Immuto Scientific, a company co-founded in 2018 by CEO Faraz A. Choudhury, Ph.D., and CTO Daniel Benjamin, Ph.D., both electrical engineers by training, is revolutionizing the field of protein structure determination by integrating artificial intelligence (AI) with advanced mass spectrometry. The California-based firm is making a strategic pitch to address significant bottlenecks in traditional structural biology methods, particularly for workflows involving antibody-antigen analysis, promising a substantial acceleration in throughput and accessibility. While cryo-electron microscopy (cryo-EM) has delivered groundbreaking insights into near-atomic and atomic-resolution protein structures, its inherent challenges in sample preparation and grid optimization often extend timelines from weeks to months, creating a critical impediment in the rapid pace required for modern drug discovery. Immuto’s platform offers a powerful alternative, capable of generating data on approximately 1,000 samples per week, translating to an estimated 100 protein structures weekly, a remarkable leap forward in efficiency.
The Enduring Challenges of Traditional Structural Biology
For decades, determining the precise three-dimensional structure of proteins has been a cornerstone of drug discovery and fundamental biological research. Understanding a protein’s architecture is crucial for designing drugs that can interact with it effectively, either by inhibiting its function or enhancing it. Historically, X-ray crystallography dominated the field, providing atomic-resolution structures but requiring the often arduous task of crystallizing proteins. More recently, cryo-electron microscopy (cryo-EM) emerged as a powerful technique, earning its pioneers the Nobel Prize in Chemistry in 2017. Cryo-EM allows researchers to visualize biomolecules in their near-native states, often achieving resolutions comparable to X-ray crystallography for well-behaved samples.
Despite its transformative impact, cryo-EM presents several well-documented hurdles. The most prominent among these are the complex and iterative processes of sample preparation and grid optimization. Preparing a sample for cryo-EM involves purifying the protein to high homogeneity, achieving optimal concentration, and then vitrifying it rapidly in a thin layer of amorphous ice on a specialized grid. Any impurities, aggregation, or suboptimal ice conditions can severely compromise data quality and resolution. This optimization phase often requires multiple rounds of trial and error, consuming significant time, resources, and highly specialized expertise. Dr. Benjamin notes that even in optimized cryo-EM pipelines, progressing from a raw sample to a fully resolved structure can take weeks, and for challenging targets, this timeline can stretch into months, as evidenced by recent publications detailing protracted optimization periods.
Furthermore, cryo-EM instrumentation represents a substantial capital investment, with high-end microscopes costing several million dollars, alongside significant operational and maintenance expenses. The technique also requires highly trained personnel to operate the equipment, process data, and interpret results. While cryo-EM excels at providing a detailed, atomic-resolution "snapshot" of a protein’s structure, it often struggles to capture the dynamic flexibility and conformational changes that are characteristic of many proteins in their physiological environments, particularly within living cells. This limitation is particularly relevant for proteins that exhibit significant disorder or undergo large structural rearrangements during their function or upon binding to other molecules.
Immuto’s AI-Driven Mass Spectrometry: A Paradigm Shift
Immuto Scientific’s approach directly addresses these challenges by leveraging the inherent strengths of mass spectrometry (MS) and augmenting it with advanced artificial intelligence. Mass spectrometry, a technique that measures the mass-to-charge ratio of ions, has long been used for identifying proteins and characterizing their modifications. However, Immuto has developed a proprietary platform that extracts structural information at a residue level, offering a unique blend of speed, versatility, and insights into protein dynamics.
"We can get data on about 1,000 samples per week with our platform, so that roughly translates to something like 100 structures per week," stated Daniel Benjamin, Ph.D., Immuto Scientific’s co-founder and CTO. This throughput represents a radical acceleration compared to traditional methods. Unlike cryo-EM, which provides a comprehensive three-dimensional structure, Immuto’s method provides residue-level information, which, when coupled with AI, can be assembled into highly accurate structural models. The "barrier to entry is quite a bit lower running a mass spectrometer," Benjamin added, implying reduced costs and a broader accessibility for researchers and drug developers.
A key differentiator for Immuto’s technology is its ability to probe proteins in living cells and track structural changes dynamically. This capability is crucial for understanding the functional biology of proteins, as their structures are not static but constantly evolving in response to their environment. This real-time, in vivo structural monitoring stands in stark contrast to the static snapshots typically obtained from cryo-EM, offering a more complete picture of protein behavior, including flexibility and disorder—aspects that cryo-EM often finds challenging to resolve clearly.
Bridging the Gap: AI Prediction Meets Experimental Validation
The advent of powerful AI models like AlphaFold from DeepMind, as well as Boltz, Chai, and ByteDance’s Protenix, has revolutionized de novo protein structure prediction. These tools can generate highly accurate models of individual protein structures, and increasingly, protein complexes. However, a significant challenge remains: when these AI models predict multiple plausible structures, especially for complex interactions like antibody-antigen binding, ranking the truly correct structure among many possibilities can be difficult.
"If you were to output, let’s say, 1,000 different possible structures, the correct structure will be in there, but it won’t necessarily be the top-ranked structure," Benjamin explained. This "ranking problem" is where Immuto’s experimental data becomes invaluable. The residue-level information derived from their AI-assisted mass spectrometry acts as empirical constraints, providing real-world validation points that guide the AI models to select and refine the most accurate structure. Benjamin asserts that this synergy produces results "almost dead on with what you would see with Cryo-EM," effectively leveraging the speed and predictive power of AI with the undeniable truth of experimental data.
This integration is particularly impactful for antibody-antigen structure determination, an area critical for therapeutic antibody development. Understanding precisely how an antibody binds to its target antigen, down to the specific epitope, is fundamental for ensuring specificity, potency, and avoiding off-target effects. Current AI models can suggest binding modes, but without experimental validation, their clinical utility can be limited. Immuto’s platform offers a rapid, high-throughput solution to validate and refine these predictions, accelerating the development cycle for novel biologics.
Strategic Growth: Internal Pipeline and Key Partnerships
Immuto Scientific is not merely a technology platform provider; the company is strategically maturing its platform into a robust drug discovery pipeline. Dr. Benjamin highlighted the company’s internal focus on oncology, a therapeutic area with immense unmet need and significant investment in antibody-based therapies. Immuto’s lead oncology program is "gearing up to enter the clinic in 2027," signaling concrete progress towards bringing novel therapeutics to patients.
Further underscoring its impact and credibility, Immuto announced a significant partnership last year with Daiichi Sankyo, a major global pharmaceutical company. This collaboration focuses on a solid-tumor program, encompassing novel target discovery and antibody development. Such partnerships with established pharmaceutical giants are crucial for validating emerging technologies and facilitating their adoption across the industry. The focus on novel target discovery within solid tumors highlights the potential of Immuto’s platform to uncover new biological insights and therapeutic avenues in challenging disease areas.
Immuto’s approach to antibody discovery itself is distinct. Instead of prioritizing binding strength (affinity) from the outset, the company adopts a structure-based approach that first focuses on the binding site. "We intentionally start with medium- or low-affinity binders just to ensure they are binding to the exact right epitope," Benjamin noted. "Once we know it binds to the right site, we can engineer all the binding affinity we need." This epitope-first strategy is designed to mitigate risks associated with off-target binding and ensure that subsequent affinity maturation efforts are directed towards the most therapeutically relevant interactions.
Versatility Across Biological Systems
The adaptability of Immuto’s platform across various biological systems further broadens its potential applications. The company began its exploration with standard human cell lines, a common starting point in biomedical research. However, it has progressively expanded its capabilities to more complex and physiologically relevant models, including single-cell suspensions, 2D and 3D cultures, tumors, and even organoids. "We’ve even used our technology to look at tissue resections," Benjamin revealed. This progression reflects a strategic move towards capturing the intricate complexity of native biology. For target discovery, especially in oncology, patient-derived models that accurately reflect native biology and tumor heterogeneity are paramount, as they offer a more predictive environment for drug efficacy and safety than immortalized cell lines. This versatility positions Immuto’s technology as a powerful tool for a wide array of research and development applications, from basic biological discovery to translational medicine.
Looking Ahead: The PEGS Conference and Beyond
The scientific community eagerly anticipates the first public performance data for Immuto’s v1 antibody-antigen model, which Dr. Benjamin plans to present at the upcoming PEGS conference. This model, trained and validated on approximately 30 to 40 structures, is specifically optimized for antibody-antigen structure determination – a task where current AI tools still face significant challenges in ranking and selection. The presentation at PEGS, a prominent conference for protein engineering and antibody development, will provide critical insights into the real-world accuracy and utility of Immuto’s platform.
The implications of Immuto Scientific’s advancements extend far beyond increased throughput. By making high-quality structural information more accessible and faster to obtain, the company has the potential to:
- Accelerate Drug Discovery: Dramatically reduce the time and cost associated with identifying and validating drug targets, optimizing lead compounds, and developing biologics.
- Democratize Structural Biology: Lower the barriers to entry for structural biology studies, enabling more researchers and smaller labs to gain critical insights into protein function.
- Enable Novel Therapeutic Modalities: Facilitate the development of drugs targeting flexible, dynamic, or membrane proteins that are difficult to characterize using traditional methods.
- Enhance Precision Medicine: Support the discovery of patient-specific biomarkers and therapeutic targets by enabling rapid analysis of proteins from patient-derived samples.
- Bridge AI and Experimentation: Solidify a crucial workflow where powerful AI predictions are rigorously validated and refined by rapid experimental data, ensuring higher confidence in in silico models.
In an era where the speed of innovation dictates success in drug discovery, Immuto Scientific stands at the forefront of a new wave of structural biology. By harmonizing the predictive power of artificial intelligence with the empirical insights of mass spectrometry, the company is poised to redefine how protein structures are determined, ultimately accelerating the journey from scientific discovery to life-saving therapies.
Leave a Reply