Telling friend from foe in the bacterial immune system

From the Stoddard lab, Basic Sciences Division

If you aren't destroying your enemies, it's because you have been conquered and assimilated, you do not even have an idea of who your enemies are. You have been brainwashed into believing you are your own enemy, and you are set against yourself. The enemy is laughing at you as you tear yourself to pieces. That is the most effective warfare an enemy can launch on his foes: confounding them. -Bangambiki Habyarimana


The cell is a fortress, built to defend its residents from outside threats. Its first line of defense: the plasma membrane, a sturdy wall that encircles the cell, allowing through friendly visitors and repelling those with ill intentions. Among bacteria, the enemy at the gates is often bacteriophage, infectious viruses that aim to breach the walls and enslave the cell’s residents to do their own nefarious bidding. As the phage approach, the fortress wall looms large and intimidating, but the viruses have learned, over billions of years of evolutionary experience, what to expect, and have strategized accordingly. Each virus is equipped with a built-in battering ram, allowing it to punch a hole in the wall just large enough for its DNA to slip through. Once the phage DNA is inside, the conflict turns to hand-to-hand combat. A legion of cellular defenders set out to find and destroy the invaders before they can do the community harm. Within this defense force are restriction enzymes - nucleases that can slice the invading DNA to pieces. But finding their targets is no easy task, for the cell contains its own resident DNA molecules, which the restriction enzymes are tasked to protect but which are nearly indistinguishable from their foreign counterparts. How then, do these enzymes know which DNA to defend and which to attack? In a recent paper in the journal Structure, the group of Dr. Barry Stoddard, a Professor in the Basic Sciences Division at Fred Hutch, examined the process by which restriction enzymes distinguish friend from foe, and how they make the fateful decision whether to execute a suspected intruder.

There is a simple method by which a bacterial cell can distinguish its own DNA (termed host DNA) from an invading virus’s DNA (termed foreign DNA). The bacterium methylates adenines in its own DNA, providing an identifying mark, which the foreign DNA lacks, that a restriction enzyme can check to determine the DNA’s identity. But this system has a flaw – newly replicated host DNA is born without protective methylation marks and, until it receives them, it risks being mistaken as an invader. How, then, does the cell distinguish nascent host DNA from foreign DNA?  To answer this question, Dr. Stoddard’s lab, led by staff scientist Dr. Betty Shen, focused on a particular enzyme, DrdV, which is responsible for this task. DrdV is a dual restriction-modification enzyme, meaning it has the ability to either methylate or cut DNA. Thus, when DrdV encounters unmethylated DNA, it must reliably determine whether that DNA should be protected, in which case it adds the protective methyl mark, or destroyed, in which case it cuts the DNA. There is, Dr. Stoddard described, one key difference between unmethylated host and foreign DNA that informed his group’s hypothesis about how this process works: “[replicating host DNA] transiently present only one unprotected target site to the enzyme at a time…[viral DNA] suddenly presents many unprotected target sites to the enzyme all at once, when the virus enters the cell.” Thus, the authors explain, in order to distinguish between DNA containing one vs. many unprotected sites, “one reasonable (and frequently postulated) solution to that challenge is to (1) ensure that cleavage is significantly faster than methylation, while also requiring that (2) multiple unmethylated DNA target sites be brought together into an enzyme-DNA complex before cleavage is licensed to occur.”

DrdV structure
Top: When DrdV encounters unmethylated Adenine (blue) in DNA, it must decide to methylate (top branch) or cleave(bottom branch) that DNA. Bottom: Cryo-EM map (left) and corresponding structural model (right) of the DrdV bound to DNA. Image provided by Barry Stoddard.

To test the two components of their hypothesis, the authors first purified DrdV and examined its enzymatic properties. By incubating DrdV with unmethylated DNA, they found that it methylates very slowly (it took between 4 and 16 hours to achieve full methylation), but cuts very quickly (it achieved 90% cleavage within 5 minutes). They then presented DrdV with DNA containing one or a pair of target sites, and found that it cuts the paired-site DNA much more effectively than the single-site DNA, consistent with their proposed model. To gain further insight into its function, the group then used Cryo-Electron Microscopy (Cryo-EM), a powerful new tool to visualize the structure of protein complexes at atomic resolution, in order to directly observe DrdV as it interacts with DNA. In viewing how DrdV proteins organize around one or multiple target sites, they concluded that “not less than three individual protein subunits and two bound DNA targets are required in order to form all contacts necessary to cleave a single bound DNA duplex…and four protein subunits are required in order to simultaneously cleave two DNA duplexes,” confirming their theory and their biochemical observations. In summary, the group concluded that the balance between methylation and cleavage is a race against time. Once a DrdV molecule binds a DNA target, a timer starts ticking. If it quickly encounters another DrdV bound to another target, it infers that the DNA is foreign and makes a cut. If, alternatively, the timer runs out before it can find a partner, DrdV infers that the DNA is host and methylates it. An elegant solution to a challenging molecular problem.

The battle between bacteria and phage may seem far removed from the concerns of daily life, but Dr. Stoddard doesn’t see it that way. He is enthusiastic about the potential of studying this conflict to reveal larger fundamental themes in biology that affect us all. “The battle between bacterial and the viruses (phage) that infect them represents one of the most ancient biological conflicts in existence, literally extending back billions of years. Therefore, it’s not surprising that viral defense systems in bacterial exist in an amazing and bewildering variety and diversity. I believe that the lessons learned about how the bacterial-viral battle is being played out will have significant implications for more immediate questions of viral restriction in multi-cellular organisms including humans.”

This work was supported by the National Institutes of Health, Amazon Cloud Credit, the Fred Hutchinson Cancer Research Center, and New England Biolabs. Part of this work was conducted at the Pacific Northwest CryoEM Center at Oregon Health and Science University and accessed through the Environmental Molecular Sciences Laboratory, a Department of Education Office of Science User Facility sponsored by the Office of Biological and Environmental Research.

Fred Hutch/UW Cancer Consortium member Barry Stoddard contributed to this work.

Shen BW, Quispe JD, Luyten Y, McGough BE, Morgan RD, Stoddard BL. (2021) Coordination of phage genome degradation versus host genome protection by a bifunctional restriction-modification enzyme visualized by CryoEM. Structure. S0969-2126(21)00085-X. doi: 10.1016/j.str.2021.03.012.