The team, led by Akash Kumar and Deepak Sehgal of the Shiv Nadar Institution of Eminence, with collaborators at the All India Institute of Medical Sciences, the University of Pittsburgh School of Medicine and King Saud University, began with computational work rather than wet-lab experiments. Using AlphaFold to predict the three-dimensional structure of the membrane protein from the Delta variant, refined through the YASARA energy-minimization server, they docked two copies of the protein together with ClusPro and analyzed the resulting dimer interface with PDBSum. The interaction analysis revealed one salt bridge, seven hydrogen bonds and 215 non-bonded van der Waals contacts holding the homodimer together. Four residues emerged as particularly important: glutamate 115, tyrosine 39, lysine 50 and glutamate 135. Glutamate 115 sits in the hinge region that governs the switch between the long and short conformations of the membrane protein dimer, a transition that cryo-electron microscopy studies had previously shown to be essential for virus assembly. Glutamate 135 lies in the C-terminal tail, where acidic residues help stabilize the intermolecular contacts needed for higher-order particle structure.

Molecular dynamics simulations then put the docked dimer through its paces. The team ran a 500-nanosecond simulation at 300 Kelvin and 1 bar using the OPLS force field in Desmond, tracking root-mean-square deviation, residue fluctuations, radius of gyration and solvent-accessible surface area. The results painted a picture of a rigid core with flexible edges: the C-alpha RMSD rose sharply at first, then plateaued at a stable value of roughly 8 to 9 angstroms, indicating that the complex relaxed into a new but durable conformation. The radius of gyration decreased and stabilized, and the solvent-accessible surface area declined, both signs that the dimer was packing more tightly over time. Only the peripheral helices and terminal tails showed meaningful mobility, consistent with the idea that membrane protein dimers act as stiff scaffolds that impose curvature on the viral membrane and provide a platform onto which the spike, envelope and nucleocapsid proteins are organized.

With the computational groundwork laid, the researchers turned to the baculovirus expression vector system, a workhorse of industrial protein production. They cloned the membrane gene into a recombinant bacmid and transfected Sf-21 insect cells, generating a P0 viral stock that was amplified and used to infect 500 million cells at a multiplicity of infection of 10. After 96 hours of incubation at 27 degrees Celsius, the cells were lysed with a Dounce homogenizer and sonication, and both the lysate and the concentrated culture medium were loaded onto a 20 to 80 percent sucrose gradient for ultracentrifugation at 27,000 rpm in an SW41 Ti rotor. A visible ring appeared at the junction of the 40 and 50 percent sucrose layers, corresponding to a particle density of 1.16 to 1.18 grams per milliliter, closely matching the density reported for purified SARS-CoV-2 virus-like particles and infectious coronavirions in earlier studies. SDS-PAGE and Western blotting confirmed the presence of the roughly 25-kilodalton membrane protein in these fractions, while uninfected Sf-21 cell controls produced no comparable structures.

The biophysical characterization confirmed that these were genuine, discrete particles rather than protein aggregates. Dynamic light scattering on a Horiba Zeta Sizer measured a hydrodynamic diameter of approximately 225 nanometers and a zeta potential of around minus 15 to minus 25 millivolts, the negative surface charge typical of enveloped virus particles. Field-emission scanning electron microscopy, at 40,000-fold magnification, revealed spherical to slightly pleomorphic particles of roughly 180 to 200 nanometers, and transmission electron microscopy of negatively stained samples showed numerous electron-dense spherical particles of about 200 nanometers embedded in a background of amorphous material. Taken together, the DLS, FESEM, TEM and immunoblotting data support the conclusion that expression of the membrane protein alone is sufficient to drive the budding of virus-like particles in this system, a capability that had not previously been demonstrated for the SARS-CoV-2 membrane protein in isolation.

The immunological evaluation was carried out in female BALB/c mice, aged four to six weeks, under protocols approved by the Institutional Animal Ethics Committee of Rodent Research India and compliant with Indian CCSEA standards. Each mouse received 100 micrograms of purified particles subcutaneously with 0.5 percent aluminum hydroxide adjuvant, while control animals received sterile phosphate-buffered saline. Blood samples collected at days 7, 14, 35 and 45 told a clear story. Serum IgG in the immunized group was already elevated by day 7, remained above control levels at all subsequent time points, and showed a gradual decline at days 35 and 45, a pattern consistent with a strong primary immune response rather than a loss of immunological memory. IgM peaked around day 7 and then declined but stayed above control levels through day 45. IgA, typically associated with mucosal immunity, was higher in the immunized group at every time point, indicating that the particles could stimulate class-switched antibody responses in addition to the early IgM wave. Isotype profiling on the final serum collection showed a broad antibody repertoire, with measurable IgG subclasses, strong IgM reactivity, and robust signals for kappa light chains and total heavy-plus-light chains, confirming active production of functional immunoglobulins.

The cellular arm of immunity proved equally responsive. Sandwich ELISA measured serum interferon-gamma, which was consistently higher in the immunized group with peak levels around day 14 and sustained elevation at days 35 and 45, a signature of T helper type 1 activation that mirrors findings from earlier SARS-CoV and SARS-CoV-2 structural protein vaccine studies. Quantitative real-time PCR on splenocytes deepened this picture. Using the 2^-ΔΔCt method on TRIzol-extracted RNA converted to cDNA, the team found markedly elevated expression of interferon-gamma, interleukin-2 and interleukin-12, together indicating a dominant Th1 pattern of the kind observed in SARS-CoV-2-specific T cells during infection and vaccination. Importantly, the particles also raised interleukin-4 and interleukin-13, showing that Th2 responses supporting antibody class switching were engaged as well, and a modest rise in interleukin-10 suggested a regulatory brake that could prevent excessive inflammation while preserving protective immunity. Th17 and transforming growth factor beta components rounded out a balanced cytokine profile. Statistical significance was assessed with two-way ANOVA and Tukey’s post-hoc test for the ELISA data and Mann-Whitney tests for the qRT-PCR data.

The question of neutralization, however, received a sober answer. Because the membrane protein has only a short N-terminal ectodomain and sits mostly embedded in the viral envelope, antibodies directed against it are not expected to block viral attachment or entry, and indeed the M-protein particles alone did not induce detectable neutralizing activity. This distinguishes them sharply from spike-targeted vaccines, which elicit the receptor-blocking antibodies that prevent infection. Yet the team argues that this is not a disqualifying limitation. The membrane protein has remained highly conserved across the many SARS-CoV-2 variants that have emerged since the pandemic began, including the Delta variant that originated in India, whereas spike has mutated relentlessly under immune pressure. A vaccine component built on the membrane protein could therefore provide durable immune memory and T-cell help that remains effective regardless of how the spike evolves, potentially serving as a conserved backbone in combination with variant-matched spike antigens. The researchers’ previous work on Membrane-Envelope particles had already shown antigenicity and neutralization activity, and co-expression of M and E proteins is known to enhance virion assembly and produce particles more faithful to the native virus, suggesting that head-to-head comparisons of M-only versus M-plus-E particles in larger animal cohorts are a natural next step.

The study also charts a path for refining the platform itself. Targeted mutations at glutamate 115 or glutamate 135 could probe how changes at the dimer interface affect particle formation, stability and immunogenicity, while tuning the lipid composition of the production system to favor lipids such as ceramide-1-phosphate, which is known to stabilize assembly-competent conformations of the membrane protein, might yield higher-quality particles. More definitive structural work, including immunogold labeling, cryo-electron microscopy and cryo-electron tomography, will be needed to visualize exactly how the membrane protein is oriented within the assembled particles. And before any translational claims can be made, neutralization assays, virus challenge studies and protection-efficacy experiments will be essential. For now, the finding stands as a striking demonstration of molecular self-assembly: one small, 25-kilodalton protein, acting alone in insect cells, can build a virus-like shell that the mouse immune system recognizes and attacks with vigor, opening a new avenue in the search for coronavirus vaccines built not on the shifting spike, but on the stable scaffold beneath it.

Asgard archaea have attracted significant scientific attention due to their unique position in the tree of life, nestled at the intersection between prokaryotes and eukaryotes. Unlike their bacterial and conventional archaeal cousins, Asgard members possess genes previously thought exclusive to eukaryotic cells. Yet, the precise extent and functionality of their complex cellular components remained elusive. This new study transcends mere genomic analysis, integrating sophisticated structural prediction algorithms to unravel the three-dimensional conformation of protein complexes that are foundational to eukaryotic cell biology.

The research team applied state-of-the-art deep learning models, including AlphaFold and RoseTTAFold, to predict protein structures from Asgard archaeal sequences with unprecedented accuracy. By simulating the spatial arrangements of these proteins, they reconstructed multiprotein assemblies key to cellular processes such as cytoskeleton formation, membrane trafficking, and intracellular signaling. Remarkably, several of these complexes show striking parallels to their eukaryotic counterparts, suggesting functional conservation and ancestral origins. This finding challenges the conventional binary classification of life into simple prokaryotes and complex eukaryotes, instead emphasizing a continuum of cellular sophistication.

One of the most astonishing revelations is the identification of Asgard-encoded homologs to eukaryotic cytoskeletal proteins, such as actin and tubulin analogues. The cytoskeleton is central to maintaining cell shape, enabling motility, and orchestrating intracellular transport in eukaryotes. Previously, such elaborate structures were considered absent in archaea. Through precise structural modeling, the study indicates that Asgard proteins could polymerize into filamentous networks similar to those in eukaryotic cells. These networks might underpin processes critical to cellular organization and division, hinting at a primordial cytoskeletal toolkit preceding the rise of true eukaryotes.

The research also delves into the membrane remodeling machinery of Asgard archaea, uncovering predicted structural homologs to eukaryotic ESCRT (Endosomal Sorting Complex Required for Transport) proteins. ESCRT complexes regulate membrane scission events vital to vesicle formation and trafficking, which are fundamental for intracellular compartmentalization. The presence of such proteins in Asgard archaea signals potential capabilities for primitive membrane dynamics, potentially foreshadowing the complex endomembrane systems characteristic of eukaryotic cells. This discovery underscores the possibility that key cellular innovations emerged incrementally within archaeal ancestors.

Furthermore, the study sheds light on the signaling networks within Asgard archaea, identifying structural motifs resembling those involved in eukaryotic signal transduction pathways. Signal transduction enables cells to respond dynamically to environmental cues, coordinating growth and adaptation. The predicted protein structures include domains that can mediate protein-protein interactions and phosphorylation events, fundamental to intracellular communication. This suggests that rudimentary signaling cascades might have operated in the archaeal lineage, providing a proto-framework upon which eukaryotic complexity could build.

The implications of these findings reach beyond the realm of evolutionary biology. Understanding the cellular complexity of Asgard archaea could inform synthetic biology efforts aimed at engineering minimalist versions of eukaryotic cells, advancing biotechnology and medicine. Additionally, revealing the molecular underpinnings of early eukaryogenesis aids in interpreting the evolutionary pressures and innovations that led to multicellular life, thereby enriching our comprehension of life’s history on Earth.

Significantly, the study emphasizes the utility of integrative structural modeling in bridging gaps left by traditional genomic and proteomic methods. Genomic data alone often cannot predict protein folding and complex assembly, especially for uncharacterized or divergent sequences. By employing computational tools that capture three-dimensional conformations, the researchers have unlocked functional predictions that traditional homology-based annotations miss. This methodological advance paves the way for future inquiries into other enigmatic microbial lineages.

Crucially, the work highlights the mosaic nature of cellular evolution. Rather than a sudden leap, the emergence of eukaryotic complexity likely involved the gradual accrual of modular components. Asgard archaea exemplify this intermediate stage, possessing a suite of proteins that were co-opted and elaborated upon during the evolution of eukaryotes. These insights align with the symbiogenesis theory, wherein a merger between archaeal hosts and bacterial endosymbionts catalyzed the origin of eukaryotic cells.

The authors acknowledge current limitations and avenues for further validation. Experimental structural studies, such as cryo-electron microscopy of Asgard proteins, will be indispensable to confirm the computational models. Moreover, culturing Asgard archaea remains a formidable challenge, constraining direct biochemical probing. Nonetheless, the predictive power demonstrated here sets a robust framework for future empirical investigation.

In conclusion, the study by Köstlbacher et al. represents a monumental step forward in decoding the molecular complexity of Asgard archaea and their evolutionary significance. By harnessing the power of structural prediction, it redefines our perspective on the prokaryote-eukaryote boundary, illuminating the ancient roots of cellular architecture. This work not only deepens our understanding of microbial diversity but also inspires a reevaluation of life’s grand tapestry, reminding us that complexity arises through countless incremental adaptations etched in molecular form.

As scientific exploration continues to push the envelope of what is known about life’s origin, the revelations from Asgard archaea underscore a captivating narrative: the story of how life’s complexity unfolded was encoded in the very folds of proteins long before true eukaryotic cells flourished. Studies like this promise to reveal more about our cellular heritage and spotlight the ingenious simplicity from which complexity emerges.

Researchers and enthusiasts alike anticipate that these findings will stimulate interdisciplinary collaborations, blending molecular biology, bioinformatics, evolutionary theory, and systems biology. The insights gleaned may also resonate with astrobiology, offering clues about possible evolutionary trajectories for life beyond Earth. As such, the ramifications of this research extend far beyond a single microbial lineage.

Ultimately, this pioneering study serves as a testament to the power of combining computational innovation with evolutionary inquiry. By unveiling a structural blueprint for eukaryotic precursors encoded in Asgard archaea, it propels the quest to unlock the mysteries of cellular evolution into an exhilarating new chapter. The evolutionary saga, long obscured in the depths of ancient microbes, has begun to reveal its secrets with unprecedented clarity.

Subject of Research: Prediction of eukaryotic cellular complexity in Asgard archaea using structural modelling.

Article Title: Prediction of eukaryotic cellular complexity in Asgard archaea using structural modelling.

Article References: Köstlbacher, S., van Hooff, J.J.E., Panagiotou, K. et al. Prediction of eukaryotic cellular complexity in Asgard archaea using structural modelling. Nat Microbiol 11, 747–758 (2026). https://doi.org/10.1038/s41564-026-02273-y

Image Credits: AI Generated

DOI: March 2026

A cornerstone of this transformation is the legacy of AlphaFold, a revolutionary AI model developed by Google DeepMind that accurately predicts the three-dimensional structures of proteins based on amino acid sequences. Enabled by extensive open data shared by EMBL-EBI and global collaborators, AlphaFold has catalyzed a paradigm shift in structural biology, ensuring that protein structure predictions are freely accessible to researchers worldwide. This accomplishment underscores EMBL’s critical role as a facilitator and innovator in the AI life sciences ecosystem.

Expanding beyond structural biology, EMBL is pioneering novel AI-driven methodologies that apply to diverse biological datasets. Leveraging machine learning for cellular imaging allows for enhanced resolution and throughput beyond traditional microscopy techniques, reducing reliance on manual image analysis and improving experimental consistency. Furthermore, the integration of heterogeneous biological datasets—such as genomics, proteomics, and metabolomics—is enabling a systems-level understanding of biological processes, facilitating biomarker discovery and disease characterization with unprecedented precision.

Central to EMBL’s AI vision is the transformational funding from the German Hector Foundation, which has committed long-term support earmarked for building dedicated AI research groups, advancing data engineering capabilities, and deploying state-of-the-art computational infrastructure. This philanthropic investment not only provides the resources necessary for sustained innovation but also supports fellowship programs designed to cultivate multidisciplinary expertise that bridges computational and biological sciences—ensuring a pipeline of talent equipped to tackle tomorrow’s scientific challenges.

Oliver Stegle, EMBL’s Acting Head of AI, emphasizes that the true power of AI is realized through collaborative, cross-disciplinary efforts spanning geographical and institutional boundaries. AI’s ability to rapidly process massive biological datasets — ranging from genomic sequences to clinical health records — enables hypothesis generation and experimental design at scales and speeds unattainable by traditional methods. However, meaningful breakthroughs emerge from synergistic partnerships that integrate domain expertise and computational innovation.

EMBL envisions the future of life sciences research as inherently interdisciplinary. Machine learning models deployed for decoding genomic complexity continue to evolve, harnessing long-read sequencing technologies to uncover structural variants and somatic mutations critical in cancer genomics. Concurrently, AI methods enrich proteomics by predicting protein structures and dynamic interactions, contributing to a nuanced understanding of cellular machinery and pathophysiology. These advances offer promising avenues for precision medicine and therapeutic development.

In cellular microscopy, AI-driven image analysis algorithms improve the resolution and quantitative interpretation of cellular and subcellular structures. Automating traditionally laborious processes reduces human bias and enhances reproducibility, facilitating large-scale experiments that chart developmental pathways or disease progression. This shift from manual curation to computational inference supports high-throughput phenotyping and accelerates biological discovery.

Drug discovery is undergoing a radical transformation through AI-powered molecular simulations. These methods integrate physics-based models with machine learning to predict molecular interactions and prioritize pharmacological targets efficiently. By significantly compressing research timelines and resource requirements, AI accelerates the path from molecular hypothesis to viable drug candidates, enhancing lead optimization and toxicity prediction with increasing accuracy.

The sheer volume and diversity of biological data necessitate sophisticated data management systems to ensure accessibility and interoperability. EMBL’s AI-guided platforms improve data annotation, curation, and synthesis, fostering open science and enabling researchers to navigate vast datasets effectively. This democratization of data resources facilitates a global research community working collaboratively and building on shared knowledge.

Anna Kreshuk, senior scientist at EMBL, reflects that artificial intelligence is not merely a tool but is fundamentally reshaping the scientific process. AI influences how research questions are formulated, strategies are devised, and experiments are integrated with computational models. This paradigm shift brings together theoretical insights and empirical evidence in a tighter dialogue, accelerating iterative cycles of hypothesis testing and validation.

To fully leverage AI’s transformative potential, EMBL is intensifying efforts to create a pan-European AI ecosystem through strategic partnerships with academic institutions, industry stakeholders, and policy makers. By assembling a critical mass of expertise, resources, and infrastructure, EMBL fosters an environment of rigorous, open, and collaborative science. Training initiatives ensure that emerging scientists develop the computational literacy and interdisciplinary skills required to lead in this evolving landscape.

Ethical considerations are integral to EMBL’s AI strategy, addressing privacy, reproducibility, and societal impact. Responsible AI deployment ensures that advances in computational biology contribute positively, maintaining transparency and trustworthiness in scientific outputs. EMBL’s leadership extends beyond technology, promoting frameworks that guide the ethical conduct of AI-driven research aligned with societal values.

The Hector Foundation’s visionary philanthropy catalyzes EMBL’s capacity for sustained leadership at the interface of AI and life sciences. This investment not only amplifies EMBL’s innovative research programs but also creates momentum for attracting additional funding and forging collaborative networks across Europe. Dr. h.c. Hans-Werner Hector emphasizes that AI represents a new scientific epoch, one in which computational ingenuity drives breakthroughs that benefit medicine, research, and society holistically.

Together, EMBL’s strategic vision, scientific excellence, and collaborative ethos establish a global benchmark for AI-integrated life science research. By empowering researchers with advanced computational tools, multidisciplinary expertise, and ethical rigor, EMBL accelerates the pace of discovery and fosters innovations that transcend disciplinary and geographic boundaries. The integration of AI into the fabric of biological research heralds an era of unprecedented insight into life’s fundamental mechanisms and transformative applications for human health.

Subject of Research: Artificial Intelligence Integration in Life Sciences Research at EMBL
Article Title: EMBL’s Visionary AI Strategy: Revolutionizing Life Sciences Through Advanced Computational Research
News Publication Date: Not explicitly provided
Web References:

• https://www.embl.org/topics/ai-at-embl/
• https://www.embl.org/news/science/alphafold-using-open-data-and-ai-to-discover-the-3d-protein-universe/
• https://www.embl.org/editorhub/wp-content/uploads/2025/02/EMBL_AI-Strategy_Feb2025_Accessible.pdf
• https://www.ebi.ac.uk/about/news/perspectives/leveraging-long-read-sequencing-for-cancer-genomics/
• https://www.embl.org/news/science/puzzling-out-the-structure-of-a-molecular-giant/
• https://www.embl.org/news/science/charting-a-multi-omic-universe/
• https://www.embl.org/news/science-technology/follow-the-cellular-road/
• https://www.embl.org/news/science/machine-learning-to-identify-and-prioritise-drug-targets/
• https://www.embl.org/news/science/ai-annotations-increase-patent-data-in-surechembl/
  Image Credits: Creative team/ EMBL
  Keywords: Life sciences