E. coli aficionados, do your strains have the wild-type rpoS gene?

by Roberto
 

For many of you, the rpoS allele in your E. coli strains might not matter. But if you are interested in the general stress response or adaptations to stationary phase, you might want to know if you are starting out with a wild-type rpoS gene. After all rpoS encodes the sigma factor σ^S, a global regulator that directs RNA polymerase to transcribe genes essential for adapting to numerous stresses among them starvation conditions.

 

A recent thread on Twitter (now 𝕏) by Alexander Harms prompted me to write this post. In short, he expressed concern that E. coli strains are often shipped in LB agar stabs after growth in LB medium, citing a nice paperby Spira et al. who explored the subject in detail. Back in 2009, Horishi Nikaido warned STC readers – in what is probably our most viewed post – of several problems with using LB. But the problem that Harms is concerned about is not one that arises during growth but rather in keeping LB cultures in stationary phase for long periods. That's something that takes me back more than thirty years, as we worked on this in my lab. In brief, in LB the cells catabolize amino acids to grow and thus release ammonia and alkalinize the medium. The high pH of stationary phase cultures kills the cells. After a couple of days, ~99% of population is dead. Surprisingly, the survivors are not the parental type. In that short time the population will have been taken over by mutants with a "GASP" phenotype (Growth Advantage in Stationary Phase). With widely varying dynamics, GASP appears to be a general phenomenon of bacterial evolution. But in the particular case of E. coli grown in LB, if you start with a strain that has a wild-type rpoS, the first GASP mutants that will take over the culture almost invariably carry attenuated or "hypomorph" rpoS alleles whose partial loss of function leads to the entire RpoS regulon being down regulated. One unexpected consequence of these alleles is that E. coli that harbor them grow better on the detritus of dead cells (first and foremost amino acids), thus expressing the GASP phenotype. Along the decades and across the world, many investigators appear to have kept their E. coli strains on LB for long periods because many lab strains indeed carry attenuated rpoS alleles.

 

Fig. 1. Assay for Hydrogen Peroxidase II (catalase) from E. coli. Strains bearing the wild-type (A), rpoS819 (B), or rpoS::Km (C) alleles were grown on LB agar plates and assayed for catalase activity by direct addition of H₂O₂ to the cells. Source Frontispiece: Most of the cells have died in 3-day-old culture of E. coli grown in LB liquid medium. Cells were treated with the vital stain acridine orange. Dead cells are green, live cells orange. Source: Kolter Lab Micrograph, ca. 1990.

Which brings me back to the original point, you might want to know what rpoS alleles are present in your strains. Should you go through the trouble of sequencing their rpoS genes? I say no, at least not as a first step. As easy as sequencing has become, I have an easier approach. This is an assay that will take you seconds, provided you have pre-grown your strains overnight. Because rpoS controls the expression of catalase (hydrogen peroxide II), the activity of this enzyme can serve as a proxy for the allelic state of rpoS. After growing patches of the strains in question overnight (yes, LB plates are okay for this) simply apply a drop of hydrogen peroxide solution on top of each patch. The breakdown of hydrogen peroxide will result in bubbles of oxygen gas. You'll quickly see robust bubbling from the strains with the wild-type allele and much diminished bubbling from strains that carry any one of the many attenuated alleles (virtually no bubbles from the rpoS null strain). That is truly simple, and simple I like. (Note: having strains with the wild-type and null rpoS alleles as controls helps!)

 

Do I have any advice on how to ship E. coli strains? Here's a simple suggestion that might be worth trying. Why not send them in agar stabs but using 0.05X LB (or some other rich medium but similarly dilute). With the lower concentration of nutrients, the pH will not shift much, there will be much less death and the takeover by GASP mutants will be much slower. Your strains should arrive okay.

Do you want to comment on this post? We would be happy about it! Please comment on Mastodon, Bluesky, or on 𝕏 (formerly Twitter).