, such as valid valency rules and realistic interatomic distances, preventing the generation of chemically implausible structures. Moreover, BInD utilizes an iterative optimization strategy that reuses superior binding patterns discovered in prior generation cycles, fostering the continual improvement of candidate molecules without necessitating additional retraining of the model.

One of the most compelling demonstrations of BInD’s effectiveness is its success in generating molecules that selectively target mutated residues of the epidermal growth factor receptor (EGFR), a critical oncogenic protein frequently altered in various cancers. By tailoring drug candidates to the unique structural aberrations presented by mutated EGFR, the AI model offers a promising pathway toward highly selective cancer therapeutics with potentially reduced off-target effects, addressing one of the paramount challenges in oncology drug design.

This research heralds an evolution beyond the group’s earlier efforts, which required explicit prior knowledge of molecular interaction conditions to inform binding patterns. The current system’s ability to autonomously learn and internalize the key features for robust target binding—absent any molecular priors—marks a substantial stride toward genuinely autonomous drug design. Professor Woo Youn Kim emphasized that this AI model “can learn and understand the key features required for strong binding to a target protein, and design optimal drug candidate molecules—even without any prior input,” highlighting the transformative potential of this technology to reshape pharmaceutical innovation.

The implications of this work extend beyond accelerated drug discovery; by embedding fundamental chemical interaction principles into the generative process, BInD promises heightened reliability and reduced attrition in downstream development phases. The resultant acceleration not only reduces costs associated with lengthy trial-and-error synthesis and screening but also opens avenues for tackling previously “undruggable” targets lacking extensive molecular data.

This innovative study was carried out by a research team led by Professor Woo Youn Kim in KAIST’s Department of Chemistry and includes co-first authorship by Ph.D. candidates Joongwon Lee and Wonho Zhung. Their findings were published in the prestigious international journal Advanced Science on July 11, 2025. This work received financial support from the National Research Foundation of Korea and the Ministry of Health and Welfare.

As artificial intelligence continues to penetrate every facet of biomedical research, models like BInD exemplify the convergence of computational sophistication and chemical intuition necessary to surmount the persistent bottlenecks in drug design. The emerging capability to expediently generate chemically viable, multi-objective optimized drug candidates tailored to protein structures holds immense promise to accelerate therapeutic discovery, particularly in complex disease areas like cancer where mutation-specific targeting can offer profound clinical benefits.

The next steps for this line of research include experimental validation of the AI-designed molecules, expansion to a broader spectrum of protein targets, and integration into automated synthesis and screening platforms. Should these developments proceed as anticipated, BInD and similar AI-powered diffusion models stand poised to usher in a new era where drug discovery operates at the fusion of data-driven design and fundamental chemical principles, ultimately enabling more precise, effective, and rapidly developed medicines for patients worldwide.

Subject of Research: AI-driven structure-based drug design using diffusion models for cancer-targeting mutations

Article Title: Bond and Interaction-Generating Diffusion Model for Multi-Objective Structure-Based Drug Design

News Publication Date: 11-Jul-2025

Web References: DOI: 10.1002/advs.202502702

Image Credits: KAIST

Keywords: Health care