Lula in the Laboratory: How a Phage Has Contaminated Many E. coli Lab Strains

When I first saw the title of this PloSOne article, "Unauthorized Horizontal Spread in the Laboratory Environment: The Tactics of Lula, a Temperate Lambdoid Bacteriophage of Escherichia coli", I thought, "Hunh?!? You can actually publish articles about laboratory contamination?", but it's actually a very interesting article. In short, the article describes the discovery of a bacteriophage ('phage')--a bacterial virus--that is very well suited for survival and spread in microbiology laboratories.

To be successful, Lula faces several problems:

1) It has to successfully hide from researchers who don't want this phage mucking up their experiments.

2) E. coli K-12, the ubiquitous lab strain and workhorse of molecular genetics, already contains a phage, Lambda, against which Lula has to compete.

3) Researchers often use another phage, P1, in a process known as transduction. Basically, P1, when it enters infectious mode (known as lysis or the lytic phase), will sometimes take up some of its previous host's DNA. When it infects a new cell, and then is induced to become lysogenic (when lysogenic, the phage is quiescent, does not produce viral particles, and only replicates when the bacterium divides), the previous host's DNA is now in the newly infected cell. During this process, Lula has a difficult balancing act: Lula not only has to reproduce enough phage particles so that it won't go extinct during the transfer, it also has to not become noticed (if it is, the researcher will destroy the culture, and no more Lula).

Life is hard for Little Lula. So what's Lula's secret? Well, Lula has five secrets, actually:

1) When Lula infects a cell, it enters lysogeny much faster than other phage. This makes Lula difficult to spot.

2) Lula is rarely lytic at 37ºC, the temperature at which E. coli lab experiments are usually done (lysis involves taking over the bacterium, and turning it into a factory to make more phage, which then infect other bacteria, take them over, and so on). In fact, the only reason the authors discovered Lula is that they typically do their experiments at 28ºC--a temperature that makes Lula much more lytic.

3) Lula doesn't attach to cells that aren't growing. This probably seems obscure, but it helps to think of a phage as a hypodermic needle with DNA. When the phage bumps into the right part of the bacterial cell, it injects its DNA (and then either becomes lytic or lysogenic, depending on the circumstances). Phage aren't very smart, so many phage will inject their DNA into any old cell, including those that aren't growing. If the bacterial cell isn't growing, then the phage, regardless of whether it would have lytic or lysogenic, isn't reproducing. To put it crudely, Lula doesn't shoot its wad unnecessarily.

4) Even minor DNA damage causes Lula to leave lysogeny and become lytic. This is very important. To produce high concentrations of phage, for example, during P1 transduction, the bacteria are exposed to DNA damaging agents (e.g., mitomycin C) which induce the phage to enter lysis. Lula's 'hair trigger' for DNA damage gives it a head start over P1, which means it doesn't get lost when a small amount of the culture containing the phage (both P1 and Lula) is transferred to a new culture.

5) Lula kicks other phages' asses. Sometimes, two different phage will infect the same cell. In that situation, Lula is really good at inhibiting other phage from replicating. Win Lula.

So what is Lula? Some phage we've never seen before? Nope:

Sequencing of the Lula genome confirmed both its length (46,150 bp) and that it indeed had not been published before (E.R. and A.K., unpublished). However, several genes of Lula matched exactly the few sequenced genes of phi80, a lambdoid phage isolated by Matsushiro in 1961 and widely used in the 1970s and 1980s in phage studies. Eventually we tracked down a completed, but never published, phi80 genome sequence to the Blattner laboratory at the University of Wisconsin (Guy Plunkett III, personal communication). Comparison of the two genomes, -- Lula from Illinois versus phi80 from Wisconsin, -- showed that they were identical, so the cross-contaminating prophage turned out to be the original phi80.

It's just good ol' phi80, a once-widely used phage. So is this contaminant unique to this lab--maybe they just suck at microbiology? Not exactly (italics mine):

We then discovered that quite a few clones in our laboratory collection were contaminated with this commensal, attesting to its spreading powers. However, only when we found that some newly-arriving strains from the E. coli Genetic Stock Center or from other laboratories were also infected with the exact same prophage, did we comprehend both the scope of the infection and the uniqueness of Lula.

Oh dear. I think people need to start looking for phi80 in culture collections.

And now, a tangentially related music video:

An aside: I'm involved with a project that is sequencing the genomes of a bunch of commensal E. coli. Didn't find any phi80.

Cited article: Rotman E, Amado L, Kuzminov A, 2010 Unauthorized Horizontal Spread in the Laboratory Environment: The Tactics of Lula, a Temperate Lambdoid Bacteriophage of Escherichia coli. PLoS ONE 5(6): e11106. doi:10.1371/journal.pone.0011106