How many bacterial cells are infected by viruses in the oceans?

Microorganisms, including bacteria, are essential for life in the oceans, playing a crucial role in all major elemental and energy fluxes. Growing on dissolved substrates, their community is shaped by grazing and by viruses that infect them. We know since a long time that viruses outnumber their host cells by about 10-fold in surface seawater. However, the precise number of bacterial cells infected by viruses at any given time remains an intriguing question. In our quest to answer this for the most abundant oceanic bacteria, we discovered a significant proportion of virus-infected cells devoid of ribosomes, which we termed ‘zombie cells’.

Studying bacterial cell division rates

In a previous study, we looked at the cell division rates of abundant groups of bacteria in the ocean and found that the so-called SAR11 bacteria divide most rapidly, when the cell numbers decrease¹. This is counterintuitive and does not make sense at first glance.

But let’s start from the beginning: Phytoplankton spring blooms are annual events in temperate oceans, where increasing light availability fuels the exponential growth of phytoplankton (algae). They release large amounts of organic matter into the water column, providing substrates for various groups of microorganisms. After ~20 years of research, we observe the same microbial groups benefitting from the phytoplankton bloom and growing to high cell numbers.

To gain a better understanding of the role of individual microbial groups, we employ fluorescence in situ hybridization (FISH) and high-throughput microscopy. FISH involves the specific staining of individual microbial groups with a fluorescent dye, allowing us to identify and count them. We also fluorescently stain the genome of the microorganisms. Then, we used a little trick: when a microbial cell duplicates, it first replicates its genome, before it is being divided into the future daughter cells. We can see this in the microscopy images (Fig. 1) and, combined with another trick, we are able to calculate cell division rates for specific microbial groups in the environmental samples.

Using this approach, we found that most bacterial groups behaved as expected: Their cell division rates increased, which we could subsequently observe in increasing numbers. However, SAR11 bacteria showed a different picture: While growing faster and faster their cell counts were dwindling (Fig. 2). This puzzled us and raised the question: What specifically kills the SAR11 bacteria? We hypothesized that viruses that infect and kill SAR11 might cause this discrepancy between fast cell division and decreasing cell numbers. 

 Visualizing virus-infected bacterial cells

While we know viruses play a crucial role in controlling microbial communities, we do not know how many microbial cells in the oceans are infected by viruses at any given time (with one exception²). In our recent study, we used direct-geneFISH to fluorescently label individual genes to specifically visualize virus-encoded genes. Combined with the FISH method described above, we were able to identify virus-infected host cells. After establishing this method in the laboratory, we screened our spring bloom samples and found the solution to the conundrum mentioned earlier: the highest number of virus-infected cells occurred when SAR11 cells divided the fastest and their cell abundance decreased (Fig. 3) ³.

We were quite excited, as we contributed the first direct observation of a virus infection on the SAR11 community in the ocean. However, we made another observation, which led us to discover zombie cells.

The discovery of zombie cells: virus-infected but ribosomes-deprived cells

When I established the virus-staining method with pure cultures in the laboratory, I noticed that a substantial proportion of cells were virus-infected but the host cells could not be stained with FISH (Fig. 4). At this point, we should take a closer look at the method to identify the host cell. FISH is based on short DNA probes, which bind to the ‘genetic fingerprint’ of the microbial cells. It thereby allows to stain individual bacterial groups. The ‘genetic fingerprint’ is the 16S ribosomal RNA. Ribosomes and its RNA is highly conserved across different bacterial groups, as they are essential for protein synthesis in any cell.

The cells that were virus-infected but not identifiable with FISH obviously lacked ribosomes and, therefore, could not produce proteins. We termed these zombie cells because they exist in a state between life and death, likely producing more viruses that will kill additional cells.

We observed zombie cells (virus-infected but ribosome-deprived) not only during the phytoplankton bloom described above, but also in samples across the Atlantic, Southern, and Pacific Ocean. They are a globally occurring phenomenon, contributing to, on average, 4% of all microbial cells in the ocean.

How do zombie cells develop? Using genetic information and other hints, we hypothesize that the viruses use the nucleotides of the hosts ribosomal and other RNA to produce more virus genomes.

In summary, we study the microbial community using microscopy techniques. In a first part, we observed the conundrum of rapidly dividing SAR11 cells (the most abundant bacterium in the ocean), while their cell counts decreased. As a consequence, we studied virus-infected SAR11 cells in a second part and found the missing link: Virus infection was greatest, when cell division was high and cell abundances decreased. We additionally discovered zombie cells, which are virus-infected but ribosome deprived cells. We could show that virus-infected cells with ribosomes and zombie cells occurred globally. We speculate that zombie cells are not limited to SAR11. This and the ecological implications require further research in the future.

¹Brüwer, Jan D., et al. "In situ cell division and mortality rates of SAR11, SAR86, Bacteroidetes, and Aurantivirga during phytoplankton blooms reveal differences in population controls." Msystems 8.3 (2023): e01287-22

²Mruwat, Noor, et al. "A single-cell polony method reveals low levels of infected Prochlorococcus in oligotrophic waters despite high cyanophage abundances." The ISME Journal15.1 (2021): 41-54.

³ Brüwer, Jan D., et al. "Globally occurring pelagiphage infections create ribosome-deprived cells." Nature Communications 15.1 (2024): 3715.