Paul E. Turner, PhD, is the Rachel Carson professor of ecology and evolutionary biology at Yale University

Paul E. Turner, PhD

As antimicrobial resistance continues to outpace drug development, hospitals are increasingly forced to consider alternatives that once lived on the fringes of clinical practice. Among them, bacteriophages have re-emerged as a serious area of scientific and clinical interest, not as naturally occurring viruses alone, but as engineered tools for the precision targeting of problematic pathogens.

The key question is no longer whether phage therapy is scientifically interesting, but where it might realistically fit into hospital practice. Should engineered phages be considered first-line treatment options for multidrug-resistant (MDR) infections, such as 

? Do they have a role in decolonization strategies? Or could their greatest impact come from environmental or systems-level applications?

, senior principal investigator at New England Biolabs, and 

, the Rachel Carson professor of ecology and evolutionary biology at Yale University, researchers working at the intersection of phage biology and engineering to explore what is possible now and what remains aspirational. The discussion examines near-term use cases for hospitals, the limitations imposed by narrow host range, the continued need for phage cocktails, and how engineered phage libraries could change the current model of patient-by-patient phage hunting.

For infection prevention and control (IPC) personnel navigating rising resistance, vulnerable patient populations, and increasing pressure to innovate responsibly, this conversation offers a grounded look at how engineered phages may evolve from experimental tools into clinically relevant options, and why careful stewardship, surveillance, and regulatory planning will be essential as the field moves forward.

: From an IPC standpoint, where do you see engineered phages fitting first: treatment of difficult infections (eg, MDR 

), decolonization, or environmental bioburden control?

 I think the first area engineering will impact is the biological study of phages and their interactions with hosts. This is still an active area of research, even though phages were discovered over 100 years ago. The power of our system is to make it possible to do high-throughput data generation, which is currently straightforward to do with proteins, but much more difficult to accomplish with phages. 

A systematic study of genotype/phenotype relationships within host recognition factors is key to uncovering proteins that improve phage fitness. In combination with the sorts of machine learning that is accelerating protein engineering, I am optimistic the field will be able to rapidly learn how to push phages in the direction of whatever application a given lab is exploring, but I can not predict the specific applications that will make it to the clinic first.

Decolonization is a difficult and perhaps unrealistic target for some individuals in the clinic, especially in patients that are immunocompromised or otherwise vulnerable to polymicrobial infections. Here, the goal is to use phage therapy to reduce the population of infecting bacteria, potentially allowing the human immune system (if active) to further reduce the infection. A great example is the lungs of people with cystic fibrosis, which are often colonized by multiple pathogens. When a patient experiences an exacerbation, our goal has often been to reduce the population of infecting bacteria, without expecting to clear the infection. Thus, engineered phages could improve disease management when decolonization is difficult or impossible.

What’s the most realistic near-term use case in hospitals, and why?

 A continued concern is that modern phage therapy tends to focus on environmentally sourced phage candidates, but this discovery process is very uneven among bacterial pathogens, where antibiotic resistance is on the rise. For target bacteria where ‘phage hunting’ fails to isolate genetically diverse virus strains from nature, perhaps libraries of engineered phages can jumpstart these lagging efforts, enabling experimental compassionate treatment of patients and ultimately clinical trials, focusing on engineered phages.

 My hope is that having libraries of engineered phages will supplement, or even supplant, screening of environmental phages to find strains that work on a particular patient infection

: You highlight host-range limits and the need for “phage cocktails.” What did your tail fiber gene swapping teach you about reliably expanding host range in

 Through our study, we demonstrated that we can alter the host range of a bacteriophage by altering tail fiber genes. This is an important proof-of-principle, but we still cannot predict the outcome of a given sequence change. To inform more strategic host-range engineering, the field now needs a way to tie genotype to host range. This can now be accomplished with a high-throughput study that resolves how nuclease-level differences in tail-fiber genes translate into phenotypic differences in receptor specificity. Our method makes such a study feasible.

Even once we can rationally design broader host ranges, phage cocktails are likely to remain important. Engineered phages are still far more specific than antibiotics, and combining phages that use different receptors has the potential to broaden coverage and reduce the likelihood of resistance emerging.

: For IPC readers: how close are we to a “broad-coverage” phage product that doesn’t require patient-by-patient phage hunting?

 We are not close to a “broad-coverage” phage. One of the key advantages of phage therapy is its host specificity, which allows targeting of a specific pathogen while leaving the microbiome alone. On the flip side, host-specificity is also a major challenge. Even closely related strains of 1 pathogenic species will respond entirely differently to the same phage. We still have a lot to learn about these phage-host interactions, and we are a long way from being able to have specific control of the host range of an engineered phage.

Looking forward, I think there are a couple of main avenues for development – one is learning enough about host specificity determinants to target common surface antigens, which would increase phage host range. However, this approach may lead to off-target effects or reduce efficacy. Another strategy would be to engineer a library of phages that contains enough diversity to cover most pathogens of a given genus. This could replace the need to hunt for infection-specific phages in environmental samples, speeding up the current phage-selection process. In any scenario, the goal will be to find a balance between broadening host range, maintaining efficacy, minimizing off-target effects, and improving scalability.

: Resistance is the elephant in the room. In your view, how should hospitals think about phage resistance compared with antibiotic resistance: surveillance, stewardship, and operational response?

 The evolution of phage resistance is likely to be an ongoing challenge for phage therapy development. Fortunately, our research and that of others show that certain phages can ‘steer’ their target bacteria to evolve phage resistance that compromises clinically relevant virulence and drug-resistance traits. This should be especially true if therapeutic phages bind to cell-surface structures responsible for pathogenicity, because bacteria tend to evolve phage resistance by modifying (or even deleting) these mechanisms. It may be optimal to engineer phage candidates to produce libraries of virus strains that more readily cause these pathogenicity trade-offs when target bacteria evolve phage-resistance

: Your platform enables adding payloads (like fluorescent reporters) and making precise edits with high success. What kinds of engineered functions matter most for infection control?

 There are 2 main regions that are likely to be good engineering targets. One is the tail fiber and baseplate proteins that control host recognition and entry. The second is the “antidefense” region of phage genomes, which encodes proteins that suppress and evade host bacterial immunity. These are functions that phages already perform, which we can harness and modify, rather than developing entirely new ones. We have databases containing millions of phage genome sequences with substantial diversity, and the task now is to design experiments to analyze and learn from that diversity.

: Implementation question: What are the biggest barriers between this engineering approach and real-world clinical deployment in high-risk settings, such as intensive care or burn units?

 As news spreads of a ‘looming’ antibiotic resistance crisis that threatens global public health, we should be reminded that immunocompromised and other vulnerable people are already on the frontlines of the battle, struggling to overcome life-threatening bacterial infections that standard antibiotics fail to control. This immediately suggests that further basic research should address the safety and efficacy of engineered phages relative to naturally occurring viruses, to ensure these options are equally safe and effective to deploy in high-risk settings.

 Similar to natural phages, synthetically engineered phages will need to demonstrate efficacy and safety in clinical settings. They will also need to achieve regulatory approval, which may prove challenging. There isn’t a pathway for therapies like this. Developing one will require careful study and cooperation between biologists, clinicians, and regulators.

Articles

Could engineered bacteriophages help hospitals tackle drug-resistant infections when antibiotics fail? In this ICT Q&A, researchers discuss where phage engineering may realistically fit first in hospitals, from treatment of MDR infections to environmental control, and why broad-coverage phage products remain a longer-term goal.

Paul E. Turner, PhD

Articles

Engineered Phages in Infection Prevention: What Is Realistic for Hospitals Now