The bacteriophage λ rex-centric mutualism phenotype, conditional rex, and other novel rex phenotypes
Slavcev, Roderick Adrian Ernest
Rex-Centric Mutualism The Rex Exclusion Phenotype, encoded by the rexA-rexB genes of bacteriophage lambda is defined as the inability of the mutant bacteriophage T4rll to generate plaques on a lysogenized E. coli(λ) host. Although the phenotype was first observed more than four decades ago, few advancements have been made in the elucidation of the Rex mechanism. The current model for Rex exclusion proposed in 1992 by Parma et aI., states that the Rex system functions as an altruistic bacterial apoptotic module triggered by T4rll infection of the lysogenized host. We asked whether Rex exclusion confers a protective or a cell-killing phenotype to the host and found that following T4rll infection, the Rex system can channel lysogenic cells into a temporary arrested growth phase that gives an overall protective effect to the host even at enormous multiplicities of infection compared to nonlysogens, despite some associated killing. We termed this phenomenon Rex-centric mutualism. In 1989, Snyder and McWilliams demonstrated that the Rex-mediated arrest of cell growth can be triggered in the absence of infection by over-expressing rexA relative to rexB. We noted that plasmid expression of rexA in Rex⁺ cells in the absence of infection resulted in similar cellular viabilities as that observed following T4rll infection. We visualized lambda Rex⁺ lysogens, infected by T4rll and found that they were much delayed in colony formation, contracted in length, formed aggregates with adjacent cells, and released flagella. These phenotypes were accentuated in nonlysogenic cells carrying a specific multicopy rexA-rexB plasmid: cells were about two-fold contracted in length, expressed membrane-bound and secreted flagellar structures, were odorous, were insensitive to infection by a variety of phages, and they extensively clumped/adhered when grown up in culture. Lysogenic cells mutated for rpoS stationary phase sigma factor were abrogated for Rex-centric mutualism, exhibiting more than 400-fold lower viability compared to the wild type, following infection by T4rll. These phenotypes show that the Rex system can impart a stationary phase like response that protects the host from T4rll killing. RexB Inhibition of T4rll Lysis Inhibition Phenotype We add to the activities and phenotypes of RexB. Expression of rexB from either a multicopy plasmid, or a rexA⁻-rexB⁺ phage is capable of suppressing the Lysis Inhibition Phenotype (LIN) seen upon infection of E. coli K strains by T4rll at high MOl. We also show that host mutations in either the periplasmic "tail-specific protease" tsp, or the 10Sa RNA ssrA completely abrogates the establishment of LIN inT4rll and T4 alike. We found that over-expression of rexB in cells suppresses both T4t⁻ and λS⁻ holin mutations, increasing the plating efficiency of the mutant phages by up to 10⁵ fold. Prophage level expression was noted to suppress the λS⁻ mutation to a lesser degree, but only from a rexB⁺-rexA⁻ prophage and co-expression of rexA with rexB on the prophage, or on a multicopy plasmid inhibited this activity of RexB. We also determined that nonlysogenic cells carrying the multicopy rexB plasmid are leaky for cytoplasmic proteins, whereas lysogenic (λrexB⁺-rexA⁻) cells are not leaky but did reveal an unusually high concentration of cytoplasmic β-galactosidase in the periplasm. Electron microscopy was used to visualize cells transformed with a rexB, or rexA-rexB multicopy plasmid. The rexB plasmid conferred gross distortions to the outer surface of the cell, while the rexA⁺-rexB⁺ plasmid imparted a shrunken, but otherwise normal appearance to cells. Our findings are consistent with RexB function as a pore forming unit, but RexB activity is inhibited by RexA. We propose a model for RexB suppression of T4rll lysis inhibition and the involvement of Tsp and 10Sa RNA in lysis timing and the establishment of LIN. The Conditional Rex Exclusion Phenotype The cl-rex operon of bacteriophage λ is expressed from the PM maintenance promoter of the prophage as PM-cI-rexA-rexB-timm message and confers a T4rII mutant phage exclusion phenotype to the lysogen (Rex exclusion). Derepression of the prophage results in very strong PE-cI-rexA-rexB-timm transcription, terminating at timm. Replication and excision defective, cryptic λcI[Ts]857cro27 lysogens exhibit a conditional Rex[Ts] exclusion phenotype. At temperatures where the C1857 repressor is functional, rexA-rexB expression from PM confers full Rex exclusion. However, upon thermal inactivation of the repressor little or no Rex exclusion is observed, despite a much higher level of transcription from PE, stimulated by CII. The same conditional Rex exclusion phenotype was observed in cells harbouring a low copy plasmid encoding a PTet-PM-c1857 -rexA-rexB-timm fragment, but not with a cl⁺ derivative plasmid, that imparted a Rex⁺ phenotype. Thermally derepressed λcl[Ts]857cro27 lysogens exhibited very high PE transcription levels that abated dramatically toward the C-terminal of rexA, showing a powerful polar effect on downstream rexB. Renaturation of C1857 following prophage induction did not reestablish repressor activity, although there was a 20-30 fold increase in transcription compared to that seen from the repressed prophage. Introduction of a rho mutation into our conditional Rex[Ts] strains partially suppressed Rex thermosensitivity, increasing Rex exclusion at 43°C by up to 10⁴ fold, while mutation of hflA to stabilize CII and heighten CII-dependent PE transcription conferred only a slight increase. Partial suppression of the conditional Rex[Ts] phenotype was also imparted by ssrA⁻and clpP⁻ null mutations, which suggests that Rex may be subject to 10Sa RNA tagging and ClpP(X) degradation. We propose two possible models to account for cI-rex polarity and correlation between CI activity and Rex exclusion. RexA:RexB Stoichiometry and the Rex Exclusion Phenotype We examined the influence of disrupting Rex stoichiometric balance on Rex activity to account for how polarity in the PTet-PM-cl857-rexA-rexB-t imm operon can abrogate the Rex exclusion phenotype. Cultured λ rex⁺ lysogenic cells were transformed with low-copy, and multicopy plasmids constitutive, or inducible for rexA⁺, rexB⁺, or rexA⁺-rexB⁺ expression. Lambda rex⁺ lysogenic cells transformed with a low-copy plasmid constitutively expressing rexA⁺, conferred only a minor attenuation of Rex exclusion, while transformation of rex⁺ cells with a constitutive, or induced multicopy plasmid expressing rexA⁺, or rexB⁺ completely suppressed the Rex exclusion phenotype. In contrast, multicopy and low-copy rexA⁺-rexB ⁺ and rex⁻ plasmid derivatives did not abrogate Rex exclusion in transformed rex⁺ cells. Furthermore, phage T4rll exhibited large rapid lysis plaques on the thermally induced Cro⁺ conditional Rex exclusion phenotypic lysogen, resembling plaques formed on λ lysogens carrying the multicopy constitutive rexB plasmid. Plaques formed on the isogenic cro⁻ derivative were tiny and nonsymmetrical; identical to T4rll plaque morphology on λ lysogens carrying the multicopy constitutive rexA⁺ plasmid. Our results suggest that an induced Cro⁺ λ prophage escapes Rex exclusion by over-expressing rexB relative to rexA, while a derepressed λcro⁻ prophage suppresses Rex exclusion phenotype by over-expressing rexA relative to rexB. CI Repressor Modulation of the Rex Exclusion Phenotype Bacteriophage λ mutants defective for ren and red (exo or bet) are sensitive to restriction by λ rex genes, but exclusion is modulated by the cl repressor allele of the prophage. λspi156nin5 forms plaques with 10⁵ fold higher efficiency on a cl⁺-rex⁺ lysogen than on cl[Ts]857, or cl[Ts]2 derivatives. Exclusion in cl[Ts] lysogens is suppressed by complementation with cl⁺ plasmid.