Viruses are tiny infectious agents that invade host cells and hijack machinery to replicate. Most of us have encountered viruses in our lifetimes as the common cold, the flu, or SARS-CoV-2, which causes COVID-19. These bugs are specialized to invade and reproduce in human cells. Similarly, bacteria cells can be invaded and hijacked by viruses called bacteriophage. Like respiratory viruses that target specific cells in the human respiratory tract, different types of phage target unique bacterial species. Once a bacterium is infected with lytic phage, it rapidly dies.
Bacteria have evolved defense systems like structural barriers and immunity systems to combat phage infections, but attempts to study how phage interact with these bacterial defense systems have major limitations. For one, isolating and culturing phage is a slow, labor-intensive process. Additionally, most defense systems are studied by overexpression in model organisms rather than their native hosts. Constant high levels of defense system genes do not reflect natural expression patterns, limiting the real-world translatability of these studies.
Bacteroides fragilis is a pathobiont found in the gut bacteria. This means that normal guts have B. fragilis, but under certain conditions, the bacteria can cause disease. B. fragilis has been associated with colorectal cancer, abscesses, undernutrition, and inflammatory bowel disease. Previous research has shown that Bacteroides species use structural barriers to defend against phage infection. These bacteria alter the expression of capsular polysaccharide (CPS) genes to produce multiple CPS structures that are expressed on the cell surface to block phage entry. However, not every individual bacterium responds to phage exposure in the same way. To explore how different responses to phage shape how susceptible B. fragilis is to phage, Dr. Neel Dey in the Translational Sciences and Therapeutics Division at Fred Hutch teamed up with Dr. Anna Kuchina at the Institute of Systems Biology. Together, they used a bacterial single-cell RNA sequencing technology to capture transcriptional changes over the course of a phage infection cycle in B. fragilis.
The first challenge to this project was actually finding a phage specific to B. fragilis. To do this, the researchers turned to a phage-rich source: King County wastewater. They added the wastewater to a culture of B. fragilis, incubated the culture overnight, filtered out the bacteria, and then plated the filtrate on an agar plate of B. fragilis. Phage isolates were picked from zones where bacteria were cleared. After some purification and sequencing steps, the researchers were left with Bacteroides phage Bf12P1.
Next, they treated a culture of B. fragilis with Bf12P1 and collected the cells just before phage-induced death could occur. RNA sequencing showed that the phage-treated bacteria separated into two clusters. One cluster contained phage transcripts (the “infected cluster”), and the other contained no phage transcripts (the “uninfected cluster”). To understand transcriptional changes in B. fragilis upon infection, they filtered out the phage transcripts. They found that infected bacteria had more RNA polymerase, replication proteins, and ribonucleoside reductase transcripts compared to uninfected cells. All of these changes are expected and reflect a natural host response to viral infections, but the expression data is from bacteria at all stages of the infection cycle, limiting the resolution of this data.
To gain insights into how the infection timepoint impacted transcriptional changes, the team constructed a pseudotime trajectory for their samples. Pseudotime assigns each cell a value between 0 and 1 based on how similar the transcripts in a particular bacterium are to those in another bacterium. In this case, a pseudotime value of 0 would represent the earliest stage of infection, and a pseudotime value of 1 would represent the latest stage of infection. They found that the frequency of transcripts from B. fragilis decreased at higher pseudotime values as phage transcripts increased. In early pseudotime, B. fragilis expressed genes encoding transcription and translation machinery. In mid pseudotime, gene expression shifted towards dNTP synthesis, likely to promote phage replication. In late pseudotime, bacteria expressed genes indicating the cells were damaged and likely to die. Together, this approach indicates that pseudotime can effectively separate bacteria at different stages of infection and provide valuable insights into the genetic changes that happen over the course of infection.
Dey and Kuchina also wanted to understand how CPS genes regulated susceptibility to phage. CPS subtypes are encoded by 8 operons termed PSA-H that form distinct structures to help bacteria resist phage. CPS genes were differentially expressed between infected and uninfected bacteria. Uninfected B. fragilis had higher expression of PSB, PSF, and PSG, suggesting that these CPS types may protect phage from infection. They calculated phage infection chances based on CPS expression and found that PSB and PSG-expressing bacteria were least likely to succumb to infection, while PSA, PSC, PSE, and PSH-expressing bacteria were most likely to be infected. They tested this prediction in the lab by infecting B. fragilis that only expressed either PSA, PSC, PSG, or PSH with their phage. In line with their prediction, PSG-only bacteria were resistant to phage killing, while PSA, PSC, and PSH-only bacteria were quickly eliminated by the phage. Together, these results show that CPS subtype underlies B. fragilis susceptibility to phage.
Overall, Dey and Kuchina demonstrated that bacterial single-cell RNA sequencing can be used to understand phage-bacteria interactions, model the cycle of phage infections, and dissect differences in phage response within a single species of bacteria. This work paves the way for designing phage therapies that could treat bacterial infections and prevent bacterial resistance.
This work was supported by funding from the Department of Energy Office of Science, the Biological and Environmental Research Program, the National Institutes of Health, the Fred Hutch Electron Microscopy Shared Resource, and the Fred Hutch/University of Washington/Seattle Children’s Hospital Cancer Consortium.
Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium Members Drs. Georg Seelig and Neelendu Dey contributed to this work.
Gupta A, Morella N, Sutormin D, Li N, Gaisser K, Robertson A, Ispolatov Y, Seelig G, Dey N, Kuchina A. 2026. Dynamics of phage-host interactions in Bacteroides fragilis resolved by single-cell transcriptomics. Nat Commun. 2026 Mar 16;17(1):4035. doi: 10.1038/s41467-026-70381-8.
Kelsey Woodruff is a PhD candidate in the Termini Lab at Fred Hutch Cancer Center. She studies how acute myeloid leukemia cells remodel the sugars on their membranes to reprogram cancer cell signaling. Originally from Indiana, she holds a bachelor's degree in Biochemistry from Ball State University. Outside of lab, you can find her crocheting and enjoying the Seattle summers.