PMC

Site-specific endonucleases and AZT have disparate effects in this system. (A) Recombination efficiency: mating of the ΔrecA donor (ER3435) with and without rhamnose induction of the ΔrecA recipient (control, ER3473) or inducible rpnA (ER3514), mcrA (ER3533), asiSI (ER3535), or bsrDIB (ER3541) and with azidothymidine during mating (AZT; 2.5 ng/ml) (ER3435 × ER3473). (B) Toxicity of the expressed proteins/treatments during these matings.

YhgA-like protein overexpression induces an SOS reporter in rec-positive E. coli. The expression magnitude and dynamics of the indirect induction of dinD::lacZ by Rpn proteins were compared with the direct induction of rhaBp-lacZ by rhamnose. Strains with inducible rhaBp-rpnD (ER3560), rhaBp-rpnB (ER3561), rhaBp-rpnC (ER3562), rhaBp-rpnE (ER3563), or rhaBp-rpnA (ER3564) and rhamnose-inducible lacZ (rhaBp-lacZ, ER3245) were grown in parallel and induced when the OD₆₀₀ was 0.2. Aliquots were tested for β-galactosidase activity with time.

YhgA-like proteins increase recombination efficiency and are toxic. (A) Frequency of recombination during matings between the standard ΔrecA donor (strain ER3435) and either the ΔrecA recipient (control, ER3473) or a ΔrecA recipient with a rhamnose-inducible overexpression construct (rpnD, ER3481; rpnC, ER3512; rpnE, ER3513; rpnA, ER3514; and rpnB, ER3511) that was uninduced (no rhamnose) or induced (0.2% rhamnose). (B) Toxicity of expression was measured as the percent reduction in the number of recipient CFU per milliliter during rhamnose treatment relative to that for the control during the matings indicated in panel A.

Phenotypes of mutated RpnA proteins in vivo. (A) The RpnA domain structure, PD-(D/E)XK core motifs, and active-site residues predicted by Knizewski et al. (). RpnA variants with mutations in these active-site residues were created; an additional mutation in which a conserved PDDE motif was converted to PDAE (D165A) was made. (B) Recombination efficiency of matings between the ΔrecA donor (ER3435) and either the control ΔrecA recipient (ER3473) or ΔrecA recipients with rhamnose-inducible rpnA or its mutants (WT, ER3514; D11A, ER3552; D63A, ER3553; E82A, ER3554; Q84A, ER3556; Q84K, ER3555; R94A, ER3557; and D165A, ER3558) with and without induction. (C) Toxicity of induction during these matings, reflected in viability decline.

Alignment of the five E. coli K-12 YhgA-like proteins and their predicted domain structures. The RpnA (YhgA), RpnB (YfcI), RpnC (YadD), RpnD (YjiP), and RpnE (YfaD) sequences were derived from the E. coli K-12 genome and aligned. The chart on top gives the percent identity at each position of the alignment. Green, 100% identity; yellow, 30 to 70% identity; red, <30% identity. For each individual protein sequence, black bars denote 100% similarity to the consensus sequence, with progressively lighter bars indicating less similarity and gaps being represented as light gray lines. The location of the transposase_31 domain, as reported by the Pfam database (), and the PD-(D/E)XK structural core identified by Knizewski et al. () are shown as boxes under the alignment.

Distribution of genomic exchanges in recombinants. (A) Diagram of markers that distinguish the donor and recipient genomes and distances from the selected tetRA cassette. The chromosome segregation site dif is also shown. ICR, restriction enzyme gene cluster known as the immigration control region. (B) Recombinants were screened for the cat, npt, mrr, fhuA, and lacZ markers. Recombinants containing both tetRA and mrr were classified as having genome additions, and the results for these recombinants are not shown. Horizontal bars indicate the extent of donor DNA (orange) that we inferred replaced the recipient genome (blue) during the recombination event. NA, not available. (C) Proportion of recombinants in each class from a basal mating (WT; ER3435 × ER3473) with or without AZT treatment or from a mating in which rpnA was overexpressed (RpnA; ER3435 × ER3514). For the WT, most recombinants were created by large replacements of over 400 kb of genomic DNA. In the RpnA and AZT matings, large replacements were less frequent, and over half of the genomic replacements were within the 236-kb segment between the npt and fhuA markers flanking the selected marker, tetRA.

In vitro analysis of RpnA endonuclease activity. (A) WT RpnA cleaves pUC19, RpnA-D63A does not cleave pUC19, and RpnA-D165A is more active on pUC19. The pUC19 DNA (29 nM, 50 μg/ml) is initially supercoiled but can be relaxed by nicks, linearized by double-strand cleavage, or cleaved further. The supercoiled (control), relaxed (Nb.BtsI), and linear (HindIII) forms are indicated. pUC19 was treated with RpnA-inactive RpnA-D63A or hyperactive RpnA-D165A (15 μM, 45 min). (B) Time course of an RpnA (7.5 μM)-pUC19 (29 nM) digest. Band intensity was compared to determine the relative amounts of supercoiled, nicked, and linear pUC19 at each time point. Over 90% of the supercoiled pUC19 was digested within 180 min. (C) RpnA endonuclease activity depends on divalent cation and is stimulated by Ca²⁺. The reaction buffer was 50 mM NaCl and 10 mM Tris, pH 8.0; the indicated additives were at 10 mM each. RpnA at 3.8 μM was added for 18 h. (D) RpnA cleavage products provide a DNA polymerase primer. pUC19 was digested with RpnA, DNase I, or micrococcal nuclease (MNase) to produce similar smears and then incubated with fluorescein-labeled dNTPs and the Klenow fragment of DNA polymerase. DNA was visualized by ethidium bromide (EtBr; left) or fluorescein (middle), with the two signals being merged at the right. RpnA- and DNase I-digested DNAs were effectively labeled, but micrococcal nuclease-digested DNA was not.