From Molecules to Modalities: Engineering the Future of Therapeutics
By Professor Andreas Plückthun, University of Zurich
Over the past several decades, the life sciences have witnessed repeated waves of innovation—each promising to transform medicine. Genomics, proteomics, combinatorial chemistry, biologics, gene therapy, artificial intelligence. Some delivered quickly, others more slowly. Yet what has truly changed the therapeutic landscape is not a single breakthrough, but the steady expansion of what is possible.
At the center of this expansion lies protein engineering.
Today, protein engineering is no longer confined to a single class of therapeutics or technologies. It has become a foundational discipline that underpins nearly every modern modality—from antibodies and engineered binding proteins to vaccines, diagnostics, and gene delivery systems. This journey, from molecules to modalities, is not about replacing one approach with another. It is about broadening the therapeutic portfolio and learning how to deploy the right tools for the right biological problems.
Protein Engineering Is Everywhere
Protein engineering now permeates the entire drug discovery and development ecosystem. Engineered proteins are not only the therapeutic agents delivered to patients; they also enable discovery itself. In many cases, receptors used to identify or validate small-molecule drugs must first be engineered. Vaccines are frequently based on stabilized or redesigned proteins. Virus-like particles used for DNA and RNA delivery rely on engineered protein components that control cell recognition and uptake.
Even small-molecule drug discovery—often viewed as a separate discipline—depends heavily on protein engineering. Without suitable binding pockets or engineered targets, many small molecules simply cannot be discovered or optimized.
Rather than signaling a shift away from traditional approaches, this reflects a deeper reality: protein engineering has become a universal enabler across modalities.
Expansion, Not Replacement
Much has been made of the movement from small molecules toward biologics and engineered proteins. While this trend is real, it is often mischaracterized.
There is no reason to believe that small molecules will disappear. They remain uniquely effective at penetrating cells and inhibiting intracellular targets—provided those targets present suitable binding pockets. What has changed is that we now have additional modalities capable of addressing targets that small molecules cannot, particularly cell-surface receptors involved in cancer and immune regulation.
Antibodies filled part of this gap. Engineered protein scaffolds have extended it further.
These scaffolds make it possible to assemble highly complex, multifunctional therapeutic systems—sometimes combining four or five distinct activities within a single molecule. Such architectures enable mechanisms of action that were simply not achievable before, while also allowing more flexible coupling to small molecules or other functional components.
The result is not a narrowing of options, but a dramatic expansion of therapeutic possibilities.
Engineered Binding Proteins: Strengths and Limitations
Engineered binding proteins offer clear advantages over traditional antibodies in many contexts. Their robustness, predictable expression, and favorable manufacturability allow for rapid development and scale-up. This reliability makes them particularly well suited for applications requiring complex molecular assemblies or unconventional designs.
These properties have enabled therapeutic strategies that would have been extraordinarily difficult—if not impossible—to realize with antibodies alone.
Yet important challenges remain. Chief among them is immunogenicity. While empirical solutions exist for specific cases, there is still no universal, predictive method to eliminate immune responses against engineered proteins. This is not a solved problem, and it remains one of the key barriers to broader application.
Diagnostics: An Underappreciated Opportunity
One of the most promising—yet often overlooked—applications of engineered proteins lies in diagnostics.
Traditional monoclonal antibodies used for detection have a long and uneven track record. Reproducibility issues, variability between batches, and high costs continue to limit their reliability and scalability.
New protein engineering technologies now make it possible to generate detection reagents rapidly, cost-effectively, and with far greater consistency. If widely adopted, these approaches could significantly improve diagnostic quality across research, clinical testing, and public health—an impact that may rival that of many therapeutic advances.
The Real Bottleneck: Biology, Not Engineering
Despite remarkable progress in molecular engineering, the greatest obstacle in translating innovation from the laboratory to the clinic remains our incomplete understanding of biology.
Biological systems are highly redundant. Inhibiting one pathway may produce no clinical benefit if parallel pathways compensate. Animal models, while invaluable, are not identical to humans. Subtle differences in wiring can lead to unexpected failures—or unexpected toxicities—when therapies enter clinical trials.
These challenges are not unique to any modality. They affect small molecules, biologics, gene therapies, and cell therapies alike.
The lesson is humility. We must acknowledge that we understand only a fraction of the system we are attempting to manipulate.
Designing with Reality in Mind
Given these complexities, considerations such as developability, manufacturability, and delivery must be integrated from the very beginning of therapeutic design.
One effective strategy is to construct molecular libraries where every member already meets baseline criteria for stability and manufacturability. Another is to use computational tools—including AI—as filters to identify liabilities early in discovery.
Artificial intelligence has a role to play here, but it should be viewed as an addition rather than a replacement. Like genomics, proteomics, or directed evolution before it, AI expands what we can do—but it does not eliminate the need for human judgment, experimentation, and critical thinking.
The Future of Therapeutic Modalities
Cell and gene therapies have recently faced economic and investment challenges, but their scientific potential remains immense. While early successes focused on rare genetic diseases, these modalities are increasingly being explored for complex, large-scale indications such as cancer and metabolic disorders.
Progress will not be fast, and it will not be easy. But the potential to go beyond what traditional drugs can achieve ensures that cell and gene therapies will remain an essential part of the future therapeutic landscape.
Again, this is not about replacement—it is about addition.
The Role of Academia: Thinking Far Ahead
Academic research plays a critical role in shaping future modalities precisely because it allows for risk-taking. Academia is where ideas can be explored long before their clinical or commercial relevance is clear.
Many of the technologies underlying modern smartphones were developed in academic laboratories decades before they became viable products. The same principle applies to biotechnology. Academia must remain a place where researchers think far ahead, identify limitations, and open new conceptual spaces—while industry focuses on integration, scaling, and execution.
Advice for the Next Generation of Leaders
For scientific and biotech leaders looking ahead, two capabilities stand out as essential.
First, a willingness to engage deeply with the details. Reading original literature—sometimes tedious, often demanding—cannot be replaced by summaries or automated tools.
Second, intellectual openness. Attending sessions outside one’s immediate expertise and engaging with unfamiliar ideas is often where the most transformative insights emerge.
The future of therapeutics will not be built by any single technology or discipline. It will be engineered—carefully, collaboratively, and iteratively—across molecules, modalities, and minds.
