XJ Autolysis™ Strains From ZymoResearch

XJa and XJb, the two conditionally autolytic E. coli strains marketed by Zymo Research, are JM109 and BL21 derivatives, respectively. These strains allow for gentle and (nearly) complete cell lysis after one freeze-thaw cycle. To achieve the increased propensity for lysis, parental strains were genetically modified to introduce arabinose-inducible λ-endolysin gene. Endogenously produced (in presence of 0.2% arabinose) λ-endolysin combined with a freeze-thaw cycle results in 90% or better lysis efficiency without addition of lysozyme, or physical disruption (through use of ultrasonicator, French press, or bead-beater). Both strains are also available as λDE3 lysogens for T7 RNA polymerase-driven gene expression.

In our lab, we used both XJa(DE3) and XJb(DE3) strains for protein production using pET- and pTYB-based vectors. XJa(DE3), which was introduced first, appears to be poorly suited for routine protein production. Because it is based on the JM109 strain, it has higher levels of endogenous proteases, and contains an amber suppressor gene supE44. (In my hands, one construction utilizing amber stop codon exhibited ~50% suppression). The XJa strain is said to have some utility in purification of nucleic acids, but in my experience, the lysis of cells is hardly a problem when using lysis reagents from a number of plasmid purification kits (including those from Zymo Research). The only realistic application of this strain I can envision is some cytotoxic protein production from an expression vector that is liable to undergo RecA-mediated inactivation in BL21(or like) strain. In this case, the recA- status of XJa(DE3) can be a solution to a plasmid stability problem. Overall, I found XJa(DE3) to be of very limited use.

XJb(DE3) appears to have retained all the positive features of the classic expression strain BL21 it was derived from. I have used this strain in production of more than 30 different proteins from pET- and pTYB-based vectors using conventional IPTG induction as well as auto-induction (Novagen’s Overnight Express™) protocols. The only modification for these protocols was the addition of arabinose to 0.1-0.2% at the time of supplying the inducer as required for protein expression (1 mM IPTG or Overnight Express™ reagents). Harvested cells were frozen at –80ºC for at least 2 hours and exhibited significant lysis (~60-90% based on soluble protein estimates) upon thawing. One round of sonication usually allows for nearly complete lysis. Addition of extra lysozyme and/or detergent-based lysis reagents did not result in detectable increase in lysis efficiency.

The single biggest improvement achieved through the use of the XJb(DE3) autolytic strain is the greatly reduced time of ultrasonication when used for expression of protein from pTYB-series of plasmids (New England Biolabs). Proteins produced using these vectors (in turn being a part of the IMPACTTM-CN expression system) are synthesized as chitin-binding-domain-intein fusions and require chitin beads for affinity purification. Since chitin is readily degraded by the lysozyme, the latter used in connection with the IMPACT™ system is not recommended. Consequently, the omission of the lysozyme from the cell lysis protocol entails more rounds of mechanical disruption (e.g. ultrasonication) when using conventional BL21(DE3) cells. In my experience, the levels of λ-endolysin produced in XJb(DE3) cells in presence of 0.05-0.1% of arabinose are sufficient for ~90% complete cell lysis after one round of sonication without significant degradation of the chitin resin. As a result, the preparation of the cell extract is greatly simplified, hand-on time is significantly reduced as are the chances of protein denaturation/degradation.